o g
a i
ui s
□ "
t-1 1
a{
a ?
ru i
tr ;
_a !
o= I
IS
HANDBOOK OF PHYSIOLOGY
section 2: Circulation, volume ii
HANDBOOK EDITORIAL COMMITTEE
Maurice B. Visscher, Chairman
A. Baird Hastings
John R. Pappenheimer
Hermann Rahn
HANDBOOK OF PHYSIOLOGY
A critical, comprehensive presentation
of physiological knowledge and concepts
SECTION 2:
Circulation
VOLUME II
Section Editor : W. F. HAMILTON
Executive Editor: PHILIP DOW
American Physiological Society, Washington, d. c, 1963
©Ipopyright !9^3> American Physiological Society
Library oj Congress Catalog Card No. 60-4587
Printed in the United States of America by Waverly Press \ Inc., Baltimore, .Maryland 21 202
Distributed by The Williams & Wilkins Company, Baltimore, Maryland 21202
Contents
23. Functional anatomy of cardiac pumping
GERHARD A. BRECHER
PIERRE M. GALLETTI 759
24. The physiology of the aorta and
major arteries
JOHN \V. REMINGTON 799
25. Pulsatile blood flow in the vascular
system
MERRILL P. SPENCER
ADAM B. DEMSON, JR 839
26. The anatomy and physiology of the
vascular wall
HERMANN BADER 865
27. Patterns of the arteriovenous pathways
MARY P. WIEDEMAN 89 1
28. Resistance (conductance) and capacitance
phenomena in terminal vascular beds
HAROLD D. GREEN
CARLOS E. RAPELA
MARGARET C. CONRAD 935
29. Exchange of substances through the
capillary walls
E. M. LANDIS
J. R. PAPPENHEIMER 96 1
30. The physiologic importance of lymph
H. S. MAVERSON IO35
3 1 . The peripheral venous system
ROBERTS. ALEXANDER I°75
32. Venous return
ARTHUR C. GUYTON I O99
33. Effects of ions on vascular smooth muscle
SYDNEY M. FRIEDMAN
CONSTANCE L. FRIEDMAN I I 35
34. Lipid metabolism in relation to physiology
and pathology of atherosclerosis
SAMI A. HASHIM
WILLIAM C. FELCH
THEODORE B. VAN ITALLIE I 1 67
Note: For coverage of blood flow to the brain, see "The
Cerebral Circulation" by S. S. Kety, chapter LXXI, pp. 1 751 —
1760, Neurophysiology section of the Handbook series.
35. The role of endocrines, stress, and
heredity on atherosclerosis
L. N. KATZ
R. PICK I I97
36. Peripheral vascular diseases —
diseases other than atherosclerosis
GEORGE E. BURCH
JOHN PHILLIPS 1215
37. Situations which lead to changes in
vascular patterns
AVERILL A. LIEBOW 1 25 1
38. Methods of measuring blood flow
KURT KRAMER
WILHELM LOCHNER
E. WETTERER 1277
39. The circulation through the skin
A. D. M. GREENFIELD !325
40. Circulation in skeletal muscle
HENRY BARCROFT 1 353
41. The hepatic circulation
STANLEY E. BRADLEY '387
42. The flow of blood in the mesenteric vessels
EUGENE GRIM 1439
43. The renal circulation
EWALD E. SELKURT I457
44. Blood supply to the heart
DONALD E. GREGG
LLOYD C. FISHER I 51 7
45. Maternal blood flow in the uterus and
placenta
S. R. M. REYNOLDS 1 585
46. The fetal and neonatal circulation
MAUREEN YOUNG 1619
47. The flow of blood through bones and joints
WALTER S. ROOT I 65 I
48. Dynamics of the pulmonary circulation
ALFRED P. FISHMAN 1 667
Index 1745
82264
CHAPTER 23
Functional anatomy of cardiac pumping1
GERHARD A. BRECHER
PIERRE M. GALLETTI
Department of Physiology, Emory University
School of Medicine, Atlanta, Georgia
CHAPTER CONTENTS
Macroscopic Structures
Composition of Cardiac Tissues
Architecture of the Ventricular Myocardium
Architecture of the Atrial Myocardium
Pressure and Flow Effects During the Cardiac Cycle
Correlation of Other Cardiac Events With the Cardiac Cycle
Atrial Pressures
Electrocardiogram
Vibrocardiogram (Apex Cardiogram I
Function of the Heart Valves
Veno-Atrial Junction
Atrioventricular Valves
Arterial Valves
Ventricular and Atrial Volumes in Various Activities
Ventricular Volume
Atrial Volume
Atrial Filling
Ventricular Filling
Differences Between Right and Left Cardiac Cavities
The Pericardium
Closing Remarks
ALTHOUGH THE PURELY MECHANICAL NATURE of
cardiac pumping is taken for granted by modern
scientists, this view has not always been accepted in
the past. Only during the last hundred years were the
forces of muscle contraction finally stripped of the
'vis vitalis' and ascribed exclusively to energy trans-
formation according to the laws of physics and
chemistry. In this historical process, the heart which
had been formerly thought of as the seat of emotions,
was deprived of all metaphysical connotations and
became an organ of purely mechanical function just
1 The results of some recent experiments of the authors and
their colleagues are quoted in this paper. This work was sup-
ported in part by USPHS grant H-3796, and grants from the
Life Insurance Medical Research Fund and the Georgia
Heart Association.
as the skeletal muscle. It is of interest to trace briefly
the emergence of this concept (160, 161).
During the age of the pyramids (3000-2500 B.C.)
an unknown Egyptian clearly recognized the heart
as the center of a system of distributing vessels and
associated the pulse with the cardiac beat. The Greek
philosopher Alcmaeon of Croton (about 500 B.C.)
distinguished the veins from the arteries and asserted
that the seat of sensation was not in the heart but in
the brain. The function of the heart as a pump was
apparently expressed for the first time by Plato (427-
347 B.C.) when he stated: it "pumps particles as from
a fountain into the channels of the veins, and makes
the stream of the veins flow through the body as
through a conduit." Hippocrates (493-423 B.C.)
had described the cardiac valves, the ventricles and
the great vessels, but he did not refer to the pumping
action, which he might have taken for granted. For
Aristotle (384-322 B.C.) the heart was the seat of
"innate heat" and also of the soul. This notion was
probably based on the observation that death results
from dissection of the beating heart. However, from
his studies on the embryonic chick heart Aristotle
may have had knowledge of the pumping function.
Erasistratus (310-250 B.C.), who described the
aortic valves, pulmonary valves, and chordae
tendineae, and Galen of Pergamon (131-201 A.D.)
both stated that the heart is a pressure-suction pump.
Their view was founded mainly on the assumption
that during diastole blood was sucked into the
ventricles by active enlargement of the cardiac walls
[discussed by Ebstein (40), Bohme (14)]. They also
believed that blood is expelled backward into the
caval veins during ventricular systole. The first
definite statement concerning the continued forward
flow of blood from the right ventricle through the
lungs into the left heart was made by Ibn an-Nafis
(1210-1299 A.D.). The first scientist of the Renais-
759
760
HANDBOOK OF PHYSIOLOGY
CIROl'LATION II
sancc who recognized the heart as a hollow muscle
and probably as a pump was the artist-engineer,
Leonardo da Vinci (1452-1519 A.D.), who stated:
"The heart is a principal muscle, in respect of force,
and it is much more powerful than the other muscles"
[Keele (90)]. However, it remained to William
Harvey ( 1 578—1657) to prose that the heart, and not
the liver, is the center of the vascular system and
that it propels the blood unidirectionally by its
rhythmical contractions as would the repeated
strokes of a man-made pump. The microscopic proof
of the muscular nature of the heart was brought by
Niels Stenson (1 638-1 686), who demonstrated that
the substance of the heart is composed of fibers,
membranes, arteries, veins, and nerves just as is the
substance of other muscles. Once this important point
had been firmly established, it became customary to
consider the heart as a pump, to develop analogies
with mechanical systems of fluid transfer, and to
apply to the myocardium the increasing knowledge
about skeletal muscle contraction. The present
chapter is a rather general and classically oriented
treatment of the mechanical function of the heart. It
attempts to provide an understanding of the anatomi-
cal structures, while avoiding teleological oxer-
simplification as well as useless controversies about
functions.
The role of the heart consists of providing the body
tissues with a continuous stream of blood. The heart
fulfills this function by converting potential energy
(primarily chemical energy, secondarily energy of
position) into kinetic energy, as movement is imparted
to the blood ejected from the ventricular cavities.
From the standpoint of cellular function at large, it
does not matter whether tissue perfusion is brought
about by alternate contraction and relaxation of
myocardial cells, or by the action of an artificial
pump. This concept has been established on a firm
experimental basis by the advent of extracorporeal
circulation techniques, whereby a mechanical pump
substituted for the human heart can fully support the
circulation. Thus the heart can be looked upon as a
pump inserted in the circulatory system and its
function can be described by analogy with purely
mechanical systems.
Mechanical pumps are divided into two main
classes: kinetic pumps and positive displacement
pumps. In the former class, kinetic energy is added to
the fluid by the forced rotation of an impeller (fig.
lA). In the latter class, the fluid is progressively
displaced from a suction inlet to a discharge opening.
Two kinds of positive displacement pumps need to
fig. i . Mechanical analogues for some pumping principles
embodied in the heart. A: kinetic pump in which energy is
added to the fluid by the rotation of an impeller. B: rotary
pump in which fluid is propelled through squeezing a resilient
tube by means of rollers mounted on a rotating arm. C: recipro-
cating pump in which fluid is displaced by the back and forth
movement of a diaphragm while valves give direction to the
stream.
be mentioned here. In rotary pumps (fig. \B),
moving members trap a portion of the fluid in a
chamber of pliable tubing and conduct it toward the
outlet. The segment of tubing occluded acts as a
valve to prevent backflow. In reciprocating pumps
(fig. 1 C) a cavity limited by two valves is subjected to
the action of a piston or diaphragm. As the piston
moves back and forth, fluid is drawn in through the
suction valve and forced out through the discharge
valve.
The action of the heart in some invertebrates can
be compared to that of rotary pumps, since forward
movement of fluid is obtained by peristaltic move-
ments of the walls. In the mammalian heart also
some degree of blood propulsion may be accomplished
on the ''progressive cavity principle" as in rotary
pumps, particularly the displacement caused by the
wringing action of the ventricles. However, cardiac
action in vertebrates most closely resembles that of
reciprocating pumps. It is characterized by pulsatile
action, by the presence of valves, and by the capa-
bility of the pump to be adjusted in terms of either
speed, or volume displacement, or of speed and
volume displacement simultaneously. Although the
design of the heart has nothing in common with that
of kinetic (centrifugal) pumps, its control displays
two characteristics for which kinetic pumps are
appreciated in technology: namely that the volume
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
76.
output is directly related to the input pressure, and
is inversely related to the pressure head against which
the pump works. Like centrifugal pumps, the heart
has the tendency to deliver a higher flow as more
blood is fed into it at the atrial level; it also provides
a lower flow when the resistance to ejection in the
vascular system increases.
A close look at mechanical pumps for cardiac
substitution throws a light on built-in features of the
natural heart that one easily takes for granted.
Adequate perfusion of an adult human organism
under all possible conditions requires that :
/) The heart be able to move blood volumes
ranging from 3 to 30 liters per min and to pump
against pressures up to 300 mm Hg.
2) Even at maximal cardiac output, the flow
velocity must not exceed the limit of tolerance for
mechanical trauma to blood corpuscles through
turbulence, friction, or cavitation (1-2 m/sec).
3) The relationship between stroke volume and
stroke rate must not deviate much from an optimum
which is set by the elastic properties of the cardiac
walls, the time needed for efficient transformation of
potential into kinetic energy and by the lowest flow-
velocity compatible with the output required.
4) The valves must easily open during their flow-
phase, yet be competent and prevent regurgitation
of blood during their holding period.
5) The regulation of the pumping action must be
automatically controlled through sensing elements
with feedback mechanisms which adapt the output
to the tissue demands [see also Wagner (153)]- These
control mechanisms must integrate hemodynamic
data (e.g., perfusion flow, arterial and venous pres-
sures) and metabolic data (e.g., arteriovenous oxygen
difference) to maintain viable conditions.
Considering these points in more detail, one must
first emphasize the pumping capacity of the heart.
As 3 to 30 liters per min of blood is pumped by the
left ventricle into the systemic circulation, practically
the same amount is ejected by the right ventricle into
the pulmonary vascular bed. Furthermore, the atria
have some pumping function of their own, so that
the combined pumping of all the chambers of the
human heart is in the order of 7 to 70 liters per min,
depending upon the state of muscular activity. A
range of this magnitude (1:10) is not easily obtained
in artificial pumps and, when it is reached, it is at the
price of considerable sacrifices in mechanical efficiency
(ratio of work produced to fuel consumed). On the
contrary, the mechanical efficiency of the heart does
not seem to be very closely related to cardiac output.
The extended scale of activity over which the heart
can perform is certainly facilitated by the elastoviscous
properties of the cardiac walls. The cavities are
distensible over a wide range of volume increments
without much increase in intraventricular or intra-
atrial pressures [see fig. 2, and Little (99)]. Therefore
the heart can easily accommodate and deliver
varying stroke volumes even if the stroke frequency
remains unchanged. Furthermore the time needed
for the transformation of chemical into mechanical
energv apparentlv comprises only a fraction of the
systole. At a constant stroke volume the heart can
increase its minute output simply by beating faster
and shortening the pause between the strokes without
affecting the energy conversion processes. The limiting
factor of cardiac output at high heart rates is not an
encroachment on the time needed for energy con-
version but an encroachment on the time needed for
filling the pump chambers (ventricular filling phase).
Another fundamental difference between artificial
pumps and the heart is that in the former a force is
applied from the outside to activate a part or the
entire wall of the pump chamber, whereas in the
latter the force is developed within the wall of the
pump chamber itself by small elements, the muscle
fibrils, which alternately shorten and lengthen.
Furthermore, since the heart is surrounded by other
resilient structures in the thorax, there is an inter-
action of the physical forces developed in the myo-
cardium and those developed either passively or
actively in these structures [Pfuhl (129, 130), Blair &
Wedd (12)]. For example, during ventricular con-
traction and ejection the elastic forces of the lungs
oppose to a small extent the diminution of the
ventricular size, whereas during ventricular relaxation
the same forces of the lunsjs enhance slightly the
expansion of the ventricles. These forces are said to
be negligible as compared with the intravenous
filling pressures (60, 64). Mechanical effects are
exerted upon the rhythmical form changes of the
heart by such structures as the pericardium, the
attachments of the heart to the large vessels, the
sternum, the mediastinal tissues, and the diaphragm
through its changes in position during respiration or
because of varying degrees of abdominal filling. The
complexity of these forces, in terms of direction and
magnitude, and their continuous changes during the
cardiac and the respiratory cycle make it presently
impossible to evaluate quantitatively the contribution
of extracardiac structures to cardiac pumping.
Nevertheless, their importance is demonstrated by
the possibility of pumping blood solely by the action
762
HWUBOOK OF PHYSIOLOGY
CIRCULATION II
of external forces on the heart [Hosier (81), Stephen-
son (150)]. In closed-chest cardiac massage, vigorous
pressure on the lower part of the sternum causes
ejection of the ventricular content into the large
arteries. Conversely, when pressure is released, the
recoil is sufficient to permit the venous pressure to fill
the ventricles again [Kouwenhoven et al. (92)]. In
this manner, a sufficient, though subnormal, cardiac
output can be maintained in the absence of any
myocardial activity. This points again to the fact
that, in principle, it does not matter whether the
propulsion of blood through the body is brought
about by the contraction of cardiac fibers or by any
other suitable forces applied to the blood contained
in the ventricles.
MACROSCOPIC STRUCTURES
A great deal of commonly accepted knowledge
about cardiac pumping is derived from purely
morphological considerations. Although conclusions
reached in this manner have occasionally proved to
be correct, morphological reasoning often leads to
fallacious lunctional interpretations of structural
findings. In the case of the heart, physical vector
analvsis of all the mechanical forces involved is
especially difficult because of the great complexity of
the anatomical structures and of the perplexing
geometry of cardiac filling and emptying. We have
only a limited knowledge of the sequence of events
as they occur during muscular contraction and
relaxation within various parts of the myocardium.
In this particular section an attempt is made to
describe the macroscopic structures of the heart with
reference to their probable function as deduced from
the anatomical observations. A topographic anatomi-
cal description of the heart is available in standard
texts (51, 95, 98, 101).
Composition nj Cardiac Tissues
The myocardium is the most important structure
of the heart because its contraction causes the blood
to flow. However, it should be realized that only
part of the cardiac walls consists of muscle fibers, and
that within the muscle fibers, the contractile substance
is limited to the fibrils. Indeed about half of the
heart's weight is made of noncontractile material
such as the sarcolemma in the muscle fibers, con-
nective tissue in the heart skeleton, tendons and
valves, and finally blood vessels, lymphatics, and
nerve fibers. All these elements are interwoven with
the muscle fibers or closely connected to them (45,
59). During cardiac contraction or relaxation, they
are deformed and resist to some degree the shortening
or lengthening of the myofibrils.
Little is known about the mechanical effects of the
coronary vessels upon the function of the ventricles.
Though relatively inconspicuous in a "dead" heart,
they appear heavily engorged with blood in the live
organ. In fact, since 5 to 10 per cent of the cardiac
output passes through the coronary system, a signifi-
cant mass of the beating heart consists of circulating
blood contained within the anatomical bounds of the
epicardium. During heavy exercise the coronary-
blood supply is probably so great that one might look
upon the myocardium as a spongy structure of muscle
fibers suspended like chains of islands in a lake of
blood. In the past it has been postulated frequently
that the degree of filling of the coronary vascular bed
affects in some form the ventricular contraction.
AV +
fig. 2. Left ventricular pressure-volumes curves of a dog
heart illustrating the changes resulting from coronary perfu-
sion. The freshly excised heart of a 13.5-kg dog was submerged
in Locke's solution and assumed its elastic equilibrium state
(zero transmural pressure, origin of the coordinates) upon
cessation of spontaneous contraction. The curves were obtained
by addition or reduction of the intraventricular volume
[Brecher & Kissen (22)]. The origins of the coordinates for the
perfused and unperfused heart were arbitrarily superimposed.
At negative (and up to +5 mm Hg) intraventricular pressures
the ventricle accommodated a greater volume with coronary
perfusion than without. At pressures above +5 mm He, the
ventricle accommodated less fluid with coronary perfusion
than without (Horres et al., unpublished data).
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
763
Most of these postulates were of speculative nature.
For example, Donders (39) stated that "the blood
which enters at the end of systole into the coronary
arteries seems to cause a slight active expansion of the
heart, especially of the ventricles.-' This view was
originally formulated in 1855 by Briicke (26) and
also advocated by Luciani (102). Based on X-ray
kymograph studies, a modern modification of the
same hypothesis was presented by Cignolini (34)
without conclusive evidence. However, recent work by
Salisbury et al. (141) indicates that the filling of the
coronary bed affects the ventricular distensibility.
There are indeed significant differences in the
ventricular pressure volume relationship depending
upon whether the coronaries are perfused or not
[Brecher et al. (24)]. In figure 2 the S-shaped pressure
volume curve of the ventricle with an empty coronary
bed (solid line) is different from that obtained during
coronary perfusion (broken line). This shift of the
curve when the coronary bed is perfused indicates
that the perfused heart accommodates more fluid at
low intraventricular pressures and less fluid at high
intraventricular pressures. Around the elastic equilib-
rium state (zero transmural pressure) the perfused
heart is somewhat stiffer than the nonperfused heart.
The effect of varying degrees of engorgement of the
coronary bed upon the distensibility of the beating
ventricle during the different phases of the cardiac
cvcle is still unknown.
The heart skeleton, the chordae tendineae, and
the cells of the Purkinje system are noncontractile,
yet are functional components of the myocardium.
The heart skeleton (fig. 3) is represented by four
interconnected fibrous rings of dense connective
tissue, which surround the orifices of the great
vessels. The musculature of the ventricles and atria,
the roots of the large vessels, and the heart valves are
attached to this skeleton, which also anchors the
tendinous endings of the ventricular muscle (see
below). An important function of the cardiac skeleton
is to provide a firm basis for the attachment of the
cardiac valves. Another function, though less fre-
quently mentioned, is to aid in keeping the orifices
open during the phases of blood inflow and outflow.
During ventricular activity, the orifices undergo
changes in form which probably involve also the
cardiac skeleton as indicated in the different outlines
of the orifices during systole and diastole in figures 4,
5, and 6. By inserting a finger through the atrial
appendage in the intact beating heart, one can
easily verify that the atrioventricular valve rings
become smaller during ventricular contraction and
larger during relaxation. This observation, which
has not been substantiated by precise measurements
as yet, indicates that the fibrous tissues of the heart
skeleton are passively deformed by myocardial con-
traction and thereby store energy which is released
fig. 3. Anatomic components of the heart
depicting the relation of the fibrous skeleton
to the heart chambers and arterial roots. The
trunks of the aorta and pulmonary artery as
well as the atria are fastened to the cranial
aspect of the four annuli fibrosi, whereas the
ventricles are attached to the caudal aspect.
[From Rushmer (139).]
764 HANDBOOK OF PHYSIOLOGY -^ CIRCULATION II
Yalvula scmilui ior a, pulmonale
Valvula semiltinai is
V"al\ ul.. -' mihmaris
:
■ .
Cuspis anterior \ alvulac
l'i< 11-jnd.ih-
(. uspis posh rior
i-alvulac bit u*pidalis
■
■.,--■!.:-'.'
Ivula semilunaris di -srr.i
1- pulu
- \ tlvnl.i scmilun 1
\ anti n.ir ,! ■ ■
^ \ ■ 1 ! ' ■ 1 ' l ■ :
~ ~ ant" i
aritern >r
Cn>pw f . ,
I
tncuspidalu
1 1
1
(.Anulu '
,' ' 'I- ■'
Aniilus fibroins sinister
\<-ntri< nli
fig. 4. Base of the human ventricles seen from their cranial aspect after the atria have been re-
moved. The shape of the ostial orifices in the state of contraction differs significantly from the shape
in the state of relaxation, as indicated by the dashed lines. [From Spalteholz (148).]
by elastic rebound at the beginning of muscular
relaxation.
Many strands of myocardial fibers end with
tendinous tissues. Yet one cannot compare them with
skeletal muscle, since there is no bone to provide a
fixed attachment. In reality, all myocardial fibers end
on other myocardial fibers either directly or by-
insertion of connective tissue. For instance the myo-
cardial fibers of the papillary muscles continue as
chordae tendineae, which in turn lead via the bicuspid
and tricuspid valve leaflets and the fibrous tissue
of the heart skeleton to other myocardial fibers. This
arrangement forms a circle of myocardial tissue,
although with inclusion of a tendinous segment.
Other myocardial fibers, such as many strands in the
left ventricular deep bulbospiral bundles, simply form
a circle. Since, in the final analysis, all myocardial
fibers pull directly or indirectly on other myocardial
fibers, the concerted effect of their contraction
diminishes each heart cavity more or less concentri-
cally [see also Hawthorne (67)]. It also stands to
reason that all muscle "■fiber-rings" which include a
noncontractile segment exert during their contraction
a pull on the noncontractile segment, storing in it
potential energy for release during myocardial
relaxation.
The cells of the conduction system have a special
position as far as their participation in the contractile
process is concerned. They are derived from muscle
cells, but their primary function is the fast conduction
of excitation. Yet they do contain a small number of
myofibrils and therefore must be expected to partici-
pate in the over-all myocardial activity. Since nobody-
has measured their contractile force, it remains a
matter of conjecture whether Purkinje cells contribute
to any significant extent to the force of ventricular
contraction. It may be that their contraction serves
only the purpose of diminishing the shear forces which
would develop between myocardial and Purkinje
cells if the latter remain purely passive. In the
longitudinal direction the Purkinje cells are joined to
intermediate cells which connect them to myocardial
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
765
Auricula
sinistra
Basis cordis
Ati nun sinistrum
\'. pulmoaalis dextra
V. cava superior
! A +
Sinus coronarius
Septum ventru ul< muo
Ostium vcn
sinistrum /-
\ in - ill,:
dextrum
:
(c.
fig. 5. Superficial muscle layers of the
maximally contracted human heart, viewed
from the caudal aspect after separation of the
atria (above) from the ventricles (below).
The ostia of the contracted ventricles can be
compared with their state in the relaxed
ventricles (dashed lines). The changes in
ventricular configuration during relaxation
are also indicated by dashed lines. [From
Spalteholz (148).]
Apex cordis
cells. These intermediate cells contain an increasingly
larger number of myofibrils as they approach the
true myocardial cells. Merely judging from morpho-
logical evidence, they must contribute to some
extent to the over-all contractile process.
Architecture of the Ventricular Myocardium
Since the ventricles perform more pumping action
than the atria, the architecture of the ventricular
myocardium has attracted most of the attention of
functionally oriented anatomists. Despite extensive
description by MacCallum (108), Mall (111),
Monckeberg (114), Benninghoff (10), Robb &
Robb (136), Spalteholz (148), and Lev & Simkins
(97), much confusion still prevails. Opinions vary
because it is difficult to dissect clearly the complexly
arranged, intertwined and crisscrossing discrete
muscle bands. Consequently, it is even harder to
derive from the anatomic findings a picture of the
direction of maximal pull of each muscular compo-
nent, not to mention the concerted action of several
components.
Many of the muscle bands encircle both left and
right ventricles. According to the most commonly
accepted terminology, one distinguishes four different
muscles, the course of which can be best understood
from semischematic drawings: the superficial bulbo-
spiral (fig. 7), superior sinospiral (fig. 8), deep sino-
spiral (fig. 9), and deep bulbospiral muscle (fig. 10.)
766
HANDBOOK OF PHYSIOLOGY — CIRCULATION II
V. pulmoiLilb dcxtra
\trium Binblrum
V i iv tpei
Aiin< nl i sinistra
fig. 6. Superficial muscle layers of the
maximally contracted human heart, seen
from the ventrocranial aspect after separation
of the atria (above) from the ventricles (be-
low). The position changes of the great vessels
and the ventricle outlines during relaxation
are indicated by dashed lines. [From Spalte-
holz (148).]
A. pulmonalis
i^iiuam.ilis
anterior
Ventriculus dexter
iDcfsura [apicfe ■ ordfc
Apex coidis
Vol ti & tunlii
According to Benninghoff (10), who uses a somewhat
different classification, there are three interconnected
systems which intersect rectangularly: a) the outer
longitudinal fibers which connect to the outer contour
fibers at the ostia; b) the ring fibers which encircle the
entire chamber and curve around to form fibers of
the ventricular septum; c) the internal longitudinal
fibers which run from the contour fibers toward the
apex (figs. 11 and 12). Benninghoff (10) analyzed
the function of these various bundles on the basis of
careful comparative anatomical studies and in vivo
observations. He emphasized the concept that
crossing of the fiber layers at right angles results in an
over-all reduction of the cavity size, as first postulated
by Carl Ludwig. Each of the three systems affects the
entire heart and at the same time each of the ventric-
ular cavities. They act in such a manner that a
reduction of the heart chambers does not occur
equally in all directions but in such a manner and
sequence that the cavities are emptied toward their
outflow tracts. The evolution proceeded as follows: in
lower vertebrates (fish, amphibia) there are no
tendinous elements and all muscle bundles are ring
shaped. In the mammalian heart secondary valves
(atrioventricular) are formed from which the con-
nective tissues of the fibrous rings of the atrioventricu-
lar valves and of the chordae tendineae originate and
become inserted into the course of the ring-shaped
muscle. The fibrous rings become connected to the
roots of the arteries and form the solid trigona fibrosa,
which furnish new insertions for many myocardial
fibers (see fig. 4). In the evolutionary process the
FUNCTIONAL ANATOMY OF CARDIAC PUMPING 767
FIG. 7. The superficial bulbospiral muscle as seen from the
front of the human heart. A = Aorta; M = mitral orifice;
P = pulmonary artery; T = tricuspid orifice; AT — anterior
leaflet of tricuspid valve; MT = medial leaflet of tricuspid
valve. A F-shaped section is cut from those fibers encircling the
left ventricle subendocardially, so that the mitral valve may be
seen. A similar band on the right is not sketched in. [From
Robb & Robb (136).]
fig. 8. The superior sinospiral muscle as seen from the
antreior surface of the heart. Symbols as in fig. 7. Again the sub-
endocardial layer has been cut through in order to show deeper
structures. The window in the right ventricular wall shows the
fibers from the trabeculated area running up to the anterior
and medial leaflets of the tricuspid valve. In both of these super-
ficial muscles, blood vessels follow the muscle strands as they
encircle the apex. [From Robb & Robb (136).]
fig. 9. The deep sinospiral muscle as seen from the front.
Note the division of the muscle at the posterior inter-ventricular
sulcus, with fibers passing anteriorly to form most of the basal
two-thirds of the septum; these septal fibers lie just distal to
the band of the left head of origin at the base of the aorta.
Symbols as in fig. 7. [From Robb & Robb (136).]
fig. 10. The deep bulbospiral muscle, a powerful sphincter
encircling the left ventricular base and enclosing both the
aorta and the mitral orifice within its sweep. [From Robb &
Robb (136).]
:m ,irt< n
Willi !■ ,
sinister —
fig. 11. Human ventric-
ular myocardium after re-
moval of the superficial
muscle layers (seen from
the caudal aspect). [From
Spalteholz (148).]
spongiosa (spongy network of muscle fibers) is
gradually reduced by the increasing compacta (solid
tissue of muscle fibers). The phylogenetic remainders
of the spongiosa are the muscular trabeculae, which
are only moderately developed in the mammalian
heart and are almost completely replaced by com-
pacta in the bird heart. In this respect the birds
represent the highest functional development. Accord-
ing to Benninghoff (10) the spiral course of the
muscle bundles toward the heart skeleton and the
768 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
Fihrac mua ulares anuli fibro
i :. 1 ob! qua. superl iak-» .id
v 1 :,;;:■ Lllum dl Wnilii
1 rigi mum Aorta
fibrosum :
stnisti niu
Stratum mu
medium [cin ulare
stratum
mus ularc profundui
t'ibrae circularo
?lral i tUI Ei
V'orti x 1 "i.li-
< istium a.-v.-smistmiii
\ 1 1 igi inum fibrosum
---. 1 1' strum
\Fibrac longitudi-
J-> nali - -11. .11
\ profundi
,' Chord
/ " tendim ai
/ — M. papillaris
. Vortex
fig. 12. Course of the left ventricular muscle fibers. Left: preparation of human heart after partial
removal of the superficial and medial muscle layers (seen from the dorsal aspect 1. Riaht: schematic
presentation of the course of the muscle fibers as viewed from the dorsal aspect. [From Spalteholz
(148).]
vortex formation near the apex are more pronounced
in the mammalian than in reptile and bird hearts
(see fig. 1 2, right).
Rushmer (139) points out that the division of the
heart musculature into "sinospiral" or "bulbospiral"
bundles is rather arbitrary and complicates the
functional analysis. He suggests the division of the
ventricular musculature into two groups of myocardial
bundles, the spiral muscles and the deep constrictor
muscles (fig. 13). He states in his unsurpassed descrip-
tion, the "functional anatomical analysis points to
the direction physiological experimental work should
pursue to verify . . . postulations and to obtain
quantitative measurements."
Architecture of the Atrial Myocardium
The atria supply blood to the ventricles through
three mechanisms: /) passively, during the first part
of their diastole, by serving as blood collecting
chambers as long as the atrioventricular valves are
closed by the high ventricular pressure; 2) still
passively, during the second part of their diastole, by
serving as channels to permit the passage of blood
from the systemic or pulmonary veins into the ven-
tricles once the atrioventricular valves are opened;
3) actively, during atrial systole, by contracting and
thereby pushing some blood into the ventricles shortly
before the ventricular myocardium begins to con-
tract. Since usually only a small fraction (10-30%)
of the blood for ventricular filling is actively propelled
by the atrial musculature and the resistance to inflow
into the ventricular cavity is negligible, the normal
atrial myocardium does not need to be thick walled.
The arrangement of the muscle fibers in the atria
is much simpler than that in the ventricles. Two
groups of fibers can be distinguished : /) those which
belong to one atrium only, and 2) those which are
common to both atria (151).
Group 1: The fibers which lie in the wall of each
atrium form muscle rings around the entrance
orifices, i.e., the pulmonary veins in the left atrium
and the coronary and caval veins in the right atrium.
These annular fibers may act as sphincters, possibly
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
769
0ORIENTATION OF MYOCARDIAL FIBERS IN VENTRICULAR WALLS
RIGHT VENTRICLE — •
LEFT VENTRICLE
(|)FUNCTIONAL COMPONENTS OF VENTRICULAR MUSCULATURE
fig. 13. Muscular structures of the ventricles diagram"
matically arranged so as to reveal their functional components.
A: blocks of tissue removed from the walls of the ventricles are
composed of three layers of muscle. The myocardial fibers in
these layers are oriented roughly in the three general directions
indicated by the arrows. B: from a functional point of view,
the ventricles are formed of two sets of myocardial bundles:
a, the internal and external layers of spiral muscle, which en-
close b, the ventricular constrictor muscles. The internal and
external investments of the ventricular chambers are composed
of the same muscle bundles, which are strongly twisted at the
vortex and spiral in opposite directions from the apex toward
the base. [From Rushmer (139).]
impeding, though not completely blocking, the
backflow of blood into the veins during atrial systole.
Looped fibers are also found which run from the
anterior to the posterior segments of the atrioventric-
ular junction, directly beneath the endocardium. At
many places these fibers bulge into the atrial cavities
forming various ridges which are most conspicuous
at the inner walls of the atrial appendages, where they
are named musculi pectinati from their resemblance
to a comb.
Group 2: The fibers common to both atria are less
numerous and lie superficially with respect to the
proper fibers of each individual atrium. They consist
of two thin muscle sheets which extend in a transverse
direction from one atrium to the other. They can be
subdivided into anterior and posterior fascicles. The
muscle fibers of the atria and ventricles are separated
by connective tissue except at one place, known as the
atrioventricular bundle or bundle of His.
The atrial cavity is surrounded by the thin myo-
cardial fibers of both groups arranged in layers which
are partly parallel and partly crisscrossed. The con-
certed action of all fibers is that, upon their con-
traction, they diminish the size of the atrial cavity
and push blood into the region of least resistance, i.e.,
primarily into the ventricles, secondarily into the
venous orifices. In addition to the main atrial cavity,
there is an adjoining cavity formed by the lumen of
the atrial appendage, also called "auricle" because
of its resemblance to a little ear. The function of the
auricles is unknown. Excision of the auricles in various
operative procedures does not influence the circulation
noticeably. Yet one cannot state bluntly that the
atrial appendages have no function at all, since in a
complex system, such as the heart, the function of a
missing part may often be taken over or substituted
by increased activity of other components. The
mere presence of the atrial appendages results in an
increase in the cardiac reserve. According to Benning-
hoff (10) and Bohme (14), the atrial appendages fill
the space which is created within the pericardial sac
during ventricular systole, as the ventricles eject
blood into the large arteries and decrease in size.
During this period the atrial appendages accommo-
date a considerable amount of blood. This blood is
immediately available at the beginning of the rapid
ventricular filling phase to be transferred into the
ventricular cavities.
PRESSURE AND FLOW EVENTS DURINO
THE CARDIAC CYCLE
Historically the cardiac cycle was first divided into
"systole," or period of contraction, and "diastole,"
or period of relaxation of the ventricles. It was soon
recognized that the terms systole and diastole should
refer equally to the atrial contraction and relaxation,
although the ventricular events were most con-
spicuous in the gross observation of cardiac activity.
Since the atrial contraction precedes that of the
ventricle, terminological difficulties arose as to which
systole was meant in describing the time sequence of
cardiac events. As knowledge about the heart's
action increased, it was also deemed necessary to
subdivide the cardiac cycle in greater detail [see also
Mackenzie (no)]. With the advent of methods for
precise pressure recording from the cardiac chambers
and great vessels, the ventricular pressure tracings
77°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
became the deciding guidelines for characterizing the
phases of the cycle. The generally adopted sub-
divisions of Wiggers (156) stem from this era. Since
other landmarks of cardiac activity such as flow,
volume changes, or biochemical processes were
difficult to record adequately, they were only corre-
lated with the pressure curves at a later date.
It is still impossible to subdivide the cardiac cycle
according to the most important physiological
events : the blood flow into and out of the cavities.
The approximate beginning and end of systolic
ejection can be determined from simultaneous pres-
sure tracings in a ventricle and in an arterial outflow
tract. However, the precise timing of flow is only
possible through direct recording of flow either at the
root of the aorta or at the pulmonary artery [see also
Moscovitz & Wilder (117)]. The recent advent of
refined flowmeters will probably necessitate some
adjustments in the original Wiggers scheme of the
cardiac cycle. For the time being it is still preferable
to retain the well-established scheme and to fit
modifications into it, rather than to advocate a com-
pletely new one [see also Horowitz (80)].
Figure 14 [modified from Wiggers (156, 159)]
illustrates in schematic form the sequence of pressure
events during the cardiac cycle in the left ventricle,
left atrium and aorta, and the volume changes in the
combined ventricles [from Henderson (69)]. For
time correlation, tracings of the heart sounds and
of the electrocardiogram are added. This composite
chart is mainly based on curves obtained in animal
experiments.
The cycle is divided into two periods, systole and
diastole. The former begins with the rise of ventricular
pressure caused by ventricular contraction (fig. 1 4, 1 )
and ends at the onset of myocardial relaxation, 4,
at the point when ejection actually ceases. This point
then also represents the beginning of the diastole.
The systolic period is subdivided into 1-2, isovolu-
metric ventricular contraction (50 msec); 2-3,
maximum ventricular ejection (90 msec); and 3-4,
reduced ventricular ejection (130 msec). The diastolic
period is subdivided into 4-5, isovolumetric ventricular
relaxation (120 msec), which includes a phase
occurring just prior to the incisura and formerly
called protodiastole (40 msec), plus the phase formerly
known as isometric relaxation (80 msec); 5-6, rapid
ventricular filling (no msec); 6-7, slow ventricular
filling or diastasis (190 msec); and 7-1, ventricular
filling by atrial contraction (60 msec).
Numerous other cyclical events occur with each
fig. 14. Scheme of the cardiac cycle. Time, totaling 1 sec,
on upper margin. Numbers under lower margin indicate be-
ginning and end of phases. Period of- ventricular systole lasts
from 1 to 4, period of ventricular diastole lasts from 4 to 1 .
Detailed description in text. [Figure (but not numbers in text)
slightly modified from Wiggers (159).]
heart beat. They are correlated timewise with the
phases of the pressure-volume cycle as follows.
1-2: Isovolumetric ventricular contraction. During this
phase the myocardium builds up tension and this
gives a fast rise of intraventricular pressure without
change in the volume of blood contained in the
ventricular cavity. The intraventricular pressure
must rise to the level of the diastolic pressure pre-
vailing in the aorta (or pulmonary artery) before
blood can be ejected from the ventricles during the
next two phases. The term "isovolumetric contrac-
tion" suggested by Rushmer (139) should supersede
the older term "isometric contraction," since at the
beginning of this phase there is an actual shortening
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
771
of the fibers of the papillary muscles and trabeculae
carneae which results in a tension of the chordae
tendineae, and an approximation of the atrioventri-
cular valves (139)- Simultaneously, there is a passive
stretching of the other still relaxed myocardial layers,
mainly those of the outer walls of the heart [see also
Hawthorne (67), Anzola (4), and Burton (29)].
The older term ''isometric contraction'' had the
misleading implication that all myocardial fibers
contract simultaneously and isometrically from the
very start. Since in fact some muscle fibers shorten
whereas others are passively lengthened during this
phase, while the intraventricular volume remains
constant, the term isovolumetric contraction provides
a more accurate description than isometric con-
traction. Apparently instrumentation has not yet been
refined sufficiently to decide whether or not there is
in this phase a brief "latent relaxation" of cardiac
muscle fibers as there exists in skeletal muscle fibers.
The shortening of the ventricle in the longitudinal
axis results in a descent of the atrioventricular
junction which in turn expands the atrial cavities.
This leads to a precipitous lowering of the atrial
pressure (fig. 14) which is often observed even before
' the ventricle ejects blood. The ventricular muscle
fibers contract in a successive order, probably follow-
ing the same time sequence as their depolarization
(75, 142). As a consequence the blood contained in
the ventricular cavity is pushed from the apex region
toward the center of the ventricle and moves thereby
closer to the outflow tract. The subsequent ejection
from the ventricles can be looked upon as a con-
tinuation of the intraventricular movement of blood
which already starts before the semilunar valves open.
At the same time the ventricular cavity changes from a
cylindrical to a more spherical shape, which from
the energy standpoint represents a more economical
way of discharging the ventricular content, once the
aortic diastolic pressure is overcome. As pointed out
by Rushmer (139), the asynchronous contraction of
the ventricular myocardium readily explains the
brief upward deflection at the beginning of iso-
volumetric contraction in the ventricular volume
curve described by Wiggers (156) in fig. 14.
This was formerly interpreted as an artifact in the
recording.
Some arbitrariness is invoked in determining
accurately the start of isovolumetric contraction. In
all pressure tracings the upward movement begins
slowly in the form of a rounded curve. There is no
abrupt beginning, inflection, or break. This becomes
especially evident if one records the pressure events
by drawing out the time axis with fast moving paper
as can be easily done today with electrical recording
apparatus. The rounded beginning of the upward
limb results from the combined effect of a) the
contraction of the papillary muscles, and b) the
simultaneous passive distention of some of the muscle
fibers in the ventricular wall. Whenever the trans-
figuration of the ventricle causes a detectable rise of
intraventricular pressure, then by convention the
ventricular isometric contraction is said to begin. The
fact that the different strands of myocardial fibers
contract in sequence rather than simultaneously may
also explain the great variability of the slopes of the
pressure tracings in the early part of isovolumetric
contraction.
The steepness of the slope during isovolumetric
contraction is predominantly determined by the
forcefulness of the fiber contraction. If the difference
between the end-diastolic ventricular and end-
diastolic aortic pressure remains unchanged, the
duration of the ventricular isovolumetric contraction
is shortened by sympathetic or sympathomimetic
stimulation and lengthened by agents or conditions
which depress the sympathetic control of the heart
[Cotton & Maling (35), Gleason & Braunwald
(54); see also Reeves et al. (133)]. Thus in forcefully
contracting ventricles, the slope will be steeper than
in feebly contracting preparations.
The atrioventricular valves close approximately at
the beginning of isovolumetric contraction; the
opening of the semilunar valves marks the end of this
phase. The precise moment of the valve actuation is
difficult to establish experimentally (discussed in the
section on heart valves). In the interval between
closure of the atrioventricular valves and opening of
the aortic and pulmonary valves, the blood contained
in the ventricular cavities is temporarily isolated from
the fluid columns in the atria and arteries. However,
the ventricular content does not remain still (10).
In fact the blood which rushed into the ventricles at
high velocity during diastole may aid in expanding
the ventricular cavities. Since the inflow is primarily
directed toward the apex, it is this part of ventricular
wall which could be preferentially expanded. As the
papillary muscles and trabeculae carneae begin to
contract, the movement of the blood is deviated
toward the outflow tract. This change in direction of
flow is favored anatomically by the fact that the axis
of the inflow tract and that of the outflow tract form
an angle. In other words, the inflowing blood prob-
ably does not come to a complete standstill in order to
reverse its direction of flow for ejection into the
772
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
arteries, but rather it keeps flowing in a curve from
the main direction of the inflow tract toward the
outflow tract. This translocation of blood within the
ventricle during the isovolumetric phase is energy pre-
serving. In fact, there seems to be rather little turbu-
lence and not always complete mixing of blood during
this "intraventricular" streaming from the inflow side
to the outflow region. This explains why the streamlin-
ing of flow in the venous circulation is not always com-
pletely interrupted by the passage of blood through the
ventricle. For example, the systemic venous blood is
transferred into the pulmonary arteries in such
manner that superior caval blood reaches predomi-
nantly the right lung and inferior caval blood the left
lung [see also Bucher et al. (27)]. Obviously, the
possibility of incomplete mixing deserves attention
when samples of so-called mixed venous blood are
drawn.
How much does the velocity of the blood flow
decrease during the transit in the ventricle? In the
resting organism with a slow heart rate, the velocity
of blood streaming into the ventricle toward the end
of diastole is rather small, as may be surmised from
the fairly flat portion of the ventricular volume curve.
When the cardiac output is elevated, the velocity of
the intraventricular flow during isovolumetric con-
traction will probably increase for two reasons: /)
the velocity of end diastolic ventricular inflow in-
creases through a shortening of diastole due to high
heart rates and through a more forceful atrial con-
traction; 2) the transit time through the ventricle is
shortened by the more powerful and often shorter
myocardial contraction. Such higher intraventricular
flow velocities under sympathetic activity could then
result in a better energy conservation by not letting
the speed of blood flow slow down too much before
ventricular ejection begins again.
[It is the feeling of the editors that there is not
sufficient evidence to show that continued transloca-
tion of blood within the ventricular cavity during iso-
volumetric contraction could contribute significantly
to the subsequent ejection. Ed.]
2-4: Rapid and reduced ventricular ejection. As soon as
the pressure in the ventricular cavities exceeds that
in the aorta or the pulmonary artery, the blood is
suddenly ejected. Although flow is created by a
difference between the intraventricular and arterial
pressures, an inspection of pressure curves alone,
simultaneously recorded from the ventricle and the
root of the artery, furnishes only meager information
about the rate of volume flow and its time course.
However, from simultaneously recorded flow and
fig. 15. Phase relationships between pressure and flow as
revealed by simultaneously recorded curves from the ascending
aorta of a conscious dog. Upper tracings: rate of volume flow
measured with a permanently implanted electromagnetic
flowmeter. Lower tracings: aortic pressures obtained through a
permanently implanted cannula leading to a strain gauge
manometer. A: curves from the quiet reclining animal. B:
curves from the animal running behind a car during moderate
exercise. [Original curves by the courtesy of Frederick Olm-
stead, Cleveland Clinic, Cleveland, Ohio (personal communi-
cation, 1961 ).]
pressure curves in the aorta or in the pulmonary
artery, the process of ventricular ejection is now
fairly well understood [VVetterer (155)]. The ejection
starts abruptly (fig. 15). The blood column in the
root of the aorta, which is practically stationary at the
end of diastole and during isovolumetric contraction,
is rapidly accelerated and pushed toward the periph-
ery. The greatest flow acceleration occurs during
the steeply ascending limb of the aortic pressure
curve, so that the highest flow rate (peak of the
flow curve) is actually reached prior to the summit
of the pressure curve. When the flow then becomes
less rapid, the phase of reduced ejection is said to
begin. The border between rapid and reduced
ejection is quite arbitrary. When only pressure and
cardiometer curves were available [Wiggers (156)],
it was difficult to determine from the gradual leveling
off of the downward limb of the volume curve when
the rapid ejection started to slow down. The summit
of the ventricular pressure curve was thought to
indicate the end of rapid ejection (fig. 14). It is now
known that the flow slows down earlier, since the peak
of the flow curve definitely precedes the peak of the
ventricular or aortic pressure curve (upper tracings
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
77 1
in fig. 15). The fact that the ventricular and aortic
pressures continue to rise even after the flow rate
starts to drop is not surprising, since during the period
of reduced ejection, blood continues to accumulate
in the aortic arch. Because aortic pressure at any one
instant is determined both by the distention of the
arterial walls with blood coming from the ventricle
and by the runoff into the periphery, the pressure
continues to rise as long as more blood enters the
aortic arch than runs off toward the periphery. There
is no fixed and easily definable time relation between
the summits of the flow and pressure curves, because
the factors determining the position of each of the
summits are numerous and variable (149).
Some aspects of the time relation between the flow
and pressure curves A and B are illustrated in figure
15. During exercise the rapid ejection occupies a
relatively shorter portion of the total ventricular
ejection. In this example the whole ventricular
ejection lasts 374 msec at rest and only 234 msec
during exercise. However, the delay between the
summits of the flow and pressure curves is 109 msec,
in both cases, or 29 per cent of the whole ventricular
ejection at rest, and 46 per cent during exercise. In
other words, during exercise the aortic pressure
continues to rise relatively longer after the aortic flow
rate has started to drop, indicating that there is
relatively more blood accommodated in the central
arteries (arterial compression chamber) during
systole. This also means that, when ventricular
ejection ceases, there is a higher aortic pressure and
consequently a larger amount of peripheral runoff
during early diastole. Such conditions help to main-
tain greater tissue perfusion in the active organism.
The configuration of the aortic flow curve is not
the same in the organism at rest and during exercise
or sympathetic stimulation. At rest, the descending
limb of the flow curve first declines gently, then
progressively faster, forming thereby a shallow hump
(see fig. 15.4). During exercise the ascending limb is
steeper and the descending limb declines precipi-
tously (fig. 15B). This pattern indicates: /) an increase
in the myocardial contractile force, which continues
to exert its strong effect after the end of the iso-
volumetric contraction phase and achieves a more
rapid flow acceleration; 2) a longer duration of flow
at near maximal velocity (Olmstead, personal
communication); and 3) a faster return to the begin-
ning of myocardial relaxation. Despite shortening of
the ejection phase, the stroke volume, which can be
calculated from the area under the curve, is approxi-
mately the same during moderate exercise as it is at
rest. However, during strenuous exercise (running of
dog at 16 miles per hour over rough terrain) the
stroke volume appears to be increased by approxi-
mately 25 to 40 per cent (Olmstead, personal com-
munication). Whether or not there is always an
increase in stroke volume during exercise is still a
matter of debate among various investigators [see also
Rushmer (139)].
Toward the end of the reduced ejection phase the
intraventricular and aortic pressures drop quickly.
The ejection stops and forward flow in the ascending
aorta ceases shortly after closure of the semilunar
valves as seen by the crossing of the flow curve through
the horizontal zero flow line in figure 15. Flow in the
root of the aorta near the valves momentarily reverses
its direction, because of a translocation of blood into
the sinuses of Valsalva and the coronary vessels,
which helps to close the aortic valves. Although there
is a brief backflow near the valves, in the more
distal part of the aorta the flow continues forward for
a while, since the energy momentarily stored in the
distended aortic arch propels the blood to the area of
lower pressure, i.e., the peripheral vessels (compres-
sion chamber effect). The precise moment of valve
closure cannot be easily correlated with the pressure
and flow curves. It can be stated from the flow curve
that the valves must have closed at least by the time
when the downward deflection is suddenly stopped
(fig. 1 5) and after which blood again is propelled
forward in the ascending aorta. How much blood is
regurgitated into the ventricles and how much flows
into the coronary arteries while the valves are in the
process of closing, has not yet been determined.
4-5: Isovolumetric relaxation. It is also difficult to
establish the exact moment when the myocardial
fibers start to relax after maximal shortening. Wiggers
(156) took the steepening of the decline in the ventric-
ular pressure curve prior to the deepest point of the
aortic incisura as the beginning of the relaxation
process and referred to the brief interval from the
beginning of muscular relaxation to semilunar valve
closure as the protodiastolic phase. Since little is
gained by singling out this interval, which cannot
be accurately measured, the phases of protodiastole
and isovolumetric relaxation will be treated here as a
single process as Wiggers ( 1 59) also suggested in his
recent discussion on this subject.
It appears reasonable to assume that, just as the
contraction began asynchronously, some myocardial
fibers will begin to relax earlier than others. However,
no direct measurements are available to document
this hypothesis. At the end of isovolumetric relaxation,
774
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 1 6. Phase relationships between aortic
pressures, Ao; left ventricular volume, LV,
(dots); atrial volume, LA, (open circles);
and electrocardiogram, ECG, in an anesthe-
tized normal dog with a spontaneous heart
rate of iio/min. The pressure tracings were
simultaneously recorded and correlated with
the volume measurement from the kinemato-
graphic frames. [Original curves and labeling
by courtesy of Peo Gribbe, Wenner-Gren
Research Laboratory, NorrtulPs Hospital,
Stockholm, Sweden (personal communica-
tion, 1 961).]
PROTODIASTOLE
EJECTION
ISOMETRIC
CONTRACTION
ISOMETRIC RELAXATION
RAPID INFLOW
DIASTASIS
/ ATRIAL SYSTOLE
ORMOTHERMIA 36C
EART RATE 110
0.5 sec
the intraventricular pressure drops to the level of
the atrial pressure. The pressure decline, like the
pressure rise, is more rapid under the action of
epinephrine (158, 124) and apparently also in
exercise. Therefore, with epinephrine and during
exercise the duration of the isovolumetric relaxation
phase is shorter for the same pressure difference
between the incisura and atrial pressure. Neither the
precise moment of the valve opening at the end of
isovolumetric relaxation nor the pressure difference
necessary to actuate them has been satisfactorily
determined as yet. One usually takes the decline in
atrial pressure after the V point as an indication that
the atrioventricular valves have just opened and flow
through the orifice has begun. The crossing over of the
atrial and ventricular pressure tracings is therefore
taken as the end of isovolumetric relaxation. How-
ever, even with the most careful recording, it is
difficult to establish such a crossing over without
artifacts. Since at this part of the cardiac cycle the
heart has become a low-pressure system, instru-
mentation errors are commonly experienced with
positioning of pickup devices, movement artifacts, lack
of sensitivity, Bernoulli effect, and lack of a common
and reliable reference zero pressure level.
5-7 : Rapid and slow ventricular filling. There is no
satisfactory procedure to measure directly the inflow
of blood from the atrium into the ventricle. The best
information stems from X-ray kinematographic
studies such as those of Rushmer (139), Chapman
et al. (31, 32), Gribbe et al. (56). A good time resolu-
tion was obtained by Gribbe, who took 40 to 50 frames
per sec using an image intensifier. The individual
frames of film were projected and the volume was
calculated from the contrast silhouette of the left
atrium and ventricle, assuming that the left ventricu-
lar cavity resembles an ellipsoid of rotation. Figure 16
illustrates the steep upward slant of the curve during
the phase of rapid ventricular filling. The incline of
this part of the curve is even steeper than the decline
of the curve during rapid ventricular ejection,
indicating that blood actually rushes into the ven-
tricle faster than it is ejected from the ventricle. This
observation has an important bearing upon the
concepts of the forces which bring about ventricular
filling (see later section). After the rate of ventricular
inflow has reached its maximum, it begins gradually
to slow down until finally the curve tends to level
off. There is no distinct break which could serve as a
criterion for precise determination of the end of the
rapid filling phase and the beginning of the slow
phase. Nevertheless, the distinction between these
two phases remains useful at slow heart rates, as for
instance under strong vagotonic influence, because
the slow phase of ventricular filling then lasts much
longer than depicted in figure 16.
7-1 : Filling by atrial contraction. With the contraction
of the atrial myocardium an additional volume of
blood is pushed into the ventricle, as shown by the
sudden final incline of the curve in figure 16. The
contribution of atrial contraction to ventricular
filling has been much debated [see Mitchell et al.
(113)]. According to the measurements of Gribbe
et al. (56), it should amount to about 20 to 25 per cent
of the volume entering the ventricle. Atrial pressure
drops after the peak of atrial systole but seemingly
without a measurable decrease in ventricular volume
by backflow through the atrioventricular valves. 11
FUNCTIONAL ANATOMY OF CARDIAC PUMPINC
775
ventricular pressures are recorded with instruments of
sufficient sensitivity, the transfer of the atrial pressure
rise can be observed on the ventricular pressure trac-
ing, since atrium and ventricle form a common cavity
during atrial systole. Similarly, after the peak pressure
of atrial systole has been reached, the pressure drops
not only in the atrium but also in the ventricular
cavity. When the atrioventricular valves actually
close is still a matter of debate (see also later section).
It may well be that the large valve leaflets begin to
approximate each other at the moment when the
atrial pressure starts dropping and that they continue
to move toward each other because ventricular blood
flows into the large spaces behind the closing leaflets.
Therefore, the valves may start to close at a time
when the ventricular pressure is decreasing slightly.
Complete closure would then be achieved when the
dropping atrial and intraventricular pressures level
off (Z point in atrial pressure curve).
As the heart rate becomes faster under sympathetic
stimulation or in exercise, the period of slow ventric-
ular filling is progressively shortened by an earlier
onset of the atrial systole [see Mitchell ct a/. (113)].
At heart rates above 1 20 per min the phase of slow
ventricular filling is more or less abrogated (56).
Figure 17 illustrates how at a heart rate of 160 per
min the phase of rapid ventricular filling is directly
followed by the inflow due to atrial contraction. In
these curves one cannot discern the usual hump in the
upstroke of the filling curve which occurs when rapid
ventricular filling changes to slow ventricular filling
before the atrium adds its contribution. It is likely that
at still faster heart rates, the atrial component of the
curve blends completely with the inclined tracing
characteristic of rapid ventricular filling. At extreme
degrees of tachycardia even the phase of rapid ventric-
ular filling may be encroached upon. This would
explain why the stroke volume decreases at very-
rapid heart rates since there is not sufficient time
for adequate filling of the ventricle.
The force of atrial contraction usually varies
concomitantly with that of ventricular contraction.
Therefore, the percentile contribution of atrial
systole to ventricular filling is lower under vagal
influence and higher under sympathetic excitation.
In extreme tachycardia, when atrial systole begins
during the phase of rapid ventricular filling, the
actual contribution of the atrium to the filling of the
ventricle may be as high as 30 to 40 per cent. The
atrial contraction would then serve to increase the
pressure difference between the atrium and ventricle
in the later part of the rapid filling phase and thereby
0.5 sec
Fig. 17. Phase relationships among aortic pressure, [eft
ventricular volume (dots), atrial volume (open circles), and
electrocardiogram (ECG), in an anesthetized normal dog with
a spontaneous heart rate of 160/min. The pressure tracings
were simultaneously recorded and correlated with the volume
measurement from the kinematographic frames. [Original
curves and labeling by courtesy of Pco Gribbe, Wenner-Grcn
Research Laboratory, Norrtull's Hospital, Stockholm, Sweden
(personal communication, 1 96 1 ).]
produce a maximal velocity of inflow throughout the
entire, though brief, phase of rapid ventricular
filling.
CORRELATION OF OTHER CARDIAC EVENTS
WITH THE CARDIAC CYCLE
The time sequence of cardiac events originally
described by Wiggers (156) was based upon those
pressure changes in the circulatory system which were
measurable at the time. However, during the last
three decades our trend of thinking about cardiac
events has been greatly affected by the progress of
electrocardiography. At the very beginning of the
investigations of the field (about 1910-1920) the
electrocardiogram could only be correlated second-
arily with the time course of the more easily meas-
urable pressure events (138). Nowadays one can
record electrocardiograms with a higher degree of
time resolution than intracardiac pressures. For this
reason it is rather common to use the electrocardio-
gram as the basis or guideline for dividing the cardiac
cycle into phases and then to fit secondarily the
pressure and flow events into the patterns of the
electrical events (17, 157). However, there is a varying
time lag between electrical and mechanical events
under different experimental conditions [see Luisada
& Liu (104)], so that such a correlation system is not
entirely satisfactory.
77(»
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Atrial Pressures
The recording of atrial pressures is beset with
considerable experimental difficulty, as is the case
with all fast changing phasic events in low pressure
s\ stems. The finer details of atrial pressure contours
are therefore often affected by artifacts from impacts
or vibrations which make it difficult to arrive at
accurate deductions concerning atrial flow dynamics.
Indeed the atrial pressure pulse contour depicted in
figure 14 is highly schematized.
Atrial pressure begins to rise at the onset of atrial
systole (A wave). Since at this moment the atrium
and ventricle form a common cavity, the height which
the atrial pressure attains is influenced by the volume
distensibility characteristics both of the atrium and
of the ventricle, in addition to the rate of translocation
of the fluid from one part of the cavity into the other.
The resistance to flow through the normal atrio-
ventricular orifice is so low that it cannot be measured
with presently available techniques.
Little is known about the synchronicity or asyn-
chronicity of the contraction of atrial muscle fibers.
The excitation wave spreading over the atrial walls
proceeds from the sino-atrial node. Therefore it can
be assumed that the contraction, which follows
depolarization after a brief interval, similarly pro-
ceeds in a wave. The concept of asynchronous atrial
muscle contraction is based on the premise that the
delay between depolarization and contraction is the
same for all atrial muscle fibers. It is not certain that
diis is the case. A contractile wave is difficult to
demonstrate conclusively and consistently by means
of slow motion pictures.
The drop in atrial pressure, after the pressure has
reached a peak during atrial contraction, is probably
caused by the beginning of atrial muscle fiber relaxa-
tion. Again, it is not possible to state whether this
occurs synchronously or in a sequential order and to
state what effect diis process has on the flow dynamics.
From the configuration of the atrial pressure curve
one cannot necessarily infer the exact onset of atrial
muscle relaxation. Just as the ventricular pressure
curve can still rise although the rate of ejection already
declines (see fig. 15), the atrial pressure could well
start to decrease before or after the relaxation in the
atrial musculature actually begins. The convention of
calling the leveling off of the declining atrial pressure
curve the "end of atrial systole" is also arbitrary
[Opdyke el al. (126); Opdyke & Brecher (125)].
There is often a brief period (Z point) during which
the atrial pressure curve remains level after its decline
from the systolic rise. This is the last moment at which
the atrial and ventricular cavities are probably still
in communication before complete closure of the
atrioventricular valves. The atrial Z point pressure
is almost equal to the ventricular end-diastolic
pressure because the rate of ventricular inflow has
become minimal at this moment. Therefore it is
fairly safe to take the Z point as a representative of
end-diastolic ventricular pressure. It is definitely more
accurate to use the Z point than the mean atrial
pressure, which depends upon numerous factors
unrelated to the end-diastolic ventricular pressure,
such as integration of artifacts and peaks at the A, C,
V points ( 126).
After the Z point the atrial pressure often rises
briefly and precipitously (C wave). This pressure
rise, frequently accompanied by vibrations, is ascribed
to the bulging of the atrioventricular valves into the
atrial cavity during ventricular isovolumetric con-
traction. Immediately following the sharply peaked C
wave, atrial pressure usually declines to a level
corresponding to atmospheric zero in an open-chest
preparation (see fig. 14), or to near-zero transmural
pressure in a closed-chest organism. It is believed that
this pressure drop is caused by the pull of the papillary
muscles on the atrioventricular valve leaflets and by
the descent of the atrioventricular junction which
suddenly enlarges the atrial cavity. The bottom of the
pressure drop is called the X point (or wave). There-
after atrial pressure rises slowly up to the V point (or
wave) located at the end of ventricular isovolumetric
relaxation. The pressure rise from the X point to the
V point is probably caused by an inflow of blood
which distends the atrial walls. The atrial pressure
drop (V point or wave) after the opening of the
atrioventricular valves results from the rapid transfer
of blood into the ventricular cavity in which a lower
pressure prevails. While it is assumed that the actual
opening of the atrioventricular valves occurs at the
summit of the V wave, there is some debate whether
or not it occurs slightly afterward [see also Xixon
(120)].
A minor change in the conventional labeling of
atrial pressure tracings has been used by Kaplan (88).
Without mentioning the Z point, he refers to the
small decline in pressure which frequently follows
atrial systole, before the C wave, as the X wave. Then
he designates the pressure decline after the C wave
as the X1 wave. There is no conspicuous advantage
to this svstem of notation.
Electrocardiogram
Since the electrocardiogram is easily obtainable and
serves as an important diagnostic tool, considerable
efforts have been made by theoretical and clinical
scientists to correlate timewise electrical and mechani-
cal events of cardiac contraction. Nearly half the
large volume of the Physiology of the Heart, by Schiitz
(144), is devoted to this subject and should be referred
to for detailed information.
From the standpoint of cardiac pumping, the
electrocardiogram is principally of interest insofar as
it may furnish a convenient method of determining
precisely the course of mechanical events without
resorting to surgical interventions, cannulations, etc.
Present attempts along these lines are still inadequate.
They are also theoretically limited for the following
reasons: a) although the electrical event always
precedes the mechanical event, it is not known
whether the time intervals between depolarization
and beginning of contraction are identical in all
heart muscle fibers, b) There may be differences in
the rates of impulse propagation along various fibers
resulting in varying rates of contraction once de-
polarization has started at one point, c) As repeatedly
emphasized by Rushmer (139) and Scher (142),
there is considerable mechanical asynchronicity in
the contraction of cardiac muscle fibers, which cannot
be completely unraveled by recording the over-all
electrical changes from a large mass of tissues such as
the heart, d) There are possibly time differences
between the right and left heart depolarizations and
contractions, but these differences are inconstant
and change with various factors.
An example of the difficulties encountered in
attempting to establish empirical time correlations in
the cardiac cycle is shown in figure 18. Using normal
anesthetized dogs, Gribbe et al. (57) studied the
volume changes in the cardiac chambers with
cineradiography and timed the events with the
conventional electrocardiogram. Comparison of the
time relations in figure 19 with those shown in
figures 14 and 18 reveals considerable discrepancies.
Vibrocardiogram (Apex Cardiogram)
The intracardiac pressures are not easier to corre-
late with the electrocardiogram than with a number
of other mechanical events [e.g., Harrison et al.
(65)]. For instance, the classical mechanocardio-
gram or apex cardiogram, now often referred to as
precordial vibrocardiogram, offers a good example of
the present limitations in the description of the
FUNCTIONAL ANATOMY OF CARDIAC PUMPING 777
I 2 3 4 5
RIGHT
-v
LEFT
.,...,..,.,. .. ... 1 ,, j..
■I I I II ! I I I
10
+4-
fig. 18. Schematic presentation of the relationship between
electrical and mechanical events. Heart rate 120 beats/min.
The markings in lower part of the figure indicate the picture
frequency at an exposure rate of 48 frames/sec. 1 , Onset of
right atrial contraction; 2, onset of right ventricular contrac-
tion; 3, onset of right ventricular ejection; 4, end of right
ventricular ejection; 5, onset of right ventricular filling; 6,
onset of left atrial contraction, 7, onset of left ventricular con-
traction; 8, onset of left ventricular ejection; g, end of left
ventricular ejection; 10, onset of left ventricular filling. The
striped areas represent the phase of ventricular isovolumetric
contraction. The stippled areas represent the phase of ventricu-
lar ejection. [From Gribbe el al. (57).]
cardiac cycle. Figure 1 9 shows a composite chart made
up of superimposed schematized tracings as they are
observed in normal man. The chart has been con-
structed by Agress et al. (2) from the most appropriate
tracings which they could find in the recent literature.
The authors state: "Although this is a composite
graph, an accuracy of 0.005 second per scale division
was made possible by using a simultaneouslv in-
scribed electrocardiogram as the time base." It is
evident that claims for an accuracy of 5 msec can be
only referred to the electrocardiogram, since the
time definition of intracardiac pressure recording
through long catheters is usually poorer than 0.005
sec. This fact is not pointed out in criticism of the
well-deserving attempts to correlate the various
events of the cardiac cycle in man, but only for the
purpose of cautioning against hasty conclusions.
Agress et al. (2) divide the cardiac cycle into
phases, which differ slightly from those customarily
accepted in the past. The curves (fig. 19) are inter-
sected by vertical lines based on the time relation of
left atrial and left ventricular pressures, as indicated
by the upper margin band, /., of the graph and the
small heart schemes above it. The phases are labeled
778 IIWDBOOK OF PHYSIOLOGY ^> CIRCULATION II
a £r o
FIG. 19. Compos-
ite graph of the
events of the cardiac
cycle in the human
heart. For discus-
sion see text. [From
Agress et ai (2).
PHHBBBH
] H»H[ STSTOlt
P s
ISO CON
RJPIO
IECTI0"
REDUCED
lEClICK
P D ISO »IL
RAPID F
IllINQ
onsiisis
45
J 1
™ 1
150
mm fi j
30
J/o
iter
100
UVo
50
TV
b
/ /'
'e
j
W°
i
' ' - u
FUNCTIONAL ANATOMY OF CARDIAC PUMPINC
779
in the next lower band, where P.S. stands for proto-
systole and P.D. for protodiastole. The lowest band
of the upper margin, R, illustrates the time difference
between the activities of the right and left heart. A
number of points warrant brief comments. For all
intracardiac and arterial pressures a common zero is
used. The "low" pressure events (left atrium, right
ventricle, and pulmonary artery) are plotted on a
scale from o to 45 mm Hg (left), whereas the "high"
pressure events (left ventricle and aorta) are graphed
on a scale from o to 150 mm Hg (right). The use of
two different pressure scales for correlating simul-
taneous events on the same time basis results in
different slopes. The casual viewer may hastily con-
clude that the rate of pressure rise during isovolu-
metric contraction is greater in the right ventricle
than in the left ventricle, which is actually not the
case. Correspondingly, the pressure drop during
ventricular isovolumetric relaxation appears to occur
faster on the right than on the left side, which does not
happen either. The ascending limbs of the aortic and
pulmonary arterial pressure curves have been obtained
from overdamped recording systems, such as often
happens with long catheters. This might explain
why the tracings of the arterial pressures have gentle
slopes and remain considerably below the summits of
the ventricular pressure curves. An interesting
innovation is the protosystolic phase (P.S.) which
apparently extends from the leveling off of the left
atrial A wave until the beginning of the left ventricu-
lar pressure rise. It seems to correspond to the con-
ventional Z point. According to the tracings in figure
19, the protosystolic phase appears to be timewise
closely related to the electrocardiogram. It seems to
last from the beginning of the Q wave until the tip of
the R wave. The usefulness of introducing this
distinctly different phase seems to lie in the easy
correlation of certain characteristics of the phono-
cardiogram and vibrocardiogram with the cardiac
cycle during this time interval.
A number of other cyclical events occurring in the
circulatory system also can be more or less accurately
correlated with the cardiac cycle. Examples would
be: various peripheral arterial and venous pressure
pulse curves as well as flow pulse curves; intra-
myocardial pressures (96); the ballistocardiogram;
electrokymogram (42, 163, 135); angiokymogram
( 147); angiocardiogram; cardiorheogram (3, 68); and
heart sounds [Luisada et al. (105)]. In all cases the
previously mentioned difficulties in precise timing
must be given serious scrutiny. A discussion of all
events which lend themselves to correlation would
exceed the scope of this chapter.
FUNCTION OF THE HEART VALVES
Valves are essential for efficient action of all
reciprocating pumps in order to maintain unidirec-
tional flow. Valves must offer a minimal impedance
to flow, yet be able to close abruptly with minimal
leakage and minimal displacement. The heart dif-
fers from a mechanical pump in that a perfect seal
must be obtained in orifices which are continuously
changing in shape, size, and position throughout
the cardiac cycle. Therefore, the valves must be
somewhat larger than the area to be covered in order
to remain competent under all normal working
situations. The heart valves are also located in
orifices beyond which the blood enters wider
chambers. This provides for a rapid stream along
the axis of the orifice, with decrease in lateral pressure
at, and just beyond, the restricted valvular plane
(Bernoulli effect) and possibly the production of
eddy currents. This mechanism keeps the valves
floating in the blood stream and insures rapid ap-
proximation of the valve leaflets as soon as the axial
stream of blood ceases.
The movements of the atrioventricular and semi-
lunar valves are passive since the valve leaflets do
not contain muscle fibers [see also Moritz (116)].
The consideration of this aspect is important because
it is easier to replace passive structures by protheses
than it is to create protheses for active structures
such as muscles. It is also possible to investigate the
forces involved in the movement of passive structures
by observing the function of a prothesis which simu-
lates a natural organ in suitable physical analogues.
Davilla (36) summarizes this trend of thought as
follows: "The most successful protheses have been
those which fulfill a passive role in the functional
complex: a metal plate on the skull, a nail in a long
bone, steel or plastic mesh in a weak abdominal wall,
a tube of cloth to replace a blood vessel. . . . Fortu-
nately, the role of the cardiac valves in hemodynamics
is a passive one. They are not parts that move but
parts which are moved. Their role is identical to that
of a simple check valve. It is their environment which
complicates the matter. They are immersed in flowing
tissue which is chemically unstable but which must
not be subjected to extreme unbalance; which pos-
sesses a clotting mechanism that must not be activated
by the valve; and which transports vital cells that
780
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
must not be traumatized. The valves must open and
close an orifice which continuously changes in size
and provides scant room for support between pul-
satile and irregular chambers. The motion of the
valves depends upon cardiac action and must be
coordinated with those structures which impart it.
And, finally, the valvular apparatus must withstand
the stress of beating more than forty million times a
year during the life span of the subject."
In a multichambered pump such as the heart,
valves or valve-like mechanisms are expected at each
boundary between functionally separate chambers.
We shall therefore discuss the function of the struc-
tures which prevent backflow: a) at the veno-a trial
junction; b) at the atrioventricular junction; c) at
the ventricle-arterial junction.
I < no- A/rial Junction
In adult mammals the sinus venosus is completely
incorporated in the atrium and there are only
remnants of "valves" which are incapable of prevent-
ing backflow into the superior caval vein (valvula
Eustachii) or in the coronary sinus (valvula Thebesii).
In the left atrium there is not even the most rudi-
mentary valve to prevent backflow in the pulmonary
veins. Nevertheless, there are some structures which
may contribute to prevent backflow from the atria,
such as circular muscle fibers around the pulmonary
veins and the coronary sinus, and a complex system
of more or less discrete muscular bundles around the
orifices of the caval veins. It has been speculated that
these fibers contract very early in atrial systole and
that by a narrowing of the orifices a valve-like or
sphincter-like action occurs [see also Kjellberg &
Olsson (91), Burch & Romney (28), Campeti et al.
(30)]. This would prevent or at least diminish back-
flow of blood from the atria into the venous trees at
the beginning of atrial systole. Although backflow
from the atria into the caval veins can be frequently
observed and quantitatively registered with flow-
meters (18) and contrast media (14), the amount of
backflow is surprisingly small as compared to the
amount of blood simultaneously pushed by the atria
into the ventricles. The precise mechanism by which
backflow is kept so small is still enigmatic (144). An
interesting light has been cast upon this problem
recently by Little (101). He determined pressure-
volume curves in the left atrium of dogs during tem-
porary ventricular asystole. His findings suggest that
upon a slight rise in atrial pressure above the pul-
monary venous pressure there is a closure of the
pulmonary veins near their atrial junction. This
closure, apparently brought about by collapse of the
vein in a critical region, prevents regurgitation of
blood from the atrium into the pulmonary bed.
However, at high atrial pressures the closed segment
opens and blood flows into the pulmonary veins.
Atrioventricular Valves
The atrioventricular valves are funnel-shaped
structures inserted on a fibrous ring. They are de-
veloped as an ellipsoidal diaphragm and separated
by commissures into somewhat independent cusps,
the edges of which delineate the valvular orifice
(37, 38). The commissures do not extend all the
way to the valve ring (see fig. 3). Traditionally, one
distinguishes two cusps on the mitral valve and three
on the tricuspid valve, although both valves es-
sentially consist of two large opposite cusps and a
variable number of small intermediate cusps at
each end of the ellipse [see also Rusted et al. (140)].
The strands of collagenous fibers known as chordae
tendineae extend from the papillary muscles either
to the free edge of the cusps (first order chords) or to
a few millimeters beyond the edge (second order
chords), or even quite far back into the substance of
the valve through a kind of ''goose foot" forked
insertion (third order chords). The anatomy of
papillary muscles is quite variable. One usually
recognizes in the right ventricle three groups of
papillary muscles to which the tricuspid valve is
fastened, whereas there are usually two such groups
to fulfill the corresponding function in the left ven-
tricle. The chords are of unequal length, so that
probably the same tension is exerted on each at the
time the valve closes. The chords from adjacent
regions of opposite cusps are inserted on the same
or adjacent papillary muscles, in order to insure
leakproof closure [see also Brandt (15), Hubacher
(82)].
The exact mechanism of closure of the atrioven-
tricular valves has been the subject of much debate
[Kantrowitz et al. (87)]. The old theory of closure
mainly by active contraction of the papillary muscles
has been abandoned, and the role of active contraction
of muscular fibers at the base of the valves just after
atrial systole is taken as either minor, or nonexistent
[see also Little (100)]. The decisive factor is prob-
ably the onset of ventricular contraction, which
establishes a higher pressure in the ventricle than
in the atrium. It can be shown that whenever the
ventricles begin to contract, there is a retrograde
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
78.
flow of blood toward the atria which catches the
valves like a pair of sails and flings them into apposi-
tion. As pointed out by Rushmer ( 1 39) this mechanism
inevitably involves a leak before the orifice is closed.
The occurrence of such regurgitation is widely
acknowledged when the atrioventricular valves are
closed by a ventricular systole which is not preceded
by atrial contraction, i.e., "premature ventricular
contraction" [see also Paul et al. (128)]. Since the
normal wave of excitation propagated by the Purkinje
fibers enters the ventricular myocardium over the
endocardial surface, at the roots of the papillary
muscles, the early contraction of papillary muscles
draws the valve edges toward the apex and thereby
produces some shortening of the ventricular chamber
with a resulting passive lateral displacement of the
ventricular walls. During ventricular ejection, the
decrease in ventricular volume is accompanied by a
further shortening of the papillary muscles, taking
up any slack in the valves which might develop.
It has been suggested that under normal condi-
tions, the valves approximate, i.e., begin to close
before the onset of ventricular contraction, although
it is difficult to substantiate this view by chrono-
logically precise measurements. Many explanations
have been advanced for this: passive, upward move-
ment of the valves at the end of diastole caused by
retrograde flow of blood along the ventricular wall,
or by elastic recoil of the ventricular wall to the
strain of atrial systole (8); eddy formation beyond
the valves during atrial systole (102); or develop-
ment of a wave of negative pressure as atrial inflow
abruptly ceases [Henderson & Johnson (70)].
Through these mechanisms, the atrioventricular
valves are approximated or almost closed just before
ventricular systole. Then, when the ventricular
myocardium contracts, the valves are completely
closed to prevent backflow into the atria.
The concept of a presystolic approximation of the
atrioventricular valves received considerable impetus
when Dean (37) succeeded in obtaining direct re-
cording of the valve movement in an isolated, sur-
viving heart. Dean demonstrated that when the
interval between atrial and ventricular systole is
sufficiently long, there is indeed a rapid movement
of the valves toward the atrium at the end of atrial
systole, followed by a second period of separation of
the cusps just before the onset of ventricular systole.
When the period of ventricular filling is shorter (at
faster heart rates) there is no time to observe sepa-
rately the effects of atrial systole and of ventricular
systole. Then there is '"only a single closure move-
ment beginning before ventricular systole, a single
movement due in part to auricular contraction and
in part to ventricular contraction" (37).
The extent to which the valves and their attach-
ments move in the intact organism has been recently
questioned by Rushmer (139). Having observed that
exposed or excised hearts tend to shrink, he and his
associates surmised that the valves might have much
less slack under their normal operating conditions
than reported by previous investigators. They used
cinefluorography to observe the movements of the
mitral valves which had been marked with tiny metal
clips at a previous operation. These studies demon-
strated that the excursion of the valves, at least where
the metal clips were placed, toward the atrium is
remarkably small, and pointed to a more or less con-
tinuous restraint by the chordae tendineae.
Arterial Valves
The aortic and pulmonary valves consist of three
symmetrical cusps attached, similarly to suspension
bridges, around the circumference of the valve orifice
(see fig. 3). When the cusps are approximated they
form a starlike figure; when open, they delineate a
nearly rounded but still somewhat triangular orifice
of an area slightly smaller than that of the artery. At
the tip of each triangular valve leaflet, where the
three valve leaflets come in contact, there is a dis-
crete thickening called the nodulus Arantii. There
are also thin, membrane-like structures (lunulae)
on the free edges of the valves on either side of the
noduli. Normally the free edges come into contact
surface against surface rather than border to border.
Toward the valvular orifice, the ventricular muscula-
ture assumes the shape of a funnel (conus arteriosus),
whereas bevond the valves at the origin of the aorta
and pulmonary artery there are three outpouchings
which provide some free space behind the valve
cusps even when they are maximally opened.
The mechanism of action of the arterial valves can
be described as follows. During ventricular ejection,
the blood stream opens the valves from below and a
rapid flow is established along the axis of the valvular
orifices. However, the valves are not pushed flush
against the arterial wall. On the contrary, eddy cur-
rents, generated by the axial jet of blood, swirl in the
spaces behind the cusps. Indeed, the action of turbu-
lent eddies is such that the faster the ventricular
ejection, the closer to the center of the axial stream
the valve edges are brought [Hochrein (76)]. Thereby
the valves are prepared to close almost instantane-
782
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
ously as soon as ventricular ejection ceases. It is
likely that contractions of muscle fibers in the conus
arteriosus tend to make the valve rings more narrow
during ejection, as discussed in an earlier section. In
this manner axial velocity of flow is increased and
turbulence, which may prepare the valves for closure,
is enhanced. At the end of systole when ejection
ceases, the forward movement of blood in the root
of the artery continues for a very brief period. Then,
the action of the eddies on the upper surface of the
valves prevails over the force exerted from below
(rapidly falling intraventricular pressure). Hence
the retrograde surge of blood toward the ventricle
(see fig. 15) is arrested by valve closure, and marked
pressure differences between the relaxing ventricle
and the elastically distended aorta can develop.
VENTRICULAR AND ATRIAL VOLUMES
IN VARIOUS ACTIVITIES
Certainly one of the most important features of a
pump is the volume which can be propelled per
stroke. This is easy to measure in a mechanical pump,
but it requires complex and sophisticated instru-
mentation to determine the stroke volume of the
intact heart. Successful attempts in this direction have
recently been reported, e.g , by Rushmer ( 1 39) and
his school, by Hawthorne (67), and by Olmstead
(personal communication). Nevertheless, numerous
questions remain still unanswered, namely, a) the
quantitative correlation of the stroke volume with the
other parameters of cardiac activity, and b) the
relationships of the ventricular stroke volume to the
volumes remaining in or passing through the atria,
the ventricles, and the large vessels. At the present
time most of these questions can only be approached
under highly controlled situations which limit the
significance of the experiment. Discrepancies are
therefore encountered depending upon the method of
approach used. At this point, it must also be re-
marked that heart volumes have traditionally been
measured by X-ray or cardiometer techniques which
include the volumes of the walls. Only recently have
radiopaque dyes and other media been developed
which permit measuring the content of the cardiac
cavities and not their over-all volumes (56). Conse-
quently a large part of the data incorporated in the
literature require critical attention. In this chapter,
the word "volume" refers exclusively to the liquid con-
tent of the cardiac cavities and excludes the volume oc-
cupied by the walls and by the blood-filled vessels or
channels in the walls.
I 'entricular Volume
To describe the changes in ventricular volumes
under dynamic conditions, it is advisable to review
the modern terminology introduced bv Rushmer
(139). This terminology is illustrated and somewhat
expanded in figure 20 by drawing a parallel between
the familiar lung volumes (left) and the ventricular
volumes (right).
The stroke volume of the organism at rest corre-
sponds to the tidal volume of respiration. In exercise
the ventricle can also eject some of the blood which
at rest would remain in the ventricular cavity at the
end of systole. Rushmer suggests the term "systolic
reserve volume" for that additional amount of blood
which is not ejected under resting conditions, but can
be maximally ejected with a more forceful contrac-
tion. This corresponds to the expiratory reserve
volume of the lungs. The volume of blood left in the
ventricle after a normal systole used to be called
"residual volume." Rushmer restricts the term
residual volume to that amount of blood remaining
in the ventricle after maximal ejection. Then the
term corresponds truly to the lung residual volume.
The ventricle can also increase its stroke volume by an
augmented venous return during diastole and sub-
sequent ejection of this extra volume in addition to
the resting stroke volume. The term "diastolic
reserve volume" defines the maximal amount of
blood which the ventricle can receive and then eject
in addition to the normal diastolic inflow. This
volume corresponds to the inspiratory reserve volume
of the lungs. The resting stroke volume, systolic
reserve volume, and diastolic reserve volume together
define the maximal stroke volume, which corre-
sponds to the vital capacity of the lungs.
The parallelism can be carried even further (fig.
21). In the resting organism, the amount of blood
remaining in the ventricle at the end of ejection
(called by some authors the "end-systolic volume")
would best be referred to as functional residual capac-
ity since it is now customary to use the term capacity
for the sum of two or more "volumes." ("Capacity"
does not imply something that is absolute or fixed,
despite the unfortunate analogy suggested by the
age-old and uneradicated expression "vital capacity.")
The functional residual capacity of the ventricle
comprises the systolic reserve volume plus residual
FUNCTIONAL ANATOMY OF CARDIAC PUMPING 783
fig. 20. Scheme of lung volumes {left) and ventricular volumes (right) at rest and during exercise
or sympathetic stimulation. For details see text.
fig. 31. Scheme of lung volumes (left) and ventricular volumes (rig/it) at rest to illustrate the
parallelism between respiratory and ventricular volumes.
784
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
volume, much as the functional residual capacity of
the lungs comprises the expiratory reserve volume
and the residual volume. Correspondingly, the
amount of blood accumulated in the ventricle at the
end of ventricular diastole (often called the end-
diastolic volume) could be referred to as '"diastolic
capacity," including the stroke volume plus systolic
reserve volume plus residual volume. Since the stroke
volume varies with changes in activity, the diastolic
capacity is not a fixed amount but functionally
variable. It becomes larger as the stroke volume
increases and smaller as the stroke volume decreases.
There are two more terms which can be described in
parallelism with the nomenclature used in respira-
torv physiology. The volume level reached at the end
of systolic ejection is termed '"end-systolic level" and
corresponds to the expiratory level of the lungs. The
volume level reached at the end of diastolic filling is
called ""end-diastolic level" in analogy to the in-
spiratory level of the lungs.
The importance of a clear terminology for the
description of the dynamic shifts of ventricular
volumes under varying conditions of activity has
been pointed out by Rushmer (139) and many
others have followed his lead. The parallel with the
lung volumes also permits useful analogies. For in-
stance, an increase in residual volume of the lungs
in emphysema diminishes the ventilatory efficiency.
In a somewhat similar manner an increase in the
ventricular residual volume, such as occurs in ex-
cessive ventricular dilatation, diminishes the pump-
ing efficiency of the heart.
To compare with the cardiac ventricles, a piston
pump would need the following features. The course
of the piston, which defines the stroke volume, would
have to be limited in order to leave fluid in the pump-
ing chamber at the end of ejection (functional
residual capacity). If a greater output were needed,
the piston would have to push farther and increase
its stroke volume by encroaching upon the systolic
reserve volume. Yet the volume filling the dead space
of the pump (residual volume) could never be ejected.
The diastolic reserve volume would be represented
by a farther pulling back of the piston to allow greater
filling of the pump chamber. In this mechanical
system, the need for a greater output could be met
instantaneously by the ejection of part of systolic
reserve volume. However, the diastolic reserve
volume could not be utilized instantaneously because
the pump chamber has first to be filled to a greater
extent before more can be ejected. The situation
seems to be the same in the heart. The left ventricular
stroke volume can be increased from one heart beat
to the next by drawing upon the systolic reserve
volume, as occurs for instance when the organism
passes abruptly from rest to exercise (33, 139, 146).
On the contrary, the mechanism of greater diastolic
filling (Starling'., law) always involves a brief delav
brought about by the need for greater venous return
before increased ejection. Apparently, onlv in
strenuous exercise and in some pathological condi-
tions is the diastolic reserve volume called upon.
Indeed the ventricular stroke volume varies almost
continuously and is not identical from beat-to-beat
even under resting conditions. Some of the variations
are probably caused by the fluctuating play of poorly-
known neural feedback processes. Others are caused
by mechanical forces such as those which accompanv
respiration. In fact the respiratory variations of the
ventricular stroke volume are remarkable even under
resting conditions (21). Figure 22 shows the typical
changes in right ventricular output during five
heartbeats modified by the action of one respiratory
cycle. After the onset of inspiration, .-1, there is first
an increase in venous inflow (second heartbeat) and
then in ventricular stroke volume (third heartbeat).
Similarly, the drop in venous return during expira-
tion (at the fourth heartbeat) is reflected by a de-
crease in stroke volume one beat later. From this
record it appears that the right ventricle temporarily
accommodates part of the large inspiratory inflow of
venous blood, and releases it into the pulmonary
circulation during the respiratory pause. This in-
dicates that with respiration not only the stroke
volume varies but also the functional residual capacity
(or end-systolic level ) .
The functional residual capacity cannot be meas-
ured directly in the intact organism. Most of the esti-
mates obtained with indirect methods display con-
siderable variation according to the technique
emploved. The volume curves shown in figures 16 and
1 7 were obtained with multiple plane high-speed
X-ray cinematography. They show not only the
volume changes throughout the cardiac cycle, but
also the end-systolic level. The functional residual
capacity of the left ventricle of these 12-kg dogs
amounts to approximately 5 to 6 ml. In unanes-
thetized, quiescent dogs, the values measured were
of the same order of magnitude. Gribbe et al. (56, 57)
estimate that on the average the stroke volume of
dogs is 60 per cent of the diastolic capacity. Thus
functional residual capacity amounted to 40 per
cent of the diastolic capacity. It should be pointed out
that Gribbe's values are much smaller than tht
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
785
fig. 22. Effect of spontaneous respiration on right ventricu-
lar stroke volume, measured by the directly recorded pulmonary
blood flow in an anesthetized normal dog. The simultaneously
recorded pattern of right atrial filling (represented by superior
vena cava flow), arterial, venous, and intrathoracic pressures
permit a time correlation. Tracings from top to bottom: time
and base line, aortic pressure in mm Hg, pulmonary artery,
superior vena caval and intrathoracic pressures in mm water,
pulmonary arterial and superior vena caval flows in ml/min.
A = beginning of inspiration; S = acceleration of superior
vena caval flow during ventricular systole; D = acceleration
of superior vena caval flow during ventricular diastole. Stroke
volume (in ml) entered under pulmonary arterial flow curve.
Flow (in ml) through superior vena cava during each cardiac
cycle entered at bottom of record. Electrical frequency re-
sponse of both flowmeters reduced from 400 to 40 cycles/sec.
Superior vena caval pressure curve damped. [From Brecher &
Hubay (21).]
estimates of Holt (77) with dye dilution techniques,
which range from 30 to 76 ml for a dog of 1 5 to 1 6 kg
in weight. This enormous discrepancy cannot be
reconciled at present. Simultaneous determinations
under rigidly controlled conditions with both methods,
the cineangioradiography and the indicator dilution
technique, may elucidate this point.
That the situation is equally unsettled for measure-
ments in man is shown by the work of Rushmer
(139), Chapman et al. (32), Nylin (122), Reindell
et al. (134), Folse et al. (47), Luthy (personal com-
munication, and 106). Generally, determinations
using roentgenologic technique furnish smaller values
for the functional residual capacity than measure-
ments with dye dilution techniques. For instance,
Folse et al. (47), employing radio-iodinated Diodrast,
found in 20 resting persons that the left ventricular
stroke volume averaged 42.2 ± 8.8 ml per m2 of body
surface area. The diastolic capacity averaged 90 ±
26 ml per m2 (functional residual capacity 48 ml/m2).
On the other hand, Luthy (106) found with thermo-
dilution techniques that in normal patients the left
ventricular stroke volume amounted to 45 ml per
m2 of body surface (range 39-57 ml/m2) but the
diastolic capacity to 145 ml per m2 (range 128-173
ml/m2). According to these data the stroke volume
would be only one-third of the diastolic capacity
[39 %, Folse et al. (47); 31 ';, Luthy (106)] whereas,
in the dog it is apparently about two-thirds (60%,
Gribbe).
The problem is further complicated by the fact
that the ratio of stroke volume to functional residual
capacity changes markedly under various normal and
pathological conditions. This is well illustrated by
the observation that a great increase in resistance to
ventricular ejection (e.g., in extreme hypertension)
causes the heart size to become much larger while
the stroke volume decreases. This implies a large
increase in the functional residual capacity. Direct
evidence for an increase in functional residual
capacity under this condition is the observation that,
when the aortic resistance is suddenly reduced by
opening of an arteriovenous shunt, the first stroke
volume is twice the normal size [Hamilton (61)].
From the foregoing it is obvious that much more
information based on direct measurements of heart
volumes under various conditions is needed.
Atrial Volume
The volume of blood contained in the atrium at
any time has evoked much less interest than the
ventricular volume. No quantitative information
has been available until recently. Even the termi-
nology of atrial blood volumes is more difficult to
define than that of ventricular volumes. During two
phases of the cardiac cycle (isovolumetric contraction
and isovolumetric relaxation), the ventricle contains
a definite volume because the atrioventricular and
semilunar valves lock the ventricular content. The
atria, however, are always open on the venous inflow
side. On the outflow side they are closed only from
the beginning of isovolumetric ventricular contraction
to the end of isovolumetric ventricular relaxation.
Consequently, the volume contained at any one
instant represents the balance of almost continuously
changing inflow and outflow.
The changes in atrial volumes can be understood
by following the atrial volume curves (open circles)
;86
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
in figures 16 and 17. The left atrial volume is greatest
during isovolumetric relaxation of the left ventricle.
During the rapid ventricular inflow* phase the atrial
volume decreases rapidly, but not so fast as the
ventricle fills. This indicates that some blood enters
the atrium from the veins while at the same time a
greater amount leaves it toward the ventricle. During
the phase of slow ventricular inflow (diastasis) the
atrial volume remains practically unchanged, point-
ing out that inflow from the veins and outflow toward
the ventricle are approximately in balance. Inci-
dentally, this gives a measure of the rate of venous
return to the atrium during this phase by simply
calculating the increase in ventricular volume.
During atrial systole (approximately end of P wave
of electrocardiogram, fig. 1 7) the atrial volume
decreases precipitously. The rate at which the atrial
volume decreases and the ventricular volume simul-
taneously increases speaks in favor of a negligible
backflow of atrial blood into the veins during atrial
systole. At the peak of atrial systole the volume of
blood contained in the atrium is minimal, but still
amounts to approximately 4 ml in dogs. During the
phase of isovolumetric ventricular contraction, the
atrial volume already begins to increase owing to
an accelerated venous inflow. It continues to increase
at a rather fast rate during rapid ventricular ejection
and at a slower rate during the phase of reduced
ventricular ejection. Whereas the ventricular stroke
volume of the 1 2-kg dog amounts to about 8 ml,
the difference between the largest and smallest
atrial volume amounts to only 5 ml. This indicates
that during the ventricular rapid and slow filling
phases approximately 3 ml passes from the veins
through the atrium into the ventricle without being
recorded as an atrial volume increase. It further
indicates that during ventricular isovolumetric con-
traction, rapid and reduced ventricular ejection,
about 5 ml pass from the veins into the atrium
while the atrioventricular valves are closed.
Obviously, one expects during exercise a greater
accommodation of blood in the atrium during ven-
tricular ejection and a greater outflow of blood from
the atrium into the ventricle during the ventricular
rapid and slow filling phase. There are still no meas-
urements available concerning such physiological
adaptations. It appears reasonable to suggest that
in exercise the atrium ejects during its own systole a
greater volume, thereby drawing upon the amount
of blood usually remaining at rest in the atrium at
the end of atrial systole. This might be termed the
"atrial systolic reserve volume."
ATRIAL FILLING
The phasic changes of venous return which bring
about atrial filling are still a subject of debate. It is
often stated that venous blood returns to the heart
solely as a result of the force imparted to it on the
arterial side of the circulation (vis a tergo). Yet there
are reasons to believe that the systolic contraction
of the ventricular myocardium also contributes to
atrial filling by causing an expansion of the atria
[see also Hamilton (60) and Holzlohner (79)].
This view- was originally advocated by Purkinje
(132) who observed that during ventricular systole
the atrioventricular junction (the plane of the heart
valves) descends toward the apex and pulls on the
atrial walls. The atrial cavity is then passively ex-
panded and the pressure in it drops, causing an
acceleration of blood from the veins into the atria.
Among functional anatomists, the concept of the
attraction of blood into the atrium by the descent of
the valvular plane toward the apex during ventric-
ular systole has gained great favor. By injecting
drops of radiopaque contrast material into peripheral
veins and taking X-ray cinematographic pictures,
Bohme (14) could demonstrate a remarkable acceler-
ation of central venous flow during ventricular sys-
tole. Records obtained from direct measurements
of blood flow in the superior and inferior venae cavae
with a high fidelity flowmeter by Brecher & Praglin
(25) and by Brecher (19) confirmed Bohme's ob-
servations. It appears now that ventricular contrac-
tion does cause a sudden expansion of the atrium.
This mechanism lowers the pressure in the atrium
and produces the X wave, much as a plunger with-
drawn in the barrel of a syringe lowers the pressure
therein. The expansion of the atrium probably begins
with the asynchronous contraction of the papillary
muscles during the early part of the isovolumetric
phase and continues during rapid ventricular ejec-
tion. In the hands of Rushmer (136, 139) the lipiodol
injection technique indicated only a moderate accel-
eration of caval blood flow during early ventricular
systole [see also Lynch (107)]. However, Rushmer's
findings can be reconciled with those of Bohme and
Brecher, if one considers the differences in the various
experimental conditions (open or closed chest,
anesthetized or awake animals, slow or fast heart
rate, inspiration, expiration, volemic status, etc.).
The measurements of Gribbe (56, 57) in intact closed-
chest animals definitely indicate an increased atrial
inflow beginning at isovolumetric ventricular con-
traction and continuing during the rapid phase of
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
787
ventricular ejection (see atrial volume curve in
fig. 16 and superior vena cava flow curve in fig. 22).
In Chapter 1 7, vol. I, of this Handbook, evidence is
given that in normal man the venous stream toward
the heart pulsates reciprocally to the aortic stream
leaving the chest. The exactitude of the reciprocal
relationship is said to be measured by the very small
volume equivalent of the cyclic changes (cardiac) in
intrathoracic pressure after making allowance for
the elasticity of the chest walls (60, 64).
In conclusion, the filling of the atrium during ven-
tricular systole depends not only upon the pressure
of blood in the venous reservoir which is available
for passive filling from behind; it depends also on
the vigor with which ventricular systole moves the
atrioventricular junction. Thus a more forceful con-
traction (commonly associated with a larger stroke
volume) ensures additional inflow into the atrium
and therefore facilitates the next ventricular ejection
without the need for further decreasing the systolic
reserve volume to maintain a large stroke volume.
The increase in ventricular outflow and in atrial
inflow are mediated by the same force, ventricular
contraction, and both favor a more thorough filling
of the ventricle during the next diastole.
The central veins are most suitable for this reser-
voir function because through partial collapse of
their walls, their content can change rapidly without
much change of pressure. They form a collapse
chamber which is the functional counterpart of
the aortic compression chamber (18). On the ar-
terial side, the compression chamber based on the
elastic distensibility of the walls assures the trans-
formation of the discontinuous cardiac ejections into
steady flow to the tissues. On the venous side, the
collapse chamber based on the pliability of the walls
assures at the atrial entrance the transformation
of the steady flow from the tissues into the pulsatile
flow which is needed for the discontinuous cardiac
filling [see also Irisawa et al. (83)].
The filling of the venous collapse chamber is in
turn aided by the atrial systole. From a hemodynamic
standpoint the function of atrial contraction is two-
fold: /) It ejects some blood into the ventricular
cavity, a well-established fact, and 2) it passively
enlarges the central venous reservoir by briefly-
slowing down or stopping atrial inflow. The small
amount of backflow which is often recorded during
atrial systole at the caval-atrial junction normally
does not extend far into the periphery (18). It is
readily taken up by a widening of the collapse
chamber and, together with the continued inflow
from the periphery, creates the pool from which the
next ventricular filling derives its supply.
Any force which lowers pressure in a region toward
which flow occurs, is called suction, whether or not
the pressure developed in that region drops below
atmospheric zero. Physically "suction" is the same
as pressure (force per area). It is a reduction of pres-
sure at some point in a system by the application of a
force which results from an energy conversion proc-
ess, e.g. muscular contraction, elastic recoil, pulling
of a plunger. Since blood is attracted into the atrium
by ventricular contraction, one may therefore state
that atrial filling is at least in part brought about
by suction upon the venous blood mediated through a
stretching and enlargement of the atrial cavity by
the contracting ventricular muscles which cause a
descent of the atrioventricular junction. This phe-
nomenon can be termed ''ventricular systolic suction"
upon the atrial content.
[A semantically more rigid definition of the concept of
"suction" holds that suction can be thought of only in locations
where the transmural pressure is negative (142). Accordingly
suction cannot be transmitted through viscera (atria, veins)
which have collapsible walls. As long as these viscera contain
blood at a greater pressure (including equivalent kinetic energy)
than the extra visceral pressure, the dominant force is "pressure"
from upstream rather than "suction" from downstream. This
pressure maintains the walls of the atria and intrathoracic veins
under elastic tension during the entire cardiac cycle. Ed.]
When the atrioventricular valves open during
ventricular diastole, there is another decrease in
atrial pressure (Y wave), which results once more
in an acceleration of venous blood inflow into the
atrium. This second acceleration is more pronounced
at slow heart rates and is usually greater in closed-
chest than in open-chest animals. Apparently the
expansion of all the cardiac cavities through the
pulling force of the lungs [Pfuhl (129, 130)] helps
to make the atrial inflow during ventricular diastole
slightly greater than during ventricular systole. The
increase in atrial inflow during ventricular diastole
may also be a consequence of the attraction of atrial
blood into the ventricle through the forces which
expand the ventricle during diastole (ventricular
diastolic suction, see following section). These forces
not only affect the blood contained in the atrium
but in turn even affect the adjoining veins by lowering
the atrial pressure, particularly during the phase of
rapid ventricular filling. Figure 22 illustrates the
phasic increases in superior vena caval flow during
ventricular systole (S) and during ventricular diastole
(D) in an anesthetized closed-chest dog.
788
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
In summary, ventricular myocardial contraction
could have a threefold effect upon atrial filling:
/) It imparts so much energy to the arterial stream
that even after passing through the capillaries, the
blood continues to flow in the veins and fills the cen-
tral reservoirs; 2) during ventricular systole blood
is drawn actively into the atrium through an expan-
sion of the atrial cavities by a movement of the atrio-
ventricular junction toward the apex ("ventricular
systolic suction"); j) during ventricular diastole
the elastic forces, created by the preceding systole
in the ventricular walls, can aid in drawing blood
from the atrium into the ventricle and even in attract-
ing blood from the veins into the atrium ("ventric-
ular diastolic suction" upon the atrial content).
VENTRICULAR FILLING
While the forces causing ventricular ejection are
unquestionably those originating from myocardial
contraction, there is considerable debate concerning
the forces responsible for ventricular filling [for de-
tails see Brecher (19), Krug & Schlicher (95), El-
dridge & Hultgren (41), Bauereisen et al. (7)].
The ejection of fluid from a pump is more spec-
tacular than the filling phase. This may be the reason
why the forces dealing with ventricular contraction
have received considerably more attention than the
forces dealing with ventricular filling. For a long
time it has been believed that the heart is filled ex-
clusively by a force which pushes blood into the ven-
tricle from behind [vis a tergo, Galli (50)]. This
force results from the preceding ventricular contrac-
tions and is imparted to the blood for circulating it
through the arteries, capillaries, and veins. Others
have maintained that some part of ventricular filling
is produced by a force from the front (vis a fronte)
which attracts or sucks blood from the atrium into
the ventricle. This force would manifest itself in
the ventricle during diastole and would probably be
caused by an elastic recoil of the ventricular walls.
It would lower the intraventricular pressure below
the level which would prevail if such a vis a fronte
did not exist. The history of the issue between vis a
tergo and vis a fronte is treated in detail by Ebstein
(40), Hamilton & Lombard (64), Bohme (14),
Brecher (19, 20), Krug & Schlicher (95). One can
summarize as follows the evidence in favor of the
existence of a vis a fronte: Ventricles of cold-blooded
animals, in the observations of Kraner & Ogden (94),
Kraner (93), Hennacy & Ogden (73), Hennacy (72),
Peiper & Weigand (131), and of mammals according
to Bloom (13), Fowler et al. (48, 49) Brecher (19), and
O'Brien (123), can definitely suck in blood when the
filling pressure at the atrioventricular orifice is atmos-
pheric (zero) or subatmospheric (negative). Whether
or not ventricular suction also contributes to ventricu-
lar filling in the presence of a positive filling pressure
at the atrioventricular orifices has not yet been experi-
mentally established.
It has been objected that the ventricles in which
suction forces were demonstrated had an abnormally
small functional residual capacity. The question there-
fore arises whether or not ventricles with physiological
volumes would also exert a suction force. Scheu &
Hamilton (143), using the intact spontaneously
breathing anesthetized dog, made simultaneous re-
cordings of the intraventricular and thoracic pres-
sure and thus established the transmural ventricular
pressure gradient. They held that "suction," prob-
ably by the elastic recoil of the ventricular walls,
could be demonstrated only if and when the trans-
mural pressure was negative. They concluded that
suction did not occur during normal diastole but
could be brought about by compressing the mitral
orifice or by hemorrhage. These two maneuvers
made the diastolic ventricular shadow smaller and
were thought to have reduced the residual blood to a
subnormal figure.
Brecher & Kissen (23) demonstrated that dog
ventricles of an approximately normal functional
residual capacity filled by suction at zero ventricular
inflow pressure. Nevertheless, as long as there is not
unequivocal experimental proof of the existence of
ventricular diastolic vis a ironte in the unanesthe-
tized intact mammalian organism, one should be ex-
ceedingly cautious with any statement concerning
the role of diastolic suction in ventricular filling
[Brecher (20)].
Horres and his group (unpublished observations)
determined the average left ventricular volumes of
excised submerged hearts at equilibrium state and
found it to be 17 (±6) ml for dogs weighing 12 kg
(fig. 23). If one assumes that the elastic equilibrium
state of the relaxed ventricle in vivo is the same
as that of the freshly excised, and still responding
ventricle in vitro, then diastolic ventricular suction
could occur at any ventricular volume below the
equilibrium point (i.e., less than 17 ml). Unfor-
tunately the values of the functional residual capac-
ity reported for the dog heart vary too much to per-
mit an unbiased conclusion about the role of suction
in ventricular filling. According to the data of Holt
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
789
40-
•
30-
e
0
>
•
• **
^^^LEFT
^/ventricle
*^r V V LEFT =
^r -0.80*1.68 BW
20-
UJ
CL
■z
>
•£
0 +S
m / •
B9^9 • RIGHT
VENTRICLE
^,''''v.V. RIGHT =
-.^-" 6.26 +0.67 BW
0 0
0
10-
c
0
0
o°
BODY WT kg
1
6
10 14
18
1 1 ' I
22 26
fig. 23. Relationship of ventricular volume (VV) and body
weight (BW) in dogs. Ventricular volume measured at the
equilibrium state (zero transmural pressure). [See also fig. 2
(Horres et al., unpublished obser%'ations).]
(77) the average functional residual capacity of the
left ventricle would be approximately 30 ml for
dogs weighing 12 kg. According to Gribbe et al.
(57), it is only 5 ml. The discrepancy between Holt's
and Gribbe's measurements in terms of ventricular
diastolic suction is illustrated in figure 24. If Holt's
data are correct, ventricular diastolic suction never
occurs under normal conditions. On the contrary
Gribbe's figures speak for the occurrence of diastolic
ventricular suction during all phases of diastole.
Obviously the controversy cannot be resolved on
the basis of presently available data.
There has been some speculation about the pos-
sible nature of the frontal force, particularly whether
it originates from an active or a passive process.
An active process would be the contraction of muscle
fibers which, owing to their anatomical arrangement,
could widen the ventricular cavity during diastole
[Guasp (58)]. There is no experimental evidence to
support such a view. Another active process would
be the development of a force acting to lengthen the
muscle fibers upon completion of their contraction
("active decontraction"). However, it has never
been satisfactorily demonstrated that processes of
energy conversion from chemical to kinetic energy
occur during muscular relaxation [Villa (152);
for review, Brecher (19, 20)]. The most acceptable
evidence is, at present, that the diastolic ventricular
vis a fronte is caused by passive processes, such as
one of the following, a) During systole an interfas-
cicular tension develops through shear forces be-
tween myocardial strands which contract to different
extents and asynchronously [details in Rushmer
(•39)]- f>) During systole noncontractile elements in
the heart and possibly also some components of the
muscle fibers are elastically deformed beyond their
equilibrium state, thereby storing potential energy
which is released through elastic recoil during diastole
(see fig. 2). c) In the closed-chest mammal, additional
external forces residing in the elastic recoil of the
lungs exert their effect upon the heart by tending to
expand the cardiac cavities beyond the size these
cavities would assume in the absence of the lung
forces.
In conclusion, some of the classical views concern-
ing the filling of the heart may need revision. The
ventricle acts as a reciprocating pump in which the
output stroke simultaneously provides energy for
the filling of the pump for the next stroke. In other
words, the heart does not act merely as a pressure
pump as William Harvey (66) believed, but it actu-
ally functions as a pressure-suction pump [see also
Gauer, (52, 53)]. The amount of energy necessary
for pump filling is, however, only a fraction of that
needed for ejection, since the filling occurs through a
fluid transfer into a low resistance system in which
small pressure differences will cause a rapid flow of
large amounts of blood.
DIFFERENCES BETWEEN RIGHT AND
LEFT CARDIAC CAVITIES
Functional differences between the right and left
cardiac cavities can be expected from their anatom-
ical characteristics. Yet it had long been tacitly
assumed that the two atria and the two ventricles
initiate and terminate their contraction simultane-
ously, and that a description of cardiac events on
both sides would be redundant.
In fact, there are significant differences between
the left side and the right side chambers [see also
Katz (89), Hamilton et al. (62), Luisada & Fleischner
(103), Segers (145), Braunwald et al. (16), McKusick
(109)]. For instance, at equal pressures the right
atrium has a volume twice that of the left atrium,
which is thicker and less distensible than the right
(99). Experimentally, the volume-pressure curve
relationship in the left atrium has been found to be
linear only as long as the pressure remains within
the normal limits (pressure below 150 mm H20).
When this limit is exceeded, a slight increase in
volume causes a much larger increase in pressure.
The normal level and patterns of pressure also
differ somewhat between the right and the left
atrium. The A wave of the right atrium, produced
79°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 24. Functional residual capacity
of the clog's ventricle and its relation to
ventricular diastolic suction. Lejt: data
derived from Holt (77), assuming that a
functional residual capacity of 50 ml in
a 20.9-kg dog corresponds to 30 ml in
a 12-kg dog (regression line of fig. 23).
Right: data of Gribbe et al. (56) for a
12-kg dog. Center: elastic equilibrium
state of the ventricle in a 12-kg dog
(Horresf/ al., unpublished observations).
FUNCTIONAL RESIDUAL CAPACITY
HOLT
(in vivo)
t
NO
SUCTION
I
I
I
NEUTRAL POSITION
I
(ELASTIC EQUILIBRIUM;
I
I
SUCTION
HORRES et al,
(in vitro)
FUNCTIONAL RESIDUAL CAPACITY
GRIBBE et al .
(in vivo)
ml
30
by atrial systole, is often not so steep and tall as that
of the left atrium, and normally precedes it slightly.
The peak of the right atrial V wave is usually lower
than that of the left atrial V, and the mean right
atrial pressure is usually less than the left atrial
pressure.
At the level of the ventricles the bundles of myo-
cardial fibers which encircle the two cavities, much
as the windings of a turban, belong to a common
anatomical structure. The combined effect of their
contraction is to wring blood out of the ventricular
chambers into the respective arteries. Yet the mus-
cular arrangement is such that contraction of the
left ventricle produces primarily a reduction in the
lateral diameter with only a moderate shortening
along the vertical axis, whereas on the right side
there is much ventricular shortening between apex
and base with relatively less pulling of the free wall
toward the septum. The mechanical effects of left
ventricular contraction occur a trifle earlier than
those of right ventricular contraction, since the rise
in right ventricular pressure usually lags by 0.01
to 0.02 sec behind the rise in left ventricular pressure
(see also fig. 19). It is, therefore, understandable
that mitral valve closure usually precedes tricuspid
valve closure. Nevertheless, it is by no means es-
tablished whether the later start of right ventricular
contraction is the only cause of asynchronicity.
Other factors could also operate, such as a faster
rate of contraction of the left ventricular wall or a
quicker reaction of the mitral cusps to the rising
wave of pressure in comparison with the tricuspid
valve. The characteristics of the vascular bed into
which ejection proceeds cause differences in the se-
quence of pumping events on the right and on the
left side. Since the pressure in the pulmonary artery
is low (low resistance to flow in the pulmonary vas-
cular bed), the pulmonary valve opens first. Indeed,
for an interval of about 0.02 sec, the ventricular
contraction produces an ejection on the right side
while there is still an isovolumetric pressure rise
on the left side. Similarly, right ventricular ejection
continues well after cessation of left ventricular ejec-
tion. In other words, the end of left ventricular
systole precedes that of the right and closure of the
aortic valve precedes closure of the pulmonary valve.
On the other hand, because of a much longer phase
of isovolumetric relaxation in the left ventricle as
compared to the right, the opening of the mitral
valve is thought to follow that of the tricuspid, as
seen in table 1, which is borrowed from Luisada &
Liu (104) and summarizes the sequence of events of
right and left ventricular contraction (see also figs.
18 and 19).
Because of these differences in the pumping action
of the ventricles, reference is often made to right
and left ventricles as being a ''volume pump" and
"pressure pump," respectively. Implied in this no-
menclature are the facts that the right ventricle
can easily handle an increase in volume output with-
out apparent strain, whereas it is not so well equipped
for raising its pressure to a high level. On the con-
trary, the left ventricle, with a much larger mass
of active musculature and a more nearly spherical
geometry is better able to face an increase in out-
flow resistance than the right ventricle. What is
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
79'
Interval,
in Sec
0.06-0
.07
O.OO-O
.02
0.01-0
•03
0.01-0
.02
0.02
0.03-0
,04
0.04-0
.08
0 . 00-0
.04
0 . 04-0
.08
table i . The Cardiac Cycle. Time Intervals Between
Valvular Motions (Normal Dogs)
Event
Q wave (ECG)
Closure of mitral valve
Closure of tricuspid valve
Opening of pulmonic valve
Opening of aortic valve
Closure of aortic valve
Closure of pulmonic valve
Opening of tricuspid valve
Opening of mitral valve
Rapid filling of right ventricle
Rapid filling of left ventricle
Each peak of rapid ventricular filling follows A-V valve
opening by 0.08-0.10 sec, and the closure of the respective
semilunar (i.e., pulmonary or aortic) value by 0.12-0. 18
sec. This table shows the sequence of events in normal,
large dogs, whose figures are probably very close to those of
normal man. [From Luisada & Liu (104).]
meant by the terms "volume pump" and "pressure
pump" is actually "low-pressure head pump" and
"high-pressure head pump."
There is no fundamental difference in the pump-
ing action of the two ventricles before birth. They
both receive blood from a common atrial chamber,
their walls are of the same thickness, they have the
same capacity, and they both eject their contents
into a common aortic chamber via either the ductus
arteriosus or the ascending aorta. The pressure against
which the fetal ventricles eject their contents is lower
than that in the adult systemic circuit, but after
birth the resistance in the lesser circuit drops sud-
denly with the first breath, whereas that in the sys-
temic circuit gradually increases. The difference
in pumping action that prevails in the normal adult
exists essentially because of the difference in resist-
ance to flow.
In the adult the resistance to flow in the pulmonary
vascular bed is estimated to be only about one-
eighth of that in the systemic circulation. On the
other hand, it is estimated that a sizable amount of
the mechanical energy, both pressure and kinetic,
imparted to the blood by left ventricular ejection
is still available at the point of venous inflow into
the right ventricle, and is partly responsible for right
ventricular filling and distention during diastole
(vis a tergo). Connecting these two observations,
one may wonder whether right ventricular contrac-
tion is necessary at all, or whether the left ventricle
alone could not only circulate the blood through the
systemic vascular bed but also through the pulmonary
vascular bed. The problem has been approached in
different ways. It was first observed that major de-
struction of the right ventricular wall (by cauteriza-
tion, for instance) causes but slight changes in sys-
temic venous and arterial pressure [Bakos (5),
Kagan (86)]. Yet a doubt remains, since some inner
layers of the right ventricular wall are left intact
in such experiments, and conceivably contractions
of muscular bundles, which belong to the left ven-
tricular wall, could still pull on passive strands of
the remaining right ventricular wall and thus in-
directly eject blood through the pulmonary ostium.
In such a case, the noncontracting cauterized re-
mainder of the right ventricular wall would passively
compress the half-moon-shaped right ventricular
cavity. Also, the still intact, powerful ventricular
septum could contribute to the right ventricular
systolic pressure rise. Acute experiments in which the
entire heart is arrested and only the left ven-
tricle, but not the right, is replaced by a mechanical
pump, indicate that apparently stable circulatory
conditions in both the systemic and the pulmonary
vascular beds can be maintained using a single "left
ventricle" pump [see also Rodbard & Wagner
(137), Jamison et al. (85), Warden et al. (154),
Patino (•/ al. (127), Xuland et al. (121), Monod-
Broca (115), Glenn (55)]. Obviously, the pressure
in the vena cava is then raised to maintain sufficient
pressure for the pulmonary circulation, and there
may be an impairment to cerebral, coronary, and
hepatic venous outflow. However, the conditions are
compatible with survival of the animal. The question
is not solely of academic interest, since it is not im-
possible to envision that some day surgical tech-
niques will be devised to drain the entire systemic
venous return directly into the pulmonary artery,
thereby placing the entire load of circulation on the
left ventricle.
the pericardium
The function of the pericardium in the cardiac
pumping process has been the subject of much de-
bate. Some authors consider that the pericardium
does not affect cardiac performance, because con-
genital absence of this structure in man is compat-
ible with maintenance of a seemingly normal cardiac
function [see Ellis et al. (43) and Hering et al. (74)].
Others have speculated that the primary function of
the pericardium is to confine in space the pumping
structures which are characterized by their expansi-
792
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
bility, pliability, and limited anchorage [see Pfuhl,
(129) and Nelemans (118)]. A number of additional
functions have been ascribed to the pericardium,
namely : protection of the entire heart from over-
dilatation; protection from overdilatation of the left
ventricle only; protection from overdilatation of the
right ventricle only; increase of cardiac performance
because of higher filling pressure; better harmonic
coordination of right and left ventricular contrac-
tions; facilitation of atrial filling; and facilitation of
the gliding of the epicardium through lymph lubrica-
tion.
Most experimental studies on the function of the
pericardium concern abnormal situations such as
are met in pericardial effusion and tamponade
from various other causes [Feinberg (46), Adcock
et al. (1), Evans et al. (44), Nerlich (119), Metcalfe
et al. (112), Isaacs et al. (84), Bencini & Parola
(9)]. The information from these studies, though in
some respects limited to abnormal hemodynamic
situations, has contributed much to the elucidation
of normal functions of the pericardium.
Since some of the functions are implicit from the
architecture of the pericardium, a brief anatomical-
histological review will be helpful (118). The parietal
pericardium, generally just called pericardium,
forms a thin, but firm sac of connective tissue en-
veloping the ventricles and atria. At the base of the
heart, near the entry of the veins into the atria,
the parietal pericardium joins the visceral pericar-
dium or "epicardium." Pericardium and epicardium
are separated by a thin fluid film of pericardial
liquor which is similar to the fluid filling the intra-
pleural space. The outer aspect of the pericardium
is covered with a thin layer of loose connective tissue
which constitutes the pericardial, or parietal, pleura.
The pericardium is attached to the diaphragm with
two septa (right and left) to the sternum and to the
mediastinum. This anchorage limits the mobility of
the sac and in turn confines the heart to a definite
space within the thorax, especially in primates.
Histologically the pericardium consists of three
layers of regularly oriented collagenous and elastic
fibers, each oriented in a different direction. The
fibers of the outer and middle layer form a thicker
structure than those of the inner layer. In a prepara-
tion which is not submitted to stretch, the collage-
nous fibers appear wavy and the elastic fibers appear
straight. Nelemans (118) showed that with an acute
dilatation of the normal heart, the pericardium
extends itself by approximately 20 per cent until
the elastic fibers are markedly stretched and the
120
150 VOLUME (ml) 300
fig. 25. Pericardial pressure-volume curve in a dead dog,
determined after removal of heart. The volume of the heart
is replaced by fluid injected into the pericardial space. [From
Holt (78).]
rather tense collagenous fibers hinder further ex-
pansion. This elastic stretching is quickly reversible,
since the sac will again fit snugly when cardiac dila-
tation is abrogated. If, however, the heart dilates
beyond the limit of elastic stretch, which obviously
can only happen if venous return and end-diastolic
pressures rise substantially, then the pericardium
will "give" or yield. However, this additional stretch-
ing is not quickly reversible and persists for a long
time after the heart has returned to its normal shape,
as evidenced by the slackness of the pericardial sac.
The additional stretching is of "plastic" nature and
can be attributed to the collagenous fibers which
return only very slowly to their previous length
after stretching. This dual mechanism of "elastic"
and "plastic" stretch hinders acute overinflation
(elastic limitation) but permits long term dilatation
of the heart (plastic adaptation).
Nelemans' (118) views concerning the "plastic"
behavior of the pericardium are supported by ob-
servations made in experimental pericardial effusion :
following the infusion of 50 ml of saline into the peri-
cardial sac in an open-chest dog, the mean intra-
pericardial pressure first rises markedly but then
gradually decreases to a lower level. Withdrawal
and immediate reinfusion of the same amount of
fluid results in a lower pressure than the one initially
obtained. Withdrawing the saline, but then waiting
for about 2 hours before reinjecting it, results in
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
793
approximately the same pressure rise as initially
recorded upon the first infusion.
The pressure-volume relationships of the peri-
cardial sac without the heart were recently studied by
Holt et al. (78) in experiments which emphasize the
relatively nondistensible nature of the pericardium.
The curve in figure 25 illustrates that, as the fluid
volume in a dog's pericardial cavity increases, the
pressure remains at zero until the volume has reached
about 200 ml. Further volume increments cause a
rather steep rise in pericardial pressure. The results
of these and other experiments [Isaacs et al. (84),
Berglund et al. (11)] point to an important function
of the pericardium, i.e., to restrain the heart's cav-
ities from overdistention.
In 1 91 4, Henderson & Prince (71) showed that
the filling-force relationships which later became
known as Starling's law were such in the right and
left ventricles as to prevent the engorgement or de-
pletion of the lung blood. The lungs were further
safeguarded against congestion by the fact that a
sudden dilation of the left ventricle within the peri-
cardium would prevent the filling of the right heart,
limit the amount of blood that could be pumped into
the lungs and thus prevent their engorgement.
There has been much debate as to whether during
diastole the heart normally fills the entire pericardial
sac [see also Wilson & Meek (162)]. Nelemans
(118) concluded that the heart fills the pericardium
completely during diastole and that the sac has a
restraining influence upon the expansion of the heart.
However, this question has not been studied ex-
tensively until modern recording techniques enabled
Holt et al. (78) to follow the phasic changes of intra-
cardiac and intrapericardial pressures during the
cardiac cycle under various filling conditions ranging
from hypovolemia to plethora. It was found that in
an open-chest dog any increase in ventricular end-
diastolic pressure above approximately 1 mm Hg
causes a nearly equal rise in pericardial pressure.
Since end-diastolic pressures of this order of magni-
tude are found under normal circulatory conditions,
it appears that the ventricle does occupy the peri-
cardial sac completely and even stretches it slightly
at the end of the filling phase. Since under conditions
of plethora a positive pressure is maintained in the
pericardial space throughout the cardiac cycle, the
transmural ventricular or transmural atrial pressure
must then be taken as the difference between intra-
cardiac and pericardial pressures rather than as the
difference between intracardiac and intrapleural
pressures.
In comparing the phasic changes in intra-atrial,
intraventricular, and pericardial pressures, Holt et al.
(78) also made observations which cast light on the
contribution of the pericardium to the pumping
action of the heart by facilitating atrial filling. The
pericardial pressure drops markedly during the early
part of ventricular systole. "Since the atria are lo-
cated within the pericardial sac . . . , the pressure
in the right atrium decreases in early systole and
the atrium becomes distended by blood rushing
into it from the great veins. A measure of the degree
of this atrial 'filling pressure' is the difference be-
tween right atrial end diastolic pressure and the
pericardial pressure in early systole." When the atrial
pressure drops, "the pressure gradient from the
great thoracic veins to the right atrium is markedly
increased. This appears to be a mechanism by which
blood is drawn into the atrium during ventricular
systole, and in this way blood is ready to fill the ven-
tricles immediately on cessation of ventricular systole.
Thus, with the pericardium intact, the act of ven-
tricular systole draws blood to the ventricle [sic] and
insures ventricular filling in early diastole. These
results are in agreement with those of Bohme and
Brecher who showed that there was a large sudden
flow of blood through the superior vena cava toward
the heart during early ventricular systole. This
has been attributed by several investigators, and
most recently by Brecher, to the sudden piston-like
downward movement of the atrioventricular junction
attracting blood from the central veins into the right
atrium. Our data indicate that the increased flow
into the right atrium is caused by the sudden de-
crease in pericardial pressure with ventricular sys-
tolic ejection, and that this factor becomes greater
with higher ventricular diastolic pressures. Confirma-
tion of the importance of the pericardium in this
connection is the observation of Brecher that the
acceleration of venous flow toward the right atrium
during ventricular systole is decreased by opening
the pericardium. It would appear that the increased
flow into the right atrium during ventricular systole
was caused in large part by the decrease in peri-
cardial pressure during early ventricular systole.
The question as to how much of this flow is caused
by a downward movement of the atrioventricular
junction remains unanswered. Quantitative data on
this point could be obtained by measuring the flow
into the right atrium in the open-chest dog, with
the pericardium intact and after complete removal of
the pericardium" (78).
Obviously, both the piston-like downward move-
794
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
ment of the atrioventricular junction and the drop
of pericardial pressure during early ventricular
ejection are caused by the same force, i.e., the ven-
tricular myocardial contraction. Which of the two
factors is predominant in facilitating atrial filling
during ventricular systole will not be easily decided.
From the presently available evidence one can
conclude that the pericardium aids the pumping
function of the heart. In contradiction to the wide-
spread opinion that the absence of the pericardium
does not have a noticeable effect upon the circulation,
there is some experimental evidence that a large
percentage of pericardiectomized animals develop
heart hypertrophy and perform poorly on the tread-
mill [see Nelemans (118)]. It is quite possible that
under conditions of rest or mild exercise the heart
without pericardium can satisfy the metabolic
demands of the tissues. However, under conditions of
strenuous exercise the reserve power of the heart
without pericardium is probably diminished.
CLOSING REMARKS
In recent years the opinion has been often voiced
that hemodynamics is a dead science, in which no
more essential work needs to be done. This view refers
primarily to the mechanical features of the cardio-
vascular system which are supposedly well known.
It applies less to the regulatory aspects which ad-
mittedly require further clarification. However, the
analysis of such a presumably simple function as the
heart's normal pumping, even without consideration
of any neural or hormonal regulatory processes,
reveals wide gaps in our knowledge.
There are numerous reasons for these shortcomings.
Many of the measurements on which present con-
cepts are based were obtained under highly artificial
conditions, such as excised heart preparations and
open-chest animals, which limit the applicability of
the results to intact normal organisms. Findings in
one animal species often cannot be transferred or
extrapolated to other species. For instance, the me-
chanics of cardiac pumping in man differ from those
in other animals because of his upright position,
minimal splanchnic pooling, and several other fac-
tors. Finally, measurements are often performed with
inadequate instrumentation. For example, errors
caused by insufficient sensitivity and time resolution
make it difficult to correlate simultaneous events in
the cardiac cycle.
Only rather guarded conclusions can therefore be
drawn from the available experimental evidence
as to the exact nature of the heart's pumping func-
tion. Wide discrepancies of information obtained
with different methods need to be reconciled. For
example, measurements of the ventricular volumes
by various methods differ so greatly in their order of
magnitude that it is today still impossible to state
how large is the normal functional residual capacity
of the heart. Obviously, views based on insufficient
data must remain in the realms of speculations and
postulates. This is the state of knowledge about many
age-old problems such as cardiac filling and the
precise moment of valve closure. How much more
complex these problems become under various patho-
logical conditions does not need to be elaborated
upon. In view of the recent progress in the bio-
medical sciences, it is shocking to observe how a
seemingly simple mechanical process such as cardiac
pumping still remains so enigmatic.
The help of Dr. Robert C. Schlant in reviewing the manu-
script is gratefully acknowledged.
REFERENCES
i. Adcock, J. D., R. H. Lyons, and J. B. Barnwell.
The circulatory effects produced in a patient with pneu-
mopericardium by artificially varying the intrapericardial
pressure. Am. Heart J. ig: 283-291, 1940.
2. Agress, C. M., L. G. Fields, S. Wegner, M. Wilburne,
M. D. Shickman, and R. M. Muller. The normal vibro-
cardiogram. Physiologic variations and relation to cardio-
dynamic events. Am. J. Cardiol. 8: 22-31, 1961.
3. Akman, L. C, A.J. Miller, E. N. Silber, J. A. Schack,
and L. N. Katz. The ventricular electrokymogram.
Circulation 2 : 890-899, 1 950.
4. Anzola, J. Right ventricular contraction. Am. J. Physiol.
184: 567-57', '956-
5. Bakos, A. C. P. The question of the function of the right
ventricular myocardium: An experimental study. Circula-
tion 1: 724-732, 1950.
6. Bauereisen, E., H. Bohme, H. Krug, U. Peiper, and
L. Schlicher. Der Einfluss der Inspiration auf den
Effektivdruck der intrathorakalen Kreislaufabschnitte.
Pfliigers Arch ges. Physiol. 266: 499-51 I, 1958.
7. Bauereisen, E., U. Peiper, and K. H. Weigand. The
diastolic suction effect of the cardiac ventricles. Z.
Kreislaufforsch. 49: 195-200, i960.
8. Baumgarten (1843). Quoted by Tigerstedt, R. In: A
Textbook of Human Physiology. New York: Appleton. 1906.
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
795
9. Bencini, A., and P. L. Parola. The "pneumomassage"
of the heart. Surgery 39: 375-384, 1956.
10. Benninghoff, A. Die Architektur des Herzmuskels. Eine
vergleichend anatomische und vergleichend funktionelle
Betrachtung. Morphol. Jahrb. 67: 262-317, 1 931.
11. Berglund, E., S.J. Sarnoff, and J. P. Isaacs. Ventricu-
lar function. Role of the pericardium in the regulation of
cardiovascular hemodynamics. Circulation Research 3: 133-
139. '955-
12. Blair, H. A., and A. M. Wedd. The action of cardiac
ejection on venous return. Am. J. Physiol. 145: 528-537,
1946.
13. Bloom, \V. L. Diastolic filling of the beating excised heart.
Am. J. Physiol. 187: 143-144, 1956.
14. Bohme, VV. Uber den aktiven Anteil des Herzens an der
Forderung des Venenblutes. Ergeb. Physiol. 38: 251-338,
I936-
15. Brandt, W. The closing mechanism of the tricuspidal
valve in the human heart. Acta Anal. 30: 128-132, 1957.
16. Braunwald, E., A. P. Fishman, and A. Cournand. Time
relationship of dynamic events in the cardiac chambers,
pulmonary artery and aorta in man. Circulation Research,
4: 100-107, !956-
17. Braunwald, E., H. L. Moscovitz, S. S. Amram, R. P.
Lasser, S. O. Sapin, A. Himmelstein, M. M. Ravitch,
and A. J. Gordon. Timing of electrical and mechanical
events 01 the left side of the human heart. J. Appl. Physiol.
8: 309-3H. '955-
18. Brecher, G. A. Cardiac variations in venous return
studied with a new bristle flowmeter. Am. J. Physiol. 176:
423-430. !954-
19. Brecher, G. A. Experimental evidence of ventricular
diastolic suction. Circulation Research 4: 5 '3-5 1 8, 1956.
20. Brecher, G. A. Critical review of recent work on ventricu-
lar diastolic suction. Circulation Research 6: 554-566, 1958.
21. Brecher, G. A., and H. A. Hubay. Pulmonary blood flow
and venous return during spontaneous respiration.
Circulation Research 3: 11 0-2 1 4, 1955.
22. Brecher, G. A., and A. T. Kissen. Relation of negative
intraventricular pressure to ventricular volume. Circula-
tion Research 5: 157-162, 1957.
23. Brecher, G. A., and A. T. Kissen. Ventricular diastolic
suction at normal arterial pressures. Circulation Research
6: 100-106, 1958.
24. Brecher, G. A., H. Kolder, and A. D. Horres. Form
elasticity of the heart. Physiologist 3 (No. 3): 28, i960.
25. Brecher, G. A., and J. Praglin. A modified bristle
flowmeter for measuring phasic blood flow. Proc. Soc.
Exptl. Biol. Med. 83: 155-157, 1953.
26. Brucke, E. In: Vorlesungen uber Physiologic. Vienna: vol. 1,
1872. Publishing Company unknown (apparently pri-
vately printed).
27. Bucher, K., L. Dettli, K. Weisser, and D. v. Capeller.
Uber primar kardiale Regulationen bei der gegenseitigen
Anpassung von Lungen- und Korperkreislauf. Helv.
Physiol. Pharmacol. Acta 13: 79-88, 1955.
28. Burch, G. E., and R. B. Romnev. Functional anatomy
and "Throttle Valve" action of the pulmonary veins. Am.
Heart J. 47: 58-66, 1954.
29. Burton, A. C. The importance of the shape and size of
the heart. Am. Heart J. 54: 801-810, 1957.
30. Campeti, F. L., G. H. Ramsey, R. Gramiak, and J. S.
Watson, Jr. Dynamics of the orifices of the venae cavae
studied by cineangiocardiography. Circulation 19: 55-64
'959-
31. Chapman, C, O. Baker, and J. Mitchell. Left ventricu-
lar function during rest and exercise. J. Clin. Invest. 38:
1 202-1 2 1 3, 1959.
32. Chapman, C. B., O. Baker, J. Reynolds, and J. Bonte.
Use of biplane cinefluorography for measurement of
ventricular volume. Circulation 18: 1105-1117, 1958.
33. Chapman, C. B., J. N. Fisher, and B. J. Sproule. Be-
havior of stroke volume at rest and during exercise in
human beings. J. Clin. Invest. 39: 1208-12 13, i960.
34. Cignolini, P. Contributo roentgenchimographico alia
dottrina dell'attivita diastolica. Folia Cardiol. 13: 27-41,
'954-
35. Gotten, M. deV, and H. M. Maling. Relationships
among stroke work, contractile force, and fiber length
during changes in ventricular function. Am. J. Physiol.
189:580-586, 1957.
36. Davilla, J. C. The mechanics of the cardiac valves. In:
Prosthetic Valves for Cardiac Surgery, edited by K. A. Meren-
dino. Springfield, 111.: Thomas, 1961, p. 3-47.
37. Dean, A. L., Jr. The movements of the mitral cusps in
relation to the cardiac cycle. Am. J. Physiol. 40: 206-217,
1916.
38. DeBrunner, H. U. Der funktionelle Bau der Atrioventri-
kularklappen des Menschen. Acta Anal. 7: 132-153, 1949.
39. Donders, F. C. In: Physiologie des Menschen. Leipzig:
Hirzel, 1859.
40. Ebstein, E. Die Diastole des Herzens. Ergeb. Physiol. 3:
123-194, 1904.
41. Eldridge, F. L., and H. N. Hultgren. A study of ven-
tricular filling in complete heart block. Stanford Med. Bull.
12:257-262,1954.
42. Ell.nger, G. F., F. G. Gillick, B. R. Boone, and W.
E. Chamberlain. Electrokymograph^ studies of asyn-
chronism of ejection from the ventricles. Am. Heart J.
35 : 97 1 -979. I948-
43. Ellis, K., N. E. Leeds, and A. Himmelstein. Congenital
deficiencies in the parietal pericardium. Am. J. Roent-
genol. 82: 125-137, 1959.
44- Evans, J. M., C. W. Walter, and H. K. Hellems.
Alterations in the circulation during cardiac tamponade
due to pericardial effusion. Am. Heart J. 39: 181-187
!95°-
45. Faller, A. Die fibrillaren Strukturen des menschlichen
Epikards und ihre Bedeutung fur die Verformung des
Herzens. Cardiologia 9: 337-372, 1945.
46. Feinberg, M. H. Functional capacity of the normal peri-
cardium. Am. Heart J. 1 1 : 748-751, 1936.
47. Folse, R., E. Braunwald, and M. M. Aygen. Clinical
technic for determining the fraction of left ventricular end-
diastolic volume ejected per beat (P). Circulation 24:934,
1 961.
48. Fowler, N. O., W. L. Bloom, and E. B. Ferris. Systolic
and diastolic pressure relationships in the isolated rat
heart. Circulation Research 5: 485-488, 1957.
49. Fowler, N. O., C. Couves, and J. Bewick. Effect of in-
flow obstruction and rapid bleeding on ventricular dia-
stolic pressure. J. Thoracic Surg. 55:532-537, 1958.
50. Galli, G. Aktive Erweiterung der Herzkammer durch
die -vis a fronte'. Munch. Med. Wochschr. 101 : 356-358,
'959-
796
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
51. Gardner, E., D. J. Gray, and R. O. O'Rahilly.
Anatomy. Philadelphia: Saunders, i960.
52. GaueR, O. H. Volume changes of the left ventricle during
blood pooling and exercise in the intact animal. Their
effects on left ventricular performance. Physiol. Rev. 35:
'43- '55. '955-
53. Gauer. O. H. Kreislauf des Blules. Berlin: Urban and
Schwarzenberg, i960.
54. Gleason, W. J., and E. Braunwald. Studies on the first
derivative of the ventricular pressure pulse in man. J.
Clin. Invest. 41 : 80-91, 1962.
55. Glenn, W. W. L. Circulatory bypass 01 the right side of
the heart. New Engl. J. Med. 259: 1 17-120, 1958.
56. Gribbe, P., L. Hirvonen, J. Lind, and C. Wegelius.
Cineangiocardiographic recordings of the cyclic changes
in volume of the left ventricle. Cardiologia 34: 348-366,
'959-
57. Gribbe, P., J. Lind, E. Linko, and C. Wecelius. The
events of the left side of the normal heart as studied by
cineradiography. Cardiologia 33: 293-304, 1958.
58. Guasp, F. T. El ciclo cardiaco, consideracwnes aiticas sobre la
inter pretacion clmica y nuevas ideas sobre el mismo (Monograph)
Madrid : Medical Faculty of the University of Salamanca,
'954
59. Ham, A. W., and T. S. Leeson. In: Histology (4th ed.),
Philadelphia: Lippincott, 1961, p. 416, 533,
60. Hamilton, W. F. Filling of the normal human heart in
relation to the cardio-pneumogram and abdominal
plethysmogram. Am. J. Physiol. 91 : 712-719, 1930.
61. Hamilton, W. F. The physiology of the cardiac output.
Circulation 8: 527-543, 1953.
62. Hamilton, W. F., A. M. Attyah, D. W. Fowell, J. W.
Remington, N. C. Wheeler, and A. C. Witham. Do
the human ventricles eject simultaneously? Proc. Soc.
Exptl. Biol. Med. 65: 266-268, 1947.
63. Hamilton, W. F., Jr., P. Dow, and W. F. Hamilton.
Measurement of volume of dog's heart by x-ray: effect of
hemorrhage, of epinephrine infusion, and of buffer nerve
section. Am. J. Physiol. 161: 466-472, 1950.
64. Hamilton, W. F., and E. A. Lombard. Intrathoracic vol-
ume changes in relation to the cardiopneumogram. Cir-
culation Research 1 : 76-82, 1953.
65. Harrison, T. R., J. A. Lowder, L. L. Hefner, and
D. C. Harrison. Movements and forces of the human
heart. V. Precordial movements in relation to atrial
contraction. Circulation 18: 82-91, 1958.
66. Harvey, W. Exercitaiio anatomica de motu cordis et sanguinis
in animalibws. Frankfurt: Sumptibus Gulielmi Fitzeri,
1628. English translation by C. D. Leake, Springfield, 111.:
Thomas, 1928 and 1947.
67. Hawthorne, E. VV. Instantaneous dimensional changes
of the left ventricle in dogs. Circulation Research 9: 1 lo-
ng, 1 96 1.
68. Heeoer H., K. Polzer, and F. Schuhfried. Rheo-
kardiographic und Rcographie. Eleklromedirjn 4: 63-69,
'959-
69. Henderson, Y. The volume curve of the ventricles of
the mammalian heart, and the significance of this curve
in respect to the mechanics of the heart-beat and the
filling of the ventricles. Am. J. Physiol. 16: 325-367, 1906.
70. Henderson, Y., and F. E. Johnson. Two modes of closure
of the heart valves. Heart 4: 69-82, 191 2.
71. Henderson, Y., and A. L. Prince. The relative discharges
of the right and left ventricles and their bearing on pul-
monary congestion and depletion. Heart 5: 217-226, 1914.
72. Hennacy, R. A. Effects of epinephrine on frog ventricle.
Circulation Research 8: 831-836, i960.
73. Hennacy, R. A., and E. Ogden. Factors affecting the
filling of the frogs ventricle after isotonic contraction.
Circulation Research 8: 825-830, i960.
74. Hering, C. A., S. J. Wilson, and E R. Ball. Congenital
deficiency of the pericardium. J. Thoracic Cardiovascular
Surg. 40: 49-55, i960.
75. Hesse, H., and R. Minkus. Intrathorakale Bewegungs-
studie am Herzen im Selbstversuch. Z. Kreislaujforsch.
38: 613-616, 1949.
76. Hochrein, M. Der Mechanismus der Semilunarklappen
des Herzens. (Zugleich ein Beitrag zur Frage eines vollig
"verlustolosen" Schlusses derselben.) Deut. Arch. Klin.
Med. 154: 131-164, 1927.
77. Holt, J. P. Estimation of the residual volume of the
ventricle of the dog's heart by two indicator dilution
technics. Circulation Research 4: 187-195, 1956.
78. Holt, J. P., E. A. Rhode, and H. Kines. Pericardial
and ventricular pressure. Circulation Research 8: 1 171-1 181,
i960.
79. Holzlohner, E. Die Volumenanderungen in mensch-
lichen Thorax wahrend der Herzkation. Z. Biol. 92 : 293,
■932-
80. Horwttz, O. Contraction of cardiac muscle with respect
to time and its probable relationship to the ejection curve.
Am. J. Physiol. 165: 285-287, 1 951 .
81. Hosler, R. M. A Manual on Cardiac Resuscitation. Spring-
field, 111.: Thomas, 1954.
82. Hubacher, H. Die Darstellung der Bewegung des Mitral-
ringes mit phasengezielten Herzaufnahmen. Acta Radiol.
28: 386-390, 1947.
83. Irisawa, H., A. P. Greer, and R. F. Rushmer. Changes
in the dimensions of the venae cavae. Am. J. Physiol.
196: 741-744, 1959.
84. Isaacs, J. P., E. Berglund, and S. J. Sarnoff. Ven-
tricular function. III. The pathologic physiology of acute
cardiac tamponade studied by means of ventricular func-
tion curves. Am. Heart J. 48: 66-76, 1954.
85. Jamison, W. L., W. Gemeinhardt, J. Alai, and C. P.
Bailey. Artificial maintenance of the systemic circulation
without participation of the right ventricle. Circulation
Research 2: 315-318, 1954.
86. Kagan, A. Dynamic responses of the right ventricle
following extensive damage by cauterization. Circulation
5: 816-823, 1952.
87. Kantrowitz, A., E. S. Hurwitt, and A. Herskovitz.
A cinematographic study of the function of the mitral
valve in situ. In : Surgical Forum-Clinical Congress, Am.
College of Surgeons. Philadelphia: Saunders, 1 95 1 , p.
204-206.
88. Kaplan, S. In: Intra-V oscular Catheterization, edited by
H. A. Zimmerman. Springfield, 111.: Thomas, 1959, p.
80-139.
89. Katz, L. N. The asynchronism of right and left ventricu-
lar contractions and the independent variations in their
duration. Am. J. Physiol. 72: 655-681, 1925.
90. Keele, K. D. Leonardo da Vinci on Movement of the Heart
and Blood. Philadelphia: Lippincott, 1952, p. 62.
91. Kjellberc, S. R., and S. E. Olsson. Roentgenologic
FUNCTIONAL ANATOMY OF CARDIAC PUMPING
797
studies of the sphincter mechanism of the caval and pul-
monary veins. Acta Radiol. 41 : 481-497, 1954.
92. KOUWENHOVEN, W. B, J. R. JUDE, AND G. G. KnICKER- I 1 4.
bocker. Closed-chest cardiac massage. J. Am. Med. Assoc.
173: 1064-1067, i960.
93. Kraner, J. C. Effects of increased residual volume, in- 115.
creased output resistance and autonomic drugs on ventric-
ular suction in turtle. Circulation Research 7: 101-106, 1959.
94. Kraner, J. G, and E. Ogden. Ventricular suction in the 1 16
turtle. Circulation Research 4: 724-726, 1956.
95. Krug, H., and L. Schlicher. Die Dynamik des venbsen
Riickstromes. Leipzig: Thieme, i960, pp. 1-209.
96. Laszt, L., and A. Muller. Der myokardiale Druck. Helv.
Physiol. Ada 16:88-106, 1958. 117
g7. Lev, M., and C. S. Simkins. Architecture of the human
ventricular myocardium; technique for study using a
modification of the Mall-MacCallum method. Lab. Invest. 118
5: 396-409, 1956.
98. Licata, R. Anatomy of the Heart. Cardiology 1: 30-60, 119
■959-
99. Little, R. C. Volume elastic properties of the right and
left atrium. Am. J. Physiol. 158: 237-240, 1949. 120
100. Little, R. C. Effect of atrial systole on ventricular pressure
and closure of the A-V valves. Am. J. Physiol. 166: 289-
295. '951- I21
1 01. Little, R. C. Volume pressure relationships of the pul-
monary-left heart vascular segment. Evidence for a
"Valvelike" closure of the pulmonary veins. Circulation 122
Research 8: 594-599, i960.
102. Luciani, L. Human Physiology. London: Macmillan, vol. 1
(English Translation by F. A. Welby), 191 1. 123
103. Luisada, A. A., and F. G. Fleischner. Temporal rela-
tion between contraction of right and left sides of the
normal human heart. Proc. Soc. Exptl. Biol. Med. 66 : 436- 1 24.
440, 1947.
104. Luisada, A. A., and C. K. Liu. Intracardiac Phenomena in
Right and Lejt Heart Catheterization. New York: Grune & 125
Stratton, 1958.
105. Luisada, A. A., C. K. Liu, C. Aravanis, M. Testelli,
and J. Morris. On the mechanism of production of the
heart sounds. Am. Heart J. 55: 383-399, 1958. 126
106. Luthy, E., and W. Rutishauser. Die "Thermodilution"-
Methods. Cardiologia 38: 183-189, 1961.
107. Lynch, P. R., B. L. Carter, J. Gimenez, and R. Krisch.
Venae cavae flow pattern in cats: as studied with high-
speed cinefluorographic technique. Am. J. Physiol. 199: 127
1139-1142, i960.
108. MacCallum, J. B. On the muscular architecture and
growth of the ventricles of the heart. Johns Hopkins Hosp.
Rept. 9: 307-335, 1900. 128.
109. McKusick, V. A. Cardiovascular Sound in Health and Dis-
ease. Baltimore: Williams & Wilkins, 1958, p. 1-570.
1 10. Mackenzie, J. The Study of the Pulse Arterial, Venous, and
Hepatic and of the Movement of the Heart. Edinburgh: Young 129.
J. Pentland, 1902.
ill. Mall, F. P. On the muscular architecture of the ventricles
of the human heart. Am. J. Ana/. 11:21 1-266, 1910/1 1. 130,
112. Metcalfe, J., J. W. Woodbury, V. Richards, and C. S.
Burwell. Studies in experimental pericardial tamponade : 131.
effects on intravascular pressures and cardiac output.
Circulation 5: 518-523, 1952.
113. Mitchell, J., J. P. Gilmore, and S. J. Sarnoff. The
transport function of the atrium. Factors influencing the 132
relation between mean left atrial pressure and left ventricu-
lar end diastolic pressure. Am. J. Cardiol. 9: 237-247, 1962.
Monckeberc, J. G. Der funktionelle Bau des Sauge-
tierherzens. Handbuch der normalen und pathologischen
Physiologic 7: 85-113, 1926.
Monod-Broca, P. Recherches experimentales sur la
circulation pulmonaire apres exclusion du coeur droit.
Arch. mal. coeur. 51 : 841-846, 1958.
Moritz, F. Physiologie und Pathologie der Herzklappen.
III. Spezielles iiber den Herzklappenapparat bein den
hochstehenden Saugern einschliesslich des Menschen.
Handbuch der normalen und pathologischen Physiologie 7: 168-
199. '926-
Moscovitz, H. L., and R. J. Wilder. Pressure events of
the cardiac cycle in the dog. Normal right and left heart.
Circulation Research 4: 574-578, 1956.
Nelemans, F. A. Die Funktion des Perikards. Arch, neerl.
physiol. 24: 337-390, 1940.
Nerlich, W. E. Determinants of impairment of cardiac
filling during progressive pericardial effusion. Circulation
3: 377-383, i95!-
Nixon, P. G. F. Time relationships of the left atrial V
wave in mitral valvular disease. Brit. Heart J. 23: 637-
642, 1 96 1.
Nuland, S. B., W. W. L. Glenn, and P. H. Guilfoil.
Circulatory bypass of the right heart. III. Some observa-
tions on long-term survivors. Surgery 43: 184-201, 1958.
Nylin, G. The clinical applicability of roentgenological
heart volume. Determination with special reference to
the residual blood. Acta Cardiologica 12: 588-614, 1957.
O'Brien, L. J. Negative diastolic pressure in the isolated
hypothermic dog heart. Circulation Research 8: 956-960,
i960.
Opdyke, D. F. Effect of changes in initial tension, initial
volume and epinephrine on ventricular relaxation process.
Am. J. Physiol. 169: 403-41 1, 1952.
Opdyke, D. F., and G. A. Brecher. Effect of normal and
abnormal changes of intrathoracic pressure on effective
right and left atrial pressures. Am. J. Physiol. 160: 556-
566, 1950.
Opdyke, D. F., J. Duomarco, W. H. Dillon, H.
Schreiber, R. C. Little, and R. D. Seely. Study of
simultaneous right and left atrial pressure pulses under
normal and experimentally altered conditions. Am. J.
Physiol. 154: 258-272, 1948.
Patino, J. F., W. W. L. Glenn, P. H. Guilfoil, M.
Hume, and J. E. Fenn. Circulatory bypass of the right
heart. II. Further observations on vena caval-pulmonary
artery shunts. Surg. Forum 6: 189-193, 1 955.
Paul, R. E., Jr., M. J. Oppenheimer, P. R. Lynch,
and H. M. Stauffer. Regurgitation of radiopaque con-
trast material through normal mitral valves in cinefluoro-
graphic studies of dogs. J. Appl. Physiol. 12: 98-104, 1958.
Pfuhl, W. Die mechanischen Aufgaben des Herzbeutels
und seine Rolle bei der Wechselwirkung von intrathoraka-
ler Saugkraft und Herzkraft. Ana/. Anz- 67: 337-353, 1929.
Pfuhl, W. Die Herzoberflache und ihre praktische
Bedeutung. Anal. Anz- 68: 20-38, 1929.
Peiper, U., and K. H. Weigand. Die Bedeutung der
Kraft der Kontraktion fiir die diastolische Ansaungung
des isolierten Froschherzens. Pflugers Arch. ges. Physiol. 273:
407-409, 1 96 1.
Purkinje, J. E. Ueber die Saugkraft des Herzens. Jahres-
798
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
berichl der schlesischen Gesellschaft fur valerldndische Kultur,
Breslau, 157-164, 1843, Also in: In Memonam, Joh. Er.
Purkyne, 1787-1937; Operaonmia 2: 97-103, 1937, Sbornik
Stati, Prague.
133. Reeves, T. J., L. L. Hefner, W. B.Jones, C. Coghlan,
G. Prieto, and J. Carroll. The hemodynamic deter-
minants of the rate of change in pressure in the left
ventricle during isometric contraction. Am. Heart J. 60:
745-75'. i960.
134. Reindell, H., R. Weyland, H. Klepzig, E. Schildge,
and K. Musshoff. Uber Anpassungsvorgange und
Schadigungsmoglichkeiten beim Sportherzen. Schweir.
Z. Sportmed. 1 : 97, 1953.
135. Ring, G. C, M. J. Oppenheimer, H. N. Baier, J. H.
Long, A. Sokalchuk, L. L. Bell, D. W. Ellis, P. R.
Lynch, L. J. Shapiro, and L. D. Ichtiarowa. Estima-
tion of heart output from electrokymographic measure-
ments in human subjects. J. Appl. Physiol. 5: 99-1 10, 1952.
136. Robb, J. S., and R. C. Robb. The normal heart. (Anat-
omy and physiology of the structural units.) Am. Heart J.
23 : 455-467. >942-
137. Rodbard, S., and D. Wagner. By-passing the right ven-
tricle. Proc. Soc. Exptl. Biol. Med. 71 : 69-70, 1949.
138. Rothberger, C. J. Physiologie der Rhythmik und Ko-
ordination (abbr.) Ergeb. Physiol. 32: 472-820, 1 931 .
139. Rushmer, R. F. Cardiovascular Dynamics (2nd ed.). Phila-
delphia: Saunders, 1961.
140. Rusted, I. E., C. H. Scheifley, and J. Edwards. Studies
of the mitral valve. I. Anatomic features of the normal
mitral valve and associated structures. Circulation 6: 825-
83', '952-
141. Salisbury, P. F., C. E. Cross, and P. A. Rieben. Influence
of coronary artery pressure upon myocardial elasticity.
Circulation Research 8 : 794-800, 1 960.
142. Scher, A. M. In: Medical Physiology and Biophysics (18th
ed.), edited by T. C. Ruch andj. F. Fulton. Philadelphia:
Saunders, i960, p. 570-642.
143. Scheu, H., and VV. F. Hamilton. Evidence for left ven-
tricular suction in closed-chest dogs. Am. J. Physiol. 197:
"52. >959-
144. Schutz, E. Physiologie des Herzens. Berlin: Springer-
Verlag, 1958.
145. Segers, M. Le delai d'ejection ventriculaire droit et
gauche chez l'homme. Compt. rend, soc- biol. 143: 570-571,
1949-
146. Sjostrand, T. Volume and distribution of blood and their
significance in regulating circulation. Ann. Rev. Physiol.
33:202, 1953.
147. Sousa, A. de. Angioquimografia. Lisboa : Livraria Portu-
gal, '95', P- 1-240-
148. Spalteholz, W. Hand Atlas and Textbook oj Human Anat-
omy (revised by R. Spanner). Boston: Little, Brown, 1954.
149. Spencer, M. P., and F. C. Greiss. Dynamics of ventricu-
lar ejection. Circulation Research 10: 274-279, 1962.
150. Stephenson, H. E. In: Cardiac Arrest and Resuscitation.
St. Louis: Mosby, 1958.
151. Testut, L. In: Traite d'Anatomie Humaine (2nd ed.). Paris:
Librairie Octave Doin, 1921, p. 47.
152. Villa, L. Passivite ou activite diastolique? Semaine hop.,
Paris 30: 617-622, 1954.
153. Wagner, R. Feedback principle in regulation of the circu-
lation. Circulation Research 5: 469-471, 1957.
154. Warden, H. E., R. A. DeWall, and R. L. Varco. Use
of the right auricle as a pump for the pulmonary circuit.
Surg. Forum Proc. 40th Congr., Am. Coll. Surgeons 1954-1955,
p. 12-22.
155. Wetterer, E. Die Wirkung der Herztatikgeit auf die
Dynamik des Arteriensystems. Verhandl. Deut. Ges. Kreis-
laujforsch. 22: 26-60, 1956.
156. Wiggers, C. J. Studies on the consecutive phases of the
cardiac cycle. I. The duration of the consecutive phases of
the cardiac cycle and the criteria for their precise deter-
mination. Am. J. Physiol. 56: 415-438, 1921.
157. Wiggers, C. J. The independence of electrical and me-
chanical reactions in the mammalian heart. Am. Heart J.
1:3-20, 1925.
158. Wiggers, C. J. Studies on cardiodynamic actions of drugs;
mechanism of cardiac stimulation by epinephrin. J. Phar-
macol. Exptl. Therap. 30: 233-250, 1927.
159. Wiggers, C. J. Circulatory Dynamics. New York: Grune &
Stratton, 1952.
160. Willius, F. A., and T. J. Dry. A History of the Heart and
the Circulation. Philadelphia: Saunders, 1948.
161. Willius, F. A., and T. E. Keys (editors). Cardiac Classics.
St. Louis: Mosby, 1941.
162. Wilson, J. A., and W. J. Meek. The effect of the peri-
cardium on cardiac distention as determined by the X-ray.
Am. J. Physiol. 82: 34-46, 1927.
163. Zinsser, H. F., C. F. Kay, and J. M. Benjamin, Jr.
The electrokymograph: studies in recording fidelity.
Circulation 2: 197-204, 1950.
CHAPTER 24
The physiology of the aorta and major arteries1
JOHN W. RE MINGTON Department of Physiology, Medical College of Georgia, Augusta, Georgia
CHAPTER CONTENTS
Measurement of Aortic Distensibility
General Characteristics of the Tension -Length Curve
The Hysteresis Loop
Selection of Representative Curves
Histological Considerations
Effects of Active Muscular Contraction on Distensibility
Effects of Aging on Arterial Distensibility
Expression of Extensibility in Terms of Moduli
Changes in Length and Wall Thickness of Arteries
Action of the Aorta as a Conduit
Pulsatile Flow in Rigid and Distensible Tubes
Quantitation of Fluid Displacement and Wall Distensibility
Relationships
Phase Lag and the Harmonics of the Arterial System
Construction of a Hypothetical Ejection Curve
The Aorta as a Blood Reservoir
Changes in Central Pulse Contour During Propagation
Resonance and Standing Waves
Other Factors Which May Alter the Central Pulse Contour
Calculation of the Stroke Volume From the Central Pressure
Pulse
AT outset, may I say that this article on the function
of the aorta and the major arteries makes no pretense
of being an authoritative review of the literature,
and is not only a generalized treatment but is written
with a bias. I came to cardiovascular study via Biol-
ogy, in an era when Physics was not such a firmly
trothed bride of Physiology- I do not think glibly in
terms of abstract formulas, of electrical analogues, or
the other erudite devices now so commonly used to
clarify the complex problems which underlie pressure
wave formation and propagation. I believe that I
understand a process only when I can construct some
1 This manuscript was completed January 15, 1 961 , and its
references include only papers I had read in published form at
that time.
sort of a visual image of just how it operates. In many
aspects of the subject such a visual model is at present
impossible. I can only describe what I have been able
to gather about the function of the major arteries, and
speculate about what trends future research will
follow.
These large arteries serve two clear functions. First,
they comprise a network of conduits through which
blood is moved from the centrally located cardiac
pump to the various capillary beds. It is important
that this transfer be made with a minimal loss of
energy. The problem of proper conduit design became
more acute when the body form became elongated,
instead of remaining spherical. Second, the distensible
wall of the vessels allows a temporary storage of blood
during the ejection phase of the pump cycle, which
allows a buffering of the oscillatory pressure changes.
This aspect will be spoken of as the reservoir action of
the vessels, admittedly an inadequate label. While
the buffering action might serve to protect the small
vessels from large pressure changes, it also involves
considerable change in the conduit properties.
We cannot yet form a definitive analysis of the
effectiveness of aortic design in meeting these funda-
mental requisites. Progress has been handicapped
because it has been so difficult to make critical studies
on intact vessels. Pressure changes at various points
along the arterial system have been measured often
over the past 50 years (1, 16, 18, 28-30, 40, 42, 99,
116, 132, 134, 135). This has given us knowledge,
still far from complete, about the speed of pulse-wave
propagation, the contour of the pressure pulse as
formed in the upper aorta, and the changes in this
form as the pulse moves into the distal aorta and large
arteries. Itisdoubtful that any great progress toward an
understanding of arterial dynamics will be made by
799
8oo
HANDBOOK OF PHVSIOLOGY
CIRCULATION II
further studies on pressure values alone, without
a simultaneous recording of vessel diameter or flow.
A study of two parameters has often been at-
tempted, but not too successfully. It has not proven
practical to remove the aorta from the body and
insert it into an artificial system where the volume
change and the flow through might be measured
directly, as from a calibrated stroke of a pump. The
problems of coupling this distensible tube to rigid
fittings without having an orifice that will severely
distort the flow pattern are considerable. No effective
means has been devised to occlude completely all
exit vessels, including the vasa vasorum, and thus
prevent loss of fluid along the length of the vessel.
Sometimes a rubber insert has been used (137) but,
since most rubber tubes are less extensible than the
aorta, this stratagem may have confused matters
more than it helped. And we still have no pump which
has an ejection pattern like that of the ventricle.
Curiously enough, an artificial pump capable of
producing a pressure rise similar to that seen with a
natural pulse invariably produces turbulence and
vibrations of the pressure recorder sufficient to obscure
the pulse contour being formed.
A recording of aortic flow in vivo by a technique
which requires vessel cannulation causes enough dis-
tortion of the pressure pulse contours that one must
be cautious in inferring a direct pertinence to the
intact system. Fortunately, several techniques are now
in use for the recording of flow (27, 31, 55, 82, 110,
120, 131) and the registration of diameter change (88,
91, 113) which do not require cutting the vessel.
Most flow recorders do require a crimping of the
vessel in the region where flow is being measured,
which may not be without effect on the flow profile.
As yet, the frequency of many such devices usually
does not approach that of a good pressure recording
system, so that they may not be able to give a faithful
picture of rapid change.
An engineer faced with the problem of designing a
conduit system for the most efficient movement of
blood would start with some basic equations. First,
there would be the Poiseuille formula which states
that the pressure fall (for frictional energy dissipation)
will be directly related to the flow rate, the fluid
viscosity, and the length of pipe, and inversely related
to the fourth power of the radius. The last is because
adsorptive forces between the fluid and the wall prevent
or retard longitudinal movement of the outermost fluid
layer. This in turn forces the adjacent shell to shear
past it, retarding it with a frictional dissipation of
energy, which in turn slows the next shell of fluid, and
so on to the middle of the pipe. For a given volume
flow, the greater the pipe diameter, the less is the
total fluid frictional loss. He would also include in his
formulas factors relating to the smoothness of the
wall and the material of which his conduit will be
made, since these condition the size of the boundary
layer. He must make corrections if the fluid does not
have a constant viscosity at all flow rates, as blood
apparently does not (9, 86). He also knows that an
equation which applies to laminar flow will not be
correct if fluid molecules whorl laterally across the
fluid shells, i.e., when the flow becomes turbulent.
The frictional cost increases whenever this happens.
Finally, when he is required to use pipes of different
sizes, he must carefully design the transition areas so
that turbulent eddies will not form. Tapered changes
in diameter are less conducive to turbulence than
abrupt shoulder joints.
All these formulas, which may be found in texts on
hydraulics, are based on the assumption that flow is
being maintained steady, and that the conduits have
rigid walls. But blood flow is not steady, for it shows
several accelerations and decelerations with each
pump stroke. A calculation of the energy loss accom-
panying such rapid changes in flow rate must be
complicated. A whole new set of equations will be
required and, to check them, we must be able to
measure precisely the amount of acceleration in all
parts of the arterial system. Quantitative evaluations
of the degree of smoothness or of the absorptive forces
for plasma on the endothelial lining cannot be given.
Microscopic observation of moving blood in tiny
vessels has shown that the red cells congregate in the
center of the stream, which is presumptive evidence
for a greater axial velocity. Whether we can assume
from this that a parabolic flow distribution would be
found in the large arterial vessels is open to question.
While streamline flow has been described for the
aorta (92), there remains some question as to whether
a normal flow pattern could be said to have persisted
during the measurements, and whether the streaming
would apply over the whole cardiac cycle (84).
The large arteries are not rigid, so that any equa-
tion relating energy dissipation to tube radius will be
complex. Further, the arterial system seldom has any
tubes which continue uninterrupted for an apprecia-
ble distance. Each vessel has frequent branchings.
With the exception of the ascending aorta (44) and
the main pulmonary artery (88), with each branching
there is an increase in the aggregate cross-sectional
area. These junctions appear smooth and tapered, so
that the orifice problem is probably at its simplest.
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
80 I
Further, most trunk vessels show a gradual taper
through their length. Exceptions to this, that come to
mind, are a region of the descending thoracic aorta
and one of the carotid artery, which appear to be
more nearly true cylinders.
Too detailed a particularization of the various
factors which give rise to frictional resistance may be
nonessential. The total effect of them all should be
measurable by a decrease in mean pressure, over a
whole cardiac cycle, from the upper aorta to the
peripheral arteries. Several studies have shown that
in the aorta such a decrease is so small as to be within
the error of measurement (42, 68). In fact, there is no
clear loss in mean pressure in man until the brachial
or femoral arteries are reached.
Despite this small frictional dissipation of energy
attending propagation, there is a very clear differ-
ence between the pressure energy developed in systole
and in diastole. Except under rare circumstances, the
mean systolic pressure is greater than the mean
diastolic pressure. This excess of pressure energy
could denote an inability of the stretch of the extensi-
ble wall to keep pace with the force applied by cardiac
ejection, so that energy is stored in potential form in
the visco-elastic walls, or it could indicate a different
pressure-flow relationship in the large vessels being
filled during systole, and that which marks "drainage"
through the peripheral arterioles.
To go from generalities to the specifics, an analysis
of aortic function could be focused upon three large
questions: /) What are the essential features of the
tension-length curves shown by the walls and the
derived pressure-volume curves, and to what extent
are these curves subject to physiological and patho-
logical change? 2) How does wall distention affect the
conduit properties of the vessels? j) What factors
influence the capacity of the arteries to serve as a
blood reservoir?
MEASUREMENT OF AORTIC DISTENSIBILITY
General Characteristics of the Tension-Length Curve
Until quite recently, measurements of the extensi-
bility ot blood vessels were made on isolated tissues,
using two procedures (11, 15, 22, 37, 51, 62, 65-67,
76, 107, 118). Usually a ring (for circular stretching
to produce an increase in circumference) or a cut
strip (for measuring longitudinal change) was sub-
jected to weight loads, the changes in length being
recorded. In a few cases, volume was injected into a
tied-off vessel, recording pressure. Any change in the
other dimension, e.g., a longitudinal change during
circular stretching, was either inadequately measured
or ignored. Although the specific techniques for in-
creasing load have varied, the stretches were made
rather slowly so that the vessel could approach, if
not attain, a stable length value, i.e., a "static" value.
Whether the load was applied in a continuously in-
creasing manner or stepwise, the data were generally
presented as a single tension-length curve covering
the whole physiological range. All workers agree that
such a curve is not linear, but shows a relatively
great extensibility at low tension settings and a
progressive wall stiffening as the load increases (fig.
1). This curve is therefore different from that shown
by metals, even those that obey Hooke's law over the
greatest part of their extension, or by rubber, where
the length change becomes relatively greater at high
tension levels (46). A rubber tube wrapped with a
fibrous jacket, such as a garden hose, shows the same
type of curve as does the aorta (14). Rings taken from
the aorta or from arteries appear to differ only quan-
titatively. Further, the longitudinal stretch curves are
qualitatively similar to those obtained with a circular
stretch.
The tension given in figure 1 A is the weight load
divided by the product of the length of the ring of the
thoracic aorta and the wall thickness. This can be
converted to internal pressure by dividing by the
radius. To express pressure in the usual physiological
terms of mm Hg, the obtained value is divided by 1 3.5.
We can calculate the volume per unit length of vessel
as tit2. Both pressure-diameter and pressure-volume
curves show two inflections, to give the curve a some-
what sigmoid appearance (figs. \B and C). The
pressure level at which these inflections are seen varies
with different regions of the aorta. Hence both inflec-
tions are set at a higher pressure in the upper aorta
than in the lower, and the lower inflection may not
be seen at all in the arteries (46). There is no simple
formula which will fit this sigmoid type of curve, or
even that portion of greatest physiological significance.
At outset, it is clear that any comparison of vessel dis-
tensibility from time to time, or between animals, will
require the use of the same arterial region and the
same pressure span.
There are several inadequately explained properties
of the isolated specimens which seriously affect the
recorded extensibility curves. First, as the vessel is
dissected out of the body, there is an immediate
shortening of its length and a tensing of its walls.
This is true whether the animal has just been killed,
802
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
1000-1
800
600 - oj
500-,
500-,
400
200-
i !0 i0 40 50 20 J5 76 ?0
Circumference, mm Ci rcumference, m m
fig. I . Stretch curves for a ring of thoracic aorta of a dog. In situ length = 10.5 mm. Curves
I represent the first stretch curve, made by continuous tension increase over 1 min. Curves 2
are the results of a second stretch identical in load and timing to the first. Curves R show the
curves taken during the gradual release of tension, over 1 min. Curves C show the effect of
muscle contraction by immersion of the ring in epinephrine.
I 2
Volume, cc
or has been dead for some time, and whether the
voluntary muscles are in a state of rigor mortis or not.
The change develops no matter how* carefully the
removal is done. The interpretation placed upon this
change in the past is that it reflects a strong contrac-
tion of the smooth muscle contained in the wall.
Aside from the speed of its development, which con-
trasts with the slower time courseof muscle contraction,
and the lack of correlation between the amount of
longitudinal shortening and the proportionate amount
of muscle in the wall (ascending aorta, for example,
contracts to the greatest degree), there are other
features which do not fit too well with this interpre-
tation.
When such a tensed ring is subjected to stretch,
and the load is then removed, the walls are no longer
so tense, and the circumference is about 30 per cent
greater than before the stretch. What was not real-
ized in the earlier studies was that the amount of this
diameter increase bears a direct relation to the total
stretch imposed (96). Now, if a second stretch of the
same size is made, the stretch curve starts from the
greater initial value and courses almost parallel to
the first through the region of greatest extensibility
(fig. 1 A). Then, as the wall stiffens, the second curse
becomes enough steeper that it merges with the first
some time before the peak load is reached. This
merging argues against a conclusion that the first
stretch had caused some irreparable tissue damage,
such as an internal tearing. If, after the first stretch
and stretch release is completed, the tissue is allowed
to remain unloaded for a long period (several hours),
the original small diameter may be almost, if not
fully, restored. If the first stretch had ""pulled out" an
existing muscle spasm, why should it reform and take
so long in doing so? Experiments designed to test this
"pulling out" idea with smooth muscle organs, such as
the gut, have given little indication that an active
contraction itself is eliminated by an extension of the
whole tissue under load (6, 101).
With a relatively large load, a third stretch made
after the second usually gives a stretch curve identical
to that of the second. This fact has been recognized
by many past workers who were aware that a large
initial stretch, which was not quantitated, made
subsequent stretch curves more reproducible. Such
reproducibility does not mean that they are necessarily
more descriptive of the behavior of the vessel in vivo.
With a similar load, successive stretches may start
Irom progressively increasing initial diameters, but a
reproducible stretch curve is reached within 5 to 10
stretches. Unfortunately, when examining past
studies, it is impossible to be sure whether a prelimi-
nary stretch was used, and, if so, how much tension
was involved and how long a time interval was al-
lowed before the recorded stretch was made.
The Hysteresis Loop
Until recently, too, little attention was paid to the
fact that as the loads were removed the length curve
did not follow, during this stretch release, the previous
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
8o3
stretch curve (fig. 1). Hence for any given stretch two
different tension-length curves must be considered,
one for the extension and the other for the elastic
recoil. The difference between the two comprises a
hysteresis loop. The amount of hysteresis is always
greatest with the first stretch done after a prolonged
rest period. If this first stretch is followed by repetitive
stretches of the same size, the hysteresis is progressively
reduced until it becomes relatively constant. The
number of successive stretches required to achieve
this stable loop varies among vessels, for two or three
stretches will suffice with the aorta, but more may be
needed with a muscular artery. In the gradual re-
duction of hysteresis, the stretch-release curve
remains almost or completely constant, its values
being set by the peak load used (96). It is the exten-
sion curve which shows progressively larger length
values at any given tension value. Hysteresis is present
in some metals, too, although its amount is relatively
small as compared to that seen with vascular tissues.
The presence of hysteresis is often taken to indicate
simplv a viscous retardation of the extension of elastic
elements, and handled in formulas as though it were
simple viscosity (48). This would mean that the size
of the loop is an index to the frictional energy dis-
sipation, which, in turn, would be directly related
to the rate of the imposed stretch. Hysteresis of the
vessel wall is not so easily formulated. We can sum-
marize its main features by saying that at least three
factors seem involved.
/) While a viscous retardation is present, it can be
demonstrated for the aorta only at very rapid rates
of stretch (96). A rate dependency has not been seen
for the stretch-release curve. The muscular arteries
have more rate-dependent hysteresis than does the
aorta.
2) When a stretched length is held constant for a
period of time, some internal elongation of elements
still continues, so that the tension falls. This decline
is called stress relaxation. Or if a tension value is
maintained, the length will continue to increase
slowly, which process is known as "creep." The
amount of creep is a function of time, but the relation
is not easily formulated in quantitative terms.
Muscle physiologists have frequently referred to
this slow continued elongation under stress as plasticity
(13, 93). This use of the term "plastic" is not very
appropriate. With metals, when an applied increasing
stress reaches a certain critical value, the material
becomes deformed. The length change accompanying
this deformation may show the properties of viscosity,
but the term plastic does not denote the presence or
absence of such viscosity. Once deformed, the material
does not return to the original length upon removal
of the stress, but retains the increased length. The
choice of the word plasticity for the slow elongation
of muscle lay with the belief that any reversibility
could be brought about only by an active muscle
contraction. However, the process which underlies
stress relaxation is spontaneously and completely
reversible, if enough time is allowed, whether the
muscle is alive or dead (101). Muscle contraction
may, of course, hasten the return to the original
length. Stress relaxation involves a complicated type
of internal viscosity which is so arranged that the
driving force for length return lies with some parallel
elastic unit which is under stretch.
Just as with rate-dependent viscosity, the stress-
relaxation component is but a minor part of hysteresis
as seen in the aorta (96). Its influence is more evident
the longer the vessel is kept at a peak load, or the
longer the vessel remains under no load, so that creep
recovery, or the reversible phase of stress relaxation,
can continue.
3) When, with successive stretches, a final "stable
hysteresis loop" is obtained, neither the values from
the stretch phase nor those from the release phase
show any dependency upon the rate of stretch, and a
dependence upon time cannot be easily described.
For want of a better term for the remaining factor,
which seems to dictate the greatest part of the hyster-
esis behavior, I have called it simply an architectural
rearrangement. The change is certainly dependent
upon the amount of stress and involves a reversible
change in length. While there must be some time
dimension to this internal rearrangement, the change
presumably is very rapid. It may be at a molecular
level or at a tissue fiber level.
What should be firmly emphasized is that a tissue
probably has a great many different viscous elements
with different time constants. When we refer to such
a tissue as being visco-elastic, it does not mean that
all the different viscosities can be lumped to give a
single viscosity, with an easily definable rate de-
pendency.
In view of the complexities that influence a tension-
length curve, it is possible that we should think of the
firming of the aortic wall, upon removal from the
body, in terms quite different from that of a muscle
contraction. The vessel is held in situ under con-
siderable longitudinal restraint (104, 107). When a
segment is cut, elastic elements held lengthwise under
stretch should recoil and make the wall thicker. When,
with a circular stretch, no attempt is made to restore
8o4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the original length, there could be a reorientation of
these elements into the circular plane, leaving a ring
with a larger circumference. This indicates, as is
probable, that both longitudinal and circular elastic
elements are part of a linked network, and therefore
not independent. In our studies on isolated rings,
when we converted the actual tension-length values
to pressures and volume, we used in situ length
rather than the actual one. This mathematical step
was better than using the actual lengths, but is not
necessarily sufficient as a correction if elements
previously oriented longitudinally were participating
in the circular stretch.
This is not to say that all the hysteresis phenomenon
could be due to a progressive recruitment of longi-
tudinal fibers into a circular plane. The loop still
present, despite many consecutive stretchings, prob-
ably denotes a structural rearrangement of elements
under load that were already in the plane of the
stretch. It is admittedly strange that such a rear-
rangement would have no clear time dimension.
An occluded, in situ aorta being pulsed by volume
injections shows a similar pattern of hysteresis and
change in extensibility curve with successive stretches
(5). We attempted a quantitative assessment of
possible differences between total aortic distensibility,
as measured by injections of saline into dead but in
situ aortas, and as compiled from stretch data made
on rings cut serially from the same vessels (103). The
volume required to produce a given rise in pressure
was greater than that estimated from the first stretch
curves of the isolated rings. It was also a little greater,
although perhaps not significantly so, than that
expected from the second stretch curves.
On hindsight, this comparison may not have been
so definitive as we supposed. At the time the experi-
ments were done, we were not so aware of the large
effect of the time interval between successive stretches,
of the initial pressure level, and of the peak pressure
reached, on the contour and values of the extension
curve.
Selection of Representative Curves
Historically, an interpretation of vessel wall archi-
tecture has been based on the contour of a first, or a
second, large and continuous extension curve. It
should now be obvious that this contour should not
be regarded as characteristic of wall extension during
unceasing, repetitive stretches, such as would be
present during life. How large the differences between
the two curves might be will be elaborated more
fully later in this paper. Ultimately, a study of vessel
wall behavior must be based on pressure and diameter
or volume measurements made during life. Such
studies will be technically difficult, and interpretation
of the records will be difficult, since in life the heart
rate and the pulse pressure are continually changing
from beat to beat. Work on this problem is in progress
in a number of laboratories, and some results have
been published (87, 91, 113). Unfortunately, these
show some differences, and the adequacy of the
various techniques has yet to be firmly established.
But taking the data as they now stand, it appears
possible that while the amount of hysteresis varies
among arteries, it is probably less than that shown
by the isolated rings. This may be because the living
aorta is being cyclically stretched without pause.
The suggestion, which needs further corroboration,
that the diameter change of a living vessel for a given
pulse pressure is less than that given by an isolated
ring (61, 78, 87, 91) could indicate that, for some
quite unknown reason, a very different distensibility
is present in the living vessel. In one study (91) the
reported diameter change is so much smaller that our
whole concept of the functional character of the vessel,
as will be developed in this paper, would have to be
changed. We are not satisfied yet that the instrument
used could record a change as large as that expected
from the ring stretch data. Reconciliation between
the various sets of data should not be long delayed.
Until the properties of the in vivo aorta are known
more surely, we must base a description of the factors
which might condition the extensibility curve on
data taken from isolated rings. If these data should
eventually prove quantitatively wrong, we can only
hope that the fundamental properties of the vessel
would nonetheless be qualitatively the same. First,
the nonlinearity of the continuous stretch curve is
evidence for an internal architecture more com-
plicated than that seen even with rubber or other
polymers. In the range of maximal extensibility, the
aorta shows more length change per unit tension
increase than anv other material of comparable wall
thickness. Nature seems to have created a far better
volume reservoir than man can duplicate.
Histological Considerations
We know from histological evidence that the large
vessels have four general types of tissue — endothelial
cells (with associated intracellular materials), smooth
muscle cells, elastic fibers, and collagenous fibers.
Because histology texts tend to emphasize a collection
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
80:
of these tissues into more or less well-demarcated
layers, physiologists have taken what probably is an
oversimplified approach to an analysis of wall ex-
tensibility, and have considered each of these tissue
types to be unconnected and arranged in parallel.
But while elastic tissue does appear to be condensed
into layers, it also is interspersed between muscle and
collagenous fibers. And there seem to be connections
between the gross layers themselves, which means
that we can hardly consider the influence of any one
tissue type alone. Nor are we at all certain about the
elastic characteristics of the different tissues, even if
they were to behave independently. The extensibility
of the endothelial cells is relatively unknown in any
quantitative sense. Since they comprise but a very
small part of the wall, and since they certainly are
not stiff enough that they are torn by even large
stretches, they probably are relatively extensible, and
their influence can probably be neglected for the
present. An estimate of the extensibility of collagenous
fibers has been taken from that shown by tendon,
although the collagenous fibers of the latter are larger
and more densely packed. As opposed to a tissue
containing elastic fibers, tendon is quite inextensible
and shows a linear stress-strain relation with no
discernible visco-elastic properties (21, 52, 67, 97,
103, 112). The aortic wall stiffening seen when the
pressure rises above about 100 mm Hg has thus been
attributed to the resistance to stretch offered by the
enclosing jacket of collagenous fibers (21, 103). Aortic
walls from which elastic tissue and muscle have been
digested show a similar stiff wall (50, 1 1 1 ). This
jacket must fit loosely and be considered in parallel
to the underlying elastic tissue. Once it starts to
participate in the stretch, it will assume the bulk of
the applied load.
On the basis of stretch curves shown by ligamentum
nuchae, which is predominantly elastic tissue, elastic
fibers are much more extensible than collagenous
fibers. Chemically treated aortas which retain only
their elastin show an increased extensibility, one not
greatly different from that of the whole vessel in the
lower stretch range (51, 52, 74, 111). Most workers
ascribe a linear stress-strain relation to these fibers,
too. Hence my work (97) seems to stand alone in the
claim that ligamentum nuchae shows a nonlinear
curve not unlike that of the arteries (with a stiffening
in the upper tension range despite the absence of any
clear collagenous fibrous coat), and also has visco-
elastic properties similar to those of the aorta. Since
the elastic fibers in both organs seem arranged in a
reticulum, and since the visco-elastic properties of
dog ligamentum nuchae seem clear, the following
analysis will be based on a similarity in stretch be-
havior of the two tissues.
Assessment of the extensibility of smooth muscle
cells is on most insecure ground. Studies have been
made of the stretch properties of muscular organs,
such as the bladder or gut, for this purpose (6, 97,
101, 102), although the muscle fibers here are en-
meshed in a loose weave of collagenous and even some
elastic fibers too. If this muscular tissue is subjected
to a moderately rapid stretch, its extensibility is only
about a third as great as that shown by ligamentum
nuchae (97). But because these tissues have such a
pronounced time-dependent creep, if one waits for
the length to approach a final value under a given
load, the total extensibility is greater than that of
elastic tissue. This has been the procedure used when
elastic moduli for muscular structures have been
derived. But it seems unreasonable that the muscle
contained in the aortic wall could show such a pro-
longed creep. If the muscle cells are coupled to the
elastic fibers, creep would be effectively prevented by
the resistance these fibers would offer to an elonga-
tion. Of course, the greater the amount of muscle
involved in the vessel wall, the greater would be the
creep. At one extreme, the umbilical artery, which is
almost solely muscle, shows a very pronounced stress
relaxation under load (122, 141). In most large
arterial vessels the relaxation is of more limited
degree. Hence the muscle contained would probably
be stiffer than the elastic fibers, which would remain
the most extensible part of the wall.
It should be noted that the pulmonary artery shows
a difference in distensibility behavior from the aorta.
The form of the stretch curve is more akin to that of
a large vein (105). The vessel shows more creep than
does the aorta (32). The form and the total length
change of the stretch curve are different in pulmonary
vessels that have been frozen and thawed than when
simply kept in Ringer-Tyrode solution (a reflection of
the effect of viable muscle?) (32).
There is histological evidence that, in the aortic
wall at least, muscle cells, elastic fibers, and some
collagenous fibers are interlinked into a three-
dimensional network (11, 109). The elements in this
network could be partly in series and partly in
parallel. The extensibility of the whole tissue could
be a reflection of the form of the net just as well as
it could be conditioned by the individual tissues. For
example, Bull (20) showed that while a single nylon
thread obeyed Hooke's law, and had no visco-elastic
behavior, a stocking woven from such a thread
8o6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
showed a bowed extensibility curve not unlike that
of the aorta or of ligamentum nuchae, and also
showed a pronounced hysteresis loop. When f
stretched a stocking by the techniques used for an
isolated ring of aorta, the dry specimen showed an
appreciable rate-dependent element (viscosity) in the
tension response to a given strain. When the stocking
was wet, this viscous element was relatively reduced,
and there was unmasked both a prolonged creep and
the "architectural dependency" which is so con-
spicuous for the hysteresis behavior of the aorta.
If the analogy of the stocking is valid, the first
part of the stretch curve of the aorta would reflect
only a geometrical rearrangement of the net. The
resistance to stretch would be a function of the loose-
ness of the "weave" and the presence of a lubricant
(as in the wet stocking); there also could be a "set"
of the net, which could be subject to change with
time, with muscular activity, and, very definitely, be
influenced by the size of a previous stretch. When,
under applied load, the net lost its form, the ex-
tensibility would progressively decrease, both because
the mechanical advantage of the fibers in resisting
the stretch would be increased and because the fibers
themselves would now be involved in the extension.
If our ideas of the relative extensibilities of the
different components is correct, and if they were
arranged in the net in series, the elastic fibers, being
most extensible, would condition the extension of the
whole wall. With more load, these elastic fibers would
become stiffer (as they do in ligamentum nuchae),
and other components of the net could be increasingly
involved. Probably the idea of a parallel outer
collagenous jacket should still be retained to con-
tribute to the final wall stiffness.
In an earlier analysis (103) we treated the aorta
as though it contained the three tissue types as
arranged in parallel. Since muscle had to be able to
reduce the vessel diameter below its normal unloaded
size, we conceived of the elastic jacket as fitting
loosely over the muscle coat. This would mean that
muscle alone would be involved in the very first part
of the stretch curve, and that only later in the stretch
would the elastic fibers start to participate. Such an
arrangement seemed amply supported by evidence
obtained with stretchings repeated daily, using rings
as they were allowed to putrefy. In this process muscle
cells lost their integrity first and the unloaded diameter
increased while the initial slope of the stretch curve
became steeper. Much later, the elastic fibers softened
and their continuity became disrupted. Now the
unloaded diameter had again increased, and the aorta
showed a stiffness not unlike that seen at high load
levels in the normal state, which was attributed to the
collagenous fibers still present. The net model would
fit these putrefaction studies equally well, for loss of
muscle could partly disrupt the net to give an increase
in unloaded diameter and, at the same time, leave the
wall less extensible.
There still remain several features of the visco-
elastic behavior of the aortic wall which would not be
easily explained on the basis of the net. And the
details of net construction are left purposefully vague.
The general concept has much in common with the
model proposed by Burton (21), except for its de-
emphasis of the specific location and role of the
muscle fibers themselves. He was much concerned
that the muscle be afforded a great mechanical
advantage, so that it could always effect a diameter
change. Hence he placed these fibers across the plane
of a fibrous net, which would protect them from
elongation. In muscular tissues it remains uncertain
that the contractile ability of muscle fibers is neces-
sarily impaired when they are elongated, even by the
amount that may normally be developed in an organ
such as the urinary bladder. Further, it may be that
even in smooth muscle organs the muscle cells are
arranged into a somewhat similar net (101, 102). It
may be that the muscles in the aortic wall are at-
tached to adjacent loops of the net (which would
give them more of a parallel arrangement than a
series one with the elastic and collagenous fibers), so
that they could, by shortening, act to "open the
weave," and perhaps increase its "set." This need not
mean that the stiffer muscle would now condition the
extensibility curve, for we could have, with extension,
a warping of the net toward these muscle links. Hence,
the internal architecture could be quite different
when muscle was contracted than when relaxed,
even though the diameter values under a given load
might be the same.
Effects of Active Muscular Contraction on Distensibility
Whether it is necessary that a model provide
muscle with a large mechanical advantage cannot be
answered. We are not sure just how effective muscle
contraction really is in a vessel under a load equivalent
to that of the usual physiological pressure values.
Much work is currently being done in which strips
of aorta are used as conveniently long tissues to test
the effect of drugs, or changed electrolyte environ-
ment, on muscle contraction (35, 79, 80, 121). To be
useful as a bio-assay material, such strips must be
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
807
under minimal load. Isolated aortic rings will, when
unloaded, respond to immersion in epinephrine, for
example, by a shortening of both their diameter and
length, to produce a pressure rise of some 3 to 5 mm
Hg. But if these contracted rings are subjected to
stretch, the decreased diameter is lost rather early
(fig. 1A), so that by the time loads equivalent to the
usual working pressures are reached, the contracted
and relaxed rings show identical extensibility curves.
With stretch release, the diameter does not return
to the contracted size. Either the muscle loses its
contraction early in the stretch, or the other parts of
the net have contributed more than usual to the total
extension. Even in muscular organs, such as the
bladder (102), the effect of contraction on the ex-
tensibility curve is small, and the contraction itself
seems not to be eliminated by the imposed stretch
(97, 1 18). The extensibility of a muscular organ is not
very different in the contracted or relaxed state (6,
'3)-
What is more disquieting is that if an aortic ring
is first subjected to a load equivalent to a pressure in
the usual physiological range, immersion in epi-
nephrine will no longer produce a discernible diameter
or pressure change. It is hard to accept this finding
as rational. Yet neither viable isolated specimens nor
a temporarily occluded aorta in situ (5) has been
shown to have a more powerful muscle action. This
is not to say that a contraction in muscular arteries,
where the ratio of wall muscle to internal diameter is
greater, could not influence the diameter at the
higher pressure levels.
Attempts have been made to record the effects of
muscle contraction in the intact aorta while it is
being pulsed by the heart. Most of these I have
learned of through conversations, since there is
reluctance toward publication of negative findings.
In the literature are the older experiments of Wiggers
& Wegria (138) in which an aortagraph was placed
around the thoracic aorta of a dog. After an intra-
venous injection of epinephrine or elicitation of a
strong pressor reflex, there was a recorded decrease
in diameter (the actual values not being given) at a
time when the aortic pressure was not changing. For
many years these results stood unchallenged and yet
unsupported. More recently, Patel and co-workers
(88) found a change in both diameter and wall
stiffness in the main pulmonary artery with muscle
contraction, a change persisting through several
normal pulsations. The pulmonary pressure is, of
course, much lower than that of the aorta and the
wall architecture is not the same. Then Peterson and
co-workers (91) showed a change in diameter and an
increased stiffness, with arterial pressure unaltered,
when the femoral artery or carotid artery was painted
with norepinephrine. Opposite results were obtained
with acetylcholine. The authors claim a similar
directional change, but furnish no supporting figures,
for the aorta. These results, and particularly the
claim for the aorta, must be amplified and confirmed.
Diameter and extensibility changes in the aorta of
living animals following the use of constrictor or
dilator drugs have been recorded (78) which do not
appear to fit with the stretch data obtained with
isolated rings. Since the arterial pressure also changed,
and since the physiological distensibility curve for
the intact aorta, quite aside from any muscle action,
remains to be formulated, an attempt to interpret
these changes on the basis of muscle contraction
would be premature.
We are not yet in a position, then, to answer the
long-debated question as to whether muscle con-
traction should increase or decrease the wall ex-
tensibility. The effects of such contraction on the
stretch curves shown by isolated vessels are difficult
to phrase in terms of generalities. With an unloaded
vessel, contraction is not followed by relaxation.
Instead, over a period of time, the wall gradually
becomes stiffer, as though the fibers had been reset
to the shorter length. Whether the contraction had
any essential role in this resetting, aside from the
first reduction in length, remains uncertain. De-
pending upon the amount of this resetting, a small
stretch, starting from zero tension, will reflect this
initial stiffness of the wall. With more stretch, the
distensibility suddenly increases, so that the stretch
curves of the once contracted and the relaxed ring
are now parallel. At higher loads the two curves
merge. If one assays distensibility, it must always be
with respect to the amount of load used. Thus, if the
stretch is sufficient to cause the shift toward the
increased distensibility, one would conclude that the
muscle contraction had rendered the wall more
extensible. If the assay were on the basis of the very
first part of the stretch curve only, one would reach
the opposite conclusion of a lessened extensibility.
Whether the effects shown by an unloaded tissue
have any pertinence to what might happen at physio-
logical pressure levels remains to be shown. If, for
some reason, the muscle in the living aorta were more
powerful, or if the imposed stretch were made quite
small, we might anticipate a decreased wall dis-
tensibility. Even so, any interpretation of the effect
of muscle contraction would have to be phrased in
8o8 HANDBOOK OF PHYSIOLOGY -i CIRCULATION II
terms of the shape of the stretch curve, and the total
stretch employed.
Effects of Aging on Arterial Dislensihility
The effects of aging on arterial distensibility are
supposed to be well established — at least the textbooks
so report. The actual evidence leaves much to be
desired. On the one hand is the story of pathologists
that aging is accompanied by a reduction in muscle
mass and in elastic tissue, with a replacement by
collagenous fibers. The reduction of muscle mass needs
documentation by actual cell counts. Chemical
digestion of the aortic walls, to leave only elastin,
left Lansing (74) unwilling to accept the dictum that
the elastic fibers had been reduced in number with
age. He would, of course, accept the possibility of a
chemical change which might influence the wall
extensibility.
Extensibility studies made on isolated vessels taken
from humans of different ages suffer from our un-
certainty about how to compare extensibility among
different specimens. The repeatedly quoted studies
of Hallock & Benson (37), based on a small series,
in which only the average results of a given age group
were presented, showed some decrease in extensibility
with age, with the only truly large change seen in
individuals over 70 years. The comparative data were
expressed in terms of an elastic modulus (AP/(AV/V).
Here, as in all other reports (10, 62, 66, 81, 107, 127),
there was a progressive increase in unloaded diameter
with age, which in itself could increase the value of
this modulus. In a study of a larger series of human
aortas (107) we presented results taken from the
second of two consecutive stretch curves. Variations
within the age groups were large. While the group
averages showed a progressive increase in initial
diameter, it was also true that a man of 68 showed
the same diameter as a girl of 18. All these aortas
were screened so as to include none showing athero-
sclerosis, and any from individuals with a history of
hypertension were placed in a separate category.
The diameter increase was especially noticeable in
these hypertensives. The slopes (AP/AV) given by
the stretch curves also showed intraindividual varia-
tion, but they were very much more constant than
were the initial diameters, and there was no clear
trend for this slope to change with age. A changed
modulus value with aging was, then, predominantly
conditioned by a change in the initial diameter.
Expression of Extensibility' in Terms of Moduli
This raises the question as to just how meaningful
a modulus value is in expressing extensibility data.
Certainly having to present a whole stretch curve for
each specimen studied is cumbersome, and such data
are difficult to handle statistically. But a modulus is
supposed to afford insight into the architecture of the
specimen. Thus when a physicist wishes to describe
the extensibility as a property of a material, he uses
Young's modulus, or a related one, which is simply
the ratio of the applied extending force or stress, as
expressed per unit area of material, to the proportion-
ate change in length from the unloaded value. Most
of his materials are so stiff that the strain is small.
Further, the material promptly returns to the initial
length upon removal of the stress, and in measuring
extensibility he obtains a clue to the force of this
return. He therefore calls his modulus one of elasticity,
despite the fact that he is measuring extensibility and
not elasticity at all. Any time delay in effecting the
strain is usually so brief as to be inconsequential. When
the ratio of stress to proportional strain is constant
(he carefully avoids a load sufficient to cause per-
manent yield or plastic deformation), this ratio can be
calculated by using any convenient load. Because
most materials do not have a constant ratio, and some,
as cast iron, depart quite significantly from a linear
relation, he tries to make his applied stress as small
as feasible.
When we turn to materials such as the polymers,
the stress-to-strain ratio is definitely not linear.
Further, the recorded strain is a function of the time
allowed under load; and the material may not
promptly return to the same unloaded length. Be-
cause of the last, it is definitely not proper to call the
modulus one of elasticity. As a substitute, one could
construct a modulus of extensibility, which would be
the reciprocal of what is usually called the modulus
of elasticity. On the thesis that there has been no
internal change in the material because of the stress,
and that the original internal architecture will be
precisely restored after load release, the physicist
continues to use a modulus as a proper expression of
extensibility even with polymers. To emphasize the
fact that the modulus calculations are based only on
the values seen during extension, rather than during
the elastic recoil, I have used the symbol S rather
than the conventional E in all the equations pre-
sented below. It should be obvious that one must
append to any modulus calculation a careful descrip-
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
8o9
tion of the stress used and the time over which it
acted. This has seldom been done. A derived slope
obtained by the use of a conveniently large stress,
because of the nonlinearity of the stretch curve, often
has no counterpart in this curve — the modulus
represents a mathematical figure only.
Evidence presented above leaves room for doubt
that the internal structure of a vascular tissue is
necessarily the same, just because the unloaded
diameter might be restored. This restoration might
involve muscle contraction, or it might be due to a
passive creep recovery. Even in the latter case, an
architectural rearrangement conditioned by the
stretch might still be present. We have seen that, with
enough load, the contracted ring may start from a
smaller diameter but reach the same stretched value
as an unstimulated one. Shall we, from the calculated
modulus value, deduce that the tissue has been
weakened because of the muscle contraction? We
have seen that with aging the unloaded diameter
tends to increase — an increase which may or may not
be at the expense of the elements which condition the
major portion of wall extensibility. Here, a calculated
modulus value might seem to be evidence for a wall
stiffening, which may or may not have developed.
In those cases where the slope of the pressure-volume
curve remains unchanged, we may seriously doubt
that the increased modulus value is an overly mean-
ingful index to wall stiffness.
The claim is often made that the increase in
diameter is a way of compensating for a wall stiffening
with age. This statement arose from the modulus
formula itself. If the ratio of AP/AV remains constant,
the volume uptake for a unit length of vessel remains
constant, and the change in initial diameter can
hardly be said to be a compensation at all. Actually,
as will be discussed later, the diameter change will
affect the propagation velocity of the pulse wave,
which will affect the length of vessel that is receiving
volume at any given time interval. Thus, indirectly,
some compensation for a wall stiffening might be
effected, but it is questionable that this effect can be
stated in quantitative terms, and any such formulation
certainly would not use the same equation as is used
for a modulus calculation. For the time being, it
appears essential that before we can talk in meaning-
ful fashion about changes in stiffness of the wall, a
change in the actual slope of the pressure-diameter or
pressure-volume curve alone must be shown. It is on
this last point that the evidence on the effect of aging
seems to be weakest.
Since, when one is working with a vessel during
life, the stretch does not start from an unloaded size,
another modulus has often been substituted, in which
the diameter change is related to the real size seen
just before the new increment in stress was applied,
i.e., the diastolic diameter. This modulus is just as
justifiable as that given above, but its value must be
quite different. One can be converted to the other
arithmetically only if the tension-length relation is
linear. Unfortunately, the two moduli have too often
been treated as interchangeable. Finally, since
changes in pressure and volume are usually the
primary data in the living aorta, a modulus based on
the pressure-volume relation has been substituted.
Conversion of this modulus to that using tension and
length is quite complicated. Perhaps the interrelations
could be best expressed in terms of their derivations:
Young's modulus (length) is the applied force per
unit area divided by the proportionate length change.
For a circumference increase, the area over which a
given load is applied will be the length of the ring
(/) times the wall thickness (a). The strain will then
be the relative increase in circumference, i.e., 2irAr/
2irr, or Ar/r. Since the material is being stretched
from an unloaded state, the applied tension will be
AT, and r will be ra.
Thus
AT
Ar
*Tr0
Ar
:d
If the change in radius and in tension are small
enough that they lie on the actual stretch curve, the
equation can be written as
S.OIr*-
dr
(2)
This derivation assumes that there will be no signifi-
cant change in length or in wall thickness accompany-
ing the radial stretch, which, with large diameter
changes at least, is certainly not true, as will be
discussed later.
Now let us suppose that the initial radius is not the
unloaded value, but is taken when the tissue is
already under stretch. The basic equation would not
be altered:
S -^
** Ar
(3)
8 m
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
but the value for Sa would not be the same as that
for S0. It the changes are very small,
and
„ - Ilia
dr
(4)
Next, let us express the stress in terms of pressure.
This is usually clone by taking the pressure (P) as
equal to T/r (21). As pointed out by Frank (29), and
stressed recently by Peterson et al. (91), when the wall
is relatively thick in relation to the internal radius,
the more proper expression would be P = Ta/r,
where a is wall thickness. With an imposed stretch:
P + AP = a (T + AT), '(r + Ar). When P = o
and T = o, this becomes AP{r„ + Ar)/a = AT.
Substituting for AT in equation 1 :
__ AP(r„+ Ar)r0
o a Ar
(5)
If Ar and AP are very small, their product will be
infinitesimal, so that
5 - dPro
0 a dr
And from equations 3 and 4 we obtain:
AP(rd + Ar)rd
a Ar
and
dPr/
" a dr
(6)
(7)
- (8)
Now to express the radius change in terms of
volume:
V= Ivr2, and V + AV * lir(r + Ar)Z
' lir(r2 + 2rAr + Ar2), or
AV = lrr(2rAr + Ar2), or
AV
Ar
Iw(2r + Ar)
Substituting for Ar in the denominator and then for
lief- in the numerator of equation 5:
_ AP(r„ + Ar)r0l*-(2r0 + Ar)
o
AP(2lri
a AV
/ + 3/jrr2Ar v- lvr0Ai
*)
a
AP(2Var„
0 0
AV
+ 3V0Ar
. "■*'',
a
APV0(2rg
AV
+ 3Ar +
*£>
a AV
(9)
.. 2dPVarn
0 adV
(10)
Substituting for Ar in the denominator of equation 7,
and following through as in equations 9 and 10:
AP(rd +
Ar)rdl
w(2rd i
■ Ar)
a
APV„(2rd
AV
+ 3Ar
+ Ar2
* rd
-)
and
a AV
2dPVdrd
0 a dV
(II)
(12)
A commonly used, but incomplete, modulus (see
equation 7) is:
. APr,
5, --
d Ar
And another (see equation 11) is:
b< AV
(13)
(14)
The relationship between these moduli could be
illustrated by taking a hypothetical tube with an
unloaded radius of 10 mm, in which the radius
increased 1 mm for each 1 g per cm2 increase in
tension. The stress-strain relation will then be as
shown in figure iA. The value for the modulus
(equation 1 ) would be 10. The calculated pressure-
radius and pressure-volume curves would not be
linear (fig. 2P and C).
A modulus calculated on the basis of a loaded
initial radius (equation 3) will increase as rrf increases
(table 1). A modulus based on pressure change would
give the constant value of 10 if equation 5 is used, but
if equation 6 is employed, the S0 value decreases as
the strain becomes larger, so that even a 1 per cent
change in strain produces a decreased value. Curi-
ously enough, the value of Sd calculated from equa-
tion 7 shows a constant value, while that from equa-
tion 8 progressively decreases.
Converting the radius changes to volume increases
makes the formulas very cumbersome. Once again,
equation 10 gives a changing modulus for Sot as
does equation 1 2 for Sj. There is no simple way of
converting the moduli values obtained from the
different formulas to each other.
Based on studies with rubber, King (62, 63)
introduced another measure of extensibility, /3, which
is the ratio of the unstretched length L0 to the maxi-
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
8ll
®
3-
E
2-
1-
UJ
tr
l/>
LU
a
RADIUS
m m
1 1
1
©
3n
o
UJ
tr
2-
3
in
UJ
1 -
Q.
VOLUME cc
■ T 1 1
fig. 2. Extensibility and derived distensibil-
ity relations for a hypothetical tube showing a
linear tension-length relationship.
10 II
12 13
TABLE
I
T
g/crn*
r
mm
P
cm X 1 o-a
V
cm3
(3)
So
(5)
So
(6)
Sd
(7)
Sd
(8)
So
(9)
Sd
(12)
Sd'
(.3)
Sd"
(>4)
0. I
IO. I
9-9
3'7
10.0
IO.O
9-9
10.0
10.0
21 .0
20.6
9-9
IO.3
0.2
I0.2
■9
6
326
IO. I
10. 0
9.8
10.0
9-8
10.6
6.9
9.8
3-3
o-3
IO.3
29
i
333
10.2
10. 0
9-7
10.0
9-9
10. 1
9.0
9-7
4.4
0.4
IO.4
3B
5
34o
10.3
10.0
9.6
10.0
9-8
9-7
9.2
9-7
4-5
o-5
IO.5
47
6
346
10.4
10.0
9-5
10.0
9.8
10.0
10.6
9-5
5-2
1 .0
I I .O
90
9
382
10.5
10. 0
9.1
10.0
9-5
9-7
8.8
9-i
4.2
2.0
I2.0
166
7
452
I I .0
10.0
8-3
10.0
9-2
10.0
9.0
8-3
4.1
3-0
13.O
230
8
532
12 .O
10.0
7-7
10.0
9-2
10.0
8.2
7-7
3-6
mum extension possible. Hence /3 = L0/(Lm!xx — L0).
If a proper Lmax can be determined, this ratio has
several advantages, but it also suffers from the same
uncertainty as to what a proper L„ value should be,
particularly if, in constructing the relationship, the
tissue has been stretched so as to approach Lmax.
It has become common to compare a "dynamic"
modulus, as obtained with rapidly repeated small
stretches, to a "static" modulus. For example, Lawton
(77) and Cope (22) reported a small increase in the
dynamic value over the static for the aorta, which
presumably reflects the influence of the rate-de-
pendent factors involved in the visco-elastic behavior.
But there is confusion as to how a static value should
be determined. Sometimes values taken from a
single continuous stretch curve covering the whole
range of physiological pressures are used if the in-
volved stretch has been done slowly. In other cases,
a pressure-length value representing the center of
the dynamic loop is taken as indicating the static
value. Only rarely does this give a value different
from one based on the peak values of the loop, and it
would appear to strain the definition to take this as a
static value at all. A third method is to hold a peak
load constant until, through creep, the length has
approached a final value. All three methods give
different values, which simply indicates again that
more than viscosity is concerned in tissue hysteresis.
This can be illustrated by an experiment shown in
figure 3. An isolated ring of dog thoracic aorta was
first subjected to a continuously increasing stretch.
over 2 min, to a high tension. Tension was converted
to pressure, and half-circumference to volume. The
peak tension thus represented a pressure of 350 mm
Hg. The load was then slowly released, over 2 min,
and, as before, the ring did not return to the same
initial volume setting. A second identical stretch (in
terms of tension) was then made. The relations
obtained during this second stretch and stretch
release are plotted as the solid line in the figure. This
stretch curve is not different from those we have used
in the past to classify the distensibility of aortic rings.
Now the ring was allowed to remain in Locke-
Ringer's solution for 2 hours, during which time the
unloaded volume was very slowly decreased. It was
mounted on the stretching apparatus, care being
taken not to stretch it in the process. A small length
change was then made rather rapidly (0.1 sec), and
the stretch repeated in rapid succession ten times.
Stable stretch and stretch-release curves were es-
tablished by this time. The pressure-volume relations
of this stable loop are given in the figure as loop A.
The same stretch was then performed an 11th time,
but the peak value was held constant for 5 min,
allowing the pressure to fall to its static value, some
2 mm Hg lower. The ring was then returned to a
volume setting part way up the original loop, and a
new series of rapid stretches made, the last loop
being shown as B in figure 3. Again a static pressure
was obtained. The whole process was repeated 13
times.
The initial volume for the loops was first smaller
8l2 HANDBOOK OF PHYSIOLOGY -~ CIRCULATION II
20C-
fig. 3. Pressure-volume relations for a ring
of dog thoracic aorta, in situ length 10 mm,
with stretch done by continuous stretch (curve
S) and by successive, repeated dynamic
stretches. The broken line, curve R, is the
stretch-release curve for the continuous
stretch. The crosses mark the mid points of the
successive hysteresis loops. The solid points
represent the postdecay (static) tension values
reached 5 min after completion of the stretch.
150
100-
50
0-
V0LUME cc
1.3
1.5
2.1
2.3
than that given by the continuous curve, but crossed
it and became greater at high pressure levels. The
continuous curve reflected the large stretch which
had preceded it. If less load had been used for this
stretch, the curve would have differed less at low
pressure settings, but even more at high pressures.
It might seem that a pressure-volume curve obtained
by joining the midpoints of the respective loops might
give a better measure of aortic distensibility. But, in
life, the aorta is never free from stretches, and any
departure from normotensive pressure levels is but
temporary. We would expect, then, that when the
pressure did fall below normal, the aortic volume
would be greater than indicated by this curve con-
structed from the loops.
More important, the volume change (AF) for the
different pulse pressures was almost the same for the
different loops as when taken from the continuous
curve. This is particularly true in the normotensive
pressure range. Hence, the very different methods of
stretching produced some, but not large, changes in
the AP/AF value. Now let us express these distensi-
bility curves in terms of moduli, using equation 1 4.
As shown in figure 4, the dynamic S& for each of the
loops, using the peak value only, was greater than
the static by an average of 10 per cent. The con-
tinuous stretch curve gave modulus values varying
from — 1 5 to +12 per cent of the static, with an
average difference of +2 per cent. Also shown in
figure 4 are the modulus values calculated for only
the very first part of the stretch curve for each of the
loops. The fit with the other moduli is erratic, but the
values are considerably greater than those based on
peak values. These results are given in detail only to
illustrate how difficult it is to classify the behavior of
the aorta on the basis of any single technique of
performing stretches.
Changes in Length and Wall Thickness of Arteries
In all modulus calculations, it is unrealistically
assumed that length and wall thickness remain
constant. Lawton (76) presented evidence that the
volume of the aortic wall remained unchanged during
a stretch. This means that as the circumference
increases, there should be either a shortening in
length or a decrease in thickness. Fenn (26) and
Fawcett calculated that if the wall is isotropic, there
should be no length change, so that only wall thick-
ness would be involved. A direct recording; of the
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
8l3
180-
150-
/ '/ '
////
////
/?/<.
100-
1 '/ s
£
1/ *
E
/ \
V
a
(»/
ui
fc ""
a>
T ''
a.
' /
*P
/
t
/
50-
tL
\
it I
\
/
9r\-
U/
Dd"
S
10
20
30
40
50
fig. 4. Distensibility modulus (eq. 14) calculated from the
data of fig. 3. Solid line, from continuous stretch curve. Dotted
line with dots, from peak values for hysteresis loops. Dotted
line with triangles, from postdecay (static) values for loops.
Broken line, from initial slopes of stretch curves of hysteresis
loops.
change in thickness during a stretch has not been
made. There are some sparse references to the relation
of the unloaded thickness (a) to the outside diameter
(£)). Thus King (62) found an a/D ratio of .09 for
human aortas. McDonald (84), in a survey of many
arteries from the dog, found a constant ratio of
.08. In studying the effect of age on the human
aorta, King (64) found a progressive decrease in
thickness, so that the product of thickness and radius
was nearly constant. On the other hand, young aortas
show more longitudinal retraction upon excision
than do those from older people (107), which might
account for part, at least, of the difference in wall
thickness. The question of how much the wall thins
during a stretch needs documentation, since this
factor will affect the derived modulus value.
In isolated vessels subjected to a volume increase,
Fenn found a lengthening, from which he concluded
that the wall was anisotropic (26). McDonald (84)
is quite correct in emphasizing that the longitudinal
extensibility observed in isolated vessels may not be
a measure of changes that might take place in the
in situ vessel under longitudinal restraint. Hence if
the intact vessel is in the steep portion of the longi-
tudinal extensibility curve, its length changes with
each pulse would not be large. The presence of the
aortic sheath might also reduce length change in the
in situ aorta. It is of interest here that a pulmonary
artery freed from surrounding connective tissue showed
a longitudinal thrust with a volume injection, while
one still bound showed but minor change (32). Yet
the vessel wall should not become anistropic simply
because it was released from its longitudinal restraint.
Length changes in living animals have not been
completely measured. Lawton (78), working with
serial photographs of a dog abdominal aorta, found a
small shortening in early systole and a lengthening
in diastole. This made the length and circumference
changes almost 180 degrees out of phase. Similar
length changes for the abdominal aorta were found
by Patel and co-workers (87). In contrast, they found
length and diameter changes to be in phase in the
thoracic aorta.
The small length changes recorded seem in sharp
contrast to the sometimes rather striking longitudinal
thrusts seen in the aortic arch. And at times a freed
carotid artery, or more rarely a femoral artery,
visually seems to be showing a length change. These
thrusts might reflect factors other than a distention
upon invasion by the pulse wave, however. The heart
is anchored in the chest by the large vessels. It has
long been known that the base of the heart is lowered
in contraction, which must serve to lengthen the
aorta and pulmonary artery (47). Rushmer (1 13) has
described this movement as starting in the period of
isometric contraction. The motion of the arch, and of
the brachiocephalic arteries which serve as anchor
points for the arch, would reflect not only the geom-
etry of the vessels but the firmness of attachment of
the descending arch to the body wall. Further,
respiration displaces the aorta, which acts as though
it is bound rather firmly to the diaphragm. These
longitudinal thrusts would bear no necessary time
relation to the arrival of the pulse wave, and a de-
ciphering of the origin of length changes in a vessel
may not be easy.
Considerable confusion was raised by a report (113)
that when diameter and pressure were simultaneously
recorded in the thoracic aorta, an unorthodox
hysteresis loop was obtained in which, during stretch,
the diameter change was proportionately greater
than the pressure change. These loops were taken
from an oscilloscope. Inspection of the individual
diameter and pressure records indicates that the
whole of the diameter curve simply preceded the
pressure curve (126). If the two were superimposed,
814
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
ignoring the time lag, the expected hysteresis loop,
although small in magnitude, was seen. Maintenance
of strict identity between the site of measurement of
the two variables is difficult at best. If there has been
a longitudinal displacement of the aorta, and hence
of the circumference recorder, through influences
other than the arrival of a volume pulse, a seeming
"phase lag" between the two recorders could be
produced.
Curiously enough, with isolated strips of arteries,
the time lag is reported in the opposite direction for
pressure leads. From this lag is calculated the viscous
component of a dynamic modulus (48, 84).
Summarizing, we can say that despite many
studies on the extensibility of the aorta and large
vessels, it is still uncertain whether the presented
stretch curves may be reflecting to such a great
degree the techniques used that they are not readily
illustrative of the characteristics of the wall. Work of
the future will certainly be concentrated on measure-
ments made on living vessels, that will include not
only diameter change but changes in length, and
perhaps in wall thickness. There are not sufficient
data to allow a well-based speculation as to how the
in vivo measurements might fit with those obtained
from isolated specimens. The question of how muscle
contraction might affect tissue extensibility, for the
aorta and for the muscular arteries, is yet to be defini-
tively answered. Whether an expression of extensi-
bility in terms of a modulus is the most satisfactory
tool remains questionable.
ACTION OF THE AORTA AS A CONDUIT
Pulsatile F/01
Rigid and Distensible Tubes
Since the aortic flow is never steady, we can turn
immediately to a consideration of pulsatile accelera-
tions and decelerations rather than deal further with
the classic hydrodynamic equations. As a start, let us
visualize a piston pump connected to a rigid pipe of
uniform bore, with the piston being driven by a large
force. Let us leave the distal end of the pipe open, so
that a flow through can be established. Also, let us
imagine a valve system so constructed that the barrel
of the pump can be filled, during piston withdrawal,
from an external reservoir. To start a pump cycle, the
first tiny forward movement of the piston will produce
a compression of the adjacent fluid. This initial
compression will represent a high pressure — one that
cannot be recorded, since anv manometer used would
of necessity have a membrane, the resistance of which
toward displacement would be less than that of the
fluid or the rigid pipe walls. Once the involved force
is sufficient to overcome the factional resistance to
fluid displacement, or to overcome the inertia of the
fluid column, flow can start. While the time interval
between may be short, we can say that there will
always be a temporal separation between the creation
of the pressure force and flow through the tube. This
is commonly spoken of as a phase lag, with pressure
leading. The definition of the physical forces and the
quantitation of such lags, for both rigid and dis-
tensible systems, have occupied the attention of many
physically minded workers of late (33, 48, 54, 60, 70,
^5> '39)- I do not consider myself qualified to judge
the relative contributions of these papers.
Now let the piston complete its stroke, and reverse.
The pressure in the pump will show a sharp fall, the
amount depending upon the speed of inflow from
the side reservoir. Since a pressure gradient has been
previously constructed in the pipe to produce dis-
placement toward the open end, or, if preferred,
since fluid has already been accelerated toward this
end, flow will continue for a brief interval despite
the pressure fall in the pump. Once again, then, we
have a phase lag, and the fluid column can be said
to have an inertial force. If the pump strokes are
repeated at a rapid frequency, the flow per cycle will
be related to how well matched the duration of each
phase of the pump cycle is to the phase lag, as set by
the factional and inertial characteristics of the tube.
This principle of matching can be illustrated by
another model. Suppose a U-tube mercury manom-
eter is made to oscillate by a periodic blowing of
air into a side arm on one side of the U-tube. The
first buildup in air pressure will displace the mercury,
and after this the mercury column will oscillate back
and forth, the period being conditioned by the size
of the tube and the other components of fluid re-
sistance to flow. If the frequency of the air puffs
matches that of the mercury column, the excursions
will be reinforced. Conversely, if the generating
frequency is out of phase with the mercury oscilla-
tions, movement of the mercury, or "flow," will be
minimal.
Equations which relate flow to pressure usually
express the phase lag in terms of a component of the
frequency of the repeated strokes. This is simplest if
the pressure buildup by the pump has a sinusoidal
form. If the stroke is of a different form, the pressure
curve is broken down into terms of a fundamental
sine wave and a number of superimposed harmonics.
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
8l5
Matching with the resonant characteristics of the
fluid-filled tube could be either with the fundamental
wave or one of the prominent harmonics.
Use of the same pump coupled to a distensible tube
of uniform bore and wall extensibility will present a
somewhat different pattern. Because the wall can
yield, a large part of the energy imparted by the
piston can cause an increased tension in the tube wall.
It is no longer necessary to construct a pressure
sufficient to overcome the resistance of the whole
fluid column, for as soon as the fluid resistance to
displacement in the first small segment of tube is
overcome, piston movement can displace volume
into this segment. The pressure energy of this fluid
will go into a stretching of the walls of the segment.
If the wall extensibility is great (as with condom
rubber), the first segment could accomodate all the
fluid displaced from the pump, and there would be
no appreciable pressure rise in the tube and no flow-
through its length. It might be pointed out that the
molecular movements in this wall stretching are
directed toward the side of the tube, so that the
displacement pattern is more like that of turbulent
flow in a rigid pipe than that of streamline flow.
If the tube is less distensible, only part of the fluid
compression transmitted to the first segment will
go to produce wall extension, for the fluid must
retain enough pressure to prevent the elastic recoil of
the stretched walls. This erects a pressure differential
between the first and next segment of tube, a differ-
ential related to wall elasticity (which need not be
identical with wall extensibility) and the fluid re-
sistance of the second segment. When the differential
becomes larger than the resistance, fluid displacement
will follow. In a tube of uniform distensibility, then,
except for the frictional energy dissipation, the same
volume will be accepted, per unit length of time, by
each successive tube segment as the first part or front
of the wave moves through the tube. Hence, if the
piston displacement is linear against time, the pressure
in the pump and the upper part of the tube will
simply remain constant, since all the pump outflow
will be taken to establish the wave front as it moves
from segment to segment through the tube. This
pattern of a constant pressure can be demonstrated
in a rubber-tube model. As the pressure front moves,
flow through the stretched segments behind it will
be streamlined. While the frictional cost of such
movement will be small, the further the wave pro-
gresses the greater will be the cumulative energy-
dissipation. This analysis also means that once fluid
displacement into the first tube segment occurs, the
first part of a pressure wave has been created. This
wave will continue to move through the tube whether
piston movement continues or not. Further piston
movement does act to support the later parts of the
wave, or to broaden it in time.
A sinusoidal piston movement leads to a rising and
falling pressure in the upper end of the tube. This
produces a pressure wave, positive or negative, which
is propagated back and forth through the tube. No
matter in which direction waves may be traveling
through the tube, the pressure in any one tube segment
at a given time simply reflects the balance between
the amount of fluid entering it and that leaving it.
The extensible tube should show a phase lag, too,
but since only tiny segments of tube, acting more or
less independently, presumably are involved, the
resistance to fluid movement out of the pump should
be very small. Hence a phase lag should also be
small. It is well to note here that the loss in pressure
in the aorta, due to frictional dissipation, is within
the error of recording.
Our model of a tube with uniform distensibility has
no counterpart in the arterial bed. Figure 5 shows
four drawings taken from an earlier analysis of this
problem (04), based primarily on the extensibility
values given by isolated rings. That on the upper
left depicts a part of the arterial bed of a dog, drawn
to scale in respect to anatomical length and cross-
sectional area at a pressure of 100 mm Hg. But in
describing fluid displacements, we are more concerned
with the propagation time of the pulse wave through
a region than we are with actual length. The natural
pulse wave moves slowly in the upper aorta, more
rapidly in the lower aorta, and faster yet in the
large arteries (10, 24). Suppose we redraw the figure
so that the length now represents the distance tra-
versed by the wave in a unit length of time (lower
left). Next, since the pressure rise in the parts of the
bed will be set by the segmental distensibility (ne-
glecting wall hysteresis), let us redraw the figure
(upper right) letting the assigned width represent the
distensibility, expressed as AF/AP, rather than the
cross-sectional area. Lastly, if frictional resistance is to
be discounted as of small amount, we can neglect the
effect of tube diameter per se, and group together
into a single composite tube all vessels which might
lie at the same time distance from the heart (lower
right). Such a theoretical tube has a funnel shape,
distensibility being great in the top (ascending aorta)
and tapering down gradually to the stiff vessels that
are farthest from the heart, those of the hind legs.
A linear piston displacement into such a tube will
8l6 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
fig. 5. A reconstruction of the arterial reservoir of the dog. [From Remington (94).] For
description, see text.
no longer produce a constant pressure in the upper
end. The lower part of the tube will require less fluid
to construct the same pressure rise. Hence the pressure
in the pump will continue to rise as the wave front
moves through the tube, the pressure rise being a
function of the distensibility of the lower part of the
funnel. This increased pressure at the upper end will
also be propagated, so that we are now creating a
whole pressure wave.
Quantitation of Fluid Displacement and ]Vall
Distensibility Relationships
The propagation of the wave front is really the
same as the initial displacement of fluid from segment
to segment in the tube. The rate of this displacement
should be a function of wall distensibility, and it
should be possible to formulate the relationship in
quantitative terms. This is another case where the
textbooks have such a formula so well established
that it has assumed the nature of a physiological law.
The supporting evidence is far from adequate.
Discounting all friction and other resistance factors,
Korteweg presented a theoretical formula, and Moens
(see 46), working independently, arrived at almost
the same formula on the basis of experiments done
with various distensible tubes. The latter used arti-
ficial waves which were relatively slow in their rate of
pressure rise, and he used not the first part of the
pressure wave for his measurements of wave velocity,
but the time interval between successive peaks as the
whole wave was propagated back and forth through
his closed end system. His formula differs from that
of Korteweg in that he had a constant of 0.9. We
have shown (46) that the velocity of the peak of such
artificial wave is less than that of the start, by a
factor not greatly different from Moens' constant.
Using the Korteweg formula, then, the velocity of the
wave foot (v) is related to an "elastic modulus" (E)
of the tube, and the density of the contained fluid
(p), thus:
. g_Ea_
2r/>
where g is the gravitational constant and a is the wall
thickness. Neglecting hysteresis, E would be the same
as Sd of equation 4 derived above. Hence by sub-
stitution:
,2 -
gadTrd gadT
2t ■ />dr ' 2/>dr
(15)
If p is regarded as a constant, and given the value of
1 .055 for blood, and a is taken as unity, then
2 . 9.3dT' _ 4.65 dT'
' 2dr dr
where T' is the tension per unit length of tube.
Similarly, from equation 8:
s _ godPrf 4.65 d Prd
and from equation 1 2 :
(16)
dr
(17)
.2 =
ga2dPVdrd _ 9.3dPVd
2rd/>adV dV
(18)
This equation 18 is the same as that derived by
Bramwell & Hill (15) and has since borne their
name. They further corrected the constant by multi-
plying it by the weight of mercury, so that pressure
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
would be in terms of mm Hg. Hence their formula
reads :
# 12.7 VdP
dV
Bramwell and Hill did not use the first slope to
determine dP and dV, but appreciable increments in
pressure and volume instead. Commonly the pressure
increment is taken to be the pulse pressure, which
strains the use of even AP. Because our methodology
is not adequate to give dP and dV values, we have
no right to use the above equations. If, instead, the
formula is derived from equation 1 1 :
e gaAPvj2rd + 3Ar + %.)
2rd/,aAV
. 12.7 AP Vd(rd+ 1.5 Ar + 0.57T )
AV
(19)
Actually, Arz/rd is so small it can be practically-
neglected, so that
.i -
Ar
12.7 APVd(l + l.5~)
AV
(20)
Validation of these formulas has centered on the
Bramwell and Hill equation. The earlier results,
which have been reviewed (46), offer no clear evi-
dence that the velocity of artificially generated or
natural pressure waves shows either a quantitative or
qualitative agreement with values predicted by the
formula, when it is applied to stretch data taken from
isolated vessels. The solid line of figure 6 shows an
average relation of pulse wave velocity to diastolic
pressure for some 200 pulses of a living dog, taken
from the aortic arch to the diaphragm. The broken
line shows the velocity calculated, using equation
19, from the continuous second stretch curve given
in figure 3. Agreement is certainly not good. If the
mean slope for each loop given in figure 3 is used
instead, then, as shown by the dotted lines, agreement
with the actual becomes qualitatively better, with
equation 20 giving a better fit than ig. But the mean
slope can have little significance as far as the propaga-
tion velocities of the parts of a wave are concerned.
The speed of the wave front should be dictated by
the slope taken at the beginning of the stretch phase
of the loop. Calculation from these initial slopes,
using equation 19 (which is here valid), gives the
125-
100'
50
Pulse wave velocity, M / Sec
1 r
5 6 7 8
fig. 6. Relation of pulse wave velocity to diastolic pressure.
Solid line, actual values from a living dog. Broken line, calcu-
lated (eq. 19) from continuous stretch curves of fig. 3 Dotted
line, closed circles, calculated (eq. 19) from mean slopes of
loops shown in fig. 3. Dotted line, open circles, calculated (eq.
20) from mean slopes of loops of fig. 3. Crosses, calculated (eq.
19) from initial slopes of stretch phase of loops of fig. 3.
unconnected crosses of figure 6. These velocities are
greater than the actual by 10 to 20 per cent.
In our earlier study (103), in which we compared
a curve such as the broken line of figure 6 (based,
however, upon a careful compilation of the stretch
curves of all rings, taken in sequence, from the aorta
being studied) with the actual, we believed that the
underestimation would be correctable by using the
slopes resulting from a hysteresis steepening of the
first part of the stretch curve. The loops obtained in
this earlier study were not numerous, and we did not
attempt any quantitative verification of this belief.
Further, and unfortunately, in this study we used
both a rubber tube and the excised aorta, leaving the
implication (although it very definitely was never
stated) that the two behaved similarly. With rubber,
the initial slope of the stretch phase of a loop proved
clearly dependent on the rate of stretch. In keeping,
the propagation velocity of artificial pulses through a
rubber tube was found to be directly related to the
rate of initial pressure rise. But with the aorta, using
either artificial or natural pulses, there was no similar
relation between the rate of pressure rise and the
wave velocity. My more recent evidence (96) that a
dependency of the aortic stretch curve upon the rate
of stretch is minor is quite compatible with this
finding.
A calculated velocity for the wave start, in excess
of the actual velocity, may be explained by four
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
things. /) The loops given by an isolated ring may be
much wider than those found in the living aorta.
Evidence was cited above that this might be true.
2) The velocity may be dictated not by the wall
extensibility, but rather by the force of elastic recoil.
Calculation from a stretch-release curve would indeed
give smaller values. The question is what amount of
stretch should be used to produce pertinent stretch-
release curves, j) The distensibility slope of the aorta
during life may be entirely different from that indi-
cated by the isolated rings. The data of Peterson
(91), for example, seem to have it greatly different.
4) Factors which act to slow the pulse wave should
be introduced into the formula. To the extent that
the aorta acts as a rigid tube, fluid resistance toward
flow can act in this manner, as will a phase lag be-
tween the pressure pulse and the corresponding fluid
displacement. Many workers now accept the presence
of an appreciable lag. As will be seen, acceptance of
such a lag cannot be readily reconciled with the
failure to find a correlation between wave velocity and
the rate of pressure rise, as mentioned above.
Phase Lag and the Harmonics of the Arterial System
Since phase lag is formulated in terms of the
frequency of harmonic components, the first step is
to perform a Fourier analysis of the pressure wave.
For this, a sinusoidal fundamental wave must be
selected. Since such a fundamental is not readily
apparent in the contour of the natural pulse wave, it
is selected on the basis of a time duration (123, 124).
Usually the length of the pulse cycle is used. As
emphasized by McDonald in the introduction to his
book (84), such a mathematical analysis can start
from one of two premises. First, we can assume that
each pulse is an isolated or transient phenomenon,
with the aortic volume being almost static when a
new cardiac ejection and sudden flow acceleration
are begun. Second, we can say that the heart rate is
virtually stable, so that the ventricle is repeatedly
pulsing the arteries at a set frequency. The latter
premise makes the cycle length a true measure of the
wave fundamental, makes the harmonics relatively-
reproducible from beat to beat, and makes all the
mathematical compilations very much easier. A
change in heart rate will vary the contained har-
monics and alter the phase lag between pressure and
fluid displacement. It will, then, alter the wave
velocity.
But the fact that an analysis on the basis of a
uniform heart rate is easier to make does not mean
that the premise is correct. Much evidence can be
quoted for the stand that each pulse is indeed an
independent event. Strict regularity of the pulse rate
is infrequent, usually being found only in rather
prolonged experiments in animals under deep anes-
thesia. In the unanesthetized dog or human, variation
from cycle to cycle is clear. In this variation the
diastolic period is affected predominantly or ex-
clusively. The pulse contour during systole, and its
duration, is affected but little. Further, when the
heart rate changes outside the limits of such beat-to-
beat variations, systolic durations and contour are
altered far less than is the cycle length (100). If the
fundamental is reset each time this cycle length
changes, a different harmonic picture will be re-
quired to construct the same systolic pressure contour.
If the fundamental is taken as the average cycle
length for a number of pulses, then we must proceed
cautiously in interpreting the influence of the
harmonic pattern on the contour of any single pulse
of the group.
The whole approach seems more hazardous, too,
when it is recalled that the length of systole almost
never equals half the cycle length. This makes the
heart quite unlike most pumps. Perhaps it would be
more logical to use twice the length of the systolic
period as the fundamental wave. This might be done
for a central pulse, but certainly not for a peripheral
one, where the incisura has been lost through damp-
ing.
Believing in the principle of a stable heart rate,
McDonald (84) would have the velocity of the wave
foot increasing with the heart rate. He offers no
experimental support for the claim. We have looked
often for evidence of such a dependency on heart
rate and, with the single exception presented below,
have not found it. However, McDonald has calculated
that in a vessel the size of the aorta neither the viscous
resistance factors nor the pulse frequency would
affect the velocity to significant degree. In a smaller
vessel, such as the femoral artery, he calculates that
the viscosity would slow the velocity by about 10
per cent, and an increased heart rate might restore
it to the value expected from the Bramwell and Hill
formula. Much larger changes than these would be
needed to correct the formula if the data given by
the crosses in figure 6 are correct.
We did offer evidence (46) that a slower foot
velocity was seen in the early part of the response
of an animal to an injection of acetylcholine, when
the heart rate was slow, than was seen later when,
at the same diastolic pressure levels, the heart rate
increased. A similar effect at higher pressures has
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
8l9
not been observed. We then attributed the velocity
change not to the heart rate but to a changing
hysteresis loop behavior of the wall. A long diastolic
interval could allow more time for recovery after a
stretch, and the diastolic size, and hence the distensi-
bility modulus, would be thereby reduced. It is less
clear now (96) that the difference in this size could
be sufficient to account for the difference in wave
velocity. Yet it remains possible that the correlation
with heart rate was still only coincidental, and that
when the pressure was abruptly lowered the slope of
the stretch curve could be shallower for the first few
beats than it would be after the pressure had re-
mained low for some time.
Most especially, if we regard the pulse wave as an
independent phenomenon, the velocity of the wave
start would be affected least by a change in harmonic
frequency. The upper parts of the pressure pulse
could have their propagation speed affected to
greater degree by these frequencies or by the speed
of the pressure upstroke. A different velocity for the
parts of the wave was fully accepted by Bramwell
and Hill simply on the basis of their formula. They
went further (14) and held that the difference in
velocity could be such that the anacrotic rise of the
pulse would progressively steepen in transit until
finally the wave force would become unstable, and a
"breaker" (like that seen when an ocean wave enters
shallow water) would form. Evidence of such breaker
phenomena was seen when pulses were generated in
a bicycle tire. As will be described later, evidence is
not clear that the natural pulse does so progressively
steepen during propagation, and there is no evidence
at all for sudden pressure vibrations that would mark
a breaker. However, the calculated differences in
velocity between the start and the peak of a natural
wave are not large enough to create a breaker within
the length of the aorta.
If there is a velocity differential between the parts
of the wave (and it would appear to be quite small
if present),2 it could reflect the progressive increase
2 There is an obvious discrepancy between the statement that
there is no clear evidence for a difference in propagation
velocity of the parts of a natural wave and our published
results (103) which showed clear differences in transit time for
the parts of an artificial wave. There are few inflections on the
natural pulse form which can be measured with the necessary
precision to establish a difference in propagation velocity. The
start of the wave and the incisural notch can be so timed, and
these two parts of the pulse contour appear to move with the
same velocity. Since we have no clear idea as to which tension-
length slope should be used to predict the velocity of the in-
cisura, or to which volume on the stretch-release curve this
slope should be referred, this identity of velocity with that of
in the stiffness modulus as the reference volume in-
creases, without requiring a dependency upon the
harmonic frequencies. Landowne (72, 73) did show
that when small impact waves were formed at a
point on the human brachial artery, the speed of
their propagation was faster if they fell during the
systolic portion of the pressure pulse than during the
diastolic portion. The propagation velocity of these
small waves was much greater than that of the
natural wave. Van Citters (125) believes that the
velocity is of the order to be expected if they were
being transmitted by longitudinal strain through the
wall itself, rather than by fluid accelerations within
the artery.
Landowne (71) has also shown that, with a rubber
tube or umbilical artery, either small impact waves or
rapidly repeated sinusoidal waves moved at a velocity
which bore a direct relation to the frequency. Our
experiments showing a dependency of wave velocity
upon the rate of stretch of rubber fit with this (46).
The umbilical artery has a uniquely large time-
dependent factor in its visco-elastic behavior (141),
so that it would not be at all unreasonable that the
velocity could also show a clear rate dependency.
These results should not be regarded as transferable
to the aorta, and perhaps not even to arteries such
as the femoral or carotid. We are left, then, with the
conclusion that the actual pulse wave velocity remains
to be explained in a quantitative way. A mathe-
matical analysis of the determinants of pulse wave
velocity is presented in the chapter by Hardung (49).
We still have the fundamental question as to
whether there would be an appreciable time lag
between the pressure pulse and the fluid displace-
ment, or the movement of the pulse wave from
segment to segment through the tube. The idea of a
large lag was presented first in the papers of Peterson
(89, 90). He perfected a mechanism which could
produce a very rapid input of fluid into the ascending
aorta, and thereby generate pressure curves, of rather
strange form, which were propagated. The shape of
these curves was explained on the basis of a sum-
mation of three forces. First, a very small amount of
fluid would be driven into the aorta more rapidly
than the walls could stretch, so that, just as in a
rigid pipe, there would be a sudden rise in pressure.
the wave foot may be coincidental, and not be evidence for or
against a dependency of wave velocity upon frequency. It
should also be stressed that while the wave parts of the artificial
wave moving through an excised aorta showed different
transit times, these times were not conditioned by the rate of
pressure rise or fall.
820
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
This initial peak he labeled the acceleration transient,
and its force was equated to the small fluid mass
involved times the acceleration. Next, he added a
force which increased with the velocity, representing
the resistance offered to fluid displacement through
the tube. This reached significant proportions only as
the volume displacement did, which placed its con-
tribution later in time than the acceleration transient.
Finally, after a time lag, he added a straight line
increase in pressure to represent the force necessary
to prevent an elastic recoil of the walls as they were
stretched.
Certainly the presence of these three forces in
constructing an aortic pulse cannot be denied. The
problem is to ascertain how large a role each of them
plays, and how much of a time delay between them
exists. Peterson's acceleration transient lasts for many
milliseconds. While wall hysteresis could, to use a
term employed long ago (46), make the vessel seg-
ment show a "reluctance to stretch," no studies on
isolated rings indicate that the reluctance could last
nearly this long. Just as crucial is his claim that the
same pressure excess which marks the acceleration
transient would persist through the whole systolic
period, so that all the pressure pulse would have a
higher value than would be predicted from a pressure-
volume diagram taken from stretch data. The fact
that he dealt with the whole arterial bed as a lumped
system has made it difficult to follow his argument.
Rather than a discrete time lag of this sort, other
workers are supporting the presence of a sinusoidal
phase lag (34, 36, 83, 120, 128, 129). Using an
electrical analogue, the aorta is said to have an
inductive, capacitative, and a resistive impedance to
flow. Of these, the inductive and resistive factors
would be in phase, but the capacitative would lag up
to 90 degrees. In hydraulic terms, the first of these is
called inertance, which represents the mass of blood
displaced into the tube segment times its acceleration.
Opposing this inertance is the compliance (capaci-
tance) reflecting the volume taken to accomodate the
wall stretch, and the resistance, which represents all
fluid and wall factors that cause dissipation of energy
as heat. In an actual vessel subjected to pulsatile flow,
the interrelation of the three would be dependent
upon the rate of change in the driving pressure,
usually expressed in terms of the frequencies of the
harmonics. The better matched these frequencies are
to the inherent frequency of the vessel compliance,
which is a function of the visco-elastic properties of
the wall, the greater is the flow into and through a
segment. The most proper match would be at the
"resonant" frequency of the vessel.
If an isolated vessel is suddenly stretched and then
allowed to vibrate, it will show a definite period of
oscillation (77). This frequency will be different at
various pressure levels and with different parts of
the arterial system. It also can be changed by any
factor which influences the visco-elastic properties.
The matching frequency between a segment and the
driving pressure is therefore subject to considerable
variation.
But it remains uncertain why such factors should
play a significant role in a distensible tube composed
of tiny segments. Certainly any final analysis of the
pressure-flow relation must reconcile the recent
data, based on the dictum that a phase lag must be
present, with the older descriptive work, which
includes evidence of a general absence of effect of
any physiological factor other than the diastolic
pressure level on pulse wave velocity, the details of
pulse contour change which takes place during
propagation (to be treated later), and the actual
time relation between flow and pressure curves. The
last of these has been least well covered. Records of
the flow pattern seen at different parts of the aorta
have been presented, and such records have, super-
ficially at least, much in common. But discrepancies
exist between them in regard to quantitation, timing
of peaks, and amount of end-systolic backflow.
(27, 53, 56, 84, 120). Unfortunately, a simultaneously
recorded pressure pulse is so rarely given that one
can never be sure whether the cardiodynamic con-
ditions were enough alike in the different experiments
that one should expect similar flow curves.
In the ascending aorta, the flow rises sharply to a
peak reached in early systole and then falls more
gradually to reach a zero value, or below, at the
time of the aortic valve closure (fig. 7). The flow then
remains negligible throughout the diastolic period
(119), or may show a small sinusoidal increase in
diastole (131). In records taken from other parts of
the aorta, the amount of retrograde flow seen just
after the end of systole progressively increases as one
moves out the vessel, and the diastolic wave also
increases in magnitude (119).
It may be well to digress into a semantic problem
that continually proves worrisome to students. The
point is frequently made or implied that there is a
clear distinction between the fluid displacement that
accompanies the movement of a pulse wave and the
"stream flow'" through the vessel. In the aorta there
really is no stream flow as such, and fluid displace-
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
82 1
120-t
100-
80J
200
100
100-
100
50-
o-1
1 01 SEC 1
fig. 7. Carotid pressure pulse (.4) and ascending aorta flow
(C). [From E. Wetterer (131).] B = change in volume uptake
curves for arterial bed regions. Broken line of C = the summed
uptake values taken from B. D = change in volume uptake as
calculated from a hysteresis loop, as taken from fig. 3. The
summed uptake values are given in C as the dotted line.
ment simply accompanies the movement of the pulse.
This displacement is toward the periphery in systole,
but some may be toward the heart for a period in
diastole. It should not be difficult to understand that
the molecules involved in such displacements in the
lower aorta, for example, are not the same ones as
left the ventricle during the corresponding ejection.
The stroke volume is of the order of a fourth of
the aortic volume. In contrast, a stream flow is
established in the stiffer resistance vessels, which
approach more nearly the characteristics of a rigid
tube. Another way of saying this is that flow through
the aorta starts and stops, rather than being con-
tinuous.
At present, the aortic flow curves available offer no
clear indication of the amount oi time lag between
pressure and fluid displacement. Spencer (119)
makes the statement that in the upper aorta pressure
and flow start together, but he offers no supporting
figure. If this is true, any phase lag will be based
simplv on a relatively slower increase of flow than
of pressure. On the other hand, the left ventricle
usually develops a pressure above the aortic level
before ejection apparently begins, which excess is
then gradually lost (98). Thus there appears to be a
true time lag of about 5 msec, similar to that en-
visioned by Peterson. But a study of pressure pulses
taken from adjacent parts of the aorta offers no clear
evidence that a similar excess and time lag exist
there. Thus, after the initial delay between ventricle
and ascending aorta, the pressure pulse seems to be
propagated at a steady rate through the aorta (98,
99)-
At present, a major obstacle in the interpretation
of the presented flow curves is a lack of a reference
standard against which they can be compared.
Quantitatively, the curve from the ascending aorta
should integrate to the stroke volume less the coronary
flow. But our knowledge of the time contour of
cardiac ejection rests only on cardiometer curves,
which come from open-chest animals and bear
distortions that make one question the value of a too
detailed study of their time-flow dimensions. Flows
taken from other aortic regions can be related to the
stroke volume only if one assumes a distribution of
volume between the parts of the arterial bed.
Construction of a Hypothetical Ejection Curve
It might be of interest to construct a hypothetical
ejection curve, derived from the contour of the
central pressure pulse (104). This requires that all
animals be assigned the same wave transmission time
and the same arterial distensibility, the latter taken
from an average of stretch curves of isolated rings.
Certainly no claim can be made for the accuracy of
such curves. All we do know is that the total stroke
volume derived in this way usually agrees reasonably
well with that given by a direct measurement (94).
In this construction the arterial funnel, as shown in
figure 5, is divided into segments, the lengths of which
are approximately 10 msec of transmission time. The
total volume uptake of the arterial region is then
divided by the number of segments included, with the
various segmental uptake curves starting in sequence
every 10 msec. This derivation assumes that: a) the
aortic pressure pulse, as taken from the ascending
aorta, has no distortion because of a contained
acceleration transient; b) the wall stretch shows no
hysteresis lag; c) the control pulse is propagated as an
entity, without damping and without augmentation;
and d) there is no time lag between pressure change
and the corresponding fluid displacement.
Suppose we take first the pressure pulse presented
by Wetterer (131) corresponding to his ascending
aorta flow pulse shown in figure 7. This pulse is ob-
viously from an open-chest animal, the length of
systole probably indicates that the animal was cold,
822
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and the contour is not one we would regard as repre-
sentative of that to be obtained from an animal in
good circulatory condition. The ascending aorta and
arch are taken as the first tube segment. From our
tabulated pressure-volume tables (104) a volume up-
take curve can be constructed for this segment in 10-
msec intervals, starting at the time the central pressure
pulse begins its upstroke. Because, in the Wetterer
experiment, the ascending aorta had a flowmeter
attached, we have arbitrarily reduced the volume
uptake of this segment by one half. Rather than
plotting the total uptake, only the net gain or loss of
volume for each time interval is given as the solid
line curve of figure yB. While the pressure in this
segment is still rising, the volume will be increasing.
When the pressure falls in late systole, there will be
a net loss of volume.
The thoracic aorta and head and foreleg arteries
are grouped together in the next part of the funnel,
and it takes the wave some 30 msec to move through
the whole. Hence, for uptake calculation the total
region is divided into three parts, displaced 10 msec
behind each other. The summed net volume change
for all three is given by the broken line, labeled H,
in figure yB. Next, the pressure wave invades the
abdominal aorta and visceral arteries, which takes
another 30 msec. The summed volume change of
the three units involved is given as the dotted line
(V). Finally, the summed changes of the three leg
vessel units are given in curve L. Flow through the
ascending aorta must not only accomodate the volume
acceptance of more distal arteries, but must supply
systolic drainage through the arterioles as well. The
calculation of this latter will not be gone into here
(see ref. 44), but it is indicated in figure ~]B by curve
D. Ascending aorta flow now should equal the alge-
braic sum of all these curves at any given time in-
stant. The value obtained is per square meter of
body surface. It is assumed that the dog Wetterer
used was medium size, i.e., had about 0.6 m2 surface
area. The use of this assumed value means that we
should not expect quantitative agreement between
the derived curve and the actual one, but only
qualitative agreement. The total flow calculated in
the above manner is given as the broken line in figure
yC. The actual curve presented by Wetterer is given
by the solid line. The calculated values therefore
indicate a flow increasing more steeply in early
systole, and decreasing sooner and more sharply after
the peak. This discrepancy in flow might have four
causes: /) the flowmeter might be slurring the actual
curve; 2) there might be a distortion of the aortic-
flow curve because of vessel constriction produced by
the meter; 3) there might be a true time lag, of ap-
preciable proportions, between the pressure curve and
the flow curve; and j) wall hysteresis might change
the form of this calculated flow curve. The influence
of the last of these can be directly tested. If the volume
uptake values are calculated from a hysteresis loop of
the same pattern as those given in figure 3, the
rate of volume gain in early systole would be de-
creased, and there would be little volume change
while the pressure first starts its fall in late systole.
The summed flow curve, as given in figure 7C by
the dotted line, differs but little in form from that
given by the broken line. Hence it would be difficult
to reconcile the calculated curve with the actual with
even a large amount of vessel hysteresis.
A similar calculation was done for the only three
pulses presented by Spencer and co-workers (119,
120) which have accompanying pressure pulses.
The flow recorder here was on the upper thoracic
aorta, so that the uptake of the arch, head, and
foreleg vessels were omitted when the volume changes
were summed to give the flow curve. The same type
of discrepancy between the calculated and the re-
corded flow pattern is again seen (fig. 8). It might
be noted that these pressure pulses are unusual in
that they have a very steep initial rise in pressure,
with a relatively flat systolic crest. This might be
evidence of an effective aortic constriction by the
meter. If so, the flow profile in the lower aorta would
not be expected to match the form of this pressure
pulse.
On the premise that flow should lag behind the
instantaneous pressure, Fry and co-workers (34)
derived an equation in which the pressure difference
between two points in the aorta was equated to the
sum of an inertial term and a factional resistance.
Solution of their equation was achieved by use of
an electronic computer. After checking their equa-
tion by use of a sinusoidal pump with a tube (rigid?),
they proceeded to construct a flow velocity curve
for the upper aorta, using catheter tips 6 cm apart
for the pressure recordings. Xot knowing the exact
positions of the catheters, one is uncertain as to
just what vessel segments should be included in an
attempt to construct a similar flow curve on the
basis of vessel distensibility. We used the whole
aorta, as though we were calculating a cardiac
ejection curve. Xot knowing dog size or diastolic
aortic dimensions, the calculated peak flow value was
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
823
120-
fig. 8. Pressure and flow values given
by Spencer. Pulse a from Spencer &
Denison ( 1 20) ; pulses b and c from Spen-
cer el al. (119). Dotted line, summed
arterial bed uptake values, as described
in text.
arbitrarily made to coincide with the presented
values. The agreement in contour between the curve
presented by Fry (solid line, fig. 9) and that calculated
(broken line) bears a good deal of resemblance to
those seen with the actual flow curves. It should be
mentioned that the differential pressure recording
presented by Fry indicates a peculiarly long delay
period between the two recording catheters, with a
slow transmission velocity through that particular
aorta (about 3 M/sec).
These constructions provide presumptive evidence,
then, that there is a delay between the pressure and
fluid displacement curves. There will certainly con-
tinue to be interest in the factors which contribute to
this lag. Whether harmonic analysis of the pressure
pulse curves may be the most profitable tool for this
assessment remains to be decided. It is important
that we do not let sophisticated mathematics allow
us to lose sight of the basic processes by which a
distensible tube seems to be filled. Volume is displaced
from segment to segment, establishing and maintain-
ing a moving pressure wave. Since the distal parts of
the aorta are stiffer than the proximal, we would not
expect that the pattern of fluid displacement out of
the arch would be qualitatively similar to that of the
pressure curve, for the amount of fluid leaving the
upper aorta would be decreasing when the pressure
was rising. Toward the end of systole, when the wave
front has invaded the whole network of distensible
vessels, flow would fall sharply to a low level which
represents mainly the drainage loss from the bed.
At this time, the aorta would be behaving more like
a rigid tube.
Judging from cardiometer curves, ventricular ejec-
tion starts slowly, then rapidly attains a maximal and
constant rate which lasts through the first part of
systole. The outflow then slows, reaching a small
value some time before the valves actually close.
Because the first outflow, although slow, is confined
to the ascending aorta, the pressure rise produced
must be relatively large. As the wave moves through
the aorta, an even faster ejection rate will produce less
rise in ascending aorta pressure. This tendency is in
part offset by the stiffer walls of the more distal
vessels. But in any aortic segment a pressure rise
simply means that more blood is entering than is
leasing for the more distal regions.
A pressure difference curve based upon a subtrac-
tion of pulses taken at two different sites can be mis-
leading. Even assuming no change in contour, until
the wave reaches the distal recorder the difference
will be but a replica of the proximal pulse. When the
pressure upstroke in the peripheral recording begins,
this difference curve must show a sharp inflection
and a fall, depending upon the duration of the first
steep pressure rise and the separation of the recorders.
As long as pressure is still rising in the proximal seg-
ment, the difference should remain slightly positive.
824
HANDBOOK OF PHYSIOLOGY -^ CIRCULATION II
160-
140"
I20J
<
.1 SEC
fig. g. Pressure and flow values given by Fry et at. (34).
Broken line of B, calculated flow values as described in text.
When it starts to fall, the difference should swing to a
negative value. This does not mean that flow down
the aorta will then cease. The volume displacement
is part of a wave movement, and the pressure differ-
ential simply reflects the time lag between the wave's
arrival at two points in the tube. Such a continuation
of fluid displacement toward the periphery, despite a
negative pressure differential, could properly be
called an inertial property of the fluid. What all
workers are seeking is a complete description of
what we mean by a wave, and what factors con-
tribute toward its progression through the tube.
Before leaving the descriptive model, it should be
pointed out that no length dimensions were placed on
the tube segments that were acting independently.
With wall fibers distributed longitudinally as well as
circularly, there cannot be such an independence of
action of tube segments. A unit of a distensible tube
must have a finite length, which, however, has not
been defined. This tying of segments to each other
must give a distensible tube some of the characteristics
of a rigid tube. However, it remains rather incon-
ceivable that a whole aorta could act as a single
bound entity, and could be given a single lumped
resistance value.
To summarize this section on the behavior of the
aorta as a conduit, the initiation of flow through a
rigid system certainly requires the acceleration of a
whole column of fluid, an overcoming of fluid re-
sistance for the whole length of the tube, and a phase
lag between pressure built up at the generating source
(pump) and the flow out the end of the tube. In
such a rigid system, resistance factors can certainly
be treated as a unit. With a distensible tube, how-
ever, depending upon the stiffness of the wall, only
a small segment of fluid need be accelerated in any-
given unit of time, and the fluid resistance and the
phase lag can be relatively small. A model has been
presented in which a pressure wave is propagated
from a minute segment of such a tube to the next
adjacent segment. Of course, the tube is linked lon-
gitudinally by extensible fibers, and the length of what
is being called a tube segment cannot be defined.
But it is not clear that the current trend of treating
pressure-flow relations in the aorta as though re-
sistance was lumped and as though there were an
appreciable phase lag between pressure and flow is
helping our understanding. The propagation velocity
of the wave must be linked in some way to tube
dimensions and to wall distensibility, but no com-
pletely satisfactory formula for quantitating this
relation appears yet to have been presented.
THE AORTA AS A BLOOD RESERVOIR
Changes in Central Pulse Contour During Propagation
In the description of how fluid displacement
through the arterial bed might be calculated from
the distensibility values of the various vessels, and
the course of pressure change in the ascending aorta,
the assumption was made that the central pulse
would be propagated intact. This assumption is
clearly false. The pulse contour is modified during
transmission, this modification resulting perhaps
from damping, or from a poor matching between the
frequencies of the volume input curve and those set
by the clistensibilities and flow resistances of each
arterial segment, or from an augmentation of
"matched" frequencies, or even from superposition
of a wave reflected from the periphery upon the
incident wave. It should not be implied that, because
pulse contours change, the whole method of calcula-
tion of the form of the cardiac ejection curve is
invalid. If the contour changes do not appreciably
alter the total displacement of fluid out of the as-
cending aorta, the total quantitation need not be
greatly in error.
Contour differences between a central and a
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
825
fig. 10. Reconstruction of aortic-
pressure pulses, showing comparison
between control in aortic arch (dotted
lines) and records taken simultaneously
with their controls at indicated distances
down the aorta from the arch (solid
lines). Below, five of the above ten, semi-
diagrammatically superimposed on a
somewhat larger scale, with a repre-
sentative control. [From Hamilton &
Dow (42).]
peripheral pulse were recognized even in the days
when pressure recordings were made using low
frequency manometers. When Frank (28) developed
his high fidelity manometer, he established this
difference in precise terms. This was verified and
amplified by the work of Wiggers and his associates
(1, 135, 136) and Hamilton and his group (40, 42,
140) in this country, as well as by continued work in
Europe (17, 29, 57, 58, 75, 115, 117, 133). In con-
trast to the broad systolic crest of the central pulse,
the femoral artery pulse, for example, shows a high,
narrow systolic profile. Sudden slope changes, such
as the shoulder of the central pulse and the incisural
notch, are no longer present in the distal vessel,
having been lost through damping (fig. 10). Such
damping is most obvious when the aortic pressure is
low, and least obvious when the pressure is at hyper-
tensive levels. This is probably related to the fact
that the visco-elastic properties of the wall are more
prominent at low pressure levels.
The changes in contour are similar to those that
would be obtained if a central pressure pulse were
recorded by a slow-frequency manometer system,
which would allow an overswing of pressure in
systole, and an exaggerated fall to a low level in early
diastole. The German workers, after Frank, have
therefore thought of the portion of the arterial bed
which stores blood in systole, i.e., the arterial reservoir
or Windkessel, as having a lumped distensibility
value, like a manometer (12, 17, 58, 132). It should
be remembered, of course, that the distensibility of
the arterial bed is not that of a single membrane,
and it does not follow that the arterial reservoir could
vibrate as a single unit as a manometer system does.
Despite much descriptive work on the contour
changes which attend propagation of the pulse, our
basic knowledge of the underlying principles remains
incomplete. In their classic paper on this subject,
Hamilton & Dow (42) presented for the first time a
mapping of the changes in pulse form in the dog as
recorded serially from various points in the aorta
(fig. 10). This mapping reveals that as the wave
moves toward the periphery the steep initial anacrotic
rise remains unchanged in slope, but persists for a
longer time. Hence the deflection marking its end,
or the shoulder, comes at progressively higher pressure
levels. The systolic peak becomes gradually narrower,
so that the time from the start of the pulse to the
peak is reduced. Hence, in spite of the transmission
delay of the start of the wave, the peak is reached at
826
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
just about the same time in all pulses taken from the
lower part of the aorta. It is most difficult to time the
peak of a pulse exactly, but this approximate identity
was taken as evidence that this peak was "standing."
This suggests that the aorta was achieving a "reso-
nance" with the first transit of the pulse wave.
Resonance and Standing (1 m < s
To explain the resonance concept, let us visualize
a somewhat elongated rubber balloon, filled with
fluid, and connected at one end to a syringe. A sudden
imput of fluid would start the bag oscillating, due
to a sloshing of fluid from one end to the other with
a reversal of movement, or a reflection, taking place
at each blind end. The period of such oscillations
must reflect the time required for the fluid slosh to
traverse the balloon, and therefore is related to the
conduction velocity of the fluid wave and the length
of the bag. The first of these is a function of the
distensibility of the part ot the bag through which
the wave is moving, as described earlier. If the wave
length of the slosh is just twice that of the transmission
time through the bag (or a simple multiple of it),
we could say that the bag was resonating, for a) the
pressure changes at the ends would be just 180 degrees
out of phase; b) the peak pressure, produced by a
summing of the incident wave with the reflected
wave, would be reached at the same time through
half the length of the tube. This means that there
would be a point of minimal pressure oscillation, or
a "node," at the mid point of the tube, and all peaks
and pressure troughs seen on either side of this node
would be "standing" through half the tube; c) the
time interval between two successive pressure peaks,
as recorded from any point in the tube, should be a
constant, and be an index to the length of tube and
the wave velocity.
The records of Hamilton and Dow suggest that
all three criteria can be met in the arterial system.
There are pressure oscillations at the two ends of
the dog aorta which seem 180 degrees out of phase,
and which maintain approximately (but not exactly)
the same period until they are damped out. The
amount of pressure change with such oscillations is
much smaller in the arch of the aorta than in the
abdominal aorta, just as the distensibility of the two
regions is different. There are times when all three
criteria are not met in the dog, but more recent map-
pings indicate that what seems to be a true resonance
very often is achieved (4, 101).
The carotid artery (or whole head system?) shows
no similar oscillations, or even any great change in
pulse pressure with outward propagation of the
pulse (39). In man, the arm system shows augmenta-
tion of the pulse pressure but not resonance as defined
above (108). Records made in this laboratory indicate
that the foreleg system of the dog produces pulse con-
tour changes similar to those seen in the human arm.
Further, the aorta-femoral system does not show such
resonance or even clear oscillations in very small
animals (140). There is question whether resonance
occurs in an animal as large as man (108). The
German workers do believe the human aorta to show
resonance, but, as will be discussed later, their con-
clusion is not based on a standing peak for the periph-
eral pulses.
Attempts to design a model that could illustrate
the prompt achievement of resonance, as occurs in
the dog, have not been successful. Certainly, ex-
periments in which independent pulses were gen-
erated in a closed and moderately long rubber tube
(46) provided little insight into how it would be
possible to make a previously quiet bag resonate
with the first propagation of a pressure pulse through
it. Granted that if the time of volume injection was
made identical to the transmission time of the wave
peak through the tube, reciprocal oscillations at the
ends would be seen from the time of completion of
the injection. But if the injection period was ap-
preciably longer or shorter than this, there was no
such immediate resonance. Instead the formed wave
peak could be followed back and forth through the
tube, as it was propagated at a steady rate, and
reflected at each blind end. Because the wave length
changed during these propagations, the foot moving
more rapidly than the peak, which in turn moved
faster than the "tail," after several trips through the
tube the wave could finally achieve a length equal
to that which would make the tube resonate. Whether
a given wave ever attained such resonance would
depend upon the number of trips required to 1 hange
its wave length, and the number that were possible
because of incomplete reflection and continued
damping. The change in wave length attending
propagation was attributed to the hysteresis behavior
of the wall.
Similar changes in the length of an artificial pulse
were seen in a tied off but in situ dog aorta (46).
This change is directly opposite to that predicted by
the Bramwell and Hill formula, which would have
the peak moving faster than the foot. Of course,
artificial waves never attained the same rate ot
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
827
pressure rise seen with a natural pulse, and they
were truly independent phenomena.
McDonald (84) believes that because the natural
pulse is but one of a continuous train of waves with
virtually identical wave lengths, each ejection could
serve to reinforce the component frequency which
happens to match the transmission time through the
resonating part of the reservoir. This premise would
permit development of resonance with the first transit
of each wave. But, by extension, this premise would
also require that the pulse pressure augmentation
be a function of the heart rate. Again, all we can say
is that neither the pressure augmentation nor the
period of the diastolic oscillations has been shown to
have any relation to heart rate per se when the
diastolic pressure remains the same. The reciprocal
oscillations between aortic arch and abdominal aorta
pulses in the dog appear to be the rule and not the
exception. They appear with the first beat after a
prolonged cardiac arrest, as with vagal stimulation;
they are not clearly accentuated at any given heart
rate; it is most difficult to so alter the cardiovascular
status through nerve stimulations or injected drugs as
to make them disappear.
The prompt achievement of resonance by the
aorta would seem to require, then, that the whole
vessel could act as a unit, and "mold" any ejection
wave into a pattern consistent with its own resonant
properties. Alexander (7) has used the analogy of
an orchestral chime, which, when struck, vibrates
at a frequency set by its own geometry, unaffected
by the characteristics of the impacting force. Use of
this analogy is not easily reconciled with the theorem
that wave propagation is from tiny tube segment to
adjacent segment. Instead the pressure rise in the
upper end of the aorta would have to be aisle, by
some mechanism, to throw the whole aorta into vibra-
tions. Yet there is no evidence that this pressure rise
"signals ahead" of the propagated pulse wave. There
is no pressure change in the lower aorta at the time
the central pulse is first being ejected.
One question which must be decided is whether it
is the propagated pressure wave itself which sets the
aorta into resonance. No alternative suggestion has
yet been advanced, unless one can read into a paper
describing the genesis of the ballistocardiographic
waves the notion that whole body thrusts might in-
duce this resonance pattern within the vessels (43).
The propagated wave in an aorta has much in com-
mon with an artificial wave being propagated
through a stoppered rubber tube, although the latter
does not readilv create immediate resonance. The
aorta should be even less conducive to the attainment
of resonance than the rubber tube. There certainly
is no single reflection point, for exit vessels are
distributed along the whole length of the system. One
would expect, then, innumerable returning waves
bearing no necessary time relation to each other.
Further, the exit vessels are not blind end tubes, bur
continue on to become the resistance vessels of the
arterial tree. This has led some to the conclusion
that the aortic reservoir should be considered as more
comparable to an open-end tube, the resonant wave
of which would then be twice as long as that of a
closed-end tube (60, 134). On the other hand, Hamil-
ton (38) has maintained that the sudden increase in
the resistance to flow in these vessels will serve to
produce the positive reflection. While such reflections
could take place wherever the flow pattern is changed,
as at a vessel bifurcation, or even in the curvature of
the aortic arch, these reflections within the tube
would be small when compared to those arising from
the small resistance vessels. He has documented this
belief by experiments done on a rubber tube model
fitted with many small rigid tubes of high-flow re-
sistance, but with a greater aggregate cross-sectional
area, placed in series with the rubber tube (41).
Alexander (1, 4, 8) recorded pulses from the arch,
the abdominal aorta, and the femoral artery of dogs
under a variety of physiological conditions. Usually
the two peripheral pulses showed simultaneous
peaks, although at times they did not. If the central
pulse was subtracted from the peripheral one, to
give the contour of the reflected wave, two different
waves in the subtraction curve could be seen. The
first of these was taken to represent the propagated
peak of the incident wave, •"distorted" by damping
and the other factors which may give rise to contour
change during propagation. The second, a swell of
more sinusoidal form, was the first of the resonant
oscillations. When the two waves coincided, the
femoral pulse pressure was at its greatest. In some
central pulses, a late systolic trough could be seen
that appeared simultaneously with the distal resonant
swell. When the length of the ventricular ejection
period was slowed through induced hypothermia
(7), this trough came far enough ahead of the incisura
to be clearly recognizable.
The resonant swell obtained by such a subtraction
did not have the same wave length as the central
pressure pulse. In fact, there is no real evidence that
the whole of the incident pulse is reflected. When
subtraction curves of the same tvpe were obtained
for human subclavian pulses (108), a reflected wave
828
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
seemed to be present, but it was small in magnitude
and short in duration. It was as though only the
first sudden acceleration of flow produced a reflected
wave. Whether this should be regarded as a common
finding, true for the whole arterial tree and also for
a closed-end rubber tube, is not clear.
Hamilton & Dow (42) showed that when the aorta
was occluded, the frequency of the pressure oscilla-
tions seen on a pulse was more rapid than that usually
found in a normal dog. By moving the point of
occlusion distally, they concluded that the "end" of
the resonating system must lie outside the aorta.
Taking the time from the start of the central pulse
to the peak of a lower abdominal aortic pulse as
equal to half the total resonant wave length, they
calculated that the end should be near the knee, and
the node should be in the lower thoracic aorta.
Wezler & Boger (134) placed the end, which they
took as the point of negative reflection, in the femoral
artery near the inguinal ligament in the human.
Schmitt (115) located the node in the abdominal
aorta in man, and the end in the distal part of the
tibial artery. This was based simplv on transmission
times of the wave, for the time delay from the heart
to the node should also equal the time from the node
to the end, and equal a fourth of the total interval
between successive pressure peaks of a peripheral
pulse. Similar studies by Wetterer and co-workers
(58, 132) placed the end of the system beyond the
ankle in the foot. An occlusion by cuff inflation of
the legs shortened the interval between the systolic
and the postincisural pressure peaks, which they
reasoned could be true only if the cuffs were still
proximal to the end of the system (59). There are
two aspects of studies such as these that give room
for concern. First, the pressure peaks of pulses taken
from the leg arteries are not coincident with those
of the femoral artery (fig. 1 1), nor are they timed to
reciprocal oscillations of the central pulse. Of the
Subclavian
Popliteal
Femoral
Tibial
fig. 11. Pulse contours from peripheral arteries. [Redrawn
from Kapal et al. (59).]
three criteria listed above for resonance, they satisfy
only one, i.e., the time interval between peaks re-
mains about the same as that seen more proximally.
It remains possible that the truly resonant pulse form
could be propagated with but little distortion of time
relations through the leg arteries. This seems to be
what happens in the arm vessels (108). Second, oc-
clusion of the aorta seems to make it behave as a
blind-end rubber tube would, and the waves which
are propagated back and forth in this occluded length
of vessel do not have the same characteristics as the
natural wave. It may be that leg occlusion could
introduce a reflection of the wave, and change the
timing between peaks, but that use of the occlusion
technique to identify the end of the resonant system
is not theoretically sound. It should be repeated that
our studies on aortic pulses in man (108) gave no
evidence of a standing peak.
Alexander (3) believed that the node for the clog
was in the upper abdominal aorta where the large
visceral arteries exit. This is a point of sudden increase
in total cross-sectional area, where the flow rate,
relative to the vessel size, accounts for a large fraction
of the total cardiac output. When he occluded the
visceral arteries, the frequency of the resonant waves
seen in femoral pulses was increased, and the pre-
incisural trough of the pulse of the ascending aorta
became less conspicuous. Unfortunately, the pressure
was also raised by this maneuver, so that the changes
evoked are not indisputable evidence for his hy-
pothesis. Converselv, an intra-arterial injection of
histamine into the visceral arteries decreased the
frequency of oscillations (and also lowered the
systemic pressure). Ryan and co-workers (114)
repeated the occlusion experiments, and concluded
that the pressure rise might have been sufficient to
explain the changed frequency of the resonant waves
in the femoral pulse, but observed that the occlusion
did eliminate the preincisural trough of the central
pulse. In general, however, occlusion of exit arteries
has very little influence on the timing of the pressure
oscillations (57 ).
Alexander (4) postulated that the aorta-femoral
system was essentiallv two open-end systems in
series, the region of visceral artery exit marking
an open end common to both. Arrival of the incident
wave at this area would be followed by a reflection
of a negative wave back toward the arch to produce
the preincisural trough. The incident wave would
also be propagated into the lower aorta as a positive
wave. Hence the two systems would be effectively
resonating with each other.
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
829
117/80
22-R
0.1 SEC
fig. 12. A mapping of the change in pulse form in the dog
aorta. Pulses from ascending aorta (/t), descending arch (4),
upper thoracic aorta (7), mid thoracic aorta (10), lower
thoracic aorta (14), abdominal aorta (18) and iliac artery (22).
[From Remington (100).]
The German workers have also been concerned
about the effect the large visceral arteries might have
on the resonant wave. They (58), like Hamilton,
would locate the only significant reflection point in
the small arteries. The influence of any single aortic
branch as an independent reflection unit (particu-
larly that of a vessel so far proximal to the "end" of
the system as a visceral artery) causes them no great
concern. Assuming the resonant frequency to be
already established (they have presented no analysis
as to how this might be achieved), they conclude that
reflections in this branch would serve to augment the
pressure excursion without altering the fundamental
wave length of the incident wave. An example of
this is seen when two manometers of different fre-
quency response record in parallel a rapid pressure
change; the records from both will be the same and
reflect the response characteristics of the slower
manometer. This would not explain Alexander's
preincisural trough seen on the central pulse.
In a number of mappings of the dog aorta (99)
I found that while in some pulses the late systolic
trough could be seen, it was not present in pulses
taken from the descending arch of thoracic aorta
(fig. 12). Hence this trough apparently is not prop-
agated back from the upper abdominal aorta, but
rather appears de novo in the ascending aorta pulse.
Although there were some time discrepancies be-
tween the systolic peaks of the peripheral aortic
waves, there was a general tendency for a standing
wave to occur. But this standing wave seemed to
develop as a sinusoidal swell taking off from the
broad crest of the propagated wave, and appeared
first in the aortic arch or at least high in the thoracic
aorta. Its size progressively increased as the wave
moved toward the periphery. If this swell represented
the first of a resonant wave, we would have to con-
clude that the node for this first peak was within the
aortic arch itself. This swell developed at about the
same time that the foot of the incident wave reached
the femoral artery. Later reciprocal oscillations be-
tween ascending aorta and lower abdominal aorta
could be seen, with minimal pressure change in the
upper thoracic aorta. The node of these oscillations
would thus appear to be more distal than that for
the systolic peak. All that we can conclude is that
the genesis of aortic resonance remains obscure.
Other Factors Which May Alter the Central Pulse Contour
Fascinating as this whole problem of resonance
may be, it certainly is not the sole factor which may
produce contour change and pulse pressure change
when the pulse of the ascending aorta is propagated
to the lower aorta or to the brachial artery. Possible
factors which may bear on these changes are:
a) A loss of sharp inflections and an attenuation of
the pulse pressure might result from damping. In a
distensible tube such damping is due in part to fluid
friction, but probably much more to a conversion of
energy from kinetic to potential form because of the
extension of the walls, with a delayed recoverability
of this energy because of the visco-elastic properties
of the wall. Clear illustrations of such a reduction in
pulse pressure and lengthening of the systolic wave
contour during propagation can be seen in dogs with
a deteriorated circulation, or at least a weakened
heart, after the use of a strong vasodilator agent (94)
and when the rate of flow from the upper aorta to
the lower is severely reduced, as by a partial occlu-
sion (23).
b) A peaking of the pulse contour could follow a
83o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
speeding of the upper portions of the pulse as the
modulus of extensibility is increased. This would
probably cause an augmentation of the pulse pressure.
This effect seemingly would be operative only if wall
hysteresis were of minimal importance. However, I
am not convinced that there is any difference in
transit time of the wave foot and of the incisura, for
example.
c) The same sort of peaking could indicate an
attenuation of mismatched harmonic frequencies.
If the electrical analogy is apt, there would be no
augmentation of the pulse pressure in this case, how-
ever, but simply less attenuation of the matched
frequencies. In a hydraulic system, of course, it
might be that attenuation of one part of the pulse
might yield fluid and energy for another frequency,
which conceivably could produce pulse pressure
augmentation. Such a redistribution of energy has
not been shown to be true.
d) There could be a reflection of the whole or a
part of the incident wave, whether the vascular
bed did or did not achieve resonance as it has been
described. The maximum possible increase in pulse
pressure by reflection would be to twice the original
value, which would be realized only if the whole
pulse pressure showed complete reflection, and the
pulse recording was very near the reflection "end"
of the system. Usually the pulse pressure in a femoral
artery is less than twice that of the central pulse.
However, Alexander has shown femoral pulse pres-
sures in the dog which are greater than twice the
value.
e) As an extension of d), or perhaps as a result of
another property of the bed entirely, when the aorta
does show resonance, augmentation of the pulse
pressure is appreciably greater than when a standing
wave is not seen.
/) In some cases where the cycle length is short,
the diastolic pressure swell (presumably the reflected
wave) may begin late in the diastolic period. If a
new systolic upstroke coincides with the upswing of
this swell, very high pulse pressure values can be
obtained. This mechanism of augmentation by "su-
perposition" was illustrated by pulses recorded from
the system of man ( 1 08 ) .
Aside from changes in pulse pressure and the form
of the systolic peak, there are other aspects of pulse
contour transformation which have no clear explana-
tion. The pulse formed in the ascending aorta shows,
after a variable but short period in which the pressure
rise from the diastolic level is slow, a rather abrupt
assumption of a steep and constant slope of pressure
rise. This anacrotic rise is maintained unchanged for
at least 30 msec. It is then usually lost rather abruptlv,
often with a temporary interruption of pressure
rise. This halt is called the shoulder of the pulse.
The steeper the preceding slope, the more con-
spicuous is this shoulder. The rate of anacrotic pres-
sure rise is clearly related to the amount of sympathetic
stimulation of the left ventricle, which serves to speed
the whole contractile process. Thus, with such stimu-
lation, maximal outflow is reached earlier in systole,
the shoulder tends to be at a relatively high pressure
level, and the systolic peak of the pulse occurs earlier.
A shortening of the length of the ejection period
can be used as the basis of an assay method for
sympathomimetic stimulation (100). With extreme
cardiac stimulation, particularly when the stroke
volume is reduced because of inadequate venous
return, the shoulder may be so abrupt as to throw
the whole aorta into vibrations. Under such circum-
stances, the height of the shoulder may be greater
than that of any other part of the pulse, which
makes the shoulder height represent the systolic
pressure. In such cases, the pulse pressure may have
higher values than would be anticipated from the
stroke volume (94), in the peripheral vessels as well
as in the ascending aorta.
The slope of pressure rise preceding the shoulder
was found by Hamilton & Dow (42) to be propagated
unchanged through the aorta. Alexander (4) showed
some loss of steepness in the abdominal aorta, while
I (99) found it to remain constant in the thoracic
aorta and then to steepen in the abdominal aorta.
The slope change is never marked, however, so that
all three studies are compatible with the general
conclusion that this first part of the pressure wave
seems to move as an unchanged unit. All three also
agree that the steep upstroke continues for a longer
time interval the further from the heart the recording
is made. This might lead one to the conclusion that
this early part of the wave cannot be thought of as
being propagated by repetitive accelerations of tiny
segment volumes. It was pointed out earlier that the
length of what is called a "small segment" of the
aorta is undefined. The segment may have an ap-
preciable length, and the volume contained, which
is accelerated as a unit, have an appreciable mass.
Thus it may be that the propagation of the first
part of the pulse wave would involve fluid accelera-
tions which would have many of the physical prop-
erties of a volume surge with inertia. If so, one
would expect the "surge" to produce a progressiveh
greater pressure rise in the lower regions of the aortic
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
83I
funnel, where the distensibility is reduced. The
speed at which the wave front would move would
be dependent in part upon the force of the drive.
For under these conditions, the ascending aorta
would be "driving" the fluid through the various
exit branches, one of which would be the thoracic
aorta. These statements are similar to those made
previously in the discussion of the relation between
the pressure curve and the fluid displacement. It
remains for future work to reconcile evidence which
seems to favor the presence of a fluid surge with that
which supports the proposed model, having wave
propagation based on fluid displacement from one
tiny vessel segment to the next.
A study of a great number of pulse forms leaves
the impression that the volume uptake of the aorta
in the period when the pressure shows this initial
fast rise is not so large as would be expected from
volume-pressure relations taken from a static stretch
curve. This impression has not been proven. A rapid
pressure rise at a time when the volume input is
small was a pillar of the acceleration transient story
of Peterson (90). One would like to explain an excess
pressure height, if present, on the basis of wall hystere-
sis. If, in studies with rapid stretches of isolated
vessels, there had ever been a considerable overfling
of pressure at the end of a stretch, I would feel happier
about this possible answer. If the impression is correct
that pressure rise exceeds the expected volume gain,
then it could also be true that in the interval of the
shoulder of the pulse, the volume gain would con-
tinue, and thus "catch up" with the pressure.
When the pulse enters either the arm system or the
aorta-leg system, the height of the shoulder is in-
creased. In the human arm system, this elevation of
the shoulder takes place largely in the subclavian
arteries. The brachial pulse then shows two systolic
waves, one representing the shoulder, and the other
the later systolic part of the entering wave (12, 69,
108) (fig. 13). Very often the first is higher than the
second, and hence sets the pulse pressure. This is
particularly true when the anacrotic rise formed in
the ascending aorta is steep and the shoulder is
high. Late in a Valsalva maneuver, for example, the
aortic pulse shows a steep anacrotic rise and high
shoulder, but the rest of the pulse tends to collapse
toward a low incisura. This contour is in keeping with
the much reduced stroke volume. But in the brachial
pulse the shoulder may remain at almost the normal
height, which means that a pulse pressure measured
from this height would have no relation to the stroke
volume (108).
II 1 1
1 1 1
1 1
-
ISO
2 0(-> •■'"»v'
ATAP
2pv
(22)
AV
ALAP
pv*
German workers have continued to use the Frank
formula, or modifications of it. These formulas have
received only restricted support in this country (124).
While they may, perhaps, predict in reasonable
degree the volume input into a rubber tube where
the distensibility is uniform through the tube length,
their use with the complicated arterial bed requires
very large assumptions. First, there is the inference
given above that the AP/AV relation for any single
vessel is indicative of the relation for the whole
reservoir system. Second, A does not represent the
area of any single vessel, but rather that of a hy-
pothetical tube which happens to have the same
dimensions as the mean of the whole reservoir net-
work. Attempts have been made to take values for
A from autopsy data, using the upper aorta, which
certainly would not have the same dimensions as
this mean. Further, autopsy data give a diameter at
near zero pressure and not that at a physiological
pressure. Third, since AV is the volume stored in the
Windkessel during systole, it is not directly measur-
able. If a calculation is made for the drainage loss
during systole (and various formulas have been pro-
posed for calculating this loss), then the stored volume
plus the calculated drainage loss would equal the
stroke volume, which can be measured directly only
under restricted conditions, but which is usually
taken from a cardiac output determination. Fourth,
L, the length of the reservoir network, cannot be
directlv measured. As described earlier, it has instead
been calculated from the length of the resonant
wave, as indicated by the time interval between
successive pressure peaks of a peripheral pulse. This,
of course, assumes that the reflecting end of the
system is also the end of the Windkessel. The estima-
tion of wave velocity and of the time interval between
pressure peaks, by the techniques employed, leaves
room for doubt as to the validity of any strict quan-
titation.
If the stroke volume could be directly measured,
it might be that the various unknowns could be
combined into a single constant. Its value, however,
would apply only at the diastolic pressure for which
it was derived, only if neither A nor L was subject
834
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
to physiological change, and only if a constant so
derived for one individual could be applied to
another. All these assumptions have seemed so
precarious that the German formulas have not
received favor in this country.
Yet the hope remains that some means can be
devised by which the stroke volume can be predicted
from the values of the pressure pulse. This would
allow a quantitation of beat-to-beat changes, and
also of the acute change in ejection volume that occur
as the cardiovascular status is rapidly changed. We
have no direct method applicable to closed-chest
animals that can measure these stroke volume changes.
Until we do, an indirect approach can serve a limited
but useful purpose.
Another attempt at making this sort of indirect
calculation was made by Bazett and co-workers (10).
They divided the arterial reservoir into four parts:
/) the aortic arch and its large branches; 2) the
whole of the descending aorta through the iliacs;
3) the subclavian-brachial systems; and 4) the
femoral-leg system. They recognized that the pulse
pressure would be different in these regions, and
therefore concentrated instead on the pressure change
taking place during diastole, when the previously
stored blood was being discharged through the re-
sistance vessels. They assumed that, by the time of
the incisura, the whole arterial reservoir would be
draining as a single unit, and that the pressure
change could therefore be that from the level of the
incisura of a central pulse to the end-diastolic value.
Unfortunately, their central pressure pulses were
rather inadequately recorded. Next, using calcula-
tions based on the size of the larger vessels of each
arterial region as taken from autopsy data, and
using assumptions and empirical adjustment of the
derived diastolic volume values, they arrived at
figures for the total diastolic volume for each region.
The change in volume from this level was then
equated as a function of the pulse wave velocity
through the region, or
,v. v V, v4 ,
V
V3
where AC is the total stored volume, the I"s are the
calculated diastolic volumes of the four parts of the
reservoir, and the v\ the respective wave velocities
(which could be measured only rather crudely).
With more modern technology, the basic data could
be much more accurately recorded. The formula
would still be rather cumbersome, and the necessarv
measurements many. Two of the major weaknesses
still present are that V cannot be directly obtained,
any more than the A of Frank's equation can be,
and that the v values of necessity must be taken from
a single large artery in each of the regions. This
assumes that this artery can fairly represent the whole
system, and also that this velocity gives a true indica-
tion of vessel distensibilitv.
We presented another approach to the problem,
worked out on the dog rather than on the human.
For reasons which have been covered previously,
we regarded the wave velocity as a most dubious
measure of vessel distensibilitv. Instead, we sub-
stituted volume-pressure relations taken from data
obtained by stretching isolated rings, and by inject-
ing saline into occluded arteries of dead animals.
We necessarily assumed that the values obtained
would be practically the same for different animals.
To make some correction for differences in body size,
all values were expressed per square meter of body
surface area. We also assumed that the transmission
time through the various arterial beds would be the
same for all animals, at the same diastolic pressure
level. After making the studies described above, in
which we calculated the presumed cardiac ejection
curve on the basis of a summation of volume uptake
values of the arterial regions taken serially as they
were invaded by the pulse wave, as described above,
we settled on the premise that such a summed total
uptake, with a calculated systolic drainage added,
should equal the stroke volume at the time of valve
closure. Hence we could work with a single central
pressure pulse, laying back the transmission time to
each of the four major divisions of the arterial bed,
from the incisura. The pulse pressure to be quan-
titated for each bed would then simply be the pres-
sure shown at this interval before valve closure. In
other words, the point at which this time interval
intercepts the pressure pulse curve indicates the
pressure developed in the bed in question at the end
of systole. This assumes that there was no change in
pulse contour during propagation. More rightly, it
assumes that any contour change present during
propagation would be constructed by a redistribution
of the volume of blood ejected into the ascending
aorta to create the given central pressure pulse. At-
tempted modifications based on actual pulse contours
taken at various points in the aorta did not alter the
value of the calculated stroke volume to significant
degree.
Without introducing any empirical correction
factor, the agreement between the predicted and
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
83!
the stroke volume derived from the dye injection or
Fick's procedure was within 12 per cent (44). Further
work brought to light two areas of discrepancies.
First, the predictions tended to underestimate the
actual stroke volumes at high pressure ranges. The
volume-pressure values were then empirically adjusted
to take care of this (94). The correction used is
almost identical to the difference between the single
continuous stretch curve of figure 3 and the curve
connecting the midpoints of the consecutive loops,
which fact suggests that there may be a theoretical
foundation for the empirical correction. The second
and more serious failure of the method is that it
yields a definite overestimation of the actual stroke
volume in some shock states in which the pressure
shows a brisk anacrotic rise with a high shoulder,
but is then poorly sustained later in systole. This
type of pulse has been described above. The possible
causes of the failure in prediction for these rare
pulses were discussed rather fully in a symposium
presented in 1952 (86a) and little more light has
been shed on the problem since.
A modification of our method to make it applicable
to the human has been presented (19, 130). Omitted
from the description here are the attempts to equate
the stroke index with the pulse pressure. Most es-
pecially when the pulse pressure is taken from the
brachial artery, such a prediction can be relatively
gross (95).
In summary, the search for an understanding of
the nature and functions of the distensible vessels
through which blood passes on its way from the heart
to the periphery of the circulation is a fascinating
pursuit. The physical basis of wave propagation and
of the changes in pulse form, attending such propaga-
tion, remain indefinite. Some of the possible factors
which may contribute to such contour changes have
been described. It remains clear that we are in no
position to predict a stroke volume from the form or
pressure values of a peripheral pulse. The form of the
pulse in the upper aorta, however, can reveal much
about cardiodynamics. It can also be used for the
only practical, if indirect, technique yet developed
for reasonably accurate stroke-by-stroke quantitation
of the cardiac output.
REFERENCES
1. Alexander, R. S. Transformation of the arterial pulse
between the aortic arch and the femoral artery. Am. J.
Physiol. 158: 287, 1949.
2. Alexander, R. S. Arterial pulse dynamics in aortic insuf-
ficiency. Am. J. Physiol. 158: 294, 1949.
3. Alexander, R. S. Factors determining the contour of
pressure pulses recorded from the aorta. Federation Proc. 1 1 :
738> '952-
4. Alexander, R. S. The genesis of the aortic standing wave.
Circulation Research I : 145, 1953.
5. Alexander, R. S. Influence of constrictor drugs on the
distensibility of the splanchnic venous system, analyzed on
the basis of an aortic model. Circulation Research 2: 140,
'954-
6. Alexander, R. S. Elasticity of muscular organs. In :
Tissue Elasticity. Washington, DC. : Am. Physiol. Soc,
1957, p. in.
7. Alexander, R. S. Standing wave components in arterial
pulses of hypothermic dogs. Circulation Research 6: 580,
1958.
8. Alexander, R. S., and E. A. Webb. An analysis of
changes in contour of the femoral arterial pulse in hemor-
rhagic shock. Am. J. Physiol. 150: 272, 1947.
9. Bayliss, L. E. Rheology of blood and lymph. In: Deforma-
tion and Flow in Biological Systems. Amsterdam : North-Hol-
land Publ., 1952
10. Bazett, H. C, F. S. Cotton, L. B. LaPlace, and J. C.
Scott. The calculation of cardiac output and effective
peripheral resistance from blood pressure measurements
with an appendix on the size of the aorta in man. Am. J.
Physiol. 113:312, 1935.
1 1. Benninghoff, H. Uber der Beziehungen zwischen elasti-
schen Geriist und glatter Muskulatur in der Arterienwand
und ihre funktionelle Bedeutung. Z. Zellforsch. mikroskop.
Anat. 6:349, 1927.
12. Bleichert, A., R. Lazgus, and F. Martini. Uber die
Lange der stehenden Wellen in der Armarterie des
Menschen. Z. Biol. 105: 141, 1952.
13. Bozler, E. Extensibility of contractile elements. In:
Tissue Elasticity. Washington, DC. : Am. Physiol. Soc,
1957. P- "02.
14. Bramwell, J. C. Change in form of pulse wave in course
of transmission. Heart 12: 23, 1925.
15. Bramwell, J. C, and A. V. Hill. The velocity of the
pulse wave in man. Proc. Roy. Soc, London, B, 93 : 298, 1922.
16. Brewer, G., W. F. Hamilton, and I. Brotman. Pressure
pulse contours in the pulse propagated through the aorta.
Am. J. Physiol. 1 07 : 436, 1 934.
1 7. Broemser, P. Uber die Grundschwingung des arteriellen
Pulses. Z. Biol. 100:88, 1940.
18. Broemser, P., and O. F. Ranke. Uber die Messung des
Schlagvolumens des Herzens auf unblutigem Weg. Z. Biol.
90:467. '93°-
1 9. Brotmacher, L. Evaluation of derivation of cardiac out-
put from blood pressure measurement. Circulation Re-
search $: 589, 1957.
20. Bull, H. B. Protein structure and elasticity. In : Tissue
Elasticity. Washington, DC. : Am. Physiol. Soc, 1957, p. 33.
21. Burton, A. C. Relation of structure to function of the
tissues of the wall of blood vessels. Physiol. Rev. 34: 619,
'954-
22. Cope, F. W. Elastic characteristics of isolated segments of
human aortas under dynamic conditions. J. Appl. Physiol.
"4-55. [959-
836
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
23. Dow, P. The development of the anacrotic and tardus
pulse of aortic stenosis. Am. J. Physiol. 131 : 432, 1940.
24. Dow, P., and W. F. Hamilton. An experimental study
of the velocity of the pulse wave propagated through the
aorta. Am. J. Physiol. 125: 60, 1939.
25. Dow, P., and W. F. Hamilton. Analysis of the emptying
of segments of the arterial reservoir. Am. J. Physiol. 127:
785, '939-
26. Fenn, W. O. Changes in length of blood vessels on infla-
tion. In: Tissue Elasticity. Washington, D.C. : Am. Physiol.
Soc, 1957, p. 154.
27. Ferguson, D. J., and H. S. Wells. Frequencies in pulsa-
tile flow and response of magnetic meter. Circulation Re-
search 7 : 336, 1 959.
28. Frank, O. Kritik der elastischen Manometer. Z. Biol. 44:
445- '9°3-
29. Frank, O. Die Puis in den Arterien. Z. Biol. 46: 441, 1905.
30. Frank, O. Die Theorie der Pulswellen. Z. Biol. 85: 91,
'927-
31. Franklin, D. L., R. M. Ellis, and R. F. Rushmer.
Aortic blood flow in dogs during mechanical exercise.
J. Appl. Physiol. 14: 809, 1959.
32. Frasher, W. G., and S. S. Sobin. Distensible behavior of
pulmonary artery. Am. J. Physiol. 199: 472, i960.
33. Fry, D. L., A. J. Mallos, and A. G. T. Caspar. A cathe-
ter tip method for measurement of the instantaneous
aortic blood velocity. Circulation Research 4: 627, 1956.
34. Fry, D. L., F. W. Noble, and A. J. Mallos. An electrical
device for instantaneous and continuous compilation of
aortic blood velocity. Circulation Research 5: 75, 1957.
35. Furchgott, R. F. Spiral -cut strip of rabbit aorta for in
vitro studies of response of arterial smooth muscle. In :
Methods in Medical Research. Chicago : Yr. Bk. Publ., i960,
vol. 3, p. 177.
36. Hale, J. F., D. A. McDonald, and J. R. Womersley.
Velocity profiles of oscillating arterial flow, with some
calculations of viscous drag and the Reynolds number.
J. Physiol., London 1 28: 629, 1955.
37. Hallock, P., and I. C. Benson. Studies on the elastic
properties of isolated human aorta. J. Clin. Invest. 16: 595,
■937-
38. Hamilton, \V. F. The patterns of the arterial pulse. Am.
J. Physiol. 141 : 235, 1944.
39. Hamilton, W. F. Textbook of Human Physiology (2nd ed.j.
Philadelphia: Davis, 1949, p. 361.
40. Hamilton, W. F., G. Brewer, and I. Brotman. Pressure
pulse contours in the intact animal. I. Analytical descrip-
tion of a new high-frequency hypodermic manometer with
illustrative curves of simultaneous arterial and intracar-
diac pressure. Am. J. Physiol. 107: 427, 1934.
41. Hamilton, W. F., and W. J. Brown. Positive wave reflec-
tion in an elastic model from a wider segment with higher
resistance. Am. J. Physiol. 197: 730, 1 959.
42. Hamilton, W. F., and P. Dow. An experimental study of
the standing waves in the pulse propagated through the
aorta. Am. J. Physiol. 125:48, 1939.
43. Hamilton, W. F., P. Dow, and J. W. Remington. The
relationship between the cardiac ejection curve and the
ballistocardiographic forces. Am. J. Physiol. 144: 557,
■945-
44. Hamilton, W. F., and J. W. Remington. Measurement of
the stroke volume from the pressure pulse. Am. J. Physiol.
148: 14, 1947.
45. Hamilton, W. F., and J. W. Remington. Some factors in
the regulation of the stroke volume. Am. J. Physiol. 153:
287, 1948.
46. Hamilton, W. F., J. W. Remington, and P. Dow. The
determination of the propagation velocity of the arterial
pulse wave. Am. J. Physiol. 144: 521, 1945.
47. Hamilton, VV. F., and J. H. Rompf. Measurements of the
base of the ventricle and the relative constancy of the
cardiac volume. Am. J. Physiol. 102:559, '932-
48. Hardung, V. Vergleichende Messungen der dynamischen
Elastizitat und Viskositat von Blutgefassen, Kautschauk
und synthetischen Elastomeren. Helvet. Physiol, el Pharma-
col. Acta 11 : 194, 1953.
49. Hardung, V. Propagation of pulse waves in visco-elastic
tubings. In : Handbook of Physiology. Washington, D. C. : Am.
Physiol. Soc, 1962, Sect. 2, Chapt. 7.
50. Harkness, M. L., D. R. Harkness, and D. A. Mc-
Donald. The collagen and elastin content of the arterial
wall in the dog. Proc. Roy. Soc, London B, 146: 541, 1957.
51. Hass, G. M. Elasticity and tensile strength of elastic
tissue isolated from the human aorta. A. MA. Arch. Pathol.
34: 97'. '937-
52. Hass, G. M. Relations between structure of the ageing
aorta and properties of isolated aortic elastic tissue. A. MA.
Arch. Pathol. 35: 29, 1943.
53. Inouye, A., and H. Kosaka. A study of flow patterns in
carotid and femoral arteries of rabbits and dogs with an
electromagnetic flowmeter. J. Physiol., London 147: 209,
■959-
54. Jacobs, R. B. Propagation of a disturbance through a
viscous fluid flowing in a distensible tube of appreciable
mass. Bull. Math. Biophys. 16:1 03, 1 954.
55. Jochim, K. E. Electromagnetic flow meter. In : Methods in
Medical Research. Chicago: Yr. Bk. Publ., 1957, vol. i, p.
108.
56. Jones, W. B., E. L. Hefner, J. R. Bancroft, and W.
Klip. Velocity of blood flow and stroke volume obtained
with the pressure pulse. J. Clin. Invest. 38: 2087, 1959.
57. Jungmann, H., and H. Rohr. Uber die Form des Femo-
ralispulses und ihrer Veranderungen unter dynamischer
und mechanischer Beeinflussung. Pfliigers Arch. ges.
Physiol. 258:38, 1953.
58. Kapal, E., F. Martini, and E. Wetterer. Untersuchun-
gen uber die Lange der stehcnden Wellen in arteriellen
System des Menschen. Z. Biol. 104: 256, 1 951 .
59. Kapal, E., F. Martini, H. Reichel, and E. Wetterer.
Uber die Lange der stehenden Welle bei kiinstlicher
Verkiirzung des Arteriensystems. Z. Biol. 104: 430, 1951.
60. Karrerman, G. Reflections of pressure waves in the
arterial system. Bull. Math. Biophys. 14: 327, 1952.
61. Katz, L. N., M. R. Malinow, B. Kondo, D. Feldman,
and H. Grossman. The aortic volume elasticity in the in-
tact dog. Am. Heart ./. 23: 319, 1947.
62. King, A. L. Elasticity of the aortic wall. Science 105: 127,
63
■947-
King, A.
L. Some studies in tissue elasticity. In : Tissue
Elasticity. Washington, D.C: Am. Physiol. Soc, 1957, p.
123.
64 King, A. L., and R. W. Lawton. Elasticity of body tis-
sues. In: Medical Physics. Chicago: Yr. Bk. Publ., 1950, p.
3°3-
65. Krafka, J., Jr. Mechanical factors in arteriosclerosis.
A.M.A. Arch. Pathol. 23: 1, 1937.
PHYSIOLOGY OF AORTA AND MAJOR ARTERIES
837
66. Krafka, J., Jr. Changes in elasticity of the aorta with
age. A.M. A. Arch. Pathol. 29: 303, 1940.
67. Krafka, J., Jr. Comparative study of the histophysics of
the aorta. Am. J. Physiol. 125: I, 1939.
68. Kroeker, E. J., and E. H. Wood. Comparison of simul-
taneous recorded central and peripheral arterial pressure
pulses during rest, exercise and tilted positions in man.
Circulation Research 3: 623, 1955.
69. Kroeker, E. J., and E. H. Wood. Beat-to-beat altera-
tions in relationship of simultaneously recorded central
and peripheral arterial pressure pulses during Valsalva
maneuver and prolonged expiration in man. J. Appl.
Physiol. 8: 483, 1956.
70. Lambossy, P. Oscillations forcecs d'un liquide incompres-
sible et visqueux dans un tube rigide et horizontal. Calcul
de la force de frottement. Velvet. Physiol, et Pharmacol. Acta
25: 37i. '952-
71. Landowne, M. Pulse wave velocity as an index of arterial
elastic characteristics. In: Tissue Elasticity. Washington,
D.C. : Am. Physiol. Soc., 1957, p. 168.
72. Landowne, M. A method using induced waves to study
pressure propagations in human arteries. Circulation Re-
search 5:594, 1957.
73. Landowne, M. Characteristics of impact and pulse wave
propagation in brachial and radial arteries. J. Appl.
Physiol. 12:91, 1958.
74. Lansing, A. I. Elastic tissue. In: The Arterial Wall. Balti-
more: Williams & Wilkins, 1959, p. 136.
75. Laszt, L., and A. Muller. Uber dem Druckverlauf im
Bereiche der Aorta. Helvet. Physiol, et Pharmacol. Acta. 10:
1, 1952.
76. Lawton, R. W. The thermoelastic behavior of isolated
aortic strips of the dog. Circulation Research 2 : 344, 1 954.
77. Lawton, R. W. Measurements of elasticity and damping
of isolated aortic strips of the dog. Circulation Research 3:
403. '955-
78. Lawton, R. W. Some aspects of research in biological
elasticity. In: Tissue Elasticity. Washington, D.C: Am.
Physiol. Soc, 1957, p 1
79. Leonard, E. Alteration of contractile response of artery
strips by a potassium-free solution, cardiac glucosides and
changes in stimulation frequency. Am. J. Physiol. 189: 185,
'957-
80. Mallov, S. Effects of sodium ion and solution tonicity on
reactiveness of hypertensive rat aortic strips. Am. J.
Physiol. 198: 1019, i960.
81. MacWilliam, J. A. Properties of the arterial and venous
walls. Proc. Roy. Soc, London, B, 40: 109, 1902.
82. McDonald, D. A. The velocity of blood How in the rabbit
aorta studied with high-speed cinematography. J. Physiol.,
London 1 18: 328, 1952.
83. McDonald, D. A. The relation of pulsatile pressure to
flow in arteries. J. Physiol., London, 127: 533, 1 955.
84. McDonald, D. A. Blood Flow in Arteries. London : Arnold,
i960.
85. Morgan, G. W., and W. R. Ferrante. Wave propaga-
tion in elastic tubes filled with streaming fluid. J. Acoust.
Soc. Am. 27: 715, 1955.
86. Muller, A. Uber des Druckgefalle in Blutgefassen,
insbesondere in den Kapillaren. Helvet. Physiol, el Pharma-
col. Acta 6: 181, 1948.
86a.OpDYKE, D. F. Genesis of the pressure pulse contour
method for calculating cardiac stroke index. Federation
Proc. 11 : 733-773, 1952.
87. Patel, D. J., A. J. Mallos, and D. L. Fry. Aortic pres-
sure-length-diameter relationship. Federation Proc. 19: 104,
i960.
88. Patel, D. J., D. P. Schilder, and A. J. Mallos. Me-
chanical properties and dimensions of the major pulmo-
nary arteries. J. Appl. Physiol. 15: 92, i960.
8g. Peterson, L. H. Certain physical characteristics of the
cardiovascular system and their significance in the problem
of calculating stroke volume from the arterial pulse.
Federation Proc. 1 1 : 762, 1952.
go. Peterson, L. H. The dynamics of pulsatile blood flow.
Circulation Research 2: 127, 1954.
91. Peterson, L. H., R. E. Jensen, and J. Parnell. Me-
chanical properties of arteries in vino. Circulation Research
8: 622, i960.
92. Ralston, H. J., and A. N. Taylor. Streamline flow in
the arteries of the dog and cat. Am. J. Physiol. 144: 706,
'945-
93. Reichel, H. Die elastischen Eigenschaften des glatten
Schliessmuskels von Pinna noblis bei verschiedenen
Tonuslangen unter plastischen und dynamischen Bedin-
gungen. Z. Biol. 105: 162, 1952.
94. Remington, J. W. Volume quantitation of the aortic pres-
sure pulse. Federation Proc. 1 1 : 750, 1952.
95. Remington, J. W. Relation between the stroke volume
and the pulse pressure. Minn. Med. 37: 105, 1954.
96. Remington, J. W. Hysteresis loop phenomenon of the
aorta and other extensible tissues. Am. J. Physiol. 180: 83,
!955-
97. Remington, J. W. Extensibility behavior and hysteresis
phenomenon in smooth muscle tissues. In : Tissue Elasti-
city. Washington, D.C. : Am. Physiol. Soc, 1957, p. 138.
98. Remington, J. W. Unexplained features of the left ventric-
ular pressure pulse. Am. J. Physiol. 199: 328, i960.
99. Remington, J. W. Contour changes of the aortic pulse
during propagation. Am. J Physiol. 199: 331, i960.
100. Remington, J. W., and R. P. Ahlquist. Effect of sympa-
thomimetic drugs on the Q-T interval and on the duration
of ejection. Am. J. Physiol. 174: 165, 1953.
101. Remington, J. W., and R. S. Alexander. Stretch behav-
ior of the bladder as an approach to vascular distensibil-
ity. Am. J. Physiol. 181 : 248, 1955.
102. Remington, J. W., and R. S. Alexander. Relation of
tissue extensibility to smooth muscle tone. Am. J. Physiol.
185: 382, 1956.
103. Remington, J. W., W. F. Hamilton, and P. Dow. Some
difficulties involved in the prediction of the stroke volume
from the pulse wave velocity. Am. J. Physiol. 144: 536,
1945-
104. Remington, J. W., and W. F. Hamilton. Quantitative
calculation of the time course of cardiac ejection from the
pressure pulse. Am. J. Physiol. 148: 25, 1947.
105. Remington, J. W., and W. F. Hamilton. The evaluation
of the work of the heart. Am. J. Physiol. 150: 292, 1947.
106. Remington, J. W., and R. H. Hugcins. Relation of the
left ventricular ejection period to the Q-T interval of the
electrocardiogram. Am. J. Physiol. 175: 185, 1953.
107. Remington, J. W., C. R. Noback, W. F. Hamilton, and
J. J. Gold. Volume elasticity characteristics of the human
aorta and prediction of the stroke volume from the pres-
sure pulse. Am. J. Physiol. 153: 298, 1948.
838
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
I 08.
109.
"3-
114.
I 15-
116.
117.
118.
119.
I23-
124.
Remington, J. W., and E. H. Wood. Formation of the
peripheral pulse contour in man. J. Appl. Physiol. 9:
433. !956- I25-
Reuterwall, O. P. Die Elastizitat der Gefasswande und
die Methoden ihrer naheren Priifung. Acta Med. Stand. 126.
Suppl. 2, 1921.
Richards, T. G., and T. D. Williams. Velocity changes
in the arterial and femoral arteries of dogs during the 127.
cardiac cycle. J. Physiol., London 120: 257, 1953.
Roach, M. R., and A. C. Burton. The reason for the 128.
shape of the distensibility curves of arteries. Can. J.
Biochem. and Physiol. 35: 681, 1957.
Roy, C. S. Elastic properties of the arterial wall. J. 129.
Physiol., London 3: 125, 1880.
Rushmer. R. F. Pressure-circumference relations of the
aorta. Am. J. Physiol. 183: 545, 1955. 130.
Ryan, J. M., R. W. Stacy, and R. N. Watman. Role of
abdominal aortic branches on pulse wave contour genesis.
Circulation Research 4: 676, 1956.
Schmitt, F. Beitrag zur Frage der Reflexionsbedingungen 131.
und Existenz stehender VVellen im arteriellen Kreis-
laufsystem. Z. Biol. 101 : 259, 1943.
Schnabel, T. G., H. F. Fitzpatrick, L. H. Peterson, 132.
W. J. Kashkind, D. Talley, and R. L. Rapharal. A
technique of vascular catheterization with small plastic
catheters. Circulaltm 5: 257, 1952. 133.
Sinn, E. Die Elastizitat der Arterieren und ihre Bedeutung
fur die Dynamik des arteriellen Systems. Akad. Wiss. 134.
Lit., Main.; 1956, p. 647.
Speden, R. N. The effect of initial strip length on the 135.
noradrenaline-induced isometric contraction of arterial
strips. J. Physiol., London 154: 15, i960. 136.
Spencer, M. P., F. R. Johnston, and A. B. Denison.
Dynamics of the normal aorta. Circulation Research 6: 137.
491. >958-
Spencer, M. P., and A. P. Denison, Jr. The aortic flow
as related to differential pressure. Circulation Research 4: 138.
476, 1956.
Smith, D. J. Immediate sensitization of isolated swine
arteries and their vasa vasorum to epinephrine, acetyl- 139.
choline and histamine by thyroxine. Am. ./. Physiol. 177:
7. '954-
Stacy, R. W. Reaction rate kinetics and some tissue
mechanical properties. In: Tissue Elasticity. Washington, 140.
D. C: Am. Physiol. Soc, 1957, p. 131 ■
Stacy, R. W., and F. M. Giles. Computed analysis of 141
arterial properties. Circulation Research 7 : 1 03 1 , 1 959.
Starr, I., and A. Schild. A test of the aortic compression
chamber hypothesis and of two stroke volume methods
based on it. J. Appl. Physiol. 1 1 : 169, 1957.
Van Citters, R. L. Longitudinal waves in the walls of
fluid-filled elastic tubes. Circulation Research 8: 1145, i960.
Van Citters, R. L., and R. F. Rushmer. Longitudinal
and radial strain in pulse wave transmission. Federation
Proc. 19: 104, 1960.
Wagner, R., and E. Kapal. Uber Eigenschaften des
Aortenwindkessels. Z. Biol. 104: 169, 1951.
Warner, H. R. Synthesis of central arterial pressure
pulse contour from recording of radial artery pressure in
man. Am. J. Physiol. 183: 670, 1955.
Warner, H. R. A study of the mechanisms of pressure
wave distortion by arterial walls using an electrical
analog. Circulation Research 5: 79, 1957.
Warner, H. R., H. J. C. Swan, D. C. Connolly, R. G.
Tompkins, and E. H. Wood. Quantitation of beat-to-beat
changes in stroke volume from the aortic pulse contour
in man. J. Appl. Physiol. 5: 495, 1 953-
Wetterer, E. Flow and pressure in the arterial system,
their hemodynamic relationship, and the principles of
their measurement. Minn. Med. 37: 77, 1954.
Wetterer, E. Die Wirkung der Herztatigkeit auf die
Dynamik des Arteriensystems. Verhandl. deut. Ges. Kreis-
laufforsch. 22: 26, 1956.
Wezler, K. Der Ruhezustand des Kreislaufs. Z. Biol.
98:438. '938-
Wezler, K., and A. Boger. Die Dynamik des arteriellen
System. Ergeb. Physiol. 41: 292, 1939.
Wiggers, C. J . The Pressure Pulses in the Cardiovascular
System. New York: Longmans, 1928.
Wiggers, C. J. Circulation in Health and Disease. Phila-
delphia: Lea & Febiger, 1923.
Wiggers, C. J. The influence of vascular factors on mean
pressure, pulse pressure and phasic peripheral flow. Am.
J. Physiol. 123: 644, 1938.
Wiggers, C. J., and R. Wegria. Active changes in size
and distensibility of the aorta during acute hypertension.
Am. J. Physiol. 124: 603, 1938.
Womerslky, J. R. The mathematical analysis of the
arterial circulation in a state of oscillatory motion. WADC
(Wright Air. Develop. Center), Tech. Rept. No. 56/614,
1958.
Woodbury, R. A., and W. F. Hamilton. Blood pressure
studies in small animals. Am. J. Physiol. 119: 663, 1 937.
Zatzman, M., R. W. Stacy, J. Randall, and A.
Eberstein. Time course of stress relaxation in isolated
arterial segments. Am. J. Physiol. 177: 299, 1954.
CHAPTER 25
Pulsatile blood flow in the vascular system
MERRILL P. SPENCER
ADAM B . DENISON, JR.
Department of Physiology and Pharmacology, Bowman Gray
School of Medicine, Winston-Salem, North Carolina
CHAPTER CONTENTS
Methods of Measurement
Properties and Principles of Flowmeters
Cognate Phenomena
Elements of Vascular Hydraulics
Resistance
Inertance
Compliance
Axial Flow
Radial Flow
Hydraulic Impedance
Flow Source Versus Pressure Source
The Analogy Approach
Systemic Arterial Flow
Blood Flow in the Ascending Aorta
Pressure-Flow Relationship in the Ascending Aorta
Ventricular Ejection Gradients
Windkessel Model of the Arterial System
Aortic Transformation of Flow and Pressure Pulses
Resonant — Network Model of the Arterial System
Transmission Line Model (Distributed System)
Distribution of the Blood in the Aortic Arch
Abdominal Aorta and Its Terminal Branches
Function of the Resonant Wave
Renal Blood Flow
Carotid Artery Flow
Coronary Blood Flow
Flow in the Systemic Veins
Effect of the Heart's Action on Vena Caval Flow
Effect of Normal Respiration on Vena Caval Flow
Pulmonary Flow
Right Ventricular Ejection Pulse
Pulsatile Flow in the Pulmonary Capillary Bed
Nonlaminar Flow and Murmurs
Normal Murmurs
Relationship Between the Murmur of Coarctation Stenosis
and Blood Flow Through the Stenotic Area
The Murmur Envelope and Contour Rule
General Rules Relating Murmurs to Nonlaminar Flow
Normal and Pathological Flow Pulses in Humans
Flow in the Ascending Aorta
Descending Thoracic Aorta
Tricuspid Valve
THE HIGHLY PULSATILE NATURE of the blood flow in
both the systemic and pulmonary circuits primarily
arises from the intermittent action of the heart as a
pump. Each ventricle has a valve at its exit and
entrance such that the blood flow and velocity
oscillate from near zero, when the valves are closed,
to relatively great values during the time when the
valves are open. Great changes in the velocity arise
from the starting and stopping of the blood stream
with the opening and closure of these valves. Second-
ary causes of flow pulsations, particularly in the veins,
arise from the respiratory fluctuations and muscular
contractions.
I. METHODS OF MEASUREMENT
Methods for the detection of blood flow and pres-
sure oscillations within this system require a frequency
response flat to at least ioo cps and without phase
shift. Present day pressure systems achieve this ideal
quite well if one does not introduce long elastic
catheters between the pressure tap and the transducer.
Present day blood flowmeters, however, have not
achieved this degree of perfection, but recently great
progress has been made. In addition to the frequency
and phase characteristics mentioned, a blood flow-
meter should be capable of detecting the blood flow or
839
840
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
velocity from either the surface of a surgically exposed
vessel or by means of a catheter tip introduced into
the flow stream, but without causing significant dis-
tortion of the flow dynamics.
All the blood flow recordings gathered by the
authors for this chapter, if not otherwise indicated,
have been made with a 240-cycle square-wave elec-
tromagnetic flowmeter (11) introduced in 1953 as the
first practical instrument for measuring blood flow in
any of the body's arteries and veins which have been
surgically exposed. As used here, this instrument has a
flat frequency response to 40 cps, and is down by 50
per cent at 100 cps, and, although the principle is
capable of an infinite frequency response, these
limitations are necessary in a practical instrument
primarily because of the carrier frequency residual
which would otherwise appear on the flow record.
The magnetic probes applied to the blood vessels
restrict pulsations and encroach on the lumen to the
extent of reduction in cross-sectional area by approxi-
mately 5 to 10 per cent. Such slight constriction
assures firm contact of the electrodes to the arterial
wall. Experimental testing showed that this amount of
constriction caused no perceptible change in the re-
corded flow pulse (45).
Properties and Principles oj Flowmeters
For recording of vascular flow velocity pulses an
ideal flowmeter should possess several properties: a
linear response to forward and backward flows, a
stable zero reference, and a frequency response ade-
quate to follow the phasic phenomena being re-
corded. It should also be unaffected by nonrelated
phenomena such as blood pressure, internal noise,
and muscle action potentials. Furthermore, its opera-
tion should not modify the phasic flow patterns, mean
flows, or blood pressure. To meet this last requirement
completely would mean that not only must the blood
vessel under consideration be unobstructed and non-
cannulated, but also that the recording be done
without anticoagulants or anesthesia and without
psychic trauma to the experimental subject (59).
Obviously, a practical flow-recording system re-
quires some compromise with the above ideals; also,
such an elegant device would not be necessary for
most research work. If one knows the general charac-
ter of the quantities to be recorded, he may use with
confidence an equipment the characteristics of which
are considerably more restricted than the ideal. For
instance, a frequency response of zero to 50 cps is felt
to be adequate for cardiovascular work (6, 46); also,
when recording ascending aorta flow, zero drift is not
serious since the flow can be taken to be zero at the
end of diastole, thus giving a continuously repeated
zero check. Therefore, an instrument used for cardiac
output measurements may have considerable drift of
zero and still be satisfactory for the purpose if it meets
the other requirements, although it might be unsatis-
factory for other situations where a stable zero refer-
ence is essential (14).
Many different principles have been used for flow
recording; all have inherent potentialities for errors in
application or interpretation. These principles and
the instruments which embody them are discussed in
detail in Chapter 38. Here it will be necessary only to
list the different types of instrument and certain
references to the literature which are not found else-
where either in this chapter or in Chapter 38.
a) Electromagnetic flowmeters (1, 3, 5, 8, 9, 16,
24> 53> 57> 63). b) Ultrasonic flowmeters (25). c)
Nuclear magnetic resonance (4, 23). d) Pendulum or
bristle flowmeters, e) Catheter tip pickups (36, 38).
/) Turbinometers (40, 41). g) Differential pressure
flowmeters (10, 17, 39).
Cognate Phenomena
LATERAL AND DIFFERENTIAL PRESSURES. Ill any Critical
study of the relationship of the dynamics of pulsatile
flow it is necessary that pressure and flow be measured
simultaneously, and that the pressure be picked up
from a pressure tap the orifice perimeter of which is
in a plane parallel to the flow stream. One highly
practical system is to use a "clip needle" which by
means of a flexible clip holds the end of the needle
against the inside of the blood vessel wall (49). If a
Huber point is used on the clip needle, the recorded
pressure can be a true lateral pressure.
The use of differential pressure measurements has
greatly enhanced our interpretation of the phenomena
occurring simultaneously within the vascular system.
Such a method usually takes the form of two pressure
taps conducted separately to either a differential
pressure transducer or to two individual pressure
transducers the amplified signals of which are elec-
trically subtracted from one another continuously.
The latter system has the advantage of being able to
view the individual pressures which make up the
differential pressure recording. These individual re-
cordings are useful in identifying artifacts which may
arise.
computer techniques. For a proper understanding of
the hemodynamics of the cardiovascular system, a full
PULSATILE BLOOD FLOW
84I
appreciation should be had for the relationship be-
tween flow velocity, volume, and displacement. These
relationships may be expressed best by means of
calculus symbology, as follows:
Displacement (cm) = / Velocity (cm/sec) dt
Jo
Velocity (cm/sec) = / Acceleration (cm/sec2), dt and
Jo
Volume (cm1)
Flow (cm3/sec)
= / Flow (cm3/sec) dt
Jo
f
Jo
Volume acceleration (cm3/sec2) dt
For example, the cardiometer tracings during the
systolic ejection period may be said to be the negative
integral of flow through the aortic and pulmonary
valves, and the diastolic cardiometer tracing is the
integral of the flow through the A-V valves. Also, the
radial displacement of the arteries and veins may be
said to be the integral of the radial velocity of blood
flow within the lumen.
Analogue computer techniques, useful in the study
of vascular hemodynamics (50), allow one to move
from volume to flow to acceleration by means of
integration, or the reverse, through differentiation.
Two types of integration are currently in use: /) the
true time integral which is an instantaneous sum of a
given function, beginning from any given time; and
2) damping or "meaning," an older usage of the word
which implies a mean value of a periodic function.
Damping may be accompished either mechanically or
electrically. The most practical way to perform this
mechanically in a pressure recording system is to
introduce compliance or resistance into the trans-
mitting system, e.g., by means of a bubble in the
gauge or a partial occlusion clamp on the catheter
tubing. Damping in an electrical system amounts to a
fully charged integrating circuit in which the rate of
current inflow into the integrator over one pulse
cycle equals the rate of current outflow.
Resistance (R)
This arises from the friction between shearing mole-
cules flowing through the segment. Expressed in
terms of the pressure difference (15) across the re-
sistance, APR , in dynes per square centimeter1; the
blood velocity, u, in centimeters per second; the cross-
sectional area, A, in square centimeters; and the flow
(FR = uA), in cubic centimeters per second,
APP APR
u-A
F*
(I)
After Poiseuille, in terms of vessel dimensions, length
(/) in centimeters, radius (r) in centimeters, and blood
viscosity (77) in dynes • second per square centimeter,
8t,1
(2)
The vessel wall also has a small resistance opposing
radial distention and collapse. The inverse of resist-
ance or conductance (1//?) is often a useful term.
The symbol for hydraulic and viscous resistance is
taken from electronics ( A/WV~).
Inertance (L)
This resides primarily as the mass of blood and
secondarily as the mass of the arterial wall. It is ex-
pressed in terms of acceleration (a), and attendant
pressure difference across the inertance, AP;, , in
dynes per square centimeter.
AP,
AP,
a- A dF,
(3)
L/dt
where FL is the flow through the inertance, and
where L = m/A2, m is the mass of the blood in the seg-
ment of artery under consideration expressed in grams,
and A is the cross-sectional area in square centimeters.
In terms of vessel dimensions (55), / and r, in centi-
meters and blood density (pH) in grams per cubic
centimeter,
II. ELEMENTS OF VASCULAR HYDRAULICS
The arterial system is a many-branched elastic
conduit for distribution of blood from the heart to all
body tissues. The caliber ranges from 35 mm for the
human aorta to 7 n for the capillaries. Over this wide
range each vascular segment may be described by
three fundamental physical properties: resistance,
inertance, and compliance.
Pf, I
(4)
1 Pressure in dynes per square centimeter should be used
instead of the conventional pressure in millimeters of mercury.
The following expression is used to convert from millimeters of
mercury (h) to pressure in dynes per square centimeter (P) :
P = 0.1 g pneh = '323 X mm Hg
where g is the acceleration of gravity in cm/sec2, and pnK is the
842
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
The symbol for inertance is that for electrical induc-
tance (13£R5T~"). Because inertance is defined in terms
of volumetric acceleration, the larger the cross section
of the vessel lumen, the smaller is the inertance in
a vessel of given length.
-'ffOTffWW^
)
Compliance (C)
Compliance is a property of the arterial wall
arising from its distensibility and chiefly residing in the
elastic fibers. The contribution of smooth muscle and
fibrous tissue has not been determined. It is expressed
in terms of blood volume (V) in the segment and the
attending pressure difference across the vascular wall
APC.
APC APC J
(5)
where Fr is the flow into the compliance, or in terms of
dimensions (22) length (/), radius (r), wall thickness
(0), and the modulus of elasticity (£) :
C°
2irr-
Eo-l
(6)
The symbol for compliance is that of electrical
capacitance ( — I 1 — ).
Axial Flow
In a segment of rigid pipe axial flow is analogous to
the current in the diagram of figure 1 . Where APaxiai
= pl — P2 , APa at any instant in time will be equal
to the sum of the pressure differences due to the R and
L components. Thus :
A P.
axial
= AP,
*■ AP
Rtsislono r " r Inertanca
(7)
(Any pressure gradient resulting from gravity
cancels if the pressures are referred to the same level.)
Substituting from equations 1 and 3, and considering
FR = FL = F,
AP=RF+L
dF
dt
(8)
integrating with respect to time we have
F-j-f(AP0-RF)dt (9)
This equation may be solved continuously by an
analogue computer and has some practical applica-
density of mercury at the existing experimental conditions in
g/cm3.
fig. I. Electrical analogue of axial flow and pressure in a
rigid tube.
i
Fr I
J'
fic. 2. Electrical analogue of radial How in an elastic tube.
tion in the ascending aorta (12, 13). The procedure
is to subtract P> from Pi to obtain i\P, and then to
subtract RF from AP and integrate the result. If
j/L is known or is chosen arbitrarily, R may then be
adjusted until F achieves some known boundary
condition such as F = o during diastole. Figure 3.Z?
graphically illustrates the procedure. If accurate
values for vessel dimensions, blood density, and
viscosity are available to calculate L and R, the
result can be obtained in terms of actual flow in
cubic centimeters per second, otherwise the answer
only yields the velocity in centimeters per second.
Radial Flow
In a visco-elastic artery, radial flow is analogous to
the current in the diagram of figure 2.
4Pm*m"A%*AIk
(10)
where APradiai represents the pressure difference across
the arterial wall (PT\ — Pr2). Substituting from equa-
tions 1 and 5, where Fr = Fc = FR ,
AP,'7-fFdt + RF
(II)
and differentiating,
PULSATILE BLOOD FLOW
843
F = C
dAPr
dt
R
dF
dt
(12)
or rearranging equation 1 1
w [Apr-~rfFdt] <">
Since measurements of pressure and vessel diameter
are very similar, friction within the arterial wall and
radial inertance are apparently quite small, although
in the final analysis, as clearly indicated by Peterson
(35), one must consider acceleration along with dis-
tensibility and friction.
When the total flow (FT) in an elastic pipe is con-
sidered, both radial and axial flow equations must be
combined as follows for instantaneous flow: FT =
!'„,,,, 1 + FTmiial, and, from equations 9 and 12,
V Jrf(APo-Ralra)dt+ (14)
In analogue computer language this equation is
solved as in figure 3. Patel et al. (33) have found
negligible degrees of inertiance and resistance in the
pulmonary artery wall.
A more complete hydraulic diagram of an arterial
segment may be well shown as in figure 4. L, , Ri ,
and C'i represent its most important elements, with
R2 and R3 representing radial and axial resistance,
and C-i representing axial compliance. The complete
arterial system may be viewed as a continuous linkage
of such segments, each branch and segment having
quantitative differences in magnitude of the individual
physical elements. At the same time, the physical
elements of any segment or group of segments may be
described by over-all "lumping" of the elements.
The arterial system is not a passive network because
the elements may be influenced by the nervous
system, endocrine system, metabolic processes in the
wall, and changes in the physical properties of the
blood. In addition, the values of the elements are
nonlinear functions of pressure, vascular dimensions,
velocity profile and many other influences. In spite of
these complications, much can be learned by linear
analysis of the pressure and flow pulses at various
fig. 3. Analogue computer diagram for solution of the equa-
tion of liquid flow in an clastic tube. P, and ft represent the
lateral pressures from two stream points. AP is the independent
variable (AP = Pl - ft).
U R,
L,
-VWA
R,
fig. 4. Elaborate electrical analogy of a vascular segment.
points within the arterial network studied under
reasonably steady-state conditions.
Hydraulic Impedance
This is the concept of total opposition to pulsating
and constant flow. Drawing on the electrical symbol-
ogy, we have Z, XL, A'c, and XR, where Z represents
the total impedance, and A'L, Xc, and XR equal the
inertial, compliant, and resistive impedances. XR is the
opposition to flow, XL is the opposition to change in
flow, and A'c is the opposition to change in volume. Both
844
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
X, and Xc depend on the frequency (v) as follows:
/
Xj2iryL
XC 2wvC
The impedance to blood flow through L and C
elements will therefore be expected to be frequency
dependent, and may be termed hydraulic reactance
in contradistinction to resistance which is not fre-
quency dependent. Hydraulic impedance mayr be
expressed as dP/dF, i.e., the rate of change of pressure
with respect to simultaneous rate of change of flow.
On a pressure-flow diagram, impedance would be
represented by the tangent to the curve at any given
time.
Flow Source Versus Pressure Source
The impedance or "stiffness" of the flow source
and pressure source may be expected to influence the
response of a vascular segment. For example, a small
branch, such as a renal artery arising directly from
the aorta, is fed by a stiff pressure source, inasmuch as
great changes in renal vascular impedance encoun-
tered within extreme physiological ranges have no
effect on the abdominal aorta pressure.
On the other hand, the left ventricle without
external controls behaves as a flow source because
relatively great changes in systemic arterial imped-
ances (viz., aortic stenosis, hypertension, vasodila-
tion) cause small changes in the cardiac output. On
a beat-to-beat basis, therefore, the left ventricle may
be considered to be a stiff flow source or volume pump,
and the response of the arterial system is greatly
influenced by this fact. The performance of the left
ventricle as a volume pump is illustrated in figure 5.
The carotid sinus feedback loop tends to make the
heart rate and strength of contraction vary inversely
with the arterial pressure, but requires several beats
for its correcting action. Hormonal negative feedback
loops also act on the heart to cause it to perform as a
stiff pressure source, but act even more slowly than
the reflexes.
The Analogy Approach
This is: a) to diagram an electrical network model
of specific segments of the arterial system based on
qualitative facts available from physiology by identi-
fying blood pressure and blood flow with electrical
voltage and current; b) to test the model against
conditions in an experimental animal by pulsing a
direct analogue or an analogue computer with
electrical voltage or current transduced from the
100
80
60
8
0
+ 100
+ 50
0
-50
BLOOD PRESSURE
5
AORTIC BLOOD
FLOW
AP
(LV.P-A.A.P)
0.5 sec
fig. 5. Response of the dog's left ventricle to sudden increase
in outflow resistance caused by partial occlusion of the as-
cending aorta. — ■ — Control; moderate obstruction;
■ • • • severe obstruction.
blood pressure or flow. The computer is programmed
to solve the equation of the electrical network, but in
this case in terms of pressure and flow instead of
voltage and current. Several general considerations
of the analogue approach are available (32, 34, 54,
60).
Considerations of this section approach the vascular
system from the standpoint of a transient response as
distinguished from the usual use of steady-state
oscillation in which the harmonic content must be
known to reach a solution (20, 26, 51, 61, 62). The
transient response method has the advantage of
giving an instantaneous solution while in addition
each term of the equation has physiological meaning.
To regard the arterial pulse as a steady-state oscilla-
tion is to fail to recognize the input pulse and the
response of the vessels as two independent phenomena
PULSATILE BLOOD FLOW
845
and overemphasize the regularity of the heart rate.
Also, the terms of a series such as the Fourier have no
real physiologic meaning and in fact may fail to show
a dominant and important frequency such as the
arterial resonant wave.
III. SYSTEMIC ARTERIAL FLOW
Shipley et al. (44) and Pritchard et al. (37) made
one of the most comprehensive recordings of the
arterial flow pulses using the differential pressure
flowmeter. They offered no fundamental theory to
explain the recorded phasic pressure-flow relationship.
Although some exception may be taken to their flow-
meter, the general form of the flow pulses agrees well
with more recent electromagnetic noncannulating
recordings.
Blood Flow in the Ascending Aorta
The arterial network is pulsed by a flow pulse from
the left ventricle normally of the configuration in
figure 6. This recording is taken with the square-wave
electromagnetic flowmeter on the ascending aorta, 3
to 5 cm distal to the aortic valve. What were ap-
parently the first accurate phasic recordings were
made by Wetterer (58). The linear acceleration of the
blood by the left ventricle is remarkable, reaching
greater than 8000 cm per sec per sec in an anesthe-
tized open-chest dog (47). At the end of acceleration,
the velocity of the blood in the ascending aorta may
easily exceed 100 cm per sec in the resting state.
Deceleration takes place at a rate approximately one-
sixth of acceleration until closure of the aortic valve
when a sharp notch of deceleration and acceleration
brings the flow to nearly zero for the duration of
diastole.
For many practical purposes this flat "uneventful"
tracing during diastole in the ascending aorta may be
used as a zero flow reference to compute the stroke
volume. The fact that coronary flow is not included
may produce a small unknown error. Apparently, the
diastolic flow curve in the ascending aorta is flat at
nearly zero because the reversing effect of coronary-
flow is balanced by the forward effect of decompres-
sion of the first portion of the ascending aorta. The
left ventricular ejection velocity at the root of the
aorta recorded by Pieper (fig. 7) is similar to the flow
pulse throughout the ascending aorta. Since this
instrument records the axial velocity, it appears that
the velocity profile of the ascending aorta is relatively
0.5 sec
10-
c
—
5-
^
0-
2.5-
fig. 6. Flow pulses in the ascending aorta of an unanesthe-
tized dog. C-core electromagnetic probe was implanted 6 weeks
prior to this record on the ascending aorta. Electrical connec-
tions were made by means of implanted subcutaneous wires,
brought to the surface through a small superficial incision.
Aortic Pressure
133 mm Hg
fig. 7. Axial flow pulse in the ascending aorta recorded by
means of velocity probe situated in midstream. [From Pieper
(36).]
flat. More backflow occurs here during early diastole
presumably because of diastolic coronary flow.
The effect of exercise on the ventricular ejection
pulse is illustrated by a remarkable experiment by
Olmsted (personal communication), figure 8. The
animal had a magnetic probe implanted on the as-
cending aorta and an arrangement for remote pres-
sure recording. After one month's recovery from the
surgical procedure he was exercised by running in a
harness behind a station wagon carrying recording
equipment. Suitable wiring carried the electrical
signals between the automobile and the dog. The
course was one-half mile over rough terrain at an
average speed of 10 mph. Upon standing, the cardiac
output increased primarily because of increased
heart rate without change in stroke volume and with
little change in form of the ejection pulse. Running
at 5 mph increased cardiac output by increasing both
846
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
T~ig """TH
table i . Ascending Aorta Flow During Exercise
ID
I — * — zz — ~~
r_^^i"_T~U I
o
a
O'S
x.
a
5
dt
fig. 9. Computer solution to blood flow in the ascending
aorta using the analogue computer of fig. 3. Femi = the flow
measured by the electromagnetic flowmeter. Ft = computed
How. C(dP/dt) = the radial flow pattern derived from the time
differential of the aortic arch pressure P.
flow, is F = C (dP/dt) + (i/R) P. When this electrical
model is experimentally pulsed by a current trans-
duced from the flow in the ascending aorta (F), the
voltage form (V) is obtained. By comparison, the
actual pressure pulse in the aortic arch (P) deviates
in several details from 1": 1) P has a superimposed
3 to 6 cps oscillation apparent from midsystole
throughout diastole, and 2) P has a more prominent
"incisura" marking aortic valve closure and a more
abrupt rise, often with an anacrotic wave. In addition,
the windkessel model fails to explain the changes in
form occurring along the arterial network. Detail 2
appears if the analogy is elaborated by the placement
of some restraint on the distensible element, i.e.,
taking into consideration the friction in lateral
expansion of the arterial wall, as in equations 1 2 and
13. Detail / requires a concept of reflections or
resonant network filter as explained in the succeeding
paragraphs. Cope (7) has attempted new use of the
windkessel concept using empirical constants.
Aortic Transformation of Flow and Pressure Pulses
Figure 1 2 illustrates the changes in form and
magnitude of the flow pulses between the ascending
aorta and the abdominal aorta. The flow in the
descending thoracic aorta represents an intermediary
form and well illustrates the superimposition of a
prominent smooth 3 to 6 cps wave decreasing in
amplitude throughout diastole. This wave referred to
as the "resonant" wave frequently causes backflow in
diastole throughout the aorta and many of its
branches. Considered as a whole, the arterial system
is a low-pass filtered hydraulic supply, i.e., it is
designed to offer negligible impedance to steady flow
and frequencies up to 10 cps. There is normally one
frequency between 3 and 6 cps to which it offers
lowest impedance, and resonates at that frequency
with each beat of the heart. Early physiological
workers recognized this resonant system as analogous
to a low-frequency underdamped manometer system.
Resonant-Network Model of the Arterial System
This represents an improved concept to explain
the transformation of arterial pressures and flow
pulses. It is diagramed in figure 13. C\ roughly
represents the lumped compliance of the aortic arch
and its branches, and C2, the lumped compliance of
the abdominal aorta and its branches. L represents
the lumped inertance of the blood in the descending
aorta. FH represents the forcing function of the left
848
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
fig. 10. Relationship between
the differential pressure and flow
through the normal aortic valve.
The ordinates refer to the tops of
these simultaneous tracings. AP
= left ventricular pressure {LVP)
minus ascending aorta pressure
(AAP). The pressure gradient
is against the direction of flow
in the latter part of systole.
Stroke Vol. = 18.2 cc
**A»«»
^-Vvj"^ -_\w\xv^t^ ^-L^V "yJ
Bockflow Vol. = 0.58 cc
ventricular ejection pulse. R\ and R2 represent periph-
eral resistances which may be adjusted relative to
Ci and C2 to give any desired ratio of pulse pressure
to mean pressure at Pi.
Ci, C2, and L form a series resonant circuit and may-
be adjusted to give any resonant frequency, and Rz
and Rt are chosen to provide the proper damping
ratio of the observed resonant wave in the arterial
system, as well as the high frequency details. When
C2 is smaller than Ci, the P> pulse pressure is greater
than Pi pulse pressure, thus explaining a time-
honored observation that the arterial pressure in the
legs rises higher than in the arms during systole.
Figure 14 demonstrates the degree of accuracy
with which such a grossly lumped electronic model
may reproduce the observed set of pressure and flow-
values in the arterial network. The resonant-network
model embodies several concepts which provide a
rational explanation of the major hydraulic features
of the arterial system.
/) The over-all frequency response characteristics
of the arterial system may be taken as that of an
analogous filter network (56), figure 15.
2) The resonant frequency (vn) varies from 2 to 10
cps and is increased by hypertension produced by
increased cardiac output and sympathetic con-
strictor agents. It is also increased in cardiac failure
due to mitral stenosis. Hypotension from decreased
fig. 11. Electrical analogue of the windkessel model of the
arterial system with experimental testing. P represents the
pressure in the aortic arch. V represents the voltage across the
parallel resistor and condenser. F represents the measured
blood flow in the ascending aorta and the electrical input cur-
rent forcing the analogy.
PULSATILE BLOOD FLOW
849
0.5 sec
H
y I 26 mm Hg
Abdominal Aorta
table 2. Distribution of Arterial Flow Pulse
Mean Peak Velocity
fig. 12. Transformation of the aortic blood flow between the
ascending aorta and abdominal aorta. High frequencies are
attenuated and a resonant frequency is superimposed.
cardiac output decreases the resonant frequency.
Presumably the over-all compliance changes more
than the inertance in these conditions.
3) The amplitude of the resonant wave is increased
when, in tachycardia, the systolic flow pulse is in
phase (48) with the resonant wave.
4) The pressures and movements in the arterial
system represent, at any steady state of the hormonal
and nervous controlling conditions, transient re-
sponses to the flow input from the left ventricle.
5) The augmentation of the pressure pulse, as it is
transmitted to the abdominal aorta, results from the
lower gross compliance of the abdominal arterial bed
as compared to the aortic arch vascular bed.
Artery
Circumfer
ence, cm
Mean Peak
Flow, ml min
cm min
cm/sec
Ascending aorta
6.28
II ,870
3.780
63
Descending
5°
3.243
1,621.5
27
thoracic aorta
Abdominal
3-4
1,108
1 ,256.2
21
aorta
I , 260
■ -429
23
Iliac
1.8
750
3.275
54-6
Femoral
0.7
182
2,045
34-1
Renal
1.4
176
1 ,401
23-4
Carotid
'•5
'93
1 ,162
■9-4
Brachial
0.7
94
'.774
29.6
6) The 30-100 cps components prominent in the
central aortic pressure pulse, as in the anacrotic
wave and the incisura, result from the stiffness of the
arterial walls and are damped out as they proceed
away from the heart. The dicrotic wave so prominent
in the peripheral pulse does not arise from this
source but is an expression of the resonant wave
phenomenon.
Transmission Line Model {Distributed System)
This is a useful concept in the arterial system, as in
any hydraulic continuum. It is represented by van der
Tweel (55) in figure 16. No matter how short or how
long a given segment may be, there is always present
some combination of inertance, compliance, and
resistance which may be lumped in a close approxima-
tion of the behavior of that particular segment.
The performance of the transmission line is greatly
affected by the relation of the terminating impedance
to the characteristic impedance (52) of the line. If
the terminal impedance is equal to the characteristic
impedance all the energy will be absorbed and no
reflections occur. The characteristic impedance,
however, is frequency dependent, increasing with
frequency. If the terminal Z is greater than the line
Z, positive reflections will occur. Negative reflec-
tions will occur if the terminal Z is less than the
characteristic Z.
Distribution of the Blood in the Aortic Arch
This is shown in figure 17. In a manner analogous
to Kirchoff's current law, the flow into the arch at
any given instant from the ascending aorta is equal
to the sum of the flows into the brachiocephalic and
left subclavaian arteries, and the flow into the de-
scending thoracic aorta plus the uptake rate of the
85o
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
€>
R
I
IOOK
I5MFD.'
/ '
/
/
/ 9
M FH
/ -i-
/
Peripheral /
Resistance r.' ,„,„
u o m p 1 1 a n c e
nppnr^
©
200h / 0.5K
/
/
Inertiance
BLOOD MASS
fig. 13. The resonant network model of the arterial system. Component values indicated are those
found in one typical experiment on an anesthetized open-chest dog.
Adjust Restraint (R4) of Compliance (C2) Adjust Restraint (R3) of Compliance (C,)
P,
1 1 '
■ m
■ ?= t i
-J J.
\ 1 \
r^/. — v" ' „^. — ,v-
\
— *-V~~~-fS* '■ —wu.— —,'■-'
V,
R3' 0 R4-O
t. !
1
t
R3*0 R4 1.5 K
R3-0.6 R4*I.5K
v2
1 1
fig. 14. Experimental testing of the resonant network model of the arterial system. Pi = pressure
in the aortic arch. Pi = pressure in the femoral artery. F = flow in the ascending aorta. l\ = the
voltage across the capacitor- C, to ground. V« = the voltage across the capacitor Ci to ground. In
the first panel the values of G and Co have been adjusted to give the correct resonant frequency as
represented in the arterial pulses. Between the first and second panels, R4 was adjusted to give the
proper damping ratio of the series resonant elements C\ , C« , and L. The third panel shows the
closest equivalent achieved by adjusting R:i to reduplicate in Vi the high components of Pl . Delay in
transmission time is not present because the simplicity of the analogue limited the number of L-C-R
transmission line segments.
I 2 3 4 5 6
fig. 15. Transfer function of the arteries computed by War-
ner (56). i/n is equal to a resonant frequency.
40 KS =4=
1 H
0 Olp
0 005fJ
0 Oil
110*0 01 11)
-nST^~
0005JJ
=r=0SC
fig. 16. Transmission line model of the arterial system,
showing the stacking of L-C transmission line segments. [From
van der Tweel (55).]
compliance of the arch (CAA), or
dP
Faa = FBroch + F5lJbcl + FDTA + CAA — (15)
figure 1 7 demonstrates this fact experimentally by
comparing a plot of the instantaneous sums of -FBrach.>
FDTA., FSubci., and CAA (dP/dt) to FAA. The value of
CAA was adjusted arbitrarily.
Abdominal Aorta and Its Terminal Branches
By the time the pressure and flow pulses reach the
abdominal aorta, the highest frequency components
are so attenuated that the flow pulses are dominated
by a strong resonant wave superimposed on the mean
forward flow (fig. 18). The resonant flow wave in the
abdominal aorta is in phase with that in the de-
scending thoracic and the resonant pressure wave
[standing wave of Hamilton & Dow (21)] of the
abdominal aorta is 180 degrees out of phase with
z
= 5-
in
a.
u
1-
PULSATILE BLOOD FLOW 85 1
125.
Aortic Arch
Pressure (P)
100^3"
\\ ^Flow Out of Aortic Arch
12 F2,F3, F4, F5)
Flow into Aortic Arch (F,)
(Left Ventriculor Outflow)
H
0 5 sec
0— V
5—1
5 —
Aortic Arch Radial Flow (F2)
* —
Flow out Arch Branches
< S F3, F4, F5 )
Descending Thoracic Aorta Flow (F3)
Brachiocephalic Artery Flow (F4 )
Subclavian Artery Flow (F5)
fig. 17. Distribution of blood flow in the aortic arch in a
manner analogous to KirchhofF's current law. It is shown that
the volumetric flow of blood into the arch of the aorta is equal
to the sum of the instantaneous flows into the subclavian artery,
brachiocephalic artery, descending thoracic aorta, and the
radial flow uptake in the aortic arch. The dotted line in the
lower section illustrates the branch outflows (not including
radial flow).
that in the upper aorta (48). These phase relation-
ships are similar to those of the series resonant circuit
of figure 13. When the resonant flow wave reaches a
maximum moving down the aorta, the attending
pressure wave in the arch is falling most rapidly.
When the resonant flow wave reverses, and flows at
maximal rate headward in the aorta, the pressure
falls rapidly in the lower aorta while it rises rapidly
in the arch. There is a nodal area in the descending
thoracic aorta where the pressure wave is minimal
(2) and the flow wave is maximal. These findings
support the resonant-network model of the arterial
system.
Figure 18 demonstrates the remarkable simul-
852
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
I 25 mm Hg
fig. 1 8. Blood flow in the aortic branches. All flow ordinates
are scaled equally. The contours of the various flow pulses here
may be considered characteristic of the flow in the indicated
branches. Carotid flow and renal flow characteristically pulsate
around a mean value representing considerable continuous
forward velocity. Blood flow in the femoral artery, iliac artery,
abdominal artery, and brachial artery may, in the resting condi-
tion, oscillate through zero in early diastole but are also, under
the conditions of muscular exercise, or metabolic demands, or
vasodilator drugs, raised to a level corresponding to considera-
ble mean forward velocity.
taneity of the peaks and troughs of the abdominal and
descending thoracic aorta, iliac, and femoral flow
pulses. The time of the initial rise is delayed according
to the transmission time between the two points under
comparison.
Function of the Resonant Wave
The finding of a large backflow component to the
flow wave in the descending aorta and vessels of the
extremities is at first surprising when viewed from
the point of efficiency needs of the circulation. This
finding, however, observed in the resting state of
dogs (31), sheep (F. C. Greiss, unpublished observa-
tions), and in humans (48), disappears upon exercise
of the extremities as the muscle vascular beds dilate
to accommodate a greater flow.
The normal terminal impedance (peripheral
resistance) of the arterial transmission line is ap-
parently greater than the characteristic impedance
during the resting state. The vasodilator mechanisms
of exercise bring the terminal impedance down to and
below that of the line, thereby eliminating positive
reflections. Negative reflections do not arise because
they are damped out by the resistance of the larger
channels made more effective by increased flow. The
circulation is thus brought up to more efficient
operating conditions when the demands are increased.
All the pulse energy passing to the periphery is
completely absorbed without reflections when the
peripheral resistance is decreased by exercise, injection
of vasodilator drugs, and in peripheral A-V fistulas.
Figure 1 9 illustrates the action of lowered terminal
impedance in increasing the more efficient transfer
of energy. Reflection from the bed beyond, seen in
the control blood flow of a small artery in the dog's
paw, disappears under vasodilation conditions caused
by an intra-arterial injection of acetylcholine. The
resonant flow wave disappears and the flow is a
simpler function of the arterial pressure. Okino
[see (42)] has also recorded these changes.
Renal Blood Flow
The renal vascular circuit may be, as a first approxi-
mation, compared to a simple parallel RC circuit
(30). The dominant hydraulic elements of the renal
artery flow are resistance and compliance, and the
equation relating abdominal aorta pressure (Pi), and
renal artery flow (F) is:
dt R
(16)
PULSATILE BLOOD FLOW
853
^ ,0
CONTROL
FLOW
ACETYLCHOLINE
FLOW
ISO
glOO
6
SO
AP
VP
VP
-"•9--
fig. 19. Blood flow in a small peripheral artery and the effect
of vasodilation. According to the definitions of the text, the
control flow may be considered a resonant flow form which is
converted to resistant flow form by the injection of acetylcholine
into the arterial channel. AP represents the arterial pressure,
immediately proximal to the flowmeter probe applied to a small
artery in the dog's paw. VP represents the venous pressure in a
small vein of the dog's paw. Conversion of the flow from reso-
nant flow to resistant flow by the action of acetylcholine lowers
the arterial pressure and raises the venous pressure. (Tracings,
courtesy of M. C. Conrad and H. D. Green.)
i/R represents the conductance at the existing pres-
sure, and A' represents the fact that the pressure-
resistive-flow relationship (excluding the dynamic
compliant flow term, C(dP/dt)), is not constant and is
a nonlinear function of pressure. Presumably this
results from the fact that i/R is directly dependent on
the pressure in a manner similar to that shown by
the vascular beds of the skin. If this is true, then the
relationship is:
°dt +
I
R(v) + R(P)
(17)
where R{v) equals resistance controlled by vaso-
motor tone, and R(F) equals resistance controlled by
intraluminal pressure P. At present, the coefficients
C], Riy, and R(P) are obtained only by measuring
the flow and pressure without any means of indirect
evaluation. Figure 20 illustrates one example of how
C and R of equation 16 were adjusted until the
dynamic flow pulse was computed from P (30). In
this case flow was alreadv known from simultaneous
measurement with the square-wave electromagnetic
flowmeter.
There are apparently no positive reflections from
the normal renal bed, hence the flow computed for
the total renal circuit according to equation 16 and
without an inertance term represents the flow in the
renal artery. To compute, however, the instantaneous
flow in other arteries from whose bed there are
reflections one must use the difference in pressure
along the artery (i.e., two pressure sources in the
artery itself) and the equations 9 and 12. The most
important term is then the inertial one of equation
9 although, as explained earlier, a further step in
precise computation brings in the compliance of
equation 12.
Carotid Artery Flow
This is illustrated in figure 18. Like the renal flow
there is a large constant flow component upon which
there is superimposed a dynamic component. It is
related to the carotid pressure by equation 1 6 with
the conductance term i/R being the largest by far.
Because the carotid and renal flow patterns are
governed largely by the resistance, they are called
viscous or resistance flow patterns. The resting flow
patterns of the entire aorta, iliac, femorals, and
subclavian arteries are called reactance flow patterns
because inertance and compliance are dominant.
They have a small constant flow when compared to
their dynamic component and this frequently demon-
strates a period of negative flow in early diastole.
Coronarx Blood Flow
Coronary blood flow in the unopened artery was
first measured by Marston and Spencer with the
square-wave electromagnetic flowmeter (24). The
resting patterns differ little from those of Gregg (18)
who used a cannulating system and orifice meter.
The equation relating coronary flow to the vascular
pressures is a modification of equation 16:
'">§++"'
(18)
where AP equals the aortic pressure minus the
ventricular pressure minus the right atrial pressure.
Figures 21 and 22 (27) illustrate the measured left
anterior descending coronary flow and the circumflex
coronary artery flow.
These coronary inflow curves display a marked
dependence on both aortic pressure and intra-
854
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 20. Computed renal blood flow.
P = the abdominal aorta pressure at
the level of the renal artery. dP/dt =
the first derivative of the abdominal
arterial pressure. Fc = computed blood
flow using equation 16. The lower
tracing is measured blood flow using
the square-wave electromagnetic flow-
meter on the renal artery.
dt
Fc
v\. w Vy
500
Femf
0. 5 sec
-^vv'
ventricular pressure where coronary resistance
is a complex function of vasomotor tone, intra-arterial
pressure, and intramuscular pressure. The function of
intramuscular pressure which reduces inflow during;
systole increases venous outflow during systole and
may also increase capillars- flow at the same time.
IV. FLOW IN THE SYSTEMIC VEINS
Phasic variations in venous blood flow result from
three principle sources: /) the beat of the heart;
2) the respiratory fluctuations; and j) the contraction
of skeletal muscle. Severe changes in position and
acceleration of the body also may have profound
effects on venous flow. Pulsatile flow originating from
the heart beat may occur in the small peripheral
veins, as a result of transmitted oscillations from the
arterial system. These pulsations are generally small
in the normal condition because of the damping
action of the resistance of the small arteries and
arterioles and the elasticity of the capillary bed. They
may be accentuated, however, by vasodilatation,
either by reactive hyperemia or by means of drugs,
such as acetylcholine. Flow in the renal vein, normally
phasic presumably because of the low renal vascular
resistance, causes less damping than in most vascular
beds. Great variations in blood flow within the
thoracic vena cava have been recorded by Brecher
(5) and others.
Effect of the Heart's Action on Vena Caval Flow
It has been shown by Gauer and Sieker [quoted in
(5)] that there is an almost immeasurable gradient
in the mean blood pressure along the venae cava
toward the heart. Since there is a net movement of
blood in that direction, some small gradient must be
present which may be sufficient, in view of the large
size of the channels, to move considerable blood. It
is also true, according to the principles of vascular
hydraulics discussed in sections II and III, this
chapter, that considerable blood may be moved by
an oscillatory pressure gradient without consideration
of a mean frictional gradient (fig. 23).
Atrial contraction injects a late diastolic quota of
blood through the tricuspid valve and also causes a
pressure transient to pass along the vena cava away
from the heart. This pressure transient produces a
sharp reduction in flow which may or may not cause
a reversal depending upon its amplitude and the level
of mean flow (fig. 24). This impediment or reversal
is, however, overcome immediately by a large for-
ward flow caused by ventricular contraction. This
"vis a fronte" which draws blood toward the heart
during ventricular systole arises from movement of
the base of the heart (tricuspid valve closed) toward
its apex, producing a transient pressure gradient in
favor of flow toward the heart. Flow may be expected
to follow this differential pressure transient, approxi-
mately 90 degrees out of phase.
As the base of the heart moves away from the apex
during diastole, the tricuspid ring dilates and filling
of the heart takes place as much by the ventricle
sliding over the atrial blood as by the atrial blood
flowing into the ventricle. From direct observations
of the heart and from slow motion movies, it can be
seen that the ventricle fills by a) dilation of the tri-
PULSATILE BLOOD FLOW 855
••
r ■ _3 t 1 [
JlJo^Ci: i«4&iT-eV t J«— 2« "= :Hi 3Cp" *— , V- ~ -" "V — - — "
n —
:i:.._.....± 4=
LEFT VENTRICULAR PRESSURE
200
mm
Hg
0 -
\ "^ ^ \, "L
fed— H ffV ^r-
±■"+3 J-" :
:r t-1 ±:
±Tt -u i 1-
-1 e -_j i
_ l/w' /V^ tv--7
I^/~ ht? 3'
A(
)R
ru
: i
XR
:h
p
RE
SS
UF
E
V
V
1
^
*•
/
*\
1
**.
, 1
*"■>
^s
J
ng
0
;
LEFT ANTERIOR DESCENDING
133 -
b
^
i
fe
^
1
T Stt
^
g
E
1
fig. 21. Coronary blood flow in the left anterior descending
branch. The upper tracing represents the electrocardiogram
taken simultaneously with the left ventricular pressure, aortic
arch pressure, and coronary flow. [From Schenk (43).]
cuspid ring, b) engulfing of the atrial blood by the
right ventricle, and c) a final quota of blood delivered
by atrial contraction. The question of whether or not
the ventricle produces a sucking force during diastole
is unresolved. The answer will await definitive
differential pressure measurements made across the
ventricular wall.
Effect of Normal Respiratio
Carol Flow
This is illustrated in figure 25 (28). Inspiration
greatly increases the venous return as shown in the
thoracic vena cava and abdominal vena cava caudal
to the renal veins.
V. PULMONARY FLOW
Right Ventricular Ejection Pulse
The form of the right ventricular ejection pulse
(fig 26) differs from that of the left ventricle in
! 1 i i 1 1
i. !
■ i
1
r ' a
A
Hi
f
i
"— 1 ,. _J .^
-'%-
- - ■ ■• -4 1
r - H±r
—
LEFT VENTRICU
lAr
prYssur
E
200
ASCENDING 'AORTA PRESSURE
200 —
mm
Hg I
est
fcs:
ci rcumfl ex to ro nAry "artery
*l
&si:
FLOW
f^
232
I SEC
fig. 22. Coronary blood flow in the circumflex branch.
Tracings taken from the same animal as in fig. 2 1 . [From
Schenk (43).]
general by an over-all lack of the higher frequency
components. The initial acceleration is slower, peak
more rounded, at a somewhat lower velocity; and the
reverse flow due to pulmonary valve closure forms a
more rounded notch followed by a lower frequency
aftervibration. There are also fewer random frequency
vibrations throughout diastole. Presumably these
differences arise from a slower rate of contraction of
the right ventricle versus that of the left, and a
greater compliance per unit of arterial wall in the
pulmonary artery than that of the aorta, the latter
arising perhaps from the lower distending pressure.
When flow recordings are taken off the pulmonary
trunk near the bifurcation, one frequently notices a
low-frequency vibration during the diastolic period
which may be due to some reflections from the pulmo-
nary periphery.
Measurements of differential pressure across the
pulmonary valve between the right ventricle and
pulmonary artery display less of a tendency for the
856
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
A P = P, - P2
fig. 23. The "raking in" or "vis a fronte" action of the right
ventricle and tricuspid valve ( diastole ; systole), and
the principle of blood movement without net gradient.
SURCAVA PRES.
fig. 24 Pulsatile How in the superior vena cava [Brecher (5)]
as measured by the bristle flowmeter, showing the effect of
changes in heart rate on inflow into right atrium. Venous
return is phasically recorded with the bristle flowmeter in the
superior vena cava (open chest). From above downward the
tracings are time, aortic pressure in mm Hg, superior vena
caval flow in ml/min, and superior vena caval pressure in mm
HjO. PAS denotes peak of atrial systole.
systolic pressure reversal demonstrated already across
the aortic valve. This difference is possibly due to two
causes: first, there may be more effective resistance in
the pulmonary artery due to the sharp turn that it
makes immediately after arising from the ventricle,
and, second and more importantly, since there is a
smaller acceleration and more sustained peak flow,
the differential term of the equation relating flow and
differential pressure is relatively small. This equation
rewritten for convenience is as follows: AP = L/
dF dt + RF.
Pulsatile Flow in the Pulmonary Capillary Bed
Direct observations through the microscope of
pulmonary capillaries in vivo clearly indicate a
markedly pulsatile character of the blood flow. In
addition, nitrous oxide uptake curves in the body
plethysmograph which represent an integral relation-
ship to the pulmonary capillary flow demonstrate
pulsatile flow through the vessels perfusing the al-
veoli. These pulsations are relatively small, however,
and are superimposed upon a strong mean flow
through the same vessels. Pulmonary capillary
pressures, taken through the wedged cardiac catheter,
generally demonstrate a marked pulsatile pressure
in the pulmonary capillaries, and thus represent
indirect evidence of phasic flow in these vessels.
VI. NONLAMINAR FLOW AND MURMURS
Normal Murmurs
Careful evaluation of the normal circulation for
murmurs by means of sensitive microphones, in-
cluding the application of a barium titanate phono-
catheter directly to the surface of the heart and
blood vessels, has been made. The assumption is
made that the presence of a murmur indicates a
nonlaminar and turbulent flow pattern. Frequently,
one can detect a brief systolic murmur in the arch of
the aorta corresponding in time to the peak of the
ejection pulse. Groom (19) has also shown con-
siderable indirect evidence that systolic murmurs can
frequently be recorded from normal humans with
sensitive microphones on the body surface under the
low-background noise conditions of a soundproof
room. In addition, low-frequency vibrations can be
recorded from the normal cardiac chambers during
diastole by means of intracardiac phonocatheters.
Relationship Between the Murmur of Coarctation
Stenosis and Blood Flow Through the Stenotic Area
A study of the pressure-flow-murmur dynamics in
coarctation of the aorta illustrates many principles
applicable to stenosis of the larger arteries as well as
flow through the pathological heart valve orifices of
stenosis and regurgitation.
Coarctation was produced by progressively con-
stricting a wire loop passed around the descending
aorta of an experimental animal (49) (fig. 27).
Changes in the flow contour were noted during
constriction from a normal diameter of 6.5 mm down
PULSATILE BLOOD FLOW
857
rO
-100
-200
IT
IA
■ ■ I —
1.0
SECONDS
20
90
■70
L50
30
fig. 25. The effect of natural
breathing on flow in the thoracic
and abdominal vena cava of the
dog (closed chest). Tracings
represent from above downward.
AO = the aortic pressure in mm
Hg; IT = intrathoracic pressure
in mm H20; VC-\ = thoracic
inferior vena caval blood flow in
ml/sec; VC-i = flow in IVC
below renal veins in ml/sec;
RA = right atrial pressure in
mm H20; FV = femoral vein
pressure in mm H20;and IA =
intra-abdominal pressure in mm
H,0. [After Mixter (28).]
PULMONARY BLOOD FLOW
AORTIC BLOOD FLOW
-i ■■■>■ AfttfH
fig. 26. Left and right ventricular
ejection pulses in congenital atrial
septal defect before and after repair
(7 -year-old boy). Repair consisted of
an open-heart procedure using total
cardiopulmonary bypass while closing
the defect by means of a suture tech-
nique. Figures between the flow pulse
tracings represent the mean output of
the ventricles averaged over several
heart cycles. The contour of the trac-
ings is typical of normal tracings found
in humans and dogs. Stroke volume
differs markedly, however, between the
two ventricles before and after repair.
The repair diminished pulmonary flow
and increased the aortic flow. Pre-
sumably the difference between the
pulmonary and aortic flow after repair
results from either /) incomplete
closure, or 2) actual difference in the
cardiac output between the measure-
ments which were not taken simul-
taneously.
to an internal diameter of 3 mm. Blood flow through-
out these degrees of experimental coarctation was
maintained at the normal level and the stroke flow
was maintained primarily by flattening the peak flow
and broadening of the systolic area. During this time
a systolic murmur began softly and increased in
loudness and duration; its envelope maintained a
contour similar to the contour of the peak of the flow
858
HANDBOOK OF PHYSIOl I » , -i
CIRCULATION II
AP j'ISOmmHg
.124
752
ml/min
-446
4,059
CONTROL
AP .-l26mmHg
AF
AS
iFA^/ID. 2mm A;2,87l i
/ V / \ml/min
«•»*
m>
fig. 27. Experimental graded coarctation of the descending
thoracic aorta. ID = internal diameter; AP = aortic pressure;
AF = aortic flow in the descending thoracic ; and AS = aortic
sounds. The sounds are taken by means of a barium titanate
phonocatheter downstream to the point of constriction. During
control, systolic pressure was 1 30 mm Hg, diastolic 86 mm Hg.
The peak systolic flow was 4,752 ml/min, while the mean flow
in the descending thoracic aorta was 446 ml/min. In spite of the
reduced peak flow, there was little reduction of mean during the
early stages of constriction because of change of contour of the
flow pulse. The pressure gradient, flow pulse contour and mur-
mur envelope follow the "contour rule."
pulse. Presumably, this remarkable reduction in
cross-sectional area without reduction in flow is
attributable to the progressively increasing gradient
across the stenotic area.
Beyond this degree of obstruction any further
change in the internal diameter becomes extremely
critical as far as blood flow is concerned. With an
internal diameter of approximately 2 to 3 mm, the
murmur consistently filled systole throughout, and
further reduction caused the murmur to increase in
duration beyond the second sound and extend into
the diastolic period. Internal diameters of 1 mm or less
frequently caused a continuous, high-pitched, blow-
ing murmur distal to the site of coarctation in both
experimental animals as well as in congenital lesions.
The Murmur Envelope and Contour Rule
The "envelope" of a murmur is defined as the
amplitude of the full wave rectified murmur averaged
over several heart cycles. The term envelope is
similar to the "shape" of a murmur which itself
means amplitude of the unrectified murmur. The
envelope of a blowing murmur follows a "contour
rule," which means that it corresponds closely to the
contour of the flow pulse originating the murmur.
This is true because apparently once the critical
velocity is reached where turbulent flow begins
(turbulence is used in a general sense to indicate
nonlaminar flow), the amplitude of the resultant
turbulence or lateral velocities of the nonlaminar
flow is proportional to the mean axial velocity.
Further, the flow under these conditions is principally
viscous in nature and therefore the extant resistant
pressure gradient contour parallels the flow pulse
contour. The correspondence of the murmur envelope
and the pressure gradient to the flow pulse exists only
when the flow is highly viscous in nature (having no
significant reactance flow term), and may occur
without stenosis in a normal vessel under high-
velocity conditions producing nonlaminar flow or in
an aneurysm where nonlaminar flow may be achieved.
Some examples of the contour rule are given in
section VII.
General Rules Relating Murmurs to Nonlaminar Flow
From this and other studies in section VII, general
rules concerning the interpretation of frequency band
width and envelope (amplitude and duration) of
"blowing" murmurs may be made relative to the
functional anatomy of the source as follows:
/) Blowing murmurs with high pitch and low
intensity are associated with small orifices through
which blood is flowing at high velocity, driven by
large pressure gradients.
2) Loud, blowing murmurs of relatively low-
frequency spectrum, generally sounding coarse to
the ear are associated with relatively large orifices
through which large volumes of blood flow under
relatively high-pressure gradients.
3) Very low-frequency (rumbling) murmurs of low
intensity are associated with turbulent flow beyond
large orifices through which blood flows under low-
pressure gradients.
PULSATILE BLOOD FLOW
859
4) The contour of the time-intensity pattern or
"envelope" of a murmur corresponds to the contour
of the flow pulse passing through the region at the
time of murmur production.
5) Musical murmurs, that is, murmurs with periodic
reproductions in the frequency pattern as opposed to
the random vibrations of blowing murmurs, arise
from tissue structures, or other coherent material,
set into oscillation by blood flow of high velocity.
Examples of general rule / are found in the con-
tinuous aortic murmur of severe degrees of coarcta-
tion and the diastolic murmur of minimum aortic re-
gurgitation. Examples of general rule 2 are moderate
degrees of coarctation, aortic and pulmonary valve
stenosis, Korotkoff's sounds, patent ductus arteriosus,
and the murmurs of most arteriovenous fistulae.
Examples of general rule 3 are mitral stenosis, tri-
cuspid stenosis, and occasional right atrial murmur
of an interatrial septal defect. Examples of general
rule 4 are really found among all murmurs wherever
one compares the murmur envelope with the flow
pattern as illustrated in section VII on flow in patho-
logical conditions. Examples of general rule 5 are the
vibrations of a vein wall giving rise to a "venous
hum," the "sea gull" murmur arising from vibration
of aortic valve cusps in aortic regurgitation, and the
"moaning" systolic murmur of retroverted mitral
cusps, or arising from vegetation on the mitral cusps
giving rise to a systolic low-frequency periodic
murmur in mitral regurgitation. In addition, musical
qualities may be heard in arteriovenous fistulas which
are presumably due to the vibration of the vascular
wall, and are usually superimposed, like most musical
murmurs, upon blowing or random noise vibrations.
VII. NORMAL AND PATHOLOGICAL FLOW
PULSES IN HUMANS
The normal human flow pulses closely resemble in
pattern those which have been found in the cor-
responding vessels of the dog and sheep. Most records
of human flow pulses have been made at the time of
a surgical procedure indicated because of some
pathological condition. The pulses in this section
presented as "normal" tracings are so called because
there was no physiological reason to doubt their
normalcy and second, because they correspond to
those found in the experimental animal.
Flow in the Ascending Aorta
Normal blood-flow patterns are shown in figure
26. As in the dog, they show a rapid acceleration
phase, a slower deceleration phase with a high-
frequency (40-60 cps) backflow coincident with aortic
valve closure. The volume of the flow pulse has little
effect on these contour features (fig. 26) except from
hypodynamic ventricles and severe degrees of exercise.
Diastole is relatively uneventful with the resonant
wave not appearing. The ejection pattern of the left
ventricle in the presence of aortic valve stenosis is
shown in figure 28. The principal deviations from
normal contour seen here are a more flattened and
delayed peak, with flow vibrations superimposed.
In addition, there is less prominence of the valve
closure backflow wave. The backflow wave incident
20
16 -
t
12 -
c
, m ,»f 1
A
ekgT
JL
fig. 28. Aortic valve stenosis of rheumatic origin without
regurgitation in a 1 4-year-old boy. Measurements were made
during thoracotomy prior to repair. Flow pulse shows the
rounded irregular top of turbulent blood flow, corresponding to
a diamond-shaped murmur and a pressure gradient between the
left ventricle and ascending aorta which follows the contour rule
86o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 29. Blood flow and pressure gradient in aortic valve
regurgitation without stenosis in a 19-year-old girl. In ascending
aorta flow, the zero reference is estimated. The regurgitant
blood flow, the envelope of the diastolic murmur, and the pres-
sure gradient between ascending aorta and left ventricle follow
the contour rule. The blood flow pattern during systole is not
greatly altered from that of the normal, in spite of increased
stroke volume (see also atrial septal defect, fig. 26).
to valve closure is not, per se, dependent upon
absence of aortic stenosis, but rather upon the degree
of flexibility of the valve. If the valve is stiff and
immobile, this wave will disappear. If the valve, as
in congenital stenosis, is flexible, this wave will
persist and also one may differentiate subaortic
stenosis or outflow stenosis from valvular stenosis by
the presence of a stenosis pattern during systole which
retains the normal amplitude of the valve closure
wave.
The severe pressure gradient across the stenotic
valve, as shown by the difference in the aortic pressure
and ventricular pressure when recorded directly, is
quite similar in contour to the flow-pulse contour
(fig. 28). The murmur, which is harsh, loud and
blowing, and is located in the ascending aorta and
arch, has an envelope with a contour closely paral-
leling that of both the peak of the flow curve and
differential pressure curve. Experimental aortic
stenosis produced by means of a wire tightened about
the ascending aorta at the sinus of Valsalva is shown
in figure 5. The effect of varying degrees of stenosis
is demonstrated on the flow curve, the aortic arch
pressure, and the differential pressure between
ventricle and ascending aorta. Also of note here is
the heart's ability to maintain a stroke volume in the
face of this severe increase in load impedance. Under
the conditions of this experiment, in which several
heartbeats were allowed for cardiac compensatory
mechanisms to act, the left ventricle functions as a
constant flow source. Further compensatory mech-
anisms brought into play over long periods of time,
particularly allowing hypertrophy of the left ventric-
ular wall, further enhance the heart's ability to
maintain a constant stroke with a severe increase in
load impedance.
Aortic regurgitation causes changes both in the
systolic and diastolic contour of the left ventricular
ejection pulse (fig. 29). The systolic stroke volume
exceeds the normal volume by the amount necessary
to compensate for regurgitation during diastole. As a
result the flow pulse tends to be somewhat more
rounded, but otherwise maintains the general shape
of the normal pulse. However, because of the greater
stroke volume, there is necessarily a greater accelera-
tion and deceleration at the onset and termination of
ejection. The valve closure notch disappears, and in
its place one sees a sustained backflow deflection.
The backflow, diminishing throughout diastole, is a
function of the diastolic pressure gradient across the
valve, and the size of the regurgitant orifice. The
murmur, which has a wide frequency band extending
above 100 cps, sounds high pitched and blowing,
begins with reversal of the pressure gradient across
the valve, usually builds up early in diastole to a
maximum, and then follows a decrescendo pro-
portional to the backflow. As seen from figure 29,
the envelope of the diastolic murmur is similar to the
pattern of the diastolic backflow.
Descending Thoracic Aorta
Figure 30 illustrates flow in the descending thoracic
aorta immediately distal to a ductus arteriosus, before
and after the closure of the ductus. The descending
PULSATILE BLOOD FLOW
86 1
Ductus Open
Ductus Occluded
Flow f
"J?
Back +
AGE 3 YR
28 FEB 1958
fig. 30. Flow in the descending
thoracic aorta distal to a ductus arterio-
sus while patent and after the ductus
was occluded. Simultaneous recording
of the descending aorta pressure was
made by means of a needle inserted
near the square-wa%'e electromagnetic
probe.
thoracic flow following closure of this ductus may be
taken as the shape of the normal flow in the human
descending thoracic aorta. Variations in contour
show basic similarities to those of the dog's descending
thoracic aorta of section III. It is of interest to note
that the mean forward flow in the descending thoracic
aorta distal to a patent ductus arteriosus is not
affected by closure of the ductus. This finding indi-
cates what is confirmed by flow in the descending
thoracic aorta proximal to the patent ductus, namely,
that the left ventricle compensates for a ductus
arteriosus by increasing its output just a sufficient
amount to make up for the flow passing through this
-JWtESSUaE
180-1 A«t<»
fig. 31. Blood flow through a patent ductus arteriosus. Con-
tour of the flow pulse follows closely that of the contour of the
aortic pressure thereby indicating strong predominance of
resistant blood flow in this situation. The murmur was con-
tinuous and had an envelope the contour of which followed the
contour rule. Pulmonary pressures were normal.
fig. 32. Blood flow through a patent ductus arteriosus in a
patient with pulmonary hypertension. There was a net flow of
blood into the aorta. Flow from pulmonary artery into the
aorta took place during systole, and from aorta into the pulmo-
nary artery during diastole. The greater pulse pressure in the
pulmonary artery as compared to that in the aorta undoubtedly
resulted from the lower compliance of the pulmonary tree as
compared to that of the systemic arterial tree. The three records:
aortic pressure, pulmonary artery pressure, and patent ductus
flow, were taken at different times in rapid succession.
shunt to the pulmonary artery. The flow through the
ductus is predominantly viscous in type, as its contour
follows closely that of the contour of the differential
pressure between the aorta and pulmonary artery
(fig. 31). The envelope of the murmur follows the
contour rule of the differential pressure and flow
pulse in a viscous flow situation. The murmur
envelope and flow contour are closely represented
862
HANDBOOK OF PHYSIOLOGY ^- CIRCULATION II
BEFORE RESECTION
.186
#
mean flow 660 ^„^
475 320
AP "
(mmHg)
AFTER RESECTION
,140
I D 6mm^^^F
>4 100 mean "ow
>l.800
^
80
flow
-Q—
^ i i ye «ijy 'i Ifui'mv ,V
I '
< ES > «
Medium Frequency
*****
300'
mim
WtHm
«*-
niMii'i «\m*
600~
I I
I 2
« — - ES >
High Frequency
Transcription
Wtt
■WW
w>
I I
I 2
fig. 33. Congenital coarctation. Tracings from above downward both before and after resection
are : descending thoracic aorta pressure, blood How through the coarctation area, low-frequency
phonocardiogram, and high-frequency phonocardiogram. Blood flow records both before and after
resection were made at the same sensitivity setting of the electromagnetic flowmeter. Zero reference
after resection was not obtainable because of danger of rupture of the suture line. On the basis that
zero was somewhat lower than the lowest flow point, the mean flow was estimated to be greater than
1800 ml/min after resection, while the mean flow before resection was 475 ml/min. Some constric-
tion remained after resection and suture.
fig. 34. Tricuspid regurgitation after the differential pressure
method of Miiller & Shillingford (29). Blood flow is recorded
between the superior vena cava and the right atrium in a
normal subject in record A, and in a patient with tricuspid
incompetence and high venous pressure in record B.
by the contour of the aortic pressure pulse alone,
because normally it greatly exceeds the pulmonary
artery pressure at all times throughout the cardiac
cycle. Because of this situation, the murmur is con-
tinuous throughout the cardiac cycle.
Figure 32 illustrates a flow pulse through a ductus
of an unusual type. In this situation, chronic pulmo-
nary hypertension had developed until the pulmo-
nary pressure exceeded that in the aorta during
systole, and was less than that in the aorta during
diastole. (The larger pulse pressure in the pulmonary
tree than in the systemic arteries probably resulted
from the smaller compliance of the pulmonary tree.)
As a result, ductus flow was from pulmonary artery
to aorta during systole, and from aorta to pulmonary
artery during diastole.
Coarctation of the aorta also produces a viscous
type flow through the stenotic region (43, 49). Most
patients with coarctation have a severe degree of the
type illustrated in figure 33. The differential pressure
and murmur envelope both follow the contour rule
with reference to the flow pulse. Flow pulses in the
aortic branches are considerably altered (43).
Tricuspid Valve
Blood flow between the superior vena cava and the
right atrium was measured by a pitometer by
Miiller & Shillingford (29). This record probably
represents a close approximation of the flow pulse at
the tricuspid valve except for the period of atrial
contraction which is inverted to show forward flow
(fig. 34)-
PULSATILE BLOOD FLOW
863
REFERENCES
1. Albertal, G., R. H. Clauss, A. M. Fosberg, and D. E.
Harkens. Flowmeter for extracorporeal circulation. IRE
Trans, on Med. Electronics ME-6 : 246, 1 959.
2. Alexander, R. S. The genesis of the aortic standing wave.
Circulation Research 1 : 145, 1953.
3. Assali, N. S., K. Dasgupta, A. Kolin, and L. Holms.
Measurement of uterine blood flow and uterine metab-
olism. Am. J. Physiol. 195: 614, 1958.
4. Bowman, R. L., and V. Kudravcev. Blood flowmeter
utilizing nuclear magnetic resonance. IRE Trans, on Med.
Electronics. ME-6: 267, 1959.
5. Brecher, G. A. Venous Return. New York : Grune & Strat-
ton, 1956.
6. Cooper, T., and A. W. Richardson. Electromagnetic
flowmeters. Comparative pulsatile blood flow contours
demonstrating the importance of RC output circuit
design in electromagnetic blood flowmeters. IRE Trans, on
Med. Electronics ME-6: 207, 1959.
7. Cope, F. W. An elastic reservoir theory of the human
systemic arterial system using current data on aortic elas-
ticity. Bull Math. Biophys. 22: ig, i960.
8. Cordell, A. R., and M. P. Spencer. Electromagnetic
blood flow measurements in extracorporeal circuits. IRE
Trans, on Med. Electronics ME-6: 228, 1959.
g. Cordell, A. R., and M. P. Spencer. Electromagnetic
blood flow measurement in extracorporeal circuits : its
application to cardiopulmonary bypass. Ann. Surg. 151:71,
i960.
10. Crittenden, E. E., Jr. An electronic recording flowmeter.
Rev. Sci. Inslr. 15: 343, 1944.
11. Denison, A. B., Jr., and M. P. Spencer. Square-wave
electromagnetic flowmeter design. Rev. Sci. Instr. 27: 707,
I95°-
12. Fry, D. L. The measurement of pulsatile blood flow by the
computed pressure gradient technique. IRE Trans, on Med.
Electronics ME-6: 259, 1959.
Fry, D. L., F. W. Nobel, and A. J Mallos. An electric
device for instantaneous and continuous computation of
aortic blood velocity. Circulation Research 5: 75, 1957.
Fry, D. L. Physiologic recording by modern instruments
with particular reference to pressure recording. Physiol.
Revs. 40: 753, i960.
Green, H. D. Circulatory system: physical principles. In:
.Medical Physics, edited by Glasser. Chicago: Yr. Bk. Pub.,
1950, vol. 2, pp. 228-251.
16. Green, H. D., K. Ottis, and T. Kitchen. Autonomic
stimulation and blockade on canine splenic inflow, outflow
and weight. Am. J. Physiol. 198: 424, i960.
17. Green, H. D., A. W. Richardson, and A. B. Denison, Jr.
A direct reading differential pressure flowmeter composed
largely of commercially available components, and having
a linear calibration. J. Lab. Clin. Med. 39: 314, 1952.
18. Gregg, D. E. Coronary Circulation in Health and Disease.
Philadelphia: Lea & Febiger, 1950.
19. Groom, D., W. Chapman, W. W. Francis, A. Bass, and Y.
T. Sihvonen. The normal systolic murmur. Ann. Internal
Med. 52 : 134, i960.
20. Hale, J. F., D. A. McDonald, and J. R. Womersley.
Velocity profiles of oscillating arterial flow, with some
calculations of viscous drag and the Reynolds number. J.
Physiol ., London 128:629, 1 955.
'3
14
>r>
23
24-
25-
26.
27.
29-
3°-
3i-
32-
33-
34-
35-
36.
37-
38.
39-
40.
41-
Hamilton, W. F., and P. Dow. An experimental study of
the standing waves in the pulse propagated through the
aorta. Am. J. Physiol. 125: 48, 1939.
Hardung, V. Die nichtstationare Stromung undehnbarer
Rohrleitungen. Proc. II Intern. Congr. on Angiology, edited by
Lazst, Meier, and Miiller. Fribourg, Switzerland: Univ.
Fribourg Press, 1956, p. 384.
Kudravcev, V., and R. L. Bowman. Utilization of nuclear
magnetic resonance for flow rate measurement. Proc. 13th
Ann. Conf. on Electromedical Techniques in Med. and Biol.,
Washington, D.C., i960, p. 21.
Marston, E. L., C. A. Barefoot, and M. P. Spencer.
Non-cannulating measurement of coronary blood flow.
Surg. Eorum 1 o : 636, 1 960.
Maxson Instruments. Rev. Set. Instr. 27: 116, 1956.
McDonald, D. A. Blood Flow in Arteries. Baltimore:
Williams & VVilkins, i960.
Menno, A D., and VV. G Schenk, Jr. Dynamics of coro-
nary arterial flow: flow alterations resulting from certain
surgical procedures and drugs of surgical importance
Surgery 50: 82 ; 196 1.
Mixter, G., Jr. Respiratory augmentation of inferior
caval flow demonstrated by low-resistance phasic flow-
meter. Am. J. Physiol. 172: 446, 1953.
Muller, O., and J. Shillingford. The blood flow in the
right atrium and superior vena cava in tricuspid incompe-
tence. Brit. Heart J. 17: 163, 1955.
Okino, H., and M. P. Spencer. Analysis of the dynamic
pressure-flow relationship in the renal artery. Federation
Proc. 20 (No. 1): 109, 1961.
Okino, H., K. Fujisaku, D. Sakaguchi, and H. Sasamoto.
Pulsatile blood flow in the arterial system. Respiration &
Circulation 8 : 4g, 1 g6o.
Olson, H. F. Dynamical Analogies (2nd ed.). New York:
Van Nostrand, 1958.
Patel, D. J., D. P. Schilder, and A.J. Mallos. Mechani-
cal properties and dimensions of the major pulmonarv
arteries. J. Appl. Physiol. 15: 92, i960.
Paynter, H. M. Hydraulics by analog: An electronic
model of a pumping plant. J. Boston Soc. Civil Eng., July
'959-
Peterson, L. H. The dynamics of pulsatile blood flow.
Circulation Research 2: 127, 1954.
Pieper, H. P. Registration of phasic changes of blood
flow by means of a catheter-type flowmeter. Rev. Sci. Instr.
29:965. <958-
Pritchard, W. H., D. E. Gregg, R. E. Shipley, and A. S.
Weisberger. A study of flow and pattern responses in
peripheral arteries to the injection of vasomotor drugs. Am.
J.Physiol. 138:731, 1943.
Richards, A. M., and F. W. Kuether. A new velocity
probe for sensing pulsatile blood flow. IRE Trans, on Med.
Electronics ME-6: 286, igsg.
Robiscek, F. Orifice-plate flowmeter for extracorporeal
circuit. IRE Trans, on Med. Electronics ME-6: 249, ig5g.
Sarnoff, S. J., and E. Berglund. The Potter electroturbi-
nometer; an instrument for recording total systemic blood
flow in the dog. Circulation Research 1 : 331, 1953.
Sarnoff, S. J., and E. Berglund. The Potter electroturbi-
nometer; an instrument for recording total systemic blood
864
HANDBOOK OF PHVSIOLOGV
CIRCULATION II
flow in the dog. IRE Trans, on Med. Electronics ME-6: 270,
■959-
42. Sasamoto, H., H. Okino, K. Fujisaku, and D. Sakaguchi.
The blood flow in the arterial system; asynchronism of the
electrical and mechanical phenomenon of the heart.
Thoracic Surg. 13: 230, i960.
43. Schenk, W. G., Jr., A. D. Menno, and J. W. Martin.
Hemodynamics of experimental coarctation of the aorta.
Ann. 'Surg. 153: 163, i960.
44. Shipley, R. E., D. E. Gregg, and E. F. Schroeder. An
experimental study of flow patterns in various peripheral
arteries. Am. J. Physiol. 138: 718, 1 943.
45. Spencer, M. P., and A. B. Denison, Jr. The square-wave
electromagnetic flowmeter; theory of operation and design
of magnetic probes for clinical and experimental applica-
tion. IRE Trans, on Med. Electronics ME-6: 220, 1959.
46. Spencer, M. P., and A. B. Denison, Jr. Square-wave
electromagnetic flowmeter for surgical and experimental
application. Methods 111 Medical Research, edited byBruner.
Chicago: Yr. Bk. Pub., 1960, vol. 8, p. 321.
47. Spencer, M. P., and F. C. Greiss. Dynamics of ventricular
ejection. Circulation Research 10: 274, 1962.
48. Spencer, M. P., F. R. Johnston, and A. B. Denison, Jr.
Dynamics of the normal aorta: "Inertiance" and "Compli-
ance" of the arterial system which transforms the cardiac
ejection pulse. Circulation Research 6: 491, 1958.
49. Spencer, M. P., F. R. Johnston, and J. H. Meredith.
The origin and interpretation of murmurs in coarctation
of the aorta. Am. Heart J. 56: 722, 1958.
50. Stacy, R. W. Computers: Analog. In. Medical Physics,
edited by Glasser. Chicago: Yr. Bk. Pub. i960, vol. Ill, p.
"93-
51. Taylor, M. G. An experimental determination of the
propagation of fluid oscillations in a tube with a visco-
elastic wall; together with an analysis of the characteristics
required in an electrical analog. P/iys. Med. Biol. 4: 63, 1960.
52. The Radio Amateur's Handbook (35th ed.). West Hartford
The American Radio Relay League, 1958, p. 335.
53. Thornton, VV., and B. Bejack. Performance and applica-
tion of a commercial blood flowmeter. IRE Trans, on Med.
Electronics ME-6: 237, 1959.
54. Usher, T., Jr. Dynamics of lumped-parameter mechanical
systems, I In: Vibration Topics. Hamden, Conn.: Unholtz
Dickie, i960, vol. 1.
55. Van der Tweel, L. H. Some physical aspects of blood
pressure pulse wave, and blood pressure measurements
Am. Heart J. 53:4, 1957.
56. Warner, H. R. A study of the mechanism of pressure wave
distortion by arterial walls using an electrical analog.
Circulation Research 5: 79, 1957.
57. Westersten, A., G. Herrold, E. Abbott, and N. S.
Assali. Gated sinewave electromagnetic flowmeter. IRE
Trans, on Med. Electronics ME-6 : 2 1 3, 1 959.
58. Wetterer, E. Flow and pressure in the arterial system,
their hemodynamic relationship and the principles of their
measurement. Minn. Med. 37: 77, 1954.
59. Wilhelmj, C. M., E. B. Waldmann, T. F. McGuire, and
J. McDonough. Emotional blood pressure responses of
trained normal dogs. Federation Proc. 2: 173, 1952.
60. Wolff, J. Electrical analogues of mechanical systems.
Electronic Equipment Engineering 8: 75, iq6o.
61. Womersley, J. R. Method for the calculation of velocity,
rate of flow and viscous drag in arteries when the pressure
gradient is known. J. Physiol., London 127: 553, 1955.
62. Womersley, J. R. Oscillatory motion of a viscous liquid in
a thin-walled elastic tube — I : The linear approximation for
long waves. Phil. Mag. 46: 199, 1955.
63. Yanof, H. M., and P. Salz. A new trapezoidal-wave elec-
tromagnetic blood flowmeter. Abstr. of jth Ann. Meeting
Biophysical Soc, St. Louis. 1 961, No. FE-5.
CHAPTER 26
The anatomy and physiology of the vascular wall
HERMANN BADER Physiologisches Inslilut, Universitiit Wurzburg, Wiirzburg, Germany
CHAPTER CONTENTS
Elements of the Vascular Wall
Endothelium
Collagen Tissue
Ground Substance
Elastic Tissue
Smooth Muscle
Different Types of Vessels
Arteries of the Elastic Type
Vessels of the Muscular Type
Capillaries and Arteriovenous Anastomoses
Veins
Nutrition of the Vascular Wall
Diffusion from the Inside
Vasa Vasorum
blood vessels serve as a conducting system for the
blood. They carry the blood, forced by the heart,
throughout the whole body and back again to the
heart. To make this possible there must be a pressure
gradient with its highest values in the aorta and its
lowest values in the large veins. The circulation there-
fore withstands a much higher pressure on the arterial
side than on the venous side, a difference which is
reflected in the architecture of the wall.
The stress on the vessel wall is, according to Frank
(30), proportional to the blood pressure and the ratio
of radius to wall thickness
/vaaaMa
Hut
~ ' ~" — '
-=.
Fib.
Trul
Capillary
*M
Van
20
e
lit
M
U
VAAA
EG
Varta Cava
End. i
Ela.
Mat.
Flfc.
XOOji
>4S*
fig. I . Variety of admixture of the four tissues in the wall
of different blood vessels. The figures under the name of the
vessel are the diameter of the lumen and below it the thickness
of the wall. [Burton (20).]
rest. This is best done, according to the law of Poi-
seuille, by changing the radius of the small supplying
vessel by relaxation of the smooth muscles in the
arterial wall.
Any living organ must be nourished. All but the
smallest blood vessels have their own circulatory-
system, the vasa vasorum, which supplies blood to the
vessels from the outside. In addition, simple diffusion
from the inside transports nutrients and oxygen to the
inner avascular layer of the blood vessel. This outward
fluid shift may be aided by the radial pressure gradient
through the vessel wall.
The purpose of this chapter is to discuss and to
interpret these qualities and functions of the vascular
wall and to explain the performance of the various
wall elements in the different types of vessels.
ELEMENTS OF THE VASCULAR WALL
The terms used in this article are, for the most
part, those denned by Landowne & Stacy (50). Here
we will consider some of these terms in detail. Collagen
tissue, elastic tissue, and smooth muscle have three
qualitites in common, which appear differently. These
qualities are elasticity, visco-elasticity, and plasticity.
Elasticity is that property of a material which
determines its tendency, when stressed, to return to
its unstressed geometrical configuration without loss
of energy. If a material is completely elastic, all energy
applied to it by an external straining force can be
recovered as mechanical energy. Figure 2a shows an
extension release cycle of such a perfectly elastic
material, illustrated by a spring. Any given length
has its particular tension. The extension curve and
release curve are the same. It can be linear, as in
figure 2(7, or convex or concave to the abscissa, de-
pending on the material stretched. Tension-length
diagrams of organic materials usually show a curve
which is convex to the abscissa. A perfectly elastic
c
o
l/l
c
length
fig. 2. Behavior with stretch of different materials. Tension
and length are taken as arbitrary, a: Elastic material, demon-
strated by a spring. Extension and release curve are the same.
b: Visco-elastic material demonstrated by a spring, which has
a brake disc on the top and which moves in a liquid. Extension
and release curves inscribe a hysteresis loop, the size of which
depends on the velocity of the extension and release. Outer
curve: fast stretch; inner curve: slow stretch. An infinitely slow
stretch gives the same curve as a. c: Viscous or plastic material
demonstrated with a brake disc, which moves in a liquid. The
material keeps every length to which it was brought by an
external force. The force depends on the velocity of the exten-
sion. Upper curve: fast stretch; lower curve: slow stretch. In each
the rate of viscous deformation is constant.
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
867
material can maintain a constant tension at any given
length for an indefinite time.
Thus, the term visco-elasticity applies to materials
having the combined properties of elasticity and vis-
cosity, the elastic action being damped by a viscous
one. Such a system is most easily illustrated by a spring
which has a brake disc at the top and moves in a
viscous fluid. The tension of such a system depends
not only on the length but also on the velocity with
which it is extended. The tension will be higher the
faster it is stretched, and also lower, the faster the
stretch is released. When the extension-release cycle
of such a system is plotted with tension on the ordinate
and length on the abscissa, the graph forms what is
called a ''hysteresis loop.'' Two such cycles are shown
in figure 2b. The large loop results from a quick
stretch cycle, with immediate release, the small loop
from a slow stretch cycle. The area between the exten-
sion curve and the abscissa is always larger than the
area below the release curve. This behavior indicates
loss of energy increasing with the velocity of the
stretch. More energy is required to stretch such a
visco-elastic material than can be recovered during
release. The area within the hysteresis loop can be
expressed as percentage of the area under the exten-
sion curve. It depends only on the velocity with which
the system is stretched. Rapid cyclic stretches are
called "dynamic stretches," and the shape of the
extension-release curves depends on the frequency of
the cycles. The hysteresis loop of a pure visco-elastic
element will be larger in area, the more frequent the
cycles. The hysteresis should vanish if the stretch is
made slowly enough, and this is called "static stretch."
If the stretched material is purely visco-elastic, it
returns, after an extension-release cycle, to its original
length. But if it is kept at a constant stretched length,
the tension will decrease with time in a hyperbolic
manner until it reaches an equilibrium. This process
is similar to that shown in the two curves in figure 5.
A material is called plastic when it shows the tend-
ency to retain its new shape after deformation. Plas-
ticity is usually understood as the quality of a material
which allows it to withstand stresses of less than a
critical or yield magnitude without suffering a per-
manent set, but which will then allow a viscous
deformation with stress above this yield value. The
appearance of plastic yield is not time-dependent.
Viscous or plastic behavior is illustrated in figure
ic by a disk which is moved in a viscous fluid. The
force required depends upon the velocity with which
the disc is moved. The top curve in figure 2c is derived
by a quick movement of the disc, the bottom curve by
a slow movement. As long as the velocity remains
constant, the stress will be constant too. If the applied
force is removed, the stress decreases without reducing
the length. In contrast to the behavior of a visco-
elastic element, a viscous or plastic element will never
go back to its original shape by itself.
The systems shown in figure 2a, b, and c are very
much simplified models to describe the physical defini-
tions of elastic, visco-elastic, and plastic properties.
These properties reflect, in organic materials, their
complicated molecular structure. In organic materials
there is usually a combination of the three qualities
described, with elastic, visco-elastic, and plastic prop-
erties behaving as though arranged in series, and
present in different amount. Such a combination is
described by the term "elastic incompleteness."
For instance, if an elastic and a visco-elastic element
are in series, then the element which offers the smaller
resistance to extension will dominate the stretch
behavior. Since the resistance of the visco-elastic
element is greater at high rates of stretch, the prop-
erties of this series combination is determined more
by the elastic element. The more frequent are the
stretch cycles, the less is the hysteresis. If there is also
a purely viscous or a plastic element in series, after
every stretch cycle the material assumes a greater
length. There is also the possibility that many visco-
elastic and viscous elements may be in series, each
having a different rate- and time-dependency. Such
combinations of elastic, visco-elastic, and plastic
elements can show a very slight rate-dependency, if,
for instance, the elastic element offers the smallest
resistance to stretch when compared with the other
elements present. Because of the visco-elastic or plastic
units, the system may show a great time-dependency,
which occurs as an elastic aftereffect or a relaxation
when the material is kept on a constant stress or
length.
The viscosity of organic materials may not only
derive from a viscous flow within the tissue, but also
from an architectural reanangement involving the
uncoiling or slippage of twisted elements. Such proc-
esses may be involved in the phenomenon of the
"stable loop" seen in elastic arteries after a number of
stretches, which does not show any rate or time-
dependency but does depend on the existing tension
level [Remington (73); see also Chapter 24]. The so-
called viscosity or plasticity of organic materials may
be complicated, and thus not follow the physical
definitions. Further, there is usually a certain polarity
to these tissues in that tension-length relations are
different in various directions. Most organic materials
868
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
cannot be extended ad infinitum, but tear at a certain
length.
Smooth muscle is a special case. It can be elongated
like a purely plastic material and can behave at any
given length like an elastic or visco-elastic material.
But it can also recover its original length by contrac-
tion. This means that the plastic property of smooth
muscle can be neutralized or hindered, leaving only
the elastic or visco-elastic elements under stress, as a
result of the action of the contractile element. [For
further details see Reichel (71)-]
Endothelium
The circulatory system is lined almost completely
by a single layer of very thin polygonal-shaped cells,
the endothelium. This forms a tight, smooth surface
on the inside of the vascular wall and serves as a
semipermeable membrane for the interchange of
materials between blood and tissues. It has a high
distensibility. Its ability for regeneration is very good.
For instance, 3 weeks after implantation aortic grafts
show a smooth continuous lining of endothelial cells,
which presumably are built from fibrocytes [Petry &
Heberer (67)]. A detailed discussion of the qualities
of the endothelium is given in Chapter 29.
Collagen Tissue
Collagen tissue is produced by fibroblasts, which are
located in all connective tissues. The probable pre-
cursors of collagen fibers are the reticular fibers. These
are argyrophilic fibers which are found especially in
places where collagen fibers are forming, as around
aortic grafts. They both show a banded appearance
under the electron microscope [see Wassermann (95)].
The collagen fibers consist of a network of long
protein chains which are linked by H bonds and ionic
bonds. This network is filled with an amorphous
substance (mucopolysaccharide). Smaller fibers are
glued together to larger fibers by a cement substance
which is continuous with the ground substance.
This structure gives the collagen fibers a very high
elastic modulus and also makes them very flexible [see
Harkness (36)]. Collagen fibers are nearly 25 times as
strong as elastic tissue but 15 times less extensible
(table 1). Collagen fibers are found in all vessels,
spread over the whole wall. They appear in the un-
stretched vessel wall as wavy bundles, but some of
them become straightened if the pressure within the
vessel is raised above the mean blood pressure [Reuter-
wall (74)]. This anatomical design, together with the
table 1. Elastic, Visco-E/astic and Plastic Behavior
of Collagen and Elastic Tissue (102)
Maximal Ten-
sile Strength,
kg/cm!
Maximal Ex-
tension, in ','
Irrevers-
ible
Elonga-
tion in %
of Total
Elonga-
tion
Hysteresis
IN ' J Ml
Area
Under
Stretch
Curve
Collagen tissue
660
10
67
57
(tendon)
(250-750)
(5-12)
Elastic tissue
25
150
'9
60
(ligamentun
(■2-45)
(I2O-220)
nuchae)
high elastic modulus, enables them to form a "jacket"
[Burton (20)], which is put in action only when the
fibers are straightened, by increased intraluminal
pressure. Thus the collagen fibers are not strained by
the normal blood pressure; they serve as a safety factor
for the vessels and keep them from ''blowing out" at
high pressure.
Only the collagen fibers increase in number with
aging, replacing frayed elastic fibers and degenerated
smooth muscle cells [Meyer (58), Kobayashi (46)].
Since the elastic and muscular tissues originally sup-
ported the wall tension at normal blood pressures, the
replacing collagen fibers must take over this task
[Bader & Kapal (7)] with the result that the wall
becomes less distensible. This is compensated until
the sixth decade of life by enlargement of the diame-
ters of the vessels involved [Simon & Meyer (86)].
However, collagen tissue, when overloaded, does show
elastic incompleteness, which means that it is to some
extent a plastic material (table 1). The fibers do not
return to their original length immediately after exten-
sion and the residual elongation can be 67 per cent
of the total elongation. This plasticity may produce a
large hysteresis. It may be that such plastic property
can account for the increase in vessel volume seen in
aging.
Ground Substance
The ground substance has the properties of a col-
loid— it is water-insoluble, but can bind water. It
consists of the mucopolysaccharides: hyaluronic acid
and chondroitin sulfate. It is likely that chondroitin
sulfate forms the cement substance which binds
collagen fibers together, and hyaluronic acid serves
as a lubricating material [see Harkness (36)]. Such
a lubricating substance is necessary in the vessel wall,
since the fibrous elements of the wall (collagen fibers,
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
869
elastic fibers, and smooth muscles) must be able to
slide past each other with minimal friction during the
pulsatile expansion of the vessel. The ground substance
is a very viscous material and it probably contributes
to the typical visco-elastic behavior of distensible
vessels.
Elastic Tissue
Elastic tissue is a rubberlike material with high
extensibility. It contains the protein elastin without
any detectable amount of carbohydrate [Lansing
(52)]. In contrast to collagen tissue, it is an extremely
insoluble material, and is not influenced by boiling or
autoclaving. X-ray diffraction and electron micros-
copy do not ordinarily show any internal organization
in elastic fibers. It is therefore assumed that the protein
fibrils lie without orientation within the fibers [see
Lansing (52)]. This disorder gives the elastic tissue its
high extensibility. It can be extended to twice its
original length, but its tensile strength is ^o to H$0
that of collagen tissue (table 1). This explains why it
must be protected from excessive elongation and
tearing in the vascular wall by the much stronger
collagen fibers.
Elastic tissue forms fenestrated membranes which
lie one over the other in elastic vessels. These mem-
bra nes serve as footholds for the tension muscles
(fig. 3). There is less elastic tissue in the more periph-
eral muscular vessels. It is only a very minor com-
ponent in the arterioles and precapillary sphincters
(fig. 1). Elastic fibers appear in the veins, increasing
in amount as they near the heart. They are partly-
straight and partly wavy in unstretched vessels. The
wavy ones become straight before the collagen fibers
straighten out as the pressure rises [Reuterwall (74)].
At ordinary pressures the elastic tissue supports most
of the tension in elastic vessels, whereas this task is
performed by smooth muscles in the muscular vessels.
Elastic tissues usually fray with age. This is a normal
change which appears in all vessels of old people.
Calcification of the fibers is also progressive with age.
In addition, the fibers undergo fragmentation, which
finally leaves little more than dispersed granular
material [see Lansing (52)]. Calcification is especially
great in arteriosclerosis. However, Lansing (52) has
shown that the percentage of elastin in the vessel wall
does not decrease with age, while the calcium content
rises in the human aorta from 0.4 per cent in the
second decade to 7 per cent in the eighth decade.
Frayed and fragmented elastic fibers remaining can-
not support the wall tension at normal pressures. This
®
M
1 -I
1
fig. 3. Axillary artery (human). Irradiation of tension
muscles in the elastica externa — a: in situ; b: elastica externa
artificially lifted off. Muscle endings fasten on the elastic
membrane. [Benninghoff (10).]
task is taken over in old age by collagen fibers, which
are under stress with ordinary blood pressure [Bader
& Kapal (7)]. Thus, the distensibility of the vessels
decreases, but the volume of the aorta increases and
its total elastic uptake may remain within normal
limits so long as the increase in volume keeps pace with
the decrease in distensibility [Kapal & Bader (44),
Simon & Meyer (86)].
Smooth Muscle
The smooth muscle cell is an elongated spindle with
a single elongated nucleus in the thickest part of the
cell. The cells vary very much in size. In the vascular
wall they are between 20 and 50 n in length, with their
greatest diameter between 5 and 10 n. There are two
types of smooth muscles in the vascular wall: "Spann-
muskeln'' (tension muscles), which are described in
detail by Benninghoff (10, 11), and ring muscles.
The tension muscles are connected to elastic fibers
and membranes, using them as tendons (fig. 3). They
can thus raise the tension on the elastic tissue in the
vessel wall by contraction (fig. 7) and so affect the
blood pressure (see Arteries of the Elastic Type,
870
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
below). The smooth muscles of the aorta and the
pulmonalis are almost exclusively tension muscles.
But the proportion of tension muscles diminishes
toward the periphery. The smallest arteries and
arterioles have almost no tension muscles. They may
appear again on the venous side of the circulatory
system, but not in as large numbers as in the elastic
arteries [Grau (34)].
The ring muscles are connected with each other.
How they are related to the elastic and collagen tissue
is not certain, but it is very likely that they have slack
connections with both tissues. Since smooth muscles
are almost completely surrounded by reticular fibers,
it is possible that these fibers bind them together. The
ring muscles form the greatest part of the wall in the
muscular vessels, where they form a helical arrange-
ment [Fischer (23), Schultze-Jena (84)]. Arterioles
and precapillary sphincters consist mostly of ring
muscles.
Smooth muscles have the general quality of sponta-
neous activity and self-conduction [see Biilbring (17)].
Bozler (16) has concluded from this behavior that
they form a syncytium, which would mean that the
individual muscle cells are interconnected by proto-
plasmic bridges. But in reality they form a network
in which every muscle fibril is surrounded by its own
membrane, so that there are double membranes at
places where the cells are in contact with each other
[see Prosser (68)]. This network acts like a functional
"syncytium," since an excitation can be conducted
over the double membranes. This double membrane
has a high resistance, and therefore the conduction in
smooth muscles is much slower than that in nerve
fibers. The conduction can be propagated over the
whole organ, as in the uterus or the ureter (single-unit
smooth muscles), or it can be limited to a certain area,
as in the intestine [Bozler (16), Greven (35), Biilbring
(17)]. The limitation results from the presence of a
higher resistance of the double membranes at certain
places. The resistance can be changed so that the area
which responds to a stimulus can be increased or
decreased. Another characteristic of the syncytium
is its response to stretch [Biilbring (17)]. If smooth
muscles of the intestine are stretched, the membrane
depolarizes and spikes are produced (fig. 4). Contrac-
tion occurs and tension rises in direct proportion to the
increase in spike frequency. But in addition to having
independent conductivity and excitability, smooth
muscles also receive innervation from both the sympa-
thetic and parasympathetic nerve system. There are
ganglion cells around the adventitia [Leontowitsch
(55)] and nerve fibers in the media [Boeke (15)] of the
56-
mv
54
52-
A
/
vWv
48-
46-
r
Resting Potential
A /• *
. , <} tx a "00 d-o
' d
, a 1 o-cr
; "3 ; Spike Frequency
7-5-
g
\ 6-6
1 \ / Tension
7-0
6-5 +
1 1
M IN
15
o
uu
1/1
IO p
—
in
c 111
O 2 4 6 8
fig. 4. Graph showing correlation between membrane
potentials, spike frequency, and tension during spontaneous
pendular rhythm recorded for 10 min. [Biilbring (17).]
blood vessels [see also Staubesand (87)]. The auto-
nomic nervous system can change the spontaneous
activity of the smooth muscles by changing their
membrane potentials.
In addition to the syncytium-like smooth muscles,
there are also multiple-unit smooth muscles which are
neither self-conducting nor spontaneously active. They
receive extensive innervation and appear to be or-
ganized in some tvpe of motor-unit plan [see Prosser
(68)].
Little work has been done on the problem of the
excitation and conduction of vascular smooth muscle.
Therefore it is hard to say whether it represents a
multiple-unit system or a syncytium. Monnier (62)
has shown that the conduction of excitation in the
mesenteric artery of cattle (an artery of the muscular
type) is very slow (only about 2 mm/sec). This is
much slower than the conduction of any known nerve,
and in a range similar to other syncytium-like smooth
muscles [see Biilbring (17)]. The mesenteric artery
also responds to stretch with a contraction. It may be
assumed therefore that the muscles in the peripheral
arteries behave as a syncytium. This is likely, in view
of the relationship of blood pressure to flow. For in-
stance, Thurau & Kramer (91) have shown that the
flow in the kidney becomes constant if the pressure is
raised above 90 mm Hg. This special flow-pressure
relationship is due to an increase of resistance, effected
by contraction of the smooth muscles in the pre-
glomerular arteries. Similar behavior in the arteries of
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
87 ■
the extremities is reported by Folkow (26). This idea
has a long history, beginning with the contribution of
Bayliss (9) in 1902.
It may be assumed that the contraction of smooth
muscle is caused by the increase of tension in the
vascular wall as a result of the rising blood pressure.
There is no adaptation to the tension stimulus, which
agrees with the results of Bulbring (17) on the in-
testinal smooth muscle. Folkow (28) suggests that the
tonus of the resistance vessels is maintained by myo-
genic activity of their smooth muscles, which are
excited by the tension of the wall (similar to the case
for intestinal smooth muscles shown in fig. 4). How-
ever, this autoactivity, in both types of smooth
muscle, is controlled by the autonomic nervous system.
Since the smooth muscles in the peripheral arteries are
mostly ring muscles, it may be that the ring muscles
can behave like a syncytium. The tension muscles are
always interrupted by elastic fibers, and Prosser et al.
(69) have found a much larger intercellular distance
between the individual muscle cells in vessels of the
elastic type than in other organs (1000 nm in the pig
carotid artery as against 120 nm in the cat intestine).
It is therefore very likely that they form a multiunit
system in which the muscle cells receive extensive
sympathetic and parasympathetic innervation. This
impression is confirmed by Burnstock & Prosser (18),
who got no response to stretch from the carotid artery,
a vessel of the elastic type, or from the renal vein.
Over and over the idea appears in the literature that
arteries may contract and relax as quickly as the
heart and so force the blood to the periphery just as
the intestine propels a bolus to the colon. One of its
newer proponents, Dickinson (22), shows a contrac-
tion curve of a sheep's hepatic artery which develops
its peak tension in about 3 sec after an unphysiological
stimulus of 120 v. The slowness of contraction and the
long latency speak against the possiblity of the propul-
sion of blood by arterial contraction. This latter
attitude is shared by Fleisch (25) and Wetterer &
Kapal (99).
If smooth muscles are extended slowly they behave
like a plastic material. They can maintain a given
length, either short or long, for protracted periods
with very low metabolism. However, this length
maintenance does depend upon repeated stimulations
of constant magnitude. If the stimulation is increased,
these muscles respond by contracting, regardless of
their initial length (except, of course, if already maxi-
mally contracted). From this it follows that there must
be some mechanism which enables smooth muscle to
shift its behavior from plastic when "set" in length to
visco-elastic, when contracting. Uxkiill (93) has postu-
lated for this a "Sperrung" (catch mechanism),
signifying that the protein filaments within the muscle
fibers "catch" at a certain length so that they cannot
slip apart when tension is applied.
Three possible explanations have been offered for
this behavior. The first, suggested by Reichel, is that
the smooth muscle consists of two elements in series,
an elastic element and a contractile element, where
the contractile element can behave with either
plasticity or contractility (70). If this is true, the
"catch mechanism" could be described as a trans-
formation of plasticity to contractility, where the
element is "caught" at any length and thus is able to
keep a given tension with a low metabolism or to
contract. An alternative to this theory, suggested by
Lowy & Hanson (56a), is called the sliding filament
mechanism. They assume that thin discontinuous
actin-containing filaments move relative to thick
discontinuous paramyosin-containing filaments, as in
striated muscles. Linkages are presumably formed
during contraction between both filaments all of one
type, with one rate of formation. The rate of breaking
can vary from slow (tonic contraction, visco-elastic)
to fast (phasic contraction, plastic) depending on the
concentration of a relaxant present (i.e., 5-hydrox-
ytryptamine). Repeated excitatory stimulation could
maintain these linkages, whereas stimulation of
inhibitory nerves could increase the rate at which they
break [Lowy & Millman (57)]. A second possibility
is that the plastic and the contractile elements are in
parallel, with an elastic element in series. In such an
arrangement the catch mechanism could be in the
plastic element, whereas the contractile element
could cancel any plastic deformation by contraction.
Such a parallel arrangement is postulated by Johnson
(41a). He assumes that the contractile system is
formed by the actomyosin, and that paramyosin is
situated parallel to it as the plastic element. Laszt
(54) assumes a similar mechanism in the vascular
smoothVmuscle. A third possibility is that the plastic
and contractile elements are in series. In such an
arrangement the contractile element could work
only if the catch mechanism were put in action.
But it would then be necessary to have a special
mechanism to cancel the plastic deformation, such as
the presence of both fast and slow contractile elements
within the smooth muscle, the slow elements being
virtually "plastic."
Whether any of these three mechanisms may be
the real one is not clear. It is possible, too, that one
smooth muscle may work by one mechanism and
872
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
FIG. 5. Average stress-relaxation
curves of carotid and umbilical arteries.
Vertical coordinate is given in per-
centage of total pressure rise, following
injection. Upper curve, common carotid
artery of the dog. Lower curve: umbilical
artery of the human. [Zatzman el al.
(103).]
OO -
80
6O-
40-x
20-
Secorid t
others by another, since, for example, the uterus and
the bladder are very different in properties and action
[Bader (3)].
If a smooth muscle is stretched quickly to a certain
length, it will show a tension increase. If this length
is held for a longer time, the tension will decrease, at
first quickly, later more slowly. This typical stress
relaxation is a result of the visco-elasticity of the
smooth muscle, which may be due to breaking of
the linkages within the myofilaments. A typical
stress-relaxation curve of smooth muscles is like
that of the lower curve of figure 5. For further
details of the mechanical properties of muscles, see
Reichel (71).
If these mechanical properties of smooth muscles
are to be compared with those of the vascular wall,
one must keep in mind the modifying effects of col-
lagen and elastic tissue [Remington (73)]. Another
point is that most of the experiments with smooth
muscles are made on organs other than blood vessels,
in vitro, and without innervation. Smooth muscles
in vivo are under a continuous stimulation, and they
are also under constant contraction and tension in the
vascular wall. It is therefore very likely that the
smooth muscles of the vessel wall in vivo would show
different visco-elastic and plastic behavior from those
found during in vitro experiments.
Zatzman et al. (103) have shown that there is a
great difference in the stress-relaxation behavior of
the elastic carotid artery and the muscular umbilical
artery. After 10 sec of stress, relaxation of the carotid
artery amounts to about 20 per cent of the original
tension, whereas in the umbilical artery it is about 95
per cent (fig. 5). This indicates that in the umbilical
artery the tension is applied mostly to smooth muscle
with its large visco-elasticity and plasticity. It is not
easy to say to what degree smooth muscle is respon-
sible for the relaxation of the carotid artery, since
elastic and collagen tissues may each have both a
visco-elastic component (hysteresis loop during stretch
cycle) and a plastic component [irreversible elonga-
tion (table 1)]. This passive behavior is responsible
in part for the stress relaxation of elastic arteries
[see Kapal (42)].
Smooth muscle tissue degenerates and decreases in
amount with age, and is replaced by collagen fibers
[Meyer (58), Kobayashi (46)]. This tends to render
the arterial tree more rigid and to explain the systolic
hypertension of old age. The high systolic pressure in
this condition puts an extra burden on the heart.
[See Bader & Kapal (5) and the paragraphs on elastic
arteries below.]
Smooth muscles have the ability to regenerate.
For example, Petry & Heberer (67) have described
cell formations which are found on the inside of aortic
grafts some weeks after implantation. These cells seem
to be muscle cells, and are assumed to originate from
fibroblasts.
DIFFERENT TYPES OF VESSELS
Vessels differ in their architectural structure and
their behavior according to their varied tasks. There
are, in general, four different vessel types: on the
arterial side are elastic arteries and muscular arteries,
but it is hard to say where the one ends and the other
begins, since the structural changes are gradual.
Usually the aorta, the pulmonary artery, the common
carotid artery, the subclavian artery, and the common
iliac artery are regarded as elastic arteries. Arteries
more peripheral than the above, down to the ar-
terioles, are classed as muscular arteries. After these
are the capillaries, which consist mostly of endothe-
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
873
lium. Then there are veins, which are built to some
extent like the arteries. The most striking difference
between them is their mounting, for most arteries
have a slack connection with the surrounding tissues,
whereas the veins are more intimately bound up with
the surrounding tissues to make a functional svstem.
Arteries of the Elastic Type
The wall of the elastic arteries is characterized by
a high percentage of elastic tissue (fig. 1), which may
be 40 per cent of the wall in the thoracic aorta, but
decreases toward the periphery. The elastic tissue
is mostly present as fenestrated membranes — up to
50 membranes located one upon the other. There
are also star-shaped membranes in the wall of the
pulmonary artery [Meyer (59)]. There is a network of
elastic fibers between all these membranes. The
membranes are connected by smooth muscles, the
tension muscles, described by Benninghoff (10, 11).
These tension muscles use the elastic membranes as
footholds (fig. 3). There are no ring muscles in the
thoracic aorta, but they appear in increasing numbers
in the more peripheral arteries. Where the ring mus-
cles exceed the tension muscles in amount the arteries
are called muscular arteries. The collagen fibers are
distributed over the entire wall. They lie there in
wavy bundles, which become straight if the blood
pressure rises over the normal mean value [Reuter-
wall (74)].
The large amount of elastic tissue and looseness of
the collagen fibers give elastic arteries high disten-
sibility. For instance, the aorta can be distended to a
threefold increase in contained volume over that at
300
ZOO
fig. 6. Pressure-volume diagram of the thoracic aorta of
the pig. a: Extension curve made shortly after sacrificing.
The aorta was stimulated with epinephrine, b: Extension
curve made 8 days later. Smooth muscles were dead. [Bader &
Kapal (5).]
zero pressure. This high distensibility enables an
elastic vessel to act as would an air chamber (Wind-
kessel). The aorta contributes over 50 per cent to the
total vascular air chamber action [Wetterer (98)]. If
an elastic artery is stretched, it shows a typical S-
shaped pressure-volume diagram, like that in figure
6 where "static" stretch curves are given for the
thoracic aorta of a pig. [See Chapters 7 and 24 for
the explanation of the typical S-shaped pressure-
volume diagram of elastic arteries and its relation
to the tension-length diagram. [See also Frank (30,
31).] A similar S-shaped curve may be obtained from
a rubber tube within a nylon tube, where the nylon
tube serves as a "jacket" [Bader & Kapal (6)]. In
such a pressure-volume diagram the rubber tube is
responsible for the curve below the inflexion point
and which appears concave to the abscissa; the nylon
jacket for the convex part above the inflexion point.
This fact, together with the finding of Reuterwall (74)
that when elastic tissue becomes straight collagen
tissue is still wavy (i.e., still relatively unstretched)
and the study of Roach & Burton (75) which in-
volved differential digestion of collagen and elastic
fibers of the iliac artery, indicates that the part of the
pressure-volume diagram from zero pressure to the
inflexion point reflects the extension of elastic tissue,
whereas the part above the inflexion point is due to
the collagen tissue [see Bader & Kapal (7)].
The upper curve of figure 6 is derived from a stretch
curve made shortly after death, after the aorta was
stimulated with epinephrine; the lower curve was
made 8 days later when the smooth muscles were dead.
Schonenberger & Miiller (83) got similar results on
cow aortas with dynamic stretches. Millahn & Jaster
(61) stretched pig and cow aortas after relaxing the
smooth muscles with acetylcholine, finding that the
stretch curve lay below the curve given by the stimu-
lated vessel. The pressure-volume diagram can shift
to higher or lower pressures depending on the con-
tractile state of the smooth muscle, but the shape of
the curve never changes. This proves that smooth
muscles can increase the wall tension without chang-
ing the elastic properties of the vessel, a finding which
Benninghoff (10, 11) had proposed as a result of his
microscopic studies. Bader & Kapal (5) concluded
from their experiments that smooth muscle can be
arranged neither in series with the elastic elements
nor in parallel. Both arrangements would give, with
stimulation, not only a shift of the stretch curve to
higher pressures, but also a change of the shape of
the curve.
Since the tension muscles are attached to the elastic
874
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
©
fig. 7. Model for the arrangement of tension muscles and
elastic tissue. Stretch in vertical direction, a: Tension muscles
relaxed; b: tension muscles contracted. [Kapal & Bader (43).]
fig. 7A. Baroceptor in the adventitia of the human aortic
ar< li a: End fiber of the aortic depressor nerve; b: network;
c: end network; d: neurofibrils. Method after Biclschowsky.
X 1100. [Abraham (1).]
membranes, and the contraction of the smooth mus-
cles seems to influence only the part of the curve
which is ascribed to the extension of elastic tissue,
Kapal & Bader (43) have designed a model which is
similar to an arrangement which Burton (20) had
published in 1953 in his highly stimulating review
article. It shows the action of the tension muscles in
elastic arteries.
In figure ~ja the smooth muscles are relaxed; in
yb, thev are contracted. The smooth muscles fasten
on the elastic fibers or membranes at right angles to
the direction in which the elastic fibers are stretched.
These fibers become elongated by contraction of the
smootli muscles, but the circumference of the vessel
is not changed (fig. jb). The consequence is a rise
in the tension of the elastic fibers. Since the model will
be involved in a stretch in the direction of the elastic
fibers, the tension muscles do not need to develop
tension as great as the total wall tension, but can
increase the stress on the elastic tissue with relatively
little work.
This model has an advantage to neurophysiologists
as well as to muscle physiologists. Heymans & Delau-
nois (37) have shown that the blood pressure decreases
if the smooth muscles of the carotid sinus are stimu-
lated by noradrenalin. Heymans et al. (38) obtained
the same results by elongating the carotid sinus. They
concluded from these experiments that the pres-
soreceptors located in the carotid sinus, which cause a
decrease of the mean blood pressure after stimulation,
do not respond to the blood pressure, but rather to the
wall tension [see Heymans & van den Heuvel-Hey-
mans (39)]. It is very likely that the pressoreceptors
situated in the aorta work in the same way.
The pressoreceptors appear as a very fine network
of neurofibrils (fig. J A). They are mainly located in
the adventitia of the carotid sinus [Sunder-Plassmann
(89)] and in the adventitia and the outer part of the
media of the aortic arch [Seto (85)]. Stohr (88) has
the impression that this network of neurofibrils shown
in figure yA may be only a part of the whole
neurofibril mass of which the pressoreceptor is
constituted. He assumes that smaller fibrils exist but
are not visible because of limitations in the staining
method and in the optical properties of the light
microscope.
There are very few clues as to how the network of
the pressoreceptor is related to the surrounding tissue.
Sunder-Plassmann (89) has shown that the media of
the carotid sinus is thinner than that of the nearby
vessel, but the membrana elastica externa is thickened.
The elastica externa in the carotid sinus shows a
sharp boundary separating it from the media, but a
more gradual merging with the adventitia. In this
diffuse zone, which shows collagen fibers and large
elastic membranes, the pressoreceptors of the carotid
sinus are located, and the neurofibril networks show
a certain degree of adaptation to the shape of the
connective tissue. Abraham (1) describes the neuro-
fibril networks of the aortic arch as nestling flat
against the vessel wall. Their position follows the direc-
tion of the fibrous elements.
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
87:
It may very likely be that in both the carotid sinus
and the aortic arch, the neurofibril networks of the
pressoreceptors are in one way or another attached
to the connective tissue, especially the elastic mem-
branes or fibers. Thus the pressoreceptors are assumed
to be parallel to the elastic membranes or fibers, an
assumption which agrees with the facts now available.
They will be stimulated if the smooth muscles increase
the tension of the elastic tissue by contraction. But
now the stimulated pressoreceptors reflexly lower the
blood pressure until the tension of the elastic fibers,
and with them that of the pressoreceptors, decreases
again to the normal value (equation 1). The tension
muscles are thus able to change the blood pressure,
as shown in the work of Bader & Kapal (5). The
model of Kapal & Bader (43), and the results of
Heymans et al. (37-39) agree very well. It also fits very
well with the idea that the tension muscles are mul-
tiple-unit muscles, since they are a type of control
organ which does not depend on the wall tension.
The model may also explain the higher resistance
and blood pressure of older people in contrast to
younger people. Smooth muscles degenerate with age
and are replaced by collagen fibers [Meyer (58),
Kobayashi (46)]. This means that the smooth muscles
are no longer able to stretch the elastic tissue as much
as in younger individuals. But if the tension of the
elastic tissue is lowered by lack of smooth muscle
function, the blood pressure will rise until the tension
reaches a physiological value for the pressoreceptors.
As the muscles continue to degenerate, the pressure
needed to stimulate the receptors continues to rise,
and this may be one of the various mechanisms which
cause essential hypertension. Such a hypertension
must be called "essential," since the weakness of the
tension muscles cannot be diagnosed and there may
be no clearly diagnostic anatomical change of the
arterial wall. The only evident sign of such muscle
weakness would be the hypertension.
However, in the aging process degeneration of
smooth muscles, fraying of elastic tissue, and increase
of the collagen tissue are accompanied by a decrease
in the distensibility of the arteries. A 20-year-old
aorta can be distended to 300 per cent of its zero-
pressure volume, but a go-year-old aorta can be
distended only about 25 per cent [Simon & Meyer
(86)]. If the inflexion point of the volume pressure is
high and the curve reaches its slope of maximum
distensibility at about 100 mm Hg (the normal mean
blood pressure) (cf the 13-year-old aorta in fig. 8a),
the work required of the heart is reduced in maintain-
ing a physiological pressure level. If the inflexion
occurs at a lower pressure level, as after an increase
of collagen tissue (older aortas), the mean blood pres-
sure falls on a steeper slope of the pressure-volume
curve. Under this condition the heart would have to
work more were this disadvantage not compensated
by enlarging the volume of the aorta. Figure 8a shows
a fivefold increase in the volume at 100 mm Hg
between the 13-year-old and the 85-year-old aorta.
The volume change of the elastic chamber with each
heart beat, that is, the volume which can be injected
to give the physiological pulse pressure amplitude,
remains nearly constant until about the sixth decade
of life, as a result of the initial volume increase. This
means that the heart work need not increase with the
decrease of the wall distensibility [Simon & Meyer
(86)]. The aortas over 60 years do not show any
inflexion point; they are convex to the abscissa from
the very beginning. This signifies that collagen tissue
is already stressed near zero pressure. The pulmonary
artery shows similar behavior, but the inflexion point
occurs at a lower pressure, just as the pulmonary
pressure is lower [Meyer & Simon (60), Frasher &
Sobin (32)].
The ratio of radius to wall thickness, which is im-
portant in the relationship of the wall stress to the
blood pressure (equation 1 ), is the same at zero
pressure in aortas of different ages [Hieronymi
fig. 8. a: Pressure-volume diagrams of the thoracic aorta of
the human at different ages, b: The relationship of radius to
wall thickness of the same aortas in relation to the pressure.
[Bader (4).]
8y6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 9. Schematic presentation of the
behavior of the different tissues in the
wall of elastic-type vessels at different
degrees of extension. Description in the
text. [Bader & Kapal (7).] The stretch
is in the vertical direction. The S-
shaped line is the pressure-volume
curve of a young human aorta.
(40)], but at 100 mm Hg this ratio is age-dependent
(fig. 86). It increases from the first decade of life until
the end of the third decade and decreases from then
until the end of life. The shift of this ratio, like the
shift of the inflexion point, is caused by the increase
of the collagen tissue. The arteries become more
and more rigid with age. A rise of the blood pres-
sure in older people does not change the tension
on the elastica, and thus prevents stimulation of the
pressoreceptors. This indicates that the regulation of
the blood pressure of old people would become more
and more unstable [Bader (4)]. Since the decrease
of the ratio of radius to wall thickness in the aging
aorta simultaneously decreases the wall tension (equa-
tion 1), the tension of the pressoreceptors must also
become lower and lower. But this process may re-
sult in an increase of the mean blood pressure to
get the pressoreceptor again on the normal tension
level. A similar change has already been mentioned
with the degeneration of the tension muscles. But
since the changes in smooth muscles, and in the ratio
of radius to wall thickness, with aging, are greater in
proportion than the usual increase of the mean blood
pressure, one may assume also a change in receptor
sensitivity.
The interaction of the three wall elements in the
elastic arteries, elastic tissue, collagen tissue, and
smooth muscle, may be illustrated by the scheme of
figure 9. Sections 1 through 4 represent different
stretch phases. The stretch takes place in vertical
direction. Both halves of each phase must be regarded
together. The element a represents two elastic fibers
which are connected by smooth muscles as in figure 7.
Both elastic fibers are already straight at zero pres-
sure. The element b is an elastic fiber which is still
wavy (unstressed). Both a and b, in the upper and the
lower half, are under minimal stress. Element c is a
collagen fiber. In the upper half it is less wavy than
in the lower half. Phase 2 will be reached after the
stress has begun. Element a is already stressed, whereas
element b is just straightened. By this means the
recruitment of the elastic fibers is represented. The
stretch proceeds in phase 3, so that collagen fibers
are partly straight and included in the stress (upper
half). The collagen fiber in the lower half is still wavy.
Elastic fibers only are stressed in phases 1 and 2,
whereas elastic and collagen fibers are functional in
series in phase 3. Now the length available for further
stretching of the elastic fibers is only half as much as
in phases 1 and 2, since a further stretch of the elastic
fibers in the upper half is prevented by the collagen
fibers. This means that the increase in extension, per
unit rise in pressure, becomes less and the pressure
volume diagram, which until now was concave to the
abscissa, becomes convex. At last a point is reached
where all collagen fibers are straight: phase 4. Elastic
and collagen fibers are straight and parallel. Since
collagen tissue is much less distensible than elastic
tissue, the wall distensibility at this point depends only
on collagen tissue. This model illustrates the wall
architecture of a proximal elastic artery. The more
peripheral the vessel, the more the ring muscles
participate and the more the model of figure g will be
combined with the model of figure 13 (see below).
The effect of changes with age in the arterial wall
can be illustrated by having the stretch start in phase
2, 3, or 4. Thus unextensible elements are put in
action at lower pressures than in young arteries
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
877
j 250
200
^.'50
% 100
I
so
i
~1\
^
s£^
s
V
^
5 10 15
Vol urn (cm1} —
20 25 jo 35
®
200
150
E$ieo
6
;
/
1
/
II
II
!'/
20 (0
Vol urn (cm'' I —
60 SO WO
®
!S0
200
o>7S0
£ 50
1
1
h
'
/
-**!—-
^^"^ *"
A'
5 W 15 20 25 30 35 <0
Volum (cm>) —
®
200
150
$100
<5
50
^
^
1
1
1
y
r
20 io so bo ion
Volumlcm') •-
fig. 10. Extension-release curves of the
thoracic aorta of the human at different ages
and with different numbers of stretches. I:
14 years old — (a) 4th stretch cycle; (b) 100th
stretch cycle. II : 63 years old — (a) 3rd stretch
cycle, (4) 15th stretch cycle. [Wagner &
Kapal (94).]
[Bader & Kapal (7)]. The unextensible fibers might
be collagen fibers or calcified elastic fibers which are
under stress with small extension of the arterial wall
[Roach & Burton (76)] or even with no extension at
all.
Elastic vessels, like any tissue, show typical visco-
elastic and plastic behavior. An extension-release cycle
gives a hysteresis loop which depends in part on the
velocity with which the stretch was applied [see
Remington (72)]. There is also a shift, on repeated
stretching, of the pressure-volume diagram toward
greater volume at the initial pressure level, indicating
some plasticity. Wagner & Kapal (94) have found
with experiments on the human aorta that hysteresis
is not only dependent on the stretch velocity, but also
on the age of the vessel (fig. 10). It becomes smaller,
the older the vessel is. The same effect appears if an
aorta is stretched repeatedly. The more frequently
the artery is stretched, the smaller is the hysteresis.
The hysteresis is greater above the inflexion point of
the pressure- volume curve than below [Wagner &
Kapal (94)].
In large elastic vessels, contraction of smooth
muscles does not influence the hysteresis [Remington
(72)]. Kapal (42) has shown that the aorta responds
to dynamic stresses as would collagen and elastic
tissue, but not like smooth muscle. Therefore, it seems
that the visco-elastic and plastic behavior of elastic
arteries depends mostly on elastic tissue, collagen
tissue, and ground substance, but only to a small
degree on smooth muscles. This is confirmed by the
curves of figures 6 and the model in figure 7, where
smooth muscles do not affect the mechanical prop-
erties of the vessel (see also fig. 5). However, in the
more peripheral vessels the increasingly plentiful ring
muscles have a correspondingly greater effect on visco-
elastic behavior [see Peterson et al. (66), Bergel (12,
13)]. The greater elastic incompleteness of collagen
fibers, as compared to elastic fibers (see table 1),
agrees very well with the larger hysteresis in the upper,
collagen-dependent part of the pressure-volume
diagram. But with both collagen and elastic tissue,
the elastic incompleteness seems to diminish as more
stretch cycles are made. The similarity between
decrease of the hysteresis with age and with repeated
cycles has led to the assumption that, as a result of
their elastic incompleteness, the vessels are distended
more and more by the pulse pressure during their
life, until they reach a stable state, eliminating the
visco-elastic and plastic elements [Wagner & Kapal
(94)]-
Vessels of the Muscular Type
The more peripherally the arteries are located, the
higher is the percentage of smooth muscles in the wall
(fig. 1). In elastic arteries one cannot distinguish easily
between intima, media, and adventitia, whereas in
muscular arteries there is a clear separation of these
layers. The media consists mostly of smooth muscles,
the ring muscles. Between them are collagen and
elastic fibers. The elastic membranes, typical for the
878
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
o
o
fig. 10A : Longitudinal muscles in the intima of the branches
of the bronchia] artery. Schematic presentation of their differ-
ent arrangements. [Weibel (97).]
elastic vessels, are concentrated in the elastica interna,
which separates the intima from the media, and in
the elastica externa, which separates the media from
the adventitia. Attached to these membranes are the
tension muscles, but they account for but a small per-
centage of the total vascular smooth muscles [Ben-
ninghoff (10, 11)]. The ring muscles are arranged in
the wall in a helical structure [Schultze-Jena (84),
Fischer (23)].
Arteries which are frequently extended in the
longitudinal direction, like the branches of the bron-
chial artery of the lung, possess longitudinal muscles
in addition to the ring muscles [Weibel (96)]. These
longitudinal muscles are situated in the split mem-
brana elastica interna and can be arranged either in
fairly thick one-sided bundles, or as concentric shells
which surround the whole lumen (fig. 10A).
The mechanical behavior changes in the same way
as the anatomical picture, smooth muscle forming
the major support of muscular arteries, elastic tissue
of elastic arteries. In contrast to the elastic vessels,
the muscular arteries can change their radius over a
wide range. The smallest vessels, like the arterioles
and the precapillary sphincters, can even close their
lumens completely. Figure 1 1 shows pressure-diame-
fig. 1 1 : Pressure-diameter diagrams of a small branch of the
mesenteric artery of the horse — a: 1st stretch cycle; b: 2nd
stretch cvcle; c: 6th stretch cycle. [After Wezler & Schliiter
(100).]
fig. i 2 : Pressure-volume diagrams of the vessels of the hand :
an in vivo experiment, a- Temperature in the plethysmograph :
25.5 C (vessels contracted); b: temperature 31.5 C (vessels
normal); c: temperature 36.0 C (vessels relaxed). [After Thron
et al. (90).]
ter diagrams of a small branch of the mesenteric
artery of the horse [Wezler & Schliiter (100)]. Six
extension-release curves are made successively, the
first, second, and sixth being shown. During the first
stretch cycle the smooth muscles are assumed to be
contracted, during the following cycles they are more
and more relaxed. The first extension-release curve
shows a large hysteresis, where the extension curve is
concave to the abscissa and the release curve is con-
vex. This indicates a very large visco-elasticity of the
vessel wall, and very different behavior from that of
elastic-type vessels for which the extension and the
release curves have a similar shape. Later extension-
release cycles show smaller hysteresis, and the exten-
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
879
n
fig. 13. Schematic presentation of
the behavior of the different tissues in
the wall of muscular vessels with differ-
ent degrees of extension. Description
in the text. [After Wezler & Schliiter
(.00).]
sion curves also become convex to the abscissa.
Schliiter & Wezler (80) have described other curves
where the first extension curve had an S-shape or was
convex to the abscissa. In all these cases the shape of
the diagram seems to depend very much on the state
of contraction of the smooth muscles. The extension
diagram will be concave to the abscissa if the smooth
muscles are contracted, and the slope of the curve
will be steeper the stronger the contraction. If the
smooth muscles are less contracted, the extension
curve will show an S-shape or be concave to the
abscissa. The whole diagram moves with frequent
stretches to larger diameters or volumes, but comes
back to the original place if the smooth muscles are
stimulated. Thron et al. (90) obtained similar results
in vivo. Figure 12 shows plethysmographically ob-
tained pressure-volume diagrams of the human hand
vessels at different states of contraction of the vascular
muscles, due to different temperatures. The extension
curve of the constricted vessels (a) is nearly straight
and is followed during release by a large hysteresis.
The less constricted vessels have less hysteresis with
both distention and release convex to the abscissa.
This agrees very well with the diagrams of Wezler &
Schliiter (100), which were made in vitro (fig. 1 1).
The diagrams shown in figures 1 1 and 1 2 are in
many ways different from the diagrams in figures 6,
8, and 10, which were made from the elastic aorta.
The most striking differences are: first, the shape of
the aortic diagram remains the same whether the
smooth muscles are contracted or relaxed; second,
the hysteresis of the elastic vessels is smaller than that
of the muscular vessels. This indicates that the ar-
rangement of the different wall elements must differ in
the two types of vessels. The smooth muscles, which
play only a minor role in the elastic vessels, take a
major one in the stretch curve of muscular vessels.
Wezler & Schliiter (100) have designed a model which
may give the action of the three wall elements and in
different contracted states. This model is shown (sim-
plified) in figure 13. It is distinguished from the
model in figure 9 by the parallel arrangement of the
smooth muscles to the other elements. Sections 1
through 3 represent, as in figure 9, different stretch
phases. The stretch takes place in the vertical direc-
tion. The element a represents an elastic fiber, the
two elements b are smooth muscles, w-here the in-
dividual muscle fibers are in series, and element c is
a collagen fiber. Both the elastic and the collagen
fibers are wavy (unstressed) in phase 1. At the begin-
ning of the stretch, near zero pressure, only the smooth
muscles bear the stress. If the muscle fibers are con-
tracted the slope of the pressure-volume diagram
will be steep in the beginning and concave to the
volume abscissa. Since the contracted smooth muscles
behave in general like a visco-elastic material, the
pressure-volume diagram will show prominent hys-
teresis as described in figure o.b. If the muscle fibers
are relaxed, the slope of the pressure-volume diagram
will be flatter. As extension proceeds, phase 2 will be
reached, in which the elastic fibers are straightened.
This will be at a higher pressure if the muscles are
contracted than if they are relaxed. Finally, the
collagen fibers are involved in the stretch (phase 3).
If the smooth muscles are in strong contraction, the
whole diagram within physiological pressure limits
is concave to the volume abscissa. If the contraction
is less, the elastic and collagen fibers come into play
at lower pressures and the pressure-volume diagram
shows an inflexion point and an S-shape. The pres-
sure at which the inflexion point is located depends
on the intensity of the contraction. If the smooth
muscles are relaxed, the vessels will show in the
beginning only a plastic elongation without a rise of
pressure, but the pressure will increase when the
elastic and collagen fibers are involved in the exten-
sion. The pressure-volume diagram is, from the very
beginning, convex to the abscissa. The collagen fibers
88o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
serve as a "jacket" just as they do in the elastic vessels.
They provide a safety factor to prevent overstretching
the smooth muscles.
The model of figure i 3 gives only the arrangement
of the ring muscles in relation to the other elements.
The more centrally the arteries of the muscular type
are located, the greater is their amount of tension
muscles [Benninghoff (10, 11)]. We must then assume
a mixture of models shown in figures 9 and 13. The
change over from the pure elastic-type model (fig.
9) to the pure muscular-type model (fig. 13) is grad-
ual. The different behavior of the arteries may be
the reason why different authors have different opin-
ions about the architecture of the same wall. For
instance Burton (20) suggests an arrangement similar
to figure 9, whereas Bergel (12, 13) speaks of an
arrangement similar to figure 13. Both may be right.
The smooth muscles in the muscular arteries are,
during life, under a continuous stress since they are
in parallel. Therefore they must have a certain basic
tone to withstand the stress of the blood pressure.
There is strong evidence that this basic tone may
derive from myogenic activity. Bayliss (9) suggested,
in 1902, that the blood pressure might act as a me-
chanical stimulus to the vascular wall. Lately it has
become more and more evident that the smooth
muscle possesses the capability of spontaneous activity
(see above). For instance, denervated intestinal
smooth muscles respond to a stretch with a contraction
(fig. 4). The tonus of the vascular smooth muscles
may be assumed to depend on the tension of the wall
and consequently on the pressure within the vessel.
Folkow (26), Thurau & Kramer (91), and earlier
workers have found that the blood flow becomes
constant above a certain pressure which may mean
that the pressure or, rather, the wall tension serves
as a stimulus for contraction of the smooth muscles,
and so causes an increase in the peripheral resistance
(see above). This autoregulation of flow results in a
homeostasis of wall tension for, in contracting, the
smooth muscles increase the thickness of the wall and
reduce the radius of the lumen. Both of these changes
reduce tension on individual muscle fibers. So the
vascular smooth muscles may keep their tension near
a constant level by contraction, when the pressure
rises.
The suggestion has been made that basal tone may
derive from locally released constrictor agents or
regional reflex arcs of independent nerve plexuses in
the vascular wall. However, Folkow & Oberg (29)
have recently published experiments which eliminate
these possible mechanisms. These experiments show
that the basic tone of precapillary resistance vessels
and autoregulation of flow is not due to nerve plexuses
or vasoconstrictor agents, but to myogenic activity.
The task of the autonomic nervous system, which
innervates the vascular muscles, would then be to
control the myogenic activity and adjust it to the
appropriate situation of the circulatory system (see
above). In the same way tone may be controlled by
chemical agents. This matter is also discussed in
Chapter 37.
As the pressure perfusing a vascular bed is gradually
reduced, the flow becomes less in proportion, the
exact nature of this relation changing under different
circumstances and with various vascular beds, as
discussed in Chapter 28. The flow stops before the
arteriovenous pressure difference reaches zero. The
pressure at which this stoppage occurs has been
called the "critical closing pressure" by Burton (19).
The physical and physiological factors which deter-
mine the height of this pressure are discussed in
Chapter 6.
Capillaries and Arteriovenous Anastomoses
The arterial side of the circulatory system is con-
nected with the venous side by two types of vessels:
capillaries and arteriovenous anastomoses (see also
Chapter 27).
Capillaries are the tiny vessels through the walls of
which materials are exchanged between blood and
the tissues. They consist of a thin layer of endothelial
cells, which sit on a basal membrane. On the outside
of the capillaries are found the pericytes, which are
cells with many irregular branches. Capillaries have
no smooth muscles and, in spite of earlier contentions
to the contrary, it is the current consensus that the
pericytes cannot constrict mammalian capillaries
(see Chapter 27); nor can the swelling of endothelial
cells cause stoppage of flow [for additional references
see Illig (41)].
The arteriovenous anastomoses are vessels the
walls of which consist almost entirely of smooth
muscle. They serve as a direct connection between
the arteries and veins, bypassing the capillaries. The
large amount of smooth muscle enables the arterio-
venous anastomoses to keep their lumens closed over
long periods of time. It is not impossible that these
anastomoses regulate the capillary blood flow- through
the several organs, according to their activity. If an
organ is active the anastomoses close and the blood
may flow through the capillaries, whereas in a resting
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
7 Z 3 t S S 7 8 9 10 It IZ 13 IV IS 16 17 IB '9 20 21 22 23 29 ZS 26 27 26 29 30 31 32
Obere Exfreraitat \ Thorax \ Bavchttoh/e J \ Unterz Extremitat
fig. 14. Percentage of transverse (circular) muscles (black columns), longitudinal muscles (white
columns), and collagen and elastic tissue (hatched columns) of the human veins at different sites
1 : Skin vein of the forearm; 2 : v. mediana cubiti; 3 : v. basilica; 4: v. comitans of the a. brachialis
5: v. brachialis, proximal part; 6: v. comitans of the a. circumHexa humeri dorsalis; 7: v. axillaris
8: v. brachiocephalica dextra; 9: v. thoracica interna; 10: v. thoracica longitudinalis dextra; 1 1 : v
cava cranialis; 1 2 : v. cava caudalis; 13: v. portae; 14: v. coronaria ventriculi; 15: v. lienalis; 16: v
renalis sinistra; 1 7 : v. renalis dextra; 18: v. mesenterica caudalis; 19: v. cava caudalis, most distal
part; 20: v. spermatica; 21 : v. iliaca communis sinistra; 22: v. iliaca communis dextra; 23 : v. dorsalis
penis subcutanea; 24: v. saphena magna of the thigh; 25: v. femoralis; 26: v. poplitea; 27: v. saphena
of the shank; 28: v. comitans of the a. tibialis posterior; 29: v. comitans of the a. tibialis anterior;
30: v. comitans of the a. dorsalis pedis; 31 : skin vein of the back of the foot; 32 : v. comitans of the
a. plantaris fibularis. [v. Kiigelgen (48).]
organ, the anastomoses may open and let the blood
bypass the capillaries.
Anastomoses in the lung possess longitudinal
muscles. Weibel (96, 97) has demonstrated that these
muscles always appear in those vessels which have
to withstand longitudinal elongation. He assumes
that the longitudinal muscles support this stretch.
Veins
The veins, in contrast to the arteries, are verv
variable in their wall structure. Usually they have a
larger percentage of collagen fibers than the arteries,
but there are veins in which the muscular mass exceeds
by far that of the collagen fibers. Veins have little
elastic tissue (fig. 1). The arrangement of the wall
elements is both circular and longitudinal in varying
proportions in different veins. Tension muscles seen
attached to elastic fibers in arteries seldom appear
in veins. Grau (34) described elastic-muscular sys-
tems in the large veins of the cow, similar to the
tension muscles described by Benninghoff (10, 11).
However, v. Kiigelgen (47, 49) could never find such
tension muscles in human veins. He described, rather,
muscles like the arterial ring muscles, the individual
muscle cells being connected together as a network.
This network of smooth muscles is tied to the collagen
fibers and the intima.
Figure 14 shows the percentages of smooth muscles
and collagen and elastic fibers in different human
veins. The smooth muscles are separately graphed as
transverse (circular) and longitudinal muscles. The
longitudinal muscles of the veins are not arranged
in bundles in the intima, as are those of the arteries.
Rather, they form a network in the wall with the
circular muscles, the smooth muscles being either
longitudinal or transverse, or at any other angle.
The circular muscles are mainly in the veins of
the leg, whereas the longitudinal muscles predomi-
nate in the abdominal veins. Figure 1 5 shows that
the proportion of circular muscles parallels the
pressure in the veins in the erect posture, there then
being considerable hydrostatic pressure in the human
leg veins. Since the wall tension in a tube, in the
transverse direction, is twice that in the longitudinal
direction [Frank (30)], the percentages of circu-
lar muscles are higher in veins subject to higher
pressures. Hydrostatic pressure and wall tension vary
with posture; therefore this variable pressure load
can be supported better by muscles capable of
myogenic activity, as in arteries, than by collagen or
elastic tissues (see above). Along with the higher
amount of smooth muscles in the leg veins the
relationship of radius to wall thickness is less than that
of other veins [v. Kiigelgen (48)]. Veins of the thorax,
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 15. The relationship of the
abundance of transverse (circular)
muscles in human veins {smooth line)
to the venous pressure in the erect
posture {dotted line). The numbers
indicate the same veins as in fig. 14.
[v. Kiigelgen (48).]
ICO
%
SO
150
cm
flachenanteil der Quermuskulalur
Wahrer Innendruck
Obere Extremitat { Thorax i-
16 18 20
Bauchhohle
£. tT- ZU CO ./I
-I iUntcre Extremiiat
abdomen, and neck are not under this hydrostatic
stress and have less circular muscle tissue.
The mechanical properties of the veins are similar
to those of the arteries. Smaller veins show different
pressure-volume diagrams, depending upon the state
of contraction of their smooth muscles, as do those of
muscular arteries [Alexander (2)]. The only differ-
ence between the diagrams of the muscular arteries
and the veins is that the diagram of the veins is
located at much lower pressures. This indicates that
the elasticity of small veins depends to a high degree
on smooth muscle. The large veins, like the vena cava,
give a pressure-volume diagram more like that of
elastic arteries [Blomer (14)]. It shows an S-shape,
like the aorta, with an inflexion point at the low
pressure of about 7 mm Hg. This S-shape does not
depend on the activity of the smooth muscles. The
main support of the wall tension of the large veins,
at least above 7 mm Hg, is the collagen tissue rather
than the elastic tissue, as is the case for elastic arteries,
since the amount of collagen exceeds by far the
amount of elastic tissue in the vein wall. The relatively
high distensibility of the vena cava, in spite of the
collagen fibers, may depend on a gradual recruit-
ment of these fibers, as shown in figure 9, phases 3
and 4, or it may be due to a reorientation of the net-
work formed by the collagen fibers in the venous
wall [see v. Kiigelgen (49)].
The most striking difference between arteries and
veins are that the veins possess valves and are securely
embedded in the surrounding tissue (33, 51, 53, 78),
whereas the arteries never have valves and they are
loosely connected with the surrounding tissue.
The valves of the veins are folds of the intima. They
consist of collagen and elastic fibers but not of smooth
muscles. Around the vein at the base of the valve is a
thickened band of collagen fiber (51). Usually two
valves face each other [Bardeleben (8)]. The leg veins
are best guarded by valves. Very small veins are said
to be free of valves [Klotz (45)], as are the venae
cavae. The valves minimize postural hydrostatic
pressure changes in the leg veins, protecting the
capillaries and the veins themselves from unphysio-
logical pressures.
The action of the skeletal muscles, compressing,
stretching, and releasing the veins, and even arterial
pulsation (53, 78), cause periodic changes in venous
capacity. Since the valves open toward the heart,
these movements cause the veins to act as pumps,
promoting return of blood to the heart and main-
taining low capillary pressures (for further details
see Chapter 32).
The numbers of valves in the veins depend very
much on the age of the individual. Many valves
degenerate with aging. Bardeleben (8) has ascer-
tained, for example, that the greater saphenous vein
of a child has, on the average, 1 3.6 valves, whereas
that of an adult has only 10.7 valves. Klotz (45) has
even found up to 70 per cent of atrophied valves at
age 70. The first sign of degeneration is a functional
insufficiency of the valves permitting leaking at higher
pressures when the vein is distended, although they
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
883
still are tight at low pressures [Schliiter (79)]. Further
degeneration makes them leaky at low pressures and
later they are so degenerated that only a small margin
remains or the valve leaflets are broken through. At
last they vanish completely. The effects of venous
valvular insufficiency are discussed in Chapter 36.
The firm anchorage of the proximal part of the
veins may facilitate transmission of arterial pulsation
from arteries to their venae comitantes [Schade (78),
v. Lanz et al. (53)], but it seems also very convenient
for another task. Any distensible tube which stands
upright and is filled with fluid tends to pull down-
ward. The radius in the proximal part is then small
and the wall is stretched mainly in a longitudinal
direction. It is therefore necessary that the tube be
supported in the proximal part in the longitudinal
direction and be fixed to its surroundings. Such a
support may be formed in the veins by the bracing
straps (33, 53) and also by the longitudinal muscles.
In the distal part of such an upright tube the radius
is enlarged and the wall is stretched mainly in a
transverse direction. As indicated above, this is
countered by the increasing amount of circular
muscles.
NUTRITION OF THE VASCULAR WALL
The vascular wall is a living organ and its smooth
muscles need a source of energy. Their nutrition is
accomplished by two different means: diffusion from
the circulating blood from the inside of the vessel
toward the outside, and from the vasa vasorum
vessels which dip into the vascular wall from the
outside. These two supply routes meet in the vascular
wall. Miiller (64) has demonstrated a model in which
ten coaxial thin rubber tubes with increasing diam-
eters were telescoped and fixed so that the fluid
between the different sheets could not escape during
distention. This model satisfies very well the situation
in the vascular wall, which is also built from different
layers consisting of different materials. The pressure
between the sheets is equal to the negative radial
stress, and the tension of the different sheets of the
model decreases nearly linearly from the inside to
the outside at any given internal pressure. The inner-
most sheet has almost the same pressure as the filling
fluid, whereas the outermost sheet is at the ambient
pressure. This means that the vessels which supply
the vascular wall meet progressively higher pressures
the further they penetrate the wall. On the other
hand, the pressure gradient from the inside toward
the outside facilitates the movement of materials
directly from the circulating blood through the wall.
The border between diffusion and vascular supply in
the vascular wall depends on the thickness of the
wall. The limit for diffusion is set by oxygen, which
is transported in the blood by the hemoglobin and
which can supply tissues adequately if the distance
from the hemoglobin, which stays in the blood, to
the tissue cells is not too great. This distance is, in
the vascular wall, about 500 ju [Linzbach (56)]. The
limit for the vasa vasorum is set by the pressure in
the wall. Since the vasa vasorum come mostly from
the adventitia, the pressure fall over the length of
the vasa vasorum allows them only to penetrate as
far as the pressure in the wall is less than the pressure
of the intramural capillaries.
Diffusion from the Inside
The whole circulatory system is lined with a single
layer of endothelium. This lining prevents extra-
vasation of blood even if the pressure in the vessels
exceeds by far the surrounding pressure. Any nutrient
material entering the vascular wall from the inside
must pass across this endothelial lining. Such pene-
tration is rendered possible either through the pres-
sure and concentration gradient between the blood
and the wall tissue or by means of active transport.
Chambers & Zweifach (21) assume that the individual
endothelial cells are held together by a cement
substance and that this cement substance makes
penetration possible. However, Linzbach (56) could
never find such a cement substance. He describes
cell branches which are near the basal side of the
endothelial cells, and with which the endothelial
cells are very tightly connected. The boundary
between the cells may form fissures, where capillary
attraction may be effective and render penetration
possible. Pappenheimer (65) suggests channels
between the endothelial cells with a diameter of 30
to 45 A, through which the materials can enter the
wall. All these mechanisms depend on a pressure
gradient between the blood and the wall tissue or an
osmotic concentration gradient of the different
materials.
However, there may also be active transport.
Moore & Ruska (63) described small vacuoles on
the surface of the endothelial cells which contain
blood plasma. These vacuoles separate themselves
from the surface and wander through the cell sub-
884
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 1 6. Vasa vasorum in the
wall of the aorta of the horse.
China ink, thick cuts. Left:
longitudinal section; right: cross
section. [Straubesand (87).]
1.5mm
6.0mm <
Interna
> Media
\ fig. 17. Schematic longitudinal section through the aortic
wall at the origin of an intercostal artery. / = intercostal
artery; V = vas vasis externum; a = outer branch; ; = inner
branch. [Schonenberger & Muller (82).]
stance to the basal side of the endothelial cells, where
thev release their contents to the wall tissue. This
transport is called cytopempsis. Another possibility
of active transport through the endothelium may be
by a similar mechanism which was described by
Ussing (92) for the frog skin. He demonstrated that
sodium is actively transported across the skin cells
l>\ a carrier system located in the cell membrane. This
would mean that ions or other materials enter the
endothelial cell passively through the surface mem-
brane, along a concentration gradient, and are then
actively transported out of the cell and into the wall
tissue against a concentration gradient. Sawyer &
Yalmont (77) have published evidence for such a
mechanism in the canine thoracic aorta and vena
cava, where the net flux of sodium or chloride ions
fig. 18. Cross section through the thoracic aorta of the dog,
showing a vas vasis internum. There is no branching of capil-
laries in the intima and the innermost part of the media. The
dark masses in the outer third of the media are accumulations
of injected material that has broken out of the capillaries.
[Woerner (101 I
in the aorta is from the inside to the outside. In the
vena cava it is in the opposite direction (for a possible
explanation of this contrasting behavior see below).
The variety of theories about the transport of material
across the endothelial lining shows that much work
remains to be done. Most of these theories are deduced
from experimental results on capillary endothelial
cells but it may be that transport differs in the capil-
lary endothelium and the endothelium of larger
vessels. There is also the possibility that different
tissues use different transport mechanisms. (See also
Chapter 29.)
The further transport through the intima and the
media may be passive in the intercellular space,
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
885
fig. 19. Schematic drawing of the vascularization of a
middle-sized artery with both its venae comitantes. a: Artery;
b: venae comitantes; c: arterial vasa vasorum; d: venous vasa
vasorum; e: lymphatic vessel; /.■ capillaries of the arterial
sheath; g: capillaries in the stratum longitudinale fibroelasti-
cum of the artery = wall capillaries; h: small channel, less
than 3 fi thick; i: vascular network in the venous wall; j:
capillary network in the venous wall; k: nerve; /.• stratum
longitudinale fibroelasticum = adventitia of the artery;
m: media of the artery; n: circular sheath of the vessel group;
0: conjunctiva of the sheath. [Lang (51).]
effected by the pressure and concentration gradient
across the vascular wall.
Vasa Vasorum
The vasa vasorum penetrate the vessel wall to
different depths, depending on the thickness of the
wall and the type of vessel. The thicker the wall,
the greater the part of the wall tissue they supply.
In general, veins have greater vascularization than
arteries. The vasa vasorum penetrate the aortic wall
as far as the inner third of the media (fig. 16). The
innermost part of the media and the intima are
always free of capillaries. Only arteriosclerotic vessels
with a thickened intima show vascularization of the
innermost part of the wall [Woerner (101)].
The vasa vasorum of the aorta can be classified as
vasa vasorum externa and vasa vasorum interna.
The vasa vasorum externa originate near the origin
of arterial branches, such as the intercostal arteries.
They soon divide into an outer branch and an inner
branch (fig. 17). The outer branch goes into the
adventitia and from there sends branches into the
wall, whereas the inner branches remain within the
wall. Their branching is mostly trichotomous. The
vasa vasorum interna originate directly from the
lumen (fig. 18) far away from branching vessels,
that is, in the aorta on the ventral side [Schonen-
berger & Miiller (82)]. They are not very numerous;
according to Woerner (101), there are never more
than two per square centimeter. They are mostly-
found in the proximal part of the aorta and very
seldom in the distal part. The vasa vasorum externa
and interna anastomose in the aortic wall. The vasa
vasorum externa are about 65 to 70 mm in length,
the vasa vasorum interna about 30 to 50 mm.
The more peripherally the arteries are located,
the less vascularized is their wall. Figure 19 shows
a schematic drawing of an artery with its two venae
comitantes from a human shank. The vasa vasorum
of the peripheral arteries arise at smaller branches
of the artery, similar to the vasa vasorum externa of
the aorta. There are no vasa vasorum interna in these
arteries. The vasa vasorum never dip into the media.
They are located in the stratum longitudinale fibro-
elasticum, the innermost part of the adventitia. They
form their capillary loops mostly in the longitudinal
direction. In addition to these capillary loops there
are still smaller vessels. Lang (51) has described two
types of such small vessels. The first type is a small,
blind-ending channel about 3 mm in length and
1 to 3 ju diameter (fig. 19) which is much too small
for blood cells to pass. These small channels run into
the tip or the venous part of the capillary loop. The
second type is a network of small channels of about
the same diameter (fig. 20).
The same vas vasis which supplies the artery also
supplies its venae comitantes (fig. 19). In contrast
to the case for arteries, the capillaries in the venous
wall form a dense network which extends to the media.
The segment around the valves is usually not vascular-
ized. The venous vasa vasorum do not drain directly
into the large vein along which they lie; rather they
and their counterparts in the arterial wall drain into a
small venous branch [Lang (51)].
Schonenberger & Miiller (82) have calculated the
drop of pressure in the vasa vasorum externa of the
aorta, finding that the capillary pressure within the
wall can be sufficiently high only if the origin of the
main vas vasis is very near the inner surface of the
aortic wall (fig. 17). The intramural capillaries must
be quite near to the origin of the vasa vasorum externa
886
HANDBOOK. OF PHYSIOLOGY
CIRCULATION II
fig. 20. Net-shaped small channels with a diameter of
1-2 n in the adventitia of the peroneal artery. Capillaries of the
sheath and the arterial sheath partly removed. The small
channels were connected with the capillaries of the sheath.
a: Wall capillaries in the stratum longitudinale hbroelasticum;
b: network of small channels; c: arterial sheath and capillaries
of the sheath, in situ. [Lang (51).]
to maintain a pressure which is necessary to exceed the
tissue pressure. If the capillaries are too far from the
origin, the pressure will be too low to supply the wall
efficiently. At greater distances from the intercostal
arteries, the vasa vasorum interna, with their short
delivery system, may provide sufficient blood supply.
Schonenberger & M tiller (82) have also deter-
mined the flow and resistance in the vasa vasorum of
the cow's aorta. Flow increases and resistance
decreases, with rising pressure, with a maximum at
about 140 mm Hg (distended vessels). At higher
pressures, a decrease of flow and an increase of
resistance occurs (collapsed vessels). The maximal
flow seems to occur at the systolic blood pressure
level. This indicates that nutrition of the vascular
wall may be problematic in hypertension, if the
diastolic pressure exceeds the physiological systolic
pressure, since the flow in the vasa vasorum may never
reach the maximum. The inner region of the wall,
which is not nourished by the vasa vasorum, also
suffers ischemic changes, including a compensating
increase in vascularity.
The lymphatics certainly play no role in the circu-
lation of the tissue fluid within the vessel wall itself,
as they do in other organs. The very small channels
which arise from capillaries (see above) and which
have a diameter of 1 to 3 fi may function like the
lymphatics with local drainage (51). The mechanical
or hydrostatic pressure gradients are irrelevant to
this diffusion transport and determine only the
direction of flow in the vasculature of the vessel wall —
whether the supply to a capillary comes from an
internal or an external arterial branch.
The situation in the veins is quite different from
that in the arteries. There are no channels of supply
from the lumen of the vein as is the case in the artery
and, since the venous blood is depleted of oxygen and
nutrients, the supply by diffusion through the intima
is nonexistent or very limited. The pressure gradient
is from the external arterial plexus to the capillary
plexus, extending as far as the intima. Venous
drainage is into small venae vasorum rather than into
the lumen of the large vein. The interstitial space is
probably drained by "lymphatics", although the fluid
may pass directly across the intima and into the
lumen. The pressure gradient is favorable for this
movement of fluid and it might nourish the non-
vascular inner wall. This concept agrees very well
with the experiments of Sawyer & Valmont (77), who
have found a net transport of Na and CI from the
outside to the inside in the canine vena cava.
REFERENCES
Abraham, A. Uber die Struktur und die Endigungen
der Aorticusfasern im Aortenbogen des Menschen mit
Berucksichtigung der Cholinesterase-Aktivitat der Presso-
receptoren. Z. mikroskop.-anat. Forsch. 62: 194-228, 1956.
Alexander, R. S. The participation of the venomotor
system in pressor reflex. Circulation Research 2 : 405-409,
'954-
Bader, H. Uber die Reversibilitat der plastischen Deh-
nung des glatten Muskels. Z. Biol. 1 10: 347-355, 1958.
4. Bader, H. Die Abhangigkeit des Verhaltnisses von
Radius zu VVanddicke in der menschlichen Brustaorta
vom Alter und vom Druck. In preparation.
5. Bader, H., and E. Kapal. Uber die Bedeutung der
Wandmuskulatur fur die elastischen Eigenschaften des
Aortenwindkessels. Z. Biol. 109: 250-261, 1957.
6. Bader, H., and E. Kapal. Experimentelle Untersuch-
ungen uber die Druck-Volumenbeziehung von Gum-
mischlauchen. 2. Mitteilung. Z. Biol. 109: 325-331, 1957.
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
887
7. Bader, H., and E. Kapal. Altersveranderungen der
Aortenelastizitat. Geronlologia 2: 253-265, 1958.
8. Bardeleben, K. Das Klappengesetz. Jenai. Z. Naturw.
14: 467, 1880.
g. Bavliss, W. M. On the local reactions of the arterial
wall to changes of internal pressure. J. Physiol. 28 : 220,
1902.
10. Benninghoff, A. Uber die Beziehungen zwischen ela-
stichem Geriist und glatter Muskulatur in der Arterien-
wand und ihre funktionelle Bedeutung. Z. Zellforsch. 6:
348-396, 1927.
11. Benninghoff, A. Blutgefafe und Herz. In: Handbuch
der mikroskopischen Anatomic, Berlin: Springer-Verlag,
1930, vol. vi/i pp. 1-225.
12. Bergel, D. H. The static elastic properties of the arterial
wall. J. Physiol. 156: 445-457, 1961 .
13. Bergel, D. H. The dynamic elastic properties of the
arterial wall. J. Physiol. 156: 458-469, 1961.
14. Blomer, H. Dehnungsversuche an iiberlebenden groBen
Venen. Z. Biol. 107: 468-480, 1955.
15. Boeke, J. Innervationsstudien IV. Die efferente Gefafiin-
nervation und der sympathische Plexus im Bindegewebe.
Z. mikroskop.-anat. Forsch. 33: 276-328, 1933.
16. Bozler, E. Conduction, automaticity and tonus of
visceral muscles. Experientia 4: 213-218, 1948.
17. Bulbring, E. Physiology and pharmacology of intestinal
smooth muscle. Lectures on the Scientific Basis of Medicine.
Univ. of London 7: 374-397, 1957-1958.
18. Burnstock, G. and C. L. Prosser. Responses of smooth
muscles to quick stretch; relation of stretch to conduction.
Am. J. Physiol. 198: 921-925, i960.
19. Burton, A. C. On the physical equilibrium of small blood
vessels. Am. J. Physiol. 164: 319-329, 1 95 1 .
20. Burton, A. C. Relation of structure to function of the
tissues of the wall of blood vessels. Physiol. Revs. 34: 619-
642. 1954-
21. Chambers, R., and B. W. Zweifach. Intercellular
cement and capillary permeability. Physiol. Revs. 27 : 436,
1947.
22. Dickinson, C. J. Rapid contractile properties of isolated
arteries. Nature 185:620-621, i960.
23. Fischer, H. Uber die funktionelle Bedeutung des Spiral-
verlaufes der Muskulatur in der Arterienwand. Morphol.
Jahrb. 91:394-446, 1951.
24. Fleisch, A. Gestalt und Eigenschaften des peripheren
Gefaliapparates. Handbuch der normalen und palhologischen
Physiologic, Berlin: Springer-Verlag, 1927, vol. vii/2/2
pp. 865-888.
25. Fleisch, A. Die aktive Forderung des Blutstromes durch
die GefaCe. Handbuch der normalen und palhologischen
Physiologic, Berlin: Springer-Verlag, 1927, vol. vii/2/2,
pp. 1071-1087.
26. Folkow, B. A study of the factors influencing the tone of
denervated blood vessels, perfused at various pressures.
Acta Physiol. Scand. 27: 99-117, 1953.
27. Folkow, B., and B. Lofving. The distensibility of the
systemic resistance blood vessels. Acta Physiol. Scand. 38:
37-52, I956-
28. Folkow, B. Role of the nervous system in the control of
vascular tone. Circulation 21 : 760-768, i960.
29. Folkow, B., and B. Oberg. Autoregulation and basal
tone in consecutive vascular sections of the skeletal
muscles in reserpine treated cats. Acta Physiol. Scand. 53:
105, 1961.
3'
33
30. Frank, O. Die Elastizitat der Blutgefafie Z. Biol. 71 :
255-272, I92°-
Frank, O. Das Aufblahen von Schlauchen und kugel-
formigen Blasen. Z. Biol. 88: 93-104, 1928.
Frasher, W. G., and S. S. Sobin. Distensible behavior of
pulmonary artery. Am. J. Physiol. 199: 472-480, i960.
Goerttler, K. Uber den Einbau der grofien Venen des
menschlichen Unterschenkels. Z. Anat. Entwicklungs-
geschicte 116:591-609, 1953.
34. Grau, H. Zur Frage des "elastisch-muskulosen Systems"
in der Venenwand. Morphol. Jahrb. 67: 745-750, 1931.
35. Greven, K. Die Aktionsstrome der glatten Muskulatur
und ihre Beziehung zur Erregungsbildung und Erregungs-
leitung. Klin. Wochschr. 33: 241-247, 1955.
36. Harkness, R. D. Metabolism of collagen. Lectures on the
Scientific Basis of Medicine. Univ. of London, 5: 183-216,
1955-56
37. Heymans, C, and A. L. Delaunois. Action of norepi-
nephrine on carotid sinus arterial wall and blood pressure.
Proc. Soc. Exptl. Biol. Med. 89: 597, 1955. Cited in: Hey-
mans and van den Heuvel-Heymans (39).
38. Heymans, C, A. de Schaepdrijver, and T. O. King.
Actions of heart rate and blood pressure of mechanical
tension on carotid sinus arterial wall. XX" Congres
International de Physiologic Resumes des Communications.
Bruxelles, 1956, pp. 424-425.
39. Heymans, C, and G. van den Heuvel-Heymans.
Homoostase des Blutdrucks und Hypertonic Ciba Symposia
5:66-72, 1957.
40. Hieronymi, G. Uber den altersbedingten Formwandel
elastischer und muskularer Arterien. Osterr. Akad. Wiss.
Math.-naturw. Kl. Sitzber. pp. 221-352, 1956.
41. Illig, L. Capillar "Kontraktilitat", Capillar "Sphinkter"
und "Zentralkanale" ("A. -V. -bridges"). Ein tierexperi-
menteller Beitrag zur motorischen Funktion und zum
Aufbau des Capillarbettes mit Schrifttumsubersicht. Klin.
Wochschr. 35: 7-22, 1957.
41 a. Johnson, W. H. Tonic mechanisms in smooth muscles.
Physiol. Revs. 42: suppl. 5: 1 13-143, 1962.
42. Kapal, E. Die elastischen Eigenschaften der Aortenwand
sowie des elastischen und kollagenen Bindegewebes bei
frequenten zyklischen Beanspruchungen. Z. Biol. 107.
347-404, 1954.
43. Kapal, E., and H. Bader. Ein Modell fur die Wirkungs-
weise der glatten Muskulatur in der Aortenwand. Z. Biol.
1 10: 236-240, 1958.
44. Kapal, E., and H. Bader. Uber die elastischen Eigen-
schaften des Aortenwindkessels. Untersuchungen an
ganzen menschlichen Aorten. Z. Kreislaufforsch. 47: 66-73,
'958-
Klotz, K. Arch. Anat. Entwicklungsgeschichte 1887, p. 159,
cited in Fleisch (24).
Kobavashi, Y. Veranderungen der Struktur der Brust-
aorta des Menschen wahrend der pra- und postnatalen
Entwicklung und im Senium. Arch. hist. jap. 13: 503-516,
1957-
47. Kugelgen, A. v. Uber den VVandbau der grofien Venen.
Morphol. Jahrb. 91 : 447-482, 1951.
48. Kugelgen, A. v. Uber das Verhaltnis son Ringmusku-
latur und Innendruck in menschlichen groBen Venen.
Z. Zellforsch. 43: 168-183, 1955.
49. Kugelgen, A. v. Weitere Mitteilungen uber den VVand-
bau der grofien Venen des Menschen unter besonderer
45-
46.
888
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
"»"
r.'
57
58
Beriicksichtigung ihrer Kollagenstruktur. Z. Zellforsch.
44: 121-174, '956-
Landowne, M., and R. W. Stacy. Glossary of terms. 69.
In: Tissue Elasticity, edited by J. W. Remington, Wash-
ington, D.C. : Am. Physiol. Soc, 1957, pp. 191 -201.
Lang, J. Uber die Vascularisation der Wand und des 70.
Einbaugewebes mittelgrolier Gefafie des Unterschenkels.
Z. Aunt. Entuicklungsgeschichte 122:482-517, 1961.
52. Lansing, A. I. Elastic tissue. In: The Arterial Wall.
Baltimore: Williams & Wilkins, 1959, pp. 136-160. 71.
53. Lantz, T. v., A. Kressner, and R. Sghwendemann.
Der Einbau der oberflachlichen und der tiefen Venen am 72.
Bein, morphologisch und konstruktiv betrachtet. Z. Anat.
Entwicklungsgeschichte. 108:695-718, 1938.
54. Laszt, L. Uber die Eigenschaften der GefalSmuskulatur 73.
mit besonderer Beriicksichtigung der Kalium-VVirkung.
Arch. Kreislaufforsch. 32: 220-244, '960.
55. Leontowitsch, A. W. Uber die Ganglienzellen der 74.
BlutgefalSe. Z. Zellforsch. Ill 23-45, !93°-
56. Linzbach, A. J. Die allgemeine Pathogenese der GefalS-
krankheiten. In : Angiologie, edited by M. Ratschow. 75.
Stuttgart: Thieme-Verlag, 1959, pp. 140-164.
56a.Lowv, J. and J. Hanson. Ultrastructure of invertebrate
smooth muscles. Physiol. Revs. 42: suppl. 5: 34-42, 1962. 76.
Lowv, J., and B. Millman. Contraction and relaxation
in smooth muscles of Lamellibranch Molluscs. Nature 1 83 :
1 730-I 731, 1959. 77-
Meyer, W. W. Die Lebenswandlung der Struktur von
Arterien und Venen. Verhandi. dent. Ges. Kreislaufforsch. 24:
15-4°. I958 .. 78-
59. Meyer, W. W. Uber die eigenartige Beziehung des
elastischen Geriistes zur glatten Muskulatur im extra-
pulmonalen Abschnitt der Lungenarterie des Menschen. 79.
Z. Zellforsch. 43: 383-390, 1955.
60. Meyer, W. W., and E. Simon. Die phasenartige Abwand-
lung der Pulmonalis-Volumendehnbarkeit im Verlauf des 80.
Lebens in ihrer Beziehung zur Struktur der Arterienwand.
Arch. Kreislaufforsch. 31 : 95-112, 1959.
61. Millahn, H. P., and D. Jaster. Der EinflulS von Nor-
adrenalin und Acetylcholin auf das Druckvolumdiagramm
und die Elastizitat isolierter Rinder-und Schweineaorten. 81.
Z. Biol. 1 1 1 : 35 1 -356, 1 960.
62. Monnier, M. Die funktioncllen Potenzen der isolierten
Arterie (Erregbarkeit, Reizbildung, Erregungsleitung,
autonome Anpassung). Helvet. Physiol, et Pharmacol. Acta 82.
2:533-539. '944-
63. Moore, D. H., and H. Ruska. The fine structure of
capillaries and small arteries. J. Biophys. Biochem. Cytol. 83.
3:457. 1957-
64. Muller, A. Die mehrschichtige Rohrwand als Modell
fur die Aorta. Helvet. Physiol, et Pharmacol. Acta 17: 131-
■45. !959- 84-
Pappenheimer, J. R. Passage of molecules through capil-
lary wall. Physiol. Rev. 33: 387, 1953. 85.
66. Peterson, L. H., R. E. Jensen, and J. Parnell. Me-
chanical properties of arteries in vivo. Circulation Research
8 : 622-639, 1 960.
Petry, G., and G. Heberer. Die Neubildung der 86.
GefalJwand auf der Grundlage synthetischer Arterien-
prothesen. Langenbecks Arch. u. Dtsch. Z. Chir. 286: 249-
290. !957-
Prosser, C. L. Comparative physiology of activation of
muscles, with particular attention to smooth muscles. In : 87
•'.I
67
68.
Structure and Function of Muscle, edited by G. II. Bourne.
New York : Academic Press, 1 960, pp. 387-434.
Prosser, C. L , G. Burnstock, and J. Kahn. Conduction
in smooth muscle : comparative structural properties. Am.
J. Physiol. 199: 545-552, i960.
Reichel, H. Die elastischen Eigenschaften des glatten
SchlieBmuskels von Pinna nobilis bei verschiedenen
Tonuslangen unter plastischen und dynamischen Bedin-
gungen. Z Biol. 105: 162-169, ' 952-
Reichel, H. Muskelphysiologie. Berlin: Springer- Verlag,
i960.
Remington, J. W. Hysteresis loop behavior of the aorta
and other extensible tissues. Am. J. Physiol. 180: 83-95,
!955-
Remington, J. W. Extensibility behavior and hysteresis
phenomena in smooth muscle tissues. In : Tissue Elasticity,
Washington D.C. : Am. Physiol. Soc, 1957, pp. 138-153.
Reutervvall, O. P. Uber die Elastizitat der Gefafiwande
und die Methode ihrer naheren Priifung. Acta Med.
Scand. Suppl. 2, 1-175, 1921.
Roach, M. R., and A. C. Burton. The reason for the
shape of the distensibility curves of arteries. Can. J.
Biochem. Physiol. 35: 681-690, 1957.
Roach, M. R., and A. C. Burton. The effect of age on
the elasticity of human iliac arteries. Can. J. Biochem.
Physiol. 37:557-570. [959-
Sawyer, P. N., and I. Valmont. Evidence of active ion
transport across large canine blood vessel walls. Nature
189:470-472, 1961.
Schade, H. Die Pulsationsiibertragung von der Arterie
auf die Vene und ihre Bedeutung fur den Blutkreislauf.
Z Kreislaufforsch. 28: 131-144, 153-172, 1936.
Schluter, F. Die SchlielSfahigkeit der Venenklappcn
unter dem EinflulS funktionell und morphologisch wirk-
samer Faktoren. Z. Kreislaufforsch. 50: 1-15, 1961.
Schluter, F., and K. Wezler. Die Wirkung konstringie-
render und dilatierender Stoffe auf die Querdehnbarkeit
isolierter kleiner Arterien vom muskularen Typ. Abhandl.
Akad. Wiss. Lit. Main-.. Mnlh.-Naturw. Kl. pp. 71-140,
'955-
Schonbach, G., and H. Langendorf. Das Verhaltnis
von Innenradius und Wandstarke in den kleinen Blutge-
falien. Abhandl. Akad. Wiss. Lit. Mainz, Malh.-Nalurw. Kl.
PP. '55-l85, '955-
Schonenberger, F. and A. Muller. Uber die Vaskulari-
sation der Rinderaortenwand. Helvet. Physiol, et Pharmacol.
Acta 18: 136-150, i960.
Schonenberger, F., and A. Muller. Uber die Elastizitat
und Reaktionsfahigkeit der extrakorporalen im physi-
ologischen Zustand erhaltenen Rinderaorta. Helvet.
Physiol, el Pharmacol. Acta 18: 151-173, i960.
Schultze-Jena, B. S. Uber die schraubenformige Struktur
der Arterienwand. Morphol. Jahrb. 83: 230-246, 1939.
Seto, H. Uber die efferenten Nerven im Aortenbogen
und im Herzen beim Menschen im Hinblick auf den
Aorten- und Herzrerlex. Arb. anat. Inst, kaiserl.- japan.
Univ. Sendai 20: 1-16, 1937.
Simon, E., and W. W. Meyer. Das Volumen, die
Volumdchnbarkeit und die Druck-Langen-Beziehungen
des gesamten aortalen Windkessels in Abhangigkeit von
Alter, Hochdruck und Arteriosklerose. Klin. Wochschr. 36:
424-432, 1958.
Staubesand, J. Funktionelle Morphologic der Arterien,
ANATOMY AND PHYSIOLOGY OF THE VASCULAR WALL
889
Venen und arterio-venosen Anastomosen. In : Angiologie,
edited by M. Ratschow. Stuttgart: Thieme-Verlag,
pp. 23-72, 1959.
88. Stohr, P., Jr. Mikroskopische Anatomic des vegetatisen
Nervensystems In: Handbuch der mikroskopischen Anatomic,
Berlin: Springer-Verlag, 1957, vol. iv/5, p. 215.
89. Sunder-Plassmann, P. Untersuchungen iiber den Bulbus
carotidis bei Mensch und Tier im Hinblick auf die
"Sinusrerlexe*' nach H. E. Hering; ein Vergleich mit
anderen Gefalistrecken; die Histophysiologie des Bulbus
carotidis; das Glomum caroticum. Z. Anal. Entwick-
lungsgeschichte. 93:567-622, 1930.
90. Thron, H. L., K. D. Scheppokat, A. Hevden, and
O. H. Gauer. Das Verhalten der kapazitiven und der
Widerstandsgeiafie der menschlichen Hand in Abhangig-
keit von thermischen Einfliissen. Pfliigers Arch. ges. Physiol.
266: 150-166, 1958.
91. Thurau, K., and K Kramer. Die Reaktionsweise der
glatten Muskulatur der Nierengefalie auf Dehnungsreiz
und ihre Bedeutung fiir die Autoregulation des Nieren-
kreislaufes. Pfliigers Arch. ges. Physiol. 268: 188-203, '959-
92. Ussing, H. H. The frog skin potential. J. Gen. Physiol. 43:
'.35-147, i96°-
93. Uxkull, J. von. Studien iiber den Tonus. Z. Biol. 44:
269-344- <9°3-
94. Wagner, R., and E. Kapal. Uber die Eigenschaften des
Aortenwindkessels. 2. Mitteilung. Z. Bud. 105: 263-292,
'952-
95. Wassermann, F. The intercellular components of con-
nective tissue. Ergeh. Anat. it. Entwicklungsgeschichte 35 :
240-333. >956-
96. Weibel, E. Die Entstehung der Langsmuskulatur in den
Asten der a. bronchialis. Z. Zelljorsch. 47: 440-468, 1958.
97. Weibel, E. Die Blutgefafianastomosen in der mensch-
lichen Lunge. Z. Zelljorsch. 50: 653-692, 1959.
98. Wetterer, E. Die Wirkung der Herztatigkeit auf die
Dynamik des Arteriensystems. Verhandl. dent. Ges. Kreis-
laujforsch. 22 : 26-60, 1 956.
99. Wetterer, E., and E. Kapal. Druck-Umfang-Bezie-
hungen pulsierender Arterien in situ. 27. Tagung der
Deutschen Physiologischen Gesellschaft 23.-26. Mai 1961,
Zurich. Pfliigers Arch. ges. Physiol. 274: 39, 196 1.
100. Wezler, K., and F. Schluter. Die Querdehnbarkeit
isolierter kleiner Arterien vom muskularen Typ. Akad.
Wiss. Lit. Mainz, Ablwndl. math.-nat. Kl. Jg. 1953, pp.
413-492.
101. Woerner, C. A. Vasa vasorum of arteries, their demon-
stration and distribution. In: The Arterial Wall, edited by
A I. Lansing. Baltimore: Williams & Wilkins 1959, pp.
1-14.
102. Wohlisch, E., R. du Mesnil de Rochemont, and H.
Gerschler. Untersuchungen iiber die elastischen Eigen-
schaften tierischer Gewebe I. Z. Biol. 85: 325-341, 1927.
103. Zatzman, M., R. W. Stacy, J. Randall, and A. Eber-
stein. Time course of stress relaxation in isolated arterial
segments. Am. J. Physiol. 177: 299-302, 1954.
CHAPTER 27
Patterns of the arteriovenous pathways
MARY P. WIEDEMAN
Department of Physiology, Temple University
School of Medicine, Philadelphia, Pennsylvania
CHAPTER CONTENTS
Definitions
Techniques for Microscopic Observation of Small Blood Vessels
Hamster Cheek Pouch
Transparent Chamber
Fused Quartz Rod
Bulbar Conjunctiva
Rat Mesoappendix
Bat Wing
Structure of Terminal Vascular Beds
Microcirculation in the Bat Wing
Microcirculation in the Rabbit Ear
Microcirculation in the Mesentery
Microcirculation in the Hamster Cheek Pouch
Microcirculation in Skeletal Muscle
Microcirculation in Myocardium
Microcirculation in Skin
Microcirculation in Stomach and Intestine
Microcirculation in the Bulbar Conjunctiva
Microcirculation in the Spleen
Microcirculation in the Lung
Microcirculation in the Cochlea
Preferential or Thoroughfare Channel
Arteriovenous Anastomoses
Blood Flow Through Terminal Vascular Beds
Capillary Contractility
Vasomotion
Summary
although capillary vessels in living animals have
been observed microscopically for three hundred
years, there is a great diversity of opinion regarding
the structure and function of minute vessels in terminal
vascular beds. Actually, a survey of the descriptions
of patterns formed by capillary networks and the
flow of blood through them in a wide variety of tissues
and organs reveals great similarity in vascular
patterns and also in the manner in which blood flows
from arterioles through capillary nets on to collecting
venules. This similarity of structure and function
leaves the impression that acceptable generalizations,
applicable to these terminal beds, must be developed
in order that future studies may prove profitable.
It is well known that unnecessary disagreement
arises from lack of uniformity in terminology. It is
equally obvious that unnecessary confusion arises
from assigning complex functions to isolated com-
ponents of a specific vascular bed when, in truth,
both the activity and the structure are common
features of small blood vessels anywhere.
DEFINITIONS
Any attempt to supply a list of universally accept-
able definitions of vascular structures would be
useless, and yet it is necessary to present some general-
izations regarding current usage of terms before
describing the types and variations of structural
patterns that connect distributing arteries and col-
lecting veins.
The term "microcirculation" is used to designate
blood flow through small vessels at the capillary level
(48). The microcirculatory bed is the ultimate portion
of the cardiovascular system which is generally
accepted as being concerned with the transfer of
gases and nutrients and the removal of metabolic
waste products. Minute precapillary arterioles and
postcapillary venules are included with the capillaries
as major components of the microcirculation (53).
Terminal arterioles are the final arterial ramifica-
tions, the branchings of which continue as non-
muscular capillary vessels (88). They are further
891
892
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
defined (152) as vessels which have a single layer of
smooth muscle and very little supporting connective
tissue. The "metarteriole," a term introduced by
Chambers & Zweifach (20), is defined by Zweifach
(148) as a primary structural unit which serves as a
framework for the distribution of capillaries.
The term "precapillary sphincter" was first used
by Chambers & Zweifach (20) to designate the
muscular investment at the origin of the outflowing
branches of the preferential channel (distal con-
tinuations of arterial vessels that go directly to the
venous side). These outflowing branches lead into
true capillaries. In more general usage, a precapillary
sphincter is the last smooth muscle cell along any
branch of a terminal arteriole (130).
A capillary may most simply be described as an
endothelial tube devoid of smooth muscle and having
a minimal amount of supporting elements (48). In
descriptions of vascular patterns and flow it is not
necessary to add any qualifications as to size, direc-
tion of flow, or function.
Venules originate at the appearance of the first
smooth muscle cell on a postcapillary vessel. Venules
merge into veins which have a double coat of circular
and longitudinal muscle cells.
Vasoconstriction is the contraction of the smooth
muscle of the vessel wall, vasodilation is relaxation of
the smooth muscle. Vasomotion refers to any active
change in the diameter of blood vessels (81, 89).
There are numerous other designations for vascular
structures, unique to specific organs or tissues, which
will be discussed where they appear in the descriptions
of arteriovenous pathways in various sites.
TECHNIQUES FOR MICROSCOPIC OBSERVATION
OF SMALL BLOOD VESSELS
In discussing the techniques used for microscopic
observation of small blood vessels, the most com-
monly used sites and methods have been included.
There are numerous adaptations of the basic tech-
niques for specific approaches, and also many highly
specialized adaptations for specific areas that will be
described in sections of this chapter where they are
pertinent.
In general, there are four basic methods: /) obser-
vation of tissues and organs in situ illuminated by the
fused quartz rod, 2) exteriorization of internal tissues
or organs which can be spread out as a thin layer for
examination, 3) preparation of tissues using trans-
parent chambers, 4) utilization of superficial structures
which can be seen with direct or transmitted light.
The brevity of the descriptive material should not
mislead the reader as to the difficulty of mastering
the technical problems associated with each method,
nor should he overlook the necessity of being com-
pletely familiar with the characteristics of the site
selected for observation. A survey of the methods
should make it clear that some sites or structures are
more suitable than others for any specific investiga-
tion and should be evaluated on that basis.
Hamster Cheek Pnuch
A method for observation of peripheral circulation
at the microscopic level in the membranous cheek
pouch of the hamster has been developed by Fulton,
Jackson, and Lutz (51, 52, 80). The cheek pouch of
the anesthetized hamster is everted and, when
properly exposed for viewing, forms a flat double-
layered preparation suitable for low power magni-
fication. The pouch is bathed in a 37 C Ringer's
solution. If higher magnifications (200 X to 1200 X)
are used, it is necessary to cut through the upper
layer to form a flap of a single layer. The originators
of the method believe that the cheek pouch is ideally-
suited for investigations on small blood vessels be-
cause the thin membrane presents a normal physio-
logical surface with blood vessels in their usual tissue
environment. A valuable feature of the pouch is also
that the same natural vascular bed can be studied
over long periods and thus changes in circulation or
other characteristics, such as growth of vessels, can be
followed. The pouch is more vascular than the
mesentery of rats or membranes in transparent
chambers. Its vascularity makes tumor transplanta-
tion extremely successful. Other investigations made
on the hamster cheek pouch include the study of
blood pressure, inflammation, hemostasis, petechial
formation, thromboembolism, bacterial and parasitic
infections, drugs, and the vascularization of tumor
transplants.
The disadvantages of the preparation are that the
animal is anesthetized, the exposed tissue must be
irrigated and kept at body temperature, and for
high magnifications the integrity of the vascular bed
is disrupted by surgery needed to obtain a single-
layered membrane. The membranous surface con-
tinuously exudes mucus, which reduces visibility of
the underlying structures. In the hands of its origi-
nators, judging by their excellent films, the method is
very satisfactory for the investigations in which it has
been used.
PATTERNS OF THE A-V PATHWAYS
H<| 1
Transparent Chamber
The development of the transparent chamber
technique and its utilization in numerous tissues have
been extensively reviewed recently (3, 24, 25). For a
detailed description of the methods for installation
and observation, the reader may refer to these papers.
Basically, the method consists of the insertion of a
glass and mica chamber fastened to the cartilage of
the rabbit ear or other applicable site. Since the first
chamber was designed by Sandison (104) in 1924,
several types have evolved, with modifications and
improvements introduced for specific purposes. The
round-table chamber, essentially the same as the
original Sandison model, was introduced in 1930 by
Clark et al. (36). The chamber is constructed to allow
new tissue to grow into an empty space from the
edges of cartilage left by a punctured hole. This
chamber has been used to study the growth and
development of blood vessels, lymphatics, and nerves.
The preformed tissue chamber (36) is one in which
the original tissues can be observed after removal of
the cartilage and skin of the inner side of the ear.
The moat chamber was developed to study the
response of the vessels to various chemical substances
(1, 2). It contains a small space or moat to permit
injection and withdrawal of fluids. The chamber has
been used to investigate absorption, diffusion, and the
reactions of vessels to chemical solutions. A remov-
able-top chamber was designed by Williams (138)
for the purpose of obtaining easy access to living
tissue of the chamber for transplantation of organs or
tissues. The most recent improvements have been
developed by Williams & Roberts (140), who de-
signed a versatile and highly useful chamber which
has the following characteristics: it has a longer life
than any other type of chamber, produces very little
irritation to the ear, is quickly and easily installed,
can be used for transplants of tissue, may be modified
to study existing or preformed vessels, and may be
adapted for the introduction or removal of fluids.
Epidermis, which invades the round-table chamber,
is never seen to grow into this new tantalum and mica
chamber.
Clark (25) points out the many advantages of the
transparent chambers, among them the fact that the
manner of growth and extension of capillaries, the
growth of nerves along arterioles, and the develop-
ment of inflammatory reactions can be observed for
long periods of time in unanesthetized animals. The
disadvantages include injury to the nerves during
installation, the rigidity of the chamber which may-
result in an abnormally high external pressure,
especially with inflammation, and the occurrence of
infection. The advantage of having an unanesthetized
animal is great, and equally helpful is the fact that
the exposed tissues need not be warmed or irrigated
as is the case for visceral or other exteriorized tissues.
A serious disadvantage of the technique is the dis-
ruption of normal circulatory patterns and behavior
by installation of the chamber in which the new
tissue must form.
Fused Quartz Rod
A lengthy discussion of this method of transillumi-
nation of living internal organs in situ for microscopic
study is given by Knisely (70). The limitations and
the applications of the method are fully covered.
The method is based on conducting intense lia;ht
to the structure to be studied by a fused quartz rod.
These rods conduct light around bends and turns by
internal reflection. Overheating and drying of the
tissue is prevented by an isotonic wash solution.
Magnifications from 20 times to 1000 times can be
used.
Transillumination with the fused quartz rod has
been carried out in a wide variety of tissues including
frog skin, tongue, brain, gastrointestinal tract,
stomach, bladder, striated muscle, lung, kidney, and
liver. In mammals, the small vessels in smooth
muscle, mesentery, uterus, spleen, and liver have been
studied.
Knisely feels that the limitations of the method
include the necessity for an anesthetic, surgery, and
the exposure of internal organs to the air. The method
is best used to examine structures at their natural
anatomical surfaces or free edges rather than at cut
surfaces. In examining a thick organ, such as liver or
spleen, one is limited in the degree of magnification
of the deeper structures due to the direct relationship
between the focal length and magnifying power of
lenses. Fulton (49) points out that this procedure does
not reveal certain details of vascular structure or
permit critical discernment of individually formed
elements. It remains, however, the only method
applicable to many types of tissues, but requires
rational selection of the problems to be studied.
Bulbar Conjunctiva
Although observations of the conjunctival vessels are
not new, recent improvements in microscopes and
«94
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
lights have made this site a popular one, especially for
observing changes in vascular patterns and flow in
diseases in humans. Specific instructions for its use
can be found in papers by Bloch (15), Grafflin &
Corddry (56), and Lee (76).
The type of microscopic and lighting equipment,
as well as the position of the patient (supine or up-
right), varies with the investigator, but, generally,
compound microscopes, routine ophthalmological
supports, and lights of moderate intensity constitute
the basic components.
Bloch (13) notes that the walls of the blood vessels
are not clearly seen because of the use of oblique
illumination, although a moving column of blood
can be clearly seen against the white background of
the sclera. Other limitations are that optical resolu-
tion is lost by the patient's inability to hold the eye
still, edema, highlights due to lacrimation, and
excessive abnormal pigmentation in some cases.
Very high magnification is difficult because of move-
ments of the eyeball and the inability of the patient to
tolerate light of high intensity. Also, high power
objectives must be too close to the eyeball if they are
to be in focus.
The advantages of the technique are that blood
vessels in an unanesthetized human can be readily
observed without any surgical intervention or any
preparation to render the vascular beds visible.
Tears supply the proper irrigation for this membra-
nous tissue. An entire vascular field can be studied
again and again in the same subject, and blood flow
can be followed from arteriole through capillary to
venule.
Rat Mesoappendix
The technique for microscopic observation of
mesenteric structures, as described in detail by Zwei-
fach (145), has been used by him in studies on dog
omentum and mesenteric structures in several ani-
mals, but primarily in the cecal mesentery (mesoce-
cum) of the rat.
Preparation of this tissue consists of exteriorizing
the cecum of the anesthetized rat and then spreading
the mesentery, which lies between the cecum and the
terminal ileum, for observation. The mesentery is
continuously irrigated with a warm Ringer's gelatin
solution.
Zweifach (147) believes that the advantages of
using this terminal vascular bed are: a) the accessibil-
ity of the vessels for direct stimulation by mechanical,
chemical, or electrical means; b) clarity of visualiza-
tion; r) minimum interference by surgical procedures
to normal vascular behavior; d) adequate display of
the entire extent of the terminal vascular bed.
The disadvantages include those which apply to
any technique using anesthetized animals subjected
to surgical procedures to expose the tissue for observa-
tion. An idea of the lability of this vascular bed may
be obtained by reading the precautions to be taken in
using the rat mesoappendix for bioassay (1 53).
Bat Wing
Microscopic observations of vascular structures in
the bat wing, a comparatively old technique (68),
was revived by Nicoll & Webb (88) in 1946. A de-
scription of the preparation and current uses may be
found in papers by Webb & Nicoll (130) and Wiede-
man (136).
An unanesthetized animal is slipped into a holder
that allows the wings, lightly held by spring clips, to
be extended over a glass plate. Magnifications up to
2500 times can be used with good resolution.
The simplicity of the preparation is one of its great
advantages, coupled with the elimination of anesthesia
and surgery which permits observation without dis-
turbing the normal circulation or subjecting the
animal to undue stress. The blood vessels and lym-
phatics are accessible for cannulation which permits
perfusion of drugs or measurements of pressure, and
the nerves can be readily stimulated or sectioned. Also,
in this mammal the two wings can be used simulta-
neously, which allows one for control and the other
for experimental procedures.
One undesirable feature is the difficulty in obtaining
bats during the entire year, and, because the animals
will not eat in captivity, their survival time in the
laboratory is limited to a few months. Also, histological
sections are difficult to prepare for study because of
the extreme thinness of the wing.
Utilization of these various techniques has resulted
in the resolution of some old controversies, e.g., the
role of the Rouget cell, and has clarified to some ex-
tent the anatomical structure and physiological func-
tion of terminal vascular beds. It has made many
investigators aware of the danger of ascribing specific
changes in blood pressure or the rate or volume of
blood flow to the activity of small blood vessels, on the
basis of indirect measurements. While the change
in systemic pressure following some experimen-
tal procedure need not be challenged, the means by
which it is brought about may be better explained
PATTERNS OF THE A-V PATHWAYS 895
if direct observations of the vessels controlling
peripheral resistance are employed.
STRUCTURE OF TERMINAL VASCULAR BEDS
From the foregoing section it is apparent that the
microscopic blood vessels which connect the venous
and arterial systems have been studied in a wide
variety of tissues with equal variety in the choice of
experimental animals. Differences in vascular pat-
terns and structural components are to be expected,
but these differences are minor compared to the more
general similarities among the various microcircula-
tory beds. It is this aspect that will be emphasized in
the following descriptions of the microcirculation.
Microcirculation in the Bat Wing
Utilization of the bat wing for studies of the struc-
ture and function of small blood vessels has a long
history. An interesting and detailed report appeared in
1852, written by T. Wharton Jones (68), who de-
scribed the impressive rhythmical vasomotion of the
veins. Scattered publications by other investigators
appeared (18, 63, 87) early in the twentieth century
when new interest in capillary circulation was at its
peak. The interest in the wing of the bat as a site for
microscopic observation of vascular beds was stimu-
lated in 1946, when Nicoll & Webb (88) published
the results of several years of observations. A descrip-
tion of the vessels and their patterns in the terminal
vascular beds of the wing follows.
The major site of peripheral resistance was found
to be in the small arteries which anastomosed to make
interconnected channels or loops. These small arteries,
which arose from the main arterial plexus and formed
arteriolar nets, had the capacity for changing their
lumen size by vasoconstriction. The smaller arterioles
of the nets usually had an inside diameter that was
equal to or smaller than that of a red blood cell.
Nonmuscular capillaries arose as branches of any of
these vessels of the arteriolar plexus, the parent vessel
of the nonmuscular capillary being designated the
terminal arteriole. The muscular coat of the terminal
arteriole became less regular as the vessels advanced
peripherally, as did the number of muscle cells on the
branches arising from it.
The pathways between the arteriolar and venous
plexuses were seen to be similar to those of the rabbit
ear, as described by Clark & Clark (34) and Sandison
(106), with no preferential channel to carry blood
To venule
fig. I. Paths of blood flow in capillary bed in small area of
the bat's wing. [From Nicoll & Webb (88).]
from the arterial to the venous side. Blood was seen
to take alternate routes through the capillary nets.
At times, especially in the terminal arterioles, there
appeared to be a major path of flow through the
capillary vessels to the venules, but this path was seen
to be inconsistent and changed to alternate routes
with modifications in arteriolar or venular circulation
in adjacent regions (see fig. 1 ).
Supravital staining made it possible to study the
arrangement of the vascular smooth muscle of the
various vessels. Arteries had both circular and longi-
tudinal muscle fibers, the latter disappearing in the
arteriolar vessels. The terminal arteriole gradually
lost its circular muscle investment until areas of bare
endothelium could be seen and finally a single coiled
muscle cell formed the precapillary sphincter. The
spiral arrangement of a muscle fiber continued for a
number of turns, presumably reaching a length of over
100 fj. if uncoiled. Postcapillary vessels acquired a
muscular coat in the region of the first valves, and thus
veins were formed. Veins had the usual double layer
of circular and longitudinal muscle fibers.
Because of the small caliber of arterioles and capil-
laries, flow was frequently seen to stop due to obstruc-
tion by a leukocyte. In some instances, an internal
pressure increase would cause the leukocyte to move
on. At other times, the leukocyte could be seen to
896
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
migrate slowly along the vessel wall until it reached a
larger vessel where it was swept forward in the blood
stream. Plugging of small vessels by leukocytes was
found in normal fields with vigorous flow, and this
obstruction determined to some extent the flow of
blood through the capillary nets.
In 1954, Webb & Nicoll (130) discussed the angle
which an arterial branch forms as it leaves its parent
vessel. The downstream angle was found to approach
45 degrees. Webb and Nicoll postulate that this helps
to insure almost equal pressure in the artery and the
branch which arises from it. A similar type of
branching is seen in the arcuate arrangement formed
by arterioles. Arterioles, however, usually leave ar-
teries at right angles. The arteriolar branches have
sphincters at their point of origin that regulate the
size of the lumen of the branching vessel as it leaves
the artery.
The arcuate system formed by the arterioles affords
collateral pathways and contributes to a uniform
distribution of blood at uniform pressure within the
capillaries.
The capillaries form an extensive anastomosing net
which is supplied by terminal arterioles arranged in
such a manner that no capillary net is very far away
from its arteriolar supply.
Active vasomotion of arteriolar vessels is, according
to Webb and Nicoll, the principal factor of a local
nature that regulates blood flow and blood pressure in
the capillary beds. The activity of the muscular wall
of the arterioles, which constitutes active vasomotion,
is independent of central nervous control. Degenera-
tion of nerves supplying an area does not affect active
vasomotion in the smaller arterioles, nor does stimu-
lation of intact nerves, although this does result in a
contractile response from the arteries or the larger
arterioles.
Blood flow and blood pressure in capillary nets,
then, are controlled by two factors, one being the
anatomical arrangement of the arterioles which form
arcades, and the second being the active vasomotion
of the arterioles which is determined by local condi-
tions.
Further discussion of the arcuate patterns formed
by arterioles in the bat wing appeared in a report by
Nicoll & Webb (89) in 1955. Arteriolar vessels form
arcuate configurations. These anastomosing vessels
are approximately equal in diameter. Several distinct
arteriolar arcuate systems can be identified arising
from either an artery or a large arteriole. Two
characteristic features were found in the manner in
*
I
I
fig. 2. Enlargement of an arteriolar branch at its point of
origin. Bat wing. X875.
which the arcuate systems began. One was the angle
of origin of the arteriolar vessels lrom the parent vessel,
and this was found to be 90 degrees or less in reference
to the forward direction of flow in the parent vessel.
The second characteristic feature is a dilatation or
enlargement of the arteriolar branch at its point of
origin compared to its diameter throughout its length.
Also, the inside diameter of the opening between the
parent vessel and branch is much smaller than the
average inside diameter of the branch (fig. 2). This
formation, described by Nicoll and Webb in the bat
wing and named "Indian Club," has not been de-
scribed in microscopic vessels in other terminal vascu-
lar beds. In view of the fact that the notable appear-
ance of the enlargement of a vessel at its junction
depends to some degree on tonus, it may not be
readily apparent in anesthetized animals in which
vessel tone is low. If the tonus of the branch is quite
low, there may be little or no apparent difference be-
PATTERNS OF THE A-V PATHWAYS
897
fig. 3. Arcuate patterns in the terminal vascular bed. [From
Nicoll & Webb (89).]
tween the outside diameter of the branch at its
junction and along its length. Terminal arterioles
originate mainly from the smallest arcuate vessels,
but may also arise from any of the arcuate arterioles
or a small artery (see fig. 3).
The capillaries form extensive nets, and the distri-
bution of blood within the nets from any particular
terminal arteriole is limited. Local conditions, which
must be considered to be a major factor of control,
constantly change the paths of blood flow through
the capillary bed.
Venous vessels show an arcuate pattern that
roughly follows that of the arteriolar vessels. At the
point where a capillary vessel joins a venule, a valve
may often be seen, although in many instances no
such structure is evident. Nicoll and Webb suggest
that since the muscular coat of the venule begins in
the immediate vicinity of the valve, this site may be
considered as the true junction between capillary
and venule.
The flow of blood through the capillary nets is con-
trolled chiefly by activity of the terminal arterioles.
When they are dilated, flow is rapid and continuous
in the capillary nets. Constriction of a terminal
arteriole necessarily stops the flow of blood through
the capillary vessels supplied by it. When the numer-
ous terminal arterioles which supply an interconnec-
tive network of capillary vessels are contracting and
relaxing intermittently and aphasically, the flow of
blood into collecting venules may be continuous. Ces-
sation of flow from venous capillaries into venules is
often produced when resistance to inflow is met be-
cause of a closed valve at the junction of the two
converging vessels. Forward flow is seen on opening
of the valve.
Nicoll and Webb offer several features of both the
anatomical arrangement and the behavior of vessels
in terminal vascular beds as the regulators of blood
flow and blood pressure at this level, a) The arcuate
pattern of arterioles provides a means for intrinsic
regulation of flow and pressure. The roughly concen-
tric organization of the arcuate systems, made up of
anastomosing vessels of the same size, serve as volume
reservoirs for capillaries. Such an arrangement assures
an adequate blood supply for capillary nets which
does not fluctuate widely with changes in flow and
pressure in single arterial vessels. The authors con-
sider such an arrangement to be necessary in a system
in which the demand for blood varies and in which
some of the distributing vessels are distensible, thus
allowing increases in pressure to be absorbed in the
stretched vessels rather than to contribute to increased
flow, b) The angle of origin formed by an arteriole in
reference to its parent vessel affords a means by
which pressure may be abruptly reduced. Also, this
manner of branching off at a 90-degree angle or more
assures an adequate pressure head for each outlet from
a given vessel. This arrangement, coupled with the
fact that a capillary bed receives blood from several
terminal arterioles, results in equal pressure in all
capillaries regardless of their distance from their
arterial supply. Capillary pressure, sufficient for
proper function, can be maintained with minimal
arterial pressure, c) Nicoll and Webb believe that the
Indian Club formation at the arteriolar origins is
most important in pressure regulation. The actual
size of the orifice of each arteriole aids in reducing
pressure from artery to arteriole. The variability in
the size of the orifice, which depends on contraction or
relaxation of the muscle cells which form it, adds
another means of control of pressure in small arteries
and arterioles. It is possible that the contraction and
relaxation of the muscular elements at arteriolar
origins is determined by intra-arterial pressure. This
myogenic response would afford another intrinsic
mechanism whereby the pressure and flow through
capillary nets could be kept at a constant level inde-
pendent of wide variations among these values in
arterial vessels, d) Neural control of larger arteries
does not seem to be important in the regulation of
capillary blood flow, e) Active vasomotion in the
terminal arterioles causes blood flow through capillary
nets to alternate between very vigorous flow and no
flow at all. Local conditions determine the degree and
898
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
extent of contractile activity of the smooth muscle
cells which encircle the terminal arterioles, and there-
fore local conditions can be responsible for controlling
capillary flow to meet the requirements of the tissues
in the immediate environment.
Microcirculation in the Rabbit Ear
Collection of new and important data on mam-
malian small blood vessels began in 1924, following
the introduction of the transparent chamber technique
by Sandison (104). He reported (106) observations on
circulation in the rabbit ear primarily concerned with
contractility of small blood vessels. Local control of
blood flow was seen to reside in the smooth muscle
cells which developed on newly formed capillaries as
they were transformed into arterioles. In observing
circulation of blood through the vessels which formed
in the chamber, Sandison saw an axial stream of cells
surrounded by a narrow, clear plasma layer. Leuko-
cytes were thrown into the peripheral layer of plasma
and slowly rolled along the vessel wall. An uneven
mixture of blood cells and plasma was observed during
sluggish or irregular flow through capillary nets, this
type of flow resulting from the aphasic and independ-
ent contraction of arterioles which causes blood to be
fed to the veins through capillaries and venules in a
broken stream. "Plasma skimming" was seen mainly
in partially contracted vessels or in capillaries con-
necting two vessels and in which there was no circula-
tion due to equal pressure at each end of the con-
necting capillary. In a capillary loop, the two ends of
which were connected to a larger vessel, plasma flow
(indicated by the passage of blood platelets) would
continue in the absence of circulation of blood cells.
An increase in the blood supply to the larger, parent
vessel often caused red blood cells and leukocytes to
be forced through the capillary loop. The blood cells
were often seen to take long narrow shapes as they
were forced through the constricted entrance to the
capillary loop.
Although capillary circulation was almost entirely
regulated by contraction of the arterial vessels sup-
plying the capillary plexus, flow was seen to be slowed
or even stopped by a single leukocyte caught in a
constricted portion of a vessel. One of the most
favorable places for plugging by a leukocyte was
found to be at the origin of the small arterioles from
their arteries. This region was normally partly con-
stricted because of the bulging of endothelial cells into
the lumen of the vessel. This site bears a close re-
semblance to the Indian Club structure described by
Cap.
fig. 4. Camera-lucida drawing of a precapillary branch of
an artery. Muse. = muscle cells; End. nucl. = endothelial
nucleus; Adv. = adventitial cell; Cap. = capillary. [From
Sandison (106).]
Nicoll & Webb (89) (see fig. 4). The leukocytes were
dislodged by an increase in force of the blood stream
or by the ameboid activity of the leukocytes. A similar
occurrence was seen by Nicoll & Webb (88) in blood
flow through comparable vessels in the bat wing.
Clark & Clark (29), in the same year, reported on
the behavior of microscopic vessels seen in the rabbit
ear using a "preformed-tissue" chamber, one in which
the original structures were present as opposed to
newly formed vessels and nerves seen in the first
studies using the transparent chamber. The Clarks
were impressed with the contraction of arteries and
arterioles; spontaneous rhythmical contractions as
well as contractions in response to artificial stimulation
(mechanical, tactile, or auditory). Spontaneous rhyth-
mical contraction was seen to play an important role
in regulation of blood flow, causing changes in the
distribution of blood to different capillary areas and
causing continuous alterations in the direction of flow.
Contractions of arterial vessels were found to be
varied. Contraction of the main artery reduced the
blood flow to the whole area, but the distribution of
blood to different portions was dependent on contrac-
tions of different arterial branches, each at a different
tempo and independent of the contraction of the main
artery and of each other. An arteriole might contract
to complete closure and thus cut off blood to the
capillaries it supplied while an adjacent vessel, a
branch from the same artery, would remain open to
allow rapid passage of blood. Arteriovenous anasto-
moses were seen to contract actively and so influence
the distribution of blood. Contractions were seen to
decrease in animals that were asleep or anesthetized.
Further studies on the activity of arterial vessels,
PATTERNS OF THE A-V PATHWAYS
899
including arteriovenous anastomoses, appeared in
1934 (30). Clark and Clark again described the
fluctuation in rate and amount of blood flow through
any given vessel, as well as the frequent reversals in
the direction of flow. A single capillary or venule was
seen to have an abundant flow of blood in one direc-
tion and a few seconds later an equally great flow in
the opposite direction. The variation in flow included
scanty flow of a few blood cells, or plasma and plate-
lets only, or stasis, or complete emptying. Such
changes were brought about by periodic active con-
tractions of arterial vessels or portions of arterial
vessels. The numerous thick-walled arteriovenous
anastomoses were most conspicuous for their active
contractility. Their contractions were usually more
frequent, quicker, and more powerful than those of
the arteries, and their effect on venous circulation was
more sudden. Definite active contraction of veins was
reported to occur near the point of entrance of a
cluster of arteriovenous anastomoses.
Clark & Clark (32) studied the growth of capillaries
into a transparent chamber and found that new
capillaries arose as endothelial outgrowths from vascu-
lar endothelium. They advanced as blindly ending
sprouts, connecting with neighboring sprouts to form
loops, and continued to advance as a plexus with a
growing edge of new sprouts. The growing vascular
network showed differentiation of vessels in the older
portions of the first-formed capillary plexus and many
of the capillaries were seen to retract and disappear.
An entire chamber was revascularized in 2 or 3 weeks
with further differentiation continuing through en-
largement of new arterioles which were receiving a
large blood supply and widening of venules draining
large amounts of blood. There was a further reduction
in surplus capillaries. After a few days, the vascular
pattern was relatively stable.
The Clarks next directed their attention to the
development of extra-endothelial cells on the walls of
peripheral blood vessels (33). Three months after
vessels had regenerated it was found that venules were
wider than capillaries, both vessels having similar
walls, while arterioles were as narrow as capillaries
and narrower than venules. The walls of the arterioles
differed in number and arrangement, and in the
form of the extra-endothelial cells. Blood flow was
seen to be steady and rapid in arteries and arterioles,
steady and slower in veins, and slow with frequent
hesitations and reversals in capillaries. Circulation in
capillaries was variable, with intervals of steady flow-
being interspersed with periods of stasis, plasma
skimming, or absence of flow during which the vessels
remained open and were filled with plasma. The
subsequent fate of the extra-endothelial cells depended
on the fate of the vessel on which they appeared. If
the vessel remained a capillary, they were occasionally
seen to increase in number by mitotic division or to
retain the same number. The cells were inert. If the
capillaries became parts of venules, the adventitial
cells increased in number, retained their longitudinal
arrangement, and remained inert. The change of a
capillary to an arteriole involved straightening of the
vessel, loss of side branches, narrowing of caliber, and
an increase in thickness of the endothelium. There
was a rapid increase in the number of extra-endothelial
cells which assumed a transverse position. Definite
active contractility was seen to develop in these cells
which became smooth muscle cells, providing they
were reached by a regenerating vasomotor nerve.
The caliber changes in minute vessels were dis-
cussed by Clark & Clark (34) in 1943. In earlier
published studies, the attention of the authors had
been on the main arteries, their branches, the arterio-
venous anastomoses, and the larger veins. Observa-
tions on newly formed arteries indicated that the
number of arteries which developed contractility, the
rate at which contractility appeared, and its final
extent on individual vessels and their branches de-
pended on the rate and extent of growth of new vaso-
motor nerves. Terminal arterioles in original vascular
beds in the preformed type of chamber were seen to
show^ spontaneous contractions which in most cases
obliterated the lumen. These vessels could sometimes
be made to contract by prodding the animal, but
their behavior was erratic. A terminal arteriole was
seen to divide immediately beyond its last muscle cell
into a capillary plexus. In some instances a terminal
arteriole was prolonged for a distance beyond the
point of the final muscle cells before forming a capil-
lary plexus. Such vessels had longitudinally arranged
adventitial cells rather than muscle cells on their walls.
Except for this, they had the characteristics of ar-
terioles, being straight, uniformly narrow, and having
a relatively thick endothelium. The region of active
contraction was confined to the portion of the vessel
which had smooth muscle cells, but the distal portion
at times showed a narrowing, with protrusions of
endothelial nuclei into the lumen, after blood flow
was shut off by active contraction of the proximal
portion of the vessel. The vessel showed an increase in
caliber following increased blood flow through it.
The Clarks refer to these vessels as arterial capillaries.
goo
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Microcirculation in the Mesentery
Early descriptions of mesenteric circulation patterns
by Zweifach (143) and Chambers & Zweifach (20)
dealt primarily with establishing a structural and
functional unit, the preferential or thoroughfare
channel, which was thought by the authors at that
time to be a representative structure of terminal
vascular beds. A discussion of the vascular compo-
nents of the mesenteric circulation and blood flow
through them appeared in 1954 (147)- Zweifach
stated that the mesentery represented a simplified
vascular structure devoid of ancillary features peculiar
to specific organs. In observing normal circulation in
the rat mesoappenctix (cecal mesentery), Zweifach
found the larger arteries along one side of the mesen-
tery to be about one-third as large in diameter as
their paired veins. The depth of anesthesia influenced
the caliber of these vessels, deep anesthesia causing
them to dilate until both vessels had the same diame-
ter. Respiratory difficulties caused venous constriction.
Terminal arterioles in the mesentery proper had a very
rapid flow of blood. Collecting veins had a steady flow
of blood with continuous forward flow without cessa-
tion or temporary reversal. Capillary circulation,
however, showed intermittent flow produced by the
contraction of precapillary sphincters, the activity of
which was irregular and unpredictable. Preferential
channels were found to be unusually prominent in
the mesentery. The most important structural compo-
nent for regulating capillary blood flow was the pre-
capillary sphincter. The precapillary sphincter was
found at the junction of all offshoots of the muscu-
lar components of the vascular bed. The true capillary
network was made up of endothelial tubes with no
perivascular muscle cells. In some areas the collecting
venules were formed by the joining of several side
branches leading from precapillary sphincters. Both
terminal arterioles and venules were seen to be inter-
connected to form a series of arcades, so extensive in
some cases that they completely circumscribed the
capillary bed. The metarterioles originated as off-
shoots of the arteriolar arcades, extended toward the
center of the tissue distributing typical precapillary
branches. The arteriolar channels terminated as one
or two short capillaries which fed directly into a
venous vessel. Zweifach expresses the opinion that the
primary mechanisms which readjust circulation
through the capillary bed are essentially of a humoral
nature. Neurogenic mechanisms, local metabolic fac-
tors, and blood-borne substances from organs con-
tribute to the local regulation.
Microcirculation in the Hamster Cheek Pouch
The use of the hamster cheek pouch for microscopic
study of the peripheral circulation was introduced by
Fulton et al. (50, 51).
Although there is no detailed description by these
authors of the basic vascular pattern that is seen in
this mucous membrane, the literature contains refer-
ences to the presence or absence of various vascular
structures which will be presented here.
The cheek pouch is exceedingly vascular compared
to rat mesentery or membranes in transparent cham-
bers. The pattern differs also from the mesentery in
that no preferential channels have been found. A rich
network of anastomoses between venous vessels and
arteriolar vessels is present. Arterioles, which supply
the capillary network, bifurcate progressively into
branches of equal significance for the distribution of
blood (fig. 5). The arterioles exhibit spontaneous
vasomotion (80, 82). Lutz & Fulton (81) state that
precapillary sphincters were seen to contract inde-
pendently of adjacent smooth muscle in cheek pouch
vessels. Intermittent flow from small veins was also
seen, but no venous sphincters were identified.
Lutz & Fulton (81) point out that there is always
variation in the demand for blood by the organs, and
this variable demand can be satisfied by vasomotor
responses without involving the heart or other large
structures. The complex anastomosing system of ves-
sels in the cheek pouch, for instance, coupled with
vasomotion, permits changes in flow. Neither the
vessel wall nor the flow are ever quiescent, the most
striking feature of the small vessels being their con-
stant activity.
More vein-to-vein than artery-to-artery anasto-
moses are seen in the hamster cheek pouch. Venules
make up the greatest amount of endothelial surface
and contain the greatest proportion of circulating
blood at any one time. Lutz and Fulton believe that
60 to 70 per cent of the peripheral circulating blood
is in the venous vessels.
Poor & Lutz (97) studied the functional anastomotic
vessels in the cheek pouch and reported that artery-
to-artery anastomoses were generally one-third to
one-half the size of the parent arteriole. These were
outnumbered by the vein-to-vein anastomoses. The
venous anastomoses were nearly the size of the veins
which they connected (fig. 6).
Microcirculation in Skeletal Muscle
The description of the distribution of minute vessels
in skeletal muscle has not changed to anv marked
PATTERNS OF THE A-V PATHWAYS 90 1
-Ut
1 'iV-"' • ' ^AI
V \ 1 \
v
■vfV \ 1
fig. 5. Vascular pattern of the hamster cheek pouch.
(Courtesy of Dr. E. P. Fowler, Jr.)
degree in the last eight decades. The early information
comes from studies of injected and fixed material, and
in recent years there have been investigations using
microscopic techniques on living animals.
Krogh (73) reviews the work of Spalteholz (116),
who depicted the vascular arrangement as follows:
freely branching arteries with numerous anastomoses
between the branches form a primary network which
in turn gives off anastomosing small arteries that form
a second network. Arterioles branch from this net-
work, usually at right angles to muscle fibers at
regular intervals. The arterioles then split up into a
large number of capillaries which run along parallel
to the muscle fibers with numerous anastomoses. The
capillaries unite into venules. The pattern of the
venous system is almost exactly that of the arterial
systems.
Clark (37) and Walls (126) state that skeletal
muscle, which is highly vascular, is supplied by
branches from neighboring arteries which invade the
epimysium and travel into the perimysium, dividing
fig. 6. Vascular network of hamster cheek pouch near
buccal end, lead chromate injection. [From Poor (97).]
as they do so. Various branches of the vessels entering
the perimysium anastomose with one another. The
finer branches lie transversely to the long axes of the
muscle fibers and give rise to the capillaries which run
parallel to the muscle fibers. These parallel capil-
laries lie in the endomysium. This, then, is the anatom-
ical sequence: arteries and veins run together until
terminal arterioles and venules are reached. The
terminal arterioles and venules then come ofF of the
parent vessels in alternate sequence. The capillaries,
running longitudinally between muscle fibers, are
connected frequently by transverse vessels which run
over or under the intervening fibers and thus form a
fine capillary network of tiny oblong meshes.
Zweifach & Metz (151, 152) have studied the
vascular supply of the spinotrapezius muscle in the
rat. Their observations were primarily of vessels in the
epimysium and the perimysium. They found two
distinct components in the capillary circulation of
muscle bundles, /) a vascular bed that was distributed
along the natural cleavage planes in the connective
tissue sheath that binds collections of muscle bundles
together, and 2) a second capillary network originated
by short muscular arterioles which penetrate into the
902
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
depth of the muscle proper and terminate by
branching into numerous capillaries.
There are impressive numbers of anastomoses be-
tween both arterial and venous vessels which form a
series of arcades. Direct anastomoses between ar-
terioles and venules are also found.
The capillary bed of the perimysium is supplied by
metarterioles which come off at right angles from the
arterial or arteriolar arcades. These metarterioles
terminate as one or two capillaries which unite with
other capillaries to form venous effluent vessels. The
capillaries lie directly on the surface of the small
muscle bundles, thus each muscle bundle is sur-
rounded by a network of arterial and venous vessels
which interconnect freely with one another within
the connective tissue separating the bundles. The
muscle fibers are supplied with blood by branches
from the arteriolar arcades which penetrate the con-
nective tissue and give rise to capillaries which run
along the length of the muscle.
Zweifach and Metz report the presence of metar-
terioles along the free margins of the skeletal muscle
which can be traced directly to the venous system.
These vessels, they believe, represent preferential
pathways which convey the most rapid stream of
blood from the arterial to the venous side.
In addition to structural features, spontaneous
vasomotor changes were seen by Zweifach and Metz in
arterial and venous vessels. The vasoconstriction was
not often intense enough to stop blood flow through
the vessels involved, except at the level of the pre-
capillary sphincters.
In investigations of red and white skeletal muscle
in rabbits (75, 113) injected preparations showed
arterial vessels which branched profusely to end in
capillaries running parallel to muscle fibers. Also
shown were numerous anastomotic connections be-
tween small vessels.
Algire (4) and Algire & Merwin (5) studied the
panniculus carnosus through a transparent chamber
in the rat's back and also saw many arterial anasto-
moses as well as arteriovenous anastomoses. Arterial
branches from the subcutaneous layer supplied the
thin striated muscle layer with blood. The arterioles
from these branches subdivided into capillaries which
ran parallel to the muscle fibers, with cross connec-
tions between them, joining other capillaries to form
collecting venules.
The capillary blood flow was noted to be inter-
mittent, the result of active vasomotion of the ar-
terioles. Algire & Merwin (5) estimated the length of
capillaries that were seen between the muscle fibers
fig. 7. Capillary vessels in skeletal muscle.
to be between 0.3 and 1.0 mm, with anastomoses oc-
curring at intervals of about o. 1 mm.
Observations of the capillary network of the skeletal
muscle bands that course through the bat wing show a
vascular pattern similar to the descriptions given
above (see fig. 7).
There are notable differences between the distribu-
tion of capillaries in the endomysium and that in the
areas adjacent to the skeletal muscle. The vessels
which run parallel to the muscle fibers are generally
longer and straighter than comparable vessels in the
surrounding connective tissue. An arteriolar branch
that crosses the muscle fibers often subdivides into
two capillaries that originate at right angles to the
parent vessel and go off in opposite directions. Thev
usually do not lie in the same plane, one going deep
between the fibers, occasionally until lost from view,
while the other vessel continues on the upper surface.
As a result of this downward, or sometimes upward,
turn it is possible to look directly down into the lumen
of a capillary rather than at the customary longitudi-
nal view. An arteriole with its accompanying venule
mav cross the muscle band without either of the vessels
PATTERNS OF THE A-V PATHWAYS
9°3
fig. 8. Red blood cells "on edge" in a capillary of skeletal
muscle.
giving off branches to contribute or to receive blood
from the underlying muscle. When capillary vessels
running parallel to the muscle fibers are confined to
the narrow space between two fibers, the capillaries
are flattened and the cells face the fibers (fig. 8), as
reported by Reynolds et al. (ioo) in similar vessels in
the myocardium. In a wider space, the cells are often
seen broadside.
Anastomoses between the capillary vessels running
parallel to the muscle fibers are numerous, the con-
nections occurring sometimes between adjacent ves-
sels and just as frequently with vessels lying some dis-
tance away. Short connections between arterial and
venous pathways are also seen.
Intermittent flow occurs in the capillaries of these
skeletal muscle bands as a result of spontaneous closure
of short duration of terminal arterioles that give rise
to the capillaries which lie in the endomysium.
Studies made thus far in these beds have established
no characteristic pattern formed by vessels supplying
the muscle bands that deviates from what is normallv
seen elsewhere, except for the parallel course of the
vessels lying between the muscle fibers.
Microcirculation in Myocardium
There is a paucity of descriptive literature on the
capillary beds in the myocardium. Although a few-
investigations on fixed material appear, no studies
have been made on circulation through the minute
vessels in the living animal, presumably because of
the difficulty of microscopic observations of an organ
in motion.
In 1928, Wearn (127) studied sections of myocar-
dium obtained from man, cats, and rabbits. The ves-
sels were filled with material injected through coro-
nary arteries. Wearn observed that almost every
cardiac muscle fiber was in direct contact with one
capillary and some fibers were touched by two or
more. A muscle fiber was completely surrounded in
some instances due to numerous anastomoses between
capillaries. These interconnecting branches ran across
the parallel muscle fibers. Capillary vessels were
found to lie between the cardiac muscle fibers and
did not actually penetrate the muscle substance.
Saunders & Knisely (107) reported having watched
through a microscope the circulation in the myocar-
dium of beating frog hearts. Blood flow was seen to
stop during systole and to flow profusely during
diastole. The cessation of flow in systole was brought
about by compression of the small vessels by the
contracted myocardial fibers.
Reynolds et al. (100) studied fixed sections of heart
muscle, the hearts having been taken from dogs
without loss of blood from the coronary vessels. They
report that capillaries had a diameter of approxi-
mately 4 m- The capillaries were seen to run along the
muscle fibers as described by Wearn (127), about one
capillary to every muscle cell. The orientation of the
red blood cells within the capillary vessels was be-
lieved to be unusual, in that the cells were often
seen edgewise, i.e., with the flat surface of the red
cell facing the parallel myocardial fibers. The authors
conclude that the capillary' vessels running between
the muscle fibers are elliptical in cross section rather
than round.
In normal hearts, more capillaries were found in
the epicardium than in the middle portion or the
endocardium. Various explanations, none conclusive,
were given for this.
Terminal arterioles were identified by the presence
of an endothelium with distinguishable smooth muscle
cells along their walls. The terminal arteriole gave rise
9°4
HANDBOOK OF PHYSIOI.n<;V
CIRCX'LATION II
to a number of capillaries that ran parallel to each
other in the same direction as their parent vessel.
A postcapillary venule was formed by the union of
capillaries which came from opposite directions along
the muscle fibers. The postcapillary venule increased
in size as it was joined by similar tributaries. These
tributaries formed venules, which were identified by
their muscular walls.
Provenza & Scherlis (99) studied sections made
from dog hearts and placed great emphasis on the
appearance of "muscle sphincters" in various small
vessels. Although the authors have used the terminol-
ogy of Chambers & Zweifach (20), it has not in every
instance been properly applied, and comparison
with other terminal vascular beds is difficult. A highly
imaginative diagram indicates the presence of arterio-
venous anastomoses, metarterioles, thoroughfare
channels, and precapillaries.
Microcirculation in Skin
Zweifach (149) has presented a description of the
cutaneous circulation in a flap of skin of the rat from
which the connective tissue had been cleaned off.
A network of arterial vessels in the connective tissue
between the skin and underlying muscle gives rise to
small arteries which enter the dermis. These small
arteries, as well as the ones from which they originate,
form a regular pattern of interconnecting links or
arcades. The capillary bed of the dermis is composed
of a secondary network lying between the inter-
arcading arterioles. This secondary network is formed
by precapillary and capillary vessels that are branches
of the interarcading arterioles.
Blood flows away from the capillary bed in wide
vessels, which join to form collecting venules. The
collecting venules form an interconnecting plexus that
is similar to, but more extensive than, the arterial
plexus. Many short arteriolar branches are seen to go
directly into the vessels of the venous plexus. Also
seen are direct connections between arterial and
venous arcades that allow blood to go from arterial
to venous side without going through a capillary net-
work. The venous vessels form the major portion of
the cutaneous vascular beds.
Zweifach believes that the branches which leave
the arterial arcades are structurally similar to metar-
terioles in that they have a thin layer of smooth
muscle and a comparatively straight course. The
vessel finally becomes part of the capillary bed after
giving off branches along its course. These offshoots,
or side branches, are precapillary vessels, having
spirally arranged muscle cells in the immediate junc-
tional region. The precapillary vessels show character-
istic spontaneous vasomotion. Yasomotion is also seen
in the deeper lying arterioles. The small venules of
the cutaneous bed show a continuous almost rhyth-
mic pattern of spontaneous activity that is unrelated
to the vasomotion of the deeper lying vessels.
The arteriolar arcades were found to be very re-
sponsive to constrictor and dilator agents. The venous
arcades showed a 20-fold increase in responsiveness to
epinephrine when the temperature was made to fall 1
or 2 degrees, indicating that they are greatly influ-
enced by temperature change. Zweifach considers the
venous network in the skin to be unique in this regard.
From his studies, Zweifach concluded that the
structural pattern of the cutaneous circulation was
atypical, since it was composed predominantly of
highly reactive venous vessels. The circulation in the
skin appeared to be regulated locally by tissue
mediators.
The description of the cutaneous vascular pattern
and its vasomotion conforms in most respects to that
of other terminal vascular beds that have been studied,
with the possible exception of the mesentery and the
omentum. The interconnecting arcades of both arte-
rial and venous vessels with a secondary network
forming the capillary bed are prominent features of
the pattern of small blood vessels in the hamster
cheek pouch and the bat wing. Such an arrangement
seems to be a common denominator in vascular
patterns of the microcirculation.
Microcirculation in Stomach and Intestine
Until a recent paper by Baez (6), descriptions
dealing with the vascular patterns of small blood
vessels in the stomach and intestine have been based
on injected and fixed material, ft is extremely difficult
to establish the paths of blood flow in a tissue
without observing the flow in living material. This
would apply especially in such a vast network of
venules and arterioles which intercommunicate so
freely by a system of arcades as is present in the muscu-
lature of the gut. Although the early investigations
briefly discussed here are not concerned with the
smallest vessels, they will serve as a background for a
more detailed description of the terminal vascular
beds of the alimentary canal.
Xoer (90) studied the vascular patterns in the
jejunum and ileum of specimens prepared by liquid
latex injections. The descriptions are of the mesenteric
circulation and the superficial vessels of the gut wall
PATTERNS OF THE A-V PATHWAYS
905
fig. 9. Types of antimesenteric
anastomoses. [From Noer (90).]
(the mural trunks). In his search for an experimental
animal which might have a vascular distribution
similar to man, he observed 14 different animals. The
basic architecture was found to be similar in all
animals in that the intestinal arteries formed mesen-
teric arcades or arches which in turn gave rise to vasa
recta, which then proceeded to the intestinal wall to
form mural trunks. Striking variations in the numbers
of mesenteric arcades were found among the species
as well as differences in the pattern of the vasa recta,
including their length and whether or not they had
anastomotic connections with one another. The mural
trunks in the human stomach were found to ramify in
two ways, a similar arrangement being seen in other
animals. A single vessel passing to the antimesenteric
area might give off lateral branches along the way,
or the vessel might break up into several branches
rather quickly and subdivide in an arboreal fashion.
Three types of anastomoses between the mural trunks
in the antimesenteric area were found to be /) direct
communication between the mural trunks of the two
sides, 2) a plexiform arrangement, 3) short vessels
joining arcuate mural anastomoses (fig. 9). Veins
were found to follow the arteries with few exceptions.
Although Noer was not the first to describe the
arcuate patterns found in the arterial and venous
vessels of the intestine, his report is extensive and de-
tailed, and contains a comprehensive review of the
literature up to that time.
Investigations of the alimentary tract during the
next few years centered around the absence or
presence of arteriovenous anastomoses, especially in
the human stomach. Barclay & Bentley(7), stimulated
by the findings of Trueta et al. ( 1 20) of vascular shunts
in the kidney, proposed that in the wall of the stomach
there were arteriovenous anastomoses in the region of
the submucous plexus, and that when these arterio-
venous anastomoses were open, active circulation
through the vessels of the mucous membrane was
excluded. Their conclusions were based on the absence
of radiopaque material in the mucous membrane
of stomach injected immediately after surgical re-
moval. They suggest that the injected material flowed
from arteries of the submucosal plexus to the gastric
veins directly through a shunt located in the sub-
mucous plexus. In 1952, Walder (125), accepting the
presence of arteriovenous anastomoses in the sub-
mucous layer of the human stomach after what seemed
to be confirmation of them by Barlow (8) through
microdissection, carried out investigations to deter-
mine their function, size, and responses to stimuli,
both physical and pharmacological. Cannulation of
the right gastroepiploic artery and its accompanying
vein permitted him to introduce glass beads, 40 to 200
11 in diameter, into the artery and to recover them in
the venous outflow. The presence of spheres, 1 40 n in
diameter, in the venous outflow was believed to be
indicative of patent arteriovenous anastomoses, be-
cause spheres of this size could not travel through
the capillary network. The results of the injection of
drugs, nerve stimulation, and varying perfusion pres-
sures to determine their influence on the size of the
arteriovenous anastomoses were inconclusive.
In an extensive study of the vascular patterns in the
alimentary canal, Barlow (8), in describing the arte-
rial supply to various portions of the stomach, notes
frequent anastomoses of the arteries in the submucous
plexus and the mucosa. The mucosal arteries give
rise to capillaries which also have anastomotic connec-
tions. He found arteriovenous anastomoses in the
stomach which consisted of an arterial end, variable in
length, a short narrow junction area, and a short
wide venous channel. This structure was demon-
strated by Barlow's double injection technique. The
arterial end may be a direct branch of a mucosal
artery or arise from a main channel in the submucous
plexus. It terminates by joining either a distant mu-
cosal vein or may double back on itself and anasto-
go6 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
fig. io. Arteriovenous anas-
tomosis in the submucous plexus.
[From Barlow (8).]
mose with a tributary of its accompanying vein (fig.
io).
In 1959, Baez (6), using the small intestine of the
rat, gave a detailed account of the vasculature at
various levels based on observations in the living
animal. In discussing the method of observations,
Baez points out the advantage of having an intact
animal in which the distribution of large supplying
arteries and accompanying veins of the wall as well
as their relation to the small vasa recta can be deter-
mined. The type of a vascular connection at the anti-
mesenteric border as well as the vessels of the submu-
cosal plexus and the final ramifications of the vessels
to the muscular coat can be established.
The description of the vessels in the submucosal
plexus is as follows: Main arteries (60-80 n), which
pierce the muscularis in the mesenteric region at
intervals, divide into two or three branches in the
submucosa. Each branch subdivides into four or six
smaller branches (30-40 //) which then proceed to the
antimesenteric border where they connect with simi-
lar arterial branches from the other side. Other arte-
rial anastomoses are formed by interconnections of
branches between neighboring recta. Baez uses a
general description for main arterial arcades with
three characteristics: /) they are located in the outer
plane of the submucosa and average 30 to 40 m in
diameter; 2) all give rise to secondary arcades, muco-
sal arteries, and vessels to nourish the muscular coat;
3) the direction of blood flow through them is
changing constantly.
The small vasa recta which arise from the last
mesenteric arcade terminate quickly by anastomosing
with secondary branches of other arcades. The small
vasa recta lie between the large vasa recta. They sup-
ply the vessels to the submucosa and muscular coats
of the gut wall near the mesenteric border (fig. 11).
These vessels go in opposite directions, some to the
outer smooth muscle coat and others to the inner
absorptive surface of the gastrointestinal tract. They
seem to have a unique type of blood flow. Unidirec-
tional flow in the large and small vasa recta is altered
in the meshwork of interconnected arterial vessels in
the submucosal plexus. A main arterial arcade may
show complete reversal of flow or, as more frequently
happens, blood may flow from both sides of an arterial
arcade into a mucosal artery. At times, when blood
is rushing into an arcade from opposite directions,
the converging currents may produce a space of
clear plasma at the point where they meet. When this
occurs at the origin of a mucosal vessel, plasma is
"skimmed" into it.
The muscular coat of the ileum is supplied by
vessels that originate from the proximal end of mu-
cosal arteries or from secondary arcades in the sub-
mucosa. Baez considers these vessels to be metar-
terioles, 18 to 24 n in diameter, which enter the
muscularis and run in the plane of cleavage between
the circular and longitudinal muscle bundles. In
the intermuscular septum the capillaries for the
circular muscle bundles stay on the same plane as the
parent vessel, while those for the longitudinal muscle
bundles turn outward. The capillaries communicate
freelv to form a network, both in the same plane and
PATTERNS OF THE A-V PATHWAYS
907
fig. 11. Photomicrograph from the anterior wall of rat
ileum. (Courtesy of Dr. Silvio Baez.)
between adjacent layers of muscle. The parent vessel
either divides in two, or arches and becomes a venule
which isjoined by venules from neighboring capillary
nets before entering a submucosal vein. Such an
arrangement, whereby a metarteriole leaves the sub-
mucosa and enters the smooth muscle coat where it
gives off a capillary network and then returns to the
submucosa as a venule, constitutes, according to
Baez, a distinctly organized terminal vascular unit.
The minute vessels of this unit are both muscular
and nonmuscular, the muscular component being the
centrally located metarteriole and the precapillary
vessels which branch from it and in turn give rise to
the nonmuscular capillary network. The muscular
vessels and their parent metarterioles are considered
by Baez to be the most highly reactive of the mural
vasculature, a fact demonstrated by vasoactive drugs
and varying intraluminal pressures. The capillary
bed of the muscular coat of the gut, which is served
by these metarterioles, shows periodic changes in
blood flow; the changes being independent of flow
through the arterial plexus of the submucosa. The
muscular coat may be devoid of circulation while
blood continues to flow through arteries of the sub-
mucosa and mucosa.
In some instances the parent vessel may begin as a
short arteriole, rather than a metarteriole, which in
turn then gives rise to several metarterioles when it
reaches the muscular coat. The pattern of distribution
is then the same as described above; the metarteriole
forms a central channel, and turns inward to become
a venule or breaks up into two or three capillaries.
The mucosal artery continues toward the muscu-
laris mucosa after having given off the vessels just
described which go to the outer muscular coat. One
or two short vessels are now seen to branch from the
mucosal artery. The short vessels subdivide into
several capillaries which reunite as a venule and
empty into a submucous vein. Deeper in the sub-
mucosa all mucosal arteries anastomose with a sim-
ilar mucosal artery and give rise to one or two
branches which in turn subdivide to form capillary
nets. The mucosal artery terminates by penetrating
the base of a villus.
Baez was unable to find any arteriovenous anasto-
moses in the submucosa of the jejunum or ileum. It
was possible to follow all the arteries and arterioles of
the submucous plexus to their finest ramifications
without observing any short cuts from the arterial
to the venous side. This was also true in vessels in the
muscular coat. He does elaborate, however, on the
direct connection between arterioles and venules at
the bases of villi. While the arterial component does
deliver arterial blood directly to the venule which
drains the villus, it cannot be called a true arterio-
venous anastomosis in that the arteriole gives off
branches to neighboring structures. The location of
these vessels is the same as the location of vessels
described by an earlier investigator (117) as arterio-
venous anastomoses. Baez points out that in an in-
jected and fixed preparation the capillary offshoots
might be closed, giving the appearance of a true
arteriovenous anastomosis.
Two or three venules from adjacent villi were seen
to converge to form a mucosal venule. The mucosal
venule also was joined by an arterial capillary which
originated as a branch of the nearest mucosal artery.
The small vein thus formed then emerged into the
submucosa where it was joined by other veins of
similar origin to form an intricate anastomosing
arcade. These submucous arcades were further en-
larged by venules from the outer muscular coat. The
flow of blood through the venules of the muscular
9o8
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
coat and the mucosal venules was rapid and unidirec-
tional. Backflow of blood was observed in veins of the
submucosal plexus.
.1//. rot initiation in the Bulbar Conjunctiva
The bulbar conjunctiva is a highly vascularized
transparent mucous membrane on the anterior surface
of the eye. It extends from the palpebral conjunctiva,
which lines the eyelid, to the cornea. The pattern of
its blood vessels were derived primarily from studies
by Grafflin & Corddry (56) and Lee & Holze (77).
A description of the arrangement of the superficial
blood vessels of the human conjunctiva is given by
Lack et al. (74) in a study designed to observe vascular
changes in hypertension. Arterioles were seen to
divide into numerous side capillaries and terminate
with an end capillary. The term "end capillary" is
not defined. The end capillary on occasion functioned
as a "through-and-through" channel. The capil-
laries appeared to be uniform in caliber. Only rare
arteriovenous anastomoses of the short type were
seen.
A more detailed description appears in a paper by
Lee & Holze (77) in 1950, in which they state that
the arrangement of terminal arterioles, capillaries,
and venules in the human conjunctiva was in accord
with the pattern of vessels as seen in the omentum
and mesentery of other animals, referring to the de-
scriptions of Chambers & Zweifach (20) and Lee &
Lee (78). Capillaries arose at intervals from end ar-
terioles to form an irregular network of vessels which
then rejoined to form the venular system. The ar-
terioles were also seen to terminate in main channels
which communicated directly with a venule. This
pattern was most often seen at the corneoscleral
junction. It was also noticed that blood continued to
flow from arterioles to venules, through the patent
arteriovenous channels, at a time when there was
widespread arteriolar and precapillary constriction.
The precapillaries were found to be more sensitive to
stimuli than their parent arterioles.
Observations of blood flow revealed active contrac-
tion of vessels. Constriction of precapillaries was seen
to occur at their point of origin from the parent ar-
teriole. Attention was directed to these precapillary
sites because of the difficulty in determining minor
changes in diameter or flow in the arterioles. Com-
plete constriction occurred at the precapillary region
lasting for 2 to 3 min. After a gradual relaxation,
blood flow continued for 1 to 5 min before the next
constriction. The periods of constriction and relaxa-
tion were found to be very irregular, with relaxation
predominating.
Arteriolar flow was rapid, capillary flow was slower,
and also intermittent due to spontaneous changes in
diameter at the precapillary sites, while venous flow
speeded up after entering the system of collecting
venules and was consistently regular.
The proposal of a definite structural and functional
unit, such as the preferential channel, as described by
Chambers & Zweifach (20) in the rat mesentery and
seconded by Lee & Holze (77) in the human conjunc-
tiva, prompted Grafflin & Bagley (55) to reinvestigate
the human conjunctiva. These investigators were im-
pressed by an endless variety of vascular patterns
with no apparent plan of organization. This paper
was followed by one by Grafflin & Corddry (56) who
reinvestigated, with improved equipment, the archi-
tecture of vascular beds in the human conjunctiva
in an effort to resolve the differences between the
earlier observations and those of Lee & Holze (77).
Once again they reported a great variety of vascular
patterns with the lack of any recognizable structural
and functional unit similar to that proposed by
Chambers & Zweifach (20). They saw, however,
vessels between arterial and venous channels that
were larger than capillaries. They believed that these
vessels were arteriovenous communications with a
functional significance different from that of capil-
laries. They do not say what the difference is. The
arteriovenous communications were seen so fre-
quently that the investigators believed that they were a
characteristic feature of the conjunctival vascular
beds. In freehand drawings at magnifications up to
80 times, a variety of vascular patterns are shown.
The arteriovenous communications, veno-venous
anastomoses and arterio-arterial anastomoses are
common features (fig. 12). Although at first glance
the vascular pattern may seem very complex, it is
comparable in its arrangement to other terminal
vascular beds which have been presented in such
detail covering a large area. A smaller area is seen
in figure 13.
A representative type of arteriovenous anasto-
mosis, as seen in vascular beds below the surface of
the conjunctiva (presumably on the episcleral sur-
face), is shown in figure 14. It bears a striking re-
semblance to both photomicrographs and diagrams
of the vascular bed in the rat mcsoappendix. The
authors do not describe the kind of blood flow through
these vessels which would qualify them as preferential
channels on a functional basis.
The vascular patterns presented by these authors
PATTERNS OF THE A-V PATHWAYS
909
fig. 12. Superficial vascular pattern, temporal quadrant,
right eye. [From Grafflin & Corddry (56).]
may be considered to contain all the blood vessels in
the areas under observation. While the walls of capil-
laries were never seen and their detection is dependent
on the presence of blood in the vessels, it is unlikely
that the same vessels would be devoid of blood con-
sistently over a period of months during which re-
peated observations were made.
Vasomotion was a prominent feature of flow in the
vessels of this mucous membrane. It was indicated by
variations in the speed of flow, alterations in the
caliber of individual vessels, and intermittent blood
flow through capillary vessels. Arterial vessels usually
fig. 13. Superficial vascular pattern, nasal quadrant, right
eye. [From Grafflin & Corddry (56).]
had a rapid and continuous flow. At times the
arterial vessels showed irregular alterations in the
rate of flow, a reduction in the speed occurring some-
times gradually, sometimes abruptly, and sometimes
stopping completely for a brief interval before surging
forward.
Concerning small arteriovenous communications,
there are three criteria to distinguish them from
true capillaries: /) a larger caliber than capillaries,
2) vasomotion, 3) continuous flow at variable speeds.
However, one or all of these criteria might be un-
satisfied on occasion. Grafflin & Corddry (56) were
unable to detect precapillary sphincters at the points
of emergence of true capillaries from the arteriovenous
channels. It may be assumed that this failure was due
to the limitations of the technique, in that the walls of
the small vessels were not seen distinctly.
Venous flow is described as being continuous at a
relatively moderate speed with irregular alterations
in flow. At times the flow stopped completely. This
does not concur with the description given by Lee &
Holze (77), who reported venous flow as consistently-
regular.
Bloch (15), in a lengthy article dealing primarily
with red cell aggregates, describes arterioles and
venules in the bulbar conjunctiva in the following
way: Arterioles in the bulbar conjunctiva do not
differ from arterioles in other tissues. As elsewhere,
giO HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
c ] mm
fig. 14. Prominent arteriovenous anastomosis lying below the superficial layer of the conjunctiva.
[From Grafflin & Corddry (56).]
arterioles are most readily identified by noting the
direction of blood flow. The direction of flow in these
vessels is toward progressively smaller vessels, and
the smallest of them empty into the capillary bed.
Bloch states that when there is a low blood volume,
only one capillary may be seen connecting the
arterial and venous systems. The arterial segment of
this single vessel has a more rapid rate of flow through
it (does not say more rapid than what), differs in
regard to the direction of taper of its walls (pre-
sumably larger at the venous than the arterial end),
and is less permeable than the venous segment (no
basis given for this statement).
Bloch feels that the difficulty in determining which
vessel is an arteriole has arisen partly because of
sudden changes in the direction of flow. According
to Bloch, changes in direction occur when an arterio-
venous anastomosis opens, causing flow in the peri-
pheral segment of the arteriole to stop while flow in
the central portion of the venule speeds up. He states
that there is no difficulty in recognizing the cause of
this directional change if the arteriovenous anasto-
mosis opens while the observer is watching and when
the site of the arteriovenous anastomosis can be
found. He further states that usually the site of the
arteriovenous anastomosis is not identified.
Linear velocity of arterioles is greater than venules
of corresponding diameter. The course of the arte-
rioles is straight compared to the relatively sinuous
course of the accompanying venules. Arteriolar
branches arise gradually from their parent vessels,
while venules branch more nearlv at right angles.
Arterioles are deeper in the tissue than corresponding
venules, their flow rates being so rapid that individual
cells cannot be recognized.
Capillaries are described as cylinders, in contra-
distinction to arterioles and venules which are cones.
Some difficulty arises in comparing the descrip-
tions of the vascular structure as given by Bloch with
that of other investigators due to the absence of
detailed diagrams. True arteriolar branches are not
represented in other vascular beds as branching
gradually, although this type of branching occurs in
vascular nets forming arcuate systems.
Microcirculation in the Spleen
Differences of opinion regarding the manner in
which blood is conveyed from terminal arterioles to
collecting venules in the spleen still exist in spite of
the continued efforts of numerous investigators to
resolve the controversy. Current histology textbooks
(9, 112) present three views. The theory of closed
circulation is that blood in the spleen flows through
completely endothelium-lined pathways from its
entrance into the spleen through the splenic artery
to its exit from the spleen through the splenic vein.
The theory of open circulation proposes that arterial
terminations in the spleen pour blood flowing through
them into the interstices of the reticulum of the red
pulp. The walls of the venous sinusoids are incom-
plete, having longitudinal slits between the reticular
cells which make up the lining of the sinuses. A third
view is a compromise between the open and closed
systems in that some of the capillaries are thought to
open into the intercellular spaces while others open
directly into the sinuses.
It was hoped that the introduction of a technique
which permitted observation of the spleen of a living
animal might settle the controversy. In 1936, Knisely
(69) studied living transilluminated spleens of mice,
rats, and cats and reported that each vessel traced in
the spleen was connected to the arterial system and
the venous system. No vessels were seen to open out
into or pour blood into intercellular pulp spaces.
The lining of the arterioles, arterial capillaries,
capillaries, venous sinuses, and venules was readily
apparent through the microscope as a narrow, clear,
sharply refractile line, visible also during periods
when no blood was flowing through the vessels.
Knisely's conclusion was that the vascular system of
the spleen consisted of a series of completely inter-
connected, preformed, lined channels. He describes
spontaneous vasoconstriction in the arterial branches
PATTERNS OF THE A-V PATHWAYS
9"
as well as in the venous sinuses, and assigned this
normal activity of vascular smooth muscle to ''physio-
logical sphincters." No significant differences were
noted in the structure or activities of living mouse,
rat, and cat spleens.
In 1941, MacKenzie et al. (83) reported that they
had been unable to confirm Knisely's findings. They
point out that the modern concensus favored a
splenic circulation that had an open component
which allowed flooding of the pulp interstices with
whole blood, and additional pathways afforded a
closed circulatory component. The reception of
Knisely's investigations had been favorable and
"offered a reasonable conclusion to an otherwise
apparently interminable discussion.'' However, they
were not able to see what Knisely had seen. In
transilluminated spleens of mice, the walls of follicle
arteries were seen as sharply refractile lines, running
parallel, the diameters of the vessels uniform except
when constriction occurred. The follicle arteries
branched two or three times to form penicilli, syno-
nyms being pulp arteries, sheathed arteries, or pulp
arterioles. They were able to see only the peripheral
portion of the follicle capillary network, the ultimate
twigs penetrating the marginal zone of the red
pulp. Terminal capillary branches, as many as eight
in number, enter the adjacent red pulp and develop
funnel-shaped dilatations. These arteriocapillary
ampullae communicate directly with the pulp
interstices by way of numerous apertures. As the
lumen of the capillary widens in the formation of its
ampulla, the refractive quality of the vessel wall is
rapidly lost. The parallel linear shadows produced
by the capillary are replaced by the contours of pulp
cells.
Venous sinuses originate in the red pulp by an
enclosure of pulp spaces. A venous sinus gradually
increases in diameter to a maximum and then joins a
vein. The wall of the venous sinus is composed of
loosely connected cells lying parallel to the long
axis. The openings between these cells, according to
MacKenzie et al., permit the free passage of blood
cells.
The interstices of the pulp provide the one and only
type of connection seen by them to link the arterial
and venous systems in the spleens of mice, rats,
rabbits, guinea pigs, and cats. They state, however,
that in all spleens there were instances when an
arterial capillary appeared to be connected by a
vessel to an adjacent venous sinus, but this proved to
be an optical illusion caused by weaknesses inherent
in the transillumination technique. Spontaneous
arterial vasoconstrictions were seen to occur inter-
mittently and were a factor in the control of circula-
tion of blood through the small vessels as were trabec-
ular and capsular contractions. MacKenzie et al.
believed that the results of their work supplied
additional confirmation of an open circulation for the
mammalian spleen.
Bjorkman (11) studied rabbit spleens following the
injection of starch granules and concluded from the
distribution of the grains that circulation through the
spleen was the open type.
A detailed and convincing report in favor of the
closed system of circulation appeared in 1 951 ,
authored by Peck & Hoerr (94). They selectively
attacked statements made in the paper by Mac-
Kenzie et al. (83), pointing out where possible
technical variances could explain the differences in
their observations. Peck and Hoerr found the inter-
mediate circulation of the spleen of the mouse to be
essentially as Knisely (6g) described it. They say that,
on arriving at the red pulp, arteries branch two to six
times to form the penicillar arteries which then
extend into the red pulp 10 to 15 u before branching
several times to form capillaries. Where more than two
capillaries arise from a red pulp artery, the termina-
tion of the artery may be ampulla-shaped (see fig. 15).
There is no discontinuity of the refractile lines from
artery to capillary, and these lines caused by the
vessel walls must be endothelium or reticular fibers.
Capillaries extend in all directions from the ampulla-
like terminations of the penicillar arteries to terminate
in venous sinuses or venules (fig. 16).
The course taken by a capillary may be straight,
curved, or tortuous. If they have a tortuous course it
is often difficult to follow them because they run
under other vessels or extend beyond the range of
focus. Capillaries may turn away from view at their
point of origin, giving the ampulla-like termination
of the penicillar artery the appearance of ending in a
sac or pouring its blood into the pulp. The capillaries
can be seen to terminate in the venous sinuses, the
walls of which (more difficult to see in the contracted
spleen) are continuous with the walls of the capillary.
Penicillar arteries have a powerful sphincter action,
although an artery may contract along its entire
length. Individual arteries may exhibit this con-
striction independently of neighboring arteries, which
continue to have a rapid flow. Fairly constant blood
flow over a period of hours is seen in straight capil-
laries which terminate in venules. In some cases red
cells may seem to wander in the extravascular tissue,
912
HANDBOOK OF PHYSIOLOGY -^ CIRCULATION II
fig. 15. An ampulla of a red pulp artery.
[From Peck & Hoerr (94).]
fig. 16. Diagram summarizing the main types of arterio "
venous connections in the mouse spleen. [From Peck & Hoerr
(94)-]
but proper focusing shows them to be within tortuous
capillaries.
Peck and Hoerr conclude from their observations
that blood in the spleen passes through lined, intact
blood vessels which join the arterial and venous
systems.
The next major reports on intermediate circulation
in the spleen were by Parpart et. al. (93) and Whipple
et al. (133), and favored open circulation. These
investigators saw three ways in which trabecular
arteries, terminating as arterioles, connected with
collecting veins. Most of the arterioles spewed blood
through funnel-shaped openings into large pulp
spaces from which the blood flowed into collecting
veins. Some of the arterioles made direct connection
with collecting veins, and such connections were
called arteriolar-venous anastomoses. A few of the
arterioles were seen to branch into a loose, irregular
capillary network which formed venules that returned
to the collecting veins. Considering each component
of the system from arteriole to veins separately,
Parpart et al. (93) state the following: Terminal
arterioles are thick and muscular and show con-
tinuous diameter changes clue to constriction and
relaxation of the vascular muscle. The terminations
of these arterioles are usually three-dimensional and
funnel-shaped with the flared ends becoming too thin
to be seen with the microscope. This indicates,
according to Parpart et al., that the flare thins out to
a condition of no endothelial covering of the blood
that flows out of the ampulla into the pulp. The pulp
space may be fed by only one arteriole or by several.
Pulp (reticular) cells are scattered throughout the
space seemingly held in position by connective tissue
strands. Red and white blood cells can be seen to
enter and leave the main stream of blood flowing
through the pulp space, remaining outside the stream
and thus stationarv for variable periods. The pulp
spaces are interconnected as shown by the passage of
blood cells between them. Collecting veins are seen
in the pulp spaces at positions opposite to the arteriole
entrance, receiving blood through end and lateral
openings in their walls. The lateral openings are
large enough in some instances to allow the passage
PATTERNS OF THE A-V PATHWAYS
9'3
of several red cells abreast. The collecting veins are
described as thin-walled structures randomly per-
forated with holes of varying sizes that are part of a
branched-treelike arrangement. Although capillary
networks supplied by an arteriole and feeding into
veins are occasionally seen, there are relatively few
of them. The capillaries are said to have holes in their
walls through which blood enters or leaves the ad-
jacent pulp space.
Parpart et al. report that they have never seen a
venous sinus of the type described by Knisely (69),
nor have they seen any activity in the venous pulp
spaces that could be regulatory to the blood flowing
through them. This is in direct opposition to state-
ments by Knisely (71) and by Peck & Hoerr (94)
regarding the regulation of blood flow through the
splenic pulp. Knisely (71) took exception to the
conclusions of Parpart et al. (93), particularly pointing
out that their optical arrangements were such that
not all structures present in the tissue would neces-
sarily be observed. With the quartz rod, which
Knisely used, it is possible to direct the light first one
way and then another and thereby make previously
unobserved structures visible (69).
In 1958, Snook (115), who believed one reason for
disagreement concerning splenic circulation was the
structural variability of the spleen among mammals,
reported on fixed rabbit spleens. He had previously
classified the mouse with mammals that had non-
sinusal spleens (114), and comparative studies
showed that the rabbit spleen was more nearly like
human spleen than the other animals observed.
Conclusions from his histological studies of the rabbit
were that rabbit spleen had the open type of inter-
mediate circulation, that white pulp capillaries
occasionally connected directly with premarginal
sinuses, and that penicillar branches terminated in
pulp cords in ampullary dilatations.
In the 1958 edition of Bailey's Textbook of Histology
(112) the authors take the stand that "there is a
fairly direct connection from the capillary to the
venous sinus in most cases, but the system is open in
the sense that the lining membrane changes from
endothelial cells to flattened reticular cells, and
contains perforations through which erythrocytes
may readily pass."
Fleming & Parpart (41) investigated the spleens of
young rats and found them to be very different from
mice. Capillary networks were seen which had a
pattern very similar to that of mesenteric circulation
(144). No venous sinus or pulp spaces were found.
Vascular walls were easily seen and very few red blood
cells were free in the intercellular space.
Fleming and Parpart suggest that such a capillary-
pattern in the rat is an infantile characteristic and
that pulp spaces develop when the animal becomes
more mature. They believe that the fact that endo-
thelial walls can be seen with such clarity in this
preparation proves that they could also be seen, if
present, in the spleens of mice. Thus the position
taken by Parpart for an open system of intermediate
circulation in the mouse spleen appears strengthened.
It is very difficult for an unbiased reader to decide
in favor of one or the other types of circulation in the
spleen because of the convincing arguments presented
by the proponents of each. It has been suggested,
however, that the burden of proof rests upon those
who favor the open type of circulation, since endo-
thelium is universally present in every other vascular
system (26).
There are one or two structural arrangements
described by Parpart et al. which would be unique if
they do exist, i.e., capillary vessels with holes in their
walls through which blood enters or leaves the ad-
jacent pulp area, and veins which have end and
lateral openings varying in size from 5 to 20 jx, the
latter openings being randomly spaced along the
endothelial lining of the veins.
Williams (139) expressed the opinion that the
entire spleen might be thought of as a modified blood
vessel with certain special structures in its lumen and,
therefore, that the endothelial lining of the internal
blood channels might have a different significance
than elsewhere.
Microcirculation in the Lung
Microscopic observations of pulmonary circulation
date back to Malpighi (84) in 1661. Occasional
reports appeared in the literature from time to time
after this, possibly the greatest concentration being
in the 1930's.
In 1930, Olkon & Joannides (gi, 92) studied the
pulmonary circulation in dogs, frogs, and alligators !
The optical magnification was quite low by present
day standards (60 X) and the fact that the animals
were on artificial respiration and their lungs in con-
stant motion must have added considerable difficulty
to their investigation. They describe what appeared
to be a large capillary lying between the walls of the
alveoli from which many smaller capillaries were
given off. They believed that the single large capillary
surrounding the alveolus was most likely a capillary
9'4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. i 7. Sketch of air sac and its vessels in the lung of a cat.
[From VVearn et al. (128).]
network seen on edge. The smaller capillaries anasto-
mosed frequently with each other and appeared to
contract and relax.
In 1933, Daly (38) reviewed pulmonary circulation
and briefly mentioned microscopic observations of
pulmonary vessels, but these studies were concerned
primarily with the response of small vessels to epi-
nephrine. There were no papers in which normal
vascular patterns were described.
Wearn et al. (128) introduced a method for the
observation of minute vessels in the lung which
utilized a quartz rod, for transmitting light, placed
inside the chest cavity. The vessels were observed
microscopically in the lung tip through a window in
the chest wall. Observations were made on both
moving and immobilized lungs. Arteriolar vessels
showed pulsatile flow and steady flow, a different
type of flow often occurring in two arteriolar branches
from the same parent arteriole. Arterioles were seen
to contract and relax, and reversal of the direction of
blood flow was common. The walls of the capillaries
were invisible, so that the caliber of the vessel and its
course were determined by the column of blood it
contained. Capillary vessels were seen to branch and
anastomose frequently (see fig. 17). Wearn et al.
report spontaneous opening and closing of capillaries,
but the fact that the differentiation between arteriole
and capillary was based on the number of blood cells
which the vessel would accommodate (three or less
for a capillary) throws some doubt on the validity of
this statement. Also, the suggestion is made later in
the report that intermittent flow in capillaries was
probably due to changes in the flow of arterioles
from which the capillaries arose and that no proof of
contraction of capillary walls was obtained. The
capillary network, depicted by Wearn et al. as covering
an alveolus, is very similar to that in many other sites.
The next detailed report of vascular architecture of
the lungs appeared in 1954, when Irwin and his
associates (66) published the results of microscopic
observations on guinea pigs and rabbits. The tech-
nique used was that of transillumination with a
quartz rod and oxygen insufflation to prevent re-
spiratory movements. These investigators traced
pulmonary arterioles to terminal pulmonary arterioles
which branched to form capillaries. Blunt termina-
tions of pulmonary arterioles, which lie in the septa
between alveoli, gave rise to capillaries which then
spread over several adjacent alveoli. Pulmonary
capillaries were seen to be completely lined cylindrical
tubes which branched and anastomosed to form
intricate networks over the surfaces of alveoli. The
capillary network was often supplied by several
terminal arterioles and each network was drained by
more than one venule. Arteriovenous shunts were
found between a pulmonary arteriole and venule that
ran side by side in an alveolar septum.
Additional observations by Irwin & Burrage (67)
were that the diameters of arterioles and venules
changed size when measurements of their walls were
made over long periods of time, affecting the flow of
blood. Alterations in the diameters of the arterioles
were more marked than venular changes. Irwin and
Burrage report that the walls of the capillaries cover-
ing an alveolus were seen to come together to ob-
literate their lumina. They suggest that, although the
intermittent blood flow in pulmonary capillaries
could be due to contraction of either arterioles or
venules, the possibility of activity in the capillary walls
must be considered. That capillary walls might
contain contractile tissue, or that the endothelium
lining the cells might swell to block the lumina are the
possible means offered by the authors for causing
intermittent flow in pulmonarv capillaries.
A surprisingly small number of investigations of the
normal vascular structures and flow of blood in
microscopic pulmonarv vessels have been carried out.
However, Wearn's, diagrams (128) and the descrip-
tions by Irwin et al. (66) of capillary networks covering
alveoli indicate that this terminal vascular bed is made
up of the same structural components with the same
basic form as that of beds in other tissues and organs.
Perhaps further investigations will explain the ap-
parent closure of "capillary" walls seen by Irwin &
PATTERNS OF THE A-V PATHWAYS
9'5
Burrage (67). The most likely explanation for inter-
mittent flow through these minute capillary vessels
is that the terminal arterioles and precapillary
vessels which supply the capillary nets exhibit spon-
taneous vasomotion as seen in other areas.
Microcirculation in the Cochlea
The general pattern of blood vessels of the cochlea
has been known for some time, the early descriptions
being obtained from injected and fixed material.
In general terms (40) the cochlea is supplied by the
cochlear artery. This vessel enters the modiolus
through the internal auditory meatus. The spiral
ganglion has a rich supply of capillaries, and many
arterioles find their way to the spiral ligament by way
of the roof of the scala vestibuli. The stria vascularis
is a rich network of small blood vessels with many
anastomotic connections. The limbus has a capillary
supply, and the tympanic surface of the basilar
membrane often has a small arteriole running along
it. Renewed interest in the blood supply of the
cochlea in the past few years has resulted from apply-
ing microcirculatory techniques to this rather in-
accessible site.
The capillary networks of several portions of the
cochlea have been studied in detail. The areas so
studied include the spiral ligament, a projection of
thickened periosteum lying on the outer wall of the
osseous canal of the cochlea; the spiral prominence,
a slight ridge which projects into the cochlear duct;
and the stria vascularis, the part of the spiral ligament
lying on the outer wall of the cochlear duct between
the spiral prominence and the vestibular membrane.
Two papers by Smith (1 10, 1 1 1), which contain a
detailed description of cochlea blood vessels obtained
from fixed material, will be considered before dis-
cussing in vivo preparations. Investigations of
capillary beds following intravascular precipitation
of Prussian blue or lead chromate in the cochlea of
guinea pigs, cats, and humans were carried out by
Smith. She felt that while large features of the circu-
latory patterns had been adequately demonstrated,
the capillary beds had been indistinctly shown and
no attempts had been made to locate them precisely
in relation to various portions of the inner ear.
In these studies, the radiating arteriole was found
to ramify into terminal branches before entering the
spiral ligament. In the cat and guinea pig four groups
of small vessels, depending on their location and the
course which they took, were designated by Smith.
The first group was the network of the upper spiral
ligament, group two was in the stria vascularis, group
three was found in the spiral prominence, while
group four was formed by the capillaries of the lower
portion of the spiral ligament. In human labyrinths
a fifth group, straight vessels in the thicker portion of
the spiral ligament, was included in the classification
(see fig. 18). The network in the upper spiral ligament
is described as follows: Small branches from the
radiating arteriole or one of its terminal ramifications
have a winding course in a spiral direction usually
above the attachment of Reissner's membrane. These
small branches are seen to anastomose with other
tributaries. They leave the upper spiral ligament bv
turning downward to the thicker part of the spiral
ligament where they join venules, or they may turn
upward and go through the bone wall to end in a
collecting vein. The capillaries in the second group,
the stria vascularis, are extensively connected with
fig. 18. Schematic drawing showing typical
distribution of small blood vessels in the spiral
ligament of the human cochlea. [From Smith &
Giovacchine (113).]
gi6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
one another and give the appearance of a network
superimposed upon the deeper vessels of the spiral
ligament. The superior and inferior borders appear
straight and parallel. A large venule drains the
network, formed by the junction of three or four
strial capillaries. The venule turns backward and
leaves the stria vascularis in its lower half where it
descends peripherally, sometimes joined by other
venules before entering into the collecting venous
system at the lower edge of the spiral ligament. The
blood supply of the spiral prominence, the third
group, is different in the guinea pig from cat and man.
In the guinea pig a single vessel is found near the
epithelial layer, with perhaps a single layer of con-
nective tissue cells interposed. The vessel courses
parallel to the network of the stria vascularis just
below its inferior border, although no vessel of the
spiral prominence is ever connected to the network of
the stria vascularis. At times the vessel is double, with
the duplicate vessel running under the edge of the
stria vascularis. The venules join the collecting
venules of the lower spiral ligament. In the human,
the vessels in the spiral prominence form a separate,
narrow, rolled network below the stria vascularis,
supplied by large arteriolar vessels and drained by
large veins. Small vessels leave the network by turning
upward and laterally into the spiral ligament before
emptying into veins. They also may enter venules in
the lower spiral ligament.
The capillary network in the lower spiral ligament
is also supplied by direct large arteriolar branches.
These arteriolar branches descend close behind the
stria vascularis and terminate in a spiral vessel in the
crest of the spiral ligament. Branches are given off
to the stria vascularis and spiral prominence on the
way. The spiral vessel marks the upper limit of a
network which originates from it. Where the spiral
ligament is quite thin, the network can be seen as a
loose mesh of vessels under the mesothelium of the
scala tympani.
The fifth group found in the human is made up of
straight vessels found in the connective tissue between
the scalae and the bone. They show variations in size
and structure, and course directly from arteriole to
venule. Some seem to be capillaries, being devoid of
perivascular cells, while others are larger and may
represent a type of arteriovenous shunt.
The radiating arteriole was found to have both
longitudinal and tangential smooth muscle cells and
a thin adventitia of two or three layers of connective
tissue cells. Capillaries were composed of endothelial
cells and infrequent smooth muscle cells. The capil-
laries of the stria vascularis were composed only of
endothelial cells, although occasionally a perivascular
cell was seen. It could not be determined whether it
was a smooth muscle cell or not. The large draining
vein was seen to have a few smooth muscle cells
arranged transversely or tangentially.
Smith concludes that the human cochlea shows a
definite arteriolar supply to the various vascular
groups of the spiral ligament. There are several
distinct capillary networks rather than one large
continuous field, and these networks are separated by
their vascular supply and drainage. She suggests that
such a vascular pattern makes it possible to have
regional circulatory variations within a small segment
of the spiral ligament.
In 1954, Weille et al. (131, 132) published two
papers describing the circulation in the spiral liga-
ment and stria vascularis of the living guinea pig.
The cochlea was first exposed and then microscopic
fenestration of either the apical or third cochlear turn
was carried out. The vessels observed included
arterioles, arteriovenous anastomoses, capillaries,
and venules of the spiral ligament and the capillary
network of the stria vascularis. Capillaries of the
spiral ligament formed an intricate network of
dividing and anastomosing vessels fed by arterioles.
Arteriovenous anastomoses were formed as a branch
of an arteriole that entered a venule with no inter-
vening capillary network. The capillaries of the stria
vascularis formed a network of branching and
anastomosing vessels that emptied into the venules of
the spiral ligament.
All arterioles, arteriovenous anastomoses, and
venules contracted and dilated independently. The
rate of blood flow varied in each vessel from time to
time, going from very rapid to no flow at all.
A more detailed description followed (65), in which
it was reported that there were two distinct types of
tiny vessels, one in the area of the upper spiral liga-
ment and the other in the area of the pigmented
cells (the cochlear duct). Branching and anastomosing
were frequent in these networks. Both received blood
from the arterioles and both drained into the venules
in the area of the cochlear duct.
Collecting venules, into which the capillaries drain,
pass transversely through the area of the cochlear
duct, and drain into the venules which are perpen-
dicular to them and lie outside this area.
Vessels which ran from arterioles to venules were
seen to give off capillaries, but no capillaries were
seen to re-enter them. For this reason they were
called arteriovenous anastomoses rather than met-
PATTERNS OF THE A-V PATHWAYS
9J7
arterioles, a name given to them by Seymour (109).
The anastomoses were seen to contract to complete
closure.
Microscopic observations of cochlear blood vessels
in living guinea pigs were reported by Perlman &
Kimura (95, 96) in 1955. Special attention was given
to the small vessels of the spiral ligament and the
stria vascularis. The quartz rod technique was used
after the cochlea was fenestrated in the fourth turn.
The fenestra was o. 1 to 0.2 mm'2 and exposed the
spiral ligament on the lateral wall of the cochlear
duct as well as the stria vascularis. Perlman and
Kimura were certain that all vessels in the field were
visible to them and that all the basic units of a vascu-
lar bed were present. The identification of the various
components was based on the diameter, the wall
thickness, shape of the vessels, the rate and direction
of flow, and the presence of smooth muscle cells and
vasomotion. Numerous anastomoses between all
types of vessels were seen, but the distribution and
direction of flow from the radiating arteriole to the
collecting venule suggested a segmental blood supply.
The arterioles of the spiral ligament were seen to
branch into a number of different vessels. A small
branch at right angles to the radiating arteriole was
seen to run parallel to the cochlear duct in the upper
portion of the spiral ligament. It anastomosed with a
similar vessel from an adjacent arteriole.
Another branch was seen that crossed the under-
lying stria vascularis and emptied into the collecting
vein below the cochlear duct. This type of vessel,
regularly seen in the area, has no branches, is narrow
and straight, and has a rapid blood flow. It has
ft. COCHLEAE
PROPRIA
H .V. SPIRALIS
'/■ POSTERIOR
fig. 19. Segment of cochlea showing relation of exposed
vessels to the cochlear duct and the main trunks in the modiolus.
[From Perlman & Kimura (95).]
smooth muscle cells regularly distributed along its
walls. The authors have called this vessel an arterio-
venous arcade. (See figs. 19 and 20.)
Another branch of the radiating arteriole with a
uniform diameter extends over the underlying stria
vascularis, has no branches, and ends at the level of
the spiral prominence just below the stria vascularis.
The vessel with which the branch connects runs paral-
lel to cochlear duct and tributaries from it join collect-
ing venules of the spiral ligament. The blood vessels in
the stria vascularis are at right angles to the radiating
arteriole, the collecting veins, and the arteriovenous
arcade.
The last branch from the radiating arteriole is the
one which enters the stria vascularis. Diameters of
the strial vessels are usually larger than the diameters
of the radiating arteriole or arteriovenous arcades.
Strial vessels do not have a regular distribution of
smooth muscle cells and were not seen to contract.
The vessel in the spiral prominence is independent
of the stria vascularis, being directly supplied by a
branch from the radiating arteriole. A large number
of tributaries leave this vessel to join the collecting
vein. Anatomically, it seems to have the qualifications
of a capillary, being small in size and devoid of
smooth muscle cells, and having a slow rate of blood
flow. It shows no vasomotion.
In commenting on the vascular pattern, Perlman
and Kimura state that the segmental type of blood
flow suggests that interference with function may be
localized. Interruption of flow in a radiating arteriole
of an arteriovenous arcade may occur while flow
continues in the underlying stria vascularis. Flow in
the stria vascularis may cease while active flow con-
tinues in the radiating arteriole, arteriovenous arcade,
spiral prominence vessels, and venules. The presence
of arteriovenous arcades in the spiral ligament sug-
gests a possible regulatory mechanism for controlling
flow in the stria as well as affording anastomotic
channels to insure continuity in blood flow along the
spiral ligament.
Perlman believes that the strial vessels, the arterio-
venous arcade, and the spiral prominence vessels
have functional roles. The strial vessels may be called
capillaries with regard to their position, the fact that
they have the lowest blood flow rate, and the fact
that they have no smooth muscle cells in their walls.
The decrease in the rate of blood flow from the radiat-
ing arteriole to these capillaries of the stria vascularis
is large and abrupt. The final exchange of diffusible
substances probably occurs in these vessels.
The role of the cochlear blood vessels in the absorp-
9i8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
RADIATING ARTERIOLE
165
fig. 20. Schematic drawing showing rela-
tions of basic vascular units exposed by the
fenestra and the average blood velocity in
micra per sec. [From Perlman & Kimura
(95)-]
STRIA
VASCULARIS
SPIRAL
PROMINENCE
SCALA
VESTIBULI
COCHLEAR
DUCT
SCALA
TYMPANI
COLLECTING VEIN
tion and secretion of perilymph and cndolymph is
still not clear, although it is believed that endolvmph
is secreted by the stria vascularis (40). It is interesting
that this fluid, which fills the scala media, differs in
ionic content from perilymph which fills the scala
vestibuli and the scala tympani. Unlike all other
extracellular fluids in the body, endolvmph is high in
potassium and low in sodium, thus more nearly
resembling an intracellular fluid. The very slow rate
of blood flow in the stria vascularis may be necessary
to allow the formation of this intracellular-like fluid.
It has been suggested, also, that the slow flow rate
and the absence of vasomotion in the strial vessels
contribute to the fact that blood flowing through the
cochlea is not heard.
Preferential or Thoroughfare Channel
The attempt to give a representative description of
vascular patterns in terminal vascular beds has not
escaped the usual criticisms aimed at generalization
from one animal to another or among different tissues
within the same animal. One of the major issues
among investigators of the microcirculation has been
the acceptance of a preferential channel as a com-
ponent of all capillary networks. A chronological
presentation of the development of the concept, the
modifications, and current status might assist in
clarifying the issue.
The first description of the arteriovenous (a-v)
bridge, later to be called the thoroughfare or prefer-
ential channel, appeared in 1937 (143). Zweifach,
in studies of the mesentery, nictitating membrane,
and undersurface of the tongue of the frog, described
two types of vessels present in these structures. One
was a continuous central trunk that connected an
arteriole and a venule. This vessel, a direct continua-
tion of an arteriole, was invested with widely sepa-
rated atypical smooth muscle cells which were less
responsive to mechanical stimulation than smooth
muscle cells of the arterioles. The a-v bridges did not
always take a direct linear course from the arterial
to the venous side, but appeared in three basic pat-
terns: /) a direct course without other terminal
branches, 2) a fountain-shaped pattern, 3) a horse-
shoe-shaped pattern, which was to become regarded
as the basic design for the preferential channel. The
a-v bridge was functionally different in that it
always had a patent lumen with uninterrupted blood
flow. The second type of vessel was the true capillary,
a nonmuscular vessel which was an off-shoot or
branch from the a-v bridge, not in the direct path of
blood flow from arteriole to venule. In a camera-
lucida drawing of the vascular pattern in the frog
mesentery it can be seen that the a-v bridge either
formed a loop to return to the venule accompanying
the arteriole from which it arose, or continued across
the capillary bed to join another venule (fig. 21).
In a second paper the same year, the arteriovenous
bridge was reported in the mesentery and ear of the
mouse (150). In 1939, Zweifach (144) extended his
studies on living vessels in the mesentery, tongue, skin,
and intestinal wall of the frog, and in the mesentery
and ear of the mouse. Little new information was
added, but the functional significance of the a-v
bridge was stressed. The main central pathways were
PATTERNS OF THE A-V PATHWAYS
9>9
fig. 21. Camera-lucida outline of
vessels in the capillary bed of the frog
mesentery. [From Zweifach (143)-]
said to have a vigorous circulation even when tissues
were in a resting or anemic state. The bridge was
regarded as a muscular capillary and was the central
pathway from which the remainder of the capillary
vessels were distributed as side channels.
In 1944, Chambers & Zweifach (20) collaborated
on a paper in which the studies were confined to the
mesenteric circulation in the dog and the rat. They
state that the fundamental architecture is the same
for both tissues. The mesoappendix of the rat differs
from other parts of the mesentery in its lack of any
major vessels coursing from the aorta or to the vena
cava, all vessels in the mesoappendix being less than
80 to ioo/x in diameter. The description of the capil-
lary bed is based primarily on observations in the
mesoappendix of the rat, in which the a-v bridge is a
prominent structure. The term "metarteriole" (Gr.
meta — beyond) is introduced to designate the proxi-
mal contractile portion of the central channel.
Beyond the metarteriole, muscle cells disappear and
the channel continues as the a-v capillary until it
joins a venule. Other contractile muscle cells, desig-
nated precapillary sphincters (the precursors of the
true capillary), were found at the proximal end of the
channel but were absent at the venular end. Each
central channel and its side branches with their inter-
posed true capillaries were said to constitute a struc-
tural unit. In a summary in 1946, Chambers &
Zweifach (21) state that the basic topography of a
predominantly nutritive type of capillary bed is
presented as a central channel of which the true capil-
laries are side branches. The different portions of the
central channel, in sequence, were the metarteriole
which exhibits vasomotion and has typical but dis-
continuous muscle cells; the proximal portion with
atypical muscle cells; the distal portion with no
muscle cells; and the nonmuscular venule. The pre-
capillaries were described as the proximal muscular
portions of the abrupt offshoots of the muscular por-
tion of the central channel, acting as sphincters and
controlling blood flow through capillaries. The true
capillaries continue from the piecapillaries and are
also direct branches of the distal portion of the central
channel and of the nonmuscular venule.
In 1947, the functional aspect of the structural unit
was again emphasized (22). It was pointed out that
in some tissues which maintain a constant level of
flow volume there is no discernible organization of
capillaries, while in tissues such as the muscular sys-
tem and the gastrointestinal tract with varying
activity the structural unit exists. This vascular
pattern allows for great expansion in the number of
vessels with an active circulation at one time and
restriction of flow to the preferential channel during a
period of inactivity. In the decade following the
introduction of the preferential channel, new ideas
and new terminology were added. For clarification
of the structure and function of the terminal vascular
bed in the rat mesentery, excerpts from a paper by
Chambers & Zweifach (22) follow. "The preferential
vessels have been termed thoroughfare or a-v
channels. The proximal portion of these channels,
termed metarterioles, together with their precapillary
sphincteric offshoots, are muscular and spontaneously
undergo periodic changes in caliber. This type of
movement has been termed vasomotion, a slow inter-
mittency of partial relaxation and constriction at
intervals of about 30 seconds to 3 minutes. . . . The
precapillary sphincteric offshoots lead into an inter-
anastomosing system of true capillaries (devoid of
muscle elements) which constitutes the bulk of the
bed. The capillaries rejoin the distal continuation of
the thoroughfare channels through inflowing tribu-
taries." The thoroughfare channels, always open, are
said to maintain a constant pressure relationship
between their arteriolar and venous ends, the flow of
blood through them being more rapid than thiough
other vessels in the bed. The channel is said to be the
site of outward filtration, while inward filtration
occurs in the true capillaries.
920
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
One significant anatomical difference noticed in
comparing microphotographs of circulatory patterns
in the rat mesoappendix with that of other tissues is
the absence of paired arteriole and venule in the
mesoappendix of the rat. An arteriole emerges singly
from its parent vessel, descends into the mesentery,
then forms a loop which in its return is joined by other
capillary vessels before emptying into the vein which
accompanies the artery of origin. In most vascular
beds, other than the mesentery, arterioles and venules
are found to be adjacent and to branch together until
the final ramification which forms the capillary net.
The small arterioles, from which the true capillaries
arise, form arcades or arcuate patterns with other
arterioles rather than continuing as a direct pathway
to the venous side. It is not uncommon to see, in the
bat wing at least, a short terminal arteriole that
quickly joins a collecting venule after giving off one
or two branches (see fig. 22). Contrary to the descrip-
tion given to preferential channels these terminal
arterioles will close down completely or may be devoid
of blood when their parent vessel is occluded by con-
traction of the circular smooth muscle which invests
them.
According to subsequent papers by Zweifach ( 1 46,
147) and Zweifach & Metz (151, 152), in which
vascular patterns are compared, it seems that the
preferential channel occurs mainly in rat mesentery,
outer edge of rat skeletal muscle, and the serosa of
the small intestine of the same animal. He states that
the preferential channel is unusually prominent in
the mesentery but is not a major structural feature of
the urinary bladder and the skin. Further reservations
as to the ubiquitousness of the preferential channel
have appeared as a result of observations in the under
surface of the skin, the skeletal muscle, urinary blad-
der, several mesenteric structures, and the serosal
surface of the small intestine. Zweifach states that "a
major variable lies in the structural organization of
fig. 22. Arteriovenous pathways in the subcutaneous area
of the bat wing.
the different vascular beds, especially the mode of
distribution of the capillary system from the arterial
vessels. In such tissues as skin and intestinal wall, the
majority of capillaries originate as direct offshoots of
larger arteries and arterioles. The distal ramifications
of the arterioles have relatively few capillary offshoots
and usually terminate by interconnecting freely with
one another in a series of arcades. This is in direct
contrast to the mesentery where the arterial subdi-
visions, the metarterioles, serve as the parent stem
from which the precapillaries and capillaries branch
out."
Other investigators who have found thoroughfare
channels in various tissues include Lutz el al. (82), who
confirmed their presence in frog mesentery but failed
to identify them in the hamster cheek pouch or retro-
lingual membrane of the frog. Baez (6) reports a short
arteriole which turns inward to become a draining
venule, thus forming a thoroughfare channel in the
muscular coat of the small intestine. Staple & Copley
(118) describe a thoroughfare channel in the labial
marginal gingiva of the mandibular incisor of the
hamster. Lee & Holze (77) observed the thorough-
fare channel in the human conjunctivae, and Lee &
Lee (78) describe the structure in the mesentery of the
guinea pig.
The preferential or thoroughfare channel, either as
a structural or functional unit, has not been seen in
some areas which have been subjected to extensive
studv by various microcirculatory investigators. Nicoll
& Webb (88) report that there are no preferential
pathways in the subcutaneous tissues of the bat's
wing. Clark & Clark (29) do not report them in the
rabbit ear. Grafflin & Bagley (55) found no such struc-
ture in the frog web and urinary bladder, nor in the
human conjunctivae. Later, Grafflin & Corddry (56),
reporting a more detailed study on the bulbar con-
junctiva of man, described vessels between arterial
and venous channels, arteriovenous communications
that seem similar to the preferential channel.
It would seem then that the preferential channel
should not be considered as a component of a typical
capillary network. Although it is possible to demon-
strate a similar anatomical arrangement in terminal
vascular beds other than in the mesentery, there is no
confirmation of the existence of a preferential channel
on a functional basis. It is possible that such a flow
pattern is necessary for the relatively avascular mesen-
tery, and therefore constitutes a special rather than a
typical entity of microcirculation.
PATTERNS OF THE A-V PATHWAYS 92 1
Arteriovenous Anastomoses
A detailed and comprehensive review by Clark (23)
in 1938, dealing with arteriovenous anastomoses,
obviates the necessity of reporting individual investi-
gations up to that time. The discussion here is mostly
confined to the results of in vivo studies.
A direct connection between arteries and veins by
passages through which blood is carried without inter-
change with extravascular fluids had been described
repeatedly since the early 1800's. Such passages were
then considered to be rare, occurring as a result of
injury or as a developmental anomaly. Their presence
in the ear of the living rabbit, as demonstrated by
Clark & Clark (29, 30) and Grant (57), established
their existence in the normal vascular bed.
Grant (57) concluded from his observations of reac-
tions of these vessels that arteriovenous anastomoses
were important in regulating body temperature.
Responses to heating the animal indicated that when
the body temperature was elevated, dilation of
arteriovenous anastomoses permitted a large amount
of blood to flow through the ear, thereby increasing
heat loss. Constriction of arteriovenous anastomoses
occurred when the animal was cooled and thus heat
was conserved. This concept was extended in studies
of the toes of birds and the fingers and toes of man (58).
Clark & Clark (30) studied the arteriovenous
anastomoses in transparent chambers in the rabbit
ear with observations over long periods at high magni-
fications. High magnifications made possible the
descriptions of structural components. Many arterio-
venous anastomoses are present in the ear of the rab-
bit, with considerable variations in their arrangement.
Some arise directly from the central arterv of the
ear, others from secondary or smaller branches, and
some form the termination of an artery or arteriole.
Most of these, however, arise from small arterial
branches. They all empty primarily into larger veins
(see fig. 23).
As to the structure, arteriovenous anastomoses are
found to be straight or coiled, with a thick muscular
wall on the arterial side and a thinner, funnel-shaped
widening on the nonmuscular venous end. Variations
from this general pattern include the absence of the
funnel-shaped venous end and a continuous muscular
wall throughout the entire vessel. The narrow inter-
mediate portion has a wall of extra thickness which
seems to be the most contractile portion. The venous
portion is noncontractile, but the large veins with
which the communicating vessel connects often have
substantial muscle walls and show definite contrac-
ART
fig. 23. Camera-lucida drawing of a plexus of regenerated
vessels in the rabbit ear. [From Clark & Clark (34).]
tility. Most of the cross connections show inside di-
ameters of 20 to 40 n during dilatation. Typical
anastomoses may be as small as 5 y. in diameter or as
large as 40 n-
Arteriovenous anastomoses are more active than
arteries and arterioles and show a greater tendency for
independent action. They contract and dilate spon-
taneously and periodically, but with a rhythm inde-
pendent of either neighboring anastomoses or even of
the artery from which they arose. They generally con-
tract more rapidly than arteries, both rhythmically
and in response to stimuli.
Clark (23), did not attempt to explain the function
of the arteriovenous anastomoses, but felt that it was
significant that they occur normally in greatest num-
bers at sites most frequently subjected to mechanical
and thermal irritations, the kinds of stimuli which
produce prolonged dilation of arteries and arterioles.
From his observations of the frog mesentery,
Zweifach (143) describes short arteriolar-venular
anastomoses between vessels only slightly larger than
capillaries. These short channels effectively divert
arterial blood directly into veins. Their caliber changes
seem related to the activity of the capillary circulation.
When most capillary vessels are open and have active
blood flow, the arteriolar-venular anastomoses remain
closed, and then open when capillary circulation de-
creases. The anastomoses differ in this respect from
arteriovenous bridges which maintain a relatively
fixed diameter. Arteriolar-venous anastomoses in the
mesentery of the mouse were described as short, tor-
922
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
tuous vessels that never branched and were not part
of the capillary bed.
Later, similar vessels in the rat mesoappendix were
described (20). The connecting passages in that tissue
join a metarteriole with a neighboring venule, or an
arteriole with a venule. They are muscular for about
two-thirds of their length from the arterial end. When
such shunts dilate, blood flow ceases in the arterial
components distal to the shunts.
Direct microscopic observations in other tissues of
living animals have revealed arteriovenous anastomo-
ses. Wakim & Mann (124) carried out microscopic
studies on the liver of frogs and various mammals at
magnifications up to 600 times, utilizing the quartz
rod transillumination technique. They found arterio-
venous anastomoses in all animals studied. They saw
anastomotic connections between the interlobular
branches of the hepatic artery and the portal vein in
both amphibian and mammalian livers. Seneviratne
(108) observed blood vessels of the livers of frogs,
mice, and rats, and described similar anastomoses.
For frog liver he described several phenomena. Many
short branches from a hepatic artery enter the ac-
companying portal vein. Arterioles cross a lobule and
enter a portal vein on the other side. Occasionally the
arteriole enters a hepatic vein. Small arterial branches
pass through the liver and anastomose with subcapsu-
lar arteries. In the mouse and rat many types of
anastomoses occur between arterial and venous
vessels, the commonest being a direct communication
by short branches between the hepatic artery and the
accompanying portal vein.
Irwin & MacDonald (64) studied guinea pig livers
using the quartz rod technicjue and found the vascular
bed to be similar to that described for the liver by
Knisely et al. (72). The Knisely group found connec-
tions between hepatic arterioles and portal venules
which they called arterioportal anastomoses. Bloch
(14) described arterioportal anastomoses (APA) as
being completely lined smooth-walled tubes that
connect hepatic arterioles with portal venules at
irregular intervals. The hepatic arteriole winds itself
around the portal venule and then sends short
branches out to form APA.
Parpart et al. (93) describe arteriovenous anasto-
moses in the spleen of the mouse as seen by micro-
scopic observation. There, about one artery in ten
makes direct connection with a collecting vein. An
arteriole may anastomose with a collecting vein at
any point on the vein, lateral connections occurring
more frequently than end-to-end anastomoses. When
the connection is lateral, the arteriole is perpendicular
to the venous wall. In the end-to-end anastomosis, the
arteriole gradually widens to become a collecting vein.
Poor & Lutz (97) found no arteriovenous anasto-
moses in the hamster cheek pouch. Irwin et al. (66)
found arteriovenous shunts in the lungs of guinea
pigs and rabbits, although they appeared infrequently.
Blood flow through the shunts was unidirectional,
going from arteriole to venule. Blood flow through
arteriovenous anastomoses in the bulbar conjunctiva
has been described by Bloch (15). Weille et al. (132)
saw them in the stria vascularis.
Zweifach (148) has said that there is little doubt
that occasional shunts between arteries and veins
exist in almost every tissue in the body, but are not a
prominent feature of most tissues. He further suggests
that pathways, not distinct anatomical shunts, go from
arterial to venous systems allowing blood to bypass
the capillary network. Communications between ar-
terial and venous vessels occur more frequently in
terminal vascular beds than in more proximal por-
tions.
It does not seem necessary to assign any highly spe-
cialized function to arteriovenous anastomoses, such
as heat regulation, although this is still done (39, 98).
Folkow (44) is of the firm opinion that arteriovenous
anastomoses in the skin are specialized structures pre-
dominantly engaged in regulation of heat loss and
are regulated by their own constrictor fibers. His
evidence, while convincing, is indirect. Van Dobben-
Broekema & Dirken (121, 1 22), in a study of the reac-
tion of rabbit ear vessels to heating, offer evidence
that there is no obvious relationship between the
temperature of the ear and the diameter of the
arteriovenous anastomoses. Zweifach (148) mentions
the possibility that selective vasoconstriction may
reduce capillary circulation and cause blood to be
shunted through passages which would offer the least
resistance to flow from the arterial to the venous side.
The information derived from the above investiga-
tions indicates that terminal vascular beds of most
tissues are supplied with short communicating vessels
between arterial and venous systems. These arterio-
venous connections allow arterial blood to be shunted
into the venous system without first passing through a
capillary network. As Zweifach (148) has suggested,
the shunts may be preferentially in use when vasocon-
striction of small arterial vessels beyond the shunts
increases resistance to flow. Arterial blood would then
be diverted through shunts which afford the path of
least resistance. The selective vasoconstriction to which
Zweifach refers might result from the response of
terminal arterioles or precapillary sphincters to
PATTERNS OF THE A-V PATHWAYS
923
changes in the local environment, and thus whether or
not blood flowed through capillary networks would be
determined by the immediate needs of the tissue.
Thus, no complex central nervous control is neces-
sary, if the postulate that terminal vasculature is
primarily under the control of local conditions is
acceptable.
BLOOD FLOW THROUGH TERMINAL VASCULAR BEDS
Capillary Contractility
Ideas regarding contractility of capillary vessels
have come full circle, beginning and ending with the
concept that capillaries are noncontractile and the
blood flow through them depends on contraction or
dilatation of the arterioles which supply them. Dur-
ing the intervening periods, investigators have pro-
moted the concept of independent contractility of
capillary vessels, first believed to be brought about by
the contraction of perivascular or Rouget cells and
later thought to be due to the contraction of endothe-
lial cells. At present it is generally accepted that true
capillaries do not contract. By definition they are
devoid of muscular elements, so that muscular contrac-
tility is out of the question, and the endothelial cells of
which they are composed are also noncontractile.
The internal diameter of capillaries may vary, how-
ever, by passive response to changes in pressure or in
the size of the endothelial cells which form the basic
structure of their walls.
Independent contractility of capillaries was a con-
troversial subject in the eighteenth century. The
opinion expressed by Haller (60) in 1 756 that capil-
laries did not contract was generally accepted by most
physiologists until early in the twentieth century [(54),
see also (87)]. At this time publications by August
Krogh (73) appeared. Krogh's belief that capillaries
really contracted is found in a description of an ex-
periment in his book "which demonstrates in a crucial
manner that the whole length of a capillary from an
arteriole to a venule can be contractile, that it cannot,
when contracted, be forced open by the available
arterial pressure. . . ." Krogh was convinced of the
independent contractility of capillaries but he also
believed that no evidence obtained thus far was con-
clusive enough to explain the mechanism by which
this was carried out.
Two possible means of decreasing the diameter of
capillary vessels had been suggested. One was that
either osmosis, or imbibition by endothelial cells, was
responsible, and the other was that active contraction
of extraendothelial cells occurred, as described by
Rouget (103) in 1873. Krogh believed that the imbibi-
tion theory was ruled out by data published by Stein-
ach & Kahn (119), showing that the outside diameter
of contracting vessels decreased, rather than remain-
ing constant or increasing as it would if the endothe-
lial cells enlarged. He believed that anatomical proof
was lacking to establish the functional role of Rouget's
cell. Because of this need for more histological infor-
mation, he encouraged Yimtrup to conduct a detailed
study of the structure of the capillary wall. Vimtrup
(123) examined stained sections of frog tongue and
found cells such as those described by Rouget. He
subsequently named them Rouget cells. He was also
successful in identifying these cells on living minute
vessels in the tail of newt larvae, and in seeing them
contract. The frequent spontaneous contractions and
dilatations of vessels seemed to occur at the location
of the nucleus of a Rouget cell. This was final proof
for Krogh that capillaries possessed independent con-
tractility, the contractile element being the Rouget
cell. He explained away the conclusions of the Clarks
(27, 28) that the Rouget cells were noncontractile by
saying that there was no proof that the cells they
described on vessels in tadpole tails were the same as
Rouget cells or that the contractions they saw were
similar to normal contractility. Krogh, convinced
that the controversy regarding capillary contractility
was settled, extended his belief in the Rouget cell to
include its occurrence on all capillaries in both
Amphibia and mammals.
In a very short time, however, the concept of the
Rouget cell as a contractile cell controlling the diame-
ter of capillary vessels was challenged by detailed
studies on small vessels in the rabbit ear, a technique
introduced by Sandison (104) in 1924, and used by
him and the Clarks, in whose laboratory he began his
work. A chamber for the rabbit ear was perfected in
which original vessels as well as newly formed ones
could be watched for many months. In 1931, Sandi-
son (105) reported that the appearance of Rouget or
adventitial cells on newly formed vessels occurred in a
few hours. Using a magnification of 400 times he could
find a clear space between these cells and the vessel
wall or an endothelial nucleus. The cells did not re-
main fixed, but wandered along the vessel wall.
Sandison stated that the function of the Rouget cell
was obscure, except that it helped form a supporting
framework for the vessels. He was able to demonstrate
in the rabbit ear vessels that a widening of the space
between the adventitial cell and the arteriolar wall
'1-4
HANDBOOK OF PHVSIOLc K ,Y
CIRCULATION II
occurred when the arteriole narrowed. Clark & Clark
(27) had previously made a similar observation in
Amphibia, and also had observed that the contraction
of small vessels on which no adventitial cells de-
veloped was the same as those in which adventitial
cells were present. These observations seemed to rule
out any possibility that the adventitial or Rouget cell
was responsible for contraction. Sandison also noticed
that, as newly formed vessels changed from capillary
to arterial forms, the adventitial cells disappeared and
circular smooth muscle cells took their place. It was
shown later by Clark & Clark (33) that adventitial
cells actually differentiated into smooth muscle cells
as new capillaries developed into arterioles.
A year later Sandison (106) stated that it was clear
from continuous microscopic observation of minute
vessels that contraction and relaxation of smooth
muscle cells of arteries and arterioles were responsible
for alterations in blood flow through capillaries, and
that neither Rouget cells nor endothelial cells played
any part in contraction of vessels.
In the same year, Clark & Clark (29), after observing
capillaries of normal ear tissue through a trans-
parent chamber, stated that if any capillary contrac-
tility did occur it was too negligible to have any influ-
ence on the circulation. In subsequent papers Clark &
Clark (33, 34) summarized the accepted ideas regard-
ing capillary and endothelial contractility as follows:
a) Studies on mammalian vessels in transparent cham-
bers, where details of the cellular structures could be
clearly seen in unanesthetized animals, gave no evi-
dence for any contractile power of either endothelial
or adventitial cells. This view was supported by other
investigators (61, 101, 102). b) The real factors re-
sponsible for the control of circulation in the minute
vessels of the mammal are smooth muscle cells on
arteries, arteriovenous anastomoses, and large veins
(105). c) No contractile activity is seen in mammalian
capillary endothelium (29, 31, 33, 86), although
definite active spontaneous contractions occur in the
capillary endothelium of Amphibia (28, 142, 143).
They point out that the experimental evidence for
contractility of mammalian capillaries was based, in
some instances, on studies of nontransparent regions in
which the structure of the wall of the minute vessels
and their true diameters could not be seen. Therefore,
conclusions as to whether they were contracted or
dilated could only be inferred from the number of red
cells present in them. Also, belief in contraction of
mammalian capillaries was often based on observa-
tions of amphibian vessels in which contractions had
been seen to occur with and without extra-enclothelial
cells.
In spite of overwhelming contrary evidence, some
investigators still held for a time to their belief in
capillary contractility. It seems unnecessary to review
the disagreements in the face of the general accept-
ance at the present time of the opinions originally ex-
pressed by Sandison (106) and extended by Clark &
Clark (29). If one accepts the definition of a capillary
as a nonmuscular endothelial tube between the ar-
terial and venular systems, one may state unequiv-
ocally that mammalian capillaries are noncontractile.
Nicoll & Webb (88) stated that observations on
capillaries in the bat wing showed that no perivascu-
lar cells, such as Rouget cells, existed in the region of
these vessels. The smooth muscle cells, at the tran-
sitional points from the terminal arteriole to the
capillary, end rather abruptly. Beyond the termina-
tion of smooth muscle cells within the walls no change
in the diameter of the capillaries, due to activity of
perivascular cells, has been observed.
The question of the role played by the endothelial
cell in caliber changes in capillaries is more unsettled.
To cite some of the recent descriptions of endothelial
cell activity, Nicoll & Webb (88) reported modifica-
tions in capillary diameter that may result from elas-
tic recoil of the endothelial wall due to pressure varia-
tions either inside or outside the vessel. The caliber
change is due neither to active contraction of the
endothelium nor to intracellular swelling. Later,
Webb & Nicoll (130) pointed out that loss or gain of
fluid through the walls of endothelial cells may result
in apparent changes in their size. Also, since capil-
laries are distensible, they may show deformation
under variable conditions (89). These responses are
usually slow in their development and give no indica-
tion of active participation by the endothelial cells.
Chambers & Zweifach (2 1 ) believe that slow spon-
taneous endothelial responses for the most part repre-
sent accommodation to changes in pressure; that
endothelial cells possess a cellular tone which gives a
degree of elasticity to the capillary wall. Lutz el al.
(82) found that endothelium did not respond to me-
chanical stimulation.
Folkow (44) summarizes current opinion in stating
that "slow swellings of the capillary endothelium are
sometimes observed, but are more probably to be
looked upon as passive osmotic effects or deformations
due to passive luminal changes, caused by variations
in intravascular pressure."
PATTERNS OF THE A-V PATHWAYS
925
Vasomotion
The word "vasomotion" has had extensive use
since its first appearance in 1944 (20). The term was
used at this time by Chambers and Zweifach to
describe the spontaneous contractions and dilations of
small arterioles (metarterioles) and the muscle cells
of their branches (precapillary sphincters) in the rat
mesentery, also called mesoappendix. It has subse-
quently come to be used to indicate observed diame-
ter changes of any blood vessel.
Reports of variations in the caliber of small blood
vessels have been in existence for almost as long as
microscopic studies of them have been carried out.
Special interest in this phenomenon was shown during
the period of controversy over capillary contractility.
In the years following the introduction of the rabbit
ear chamber for microscopic observation of small
blood vessels, numerous papers appeared in which
spontaneous alterations in small blood vessels were
described. Clark & Clark (29) spoke of the normal
occurrence of spontaneous rhythmic contractions of
arteries down to their smallest branches. Different
arteries and parts of arteries were seen to contract at
different rates (30). Sandison (106) reported rhyth-
mical contractions of arterial vessels but saw no active
contractions of veins or venules. Clark el al. (35) be-
lieved that an intact nerve supply was necessary for
spontaneous contraction of the arterial vessels. Nu-
merous other investigators reported periodic altera-
tions of small vessel diameters (17, 57, 62, 141 ).
Chambers & Zweifach (20) described vasomotion
in terminal arterioles and larger arterioles as irregu-
larly periodic dilatations that are slower and more
regular than the diameter changes seen in metarteri-
oles and precapillary sphincters. When metarterioles
were exhibiting vasomotion, they usually showed a de-
crease in diameter of about one-third, but were even
seen to reduce the diameter by one-half or more.
Other observations were that when a tissue was
hyperemic, the dilator phase was most prominent, the
constrictor phase dominating in ischemic tissue. No
synchrony in vasomotion of neighboring arterioles
was seen. Vasomotion was seen to continue in a metar-
teriole in the absence of blood flow through it. Also,
diminished blood flow was followed by an increased
dilator phase, while increased blood flow apparently
brought on an intensified constrictor phase. Vaso-
motion was affected by local environmental condi-
tions (irritation of the tissue caused vasomotion to
disappear). Vasomotion also stopped when the ani-
mal was deeply anesthetized.
The recurrent vasomotion in metarterioles was
considered by Chambers and Zweifach to be the
factor which controls the rate of flow through the
central vessels of a capillary bed while the vasomotion
of precapillary sphincters controlled the flow through
the true capillaries.
An extensive discussion of vasomotion bv Nicoll &
Webb (88) in 1946 described various types of caliber
changes seen in both arterial and venous vessels. They
suggested that the word vasomotion should be pre-
ceded by a suitable adjective to indicate a specific kind
of change in vessel diameter, e.g., if the caliber change
is brought about by contraction or relaxation of the
vascular musculature, reference should be made to
active vasomotion. If, on the other hand, caliber
changes are produced by internal or external altera-
tions of pressure not due to the activity of vascular
musculature, reference should be made to passive
vasomotion. Active vasomotion was further classified
into three groups. "Tonic active vasomotion" was
the term used to describe the maintained contraction
of arteries, considered to be a tonus response. Super-
imposed on tone was the rapid contraction and relaxa-
tion of vessels that occur in response to nerve impulses.
This was called irregular active vasomotion. The third
type of movement was called rhythmical active vaso-
motion and referred to a regular alternation of con-
traction and relaxation of the vascular smooth muscle.
An analysis of the various types of active vasomo-
tion, as given by Nicoll & Webb (88), follows. That
arteries and arterioles possess tone, or are in a con-
tinuous state of active contraction, can best be demon-
strated by noting the marked increase in their diame-
ter that follows denervation. The diameters of ar-
terial vessels in a denervated area have been shown to
increase 27 to 29 per cent following nerve section
(134). Nicoll and Webb state that the outstanding
characteristics of tonic active vasomotion are its con-
stancy and sluggishness, and suggest it may function
to correlate blood vessel volume and blood fluid
volume.
Irregular active vasomotion is characterized by
rapid changes in the caliber of arteries and arterioles.
The changes vary as to their magnitude and the
length of time they endure. Such caliber changes are
the direct result of impulses from the vasomotor
nerves, controlled by the vasomotor center. Nerve
section obliterates this type of activity. Nicoll and
Webb are of the opinion that the function of irregular
active vasomotion is to modify peripheral resistance
and also to regulate the pressure gradient in the
capillaries.
926
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
Rhythmical active vasomotion is the third type of
activity seen in vascular beds. It has been observed in
arteries, arterioles, precapillary sphincters, and veins.
This regular alternation of contraction and relaxation
of vascular smooth muscle cells has been shown to
continue after denervation (129). It becomes more
marked in the most peripheral vessel, the best example
of it on the arterial side being at the level of the pre-
capillary sphincters. It is the predominant type of
activity shown by veins. In general, the more normal
conditions are, the more outstanding is the rhythmical
active vasomotion.
Nicoll & Webb (88) offer several reasons to give
support to the hypothesis that this rhythmical ac-
tivity is the result of an inherent property of smooth
muscle cells rather than the response of vascular
muscle to a rhythmical discharge from the vasomotor
center or from humoral influences or physical condi-
tions. The reasons are these: a) Terminal arterioles,
precapillary sphincters, and veins exhibit rhythmical
active vasomotion after denervation, b) This type of
vasomotion is most highly developed in the venous
muscular coat and precapillary sphincters, neither of
which appears to be under direct control of vasomotor
nerves, c) Adjacent vessels vary independently in the
rate and magnitude of their rhythmical activitv.
Rhythmical active vasomotion in veins is fre-
quently powerful, reducing the vascular lumen to
one-third or one-fourth of its resting diameter at the
peak of contraction. The rate at which the contrac-
tions and relaxations occur is usually much faster
than that observed in arterial vessels.
In later reports, Chambers & Zweifach (19, 22) dis-
cussed vasomotion and its relation to fluid exchange
across the capillary wall. The term "vasomotion" was
still used only in reference to spontaneous contraction
and relaxation of the metarteriole and its branches.
Vasomotion in the precapillary offshoots was said to
produce alternate periods of varying hydrostatic pres-
sure, thus greatly influencing fluid exchange in the
capillary bed. When vasomotion was reduced or ab-
sent, blood pressure in the arterioles was spread
through the numerous capillaries of the bed resulting
in a slower flow through the capillaries and a subse-
quent accumulation of fluid in the collecting venules.
Such a situation would create a sufficient back pres-
sure to favor outward filtration. When vasomotion
was active, blood flow went primarily through the
arteriovenous pathways, bypassing the capillary ves-
sels and producing a rapid flow in collecting venules.
This bypassing of capillaries would favor drainage
from the capillary network, a condition which would
increase inward filtration. In summarizing the sig-
nificance of vasomotion in fluid exchange in the
capillary bed, they state that the delicate vasomotor
adjustments, which vary the surface area over which
hydrostatic pressure may cause outward filtration,
play a greater role than the differences between
hydrostatic and colloidal pressure. Osmotic uptake,
which is responsible for inward filtration, is depend-
ent upon and reinforced by adequate venous outflow,
a factor influenced by vasomotion.
Webb & Nicoll (130) refer to rhythmical active
vasomotion as being the outstanding activity in the
entire minute vascular bed and regard it as the prin-
cipal factor of a local nature that regulates blood flow,
and probably pressure, in the capillaries. All types of
anesthesia reduce or abolish active vasomotion of the
smaller vessels, and the authors suggest that this may
be the reason why such activity is overlooked in many
vascular studies. Active vasomotion is greatly reduced
in conditions in which vascular flow is sluggish or
irregular, or when the arterial pressure is low. Active
vasomotion in arterioles can be augmented by sudden
increases in intra-arteriolar pressure.
Active vasomotion was further discussed by Xicoll
& Webb (89) in a paper in which investigations to
determine the effect of environmental changes on
active vasomotion were described. Arteries and the
largest arterioles were said to show two types of active
vasomotion, one being a slowly developing diameter
change dependent on tonus and the other, a rapid
diameter change dependent on the response of vascu-
lar muscle to nerve excitation. The smaller arcuate
and terminal arterioles showed active vasomotion
independent of nerve connection, and no classifica-
tion of this activity into tonus changes or contractile
responses was possible. However, two different types
of muscular activity were seen in these small vessels,
one being peristaltic waves sweeping along the ter-
minal arterioles and the other being localized contrac-
tion of the precapillary sphincters.
The effect that active vasomotion has on blood flow
through capillary beds depends on both its intensity
and its duration. When constriction is not great
enough to close the lumen completely, plasma and
platelets continue to flow through the capillary vessels
while the cellular elements are held back. This is re-
ferred to as "plasma skimming.'" When contraction of
the vascular muscle is great enough to occlude the
lumen, blood flow into the capillary nets is necessarily
curtailed. The occlusion is normally temporary, result-
ing in intermittent flow through the capillary nets.
Veins and venules in the bat wing with smooth
PATTERNS OF THE A-V PATHWAYS
92 7
muscle as a component of their walls exhibit marked
active vasomotion. Nicoll and Webb describe the con-
traction as sharp, and one that sweeps along the vein
as a peristaltic wave in a central direction. Each wave
of contraction seems to originate at a distal valve
and die out at the next valve central to it. Since the
majority of valves are located at the confluence of
tributaries, valve action and blood flow may seem
unrelated due to asynchronous waves in two segments
which are separated by their valves.
Two major tributaries which form a vessel may con-
tract alternately. One tributary may empty into a
segment ahead while the other tributary is relaxed.
Irregular flow results when the frequency of contrac-
tion of the two tributaries is not coordinated. Single
tributaries empty into a segment of the central vessel
during its period of relaxation.
Nicoll and Webb adopt the concept that vascular
smooth muscle cells possess an inherent ability to
change their tonus or exhibit sudden contraction in
response to changes in their immediate environment.
In order to determine what environmental changes
affect vascular smooth muscle, they observed changes
in vasomotion in response to nerve stimulation, various
gas mixtures, and temperature changes. They found
that arteries and large arterioles responded to nerve
stimulation with intense constriction. The smaller
vessels, arcuate and terminal arterioles, precapillary
sphincters, veins, and venules never showed initiation
or modification of active vasomotion as a direct re-
sponse to central impulses (129). Changing the local
environment by flow of constant current between a
single fluid electrode on the wing surface and an
indifferent electrode produced alternate areas of
marked constriction and dilatation on arteries and
arterioles. Reversal of the current caused previously
constricted areas to dilate and previously dilated
areas to constrict. Nicoll & Webb (89) believe that this
observation should be taken into account when inter-
preting responses to direct excitation of nerves with
microelectrodes. Inhalation of carbon dioxide in a
specific concentration proved to be a powerful stimu-
lus of the contractile phase of active vasomotion.
Variations in temperature showed the frequency of
active vasomotion to vary directly with the tem-
perature.
Changes in internal pressure of vessels have marked
effects on vasomotion. Slow changes in pressure caused
a vessel to adjust its tone gradually. Sudden increases
in pressure, however, first caused a vessel to be dis-
tended mechanically and then to contract with great
intensity. The contraction then spread along the
vessel as a peristaltic wave. Nicoll and Webb suggest
that rhythmical variations in small arterial vessels
may originate from sudden internal pressure changes
at their origins from parent vessels.
Spontaneous changes in vascular tone, resulting
from a rise or fall in internal pressure, were demon-
strated. After blood flow to an area had been stopped
by occlusion of a small supplying artery and was then
allowed to resume, the vessels were first distended as
they filled and then were seen to contract as a response,
presumably, to the distention. Thus, blood was forced
along to the next branches. Another example of ad-
justment of tone to a change in internal pressure is seen
following denervation. The resulting dilatation of the
main arteries probably raises internal pressure in the
arterioles and increases their tone, sometimes reducing
the flow through the arterioles to the capillary beds
due to the reduction in lumen of the arterioles.
Nicoll and Webb express the opinion that the ul-
timate result of active vasomotion in terminal arteri-
oles is to establish flow through capillary beds, the
muscle cells of the terminal arterioles being the prin-
cipal targets of changes in the local environment.
Active vasomotion in venules and veins may repre-
sent the adaptation of an inherent property of vascu-
lar smooth muscle to aid venous return. Nicoll and
Webb suggest that this activity may be more wide-
spread in vascular systems than is currently recog-
nized. It may be more prominent in the veins of the
bat wing than in small veins in other mammals due to
the structure of the wing. Pressure within the veins
seems to be the principal stimulus for the action.
Experimental evidence in confirmation of this pro-
posal appears in the investigations by Wiedeman
(135), in which veins in the bat wing were observed
during elevations in venous pressure. Both diverting
excess blood into a vein by ligating other venous path-
ways and infusing dextran to increase total volume
caused a significant increase in cycles of venous vaso-
motion. Similar results were obtained when venous
pressure was elevated by direct infusion with saline
(136).
Although venous vasomotion is most prominent in
the bat wing and shows a definite rhythmicity (fig. 24),
spontaneous changes in pressure which are unrelated
to arterial pressure or respiration have been demon-
strated in small veins in hind legs of dogs (59, 1 37)
^A/VvAAAAAAAAAAA/WN
fig. 24. Rhythmical variations in the pressure in a vein
resulting from alternate contraction and relaxation.
928
HANDBOOK. OF PHYSIOLOGY
CIRCULATION II
(see fig. 25). Such changes have also been recorded in
rabbit ear veins (unpublished data). (See fig. 26.)
Recently, spontaneous changes in venous tone were
recorded from the arm veins of man (16). Folkow has
long supported the concept that rhythmic changes in
tone of vascular smooth muscle is due to myogenic
automaticity (42, 43, 45, 46). He points out (45) that
because the rhythmical reactions seem to be com-
pletely unsynchronized, even in closely adjacent
smooth muscle cells, it is improbable that they should
be due to activity in a local syncytial nerve cell plexus
in the vascular wall, as suggested by others (85). He is
of the opinion (43) that intravascular pressure in a
purely mechanical way to some degree will add "ex-
citatory drive" to myogenic activity as proposed by
Bayliss (10) and confirmed in Folkow's laboratory
(42). Further confirmation appears in recent studies of
forearm blood flow by Blair el al. (12).
At the present time then, in concurrence with Lutz
& Fulton (81), the term "vasomotion" should refer
to anv active change in the diameter of a blood
vessel. It may be seen in one form or another where
vascular smooth muscle exists, such as in arteries,
arterioles, terminal arterioles, precapillary sphincters,
venules, and veins. Any definite conclusions now as to
the actual mechanism or mechanisms which initiate
or control this vascular activity would be premature
insofar as both direct and indirect evidence indicate
that vasomotion in its various forms may be ac-
tivated or modified through the central nervous sys-
tem, reflexly or automatically, through myogenic
automaticity, or through local metabolic factors.
This activity in venous vessels, especially if pri-
marily dependent on myogenic automaticity excited
by increased intravascular pressure, could serve as an
effective aid to venous return from postcapillary
vessels. On the arterial side it could serve as the regu-
lator of blood flow through capillary nets as well as a
protective mechanism whereby capillary vessels could
not be subjected to sudden or prolonged increases in
pressure which might rupture their thin walls.
fig. 25. Spontaneous pressure waxes in a
small \ein in the hindleg of the dog. [From
Wiedeman (137)]
rYVVY
I Art. Press.
10 Sec.
TflJJjTrlT^^
fig. 26. Spontaneous pressure variations
in a small vein of the rabbit ear.
Smoll Vein
W^mfflTO^TOBfflffi
10 Sec
PATTERNS OF THE A-V PATHWAYS
9'29
SUMMARY
It is apparent, from the foregoing descriptions of
structural organization of microcirculatory beds and
regulation of the flow of blood through them, that the
investigations have revealed more similarities than
dissimilarities. Minor differences among patterns
seem to be associated with the structural organization
of the tissue in which the vessels lie, but the basic pat-
terns remain the same.
Although presentation of an anatomical pattern
that would be "typical" for terminal vascular beds
would be likely to meet some resistance, it does seem
necessary to agree on such features as arcuate or ar-
cade connections, gradual divestment of spiral smooth
muscle cells along terminal arterioles to form capil-
laries, absence of direct association and control of
capillaries through nerves, and similarity of the
courses taken by small arteries and small veins. Also,
certain functional activities which regulate blood flow
and blood pressure through these beds must be con-
sidered as universal, these being spontaneous vaso-
constriction and relaxation of arterioles, reversal of
flow paths, alternation of routes of blood flow from
arterial to venous vessels, and variations in the filling
of capillary networks depending on local conditions.
Future investigations may permit generalizations con-
cerning the angles of branching in the arterial system
and spontaneous vasomotion in the venous system.
REFERENCES
1. Abell, R. G. Quantitative studies of the rate of removal
of urea by living blood capillaries from extravascular
solutions in transparent moat chambers introduced into
the rabbit's ear. Anat. Record 6g: 11-31, 1937.
2. Abell, R. G., and E. R. Clark. A method of studying
the effects of chemicals upon living cells and tissues in
the moat chamber, a transparent chamber inserted in the
rabbit's ear. Anat. Record 53: 121-140, 1932.
3. Algire, G. H. Transparent chamber technique. In:
Laboratory Technique in Biology and Medicine, edited by
E. V. Cowdry. Baltimore: Williams & Wilkins, 1952,
PP 354-356.
4. Algire, G. H. The transparent chamber technique for
observation of the peripheral circulation, as studied in
mice. In : Peripheral Circulation in Man. Ciba Foundation
Symposium, edited by G. E. W. Wolstenholme and
J. S. Freeman. Boston: Little, Brown, 1954, pp. 56-63.
5. Algire, G. H., and R. Merwin. Vascular patterns in
tissues and grafts within transparent chambers in mice.
Angiology 6: 31 1-3 18, 1955.
6. Baez, S. Microcirculation in the intramural vessels of the
small intestine in the rat. In: The Microcirculation. Urbana,
111.: Univ. Illinois Press, 1959, pp. 1 14-129.
7. Barclay, A. E., and F. H. Bentley. The vascularization
of the human stomach. British J. Radiol. 22: 62-69, '949-
8. Barlow, T. E. Vascular patterns in the alimentary
canal. In: Visceral Circulation. Ciba Foundation Sym-
posium, edited by G. E. W. Wolstenholme. Boston :
Little, Brown, 1953.
9. Barrnett, R. J. Blood vascular system. In: Histology,
edited by R. O. Greep. New York: Blakiston, 1954, pp.
273"3°3-
10. Bayliss, W. M. On the local reactions of the arterial wall
to changes of internal pressure. J. Physiol. 28: 220-231,
1902.
11. Bjorkman, S. E. The splenic circulation with special
reference to the function of the spleen sinus wall. Acta
Med. Scand. Suppl. 191 : 1-89, 1947.
12. Blair, D. A., W. E. Glover, A. D. M. Greenfield, and
I. C. Roddie. The increase in tone in forearm resistance
blood vessels exposed to increased transmural pressure.
J. Physiol. 149: 614-625, 1959.
13. Bloch, E. H. The bulbar conjunctiva of man as a site for
the microscopic study of the circulation. Anat. Record
120: 349-361, 1954.
14. Bloch, E. H. The in vivo microscopic vascular anatomy
and physiology of the liver as determined with the quartz
rod method of transillumination. Angiology 6: 340-349,
:955-
15. Bloch, E. H. Microscopic observations of the circulating
blood in the bulbar conjunctiva in man in health and
disease. Ergeb. Anat. Entwicklungsgeschichte 35: 1-98, 1956.
16. Borch, G. E. Influence of the central nervous system on
veins in man. Physiol. Revs. 40: 50-56, i960.
17. Burton, A. C, and R. M. Taylor. Rhythmic fluctua-
tions of sympathetic tone and their modification by
temperature and by psychic influences. .4m. J. Physiol.
I26: 453-454. !939-
18. Carrier, E. B. Observations of living cells in the bat's
wing. In : Physiological Papers Dedicated to August Krogh,
edited by R. Ege, H. C. Hagedon, J. Linhard and P. B.
Rehberg. Copenhagen: Levin and Munksgaard, 1926,
PP- 1-9-
19. Chameers, R. Vasomotion in the hemodynamics of the
blood capillary circulation. Ann. N. Y. Acad. Sci. 49:
549"552> '948-
20. Chambers, R., and B. W. Zweifach. The topography
and function of the mesenteric capillary circulation.
Am. J. Anat. 75: 173-205, 1944.
21. Chambers, R., and B. W. Zweifach. Functional activity
of the blood capillary bed, with special reference to
visceral tissue. Ann. N. Y. Acad. Sci. 46: 683-694, 1946.
22. Chambers, R., and B. W. Zweifach. Intercellular
cement and capillary permeability. Physiol. Revs. 27:
436-463, 1947.
23. Clark, E. R. Arteriovenous anastomoses. Physiol. Revs.
18: 229-247, 1938.
24. Clark, E. R. Transparent chamber technique. In :
93°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Laboratory Technique in Biology and Medicine (3rd ed.),
edited by E. V. Cowdry. Baltimore : Williams & Wilkins,
1952, pp. 351-354-
25. Clark, E. R. The transparent chamber technique for the
microscopic study of living blood vessels. Anal. Record
120: 241-251, 1954.
26. Clark, E. R., and E. L. Clark. Observations on changes
in blood vascular endothelium in the living animal. Am.
J. Anat. 57:385-438, 1935.
27. Clark, E. R., and E. L. Clark. The development of
adventitial (Rouget) cells on the blood capillaries of
amphibian larvae. Am. J. Anat. 35: 239-264, 1925.
28. Clark, E. R., and E. L. Clark. The relation of Rouget
cells to capillary contractility. Am. J. Anat. 35: 265-282,
1925-
29. Clark, E. R., and E. L. Clark. Observations on living
preformed blood vessels as seen in a transparent chamber
in the rabbit's ear. Am. J. Anat. 49: 441-473, 1932.
30. Clark, E. R., and E. L. Clark. Observations on living
arterio-venous anastomoses as seen in transparent
chambers introduced into the rabbit's ear. Am. J. Anat.
54: 229-286, 1934.
31. Clark, E. R., and E. L. Clark. Observations on living
mammalian lymphatic capillaries — their relation to the
blood vessels. Am. J. Anat. 60: 253-296, 1937.
32. Clark, E. R., and E. L. Clark. Microscopic observa-
tions on the growth of blood capillaries in the living
mammal. Am. J. Anat. 64: 251-301, 1939.
33. Clark, E. R., and E. L. Clark. Microscopic observa-
tions on the extraendothelial cells of living mammalian
blood vessels. Am. J. Anat. 66 : 1 -49, 1 940.
34. Clark, E. R., and E. L. Clark. Caliber changes in
minute blood vessels observed in the living mammal.
Am. J. Anat. 73: 215-250, 1943.
35. Clark, E. R., E. L. Clark, and R. E. Williams. Micro-
scopic observations in the living rabbit of the new growth
of nerves and the establishment of nerve-controlled
contractions of newly-formed arterioles. Am. J. Anat.
55:47-78, 1934-
36. Clark, E. R., H. T. Kirby-Smith, R. O. Rex, and R. G.
Williams. Recent modifications of the method of studying
living cells and tissues in transparent chambers inserted
in the rabbit's ear. Anat. Record 47: 187-21 I, 1930.
37. Clark, W. E. Le Gros. The Tissues of the Body. New
York: Oxford Univ. Press, 1952, pp. 144-145.
38. Daly, I. deB. Reactions of the pulmonary and bronchial
blood vessels. Physiol. Revs. 13: 149-184, 1933.
39. Daniel, P. M., and M. M. L. Prichard. Arteriovenous
anastomoses in the external ear. Quart. J . Expll. Physiol.
41: 107-123. 1956.
40. Davis, H. Excitation of auditory receptors. In: Handbook
of Physiology. Washington, D. C. : Am. Physiol. Soc., 1959,
sect. 1, pp. 565-584.
41. Fleming, W. W., and A. K. Parpart. Structure of the
intermediate circulation of the rat spleen. Angio/ogy 10:
*8, 1959.
42. Folkow, B. Intravascular pressure as a factor regulating
the tone of the small vessels. Acta Physiol. Scand. 1 7 : 289-
3'°, '949-
43. Folkow, B. A study of the factors influencing the tone of
denervated blood vessels perfused at various pressures.
Acta Physiol. Scand. 27: 99-117, 1952.
44. Folkow, B. Nervous control of the blood vessels. Physiol.
Revs. 35:629-664, 1955.
45. Folkow, B. The nervous control of the blood vessels.
In: Suppl. Vol. to The Control of the Circulation of the
Blood by R.J. S. McDowall. London: Dawson 1956.
46. Folkow, B. The role of the nervous system in the control
of vascular tone. Circulation 21 : 760-768, i960.
47. Fulton, G. P. Conference on microcirculatory physiology
and pathology. Angiology 6: 281, 1955.
48. Fulton, G. P. Microcirculatory terminology (editorial).
Angiology 8: 102-104, '957-
49. Fulton, G. P. Functional aspects of the microcirculation
(editorial). Angiology 11: 146-148, i960.
50. Fulton, G. P., R. G. Jackson, and B. R. Lutz. Cine-
photomicroscopy of normal blood circulation in the
cheek pouch of the hamster, Cricetus auratus. Anat. Record
96:537. 1946.
51. Fulton, G P., R. G. Jackson, and B. R. Lutz. Cine-
photomicroscopy of normal blood circulation in the
cheek pouch of the hamster. Science 105: 361-362, 1947.
52. Fulton, G P., and B. R. Lutz. The use of the hamster
cheek pouch and cinephotomicrography for research on
the microcirculation and tumor growth, and for teaching
purposes. Boston Med. Quart. 8: 1-7, 1957.
53. Fulton, G. P., B. R. Lutz, and A. B. Callahan. In-
nervation as a factor in control of microcirculation.
Physiol. Revs. 40: 57-64, i960.
54. Fulton, J. F. Selected Readings in the History of Physiology.
Springfield, 111. : Thomas, 1930.
55. Grafflin, A. L., and E. H. Bagley. Studies of peripheral
blood vascular beds. Bull. Johns Hopkins Hosp. 92 : 47-73,
■953-
56. Grafflin, A. L., and E. G. Corddry. Studies of periph-
eral blood vascular beds in the bulbar conjunctiva of
man. Bull. Johns Hopkins Hosp. 93: 275-289, 1953.
57. Grant, R. T. Observations on direct communications
between arteries and veins in the rabbit's ear. Heart 1 5 :
281-303, 1930.
58. Grant, R. T., and E. F. Bland. Observations on arterio-
venous anastomoses in human skin and in the bird's foot
with special reference to the reaction to cold. Heart 15:
385-407. I931-
59. Haddy, F. J., A. G. Richards, J. L. Alden, and M. B.
Visscher. Small vein and artery pressures in normal and
edematous extremities of dogs under local and general
anesthesia. Am. J. Physiol. 176: 355-360, 1954.
60. Haller, A. von. Deux memoirs sur le mouvement du sang,
el sur les effets de la saignee; fondes sur des experiences faites sur
des animaux. Lausanne: Marc-Mic. Bousquet, 1756. pp.
136-139. Cited in: Fulton, J. F. Selected Readings in the
History of Physiology. Springfield, 111.: Thomas, 1930,
pp. 82-86.
61. Hartman, F., and J. L. Evans. Control of capillaries of
skeletal muscles. Am. J. Physiol. 90: 668-688, 1929.
62. Hertzman, A. B. The relative responses of the dorsal
metacarpal, digital and terminal skin arteries of the hand
in vasoconstrictor reflexes. Am. J. Physiol. 134: 59-64,
1941.
63. Hill, L. The pressure in the small arteries, veins and
capillaries of the bat's wing. J. Physiol. 54: cxliv p., 192 1.
64. Irwin, J. W., and J. MacDonald. Microscopic observa-
PATTERNS OF THE A-V PATHWAYS
931
tions of the intrahepatic circulation of living guinea pig.
Anat. Record 117: 1-13, 1953.
65. Irwin, J. W., F. L. Weille, and W. S. Burrage. Small
blood vessels during allergic reactions. Ann. Otol. Rhinol. &
Laryngol. 64: n 64-1 175, 1955.
66. Irwin, J. W., W. S. Burrage, C. E. Aimar, and R. N.
Chestnut. Microscopical observations of the pulmonary
arterioles, capillaries and venules of living guinea pigs and
rabbits. Anat. Record 119: 391-408, 1954.
67. Irwin, J. \V\, and W. S. Burrage. Regulation of micro-
circulation in the rabbit's lung. In : Factors Regulating
Blood Flow, edited by G. P. Fulton and B. Zweifach.
Washington, D. C. : Am. Physiol. Soc. 1958, pp. 55-63.
68. Jones, T. W. Discovery that the veins of the bat's wing
(which are furnished with valves) are endowed with
rhythmical contractility, and that onward How of blood
is accelerated by each contraction. Phil. Trans. Roy. Soc.
London, Part 1, 142: 1 31-136, 1852.
69. Knisely, M. H. Spleen studies. I. Microscopic observa-
tions of the circulatory system of living unstimulated
mammalian spleens. Anat. Record 65: 23-50, 1936.
70. Knisely, M. H. Quartz rod technique for illuminating
living organs. In: Laboratory Technique in Biology and
Medicine, edited by E. V. Cowdry. Baltimore : Williams &
Wilkins, 1948, pp. 291-296.
71. Knisely, M. H. The microcirculation of the spleen of the
mouse. Discussion. Angiology 6: 363-368, 1955.
72. Knisely, M. H., E. H. Bloch, T. S. Eliot, and L.
Warner. Sludged blood. Science 106: 431-438, 1947.
73. Krogh, A. Anatomy and Physiology of the Capillaries. New
Haven: Yale Univ. Press, 1929.
74. Lack, A., W. Adolph, W. Ralston, G. Leiby, T. Winsor,
and G. Griffith. Biomicroscopy of conjunctival vessels
in hypertension. Am. Heart J. 38: 654-664, 1949.
75. Lee, C.J. Vascular patterns in the red and white muscles
of the rabbit. Anat. Record 132: 597-611, 1958.
76. Lee, R. E. Anatomical and physiological aspects of the
capillary bed in the bulbar conjunctiva of man in health
and disease. Angiology 6: 369-381, 1955.
77. Lee, R. E., and E. A. Holze. The peripheral vascular
system in the bulbar conjunctiva of young normotensive
adults at rest. J. Clin. Invest. 29: 146-150, 1950.
78. Lee, R. E., and N. Z. Lee. The peripheral vascular
system and its reactions in scurvy. An experimental
study. Am. J. Physiol. 149:465-475, 1947.
79. Lutz, B. R. Microcirculation (editorial). Angiology 10:
241-242, 1959.
80. Lutz, B. R., and G. P. Fulton. The use of the hamster
cheek pouch for the study of vascular changes at the
microscopic level. Anat. Record 120: 293-309, 1954.
81. Lutz, B. R., and G. P. Fulton. Smooth muscle and blood
flow in small blood vessels. In : Factors Regulating Blood
Flow, edited by G. P. Fulton and B. Zweifach. Washing-
ton, D. C. : Am. Physiol. Soc, 1958, pp. 13-24.
82. Lutz, B. R., G. P. Fulton, and R. P. Akers. The neuro-
motor mechanism of the small blood vessels in membranes
of the frog (Rana pipiens) and the hamster (Mesocricelus
auratus) with reference to the normal and pathological
conditions of blood flow. Exptl. Med. Surg. 8: 258-287,
'95°-
83. Mackenzie, D. W., A. O. Whipple, and M. P. Winter-
steiner. Studies on the microscopic anatomy and physi-
ology of living transilluminated mammalian spleens.
Am. J. Anat. 68: 397-456, 1941.
84. Malpighi, M. De pulmombus. Observations anatomical.
Bologna, 1661. Cited in: J. F. Fulton's Selected Readings in
the History of Physiology. Springfield, 111.: Thomas, 1930,
pp. 61-67.
85. Meyling, H. A. Structure and significance of the periph-
eral extension of the autonomic nervous system. J.
Comp. Neurol. 99: 495-543, 1953.
86. Moore, R. L. Adaptation of the transparent chamber
technique to the ear of the dog. Anat. Record 64: 387-404,
>936.
87. Ni, T. G. Response of capillaries to various forms of
excitation. Am. J. Physiol. 62: 282-309, 1922.
88. Nicoll, P. A., and R. L. Webb. Blood circulation in the
subcutaneous tissue of the living bat's wing. Ann. N. Y.
Acad. Sci. 46: 697-709, 1946.
89. Nicoll, P. A., and R. L. Webb. Vascular patterns and
active vasomotion as determiners of flow through minute
vessels. Angiology 6: 291-310, 1955.
90. Noer, R. J. The blood vessels of the jejunum and ileum:
A comparative study of man and certain laboratory
animals. Am. J. Anat. 73: 293-334, 1943.
gi. Olkon, D. M., and M. Joannides. The capillary circula-
tion in the alveolus pulmonalis of the Using dog. A. M. A.
Arch. Internal. Med. 45: 201-205, 1930.
92. Olkon, D. M., and M. Joannides. Capillaroscopic
appearance of the pulmonary alveoli in the living dog.
Anat. Record 45: 121-127, 1930.
93. Parpart, A. K., A. O. Whipple, and J. J. Chang. The
microcirculation of the spleen of the mouse. Angiology
6:35°-362. !955-
94. Peck, H. M., and N. L. Hoerr. The intermediary
circulation in the red pulp of the mouse spleen. Anat.
Record 109: 447-477, 1951.
95. Perlman, H. B., and R. S. Kimura. Observations of the
living blood vessels of the cochlea. Ann. Otol. Rhinol. &
Larynol. 64: 1176-1192, 1955.
96. Perlman, H. B., and R. S. Kimura. Physiology of the
cochlear blood vessels. Angiology 6: 383-390, 1955.
97. Poor, E., and B. R. Lutz. Functional anastomotic
vessels of the cheek pouch of the hamster. Anat. Record
132: 121-126, 1958.
98. Prichard, M. M. L., and P. M. Daniel. Arteriovenous
anastomoses in the human external ear. J. Anat. 90:
309-317, 1956.
g9. Provenza, D. V., and S. Scherlis. Coronary circulation
in dog's heart. Demonstration of muscle sphincters in
capillaries. Circulation Research J: 318-324, 1959.
100. Reynolds, S. R. M., M. Kirsch, and R. J. Bing. Func-
tional capillary beds in the beating, KCl-arrested and
KCl-arrested-perfused myocardium of the dog. Circulation
Research 6: 600-611, 1958.
101. Rogers, J. B. Observations on pericapillary cells in the
mesenteries of rabbits. Anat. Record 54: 1-8, 1932.
102. Rogers, J. B. Observations in vivo on the capillaries in
the greater omentum of the cat. Anat. Record 63: 193-202,
1935-
103. Rouget, C. Memoire sur le developpement de la tunique
contractile des vaisseaux. Compt. rend. Acad. sci. 79: 559,
i873-
104. Sandison, J. C. A new method for the microscopic study
932
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
105.
ml.
107.
108.
109.
"3-
1 14.
116.
117.
118.
"9
123.
124.
■25-
126.
of living growing tissues by the introduction of a trans-
parent chamber in the rabbit's ear. Anal. Record 28:
281-287, 1924. 127.
Sandison, J. C. Observations on the circulating blood
cells, adventitial (Rouget) and muscle cells, endothelium 128.
and macrophages in the transparent chamber of the
rabbit's ear. Anal. Record 50: 355-379, 1 931 .
Sandison, J. C. Contraction of blood vessels and observa-
tions on the circulation in the transparent chamber of the
rabbit's ear. Anal. Record 54: 105-127, 1932. 129.
Saunders, E. A., and M. H. Knisely. Living mesenteric
terminal arterioles before and immediately after emboli-
zation. A. M. A. Arch. Pathol. 58: 309-344, 1954. 130.
Seneviratne, R. D. Physiological and pathological
responses in the blood vessels of the liver. Quart. J. Exptl.
Physiol. 35: 77-110, 1950. 131.
Seymour, J. C. Observations on the circulation of the
cochlea. J. Laryngol. & Otol. 68: 689-711, 1954.
Smith, C. A. Capillary areas of the cochlea of the guinea
pig. Laryngoscope 61: 1073-1095, 1951. 132.
Smith, C. A. Capillary areas of the membranous laby-
rinth. Ann. Otol., Rhinol. & Larynol. 63: 435-447, 1954.
Smith, P. E., and W. M. Copenhaver. Bailey s Textbook
of Histology. Baltimore: Williams & Wilkins, 1958.
Smith, R. D , and R. P. Giovacchine. On vascular 133.
patterns in red and white muscles. Anal. Record 118:
355-35°. '954-
Snook, T. A comparative study of the vascular arrange- 134.
ments in mammalian spleens. Am. J. Anal. 87: 31-78,
■95°-
Snook, T. The histology of vascular terminations in the 135.
spleen. Anat. Record 130: 711-730, 1958.
Spalteholz, W. Die Vertheilung der Blutgefasse im
Muskel. Abhandl. mat/i.-phys. CI. sdchs. GeseUsch. Wis- 136.
sench. 14: 509, 1888.
Spanner, R. Neue Befunde iiber die Blutwege der Darm- 137.
wand und ihre funktionele Bedeutung. Morphol. Jahrb.
69 : 394-454. >932- !38'
Staple, P. H., and A. L. Copley. Observations on the
microcirculation in the gingiva of hamsters and other
laboratory animals. Circulation Research. 7: 243-249, 1959. 1 39.
Steinach, E., and R. H. Kahn. Echte Contratilitat und
motorische innervation der Blutcapillaren. Pfliigers Arch,
ges. Physiol. 97: 105-133, 1903. 140.
Trueta, J., A. E. Barclay, P. M. Daniel, K. J.
Franklin, and M. M. L. Pritchard. Studies of the Renal
Circulation. Oxford: Blackwell, 1953. 1 41 .
Van Dobben-Broekema, M., and M. N. J. Dirken.
Reactions of the vessels of the rabbit's ear in response to
heating the body. Acta Physiol. Pharmacol. Neerl. 1: 562- 142.
583. '95°-
Van Dobben-Broekema, M., and M. N. J. Dirken. 143.
Influence of the sympathetic nervous system on the
circulation in the rabbit's ear. Acta Physiol. Pharmacol.
Neerl. 1 : 584-602, 1950. 144.
Vimtrup, B. Beitrage zur Anatomie der Kapillaren.
Z. ges. Anat. 65: 150-182, 1922. 145.
Wakim, K. C, and F. C. Mann. The intrahepatic
circulation of blood. Anat. Record 82: 233-253, 1942.
Walder, D. Arteriovenous anastomoses of the human 146.
stomach. Clin. Sci. 11 : 59-71, 1952.
Walls, E. W. The microanatomy of muscle. In : Structure
and Function of Muscle, vol. 1, edited by G. H. Bourne.
New York: Academic Press, i960, pp. 21-61.
Wi \rn, J. T. The extent of the capillary bed of the heart.
J. E.xptl. Med. 47: 273-291 1928.
Wearn, J. T., A. C. Ernstene, A. W. Bromer, J. S.
Barr, W. J. German, and L. J. Aschiesche. The normal
behavior of the pulmonary blood vessels with observations
on the intermittence of the flow of blood in the arterioles
and capillaries. Am. J. Physiol. 109: 236-256, 1934.
Webb, R. L., and P. A. Nicoll. Persistence of active
vasomotion after denervation (motion picture). Federation
Proc. 1 1 : 169, 1952.
Webb, R. L., and P. A. Nicoll. The bat wing as a
subject for studies in homeostasis of capillary beds. Anal.
Record 1 20 : 253-263, 1 954.
Wkille, F. L., S. R. Gargano, R. Pfister, D. Martinez,
and J. W. Irwin. Circulation of the spiral ligament and
stria vascularis of living guinea pig. A. M. A. Arch. Otol-
aryngol. 59: 731-738, 1954.
Weille, F. L., D. E. Martinez, S. R. Gargano, and
J. W. Irwin. An experimental study of the small blood
vessels of the spiral ligament and stria vascularis of living
guinea pigs during anaphylaxis. Laryngoscope 64: 656-665,
!954-
Whipple, A. O, A. K. Parpart, and J. J. Chang. A
study of the circulation of the blood in the spleen of the
living mouse. Ann. Surg. 140: 266-269, 1954.
Wiedeman, M. P. Effect of denervation on diameter and
reactivity of arteries in the bat wing. Circulation Research
3:618-622, 1955.
Wiedeman, M. P. Effect of venous flow on frequency of
venous vasomotion in the bat wing. Circulation Research
5:641-644, 1957.
Wiedeman, M. P. Response of subcutaneous vessels to
venous distention. Circulation Research 7: 238-242, 1959.
Wiedeman, M. P. Pressure variations in small veins in the
hindleg of the dog. Circulation Research 8: 440-445, i960.
Williams, R. G. An adaptation of the transparent
chamber technique to the skin of the body. Anat. Record
60:493-499, 1934.
Williams, R. G. The microscopic structure and behavior
of spleen autographs in rabbits. Am. J. Anat. 87: 459-503,
■95°-
Williams, R. G., and B. Roberts. An improved tantalum
chamber for prolonged microscopic study of living cells
in mammals. Anat. Record 107: 359-374, 1950.
Wyman, L. C, and C. Tum Suden. Vascular responses to
histamine in normal and in suprarenalectomized rats.
Am. J. Physiol. 99: 285-297, 1932.
Zweifach, B. W. A micro-manipulative study of blood
capillaries. Anat. Record 59: 83-108, 1934.
Zweifach, B. W. The structure and reactions of the
small blood vessels in Amphibia. Am. J. Anat. 60: 473_
5'4, '937-
Zweifach, B. W. The character and distribution of the
blood capillaries Anat. Record 73: 475-495, 1939-
Zweifach, B. W. Indirect methods for regional blood
flow. In : Methods in Medical Research, edited by V. R.
Potter. Chicago: Yr. Bk. Pub., 1948, vol. 1, pp. 131-139-
Zweifach, B. W. Basic mechanisms in peripheral vascular
homeostasis. In : Transactions of the Third Conference on
Factors Regulating Blood Pressure, edited by B. W. Zweifach
PATTERNS OF THE A-V PATHWAYS
933
and E. Schorr. New York: Josiah Macy, Jr. Foundation.
[949. PP- !3"52-
147. Zweifach, B. W. Direct observation of the mesenteric
circulation in experimental animals. Anat. Record 120:
277-288, 1954.
148. Zweifach, B. W. General principles governing the
behavior of the microcirculation. Am. J. Med. 23 : 684-696,
■957-
149. Zweifach, B. W. Structural and functional aspects of
the microcirculation of the skin. In: The Microcirculation.
Urbana, 111.: Univ. Illinois Press, 1959, pp. 144-150.
150. Zweifach, B. W., and C. E. Kossman. Micromanipula-
tion of small blood vessels in the mouse. Am. J. Physiol.
120: 23-35, !937-
151. Zweifach, B. W., and D. B. Metz. Selective distribution
of blood through the terminal vascular bed of mesenteric
structures and skeletal muscle. Angiology 6: 282-289, '955-
152. Zweifach, B. W., and D. B. Metz. Regional differences
in response of terminal vascular bed to vasoactive agents.
Am. J. Physiol. 182: 155-165, 1955.
153. Zweifach, B. W., and D. B. Metz. Rat mesoappendix
procedure for bioassay of humoral substances acting on
peripheral blood vessels. Ergeb. Anat. Eniwicklungsgeschichte
35: 176-239, 1956.
CHAPTER 28
Resistance (conductance) and capacitance
phenomena in terminal vascular beds1
HAROLD D. GREEN
CARLOS E. RAPELA
MARGARET C. CONRAD2
Department of Physiology and Pharmacology, Bowman Gray
School of Medicine, Wake Forest College, Winston-Salem, North Carolina
CHAPTER CONTENTS
Resistance Vessels
Pressure-Flow Relations in Vascular Beds
Methods
Passive curvilinear relationship of pressure-flow plots
Mathematical relationships
Effects of Viscosity on Pressure-Flow Relationships
Effects of Extravascular Pressure on Pressure-Flow Relation-
ships
Effects of Alterations of Venous Pressure
Autoregulatory Control of Resistance Vessels
Modification of passive pressure-flow relationship by
autoregulation
Artifacts induced in autoregulation studies by pump per-
fusion schemas
Autoregulation in different vascular beds
Reactive hyperemia
Mechanisms responsible for autoregulation
Interpretation of change of vasomotor tone induced by
constrictor and dilator agents in vascular beds which
demonstrate autoregulation
Chemical Effects on Resistance Vessels
Extrinsic Control of Resistance Vessels
Effects of vasomotive agents on total resistance in a
vascular bed
Segmental Resistances in Vascular Beds
Methods
Effects of Changes of Perfusion Pressure and Venous Pressure
on Large Artery and Vein, and Distal Small Vessel
Pressures
Effects of Extrinsic Agents on Segmental Resistance
1 Preparation of this chapter and most of the original work
reported herein were aided by Grant H-487, National Heart
Institute, United States Public Health Service.
2 Formerly trainee, Cardiovascular Graduate Training
Program (HTS-539.2) National Heart Institute, United States
Public Health Service; currently Fellow of the American
Heart Association.
Blood Volume in Vascular Beds (Vascular Capacity)
Methods
Effects of Various Factors on Vascular Volume
Estimation of Change of Vascular Volume Due to Extrinsic
Influences
Pulsatile Changes in Vascular Volume
Interpretation of Vascular Behavior from Measurements of
Flow, Pressure, and Vascular Volume
since most vascular beds do not permit direct
microscopic study, indirect methods have to be used
to evaluate them. In this chapter, the behavior of
vascular beds is deduced from recordings of the rate
of blood flow, the accompanying small vessel pressures,
and the changes in vascular volume that occur as the
result of varying the artery to vein pressure difference
across the bed and as a result of other intrinsic and
extrinsic influences.
Using the above measurements, the role of the
terminal vascular beds is analyzed in terms of the
behavior of those segments which determine the
resistance to flow through the bed, i.e., the resistance
vessels, and those segments which are related to the
volume of blood contained in a terminal bed at any
moment, i.e., the capacitance vessels. These functions
of the terminal vascular beds are shown to be in-
fluenced by such physical factors as arterial perfusion
pressure, presence of communication with collateral
vascular beds, viscosity of the blood, extravascular
pressure, venous pressure, by local autoregulation,
and by extrinsic factors such as vasoactive agents and
the autonomic nerves.
935
936 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
RESISTANCE VESSELS
Pressure-Flow Relations in I 'ascular Beds
methods. Pressure-flow relations in vascular beds
have been studied by two principal methods. In one,
the bed was perfused by a constant flow pump at
various flow rates while the artery to vein pressure
differences were recorded. More usually inflow or
outflow from the vascular bed was recorded while the
perfusion pressure was varied by using a constant
pressure pump or by varying the degree of com-
pression of the artery supplying the vascular bed. In
most cases, the arterial and venous pressures were
recorded together with the flow. Varying perfusion
pressure by altering the degree of compression of the
supplying artery has the advantage of providing a
more nearly physiological situation with a minimum
of complicating equipment, but does not allow
exploration of the pressure-flow relationships above
the animal's existing mean arterial pressure. Flow
has been measured by orifice meters, rotameters,
Perfusion
Aorta
cannulo — ^
Inflow
V 1
B
Arterial pressure
- meter
/\, Arterial pressure
^\ gouge
K
Metered ortery-
(cognote)
— mb
Arteria
Colloleral artery
It"
M*h one
stamotic
channel
25
— >-
1 .
1 /
/' / X
tt:
UJ
my 1 / /
15
/ ' 1/
"5
]/r
_Q
UJ
S / 1
cc
/ 1
Uj
/ '
5
— 1-
/ 1
Uj
5 ^^^
^^ n/
/ 1
1
0
g
1
Pressure in
- 5
- ;fc
O
/ C
ollateral Artery
J
1
-IB
_ U.
1 -I
1 1
1
1 1 1 1
0 40 80
PERFUSION PRESSURE
120 160
N METERED ARTERY
fig. 2. Relationship between perfusion pressure and arterial
inflow. I — curve representing the direct relationship between
arterial pressure and flow through the cognate bed when clamps
were applied to the arterial anastomotic channels as at point C
in fig. 1. II — artifactual curve obtained if clamp C remains
open when the arterial pressure supply to the cognate bed is
varied. Note that at perfusion pressures higher than that existing
in the collateral artery, the metered flow is greater than the
flow through the cognate bed, and at pressures a little below
that in the collateral artery, the inflow is less than that through
the cognate bed while at some lower perfusion pressure back-
flow is recorded. Ill — artifactual curve obtained if clamps are
applied to the arterial supply to the collateral bed as at point
B in fig. 1 . Note that while no backflow is obtained, the metered
inflow is greater than the flow through the cognate bed at all
levels of pressure, and the magnitude of the error increases with
the perfusion pressure. [Modified from Green el at. (41)-]
Cognate bed-
Venous pressure
gauge -
anastomotic channel
Outflow
meter
-Collateral bed
-Collateral vein
s Venous pressure
gauge
fig. 1 . Diagram of collateral communications between a
metered (cognate) capillary bed and collateral bed. Inflow
meter — meter used to measure flow through the metered or
cognate bed; outflow meter — meter used to measure the outflow
from the metered or cognate bed; anastomotic channels — com-
munications between cognate bed and collateral bed on the
arterial and venous sides, respectively, of the metered or cognate
bed. A — clamp used to occlude the inflow to the metered bed
when perfusing it with fluid at various pressures; B — position
where clamp might be placed to occlude the arterial supply to a
collateral bed ; C and D — positions where clamps must be placed
in order to occlude the communications between cognate and
collateral beds; E — point of occlusion to prevent outflow from
a collateral bed. [Modified from Green el al. (41).]
electromagnetic flowmeters, drop recorders, and
Gaddum-type ordinate recorders (20, 27-29, 46).
In studying pressure-flow relationships in intact
vascular beds, it is necessary that all anastomotic
communications with collateral vascular beds be
occluded. If they are not, then, during measurement
of inflow at perfusion pressures above the animal's
mean arterial pressure, part of the perfusion fluid will
leak across the anastomotic channels from the cog-
nate into collateral vascular beds (fig. i-a), giving
an inflow which is higher than the true flow through
the cognate bed (fig. 2— II). Similarly, at perfusion
pressures below the animal's mean arterial pressure,
blood will leak across the communicating channels
from the animal's collateral arteries to the metered
(cognate) vascular bed (fig. i-b); as a result the
metered inflow will be less than the actual flow
through the cognate bed. At some lower pressure
inflow in the metered artery will cease and, at still
lower pressures, backflow from the artery will be
recorded, thus giving an entirely false picture of the
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
937
behavior of the cognate vascular bed (41). Analysis
of the rate of backflow may be useful, however, as a
measure of the effectiveness of existing communica-
tions between cognate and collateral arteries.
Errors analagous to those on the arterial side may
occur when recording venous outflow, due to the
presence of postcapillary communications with
veins draining collateral vascular beds (fig. i-c, d).
PASSIVE CURVILINEAR RELATIONSHIP OF PRESSURE-FLOW
plots. The simplest relationships between pressure
and flow occur in vascular beds which do not show
autoregulation such as skin or "nonreactive" hind
limbs (46, 88, 115) or the pulmonary vascular bed
(78). All such curves are curvilinear with the con-
vexity toward the pressure axis (fig. 3). In such
studies changes of vasomotor tone have occurred
spontaneously (46) or have been induced by infusion
of epinephrine (88). With increase in vasomotor tone
the curves rotate toward the pressure axis so that for
any given level of pressure, flow is less (fig. 3). In all
such experiments, using blood as the perfusate, the
curves approach the zero flow axis asymptotically. A
'critical closing pressure," such as was described by
Burton (6) and others (10, 37) was not noted in the
above studies (see also 32, 35, 64). Curves of this
type may be said to exhibit a "passive" relationship
between pressure and flow.
mathematical relationships. When the data from
the above experiments are plotted on log-log paper,
approximately straight lines with varying foci and
slopes are obtained (fig. 3). It appears, therefore, that
the mathematical relationship between flow and
pressure is a power function
F = c X Pn
where F = flow in ml per min, c is a constant, P =
arteriovenous pressure difference in mm Hg, and n
is an exponent having a value between 1 and 3 (46).
The lowest value of n and the highest value of c were
found at "low vasomotor tone" (fig. 4, point A) and
vice versa (fig. 4, point C). Levy & Share (74) have
confirmed these findings and demonstrated that with
maximal dilation induced by a 10-min period of
ischemia and subsequent perfusion with hypoxic
blood, the value of n is 1 .0. The relationship of c to n
48
»—
— A
A
F
495
X
10
-3
p 1.768
• — .
•
B
F
4 67
X
10
-5
p 2.315
■ —
— ■
C
F
2.23 X
10
-6
p2 735
50.0
/ /
1 /' ■ J
80 120 160
PRESSURE
10 20 50 100
PRESSURE
300
fig. 3. Plots of the arteriovenous difference pressure vs. the blood flow in a cutaneous (saphenous)
bed at three levels of spontaneous "vasomotor tone." Left half, plotted linearly; right half, plotted on
log-log paper. Triangles represent the lowest level of vasomotor tone; circles represent an intermediate
level and squares, the highest level of vasomotor tone. The figures in the upper left-hand corner of the
graph represent the parameters for the straight lines reproduced in the log-log plot and for the curvi-
linear lines reproduced in the linear plot. Flow — ml/min; pressure — mm Hg. [Modified after Green
et al. (46).]
938
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
200
O ■ PRUf VS C
• ■ PRUp vs c
A . PRUF vs n
A. . PRUp vs n
x - c vs n
i i i i i i
6 8 I
i i i i ii i
4 6 8 1
4 6 8 1
i I i
4 6
XIO"
X 10"
X 10"
X 10"
fig. 4. Log-log plots of the interrelations of the parameters from fig. 3. Open circles — plot of the
relationship of the resistance at a How of 5 ml/min to the constant c; solid circles — plots of the rela-
tionship of the resistance at a perfusion of 100 mm Hg to the constant c; open triangles — plots of the
resistance at a flow of 5 ml/min to the exponent n; solid triangles — plots of the relationship of the re-
sistance at a pressure of 100 mm Hg to the exponent n; X — plot of the relationship of the constant c
to the exponent «; points labeled A, B, and C refer to curves A, B, and C, respectively, in tig. 3. Ordi-
nate scale applies to resistance at a flow of 5 ml/min (PRUf), to resistance at a perfusion pressure of
100 mm Hg (PRUp), and to n. Abscissal scales apply to n (left graph) and to the constant c (right
graph). See Table 1 for further identification of symbols.
in the above experiments (46) was also approximately
linear in a log-log plot (fig. 4).
There are at least two possible explanations for a
value of n greater than 1 . a) The apparent viscosity
of whole blood decreases as the pressure difference
across a length of rigid capillary tubing 0.3 mm or
less in diameter is increased, i.e., as flow increases
(40) ; this is due probably to the red cells being
clumped progressively closer together as a solid core
in the middle of the tube as the rate of flow increases,
leaving a sleeve of plasma adjacent to the intima of
the blood vessel and thus reducing viscous drag, b)
Other factors being constant, flow through a conduit
is proportional to the fourth power of the diameter.
If, with increasing internal pressure, a slight but
progressive increase in the diameter of the resistance
vessels occurs, then flow through vascular beds con-
taining such distensible structure will increase in
proportion to some power of P greater than 1 [Green
el al. (46); Folkow (27-29, 32)]. From the data in
figure 4 it appears that these effects become aug-
mented with increase in "tone" of the resistance
vessels. Computations, based on data compiled by
Burton (7), indicate that an increase of internal
pressure from o to 102 mm Hg in an arteriole might
increase the cross-sectional area sufficiently that the
relative conductance would be 146 per cent of that
at zero pressure, i.e., at the unstretched diameter
(fig. 5). However, Baez & Lamport (2) report that
arterioles of 34 to 42 /u diameter showed essentially no
change in diameter under considerable pressure
variation. They did note selective closing of pre-
capillary sphincters at positive pressures.
The relationship of resistance to flow, PRU
(peripheral resistance unit = P/F = mm Hg (ml/
min)), to either flow or A-V pressure difference was
also a power function in the above studies; for each
level of tone the resistance to flow varied inversely
with either flow or perfusion pressure (table 1).
Since both the constant and the exponent varied as
vasomotor tone changed, the relationship of resist-
ance at one level of tone to that at another (B A;
C/A in table 1 ) was also a power function of either
flow or pressure; this ratio, which was inversely re-
lated to pressure and flow, can be expressed as a
number if the same pressure (i.e., 100 mm Hg =
PRUPl00) or flow (i.e., 5 ml/min = PRUF.) is used
for each curve (table 1).
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
939
CC 00
£ II
5 6
4 -
Relotive
Pressure
(|-"r7 ' 0 0 01 0.09
0 4
P(mm.Hg)= 7.5xlO"4 »Y« f/r (1
Y = Young's modulus (-^f5 for
r cm'
•t'
100%
extension
)
^J
■
for Elostic Fibers = 3xl0 ,
Burfon
(1954)
.«■»•
: r/r°
= woll thickness/unstretched rodlus
■ stretched/unstretched radius
s
+ r
t
l 1
1
1
t
Pressure
(mm Hg)
Per Cent
Extension
0 II
0 30
0 I
33
102
256
10
0.5
181
565
1470
100
1.0
362
Vessel
-Aorto
1130 -Med. Art.
3940— Arteriole
»/r.
(mm)
2'l2 5
'/2
°2/0,5
Relative
Conduct. I 104 1.46
(Flow/Press.)
16
fig. 5. Relationship of internal pressure to radius in elastic tubes. Radius is expressed as ratio of
the radius at any pressure to the unstretched radius (r/r0). Pressure is plotted over the range from
zero to 1 , 1 representing that pressure at which the vessel extends indefinitely, and zero the pressure
at which the vessel is unstretched. Corresponding values of actual pressure, opposite the bracket, are
given in mm Hg; below this is given the per cent extension of the vessel at each of the levels of rela-
tive pressure, and below that is given the conductance which the vessel would have relative to that in
its unstretched state. Conductance is expressed as the ratio of Mow to pressure drop along unit length
of the vessel. These plots arc calculated from data in Burton (7).
It is of interest that c' vs. n' and c" vs. n" (table 1)
both plotted as straight lines on log-log paper, as did
c vs. n in figure 4. It is of interest also that the lines on
the log-log plot in figure 3 approach each other at
high pressures and flows. As a result, if the data
could be extrapolated to such values, a point would
be found at which the resistance in state B would
equal that in A (i.e., P = 5,000, F = 16,000) and
another point at which the resistance in state C
would equal that in A (i.e., P = 2,950, F = 5,400).
In a log-log plot these two points are so close together
that a common point of intersection could be assumed
for all three lines — A, B, and C. This suggests that a
rise of perfusion pressure acts to overcome the con-
strictor tone, and that this effect is proportionally
greater the higher the vasomotor tone.
From the above data it appears that the most satis-
factory method for defining "vasomotor tone" in
passive vascular beds is by means of a pressure-flow
plot, or by means of the equation for such plot. The
best quantitative expression for the comparison of
vasomotor tone at one moment with that at another
is to determine the plot of the pressure-flow relation-
ship during a control phase of vasomotor tone and to
compare this with a similar plot obtained in the
experimental period (see fig. 3 and table 1, columns
B/A, C/A, lines II and III). Often this mode of ex-
pression is impractical because of the difficulty in
maintaining vasomotor tone constant, particularly
in the experimental period. A more practical com-
promise for expressing change of vasomotor tone may
be to determine the pressure-flow relationship over a
suitable range of pressures and/or flows during the
control period and to compare isolated experimental
observations of pressure and flow with this control
curve (see figs. 21 and 22 and table 1, columns
B/A, C/A, lines IV, V).
From an inspection of figures 3 and 4 and table 1 ,
it appears that comparison of the experimental per-
fusion pressure with that required in the control
period to induce the same rate of flow (table 1 ,
lines III and V) may provide a ratio which approxi-
mates the apparent separation of the lines more
nearly than does the ratio of pressures at constant flow
(table 1, lines II and IV). This would provide merit
in perfusing passive (nonautoregulating) vascular
beds at a constant rate of flow while recording the
94-0 HANDBOOK OF PHYSIOLOGY -" CIRCULATION II
table i. Mathematical Relationships Between Resistance and Pressure and Between Resistance and Flow
A
B
c
B/A
C/A
I
F = c X P-
495Xio-3-P1768
4.67Xiir5P2 31S
2.23XlO-6-P2735
po.547
1.05X 102
po 967
2.22X 103
11
c'
PR Up =
P"'
2.02X IO2
1 ■
2.14X104
pi. 315
4.40X1O5
pi. 735
1.05XI02
pO .547
2.22X103
pO.967
III
PRUf = —
F°"
20.3
p 0.434
74
p0.668
Il6
3-65
|7 0.134
5-72
p 0.635
p 0.2 01
IV
PRUPlM
5-93
5°-3
152
8-49
25.6
V
PRUf„
10. 1
29.7
4I.8
2-9
4.I4
The columns correspond to the curves: A = solid triangles; B = solid circles, and C = solid squares in fig. 3. PRUp =
resistance in mm Hg/(cc/min) expressed as a function of the arteriovenous difference in pressure in mm Hg; PRUf = resistance
expressed as a function of the rate of flow; PRUp1M = resistance computed at an arteriovenous pressure difference of 100 mm
Hg; PRUfb = resistance calculated at a flow of 5 cc/min; B/A and C/A = ratios of values in columns B and C, respectively,
to those in column A. Note — columns B/A and C/A: for lines II and IV these values are also reciprocal ratios of flow at con-
stant arteriovenous pressure difference and for lines III and V they are also direct ratios of arteriovenous difference of pressure
at constant rate of flow.
F = c X P"
P
PRUp
PRUf = - =
p
C"1
c'
c-P» ~
pn-1 ~
p7'
pi/n
C-l/n
c"
Cl/n.p
pi-l/n
p„
F
The data in this table were computed from results reported by Green et al. (46).
arteriovenous difference of pressure during experi-
mentally induced changes of vasomotor tone.
On the other hand the convergence of the pressure-
flow plots at high pressures, discussed above, suggests
a secondary influence of change of perfusion pressure
on vascular distensibility and measured resistance.
On this basis there is merit in using a constant per-
fusion pressure and allowing the flow to vary with
experimentally induced changes of vasomotor tone
rather than keeping the flow constant and allowing
the perfusion pressure to be the dependent variable.
At present we cannot find sufficient grounds for a
decision between the two methods of perfusion when
studying "passive" vascular beds.
In those vascular beds which show autoregulation
(p. 948) anything which induces a change of flow,
i.e., alteration of perfusion pressure, tends to be count-
ered by an active change of vasomotor tone which
will tend to maintain flow constant. In this type of
bed it is particularly desirable to have data on the
control steady-state pressure and flow relationships
in order to make adequate quantitative comparisons
with experimental data.
Effects of Viscosity on Pressure-Flow Relationships
The viscosity of blood relative to that of water
increases nonlinearly with red cell concentration,
0 20 40 60 80 100
CORPUSCULAR CONCENTRATION (P C)
fig. 6. Mean and probable error of the apparent viscosity
of blood relative to saline in a glass viscometer and in the
hind limb of a dog at different corpuscular concentration.
[Redrawn after Whittaker & Winton (115).]
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
94'
varying from around 2 with pure plasma to around 5
at high hematocrit readings when measured in a low
velocity viscosimeter (fig. 6). Lower relative viscosities
are obtained with high velocity viscosimeters and
still lower relative viscosities are noted in perfused
organs. The relative viscosity in the latter two systems
decreases as the pressure difference and flow rate are
raised. Most of the viscosity of normal blood is due
to the suspended red cells, but the effect of the cells
is slight until the hematocrit begins to exceed 30
per cent (40, 74, 115); however, the plasma proteins,
particularly the globulins, contribute significantly
(112). In terms of effective oxygen delivery to the
tissues, a hematocrit of around 45 appears to represent
the best compromise between viscosity and O2
carrying capacity (50).
Effects of Extravascular Pressure on
Pressure-Flow Relationships
In most vascular beds extravascular pressure
exerts little effect. However, in muscle vascular beds,
a marked increase in resistance to flow occurs with
contraction. This is exemplified best in the myo-
cardium in which during svstole a rise in resistance
FLOW
PERIPHERAL PRESSURE
fig. 7. Records of systemic arterial pressure (BP), and
lateral pressure (CP) and moment-to-moment flow (F) in
the descending ramus of the left coronary artery during the
period labeled "flow." During the period labeled "peripheral
pressure" flow was interrupted by occlusion of the coronary
artery inflow proximal to the site of pressure measurement so
that the gauge recorded "peripheral coronary artery pressure."
Note that the latter pressure begins to rise during the phase of
isometric contraction that precedes the rise of systemic arterial
pressure, and that the peripheral coronary pressure begins to
fall with the onset of protodiastole, just before the incisura in
the systemic arterial pressure. [Reproduced in modified form
from Denison & Green (19).]
occurs which closely parallels intraventricular pres-
sure in magnitude and duration, as shown in figure 7
(19. 43 >•
Coles & Gough (10) applied external pressure to a
cup applied to a digit while observing the capillaries
with a microscope. They noted that arrest of capillary
flow occurred consistently at cup pressures of 32 to
60 mm Hg in subjects with mean brachial artery
pressures of 80 to 1 20 mm Hg obtained with the
sphygmomanometer. They spoke of the pressure at
which flow ceased as the critical closing pressure and
reported that it rose with the arterial pressure in
hypertensives and fell with digital vasodilation
induced by body warming. The use of the term
"critical closing pressure" in this sense seems to us to
be ambiguous. Quite possibly, in their experiments
the pressure decreased progressively from the brachial
artery to the small digital arteries. If this were the
case the vessels in the digits may have collapsed when
the extravascular pressure just exceeded the intra-
vascular pressure. However, since the true intra-
luminal pressure of the vessels which close was
unknown, the role of the elastic forces producing
critical closure (see Burton, Chapter 6, vol. 1, sect. 2,
of this Handbook) as against the role of simple mechani-
cal collapse can hardly be differentiated. This makes
it quite difficult to assign a figure for the critical
closing pressure if indeed one may use that concept
here.
Cerebrospinal fluid pressure may have a tendency
to vary directly with cerebral blood flow; however,
artificially induced changes of cerebrospinal fluid
pressure have little effect on flow unless the pressure
is elevated above arterial pressure (76 and un-
published data).
Effects of Alteration of Venous Pressure
When extremities were perfused with varying
pressures, while venous pressure was altered simul-
taneously so as to maintain artery to vein pressure
difference constant, flow still varied with the level of
the arterial pressure (89). The authors conclude that
some vascular structures were dilated as the arterial
(and total) pressure throughout the vascular bed
rose. In the supine anesthetized dog, inspiration was
accompanied by a rise of intra-abdominal pressure
and of small vein pressure in the hind leg. Widely
opening the abdomen abolished both (113). The
small vein pressure effects apparently were trans-
mitted peripherally from the inferior vena cava.
942
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
0
Verfebrol Artery Pressure
85
43
79
47 jr
ir°*
b2 5 --e^
Cerebral Venous Outflow
16
12.5
Autoperfused
Perfused with pump
J L
J_
J L
J I L
0 20 40 60 80 100
Meon between Carotid and Vertebral Pressures -
Cerebral Outflow Pressure mmHg
fig. 8. Autoregulation in the cerebral vascular bed of the
dog. Above: records representing vertebral arterial pressure in
mm Hg and cerebral venous outflow in ml/min. Below: heavy
solid line, relationship between the cerebral venous outflow and
the mean of the carotid and vertebral artery pressures minus
the cerebral outflow pressure during autoperfusion of the
brain from the carotid arteries. The perfusion pressure was
regulated by means of clamps on the carotid arteries. Light
solid line, similar relationship but during the perfusion of the
carotid arteries with an artificial perfusion system.
Alitor egulatory Control of Resistance Vessels
For the purpose of this chapter, we propose that
the term autoregulation be defined to include all
processes which operate locally in a vascular bed to
maintain some factor constant in the face of various
externally or internally induced stresses. The factor
which is kept constant, i.e., the controlled variable
(see below) may be blood flow, or the tissue concen-
tration or tension of some nutrient (O2, etc.) or some
metabolite (such as CC^)- As proposed here, the term
autoregulation would exclude extrinsic mechanisms
such as reflexes involving the central nervous system,
variation in arterial pressure, or changes in hormone
activity.
The activity of autoregulatory mechanisms has been
studied by subjecting isolated or semi-isolated organs
to stresses such as changes in arterial perfusion pressure
or blood gas content, or by altering tissue metabolism
while recording the resulting change, or lack of change
in blood flow, venous gas content, or the content of
other metabolites.
MODIFICATION OF PASSIVE PRESSURE-FLOW RELATION-
SHIP by autoregulation. In many vascular beds, the
above-mentioned power relationship between per-
fusion pressure and flow in passive vascular beds is
modified by occurrence of autoregulation. The insert
in the upper portion of figure 8 shows recordings of
blood flow in the brain obtained during autoperfu-
sion (76 and unpublished data). In these studies,
carotid artery inflow pressure was lowered abruptly
from 84 to 44 mm Hg; after flow stabilized, perfusion
pressure was returned to the control level. Immedi-
ately upon lowering perfusion pressure, flow dropped
from 12.5 to 8.5 ml per min, then rose to a stabilized
level of approximately 10 ml per min. Upon restora-
tion of control pressure, flow rose abruptly to 16 ml per
min and then stabilized at approximately its original
level of 1 2.5 ml per min. The rise in flow following the
initial decline probably was due to vasodilation, and
the secondary decline in flow following; restoration of
the original perfusion pressure probably was due to
vasoconstriction. When data from such experiments
are plotted they yield a series of curves such as are
reproduced in figure 9.
The heavy line in the graph in figure 8 corresponds
to the heavy line in figure 9 and is a plot of the
stabilized flows at each level of perfusion pressure from
85 to 15 mm Hg. This line is almost horizontal,
indicating an almost constant level of stabilized flow
over the pressure range of 80 to 35 mm Hg. In view
of the observations in figures 3 and 4, this finding
can be explained only by assuming some reactive
change in the diameter of the resistance vessels so as
to compensate for alteration of perfusion pressure
(76 and unpublished data). The mechanism respon-
sible for compensation for pressure change has been
termed local reaction of the arterial wall, reactive
vasodilation, intrinsic regulation, autoregulation and
''genuine autoregulation" (3, 27-29, 36, 46, 63, 66,
7'2, 9°. 92. 99» I05» I09> "O-
Observations, similar to those in the brain, were
recorded in artificially perfused skeletal muscle
vascular beds (fig. 10). At pressures above normal the
pressure-flow relationship was curvilinear and similar
to that described for skin (fig. 3) and the resistance to
flow, after stable flow was established, increased
progressively as perfusion pressure was lowered.
However, as perfusion pressure dropped below 100
mm Hg the curve began bending more sharply to the
left so as to become approximately horizontal to the
pressure axis, and the stabilized resistance to flow
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
943
PRESSURE
fig. 9. Sequential flow readings following a series of square
wave changes of perfusion pressure in a vascular bed showing
autoregulation. The pressure-flow points are numbered suc-
cessively As,, Ai, At, A3, and A, for the first square wave change
of pressure (on both insert and graph). The points for the
second pressure change are numbered similarly Bn-Bt and for
the third, C0-Ca. Light solid line represents the pressure-flow
relationship which would be found if vasomotor tone remained
constant at the level which existed during perfusion at a
pressure corresponding to Aq for a period of time sufficient to
establish a steady-state flow at this pressure. Dash-dot line
represents the corresponding pressure-flow plot which would
be found if the vasomotor tone were to remain constant at a
level corresponding to that found when steady-state flow was
established at a perfusion pressure corresponding to point B2-
Dash-dash line represents the pressure-flow plot which would
correspond with the steady-state flow established for perfusion
pressure A? assuming no change in vasomotor tone were to
occur with subsequent change of perfusion pressure. These
three plots then represent three levels of vasomotor tone. Heavy
solid line represents the actual "steady state" pressure-flow rela-
tionship after reactive changes have occurred in the vasomotor
tone following each change of perfusion pressure; it is the plot
characteristic of autoregulation.
decreased with further lowering of perfusion pressure
down to about 50 mm Hg. This progressively de-
creasing resistance tended to maintain flow relatively
constant over the range of arterial pressure from 90
down to approximately 50 mm Hg. Below 50 mm
Hg arterial pressure, flow dropped rapidly and
approached zero at a pressure of 10 to 20 mm Hg.
When the pressures and flows in figure 1 oA are
plotted on log-log graphs (fig. 10B), the portion
corresponding to the pressures above 90 mm Hg plots
as a straight line with a slope greater than 1 ; this por-
tion of the curve corresponds to a "passive" relation-
ship between pressure and flow (see above). On the
other hand, in the range between 50 and 90 mm Hg
the slope is less than 1. In the equation F = cPn,
the corresponding values of n are greater than and
less than 1, respectively. Values for n of less than 1 are
characteristic of autoregulation (28, 29).
ARTIFACTS INDUCED IN AUTOREGULATION STUDIES BY
pump perfusion schemas. Failure to detect auto-
regulation in vascular beds has been attributed to
occurrence of some alteration in the vascular bed as a
result of changes in the blood due to contact with
artificial structures or to traumatization of the blood
by perfusion pumps (28, 29, 66). This effect is notice-
able particularly in the cerebral vascular bed, as
shown in figure 8 (76 and unpublished data).
The heavy line in figure 8 represents the stable
pressure-flow relationship in a cerebral vascular bed
during an initial study when the pressure was regu-
lated by compressing the arteries supplying the brain.
The light line gives the stable pressure-flow relation-
ship during a subsequent period when a perfusion
pump was inserted in the arterial inflow circuit. With
the perfusion pump in operation, flow at all levels of
pressure was significantly above that with the brain
perfused directly from the heart. Furthermore, the
flow did not remain constant but increased regularly
200
delivery during the period of reactive
hyperemia underpaid the O2 debt accumulated
during the period of occlusion, if the muscle was at
rest during the occlusion the reactive hyperemia
overpaid the debt (9, 108).
Momentary overshoot of flow after a period of
occlusion was noted on occasion in the dog's paw.
However, comparison of weight changes (see below)
with the integral of the flow during the period of
overshoot suggested that the overshoot represented
refilling of small vessels, which had emptied by-
elastic recoil into the vein during the period of arterial
occlusion (114 and Rapela and Green, unpublished
data).
Reactive hyperemia following temporary occlusion
of the arterial supply is maximal in myocardium and
brain (49, 84; and Rapela et «/., unpublished data),
active in skeletal muscle, present in the mesenteric
artery bed and in kidney, but is almost absent in
spleen, skin (fig. 1 1, lower curve), hepatic artery, and
portal vein vascular beds (45), as shown in figure 12.
MECHANISMS RESPONSIBLE FOR AUTOREGULATION. Feed-
back loop. /) General concept. The engineer's feedback
loop provides a convenient way to visualize control
mechanisms (fig. 13). The controlled variable is that
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
945
measurement which the controlling mechanism is
attempting to keep constant (such as arterial pres-
sure, or tissue Os tension). The level of the controlled
variable may be disturbed by various loads, i.e.,
bleeding in the case of arterial pressure, or variations
of tissue metabolism in the case of tissue Oo tension.
The detector senses continuously the magnitude of the
variable and feeds the information to a summator or
discriminator where it is compared with the desired
value (set point) and modified by signals from other
loops. The resulting signal is then fed to an effector
which controls the activity of whatever process
is necessary to maintain the controlled variable
constant.
Input from
other loops
Set
point
Ui
I
control
signal (s)
Feedback signal
Load
Detector
fig. 13. Schematized diagram for a feedback loop (see text
or discussion).
2) Present evidence suggests that, in skeletal
muscle, a possible controlled variable in the feedback
loop for autoregulation is the tissue oxygen tension
(69). The detector is unknown; but the feedback loop
may involve adenine formed from adenosine triphos-
phate (ATP) in the presence of insufficient oxygen.
The summation point may be the receptor site on the
arteriolar smooth muscle, and the effector may be
the arteriole which controls the rate of blood flow by
means of which the controlled variable is regulated
(5)-
3) There are other possible controlled variables in
the feedback loop for autoregulation. The increased
flow that occurs in skeletal muscle during and follow-
ing a tetanic contraction may represent another
manifestation of this pattern of autoregulation
although axon reflexes in the sympathetic nerve
supply have been postulated as playing a role (62).
The vasodilation that occurs in skeletal muscle during
activation of the patellar reflex is thought to be due
to the same mechanism as that of the postcontraction
hyperemia ( 107).
The controlled variable responsible for auto-
regulation (and reactive hyperemia) may vary with
different vascular beds. Tissue oxygen tension
appears to be the controlled variable in the myo-
®
e/vjw/wt:
95"
Arterial
Pressure
mm Hg
Coronary
Flow
em'/min
-0
Control
90 sec 5% O2
Mean Flow
cm3/min 30
67
Control
Wean Flow
cm3/min 21
I mg NaCN
into Coronory Artery
fig. 1 4. Left anterior coronary artery inflows in the dog. A : in response to a 90-sec period of
breathing 5% Oo. B: effects of an intra-arterial injection of 1 mg of sodium cyanide. Note the calibra-
tion for flow is nonlinear (the deflection is approximately proportional to the square of the flow).
[Modified after Green & Wegria (49).]
946
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
cardium and skeletal muscle, since reducing the
oxygen content of the arterial blood produces a
rather large increase in coronary (fig. 14A) and in
skeletal muscle flow, whereas there is almost no change
in flow in either bed when arterial blood C02 content
is increased (12, 49). The fact that coronary flow is
influenced more by coronary artery Oo content than
tension (51) is not incompatible with the concept
that tissue O2 tension is the controlled variable. The
feedback loop in the heart must involve something
other than O2 tension, since intra-arterial injection of
cyanide causes as great an increase of coronary flow-
as does hypoxia (fig. 14Z?) (49).
The increase in blood flow in the brain in response
to decreased arterial blood oxygen content is relatively
minor compared to that which follows an increase in
CO2 content (fig. 15) (44), suggesting that brain CO2
tension may be the controlled variable for this tissue.
Elevation of arterial blood hydrogen ion concen-
tration decreases resistance to flow through cutaneous
(18, 25), renal (23), and skeletal muscle (18) vascular
beds. Depression of the hydrogen ion concentration
below normal is accompanied by increase of resistance
to flow in skin (18, 25) and kidney (23); however, in
skeletal muscle depressed hydrogen ion causes about
the same degree of decrease of resistance to flow as
104 .
Arterial Pressure
mm Hg
0
123
94
Cerebral Flow
cm*/ mm
0
II 6
4' 30" 8% 02
100
Arterial Pressure
mm Hg
0-
Cerebral Blood Flow
cm3 /mm
5' 10% C02 90% 02
fig. 15. Records of cerebral venous outflow and systemic
arterial pressure during a 4.5-min period of breathing 8% O2
(upper pair of curves) and in response to a 5-min period of breath-
ing 10% CO; in 90% O2 (lower pair of curves) in the dog. Brain
was perfused directly from the aorta.
does an elevation of hydrogen ion concentration (18).
Effects of hydrogen ion concentration on myocardial
blood flow are reported to be the reverse of those
in skin and kidney (38). These findings suggest that
hydrogen ion might be one of the controlled variables
in autoregulation.
Role of nervous system in autoregulation. Autoregulation
is prominent in denervated vascular beds. High
activity in the extrinsic nerves may even minimize or
prevent manifestation of autoregulation; for instance,
central reflex effects of hypoxia may overpower the
local dilatory effect in the anesthetized dog's inner-
vated skeletal muscle vascular bed (75) (see also
p. 943). Autoregulation in kidney is not abolished
by procaine anesthetization, adrenergic blocking
agents, or gamma-aminobutyric acid (m), suggest-
ing that the feedback loop contains no essential link
that responds pharmacologically as does nervous
tissue.
The behavior of cerebral blood flow in response to
changes of perfusion pressure (see above) cannot be
stated conclusively to represent strict autoregulation,
since in these studies a reflex neural mechanism was
not excluded. However, no influence of extrinsic
autonomic constrictor nerves upon cerebral blood
flow has been demonstrated conclusively (45), and
therefore, it is unlikely that an autonomic reflex is
involved in the cerebral studies.
Myogenic theory of autoregulation. Bayliss (3) proposed
from studies of dog's hind legs and isolated arteries,
that the arterial wall responds directly to a rise of
intraluminal pressure by an increase in its state of
contraction (or tone) sufficient to bring about a
reduction in the lumen of the vessel (and presumably
therefore, a decrease in the flow through the vessel).
This concept received support from Johnson (66)
who failed to find an appropriate correlation between
change in the O2 concentration of the venous blood
draining an isolated segment of gut and the occurrence
of a decreased resistance to flow as perfusion pressure
was lowered. Since he found, also, no correlation with
nerve activity, gut contraction, or presence of meta-
bolites, Johnson concluded that the autoregulatory
change of flow represented a myogenic response.
Waugh & Shanks ( 1 1 1 ) observed that hypothermia
(3-10 C), intrarenal infusion of chloral hydrate, or
high concentrations of procaine abolished auto-
regulation but that anoxic perfusion did not depress
the autoregulatory reaction; since nerve block also
did not interfere (see above) they concluded that
renal autoregulation is due to "•myogenic vaso-
motion." Folkow (27) also postulated a myogenic
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
947
basis for autoregulation in skeletal muscle since it
was not abolished by breathing oxygen at either high
or low partial pressure. It should be noted that the
above conclusions regarding the myogenic theory are
based solely on negative evidence. It seems quite
unlikely to us that a vascular wall could respond
appropriately to changes of intraluminal pressure
per se.
Physical factors related to autoregulation. Renal blood
flow at a given level of arterial pressure was the same
whether the perfusion pressure was steady or pulsatile.
Autoregulatory changes in resistance were observed
with both types of perfusion (93).
A rise in renal interstitial and intrarenal venous
pressure was found to parallel an elevation of arterial
pressure over the ''autoregulatory range." This
finding is proposed as the explanation for the auto-
regulatory rise in renal vascular resistance that
accompanies an elevation of renal arterial pressure
(65, 96). On the other hand, two other groups of
investigators (81, 111) could not account for the
observed autoregulation in their dogs' kidneys on the
basis of such changes in intrarenal tissue or venous
pressures.
An increase in postglomerular viscosity which
parallels glomerular filtration rate (116) has been
proposed to explain the ''autoregulatory" increase in
renal vascular resistance that accompanies a rise of
arterial perfusion pressure above 80 mm Hg. How-
ever, Selkurt et al. (100) found that arterial perfusion
pressure could be varied between 100 and 160 mm
Hg without significant change in blood flow (para-
minohippurate clearance), glomerular filtration rate
(creatinine clearance), or filtration fraction. In their
experiment, therefore, the postglomerular viscosity
remained unchanged, and the autoregulatory varia-
tion in renal vascular resistance must have occurred
solely in the preglomerular vessels. An increase in
effective viscosity of the blood flowing in the cortical
layers due to plasma skimming in the intralobular
arteries (cell separation theory) has been proposed by
Pappenheimer & Kinter (72, 87) to explain renal
autoregulation. However, Waugh & Shanks (111)
were able to demonstrate autoregulation in the kidney
using a cell-free perfusate, so long as the fluid con-
tained plasma. Evidently simple physical phenomena
will not serve to explain renal autoregulation.
Enlargement of collateral communications following
occlusion of the cognate arterial supply as a manifestation of
autoregulation. It is well known that, following occlu-
sion of an artery, collateral communications between
the cognate bed and collateral arteries enlarge rapidly
until within a few hours to weeks they can supply
almost a normal rate of flow to the cognate bed. Such
enlarged channels are demonstrated beautifully
during arteriography. The dilation of the communica-
ting channels may be considered a special case of
autoregulation, although almost nothing is known
regarding its mechanism of action. It does not appear
to be brought about by any special change in arterial
pressure proximal to the occlusion. The enlargement
of the communicating channels is more likely related
to an increased rate of flow or enhanced pressure drop
through the communicating channels (60).
Summary of present status of feedback control of auto-
regulation. Though the mechanism of the autoregula-
tion of blood flow has not been established as yet, the
following trends may be stated, a) Most likely there
is more than a single factor involved and the pre-
dominant one may vary in different organs. In the
kidney, for example, maintenance of a constant
glomerular filtration rate may be more significant
than satisfaction of the metabolic requirements of the
organ; consequently the sensing mechanism to regu-
late blood flow should be related directly or indirectly
to glomerular filtration. The juxtaglomerular appa-
ratus may serve this function (97, 111). In organs
such as a skeletal muscle and the heart, metabolism
fluctuates rapidly; in these a mechanism must be
available to allow adaptation of flow to the varying
metabolic demands. Such mechanism should be
capable of sustaining the metabolic activity in the
face of fluctuations in arterial pressure. In either case
the sensing mechanism may detect the adequacy of
supply (i.e., tissue Oa tension) or the adequacy of
removal of metabolic products (i.e., tissue CO2
tension). It appears probable that the former is sensed
in heart and muscle and the latter in the brain, b)
Autoregulation may be absent in vascular beds such
as the paw or skin which have very low metabolic
requirements, c) Whatever the mechanism of auto-
regulation, at times it appears to be dependent on the
presence of certain "normal factors" in plasma
required for maintenance of "normal vascular tone,"
and at other times to be masked by certain "abnormal
factors" which may induce either "abnormally high"
or "abnormally low vascular tone."
INTERPRETATION OF CHANGE OF VASOMOTOR TONE IN-
DUCED BY CONSTRICTOR AND DILATOR AGENTS IN VAS-
CULAR BEDS WHICH DEMONSTRATE AUTOREGULATION.
In artificially perfused beds. When studying responses to
vasoconstrictor and vasodilator stimulation in a vas-
cular bed which demonstrates autoregulation it may
948
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
be advantageous to use a constant-flow technique.
This is particularly the case if the autoregulation is
operating to maintain flow in proportion to metabolic
need. Under these circumstances the change in
arteriovenous difference of pressure probably will re-
flect the action of the infused agent with a minimum
of complication. On the other hand, if constant pres-
sure perfusion is used the autoregulatory mechanism
may mask the effect of the vasoactive agent by provid-
ing a counterdilation or constriction in an attempt
to maintain flow constant (see p. 940).
Interpretation of change of vasomotor tone in autoperfused
vascular beds. Intravenous infusion of a constrictor agent
which raises systemic arterial pressure may fail to
alter blood flow in a vascular bed showing autoregu-
lation. This lack of response may be interpreted
incorrectly as indicating that the agent has exerted a
vasoconstrictor effect. Such misinterpretation has been
made in the case of the cerebral vascular bed. In such
instances the vasoactive agent should be injected
directly into the arterial supply to the vascular bed.
If there should be no response to the intra-arterial
injection then the lack of response during the rise of
systemic arterial pressure with intravenous injection
of the agent would be due to local autoregulation. In
this case the vascular bed would be attempting to
maintain flow constant despite the rise of arterial
pressure (90b).
Chemical Effects on Resistance Vessels
Hypertonic solutions (above 5 '!c NaCl) decrease
the resistance to flow in systemic vascular beds (56)
Control
by an unknown mechanism, but increase that in the
lung. The latter seems to be due to intravascular red
cell agglutination (91).
Potassium and magnesium ions cause active limb
arteriolar dilation, calcium induces constriction (54)
while sodium has little effect (86). Acetate, among
the anions, produces arteriolar dilation (82).
Extrinsic Control of Resistance Vessels
EFFECTS OF VASOACTIVE AGENTS ON TOTAL RESISTANCE
in a vascular bed. Extrinsic control of resistance
vessels, i.e., of the peripheral resistance in the various
vascular beds is illustrated most typically by the
responses in a skeletal muscle vascular bed, since
reactions, characteristic of all beds, are present in this
bed. In a skeletal muscle bed, intra-arterial injections
of epinephrine and of levarterenol cause a marked
decrease in flow followed, often, by a secondary rise
above control level (fig. 16). Such response occurs
characteristically after all injections of levarterenol
and after injections of 1 fig or more of epinephrine.
Smaller amounts of epinephrine often induce either
no response or an increase in flow indicative of
vasodilation. Lumbar sympathetic chain stimulation
usually decreases flow (fig. 16), although occasionally
an initial increase followed by decrease or solely an
increase in flow occurs (30, 45, 1 17).
After induction of adrenergic blockade (fig. 16),
levarterenol may have no effect or may cause a slight
increase in flow while both epinephrine and lumbar
sympathetic chain stimulation increase flow. Atropine
injected intra-arterially abolishes the increase in flow
After Phenoxybenzamine 0 3rng/Kg
58
26
30
14
26
fig. 16. Curves of arterial inHow in a skeletal muscle vascular bed in the dog in response to intra-
arterial injections of 10 /jg of epinephrine, 10 /ng of levarterenol (norepinephrine), and a 1 -min period
of stimulation of the lumbar sympathetic chain, during a control period {left half) and after an intra-
arterial injection of 0.3 mg/kg of phenoxybenzamine (right half). Flow measured in ml/min; arterial
pressure remained constant through the study. [Modified after Youmans et at. (1 17).]
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
949
CORTEX
Metabolic Products (?)
No Known blockade —
5 Receptor (dilator)
Receptor (Constrictor)
No known blockade
Circulating Vasoconstrictor
(Angiotensin)
Epinephrine
*»
c c « c c
0 « t °<
0 1 0 w 0.
0 „ ^ X
l*J
— — S.oe-,
■4n £ §. S
8 5
\ 6 * "
\\ 5 ? g
0.
\\ \ -• ~
^J
•\
-- ^ •
c
*
\ *
0
\\
«s
\ '
\ O
c 0
\
\ *
§s
\
'
\
\
•s
t \
t»*
*N
*.
'n
\ \
0
-J
Control
Increased precapillary resistance —
— decreased
flow t decreased local vascular v
olume
Increased postcapillary resistance
— decreased
flow* increased vascular volume
fig. 26. Hypothetical plots of the pressure drops in various
portions of the terminal vascular bed during a control state,
solid line; during a state of increased precapillary resistance,
dash-dot line; and increased postcapillary resistance, dashed line.
of the resistance increase in the morphological pattern
of the vascular network. A lessening of resistance in
any part of the network will presumably induce oppo-
site changes in vascular volume. These structures are
affected differentially by various vasoactive agents.
Data are insufficient at present to draw significant
generalizations regarding the various vascular beds.
PULSATILE CHANGES IN VASCULAR VOLUME
Pulsatile flow through a dog's ulnar artery was
measured with a square wave electromagnetic flow-
meter together with pressure recorded in a small
branch just proximal to the flowmeter and with paw
volume pulse recorded plethysmographically (fig. 27).
The flow record showed a dicrotic notch but never
fell to or below zero during diastole. A volume pulse
calculated by integrating the flow pulse and sub-
tracting an assumed constant venous output was
essentially similar (fig. 27). Both showed two humps
of approximately similar magnitude.
In man, the normal digital volume pulse rises
relatively more rapidly than that of the dog's paw,
has a sharp peak at the end of the first quarter of the
pulse interval, and a slight notch about halfway down
the descending limb (fig. 28). If the supplying artery
is occluded but adequate collateral circulation is
available, the digital pulse shows a peak which is
rounded and delayed, and the notch on the descending
limb is absent. In the presence of vasospastic disease,
the peak may be delayed slightly, the notch raised or
may occur sooner on the descending limb, and the
area of pulse per unit amplitude increased in com-
parison with the normal. With elevation of venous
pressure or interference with venous outflow, the peak
of the pulse is sharper than normal and a second hump
Pulsatile arterial inflow Pulsatile orteriol pressure Pulsatile plethysmogroph vol Calculated pulsatile vol
fig. 27. Records of A — pulsatile arterial inflow; B — pulsatile arterial pressure; and C — pulsatile
plethysmographic volume, in the dog's paw. Arterial inflow recorded with an electromagnetic meter
on the ulnar artery. D — pulsatile volume calculated from the pulsatile arterial inflow, assuming a
constant venous outflow.
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
957
Normal
Arterial occlusion
Vosospasm
Venous obstruction
fig. 28. Averaged plethysmographic pulses from the digits of patients. A — normals; B — patients
with arterial occlusion, but with good collateral circulation; C — patients with vasospasm; D — pa-
tients with deep thrombophlebitis. All pulses are redrawn to the same amplitude and same time dura-
tion. Note : the flow to pulse ratios in the patients with arterial occlusion were 2 to 5 times those of
the normals.
follows the notch, indicating a larger than normal
reflected wave component in the volume pulse (11).
INTERPRETATION OF VASCULAR BEHAVIOR
FROM MEASUREMENTS OF FLOW,
PRESSURE, AND VASCULAR VOLUME
Flow through a vascular bed is dependent upon
arterial pressure, arteriolar inflow and venous outflow
resistances, viscosity of the blood, and extravascular
pressure. Physiologic control of flow is exerted in many
vascular beds by local autoregulation as modified by
the influence of autonomic nerves and by circulating
constrictor substances.
Analysis of the influences of autonomic nerves and
circulating vasoactive substances under conditions of
various physiological stresses is complicated, particu-
larly if there is an accompanying change of arterial
pressure. For instance, if there is a decrease in arterial
pressure to 50 per cent of the control, the measured
peripheral resistance might increase in a nonreactive
bed such as the skin, whereas the resistance might
decrease in a reactive bed such as that of kidney,
brain, or skeletal muscle. These changes would occur
in the absence of extrinsic influence. Therefore, in
order to analyze the potency of extrinsic influences
upon the resistance vessels, it is necessary to obtain
previous data on the behavior of the resistance vessels
during changes of pressure per se in the absence of
extrinsic influences, and then compare the measured
changes in resistance with these pre-established
findings before drawing any conclusions as to the
influence of extrinsic factors on the vascular bed.
Similar observations apply to measurements of
vascular volume. The latter measurements become
important, particularly in conditions such as shock
in which it is presumed that there is a stagnation and
pooling of blood in various vascular beds; however,
extensive data are not as yet available on such
changes in vascular volume.
REFERENCES
Ahlquist, R. P. A study of the adrenotropic receptors.
Am. J. Physiol. 153: 586-600, 1948.
Baez, S., and H. Lamport. On the nature of the un-
changing diameter in isolated microscopic vessels under
pressure variation. Physwlog'st 3 (No. 3): 13, i960.
Bayliss, W. M. On the local reactions of the arterial
wall to changes of internal pressure. J. Physiol., London
28: 220-231, 1902.
Beck, L., and M. J. Brodv. Physiology of vasodilatation.
Angiology 12: 202-222, 1 961.
Berne, R. M. Nucleotide degradation in the hypoxic
heart and its possible relation to regulation of coronary
blood flow. Federation Proc. 20: 101, 1961.
Burton, A. C. Laws of physics and flow in blood sessels.
In: Visceral Circulation (Ciba Foundation Symposium).
London: Churchill, 1952, pp. 70-86.
Burton, A. C. Relation of structure to function of the
tissues of the wall of blood vessels. Physiol. Revs. 34: 619-
642, 1954.
Chorobski, J., and W. Penfield. Cerebral vasodilator
nerves and their pathway from the medulla oblongata
with observations on the pial and intracerebral vascular
plexus. A.M. A. Arch. Neurol. Psychiat. 28: 1 257-1 289, 1932.
Coffman, J. D., and S. L. Javett. Reactive hyperemic
flow and oxygen usage of contracting skeletal muscle.
Federation Proc. 21 : 104, 1962.
Coles, D. R., and K. R. Gough. The critical closing
pressure of blood vessels of the fingers in hypertensive
and normal subjects. Clin. Sci. 19: 587-594, i960.
Conrad, M. C, and H. D. Green. Skin temperature
and digital plethysmography in arterial vascular diseases.
Circulation 24: 908, 1961.
958
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
12. Crawford, D. G, H. M. Fairchild, and A. C. Guyton. 33
Oxygen lack as a possible cause of reactive hyperemia.
Am. J. Physiol. 197:613-616, 1959.
13. Davis, D. L., Segmental vascular responses to sympathetic
stimulation. Federation Proc. 21 : 120, 1962. 34.
14. Davis, D. L., and W. F. Hamilton. Small vessel responses
of the rabbit ear. Am. J. Physiol. 196: 1312-1315, 1959.
15. Davis, D. L., and W. F. Hamilton. Small vessel responses 35.
of the dog paw. Am. J. Physiol. 196: 1 316-132 1, 1959.
16. Davis, D. L., and VV. F. Hamilton. Cross circulation at
the small blood vessel level in the dog paw. Am. J.
Physiol. 199: 1 169—1 1 73, i960. 36,
17. Day, S. B., and J. A. Johnson. Pressure-flow relation-
ships in the isolated perfused rabbit heart. Am. J. Physiol.
196: 1289-1291, 1959. 37.
18. Deal, C. P., Jr., and H. D. Green. Effects of pH on blood
flow and peripheral resistance in muscular and cutaneous 38.
vascular beds in the hind limb of the pentobarbitalized
dog. Circulation Research 2: 148-154, 1954.
ig. Denison, A. B. Jr., and H. D. Green. Effects of auto-
nomic nerves and their mediators on the coronary 39.
circulation and myocardial contraction. Circulation
Research 6: 633-643, 1958. 40.
20. Denison, A. B. Jr., M. P. Spencer, and H. D. Green. A
square wave electromagnetic flowmeter for application
to intact blood vessels. Circulation Research 3: 39-46, 1955. 41.
21. Driscol, T. E., T. W. Moir, and R. W. Eckstein.
Interarterial pressure gradients in concept of autoregula-
tion of coronary blood flow. Federation Proc. 21 : 106, 1962. 42.
22. Dumke, P. R., and C. F. Schmidt. Quantitative measure-
ments of cerebral blood flow in the macacque monkey.
Am. J. Physiol. 138: 421-431, 1943.
23. Emanuel, D. A., M. Fleishman, and F. J. Haddy.
Effect of pH change upon renal vascular resistance and 43.
urine flow. Circulation Research 5: 607-611, 1957.
24. Fishback, M. E., L. Burnett, and A. M. Scher. Auto-
regulation of coronary blood flow in the dog heart. 44.
Clm. Research 7: 60, 1959.
25. Fleishman, M., J. Scott, and F. J. Haddy. Effect of 45.
pH change upon systemic large and small vessel resistance.
Circulation Research 5: 602-606, 1957.
26. Fog, M. Cerebral circulation. II. Reaction of pial arteries 46.
to increase in blood pressure. A.M. A. Arch. Neurol.
Psychiat. 41 : 260-268, 1939.
27. Folkow, B. Intravascular pressure as a factor regulating
the tone of the small vessels. Acta Physiol. Scand. 1 7 : 289-
3io. >949- 47-
28. Folkow, B. A study of the factors influencing the tone of
denervated blood vessels perfused at various pressures.
Acta Physiol. Scand. 27: 99-1 17, 1952. 48.
29. Folkow, B. A critical study of some methods used in
investigations on the blood circulation. Acta Physiol.
.Scand. 27: 1 18-129, '952-
30. Folkow, B. Nervous control of the blood vessels. Physiol. 49.
Revs. 35:629-663, 1955.
31. Folkow, B. Effects of catechol amines on consecutive
vascular sections. In: Adrenergic .Mechanisms (Ciba 50.
Foundation Symposium). Boston: Little, Brown, i960,
pp. 190-200. 51.
32. Folkow, B., and B. Lofvinc. The distensibility of the
systemic resistance blood vessels. Acta Physiol. Scand.
38:37-52, 1957- 52-
Folkow, B., B. Johansson, and S. Mellander. The
comparative effects of angiotensin and noradrenaline on
consecutive vascular sections. Acta Physiol. Scand. 53:
99-104, 1961.
Fox, R. H., and S. M. Hilton. Bradykinin formation
in human skin as a factor in heat vasodilatation. J.
Physiol., London 142: 219-232, 1958.
Gilbert, R. P., L. B. Hinshaw, H. Kuida, and M. B.
Visscher. Absence of a general critical closing pressure
in the isolated perfused lung. Am. J. Physiol. 194: 160-164,
'958.
Ginsburg, M., and J. Grayson. Factors controlling liver
blood flow in the rat. J. Physiol., London 123: 574-602,
'954-
Girling, F. Critical closing pressure and venous pressure.
Am. J. Physiol. 171 : 204-207, 1952.
Goodyer, A. V. N., W. F. Eckhardt, R. H. Ostberg,
and M. J. Goodkind. Effects of metabolic acidosis and
alkalosis on coronary blood flow and myocardial metabo-
lism in the intact dog. Am. J. Physiol. 200: 628-632, 1961.
Gotoh, F. Effects of blood pressure on cerebral circula-
tion. Keio J. Med. 8: 13-29, 1958-59.
Green, H. D. Circulatory system: physical principles.
In: Medical Physics, 11. edited by O. Glasser. Chicago:
Yr. Bk. Pub., 1950, pp. 228-251.
Green, H. D., R. S. Cosby, and K. H. Radzow. Dy-
namics of collateral circulations. Am. J. Physiol. 140:
726-736, 1944.
Green, H. D., A. B. Denison, Jr., C. E. Rapela, and
G. Lin. L^se of indicator concentration curves in computa-
tion of mean rate of flow and volume of blood contained
within a segment of the vascular system. IRE Trans, on
Med. Electronics ME-d: 277-282, 1959.
Green, H. D., and D. E. Gregg. The relationship
between differential pressure and blood flow in a coronary
artery. Am. J. Physiol. 130: 97-107, 1940.
Green, H. D., B. Hfafner, and J. T. Anderson. Cerebral
circulation. Am. J. Physiol. 187:602, 1956.
Green, H. D., and J. H. Kepchar. Control of peripheral
resistance in major systemic vascular beds. Physiol. Revs.
39:617-686, 1959.
Green, H. D., R. N. Lewis, N. D. Nickerson, and A. L.
Heller. Blood flow, peripheral resistance and vascular
tonus with observations on the relationship between
blood flow and cutaneous temperature. Am. J. Physiol.
141 : 5' 8-536, 1944-
Green, H. D., K. Ottis, and T. Kitchen. Autonomic
stimulation and blockade on canine splenic inflow,
outflow and weight. Am. J. Physiol. 198: 424-428, i960.
Green, H. D., C. E. Rapela, and G. Lin. Simultaneous
determination, by dye measurements, of vascular volume
and conductance in dog's paw. Federation Proc. 20: 110,
1961.
Green, H. D., and R. YVegria. Effects of asphyxia,
anoxia and myocardial ischemia on the coronary blood
flow. Am. J. Physiol. 135: 271-280, 1942.
Guyton, A. C, and J. W. Crowell. Dynamics of the
heart in shock. Federation Proc. 20: 51-60, 1961.
Guz, A., G. S. Kurland, and A. S. Freedberg. Relation
of coronary flow to oxygen supply. Am. J. Physiol. 199:
179-182, i960.
Haddy, F. J. Vasomotion in systemic arteries, small
RESISTANCE AND CAPACITANCE PHENOMENA IN VASCULAR BEDS
959
vessels and veins determined by direct resistance measure-
ments. Minn. Med. 41 : 162-170, 1958.
53. Haddy, F. J. Peripheral vascular resistance. Am. Heart J.
60: 1 -5, i960.
54. Haddy, F. J. Local effects of sodium, calcium and
magnesium upon small and large blood vessels of the dog
forelimb. Circulation Research 8: 57-70, i960.
55. Haddy, F. J., M. Fleishman, and D. A. Emanuel. Effect
of epinephrine, norepinephrine and serotonin upon
systemic small and large vessel resistance. Circulation
Research 5: 247-251, 1957.
56. Haddy, F. J., and H. W. Overbeck. The effect of hyper-
and hypotonic solutions on small vessel resistance in the
dog forelimb. Physiologist 3 (No. 3): 71, i960.
57. Haddy, F. J., A. G. Richards, and M. B. Visscher.
Pressures in small and large veins and arteries. Am. J.
Physiol. 171 : 731, 1952.
58. Hardin, R. A., J. B. Scott, and F. Haddy. Relationship
of pressure to blood flow in the dog kidney. Am. J.
Physiol. 199: 1192-1194, i960.
59. Hartmann, H. , S. L. Orskov, and H. Rein. Die Gefass-
reaktionen der Niere im Verlaufe allgemeiner Kreislauf-
Regulationsvorgange. Pfliigers Arch. ges. Physiol. 238 :
239-250, i936-37-
60. Hasse, Von H. M., G. Rau, and W. Schoop. Die Bedeu-
tung von Druck und Durchstromung fur die Dilatation
der Kollateralgefasse bei Arterienverschliissen. Z. Kreis-
laufforsch .
1127-1133, 1959.
61. Hertzman, A. B. Vasomotor regulation of cutaneous
circulation. Physiol. Revs. 39: 280-306, 1959.
62. Hilton, S. M. Experiments on the post-contraction
hyperaemia of skeletal muscle. J. Physiol., London 120:
230-245, 1953.
63. Hinshaw, L. B., H. M. Ballin, S. B. Day-, and C. H.
Carlson. Tissue pressure and autoregulation in the
dextran-perfused kidney. Am. J. Physiol. 197: 853-855,
'959-
64. Hinshaw, L. B., and S. B. Day. Tissue pressure and
critical closing pressure in the isolated denervated dog
foreleg. Am. J. Physiol. 196: 489-494, 1959.
65. Hinshaw, L. B., R. D. Flaig, R. L. Logemann, and
C. H. Carlson. Intrarenal venous and tissue pressure
and autoregulation of blood flow in the perfused kidney.
Am. J. Physiol. 198:891-894, i960.
66. Johnson, P. C. Autoregulation of intestinal blood flow.
Am. J. Physiol. 199: 311 -318, i960.
67. Johnson, P. C, and K. M. Hanson. Effect of venous
pressure on blood volume and venous resistance in the
intestine. Federation Proc. 21 : 120, 1962.
68. Johnson, P. C, and E. E. Selkurt. Intestinal weight
changes in hemorrhagic shock. Am. J. Physiol. 193:
■35-143. I958-
69. Jones, R. D., and R. M. Berne. Skeletal muscle blood
flow regulation. Federation Proc. 20: 104, 1961.
70. Kelly, W. D., and M. B. Visscher. Effect of sympathetic
nerve stimulation on cutaneous small vein and small
artery pressures, blood flow and hindpaw volume in
the dog. Am. J. Physiol. 185: 453-464, 1956.
71. Kety, S. S., B. D. King, S. M. Horvath, W. A. Jeffers,
and J. H. Hafkenschiel. The effects of an acute reduction
in blood pressure by means of differential spinal sympa-
thetic block on the cerebral circulation of hypertensive
patients. J. Clin. Invest. 29: 402-407, 1950.
72. Kinter, W. B., and J. R. Pappenheimer. Role of red
blood corpuscles in regulation of renal blood flow and
glomerular filtration rate. Am. J. Physiol. 185: 399-406,
1956.
73. Langston, J. B., A. C. Guyton, and W. J. Gillespie, Jr.
Autoregulation absent in normal kidney but present after
renal damage. Am. J. Physiol, igg: 495-498, i960.
74. Levy, M. N., and L. Share. The influence of erythrocyte
concentration upon the pressure-flow relationships in
the dog's hind limb. Circulation Research 1 : 247-255, 1953.
75. Litwin, J., A. H. Dil, and D. M. Aviado. Effects of
anoxia on the vascular resistance of the dog's hind limb.
Circulation Research 8: 585-593, 1960.
76. Machowicz, P. P., G. Sabo, G. Lin, C. E. Rapela, and
H. D. Green. Effect of varying cerebral arterial pressure
on cerebral venous flow. Physiologist 4 (No. 3): 68, 1961.
77. Mall, F. Die Blut und Lymphwege im Diinndarm des
Hundes. Abhandl. Kgl. Sachs. Ges. Hiss. Math.-Physik Kl.
1888, vol. 14. (Quoted in Medical Physics II, edited by O.
Glasser. Chicago: Yr. Bk. Pub., 1950, 230J
78. Marshall, R. J., Y. Wang, H. J. Semler, and J. T.
Shepherd. Flow, pressure and volume relationships in
the pulmonary circulation during exercise in normal
dogs and dogs with divided left pulmonary artery. Circu-
lation Research 9:53-59, 196 1.
79. Mehrizi, A., and W. F. Hamilton. Effect of levarterenol
on renal blood flow and vascular volume in dogs. Am. J.
Physiol. 197: 1115-1117, 1959.
80. Mellander, S. Comparative studies on the adrenergic
neuro-hormonal control of resistance and capacitance
blood vessels in the cat. Acta Physiol. Scand. 50: Suppl.
176, 1-86, 1960.
81. Miles, B. E., M. G. Ventom, and H. E. deWardenkr.
Observations on the mechanism of circulatory autoregula-
tion in the perfused dog's kidney. J. Physiol., London 123:
'43-147. 1954-
82. Molnar, J. I., R. A. Renn, and F. J. Haddy. Local
effects of magnesium and acetate on vascular resistance
in the dog forelimb. Federation Proc. 20: g9, 1961.
83. Nakata, K., G. F. Leong, and R. W. Brauer. Direct
measurement of blood pressures in minute vessels of the
liver. Am. J. Physiol, igg: 1 181 -1 188, i960.
84. Olsson, R. A., and D. E. Gregg. Reactive hyperemia
characteristics of the myocardium. Federation Proc. 21 : 106,
1962.
85. Osher, W. J. Pressure-flow relationship of the coronary
system. .4m. J. Physiol. 172:403-416, 1953.
86. Overbeck, H. W., and F.J. Haddy. Acute effects of Na+,
K+, and Ca^ on vascular resistance in the dog forelimb.
Physiologist 3 (No. 3) : 122, i960.
87. Pappenheimer, J. R., and W. B. Kinter. Hematocrit
ratio of blood within mammalian kidney and its signifi-
cance for renal hemodynamics. Am. J. Physiol. 185: 377-
390, 1956.
88. Pappenheimer, J. R., and J. P. Maes. A quantitative
measure of the vasomotor tone in the hindlimb muscles of
the dog. Am. J. Physiol. 137: 1 87-1 99, 1942.
89. Phillips, F. A., Jr., S. H. Brind, and M. N. Levy. The
immediate influence of increased venous pressure upon re-
g6o
HANDBOOK OK PHYSIOLOGY
CIRCULATION II
sistance to flow in the dog's hind leg. Circulation Research: 3
357-362, 1955.
no. Rapela, C. E., E. J. Fox, S. Welborne, Jr., and H.
D. Green. Modification of pressure-flow relationship by
autoregulation. Federation Proc. 21 : 1 1 1, ig62.
qoa.RAPELA, C. E., and H. D. Green. Adrenergic blockade
by Dibozane. J. Pharm. Exper. Therap. 132: 29-41, 1961.
gob. Rapela, C. E.: P. Machowicz, and H. D. Green. Cere-
bral venous blood flow. Federation Proc. 20: 100, 1961.
91. Read, R. C, J. A. Johnson, J. A. Vick, and M.
\V Meyer. Vascular effects of hypertonic solutions. Cir-
culation Research ^8: 538-548, i960.
Riecker, G. liber die Beziehung zwischen Druck und
Stormstarke der portalen Lebergefasse. Pfliigers Arch,
ges. Physiol. 262: 37-50, 1955.
Ritter, E. R. Pressure/flow relations in the kidney:
Alleged effects of pulse pressure. Am. J. Physiol. 168:
480-489, 1952.
Roddie, I. C, J. T. Shepherd, and R. F. Whelan. The
contribution of constrictor and dilator nerves to the skin
vasodilatation during body heating. J. Physiol., London
136:489-497, 1957-
Sagawa, K., and A. C. Guvton. Pressure-flow relation-
ships in isolated canine cerebral circulation. Am. J. Physiol.
200: 71 1 -7 14, 1 96 1.
Scher, A. M. Autoregulation of renal blood flow. Federa-
tion Proc. 18: 138, 1959.
Schmid, H. E., and M. P. Spencer. Characteristics of
pressure-flow regulation by the kidney. J. Appl. Physiol.
17: 201-204, 1962.
98. Scott, J. B., R. A. Hardin, and F. J. Haddy. Pressure-
flow relationships in the coronary vascular bed of the dog.
Am. J. Physiol. 199: 765-769, i960.
Selkurt, E. E. The relation of renal blood flow to effec-
tive arterial pressure in the intact kidney of the dog. Am.
J. Physiol. 147:537-549. '946-
Selkurt, E. E., P. W. Hall, and M. P. Spencer. In-
fluence of graded arterial pressure decrement on renal
clearance of creatinine, />-aminohippurate and sodium.
Am. J. Physiol. 159: 369-378, 1949.
101. Selkurt, E. E., and P. C. Johnson. Effect of acute eleva-
tion of portal venous pressure on mesenteric blood volume,
interstitial fluid volume and hemodynamics. Circulation
Research 6: 592-599, 1958.
Selkurt, E. E., M. P. Scibetta, and T. E. Cull. Hemo-
dynamics of intestinal circulation. Circulation Research 6:
92-99, 1958.
92
93
94
95-
96.
97-
99-
100.
103.
104.
105.
106.
107.
108.
109.
"3-
114.
• 15-
116.
117.
102.
Senav, L. C.,Jr., M. Christensen, and A. B. Hertzman.
Cutaneous vascular responses in finger and forearm during
rising ambient temperatures. J. Appl. Physiol. 15:61 1-618,
i960.
Shadle, O. W., M. Zukof, and J. Diana. Translocation
of blood from the isolated dog's hindlimb during levartere-
nol infusion and sciatic nerve stimulation. Circulation
Research 6: 326-333, 1958.
Shipley, R. E., and R. S. Study. Changes in renal blood
flow, extraction of inulin, glomerular filtration rate, tissue
pressure and urine flow with acute alterations of renal
artery blood pressure. Am. J. Physiol. 167: 676-688, 1951.
Smith, H. W. The Kidney, Structure and Function in Health and
Disease. New York: Oxford Univ. Press, 1951 , p. 424.
Sonnenschein, R. R. Vasodilation in skeletal muscle
during activation of patellar reflex. Am. J. Physiol. 200:
685-688, 1961.
Stainsby, VV. N. Effect of muscle contractions on auto-
regulation of blood flow through skeletal muscle. Federa-
tion Proc. 20: 103, 1 96 1.
Stainsby, W. N., and E. M. Renkin. Autoregulation of
blood flow in resting skeletal muscle. Am. J. Physiol. 201 :
1 17-122, 1961.
VVaugh, W. H. Myogenic nature of autoregulation of
renal flow in the absence of blood corpuscles. Circulation
Research 6: 363-372, 1958.
Waugh, W. H., and R. G. Shanks. Cause of genuine
autoregulation of the renal circulation. Circulation Research
8:871-888, i960.
Wells, R. E., R. D. Perera, and E. W. Merrill. In-
fluence of plasma proteins upon blood viscosity. Federation
Proc. 21 : 94, 1962.
Wiedeman, M. P. Pressure variations in small veins in the
hind leg of the dog. Circulation Research 8 : 440-445, 1 960.
Wiederhielm, C. A., and R. F. Rushmer. Time course of
reactive hyperemia in isolated dog hind limbs. Federation
Proc. 20: 103, 1 96 1.
Whittaker, S. R. F., and F. R. Winton. The apparent
viscosity of blood flowing in the isolated hindlimb of the
dog, and its variation with corpuscular concentration. J.
Physiol., London 78: 339-369, 1933.
Winton, F. R. Hydrostatic pressures affecting the flow of
urine and blood in the kidney. Harvey Lectures 1951-52.
New York : Academic Press, series 47, pp. 21-52, 1953.
Youmans, P. L., H. D. Green, and A. B. Denison, Jr.
Nature of the vasodilator and vasoconstrictor receptors in
skeletal muscle of the dog. Circulation Research 3: 1 71-180,
'955-
CHAPTER 29
Exchange of substances through the capillary walls
E. M. LAND IS
J. R. PAPPENHEIMER1
Harvard Medical School, Boston, Massachusetts
CHAPTER CONTENTS
Filtration and Absorption; General Formulation
Capillary Blood Pressure, Pc
Methods of Measurement
Capillary Pressures in Various Tissues; Relation to the
Osmotic Pressure of the Plasma Proteins
Variability of Capillary Blood Pressures Under Control
Conditions
Functional Changes of Capillary Blood Pressure
Effects of Venous Pressures and of Venular Constriction on
Capillary Pressure
Osmotic Pressure of the Plasma Proteins, Upi
Methods of Measurement
Protein Osmotic Pressure of Human Plasma
Species Differences, Fetal Plasma
Physiological Significance of the Deviations from van't
Hoff's Law
Physicochemical Aspects of Protein Osmotic Pressure
Interstitial Fluid Pressure ('Tissue Pressure'), P,f
Proteins in Extracapillary Fluids; n,y
Capillary Filtrate from Limb Capillaries; Protein Content
Interstitial Fluid, Protein Content and n,y
Circulation of Interstitial Fluid; Circulation of Protein
Filtration Coefficients of Capillaries, kc; and of Tissues, k,
Normal Capillaries
Effects of Temperature on Filtration Coefficients
Adsorbed Plasma Protein and Filtration Coefficients
Effects of Injury on Filtration, Absorption, and Filtration
Coefficients
Capillary stasis
Filtration coefficients, kc, of injured capillaries
Capillary pressure in injury
Tissue asphyxia; relation of filtration coefficients to O2,
CO», and pH
Adrenal cortical hormones and filtration coefficients
Porosity of the injured capillary wall
Diffusion, General Principles
Free Diffusion
Diffusion Through Porous Membranes, Restricted Diffusion
Career Investigator, American Heart Association.
Diffusion and Hydrodynamic Flow, Relation to Pore Di-
mensions
Diffusion
Hydrodynamic flow
Simultaneous Flow and Restricted Diffusion; Theory of
Molecular Sieving
Distribution of Pore Sizes
Osmotic Pressure and Osmotic Flow Through Leaky Mem-
branes, Osmotic Reflection Coefficients
Transcapillary Movement of Lipid-Insoluble Molecules
Structure of Muscle Capillaries as Deduced from Permeability
Measurements and from Electron Microscopy. Quantita-
tive Aspects of Transcapillary Diffusion
Molecular Sieving of Large Molecules; Regional Differences
in Porosity
Capillary Permeability to Lipid-Soluble Molecules; Respira-
tory Gases
Capillary Permeability and Blood Flow in Relation to Exchange
of Materials Between Blood and Tissues
Blood-Tissue Transport of Oxygen
Blood-Tissue Exchange of Small, Nonmetabolized Molecules
or Ions
Nonuniform Distribution of Blood Flow in Relation to
Blood-Tissue Exchange
I . FILTRATION AND ABSORPTION; GENERAL FORMULATION
"transudation of water and solids" through the
walls of blood vessels was proposed by Bartholin
(10) in 1653 to explain the flow of lymph. This sug-
gestion was largely neglected though a somewhat
similar process was expressed vaguely by Hales
(140) in 1753 as an "insinuation of liquid" into the
wall of the intestine in connection with some of his
more prolonged perfusion experiments. Ludwig
(223) proposed a definite filtration theory in 1861
based largely upon observations made by Noll (263)
961
962
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
in his laboratory in 1850. According to Ludwig:
". . . the blood which is contained in the vessel
tends to equalize, through the porous vessel walls, its
pressure and its chemical composition with those of
the fluids which lie outside the vessels. If, for example,
the contents of the vessels increases, the pressure in
the vessels also increases, and immediately a portion
of blood passes out into the tissues, driven by a fil-
tration pressure."
But this "filtration pressure" proved unable, by
itself, to explain either the control of the volume of
lymph flow or the regulation of the constancy of
blood volume. Many of Ludwig's earlier experiments
supported his belief that this was accomplished by a
direct relationship between blood pressure, filtration,
and lymph formation, followed by return of this
lymph to the blood stream. Elevating venous pressure
in portions of the circulation of a whole animal in-
creased lymph flow, as did also elevating arterial
pressure in perfused tissues. However, others showed
very soon that elevations of blood pressure produced
by vasomotor changes did not always produce the
predicted increase of filtration. Moreover, little
lymph could be obtained from the resting limb;
whereas Ludwig's filtration hypothesis required that
even resting blood pressure should have produced
both filtration and lymph flow.
The problem became temporarily still more obscure
after 1880 when Heidenhain began studying the
abundant flow of lymph from the thoracic duct which
continued even during rest. The actions of his two
classes of lymphagogues, coupled with slight but
definite inequalities of solute concentrations in plasma
and lymph (explained now, in large part, by the
Gibbs-Donnan equilibrium) led him to postulate
active secretion by the cells of the capillary walls and
possibly by the lymphatics (145a). Heidenhain found
Ludwig's simple filtration theory adequate for some
conditions and quite unable to explain others. On
the other hand, Heidenhain's secretion theory was
supported by no direct proof. At this point Starling
measured the osmotic pressure of the plasma proteins
and added absorption to Ludwig's filtration. In 1896,
under the title "On the absorption of fluids from the
connective tissue spaces," Starling wrote:
". . . although the osmotic pressure of the proteids
of the plasma is so insignificant, it is of an order of
magnitude comparable to that of the capillary pres-
sures; and whereas capillary pressure determines
transudation, the osmotic pressure of the proteids
of the serum determines absorption." (345)
This hypothesis, despite its attractiveness, did not
find general acceptance for several decades until
improved methods were developed for measuring the
osmotic pressure of the plasma proteins and also
capillary blood pressure. Apparent exceptions to the
hypothesis became explicable as investigators learned
more about the nature of the capillary wall itself,
the hydrostatic pressure of the interstitial fluid, and
the osmotic pressure of the proteins in that fluid.
For purposes of summary, and of consecutive, more
detailed discussions of each factor, a general relation-
ship can be formulated. It must be emphasized, how-
ever, that this formulation is a composite which is
based on many overlapping experiments, each of
which dealt simultaneously with several of the varia-
bles, but not with all.
em. -- k(p - n. -p. +n.) (1.1)
+ - filtration
- - absorption
F.M. represents fluid movement through the
capillary wall, with a plus sign to indicate filtration,
and a minus sign to indicate absorption. Pr is capil-
lary blood pressure (hydrostatic); LTP;, the osmotic
pressure of the plasma proteins; P,<, the pressure in
the interstitial fluid compartment (hydrostatic);
and Hi/, the osmotic pressure of the proteins in the
interstitial fluid immediately outside the capillary
walls. The proportionality factor, k, has been called a
filtration constant or, more appropriately, a filtration
coefficient and is a measure of the permeability of
the capillary wall to isotonic fluid. Each of these
factors will be considered in succession.
2. CAPILLARY BLOOD PRESSURE, Pc
A. Methods of Measuremt ni
The pressure under which blood flows through the
capillary vessels was very much in the minds of the
earliest investigators even when pressure measure-
ments were limited to large blood vessels and to lower
animals. Thus Hales, in 1773, having determined
the first arterial and venous pressures, went on at
once to make certain assumptions and then calculated
the "force of the blood in the capillary vessels" to be
1.838 gr. with the qualification that to this "must
be added the velocity which the blood has acquired
at its first entrance in the capillary vessel, which can
be but small as appeared by the great resistance it
meets within the capillary vessels. . ." (140). In
1828 Poiseuille (286) devised the U-tube mercury
manometer and measured the gradient of pressure
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
963
in the arterial system. With smaller and smaller
cannulae he measured pressures in the aorta, carotid
artery, and even in a 2-mm branch of the crural
artery, and reported : "that a molecule of blood moved
with the same force throughout the course of the
arterial system, which a priori, with all physiologists,
we were far from thinking." It followed, therefore,
that the major fall of blood pressure must occur some-
where in the smaller vessels beyond the ones he can-
nulated. Poiseuille then turned his attention to
capillary tubes and studied the relation which volume
flow of liquids per unit time bears to pressure, viscosity,
tube radius, tube length, and wall surface (287-289).
Hence Poiseuille's equation, which underlies the
science of hydrodynamics, emerged from questions
concerning arterial and capillary blood pressure in
animals.
In 1875 von Kries (182) tried to measure capillary
blood pressure in man by an indirect method. He
placed a glass plate, 2 to 5 mm2 in area, on the skin
and hung from this plate a small scale pan on which
weights were placed until the skin blanched. Five
years later Roy & Brown (309) used a capsule fitted
with a distensible, transparent membrane to de-
termine, under the microscope, the pressures required
to modify or obstruct flows through single arterioles,
capillaries, and venules in the more or less trans-
parent tissues of experimental animals, e.g., the web
of the frog. From 1886 to the present, various modifica-
tions of these two basic methods were used for many
measurements but yielded discordant results, ranging
even in one species, man, and in one tissue, skin, from
1 to 71 mm Hg (207). Most of these studies were made
after 1900 because figures for capillary blood pressure
were necessary to prove or disprove Starling's filtra-
tion-absorption hypothesis. Even as late as 1925 no
conclusions could be reached because the lower values
were less than venous pressure and obviously ques-
tionable.
The higher values were criticized because they were
based on blanching of the skin or on arrest of blood
flow by microscopic examination, and so indicated
arteriolar rather than capillary pressure. Moreover,
no indirect method could yield information concern-
ing the presence or absence of a gradient of pressure
in the capillary network itself. When reviewed in
1934 (207) indirect methods were found inadequate
a) because of variable transmission of pressure through
overlying tissues to the capillaries beneath, and b)
because of the arbitrary and unproved criteria
adopted by various investigators to indicate when
externally applied pressure equaled the pressure
within the capillaries. Direct measurements of the
sort attempted by Poiseuille a century earlier were
still necessary.
The requirements for direct measurements of
pressure in single capillaries are basically simple,
though technically somewhat difficult (198, 203).
Figure 2.1 shows (upper left) a micropipette, 5 n
in diameter at its tip, under the microscope and
ready for use. A somewhat smaller pipette (lower
left) is shown inserted into a capillary of the frog's
mesentery. The micropipettes are first carefully
filled with a saline solution containing heparin,
mounted in a micromanipulator and connected to a
manometer and syringe (right) so that the pressure
exerted on the saline at the tip of the micropipette
can be changed rapidly and accurately to balance
the changing pressure in the capillary. The micro-
manipulator is required not only to insert the pipette
into the capillary, but also to keep the lumen of the
pipette in free communication with the lumen of the
capillary. Minute rods (fig. 2.1, upper left), each
controlled by its own micromanipulator, are fre-
quently necessary in addition to hold steady thin
tissues such as mesentery. Pressure readings from the
manometer can be made only at true pressure equi-
librium without net flow of liquid through the tip
of the pipette because orifices of 5 to 10 n interpose
considerable resistance to flow and consequent
inaccuracies. Failure to observe this precaution has
yielded fallaciously low values for capillary pressure
(36> 2°3)-
With these requirements in mind, suitable criteria
were developed for measuring mean, systolic, and
diastolic pressures in single capillaries, arterioles, or
venules in mesentery (198), skin (205), and muscle
of lower animals as well as in the skin of man (203)
with an accuracy of a few millimeters of water. Tests
showed that changes of capillary pressure induced
by graded venous congestion could be detected
promptly and accurately by the direct method (203)
but not by an indirect method (88).
B. Capillary Pressures in Various Tissues; Relation
to the Osmotic Pressure of the Plasma Proteins
Direct measurements of pressures in single capil-
laries, arterioles, and venules provided answers to
Poiseuille's questions concerning the nature and the
location of the pressure gradient in the circulatory
system. Figure 2.2 shows that in the mesenteric blood
vessels of the frog, the major decrease of pressure
(70 to 80%) occurred in the arterioles, but there
964
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
FIG. 2.1. Diagram of apparatus for measuring capillary blood pressure directly. Micro-
pipette shown before introduction {upper left) and in capillary [after introduction {lower left).
[From Landis (198).]
fig. 2.2. Curve showing aver-
age gradient of pressure through
the mesenteric blood vessels of
the frog. [From Landis (198).]
44
^Systolic Pressure
X)
' ^"""— -Ol
36
n.
w
\
s 28
\
a.
n 20
v>
UJ
K
D.
0
^Diastolic Pressure ^. 9
0
1
3
2
12
^N>^_
4
-
* 400
Artery Artery - first
Vein
-
l:5 2oo
bifurcation Venous
I5 0
H Arteriole Capillary Capillary
1 1 1 1 1 1 1
■ '
'
8 10
LENGTH
12
MM
30
25
20
15
10
- 5
18 20
was also a significant drop in the capillaries, amount-
ing on the average, with ordinary blood flows, to 20
or 30 per cent of the total (198). A somewhat smaller
gradient was also calculated from Poiseuille's equa-
tion and motion picture analyses of flow through
capillary networks (206).
The same determinations provided direct support
for the Starling filtration-absorption hypothesis.
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
96;
table 2.1. Resting Average Capillary Blood Pressure (Pc)*
and Osmotic Pressure of Plasma Proteins (Jlpi)
npi
Animal
mm Eg
Frog
5-10
Rat
16-21
Guinea pig
17-21
Cat
19-26
Man
21-29
Man, avg
25
values
Average Pc
Arteriolar
end
mm Eg
10.6
I I .O
10.7
22. I
28.3
3'-3
Venous end
mm Eg
7-4
7.0
7-4
'2-5
26.8
32.O 2of 12.1
34-3 I !2-2
3O.6 29.5! 21.9
- 22 f
32
24t
•5
Tissue (and Reference
for Pc)
Mesentery (198)
Muscle (205)
Skin (205)
Mesentery (202)
Mesentery (202)
Intestine (178)
Skin (203)
Skin (92)
Skin (89)
Skin (225)
* Direct measurements only. f Summit of the capillary
loop.
In the arteriolar end of the frog's mesenteric capillary
network pressure averaged 14.5 cm H20 or 10.6
mm Hg; in the venous end about 10 cm H20 or
7 mm Hg. Since the osmotic pressure of the plasma
proteins ranged in normal frogs from 5 to 10 mm Hg
(41 ), the approximate balance predicated by Starling
was present except when starvation, as in winter
frogs, reduced the concentration of plasma proteins
(23, 41). As shown in table 2.1, capillary blood
pressures in frog's muscle and skin were similar to
those in the mesentery (205). In the mesenteries of
rats and guinea pigs a balance was also found but at
a higher level of pressure (202). Pressures were highest
in the intestinal capillaries of the cat (178) and in
the cutaneous capillaries of man (89, 92, 203, 225),
but again in balance with the higher osmotic pressure
of the plasma proteins as shown in figure 2.3. Thus
in four tissues and in five species the pressures found
were generally compatible with Starling's view that,
on the average and at resting blood flows, these
pressures favor filtration in the arteriolar portion
of the capillary network and a balancing absorption
in the venous end of the capillary network. But
generalizations cannot be extended to tissues with
specialized functions. Capillary blood pressures may
be higher in kidney and lower in lung.
Hayman (145) found that glomerular capillary
pressure in the frog averaged 54 per cent of the
simultaneously measured aortic blood pressure. White
(377) observed pressures of similar magnitude in
Necturus. For mammalian glomeruli direct measure-
ments are lacking, but indirect estimates have ranged
from two-thirds of arterial pressure by Winton (384)
to about 50 per cent of arterial pressure by Gottschalk
& Mylle (124). The high rate of glomerular filtration
can be explained by these high capillary pressures
and the greater effective pore area of the glomerular
membranes (278). The mechanism by which 98
per cent or more of this filtrate passes back into the
blood of the peritubular capillaries cannot be ex-
plained so simply.
Postglomerular or peritubular capillary pressures
have been measured directly by Wirz (385) who
reported 17.4 ± 2.6 mm Hg for a small series of
rats and by Gottschalk & Mylle (124) who found
averages of 20.4 and 14.2 mm Hg for large and
small peritubular capillaries, respectively, under
normal conditions. These pressures increased, how-
ever, to very high levels not only during venous
100
MAN -Finger tip, heart level
100
ARTERIES ] CAPILLARIES
ARTERIOLES
fig. 2.3. Curves comparing gradient of pressure drop
(open circles) in four species with the corresponding osmotic
pressures (filled circles) of their plasma proteins. [Modified from
Landis (207).]
966
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
congestion, but also after dextrose infusions, e.g.,
to 37 mm Hg, and during ureteral occlusion, e.g.,
to 40 mm Hg, with relatively close parallelism be-
tween intratubular, interstitial, and peritubular
capillary pressures. Evidently, the hydrostatic pressure
difference across the walls of renal peritubular
capillaries is far less than across the walls of peripheral
capillaries generally. This implies that the full
osmotic force of the plasma proteins, unopposed by
hydrostatic pressure differences, may be available
for withdrawal of tubular reabsorbate from renal
interstitial fluid to blood.
Pulmonary capillary pressure presents an exception
in the opposite direction. Though direct measurements
arc not available as yet, an indirect "wedging"
method (146, 147) has made it clear that in the lung
capillary pressure is normally between 5 and 1 5
mm Hg in dog and man and is, therefore, well below
the osmotic pressure of the plasma proteins. Ab-
sorption is favored (55) and ensures a minimum of
interstitial fluid in the alveolar walls, which is an
important consideration in a tissue the prime function
of which is to permit rapid exchange of gases. In
normal subjects these "wedge pressures" are quite
constant; exercise produces elevations of not more
than 3 or 4 mm Hg. Greater elevations than this
during exercise have been found to be helpful in
detecting early left ventricular failure or slight mitral
stenosis not yet severe enough to produce clinical
symptoms or signs (291).
Retinal capillary pressure has not been measured
directly, but must be considerably higher than that
in muscle or skin in order to maintain blood flow
despite an intraocular pressure of about 20 mm Hg.
Nor are any reliable figures available for capillary
pressures in other special regions, e.g., brain, pleural
and peritoneal surfaces, joints, etc. In view of the
differences between capillary pressures in skin,
kidney, and lung, generalizations are obviously
unjustified and direct measurements are needed tor
each tissue.
C. Variability of Capillary Blood Pressures
I 'nder Control Conditions
The average figures so far given would, by them-
selves, present an erroneous idea of the potential
role of capillary pressure in the filtration and ab-
sorption of fluid. In any one tissue capillary pressure,
like the more easily observed capillary blood flow,
varies from moment to moment and from capillary
to capillar) even when they arise from the same
arteriole. This is to be expected from the responsive-
ness of the terminal arterioles and arteriocapillary
sphincters to nerve impulses, both constrictor and
dilator, to local metabolic products and also to
mild injury such as that produced by manipula-
tion, exposure to air, and cannulation itself (198,
203, 205). In the skin of frog (205) and man (203)
the mere introduction of a minute pipette sometimes
produces a brief rise of capillary pressure accom-
panying the transient vasodilatation of a "'triple
response" to injury.
It must also be emphasized that capillary pressure
has been measured directly in relatively few tissues.
In man, determinations have been limited to the
capillary loops in the nailfold where arteriovenous
anastomoses are also present and may influence
pressure measurements. As shown in table 2.1,
in one series pressure in the arteriolar loops averaged
32 mm Hg. However, the single readings ranged from
21 to 48 mm Hg; in the venous loops the correspond-
ing figures were 12 and from 6 to 18 (203). In an-
other series of control measurements Eichna &
Bordley (89) found even larger variations, much
more overlapping of values, and a smaller average
gradient, viz. 31 to 22 mm Hg rather than 32 to 12
(table 2.1). These differences may possibly be re-
lated to room temperature because the larger gradient
and lower pressures in the venous limbs of the capil-
laries were found at room temperatures of 18 to qo C
(203). The smaller gradient and higher venous capil-
lary pressures were observed in a warm room where
temperatures were 23 to 28 C (89). It seems likely
that capillary pressures in human digital skin, par-
ticularly in the venous limbs next to the subpapillary
venous plexus, can be influenced by the state of the
arteriovenous anastomoses. At higher room tempera-
tures opening of these large channels, and increased
blood flow direct from larger arterioles to larger
venules, may well increase pressure locally in the
subpapillary plexus into which the true capillaries
also discharge their blood.
Chambers & Zweifach (37) have suggested, in
addition, a division of function in the minute vessels,
viz. that higher pressures, and hence filtration, may
occur chiefly in the direct, arteriovenous channels,
with lower pressures and absorption located in the
true capillaries. For such specialization, however,
no supporting evidence in the form of pressure
measurements in direct channels is available. More-
over, very high pressures and filtration rates were
frequently found in true capillaries (200). Zweifach
(387) also suggested that "the arrangement whereby
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
967
the true capillaries come off at right angles to the
A-V vessels favors the development of "suction forces,'
especially where a rapid, continuous flow courses
only through the A-V channels." No suction forces
were encountered in any of the many direct measure-
ments of capillary pressure. The reason for this
becomes clear in considering actual rather than
apparent velocities of flow in the minute vessels.
Under the microscope, which magnifies linear velocity
as well as size, flows that seem very rapid indeed are
really between 1 and 2 mm per sec. Calculation of the
magnitude of the corresponding velocity effect by
the Bernoulli equation shows the insignificance of
any possible suction force, viz. for a linear velocity of
2 mm per sec a pressure difference of only .000015
mm Hg.
Far more important is the conspicuous variability
of pressures found throughout the entire minute
vessel system as shown in figure 2.4 for the exposed
frog's mesentery where measurements could include
larger vessels as well as capillaries. In general, high
capillary pressures were associated with very rapid
flows (if venous outflow was normally free) and with
increased pulse pressure in the capillary network.
Lower pressures, approaching venous pressure,
were associated with slower flow or absence of flow
(ig8). Hence, as arteriolar diameter or tone of
precapillary sphincters changes from moment to
moment, even under resting or control conditions,
the average balance between highly variable capillary
pressures and the much less variable osmotic pressure
of the plasma proteins often includes temporary
imbalances in single capillaries and corresponding
shifts toward periods of filtration or absorption.
McMaster (235) has suggested that such shifts ex-
plain, in part, the intermittent entry of Locke's
solution at atmospheric pressure into the skin through
a fine needle introduced carefully to avoid both
blood vessels and lymphatics.
Position of a capillary bed, relative to the heart,
affects capillary blood pressure in general accordance
with changes of hydrostatic pressure (203). In the
finger tip of man at heart level average pressure in
the arteriolar portion of the capillary loop was 32
mm Hg and in the venous portion 1 2 mm Hg, with
large individual variations in single capillaries around
these averages. When the hand was 30 era above
heart level these average pressures became 23 and
10 mm Hg, respecti\ely, further drop being ar-
rested presumably because of collapse of the thin-
walled veins in the arm. Conversely, lowering the
forearm to 40 cm below heart level increased average
arteriolar and venous capillary pressure to 45 and
33 mm Hg, respectively.
The relation between capillary blood pressure and
the osmotic pressure of the plasma proteins is there-
fore extremely labile, both as to time and the area
of capillary wall involved. Absorption may be favored
in a large segment of the capillary bed for consider-
able periods, e.g., during vasoconstriction or eleva-
tion of an extremity and filtration favored for other
periods, e.g., during vasodilatation or dependency.
Nevertheless, a net equilibrium is maintained and
favors constancy of plasma volume and interstitial
fluid volume. Under exceptional conditions, e.g.,
muscular activity, prolonged dependency of an
extremity, high temperature, injury, and inflamma-
tion, excessive capillary filtrate must be returned to
the blood stream by the lymphatic vessels. These
ancillary vessels, as described in the following section,
n _□
40
u 24
-Arterioles
, 1 rr
-Capillaries
arteriolar
venous
Artery Artery -first
bifurcation
L E NGTH - MM
J I I I L.
35
30
25
20
15
10
fig. 2.4. Chart showing vari-
ability of capillary blood pressure
and of pressure gradient in the
blood vessels of the frog's mes-
entery. The higher capillary
pressures and increased capillary
pulse pressure are characteristic
of vasodilatation. The lower
capillary pressures and absence
of measurable pulse pressure
are characteristic of vasocon-
striction. [From Landis (198).]
10
14
16
18
20
968
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
provide an important safeguard against abnormal
accumulations of capillary nitrate in the interstitial
fluid compartment.
D. Functional Changes of Capillary Blood Pressure
Hemorrhage and local application of epinephrine
produced vasoconstriction in the frog's mesentery
and reduced capillary blood pressure as shown in
the lowermost curves of figure 2.4 (198). In man the
marked vasoconstriction and cessation of blood flow
found in Raynaud's disease reduced capillary blood
pressure in the affected digits to between 5 and 8
mm Hg, i.e., to levels approaching local venous
pressure (204). During the hyperemia of recovery,
pressures in these same capillary loops rose rapidly
to between 32 and 45 mm Hg. In normal subjects,
however, local cooling and vasoconstriction reduced
capillary blood pressure only moderately and the
rise of pressure during the secondary hyperemia of
cold was likewise moderate (203). Also in man Eichna
& Wilkins (90) found that neurogenically induced
vasoconstriction reduced cutaneous capillary pressure
by 1 to 8 mm Hg in 52 of 89 observations with no
change or slight elevations of 1 or 2 mm Hg in the
remainder. Intravenous injection of 1 or 2 jug of
epinephrine reduced capillary pressure by 1.5 to 22
mm Hg in seven of ten experiments but in three
subjects elevations of 1 or 2 mm Hg were observed.
Svmpathectomy obliterated neurogenic effects, but
not those of epinephrine. In the vasoconstriction of
human hypertension capillary pressure was not
significantly elevated and minor increases found in
some subjects were independent of arterial pressure
(89); this was also true of the temporary rise of
arterial pressure produced by Paredrinol intra-
venously (87) with or without prior sympathectomy.
Conversely, vasodilatation increased capillary
blood pressure, frequently to very high levels ap-
proaching arteriolar pressure (see fig. 2.4). Capillary
pressure also rose during local vasodilatation induced
in the frog by dilute urethan (198), by injuries which
produced hyperemia and capillary stasis (199),
by a simple triple response and after muscular con-
traction (205). In human skin the hyperemias of
local heating, intradermal histamine, inflammation,
and reactive hyperemia after cold (203) were ac-
companied by elevations of capillary pressure to
maxima between 49 and 60 mm Hg. In these ob-
servations room temperatures were low, 18 to 20 C.
At higher temperatures, 23 to 28 C, Eichna &
Bordley (89) found that intradermal histamine
elevated capillary pressures much less conspicuously
and more in the venous than in the arteriolar limbs
in both normal and hypertensive subjects. It was
emphasized that arteriovenous anastomoses may
have been involved in these effects (203).
From the higher capillary pressures found in
localized vasodilatation and hyperemia it might be
thought that excessive filtration and increased lymph
flow must occur with any vasodilatation. This is not
always the case, however. The most notable excep-
tion is the repeated finding that denervation of an
extremity produces hyperemia and evidence of
increased blood flow without change of lymph flow,
or at most a very slight increase, as reviewed by
Drinker & Field (76). This failure of widespread
vasodilatation to increase the flow of lymph was, in
fact, for many years cited as evidence against the
Starling hypothesis. Eichna & Bordley (89) found that
reactive hyperemia and also indirect or reflex vaso-
dilatation in man, produced by body warming, did
not increase cutaneous capillary pressure significantly.
The reason for this may lie in the lowering of pres-
sures in the digital arteries by 1 o to 40 mm Hg during
the generalized vasodilatation produced by body-
warming (73, 114, 115, 244), by exercising the
forearm muscles (73), or by reactive hyperemia
(365). The named arteries to an extremity are ap-
parently large enough to conduct blood at resting
flow rates with little pressure drop. The lesser in-
crements of flow required by localized vasodilatation
are associated with little drop in arterial pressure
and capillary pressure rises conspicuously. However,
when vasodilatation involves the resistance vessels
of a whole extremity, and blood flow through the
large arteries is increased severalfold, the pressure
drop from brachial artery to digital artery becomes
significant. Then, arterial pressure head being much
reduced locally, the rise of capillary pressure is
limited even with maximal arteriolar dilatation.
In addition, the capillaries lie between two resistances
and it is quite possible that arteriolar dilatation
will not raise capillary pressure if the venules and
veins are simultaneously dilated in similar or greater
proportion.
E. Effects of J'erwus Pressures and of Venular
Constriction on Capillary Pressure
Elevations of venous pressure produce, as might
be expected, a rapid increase of capillary pressure
to levels above the pressure in the veins. Direct
measurements have shown this to be true of localized
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
969
obstruction of a venule in the frog's mesentery (198)
and also when the human extremity was congested
by inflating a cuff previously placed on the upper
arm (88, 203). When blood flow was normal, capil-
lary pressure equalled cuff (and venous) pressure
within 15 to 45 sec and eventually exceeded cuff
(and venous) pressure by 8 to 1 4 mm Hg in the
2nd to 4th minute of congestion (203). When blood
flow was very slow, however, as in the arteriolar
constriction of acrocyanosis (220), capillary pressure
rose much more slowly, requiring up to 8 min to
equal venous pressure and finally exceeding that
pressure by only 1 or 2 mm Hg. Other direct measure-
ments have shown that capillary pressure is elevated
in congestive heart failure (92) and also in glomerulo-
nephritis (225) whenever venous pressure is high,
returning to normal as venous pressure declines.
In all these measurements the variability of normal,
resting capillary pressures prevents making any
meaningful comparison of the increment of capillary
pressure which corresponds to any given increment
of venous pressure.
The effect of venous pressure on capillary pressure
was emphasized in 1894 on the basis of indirect
evidence by Bayliss & Starling (12) when they ob-
served that elevating venous pressure increased
lymph flow more than similar changes of arterial
pressure. In the absence of a direct method of measur-
ing capillary pressure they suggested using changes
of venous pressure to deduce changes of capillary
blood pressure. Also, from studies of lymph formation,
Drinker & Field (76) in 1933 suggested that, other
things remaining constant, the state of the veins might
modify capillary pressure and thereby influence
filtration of fluid through the capillary wall. In 1948
Pappenheimer & Soto-Rivera (282) found in the
denervated, perfused extremities of cats and dogs
that a given change of venous pressure influenced
filtration and absorption five to ten times more than
did a similar change of arterial pressure. They
formulated the dependence of mean capillary pressure
on arterial and venous pressures and resistances as
follows:
pc
r P + P
ra 'a hv
'* t
(2.1)
in which r,. and ra are, respectively, the precapillary
and postcapillary resistances, while pA and pv are,
respectively, arterial and venous pressures. From this
equation it follows that at given values of arterial
and venous pressures the mean capillary pressure
depends on the ratio of the postcapillary to pre-
capillary resistance to blood flow. Contractility of
the large veins has been well established for a long
time, but even in 1950 a general review (210) re-
vealed little information concerning reactions of small
veins or venules and only a few instances of inde-
pendence of such reactions from those of the arterioles.
Beginning in 1954 Haddy et al. (137) approached
the question of differential changes in precapillary
and postcapillary resistances by threading catheters,
outside diameter 0.2 to 0.5 mm, as far as possible into
"small veins" and "small arteries'' for measurement
of pressures. Under control conditions small artery
pressures averaged 65 ± 25 mm Hg, while small
vein pressures, under local anesthesia, averaged 13
mm Hg with a range of 8 to 25 mm. Small vein
pressure varied independently of the relatively con-
stant large vein pressure, indicating that the small
vein system must be responding independently to
nervous or humoral stimuli. Kelly & Visscher (171)
found that independent pressure changes in small
arteries and small veins were produced by stimulating
the lumbar sympathetic chain in dogs. Variability
of these changes in timing, magnitude, and even
direction was considerable and three main types or
combinations of pressure changes had to be described.
In further studies small vein pressure increased to
as much as 36 mm Hg and led to the suggestion by
Lee & Visscher (2 1 4) that edema of the skin could
have a neural origin. However, were this an impor-
tant possibility one would expect that cutaneous
edema would be observed at some stage in the pro-
gressive, neural vasoconstriction found in hemorrhage
and shock. This is, however, not the case. It must
be remembered, too, that if arterial pressure remains
constant, or especially if it falls, any constriction,
whether arteriolar or venous, tends to reduce blood
flow and this then tends to limit edema formation
to the extent that renewed volumes of blood plasma
are not available for filtration; at zero blood flow-
even the wheal of histamine does not appear (216,
217, 219).
Extending this method to humoral agents, Haddy
and others found that independent, and sometimes
opposite, reactions of arteries, small arteries, small
veins, and large veins were produced by change of
tissue temperature (135, 364), change of pH (105),
epinephrine (134), norepinephrine (134, 364),
serotonin (134, 136) and histamine (133). As sum-
marized by Haddy et al. (134), "almost every possible
combination of active and passive change in seg-
97o
II Wlll',1 H IK HI I'HYMM OCY
CIRCULATION II
120
100
bO
40
20
Histamine IA
-2-1—6-1 14—1 43-
Brachial Artery
Small Artery
r
V
V.
\
\
■*\-
SmallVein <**
/-
Subcutaneous Tmsue
25 50 75 100 125
Time in Minutes
ISO 17 5 20 0
fig. 2.5. Effect of histamine infused intra-arterially upon
vascular and interstitial pressures in the dog's foreleg. Num-
bers at top refer to /ig/min histamine base administered into
the brachial artery. [From Haddy (133).]
mental resistances and pressures has been observed
during one or another arrangement." The true signifi-
cance of these findings is correspondingly difficult to
evaluate. In the case of histamine (fig. 2.5), which
has been more thoroughly studied at controlled flow-
rates, this elevation was ascribed for small doses
simply to arteriolar dilatation, and for large doses
to an added selective constriction of small veins. This
constriction in turn was ascribed in part to the direct
action of histamine and in part to indirect effects
stemming from release of norepinephrine from the
adrenal medulla. It was suggested also that the result-
ing changes of capillary pressure might be sufficient
to explain the protein-rich edema, produced by
histamine, on hydrostatic grounds by passive con-
gestion, increased capillary pressure and stretching
of the capillary wall, without invoking injury of the
wall by histamine. The production of a protein-poor
filtrate is certainly possible, but the production of a
protein-rich filtrate seems unlikely. The small vein
pressures reported were all below 40 mm Hg, whereas
it has been shown, with venous pressures of 40 to 60
mm Hg, that capillary filtrate contained at most
0.7 g per cent of protein and averaged only 0.3 g
per cent (211), not the 4 or 5 g per cent found in
the histamine wheal (217).
The validity of conclusions based on such catheteri-
zation of small veins is doubtful for several reasons.
In addition to inescapable, even though slight, ob-
struction to venous outflow and false elevations of
"small vein pressure" there is the possibility of effects
from trauma to the intima of the venules under
study. Davis & Hamilton (65-67) stimulated the
sympathetic nerves to the rabbit's ear and the dog's
paw and found that the pressures developed in the
small veins depended upon the nerve stimulated,
upon the frequency of stimulation, upon the rate of
blood flow, and upon the presence or absence of
mechanical obstruction to venous outflow. They
found also that the highest small vein pressures oc-
curred while flow in the region had stopped. Pressures
in the small veins sometimes exceeded those in the
small arteries (fig. 2.6, right). They concluded that
when this occurred the walls of the small veins were
constricting against a static column of blood isolated
probably from the capillaries, and certainly from the
arterioles. Burch (29) observed similar elevations of
pressure in isolated segments of large veins in man.
More recently still, an isovolumetric technique has
been used by Mellander (243) to measure the effects
of sympathetic stimulation on the resistance and
capacitance vessels in cats with hind legs placed in a
plethysmograph. As shown in figure 2.7, frequency of
stimulation was kept within physiological limits,
i.e., from 0.25 to 16 stimuli per sec. Both the capaci-
tance vessels and resistance vessels constricted. The
former responded more actively at first and reached
maximum constriction at 8 stimuli per sec. The re-
sistance vessels were influenced less at low stimulation
rates and more at higher rates. Precapillary resistance
increased more than postcapillary resistance and
increasing absorption was found, with calculated
reductions of capillary blood pressure ranging from
2 to 15 mm Hg. Mellander suggested that Kelly
and Yisscher, by manipulating and cannulating the
small veins, may have produced local constriction of
their walls. In addition to the obstruction already
mentioned, it is also possible that intimal irritation,
secondary to catheterization or cannulation, may
make the small veins abnormally susceptible to
vasoconstrictor impulses. In any event it seems clear
that, under some conditions, stimulation of sym-
pathetic vasoconstrictor nerves increases arteriolar
resistance more than venous resistance, reduces
capillary blood pressure, and leads to rapid and
significant absorption of fluid and not to elevated
capillary pressure and filtration.
By the same technique Mellander showed that
epinephrine in small doses, and in muscle, relaxed
the arterioles and probably constricted the venules
slightly, producing filtration and hence indirect
evidence of a rise of capillary blood pressure. In
skin, all doses, and in muscle large doses of epinephrine
produced effects like those of sympathetic stimula-
tion, but only 20 to 25 per cent as great. Norepi-
120.
'70
A.
'm
%0
68
& '4!>
0.
.L.»n>« ■
4
4.5
5
V.
5
2.5
10/IOsec
5
E
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
A.
971
16
'140
fig. 2.6. Left: small vessel pressure and flow response to a low-frequency stimulation of the
ipsilateral lumbar sympathetic trunk (15 v, 1 stimulation/to sec for 1 min) Aortic pressure (A),
small artery pressure (a), small vein pressure (»), and small vein flow (/•') against atmospheric
pressure. Numerals on pressure tracings indicate pressure in mm Hg. Numerals on flow tracing
represent flow in ml/min. Timer set at 10 sec. Rig/it: same but with high frequency stimulation
(15 v, 10 stimulations/sec for 1 min). Symbols same except that F is small artery flow proximal to
distal segment. [From Davis & Hamilton (66).]
fig. 2.7. Effects on resistance and capacitance vessels and net transcapillary fluid shift produced
by maximal lumbar vasoconstrictor fiber stimulation at different frequencies. Changes in blood
flow reflect effects on resistance vessels (inflow and outflow pressures kept constant). The initial
and rapid decreases in volume reflect effects on capacitance vessels and the subsequent slower and
continuous decreases in volume (slopes indicated by dashed lines), transcapillary influx of extra-
vascular fluid. Reductions in mean hydrostatic capillary pressure calculated in approximate figures.
[From Mellander (243).] 3
nephrine was also constrictor and produced absorption
of fluid. Acetylcholine increased blood flow markedly
but produced less filtration than small doses of epi-
nephrine. Presumably capillary pressure increased
very little because pre- and postcapillary resistances
were reduced equally. Johnson and Hanson ( 1 68a)
have recently applied the isogravimetric technique
to a study of pre- and postcapillary resistance in the
intestine of the dog. In this preparation, the isogravi-
metric capillary pressure is only about 65 per cent of
the plasma protein osmotic pressure, probably reflect-
ing the higher permeability to protein of intestinal
capillaries. The postcapillary resistance to blood flow
through the intestine was increased markedly when
972
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
arterial perfusion pressure was decreased and evidence
was presented that this reaction depends upon sympa-
thetic innervation of the postcapillary blood vessels
(142a).
One more series of studies must be mentioned
briefly because the possible role of the arteriovenous
anastomoses on small vein pressures has not been
mentioned so far. These relatively large vessels run
parallel to the capillaries and are numerous in the
skin, particularly of the digits. Schroeder (324),
using a pressure plethysmograph similar to that used
in man by McLennan et al. (234), studied the effects
of acetylcholine, epinephrine, histamine (325),
hypoxia (327), calcium, and rutin (326) on vascular
volumes and pressures in the dog's foreleg. External
pressure was set arbitrarily at 35 mm Hg to measure
"changes of capillary pressure" and at 15 mm Hg
to measure "changes of venous pressure." No abso-
lute pressure readings could be obtained by this
method. The curves were variable and often difficult
to interpret. However, Schroeder placed considerable
emphasis upon independent reactions of the arterio-
venous anastomoses and their secondary effects upon
venous and capillary pressures. It is conceivable that
some of the variability in the observed small vein
pressures in the skin of the extremities may be reduced,
or at least explained in part, if body temperature and
environmental temperatures are adjusted to maintain
the arteriovenous anastomoses in as constant a state
as possible. In any case it is clear that pressures and
resistances in large veins and in small veins, together
with any factors which modify them, must be taken
into account when describing the mechanisms which
determine changes of capillary blood pressure.
3. OSMOTIC PRESSURE OF THE PLASMA PROTEINS, Upl
A . Methods of .Measurement
Starling's conception of a balance between capillary-
hydrostatic pressure and protein osmotic pressure
was supported by actual measurement of the pressure
required to maintain fluid balance across a semi-
permeable membrane separating blood serum from
serum ultrafiltrate. Starling's osmometer consisted
of a small glass bell, provided at the top with two side
arms. A piece of peritoneal membrane, soaked in
10 per cent gelatin, was tied over the mouth of the
bell and prevented from bulging by a perforated
silver plate. One sidearm was connected to a vertical
tube and the other side arm was used to introduce
serum into the bell. The lower end of the bell, in-
cluding the membrane, was then dipped into serum
ultrafiltrate or other protein-free salt solution. Within
a few hours osmotic flow of fluid from the salt solu-
tion through the membrane was made evident by a
rise of fluid in the vertical tube. Equilibrium was
established in 2 to 6 days; at this time the pressure on
the membrane, exerted by the fluid column in the
vertical tube, was considered equal and opposite to
the osmotic pressure of the serum proteins. In a typical
measurement Starling (345) found that serum con-
taining 7.56 per cent "proteids" caused fluid to rise
in the vertical tube to a height of 53 cm (~4i mm
Hg). This value is considerably higher than modern
estimates shown in figure 3.1, probably as a result of
bacterial degradation of protein during the long
period required to reach equilibrium. In later work
(1899) Starling (347) obtained values which were
generally lower than his first estimate and well
within the range expected for capillary hydrostatic
pressures.
Starling's measurements were of great interest to
colloid chemists as well as to physiologists. Accord-
ing to van't Hoff's analogy, in 1887, between ideal
solutions and gases (157), the osmotic pressure, II,
should be given by
IT 'CRT (30
where c is expressed in moles per liter.
fig. 3.1. Osmotic pressure-concentration curves for whole
plasma and selected plasma proteins. Based on data from
references (268, 270, 312, 313, 343), original measurements
corrected to 37 G. Experimental points for 7-globulin are
included to indicate magnitude of experimental error.
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
973
20,000
- 25,000
l-- 30,000
- 60,000
fig. 3.2. First derivative of protein osmotic pressure-con-
centration curve to show deviation from van't Hoff's law. At
infinite dilution the mean number average molecular weight of
plasma proteins is almost 100,000 but in normal plasma their
osmotic behavior corresponds to an ideal solute of mol wt
37,000.
If c is expressed in g per 100 ml, RT in liter-atmos-
pheres, and IT in atmospheres, then
Mol Wght. = 10c RT/II (3.2)
The potential application of equation 3.2 to the
determination of molecular weights led protein
chemists to investigate in detail the theory and
technique of osmotic measurements. As early as
1905, Reid (293) used Starling's technique to esti-
mate the molecular weight of hemoglobin. Subse-
quent studies by S0rensen (342), Adair (1, 2), and
others showed that protein osmotic pressure is de-
pendent upon ionic strength, net charge, and other
factors not included in van't Hoff's limiting law for
ideal solutions. Deviations from the limiting law
increase rapidly as a function of protein concentration
(figs. 3.1 and 3.2) and estimates of molecular weight
can only be made on the basis of extrapolation to
zero concentration. The chief technical difficulty con-
fronting early workers was the long period required
to reach equilibrium across artificial membranes.
In order to avoid bacterial degradation of protein
it was necessary to carry out measurements at low
temperature; days or even weeks were required for
each determination. Nevertheless, the first satis-
factory estimates of the molecular weights of serum
albumin (3), ovalbumin (342), and hemoglobin (2)
were obtained by this method.
Advances in the technique of osmometry have
reduced considerably the time required for the
equilibration process.
Equilibration across a semipermeable membrane, following
a step change in either hydrostatic or osmotic pressure, pro-
ceeds exponentially with a time constant equal to the product
of membrane resistance and volume distensibility
% Equilibrium = IOo[l-exp-f-j^ — — -. J (3.3)
ffi p fn'
where rm is membrane resistance to solvent flow and vpy vm
are the volume distensibilities of the pressure measuring device
and membrane, respectively. The resistance (rm) of membranes
capable of restraining the passage of serum albumin is seldom
less than io1 mm Hg per ml per hour per cm2 membrane.
The essential factor limiting the rate of approach to equilibrium
is therefore the volume of fluid which must pass through the
membrane in order to actuate the pressure detector and
satisfy the volume-pressure characteristics of the membrane.
For example, a typical osmometer with a membrane surface
area of 10 cm2 must have a total volume distensibility of less
than 3 X io~4 ml per mm Hg in order to achieve 95 per cent
equilibrium in 1 hour (equation 3.3).
In 1936 Hepp (151) described an osmometer in which
distensibility of the membrane (»,„) was made extremely small,
the chief volume displacement being confined to slight changes
in fluid level of the capillary tube manometer used to detect
pressure balance. Equilibration time was reduced to about 2
hours. Osmometers of the Hepp type have been widely used
by subsequent investigators and the osmotic pressure-concen-
tration curves shown in figure 3.1 are based on data obtained
with this instrument. A recent description of the construction
and use of Hepp osmometers has been published by Meschia
(248). Further reduction in volume displacement can be ob-
tained through the use of sensitive, recording pressure trans-
ducers having volume distensibilities less than io~6 ml per mm
Hg. With the aid of such transducers it is theoretically possible
to achieve 95 per cent equilibration across available protein-
impermeable membranes in less than 1 min. Recording os-
mometers of this type, having time constants of less than 5
min, have been in use in the authors' laboratory for several
years (277, 280I. Similar instruments, suitable for the rapid
estimation of protein osmotic pressure in o. 1 ml plasma, have
recently been described by Hansen (142).
B. Protein Osmotic Pressure of Human Plasma
Osmotic pressure-concentration curves for normal
human plasma, serum albumin, and two globulin
components of plasma are shown in figure 3.1.
The curves were obtained at physiological pH and
ionic strength, but the original measurements have
been corrected to 37 C. Experimental points, taken
from Oncley et al. (268), are shown for 7-globulin
in order to indicate the magnitude of experimental
error when a pure component is measured. The
smooth curves for albumin, whole plasma, and
ft-globulin are based on data in references (268,
270, 312, 313, 343)- Normal human plasma has a
974
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
protein osmotic pressure of 24 to 26 mm Hg, cor-
responding to a total protein concentration of about
7 per cent. It is impossible to give a significant mean
value because the protein concentration depends
upon procedures used for drawing blood samples
and in any given sample the value obtained depends
upon the method of measurement. Electrophoretic
measurements yield slightly lower values for total
protein than estimates based upon salt precipitation
or protein nitrogen. In the following discussion a
nominal value of 7.0 g per 100 ml will be considered
normal.
The osmotic pressure-concentration curves for
albumin and for normal plasma are described by the
following empirical equations which fit the experi-
mental data closely over the range o to 25 per cent
protein.
II albumin = 2.8 c + 0.18c* + 0.012c3 (3.4)
n plasma * 2.1c + 0.16 cz + 0.009c3 (3.5)
In each equation the first term represents the ideal
limiting law of van't Hoff. Thus the molecular weight
of albumin, calculated from the first term of equation
3.4, is 10 RT/2.8 = 69,000. The second and third
terms in each equation represent deviations from
van't HofPs law caused by Donnan effects and
protein-protein interaction.
The chief osmotically active protein in normal
mammalian plasma is albumin, which can be sepa-
rated and identified as a homogeneous component
representing about 50 per cent of the total protein in
plasma and contributing about 65 per cent of the
protein pressure. The globulins, on the other hand,
comprise a spectrum of components with molecular
weights ranging from 45,000 to 1,000,000 as shown in
table 3.1. The widely different osmotic activities
of jSj-globulin and 7-globulin shown in figure 3.1
serve to emphasize that no simple physicochemical
meaning can be attached to the osmotic pressures
developed by crude, heterogeneous globulin frac-
tions. Precipitation methods fail to separate albumin
from low molecular weight globulins which contribute
substantially to total protein osmotic pressure; for
this reason many early studies attempting to relate
total protein pressure to albumin: globulin ratios
(380) need to be reevaluated. Current estimates of
A : G ratio in normal plasma are close to 1 . 1 , in
comparison with values in the range 1.8 to 2.6 ob-
tained by classical fractionation procedures.
The osmotic pressure contributed by globulins
can be calculated from the difference between
albumin and whole plasma (equations 3.4, and 3.5),
it being assumed that the A:G ratio is 1.1 and that
osmotic interactions between globulins and albumin
are not significantly different from interaction be-
tween albumin and albumin (270, 314). The "aver-
age" globulin curve so calculated is given by
//globulins ' 1.6c + O.I5cz+ 0.006c3 (3. 6)
In normal plasma about 15 per cent of the total
protein pressure is contributed by known globulin
components and about 20 per cent by unidentified
components (table 3.1). Bennhold et al. (14) have
studied two extremely interesting cases of complete
analbuminemia; the osmotic pressure-concentration
curve of the albumin-free plasma from these unique
patients (brother and sister) conforms closely to
equation 3.6 (271). These patients have been in good
health for many years despite the fact that the protein
osmotic pressure of their plasma is less than 50 per
cent of normal. Presumably they have compensated
by permanent reduction of mean capillary pressure
to balance the low protein pressure.
C. Species Differences, Fetal Plasma
Comparative studies of colloid osmotic pressure
have been reviewed by Meyer (251) and by Keys &
Hill (175). A summary of data pertaining to plasma
of Elasmobranchs, Pisces, Amphibia, Reptilia,
table 3.1. Some Protei
n Components
of Human Plasma*
Component
Cone.
g/100
ml
% of
Total
Protein
mol wt
Approx. k
in Plasma
mm Hg
Whole plasma
7-0
100
25
Albumin
3-6
51
69,000
16.4
-y-Globulins
•7
1 1
156,000
0.9
Fibrinogen
•3
4
340,000
0.2
a-Lipoprotein
.28
4
160-400,000
0.2
(1.0 < p < 1. 14)
^-Lipoprotein
• 25
3-8
2 X IO6
(p = 1 .03 ± .02)
/3i-Metal combining
.2
3
90,000
0.7
/32-Globulins
.2
3
(150,000)
O.4?
ft-Lipid poor euglobu-
•'3
2
(150,000)
0.2?
lin
ai-Acid glycopro-
•03
0.4
45,000
0.2?
tein
Remaining known
•4
5
0 for 80 min. Interstitial
fluid pressure (O O) was calculated by dividing the cumulative decrease of observed filtration
rates by the previously determined average normal filtration coefficient (.0033 ml/ 100 ml forearm
tissue/min/cm H2O increase of venous pressure), and then correcting each value for the local in-
crease of plasma protein concentration and of n,,( produced by net filtration. As shown in A calcu-
lated interstitial fluid pressure increases steadily with time, reaching a maximum of almost 30 cm
H20 by 75 min. In B the same calculated interstitial fluid pressures are charted against the cumula-
tive volume of added interstitial fluid. As the interstitial compartment is distended by an increasing
volume of filtered fluid, interstitial pressure in the forearm tissues probably rises slowly at first and
then more rapidly. [Recalculated from data of Landis & Gibbon (209).]
of forearm tissue. As shown in figure 4.1 {left) when
venous pressure was raised to 60 cm water, the initial
filtration rate was .156 ml per min per 100 ml of
forearm tissue but declined rapidly to less than .040
after 75 min of congestion. At this time the volume of
newly filtered fluid was approximately 6 ml per 100
ml of tissue, i.e., about 60 per cent of the amount
which produces manifest edema, detectable by
"pitting on pressure." Knowing the decrease of filtra-
tion rate, the normal filtration coefficient (209), and,
approximately, the increase in the osmotic pressure
of the plasma proteins of the blood in the congested
forearm (188, 211), it is possible to calculate inter-
stitial fluid pressure, with results shown in figure 4. 1 .
Accumulations of interstitial fluid from prior filtra-
tion also increased the rate at which extravascular
fluid was removed from the forearm (188, 209), as
would be expected with higher interstitial fluid pres-
sures. It is still impossible to decide to what extent this
fluid was removed via the blood capillaries by absorp-
tion or via lymphatics by flow, though indirect evi-
dence (188) indicated that small accumulations were
probably removed by the former, larger accumula-
tions by the latter in addition. The importance of
interstitial fluid pressure seemed clear, although de-
pendable direct measurements were not available as
yet.
In a review of this topic in 1934 (207) it was neces-
sary to consider the conflicting views then current
concerning bound and free water in the interstitial
fluid compartment. It is now generally agreed on the
basis of many studies by several dilution methods
that the volume of truly "bound water" is negligible.
Yet in normal tissues interstitial fluid cannot be
identified microscopically as a distinct and continuous
compartment or layer around capillaries or between
cells except in a few locations. This is not surprising
because a simple calculation shows that if the normal
volume of interstitial fluid, approximately 1 5 per
cent of gross tissue volume, is distributed uniformly
between surfaces of cells, connective tissue fibrils,
blood capillaries, etc., the average thickness of this
layer cannot be greater than 1 n and is probably less
than 0.5 fi. This coincides with the findings of Mc-
Master & Parsons (240, 241) who injected dye solu-
tions into small lymphatic vessels and observed under
high magnification that the dye penetrated into the
tissues in the form of hair-like projections or "bristles,"
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
979
apparently between or along connective tissue fibrils
(fig. 4.2). Hemorrhage and dehydration, which they
used to diminish the volume of interstitial fluid,
delayed the appearance of these bristles of dye, but
eventually they became clearer than usual, as well as
more resistant to displacement by massage. On the
other hand, hydremic plethora, and particularly the
edema of inflammation, tended to obliterate the
bristles and permitted instead a diffuse and more
rapid distribution of dye which could easily be dis-
placed by pressure from one area into another, pre-
sumably because it dissolved in a larger and more
freely movable volume of edema fluid.
The paucity of normal interstitial fluid, its layered
distribution, and the disruptive effects of injecting
even small volumes of fluid (238) makes it necessary,
as with capillary blood pressure, to determine inter-
stitial fluid pressure by a "null point" method which
provides a balance of pressures with minimal move-
ment of fluid into, or out of, the interstitial fluid com-
partment. Wells et al. (375) used a capillary tube
placed between the manometer and the saline-filled
needle that was inserted into the tissue. By observing
the meniscus under a microscope, they saw that a
change of 2 or 3 mm water pressure sufficed to re-
verse the flow at the point of balance and hence, after
correcting for capillarity in the tube, they measured
interstitial fluid pressure with a small volume artifact.
Burch & Sodeman (30) and McMaster (238) re-
duced the volume change further, but still more re-
fined methods are needed to reduce the likelihood of
local hemorrhage and mechanical artifacts.
Table 4. 1 summarizes several representative series
of values given in mm Hg for easier comparison with
capillary blood pressure and the osmotic pressure of
the plasma proteins. In skin, McMaster (238) found
it necessary to determine "interstitial resistance" to
very slow rates of inflow of fluid because paucity of
freely movable fluid prevented determining a true
interstitial pressure. Although some of the values in
table 4. 1 may be artificially high, their order of
magnitude is consistent. P,t in skin and subcutaneous
tissues, under resting conditions, ranges from 1 to 5 or
6 mm Hg and averages about 2.5 mm Hg. In muscle,
P,f tends to be slightly higher, 1 to 9 mm Hg and
averages 4.5 mm Hg. In some comparisons Plf was
higher in the tightly sheathed muscles, e.g., soleus
and anterior tibial, than in the more loosely enclosed
/ ' yi-BM
•u^v/-r/',^ Mi _
I
fig. 4.2. Diagrammatic sketch of the extravascular interstitial movement of a 2% solution of
pontamine sky blue after its escape from the lymphatics, a: Dye first appears as colored bristles at
2-7 min. b: Color becomes more intense and bristles longer at 3-10 min. c: Colored lines become
broader at 5-12 min. d: Second phase. Diffuse blue staining between bristles which cannot be dis-
lodged by pressure. Bristles disappearing. During a to d color was apparently fixed on tissue ele-
ments and not dislodged by pressure, e; Diffuse blue cloud easily displaced with pressure, free fluid.
/.■ Dye escaping from ruptured lymphatics, no bristles. [From McMaster & Parsons (240).]
98o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
table 4 i . Interstitial Fluid Pressure, Ptf, Ranges at Rest
Pressure (Range)
Tissue, Species
mm Hg
Reference
Subcutaneous,
man
Forearm and leg
'-5-5-1
(375)
Foot
1. 1-3. 2
(30)
Pretibial
1.3-4.0
(3o)
Forearm
0.8-2.9
(30)
Hand
0 . 6-2 . 2
(30)
Foot
1.2-3.0
(23O
Leg
4.1-6.5
(23O
Eyelid
1.2-3.0
(28)
Cutaneous
Mouse*
0-4-3-7
(238)
Man*
1 .8-4.9
(238)
Muscle, man
Gastrocnem
us
0-7-3-7
(375)
Soleus
3-7-7-3
(375)
Gastrocnemius
2.6-8.9
(23O
Biceps
3.2-6.6
(231 )
Biceps
2.9-6.6
(127-129)
* "Interstitial resistance'
to very slow
flow, probab
0.5 mm above
Pi, (238).
gastrocnemius and biceps (375)- Very high pressures,
up to 85 mm Hg, were found briefly in some muscles
during powerful contractions (375), and values lower
than average were common during anesthesia, hemor-
rhage, surgical operations, and shock (127-129).
The effects of venous congestion, and consequent
filtration, on directly measured P ,; have been highly
variable, depending, as might be expected, upon the
distensibility of the tissue studied. McMaster (239)
found pressures up to 32 cm water or 23 mm Hg in the
skin of the mouse (fig. 4.3), and up to 23 cm water or
1 7 mm Hg in human skin, in good agreement with
the calculated interstitial fluid pressures found by
Landis & Gibbon (209) in the congested forearm
(fig. 4.1). Mayerson & Burch (231) found much
lower maximum pressures in subcutaneous tissue of
man during venous congestion or during quiet stand-
ing, e.g., 5.6 to 8.8 cm water, while intramuscular
pressure rose to maxima of 22 cm water in nonfainting
subjects and only 1 1 cm water in fainting subjects.
In isolated, perfused extremities Hyman (162)
and Pappenheimer & Soto-Rivera (282) found little
interference with prolonged filtration, and hence no
evidence of increased interstitial fluid pressure, until
manifest edema appeared. Hinshaw & Day (155),
however, made direct measurements of P,f in per-
fused extremities and found increases from control
values of 0.5 and 1.2 mm Hg to 8 and 15 mm Hg
when 1+ edema was present and to 10.5 and 24 mm
Hg when 2+ edema had appeared. This pressure
was enough to produce measurable collapse of blood
vessels. If this collapse involves the small veins it may
well distort fluid movement through changes in resist-
ance to flow, as well as through direct opposition to
capillary blood pressure itself.
Interstitial fluid pressures up to 25 mm Hg help
explain the slowness with which edema forms in
normal human beings, despite the high capillary pres-
fig. 4.3. Changes of interstitial fluid
pressure in the skin and lower leg of
mouse during and after venous con-
gestion of 40 mm Hg. Black dots
indicate pressure readings which yielded
neither inflow into the skin nor back-
flow into the apparatus; i.e., the pressure
of the extravascular fluid was accurately
balanced. Plus signs indicate pressure
readings at which fluid moved into the
tissues; i.e., pressure in the apparatus
was above interstitial pressure. Minus
signs show that backflow occurred into
the apparatus and that the plotted
pressure was lower than that of the
extravascular fluid. The interstitial
resistance during the control period is
shown by asterisks. [From McMaster
(239)-]
320
30.0
b
d 260
o
c 22 0
O 2
- 18.5
4)
§ 16.0
< ,Y
CIRCULATION II
at the end of absorption and about to enter the
lymphatic system. Pappenheimer & Soto-Rivera
(282) have pointed out that the diffusion coefficients
of the plasma proteins are such that in the absence of
flow or mechanical movement relatively large con-
centration gradients are possible in the interstitial
fluid compartment. "'Even if all filtration and absorp-
tion processes were stopped, some 20 minutes would be
required to reach 90 per cent equalization of protein
concentration over a distance of 50 microns" (282).
In perfused limbs of cats the average protein osmotic
pressure of interstitial fluid was 1.4 ± 0.4 mm Hg,
corresponding to an average protein concentration of
0.7 ± 0.2 g per 100 ml.
The average concentration of proteins in inter-
stitial fluid can also be estimated by a totally different
method. The dilution of labeled plasma albumins and
globulins after intravenous injection has shown that
the total mass of exchangeable plasma protein is
about twice the mass of plasma proteins in the blood
stream itself (103, 117). Sterling (352) found in man
that the average intravascular albumin averaged
117 g, the extracellular albumin, 147 g. Assuming
extravascular fluid volume to be the usual 15 per
cent of body weight, the average albumin concentra-
tion in extravascular fluid was calculated to be 1 .4 g
per 100 ml. By using Myant's figures (259) to estimate
globulin content in addition, the total average protein
concentration for extravascular fluid becomes ap-
proximately 2.1 g per 100 ml, which corresponds to
an average protein osmotic pressure, or LI,/, of 5
mm Hg. Similar calculations applied to the data of
Wasserman et al. (368, 372) yield slightly lower figures,
because in the dog the fraction of albumin and
globulin found normally in the interstitial fluid and
lymph appears to be rather less than that found by-
Sterling for albumin in man. Both estimates are
larger than those given by Pappenheimer and Soto-
Rivera (282) for the perfused leg of dogs as expected,
because the determinations made by the perfusion
method were restricted to the fluid in the immediate
vicinity of the capillaries and were limited to the
limb, both factors tending to give lower values. On
the other hand, calculations based upon exchangeable
protein mass include protein in the whole of the inter-
stitial fluid plus that in the lymphatics. In addition,
they include the extravascular fluids of the liver and
intestines where lymph is known to contain large
amounts of protein. Both factors tend to make the
figure for average LT,y greater for the whole body than
for the limb alone, but still not as high as that prob-
ably present in the liver and intestines.
With this qualification it can be concluded that
LT,, lies between 0.1 and 5.0 mm Hg, with the lower
value applying to capillary filtrate in the limbs and
the higher including the total interstitial fluid of liver
and intestines, as well as lymph. This can be com-
pared to P,, which ranges from 1 to g mm Hg, with
the lower values in subcutaneous tissues and skin, the
higher values in muscle. The formulation given in
equation 1 . 1 can now, with certain license for pur-
poses of summary, be provided with very approxi-
mate values in mm Hg for man at heart level and
under resting conditions, viz. :
FM. -
+ - filtration
k(P
c
32
pi
p„ *v
25 I to 9 0.1 to 5
- - absorption 15
With equal or greater license an average limb
capillary and lymphatic can be drawn, as in figure 5. 1 ,
to summarize the filtration-absorption process as it
may operate to produce a small volume of lymph
with relatively high protein content. Table 5. 1
provides a schematic summary of the changes that
occur in the fluids of the limb during several of the
more thoroughly studied functional states. Ranges of
determined values are given whenever possible.
Figures in parentheses are values that can reasonably
be inferred on the basis of available evidence. They
are given merely to show the probable direction of
presumed change and its order of magnitude. In
some instances even inferences are impossible, as
indicated by a question mark. The columns are
given letters to correspond with the schematic
capillary in figure 5.1.
Beginning at the top of the table with control con-
ditions and resting blood flow, the composition of
capillary filtrate has not been determined, but its
protein content may be inferred to be 0.2 to 0.4 g per
100 ml from the composition of capillary filtrate pro-
duced during mild venous congestion in man and
dog. The average protein content of interstitial fluid
ranges from 0.7 g per 100 ml in perfusion studies
(281) to 2.1 g per 100 ml by calculation from extra-
vascular protein mass. Lymph protein content range
from 1.3 to 3.3 g per 100 ml (386) and the volume
flow is small, requiring massage or passive movement
for collection of samples (76) as would be expected
with the absorption that occurs under resting condi-
tions.
Conversely, in venous congestion the protein con-
centration in capillary filtrate is known but the
average and highest concentrations in interstitial
ARTERIOLE
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
VENULE
985
30h
TTpl
PRESSURE
MM 20
HG
10
(a)
P, Capillary filtrate
(b)
Average
concentration
(c)
Highest
concentration
_"pl__
13 to 3.3
(d)
LYMPH
Protein concentrations given in g /I00 ml fluid
fig. 5.1. Schematic diagram of "an average limb capillary" to indicate approximate protein
concentrations in capillary nitrate, interstitial fluid, and lymph.
table 5.1. Protein Concentrations in Extravascular
Fluids of the Limb
Interstit
al Fluid
Capillary
Filtrate
Condition
Avg
Highest
Lymph
d
cone,
b
cone,
c
g/lOO ml
g/100 ml
g/100 ml
g/100 ml
Normal,
(0.2-O.4)
0.7-2.1
(1-3-3-3)
1-3-3-3
resting
Venous con-
O.2-O.4
(0.2-0.4)
(0.2-0.8)
0.2-0.8
gestion
Hypopro-
(0.OI-O.4)
0 . 04-0 . 5
(0.01-0.6)
0.01-0.6
teinemia
Burns
(3-5-5-o)
(3-5-5-°)
(3-5-5-o)
3-5-5-°
Muscular
?
■>
(0-5-I-5)
o-5-!-5
activity
Lymphedema
?
1-9-3-5
?
2-3-3-4
fluid have to be inferred from studies on lymph. This
inference is valid after newly formed capillary filtrate
has washed out of the interstitial compartment the
fluid which was present before congestion and while
the interstitial fluid compartment is being constantly
irrigated by newly formed capillary filtrate with no
absorption possible.
Hypoproteinemia in man, as mentioned above,
produces edema fluids with protein concentrations
ranging from 0.09 to 0.40 per cent. Weech et al. (374)
used chronic plasmapheresis to produce severe edema
of this type in dogs. Edema fluid contained between
0.04 and 0.4 g protein per 100 ml with all but a few
values below 0.25. Lymph protein in the same ani-
mals ranged from 0.01 to 0.6, with almost all values
below 0.3. In some instances the protein content of
edema fluid was slightly higher than that of lymph
collected simultaneously, indicating again the possi-
bility of imperfect mixing of the interstitial fluid com-
partment and, consequently, some sequestration of
edema fluid. Capillary filtrate, however, has not
been studied and so its protein content can only be
inferred. Lessened permeability to protein has been
suggested (374) but not proved so far. Sieving of
protein molecules may be involved (see section 10).
Massive injury in burns, produced by immersing
the extremities of anesthetized dogs in hot water
(48, 99, 100, 119, 120) increases lymph flow con-
spicuously and increases the protein in lymph to
between 3.5 and 5 g per 100 ml. In view of the known
effects of injury on capillary permeability to protein
(200) it is safe to infer that protein concentrations in
capillary filtrate and interstitial fluid are equally
high; particularly because lymph flow is rapid and
the interstitial compartment is well irrigated by
capillary filtrate.
For contracting muscle, information is still meager.
White et al. (379) found in dogs that while the flow of
lymph was much increased by exercise, its protein
content declined to between 0.5 and 1.5 per cent,
average 1 .0, and then remained constant as long as
exercise continued. The elevations of capillary blood
pressure and of interstitial fluid pressure during exer-
cise have already been described in sections 2D and 4.
Inferences concerning capillary filtrate and inter-
stitial fluid are unjustified because the lymph col-
986
HANDBOOK OF 1'HYSIOLOGY
CIRCULATION II
lected during exercise (379) contains more erythro-
cytes than control lymph does. This finding suggests
mechanical rupture of some capillaries, probably
when compressed between adjacent contracting
fibers. If this occurs, undetermined amounts of whole
plasma may accompany the erythrocytes and con-
tribute to the protein found in lymph. The possibility
of osmotic shifts of fluid produced by small molecules,
e.g., lactic acid, from contracting muscle has also
been proposed (207).
Finally, in lymphedema, the effect of obliterating
lymph flow by obstructive fibrosis of the larger
lymphatic vessels (77) is an accumulation of extra-
vascular fluid with abnormally high concentrations
of protein in both the edema fluid as well as in the
stagnant lymph. The protein content of capillary
filtrate is unknown and may be quite variable be-
cause of the tendency in lymph stasis toward inter-
mittent infection and consequent injury to capillaries
in severely lymphedematous extremities (77). It is
clear that more information is needed in all these
conditions.
C. Circulation of Interstitial Fluid;
Circulation of Protein
It has been customary in the past to say that capil-
laries '"leak" protein as if this were a useless defect of
the capillary wall. However, many lines of evidence
indicate that passage of plasma proteins through the
capillary wall is quite as important for cellular
metabolism and for defense against infection as the
retention of plasma protein is for normal fluid balance.
Whipple & Madden (376) showed that the circulating
plasma proteins within the blood vessels form a
"medium of exchange" which is an important part
of a larger nutritional pool. For example, dogs were
maintained in full nitrogen equilibrium by intra-
venous administration of dog plasma only. Drinker
(75) called attention to the benefits derived, during
infection, from the passage of globulins, including
antibodies, through the capillary wall into the inter-
stitial fluid around the cells and thence to the lym-
phatics. Still more recently several reviews have de-
scribed the binding of hormones (63, 304), fatty
acids (iog), and drugs (121) to plasma proteins. It is
significant, too, that the greatest passage of protein
through the capillary walls occurs in the liver, where
metabolic requirements are greatest and most varied,
and where albumin is synthesized.
Two paracapillary circulations (i.e., beside and
beyond the capillaries) can be identified. The first is a
filtration-absorption circulation which includes the
total capillary filtrate, the total interstitial fluid, and
finally that part of the interstitial fluid which passes
back into the capillary blood by the process of absorp-
tion. The second paracapillary circulation begins also
with capillary filtrate but then reduces to the unab-
sorbed fraction of interstitial fluid and its contained
protein, both of which, after bathing the tissue cells,
enter the finest lymphatic capillaries and are con-
ducted, via the major lymphatic trunks, back to
venous blood (see Chapter 30). Enough information
is available now to justify approximate calculations of
the magnitudes of these two circulations. Because both
depend upon the total volume of capillary filtrate
this figure can be considered first.
Continued blood flow through the resistance of the
capillaries requires, even at resting flow rates, a sig-
nificant pressure gradient in the capillary bed itself.
As indicated in table 2.1 and figure 2.3 this average
gradient lies above the osmotic pressure of the plasma
proteins in the first half of the capillary network. It
follows that, secondary to the basal pressure head
which is necessary for this resting blood flow, there is
necessarily a "basal filtration" of fluid under resting
conditions. Most of this filtrate is absorbed and the
low rates of lymph production in resting extremities
can give no indication of the rate at which the original
capillary filtrate is formed. A simple calculation sug-
gests, however, that in the resting animal capillary
filtrate is continuously produced at an average rate
which is at least five to ten times greater than average
resting lymph flow.
Landis & Gibbon (209) found in the human fore-
arm at 34 to 35 C that elevating venous pressure by
1 cm H20 increased filtrate by .0033 ml per 100 ml
forearm tissue per min. Assuming that 80 per cent of
a rise in venous pressure is transmitted to the capil-
laries, this becomes .0040 ml per 100 ml forearm
tissue per min for a 1 cm water increase of capillary
pressure. From capillary pressure measurements in
human skin, mean resting filtering pressure is (32 -
25 mm Hg)/2 or 3.5 mm Hg, or 4.8 cm H2O. Assum-
ing, for the purpose of obtaining a minimum figure,
that the unit increment of filtration given above
applies to the whole body, the total resting capillary
filtrate for a 75-kg human being is approximately 20
liters per 24 hours. To the extent that filtration
coefficients in liver and intestine may be greater than
in the forearm the volume of filtrate formed per 24
hours will be somewhat larger still.
For total lymph flow in man the most helpful data
are those of Crandall et al. (60) obtained from a
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
987
patient with a freely draining fistula of the thoracic
duct. Average basal lymph flow during fasting was
0.93 ml per min. After a heavy meal, flow from the
thoracic duct reached a peak volume of 3.9 ml per
min and remained above 1 .0 ml per min for several
hours. If allowance is made for the additional and
uncollected lymph from the right lymphatic duct by
adding an increment of one-fourth to one-third of
fasting thoracic duct flow, then total lymph flow in
fasting man at rest is approximately 2 liters per 24
hours. With allowance made for the effects of meals
and activity, it probably approaches 3 or 4 liters per
24 hours.
Figure 5.2 shows schematically the volumes of
these several '"circulations" in terms of exchanges in
24 hours. With a cardiac output of 6.0 liters per min
the first circulation, that of blood itself, amounts to
about 8000 liters per 24 hours. From this volume,
filtration in the capillary bed removes a minimum of
20 liters per 24 hours, a "filtration fraction" of 0.25
per cent. This capillary filtrate begins the second
circulation, that of interstitial fluid (fig. 5.2, F to IF
to A) with capillary absorption, during rest, of 80 to
90 per cent or 16 to 18 liters, of the original capillary
CARDIAC OUTPUT
8400 L per 24 HOURS
DIFFUSION EXCHANGE
80,000 L
20,000g glucose (400 utilized)
LYMPH
2 to 4 L
FILTRATION - ABSORPTION
LYMPH FLOW
FLUID 20 L+ - 16 to 18 L =C= 2 to 4 L
PROTEIN 80 to 200g - 5 (?) O 75 to 195 g
Glucose 20 g (see diffusion exchange)
fig. 5.2. Diagram of the ''several circulations" with ap-
proximate magnitudes of each. For explanation of diffusion
exchanges see section 9. For explanation of figures relating to
filtration, absorption, and lymph flow see text of this section.
filtrate. The remaining 2 to 4 liters, including the
unabsorbed protein of the original capillary filtrate,
then produces the third circulation, that of proteins in
lymph.
The potential magnitude of this protein circulation
can be estimated from the observations of Wasser-
man & Mayerson (370-372) on the rates at which
intravenously injected labeled albumin and globulin
disappeared from plasma and appeared in thoracic
duct lymph. The faster component of these two-
phase disappearance curves indicated a steady disap-
pearance of plasma albumin from plasma, and
corresponding appearance in lymph, at the rate of
approximately 0.1 per cent of the total circulating
plasma protein per minute. Allowing for the slightly
slower disappearance rate of globulin (372), this
amounts to the passage through the capillary wall in
24 hours of a mass of plasma protein approximately
equal to that in the circulating blood itself. This
includes passage from the more permeable hepatic
and intestinal capillaries as well as from the less
permeable limb capillaries. Courtice (49; 386, p. 87)
collected lymph simultaneously from the thoracic,
right lymphatic, cervical, foreleg, and hind leg ducts.
Expressed as percentage of total intravascular protein,
the lymph collected from these several sources con-
tained a 24-hour protein mass equaling, respectively,
47.5, 3.6, 2.4, 2.2, and 1.8 or, in total, 57.5 per cent
of the intravascular protein mass. Again this rate of
passage is a basal rate found in resting animals. Intra-
venous infusions increased the rate of protein passage
severalfold (371) and increased lymph flow from the
thoracic duct correspondingly (180), indicating that
the interstitial circulation of both protein and fluid
can be very rapid indeed. Studies during muscular
exercise would be most interesting, but have not been
done so far.
For man the magnitude of this protein circulation
can be estimated in two ways. First, the obligatory
capillary filtrate of 20 liters per 24 hours, containing
0.2 to 0.4 g per cent protein, would carry with it a
minimum of 40 to 80 g of protein per 24 hours. This
figure is unquestionably too low because it does not
include the higher protein content of capillary fil-
trates from liver and intestine. Second, collections of
thoracic duct lymph with analyses of protein content
have been carried out in two patients with accidental
fistulae (57, 60) and in patients with terminal neo-
plasm (15). As mentioned above, the data of Crandall
ct al. (60) justify an estimate of 2 to 4 liters of lymph
per 24 hours. Since the protein content of this lymph
ranged from 3.19 to 4.88 g per 100 ml the circulation
HANDBOOK OF PHYSIOloc.l
CIRCULATION II
of protein can range from 60 to 200 g per 24 hours,
depending upon conditions. Hence the third, or
protein, circulation passing from capillaries to inter-
stitial fluid to lymph (fig. 5.2, F-IF-L) involves daily
a volume of fluid which approaches the volume of
circulating plasma and a mass of protein equivalent
to a quarter or more of the mass of the circulating
plasma proteins. Cope & Litwin (45a) have recently
emphasized the compensatory importance, during re-
covery from hemorrhage, of this continuing flow of
lymph and its contained protein from the interstitial
spaces into the blood stream.
6. FILTRATION COEFFICIENTS OF
CAPILLARIES {kc) J AND OF TISSUES (kt)
A. Nc
■jl Capillaries
Measurements of fluid movement through the
capillary wall as a function of hydrostatic and osmotic
pressures have been made in single capillaries of
amphibian mesenteries (23, 200, 201, 383); in the
human forearm (24, 188, 209), in the perfused ex-
tremities of frogs (61, 74); of rats (162, 302); of
cats and dogs (281, 282); and in lung (132).
The primary measurements necessary to test the
validity of the Starling hypothesis were first obtained
by micromanipulation techniques in single capillaries
of the frog's mesentery (200, 201) with results shown
TTp| in vivo= e.6 to II 7cm H2O
higher in summer
frogs lower in
winter frogs
(hypoproteinemia)
E
ST .04
5-°2
E
o 04
.06
0
ABSORPTION
CAPILLARY PRESSURE CM WATER
,.."* Slope -kc =
filtration coefficient
.0056 juVsec/ju2/ cm H20
A(Pc-TTpl)
20 25 30
fig. 6. 1 . Relation between fluid movement through walls
of single capillaries of frog's mesentery and capillary blood
pressure as determined by micromanipulation methods. Slope
of line indicates filtration coefficient \kr) in n3 of fluid filtered
(or absorbed) /sec /m2 of capillary wall/cm H2O capillary
pressure. Intercept of line with zero axis measures effective
osmotic pressure (in vivo) of the plasma protein. [From Landis
(200).]
in figure t>. 1 . When capillary pressure exceeded 12 cm
water, fluid passed from the plasma inside the capil-
lary to outside the capillary (filtration). When capil-
lary pressure was less than 10 cm water, fluid was
withdrawn from the e.xtravascular space into the
capillary (absorption). At capillary pressures be-
tween 9 and 1 3 cm water there were many instances
in which little or no movement of fluid occurred. In
this range hydrostatic pressure was apparently
balanced by the osmotic pressure of the plasma pro-
teins. This was taken to be indirect evidence that the
walls of the mesenteric capillaries of the frog were
relatively impermeable to protein and that, at least
in these vessels, 9 to 13 cm water (average 1 1.5 cm)
represented the effective osmotic pressure of the
plasma proteins in vivo.
In addition to supporting the filtration-absorption
hypothesis of Starling these results also provided the
first measure of the permeability of the capillary wall
to isotonic fluid. When plotted against capillary pres-
sure the rates of fluid movement were directly pro-
portional to the difference between the capillary
pressure and the effective osmotic pressure of the
plasma proteins measured against the capillary wall
as a filter. The proportionality constant was com-
puted from the slope of the straight line drawn through
the observed points by the method oi least squares.
This was originally called a "filtration constant,"
but for reasons given below the term "filtration
coefficient" is preferable (276). For normal mesenteric
capillaries of the frog the filtration coefficient, kc,
derived from 70 observations, averaged .0056 /j:l of
fluid per sec per fi2 of capillary wall per cm water
difference between capillary pressure and the osmotic
pressure of the plasma proteins. Wind (383) found
great variation from capillary to capillary in the
toad's mesentery. Collectively, these figures provide a
slightly lower average figure, about .0032, during
the first 1 5 min after the mesentery was exposed and
a somewhat higher average figure, .0084, thereafter.
Deviations from these normal filtration coefficients
have proved helpful, as will be described below, in
measuring the effects of temperature (23), oxygen
lack (201), and injury (200) on the filtration-absorp-
tion mechanism in the frog's mesenteric capillaries.
To test the validity of the Starling hypothesis in
another tissue, and particularly in man, Krogh et al.
(188) studied the movement of fluid through the
capillary walls of forearm tissue in a pressure plethys-
mograph, by means of which the blood vessels could
be collapsed in order to measure small increments of
tissue volume produced by filtered fluid. As shown in
I VII \\(,1 ill SUBSTANCES I Hid 1 1 '.II CAPILLAR"* WALLS
989
figure 6.2 net filtration of fluid increased linearly
with venous pressures above 10 cm water. A unit
rise of venous pressure (1 cm water) increased the
filtration rate by .0023 ml per min per 100 ml of
forearm tissue when congestion periods of 30 min
were used (188) and by 0033 ml per min per 100 ml
forearm tissue when congestion periods of 10 min were
used (209). As described in section 4, this difference
was regarded as the result of increasing interstitial
fluid pressure as the volume of filtrate in the tissues
increased.
Brown et al. (22) have more recently studied, by a
totally different method, the filtration coefficient for
the whole body of man (except the thorax ) during a
systemic rise of venous pressure produced by repeated
Valsalva maneuvers. Though results varied slightly,
depending on the method of calculation, representa-
tive filtration coefficients for the whole body were
in the first 9.5 min, .0036 ml per min per 100 g body
wt per cm rise of venous pressure and, for a total
of 29.5 min, .0014. These values can be compared to
.0033 and .0023 for the forearm alone. The two sets
.200
.160
080
.040
-020*
fig. 6.2. Rates of filtration measured by pressure plethys-
mograph in human forearm during graded elevation of venous
pressure for 10-min periods. Plethysmograph temperature,
34-35 C. The slope of the line corresponds to a nitration co-
efficient (ki) of .0033 ml/min/100 ml forearm tissue per cm
H2O increase of venous pressure. [From Landis & Gibbon
(209).]
of figures are similar, presumably because the collec-
tive capillary beds of muscle and subcutaneous tissue
are large compared to the smaller, though more
permeable, capillary beds of liver and intestine. A
similar relationship has been found with respect to
diffusion (see section 8). The '"whole body" filtration
rate appears to decline more rapidly than that of the
forearm, owing probably to more rapid return of
capillary filtrate by way of the lymphatics, particu-
larly during the vigorous respiratory movements
required for repeated, brief Valsalva maneuvers.
Landis & Hortenstine (210) calculated from the
forearm filtration figures (188, 209) that a rise of
venous pressure, throughout the body, to 10 cm
water above normal might, in a man weighing 75 kg,
filter as much as 250 ml of fluid from the plasma in the
first 10 min. This has proved a fairly good estimate.
Brown et al. (22) observed the filtration of 333 ml
to 501 ml when systemic venous pressure was ele-
vated by 20 cm for 9.5 min. Over 29.5 min an
increase of venous pressure by 20 cm water filtered
460, 41 7, and 687 ml of fluid, calculated to contain
between 1 and 2 g of protein per 100 ml owing, pre-
sumably, in part to the very high protein content
of capillary filtrate from hepatic and intestinal
capillaries.
The pressure plethysmograph was used by Krogh
et al. (188) also to test the effect on filtration rate
of changing the osmotic pressure of the plasma
proteins. Filtration rates at given venous pressures
were measured with the subject recumbent and then
at the same venous pressures while the subject stood
quietly on a tilt table for 30 min or more. Quiet
standing increased the concentration of the circulating
plasma proteins by 0.6 to 1.1 g per 100 ml and the
protein osmotic pressure of plasma by 33 to 8.7 cm
water. At these higher protein osmotic pressures the
rate of filtration produced by a given venous pressure
was always lower. A unit rise of protein osmotic
pressure (1 cm water) was accompanied by a reduc-
tion of filtration rate ranging from .0027 to .0045 ml
per min per 100 ml forearm tissue. These values were
quantitatively similar to the effect produced by
elevating venous pressure by 1 cm water, but opposite
in sign. Within the limitations of the method these
results justified extending the Starling hypothesis to
the forearm capillaries of man, and were compatible
with the view that the capillaries of the human fore-
arm were relatively impermeable to the plasma
proteins.
In the first studies with the pressure plethysmo-
graph (188, 209) it was perplexing to find that venous
99°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
CC /MIN
+ 50
J • Normal Plasma
lAlburnin
28%
Globu
n 24%
TTp =
16.0
mm
»i
f x After
add
ng bov
ne Albumin
1 Albumin
6.7%
Globu
in 20%
TTp =
501
mm
»t
fig. 6.3. Relation of net fluid move-
ment in perfused hind leg of cat to
difference between the mean hydro-
static pressure in the capillaries {pC)
and the sum of all pressures opposing
filtration (isogravimetric capillary
pressure, pCi). The slope of the line
corresponds to a filtration coefficient
(kt) of 0.014 ml/min/100 g tissue/
mm Hg pressure difference. [From
Pappenheimer & Soto-Rivera (282).]
ABSORPTION
FILTRATION
pressure had to be elevated by 10 to 17 cm water
before net nitration could be detected (fig. 6.2).
Brown et al. (24) showed later, however, that the
regression lines relating filtration and venous pressure
passed through zero, provided a) that interstitial
fluid was carefully evacuated from the forearm prior
to congestion, and b) that a correction was made for
the volume of interstitial fluid pressed out of the fore-
arm segment during each volume measurement.
From regression lines such as the one shown in
figure 6.2 it is possible to calculate an approximate
filtration coefficient (kt) for forearm tissue if allow-
ance is made for the fact that a given elevation of
venous pressure produces a somewhat smaller eleva-
tion of mean capillary pressure. On the assumption
that the latter is 80 per cent of the former, kt for
human forearm capillaries becomes approximately
.0057 ml per min per 100 g tissue per mm Hg as
given in table 6.2. The filtration coefficient for the
whole body becomes .0061.
Among the several perfusion methods that have
been used to measure filtration coefficients, the most
precise and revealing is the isogravimetric technique
developed by Pappenheimer & Soto-Rivera (282)
in which filtration and absorption were identified by
changes in weight of an isolated limb. Arterial pres-
sure, venous pressure, osmotic pressure of the perfus-
ing fluid, blood flow, and temperature could be
varied at will and their influence on the filtration-
absorption equilibrium could be measured separately.
A detectable effect on fluid movement resulted from
a change of venous pressure by 0.5 mm Hg and some-
times less, or from a change of arterial pressure by
2 to 4 mm Hg. Capillary pressure was 5 to 10 times
more sensitive to a change of venous pressure than
to a change of arterial pressure.
Figure 6.3 shows net fluid movement, i.e., filtration
or absorption, plotted against the difference in pres-
sure across the capillary membranes themselves.
Filtration and absorption were proportional to the
difference between the calculated mean capillary
blood pressure and the isogravimetric capillary pres-
sure which is, by definition in this method, the sum of
all pressures opposing filtration. In figure 6.3 the
slope of the regression line indicates a filtration
coefficient of 0105 ml per 100 g tissue per min per
mm Hg change of capillary blood pressure. The filtra-
tion coefficient was independent of the absolute value
of the isogravimetric capillary pressure when this was
varied by diluting or concentrating the proteins in the
perfusing fluid. Similar methods have been applied
recently by Renkin & Zaun (302) to the hind legs of
the rat.
The constancy of filtration coefficients at high and
low capillary blood pressures (figs. 6.1, 6.2, 6.3)
suggest that under these conditions capillary surface
area and capillary porosity are not significantly
modified by pressure. This may be related to the con-
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
99 1
elusions reached by Burton (31) that vessels of small
diameter are relatively indistensible. Under more
severe conditions and in other tissues, however, the
permeability of the capillary walls may be increased
when capillary blood pressures are very high (211)
or when blood volume is much increased (131, 331,
361, 369).
In all the regions so far considered, resting average
capillary blood pressure is approximately equal to
the osmotic pressure of the plasma proteins. In the
lung, however, as described in section 2, average
capillary pressure is only 5 to 10 mm Hg and there-
fore less than half the osmotic pressure of the plasma
proteins. Figure 6.4 shows the rate of edema forma-
tion in the lungs of dogs plotted against left atrial
pressure (132). In contrast to other tissues, net filtra-
tion and increase of interstitial fluid volume were not
observed until, at atrial pressures of 25 to 30 mm Hg,
pulmonary capillary pressure began to exceed the
protein osmotic pressure of 25 mm Hg. Thereafter
filtration increased linearly with left atrial pressure
at a rate of 0.21 g of fluid per hour per mm Hg per g
dry wt of lung tissue or 0.065 g per min per mm Hg
per 100 g wet lung tissue. The relative ''dryness" of
lung tissue which is produced by a low capillary
pressure is indicated by the absence of filtration
between atrial pressures o and 23 mm Hg. This
margin of dryness was reduced to half normal when
the plasma protein concentration was decreased by
plasmapheresis to an average of 47 per cent of the
control protein concentration. Taken together, the
10
20
ATRIAL
30
PRESSURE
40
m m Hg)
50
fig. 6.4. Rate of edema formation (nitration) in lungs of
dogs subjected to prolonged elevations of left atrial pressure.
Significant nitration did not appear until left atrial pressure
exceeded 25 mm Hg, i.e., the osmotic pressure of the plasma
proteins. [From Guyton & Lindsey (132).]
results shown in figures 6.1 to 6.4 permit concluding
that in these four regions the net rates of filtration or
absorption through the capillary walls depend upon
the difference between hydrostatic and osmotic
forces acting across the membrane. In view of this
evidence the Starling hypothesis of 1896 (345) can
fittingly be called now the Starling filtration-absorp-
tion principle.
Progress has gone beyond this qualitative stage,
however, because the meaning of k, the filtration
coefficient, has been expanded not only by numerical
values for a number of capillary beds and membranes
(tables 6. 1 and 6.2) but also by more precise definition.
Pappenheimer (276) called attention to the fact that
the several different "filtration constants," "unit
filtration rates," or "filtration coefficients" used by
various authors can be related to the equation used
by Darcy (62) to describe the viscous flow of fluids
through inert porous or fibrous materials, viz. :
0.
(6.1)
where
Qj = quantity filtered per unit time
k = specific filtration constant of the porous material or
membrane
Am — area of membrane
AP = pressure difference across membrane (in capillaries
ap = pc- Pif - np, + n;/)
Ax = path length through membrane (for capillaries
usually assumed to be 0.3 ju)
7j = viscosity of filtrate
If the area of capillary wall can be measured
directly (23, 200, 201, 383) or computed (281, 282) as
in table 6.1, the proportionality factor or filtration
coefficient consists of k/riAx including the Darcy
"specific filtration constant," the thickness of the wall,
and viscosity of the fluid. On the other hand, for
tissues in which the capillary surface per weight or
volume of tissue is not yet known precisely, e.g., in
the human forearm, the hind quarters of the rat, and
the lung, the proportionality factor for unit tissue
weight or volume will consist of kAc/qAx including,
in addition, the area of the capillary walls, Ac- The
term "filtration constant" is certainly inappropriate
and should be abandoned for a membrane system as
heterogeneous as that in the capillary wall. Filtration
coefficient is a preferable term and it is suggested that
the symbol kc be used for cases where the area of
capillary wall is measured or computed and kt be
used for coefficients based on mass or volume of
tissue. Newer developments in pore theory have led
992 HANDBOOK OF PHYSIOLOGY -" CIRCULATION II
table 6.1. Filtration Coefficients {Hydr adynamic Conductivity) Through Various Membranes*
Filtration Coefficient,
Type of Membrane
Temperature. C
ml X io»
References
sec X cm2 X cm II 2O
Cell membranes
Arbacia egg (unfertilized)
20
0.016
(221, 222)
Fibroblasts imouse, rat, chick)
20-22
O.06-0. 16
(25)
Leucocytes (rabbit, man)
20-23
O . 05-0 . 2
(329)
Erythrocytes (man)
20
O.92
' 334 1
Capillary membranes
Muscle (dog, cat)
37
2-5
(281, 282)
Mesenteric (frog)
22-26
48-74
(23, 200, 201)
Glomerular (frog)
25
220
(281)
Glomerular (mammal)
37
3OO-60O
(281, 276, 339)
Artificial membranes^
Dialysis tubing (Visking) r = 16-23 A
25
IOO-180
(82, 298)
Cellophane (DuPont 450-PT-62) r = 30-40 A
25
35O-9OO
(82, 298)
Viscose wet gel (Sylvania) r = 75-85 A
25
32OO-420O
(82, 298)
* Modified from Renkin & Pappenheimer (301). t Thickness, 0.5 M, r = pore radius.
Pappenheimer (276) to suggest also that the term
"capillary permeability" be reserved for describing
the properties of the capillary wall with reference to
the diffusion of small molecules. Filtration coeffi-
cients, because they deal with flow of fluid through a
membrane, would then be a measure of hydraulic or
hydrodynamic conductivity of the capillary wall.
Table 6.1 is taken from the review by Renkin &
Pappenheimer (301) with inclusion of some more
recent values. It compares filtration coefficients of
cell membranes (upper section), of capillary walls
(middle section), and of certain artificial membranes
(lower section). Cell membranes have smaller filtra-
tion coefficients than capillaries, although the differ-
ence between the values for the erythrocyte and the
mammalian muscle capillary is small. The range of
filtration coefficients for capillary walls is very large,
amounting to a 200-fold difference in the mammal
between muscle capillaries and glomerular capil-
laries. The coefficients for artificial membranes,
calculated for comparable thickness, are in turn
much higher still and, with other evidence, led
Pappenheimer et al. (281) to the conclusion that the
collective area of the pores involved in the filtration
process is only a small fraction of the total capillary
surface. Support for this conclusion came from
measurements of capillary permeability to small
lipid-insoluble molecules, and will be given in sec-
tions 8 to 10.
Table 6.2 compares average filtration coefficients
(kt) for extremities of four species. In the forearm of
table 6.2. Average Filtration Coefficients for Tissues, kt
Species and Tissue
Filtration Coefficient
at 37 C, ml
min X
100 g tissue
X mm Hg Reference
Man
, forearm, intact
0.0057*
(188, 209)
Man
, whole body, intact
0.0061 *
(22)
Dog,
perfused hind leg
0.014
(282)
Cat,
perfused hind leg
0.0105
(281, 282)
Rat,
perfused hind legs
O.O33
(302)
* In the text these coefficients are described for a rise oi
venous pressure by 1 cm water. To facilitate comparison,
values given here have been corrected to 1 mm Hg rise of
capillary pressure. It is assumed that \PC = 0.8 APV.
man, on the assumption that capillary pressure is
increased by 80 per cent of given increases in venous
pressure, the average filtration coefficient becomes
.0057 ml per min per mm Hg per 100 g tissue. From
the data of Brown et al. (22) k, for the whole body is
.0061. In smaller animals progressively larger filtra-
tion coefficients are found. As Renkin & Pappen-
heimer suggest (301), this relationship is ideologically
fitting because the smaller the mammal, the more
active are its metabolic processes and therefore the
greater will be the requirement for a more extensive
capillary bed (320) and for more rapid exchanges
between blood and tissue. To obtain comparable
filtration coefficients for other tissues, e.g., liver,
intestine, lung, and brain, is far more difficult. Values
for lung have been published recently by Guyton &
Lindsey (132) and for brain, or perhaps chiefly the
arachnoidea, by Coulter (47).
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
993
B. Effects of Temperature on Filtration Coefficients
As noted above, the Uarcy equation indicates that
flow through porous materials changes with tem-
perature in inverse proportion to viscosity, and accu-
rate measurements of flow through artificial, porous
membranes have confirmed this expectation (16, 81).
The filtration coefficients for capillaries, kc, and for
tissues, kt, include the viscosity of the filtered fluid
and should change with temperature. This was found
to be clearly the case by Pappenheimer (277) in the
perfused limb in which the capillary membranes
retain their normal impermeability to plasma pro-
teins over a temperature range of 8 to 44 C. The
ratio of the filtration coefficient measured at 36 ± 2 C
to that measured at 10 ± 2 C averaged 1.68 (se =
±0.08). This value was within 10 per cent of the
ratio of viscosities of water at the two temperatures
(77 10 77 36 = 185) and the difference between the
two ratios was not significant (P > .05).
In the frog mesentery Brown & Landis (23) had
observed earlier a decrease of filtration coefficient
(kc) from .0070 to .00 1 g when temperature was re-
duced from 24 ± 2 C to o ± 2 C. However, as men-
tioned by Pappenheimer (276), the decrease in filtra-
tion coefficient was larger than that to be expected
theoretically and the scatter of values for individual
capillaries was too great for quantitative conclusions.
In the human forearm, Landis & Gibbon (209)
found the filtration coefficient (kt) almost halved as
the temperature of the plethysmograph (and of the
superficial tissues of the forearm) was changed from
44 or 45 C to 14 or 15 C. Brown el al. (24) extended
these observations to 4 to 5 C. Figure 6.5 (heavy line)
shows the filtration coefficients observed in the fore-
arm at plethysmograph temperatures over the total
range of 44 to 4 C. Using the data of Barcroft & Ed-
holm (9), the figures just above the bottom line of the
chart show the probable temperatures in muscle and
subcutaneous tissues corresponding to each surface
temperature. Starting from the filtration coefficient
at a tissue temperature of 35 to 36 C (surface tempera-
ture 34 to 35 C) the fainter, dash lines indicate coeffi-
cients of filtration calculated on the basis of deep
tissue temperature and the corresponding change in
viscosity of capillary filtrate. Observed and calculated
filtration coefficients agree fairly well at tissue tem-
peratures ranging from 35 to 17 C. Above 35 C the
observed coefficient is much higher than the calcu-
lated one, indicating that changes other than viscosity-
are involved. To be considered are such factors as
increased capillary pressure and filtering area (8)
secondary to vasodilatation or, possibly, opening of
arteriovenous anastomoses with elevation of small
vein pressure. Below a tissue temperature of 1 7 C, the
observed filtration coefficient does not decrease as
it should if viscosity alone were involved, but in-
0040
COEFFICIENTS
CALCULATED
FILTRATION
COEFFICIENT
.0035
POSSIBLE EFFECTS
OF CHANCING VISCOSITY
OF FILTRATE
ml per mm
per cm H,0 change
of venous pressure
in 100 ml of tissue
@
PROBABLE TISSUE TEMP C
"""©
38 3
i 28 21
17
talc from MUSCLE >
,, ,.„ SUBCUTANEOUS ->
Barcroft 1946
40 3
5 25 17
1 1
II
fig. 6.5. Effects of tempera-
ture on filtration coefficients, kt,
observed in the human forearm
(solid line), compared with fil-
tration coefficients to be expected
by calculation from change of
viscosity of water by reason of
temperature changes in muscle
( — — ) and subcutaneous tis-
sue ( — - — - — ) [Calculated
from results of Landis & Gibbon
(209), Brown et al. (24) and,
for deep temperatures, Barcroft &
Edholm (9).]
34 to 35 24 to 25 14 to 15
TEMPERATURE OF WATER AROUN0 FOREARM C
PRESSURE PLETHYSMOGRAPH - COMBINED RESULTS
994
HANDBOOK. OF PHYSIOLOGY
CIRCULATION II
creases considerably. Moreover, Brown et al. (24)
found that even at normal venous pressures cooling
the surface of the forearm to 4 C produced a slow but
steady increase of the reduced forearm volume pre-
sumably because of filtration and augmented inter-
stitial fluid volume. These results suggested "cold
injury" of surface capillaries and diminished effective
osmotic pressure of the plasma proteins. Passage of
protein through the capillary wall in severe cold has
been described by Lewis (218), who found up to 3 g
per cent of protein in the edema fluid. In summary, it
appears that for intact tissues the effects of moderate
changes of temperature on filtration coefficients can
be explained fairly well by the changing viscosity of
capillary filtrate. At very high and very low tem-
peratures other factors, as yet unanalyzed, become
more important.
C. Adsorbed Plasma Protein and Filtration Coefficients
The functional dimensions of capillary pores, and
hence the filtration coefficients of capillaries, are
probably determined in part, by a layer of adsorbed
plasma protein. Krogh & Harrop (186) were the
first to note that perfusion of extremities with non-
protein colloids fails to prevent edema. Their observa-
tions were confirmed and extended by Drinker (74),
Danielli (61), and Shleser & Freed (332). Kinter &
Pappenheimer (cf table 6.3) found that dextrans
failed to exert their full osmotic pressure in vivo
unless more than 0.2 per cent protein was present
in the perfusion fluid. Net filtration usually occurred
in dextran-Ringer perfused muscle at all venous
pressures; 10 to 20 min after addition of 1 per cent
plasma protein the direction of net fluid movement
was reversed as the osmotic pressure of the dextran
became effective across the capillary walls. The
phenomenon was fully reversible and could be re-
peated several times on the same preparation during
the course of a few hours. The capillary filtration
coefficient was usually more than doubled when
Ringer's solution (295) or Ringer-dextran solutions
were substituted for plasma. In nine experiments the
filtration coefficient averaged 0.016 ± .003 ml per
min per 100 g tissue during perfusion with blood,
0.037 ± .002 during perfusion with protein-free red
cell suspensions, and 0.019 ^ °°3 when protein was
restored to the perfusion fluid. The effect appears to
be nonspecific, since normal filtration coefficients
were found in cat or rat hind limbs perfused with
human or bovine serum albumin (295), cat hemo-
globin, or bovine hemoglobin (299).
table 6.3. Effective Osmotic Pressures of Clinical
Dextran in Capillaries of Perfused Cat Hind Limbs
Effective
Pressure
Osmotic
mm Hg
Perfusion Fluid
In vitro
(Hepp
osmom-
eter)
In vivo
(perfused
limb)
Experiment I
a) 3rc Dextran
1°7t Plasma protein in Ringer
24.6
21 .O
b) 3% Dextran in Ringer
22
6.6
Experiment 2
a) 3% Dextran
2.4^, Plasma protein in Ringer
30.8
28.5
b) ■>,% Dextran
0.2% Plasma protein in Ringer
23.2
'3-7
Experiment 3
a) 3% Dextran in Ringer
25.O
12.2
b) 3% Dextran
3% Plasma protein in Ringer
31-4
26.4
From unpublished experiments of Kinter and Pappen-
heimer.
The minor axis of serum albumin is about 30 A
and complete removal of albumin from the inside of
a pore might increase effective pore radius by this
amount. Given a mean pore radius of 45 A (see
sections 9 and 10), the filtration coefficient would be
expected to increase by the factor (30 + 45)4 -f- (45)4
or more than sevenfold. A reversible increase of this
magnitude was observed in only one preparation,
but it is possible that even prolonged washout with
protein-free solutions fails to remove all adsorbed
protein. The effects of adsorbed protein should be con-
sidered, however, in comparing pore dimensions
calculated from permeability measurements with pore
dimensions observed in electron micrographs.
D. Effects of Injury on Filtration, Absorption,
and Filtration Coefficients
capillary stasis. Cohnheim in 1867 postulated a
"molecular alteration in the vessel walls" and aug-
mented "porousness" to explain the transudation of
fluid, protein, and cells in inflammation (42, 43).
Since then abundant qualitative evidence has indi-
cated that injury of many types increases the perme-
ability of the capillary wall to fluid and protein (207).
Inflammation is, however, an exceedingly complex
series of reactions (246, 247, 344), of which increased
capillary permeability is only one part. Physiologists
have, therefore, tended to study simpler forms of
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
995
injury. A few quantitative measurements in terms of
filtration coefficients are available and provide esti-
mates of the increased porousness even though in-
formation on the "molecular alterations" of the
vessel walls is still completely lacking.
In considering the mechanism of simple, chemical
injury, Krogh & Harrop (187) in 1921 described
'capillary stasis" as direct microscopic evidence of
increased permeability of the capillary wall. The steps
by which chemical injury leads to capillary stasis
merit full description, because capillary pressure,
flow, filtration, absorption, and diffusion are all
affected. The changes observed in a single, damaged
capillary form a unit lesion which helps to explain
the effects of more generalized injury.
Blood corpuscles, when first entering a capillary,
are clearly separated by plasma. As long as the capil-
lary wall is normal this remains the case and flow con-
tinues. Even when capillary blood pressure is high,
filtration reduces the volume of plasma in flowing
capillary blood very slightly, because the fraction of
plasma filtered is normally less than 4 per cent of the
plasma volume at most, and usually 1 to 2 per cent.
However, as soon as injury is produced, e.g., by ap-
plying 25 per cent urethan or 10 per cent alcohol in
Ringer's solution, the corpuscles clearly begin to
move closer together as they flow along the capillary,
because plasma is lost progressively through the now
injured capillary wall. Eventually, at the venous end
of the capillary, nearly all the plasma having been
filtered off, the corpuscles become so closely packed,
and collectively so viscous, that they come to a stand-
still in the capillary and form a localized plug of cells
just short of the venule. Meanwhile plasma, with few
or many erythrocytes, continues to enter the arteriolar
end of the capillary, though much more slowly than
before and in a distinctly pulsatile fashion, because
entry is now limited by the volume of plasma being
filtered through the damaged capillary wall.
The plasma of this blood is also lost by rapid filtra-
tion. The additional corpuscles are progressively con-
centrated in their turn and finally deposited cumu-
latively on the already existing column of erythrocytes
in the venous end of the capillary. Eventually a
column of packed cells fills the whole capillary and
takes on a characteristic, transparent, bright red
color, apparently because the erythrocytes are so
closely packed that light rays are no longer refracted
as they are when the surfaces of single corpuscles are
lormally separated by intervening plasma. Flow
ceases entirely in capillaries thus filled and plugged.
If injury is severe, capillary stasis is irreversible. If
injury is mild, resumption of flow is frequently ob-
served. The first indication of beginning recovery is
the loosening of corpuscles in the packed column,
followed by slow, then more rapid, squeezing of the
column into the stream of the nearest venous capillary
or venule. Here the cells can be seen separating
easily in the plasma of the venules as they are carried
away. In this respect simple stasis differs from the
"sludged" corpuscles described by Knisely et al.
(176) for more drastic states in which the corpuscles
adhere to each other and form minute emboli.
Even after flow has returned some erythrocytes
and leukocytes usually remain adherent to the inner
surface of the damaged wall, but eventually these,
too, float free (199). Platelets may be seen adhering
to the wall for still longer periods and probably help
restore relative impermeability to protein as sug-
gested by Danielli's perfusion studies (61) in which
platelets reduced the rate of edema formation to
one-tenth that found with platelet-free perfusion
media. Platelet protein, in association with calcium,
has also been found to restore normal permeability
(380-
Chemical injury of the grade just described in-
creases capillary permeability enough to permit
passage of plasma proteins (200), colloidal dyes (107,
152, 184, 199), and colloidal starch (184) but, as
observed by light microscope, the walls of true capil-
laries still retain most of the carbon particles of in-
jected India ink (152, 184, 199). This is true also of
localized mechanical injury produced by compressing
capillaries with a glass rod (199), or by prodding
with a minute needle (37). In these simpler forms of
injury gross ruptures of the capillary wall are not
present because carbon particles, as well as erythro-
cytes, are retained as plasma is filtered off.
FILTRATION COEFFICIENTS, kc, OF INJURED CAPILLARIES.
The permeability of injured capillaries has been
measured in the frog's mesentery by determining their
filtration coefficient during stasis using the method
already described for normal capillaries (200).
Figure 6.6 shows filtration rates plotted against
capillary blood pressure, injury having been pro-
duced by irrigating the mesentery with 10 per cent
alcohol or 1'. 10,000 mercuric chloride in Ringer's
fluid. As with the normal capillary wall, filtration
increased linearly with capillary blood pressure.
Comparison of the regression lines for injured capil-
laries (above) and normal capillaries (below) indi-
cates, however, that the filtration coefficient was
996
HANDBOOK OF PHYSIOLOGY
CIRCl I .ATION II
50
FILTRATION
•
+
40
-
Injured j
copillaries /
•
RATE
• /
OF 30
-
+ /
FLUID
MOVEMENT
• • / •
/ •
20
• /
cubic micro
/ •
per second
_
per sq micron
+
10
"
•
"
• /
Normol
0
copillories
"
-in
ABSORPTION
i 1 1
i i
5 10 15 20
CAPILLARY PRESSURE - cm woter
fig. 6.6. Effects of severe chemical injury on fluid move-
ment through walls of frog's mesenteric capillaries. Slope of
lower regression line shows filtration coefficient, kc, for normal
capillaries. Slope of upper regression line indicates the 7-fold
increase of filtration coefficient found after injury. Filled circles
refer to injury by io^i alcohol in Ringer's fluid; plus signs,
to 1:10,000 mercuric chloride in Ringer's fluid. [From Landis
(200).]
increased from the normal value of 0.0056 to ap-
proximately 0.0390 y? per sec per p* of capillary wall
per cm water of capillary pressure, indicating a
sevenfold increase of hydrodynamic conductivity.
Increased permeability of the injured wall to plasma
proteins is indicated by the absence of absorption even
at low capillary pressures and by the reduction of the
in vivo osmotic pressure of the plasma proteins from
the normal value of 1 1 cm water to between 3 and 4
cm water. Thus the effects of severe injury are a) in-
creased filtration, b) absence of absorption, c) reduced
effective osmotic pressure of the plasma proteins,
and d) eventual cessation of flow in any capillary
injured to the point of stasis. Diffusion rates have not
been measured in such capillaries. Presumably, since
capillary permeability to fluid and protein is greater,
net diffusion of small molecules should be increased
as long as blood flow continues. However, since net
diffusion is flow limited, its effectiveness in exchanges
of substances will decline as flow decreases and will
soon cease in those capillaries that are filled with
stationary, closely packed erythrocytes.
capillary pressure in injury. The appearance of
edema in injured regions is due primarily to increased
capillary permeability, but is enhanced by increased
capillary blood pressure. Local injury elevates capil-
lary blood pressure by at least two mechanisms: /)
the vasodilatation and increased blood flow which
are parts of the triple response to injury described by
Lewis (217), and 2) the temporary blockage of
capillary blood flow and passive congestion produced
by stasis (199).
Application of a minute silver nitrate crystal to the
skin of the frog's web increases capillary pressure in
the neighborhood of the lesion to peak values which
are as much as double the earlier control values (205).
Within 10 to 20 min capillary pressure is again
within the normal range. In human skin the flare of
the triple response produced by histamine is accom-
panied by peak capillary pressures of 10 to 25 mm
Hg above preceding control values (203), but again
with relatively prompt return to control values. The
onset and duration of these elevations suggest that
they are a part of the flare due to the "axon reflex."
Elevations of capillary blood pressure are also
found in capillaries injured to the point of stasis.
These elevations are more important in the formation
of edema fluid during injury because they occur in
vessels, the walls of which are permeable to protein
and hence already the site of rapid filtration without
any balancing absorption. Figure 6.7 shows the cycle
of pressure changes which occurred in one experi-
ment involving stasis and recovery. Control capillary
blood pressure, with normal blood flow, ranged from
12 to 15 cm water. At the time marked A 25 per cent
urethan was applied to the mesentery and the onset
of stasis, as indicated by visible loss of plasma, was
clearly present at B. The sharp rise of capillary pres-
sure between B and C occurred as the venous end of
the capillary was filled and blocked by packed
erythrocytes. As flow ceased capillary pressure rose
rapidly to approach the pressure in the feeding
arteriole. At C a pressure of 22.5 cm water merely
stopped the advance of erythrocytes toward the
pipette. Even 30 cm water did not move the cor-
puscles away so that capillary blood pressure was
well in excess of the 22.5 cm charted. The very rapid
filtration observed during this period is due, there-
fore, to increased permeability and also to high
capillary pressure. Between C and D this enhanced
filtration of whole plasma packed erythrocytes
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
997
CAPILLARY
BLOOD
PRESSURE
CM
H,0
10 15
TIME - MINUTES
fig. 6.7. Chart indicating the changes of capillary blood
pressure in frog's mesentery during capillary stasis produced
by applying 25^0 urethan solution (A to C) and during re-
covery with resumption of capillary blood flow (C to D).
[From Landis (199)-]
tightly throughout the capillary. After D, arteriolar
pressure having been blocked by the packing of
erythrocytes up to the arteriocapillary junction,
capillary blood pressure fell to the level in the venule.
This secondary fall of pressure seems to assist recovery
from stasis because, as it occurs, one can observe that
the erythrocytes become less tightly packed and slow
movement toward the nearest venule begins. At D
sluggish flow was being resumed. Hence in injury,
capillary stasis and the rapid accumulation of rela-
tively large volumes of protein-containing edema fluid
or blister fluid depend primarily upon increased per-
meability, but also upon increased capillary blood
pressure. Recovery from stasis, while assisted by tem-
porary lowering of capillary blood pressure, cannot
occur until the permeability of the wall to protein
returns toward normal.
TISSUE ASPHYXIA; RELATION OF FILTRATION COEFFI-
CIENTS to 0>, C02, and pH. The effects of arrested
blood flow and, more specifically, of hypoxia on capil-
lary permeability and the filtration-absorption mecha-
nism are still uncertain. In general, it appears that
arrest of blood flow must be total and prolonged, and
that hypoxia must be severe, before changes in
permeability become demonstrable. Lazarus-Barlow
(213) in 1894 studied the edema of passive congestion,
and also the edema which appeared when blood flow
was restored after a prior period of complete arterial
and venous occlusion. He ascribed the latter edema
to functional modification of the vessel walls secondary
to "starvation of the tissues" and accumulated
waste products. More recently Pochin (285) found
in the rabbit's ear that occluding the circulation for 2
hours led to demonstrable edema which appeared
shortly after circulation was re-established. Occlusions
of 16 to 18 hours produced enough edema fluid to
permit collecting samples in which the protein con-
tent approached 5 g per cent. Edema alone might
conceivably have been the result of vasodilatation
and high capillary blood pressure that probably
followed this arrest of the circulation, but the high
concentration of protein in the edema fluid indicated
that increased permeability was also present.
Among the factors that might change capillary-
permeability under these conditions, the first to be
considered are those associated with continued
metabolism of tissues in the absence of blood flow,
viz. a) reduced oxygen tension, b) increased carbon
dioxide tension, and c) local decrease of pH due to
accumulation of metabolites such as lactic acid.
Table 6.4 summarizes the effects of these variables
on the filtration coefficients (kc) and on the in vivo
effective osmotic pressures of the plasma proteins
measured in single mesenteric capillaries of the frog
(201). They indicate that Ringer's solution, saturated
with C02 or acidified by HC1 to pH's between 7.0
and 5.0, had no significant effect on the permeability
of the capillary wall. Only when pH was made 4.0
or less, and hence unphysiologically low, was there
evidence of increased permeability to fluid and
protein.
Severe and, so far as possible, total oxygen lack
made the capillary wall permeable to protein and
fluid as indicated by decreased effective osmotic
pressure of the plasma proteins and by increased
filtration coefficient, respectively. It must be em-
phasized that the lowering of 02 tension in these
experiments was maximal because not only was
blood flow stopped by tightening a loop around the
mesenteric artery, but the mesentery was also ir-
rigated freely with Ringer's solution previously boiled
and kept saturated with nitrogen. The possibility
that metabolites from anaerobic metabolism were
responsible could not be excluded. The effects on
permeability were, however, still reversible because,
if the period of severe hypoxia was brief enough, e.g.,
3 min, resumption of blood flow and irrigation with
oxygenated Ringer's solution restored both the
filtration coefficient and the in vivo osmotic pressure
of the plasma proteins toward normal, as shown in
table 6.4. For comparison, at the bottom of table 6.4
998
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
table 6.4. Effects of C02, 02, f>H, and Severe
Chemical Injury on Frog's Mesenteric Capillaries
Filtration Coefficient,
Effective npi
(in vivo)
cm H2O
sec X if- X cm H2O
.004B-.0074* 8. 6-1 I. 7
.0088 11.8
.005b
"■5
.0065
11. 7
.0074
n. 4
.0152
11. 6
.0207
7.8
rapid stasis
.0231
6-5
.0080
"•5
.0390
<4.o
Capillaries Irrigated by
Ringer's Fluid with
Usual aeration
Saturated CO»
pH 8.0
HC1 to pH 6.0
5-o
4.0
3-5
3-o
()■. lack and arrested blood
How for 3 min
After 15 min recovery
10% alcohol or mercuric chlo-
ride, 1 : 10,000
* Accumulated control measurements (23, 200, 201).
is shown the still greater effect of chemical injury
severe enough to produce irreversible capillary stasis.
In contrast to the effects of extreme local hypoxia
just described, studies of graded hypoxemia have
demonstrated that the capillary wall tolerates less
severe grades of oxygen lack very well. In the human
forearm, Henry el al. ( 1 50) found that oxygen tension
of venous blood must be reduced to between 1 5 and
25 mm Hg before protein passage was increased
above normal. This corresponds to an oxygen satura-
tion of 15 to 25 per cent or an oxygen content of 4 to
6 vol per cent, assuming the blood has a normal
hemoglobin content. The method used to measure
protein passage was, however, indirect and the
protein content of capillary filtrate varied widely.
DiPasquale & Schiller (70) and Hendley & Schiller
(148) studied the effects of hypoxemia on the rate of
edema formation in limbs of rats perfused with Krebs-
Ringer solution containing 20 per cent washed red
cells of dog and 0.33 per cent gelatin. When the
oxygen content of the perfusing fluid was kept above
5 vol per cent, the rate of edema formation remained
at the control level. Reducing oxygen content to
between 0.88 and 2.60 vol per cent increased the
rate of edema formation above control levels by 42
per cent in the first 20 min, by 87 per cent in the
next 20 min, and by 151 per cent in the third 20-
min period. Blood flow having been kept constant
to exclude effects of the vasodilatation which ac-
companied this hypoxemia, they concluded that the
critical level below which hypoxemia influences
the permeability of a capillary wall was probably
about 2.6 vol per cent. No observations on protein
passage were made. In further studies Hendley &
Schiller (149) found, however, that either histaminic
(Xeo-Antergan) or adrenergic (Dibenzyline) blockade
eliminated these results on the basis either of specific
blocking action or of hemodynamic effects, and the
meaning of these studies therefore remains a chal-
lenging problem.
Systemic hypoxemia, within the range compatible
with the life of the organism, has no certain effect on
capillary permeability. Maurer (228, 229) and
Warren & Drinker (367) found, in dogs, that breath-
ing 8.0 to 1 1 .5 per cent oxygen in nitrogen augmented
the flow of lymph from the lungs and cervical region,
increased the total amount of lymph protein col-
lected in unit time, but decreased the concentration
of protein in that lymph. Although an increase of
capillary permeability was postulated, the decreased
protein concentration in lymph, taken together with
the studies of Courtice & Korner (53, 1 79) make
it unlikely that permeability to protein was changed.
In the human forearm McMichael & Morris (242)
found that breathing 9.5 per cent oxygen did not
increase filtration hum capillaries during venous
congestion. Moreover, in patients with generalized
hypoxemia sufficient to impair cerebral function,
Stead & Warren (351) observed no significant in-
crease in the protein content of edema fluids collected
from the extremities.
Only in agonal or antemortem stages of asphyxia
(34) or anoxemia ( 1 60 ) is there some slight evidence
of increased capillary permeability. In shock the
possibility that generalized hypoxemia might increase
capillary permeability has been considered on many
occasions. Careful studies with labeled plasma proteins
(45, 101, 102) have shown the expected rapid passage
of protein through capillary walls locally in burned
or crushed tissues. However, no abnormal passage
through capillary walls elsewhere in the body has
been found until just before death, again as an
agonal or antemortem occurrence.
In view of the many uncertainties already men-
tioned it is important to note that Bayliss & Lunds-
gaard fii) perfused isolated kidneys with cyanide-
containing blood and found that some tubular
functions were reduced conspicuously, but that the
glomerular capillaries and membranes remained
nevertheless normally impermeable to protein in the
two instances tested. In an earlier study Starling &
Verney (348) found that the urine contained only a
trace of protein after 1 5 min of cyanide perfusion,
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
999
though stepwise increases of proteinuria occurred
after that. Pappenheimer found (unpublished studies)
in the perfused cat's leg by the isogravimetric method
that cyanide did not, under certain conditions,
increase either the filtration coefficient or the per-
meability of the capillary wall to protein. This raises
the interesting possibility that any effects which severe
hypoxia may have on capillary permeability do not
involve the better known cyanide-sensitive oxidations,
but involve rather the 2 to 50 per cent (71, 224) of
tissue oxygen consumption which cyanide does not
inhibit even in high concentration. It is possible, too,
that the edema of prolonged ischemia arises from the
effects of anaerobically produced metabolic products
or from other substances liberated by hypoxic tissue
cells. In summary, the effects on capillary permeability
of arrested blood flow, and more specifically of
hypoxia, are still uncertain and require more carelul
studies both as to quantitative aspects and as to
mechanism.
ADRENAL CORTICAL HORMONES AND FILTRATION
coefficients. Adrenal hormones have frequently
been considered to be a factor in maintaining filtra-
tion coefficients and capillary permeability within
normal limits even after injury. Menkin (245) and
others (110, 332) observed that adrenal extracts and
some adrenal steroids inhibited or delayed the ap-
pearance of intravenously administered trypan blue
in the skin of rabbits where leukotaxine (110, 245)
or peptone (332) had been injected locally. Some
blanching of the skin was observed in the area treated
with adrenal cortical hormones (1 10, 332), suggesting
possible vasoconstriction. Hyman & Chambers
(163) found in the perfused hind legs of frogs that
the rate of edema formation was reduced by addition
of certain adrenal cortical extracts to the perfusing
fluid, but their method, like the preceding ones, did
not exclude possible changes in capillary blood pres-
sure.
Renkin & Zaun (302) applied to this problem the
isogravimetric perfusion method of Pappenheimer &
Soto-Rivera (282) which permitted measurement of
a) filtration coefficients to indicate permeability to
fluid and protein, b) osmotic transients to indicate
permeability to small molecules, and c) blood flow
to identify vasoconstriction. Addition of adrenal
cortical extracts to both normal and adrenalectomized
preparations produced vasoconstriction which was
shown to be clue to the presence of small amounts of
an easily oxidizable substance, presumably epi-
nephrine. Limbs from adrenalectomized animals
showed no increase of capillary permeability to
protein and filtration coefficients for fluid did not
differ significantly from normal. Addition of adrenal
cortical extracts to the perfusing fluid produced
slight decreases of the filtration coefficient, but evi-
dence of vasoconstriction was also present. The ad-
dition of epinephrine to the perfusing fluid in amounts
similar to that contaminating the extract produced
corresponding changes, both of resistance to flow and
of filtration coefficient. Permeability of the capillary
wall to sucrose was also normal in adrenalectomized
rats and was not affected by aqueous adrenal cortical
extract.
The lack of agreement concerning the effects of
adrenal extracts and steroids emphasizes again (208)
the necessity for devising methods which separate
a direct action of substances on capillary permeability
per se from the indirect effects of complicating
vasodilatation or vasoconstriction. Vasodilatation
can increase capillary blood pressure; thus favoring
greater filtration through an unchanged capillary
wall. Conversely, vasoconstriction can reduce capil-
lary blood flow and pressure, and also the area of
capillary wall available for diffusion and filtration.
Such hemodynamic changes will, of themselves,
modify exchanges of substances, and thus simulate a
change of permeability. The quantitative measure-
ment of increases or decreases of capillary per-
meability is still one of the most difficult problems in
physiology.
POROSITY OF THE INJURED CAPILLARY WALL. The
effect of injury on the size of possible pores in the
capillary wall was considered by Krogh (184) in
1922. From the passage of soluble starch and the
retention of carbon particles he concluded that
pore diameter was not less than 50 A nor more than
2000 A. For normal limb capillaries of cats the
present corrected estimate for mean effective pore
diameter lies between 80 and 90 A (see sections 8 to
10). If it be assumed a) that injury merely enlarges
existing pores, and b) that Poiseuille's equation holds
for filtration through these pores, then a sevenfold
increase of filtration coefficient can be explained by a
65 per cent increase of pore diameter, i.e., from the
normal 80 to 90 A up to between 130 and 150 A.
Judging from the effects of reduced pH shown in
table 6.4 a doubling of filtration coefficient does not
increase protein passage measurably, whereas a
threefold to fourfold increase does. It may be, how-
ever, that injury increases the size or number of the
larger openings postulated by Grotte (126) and by
IOOO
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Mayerson et al. (232) or that ultramicroscopic dis-
ruption of vessel architecture produces new and still
larger apertures. Electron microscopy has shown
recently, for instance, that histamine can produce
separation of endothelial cells in venules so that
carbon particles pass between endothelial cells to
rest against the basement membrane (226, 226a). Ad-
ditional evidence in favor of enlarged leaks or of
new openings in injury has been provided recently
by Courtice & Morris (50-52, 54). Concentrations
of total cholesterol and of phospholipids were studied
in order to determine the plasma to lymph gradients
of lipoproteins in the limbs of cats and rabbits before
and after injury. These gradients were compared
with those for albumin and globulins. In lymph
from normal legs the concentrations of albumin,
globulins, cholesterol, and phospholipids were,
respectively, 48, 35, 24, and 33 per cent of the plasma
levels. After thermal injury the corresponding figures
for lymph were 81, 74, 60, and 74 per cent of the
plasma levels. Increased permeability to protein was
accompanied by increased permeability to lipopro-
teins, the diameters of which have been placed
tentatively at 150 to 350 A (144). Larger fat particles,
measuring perhaps 1500 A or more, e.g., the particles
in chyle or in an artificial fat emulsion, were not
transferred through the capillary wall to lymph to
any measurable extent, even after injury. However, as
Courtice mentions (52), although the passage of
lipoproteins and lipids becomes less as the size of the
molecule or complex increases, the exact mechanism
of their passage, whether between or through endo-
thelial cells, is still obscure. In addition, the molecular
mechanism and ultramicroscopic location of capillary
damage may well differ, depending upon the type of
injurious agent involved (208). Hence the basic nature
of Cohnheim's "molecular alteration in the vessel
walls" in various types of injury remains still a prime
unknown requiring study by pathologists, physi-
ologists, and electron microscopists alike.
7. DIFFUSION, GENERAL PRINCIPLES
The extravascular circulation caused by capillary
filtration and absorption is exceedingly important for
homeostasis of blood volume and for removal of large
protein molecules via the lymphatics. However, the
magnitude of the extravascular circulation is too small
to be of significance for the metabolic exchange of
small molecules between blood and tissues (see fig.
5.2). Metabolic exchange takes place largely by
diffusion processes which are almost independent of
the magnitude and direction of net fluid movement.
Evidence to be discussed below indicates that diffusion
of lipid-insoluble molecules takes place through
aqueous channels between capillary endothelial cells.
Lipid-soluble molecules, on the other hand, diffuse
rapidly through the lipid plasma membranes of the
endothelial cells themselves and are thus free to
utilize the entire capillary surface area for the ex-
change process. Before undertaking a detailed analysis
of diffusion processes in the capillary circulation, it
will be helpful to review some physical laws governing
molecular diffusion in free solution and in simple
membranes.
A. Free Diffusion
The fundamental laws of free diffusion were first
described by Fick (96) in 1855.
Adolf Fick (1829-igoi) was Professor of Physiology in
Wiirzburg. His most numerous publications were in the field
of muscle physiology, but his several classical contributions to
science were in the form of short, single publications in unre-
lated fields. Among circulatory physiologists he is known chiefly
as the originator of the "Fick principle" for determination of
cardiac output. Among ophthalmologists he is noted for the
development of tonometry and as author of "Fick's law"'
relating deformation of the cornea to intraocular pressure.
It is probable, however, that his greatest contribution to science
was his clear formulation of the laws of diffusion based on
analogy with Fourier's description of the flow of heat. "Die
Verbreitung eines gelosten Korpers in Losungsmittel geht,
wofern sie ungestort unter dem ausschliesslichen Einfluss
der Molecularkrafte stattfindet, nach demselben Gesetze
vorsich, welches Fourier fur die Verbreitung der Warme in
einem Leiter aufgestellt hat. . . Man darf nur in dem
Fourier'schen Gesetz das Wort Warmquantitat mit dem Worte
Quantitat des gelosten Korpers, und das Wort Temperature
mit Losungsdichtigkeit vertauschen."
According to Fick's formulation, the rate of linear
diffusion (quantity, n, per unit time, /) in direction .v
and through cross-sectional area, A, is proportional
to the concentration gradient, dc dx.
dn/dt -- D A dc/dx (7. 1)
The constant of proportionality, D, is known as the
diffusion coefficient and its dimensions are Pr1. The
simplest possible application of equation 7.1 is to
steady-state diffusion where dc/dt is constant as a
function of both distance and time. In this case, the
equation 7.1 can be written
n -" DA Ac/ Ax (7. 2)
where the concentration gradient Ae A.v is constant
all along the diffusion path. Equation 7.2 is specially
applicable to diffusion through thin membranes
where the concentrations on the two sides of the
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
membrane can be maintained constant. This con-
dition is frequently encountered in the capillary
circulation where blood on the luminal side of the
capillary membrane is maintained at constant
composition by virtue of an adequate blood flow and
fluid on the tissue side of the capillary membrane is
maintained at a different constant composition as a
result of tissue metabolism. A specific example will
serve to illustrate the use of ecjuation 7.2 and at the
same time indicate the magnitude of diffusion in
systems of capillary dimensions. Consider the diffusion
of glucose across an aqueous boundary .5 /i (0.5 X
io-4 cm) thick and with a surface area of 10 cm-. The
diffusion coeflicient of glucose is 0-9 X io~~5 cm2 per
sec (see table 9.1). Let the concentration of glucose
on one side of the boundary be maintained constant
at 100 mg per cent and the concentration on the
other side of the boundary at 99 mg per cent, thus
producing a constant concentration difference of 0.01
mg per cm3. Substituting these values in equation
7.2, we have
cm2 .01 mg/cm3
n = .9 X 1 o-5 X 10 cm2 X
0.5X1 cr4 cm
= 0.018 mg/sec or 1.08 mg/min
This rate of transfer is greater than the normal
metabolic consumption of glucose in 100 g of skeletal
muscle containing more than 5,000 cm- of total
capillary surface and it is thus obvious that even a
small concentration difference operating over a
relatively small aqueous area will provide a physio-
logically sufficient diffusion flow of glucose through
distances comparable in thickness with the capillary
wall.
Fick realized that the driving force for diffusion
results from random kinetic motions of the diffusing
molecules, but he did not perceive the physical sig-
nificance of the diffusion coefficient. It remained for
Nernst (1888) to relate diffusion coefficient to osmotic
and frictional forces in solution (260). Nernst showed
that
D ' RT/fN
(7.3)
where / is the frictional force opposing unit linear
velocity of each molecule and N is the number of
molecules per mole (Avogadro's number). For the
case of large spherical molecules the frictional force
opposing diffusion is given by Stokes' law describing
the motion of a sphere falling at unit velocity in a
viscous medium
f = 6irVa (7. 4)
where r\ is the viscosity of the medium and a is the
molecular radius. In 1905 Sutherland (355) and
Einstein (91) independently noted the possibility
of combining equations 7.3 and 7.4 to obtain the
relationship between free diffusion coefficient and
molecular radius
D = RT/6vya N
(7.5)
Equation 7.5 indicates that diffusion coefficient is
inversely related to molecular radius and to the
viscosity of the diffusion medium; conversely the
equation allows calculation of molecular radius from
measurements of free diffusion coefficient. It should
perhaps be emphasized that molecules are rarely
spherical and the molecular radius calculated from
the Einstein-Stokes relation (equation 7.5) is a
virtual quantity represented by a sphere of equivalent
diffusion coefficient. Moreover, the equation is
derived on the assumption that the diffusing molecules
are large compared to the solvent molecules; for
molecules smaller than glucose it is necessary to
apply corrections such as those given by Gierer &
W'irz (116). Additional methods for estimating
molecular dimensions include calculations from
density, intrinsic viscosity, and X-ray diffraction data.
Table 9.1, based on more detailed tables published in
references 82, 281, and 2g8, shows free diffusion
coefficients and approximate molecular radii of a
variety of molecular species which have been used
in studies of capillary permeability.
B. Diffusion Through Porous Membranes,
Restricted Diffusion
The diffusion of small molecules through thin,
large-pored membranes takes place according to
Fick's law; the only effect of the membrane is to
reduce the total area available for free diffusion.
Indeed, the most accurate method of estimating the
pore area, Ap, in a membrane with large water-filled
pores is to measure the diffusion rate, ii, through the
membrane of small, uncharged molecules of known
free diffusion coeflicient. From rearrangement of
Fick's law
n x
Ax
DAc
In most practical applications the path length, A\,
through the membrane is also unknown and it is more
useful to solve for the pore area per unit path length,
Ap/Ax
1002
HANDBOOK OF I'HYSIOI.OUY
CIRCULATION II
P
Ax
n
D~Ac
(7.6)
Once the pore area per unit path length, Ap Ax, has
been established from equation 7.6 for a given large-
pored membrane, the membrane may be used to
determine free diffusion coefficients of test molecules
(233, 264, 316). Membranes employed for this
purpose generally have pores which are at least
100-fold larger than the diffusing molecules.
In the case of diffusion through membranes having
pores of molecular dimensions the kinetic motions of
the diffusing molecules are restricted by the pore
structure; in such membranes the effective pore area
per unit path length decreases as a function of mo-
lecular size, becoming zero when the test molecules are
the same size as the pores. Capillary permeability to
lipid-insoluble molecules of graded sizes can be
explained, in large part, by restricted diffusion
through aqueous channels of molecular dimension.
For this reason it will be necessary to discuss physical
aspects of restricted diffusion in some detail.
Figure 7.1 shows apparent pore areas per unit
path length for molecules of graded sizes diffusing
through a cellulose membrane of the type commonly
used for ultrafiltration or dialysis (Visking sausage
casing). It is evident that the apparent pore area for
free diffusion decreases rapidly as a function of
molecular size. The true pore area in the membrane
is, of course, constant and it is useful to think of the
apparent decrease in terms of a restricted diffusion
coefficient, D', such that
D * D A ,/A.
(7.7)
where As is the apparent pore area for the solute and
Av is the true pore area. Substitution of D' for D in
equation 7.6 would yield the true membrane pore
area per unit path length for all molecular species.
The essential theoretical problem is now to relate
the observed restriction to diffusion, D' D, to di-
mensions of the membrane pores.
The theory of restricted diffusion proposed by
Pappenheimer el al. (281) takes into account two
factors impeding the passage of molecules through
pores of molecular dimensions. The first factor is
concerned with steric hindrance at the entrance of
the pore. It is assumed that for entrance into a pore
a molecule must pass through the opening without
striking the edge as originally suggested by Ferry
(95). For the case of cylindrical pores the effective
target area, As, for the solute is then
Ap (l-a/r)*
As dn/dt
Ax"~ DAC
VISKING CELLULOSE
Mean pore radius
r = 16 S
-V—if
MOLECULAR RADIUS, A
* 5 6
H3HO
—If-,
GLUCOSE
SUCROSE
ANTIPYRINE
RAFFINOSE
fig. 7.1. Apparent pore areas per unit path length as a
function of molecular size. The smooth curve is constructed
from the theory of restricted diffusion, equation 7.9, assuming a
mean pore radius of 16 A. Mean pore radius determined on
the same membrane from combination of diffusion and filtra-
tion was 19 A (equation 7.13). Similar data for diffusion of
lipid insoluble molecules through the walls of muscle capillaries
are shown in figure 9.2. [Adapted from Renkin (298).]
where A v is the true geometrical area of the opening
and a r is the ratio of molecular radius to pore
radius.
The second factor takes account of friction between
a molecule moving within a pore and the stationary
walls of the pore. This factor, first studied by Laden-
burg (193), was employed by Friedman & Kraemer
(112) to describe the diffusion of sugars through
gelatin gels. The Ladenburg treatment of the problem
is strictly applicable only to cases where a r < .1
and it is preferable to use the more general formula-
tion of Faxen (94)
1-2.10 (£) * 2 09(f) -0.95(f)
(78)
where / ,'/o is the frictional resistance to diffusion in
the pore relative to that in free solution. Taking into
account both steric hindrance (equation 7.7a) and
wall effects (equation 7.8), the theoretical restriction
to diffusion through cylindrical pores becomes
£ - |- (I- °7f[-zJ0(f). 2.09(?)
4fl
-0.95
(7 9)
(7.7a)
The last term of the series is negligible when a/r <
0.5. During net flow through the membrane the
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IOO3
velocity of flow at the center of the pore is twice the
average velocity and the effective target area pre-
sented by the pore to the incoming molecules is
slightly increased (95). Under these conditions the
restricted diffusion equation becomes
(\):[2('-arf-(l-aT)]h'0(f)
^2.09(ff -0.95(ff]
P f/Ou '
(7. 10)
Figure 7.2 shows that equation 7.9 describes
observed diffusion through artificial porous
membranes with considerable accuracy. The data
were obtained using seven molecular species and
three membranes having porosities in the range of
interest for capillary physiology. Analogous develop-
ment of theory for restriction to diffusion through
rectangular slits, rather than cylindrical pores, leads
to a theoretical curve closely approximating that
shown in figure 7.2 (281). However, electron micro-
graphs indicate that true pore geometry of artificial
membranes is closer to the cylindrical than to the
rectangular model (27).
Study of figure 7.2 reveals that pores of sufficient
size to allow the slow penetration of plasma proteins
(i.e., 30-40 A) will nevertheless impose differential
restriction to diffusion of much smaller molecules.
Thus diffusion of glucose (a = 3.7 A) through pores
of radius 40 A will be slowed by 34 per cent, whereas
diffusion of water through the same pores will be
slowed by only 14 per cent. This differential restric-
tion to diffusion of small solutes and water is the
%
,tf/D-o-f)8['-«-«>(f)+"»e-f-ft*(?f]
a
Sylvonia
Viscose Wet Gel, r =
77A
O
Dupont
Cellophane, r = 31 A
X
Visking
Cellulose, r = 16 A
fic. 7.2. Restricted diffusion through artificial porous
membranes of various pore sizes. The smooth curve is drawn
from the theory of restricted diffusion, equation 7.9. [Adapted
from Renkin (298).]
essential factor underlying transcapillary fluid shifts
caused by transient changes in the concentration of
small molecules in either plasma or tissue fluids.
The theory of restricted diffusion provides a
method for estimation of effective pore radius, both
in artificial membranes and in living capillaries. Thus
equation 7.9 contains only two unknowns, D' and r,
and it is therefore possible to solve for r from observed
diffusion rates of two molecular species of known free
diffusion coefficients and molecular radii. Greater
accuracy can be obtained from the best fit of equation
7.9 to results obtained from several molecular species
as shown in figure 7.1. Pore dimensions calculated
from the theory of restricted diffusion agree well with
values obtained by independent methods (298).
C. Diffusion and Hydrodynamic Flow,
Relation to Pore Dimensions
Hydrostatic or osmotic forces, acting across a
porous membrane, cause net fluid movement in
proportion to the difference between hydrostatic
and effective osmotic pressure (equation 1.1). Two
different mechanisms are involved, diffusion and
hydrodynamic flow.
diffusion. The effective concentration (thermo-
dynamic activity) of water depends upon pressure,
temperature, and solute concentration. An increase
of pressure or temperature increases the kinetic
energy of the water molecules and therefore increases
the statistical probability of net movement toward a
region of lower pressure or temperature. Conversely,
the addition of solute molecules to water decreases
the probability of net diffusion of water to a region of
lower solute concentration. For an ideal semiperme-
able membrane Fick's law may be restated as follows
to take account of these variables
dn
dt
V ho Ax
h2o *
/AP-A/7)
1 RT '
(7. II)
where q is rate of net water flow (ml/sec) and FH2o
is the partial molal volume of water (18 cm3/mole).
The term (A j> — Ml)/R T replaces the concentration
term in Fick's law and represents the difference in
activity of water molecules on the two sides of the
membrane. Formal derivations of equation 7.1 1 may
be found in references (38) or (170).
hydrodynamic flow. The minimum dissipation of
energy for net water flow through a membrane
1004
HANDBOOK OF PHYSIOLOGY-
CIRCULATION II
containing A' cylindrical pores of radius r will occur
if the velocity profile assumes the parabolic distribu-
tion of Poiseuille's law.
^(ar-ah)
i.o
8yA>
" A-x-8-JAp-AJI)
(7 12)
Equation 7.12 implies that hydrodynamic stream-
ing occurs even when the hydrostatic pressure differ-
ence across the membrane is zero, i.e., during flow
caused by purely osmotic forces. The question is often
raised as to how Poiseuille flow could occur in the
apparent absence of a difference in hydrostatic
pressure. An explanation of this apparent paradox
was first offered in an important paper by Schlogl
(317) who pointed out that a hydrostatic pressure
drop accounting for hydrodynamic flow does indeed
exist along most of the length of the membrane
pores, even though the hydrostatic pressures on the
two sides of the membrane are equal. The intra-
membrane hydrostatic pressure gradient reverses
sharply near the edge of the pore where it becomes
equal and opposite to the steep gradient of diffusion
potential. A more detailed treatment of this hypoth-
esis will be found in the recent paper by Ray (292).
Comparison of equation 7.1 1 with equation 7.12
reveals that for a given total pore area, Ap, net flow
by diffusion is independent of pore radius, whereas
hydrodynamic flow varies with the second power of
the pore radius. It follows that for a given difference
in hydrostatic pressure the hydrodynamic component
of flow will increase rapidly as a function of pore
size. Figure 7.3 shows the relative importance of
diffusion and hydrodynamic flow as a function of
pore radius. For porosities in the range of interest for
capillary permeability (e.g., 20-50 A) the hydro-
dynamic component of net flow is overwhelmingly
greater than the diffusion component. The capillary
filtration coefficient discussed in section 6 is therefore
a measure of hydrodynamic conductivity rather than
diffusion permeability. Detailed discussions of the
relations between diffusion permeability and hydro-
dynamic conductivity will be found in papers by
Koefoed- Johnson & Ussing (177), Pappenheimer
(276, 277), Garby (1 13), Durbin et al. (83) Kedem &
Katchalsky (170) Katchalsky (169), Mauro (230),
and Ray (292). A recent experimental evaluation of
diffusion and hydrodynamic flow through artificial
membranes has been published by Robbins &
Mauro (303).
Combination of diffusion data with hydrodynamic
o
< .6
01 ?
1
HYDRODYNAMIC FLOW
\ 1
-V
- h
_/' \
NET DIFFUSION
/ \
/
-jC___ PORE RADIUS
20
-N-
30
40
50, A
>
REGION of uncertainty re-
STRICTIOM TOHrORODrNANIC
flOW AND DIFFUSION UNKNOWN
1MIS RANGE OF INTEREST FOR
CELLULAR PERMEABILITY
RECiil IN WHICH EVIDENCE
SUPPORTS VALIDITY OF
POISEUILLE FLOW THIS
NANCE OF INTEREST FOR
CAPILLARY PERMEABILITY
fig. 7.3. Net diffusion and hydrodynamic flow of water as
a function of pore size during flow induced by hydrostatic or
osmotic forces. For membranes with effective pore radii greater
than about 20 A the net flow of water by diffusion is negligible
compared to hydrodynamic flow. [From Pappenheimer (276).]
data leads to a solution for pore dimensions. Equation
7.11 describing hydrodynamic or osmotic flow through
cylindrical pores can be rearranged to give
A /Ax
But Q/(Ap — All) is the filtration coefficient defined
by equation 1.1 and {Au,/Ax) can be determined from
diffusion of labeled water as shown in figure 7.1.
Therefore, the effective pore radius is defined by
measurable quantities
8yKf
A. /Ax
(7 13)
Similar equations, based on diffusion and hydro-
dynamic flow, can be derived to estimate the dimen-
sions of slit pores (18, 281) or any other pore geom-
etry for which the laws of hydrodynamic flow are
known.
Equation 7.13 has been used to estimate pore size
in artificial membranes (82, 298) and in living
capillaries (281). In general there is good agreement
between pore size estimated from flow and diffusion
and pore size estimated from restricted diffusion
(fig. 7.1).
D. Simultaneous Flow and Restricted Diffusion:
Theory of Molecular Sieving
In the capillary circulation both filtration and
restricted diffusion usually occur simultaneously and
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IOO5
these two processes, operating together, are largely
responsible for observed concentrations of large
molecules in lymph (126) and renal glomerular
nitrate (194, 278, 366). From the theory of
molecular sieving described below it is possible to
make deductions concerning capillary permeability
for comparison with results obtained by independent
methods. The degree of molecular sieving of a given
solute may be defined as the ratio of its concentration
in the filtrate (r2) to its concentration in the filtrand
(ci). It is often supposed that during ultrafiltra-
tion of a monodisperse solute through an isoporous
membrane, the value of ( 2/< 1 will be zero when the
pores are smaller than the solute molecules and unity
when the pores are larger than the solute molecules.
If intermediate values are actually observed they are
said to be evidence for heteroporosity. If, for example,
the concentration of a given solute in a capillary
ultrafiltrate is 50 per cent of that in plasma it is
supposed that half the capillary pores were smaller
than the solute molecules and half were larger (21,
208, 232, 254). This reasoning fails to explain the
dependence of molecular sieving on filtration rate.
The following considerations show that molecular
sieving of a monodisperse solute through an isoporous
membrane is determined by the ratio of restricted
diffusion to rate of filtration.
If the passage of solute through a porous membrane
is restricted relative to passage of solvent, then the
filtration, c2 approaches c-y (dialysis) and at high rates
of filtration c, c, approaches the ratio of restricted
pore areas Aa/Aw. Equation 7.14 is derived on the as-
sumption that the concentration gradient through the
membrane is linear. Grotte (126), Garby (113), and
Kuhn (192) have pointed out that the concentration
gradient in the membrane will in general be an ex-
ponential function of flow velocity, but this correction
was shown by Grotte to be a small one and will be
neglected here.
The restricted pore areas, Aw and As, have been
expressed by equation 7.10 as a function of molecular
radius and pore radius. Substitution of this function
in equation 7.14 yields a cumbersome but explicit
expression for molecular sieving as a function of
filtration rate when molecular and membrane pore
dimensions are known; conversely, it provides an
independent method for calculation of pore size from
experimental measurements of molecular sieving and
filtration rate.
Figure 7.4 shows experimentally determined values
of molecular sieving as a function of filtration rate
through Visking dialysis membrane. The theoretical
curves were drawn according to equation 7.15, using
the value of Aw/Ax determined from diffusion of
tritiated water and choosing pore radii to provide the
best fits to the experimental data for each molecular
species. Satisfactory fits were obtained with pore radii
15 to 17 A. Pore radius for the same membrane
estimated irom the theory of restricted diffusion
C2
I +
Ax
/*{<-&'-{'-?/ % ][<-2'0(^) + 2.09(±-f-0.95(^)5] +°<
(7. 15)
filtrate will be diluted during filtration, thus giving
rise to a concentration difference for diffusion at a
rate determined by the restricted diffusion coefficient,
D' , through the membrane. The ultimate steady-
state composition of the filtrate relative to filtrand
(ci/ci) is therefore determined by a race between
hydrodynamic flow (Qj) tending to dilute the filtrate
and restricted diffusion tending to restore the concen-
tration difference. A quantitative expression for
molecular sieving through isoporous membranes
was derived by Pappenheimer (276)
C?
I +
J-= Of Ax
*.+$.
(7. 14)
Ax
Inspection of equation 7.14 shows that at low rates of
Ax
(fig. 7.1) was 16 A, and pore radius estimated from
combination with Poiseuille's law was ig A (equation
7.13). The internal consistency of these various
estimates of pore radius in artificial membranes
constitutes the chief evidence justifying the application
of similar techniques to biological membranes.
E. Distribution of Pore Sizes
Observed values for diffusion and molecular sieving
through artificial porous membranes are in reasonable
accord with theoretical predictions for isoporous
membranes. Equal or slightly better agreement
between experiment and theory can be obtained by
assuming certain limited distributions of pore sizes
(298). An upper limit to pore size may be determined
ioo6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
(a)VISKING SAUSAGE CASING
.a.
I5A UREA
I5A GLUCOSE
SUCROSE
r-RAFFINOSE
FILTRATION RATE Cm ^ec x 10*
0.2
0.4
0.6
0.8
fig. 7.4. Molecular sieving through an artificial, porous
membrane. The smooth curves are constructed from the theory
of molecular sieving, equation 7.15. The data fit pore radii in
the range 15-17 A. Mean pore radius for the same membrane
estimated from the theory of restricted diffusion was 16 A
and pore radius estimated from combination with Poiseuille's
law was 19 A. The internal consistency of these various esti-
mates of pore radius in artificial membranes constitutes the
chief evidence justifying the application of similar techniques
to biological membranes of comparable pore size. [Adapted
from Renkin (298).]
from the size of the largest molecule which just fails
to pass through the membrane. A less obvious limit
to any assumed distribution arises from the fact that
filtration rate varies with the fourth power of the
radius so that the total fraction of large pores must
be limited in order to satisfy the requirements im-
posed by the observed filtration coefficient. If a
membrane contains cylindrical pores of different
radii then the mean equivalent pore radius, r, for
hydrodynamic flow is given by
7'Vfc+tf
F„r;
(7. 16)
where Fn is the fraction of total pore population
having a radius rn. For example, the membrane
illustrated in figures 7.1 and 7.4 did not allow the
passage of hemoglobin (a 32 A) and therefore an
upper limit to its pore size distribution is 32 A. How-
ever, less than 20 per cent of the pores could be as
large as 30 A otherwise
7>yh.2x (30)4 >20A
which would not fit the requirement that r = 19 A
set by the observed filtration coefficient and diffusion
area for water (equation 7.13). The detailed com-
putation of possible pore distributions which would
fit the data for filtration, restricted diffusion and
molecular sieving is possible but laborious. For the
membrane illustrated in figures 7.1 and 7.4 the
broadest Gaussian distribution of pore radii com-
patible with the data is defined by a mean pore radius
of 14 A with a standard deviation of 7 A (298).
F. Osmotic Pressure* and Osmotic Flow Through
Leaky Membranes; Osmotic Reflection Coefficients
Van't Hoff's law relating osmotic pressure to
concentration was derived for a perfectly semi-
permeable membrane. Relatively little is known of
osmotic forces associated with diffusion and osmotic
flow through membranes which restrict, but do not
prevent entirely, the diffusion of solute molecules.
The quantitative significance of this problem may be
illustrated by a specific example. Consider a two-
compartment system separated by a membrane
containing pores of radius 30 A. Addition to one
compartment of an ideal solute of molecular radius
30 A will cause osmotic flow through the membrane
at a rate equal to that caused by a hydrostatic pressure
difference of cRT mm Hg (equation 7.12). However,
if the same molar concentration of a small molecule
such as urea (molecular radius 2.7 A) is added to one
compartment, it will be found that the osmotic flow-
is less than 5 per cent of that obtained by the hydro-
static equivalent (82).
In 1 951 Staverman (349, 350) introduced the
expression "osmotic reflection coefficient," a, as an
empirical descriptive term modifying van't Hoff's
law for the case of leaky membranes.
n -" CRT* (7. 17)
The value of a ranges from unity in perfectly semi-
permeable membranes to less than zero when the
mobility of the solute exceeds that of the solvent
(333)-
Very small values of a have been reported for
osmotic flow caused by small molecules diffusing
3 "Osmotic pressure" is ordinarily defined for the case of
thermodynamic equilibrium across ideal semipermeable
membranes and the term has no equivalent meaning for the
irreversible process to be considered here. Possibly a different
term should be coined to describe the transmembrane pressures
arising during restricted diffusion through porous membranes.
"Restricted diffusion pressure" would be accurate but could
only be applied to the case of zero net flow of solvent through
the membrane.
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IOO7
through large pored artificial membranes. Thus
Meschia & Setnikar (250) found that less than 2 per
cent of the ideal osmotic potential was developed by
sucrose during osmotic flow through a collodion
membrane having pores of radius no A (i.e., when
the radius of the pore was approximately 25-fold
greater than the radius of the diffusing molecule).
Similarly low values for osmotic reflection coefficient
have been reported by Grim (125) and by Shuler
et al. (333) on the basis of osmotic flow through
uncalibrated membranes.
Durbin (82) has recently completed a study of
osmotic flow caused by molecules of graded sizes
diffusing through calibrated porous membranes. His
results shows that
-('■%.)
where As is the restricted pore area available to the
solute and Au. is the restricted pore area available to
the solvent as defined by equation 7.9 and figure 7.1.
Durbin's results indicate that during osmotic flow-
through artificial membranes the value of a is less
than o. 1 when the radius of the diffusing molecule is
10 per cent of the radius of the pore.
From existing data one must therefore conclude
that only a small fraction of the theoretical van't
Hoff pressure is operative across artificial membranes
during restricted diffusion of small molecules through
the membranes. On the other hand, relatively large
osmotic forces have been observed during the re-
stricted diffusion of small molecules through biological
membranes under conditions involving little or no
net fluid movement (122, 281). Under these con-
ditions the osmotic reflection coefficient appears to
depend upon the restricted diffusion coefficient of
solute relative to that of the solvent.
uw » w
Combination of equation 7.18 with equation 7.9
allows numerical evaluation of a for the case of zero
net fluid movement when molecular radius and pore
radius are known. Thus,
rapidly in the virtual absence of net fluid movement
(281).
In experiments involving artificial membranes it is
exceedingly difficult to measure osmotic forces in the
absence of osmotic flow, as pointed out in section 3.
For this reason all investigations of osmotic reflection
coefficient in artificial systems have thus far invoked
net flow of fluid. Under these conditions the frictional
forces determining osmotic reflection coefficient will
contain hydrodynamic as well as diffusional terms as
emphasized in recent derivations by Ray (292) and
Katchalsky (169). Discussion of these derivations is
beyond the scope of this chapter but it seems fair to
say that no well-substantiated theory is yet available
to predict osmotic reflection coefficients as a function
of membrane permeability and flow rate. Since most
biological membranes allow the restricted passage of
environmental solutes, the problem remains as one of
the most important unsolved questions in contem-
porary studies of permeability.
8. TRANSCAPILLARV MOVEMENT OF
LIPID-INSOLUBLE MOLECULES
The concentration gradients which provide the
driving force for diffusion exchange between blood
and tissues are normally maintained by tissue me-
tabolism. However, the transcapillary exchange
process is so efficient that normal transcapillary
concentration differences of small molecules would be
too small to be detectable by existing methods, even
supposing it were feasible to collect, for analysis,
tissue fluid from the immediate vicinity of the capillary
wall. From an experimental point of view it is there-
fore necessary to establish abnormally large trans-
capillary concentration ratios in order to study
diffusion characteristics of the capillary walls.
Figure 8.1 summarizes data showing rates of
disappearance from the circulatory system of various
lipid-insoluble substances which distribute primarily
in extracellular fluid. It is evident that these sub-
stances leave the vascular system at rates which vary
IT.* I-
D'/0- &f*0&+ 2.09&-0.95&] I
(7. 19)
Equation 7.19 is specially applicable to capillary
membranes where osmotic forces can be measured
inversely with molecular size. Disappearance from
plasma is accompanied by simultaneous appearance
ioo8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
SERUM ALBUMIN
MW 69,000
fig. 8.1. Disappearance of substances from arterial plasma
of rabbits. Data for albumin are from Gitlin et al. (118), data
for inulin and sucrose are from Kruhoffer (190, 191), and data
for NaCl are from Morel (257). C is concentration in arterial
plasma at time /, C,„ is concentration in plasma at equilibrium;
C0 is initial concentration in arterial plasma obtained by
extrapolation to zero time.
of the test molecules in tissue spaces, again at rates
which vary inversely with molecular size. Most
tissues of the body participate in the distribution
process, though at widely different rates.
The rapid penetration of capillary walls by mole-
cules as large as sucrose or inulin implies a high order
of permeability to lipid-insoluble molecules. Cell
membranes generally (i.e., the plasma membranes
which envelop the protoplasm of all living cells) are
virtually impermeable to metabolically inert, lipid-
insoluble molecules as large as sucrose; indeed they
generally have a low order of permeability to most
ions. The behavior illustrated in figure 8.1 is more
characteristic of artificial porous membranes and this
resemblance has given rise to the hypothesis that
capillarv blood communicates directly with extra-
vascular fluid via a system of aqueous pores or
channels. Recent studies of capillary ultrastructure
(93) support earlier views (37, 276) that the structural
basis for this type of permeability is associated with
junctional regions between capillary endothelial cells.
The number, dimensions, and properties of trans-
capillary pores which would be necessary to account
for observed capillary permeability to lipid-insoluble
molecules will be considered more fully in section g.
Initial experiments with isotopic tracers led to the
suggestion that arterial disappearance curves might
provide a quantitative measure of capillary perme-
ability (106, 138). For this purpose it was assumed
that diffusion from blood to extravascular space
could be represented by a simple two-compartment
diffusion system separated by the capillary membranes
and that concentrations in each compartment would
be uniform at each moment during the diffusion
process. The theoretical equation describing diffusion
in this simplified model is easily derived from Fick's
law and leads to the expression
X-
DAS (Vi + Vg)
Ax
V V
V 2
(8.1)
where X is the (negative) slope of the observed ex-
ponential disappearance curve, V\ is plasma volume
and Fo is extravascular distribution volume (106, 233).
Reference to equation 7.2 shows that the term DA Sx
is the flux per unit concentration difference, n A< ,
and is therefore a direct measure of capillary perme-
ability to the test molecules. More complex equations
describing arterial disappearance curves have been
derived to take account of loss by the kidneys, loss
by metabolism and distribution between more than
two compartments in series or in parallel (318, 319,
33°. 34'- 363)-
Equation 8.1 is specially applicable to the case of
large, lipid-insoluble molecules such as proteins or
synthetic polymers. Diffusion of such substances
from the vascular system is so slow that their concen-
trations in arterial plasma may be taken as a close
approximation of mean concentration in capillary
plasma, i.e., the concentration gradient along the
length of each capillary is negligible at all times during
the diffusion process. In the example of figure 8.1 the
slope, \, for albumin is about — 0.1 per cent of the
initial concentration per min, or 100 per cent of the
plasma albumin every 16.6 hours. Similar values for
transcapillary exchange rates of serum albumin have
been observed in dog (370) and man (352). The free
diffusion coefficient of albumin is 0.085 X io-5 cm2
per sec. Given a normal plasma volume, V\t of 4 per
cent of body weight and a normal extracellular fluid
volume, r2, of 20 per cent then from equation 8.1
the effective capillary pore area per unit path length
available for restricted diffusion of serum albumin is
65 cm per 100 g tissue. This value may be compared
with the value of 70 cm per 100 g muscle calculated
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IOOg
from the theory of restricted diffusion through pores
of radius 43 A (table 9.1).
Arterial disappearance curves therefore provide a
method for quantitative studies of overall capillary
permeability to large molecules. It is not possible,
how ever, to extend this type of analysis to molecules
which diffuse rapidly through capillary walls. In
this case the mean concentration in capillary plasma
may be only a small fraction of that in arterial plasma,
particularly in early phases of the distribution process.
Application of equation 8. 1 to such data leads to
estimates of capillary permeability which are too low,
often by 1 or 2 orders of magnitude [see (281) and
(382) for critical review]. Factors which determine
mean concentration differences of small molecules
across capillary walls during diffusion include rate of
blood flow, diffusion rate, and the geometry and
volume of extravascular diffusion space. Several
interesting attempts have been made to take account
of these factors by mathematical techniques but the
solutions are complex and involve assumptions which
are difficult to evaluate experimentally (19, 319).
Specialized experimental methods for estimating
capillary permeability to small molecules were
developed by Pappenheimer et al. (281) for the study
of molecular exchanges in the capillary circulation of
hind limbs of cats or dogs. Results obtained by these
methods lead to conclusions of general interest
relating capillary permeability to the number and
dimensions of capillary pores which would be re-
quired to explain observed transcapillary diffusion
rates of lipid-insoluble molecules ranging in size from
D20 to hemoglobin.
9. STRUCTURE OF MUSCLE CAPILLARIES AS DEDUCED FROM
PERMEABILITY MEASUREMENTS AND FROM ELECTRON
MICROSCOPY'. QUANTITATIVE ASPECTS OF
TRANSCAPILLARY DIFFUSION
In isolated perfused tissues the rate of net trans-
capillary movement of test substances can be de-
termined from the product of blood flow and arterio-
venous concentration difference. Thus,
(9.1)
where Q is blood for plasma) flow and ca, cv are the
simultaneously measured concentrations of the test
substance in arterial and venous bloods (or plasma).
The driving force for diffusion (i.e., the mean
concentration difference across the capillary walls)
may be estimated from the partial osmotic pressure
n= Qb(ca-cJ
0.6 x 10 cm
0.4
As RTo- h
38 « 10* cm
0.2
mq /sec
per IOOg muscle
0.15
0.10
NET FLUX,
n = Q(C0-CV)
0.05
PARTIAL OSMOTIC PRESSURE
DEVELOPED ACROSS
CAPILLARY MEMBRANES
10
15
20
JlL
30
1
MINUTES AFTER ADDITION OF RAFFINOSE
no. 9. i . Diffusion of raffinose from the capillaries of a
perfused cat hind limb. At zero time 20 mM/liter raffinose was
added to the perfusion reservoir. The final distribution volume
of raffinose in perfused tissue was 19% of limb volume. The
capillary diffusion area per unit path length calculated for
raffinose was 0.38 X io5 cm2; this value was independent of
time, extravascular fluid volume, or of mechanically induced
changes of blood flow. [Adapted from Pappenheimer et al.
(281).]
exerted by the test molecules during the diffusion
process. Figure 9. 1 illustrates a typical experiment
showing the simultaneous measurement of net flux
rate and partial osmotic pressure during the diffusion
of raffinose from the capillaries of a perfused cat
hind limb. The ratio of flux rate to partial osmotic
pressure is proportional to permeability and may be
related to the restricted pore area per unit path
length in the capillary wall by combining equations
7.6 and 7.1 7.
Ax
RTcr ft
Dt * AIT
(9.2)
ioio
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
\ Ax
xlO~5cm
RESTRICTED PORE AREA
- \
PER UNIT PATH LENGTH
IN lOOg MUSCLE
-\
iX GLUCOSE
\ NaCI
\UREA
As RTo- li
flx ~ "D7*flTT
\ X SUCROSE
\ X RAFFINOSE
Theoretical curves for
membrane with cylindrical
-
pores of rodios 40 -45 A
and some filtration coef-
ficient as capillaries
^X INULIN
^5^— _ SERUM
MYOGLOBIN X -^=^~-— _ ALBUMIN
_J 1 1 1 1 1- IX
0 5 10 15 20 25 30 35
MOLECULAR RADIUS, A
FIG. Q.2. Restricted diffusion of lipid-insoluble molecules
from the capillaries of perfused cat hind limbs. Each point
represents the mean value of data from several experiments.
The curves are constructed from the theory of restricted diffu-
sion and filtration (equation 9.3) on the assumption that the
osmotic reflection coefficient is determined by equation 7.19.
The data fit theoretical restricted diffusion through pores of
radius 40-45 A in a membrane having the same filtration
fcoefficient as the capillaries in the hind limb. [Recalculated
rom the data of Pappenheimer et al. (281).]
The upper panel of figure 9. 1 shows that for
rafnnose the restricted pore area per unit path
cm2 or less than 0.1 per cent of the total capillary
surface area in 100 g muscle. This conclusion is
consistent with the view that transcapillary ex-
changes of lipid-insoluble molecules take place at
junctional regions between endothelial cells and we
have already seen that pore areas of this magnitude
can provide a physiologically sufficient flow of small
molecules under the influence of small concentration
gradients (section 7 A).
In the original analysis of Pappenheimer et al. (281)
the mean pore radius was estimated from combina-
tion of the capillary filtration coefficient with the
pore area per unit path length for a molecule the
size of water (equation 7.13). However, the latter
quantity was uncorrected for the osmotic reflection
coefficient and therefore cannot be employed for the
present analysis in which the osmotic reflection
coefficient is included as an unknown. In order to
solve for this additional unknown it is necessary to
introduce an additional equation relating osmotic
reflection coefficient to pore dimensions as suggested
by equation 7.19. This equation is cumbersome and
its use may not be entirely justified on the basis of
our present inadequate knowledge of factors de-
termining osmotic reflection coefficients. Neverthe-
less, it leads to a solution for capillary pore dimensions
which is more consistent with available data than the
dimensions originally proposed by Pappenheimer
et al. (281). Substitution of equations 7.19 and 9.2 in
equation 7.13 yields
(,-^f[i-2,0(^)+2.09(^)3-0.95(±)5]
(,- f*ffziO&) + 2.09(^-)-0.95(^
8VK, An
D„[ RTD. * n
(9.3)
length, calculated from equation 9.2, was 0.38 ± .04
X io6 cm. Results of similar measurements, made
with a variety of molecular species, are shown in
figure 9.2. It is seen that in capillaries, as in artificial
porous membranes (fig. 7.1), the restricted pore area
decreased as a function of molecular radius as pre-
dicted from the theory of restricted diffusion (equa-
tion 7.9). Extrapolation to zero molecular radius
suggests that the true pore area per unit path length
in the capillaries of 100 g muscle is approximately
0.6 X io5 cm. Since the average thickness of the
capillary walls is less than io~4 cm (fig. 9.3), this
suggests that the total pore area available for diffusion
exchange of lipid-insoluble molecules is less than 6
where
j = mean capillary pore radius, A
aw = radius of water molecule = 1.5 A
a, = radius of test molecule, A
= free diffusion coefficient of water = 3.4 X io-5
cm2 per sec-1
= free diffusion coefficient of test molecule
= viscosity of water = 0.007 dyne-sec-cm-2 at 37 C
Kf = filtration coefficient of capillaries, average value 1.8 X
io-7 cm5-dyne~'-sec~l per 100 g
All = observed partial osmotic pressure, dynes-cm-2 at
time, /
n = observed flux rate, mole-cm3 at time, i
RT = 25 X io9 dyne-cm-mol-1 at 37 C
Study of equation 9.3 in relation to the experimental
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
101
600 A QftWW«i!*5*»&iw
fig. 9.3. Diagram illustrating current concepts of fine structure in muscle capillaries. [From
Fawcett (93).] The nucleus of a single endothelial cell is shown at left. The capillary cross section
may be formed by a single cell rolled into a tube or may be made up of several cells. The inter-
endothelial region appears as a thin slit pore with direct connection from inside to outside of capil-
lary. The slit may be straight or slightly tortuous (inset and fig. 9.4) and is usually about 90 A wide
and 0.5-1 n in length. It occupies only a fraction of 1 % of the total endothelial surface. The cyto-
plasm contains the usual organelles, but in addition contains numerous small vesicles and inpocket-
ings of the surface which are characteristic of microphagocytosis or pinocytosis. Palade (273) has
suggested that this mechanism may be involved in the transcapillary passage of particles which
are too large to traverse interendothelial openings. The outer surface of the endothelium is en-
veloped by an amorphous basement membrane about 600 A in thickness and with histochemical
properties indicative of a mucopolysaccharide. The permeability properties of this membrane are
unknown. Dilute solutions of mucopolysaccharides, in contrast to gelatin-gels (112), may offer
appreciable resistance to free diffusion (266). Finally, it should be emphasized that the diagram
refers only to muscle capillaries. Morphological differences between capillaries in different vascular
beds have been reviewed by Bennett it al. (13).
data of figure 9.2 reveals that the pore radius, r, is
the only unknown quantity. The numerical evalua-
tion of r from equation 9.3 may be carried out by
successive approximation or by graphical analysis.
For the specific example illustrated in figure 9. 1 the
value of r is 41 A. Similar analysis of data for NaCl,
urea, glucose, and sucrose leads to mean pore radii of
44, 43, 45, and 41 A, respectively. These results are
in accord with values in the range 35 to 45 A esti-
mated by Grotte (126) from molecular sieving of
dextrans in hind limb capillaries (section 10 and
fig. 10. 1). They are in contrast to the value of 30 A
estimated by Pappenheimer et al. (281) from combi-
nation of filtration coefficient with uncorrected
diffusion data and to the value of 25 A estimated by
Renkin & Pappenheimer (301) from uncorrected
restriction to diffusion. The dimensions and fractional
surface area of aqueous transcapillary pores cor-
respond closely with dimensions of interendothelial
junctions as determined by electron microscopy.
Figure 9.3 summarizes pertinent aspects of pore
structure in muscle capillaries; details of a typical
interendothelial junction are illustrated in figure 9.4.
On the basis of present information we can compare
the permeability of capillaries in 100 g of muscle to
that of an artificial membrane, 7000 cm2 in total
area, 0.5 it thick and containing 5 X io12 uniform,
water-filled, cylindrical pores of radius 40 to 45 A.
Such an artificial membrane would have the same
filtration coefficient as the capillaries in 100 g muscle
and would restrict the diffusion of uncharged, lipid-
insoluble molecules to about the same degree (fig.
9.2). There are no reasons for supposing, however,
that the channels through capillary walls are either
1012
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 9.4. A portion of the wall of a capillary (heart musclel
to show details of the interendothelial junction. The junction
provides a continuous channel connecting the inside of the
capillary with the outside basement membrane. The width of
the channel is about 100 A. The interior of the capillary is
almost filled with an erythrocyte. [From Fawcett (93).]
cylindrical or perfectly uniform. Alternative models
utilizing different pore geometries or a limited distri-
bution of pore sizes could be devised to simulate ob-
served capillary permeability. The significant fact is
that both the hydrodynamic and diffusional char-
acteristics of the capillaries can be explained in terms
of a simple physical model which closely approxi-
mates the morphology of the capillary wall. The mean
pore radius calculated as above may be regarded as
analogous to the Einstein-Stokes molecular radius
(equation 7.5) which by itself tells nothing of the
actual shape of the molecule but is nevertheless
valuable for predicting kinetic behavior.
The restricted pore areas shown in figure 9.2 repre-
sent only a minute fraction of the total capillary
surface, but they nevertheless provide for extremely
rapid transcapillary diffusion of small lipid-insoluble
molecules. The pore area per unit path length avail-
able for diffusion of water through the capillary walls
of 100 g muscle is about 0.6 X io5 cm (fig. 9.2). The
concentration of water available for diffusion in either
direction is about 55 molar (0.99 g/ml) and the diffu-
sion coefficient of water is 3.4 X io5 cm2 per sec.
Substitution of these values in Fick's diffusion equa-
tion leads to a calculated diffusion rate of 2 g per sec.
Since the total volume of plasma in the capillaries of
100 g of muscle is only about 1 ml, this suggests that
plasma water exchanges 2 times per sec or 1 20 times
per min with the interstitial water immediately sur-
rounding the capillaries. Similar calculations for
NaCl, urea, and glucose yield exchange rate of 60, 55,
and 30 times the plasma content of these substances
per minute. An alternative method of expressing the
results is in terms of the ratio of exchange rate to
plasma flow. The latter is generally in the range 2 to 4
ml per min per 1 00 g tissue. Taking 3 ml per min as
an average, we would estimate that the diffusion of
water, NaCl, urea, and glucose back and forth through
the capillary wall occurs at rates which are, respec-
tively, 40, 20, 18, and 10 times the rate at which
these substances are brought to the tissues by the in-
coming blood. In contrast, the extravascular circula-
tion of fluid caused by net filtration and absorption is
only about 2 per cent of the plasma flow as indicated
in figure 5.2. For this reason the rates of exchange of
small molecules between blood and tissues are but
little affected by simultaneous net fluid movement.
For example, //-aminohippurate and related sub-
stances diffuse rapidly out of the peritubular capil-
laries in the direction opposite to net fluid flow (peri-
tubular capillary reabsorption). The rates of clearance
of NaM of I131 from skin are not appreciably affected
by concurrent edema formation (164).
The permeability of biological membranes is usually
expressed in terms of flux rate per unit concentration
difference divided by the area, Am , of the entire mem-
brane surface (specific permeability coefficient )
n . „ . DtAs
P, --
-*-A„
(9.4)
Ac ~m A Ax
m
Values of Ps for muscle capillaries are listed in table
9.1. Information of the type summarized in table 9.1
is not yet available from capillaries in regions other
than muscle, with the possible exception of renal
glomerular capillaries (section 10). Nevertheless, there
are many indications that permeability properties of
capillaries may differ greatly in different organs.
Studies of the rates at which labeled proteins or
dextran fractions appear in lymph from different
regions suggest that porosity of capillaries in visceral
organs may be considerably greater than in capillaries
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
table 9.1. Permeability of Mammalian Muscle
Capillaries to Lipid Insoluble Molecules
Substance
Mol Hi
D i7 C
cm- >ci '
X 10s
Approx.
Mol
Radius,
a cm
X 108
As/Ax
Restricted
fi.ll \lr.l
* Path
Length,* cm
X io5
in 100
g \hi>t le
S|n-< in
Permea-
bilityt
4- Am
cm sec'1
X io-
H„0
18
3-4
'•5
■55
28
NaCl
58
2.0
2-3
■5'
'5
Urea
60
'■95
2.6
■49
'4
Glucose
180
0.90
3-7
•44
6
Sucrose
34*
0.70
4.8
■39
4
Raffinose
5°4
0.64
5-7
■34
3
Inulin
5.500
0.21-.26
12-15
. 1 0- . 1 4
o-3
Myoglobin
1 7 , 000
0.17
'9
•°3
0. 1
Serum
67,000
0.085
36
.0007
0.001
albumin
. 00065 1
* From smooth curves of figure 9.2. f Mols/sec/cm2
membrane per mol/ml concentration difference. The total
membrane surface in 100 g muscle, Am, is assumed to be
7000 cm2 (281 ). t Calculated from over all bodily arterial
disappearance curves (equation 8.1).
of muscle (126, 232). Anatomical studies of visceral
capillaries also suggest a relatively high degree of
porosity (13). Exchange rates of lipid-insoluble mole-
cules between central nervous system tissue and blood
are far lower than in peripheral tissues (68, 185, 357)
although the anatomical site of the "blood-brain
barrier" has not been localized with certainty to the
capillary walls. Exchanges between blood and cere-
brospinal fluid are complicated by absorption in bulk
through large channels in the arachnoid villi and by-
specific active secretory mechanisms involving the
choroid plexuses and ependymal linings of the ven-
tricular system (279).
In spite of these regional differences in capillary
permeability, it may be said that over-all bodily
capillary permeability, determined from arterial disap-
pearance rates of large molecules, does not differ
greatly from that of isolated muscle. For example, the
average plasma clearance (n/Ac) of a dextran fraction
of known free diffusion coefficient (D = 0.10 X io~5
cm2 sec-1, a = 32 A) was found by Grotte (126) to be
about .6 ml per min per kg dog or io-3 ml per sec
per 100 g tissue. From equation 7. 6,
AS f> -3-5 5
Xx" 2HT "*" D * ,0 +'x /0 --O/x 10 cm
Application of the theory of restricted diffusion (equa-
tion 7.9) for average porosities of 40, 45, and 50 A
leads to true pore areas per unit path length of 1 .4, 0.5,
and 0.3 X io-5 cm per 100 g tissue, respectively.
These values, pertaining to over-all bodily permeabil-
ity, may be compared with the value of 0.6 X io5 cm
per 1 00 g tissue determined by the method of osmotic
transient in isolated muscle (Fig. 9.2). The close
correspondence between permeability to serum albu-
min computed from over-all arterial disappearance
curves on the one hand and restricted diffusion
through the capillaries of muscle on the other has
already been noted (table 9.1). Similarly, the over-
all bodily filtration coefficient is not greatly different
from that determined in intact extremities or isolated
perfused muscle (section 6). This correspondence be-
tween over-all capillary permeability and capillary
permeability determined in isolated muscle is not too
surprising since muscle accounts for some 65 per cent
of total body weight exclusive of skeleton and fat
which do not participate to a large extent in the capil-
lary exchange.
IO. MOLECULAR SIEVING OF LARGE MOLECULES:
REGIONAL DIFFERENCES IN POROSITY
In artificial systems it is possible to apply high pres-
sure differentials for rapid ultrafiltration, and under
these conditions even small molecules can be "sieved"
through porous membranes as illustrated in figure
7.4. In the capillary circulation, however, the trans-
membrane pressure differentials are necessarily small
and no appreciable steady-state concentration differ-
ences of small molecules can be maintained, even at
abnormally high rates of filtration. Substitution of
approximate value of Aw/Ax and r in equation 7. 15
suggests that appreciable molecular sieving should be
detectable with molecules of radius 10 to 15 A at high
rates of filtration caused by venous occlusion. This
prediction has been verified in perfused hind limbs
for the case of inulin (a = 12-15 A) during net filtra-
tion at the rate of 0.2 ml per min per 100 g tissue.
Under these conditions the steady-state concentration
of inulin in capillary filtrate was found to be 70 per
cent of that in plasma (281); the theoretical value
calculated from equation 7.15 is 77 per cent. In the
case of still larger molecules, including the plasma
proteins, the restriction to diffusion becomes suffi-
ciently great to allow a high degree of molecular
sieving, even at normal filtration rates.
Grotte (126) has carried out a detailed study of
molecular sieving in relation to steady-state concen-
trations of large molecules in leg lymph, liver lymph,
and cervical lymph. Grotte worked with dextran
polymers of known free diffusion coefficient and mo-
ioi4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
1.0
8
Cf_ [lymph]
C, " [PLASMA]
.6 -
.2 -
Theoretical curves
for 42 A pores
nT\°
\ X*
\» \
A
b .NORMAL LYMPH
\/^ FLOW
\°
DOUBLE '
LYMPH FLOW
(VENOUS CONGESTION)
\ V^^^\
-
\ S ^^^
MOLECULAR RADIUS, o= — —
6TTIJ
•\\ ^^j_
^ —
D ^N -"""•
l '
1 1^ 1
1
20
30
40
50
60 A
fig. i o. i . Molecular sieving of dextrans in leg lymph
obtained from dogs. The results are in accord with the theory
of molecular sieving ^equation 7.15) through pores of radius
42 A and for molecules up to 32 A in radius. Increased lymph
flow induced by venous congestion produced the expected
increase in sieving. The unexpected passage of dextran mole-
cules exceeding 40 A in radius suggests an additional "large
pore" system estimated by Grotte to comprise 1/30,000 of the
total population of pores. [Adapted from Grotte (126).]
lecular radius. Figure 10. 1, adapted from Grotte'
shows lymph : plasma concentration ratios as a func-
tion of molecular radius at two different lymph flows.
The theoretical curves for a capillary ultrafiltrate are
drawn from equation 7.15, assuming a pore radius of
42 A, that Aw/Ax was constant and that capillary
filtration rate was proportional to observed lymph
flow. The agreement between theoretical and ob-
served concentration ratios is surprisingly good for
molecules up to about 32 A in radius. However, the
observed lymph concentrations of dextran molecules
ranging in size from 50 to 90 A cannot be explained
on the basis of molecular sieving through pores of
radius 42 A. In order to explain capillary permeabil-
FIG. 10.2. Distribution of "leaks," "large pores" or gaps in
the walls of the minute vessels of frog's mesentery as indicated
by cinephotomicrographs of rapid, spotty passage of T-1824
(Evans blue dye). With camera running at the rate of 25 frames
per sec, the dye solution was perfused through the capillary
network from a micropipette introduced into the terminal,
feeding arteriole. From the film thus obtained single frames
have been removed to show sites and extent of dye passage at
intervals of seconds (e.g., 1", 2", 3", etc.) timed from that
frame in which the dye had first filled the capillaries (labeled
0). The frames labeled C show the network before dye entry;
those labeled 2' and C-» after the perfusion was ended to indicate
absence of stasis and hence absence of detectable injury.
ity to these large molecules Grotte postulated the
existence of large capillary leaks, corresponding to
pores of radius 200 to 350 A but comprising only 1
part in 30,000 of the total population of pores as com-
puted by equation 7.16. In cervical lymph and liver
lymph the molecular sieving curve was shifted to the
right and the relative number of calculated capillary
leaks was increased to 1 in 20,000 and 1 in 340, re-
spectively.
Concerning the locations of these leaks or large
pores along the length of the minute vessels very little
is known. There is some evidence, however, that they
may be more frequent in the walls of venous capillaries
and venules than in the walls of true capillaries. In
recent studies (Landis, unpublished) solutions of
T-1824 in Ringer's solution, with and without protein,
have been perfused by microinjection through single
vessels or through portions of peripheral networks in
the frog's mesentery. Motion pictures (25-40 frames
per sec) reveal sites at which the dye solutions pass
rapidly through the vessel wall during the first few
seconds of perfusion (fig. 10.2). In true capillaries the
loci of such early, spotty passage of dye are few in
number; the extravascular spots of dye are small in
size and distinct in outline. In venous capillaries and
venules the loci of passage are more numerous; the
extravascular spots of dye tend to be larger and, par-
ticularly around venules, often coalescent. It seems
likely, therefore, that while the small pore system is
uniformly distributed throughout the capillary net-
work, the leak or large pore system is more promi-
nent in the venous capillaries and venules. A differen-
tial distribution of this type helps explain earlier work
(reviewed in detail in ref. 207) on the spotty passage
of certain dyes through the walls of true capillaries
(200, 262) and on the gradient of permeability to
poorly diffusible dyes described by Rous and co-
workers (e.g., 161, 307, 308, 337, 338).
Results similar to those obtained by Grotte (126)
Left: perfusion of .01 M T-1824 freshly prepared in frog Ringer's
solution. Top section (magnification X 17) shows progression
of spotty passage involving true capillaries, a venous capillars'
and a minute venule. Middle and lower sections show greater
detail (magnification X 60) at 2 sec and 12 sec, respectively.
Right: perfusion of .01 M T-1824, ar'd 3 g/100 mg albumin, in
frog Ringer's solution. Top section (magnification X 35) shows
spotty passage in true capillaries and a venous capillary. Middle
and lower sections show greater detail (magnification X 120)
at 2 sec and 12 sec, respectively. Rapid, spotty passage of per-
fused dye persisted despite protein binding. In general, how-
ever, protein binding made spots of passage more discrete.
fig. 10.2. See legend on facing page.
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IOI7
have been reported by Mayerson et al. (232) who
found that dextran fractions of mean molecular
weights greater than about 100,000 (a > 60 A) appear
in hepatic, intestinal, and cervical lymph in concen-
trations which are almost independent of molecular
size. Mayerson et al. follow Grotte in ascribing these
results to the presence of large capillary leaks but
they also mention an interesting alternative possibility
that very large molecules may be transported by active
endothelial vesiculation (pinocytosis) as described by
Palade (273) and Moore & Ruska (255). In calcu-
lating pore-size distributions Mayerson et al. do not
take into account sieving effects nor the fact that
filtration varies with the fourth power of pore radius.
They assume that filtration rate through pores of
given collective area will be the same, regardless of
pore radius. If 5 per cent or more of the capillary
pores had a radius > 200 A, as suggested by Mayerson
et al., then from equation 7.16
r> ]/05(200 x id8) " > 95 A
A mean pore radius of 95 A for hydrodynamic flow
would lead to an improbably high value for the
filtration coefficient. For this reason we favor Grotte's
interpretation in terms of molecular sieving through
pores of radius 40 to 45 A combined with relatively
few large capillary leaks. Both interpretations are
subject to the criticism that lymph is not a capillary
ultrafiltrate and may well be modified by capillary
reabsorption (208), particularly at low rates of lymph
flow.
A more clear-cut application of the theory of
molecular sieving is possible in the case of renal glo-
merular membranes. Figure 10.3 shows the glomeru-
lar clearances of several proteins and dextrans relative
to creatinine in the dog. The apparent differences
between glomerular sieving of dextran molecules and
proteins of equivalent molecular radius may be
spurious because the dextran fractions were not
perfectly monodisperse and it is possible that the
lesser degree of sieving for each nominal molecular
radius represents the contribution of smaller dextran
molecules. The data agree well with theoretical curves
for molecular sieving through an isoporous membrane
having pores 35 to 42 A in radius and a total pore
area per unit path length for water of 1.6 X io5 cm
per g kidney (278). Substitution of these values in
equation 7.13 leads to filtration coefficients in the
range 3.5 to 5.0 X io-5 cm5 dyne-1 sec-1 per g or 2.7
to 3.9 ml per min per mm Hg per 100 g kidney.
These values are in excellent agreement with esti-
mates based on hemodynamic data (339, 384). Al-
most identical values for renal glomerular perme-
ability have also been derived by Lambert and his
associates (194-196) from molecular sieving of hemo-
globin as a function of glomerular filtration rate.
The greater permeability of renal glomerular mem-
branes relative to peripheral capillaries is evidently
due to a relatively large fractional pore area rather
than to large pores. Given a path length for filtra-
tion and diffusion of 0.5 X io~4 cm, the glomerular
pore area for passage of water would be 8 cm2 per g
C2/C1
GLOMERULAR CLEARANCE
RELATIVE TO CREATININE
1.0
DEXTRAN
FRACTIONS
fig. 10.3. Theoretical vs. ac-
tual molecular sieving through
renal glomerular membranes
of dogs. Molecular sieving of
myoglobin, egg albumin, and
hemoglobin calculated as in
references 278 and 194. Data
for dextran fractions are taken
from Wallenius (366).
40A
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
kidney or 5 to 10 per cent of the available glomerular
surface estimated histologically by Vimtrup (360).
Recent studies of glomerular ultrastructure suggest
that the anatomical basis for glomerular sieving is not
in the fenestrated capillary endothelium, but rather
in the epithelial cells (podocytes) covering glomerular
capillaries. According to Hall (141) these cells form
foot processes which approximate the endothelial
basement membranes in such fashion as to form inter-
cligitating "slit-pores" which appear to be 80 to 100 A
wide and occupy 2 to 3 per cent of the total surface.
ii. capillary permeability to lipid-soluble
molecules; respiratory gases
Capillary permeability to lipid-soluble molecules
has been studied by Renkin (296, 297) using the per-
fused hind-limb preparation. Urethan (mol wt 89),
paraldehyde (mol wt 132) and triacetin (mol wt 218)
traversed the capillary walls so rapidly that no os-
motic transients were detectable (figure 11.1). Glyc-
erol and acetic esters of glycerol were shown to pass
through capillary walls at high rates which varied in
order of their oil : water partition coefficients but in
order opposite to that expected on the basis of their
aqueous diffusion coefficients. The temperature co-
efficients of capillary permeability to antipyrine and
antipyrine derivatives were found to be related to the
temperature coefficients of their lipid solubilities
rather than to their aqueous diffusion coefficients.
These results suggest that lipid-soluble molecules
can diffuse through regions in the capillary wall
which are relatively impermeable to lipid-insoluble
materials. The permeability characteristics of this
additional pathway are similar to those of cell mem-
branes in general. It seems logical, therefore, to iden-
tify the diffusion pathway for lipid-soluble molecules
with the plasma membranes of the capillary endothe-
lial cells themselves, as opposed to the system of
water-filled pores penetrating through or between
these cells, which is capable of accounting for passage
of water and lipid-insoluble molecules.
The respiratory gases have relatively large oil:
water partition coefficients (212) and may therefore
be expected to utilize the entire endothelial surface
for the transcapillary diffusion process. Recent meas-
urements of pulmonary diffusing capacity (306) indi-
cate that permeability of human alveolar mem-
branes (alveolar capillaries plus alveolar epithelium)
is approximately 60 ml 02 per min per mm Hg O2
pressure difference. In terms of oxygen concentration
difference, this value becomes 0.4 X io5 cm3 sec-1
(i.e., ml/sec, ml/ml concentration difference). The
capillary surface area in the lungs is approximately
4 X 10s cm'2 (258), whence the specific permeability
coefficient for oxygen is 10,000 X io~B cm sec-1. This
value may be compared with 23 X io~5 cm sec-1,
representing the specific permeability of muscle capil-
laries to water (table 9.1). Presumably the greater
permeability to oxygen is a result of lipid solubility,
since the pulmonary capillaries resemble peripheral
capillaries in being relatively impermeable to small
lipid-insoluble molecules (378). In 100 g of quiescent
muscle containing a capillary surface area of 5,000-
10,000 cm2 the steady-state flow of oxygen across the
capillary walls is about 0.4 ml per min (fig. 12.2).
During maximal muscular activity the oxygen re-
quirements increase 20-fold to 30-fold and the avail-
able capillary surface may increase 2-fold to 4-fold.
fig. 1 1.1. Osmotic transients produced by
urethan and urea in an isolated perfused cat
hind limb. pCi = isogravimetric capillary
pressure. Pv = protein osmotic pressure in
perfusion fluid. 36 msi/liter of urea produced
a large osmotic transient owing to restricted
diffusion of urea through the capillary walls.
Urethan, despite its larger molecular size,
failed to produce a detectable osmotic effect.
The results are attributed to the greater lipid
solubility of urethan which enables it to diffuse
through the entire capillary endothelial
surface. [From Renkin (296 1.]
pCi
or
PP
mm
H*
25
20
15
1
T
pp
URETHAN
36mM/l
°«t>-c>J
-I-"-
UREA
+
URETHAN
36 mM/1 of each
UREA
36mM/l
80 100 120 140 160 180 200 220
TIME, MINUTES AFTER START OF PERFUSION
240
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IOig
If the specific permeability of muscle capillaries for
oxygen were comparable with that of alveolar mem-
branes, then transcapillary oxygen pressure differ-
ences of 0.3 to 0.6 mm Hg and 3 to 8 mm Hg would
suffice to account for observed rates of tissue oxygen
consumption at rest and during maximal work, re-
spectively. For CO i the corresponding values are 20-
fold smaller owing to its greater solubility. It therefore
seems unlikely that capillary permeability is an im-
portant factor limiting the exchange rates of respira-
tory gases, except possibly during maximal muscular
activity.
12. CAPILLARY PERMEABILITY AND BLOOD FLOW IN
RELATION TO EXCHANGE OF MATERIALS
BETWEEN BLOOD AND TISSUES
In previous sections evidence was reviewed showing
that lipid-soluble molecules and small lipid-insoluble
molecules or ions diffuse back and forth across capil-
lary walls at rates which greatly exceed rates at which
these substances are brought to or from the tissues by
the blood. For such substances capillary permeability
is clearly not an essential factor determining net rates
of blood-tissue exchange. Other more essential factors
include the distribution and rate of flow of capillary
blood, the volume and permeability of extravascular
distribution compartments, and rates of chemical
reaction in the tissues. For molecules of intermediate
size, including products of intermediary metabolism,
capillary permeability may become more important
but is still only one of the several factors determining
over-all kinetics of the exchange process. Only in the
case of relatively large molecules (e.g., inulin in-
larger) can capillary permeability be considered as a
primary factor limiting exchange with well-perfused
tissues.
Mathematical descriptions of diffusion kinetics in
the capillary circulation are included in papers by
Krogh (183), Hill (153, 154), Kety (173, 174), Opitz
& Schneider (269), Morales & Smith (256), Schmidt
(318, 319), Sangren & Sheppard (310), Renkin (300),
and Blum (19). Each of these mathematical descrip-
tions is based upon a particular model of capillary-
tissue geometry and each involves simplifying assump-
tions concerning permeability which do not apply to
all molecular species. Such models are nevertheless
useful, if only to provide a definite hypothesis with
which experimental results may be compared. Exam-
ples illustrating the use of such models are given
below.
A. Blood- Tissue Transport of Oxygen
The essential role of the capillaries in the blood-
tissue exchange of respiratory gases was considered by
Krogh (183) in terms of spatial distribution of blood
vessels relative to tissue metabolism. Krogh proposed a
simple model in which each capillary of radius, r, sup-
plied a cylinder of tissue of radius R. The intercapil-
lary distance was therefore iR and the number of
capillaries per cm2 was (1/2R)-. It was assumed that
rate of tissue metabolism would be uniform through-
out the cylinder and that the diffusion coefficients
of gases through the cylinder would be uniform and
identical with values measured in dead tissues. The
mathematical solution for steady-state radial diffu-
sion under these conditions was derived for Krogh by
Erlang (183) and has formed the starting point for
many subsequent discussions of the blood-tissue ex-
change of gases [cf (174) for contemporary review].
Figure 12.1 is a graph of the Krogh-Erlang equa-
tion for capillaries of radius 4 ft; the equation is rela-
tively insensitive to values of r and for all practical
purposes the same graph applies to capillaries of radii
3 to 5 /j. This model suggests that as few as 25 open
capillaries per mm2 would suffice to supply the oxygen
requirements of resting muscle without exceeding the
limiting diffusion pressure head set by oxygen in
venous blood (i.e., a finite oxygen pressure would
exist even in the outermost region of the diffusion
cylinder surrounding each capillary). The corre-
sponding figure for maximal muscular activity is 500
capillaries per mm2. Brain and liver would require 200
anc 100 capillaries per mm2, respectively. Estimates of
capillary density usually exceed these values by a wide
margin and suggest that the oxygen pressure head re-
quired to supply the diffusion cylinder around each
capillary is far less than that available in capillary
blood, even at maximal rates of tissue metabolism.
Capillary counts on injected muscles from anesthetized
animals lead to estimates in the range 200 to 600 per mm2
for resting muscle and 600 to 5000 per mm2 for contracting
muscle (84, 143, 183, 22-j, 272, 284, 320, 335, 353). There is
considerable variation among skeletal muscle, heart (305),
and abdominal wall muscle (184) representing examples of
high and low density, respectively. In general, muscles from
small animals have a higher capillary density than from large
animals (320). In maximal vasodilatation there is often a 1 : 1
relation between number of capillaries and number of muscle
fibers, but maximum capillary density can be increased by
exposure to high altitudes or by daily physical exercise (358).
Capillary counts made on fixed preparations tend to be high
because of shrinkage artifact; in frozen sections the muscle
fibers are larger and estimated capillary densities smaller.
In the author's experience, 150-200 capillaries per mm2 is
[020
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Q0 ,ml /min per lOOg Tissue
20 10
fig. 1 2. 1. Steady-state radial diffu-
sion of oxygen (Qo:) as a function of
capillary density (abscissa) and radial
oxygen pressure gradient from capil-
laries to tissue (ordinate). [Graph con-
structed from the Krogh-Erlang equa-
tion (183).]
90
80
70
60
50
40
30
20
- MM Hg
02 PRESSURE
/ARTERIAL*g0_
/ BLOOD 80-
/ 70-
GRADIENT
/ 60-
- PC - PR /
/ 50-
02Sot. = 72%
VENOUS
BLOOD
02Sat = 55%
/ *» /
/ * /
AY
20-
/ kr
/// /
- /*/ */
/ 0 / /
~Av /
A
/"V
7
/ 10—
/ KROGH-ERLANG EQUATION
/ 5"
II /
PC-PR =
4aD
^H**-r*}-
20
30 40 60 80 100 140 200 300 Intercopillary distarce, 2R, u.
1600
800
400
200
100
50
25
10
Capillaries per mm2
8.0
40
2.0
1.0
0.5
0.25
0.12
0.05
Capillary volume, % of tissue
400
200
100
50
25
12
6
2.5
Surfoce, cm2 per g tissue
usual in frozen sections of muscles from hind limbs of an-
esthetized cats. Many of the higher estimates imply capillary
blood volumes in the range 5 to 15 per cent of tissue volume.
In most skeletal muscles the entire blood volume is less than
4 per cent of tissue volume (328, 335) and at least half of this
may be accounted for by large blood vessels (243). Even taking
low estimates for capillary density, however, (200/mm2 at
rest, 600/mm2 in activity) the oxygen pressure gradient pre-
dicted by the Krogh-Erlang model would be less than 5 mm Hg
at rest and less than 20 mm Hg in maximum work, leading to
tissue oxygen pressures of 10 to 30 mm Hg in the outermost
regions of each diffusion cylinder.
The Krogh-Erlang model provides a theoretical
basis for analysis of the blood-tissue gas exchange, but
several lines of evidence suggest that factors other than
simple radial diffusion in a homogeneous medium
may be involved. In the case of skeletal muscle, Milli-
Pc , c opi 1 1 ory 02 pressure, mm Hg
PR .tissue 02 pressure at R
Qn , Op consumption ml /sec per ml tissue
u2
a ,0t solubility =2. 8 « 10"' ml /ml «mm Hg"1
D , tissue 02 diffusion coeff. = 1.5 x 10"' cmJ/sec ( )
R , radius of diffusion cylinder
r .capillary radius = 4 u.
kan (253) showed that intracellular myoglobin rapidly
becomes desaturated during contraction of the soleus
muscle in the cat. Since the half saturation pressure
of myoglobin at physiological pH is only 3 mm Hg
this implies that intracellular oxygen tension falls to
extremely low values during contraction. Lactic acid
increases rapidly in venous blood from contracting
muscle (7, 181), also indicating that oxygen supply
cannot keep up with demand at high rates of metabo-
lism, despite normal oxygen pressures in venous blood.
Mechanical reduction of blood flow to resting muscle
may cause substantial reduction of steady-state oxygen
consumption even when the blood vessels are dilated
and when venous oxygen pressure is sufficiently high
to meet the diffusion requirements estimated from the
Krogh-Erlang model (275, 359) (fig. 12.2). Inter-
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
1021
capillary oxygen pressures may be extremely low in
brain (64) despite normal oxygen pressures in cerebral
venous blood. The critical venous oxygen pressure at
which brain suffers a decrease in oxygen consumption
is 20 to 25 mm Hg (269); presumably, under these
conditions, oxygen pressure is zero in regions most
remote from the capillaries.
These observations suggest that the gradient of
oxygen pressure from capillary blood to tissues is
greater than predicted from the simplified model
proposed by Krogh. One factor neglected by Krogh's
treatment of the problem is the rate at which oxygen
can be released from red cells during their brief expo-
sure to the tissues in capillary blood. Roughton &
Forster (306) and Forster (108) have recently dis-
cussed evidence that chemical reaction velocity and
diffusion in the red cell account for almost one-half
the total resistance to transfer of oxygen between
alveolar gas and blood. The rate of dissociation of
oxygen from hemoglobin is slower than its rate of
combination and recent measurements by Niesel et al.
(261) and Thews (356) indicate that the intracapil-
lary component of oxygen diffusion may be a major
factor limiting the rate at which oxygen can be sup-
plied to adjoining tissues. This factor could be evalu-
ated experimentally and deserves attention in future
studies of the blood-tissue exchange of gases. Scho-
lander's recent demonstration of facilitated diffusion
of oxygen through thin films of hemoglobin or myo-
globin (32 1 ) may also be of significance for diffusion
of oxygen in muscle, especially cardiac muscle.
B. Blood- Tissue Exchange of Small,
Xonmctabolized Molecules or Ions
A simple model of blood-tissue exchange has been
employed by Renkin (299, 300) to describe diffusion
kinetics of urea, antipyrine, sucrose, and K42 in perfused
muscle. This model is particularly useful for illus-
trating the relative effects of permeability and blood
flow on diffusion kinetics in uniformly perfused tissue.
In its simplest form the model assumes two compart-
ments representing total blood volume, V\ , and extra-
vascular distribution volume, V2 ■ The compartments
are separated by a barrier of virtual area Am and per-
meability coefficient, P. V\ is allowed to flow past the
barrier at rate, Q. V% is assumed to be homogeneous
with respect to concentration of diffusing materials.
The mathematical solution for this model (300) is
given bv
d/i-e « J (12.1)
v.v,
(v, + v2)
where C = clearance from the blood compartment,
V 1 , ml per min, and X = slope of the exponential
disappearance curve from the blood compartment,
min-1. In perfused preparations the rate of blood flow-
may be varied over a wide range by simple adjustment
of perfusion pressure. Clearance, C, from the perfu-
sion reservoir can be measured accurately. It is there-
fore possible to determine over-all permeability, P X
Am, of barriers separating blood from the final distri-
bution volume, provided the original assumption of
uniform distribution in extravascular space is correct.
For many substances this assumption will not be valid
and in such cases PAm must be considered as a virtual
permeability which includes the effects of nonuniform
distribution in extravascular space.
Figure 12.3 shows capillary clearances of anti-
pyrine, K42 and urea as a function of blood flow in
widely dilated blood vessels of mammalian muscle.
The changes in blood flow were produced by change
of arterial perfusion pressure and presumably reflect
changes in flow velocity through a constant capillary
surface as required by the model. Comparison of the
results with theoretical curves drawn from equation
1 2. 1, suggest blood-tissue permeabilities (PAm) of
about 3 and 10 ml per min per 100 g muscle for urea
and K42, respectively. For antipyrine the observed
capillary clearances were equal to blood flow, indi-
cating that for this (lipid-soluble) substance perme-
ability (PAm) was large with respect to blood flow.
Blood-tissue permeabilities estimated by equation
1 2. 1 from measurements of blood flow and clearances
are compared in table 12.1, with capillary permeabil-
ity estimated from osmotic transients and the theory
of restricted diffusion. In the case of sucrose the blood-
tissue permeability is 30 to 60 per cent of capillary
permeability. Sucrose distributes primarily in inter-
stitial fluid and the only barriers to diffusion are
capillary walls and interstitial fluid volume. From
the available data (table 12.1) it appears that in
muscle about one-half the total resistance to distribu-
tion is located in the capillary wall. Cotlove (46)
has shown that distribution rates of NaCl, sucrose, and
inulin into connective tissue spaces of extremities are
limited by the long path length for diffusion along
fascial planes and by retardation of diffusion in the
interstitial matrix. Recent measurements by Ogston &
Sherman (266) indicate that diffusion of molecules as
small as glucose may be appreciably restricted in
dilute gels formed by hyaluronic acid and the action
of hvaluronidase in reducing resistance to flow-
through connective tissue has been described by Day
(69)-
1022
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
40
„ 02 CONSUMPTION
ml /min /I00g Muscle '
35
/
•
&
•
40 C
30
A-/
O,
• —■""
O' /
\
•
.25
o' /
K/ /
,-V' /
S = 39 %
Pv = 40mm Hg
.20
la! A
kfi /*X ______
£: / ^- — x
* X 31 C
.15
' // s = 42%
' // Pvrt = 28
.10
/ // V0 2
' if
' //
. // h no n — | p.
C
05
■ ;/y\
/// XS=55%
n
V Pv0!=l6
/ i i i i
02 SUPPLY ml
i i i
02/min /I00 g Muscle
i i i i
.2
1.0
BLOOD FLOW ml/min/IOOg
fig. 12.2. Steady-state oxygen consumption as a function of blood flow and tissue temperature
in the hind limb muscles of an anesthetized cat. Oxygen consumption was lowered when oxygen
supply (blood flow) was reduced below a critical value at each temperature. Critical oxygen pressures
in venous blood were 40, 28, and 16 mm Hg at oxygen consumptions of 0.3, 0.15, and 0.06 ml/min
respectively. A capillary -tissue pressure gradient of less than 5 mm Hg would suffice to supply these
rates of oxygen utilization by simple radial diffusion in tissue containing 100 perfused capillaries
per mm: (fig. 12. 1). The results indicate that the gradient of oxygen pressure from capillary blood
to tissue is greater than that predicted from the simplified model proposed by Krogh (183). Oxygen
saturation, S, measured by oximeter and gas analysis. Oxygen pressure in venous blood, Pv0i,
estimated from measured oxygen dissociation curves at each temperature. Tissue temperature ad-
justed by passing femoral arterial blood through a heat exchanger. Blood flow adjusted by variable
arterial resistance. (From unpublished experiments by Rapela et al.)
Urea and K42 distribute in intracellular water and
for these substances the chief barrier to diffusion is
probably located at cell membranes in the tissues.
Table 12. 1 shows that blood-tissue permeabilities to
these substances are far less than respective capillary-
permeabilities.
When blood-tissue permeability is large with respect
to blood flow, equation 12.1 approaches the limit C
= Q, and blood-tissue distribution is said to be flow
limited. This is the case for lipid-soluble molecules in
general (e.g., antipyrine, fig. 12.3) and provides the
theoretical basis for estimating regional blood flow
from blood or tissue clearances of these substances
(173). Johnson et al. (168) have shown that distribu-
tion of labeled water is blood flow limited in cardiac
and skeletal muscle and Sapirstein (311) has used the
blood clearances of Rbs6 or K42 as a measure of rela-
tive regional blood flow. The clearances of labeled
Xa or I from blood or interstitial space have also been
used for this purpose (72, 165, 172, 290, 362) but in
view of the interstitial component of blood-tissue
permeability this may not be justified. Several investi-
gators have measured fractional extractions (C/Q) of
test materials during single passage through vascular
beds of extremities (40, in), head (40), liver (40),
heart (44) and lungs (39). Equation 12.1 suggests
that the values so obtained reflect the exponential
ratios of blood-tissue permeability to blood flow under
the conditions of vascular tone prevailing at the time
of measurement.
For large lipid molecules the chief barrier to tissue
distribution is the capillary wall and in this case PAm
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
1023
14 CAPILLARY
CLEARANCE
ml /min per lOOg
PA
fig. 12.3. Diffusion kinetics of anti-
pyrine, K", and urea in vasodilated
muscle. Permeability to antipyrine
(lipid-soluble) is so large that its
clearance is limited by rate of blood
flow. The clearances of K.41 and urea
are limited, in part, by permeability
of cell membranes in extravascular
distribution volume. Less than 10%
of the diffusion barrier to urea is con-
tributed by the capillary wall (table
1 2.1). [Adapted from Renkin (299,
3°°)-]
BLOOD FLOW, Q, ml /min per lOOg
table 1 2. 1. Comparison of Blood-Tissue Permeability
with Capillary Permeability
Permeability.
ml/min/ioo g Muscle
Substance
Blood-tissue
PAm*
Capillary walls
D„A,/Axt
Sucrose
Urea
K42
5-i 1
4 ± 2
7 ± 3
18
54
90
* From equation 12.1 f From table 9.1.
= Ds A,/ Ax. For these substances PAm is small com-
pared to normal rates of blood flow and equation 12.1
reduces to equation 8. 1 describing arterial disappear-
ance curves of large molecules.
C. Nonuniform Distribution of Blood Flow in
Relation to Blood-Tissue Exchange
The model discussed in the previous paragraphs was
designed to simulate effects of changes in flow velocity
through a constant number of open capillaries and the
results illustrated in figure 12.1 refer to widely dilated
blood vessels. At any given over-all blood flow, the
clearance of test molecules may be very much smaller
during vasoconstriction (299, 300). In supine, anes-
thetized dogs the fractional extraction of antipyrine
or D20 from the circulation to extremities may be
only 0.6 to 0.8 (40) in contrast to values close to unity
in perfused, vasodilated muscle (168) or the intact
human forearm (111). The fraction of total blood
flow passing through true (nutrient) capillaries is
subject to wide variation according to metabolic de-
mands of the tissue or to hemodynamic demands of
the organism as a whole. In some tissues, such as skin,
liver, or intestine, the nonnutrient fraction of total
blood flow may pass through arteriovenous anasto-
moses of potentially large caliber; in other tissues,
such as mesentery or muscle, effective physiological
shunts are formed by arteriovenous capillaries (388).
Nonuniform distribution of blood flow within single
organs may also occur between regions of different
function and metabolic rate, examples being medulla
and cortex of the kidney or gray and white matter of
the central nervous system.
It is obvious that nonuniform alterations of blood
flow in the microcirculation will change the relations
between total blood flow and blood-tissue exchange
rates; conversely, it may be anticipated that quantita-
tive studies of effective tissue perfusion will depend
heavily upon information obtained from exchange
rates. At the present time, available information is
mostly qualitative and derives in large part from
observations on muscle.
A striking example of nonuniform distribution of
blood flow in skeletal muscle can be observed following
electrical or reflex stimulation of sympathetic vaso-
constrictor nerves. Closure of precapillary sphincters,
iO'->4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
innervated by the sympathetic vasoconstrictor system,
can stop nutrient blood flow through large areas of
capillary bed, leaving blood to flow through arterio-
venous thoroughfare channels or other regions of
low metabolic rate. Under these conditions, total
blood flow is reduced but oxygen saturation of venous
blood approaches that of arterial blood (26, 274, 294)
and respiratory gas exchange may be reduced to one-
half or less of its normal value. In diving mammals,
drastic vasoconstriction of this type greatly reduces
tissue gas exchange for periods of one-half hour or
more (322, 323). This is in contrast to uniform reduc-
tion of total blood flow caused by decrease in arterial
pressure or infusion of vasoconstrictor drugs (274);
under these conditions, oxygen extraction is increased
and oxygen utilization remains relatively constant
over a wide range of blood flow (fig. 12.2).
Recent investigations of the rates at which labeled
ions are removed from interstitial space in muscle also
suggest that vasomotor nerves control the distribution
of blood between nutrient and nonnutrient circula-
tions (165). There is general correspondence between
clearance of Na24 or I131 and over-all blood flow when
flow is altered by pressure, reactive hyperemia, or
exercise (72, 165, 172, 290, 362). However, activation
of the vasomotor system to muscle or skin generally
results in large changes of flow without corresponding
changes of clearance. Hyman el al. (165) have shown
that I131 clearance may actually decrease during large
increases of flow caused by activation of the sympa-
thetic vasodilator system. They suggest that vasodi-
lator nerves act primarily to increase flow through
arteriovenous thoroughfare channels.
Studies of this type are only beginning and they
point to new directions for research on the peripheral
circulation. The principal function of the circulation
is to provide for exchange of materials between blood
and tissues and it seems logical to study this function
directly in terms of exchange rates. Such studies will
only be meaningful, however, if the limitations im-
posed by over-all permeability are considered in rela-
tion to tissue perfusion. In the present article we have
provided a quantitative background for assessing the
role played by capillary permeability in the distribu-
tion process, indicating only briefly the contributions
of interstitial diffusion, cellular permeability, or
chemical reaction velocity.
REFERENCES
1. Adair, G. S. A critical study of the direct method of
measuring the osmotic pressure of haemoglobin. Proc.
Roy. Soc, London, Ser. A 108: 627-637, 1925.
2. Adair, G. S. The osmotic pressure of haemoglobin in
the absence of salts. Proc. Roy. Soc., London, Ser. A 109:
292-300, 1925.
3. Adair, G. S., and M. E. Robinson. The analysis of the
osmotic pressures of the serum proteins, and the molecular
weights of albumins and globulins. Biochem. J. 24: 1864-
1889, 1930.
4. Albritton, E. C. (editor). Standard Values in Blood. The
First Part of a Handbook of Biological Data. Dayton,
Ohio: Wright-Patterson AFB, 1 95 1 , 199 pp.
5. Amberson, VV. R. A criticism of the Hill-Hartree method
of curve analysis. J. Phvsiol., London 59: 67-80, 1930.
6. Armentano, von L., A. Bensath, T. Beres, St.
Rusznyak, and A. Szent-Gyorgyi. Uber den Einfluss
von Substanzen der Flavongruppe auf die Permeabilitat
der Kapillaren. Vitamin P1. Deut. med. Wochschr. 62:
i325-!328. '93d-
7. Asmussen, E., and M. Nielsen. Studies in the regulation
of respiration in heavy work. Acta Physiol. Scand. 12:
1 71-188, 1946.
8. Baltzer, A., H. Wuthrich, P. Schmuziger, and W.
Wilbrandt. Uber eine Registriermethode zum Studium
der Kapillarpermeabilitat. Helvet. Physiol, et Pharmacol.
Acta. 15:450-471, 1957.
9. Barcroft, H., and O. G. Edholm. Temperature and
blood flow in the human forearm. J. Physiol., London
104:366-376, 1946.
10. Bartholinus, T. Vasa lymphatica nuper in animantibus
inventa. Hafniae, 1653. Cited by E. Starling. In: Schafer's
Textbook of Physiology. London: Pentland, 1898, vol. 1,
p. 286-287.
11. Bayliss, L. E., and E. Lundsgaard. The action of cya-
nide on the isolated mammalian kidney. J. Physiol.,
London 74:279-293, 1932.
12. Bayliss, W. M., and E. H. Starling. Observations on
venous pressures and their relationship to capillary
pressures. J. Physiol., London 16: 159-202, 1894.
13. Bennett, H. S., J. H. Luft, and J. C. Hampton. Mor-
phological classifications of vertebrate blood capillaries.
Am. J. Physiol. 196:381-390, 1959.
14. Bennhold, H., H. Peters, and E. Roth. Uber einen
Fall von kompletter Analbuminaemie ohne wesentliche
klinische Krankheitszeichen. Verhandl. deut. Ges. inn. Med.
60:630-634, 1954.
15. Bierman, H. R., R. L. Byron, Jr., K. H. Kelly, R. S.
Gilfillan, L. P. White, N. E. Freeman, and N. L
Petrakis. The characteristics of thoracic duct lymph in
man. J. Clin. Invest. 32- 637-649, 1953.
16. Bigelow, S. L. The permeabilities of collodion, Gold
Beater's skin, parchment paper and porcelain membranes.
J. Am. Chem. Soc. 29: 1 675-1 692, 1907.
17. Bing, J. Investigation on the value of Landis' capillary-
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
1025
permeability test in the clinic. Acta Med. Scand. 94: 254-
257. I938-
18. Bjerrum, N., and E. Manegold. Ueber Kollodium-
Membranen, II. Der Zusammenhang zwischen Mem-
branstruktur und Wasserdurchlassigkeit. Kolloid-Z. 43 :
5-14, 1927.
19. Blum, J. J. Concentration profiles in and around capil-
laries. Am. J. Physiol. 197:991-998, i960.
20. Bollman, J. L. Extravascular diffusion of dextran from
blood. J. Lab. Clin. Med. 41 : 421-427, 1953.
21. Bott, P. A., and A. N. Richards. The passage of protein
molecules through the glomerular membranes ./. Biol.
Chem. 141:291-310, 1941.
22. Brown, E., J. Hopper, Jr., J. H. Sampson, and C.
Mudrick. The loss of fluid and protein from the blood
during a systemic rise of venous pressure produced by
repeated Valsalva maneuvers in man. J. Clin. Invest.
37: 1 465- 1 475, 1958.
23. Brown, E., and E. M. Landis. Effect of local cooling on
fluid movements, effective osmotic pressure and capillary
permeability in the frog's mesentery. Am. J. Physiol.
149:302-315, 1947.
24. Brown, E., C. S. Wise, and E. O. Wheeler. The effect
of local cooling on the filtration and absorption of fluid
in the human forearm. J. Clin. Invest. 26: 1031-1042,
'947-
25. Brues, A. M., and C. McT. Masters. The permeability
of normal and malignant cells to water. Am. J. Cancer
28: 324^333. '936-
26. Bucherl, E., and M. Schwab. Der Sauerstoffverbrauch
des ruhenden Skeletmuskels bei reflektorisch-nervoser
Vasokonstriktion. Pfliigers Arch. ges. Physiol. 254: 337-
343. 1952-
27. Bugher, J. C. Characteristics of collodion membranes
for ultrafiltration. J. Gen. Physiol. 36: 431-448, 1953.
28. Burch, G. E. Formation of edema in the eyelids of man.
Influence of local tissue pressure, skin distensibility,
lymph flow, intraorbital pressure gradient and venous
pressure. A.M. A. Arch. Internal Med. 65: 477-498, 1940.
29. Burch, G. E. Influence of the central nervous system
on veins in man. Physiol. Revs. 40, Suppl. 4: 50-56, i960.
30. Burch, G. E., and W. A. Sodeman. The estimation of
the subcutaneous tissue pressure by a direct method.
J. Clin. Invest. 16: 845-850, 1937.
31. Burton, A. C. Relation of structure to function of tissues
of the wall of blood vessels. Physiol. Revs. 34: 619-642,
'954-
32. Cachera, R., and F. Darnis. Etude de la permeabilite
capillaire chez le sujet normal. Ann. mid., Paris 51 : 509-
542, I95°-
33. Cachera, R., and F. Darnis. Les troubles de la per-
meabilite capillaire dans les hepatites infectieuses et
dans les cirrhoses. Semaine hop. 27: 1849-1862, 1951.
34. Calvin, D. B. The effect of asphyxia upon plasma volume
and protein concentration. Am. J. Physiol. 133: 233-234,
I94I-
35. Campbell, M. L., and A. H. Turner. Serum protein
measurements in the lower vertebrates. I. The colloid
osmotic pressure, nitrogen content, and refractive index
of turtle serum and body fluid. Biol. Bull. 73: 504-510,
■937-
36. Carrier, E. B., and P. B. Rehberg. Capillary and venous
H
45
pressure in man. Skand. Arch. Physiol. 44: 20-31, 1923.
37. Chambers, R., and B. W. Zweifach. Intercellular
cement and capillary permeability. Physiol. Revs. 27:
436-463, 1947.
38. Chinard, F. P. Derivation of an expression for the rate
of formation of glomerular fluid (GFR). Applicability of
certain physical and physicochemical concepts. Am. J.
Physiol. 171 : 578-586, 1952.
39. Chinard, F. P., and T. Enns. Transcapillary pulmonary
exchange of water in the dog. Am. J. Physiol. 178: ig7_
202, 1954.
40. Chinard, F. P., G. J. Vosburgh, and T. Enns. Trans-
capillary exchange of water and of other substances in
certain organs of the dog. Am. J. Phsiol. 183: 221-234,
'955-
41. Churchill, E. D., F. Nakazawa, and C. K. Drinker.
The circulation of body fluids in the frog. J. Physiol.,
London 63: 304-308, 1927.
42. Cohnheim, J. Ueber Entzundung und Eiterung. Virchow's
Arch. Pathol. Anal. 40: 1-79, 1867.
43. Cohnheim, J Lectures on General Pathology. A Handbook for
Practitioners and Students. Sect. I. The Pathology of Circula-
tion. Translated from the 2nd German ed. by A. B. McKee.
London: New Sydenham. Soc. 1889, p. 292.
Conn, H. L., Jr., and J. S. Robertson Kinetics of
potassium transfer in the left ventricle of the intact dog.
Am. J. Physiol. 181 : 319-324, 1955.
Cope, O , and F. D. Moore. A study of capillary per-
meability in experimental burns and burn shock using
radioactive dyes in blood and lymph. J. Clin. Invest.
23: 241-257, 1943.
45a. Cope, O., and S. B. Litwin. Contribution of the lym-
phatic system to the replenishment of the plasma protein
following a hemorrhage. Ann. Surgery 156: 655-667, 1962.
Cotlove, E. Mechanism and extent of distribution of
inulin and sucrose in chloride space of tissues. Am. J.
Physiol. 176:396-410, 1954.
Coulter, N. A., Jr. Filtration coefficient of the capil-
laries of the brain. Am. J. Physiol. 195: 459-464, 1958.
Courtice, F. C. The effect of local temperature on fluid
loss in thermal burns. J. Physiol., London 104: 321-345,
1946.
Courtice, F. C. Rept. Australian New Zealand Assoc.
Advance. Sci. 28th Meeting, Brisbane 28: 115-119, 195 1.
(Quoted from ref. 386)
50. Courtice, F. C. Permeability of normal and injured skin
capillaries to lipoproteins in the rabbit. Australian J.
Exptl. Biol. Med. Sci. 37: 451-463, 1959.
51. Courtice, F. C. The permeability of liver and skin
capillaries to lipids in the cat. Australian J. Exptl Biol.
Med. Sci. 37:465-471, 1959.
52. Courtice, F. C The transfer of proteins and lipids from
plasma to lymph in the leg of the normal and hyper-
cholesterolaemic rabbit. J. Physiol., London 155: 456-
469, 1961.
53. Courtice, F. C, and P. I. Korner. The effect of anoxia
on pulmonary oedema produced by massive intravenous
infusions. Australian J. Exptl. Biol. Med. Sci. 30: 511-526,
'952-
54. Courtice, F. C, and B. Morris. The exchange of lipids
between plasma and lymph of animals. Quart. J. Exptl.
Physiol. 40: 138-148, 1955.
46
47
49
1026
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
55. Courtice, F. C, and P. J. Phipps. The absorption of
fluids from the lungs. J Physiol., London 105: 186-190,
1946.
56. Courtice, F. C, and W. J. Simmonds. Absorption of
fluids from the pleural cavities of rabbits and cats. J.
Physiol. , London 109: 11 7-1 30, 1949.
57. Courtice, F. C, W. J. Simmonds, and A. W. Steinbeck.
Some investigations on lymph from a thoracic duct
fistula in man. Australian J. Exptl. Biol. Med. Sci. 29:
201-210, 1 95 1.
58. Courtice, F. C, and A W. Steinbeck. The lymphatic
drainage of plasma from the peritoneal cavity of the cat.
Australian J. Exptl. Biol. Med. Set. 28: 161-169, 1950.
59. Courtice, F. C, and A. W. Steinbeck. The effects of
lymphatic obstruction and of posture on the absorption
of protein from the peritoneal cavity. Australian J. Exptl.
Biol. Med. Sci. 29 : 45 1 -458, 1 95 1 .
60. Crandall, L. A., Jr., S. B. Barker, and D. G. Graham.
A study of the lymph flow from a patient with thoracic
duct fistula. Gastroenterology 1 : 1040-1048, 1943.
61. Danielli, J. F. Capillary permeability and oedema in
the perfused frog. J. Physiol., London 98: 109-129, 1940.
62. Darcy, H. Les fontaines publique de la Ville de Dijon.
Cited by R. D. Wyckoff, H. G. Botset, M. Muskat,
and D. W. Reed. Rev. Sci. Instr. 4: 394-405, 1933.
63. Dauohaday, W. H. Steroid protein interactions. Physiol.
Revs. 39:885-902, 1959.
64. Da vies, P. W., and D. W. Bronk. Oxygen tension in
mammalian brain. Federation Proc. 16: 689-692, 1957.
65. Davis, D. L., and W. F. Hamilton. Small vessel responses
of the rabbit ear. Am. J. Physiol. 196: 1312-1315, 1959.
66. Davis, D. L., and W. F. Hamilton. Small vessel responses
of the dog paw. Am. J. Physiol. 196: 1316-1321, 1959.
67. Davis, D. L., and W. F. Hamilton. Cross circulation
at the small blood vessel level in the dog's paw. Am. J.
Physiol. 199: 1169-73, IO-6o.
68. Davson, H. Physiology of the Ocular and Cerebrospinal Fluids.
Boston: Little, Brown, 1956.
6g. Day, T. D. The permeability of interstitial connective
tissue and the nature of the interfibrillary substance.
J. Physiol., London 117: 1-8, 1952.
70. DiPasqijale, E. L., and A. A. Schiller. Effect of hy-
poxemia on edema formation in perfused isolated rat
hind limb. Proc. Soc. Exptl. Biol. Med. 78: 567-571, 1951.
71. Dixon, M., and K. A. C. Elliott. The effect of cyanide
on the respiration of animal tissues. Biochem. J. 23: 812-
830, 1929.
72. Dobson, E. L., and G. F. Warner. Measurement of
regional sodium turnover rates and their application to
the estimation of regional blood flow. ,4m. J. Physiol.
189: 269-276, 1957.
73. Doupe, J., H. VV. Newman, and R. W. Wilkins. The
effect of peripheral vasomotor activity on systolic arterial
pressure in the extremities of man. J. Physiol., London
95 : 244^57, '939-
74. Drinker, C. K. The permeability and diameter of the
capillaries in the web of the brown frog (R. temporaria)
when perfused with solutions containing pituitary ex-
tract and horse serum. J. Physiol., London 63: 249-269,
1927.
75. Drinker, C. K. Extravascular protein and the lymphatic
system. Ann. X. V. Acad. Set. 46: 807-821, 1946.
76.
77-
78.
79-
80.
»3-
84
85.
86.
87.
9°-
9<-
92.
93-
94-
Drinker, C. K., and M. E. Field. Lymphatics, Lymph
and Tissue Fluid. Baltimore: Williams & Wilkins, 1933.
Drinker, C. K., M. E. Field, J. W. Heim, and O. C.
Leigh, Jr. The composition of edema fluid and lymph
in edema and elephantiasis resulting from lymphatic
obstruction. Am. ./. Physiol. 109: 572-586, 1934.
Drinker, C. K., M. F. Warren, and M. MacLanahan.
The absorption of protein solutions from the pulmonary
alveoli. J. Exptl. Med. 66: 449-458, 1937.
Drinker, C. K., and J. M. Yoffey. Lymphatics, Lymph
and Lymphoid Tissue — Their Physiological and Clinical
Si»nilh-956-
Duyff, J. W., and H. D. Bouman. Uber die Kapillarisa-
tion einiger Kaninchenmuskeln. Z. Zelljorsch. 5: 596-
614, 1927.
Ebbecke, U. Capillarerweiterung, Urticaria und Schock.
Klin. U'oschr. 2: 1 725-1 727, 1923.
Edsall, J. T., and J. Wyman. Biophysical Chemistry, Chapt.
II. New York: Academic Press, 1958.
Eichna, L. W. Capillary blood pressure in man. Direct
measurements in the digits during arterial hypertension
induced by paredrinol sulfate. J. Clin. Invest. 21: 731 —
734. '94-!-
Eichna, L. W., and J. Bordley, III. Capillary blood
pressure in man. Comparison of direct and indirect
methods of measurement. J. Clin. Invest. 18: 695-704,
■939
Eichna, L. W., and J. Bordley, III. Capillary blood
pressure in man. Direct measurements in the digits of
normal and hypertensive subjects during vasoconstriction
and vasodilatation variously induced. J. Clin. Invest. 21:
711-729, 1942.
Eichna, L. W., and R. W. Wilkins. Capillary blood
pressure in man. Direct measurements in the digits during
induced vasoconstriction. J. Clin. Invest. 21 : 697-709,
1942.
Einstein, A. Uber die von der molekularkinctischen
Theorie der Warme geforderte Bewegnng von in ruhenden
Fliissigkeiten suspendierten Teilchen. Ann. Physik. 17:
549-560, 1905.
Fahr, G., and I. Ershler. Studies of the factors con-
cerned in edema formation. II. The hydrostatic pressure
in the capillaries during edema formation in right heart
failure. Ann. Internal Med. 15: 798-810, 1941.
Favvcett, D. W. The fine structure of capillaries, arterioles
and small arteries. In : The Microcirculation. Symposium on
Factors Influencing Exchange of Substances Across Capillary
Wall. Urbana, 111. Univ. Illinois Press, 1959, pp. 1-27.
Faxen, H. Der Widerstand gegen Bewegung einer
starren Kugel in einer zahen Fliissigkeit, die zwischen
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
102 7
95
9°
97
99'
103.
105.
106.
107.
108.
109.
114.
"5-
zwei parallclen ebenen Wanden eingeschlossen ist.
Ann. Physik. 68: 89-1 ig, 1922.
Ferry, J. D. Statistical evaluation of sieve constants in
ultrafiltration. J. Gen. Physiol. 20: 95-104, 1936. 116.
Fick, A. Uber Diffusion. Ann. Physik. 94: 59-86, 1855.
Field, M. E., and C. K. Drinker. The permeability of 117.
the capillaries of the dog to protein. Am. J. Physiol.
97: 40-5', '93'-
Field, M. E., and C. K. Drinker. Conditions governing 118.
the removal of protein deposited in the subcutaneous
tissues of the dog. Am. J. Physiol. 98: 66-69, '931-
Field, M. E., and C. K. Drinker. The rapidity of
interchanges between the blood and lymph in the dog.
Am. J. Physiol. 98: 378-386, 1931 . 119.
Field, M. E., C. K. Drinker, and J. C. White. Lymph
pressures in sterile inflammation. J. Exptl. Med. 56:
363-37o. 1932. 120.
Fine, J., and A. M. Seligman. Traumatic shock: IV.
A study of the problem of the 'lost plasma' in hemorrhagic
shock by the use of radioactive plasma protein. J. Clin.
Invest. 22: 285-303, 1943. 121.
Fine, J., and A. M. Seligman. Traumatic shock VII.
A study of the problem of the 'lost plasma' in hemorrhagic, 122.
tourniquet, and burn shock by the use of radioactive
iodo-plasma protein. J. Clin. Invest . 23: 720-730, 1944.
Fink, R. M., T. Enns, C. P. Kimball, H. E. Silberstein, 123.
\V. F. Bale, S. C. Madden, and G. H. Whipple. Plasma
protein metabolism — normal and associated with shock. 124.
Observations using protein labeled by heavy nitrogen
in lysine. J. Exptl. Med. 80: 455-475, 1944-
Fleishman, M., J. Scott, and F. J. Haddy. Effect of
pH change upon systemic large and small vessel resistance.
Circulation Research 5: 602-606, 1957. 125.
Flexner, L. B., D. B. Cowie, and G. J. Vosburgh.
Studies on capillary permeability with tracer substances.
Cold Spring Harbor Symp. Quant. Biol. 13: 88-98, 1948. 126.
Florey', H. Observations on the resolution of stasis in
the finer blood vessels. Proc. Roy. Soc, London B 100:
269-283, 1926. 127.
Forster, R. E. Exchange of gases between alveolar air
and pulmonary capillary blood: pulmonary diffusing
capacity. Physiol. Revs. 37: 391-452, 1957. 128.
Fredrickson, D. S., and R. S. Gordon, Jr. Transport
of fatty acids. Physiol. Revs. 38: 585-630, 1958.
Freed, S. C, and E. Lindner. The effect of steroids of
the adrenal cortex and ovary on capillary permeability. 1 29
Am. J. Physiol. 134: 258-262, 1941.
Freis, E. D., T. F. Higgins, and H. J. Morowitz.
Transcapillary exchange rates of deuterium oxide and
thiocyanate in the forearm of man. ./. Appl. Physiol.
5 : 526-532. '953- '3°-
Friedman, L., and E. O. Kraemer. The structure of
gelatin gels from studies of diffusion. J. Am. Chem. Soc.
52: 1 295-1 304, 1930. 131.
Garby, L. Studies on transfer of matter across mem-
branes with special reference to the isolated human
amniotic membrane and the exchange of amniotic fluid.
Acta Physiol. Scand. 40: Suppl. 137, 1-84, 1957. 132.
Gaskell, P., and A. M. Krisman. An auscultatory tech-
nique for measuring the digital blood pressure. Can. J.
Biochem. and Physiol. 36: 883-888, 1958.
Gaskell, P., and A. M. Krisman. The brachial to digital 133.
blood pressure gradient in normal subjects and in pa-
tients with high blood pressure. Can. J. Biochem. and
Physiol. 36: 889-893, 1958.
Gierer, A. von., and K. Wirtz. Molekulare Theorie
der Mikroreibung. Z. Naturforsch. 8a: 532-538, 1953.
Gitlin, D., and C. A. Janeway. The dynamic equilib-
rium between circulating and extravascular plasma pro-
teins. Science 1 18: 301-302, 1953.
Gitlin, D., H. Latta, W. H. Batchelor, and C. A.
Janeway-. Experimental hypersensitivity in the rabbit.
Disappearance rates of native and labelled heterologous
proteins from the serum after intravenous injection.
J. Immunol. 66: 451-461, 1 95 1.
Glenn, W. W. L., J. Muus, and C. K. Drinker. Ob-
servations on the physiology and biochemistry of quanti-
tative burns. J. Clin. Invest. 22: 451-460, 1943.
Glenn, W. W. L., D. K. Peterson, and C. K. Drinker.
The flow of lymph from burned tissue, with particular
reference to the effects of fibrin formation upon lymph
drainage and composition. Surgery 12: 685-693, 1942.
Goldstein, A. The interactions of drugs and plasma
proteins. Pharmacol. Revs. 1: 102-165, 1949.
Goldstein, D. A., and A. K. Solomon. Determination
of equivalent pore radius for human red cells by osmotic
pressure measurement. J. Gen. Physiol. 44: 1-17, i960.
Gottschalk, C. W. A comparative study of renal in-
terstitial pressure. Am. J. Physiol. 169: 180-187, '952-
Gottschalk, C. W., and M. Mylle. Micropuncture
study of pressures in proximal tubules and peritubular
capillaries of the rat kidney and their relation to ureteral
and renal venous pressures. Am. J. Physiol. 185: 430-
439. '956-
Grim, E. Relation between pressure and concentration
differences across membranes permeable to solute and
solvent. Proc. Soc. Exptl. Biol. Med. 83: 195-200, 1953.
Grotte, G. Passage of dextran molecules across the
blood-lymph barrier. Ada Chir. Scand., Suppl 211: 1-84,
!956-
Gunther, L., H. Engelberg, and L. Strauss. Intra-
muscular pressure. I. During postoperative depression.
Am. J. Med. Sci. 204: 266-270, 1942.
Gunther, L., H. Engelberg, and L. Strauss. Intra-
muscular pressure. II. The venopressor mechanism in
shock-like conditions and the effects of various drugs.
Am. J. Med. Sci. 204: 271-283, 1942.
Gunther, L., L. Strauss, H. H. Henstell, and H.
Engelberg. Intramuscular pressure. III. The action of
various drugs on patients with normal intramuscular
and venous pressure. Am. J. Med. Sci. 204: 387-394,
■942.
Guyton, A. C, G. G. Armstrong, and J. W. Crowell.
Negative pressure in the interstitial spaces. Physiologist
3 (No. 3) : 70, i960.
Guyton, A. G, H. M. Batson, and C. M. Smith. Ad-
justments of the circulatory system following very rapid
transfusion or hemorrhage. Am. J. Physiol. 164: 351-359,
'95'-
Guyton, A. C, and A. W. Lindsey. Effect of elevated
left atrial pressure and decreased plasma protein con-
centration on the development of pulmonary edema.
Circulation Research 7: 649-657, 1959.
Haddy, F. J. Effect of histamine on small and large vessel
1028
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
pressures in the dog foreleg. Am. J. Physiol. i()8: 161-
168, i960.
134. Haddy, F. J., M. Fleishman, and D. A. Emanuel.
Effect of epinephrine, norepinephrine and serotonin
upon systemic small and large vessel resistance. Circu-
lation Research 5: 247-251, 195".
135. Haddy, F. J., M. Fleishman, and J. B. Scott. Effect of
change in air temperature upon systemic small and large
vessel resistance. Circulation Research 5: 58-63, 1957.
136. Haddy, F. J., P. Gordon, and D. A. Emanuel. The
influence of tone upon responses of small and large vessels
to serotonin. Circulation Research 7: 123-130, 1959.
137. Haddy, F. J., A. G. Richards, J. L. Alden, and M. B.
Visscher. Small vein and artery pressures in normal
and edematous extremities of dogs under local and general
anesthesia. Am. J. Physiol. 176: 355-360, 1954.
138. Hahn, L., and G. Hevesy. Rate of penetration of ions
through the capillary wall. Acta Physiol. Scand. 1 : 347-
361, 1940.
139. Hajen, H. Uber die Beziehung des intracutanen Gewebs-
druckes zur Quaddelbildung-Untcrsuchungen uber den
intracutanen Gewebsdruck. Z. ges. exptl. Med. 57 : 203-
213, 1927.
140. Hales, S. Statical Essays: Containing Haemastaticks; or, an
Account of some Hydraulick and Hydrostatical Experiments
made on the Blood and Blood Vessels of Animals. London :
Innys and Manby, 1733, vol. 2.
141. Hall, B. V. The protoplasmic basis of glomerular ultra-
filtration. Am. Heart J. 54: i-g, 1957.
142. Hansen, A. T. An apparatus for rapid measurement of
oncotic pressure in small samples. Physiologist 3 (No. 3) :
74, i960.
i42a.HANSON, K. M., and P. C. Johnson. Evidence for local
arteriovenous reflex in intestine. J. Appl. Physiol. 1 7 : 509-
5'3. >9D2-
143. Hartman, F. A., J. I. Evans, and H. G. Walker.
Control of capillaries of skeletal muscle. Am. J. Physiol.
90:668-688, 1929.
144. Hayes, T. L., and J. E. Hewitt. Visualization of indi-
vidual lipoprotein macromolecules in the electron micro-
scope. J. Appl. Physiol. 11 : 425-428, 1957.
145. Hayman, J. M. Jr. Estimations of afferent arteriole and
glomerular capillary pressures in the frog kidney. Am.
J. Physiol., London 79: 389-409, 1927.
i45a.HEiDENHAiN, R. Versuche und Fragen zur Lehre von
der Lymphbildung. Pfiiigers Arch. ges. Physiol. 49: 209-
301, 1891.
146. Hellems, H. K., F. W. Haynes, and L. Dexter. Pul-
monary 'capillary* pressure in man. J. Appl. Physiol.
2: 24-29, 1949.
147. Hellems, H. K., F. W. Haynes, L. Dexter, and T. D.
Kinney. Pulmonary capillary pressure in animals esti-
mated by venous and arterial catheterization. .4m. J.
Physiol. 155:98-105, 1948.
148. Hendley, E. D., and A. A. Schiller. Change in capillary
permeability during hypoxemic perfusion of rat hind-
legs. Am. J. Physiol. 179: 216-220, 1954.
149. Hendley, E. D., and A. A. Schiller. Protection against
hypoxemic edema by histaminic and adrenergic blockade.
Am. J. Physiol. 180: 378-386, 1955.
150. Henry, J., J. Goodman, and J. Meehan. Capillary
permeability in relation to acute anoxia and to venous
oxygen saturation. J. Clin. Invest. 26: 1119-1129, 1947.
151. Hepp, O. Ein neues Onkometer zur Bestimmung des
kolloidosmotischen Druckes mit gesteigerter Messgenauig-
keit und vereinfachter Handhabung. Z. ges. exptl. Med.
99 : 709-7 1 7. ' 936-
152. Herzog, F. Uber Beziehungen zwischen Dilatation,
Durchlassigkeit und Phagocytose an den Capillaren der
Froschzunge. Virchoiv's Arch, pathol. Anat. 256: 1-8,
I925-
153. Hill, A. V. The diffusion of oxygen and lactic
acid through tissues. Proc. Roy. Soc, London B 104: 39-96,
1928.
154. Hill, A. V. On the time required for diffusion and its
relation to processes in muscle. Proc. Roy. Soc, London B
■35: 446-453. '948-
155. Hinshaw, L. B., and S. B. Day. Tissue pressure and
critical closing pressure in the isolated denervated dog
foreleg. Am. J. Physiol. 196: 489-494, 1959.
156. Hitchcock, D. I. Selected principles of physical chem-
istry. In : Physical Chemistry of Cells and Tissues, edited by
R. Hober. Philadelphia: Blakiston, 1945.
157. Hoff, J. H. van't. Die Rolle des osmotischen Druckes in
der Analogie zwischen Losungen und Gasen. Z. physik.
C/nm. 1 : 481-508, 1887.
158. Holland, G.. and F. Meyer. Der Gewebsdruck beim
Odem. II. Mitteilung. Arch, exptl. Pathol. Pharmakol.
Naunyn-Schmiedeherg's. 168: 603-619, 1932.
159. Hollander, W., P. Reilly, and B. A. Burrows. Lym-
phatic flow in human subjects as indicated by the dis-
appearance of I13l-labelled albumin from the subcu-
taneous tissues. J. Clin. Invest. 40: 222-233, 1961.
160. Hopps, H. C, and J. H. Lewis. Studies on capillary
permeability as affected by anoxemia. Am. J. Pathol.
22 : 656, 1946.
161. Hudack, S., and P. D. McM aster. The gradient of
permeability of the skin vessels as influenced by heat,
cold and light. J. Exptl Med. 55: 431-439, 1932.
162. Hyman, C. Filtration across the vascular wall as a function
of several physical factors. Am. J. Physiol. 142: 671-685,
■944-
163. Hyman, C, and R. Chambers. Effect of adrenal cortical
compounds on edema formation of frogs' hind limbs.
Endocrinology 32: 310-318, 1943.
164. Hyman, C, S. I. Rapaport, A. M. Saul, and M. E.
Morton. Independence of capillary filtration and tissue
clearance. Am. J. Physiol. 168: 674-679, 1952.
165. Hyman, C, S. Rosell, A. Rosen, R. R. Sonnenschein,
and B. Uvnas. Effects of alterations of total muscular
blood flow on local tissue clearance of radio-iodide in the
cat. Acta Physiol. Scand. 46: 358-374, 1959.
166. Irisawa, A., and R. F. Rushmer. Relationship between
lymphatic and venous pressure in leg of dog. Am. J.
Physiol. 196:495-498, 1959.
167. Jepson, R. P., F. A Simeone, and B. M. Dobyns. Re-
moval from skin of plasma protein labeled with radio-
active iodine. Am. ./. Physiol. 175: 443-448, 1953.
168. Johnson, J. A., H. M. Cavert, and N. Lifson. Kinetics
concerned with distribution of isotopic water in isolated
perfused dog heart and skeletal muscle. Am. J. Physiol.
171 : 687-693, 1952.
i68a.JoHNSON, P. C, and K. M. Hanson. Effect of arterial
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IO29
pressure on arterial and venous resistance of intestine.
J. Appl. Physiol. 17: 503-508, 1962.
169. Kedem, O., and A. Katchalskv. A physical interpreta- 187.
tion of the phenomenological coefficients of membrane
permeability. J. Gen. Physiol. 45: 143-179, 196 1.
170. Kedem, O., and A. Katchalskv. Thermodynamic 188.
analysis of the permeability of biological membranes to
non-electrolytes. Biochim. Biophys. Acta 27: 229-246, 1958.
171. Kelly, W. D., and M. B. Visscher. Effect of sympathetic
nerve stimulation on cutaneous small vein and small 189.
artery pressures, blood flow and hindpaw volume in the
dog. Am. J. Physiol. 185: 453-464, 1956.
172. Ketv, S. S. Measurement of regional circulation by the 190.
local clearance of radioactive sodium. Am. Heart J.
38:321-328, 1949. 191.
173. Kety, S. S. The theory and applications of the exchange
of inert gas at the lungs and tissues. Pharmacol. Revs. 3:
'-4'. I951- '92'
174. Kety, S. S. Determinants of tissue oxygen tension.
Federation Proc. 16:666-670, 1957.
175. Keys, A., and R. M. Hill. The osmotic pressure of the 193.
colloids in fish sera. J. Exptl. Biol. 1 1 : 28-34, 1934.
176. Knisely, M. H., E. H. Bloch, T. S. Eliot, and L.
Warner. Sludged blood. Science 106: 431-440, 1947. 194.
177. Koefoed-Johnsen, V., and H. H. Ussing. The contri-
butions of diffusion and flow to the passage of D20
through living membranes; effect of neurohypophyseal 195.
hormone on isolated anuran skin. Acta Physiol. Scand.
28:60-76, 1953.
178. Koniges, H. G., and M. Otto. Studies on the filtration 196.
mechanism of the intestinal lymph and on the action of
acetylcholine on it and on the circulation of the intestinal
wall. Quart. J. Exptl. Physiol. 26: 319-329, 1937.
179. Korner, P. I., and F. C. Gourtice. The effects of acute 197.
anoxia and noradrenaline vasoconstriction on lymph
flow and protein dynamics following transfusions of
Ringer-Locke solution. Australian J. Exptl. Biol. Med. 198.
Set. 32:321-332, 1954.
180. Korner, P. I., B. Morris, and F. C. Courtice. An analy-
sis of factors affecting lymph flow and protein composition 199.
during gastric absorption of food and fluids, and during
intravenous infusion. Australian J. Exptl. Biol. Med. Sci.
32:301-320, 1954. 200.
181. Kramer, K, W. Quensel, and K. E. Schafer. Unter-
suchungen uber den Muskelstoffwechsel des Warmbl liters.
IV. Mitteilung. Beziehungen zwischen Sauerstoffauf-
nahme und Milchsaurcabgabe des Muskels wahrend der 201,
Tatigkeit. Pfliigers Arch. ges. Physiol. 241 : 730-740, 1 939.
182. Kries, N. von. Uber den Druck in den Blutcapillaren
der menschlichen Haut. Arheiten Physiol. Anstalt Leipzig.
10: 69-80, 1875. 202
183. Krogh, A. The number and distribution of capillaries in
muscles with calculations of the oxygen pressure head
necessary for supplying the tissue. J. Physiol., London 203.
52:409-415, 1919.
184. Krogh, A. The Anatomy and Physiology of Capillaries 204,
(rev. ed.). New Haven: Yale Univ. Press, 1929.
185. Krogh, A. The active and passive exchanges of inorganic 205.
ions through the surfaces of living cells and through living
membranes generally. Proc. Roy. Soc, London, B 133: 140-
200, 1946. 206
186. Krogh, A., and G. A. Harrop. On the substance re-
sponsible for capillary tonus. J. Physiol., London 54:
exxv, 1 92 1.
Krogh, A., and G. A. Harrop. Some observations on
stasis and oedema. J. Physiol., London 54: exxv-exxvr,
1921.
Krogh, A., E. M. Landis, and A. H. Turner. The
movement of fluid through the human capillary wall in
relation to venous pressure and to the colloid osmotic
pressure of the blood. J. Clin. Invest. 11: 63-95, '932'
Krogh, A., and F. Nakazawa. Beitrage zur Messung
des kolloid-osmotischen Druckes in biologischen Fliis-
sigkeiten. Bwchem. Z. 188: 241-258, 1927.
Kruh0ffer, P. Inulin as indicator for extracellular
space. Acta Physiol. Scand. 1 1 : 16-36, 1946.
Kruh0ffer, P. The significance of diffusion and con-
vection for distribution of solutes in interstitial space.
Acta Physiol. Scand. 1 1 : 37-47, 1946.
Kuhn, W. Grenze der Durchlassigkeit von Filtrier- und
Loslichkeitsmembranen. Z. Elektrochem. 55: 207-217)
195'-
Ladenburg, R. Uber den Einfluss von Wanden auf die
Bewegung einer Kugel in einer reibenden Fliissigkeit.
Ann. Physik. 23: 447-458, 1907.
Lambert, P. P., and F. Gregoire. Hemodynamique
glomerulaire et excretion de I'hemoglobine. Arch, intern,
physiol. 63: 7-34, 1955.
Lambert, P. P., F. Gregoire, and C. de H. de Brau-
court. Hemodynamique glomerulaire et excretion de
I'hemoglobine. Arch, intern, physiol. 60: 506-534, 1952.
Lambert, P. P., F. Gregoire, C. Malmendier, F.
Vanderveiken, and G. Gueritte. Recherches sur le
mecanisme de l'albuminurie. Bull. Acad. Roy. Med. Belg.
22:524-602, 1957.
Landerer, A. S. Die Gewebsspannung in ihrem Einfluss
auf die brtliche Blul- und Lymphhewegung. Leipzig: Vogel,
1884.
Landis, E. M. The capillary pressure in frog mesentery
as determined by micro-injection. Am. J. Physiol. 75:
548-570, 1926.
Landis, E. M. Micro-injection studies of capillary
permeability. I. Factors in the production of capillary
stasis. Am. J. Physiol. 81: 124-142, 1927.
Landis, E. M. Micro-injection studies of capillary per-
meability. II. The relation between capillary pressure
and the rate at which fluid passes through the walls of
single capillaries. Am. J. Physiol. 82: 217-238, 1927.
Landis, E. M. Micro-injection studies of capillary per-
meability. III. The effect of lack of oxygen on the perme-
ability of the capillary wall to fluid and to the plasma
proteins. Am. J. Physiol. 83: 528-542, 1928.
Landis, E. M. The capillary blood pressure in mam-
malian mesentery as determined by the micro-injection
method. Am. J. Physiol. 93: 353-362, 1930.
Landis, E. M. Micro-injection studies of capillary blood
pressure in human skin. Heart 15: 209-228, 1930.
Landis, E. M. Micro-injection studies of capillary blood
pressure in Raynaud's disease. Heart 15: 247-255, 1930.
Landis, E. M. Capillary pressure and hyperemia in
muscle and skin of the frog. Am. J. Physiol. 98: 704-716,
'93'-
Landis, E. M. Poiseuille's law and the capillary circula-
tion. Am. J. Physiol. 103: 432-443, 1933.
1030
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
207. Landis, E. M. Capillary pressure and capillary per-
meability. Physiol. Revs. 14: 404-481, 1934.
208. Landis, E. M. Capillary permeability and the factors
affecting the composition of capillary filtrate. Ann. .VI".
Acad. Sci. 46: 713-731, 1946.
209. Landis, E. M., and J. H. Gibbon, Jr. The effects of
temperature and of tissue pressure on the movement of
fluid through the human capillary wall. J. Clin. Invest.
12: 105-138, 1933.
210. Landis, E. M., and J. C. Hortenstine. Functional sig-
nificance of venous blood pressure. Physiol. Revs. 30:
i-32, '95°-
211. Landis, E. M., L. Jonas, M. Angevine, and W. Erb.
The passage of fluid and protein through the human
capillary wall during venous congestion. J. Clin. Invest.
11: 7 '7-734. '93'2-
212. Lawrence, J. H., W. F. Loomis, C. A. Tobias, and
F. H. Turpin. Preliminary observations on the narcotic
effect of xenon with a review of values for solubilities of
gases in water and oils. ./. Physiol., London 105: 197-204,
1946.
213. Lazarus-Barlow, W. S. The pathology of the oedema
which accompanies passive congestion. Phil. Trans. Roy.
Soc, London B 185: 779-817, 1894.
214. Lee, J. S., and M. B. Visscher. Microscopic studies of
skin blood vessels in relation to sympathetic nerve stimu-
lation. Am. J. Physiol. 1 go: 37-40, 1957.
215. Lewis, J. H. The route and rate of absorption of sub-
cutaneously injected serum in relation to the occurrence
of sudden death after injection of antitoxic horse serum.
J. Am. Med. Assoc. 76: 1 342-1 345, 1921.
216. Lewis, T. Vascular reactions of the skin to injury. Part I.
Reaction to stroking; urticaria factitia. Heart 1 1 : 1 19-137,
1924.
217. Lewis, T. Blood Vessels of the Human Skin and Their Re-
sponses. London: Shaw, 1927.
218. Lewis, T. Swelling of the human limbs in response to
immersion in cold water. Clin. Sci. 4: 349-360, 1942.
219. Lewis, T., and R. T. Grant. Vascular reactions of the
skin to injury. Part II. The liberation of a histamine-like
substance in injured skin ; the underlying cause of factitious
urticaria and of wheals produced by burning; and ob-
servations upon the nervous control of certain skin re-
actions. Heart 1 1 : 209-265, 1924.
220. Lewis, T., and E. M. Landis. Observations upon the
vascular mechanism in acrocyanosis. Heart 15: 229-246,
1930.
221. Lucre, B., H. K. Hartline, and M. McCutcheon.
Further studies on the kinetics of osmosis in living cells.
J. Gen. Physiol. 14:405-419, 1 93 1 .
222. Lucre, B, and M. McCutcheon. The living cell as an
osmotic system and its permeability to water. Physiol.
Revs. 1 -' : 68-139, 1932.
223. Ludwig, C. F. W. Lehrbuch iter Physiologic des Menschen.
2. Aufl. Leipzig: Winter, 1 858-1861 , vol. 2, p. 562.
224. Lundsgaard, E. Effect of phloridzin on isolated kidney
and isolated liver. Skand. Arch. Physiol. 72: 265-270, 1935.
225. MacLeod, M. Systemic capillary pressure in acute
glomerulonephritis estimated by direct micropuncture.
Clin. Sci. 19: 27-33, ]96o.
226. Majno, G., and G. E. Palade. Studies on inflammation.
I. The effect of histamine and serotonin on vascular per-
226a
227.
229.
230.
231.
233-
234-
235-
236.
237-
238.
239-
240.
241.
242.
243-
meability: An electron microscopic study. J. Biophys. Bio-
chem. Cylol. 11: 571-605, 1 961.
Majno, G., G. E. Palade, and G. I. Schoefl. Studies
on inflammation. II. The site of action of histamine and
serotonin along the vascular tree: A topographic study.
J. Biophys. Biochem. Cytol. 11 : 607-626, 1961.
Martin, E. G., E. C. Woolley, and M. Miller. Capil-
lary counts in resting and active muscles, ,4m. J. Physiol.
100: 407-416, 1932.
Maurer, F. W. The effects of decreased blood oxygen
and increased blood carbon dioxide on the flow and
composition of cervical and cardiac lymph. Am. J. Physiol.
J3I: 33 '-348, 194°-
Maurer, F. W. The effects of carbon monoxide anoxemia
on the flow and composition of cervical lymph. Am. J.
Physiol. 133: 170-179, 1 941.
Mauro, A. Some properties of ionic and non-ionic
semipermeable membranes. Circulation 21 : 845-858, i960.
Maverson, H. S., and G. E. Burch. Relationships of
tissue (subcutaneous and intramuscular) and venous
pressures to syncope induced in man by gravity. Am. J.
Physiol. 128: 258-269, 1940.
Maverson, H. S., C. G. Wolfram, H. H. Shirley,
Jr., and K. Wasserman. Regional differences in capillary
permeability. Am. ./. Physiol. 198: 155-160, i960.
McBain, J. W., and T. H. Liu. Diffusion of electrolytes,
non-electrolytes and colloidal electrolytes. J. Am. Chem.
Soc. 53:59-74, 1 93 1.
McLennan, C. E., M. T. McLennan, and E. M. Landis.
The effect of external pressure on the vascular volume of
the forearm and its relation to capillary blood pressure
and venous pressure. J. Clin. Invest. 21 : 319-338, 1942.
McMaster, P. D. Intermittent take-up of fluid from
the cutaneous tissue. J. Exptl. Med. 73: 67-84, 1941.
McMaster, P. D. Factors influencing the intermittent
passage of Locke's solution into living skin. J. Exptl.
Med. 73: 85-108, 1941.
McMaster, P. D. An inquiry into the structural con-
ditions affecting fluid transport in the interstitial tissue
of the skin. J. Exptl. Med. 74: 9-28, 1941.
McMaster, P. D. The pressure and interstitial resistance
prevailing in the normal and edematous skin of animals
and man. J. Exptl. Med. 84: 473-494, 1946.
McMaster, P. D. The effects of venous obstruction
upon interstitial pressure in animal and human skin.
J. Exptl. Med. 84: 495-509, 1946.
McMaster, P. D , and R. J. Parsons. Physiological
conditions existing in connective tissue. I. The method of
interstitial spread of vital dyes. J. Exptl. Med. 69: 247-
264, 1939.
McMaster, P. D., and R. J. Parsons. Physiological
conditions existing in connective tissue. II. The state of
the fluid in the intradermal tissue. J. Exptl. Med. 69:
265-282, 1939.
McMichael, J., and K. M. Morris. Acute oxygen lack
and capillary permeability in man. J. Physiol., London
87: 74 P, 1936.
Mellander, S. Comparative studies on the adrenergic
neuro-hormonal control of resistance and capacitance
blood vessels in the cat. Acta Physiol. Scand. 50: Suppl.
176, 1-86, i960.
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
IO3 I
244. Mendlowitz, M. Some observations on clubbed fingers.
Clin. Set. 3: 387-401, 1938.
245. Menkin, V. Effect of adrenal cortex extract on capillary
permeability. Am. J. Physiol. 129: 691-697, 1940.
246. Menkin, V. Dynamics of Inflammation — An Inquiry into the
Mechanism of Infectious Processes. New York: Macmillan,
1940.
247. Menkin, V. Biochemical Mechanisms in Inflammation (2nd
ed.). Springfield, 111. : Thomas, 1956.
248. Meschia, G. A rigid membrane for measurement of
colloidal osmotic pressure with the Hepp osmometer.
Yale J. Biol, and Med. 27: 206-212, 1954.
249. Meschia, G. Colloidal osmotic pressures of fetal and
maternal plasmas of sheep and goats. Am. J. Physiol.
181 : 1-8, 1955.
250. Meschia, G., and I. Setnikar. Experimental study of
osmosis through a collodion membrane. J. Gen. Physiol.
42:429-444, 1958.
251. Meyer, P. Der kolloidosmotische Druck biologischer
Fliissigkeiten. Ergeb. Physiol. 34: 1 8-1 11, 1932.
252. Meyer, F., and G. Holland. Die Messung des Druckes
in Geweben. I. Mitteilung. Arch. Exptl. Pharmakol. Pathol.
168: 580-602, 1932.
253. Millikan, G. A. Experiments on muscle haemoglobin
in vivo; the instantaneous measurement of muscle me-
tabolism. Proc. Roy. Soc, London B 123: 218-241, 1937.
254. Monke, J. V., and C. L. Yuile. The renal clearance of
hemoglobin in the dog. J. Exptl. Med. 72: 149-165, 1940.
255. Moore, D. H., and H. Ruska. The fine structure of
capillaries and small arteries. J. Biophys. Biochem. Cytol.
3:457-462, 1957.
256. Morales, M. F., and R. E. Smith. The physiological
factors which govern inert gas exchange. Bull. Math.
Biophys. 7:99-106, 1945.
257. Morel, F. F. Techniques de la mesure des echanges
capillaires a l'aide des indicateurs radioactifs. Helvet.
Physiol, el Pharmacol. Acta, 8: 52-73, 1950.
258. Muller, A. Bemerkungen zum Gasaustausch in den
Lungen. Helvet. Physiol, el Pharmacol. Acta 3: 203-213,
■945-
259. Myant, N. B. Observations on the metabolism of human
gamma globulin labelled by radioactive iodine. Clin.
Sci. 1 1 : 191-201, 1952.
260. Nernst, W. Zur Kinetik der in Losung befindlichen
Korper. Z. Physik. Chem. 2: 613-637, 1888.
261. Niesel, W., G. Thews, and D. Lubbers. Die Messung
des zeitlichen Verlaufes der O, — Aufsattigung und
Entsattigung menschlicher Erythrocyten mit dem
Kurzzeit-Spektralanalysator. Pfliigers Arch. ges. Physiol.
268 : 296-307, 1 959.
262. Nisimaru, Y. Studies Concerning the Physiological Behavior
of Blood Capillaries. Tokyo: Igakushoin, 1955.
263. Noll, F. Ueber den Lymphstrom in den Lymphgefassen
und die wesentlichsten anatomischen Bestandtheile der
Lymphdriisen. Z. Rat. Med. 4: 52-93, 1850.
264. Northrop, J. H., and M. L. Anson. A method for the
determination of diffusion constants and the calculation
of the radius and weight of the hemoglobin molecule.
J. Gen. Physiol. 12: 543-554. I929-
265. Oeff, K., and A. Konig. Lokale Kapillarpermcabilitat
und austauschbares Albumin in verschiedenen Organen
der Ratte. Experientia 12: 260-261, 1956.
266. Ogston, A. G., and T. F. Sherman. Effects of hyaluronic
acid upon diffusion of solutes and flow of solvent. J.
Physiol ., London 156:67-74, 1961.
267. Oncley, J. L. Plasma proteins and plasma fractionation.
In Hormones in Plasma, edited by H. N. Antoniades.
Boston: Little, Brown, i960, chapt. 11.
268. Oncley, J. L., G. Scatchard, and A. Brown. Physico-
chemical characteristics of certain of the proteins of
normal human plasma. J. Phys. &! Colloid. Chem. 51 :
184-198, 1947.
269. Optiz, E., and M. Schneider. Uber die Sauerstoff-
versorgung des Gehirns und der Mechanismus von
Mangelwirkungen. Ergeb. Physiol. 46: 126-260, 1950.
270. Ott, H. Die Errechung des kolloidosmotischen Serum-
druckes aus dem Eiweiss-Spektrum und das mittlere
Molekulargewicht der Serumeiweissfraktionen. Klin.
Wochschr. 34: 1079-1083, 1956.
271. Ott, H. Das Blutserum bei Analbuminamie. Z. ges. exptl.
Med. 128:340-360, 1957.
272. Paff, G. H. A quantitative study of the capillary supply-
in certain mammalian skeletal muscles. Anal. Record
46:401-406, 1930.
273. Palade, G. E. The endoplasmic reticulum. J. Biophys.
Biochem. Cytol. 2(Suppl.): 85-98, 1956.
274- Pappenheimer, J. R. Vasoconstrictor nerves and oxygen
consumption in the isolated perfused hindlimb muscles
of the dog. J. Physiol., London gg: 182-200, 1941.
275. Pappenheimer, J. R. Blood flow, arterial oxygen con-
sumption in the isolated perfused hind!imb of the dog.
J. Physiol., London 99: 283-303, 1941.
276. Pappenheimer, J. R. Passage of molecules through
capillary walls. Physiol. Revs. 33: 387-423, 1953.
277. Pappenheimer, J. R. Ultrafiltration and diffusion through
biological membranes. Annua/ Lecture, No. r, Bethesda,
Md. : Nat. Insts. Health, ig54-
278. Pappenheimer, J. R. Uber die Permeabilitat der Glom-
erulummembranen in der Niere. Klin. Wochschr. 33:
362-365, ig55.
279. Pappenheimer, J. R., S. R. Heisey, and E. F. Jordan.
Active transport of Diodrast and phenolsulfonphthalein
from cerebrospinal fluid to blood. Am. ./. Physiol. 200:
1-10, 1961.
280. Pappenheimer, J. R., and E. C. C. Lin. The rapid
measurement and recording of osmotic prersure. Science
Il8: 574. '953-
281. Pappenheimer, J. R., E. M. Renkin, and L. M. Borrero.
Filtration, diffusion and molecular sieving through
peripheral capillary membranes. A contribution to the
pore theory of capillary permeability. Am. J. Physiol.
167: 13-46, 1951.
282. Pappenheimer, J. R., and A. Soto-Rivera. Effective
osmotic pressure of the plasma proteins and other quanti-
ties associated with the capillary circulation in the hind-
limbs of cats and dogs. Am. J. Physiol. 152: 471-491,
1948.
283. Parsons, R. J., and P. D. McMaster. The effect of the
pulse upon the formation and flow of lymph. J. Exptl.
Med. 68: 353-376, 1938.
283a.PEDERSEN, K. O. Svedberg Memorial Volume. Stockholm:
Almqvist-YViksells Boktrycheri AB ig44, pp. 4go-4gg.
284. Perry, H. I. Vital injection as a method for the study
I032
II\M>K«l(.)K OF PHYSIOLOGY
CIRCULATION II
of capillary circulation. Skand. Arch. Physiol. 59: 67-74, 306.
[930.
285. Pochin, E. E. Oedema following ischaemia in the rabbit's
ear. Clin. Sci. 4: 341-347, 1942.
286. Poiseuille, J. L. M. Rechercha sur la Force du Coew Aortiqut
(Thesis). Paris: 1828.
287. Poiseuille, J. L. M. Recherches sur Us Causes du Mouvement 307.
du Sang dam les Vaisseaux Capillaries. Paris: 1835.
288. Poiseuille, J. L M. Recherches experimentales sur le 308.
mouvement des liquids dans les tubes de ties petits
diametres. Compt. rend. acad. sci. 11 : 961-967; 1041-1048,
1840. 309.
289. Poiseuille, J. L. M. Sur la pression du sang dans le
systeme arteriel. Gaz.. hebd. med. et chir. 7 : 563-565,
1 860. 3 1 o.
290. Prentice, T. C, R. R. Stahl, N. A. Dial, and F. V.
Ponterio. A study of the relationship between radio-
active sodium clearance and directly measured blood
flow in the biceps muscle of the dog. J. Clin. Invest. 34: 311.
545-558. '955-
291. Rapaport, E., and L. Dexter. Pulmonary 'capillary'
pressure. Methods in Medical Research 7: 85-93, '958- 312.
292. Ray, P. M. On the theory of osmotic water movement.
Plant Physiol. 35: 783-795, 1960.
293. Reid, E. VV. Osmotic pressure of solutions of haemo-
globin. J. Physiol., London 33: 12-19, '9°5-
294. Rein, H., and M. Schneider. Die lokale Stoffwech- 313.
seleinschrankung bei reflektorischnervoser Durchblu-
tungsdrosselung. Pfliigers Arch. ges. Physiol. 239: 464-
475. '937-
295. Renkin, E. M. Studies on the Permeability of the Capillaries in
Mammalian Muscle (Thesis). Cambridge, Mass.: Harvard 314.
Univ., 1951 .
296. Renkin, E. M. Capillary permeability to lipid-soluble
molecules. Am. J. Physiol. 168: 538-545, 1952.
297. Renkin, E. M. Capillary and cellular permeability to
some compounds related to antipyrine. Am. J. Physiol. 315.
'73: > 25-1 3°. '953-
298. Renkin, E. M. Filtration, diffusion, and molecular sieving
through porous cellulose membranes. J. Gen. Physiol. 316.
38:225-243, 1954.
299. Renkin, E. M. Effects of blood flow on diffusion kinetics 317.
in isolated, perfused hindlegs of cats. A double circulation
hypothesis. Am. J. Physiol. 183: 125-136, 1955. 318.
300. Renkin, E. M. Transport of potassium-42 from blood to
tissue in isolated mammalian skeletal muscles. Am. ./.
Physiol. 197: 1 205-1 2 10, 1959. 319.
301 . Renkin, E. M., and J. R. Pappenheimer. Wasserdurchlass-
igkeit und Permeabilitat der Capillarwande. Ergeb. 320.
Physiol. 49:59-126, 1957.
302. Renkin, E. M., and B. D. Zaun. Effects of adrenal hor-
mones on capillary permeability in perfused rat tissues. 321.
Am. J. Physiol. 180: 498-502, 1955.
303. Robbins, E., and A. Mauro. Experimental study of the 322.
independence of diffusion and hydrodynamic permeability
coefficients in collodion membranes. J. Gen. Physiol.
43 ; 523-532. ■g60- 123
304. Robbins, J., and J. E. Rall. Proteins associated with
the thyroid hormones. Physiol. Revs. 40: 415-489, i960.
305. Roberts, J. T., and J. T. VVearn. Quantitative changes 324.
in the capillary-muscle relationship in human hearts
during normal growth and hypertrophy. Am. Heart ./.
21:617 633. '94'-
Roughton, F. J. W., and R. E. Forster. Relative
importance of diffusion and chemical reaction rates in
determining rate of exchange of gases in the human lung,
with special reference to true diffusing capacity of pul-
monary membrane and volume of blood in the lung
capillaries. J. Appl. Physiol. II: 290-302, 1957.
Rous, P., H. P. Gilding, and F. Smith. The gradient of
vascular permeability. J. Exptl. Med. 51 : 807-830, 1930.
Rous, P., and F. Smith. The gradient of vascular per-
meability. III. The gradient along the capillaries and
venules of frog skin. J. Exptl. Med. 53: 219-242, 1 93 1.
Roy, C. S., and J. G. Brown. The blood-pressure and
its variations in the arterioles, capillaries and smaller
veins. J. Physiol., London 2: 323-359, 1880.
Sangren, W. O, and C. \V. Siieppard. A mathematical
derivation of the exchange of a labeled substance between
a liquid flowing in a vessel and an external compartment.
Bull. Math. Btophys. 15: 387-394, 1953.
Sapirstein, L. A. Regional blood flow by fractional
distribution of indicators. Am. J. Physiol. 193: 161-168,
'958-
Scatchard, C, A. C. Batchelder, and A. Brown.
Chemical, clinical and immunological studies on the
products of human plasma fractionation. VI. The osmotic
pressure of plasma and of serum albumin. J. Clin. Invest.
23 : 458-464. '944-
Scatchard, C, A. C. Batchelder, and A. Brown.
Preparation and properties of serum and plasma pro-
teins. VI. Osmotic equilibria in solutions of serum albumin
and sodium chloride. J. Am. Chan. Soc. 68: 2320-2329,
1946.
Scatchard, G., A. Gee, and J. Weeks. Physical chem-
istry of protein solutions. VI. The osmotic pressures of
mixtures of human serum albumin and 7 -globulins in
aqueous sodium chloride. J. Phys. Chem. 58: 783-787,
■954-
Scatchard, G., I. H. Sciieinberg, and S. H. Armstrong,
Jr. The combination of human serum albumin with
chloride ions. J. Am. Chem. Soc. 72: 535-540, 1950.
Scherp, 11. W. The diffusion coefficient of crystalline
trypsin. J. Gen. Physiol. 16: 795-800, 1933.
Schlogl, R. Zur Theorie der anomalen Osmose. Z.
physik. Chem. 3: 73-102, 1955.
Schmidt, G. W. A mathematical theory of capillary
exchange as a function of tissue structure. Bull. Math.
Biophys. 14: 229-264, 1952.
Schmidt, G. W. The time course of capillary exchange.
Bull. Math. Biophys. 15: 477-488, 1953.
Schmidt-Nielsen, K., and P. Pennvcuik. Capillary
density in mammals in relation to body size and oxygen
consumption. Am. J. Physiol. 200: 746-750, 1961.
Scholander, P. F. Oxygen transport through hemoglobin
solutions. Science 131 : 585-590, i960.
Scholander, P. F., L. Irving, and S. W. Grinnell.
Aerobic and anaerobic changes in seal muscles during
diving. J. Biol. Chem. 142: 431-440, 1942.
Scholander, P. F., L. Irving, and S. W. Grinnell.
On the temperature of the seal during diving. J. Cellular
Comp. Physiol. 19: 67-78, 1942.
Schroeder, W. Methodik der fortlaufcnden Messung
des Venen-, Kapillar- odcr Arteriolendruckes in der
vorderen Extremitat des wachen Hundes. Z. Biol. 103:
389-394. '95°-
EXCHANGE OF SUBSTANCES THROUGH CAPILLARY WALLS
I033
325. Schroeder, \V., and H. F. Anschutz. Die VVirkung
von Azetycholin, Adrenalin, und Histamin auf die
Durchblutung der Kapillaren und arteriovenosen 344.
Anastomosen in der vorderen Extrcmitat des Hundes.
Z. Biol. 103: 395-408, 1950. 345-
326. Schroeder, \V., F. Gersmeyer, and H. Freund. Die
Bedeutung der Capillardruckmessung fur die Beurteilung
der Wirkung sbg. capillarabdichtender Substanzen. 346.
Arch, exptl. Pathol. Pkarmakol. Naunyn-Schmiedeberg's 228:
566~575> !956-
327. Schroeder, \V., W. Schoop, and E. Stein. Die Durchblu- 347.
tung der Extrcmitat im akutcn Sauerstoffmangel unter
besonderer Beriicksichtigung der Funktion der arterio- 348.
venosen Anastomosen. Pfliigers Arch. ges. Physiol. 259:
124-141, 1954.
328. Shadle, O. VV., M. Zukof, and J. Diana. Transloca- 349.
tion of blood from the isolated dog's hindlimb during
levarterenol infusion and sciatic nerve stimulation. 350.
Circulation Research 6 : 326-333, 1 958.
329. Shapiro, H., and A. K. Parpart. The osmotic properties
of rabbit and human leucocytes. J. Cellular Comp. Physiol. 351.
10: 147-160, 1937.
330. Sheppard, C. W., and A. S. Householder. The mathe-
matical basis of the interpretation of tracer experiments
in closed steady-state systems. J. Appl. Physiol. 22: 510- 352.
52°> '951-
331. Shirley, H. H., Jr., C. G. Wolfram, K. Wasserman,
and H. S. Mayerson. Capillary permeability to macro- 353.
molecules: stretched pore phenomenon. Am. J. Physiol.
190. 189-193, 1957. 354-
332. Shleser, I. H., and S. C. Freed. The effect of peptone
on capillary permeability and its neutralization by adrenal
cortical extract. Am. J. Physiol. 137: 426-430, 1942. 355-
333. Shuler, K. E., C. A. Dames, and K. J. Laidi.er. The
kinetics of membrane processes. III. The diffusion of
various nonelectrolytes through collodion membranes. 35^.
J. Chem. Phys. 17: 860-865, '949-
334. Sidel, V. W., and A. K. Solomon. Entrance of water
into human red cells under an osmotic pressure gradient. 357-
J. Gen. Physiol. 41 : 243-257, 1957.
335. Sjostrand, T. On the principles for the distribution of
blood in the peripheral vascular system. Skand. Arch.
Physiol. 71 : Suppl. 5, 1-150, 1935. 358.
336. Smirk, F. H. Observations on the causes of oedema in
congestive heart failure. Clin. Sci. 2: 317-335, '936.
337. Smith, F., and M. Dick. The influence of the plasma 359.
colloids on the gradient of capillary permeability- J.
Exptl. Med. 56: 371-389, 1932. 360.
338. Smith, F., and P. Rous. The gradient of vascular per-
meability. IV. The permeability of the cutaneous venules
and its functional significance. ./. Exptl. Med. 54: 499- 361.
5'4. '931-
339. Smith, H. W. The Kidney. Structure and Function in Health
and Disease. New York: Oxford, 1951, chapt. xvm. 362.
340. Sodeman, W. A., and G. E. Burch. The tissue pressure
in subcutaneous edema. -4m. J. Med. Sci. 194: 846-850,
'937- 363-
341. Solomon, A. K. Equations for tracer experiments. ./.
Clin. Invest. 28: 1 297-1 307, 1949.
342. S0rensen, S. P. L. Studies on proteins. V. On the osmotic 364.
pressure of egg-albumin solutions. Compt. rend. trav. lab.
Carlsberg. 12:262-372, 1917.
343. Soto-Rivera, A. Relationship between protein osmotic 365.
pressure and density in plasma from cats, dogs and
humans. Proc. Sue. Exptl. Biol. Med. 71 : 184-186, 1949.
Spector, W. G. Substances which affect capillary per-
meability. Pharmacol. Revs. 10: 475-505, 1958.
Starling, E. H. On the absorption of fluids from the
connective tissue spaces. J. Physiol., London 19: 312-326,
1896.
Starling, E. H. Production and absorption of lymph.
In: Textbook of Physiology, edited by E. A. Schafer. New-
York: Macmillan, 1898, vol. 1, p. 296.
Starling, E. H. The glomerular functions of the kidney.
J. Physiol., London 24: 317-330, 1899.
Starling, E. H., and E. B. Verney. The secretion of
urine as studied on the isolated kidney. Proc. Roy. Soc,
London B 97: 321-363, 1925.
St a verm an, A. J. The theory of measurement of osmotic
pressure. Rec. trav. chim. 70: 344-352, 1951.
Staverman, A. J. Apparent osmotic pressure of solutions
of heterodisperse polymers. Rec. trav. chim. 71 : 623-633,
■952-
Stead, E. A., Jr., and J. V. Warren. The protein con-
tent of the extracellular fluid in normal subjects after
venous congestion and in patients with cardiac failure,
anoxemia and fever. J. Clin. Invest. 23: 283-287, 1944.
Sterling, K. The turnover rate of serum albumin in
man as measured by I131-tagged albumin. J. Clin. Invest.
30: 1 228-1 237, 1 95 1.
Stoel, G. Uber die Blutversorgung von weissen und
roten Kaninchenmuskeln. Z. Zellforsch. 3: 91-98, 1925.
Sugarman, J., M. Friedman, E. Barrett, and T. Addis.
The distribution, flow, protein and urea content of renal
lymph, ,4m. J. Physiol. 138: 108-112, 1942.
Sutherland, W. A dynamical theory of diffusion for
non-electrolytes and the molecular mass of albumin.
Phil. Mag. 9: 781-785, 1905.
Thews, G. Untersuchung der Sauerstoffaufnahme und
-abgabe sehr diinncr Blutlamellen. Pfliigers Arch. ges.
Physiol. 268: 308-317, 1959.
Tschirgi, R. D. Chemical environment of the central
nervous system. In: Handbook of Physiology. Washington,
D.C. : Am. Physiol. Soc, i960, Sect. 1, Vol. 111, pp. 1865-
1890.
Valdivia, E. Total capillary bed in striated muscle of
guinea pigs native to the Peruvian mountains. Am. J.
Physiol. 194:585-589, 1958.
Verzar, F. Der Gaswechsel des Muskels. Ergeb. Physiol.
15: 1-101, 1 9 1 6.
Vimtrup, B. J. On the number, shape, structure, and
surface area of the glomeruli in the kidneys of man and
mammals. -4m. J. Anat. 41 : 123-151, 1928.
Visscher, M. B., F. J. Haddy, and G. Stephens. The
physiology and pharmacology of lung edema. Pharmacol.
Revs. 8:389-434, 1956.
Walder, D. N. The relationship between blood flow,
capillary surface area and sodium clearance in muscle.
Clm. Sci. 14: 3°3-3>5. '955-
Walker, W. G., and W. S. Wilde. Kinetics of radio-
potassium in the circulation. Am. J. Physiol. 170: 401-
4'3. >952-
Wallace, J. M., and E. A. Stead, Jr. Spontaneous
pressure elevations in small veins and effects of norepi-
nephrine and cold. Circulation Research 5: 650-656, 1957.
Wallace, J. M., and E. A. Stead. Fall in pressure in
i°34
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
radial artery during reactive hyperemia. Circulation
Research 7 : 876-79, 1959.
366. Wallenius, G. Renal clearance of dextran as measure
of glomerular permeability. Ada Soc. Med. Upsallen 59:
Suppl. 4, 1-91, 1954.
367. Warren, M. F., and C. K. Drinker. The flow of lymph
from the lungs of the dog. Am. ./. Physiol. 136: 207-221,
1942-
368. Wasserman, K., J. D. Joseph, and H. S. Mayerson.
Kinetics of vascular and extravascular protein exchange
in unbled and bled dogs. Am. .1 . Physiol. 184: 175-182,
'956-
369. Wasserman, K.., L. Loeb, and H. S. Mayerson. Capil
lary permeability to macromolecules. Circulation Re-
search 3: 594-603, 1955.
370. Wasserman, K.., and H. S. Mayerson. Exchange of
albumin between plasma and lymph. .4m. J. Physiol.
165: 15-26, 1951.
371. Wasserman, K., and H. S. Mayerson. Mechanism of
plasma protein changes following saline infusions. Am.
J. Physiol. 170: 1-10, 1952-
372. Wasserman, K., and H. S. Mayerson. Dynamics of
lymph and plasma protein exchange. Cardiologia 21:
296-307, 1952.
373. Webb, R. C, Jr., and T. E. Starzl. The effect of blood
vessel pulsations on lymph pressure in large lymphatics.
Bull. Johns Hopkins Hasp. 93: 401-407, 1953.
374. Weech, A. A., E. Goettsch, and E. B. Reeves. The
flow and composition of lymph in relation to the forma-
tion of edema. J. Exptl. Med. 60: 63-84, 1934.
375. Wells, H. S., J. B. Youmans, and D. G. Miller, Jr.
Tissue pressure (intracutaneous, subcutaneous, and
intramuscular) as related to venous pressure, capillary
filtration, and other factors. J. Clin. Invest. 1 7 : 489-499,
I938-
376. Whipple, G. H., and S. C. Madden. Hemoglobin,
plasma protein and cell protein — their interchange and
construction in emergencies. Medicine 23: 215-224,
1944.
377. White, H. L. Observations on the nature of glomerular
activity. Am. J. Physiol. 90: 689-704, 1929.
378. White, 11. L. Measurement of cardiac output by a
continuously recording conductivity method. Am. J.
Physiol. 151:45-5/. >947-
379. White, J. C, M. E. Field, and C. K. Drinker. On the
protein content and normal flow of lymph from the foot
of the dog. Am. J. Physiol. 103: 34-44, 1933.
380. Wies, C. H., and J. P. Peters. The osmotic pressure of
proteins in whole serum. J. Clin. Invest. 16: 93-102,
■937-
381. Wilbrandt, W., E. Luscher, and H. Asper. Der Einfluss
von Thrombocytenprotein auf die Permeabilitat der
Blutkapillaren. Helvet. Physiol, et Pharmacol. Acta 14: C81-
84, 1956.
382. Wilde, W. S. Transport through biological membranes.
Ann. Rev. Physiol. 17: 17-36, 1955.
383. Wind, F. Versuche zur unmittelbaren Bestimmung des
Flussigkeitsaustritts aus den Blutkapillaren des Mesen-
terium und des Nierenglomerulus beim Kaltbliiter. I.
Mitteilung. Arch, exptl. Pathol. Pharmakol. Naunyn-
Schmiedeberg's 186: 161-184, 1937.
384. Winton, F. R. Physical factors involved in the activities
of the mammalian kidney. Physiol. Revs. 1 7 : 408-435,
1937-
385. Wirz, H. Druckmessung in Kapillaren und Tubuli der
Niere durch Mikropunktion. Helvet. Physiol, et Pharmacol.
Acta 13:42-49, 1955.
386. Yoffey, J. M., and F. C. Courtice. Lymphatics, Lymph
and Lymphoid Tissues (2nd ed.). Cambridge, Mass.:
Harvard Univ. Press 1956, pp. 87, 238.
387. Zweifach, B. W. The structural basis of permeability
and other functions of blood capillaries. Symposia Qiiant.
Biol. 8: 216-223, 1940.
388. Zweifach, B. W., and D. B. Metz. Selective distribu-
tion of blood through the terminal vascular bed of
mesenteric structures and skeletal muscle. Angiology
6: 282-290, 1955.
CHAPTER 30
The physiologic importance of lymph1
H. S. MAYERS OX
Department of Physiology, Tulane University
School of Medicine, New Orleans, Louisiana
CHAPTER CONTENTS
Methods of Study
Development and Structure of Lymphatic Vessels
Lymph vs. Tissue Fluid
Distribution of Lymphatic Vessels
General Anatomic Arrangement of Main Trunks
Contractility of Lymphatics
Exchange of Substances Between Plasma and Lymph
Extravascular Pool and Circulation of Protein
Lymphatic Return and Blood Volume Regulation
Transport Function
Lipids
Enzymes
Coagulation Principles
Iron
Miscellaneous
Significance of Some Regional Lymphatics
Thoracic Duct
Hepatic Lymph
Pulmonary Lymphatics and Edema
Cardiac
Renal
Lymph and Lymphatics in Shock
Anaphylactic Shock
Traumatic Shock
Burns
Permeability Factors
Permeability of Lymphatic Vessels
from a physiologic point of view, the lymphatic
system is primarily a drainage system. Its need arose
phylogenetically with the development of a high
pressure circulation. The latter development, de-
signed to insure an adequate supply of oxygen to
1 The work described as emanating from this laboratory was
supported by grants from the Research and Development
Command, U. S. Army, the American Heart Association, and
the U. S. Public Health Service.
tissues, created a situation favoring transudation of
fluid and other substances from the capillaries. An
increase in plasma protein served to counteract
partially this leakage, since the plasma proteins
exerted an osmotic pressure. There still remained,
however, the problem of clearing the tissue spaces of
substances which had leaked out of blood capillaries
or which were not absorbed into the blood stream.
In this sense, the lymphatic system must be regarded
as a homeostatic mechanism, important in the
maintenance of the constancy of the milieu interieur.
It is this point of view that will be emphasized in the
present discussion. The role of the lymphatic system
in the transport of materials from the liver and in-
testines to the blood stream will also be considered.
No attempt will be made to cover all that has been
done regarding lymph and lymphatics nor will the
extensive literature on lymph nodes and lymphoid
tissues be discussed. Various aspects of the general
subject have been treated in depth during the last
several decades in reviews and monographs (45, 58-
60, 62, 66, 88, 135-137, 185, 189, 215, 223, 227, 234)
and the reader is referred to these sources for basic
material not included in the present review. Two
recent monographs will be found most helpful (189,
234). The latter source will interest those concerned
with clinical implications of disturbed lymphatic
function. It also includes results of work in Hungarian
and Russian laboratories not readily available in the
English literature.
METHODS OF STUDY
Although lymphatics presumably had been seen
by members of the Alexandrian school (Herophilos,
■035
io36
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
300 B.C.; Erasistratus, 310-250 B.C.) the documented
study of lymphatics dates from 1622, when Asellius
(3) demonstrated "lacteals"' in the mesentery of a
well-fed dog and at a later date had the opportunity
of observing these channels in a criminal who had
been executed following a large meal. Jean Pecquet
(169) in 1651 described the cisterna chyli and the
thoracic duct. The term "lymphatics" was first used
by Thomas Bartholin (12, 13) and he and Rudbeck
(186) are usually considered to be the co-discoverers
of the lymphatic system.'2 In 1692 Nuck (160) intro-
duced the use of mercury for injection of lymphatic
vessels, a method which was used extensively by
many investigators during the eighteenth century to
describe the location and distribution of the main
lymphatic vessels. Of particular importance was the
work of Hewson (99), a pupil of William Hunter,
who made extensive dissections of the lymphatic
system and noted that lymph glands were absent in
fishes (also in the turtle), few in number in birds,
and well developed only in mammals. He also noted
the presence of lymphocytes in lymph and thought
they came from lymph glands to enter blood via the
lymph channels. Hunter himself speculated that "the
lymphatic vessels are the absorbing vessels, all over
the body" (101).
Anatomical studies during the nineteenth century
further delineated the distribution and characteristics
of the lymphatic supply of various organs (194), but
it remained for Ludwig (129) and Heidenhain (96)
to provide the stimulus for studies of function. Ludwig
developed techniques for the collection of lymph by
cannulating lymph vessels in different parts of the
body. He contended that lymph was a filtrate derived
from blood, a point of view contested by Heidenhain
who maintained that it was actively secreted by the
lymphatic epithelium. This classic controversy was
finally settled by the extensive work of Starling during
the first part of this century (203), who demonstrated
the relationships between hydrostatic and osmotic
pressures in the exchange of substances between
plasma and lymph, concepts still fundamental and
generally applicable. These relationships have been
discussed in Chapter 29.
The study of lymph and lymphatics has lagged
behind that of other parts of the circulation because
of inherent difficulties in identification and dissec-
tion of the lymphatics and their cannulation. Since
2 There is an interesting biographical note by G. Liljestrand
and a translation of Rudbeck's "Nova exercitatio anatomica"
by A. E. Nielsen in the Bulletin of the History of Medicine 1 1 :
304-339. I942-
lymph is virtually colorless, it does not help in the
visualization of these small vessels. Even the identi-
fication and dissection of the largest trunk, the thoracic
duct, is a formidable challenge to the uninitiated
investigator unless its visualization is aided by previous
injection of dyes or feeding of fats. Once identified
and dissected, cannulation of a lymphatic still
presents a problem because of the ease with which
the thin vessel can be torn. This may explain the
temptation for investigators to forsake the actual
collection of lymph for the much less frustrating
study of the effects of ligation of the vessels. Rudbeck
expressed these difficulties very well in 1653 when he
said:
"Of the many structures difficult to find in ana-
tomical dissections, these vessels, I must confess, are
by no means the least. For usually they will not
tolerate the finest blunt probe, a sharp knife, a suction
tube, or any other instrument whatever. And even
though abundantly present, they are often obscured
by fat, or are overlooked if not at the moment filled
with fluid. When seen they may disappear if not
ligated. Thus in elusiveness they rival the lacteals
and must be handled with utmost care."
Several recent developments have, however, made
the lives of the lymphatic investigators less trying
and their labors more rewarding. Availability of
nontoxic and radiopaque dyes has facilitated tracing
of lymphatic pathways and stimulated a new interest
in this aspect, particularly in surgery (20, 52, 112).
To the physiologist, the greatest boons have been
the availability of polyethylene tubing and isotopes.
The range of sizes and flexibility of polyethylene
tubing and relative freedom from clotting in this
tubing have made cannulation easier and have
made chronic experiments possible not only in un-
anesthetized experimental animals but in man (16,
31, 54, 125, 177, 201, 208). Small vessels entered
with glass cannulae only with the greatest difficulty
can now be studied (198, 199). The use of isotopes
has facilitated the study of lymphatic uptake from
subcutaneous tissues (100, 209). It has also made
possible more quantitative studies on the exchange
of substances between plasma and lymph. These
gains will be apparent in the discussions to follow.
DEVELOPMENT AND STRUCTURE OF
LYMPHATIC VESSELS
It is now generally agreed that lymphatic vessels
are derived from veins. To quote Sabin (190) "Lym-
PHYSIOLOGIC IMPORTANCE OF LYMPH
IO37
phatics are modified veins. They are vessels lined by
an endothelium which is derived from the veins.
They invade the body as do blood vessels and grow
into certain constant areas; their invasion of the
body is, however, not complete for there are certain
structures which never receive them. The lymphatic
capillaries have the same relation to tissue spaces as
have blood capillaries. None of the cavities of the
mesoderm, such as the peritoneal cavity, the various
bursae and serous capillaries, forms any part of the
lymphatic system. The lymphatic endothelium once
formed is specific. Like blood vessels the lymphatics
are for the most part closed vessels."
The lymphatic capillaries may be considered as
endothelial tubes resembling blood capillaries but
thinner. The medium-size vessels (100-200 n) have
muscle fibers, whereas the larger lymphatics are com-
posed of an endothelial layer covered by a diffuse
connective tissue sheath in which elastic and muscular
elements are irregularly scattered. Amyelinated nerves
can be traced to the muscle fibers. Valves develop
during intra-uterine life in the large vessels and are
usually unicuspid or bicuspid. These structures deter-
mine the direction of flow toward sites of emptying
into the blood stream.
LYMPH VS. TISSUE FLUID
It is now generally accepted, primarily from the
work of Sabin (191) and MacCallum (132), that the
lymphatics form a closed system. "Lymph," therefore,
is not synonymous with "tissue fluid," but is the
fluid found in lymphatics. This is more than a
semantic distinction because, as will be apparent
later, the composition of lymph is more particularly
determined by the permeability of blood capillaries
in a definite area and the consequent pcricapillary
filtrate than it is by the metabolism of tissue cells. In
this sense, lymph is pericapillary filtrate which has
mixed with tissue fluid and has entered the closed
lymphatic system.
Clark & Clark (43) showed that lymphatic capil-
laries are sometimes closely associated with small
blood vessels, with virtually nothing between the two
membranes, while in other cases they bear no rela-
tionship to such vessels. In any region, the fluid that
enters the lymphatic system to become lymph may be
that which is adjacent to the arterial end or to the
venular end of a blood capillary, or it may be fluid
that is relatively distant from a blood capillary.
Mc Master (137) studied the relative pressures within
the cutaneous lymphatic capillaries and the surround-
ing tissues in the mouse's ear. He reported that the
mean lymphatic pressure was 1.2 cm water and the
interstitial pressure 1.9 cm water. There was always
a gradient of pressure from the interstitial tissue to
the lumen of the lymphatic even in conditions of
increased lymphatic pressure. Presumably, whenever
increased amounts of fluid are present in the inter-
stitial tissues, the lymphatic vessels are kept open by
swelling of connective tissues and increase in the
tension of the fibers attached to the lymphatic capil-
laries (42, 43, 137, 175). Many more data are needed
in other tissues and species to establish firmly the fact
that a gradient of pressure is always present between
interstitial tissues and the lymphatics, and is an im-
portant factor in the formation of lymph. Particularly
disturbing in this connection is the recent report of
Guyton et al. (93) suggesting the existence of negative
pressures in interstitial spaces.
The close anatomical relationship between the
lymphatic and venous systems has raised the question
as to the relationship of lymphatic and venous pres-
sures. Little definitive information is available, how-
ever, in this area, due primarily to the difficulties in
measurement of lymphatic pressure. Many of the
pressures that have been recorded are end pressures
(234) and not particularly representative of the actual
pressures under normal conditions of flow. Thus, Lee
(124) found the average end thoracic duct pressure
in dogs to be 15 cm HjO, whereas Rouviere &
Valette (185) found side pressure at the entrance to
the subclavian vein to be 6.4 cm H>0. They also
found the pressure in the internal jugular vein of the
same animal to be 2.4 cm H20, thus demonstrating
the existence of a gradient capable of promoting
emptying of lymph from the thoracic duct to the
jugular vein. Webb & Starzl (222) found side pres-
sures of 3.5 to 5.5 cm HjO in the thoracic duct just
above the diaphragm in anesthetized dogs. At this
point, arterial pulsations affected the lymph pressure,
the difference between the pressures during systole
and diastole being 2 to 3 cm H20. Although these
values are lower than those of Rouviere and Valette,
they still permit of a gradient toward the vein.
Irisawa & Rushmer (103) recently reported on the
relationship between lymphatic and venous pressure
in the legs of dogs. Although previous investigators
had regarded lymphatic pressure of a resting dog leg
as being too low to measure, these authors, working
with unanesthetized dogs, found the leg lymphatic
pressures to range from 2.5 to 12.0 cm H«0, while
the range of pressures in ankle veins was from 5.5 to
i o3 8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
15 cm FLO. Pressure levels in veins and lymphatics
were generally very similar at rest, the venous pres-
sures being only slightly higher. It would thus seem
that pressure in the lymphatic capillaries may be
comparable to the pressure in the venous end of
capillaries at rest. These authors found the leg
lymphatic pressure to fluctuate with respiration. With
increased venous pressures, lymphatic pressure rose
slowly but never reached the level of venous pressure,
a reflection, perhaps, of the distensibility of lymphatics
(156) and collection of fluid in the tissues. Similar
experiments in our laboratory (Miller, unpublished)
on the anesthetized dog also showed a lack of direct
correspondence between leg lymphatic and venous
pressures under a variety of experimental procedures.
DISTRIBUTION OF LYMPHATIC VESSELS
Although lymphatic capillaries spread into a tissue
after the blood vessels, the density of the lymphatic
plexus does not always run parallel with the richness
of the blood supply. Furthermore, capillary plexuses
vary tremendously in richness in different organs
and tissues. For example, they are abundant in the
dermis, the conjunctiva, the periostium of bone, and
in the mucosa and submucosa of the alimentary,
respiratory, and genitourinary tracts, but are pre-
sumably absent in cartilage, bone marrow, the central
nervous system, epithelium, and fetal part of the
placenta (62). Voluntary muscle contains lymphatics
only in fascial planes. It is generally believed that
lymphatic capillaries do not actually reach the pul-
monary alveoli, but that their distribution ceases at
the beginning of the respiratory portion of the ulti-
mate lung structure, the atrium leading into the
alveolus. Likewise in the liver, the ultimate functional
unit, the lobule, is not supplied with lymphatic
capillaries. The fluid leaving the liver sinusoids passes
through capillary endothelium and, in the lobule, lies
between this endothelium and the liver cells. Lym-
phatic capillaries are found at the periphery of the
lobule, and these carry the highly proteinized liver
lvmph to collecting trunks which join the thoracic
and right lymph ducts. In the spleen, too, lymphatics
are observed only in the capsule and the thickest
trabeculae. Fluid which filters through the walls of
the capillaries and sinuses must permeate the stroma
before reaching the lymphatic vessels.
Lymphatic vessels in the kidney appear to begin
blindly in two areas (181). The first of these is near
Bowman's capsule, and the second is beneath the
mucosa of the papilla. Two networks of lymphatics
then arise, accompanying venous and arterial blood
vessels of the kidney. Those which originate in the
medulla drain upward and outward toward the
arcuate vessels, where they join with those beginning
near Bowman's capsule draining in the opposite
direction. When the junction occurs, larger trunks
then drain with the arcuate vessels toward the hilum
of the kidney. They may be seen around the renal
artery. There do not seem to be any demonstrable
lymphatic channels in the glomeruli or about the
afferent and efferent arterioles.
The lymphatic drainage of the eye has only re-
cently been clarified (165). The explanation of the
almost zero concentration of proteins in the anterior
chamber has long been a major problem since no
lymphatic drainage had previously been described.
Papamiltiades, in an anatomical study of lymphatics
at the iridocorneal angle of the eye, described lym-
phatic pathways in the neighborhood of the canal of
Schlemm (possibly connecting with the canal) ade-
quate to allow continual removal of proteins from the
anterior chamber.
General Anatomic Arrangement
of the Main Trunks
The large lymphatic trunks join the subclavian or
jugular veins near their junctions. On the left side,
the deep cervical duct, draining the head and neck,
the subclavian duct, draining the arm, and the
thoracic duct, draining the abdominal viscera and
lower extremities, enter the venous system in close
association with one another. The left broncho-
mediastinal trunk, draining the left sides of the thorax,
lung, and heart may join the thoracic duct in the neck
or open independently into the junction of the left
subclavian and internal jugular veins. Sometimes all
of the trunks empty into a sinus or dilatation from
which the lvmph then empties into the vein; at other
times, thev may form a network before entering the
vein or they may all enter the vein close together
but independently of one another. On the right side,
the right jugular trunk, draining the head and neck,
the right subclavian trunk, draining the right upper
extremity, and the right bronchomediastinal trunk,
draining the right side of the thorax, lung, and heart
and part of the convex surface of the liver, empty into
the right lymphatic duct which, in turn, ends in the
right subclavian vein at its angle of junction with the
right internal jugular vein. As on the left side, the
PHYSIOLOGIC IMPORTANCE OF LYMPH
I°39
three collecting ducts not infrequently enter the vein
separately at the junction of the two veins.
The thoracic duct is somewhat more complex
than other lymphatic vessels. It usually begins in
front of the body of the second lumbar vertebra, to
the right of and behind the aorta, by a dilatation of
the cysterna chyli. It enters the thorax through the
aortic hiatus and ascends through the posterior
mediastinum between the aorta and azygos vein.
Somewhere between the fourth and sixth vertebral
level it inclines to the left, enters the superior medi-
astinum, passes behind the arch of the aorta and
thoracic portion of the left subclavian artery into the
neck where, after passing in front of the left common
carotid artery, vagus nerve, and jugular vein, it
ends, as previously noted, by emptying into the
angle of junction of the left subclavian vein and left
internal jugular vein. The thoracic duct is the largest
lymph vessel and is composed of an endothelial layer,
a distinct subendothelial layer of elastic fibers, a
media of irregularly arranged but mainly circular
smooth muscle cells interspersed with elastic and
connective tissue fibers, which is succeeded by the
adventitia containing longitudinal and transverse
bundles of smooth muscle cells as well as blood vessels
and nerves. It contains valves which are quite efficient.
Kampmeier (106) found many more valves in the
thoracic duct of early embryos than in later stages. In
one human fetus, of 4.3 months, he found 42 valves
between the jugular confluence and renal arteries. In
older fetuses he found as few as three complete valves
with numerous vestiges present. Obviously, many of
the early valves never progress to the functional stage
and some vanish entirely. Kampmeier suggested that
the valves which did remain in postnatal life were
determined by areas of direct pressure on the duct
as, for example, in the area between aorta and
esophagus as they cross, an area in which a bolus of
food exerts pressure upon the duct.
The above general descriptions are actually subject
to more exceptions than have been indicated. Studies
of large numbers of animals or species soon demon-
strate this variability7. McClure & Silvester (134)
drew attention to this variability, as far back as 1909,
in their report of a study of 25 species involving 50
mammals (primates, carnivora, rodentia, ungulata,
and marsupialia). In the adult cat, communication
between the lymphatic system and the systemic veins
may normally occur on each side of the body, within
either one of two or within two typical districts.
These two districts include, approximately, the angle
of confluence formed by the union of the external
and internal jugular veins (common jugular angle)
and the angle of confluence formed by the union of
the external jugular and subclavian veins (jugulo-
subclavian angle). In the adult cat, neither one of
these two districts predominates as the place of com-
munication between the lymphatics and the veins;
either one of the two, or both, may serve equally in
this capacity. Their studies in other mammals showed
that these two districts were the predominant sites
of communication between lymphatics and veins, but
there was a marked variability in lymphatic arrange-
ment, more so than in veins, not only in different
species but among members of the same species. These
variations are presumably due to differences in the
establishment of these connections in the embrvo.
One factor which has been relatively neglected in
recent years has been the possible existence of direct
lymphatic communications between veins at points
other than the entrance of the main ducts. Silvester
(200) injected 89 adult monkeys and studied their
lymphatic arrangements. He made the significant
observation that "Whenever the mesenteric or
inguinal lymphatic nodes of a New World species
were injected, the injection mass never passed from
the lumbar or intestinal lymphatic trunks into the
thoracic duct or into the anterior regions of the body,
but passed directly into the postcava into the region
of the renal veins. A more detailed examination of
the vessels in this region of the body revealed the fact
that the lymphatics of the digestive organs and of the
posterior extremities invariably enter the venous
system at the level of the renal veins." Silvester found
the posterior communications between the lymphatic
and the venous system to vary from two to nine in
number and to open at almost any point on the renal
segment of the postcava and its immediate tributaries.
He examined 16 different species of Old World
monkeys and found no evidence of these communica-
tions.
It would be of great interest to know if similar
communications exist in the dog, rat, and man,
animals most frequently used in studies of the lym-
phatic system. Their existence might modify inter-
pretations based on the supposition that all lymph
from the viscera and posterior extremities finds its
way back to the blood stream only via the main
lymphatic channels.
CONTRACTILITY OF LYMPHATICS
In lower animals, as in the frog, lymph hearts
serve to actively propel lymph and distribute it to
1040
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
various parts of the organism. In the mammal, there
are no lymph hearts and lymph is moved along the
vessel wherever and whenever the vessel is compressed,
a situation analogous to that obtaining with veins and
venous flow. The presence of smooth muscle fibers
in the walls of at least the larger vessels, and nerve
fibers running to them, raises the question as to
whether lymphatic vessels have contractility or show
vasomotion. Of particular interest are their responses
to sympathetic and parasympathetic stimulation and
to the chemical mediators, epinephrine and acetyl-
choline.
Spontaneous contraction of lymph vessels was
described as far back as 1774 by Hewson (99), who
reported briefly of having seen actively contracting
lacteals in horses and dogs killed immediately after
the ingestion of food. Since then, many observers
have also reported spontaneous contraction of these
and other lymph vessels, but there seem to be species
differences (159, 160, 175, 202, 224). Definite spon-
taneous contractions have been observed in the
peripheral lymphatic vessels of the bat, rat, and
guinea pig. No spontaneous contractions have been
demonstrated in the cat, dog, rabbit, and squirrel.
The results on mice have been equivocal. The few
casual observations in man have shown none. The rate
of contraction appears to be directly proportional to
the rate of formation of lymph and the contractions
are apparently initiated by an increase in intra-
luminal pressure. They are not dependent on neural
control (225). The vasomotion in these vessels seems
to be similar to that seen in blood vessels and possibly
related to it. Baez and his co-workers (9) observed
mesenteric lymphatics during experimental hemor-
rhagic shock in rats and reported that the lymphatics
undergo pronounced compensatory and decom-
pensatory adjustments recalling those seen in met-
arterioles and precapillaries of the same region.
During the period when the animal is recoverable by
transfusion, the lymphatic vessels exhibit progressive
enhancement of spontaneous motion and of sensitivity
to topically applied epinephrine. Reversal of these
features occurs upon the prolongation of drastic
hvpotension. Lymph vessel adjustments after sublethal
drum trauma are of a compensatory type, compatible
with survival, whereas following lethal trauma, the
lymphatics invariably appear atonically distended and
resemble those seen in irreversible shock.
Experiments with drugs and faradic and other
types of stimulation have also given equivocal results,
due probably to differences in experimental proce-
dures, in species, and in a failure to distinguish be-
tween effects on rhythmicity and on caliber of the
vessels. There is a suggestion that the response of
different lymph vessels may not be uniform. Thus the
usual response to sympathetic stimulation or epi-
nephrine administration, in general, appears to be a
constriction (175, 187-189, 202), whereas the thoracic
duct is dilated by the same procedures (1). Much
more careful work needs to be done in this area.
EXCHANGE OF SUBSTANCES BETWEEN
PLASMA AND LYMPH
Much evidence has accumulated during the last
decade regarding the exchange of substances between
plasma and lymph. As discussed previously, the avail-
ability of isotopes has made possible quantitative
studies of the disappearance of labeled substances
from plasma and their subsequent appearance in
lymph. The availability of polyethylene tubing has
facilitated the collection of lymph and, with a few
exceptions, it has been collected from all areas of the
body and its contents more or less completely char-
acterized. Beginnings have been made in the study
of human lymph (97, 98, 125, 192) under a variety
of experimental conditions. The latter studies have
been concerned with thoracic duct lymph because
of the greater ease of its collection, but will unques-
tionably be extended in the near future to the investi-
gation of lymph from other areas.
Concepts of capillary permeability and factors
which influence it are discussed in detail in Chapter
29. Two concepts have influenced contemporary
thinking in the problem of interchange of substances
between plasma and lymph, /) the familiar "Starling
hypothesis," and 2) the "pore" concept of capillary
permeability.
Starling maintained that the direction and rate of
fluid transfer was proportional to the algebraic sum
of the effective hydrostatic pressure in the blood
capillaries and the osmotic pressure of the plasma
proteins. While the capillary membrane was freely
permeable to crystalloids, it did not allow larger
protein molecules to diffuse readily. The evidence in
general confirmation of Starling's hypothesis has
recently been reviewed by Yoffey & Courtice (234).
Although Starling conceived the capillary mem-
brane as being only relatively impermeable to protein,
there developed a point of view implied or stated in
textbooks that capillaries were impermeable to pro-
tein if they were healthy and that proteins leaked
only when the permeability was abnormal. As will
PHYSIOLOGIC IMPORTANCE OF LYMPH
I 04 I
be discussed in detail below, there is now no question
but that "normal," "healthy" capillaries leak protein
(and other macromolecules) and that the protein
content of lymph collected from different areas of
the body is primarily an expression of the leakage of
these macromolecules from the blood stream. Thus,
during the course of a day, 50 per cent or more of
the total circulating protein escapes from the blood
stream and is returned to it via the lymphatic system.
An additional factor in the reluctance to the ac-
ceptance of the idea that lymph was primarily
derived from capillary filtrate was literal adherence
to the pore concept of capillary permeability. It was
difficult to reconcile the appearance of proteins and
other macromolecules in lymph with the size of the
"pores" postulated for capillary membranes (118).
It is now obvious that the pore concept as originally
reviewed by Pappenheimer (166) must be modified
and reconciled with the more recent work on lymph
to permit of the possible operation of active processes
(141).
Drinker and his colleagues elaborated on Starling's
concept of capillary permeability. As a result of
analysis of lymph from different areas of the body
they concluded "that the capillaries practically uni-
versally leak protein, that this protein does not
reenter the blood vessels unless delivered by the lym-
phatic system; that the filtrate from the blood capil-
laries to the tissue spaces contains water, salts, and
sugars in concentrations found in blood, together
with serum globulin, serum albumin, and fibrinogen
in low concentrations, lower probably than that of
tissue fluid or lymph; that water and salts are re-
absorbed by blood vessels and protein enters the
lymphatics together with water and salts in the con-
centrations existing in the tissue fluid at the moment
of lymphatic entrance" (61). During the last decade,
as will be discussed below, experiments particularly
with isotope-labeled proteins and other macromole-
cules have confirmed the point of view of the Drinker
group and have shown unequivocally that "healthy"
capillaries leak plasma protein and other macro-
molecules and that these are returned to the blood
stream via the lymphatics. To date, all plasma pro-
teins have been shown to be present in lymph from
all areas studied (234).
Exlr avascular Pool and
Circulation of Protein
When a labeled protein is injected intravenously,
the specific activity (ratio of concentrations of labeled
and natural protein) of lymph gradually rises (44,
116, 149) until it reaches that of plasma in 7 to 13
hours in the case of the thoracic duct (220). Samples
of lymph and plasma analyzed after this time show
that the specific activities in the two compartments
remain equal and decline at the same slow rate. This
early growth type of curve suggests that protein
leaves the blood stream and mixes with the extra-
vascular protein pool before being taken up by the
lymph ducts. If lymph were a direct product of
plasma, the experiments should yield a "decay" type
of curve. This point of view is strengthened by experi-
ments in which large infusions were given to dogs
(116, 221). Infusions roughly equivalent to or greater
than plasma volume resulted in increased flow of
thoracic duct lymph and albumin leakage increased
significantly (fig. 1). It was obvious that the eventual
level of total circulating plasma protein was deter-
mined by a number of factors. Infusion results in the
filtration of a more dilute protein solution than that
filtered before the infusion, but one which has a
relatively greater albumin content as well as a larger
volume. This then mixes with the relatively more
concentrated preinfusion interstitial fluid so that the
concentration of albumin in lymph after the infusion
is intermediate between that of the interstitial fluid
formed before the infusion and the newly formed
interstitial fluid. The eventual effect of an infusion
on the total plasma protein level will thus depend
upon a) the degree of distention of the interstitial
space as a result of the infusion, since a greatly dis-
tended interstitial space may hold much of the protein
which ordinarily might have gone back to the circu-
lation via the lymphatics; b) the rate (and amount)
of albumin leaving the capillaries; c) the amount of
albumin present in interstitial fluid available for
mixing with the plasma filtrate; and d) the rate of
lymph flow. If the lymph does not return to the
venous system or if the amount disappearing from
the plasma is greater than the amount returning via
lymph, plasma albumin will be decreased and remain
low until the usual conditions of flow are re-estab-
lished. These results with isotopically labeled albumin
emphasize that we are not concerned with mobiliza-
tion of cell protein, as has been suggested in the litera-
ture (4, 59), but primarily with the movement of
interstitial fluid protein. Addition of new protein
from any source drained by lymph coming to the
thoracic duct would have been apparent by a lowering
of specific activity. This was never seen.
Evidence for the existence of an extravascular
albumin mass as a separate entity and in equilibrium
1042
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. I. A: Ratio of lymph volume flow
(ml/min) after the start of the infusion to
the preinfusion (control) lymph volume flow
plotted against the time after the beginning
of the infusion. The 250-ml infusions lasted
approximately 8 min, the 500-ml infusions
lasted 15 min, the 1000-ml infusions lasted
30 min, and the 2000-ml infusions lasted
approximately 1 hour. Average control
lymph flow for these experiments is 0.5 ml/min.
B: Ratio of albumin flow (ml/min) after the
start of the infusion to the preinfusion (con-
trol) lymph albumin flow plotted against
time after the beginning of the same experi-
ments as in A. Numerals to the right of the
infusion volumes are the numbers of experi-
ments which were averaged in each group.
Average control albumin flow for these
experiments is 8.5 mg/min.
16.0
I4 0 .
2 lt,U
o
i O lO.O
CE _
uj 1 a.o
K 0.
u- 5
< V
3
o
5°
2.0
0
1 1
i i
i i
i !
i !
i 1
i ' ]
i
\
! i
; i
; i
1 j
1 j
i
j
i....
1
i
i
i
..;.
i
j
rrj
■n 1
L
50 GO
90
>
Z 50
5
o
S
o
o
or
\-
_J
\eo 0
MINUTES
2.000ml (A)
IOOO ml I SI
500ml (4)
250ml (6)
50
60 90
1.20
with the intravascular mass was obtained from experi-
ments on unanesthetized, healthy greyhounds, infused
with 25 per cent albumin or bled, into which we in-
jected I131-labeled albumin and then determined the
albumin specific activities (218). We showed that
albumin specific activity curves can be altered by
changing the ratio of intravascular to extravascular
albumin masses in a manner predicted by a two-
compartment system. Increase of intravascular mass
(by infusion) relative to extravascular mass results in
a smaller initial disappearance of albumin specific
activity from the blood stream and a faster approach
to equilibrium. Decrease of intravascular albumin
mass relative to extravascular mass by bleeding shows
that 50 per cent of albumin replacement after hemor-
rhage appears to be accomplished within 24 hours.
Almost all this protein comes from the extravascular
compartment. Rapid anabolism accounts for the
replenishment of protein for the next 2 to 5 days,
during and after which there is a reduced catabolism
of the existing plasma albumin. Thus there are net
movements from the extravascular mass into plasma
when the equilibrium between intravascular and
extravascular masses is disturbed.
Benson et al. (14) concluded that under stand-
ardized resting conditions a given tissue eliminates a
nearly constant amount of protein in its lymph per
unit of time and that the protein concentration in the
lymph from the intestine or liver of the rat varies
inversely with the volume of lymph flow. The concen-
trations of protein fractions in rabbit lymphs, and the
rates of exchange of radioiodinated human serum
albumin between plasma and lymph which they ob-
served, suggested that the equilibration of plasma
proteins with lymph is rapid in the liver, intermediate
in the intestine, and slow in skeletal muscles. These
findings are consistent with our recent demonstration
(see figs. 2 and 3) of differences in blood capillary
permeability in different areas to macromolecules
(141) and the suggestion that there are several sets
of capillary pores of different sizes, large pores pre-
dominating in the liver, small pores in muscles, and
both size pores in the intestinal capillaries. Alter-
nately, the suggestion was made that cytopempsis or a
similar process may be involved.
LYMPHATIC RETURN AND
BLOOD VOLUME REGULATION
Lymph not only returns protein and other macro-
molecules from the extravascular to the vascular
system but also drains fluid representing the excess
of filtration over reabsorption through the capillary-
wall. As discussed later, the amount of lymph re-
turned to the blood stream via the thoracic duct
alone per 24 hours is roughly equivalent to the plasma
volume. It is thus obvious that the return of lymph
plays an essential role in the maintenance of the blood
volume level. However, little definitive data is avail-
able on this point. Courtice et al. (50) state that in
unpublished experiments on dogs anesthetized with
Nembutal "the rate of escape of fluid and protein
in lymph was equivalent to a daily loss of 60 per cent
PHYSIOLOGIC IMPORTANCE OF LYMPH
1043
1000
800
600
400
200
CERVICAL LYMPH
• •-
IN HOURS AND MINUTES
030
100
fig. 2. Typical experiment in anesthetized
dog showing disappearance of dextran of
average molecular weight of 35,000 from
plasma and its appearance in lymph of
various areas.
130
200
230 300
3 30
4:00
2500
2000
1500
1000
500
fig. 3. Same experiment as in fig. 2 showing
disappearance of radioactive albumin from
plasma and its appearance in lymph of various
areas.
0 30
I 00
130
200
230
300
3 30
400
of the plasma and 45 per cent of the circulating
plasma proteins." We have confirmed these observa-
tions (Magruder, Kern, and Mayerson, unpublished).
In 20 dogs, drainage of thoracic duct lymph for 8
hours resulted in an average drop of 16 per cent in
plasma volume. CoTui and his colleagues (46, 196)
found that when they bled dogs whose thoracic ducts
were ligated there was a greater drop in the hema-
tocrit level than in dogs bled but with intact lymphatic
circulation. This hemodilution lasted at least 8 days
after hemorrhage in the duct-ligated animals but
disappeared in about 48 hours in nonduct-ligated
animals.
Another aspect of the problem is the well-known
lymphagogue effect of infusions. As infusions are
made larger, lymph flow increases proportionately
so that with large infusions in dogs (2000 ml) the
thoracic duct lymph flow may reach a peak value of
about 14 times that of the preinfusion value (221).
The displacement of fluid from the circulation supple-
ments the diuresis through the kidneys and may be
considered as a fine adjustment of the blood volume
ro44
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
so that not all of the fluid is irrevocably lost from the
body. Large infusions also increase protein leakage
but here again the protein is slowly returned to the
blood stream and minimizes changes in total circu-
lating protein and loss of its oncotic effect.
It should, perhaps, be emphasized that blood is
the chief source of the water of lymph. Benson et al.
(15) measured the content of either D20 or Na'-4 in
intestinal lymph, portal venous blood, and femoral
arterial blood of anesthetized hydrated rats after
administration of the isotope into the stomach,
duodenum, or peripheral or portal vein. Little, if any,
water or sodium found its way into lymph after
absorption from the small intestine. At least 99 per
cent appeared to be carried in portal venous blood.
The amount of isotope found in intestinal lymph was
proportional to lymph volume whatever the route
of administration. Thus, even during absorption of
water or sodium ion from the small intestine, blood
is the principal source of the water and sodium in
lymph.
TRANSPORT FUNCTION
Lipids
There has been considerable interest, particularly
during the last decade, in the transport of lipids — the
physical state in which they are carried in the blood
and their exchange between blood plasma and tissue
cells. The availability of isotopes has facilitated the
design of experiments concerned with lipid transport
by lymph. It is now apparent that the lymphatic
system plays an important role in lipid transport as
it does in protein transport. This may be because the
passage of plasma lipids through the capillary mem-
brane depends on lipid-protein complexes rather than
on the physical properties of the lipids themselves.
All the different lipid-protein associations present
in the plasma have been identified in thoracic duct
lymph (162), as well as in cervical and leg lymph
(49). In the dog and cat, alpha-lipoprotein pre-
dominates. When rabbits are fed cholesterol, however,
the plasma beta-lipoprotein may increase consider-
ably with a much smaller rise in alpha-lipoprotein.
Lender these circumstances lymph contains beta-
lipoprotein. The evidence suggests that beta-lipopro-
tein leaves the blood circulation at a slower rate than
does alpha-lipoprotein. Not only do alpha- and beta-
lipoproteins appear in the lymph in the postabsorptive
state, but lymph from the cervical, hepatic, and leg
ducts — all draining tissues remote from the alimentary
tract — also contains chylomicrons (49). As Yoffev &
Courtice (234) state: "We can readily understand
how the intestinal lymph always contains chylomi-
crons even in what we call the postabsorptive state.
The presence of chylomicrons in lymph from other
tissues, however, suggests that they come either from
the blood stream by passing through the capillary
membrane or from the fat depots. The evidence
indicates that the chylomicron count in the lymph
may vary with that in the blood, which suggests that
these particles may pass through the capillary mem-
brane and so appear in the lymph. For example, the
hepatic and cervical lymph ducts were cannulated
in a fat-fed cat and chylomicron counts made on
lymph and plasma. The thoracic duct which was
pouring very fatty chyle into the blood stream was
then cannulated and the lymph collected. The
chylomicron count in the plasma fell in the next
few hours and with this fall the counts in the hepatic
and cervical lymph also fell. The fatty chyle which
had meantime been collected from the thoracic duct
was then injected intravenously making the plasma
quite milky. The chylomicron counts in the hepatic
and cervical lymph subsequently rose." Geyer et al.
(83) attempted to assess the permeability of capillaries
to serum cholesterol in humans by measuring the dis-
appearance of cholesterol from the blood in the
forearm during various degrees of venous congestion.
L'nder these circumstances, measurable amounts of
cholesterol were filtered and were related to the rate
of fluid filtration and the initial level of the serum
cholesterol. The results were similar to those of
Landis et al. (1 19) for serum proteins. It may also be
of interest to mention that the rise of plasma choles-
terol occurring in the hypothyroid state does not
appear to be due to any decrease in its ability to
diffuse out of the plasma (79). Electron microscope
studies suggest that chylomicrons can be transferred
directly across cell membranes (5, 164) by the active
process of pinocytosis. The probability that some
active process is concerned in the transfer of macro-
molecules from the capillaries to lymph is discussed
elsewhere (141).
When fatty chyle or artificial fat emulsions are
injected into the blood stream, they leave the circula-
tion very rapidly, but the amounts found in lymph
are relatively small (140, 142, 148, 151, 233). The
latter investigators injected fatty chyle collected from
fat-fed cats into postabsorptive cats and determined
the lipid disappearance from plasma and its ap-
pearance in hepatic, intestinal, and cervical lymph.
PHYSIOLOGIC IMPORTANCE OF LYMPH
1045
These calculations showed that protein left the circu-
lation at the rate of 77, 142, and 11 mg per hour in
the liver, intestine, and cervical tissues, respectively,
or a total leakage of 230 mg. In another cat, 2540 mg
of injected fat left the circulation within 2 hours
during which time the leakage of protein was 448 mg.
It is obvious that fat in chylomicron form can dis-
appear from the blood stream much faster than pro-
tein. The chylomicron fraction of lymph appears to
carry neutral fat (233).
Reinhardt el al. (182) injected biosynthesized
P32-Iabeled phospholipid into a peripheral vein of
rats with thoracic duct fistula and reported that 9 to
20 per cent of the injected phospholipid could be
recovered in the thoracic duct lymph in the succeed-
ing 3 to 6 hours. McCandless & Zilversmit (131)
obtained labeled lymph by feeding dogs with re-
labeled triolein. The labeled lymph was administered
to recipient dogs, and the rate of disappearance of
the lymph lipids from plasma was followed. Lymph
t 131 triglycerides were found to disappear from the
circulation rapidlv, with an initial half-time of several
minutes. Disappearance of I131 phospholipids was
slower as determined in the same animals, 10 to 40
per cent of the injected dose remaining in the blood
1 to 2 hours after injection. These results were similar
to those previously obtained by the same authors
using artificially prepared fat emulsions (130).
The presence or absence of bile appears to influence
the pattern of absorption and lymph transport of
dietary soaps and triglycerides in the dog. Rampone
& Sigurdson (179) recently reported that the ab-
sorption of triolein and sodium oleate was signifi-
cantly diminished in the absence of bile. In the normal
dog, 90 per cent of fed triolein and 94 per cent of
fed sodium oleate were recovered from the thoracic
duct as lymph lipid. In dogs with bile fistula only 8
per cent of fed triolein was recovered in lymph com-
pared to 40 per cent of sodium oleate.
The route of absorption of steroids from the gas-
trointestinal tract seems to be determined largely by
the chemical nature of the compounds. Methyl
testosterone, 1 7a-methylestradiol and cortisone-4-C14
acetate are absorbed in the rat by way of the portal
circulation (23, 24, 102). Studies in human subjects
have shown that testosterone, cortisone, and cortisone
acetate are also absorbed in this manner and are
virtually absent from lymph (97). In contrast, ab-
sorption of cholesterol into lymph of the rat (17, 41)
and dog (152) accounts for essentially all the sterol
that enters these species from the diet. This is also
true for man (98). Data from the different species
studied is consistent in showing that much of the
cholesterol is esterified by the intestinal mucosa (25,
41, 98, 213). A number of factors appear to influence
the lipid composition of lymph during cholesterol
absorption (212). In rats given intragastrically
emulsions containing cholesterol, oleic acid, and
sodium taurocholate, addition of albumin resulted
in a rapid increase in total lymph lipid which was
much more marked than in those animals not re-
ceiving albumin. The amount of lymph cholesterol,
however, was less for a 24-hour period. The presence
of taurocholate and oleic acid in administered emul-
sions resulted in elevation of ester cholesterol, indi-
cating increased absorption of endogenous cholesterol
(211). Addition of cholesterol to the emulsions also
resulted in further significant increase in the ester
cholesterol fraction in thoracic duct lymph. In
further studies (207), it was shown that small doses
of fed cholesterol-4-C14 lead to labeling of cholesterol
fractions of mucosa and lymph without an increase
in the level or turnover in lymph. Feeding tracer
dose with oleic acid and sodium tauracholate in-
creases the turnover rate of the pool which leads to an
increased amount of labeled and unlabeled cholesterol
in lymph. In fasting rats, the major fatty acids in
lymph are palmitic, linoleic, and oleic acids with
polyunsaturated fatty acids comprising 36 per cent
of the total cholesterol fatty acids (206). After feeding
oleic acid only 42.3 per cent of the total was present
as oleic acid. The total cholesterol fatty acid composi-
tion of lymph is evidently determined not only bv
dietary fatty acid, but by the composition of the fatty
acid pool in the mucosa from which fatty acids are
drawn for esterification of cholesterol, a suggestion
which had been made earlier bv previous workers
(30).
Bloom el al. (22) fed unanesthetized rats C14-
labeled stearic and myristic acids and found that
nearly all the absorbed C14 was recovered in intestinal
lymph. This finding, taken in conjunction with earlier
work with labeled palmitic and pentadecanoic acids,
showed that lymph is the major if not the exclusive
agent for the transport of absorbed long-chain fatty
acids. On the other hand, when similar experiments
were carried out with labeled lauric acid and decanoic
acid, recoveries of the absorbed C14 amounted to 15
to 55 and 5 to 19 per cent, respectively. Since it was
shown that the findings were not the result of bacterial
action, it would appear that the major portion of a
short-chain fatty acid is transported via the blood
stream from its site of absorption. Blomstrand et al.
(21) have extended this type of study to man. They
1046
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
found that linoleic acid-1-C14 incorporated in dietary
triglycerides, or fed as free acid, becomes esterified
with the same classes of lipids in human thoracic
duct lymph as oleic and palmitic acid. Evidently,
digestion and absorption of these fatty acids are com-
parable, as well as their transfer into intestinal
mucosa and resynthesis in lymph lipids. This is in
confirmation of work on animals by the same group.
They also confirmed earlier findings on animals that
stearic acid-1-C14 was found in a higher percentage
incorporated in lymph phospholipids than was
found for linoleic acid and for palmitic and oleic
acids. After isolation of lymph lecithins, there was a
difference in the position of the label in lecithins of
lymph according to the fatty acid used. After feeding
linoleic acid-i-C14, approximately 75 per cent of the
label in lymph lecithins was localized in the alpha-
position. With stearic acid-i-C14, however, about 80
per cent of the label was found in the beta-positions.
Their evidence indicates that there is a distinct
manner in which stearic acid-i-C14 and linoleic
acid- 1 -C14 are incorporated into thoracic duct lymph
lecithins, reflecting probable differences in their
metabolism.
Rampone (178) recently reported experiments in
which he measured phospholipids of lymph in rela-
tion to the total lipid in 16 dogs with chronic thoracic
duct fistula during the postabsorptive state and fol-
lowing the administration of various lipid types
(triolein, soya lecithin, oleic acid, etc.) in the diet.
He found that phospholipid transport related linearly
to total lipid transport under all conditions studied,
including the postabsorptive state. The percentage
of lymph lipid transported as phospholipid ranged
from 3 to 18 per cent and was independent of the
type of lipid led. Depriving the animals of phospho-
lipid precursors in the diet for as long as 90 days
previously failed to alter this relationship or the total
quantity of lipid transported. Since the phospholipids
increased linearly with the total lipid under all condi-
tions studied, Rampone believes it likely that the
phospholipids associate with the absorbed lipid in
some manner which relates to lipid transport, possibly
serving in the capacity of chylomicron emulsion
stabilizers during the transport phase. He points
out that while the plasma may be the source of the
phospholipids, the rate of filtration from plasma to
lymph would be somehow dependent on the lipid
concentration in lymph, since the phospholipids of
lymph increased in proportion to the total lymph.
Previous work by Bollman et al. (26) suggests that
the mucosa of the small intestine mav normally be
the source of phospholipids for plasma during fat
absorption.
An interesting application of the study of lymph
and its possible role in the pathogenesis of athero-
sclerosis was reported by Kellnor (1 1 1). He collected
leg lymph from rabbits rendered hyperlipemic by
cholesterol feeding, by the injection of the surface-
active agent Triton A-20, and by the injection of
alloxan. He found (as have others) that leg lymph
contained protein in a concentration equal to one-
third to one-half that of the blood serum. Electro-
phoretic analyses showed a pattern similar to that
of serum. The total lipid concentration was also about
one-third to one-half that of blood serum and the
major lipid fractions, cholesterol and phospholipid,
were present in lymph in about the same relation-
ship. In the cholesterol-fed rabbits, the leg lymph
showed a striking increase in lipids as did the serum.
On the basis of his results he concluded that: "It
seems likely that under normal conditions there is a
constant flow of fluid containing various serum lipids
and proteins across the endothelium into the walls
of blood vessels; this material normally passes through
the wall and is completely removed by way of vasa
vasorum and lymphatics. In certain conditions, how-
ever, where there are increased amounts of lipid in
the blood, or where there are excessive quantities of
certain types of lipids (beta-lipoproteins of the Sf 12-
20 molecules of Gofman), the removal of these
particles from the wall of the vessel is incomplete and
some remain behind to initiate the process of athero-
sclerosis. In hypertension, the increased hydrostatic
pressure appears to cause an increase in the quantity
of serum lipoprotein that diffuses across the vessel
wall, thereby increasing the possibility for incomplete
removal and hence for deposition of lipids. In those
areas of the vascular tree where the removal mech-
anism has been altered, as for example in syphilitic
aortitis or in experimentally produced trauma to the
vessel wall, the free transport of lipid and other
particles across the vessel wall is impeded, and in
these areas the lipid is therefore more apt to pre-
cipitate and to give rise to atherosclerosis. In this
theoretical formulation of the pathogenesis of athero-
sclerosis, the artery wall is regarded as an organ which
is constantly bathed by a serum transudate containing,
among other things, various serum lipoproteins, most
of which pass on through, some of which doubtless
are metabolized locally, and a few of which remain
behind to cause mischief. Atherosclerosis, broadly
considered, may thus result either from qualitative
or quantitative changes in the serum lipoproteins
PHYSIOLOGIC IMPORTANCE OF LYMPH
IO47
that filter constantly across the walls of blood vessels,
or from local structural changes inherent in the vessel
wall, or the result of age or disease, that serve to
hamper the normal passage of these fatty substances."
Enzymes
Many enzymes are found in lymph in small con-
centrations (234). Their concentrations are usually
higher in intestinal and liver lymph than in cervical
or leg lymph, but are usually lower than in plasma
and run parallel with the concentration of proteins
(18, 28). It is probable that, in most instances, these
substances have leaked from the blood stream and
take part in the extravascular circulation via the
lymphatics.
On the other hand, certain enzymes seem to be
transported to the blood stream from their cells of
origin via the lymph. Flock & Bollman (74) made
an interesting comparison between the activity of
rat intestinal lymph with respect to amylase and
tributyrinase. The activities of the two enzymes in
intestinal lymph are generally less than in plasma.
The 24-hour secretion of amylase in lymph is greater
in fed than in fasting rats, but much of the increase
is due to the increase in lymph volume. External
drainage of lymph for 2 days does not significantly
alter the plasma amylase level. On the other hand,
although the 24-hour secretion in lymph of tributy-
rinase is also much greater in fed than in fasting rats,
it appears to represent a specific effect of ingested
fat on the chemical composition of intestinal lymph.
External drainage of the lymph markedly decreases
the tributyrinase content of plasma. These results are
similar to those previously obtained by the same
authors with respect to alkaline phosphatase (72).
The increase of alkaline phosphatase of intestinal
lymph following the feeding of fat is abolished or
greatly diminished when the bile duct is ligated or
the bile drained away in a biliary fistula (73). The
presence of bile thus seems to be essential for the
release of alkaline phosphatase from the intestinal
mucosa.
The histaminase activity of lymph has received
considerable attention from Carlsten and his col-
leagues. They were led to these studies by their failure
to demonstrate the presence of histamine in venous
blood during reactive hyperemia and muscular
tetanus, where histamine was alleged by some in-
vestigators to be liberated. They then turned to
lymph on the grounds that lymph is closer to the tissue
cells which are thought to liberate histamine, and
therefore histamine should accumulate in greater
concentration in lymph than in plasma. They used
dogs, anesthetized with Nembutal, and collected
lymph from the thoracic duct (37). Lymph had no
histamine in detectable amounts. In contrast to
guinea pigs, rats, and rabbits, dogs show low plasma
histamine activity. The plasma histamine concentra-
tion could be raised to very high levels by histamine
infusion or by intravenous injection of histamine
liberators (curare and trypsin) without the appear-
ance of detectable amounts of histamine. Study of
the histaminolytic activity of the lymph showed it
to be more than 30 times as powerful as in plasma
when tested in vitro. In vivo, intralymphatically
administered histamine was inactivated at a very
high rate. This same group has also used cats and
have described a simple micromethod for estimation
of the small amounts found in lymph and plasma
(38, 231), and showed that the histaminolytic activity
of lymph is not changed by routine procedures such
as anesthesia, laparotomy, gentle handling of the
viscera, or by reactive hyperemia or pregnancy (35,
232). Adrenalectomy is followed by a marked increase
in histaminase content of thoracic duct lymph (but
not in plasma) which reaches a maximum within 2
hours and persists approximately 24 hours (36). In-
fusion of an adrenocortical extract will reverse this
increased activity (39). The histaminolytic activity
of cervical and leg lymph is less than that of the
thoracic duct and seems to originate from the kidneys
and gut (34).
It has been suggested that the lymphatic transport
of lipase may be concerned with the changes seen
in disseminated pancreatic fat necrosis (171). So-
called pancreatic and peripancreatic fat necrosis is
supposedly due to the splitting of neutral fat into
glycerol and free fatty acid by pancreatic lipase
which has escaped from the injured pancreas. The
free fatty acids are thought to combine subsequently
with calcium in the tissue and tissue fluids to form
insoluble calcium soaps which give rise to the opaque
white areas seen in the fat depots of the abdominal
cavity and elsewhere. Perry made intraperitoneal
injections of a mixture of pancreatin and graphite
suspension in rats and at necropsy found multiple
areas of fat necrosis in the abdominal and thoracic
cavities, closely associated with graphite-delineated
lymph channels. The evidence of the participation
of the lymphatics in this disease is quite suggestive
and indicates the desirability of further investigation
of the role of lymphatics in this and other diseases.
Reizenstein el al. (183) recentlv reported experi-
io48
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
merits on two beagles to which they gave C058-Bi_. by
stomach tube and measured its absorption and dis-
tribution between thoracic duct lymph and plasma.
They found only a very small amount of the total
dose in lymph which they believe to have leaked from
the plasma. They interpret their results as suggestive
that vitamin B12 is absorbed directly into the blood
stream as a compound with a molecular weight only
slightly higher than that of pure crystalline B]>. If it
were absorbed as the entire intrinsic-factor-molecule
(mol wt ± 70,000) more should have been found in
the lymph, since this large molecule probably cannot
easily get into the plasma.
Coagulation Principles
Lymph from all parts of the body clots, but does
so less readily than plasma. The concentrations of
fibrinogen and of prothrombin in lymph are always
less than in plasma and vary considerably in different
regions just as concentrations of other proteins vary.
Mann et al. ( 1 39) drained intestinal lymph from rats
and found that marked hypoprothrombinemia de-
veloped rapidly, usually within 24 hours. If adequate
amounts of vitamin K were administered parenterally,
a normal level of prothrombin was maintained,
despite loss of lymph. Transfusion of twice the animal's
normal volume of plasma did not maintain a normal
value for prothrombin while lymph was lost. Under
the conditions of their experiments, it appeared that
vitamin K was absorbed practically exclusively
through the lymph and very little of it was stored,
whereas the turnover of prothrombin was extremely
rapid.
The concentration of fibrinogen of canine thoracic
duct lymph is about 50 per cent that of plasma (29,
70). Brinkhous & Walker (29) found that the mean
prothrombin level, expressed as a percentage of that
in the plasma, was 93.2, 51.2, and 7.6 for hepatic,
thoracic duct, and leg lymph, respectively. These
findings are consistent with the known differences in
permeability of capillaries to macromolecules in the
leg and liver. Infusion of heparin into anesthetized
dogs (214) prolonged thrombin and prothrombin
times of plasma immediately, but the effect was
delayed in thoracic duct lymph and required larger
doses for its production. The differences between
plasma and lvmph were more marked with cervical
lvmph, which again may reflect the differences in
permeability between the capillaries in the areas
drained bv the cervical and thoracic ducts.
Langdell et al. (120) have extended and, in general,
confirmed these observations. They also found that
lymph samples are not fully active at the time of
collection. On exposure to glass surfaces in the pres-
ence of anticoagulant, the clotting time becomes
shorter during the first 20 to 40 min. Coagulating
lymph has a high residual prothrombin even after
18 to 24 hours in glass containers. Thoracic duct
lymph contains sufficient thromboplastic materials
so that adequate amounts of thrombin can form to
produce a fibrin clot, but it does not contain the
thromboplastic materials required for complete
prothrombin utilization. These authors conclude
"thoracic duct lymph in this respect might be com-
pared with platelet-poor native plasma; however,
the initial phase of relatively rapid prothrombin
utilization in clotting lymph is unlike the slower
initial utilization reported to occur in platelet-
deficient plasma systems. The nature of the thrombo-
plastic material in lymph is not known, but it would
appear that the lipid materials being transported
could furnish clot-accelerating activity. Additional
studies are needed to evaluate the role of the lipid
materials in the coagulation of lymph. Such studies
promise to furnish considerable information on the
role of alimentarv lipemia on blood coagulation since
lymph drains directly into the venous circulation."
Iron
The demonstration of iron within leukocytes of
the intestinal villi, subsequent to the oral administra-
tion of iron, led Macallum, in 1894, to suggest that
leukocytes are partially responsible for the transfer
of iron from the intestine (133). Since then, other
investigators (81, 84) demonstrated an increase of
iron within mesenteric lymphatics after oral iron
administration and suggested that lymphatics are
involved in iron absorption and transport. Histo-
chemical studies indicated that phagocytes might be
concerned in mediating the transfer of iron from the
intestine into the lymphatics (84), but more recent
evidence does not support these concepts. Thus,
Moore et al. (147) showed that iron absorbed from
the intestine of dogs passes directly into the blood
stream and only a minimal amount appears within
the intestinal lymphatics. Endicott et al. (68) showed
that the iron demonstrable in intestinal lymph of
dogs and guinea pigs was derived from sources other
than a single test meal. They showed that in the dog
iron was transported chiefly via the portal vein with
PHYSIOLOGIC IMPORTANCE OF LYMPH
I°49
only an insignificant amount appearing in thoracic
duct lymph. Similar conclusions were reached by
Reizenstein et al. (183). Koler & Mann (115) found
that the iron content of intestinal lymph of cats
maintained on a normal diet was relatively constant
over periods as long as 7 days. At lymph outputs of
1 ml per hour, there was an hourly output of 0.5 ^ig
of iron. Peterson & Mann (174), using radioiron,
found that only an insignificant portion of an orally
administered amount of radioiron appeared in the
lymph of rats with total intestinal-lymph fistulas —
less than o. 1 per cent of the total amount of radioiron
administered and only 2.0 to 5.0 per cent of the total
amount of iron absorbed from the gastrointestinal
tract after 8 hours. Everett et al. (69) confirmed
these results in the rat. The absorption of subcutane-
ous FeCl3 occurred primarily via the blood vessels,
but subcutaneous plasma-bound iron passed almost
exclusively into the lymphatics. Intravenously ad-
ministered iron appeared rapidly in the lymph.
These observations and those of previous workers are
unquestionably related to the fact that iron is nor-
mally bound to protein in plasma. Since proteins
leak slowly from blood capillaries, we would expect
to find small quantities of iron-protein compounds
in lvmph from all areas. Since the capillaries of the
intestine are more permeable to protein than those
of other areas, and since protein leakage is greatest
in the liver, larger amounts would be present in
intestinal, hepatic, and thoracic duct lymph. It is
also of interest that Everett and co-workers found no
evidence that leukocytes played more than a negligible
role in iron absorption regardless of the method of
iron administration.
Miscellaneous
Scattered reports deal with a variety of substances
transported in lymph. Thus Salter (193) reported
that the protein-bound iodine per gram of protein in
cervical lymph was concentrated relative to the
homologous serum value. Klitgaard et al. (113, 114)
found that about 3 per cent of a subcutaneously
administered dose of thyroxin-C14 appeared in
thoracic-duct lymph in rats during an 8-hour experi-
mental period. The level of radioactivity in lymph
was lower than in plasma on a volume basis but
significantly higher when calculated on the basis of
protein content. Chromatographic analysis of lymph
samples showed the radioactivity present to be from
unaltered thyroxin. It would be interesting to know
the extent to which other protein-bound hormones are
transported in lymph. We can assume that small
quantities escape from the capillaries as do other
macromolecules and are returned via the lymphatic
system.
Dietrich & Siegel (53) recently reported an in-
teresting study designed to determine whether
nucleotides or nucleotide precursors synthesized in
an organ or tissue, e.g. liver, were available to nourish
other tissues and organs. The stimulus for their
studies arose from observations that certain cell
types cannot utilize free bases and must secure the
nucleoside containing the base from an external
source, apparently other cell types. They argued that
if bases and other nucleotide precursors are secreted
by a distant organ or cell type, these compounds
may be present in both the blood and lymph which
bathes the cell or organ. Blood, however, contains
such a mass of living cells that it is difficult to deter-
mine whether intermediates found in the plasma
are derived from the cells within the blood or from
other somatic cells nourished by and yielding their
products to the blood. Since the cell population in
lymph is insignificant when compared with that of
blood, it might be assumed that metabolites found
in the lymph would reflect more closely the metabo-
lism of the tissue through which it has passed than
that of the lymphocytes. Working on rats anesthetized
with Nembutal, they injected glycine-2-C14 and nico-
tinamide-7-C14 and found adenine, guanine, cytosine,
uracil, and uric acid in measurable amounts in
thoracic duct lymph. Xo detectable quantities of
nucleosides were observed. The quantity of acid-
soluble nucleotides found was equivalent to that
which would be expected from the lymphocytes
present in the lymph samples analyzed. Lymph
collected for a 45-hour period following the injection
of carbon-labeled glycine contained no significant
amount of labeled purine derivatives. At the end of
this period, however, liver tissue still contained
appreciable quantities of labeled acid-soluble nu-
cleotides. Lymph collected for a similar period of
time after the injection of carbon-labeled nicotinamide
contained very small amounts of radioactivity. While
the results raised many unexplainable questions and
suggested the need of further work, they did confirm
previous investigations of plasma in indicating that
if these compounds are essential for the proper nutri-
tion of certain cell types, these purine derivatives are
not transported from sites of synthesis, such as the
liver, via the lymphatic ducts.
1050
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
SIGNIFICANCE OF SOME REGIONAL LYMPHATICS
Thoracic Duct
The size, high rate of flow and accessibility for
cannulation have made the thoracic duct the duct
of choice in studies on lymph. A considerable amount
of data has therefore accumulated during the last
125 years relative to its characteristics under a variety
of conditions in many species, including man (50,
51). During the last decade, as previously indicated,
the advent of polyethylene tubing made possible
chronic experiments in which the cannulas could be
left in place and the data collected in the unanes-
thetized animal and man and in a reasonably ''nor-
mal" physiological situation.
The thoracic duct receives lymph from the ab-
dominal viscera, the lower part of the trunk, and the
lower extremities, and empties it into the jugular
vein. The rate of flow in the thoracic duct is greater
than the sum total of flow from all other ducts.
Yoffey & Courtice (234) have assembled available
data on the rates of flow in the dog, cat, rabbit, rat,
horse, bull, cow, goat, and man. To this list may be
added the data of Shrewsbury on the mouse (198).
It is interesting that in spite of the many variables
involved in the experimentation (anesthesia, time of
feeding, duration of collection, etc.) the thoracic
duct lymph flow in all species studied averages about
2 ml per kg per hour in nonruminants and somewhat
more in ruminants. If we accept the average figure
for plasma volume for most animals of 45 ml per kg,
it is obvious that the amount of lymph returned to the
blood stream via the thoracic duct per 24 hours is
roughly equivalent to the plasma volume. The above
calculations are in terms of the quiet, resting animal.
Under conditions of activity, the return is consider-
ably greater.
Hepatic Lymph
The contribution of the liver to thoracic duct
flow appears to be variable. It may contribute from
one-fourth to one-half of the flow in the dog (33)
and cat (149, 150), and only about 10 per cent in
rats (138). Actually, the anatomy of the hepatic
lymphatic drainage is such that direct measurement
of total hepatic lymph flow is difficult and the data
available have been derived from indirect estimates
or from experiments where usually only one large
hepatic duct was cannulated. The values obtained
by such direct cannulation (0.4-1.2 ml kg hour)
suggest that liver lymph flow is higher than that of
any other part of the body of the clog.
Two hilar lymphatic pathways have been described
for the canine liver (184): /) a main hilar system,
draining predominantly the right lobes; and 2) an
accessory hilar system, draining mainly the left lobe.
Usually all of the hilar lymph seems to pass into one
common efferent trunk which then discharges it into
the cisterna chyli. About 80 per cent of lymph leaving
the canine liver probably travels by the hilar route
and the remaining 20 per cent by the hepatic venous
lymph route.
Not only is liver lymph flow higher than from any-
where else in the body, but it also has the highest pro-
tein concentration, equaling from 80 to 95 per cent of
plasma concentration in dogs, rats, and cats ( 141 , 149,
1 57, 158, 234). Electrophoretic analyses show that
protein distribution in hepatic lymph is similar to that
in plasma. These data are derived from acute experi-
ments on anesthetized animals, from animals with
chronic lymph fistulae, experiments in which T-1824
or other dyes have been used to label proteins, and
from isolated liver preparations.
The extraordinarily high permeability of the hepatic
endothelia involved in hepatic lymph formation has
been demonstrated by the use of dextran fractions of
graded molecular weights (92, 141 ). While the results
of these investigations differ in details, they are con-
sistent in showing that high molecular weight dextrans
appear in hepatic lymph in greater concentration than
in lymph from other areas (fig. 4) and suggest that
hepatic lymph represents a plasma filtrate formed in
a region of highly permeable capillary walls. In a
recent review of hepatic circulation (27) Brauer sug-
gests: "As a working hypothesis compatible with the
major part of the available data, one may accept the
following: Liver lymph formation involves two sites.
The first of these would appear to be the sinusoidal
portion of the hepatic vascular tree where a very large
area of endothelium with demonstrably large pores
surrounds the blood stream, and where one would ex-
pect the formation of a large volume of lymph, differ-
ing from the blood principally in the absence (in the
normal liver, at least) of erythrocytes and of the greater
part of the leukocytic elements. This primary lvmph
for the most part moves countercurrent to the blood
stream to enter the lymphatic vessels within the Glis-
son sheath. Here it passes through the peribiliary
plexus, the second site important in liver lymph forma-
tion. The principal role of this plexus in the normal
liver should be sought in the opportunity it provides
for secondary modification of liver lymph composi-
tion by exchange of soluble components between bile,
lymph, and blood."'
The high rate of flow of hepatic lymph coupled
PHYSIOLOGIC IMPORTANCE OF LYMPH
IO=,I
50
400
fig. 4. Curves illustrating relative permeability coefficients
for dextrans of different molecular weights. Relative perme-
ability coefficient is ratio of dextran between lymph and plasma
divided by ratio of albumin between lymph and plasma. See
(141).
with its high protein concentration emphasized the
importance of the hepatic lymph system in the turn-
over of plasma volume and plasma proteins. Little or
no new protein, as such, is added to lymph in the
liver (234), and the large amount of protein is that
which has leaked from blood capillaries and sinusoids.
Nix<7 al. (157, 158) estimated that, in the anesthetized
dog, the volume of lymph collected from the liver, in-
testine, and thoracic duct was equivalent in 24 hours
to 47, 39, and 95 per cent, respectively, of the es-
timated plasma volume. They found, as have others,
that more than half of the total circulating plasma
proteins passes through the thoracic duct daily. When
they produced hepatovenous congestion or cirrhosis,
the flow of hepatic lymph was two to five times that
found in normal dogs. The equivalent of 70 to 207
per cent of the total circulating plasma protein passed
through the liver lymphatics in 24 hours. Likewise,
Friedman et al. (78) collected hepatic lymph from the
rat in chronic experiments and reported an average
flow of 1.5 ml in 12 hours (12 rats). This rate of flow
was increased to an average of 5. 1 ml in 1 2 hours (6
rats) following biliary obstruction.
The role of liver lymphatics in the problem of ascites
is discussed by Yoffey & Courtice (234). They point
out that in the shifts of fluid which take place when
ascites develops, three major sets of lymphatics are in-
volved: lymphatics of the alimentary tract, liver
lymphatics, and lymphatics of the diaphragm. All
three are capable of carrying very large volumes of
lymph, much greater than lymphatics from any other
region of the body. Only in extreme circumstances
and in the presence of severe disease does gross ascites
become evident. Baggenstoss & Cain (10, 11) studied
the relationship of hepatic hilar lymphatics to ascites
in man. In various conditions associated with ascites,
they found these structures increased in size and num-
ber when ascites was caused by cirrhosis of the liver or
congestive heart failure but not when it was caused by
neoplastic involvement of the peritoneum or by renal
disease. Other clinical and pathologic conditions asso-
ciated with ascites which revealed an increase in
lymphatic vessels at the hilus were lupus erythemato-
sis, fatal virus hepatitis, and massive liver involvement
by neoplasms.
Pulmonary Lymphatics and Edema
There is a very large literature on the anatomy and
pathology of the pulmonary lymphatic system but
there is comparatively little information on the func-
tion of the widespread lymph vessels in the lungs.
Warren & Drinker (216) were the first to collect lung
lymph in 1942 when they succeeded in cannulating a
large lymphatic in the anterior mediastinum of dogs.
In 18 animals, they reported an average lymph flow of
1.1 ml per hour and an average protein concentration
of 3.7 g per 100 ml. As they realized, their experiments
were subject to the criticism that the thorax was open
and the usual intrathoracic pressure absent. To ob-
viate this difficulty, they and subsequent investigators
turned to collection of lymph from the right duct.
This procedure, although eliminating the above ob-
jection, introduces other variables. The right duct, as
usually cannulated in the dog for lymphatic studies,
not only drains the lungs but also carries lymph from
the heart, right side of the thorax, and part of the
convex surface of the liver. It is thus difficult to quan-
tify the contribution of the lungs to total right duct
lymph flow. However, since the contribution from
thorax and liver are small, it is probably valid to
assume with Drinker (60) that "in the quiescent,
anesthetized dog the amount of lymph collected from
the right duct expresses the lymph delivery from the
contracting heart and moving lungs. If cardiac ac-
tivity is kept reasonably constant, the quota of right
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
duct lymph arising from the heart is constant, and
variation in output reflects conditions in the lungs."
A second variable is introduced by the occasional
occurrence of anastomotic connections with the
thoracic duct and thus free flow between the two ducts.
Freeman (76) found anastomotic connections in 1 2
out of 25 carefully injected and examined animals.
In the living animal, the presence of anastomotic
connections is obvious if right duct lymph is milky
rather than clear or becomes milky and increases in
rate of flow after pressure upon the abdomen, a
maneuver which is very effective in increasing flow
from the thoracic duct but not from the right duct.
Drinker (60) also suggested that shifting from natural
breathing to artificial respiration through a tracheal
cannula was a useful test, since artificial respiration
increases right duct flow but reduces thoracic duct flow.
Further tests can be made by introducing T-1824 dye
into one of the lungs via a long catheter or by injecting
the dye into the paw or a leg lymphatic. The former
procedure results in coloring only right duct lvmph
if no anastomotic connections are present; the second
procedure results in coloring only thoracic duct
lymph. Using these tests, our experience has been that
fewer than 20 per cent of dogs show functional anas-
tomotic connections.
It is interesting that, in spite of the difficulties
described as inherent in the study of right duct lymph,
estimations of its flow and protein concentration are
not too different from those originally reported by
Warren and Drinker for lymph collected from the
mediastinal lymph duct, and the protein concentra-
tions are similar to those found in leg and cervical
lymph. Thus Courtice (47) found that the average
flow from the right lymph ducts of dogs was 2.3 ml
per hour and the average protein concentration was
3.7 per cent, levels similar to those found in our labo-
ratory (197). As Yoffey & Courtice (234) point out
this would amount to about 2 g of protein daily or 3.6
per cent of the total circulating plasma proteins. The
lymph flow is small in terms of the rich blood supply
of the lungs. It should be borne in mind, however,
that these experiments were done on anesthetized dogs
in the supine position and levels of flow and concentra-
tion may only at best reflect minimum values.
The pulmonary circulation is a low pressure system
and pulmonary capillary pressure is ordinarily less
than plasma colloid osmotic pressure, a situation con-
ducive to "dry" lungs. According to the Starling
principle, edema would be expected to occur either
when capillary filtration pressure was high or protein
concentration reduced. Paine et al. (163), using the
heart-lung preparation, showed this to be true in ex-
periments in which a) they lowered the plasma pro-
teins by plasmapheresis or by replacement with
Locke's solution, and b) they elevated hydrostatic
pressures by imposing a left ventricular overload. The
onset and progression of pulmonary edema were al-
ways attended by an increase in the flow of lymph
from the right thoracic duct. They conclude that
measurement of an increased pulmonary lymph flow
is a reliable indicator of the presence of pulmonary
edema. Uhley et al. (210) and Rabin & Meyer (176)
also studied the relationship between pulmonary
hypertension, lymph flow, and edema. The former
investigators devised a technique to collect pulmonary
lymph flow more completely. Instead of cannulating
the right lymphatic duct, they create a chamber
within the right external jugular vein which traps
lymph between the outside of a tube secured in the
vein and the vein wall. Lymph is removed from the
chamber by a polyethylene catheter. In 13 anesthe-
tized, open-chest dogs under artificial respiration they
found average lymph flow to be 0.3 ml per hour.
This value is considerably less than that found by
others, as indicated above. Elevation of pulmonary-
venous pressure to 30 mm Hg by introduction of a
balloon into the left atrium resulted in an increase in
lymph flow to an average of 1 . 14 ml per hour, the rise
occurring about 15 min after inflation of the balloon.
After maintenance of elevated pulmonary venous
pressure and progressive increase of lymphatic flow
for approximately 30 min, critical pulmonary edema
ensued. The protein content of lymph paralleled
lymph flow. Both increased lymph flow and pulmo-
nary edema were generally promptly decreased with
relief of the high pulmonary venous pressures. The
authors conclude that the small absolute increase in
right duct lymph suggests that lymphatics were
unable to function significantly to relieve the acute
pulmonary edema. Rabin and Meyer raised left
atrial pressures by means of previously appropriately
placed snares and found that, with this method, an
acute elevation of left atrial pressure could be precisely
controlled at any desired level up to a mean of 60 mm
Hg in dogs with an intact thorax. Right lymphatic
duct flow did not increase at acutely elevated left
atrial mean pressures below 25 mm Hg, whereas flow
increased 3- to 4-fold at mean pressures above 25 mm
Hg. The total amount of lymph at maximum flow,
however, was only 0.3 ml per min. Lymph flow re-
mained elevated for as long as 1 hour after left atrial
pressure was restored to normal. Pulmonary edema
did not occur readily when left atrial mean pressure
PHYSIOLOGIC IMPORTANCE OF LYMPH
I°53
was elevated only slightly above plasma oncotic
pressure. It was observed onlv after a considerable
elevation of left atrial pressure, above plasma oncotic
pressure, was maintained for a period of one-half hour
or more. Chronic elevation of left atrial pressure was
achieved in 15 dogs. Left atrial mean pressure varied
from 10 to 23 mm Hg. The dogs were followed for 10
months. Right lymphatic flow did not increase at
chronically elevated mean pressure below 25 mm Hg.
They did not study flow at higher pressures because
they were unable to sustain left atrial mean pressure
above 25 mm Hg in any dog in the chronic group.
The animals that were brought to left atrial mean
pressure between 30 to 40 mm Hg, and in which these
high levels were presumably maintained, were found
dead in their cages with pulmonary edema 1 to 2 days
after the snare was tightened.
Drinker, while admitting that hemodynamic
changes might be responsible for pulmonary edema,
stressed the importance of changes in capillary per-
meability due to anoxia (60). He believed that "in-
creased pressure in the pulmonary capillaries does not
readily cause recognizable pulmonary edema unless
coupled with heightened permeability, most fre-
quently due to anoxia." His conclusions were based
on a variety of experiments by him and his group (65,
216, 217). Thus, forced breathing of dogs against re-
sistance without hypoxia did not cause pulmonary-
edema, although right duct lymph flow was aug-
mented. On the other hand, when anoxia was present
under the same circumstances, increased lymph flow
and pulmonary edema were evident. Increased capil-
lary filtration, according to Drinker, results in ac-
cumulation of fluid in the alveoli, interfering with
oxygen uptake. A vicious cycle is set up as the hypoxia
further increases capillary permeability and filtration.
Courtice & Korner (48), on the other hand, failed to
observe pulmonary edema when they made animals
breathe a mixture containing 1 1 per cent oxygen.
They believe that the results of the Drinker group can
be equally well explained by postulating an increase
in filtration pressure rather than changes in capillary
permeability. There is considerable evidence to indi-
cate that the permeability of systemic capillaries does
not change at the levels of anoxia produced in the
Drinker and in the Courtice and Korner experiments,
and the latter correctly conclude that there is no
evidence to indicate that the permeability character-
istics of the pulmonary capillaries are different with
respect to anoxia. Courtice and Korner gave large
infusions of Ringer-Locke's solution to rabbits breath-
ing low oxygen. The presence of anoxia led to edema
at a lower level of infusion than when anoxia was not
present. The authors believe that this effect can be
explained by the hemodynamic changes observed
(decrease in cardiac output, systemic vasoconstriction,
etc.) without postulating an increase in permeability.
It may be pertinent in this connection, however, that
Fishman et al. (71) recently reported that acute anoxia
in human subjects produces a rise in cardiac output
and that significant changes in pulmonary arterial
pressure (average 7 mm Hg) are not found except
when the arterial blood oxygen saturation is below
85 per cent. It also appears that acute hypoxia does
not affect the thoracic blood volume (80).
The conclusions of Courtice and Yoffey should also
perhaps be modified in light of more recent findings
in our laboratory (197). We injected radioactive
iodinated serum albumin and dextran fractions of
average molecular weights of 51,000 to 255,000 into
dogs anesthetized with Nembutal and followed con-
centration changes in plasma and thoracic and right
duct lymphs for 4 to 6 hours. At this time, when a
"steady state" had been established between plasma
and lymph, we infused 40 ml per kg of 5 per cent
serum albumin in 0.9 per cent saline. This resulted
in a significant and striking increase in the concentra-
tion of the injected radioactive iodinated albumin and
dextrans in right duct and thoracic duct lymph in
spite of increased lymph flows. We interpreted these
results (and earlier ones) to be the result of "stretch-
ing" of capillary pores as a result of raised hydrostatic
pressure resulting from the expanded blood volume.
It is conceivable that the hypoxia in Courtice and
Korner's experiments may have served to exaggerate
this phenomenon of the capillary wall. It is interesting,
in this connection, that the possibility of pore stretch-
ing was suggested by Casten & Kistler (40) as an ex-
planation of the acute pulmonary edema which they
observed in mice and rats following blast injury and
irradiation. They hypothesized: "The intercellular
cement substance of capillary walls is postulated to
consist of processes having elastic properties that tie
the cell walls together. These processes are assumed
to be normally under tension. If the intracapillarv
pressure increases, but remains below a critical value,
the tension of the intercellular processes may be still
sufficient to hold the cell mosaic tightly together and
prevent gross fluid leakage. If the internal pressure
exceeds this critical value, then the elastic processes
may be stretched to a degree which causes the cell
mosaic to be separated, and thereby permit gross fluid
effusion to occur. With still greater internal pressure,
the processes may be stretched beyond their elastic
1054
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
limit, hereby producing an irreversible damage to the
capillary wall. It is assumed that in the extravasation
stage following acute radiation damage, the elasticity
and the stretchability of these processes are greatly
reduced."
It should be obvious from the brief review given
that, in discussing the pathogenesis of pulmonary
edema, we are concerned with a syndrome, like cir-
culatory shock, in which many factors may be opera-
tive. In an individual case, a single factor may produce
indeterminable effects unless it is very powerful. On
the other hand, several factors may combine to pro-
duce the syndrome at a threshold lower than that
necessary for each to act singly. In the final analysis
we must assume that in the healthy animal, fluid and
proteins which leak from the capillaries are drained
off as an equal amount of lymph. Pulmonary edema
arises when capillary filtration exceeds the point
where the lymphatic drainage is adequate to main-
tain the relatively "dry" state of pulmonary tissue.
This concept, that pulmonary edema results from a
relative deficiency in lymph drainage is further
supported by the experiments of Foldi and his col-
laborators (75). They studied dogs in which they had
experimentally produced mitral insufficiency, mitral
stenosis, or bilateral vagal section and ligated the right
and thoracic lymph ducts and lymph nodes of the
anterior mediastinum. Only tying off the lymphatic
supply failed to produce edema as did each experi-
mental procedure under control conditions. When the
procedure was combined with lymphatic ligation,
however, edema ensued in most animals. Failure to
occur under the combined procedure was correlated
with lack of lymph congestion presumably to incom-
plete cutting off of lymph drainage.
Cardiac
The lymphatic supply of the myocardium was first
described by Rudbeck in 1653 and has been debated
ever since, particularly as to whether myocardial
lymphatic capillaries exist. Patek (167) described
three plexuses, subendocardial, myocardial, and sub-
epicardial. The subendocardial vessels comprise capil-
laries in a single plane which drain into the myocardial
plexus, a profuse system of interconnected capillaries.
According to Patek, there were no efferent lymphatics
but only anastomoses with the subepicardial system.
Rusznyak et al. (188), however, review more recent
work of Zhemcherzhnikova and Zhdanov showing
that "the musculature of the ventricles in man con-
tains reticularly arranged true lymphatic capillaries.
The loops of these capillaries are situated along the
fasciculated muscles." According to these workers, the
lymphatic system of the epicardium consists of a dou-
ble, intercommunicating deep and superficial plexus.
The efferent lymphatics are chiefly situated subepi-
cardially, i.e., on the surface, and follow the branches
of the coronary artery. The lymphatic trunks of the
anterior surfaces of the two ventricles unite to form two
lymphatics which run in the anterior longitudinal
sulcus from the apex to the base of the heart. The
lymphatic which unites the lymph vessels of the
posterior surface of the left ventricle runs in the
posterior longitudinal sulcus and reaches the anterior
surface of the heart in the coronary sulcus. Uniting
below the left auricle, these lymphatics form the
heart's main collecting lymph vessel which drains into
a bifurcation or laterotracheal lymph node. The
lymphatics of the posterior and part of the anterior
aspect of the right ventricle unite in the right efferent
main lymphatic trunk which starts in the posterior
longitudinal sulcus, passes over to the anterior surface
of the aorta and runs along the surface of the right
auricle to the cranial mediastinal lymph node which
usually lies on the aortic arch at the origin of the left
common carotid artery.
Miller et al. (146) have recently raised the question
as to whether lymphatic vessels exist in the heart valves
of the dog. They examined mitral valves in three
groups of dogs: /) stock dogs killed during the course
of other laboratory experiments, 2) "sham"-operated
dogs, and 3) dogs in which surgical obstruction of
cardiac lymphatic drainage was produced and which
were then killed at varying periods of time after
surgery. Only an occasional thin-walled channel
was found on histological study in the first two groups.
However, numerous channels, presumed to be lym-
phatics, appeared in animals with obstructed lymph
flow. The authors believe that interference with lymph
flow may play a direct role in heart valve scarring and
may provide an additional clue to the mechanism of
progressive valvular fibrosis in the years following an
inflammatory insult (such as rheumatic valvulitis).
These authors also reported ventricular endomyo-
cardial pathology produced by chronic cardiac lym-
phatic obstruction in the dog (145). Their results are
similar to those previously described by Rusznyak
(187) who reported many variations in the anatomy of
the cardiac lymphatic system. They did not, however,
find cardiac lymphatics entering the thoracic duct as
Rusznyak reported. They found interstitial edema
most often with dilatation of lymph capillaries. Dis-
seminated focal necrosis in the myocardium was pres-
PHYSIOLOGIC IMPORTANCE OF LYMPH
10
3D
ent in 3 dogs and left ventricular subendocardial
hemorrhages in 7 of 1 7 dogs. They describe other
changes and point out the merit of further investiga-
tion of cardiac lymphatics and the relationships of dis-
turbances of their function to pathologic states.
Rusznyak and co-workers have also described the
results of similar experiments done by their group.
They report in detail the electrocardiographic changes
seen after cardiac lymphatic obstruction and after
ligation of the coronary sinus and cardiac lymph
nodes. They also emphasize the possibility that lym-
phatic congestion may play an important role in the
pathogenesis of mitral stenosis and other cardiac
diseases.
Drinker and his colleagues (64) are the only group
to date who have collected and studied cardiac
lymph. They reported measurements of flow in 10
dogs as varying between 0.31 to 1.65 ml per hour
(average 0.8 ml /hour) with no correlation between
dog weight, heart weight, blood pressure, and lymph
flow. Since only one lymphatic was cannulated and
since there are usually two main efferent trunks, the
total flow was probably approximately twice the
values obtained. This would mean that about 60 to
70 per cent of the total right lymph duct flow comes
from the heart. The lymph always contained a rela-
tively high concentration of protein. The average for
18 dogs was 3.69 per cent with a range between 2.50
to 4.73 per cent. Since right duct lymph contains
approximately the same concentration, this suggests
that cardiac and lung lymph have about the same
concentration of protein. This is also supported by the
values obtained by Warren & Drinker (216) in collec-
tion from a large lymphatic in the anterior mediasti-
num. Thus there is in the heart, as in all other tissues
studied, a continuous leakage of protein from the
capillaries and the rhythmic contractions of the car-
diac muscle insure its rapid removal by the extensive
lymphatic plexuses. These investigators also found
that cardiac lymph flow increased after the injection
of epinephrine and ephedrine, the rise appearing to be
correlated to the increased cardiac work. Experiments
with a Starling heart-lung preparation designed to
simulate exercise (increased input and peripheral
resistance and addition of epinephrine) showed a
marked increase in lymph flow to 24 ml per hour. As
Yoffey & Courtice (234) point out, even if we accept
a figure of 18 ml per hour, a hound engaged in a 12
hours' chase would be putting out 216 ml of cardiac
lymph. With a heart weight of 91 g, this would mean
2.4 ml of lymph per g of heart during the 1 2 hours.
Renal
Although a relatively extensive literature has ac-
cumulated concerning the distribution of kidney
lymphatics and their possible role in clinical disorders
(88), the physiological role of renal lymph is less well
documented. This is due to the difficulty in cannulat-
ing the vessels because of their location and extremely
small size. The main hilar trunks are particularly in-
accessible to cannulation, whereas capsular lym-
phatics, although more accessible, are quite small and
difficult to cannulate. This has led some investigators
to the highly unphysiologic compromise of eviscerat-
ing animals and collecting thoracic-duct lymph on
the assumption that this lymph was now derived
solely from the kidney.
The first physiological experiments on renal lym-
phatics were made by Ludwig & Sawarykin (129),
who showed that ligation of a ureter was followed by
dilatation of efferent renal lymphatics. They did not
study lymph flow or composition under these condi-
tions, but their experiments form the basis of a more
recent elaborate study by Babies and his collaborators
(7, 8), whose work will be discussed later. Sugerman
et al. (204) were probably the first to collect renal
lymph directly and begin its characterization in dogs.
They cannulated both capsular and hilar lymphatic
trunks and reported a wide fluctuation in flow and
protein concentration. The slower the lymph flow,
the greater was the concentration of protein. Their
average figure for flow from 1 1 dogs was 0.0232 g per
min (1.392 g/'hour) and for protein concentration
1 .84 g per cent. Of interest was the finding of a higher
average urea concentration in renal lymph (69.7
mg%) than in plasma (53.1 mgTt). In some animals,
the lymph urea concentration was considerably higher
than that in plasma of the renal artery or vein. These
findings posed two questions: /) Do the renal lym-
phatics drain only the larger collecting ducts of the
kidney, thus accounting for the high urea content of its
lymph? 2) Is renal lymph derived from tubular reab-
sorbed fluid, the blood plasma, or from both types of
fluid? Attempts to answer these questions were made
by Kaplan et al. (107) who determined and compared
the glucose content of renal and cervical lymph sam-
ples as well as their inulin content during an intra-
venous infusion of inulin. The average concentration
of glucose in 8 renal lymph samples was 92.7 mg per
cent and 10 1.9 mg per cent in 8 cervical lymph sam-
ples. They concluded that this high glucose concen-
tration in renal lymph suggests that renal lymph could
not be derived exclusively from the relatively sugar-
1056
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
free fluid contained in the larger collecting ducts of
the kidney. The average concentration of inulin in
renal lymph from 8 dogs was found to be 82.5 mg
per cent or 94 per cent of the plasma concentration.
The authors considered these results as evidence that
the composition of renal lymph is determined by the
character of both tubular reabsorbed fluid and renal
blood plasma. If it were derived exclusively from the
renal tubular reabsorbed fluid, its inulin content
would be practically nil and if it were derived exclu-
sively from renal blood plasma, its inulin content
would be equal to that present in plasma.
Swann and his colleagues (205) reported measure-
ments on renal capsular lymph from 5 dogs. Lymph
flow was between o. 1 to 0.3 ml per hour. They also
found urea concentration of renal lymph to be higher
than plasma and glucose concentration to be lower.
They found total protein content to be quite constant
at about 3.2 g per cent, about half that present in
plasma, and failed to observe the inverse relation of
protein content to lymph flow reported by Sugerman
et al. Electrophoretic separation of the proteins showed
that their distribution in lymph was similar to that in
plasma.
About 5 years ago, my colleague, Dr. S. J. LeBrie,
and I began a comprehensive study of renal lymph as
part of a general study of lymph and lymphatics. We
have collected renal capsular lymph in a variety of
experimental situations which will be detailed below.
We have found the lymph flow to be quite variable
and unrelated to sex, age, or size of the animal. The
average control flow for 63 dogs studied up to the
present time is 0.0128 ml per min or 0.768 ml per
hour. Total protein concentration for 40 dogs averages
2.76 g per 100 ml or about half of plasma concen-
tration.
We find protein distribution to be similar to that
of plasma, confirming the findings of Swann et al.
All plasma proteins, including fibrinogen, are present
in renal lymph. Potassium concentrations appear to
be identical in lymph and plasma. We previously
reported (121) that the average lymph sodium con-
centration for 30 dogs was 1 1.3 per cent higher than
in plasma. Addition of more data and refinement of
experimental procedures suggests that this value may
be too high and the difference may or may not be sig-
nificant. Limited data on osmotic pressure also have
been equivocal. Data on 30 dogs show lymph chloride
concentration to average 13 per cent higher than
plasma, a difference which is statistically significant.
We have some indication that bicarbonate concen-
tration is lower in lymph than in plasma. Obviously
we need more data to clarify the situation with respect
to these constituents. These data are being accumu-
lated.
It is of interest to compare renal lymph flow with
urine flow in the same kidney. It is reasonable to
assume that there are at least ten lymphatics draining
each kidney. Using the average flow which we ob-
tained from one lymphatic, the average total lymph
flow from both kidneys would amount to o. 1 28 ml
per min, which is approximately half of the average
amount of urine flow. In some animals with high
lymph flows, the amount of capsular lymph drained
equals the amount of urine formed. Similar admittedly
rough calculations for protein yield a value of 10.2 g of
protein returned to plasma per day via capsular
lymphatics of both kidneys. It should be emphasized
that these are average values for the anesthetized dog
and probably reflect minimum levels.
Renal lymph flow is markedly increased by raising
venous pressures. This was first shown by Schmidt &
Hayman (195) and later confirmed by Katz & Cockett
(109) and by us (123). Schmidt and Hayman analyzed
the changes in thoracic duct lymph flow in eviscerated,
hepatectomized, and uninephrectomized dogs follow-
ing ligation of the remaining functional renal vein.
They concluded that the increase in renal lymph flow
was responsible for the observed rise in thoracic duct
lymph flow when venous pressure was raised. Katz
and Cockett likewise concluded that changes in renal
lymph were responsible for changes in thoracic duct
lvmph which they observed. They observed an in-
crease in thoracic duct lymph flow and sodium concen-
tration as well as a decrease in urinary flow and
sodium concentration when venous pressure was
raised. The changes occurred only when the kidneys
were intact. Haddy and his colleagues (94), in ex-
periments designed to study pressure and flow rela-
tionships in the kidney, also reported that renal
lvmph flow increased as a function of venous pressure.
In our experiments, we collected capsular lymph and
raised the venous pressure by partially occluding the
inferior vena cava with a balloon catheter. We also
measured protein and electrolyte changes. Renal
lvmph flow increased about five times during the
periods of increased venous pressure and the flow
from one lymphatic equaled and often exceeded
urine flow from the same kidney. Electrolytes and
protein levels changed proportionately except at high
venous pressure levels (30-35 cm HaO) when dispro-
portionately high levels of protein were found in renal
lymph.
In discussing the significance of the changes in lym-
PHYSIOLOGIC IMPORTANCE OF LYMPH
1057
phatic flow and composition produced by increased
venous pressure, we suggested that they might have
some bearing on the problem of plasma volume in-
crease and sodium retention in congestive heart fail-
ure, a syndrome in which venous pressures not infre-
quently approximate the high levels used in our
study. We pointed out that under conditions of signifi-
cantly increased venous pressure the total lymph flow
from both kidneys may amount to as much as 2400 ml
per 24 hours. This would represent a total of ±379
meq of sodium not excreted by the kidneys but re-
tained by the lymphatic system. A study of kidney
lymph flow in postural proteinuria might also be of
interest. Bull (32) believes that "A rise in pressure in
the inferior vena cava is produced by compression of
the vessel against the spine by the posterior surface of
the liver. This pressure is conducted back to the kid-
ney, inducing passive congestion and proteinuria.
The compression occurs when the subject is in a lor-
dotic posture and when the anterior surface of the
liver rotates inferiorly. This rotation of the liver
normally occurs when the subject is lordotic and is
maximal in the erect lordotic posture." Goodwin &
Kaufman (89) suggest the possibility of thoracic duct
or cisterna chyli lymphatic obstruction and retrograde
lymph flow as a possible explanation of the proteinuria
and cite the report of Lowgren (128) as suggesting this
explanation.
If we agree that one of the primary functions of the
lymphatic system is to return to the vascular system
those proteins and other large molecules which have
leaked out of the blood capillaries, our accumulated
data emphasize that kidney lymphatics are no excep-
tion. This is a function of considerable importance for
the kidney. Maintenance of a relatively low concen-
tration of interstitial protein is necessary for the main-
tenance of the countercurrent action in the kidney.
This concept visualizes the vasa recta as a counter-
current exchanger carrying off salt and water. Gott-
schalk & Mylle (go) believe the efficiency of the coun-
tercurrent exchange in the vasa recta to be critical,
"for they probably remove not only the blood enter-
ing the medulla, but also the water that diffuses from
the thin descending limbs of the loops of Henle and
the collecting ducts. This water, with solute isosmotic
for the particular level of the medulla, presumably
moves into the vasa recta because of the gradient of
its chemical potential established by the colloid os-
motic pressure of the plasma proteins, since the hydro-
static pressure in the capillaries and interstitium are
the same." The gradient of colloid osmotic pressure
between the interstitium and the plasma can be main-
tained only if the colloid osmotic pressure of the in-
terstitium is kept well below that of the plasma. We
believe that this is an important function of renal lym-
phatics, the maintenance of a relatively low oncotic
pressure in the interstitium and thus the establish-
ment of a gradient with the higher oncotic pressure
within the vasa recta. Thus, as the lymphatics carry
off plasma protein that has pooled in the medullary
interstitium, the colloid osmotic pressure of these pro-
teins draws water with a higher or lower solute con-
centration, depending upon the level of the counter-
current gradient at which the lymph is formed. The
medullary and cortical lymph passes into lymph col-
lecting trunks, mixing the two and thus reducing the
electrolyte concentration and osmolarity. Some col-
lecting trunks lease the kidney through the cortex
while others follow the path of the artery and vein to
the hilus. When renal venous pressure is increased,
more protein is lost in both cortex and medulla due to
increased hydrostatic pressure. Thus, although filtra-
tion of protein increases as does water, the concen-
tration of filtered electrolytes does not increase over
control values. We assume that filtered electrolvtes
from cortical capillaries are isosmotic with both
plasma and cortical interstitial tissue, while filtered
electrolytes from the vasa recta are hyperosmotic to
plasma but isosmotic to the countercurrent gradient
in which they lie. Thus, if these fluids are mixed, one
might expect the concentration of sodium in milli-
equivalents per liter to remain unchanged. This is es-
sentially what we find.
In terms of the above discussion, we would expect
ligation or obstruction of lymphatic outflow to produce
edema and significantly alter kidney function. Kaiser-
ling & Soostmeyer (105) succeeded in tracing the
lymphatic vessels to the main hilar branch in rabbits
and in tying it off. The kidney began to swell immedi-
ately and had reached double its original size within
10 to 15 min as a result of massive interstitial edema.
There was a marked increase in urinary output on
the side with lymphatic ligation and, significantly,
the urine from the experimental side had a specific
gravity of 1 013, whereas the control side had a specific
activity of 1 035. Later, the urine flow diminished,
proteinuria was present, and the kidney and its
parenchymal cells degenerated 8 to 10 days after the
lymphatics were ligated.
These and other similar studies and their implica-
tions are discussed at length by Babies (7, 8) and
Rusznyak et al. (189).
Earlier mention was made of the fact that the first
experiments on renal lymph were those of Ludwig &
o=s8
105'
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Sawarykin ( 1 29) who noticed that ligation of a ureter
was followed by dilatation of the efferent renal lym-
phatics. Similar observations have since been made by
many investigators (88, 153, 154, and LeBrie and
Mayerson, unpublished) who have also shown that
renal lymph flow is significantly increased. Katz (108)
measured renal lymphatic pressure in two dogs with
pyelonephritis and found the pressures to rise when
either the renal vein or the ureters were compressed.
It is apparent that the renal parenchyma shows no
significant changes for a number of weeks after the
experimental ligation of the ureter; the only histo-
pathological symptoms that can be observed for a
considerable time are a high degree of interstitial
edema and marked dilatation of lymphatics. Babies &
Rrnyi-Yamos (7) ascribe the survival and continued
performance of the hydronephrotic kidney to the fact
that urine passes from the renal pelvis into the inter-
stitial space of the kidney where it is continuously
absorbed into lymphatics. Histamine is presumed to
be liberated, increasing capillary permeability and
transudation of protein. This protein, too, is carried
away by the lymphatics. If, on the other hand, the
lymphatics are also tied off, necrosis is seen within a
few days [see Babies & Renyi-Vamos (7) andRusznyak
el al. (189) for detailed discussion].
The general effect of diuretics is to increase renal
lymph flow. Reference has been made to the work of
Schmidt and Hayman who showed this to be true for
phosphate, sodium chloride, and caffeine. We have
collected renal capsular lymph in dogs during diuresis
produced by sodium chloride, urea, mannitol, and
mercury (LeBrie and Mayerson, unpublished). Under
the conditions of our experiments, we obtained the
most marked diuresis with urea and mannitol and the
least with mercury. Mannitol produced the greatest
increase in lymph flow (average 10 dogs = 587%),
while mercury and urea produced the least (Hg, 15
dogs, 42 %; urea, 7 dogs, 41 '", ). These experiments are
being continued and attempts are being made to eluci-
date the mechanism of the changes in flow, electro-
lyte and protein concentration, and to explain the
differences seen with the different diuretics.
We have also studied the influence of uranium-
nitrate injury on the flow and composition of renal
lymph (122). Lymph flow was increased approxi-
mately 15 times in the experimental animals. Like-
wise, the increase in lymph flow with mannitol
infusion was about twice as great in the experimental
animals (iooofT increase) as in control animals
(542% increase). The experimental animals showed
a significant proteinuria and decreased urine flows,
and the data appear to be consistent with the histo-
logic findings of primary damage to the distal seg-
ment of the proximal tubule.
The finding of an increased lymph flow when
ureters are obstructed and diuretics are administered
has given rise to the concept emphasized in clinical
literature, that the renal lymphatics act as a "safety-
valve" mechanism, capable of taking the extra load
from the kidney under conditions of overload.
Backflow from the kidney pelvis to the renal lym-
phatics has also been suggested by experiments of
Murphy & Myint (153) and Goodwin & Kaufman
(89). The former introduced glucose and the dye
T-1824 into the renal pelvis and found these sub-
stances earlier and in greater concentration in lymph
of the cisterna chyli than in renal or femoral blood.
The latter injected radioactive Diodrast into the
renal pelvis during ureteral occlusion and found
the same radioactivity at the same time in thoracic
duct lymph and in the control vascular area. More
work along these lines is needed, particularly defining
the mechanisms involved and the effects of pyelo-
lymphatic backflow in kidney disease. However, it
may be pertinent to emphasize, as has been previously
suggested, that the general function of the lymphatic
system is to act as a "safety valve" and as an acces-
sory circulation, clearing the interstitium of excess
substances which leak out of or are not absorbed
directly into the blood stream and returning them to
the blood circulation. This is not a peculiar or special
function of renal lymphatics. Thus, an overload of
the circulatory capacity as produced by a large
intravenous infusion results in an increased lymph
flow, etc. The particular importance of lymph with
respect to normal renal function lies in the fact that
the oncotic pressure of the interstitium must be kept
low in order for the vasa recta to act as a counter-
current exchanger. In the absence of adequate
lymph drainage the kidney becomes unable to
concentrate urine (105).
LYMPH AND LYMPHATICS IN SHOCK
Anaphylactic Shock
Petersen & Levinson (173) found that injection of
antigen into dogs resulted in an increased perme-
ability of splanchnic endothelium and subsequent
reaction of the hepatic parenchymal cells with the
antigen. In further work, Petersen & Hughes (172)
showed the injection of egg white into dogs sensitized
PHYSIOLOGIC IMPORTANCE OF LYMPH
I°59
to egg albumin to result in an immediate and marked
increase in thoracic duct lymph flow with increased
concentrations of calcium, amino nitrogen, and
magnesium, and decreased concentrations of sodium
and potassium. They did not measure proteins.
Dragstedt and his colleagues (55-57, 82) reported
that, in the dog at least, the vasomotor symptom
and death occurring in anaphylactic shock are
brought about by the sudden discharge into the
circulating blood of a vasodepressor, smooth muscle-
stimulating substance which is apparently histamine.
They were able to detect this substance in blood and
thoracic duct lymph for brief periods of time after
the assaulting or shocking dose of serum, and to corre-
late its appearance with varying grades of severity of
the shock in such a way as to indicate that it had a
causal relationship to the shock symptoms. More
specific results were recently reported by Logan (127)
who showed that bovine globulin is a satisfactory
antigen for sensitizing rats when given intraperi-
toneally simultaneously with Bordetella pertussi
vaccine. Intestinal lymph of animals so sensitized
contained increased amounts of histamine when
collected 6 min after intravenous injection of the
shocking dose. The amount of lymph histamine was
roughly proportional to the degree of shock. The rate
of lymph flow increased 8 to 25 times during the 12
min immediately after administration of the shocking
dose and the lymph contained o to 0.02 /ig histamine
per ml, an amount similar to that in plasma.
The lymphagogue action of histamine is well
known and can readily be demonstrated by the
intravenous injection of small amounts of the sub-
stance (95). This effect has usually been ascribed to
dilation and increased permeability of capillaries,
although the exact mechanisms involved have never
been clearly defined. Rusznyak et al. (189) review
much of the evidence and report experiments done
in their own laboratories by Szabo and Magyar. The
latter injected a dextran fraction of approximately
the same molecular weight as albumin simultaneously
with histamine and with Evans blue. In dogs, flow
of intestinal and hepatic lymph increased two to
three times and remained high; appearance of dextran
and dye-labeled albumin in lymph was much sooner
as was equilibration between plasma and lymph. The
authors believe they have ruled out the factor of
increased filtration pressure in favor of increased
capillary permeability. Repetition of the same experi-
ments in cats gave entirely negative results, i.e., no
increased lymph flow or accelerated equilibration.
The authors explain these differences as being due
to the fact that in the dog not only the capillaries
and small veins but also the arterioles are dilated by
histamine (and only the larger visible arteries con-
stricted), while in the cat this constrictor effect begins
more peripherally in the arterioles. The authors
take issue with the assumption by Krogh (117) that
dilation of capillaries leads to increased permeability,
since capillaries dilate in the cat quite as much as
in the dog without showing increased permeability.
Changes in blood coagulability are among the
important findings in the anaphylactic reaction
(67, 104) and result from the release of heparin from
the liver. The possibility that this substance may be
transported from the liver to the blood stream by-
way of the thoracic duct was first suggested by Gley
(87) who found that ligation of the liver lymphatics
prevented the incoagulability of blood following
peptone shock. White & Woodward (228-230)
showed that the incoagulability of thoracic duct lymph
following an anaphylacytic or peptone-induced
shock is due to heparin and that the main portal of
entry for this substance into the blood stream is via
the thoracic duct from the liver. The concentration of
heparin in lvmph in these situations was greater than
in arterial or hepatic venous blood and it was fre-
quently present only in thoracic duct lymph. When
heparin was given intravenouslv the concentration of
heparin was the same in thoracic duct lymph and in
plasma. Furthermore, heparin did not appear in
cervical or right duct lymph unless it first appeared in
blood. Removal of the liver prevented the appearance
of heparin in peptone-shocked dogs. These findings
confirm the liver as being the origin of heparin under
these circumstances. Heparin was not released into
thoracic duct lvmph or blood during a hemorrhage-
induced shock, thus ruling out hypotension as a factor
in the heparin release. It is suggested that the release
of the heparin may depend upon limited cellular
reactions and not upon general cellular activity.
Traumatic Shock
There has been considerable interest in studying the
participation of the lymphatic system in traumatic
shock. These studies have taken different forms.
The Hungarian workers (188) have done a consid-
erable amount of work in which they have measured
lymph flows from various areas and used the compo-
sition of lymph as a measure of capillary permeability.
There has been a recurrent interest in the presence
or absence in lymph of a "toxic" substance which may
or may not have come from the plasma.
io6o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
The question as to whether capillary permeability
is altered in shock has been a hardy perennial. A
detailed discussion of this topic would be out of
place at this point. It is, however, appropriate to
call attention to the studies of the Hungarian group
referred to above in which they have made extensive
observations on the role of lymph and lymphatics in
dogs during and after traumatic shock. They have
used the same general approach as we have used
(141, 219) of introducing dextrans of molecular
weights similar to those of albumin and globulin into
the blood stream, following their disappearance from
the blood stream and their appearance in lymph from
various regions. Since radioisotopes were not available
to them, they used the dye, T-1824, which is known
to bind onto albumin and thus constitute a label.
Their results on control animals resemble in general
those which we obtained; albumin or de.xtran of
molecular weight of about 50,000 appeared in thoracic
duct lymph within 10 min and, in their experiments,
the average de.xtran concentration was 29 per cent
of that in plasma at 1 5 min and about 75 per cent at
60 to go min. In tourniquet shock produced by
arresting the circulation of the hind legs for 5 hours,
they found a more rapid disappearance of the dextran
from the blood stream but a much slower appearance
and accumulation in thoracic duct lymph. At 90
min there was only an average of 43 per cent of
dextran in the lymph. Lymph flow was considerably
reduced. These workers also studied cervical lymph
and, while their results were not definitive, it seems
reasonably certain that the capillary permeability in
peripheral regions was not increased. Attempts to
find the reasons for the decreased thoracic duct
lymph flow and the delayed appearance of protein
and dextran were not successful. It was apparently
not due to diminished hepatic-lymph formation or to
lymphangiospasm. They interpret the faster disap-
pearance of the dextran from blood as reflecting the
increase in capillary permeability in the ischemic
area rather than a generalized increase in perme-
ability. One of the group (226) has extended some
of the work to hemorrhagic and burn shock and
again finds a decreased thoracic duct lymph flow
which parallels the severity of the shock. He also
interprets his data as denying any generalized
increase in capillary permeability.
Since edema is prominent in the ischemic areas,
these investigators went on to study the flow of leg
lymph and the behavior of albumin and dextran in
the hope of ascertaining whether the edema was due
to an inability of the lymphatic system to cope with
an increased tissue fluid formation or whether there
was injury to the lymphatics caused by the ligatures.
Although the small lymph flow in leg lymphatics
precluded quantitative data, there was no question of
the direction of change and that protein leakage was
increased in the ischemic area. The increased lymph
flow argued against lymphatic injury or occlusion as
factors in the edema production.
The occurrence of a vasoconstrictor substance in
blood during shock induced by trauma, hemorrhage,
and burns was reported in dogs by Page (161) and
denied for ischemic compression shock (91). Rapport
et al. (180) also reported the occurrence of a "toxic
factor" in tourniquet shock in rabbits.
In 1943, Blalock (ig) reported the results of
experiments in which he produced crush injury in
anesthetized dogs by applying a press to a hind leg.
Thoracic duct lymph collected from these dogs after
removal of the pressure and injected into other dogs
brought on a decrease in blood pressure and death
of some of the animals. Less marked results were
obtained when trauma was produced by striking the
legs with a blunt instrument. Blalock explained his
results as due to the presence of a toxic substance in
lymph of the traumatized animals. Katzenstein
et al. (1 10) reported similar results in shock produced
by tourniquets around the hind legs. These authors
appreciated the possible vasodepressor effects of
large doses of Nembutal which they used but showed
that when narcosis was controlled to avoid vaso-
depression, injection of thoracic lymph from normal
dogs had no effect. In contrast, a fall in blood pressure
followed in 50 per cent of the animals injected with
lymph of shocked animals. The problem was further
studied by Nathanson and his collaborators (155), who
devised a method of producing tourniquet shock in
dogs which permitted the collection of muscle
exudate. They collected the exudate, which accumu-
lated after muscle anoxia, and injected it into the
same or recipient dogs (6). Shock was produced in
only 25 per cent of the animals tested. The incon-
stancy of the presence of the toxic factor suggested
that the factor was an extraneous agent, not present
in the usual cellular constituents and metabolic
products found in all muscle exudates, and possibly
bacterial in origin. They further showed (235) that
the toxic properties of a collection of pooled muscle
exudates were contained in a nondialyzable fraction,
could be salted out between 0.25 and 0.7 saturation
with ammonium sulfate and were, therefore, prob-
ably protein in nature. The toxic substances were
tentatively classified as an aminoexopeptidase and a
PHYSIOLOGIC IMPORTANCE OF LYMPH
I06l
trypsinase which were present in the exudates. Free-
man & Schecter (77) tested leg lymph obtained
from dogs whose hind legs were traumatized or heated
and found that it produced an increase in permeability
as judged by leakage of dye when injected into re-
cipient animals. Arterial and venous serum and
plasma also contained a similar factor which increased
capillary permeability, and the authors concluded
that it was likely that the presence in lymph of a
substance capable of producing an increase in capil-
lary permeability is dependent upon the appearance,
after trauma, of blood plasma in the lymph draining
from the extremity. On the other hand, Lindner
et al. (126) failed to find any evidence of a perme-
ability factor either in lymph or plasma in shock
produced by manipulation of the intestine. Their
experience was similar to the earlier one reported
by Dragstedt & Mead (57), who produced shock by
sustained trauma with a padded hammer to one or
both hind legs, by trauma to the intestine, or by a
combination of the two methods.
Burns
Some work has been done on the study of lymph
in burns. Aldrich (2) collected leg lymph from burned,
anesthetized dogs and perfused it through rabbit
ears. Blood flow as measured by drop rate definitely
decreased when lymph from burned animals was
used as compared to lymph from healthy clogs. No
attempt was made to identify the vasconstrictor
substance.
Glenn and his colleagues, in Drinker's laboratory,
studied the changes in lymph composition after leg
burns produced with hot water in calves (85, 86,
170). Cervical and leg lymph was followed. Lymph
flow in the burned legs was significantly increased as
was the protein concentration of the leg lymph.
Cervical lymph, however, did not show the increase
in protein. Electrophoretic studies showed the occur-
rence of a new protein in the lymph from the burned
leg, a component migrating with half the speed of
7-globulin. Cope & Moore (44) also reported a signifi-
cant increase in capillary permeability following hot
water burns of legs of dogs. They injected radioactive
colloidal dyes into the blood stream and measured
their appearance in leg, cervical, and thoracic duct
lymph before and after the burn. They also injected
radioactive bromine, which they found to appear in
lymph from the three areas within 5 min and to
reach equilibrium with serum in 20 min. In contrast,
the colloidal dyes were slower in appearance in
lymph and no equilibrium was established with serum
under control conditions. Following the burn, the
concentration of radioactive colloids in lymph of the
burned leg rose abruptly and approached that
encountered after injection of radioactive bromine.
The specific activity of protein was actually higher
in lymph than in serum after the burn. In confirma-
tion of Glenn et al., they also found that the increased
capillary permeability was usually restricted to the
burned leg. A rise in colloid concentration in cervical
lymph was observed in only one dog.
Permeability Factors
During the last decade, considerable interest has
been aroused in the presence in plasma of endogenous
substances which, when activated, induce pathologi-
cal increases in capillary permeability (144). Two
classes of natural mediators have been suggested:
7) the pharmacologically active amines, histamine,
and hydroxytryptamine; and 2) proteases and prod-
ucts of proteolysis. This latter group includes the
proteases of plasma (plasmin, the serum globulin
permeability factors, and polypeptides like leuko-
taxine and bradykinin). The groups overlap in that
polypeptides may act as histamine liberators. There
is direct and indirect evidence that these substances
participate in the mediation of the response to
injury, but much more evidence is needed to define
their role in the healthy animal and in animals
suffering from hemorrhage, burn, or other trauma.
It would be interesting to extend the observations
of Miles & Wilhelm (144) to other substances,
species, and experimental conditions. These investi-
gators showed the presence in the guinea pig of the
precursor (pro-PF) and the inhibitor (IPF) of one of
the globulin permeability factors, both in intercellular
perfusates of skin and in normal lymph from the
cervical lymph ducts (143). It appears that the
proteins constituting the pro-PF/IPF system of the
blood, like other plasma proteins, pass continuously
via the extravascular tissues to the lymph, and that
the extravascular tissues, including the outer surface
of the capillary wall, are bathed in tissue fluid con-
taining pro-PF.
PERMEABILITY OF LYMPHATIC VESSELS
Although a considerable amount of work has been
done relative to the permeability of blood vessels, we
have very little definitive information regarding
1062
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
permeability characteristics of lymph vessels. Light
microscope studies suggested that lymph and blood
capillaries are morphologically similar, a suggestion
which has been confirmed by recent studies, particu-
larly those of Casley-Smith & Florey (39a). These
authors showed that there were no species differences
in lymphatics of mice, guinea pigs, and rats, and
that the lymphatics of the ear and the deep lymphatics
of the diaphragm and colon were similar. In general,
the structure of the lymphatic capillaries and lacunes
appeared to be similar to that of blood capillaries.
All the lymphatic endothelial cells contained many
vesicles and caveolae intracellulares. No fenestrations
in the endothelium were seen, but some intercellular
junctions were patent, especially in diaphragmatic
lacunes. The basement membrane was less regular
than that of blood capillaries or of mesothelium
and in many places, especially in diaphragmatic
lacunes, it appeared to be absent. These results are
similar to those of Palay & Karlin (164) and French
et al. (77a). The absence of a definable basement
membrane would not, as Casley-Smith and Florey
point out, fully differentiate lymphatic from vascular
endothelium, since the endothelium lining large
blood vessels may have at best a very tenuous base-
ment membrane. The significance of the absence of
fenestrations in the lymphatic endothelium also
remains questionable in the absence of definite
information as to the importance of their presence in
determining permeability characteristics. The signifi-
cance of their other findings will be discussed further
below.
The permeability pattern in lymphatics presents an
interesting and challenging problem. It is obvious
from the preceding discussion that proteins, chylo-
microns, and lymphocytes are normal constituents of
lymph as routinely collected from healthy animals.
Experimentally, bacteria, viruses, red blood cells,
graphite particles, etc. have been shown to penetrate
the lymphatic system with no apparent difficulty.
Lane Allen (ia) showed that every type of cell which
occurs normally in tissue fluid and blood will pene-
trate lymphatic endothelium. He felt he had identified
every cell of the hematopoietic series, except giant
cells, in diaphragmatic lymph after intraperitoneal
injection of bone-marrow suspensions. Likewise, he
found that the entire series of lymphoid cells will
enter through lymphatic endothelium. The large
amount of literature published before 1956 describing
these experiments has been thoroughly reviewed by
Yoffey & Courtice (234). The more recent publica-
tions will be discussed later.
In spite of the apparent ease with which substances
can penetrate into the lymphatic vessels, the avail-
able evidence suggests that once these substances are
in the lymphatic system, they are retained and
eventually find their way into the blood stream via
the larger ducts. Thus MacCallum (132a) retroin-
jected the lymphatics of the diaphragm and failed to
force suspended particles back into the peritoneal
cavity, except when he used pressures sufficient to
rupture the lymphatics. Hudack & McMaster
(100a) injected dyes into the ears of mice and studied
the escape of these substances from the lymphatics.
They reported that poorly diffusible dyes (pontamine
sky blue, Chicago blue 6B) which pass with difficulty
out of blood capillaries into the tissues, tend to be
retained by the lymphatic wall as well, whereas more
highly diffusible dyes (trypan red, bromphenol blue,
and Neptune blue) pass out with ease. Rusznyak
et al. (189) have more recently reported similar results
using fluorescent dyes (thiazine red, acridine yellow)
in intestinal lymphatics of cats. Hudack and Mc-
Master concluded that "all the evidence we have
obtained supports the view that permeability of the
lymphatic wall resembles the permeability of the
capillary wall in its essential features and perhaps in
its degree." Drinker & Field (61 a) retroinjected
lymphatics of the frog web with graphite acacia and
found no passage of the graphite particles until rupture
resulted from excessive pressure. Pullinger & Florey
(175) found that when they injected graphite particles
into ear lymphatics of the mouse, the fluid leaked out
but the graphite particles remained. Similarly, Lee
(124a) found that large particles of centrifuged,
dialyzed India ink were retained in lymph vessels,
whereas small particles passed through. Courtice &
Steinbeck (50a) attempted to evaluate lymphatic
permeability by injecting T-1824-labeled plasma
proteins intraperitoneally into rabbits and collecting
lymph containing the protein from the exteriorized
thoracic duct. They found that the injected proteins
were almost entirely absorbed by the diaphragmatic
lymphatics. In further work (50b) they demonstrated
that ligation of the parasternal lymph channels in
rabbits, rats, and guinea pigs prevented the dye-
labeled protein from reaching the circulation via the
thoracic duct, but instead, after entering diaphrag-
matic and mediastinal lymph channels, it proceeded
to leak into the mediastinum and pleural cavities.
They concluded from these experiments that lympha-
tics leaked protein, a conclusion open to question in
view of the obviously unnatural conditions of their
experiments. Ligation of the parasternal lymph chan-
PHYSIOLOGIC IMPORTANCE OF LYMPH
I ul, ;
nels and the resultant increased intralymphatic
pressure and distention of the lymph vessels might
conceivably permit leakage of substances which
ordinarily would be retained within the lymphatic
vessels as has been previously shown to be true in
earlier work cited above. This has been shown to be true
of blood capillaries (197). Furthermore, the fact that
Courtice and Steinbeck found no protein leakage when
the ducts were not ligated suggests that the lymphatics
of the anterior mediastinum, under normal circum-
stances, will not permit leakage of significant amounts
of the protein which is contained by them.
In a recent study (168) we attempted to obtain
answers to the questions: a) Does the capillary filtrate,
once it is in the lymph ducts, empty without loss into
the venous circulation or is protein free to pass out
of the lymphatic vessels throughout their lengths?
b) Does any protein pass into the blood capillaries?
c) Is any considerable amount of protein phagocytized
by the reticuloendothelial cells of lymph nodes?
Furthermore, to what degree, if any, is lymph
shunted to the blood stream through lymph-blood
anastomoses without returning via the thoracic
duct? Anomalous shunts, consisting of multiple
outlets of right and left ducts, have been described
as well as shunts between the right and left thoracic
ducts in about 15 per cent of dogs (234, 197). One
phase of our study involves the cannulation of a leg
lymphatic in an anesthetized dog, the infusion into it
of substances of different molecular weights, and
obtaining and analyzing samples of thoracic duct
lymph and plasma. The infused material thus travels
through a number of lymph nodes, through lymphatic
vessels of different sizes, and through capillaries.
The second approach to the problem is to utilize a
preparation developed some years ago by Drinker
and his group (63). One afferent and one efferent
duct going to and from the popliteal node are isolated
and catheterized with polyethylene tubing. All other
lymphatics are tied off. Test substances are infused
through the node from the afferent side and collec-
tions are made on the efferent side. Nodal and systemic
plasma are also obtained and analyzed.
When radioactive iodinated albumin is infused into
a leg lymphatic, there is a typical, consistent pattern
of appearance of the albumin in thoracic duct lvmph
as shown in figure 5. There is a lag time of approxi-
mately 10 min between the start of the infusion into
the leg lymphatic and the appearance of measurable
amounts of radioactive albumin in thoracic duct
lymph. This is followed by an abrupt rise to a plateau
which is maintained at an approximately constant
2000
1000 ■
ALBUMIN
DEXTRAN
(AVER M W 51.300)
/\
z
/~^\\
z
■ 5
1, \ U
<
Ct
D
1 ' i\
H
CD
\'\\
X
_l
\ \
LJ
<
1 I
Q
-
/ !1
'
Q
' \
cr
Z 1
\
UJ
a
11
5
"
\\
1-
z
N I
i\
Id
■tr '
t\
O.
- 1
ll
11
O
— . I,
ll
CD
i\
3
' -1 l
ll
U
_i I
u
\
5 '
a
(/)
■ x '
\\
5
<
c/1 1
\\
QC
•- 1
55
z '
\
_j
■ 3 1
_i
O .'
0 '
ll
ll
h
V
2
MINUTES
1 , 1 1 1
J 1 1 1 l__l 1
100
50
50
100
150
fig. 5. Concentration of dextran and I131 albumin in lymph
and plasma. Dextran and I131 albumin solutions infused cen-
trally into leg lymphatic of anesthetized dog at zero time at
rate of 0.5 ml/min. Infusions of dextran and albumin stopped
after 50 min and 0.9% saline infusion started at same rate for
next 1 00 min. All values are corrected for free iodine.
level for the 50-min duration of the albumin infusion
and for 10 min of a subsequent saline infusion. At
this time, radioactivity in thoracic duct lymph falls
sharply and continues to fall until the level approxi-
mates zero in about 140 to 150 min. Plasma radio-
activity rises to a maximum concentration after 60 to
90 min and remains at this level for the remainder
of the experiment. The maximum plasma concentra-
tion is less than 0.00 1 of the thoracic duct lymph
concentration during its 50-min plateau period.
Figure 5 also shows the similar behavior of dextran of
approximately the same average molecular weight
as albumin. It is apparent that these infused sub-
stances do not leave the lymphatic system and that
they return to the circulation primarily by the
thoracic duct. Actually less than 3 per cent of the
infused material reaches the circulation by routes
other than the thoracic duct, except in unusual cases
of right and left duct anastomoses. The experiments
1064
HANDBOOK ()!•' I'HVSIHLOGV
CIRCULATION II
5
CD
O
«_)
LLl
CO
\
or
UJ
I-
10
I-
z
ZJ
o
o
1600
.
1
1400
- , /
— 20p.c.UREA
1200
-J'i/V
1000
- 1
800
-ll
600
1 1
400
1 1
-II
1 1
200
f * \
MINUTES
1 1
"iC^>
j
50
100
150
fig. 6. Concentrations of urea in lymph. Same procedure
as in experiment shown in fig. 5.
with the isolated lymph node preparation suggest
that most of the 3 per cent or less not recovered from
the thoracic duct goes into the blood stream via the
popliteal node, and that there is little further loss in
the other nodes through which the lymph passes on
its way to the thoracic duct. We suspect that this is
due to the fact that we are infusing under some
pressure which is dissipated after the first node is
passed. Nisimaru & Irisawa (156), in studying lym-
phatics of the frog's web, found that the permeability
of injected lymphatic capillaries to particles of
increasing sizes was directlv related to increases in
intraluminal pressure when applied via the lymph
sac. Thus, patent blue dye escaped with 5 mm H»0
pressure, Congo red with 20 to 50 mm H20 pressure.
We have tested a variety of substances of different
molecular weights including dextran fractions,
radioactive sodium (Na22), urea, sodium thiocyanate,
glucose, cellobiose, raffinose, and insulin. Briefly, we
have found that all the macromolecules with molecu-
lar weight as large as or larger than 6000, the molecu-
lar weight of insulin, are retained almost quantita-
tively in the lymph ducts and are returned to the
venous system by the thoracic duct. On the other
hand, smaller molecules like sodium, urea, etc.
shuttle back and forth from lymph ducts and equili-
brate with plasma very rapidly. Recently, we were
fortunate in obtaining a dextran fraction of molecular
weight of 2300. This substance appears to leave the
lymphatic system as do the smaller substances. The
limit of permeability thus seems to be somewhere
between molecular weights of 2300 and 6000.
In these experiments, we always include radioactive
albumin with the test substance and thus are able to
assess any changes in permeability which may occur.
In early experiments with urea, we infused high
concentrations of urea (over 10%) in order to get a
sufficiently high concentration for accurate analysis
of our samples. Under these circumstances, there was
a striking escape of albumin from the lymphatic
system. Further experiments have shown that this
phenomenon occurs only with concentrations of
urea greater than 3 per cent (Fig. 6). Similar results
showing increased leakage of albumin were obtained
with the infusion of 25 ml of a solution containing 60
mg per cent sodium thiocyanate and 20 per cent
radioactive albumin in saline. The mechanism of
this apparently "toxic" effect of urea is being in-
vestigated as are the effects of high urea concentra-
tions on blood capillary permeability.
Experiments with the isolated node preparation
have given results which parallel those given above.
As indicated, the small amount of albumin that finds
its way to the plasma without going through the
thoracic duct pathway evidently gets into plasma
through capillaries of the node. The nature of this
uptake has not been clarified. This preparation, al-
though requiring patience and care in its use, should
continue to be particularly useful in studies of
uptake of other materials by nodes, and the functions
of these nodes as part of the lymphatic circulation.
Much of the evidence relating to the absorption of
substances by lymphatics has been concerned with
absorption from the peritoneum. Here absorption
occurs predominantly through those parts of the
peritoneal surface of the diaphragm which overlie
the lymphatic lacunes. To enter the lumen of a
lymphatic lacune, materials must pass through a
composite structure or "roof" consisting of a) a sheet
of mesothelial cells, facing the peritoneal cavity and
in continuity with the mesothelium of the rest of the
peritoneum; b) a layer of connective tissue which
forms a lattice of fibers; and c) an inner layer of
endothelium in continuity with the endothelium in
the walls and floor of the lacunae and ultimately
with the endothelium of the efferent lvmphatics.
The earliest and perhaps most tempting concept of
PHYSIOLOGIC IMPORTANCE OF LYMPH
1065
the mechanism by which macromolecules and parti-
cles entered lymphatics from the peritoneum was the
postulation of openings in the endothelial walls. This
concept was supported by the early work of von
Recklinghausen (181a, 181 b) and the presence or
absence of "stigmata" and "stomata" have been
debated for the last century. Cunningham (51a), in
reviewing the subject in 1926, concluded, "In general,
then, we may summarize the work which has been
done on the mechanism involved in the absorption of
particulate matter from the peritoneal cavity in the
following way: The earlier work all tended to estab-
lish the concept of the presence of actual preformed
physical openings between the peritoneal cavity and
the diaphragmatic lymphatics. This idea was gradu-
ally eliminated and in its place the concept of poten-
tial physical openings between the walls was offered.
In turn this hypothesis is being replaced by one which
assumes that most, if not all, of the particulate
material that is being absorbed from the peritoneal
cavity passes directly through the cytoplasm of the
mesothelial cells."
Lane Allen and his group have more recently
revived the concept of potential physical openings.
In experiments designed to test the upper limits of
absorption, Allen (ib) injected intraperitoneally a
variety of particles, yeast, mold, paraflin, and
paraffin-asphalt spheres, and monitored diaphrag-
matic lymph for their recovery. He recovered spheres
of mold of 10 /x, glass beads of 12.5 n, and paraffin-
asphalt spheres of up to 22.5 /x in diameter. He also
recovered red blood cells in lymph at a level of up
to 16 million per mm3. In later experiments, Allen &
Weatherford (ic) injected polystyrene spheres with a
range from chylomicron size up to 30 n into the
peritoneal cavities of mice, rats, and cats and re-
covered the particles from regional lymph nodes. The
largest recovered spheres in the mouse were 16.8 ^
in diameter, in the rat and cat, 24 /j. Allen (ib)
presents his concept of diaphragmatic lymphatic
absorption as follows: "As the diaphragm moves
upward in expiration the lymphatic plexus expands
and a relative negative pressure is established in the
lymphatic lumen. At the same time the triple-
layered membrane which separates the peritoneal
cavity from lymphatic lumen is stretched. On either
side of the fenestrations of the basement membrane
the peritoneal mesothelium and lymphatic endo-
thelium open, sometimes to form openings as great as
22.5 /x in diameter. Through these openings suspen-
sions are 'sucked' into the lymphatic lumen. As the
diaphragm contracts the tension on the lymphatic
wall is released, the openings close, and are no longer
demonstrable by usual techniques, and compression
of the plexus results in lymphatic flow."
The possible mechanisms of absorption of particles
by the lymphatics of the diaphragm have been further
clarified in a recent definitive study by French et al.
(77a) using the light and electron microscopes.
They point out that the mesothelial cells of the roofs
differ from other cells at the peritoneal surface of the
diaphragm in that they are more closely set, stain
more darkly, and separate from each other more
readily, particularly at the base of the intercellular
junctions. The cells are supported by a lattice of
coarse and fine fibers. In the meshes of this lattice,
mesothelial and lymphatic endothelial cells are
separated only by the basement membrane of the
mesothelium which may be incomplete. The authors,
using rabbits, injected India ink, thorium dioxide, and
saccharated iron oxide intraperitoneally and found
that the particles entered the intercellular spaces of
the mesothelium and spread freely within the fibers
of the fiber lattice. The particles appeared to pass
through the mesothelium by a predominantly extra-
cellular pathway and probably entered the lymphatic
lumen through temporary channels formed by
separation of endothelial cells at the intercellular
junctions. These gaps formed by separation of meso-
thelial and endothelial cells also permit the passage
of erythrocytes. The authors found that absorbed
colloidal particles accumulated in the cytoplasm of
mesothelial and lymphatic endothelial cells in the
roofs, and their observations suggested that some of
the absorbed material may be transported intracellu-
lar^- through these two layers in cytoplasmic vesicles.
In addition to uptake of particles by the endothelial
cells in the roofs, cells in other sites in the diaphragm
can also take up colloidal particles from the lumen
of the lymphatic. In this respect, their results are
similar to those of Odor (160a), who showed that
particles of mercuric sulphide or Thorotrast were
rapidly taken up from the peritoneal cavities of rats
by mesothelial cells over the mesentery and dia-
phragm. On the other hand, Felix & Dalton (70a)
found that melanin particles introduced intraperi-
toneally were actively ingested by free macrophages
but not by mesothelial cells. These differences may be
related to the difference in particle size or in the
electron microscope preparations.
The evidence accumulated from recent studies thus
suggests at least two possible pathways for the absorp-
tion of large particles (and erythrocytes) from the
peritoneal cavity: /) an extracellular pathway
io66
IIWDBOOK OF PHYSIOLOGY
CIRCULATION II
consisting of gaps between mcsothelial cells caused by
pressure or, as suggested by Allen, by aspiration;
2) an intracellular pathway developed by infolding
of the plasma membrane around particles and the
subsequent pinching off of small pinocytic vesicles
(13a). This vesicular mechanism may be concerned
not only in transport from the exterior to the interior
of the cell but also in transport through cells by a
process termed cytopempsis (147a). The upper limit
to the size of particles which can be absorbed through
the extracellular route is probably determined by
the size of the meshes in the connective tissue layer
rather than by the potential openings between the
mesothelial or endothelial cells. Smaller particles may
travel through intercellular spaces in the roofs of
the lacunes where the mesothelial cells separate
from each other more readily than they do elsewhere.
Evidence that small particles take an extracellular
route through the lymphatic endothelium is perhaps
not so convincing, but that particles can enter the
interspace between lymphatic endothelial cells has
been conclusively shown by Palay & Karlin (163a)
in the central lacteal of an intestinal villus and by
Casley-Smith & Florey (39a) in their study on lym-
phatics in ears of mice and guinea pigs, colons of
rats, and diaphragms of mice. These authors suggest
the possibility that lymphatic endothelial cells in
general are less compactly joined than those in blood
capillaries and may separate from each other more
easily. The apparent absence of a continuous base-
ment membrane to lymphatic endothelium, as dis-
cussed above, may possibly facilitate this separation
of cells and be important in determining the perme-
ability of lymphatic endothelium to macromolecules
and particles traveling from without inward.
If we accept the fact that two possible pathways,
intercellular and intracellular, are available for
movement of substances through lymphatic mem-
branes, their relative importance remains to be
determined. How, too, are we to explain the striking
difference between the ability of substances to enter
and to leave lymphatic vessels? Cunningham reviewed
the evidence available before 1926 and concluded
that the main pathway of absorption is intracellular.
Florey and his group, on the other hand, interpret
their more recent results with the electron microscope
as evidence of the possible greater importance of the
intercellular pathway. They (77a) point out that
there is no evidence that the greater permeability to
colloidal particles shown by lymphatic endothelium
when compared with blood capillary endothelium is
explained by a greater frequency of cytoplasmic
vesicles. It is not too dfficult, perhaps, to accept the
point of view that the morphological basis of this
relatively high permeability of lymphatic endo-
thelium from without inward is related to cleavage
at intercellular junctions and absence of a well-
defined basement membrane. The available evidence
from varied sources, although not always direct or
definitive, is sufficient to suggest that mechanical
factors, pressure and concentration gradients, elastic-
ity of connective tissue, etc. (74a, 52a) may operate
to move these substances from the interstitial space.
Difficulty arises, however, in visualizing the same
process as operating from within outward. Peters
(171a) in 1935 appreciated this difficulty when he
attempted to formulate a comprehensive theory of
lymphatic absorption and raised the question as to
how one could expect to hold water in a sieve by
putting a valve at its mouth. Admittedly we still do
not have sufficient information to provide an over-all
sophisticated concept of permeability of lymphatic
vessels. For the time being, it may therefore be
wise to consider the following simple concept. We
believe it to be consistent with the available evidence
and to offer an explanation of the apparent one-way
flow of materials into but not out of the lymphatics.
Lacking evidence to the contrary, we may assume
that the smallest terminal lymphatic capillaries are
freely permeable to small and large molecules and
particles moving in either direction through inter-
cellular gaps. Compression of these vessels in any
manner will force their contents in all directions.
Some of the contents can, however, be forced centrally
into larger vessels and ducts. The valves in these
vessels will prevent backflow. Once the lymph
reaches the larger vessels, it no longer loses its macro-
molecules and particles, since the walls of the larger
vessels, as previously discussed, restrict molecules
larger than molecular weight of approximately 2000
(at least in the dog).
This simple concept implies a relatively inept and
inefficient system, a "leaky pump" system about
which Peters complained. As Allen commented,
however, a leaky pump will still pump, and as
Drinker emphasized, the lymphatic system is, in the
final analysis, a rather casual system. It does a
reasonably good job under "normal" conditions. Its
ineffectiveness becomes manifest chiefly under patho-
logical situations. This aspect of the functions of the
lymphatic system, its inadequacy in various patho-
logical situations, will continue to merit careful
study.
PHYSIOLOGIC IMPORTANCE OF LYMPH
1067
REFERENCES
1 . Acevedo, D. Motor control of the thoracic duct. Am. J.
Physiol. 139: 600-604, !943-
1 a. Allen, L. A quantitative study of tissue fluid-lymph
cellular ratios. Anal. Record 92: 279-287, 1945.
ib.ALLEN, L. On the penetrability of the lymphatics of the
diaphragm. Anat. Record 124: 639-658, 1956.
ic.Allen, L., and T. Weatherford. Role of fenestrated
basement membrane in lymphatic absorption from
peritoneal cavity. Am. J. Physiol. 197: 551-554, '959-
2. Alrich, E. M. Studies on burns II. Surgery 15: go8-gi2,
1944.
3. Asellius, G. De lactibus sive lacteis venis, quarto vasorum
mesaraicorum genere, novo invento. Dissertatio . . . Milan:
Biddellium Mcdiolani, 1627.
4. Ashworth, C. T., Z. W. Hutcheson, W. T. Payne,
and A. W. Jester. The effect of crystalloidal and protein-
containing solutions on the body fluids and circulating
plasma proteins. Am. J. Physiol. 140: 589-597, 1944.
5. Ashworth, C. T., V. A. Stembridge, and E. Sanders.
Lipid absorption, transport, and hepatic assimilation
studied with electron microscopy. Am. J. Physiol. 198:
1 326-1 328, 1960.
6. Aub, J. C, A. M. Brues, S. S. Ketv, I. T. Nathanson,
A. L. Nutt, A. Pope, and P. C. Zamecnik. The toxic
factors in experimental traumatic shock. IV. The effects
of intravenous injection of the effusion from ischemic
muscle. J. Clin. Invest. 24: 845-849, 1945.
7. Babics, A., and F. Renyi-Vamos. Patho-physiology
and operations of the renal cavities. Quoted by
Rusznyak, Foldi and Szabo (188).
8. Babics, A. Lymphatic circulation of the kidneys. Acta Med.
Acad. Sci. Hung. 2: 1-20, 1951.
g. Baez, S., A. Carleton, and I. Forbes. Mesenteric
lymphatic adjustments during shock. Federation Proc.
'6-5. '957-
10. Baggenstoss, A. H., and J. C. Cain. The hilar lymphatics
of man: Their relation to ascites. New Engl. J. Med.
256:531-535. '957-
11. Baggenstoss, A. H., and J. C. Cain. Further studies on
the lymphatic vessels at the hilus of the liver of man:
Their relation to ascites. Proc. Staff Meetings Mayo Clinic
32:615-627, 1957.
12. Bartholin, T. Anatomia, ex Caspari Bartholini parentis
Institutionibus, omnium recentwrum, el propnes observationibus
tertium ad sanguinis circulationem reformata. Leyden : Hack,
1651.
13. Bartholin, T. Dubia Anatomica de Lacteis Thoracicis . . .
Publice Proposita. Copenhagen : Melch. Martzan, 42 pp,
1653-
1 3a. Bennett, H. S. The concepts of membrane flow and
membrane vesiculation as mechanisms for active trans-
port and ion pumping. J. Rwphys. Biochem. Cytol. 2:
suppl. 99-103, 1956.
14. Benson, J. A., Jr., K. G. Kim, and J. L. Bollman.
Extravascular diffusion of protein. Am. J. Physiol. 182:
217-220, 1955.
15. Benson, J. A., Jr., P. R. Lee, J. F. Scholer, K. S. Kim,
and J. L. Bollman. Water absorption from the intestine
via portal and lymphatic pathways. .4m. J. Physiol. 184:
441-444, 1956.
16. Bierman, II. R., R. L. Byron, Jr., K. H. Kelly, R. S.
Gilfillan, L. P. White, N. E. Freemand, and N. L.
Petrakis. The characteristics of thoracic duct lymph in
man. J. Clin. Invest. 32 : 637-649, 1953.
17. Biggs, M. W., M. Friedman, and S. O. Byers. Intestinal
lymphatic transport of absorbed cholesterol. Proc. Soc.
Espll. Biol. Med. 78: 641-643, 1951.
18. Biro, J., E. Grasz, F. Renyi-Vamos, and M. Renyi-
Vamos. Der Lymphtransport der amylase. Acta Physiol.
Acad. Sci. Hung. 16: 1 75-181, 1959.
1 9. Blalock, A. A Study of thoracic duct lymph in experi-
mental crush injury produced by gross trauma. Bull.
Johns Hopkins Hosp. 72: 54-61, 1943.
20. Blatt, L. J., and J. J. Cincotti. In vivo visualization of
lymphatics; experimental and clinical study with reference
to rectum. Surgery 38: 373-383, ig55.
31. Blomstrand, R., O. Dahlback, and E. Linder. Asym-
metric incorporation of linoleic acid-1-C14 and stearic
acid-i-C14, into human lymph lecithins during fat absorp-
tion. Proc. Soc. Exptl. Biol. Med. 100: 768-771, 1959.
22. Bloom, B., I. L. Chaikoff, and W. O. Reinhardt.
Intestinal lymph as pathway for transport of absorbed
fatty acids of different chain lengths. Am. J. Physiol.
166:451-455, 1951.
23. Bocklage, B. C, E. A. Doisy, Jr., W. H. Elliot, and
E. A. Doisy. Absorption and metabolism of cortisone-4-C14
acetate. J. Biol. Chem. 212: g35-939, '955-
24. Bocklage, B. C, H. S. Nicholas, E. A. Doisy, Jr.,
W. H. Elliot, S. A. Thayer, and E. A. Doisy. Synthesis
and biological studies of 17-methyl C14 estradiol. ./. Biol.
Chem. 202: 27-37, '953-
25. Bollman, J. L., and E. V. Flock. Cholesterol in intestinal
and hepatic lymph in rat. Am. J. Physiol. 164: 480-485,
■951-
26. Bollman, J. L., E. V. Flock, J. C. Cain, and J. H.
Grindlay. Lipids of lymph following feeding of fat: An
experimental study. Am. ./. Physiol. 163: 41-47, 1950.
27. Brauer, R. W. Liver circulation and liver function.
Physiol. Rev. 43: 1 15-21 3, 1963.
28. Brauer, R. W., and E. Hardenbergh. Distribution of
enterase in lymph from various regions and in relation to
lymphoid tissue. Am. J. Physiol. 150: 746-753, 1947.
29. Brinkhous, K. M., and S. A. Walker. Prothrombin and
fibrinogen in lymph. Am. J. Physiol. 132: 666-669, I94I>
30. Brockett, S. H., M. A. Apiers, and H. E. Himwich.
The lipid components of the lymph of the thoracic duct
of the dog. Am. J. Physiol. 1 10: 342-347, 1934.
31. Brown, C. S., and E. Hardenberch. A technique for
sampling lymph in unanesthetized dogs by means of an
exteriorized thoracic duct-venous shunt. Surgery 29:
502-507, 1 95 1.
32. Bull, G. M. Postural proteinuria. Clin. Sci. 7: 77-108,
1948-49.
33. Cain, J. C, J. H. Grindlay, J. L. Bollman, E. V. Flock,
and F. C. Mann. Lymph from liver and thoracic duct.
Surg. Gynecol. Obstet. 85: 558-562, 1947.
34. Carlsten, A. On the sources of the histaminase present
in thoracic duct lymph. Acta Physiol. Scand. 20: Suppl. 70,
5-26. !95°-
35. Carlsten, A. No change in histamine content of lymph
io68
HANDBOOK OF PHYSIOU M : Y
CIRCULATION II
and plasma in cats during pregnancy. Acta Physiol. Scantl.
20: Suppl. 70, 27-31, 1950.
36. Carlsten, A. Effect of adrenalectomy on lymph and
plasma histaminase. Acta Physiol. Scand. 20: Suppl. 70,
33~46> I95°-
37. Carlsten, A., G. Kahlson, and F. Wichsell. The strong
histaminolytic activity of lymph and its bearing on the
distribution of histamine between lymph and plasma in
dogs. Acta Physiol. Scand. 17: 370-383, 1949.
38. Carlsten, A ., and D. R. Wood. The assay of histaminase
using 2 methods for estimation of residual histamine.
Acta Physiol. Scand. 20: Suppl. 70, 1 19-125, 1950.
39. Carlsten, A., and D. R. Wood. Increased lymph histami-
nase in adrenalectomized cats and its restoration by
adrenocortical extract but not by adrenaline. J. Physiol.
112: 142-148, 1951 .
39a.CASLEY-SMiTH, J. R., and H. W. Florey. The structure
of normal small lymphatics. Quail. J. Exptl. Physiol. 46:
101-106, 1961.
40. Casten, B., and K. Kistler. Development of acute
pulmonary edema in mice and rats and an interpretation.
Am. J. Physiol. 178: 49-52, 1954.
41. Chaikoff, I. L., B. Bloom, M. D. Siperstein, J. V.
Kjyasu, W. O Reinhardt, W. G. Dauben, and J. F.
Eastham. Cu-cholesterol. I. Lymphatic transport of
absorbed cholesterol-4-C14. J. Biol. Chem. 194: 407-412,
'95^
42. Clark, E. R., and E. L. Clark. Further observations
on living lymphatic vessels in the transparent chamber
in the rabbit's ear— their relation to the tissue spaces
.4m. J. Anal. 52: 273-305, 1933.
43. Clark, E. R., and E. L. Clark. Observations on living
mammalian lymphatic capillaries — their relation to the
blood vessels. Am. J. Anal. 60: 253-298, 1936-37.
44. Cope, O., and F. D. Moore. A study of capillary perme-
ability in experimental burns and burn shock using
radioactive dyes in blood and lymph. J. Clin. Invest. 23:
-!4I-257. IQ44-
45. Cope, O., and L. Rosenfield. The lymphatic system.
Ann. Rev. Physiol. 8: 297-310, 1946.
46. CoTui, F., I. S. Barcham, and B. G. P. Shafiroff.
Ligation of the thoracic duct and the posthemorrhage
plasma protein level. Surg. Gynecol. Obslet. 79: 37-4°, >944-
47. Courtice, F. C. Rept. Australian New Zealand Assoc.
Advance. Sci. 28th Meeting, Brisbane 28: 115-119, 1 951.
48. Courtice, F. C, and P. I. Korner. The effect of anoxia
on pulmonary edema produced by massive intravenous
infusions. Australian J. Exptl. Biol. Med. Sci. 30: 511-526,
I952-
49. Courtice, F. C, and B. Morris. The exchange of lipids
between plasma and lymph of animals. Qiiart. J. Exptl.
Physiol. 40: 138-148, 1955.
50. Courtice, F. C, W.J. Simmonds, and A. W. Steinbeck.
Some investigations of lymph from a thoracic duct
fistula in man. Australian J. Exptl. Biol. Med. Sci. 29:
201-210, 1951-
50a. Courtice, F. C, and A. W. Steinbeck. The lymphatic
drainage of plasma from the peritoneal cavity of the cat.
Australian J. Exptl. Biol. Med. Sci. 28: 161 -169, 1950.
50b. Courtice, F. C, and A. W. Steinbeck. The effects of
lymphatic obstruction and of posture on the absorption
of protein from the peritoneal cavity. Australian J. Exptl.
Biol. Med. Sci. 29: 451-458, 1951.
51. Crandall, L. A., Jr., S. B. Barker, and D. G. Graham.
Study of the lymph How from a patient with thoracic
duct fistula. Gastroenterology 1 : 1040- 1048, 1943.
5 1 a. Cunningham, R. S. The physiology of the serous mem-
branes. Physiol. Rev. 6: 242-280, 1926.
52. Danese, G, and J. M. Howard. Surgical studies of the
lymphatics. Circulation 22: 738, i960.
52a. Day, T. D. The role of connective tissue in the filling of
lymphatics. Ouart. J. Exptl. Physiol. 44: 182-189, '959-
53. Dietrich, L. S., and G. J. Siecel. Purine derivatives in
lymph from the rat. Am. J. Physiol. 199: 198-200, i960.
54. Doemling, D. B., and F. R. Steggerda. Lymph flow
studies in unanesthetized dogs having chronic thoracic
duct-jugular vein cannulations. Physiologist 1 (No. 1):
21. 1957-
55. Dragstedt, C. A., and E. Gebauer-Fuelnegg. Studies
in anaphylaxis. I. The appearance of a physiologically
active substance during anaphylactic shock. Am. J.
Physiol. 102: 512-519, 1932.
56. Dragstedt, C. A., and F. B. Mead. Further observations
on the nature of the active substance C'Anaphylatoxin")
in canine anaphylactic shock. ./. Immunol. 30: 319-326,
>936-
57. Dragstedt, C. A., and F. B. Mead. A pharmacologic
study of the toxemia theory of surgical shock. J. Am.
Med. Assoc. 108: 95-96, 1937.
58. Drinker, C. K. The functional significance of the lym-
phatic system. Harvey Lectures 38: 89-1 I 1, 1937.
59. Drinker, C. K. Extravascular protein and the lymphatic
system. Ann. N. Y. Acad. Sci. 46: 807-821, 1946.
60. Drinker, C. K. Pulmonary Edema and Inflammation: An
Analysis of Processes Involved in the Formation and Removal of
Pulmonary Transudates and Exudates. Cambridge : Harvard
Univ. Press, 1950.
61. Drinker, C. K., and M. E. Field. The protein content
of mammalian lymph and the relation of lymph to tissue
thud. Am. J. Physiol. 97 : 32-39, 1 931 .
6ia.DRiNKER, C. K., and M. E. Field. The lymph capillaries
in the web of the frog. .4m. J. Physiol. 100: 642-649, 1932.
62. Drinker, C. K., and M. E. Field. Lymphatics, Lymph and
Tissue Fluid. Baltimoie: Williams & Wilkins, 1933.
63. Drinker, C. K. , M E. Field, and H. K. Ward. The
filtering capacity of lymph nodes. J. Exptl. Med. 59:
393-405. '934-
64. Drinker, C. K., M. F. Warren, F. M. Maurer, and
J. D. McCarrell. The flow, pressure, and composition
of cardiac lymph. Am. J. Physiol. 130: 43-55, 1940.
65. Drinker, C. K., and M. F. Warren. The genesis and
resolution of pulmonary transudates and exudates. J.
Am. Med. Assoc. 122: 269-273, 1943.
66. Drinker, C. K., and J. M. Yoffey. Lymphatics, Lymph and
Lymphoid Tissue. Cambridge: Harvard Univ. Press, 1941.
67. Eagle, H., C. G. Johnston, and I. S. Ravdin. On the
prolonged coagulation time subsequent to anaphylactic
shock. Bull. Johns Hopkins Hasp. 60: 428-438, 1937.
68. Endicott, K. M., T. Gillman, G. Brecher, A. T. Ness,
F. A. Clarke, and E. R. Adamik. A study of histochemi-
cal iron using tracer methods. J. Lab. Clin. Med. 34: 414-
421, 1949.
69. Everett, N. B., W. E. Garrett, and B. S. Simmons.
PHYSIOLOGIC IMPORTANCE OF LYMI'll
IO69
I ,\ mphatics in iron absorption and transport. Am. J. 86.
Physiol. 178:45-48, 1954.
70. Fantl, P., and J. F. Nelson. Coagulation in lymph.
J. Physio/., 122: 33-37, 1953.
70a. Felix, M. D., and A. J. Dalton. A comparison of 87.
mesothelial cells and macrophages in mice after the
intraperitoneal inoculation of melanin granules. J.
Biophys. Bioche?n. Cytol. 2: (pt. 3) Suppl., 109-113, 1956.
71. Fishman, A. P., H. W. Fritts, Jr., and A. Cournand.
Effects of breathing carbon dioxide upon the pulmonary
circulation. Circulation 22: 220-225, '960.
72. Flock, E. V., and J. L. Bollman. Alkaline phosphatase 89.
activity in the intestinal lymph of the rat. J. Biol. Chem,
1 75 ■ 439-449. 1 948. 90.
73. Flock, E. V., and J. L. Bollman. The influence of bile
on the alkaline phosphatase activity of intestinal lymph.
J. Biol. Chem. 184: 523-528, 1950.
74. Flock, E. V., and J. L. Bollman. Amylase and esterase 91.
in rat intestinal lymph. ./. Biol. Chem. 185: 903-908, 1950.
74a.FLOREV, H. Reactions of, and absorption by, lymphatics
with special reference to those of the diaphragm. But. J.
Exptl. Pathol. 8: 479-489, 1927.
75. Foldi, M., J. Kepes, I. Rusznyak, and G. Szabo. 92.
Bedcutung der lymph stromung fur den Saftekreislanf in
der lunge. Acta Med. Acad. Sci. Hung. 7: 345, 1955. 93-
76. Freeman, L. W. Lymphatic pathways from the intestine
in the dog. Anal. Record 82: 543-550, 1942.
77. Freeman, N. E., and A. E. Schecter. No demonstrable 94.
substance causing increased capillary permeability in
lymph from an injured area. Proc. Soc. Exptl. Biol. Med.
51: 29-31, 1942.
77a. French. J. E., H. W. Florev, and B. Morris. The 95.
absorption of particles by the lymphatics of the diaphragm.
Quart. J. Exptl. Physiol. 45: 88-103, 1960.
78. Friedman, M., S. O. Byers, and C. Omoto. Some
characteristics of hepatic lymph in the intact rat. Am. J. 96.
Physiol. 184: 1 1 -1 7, 1956.
79. Friedman, M., and R. H. Roseman. Effects of hyper-
and hypothyroidism on hepatic lymph cholesterol in 97.
rats. Am. J. Physiol. 188: 295-296, 1957.
80. Fritts, H. W., Jr., J. E. Odell, P. Harris, E. W.
Braunwald, and A. P. Fishman. Effects of acute hypoxia
on the volume of blood in the thorax. Circulation 22: 98.
216-219, '960.
Gabrio, B. W., and K. Solomon. Distribution of total
ferritin in intestine and mesenteric lymph nodes of horses 99.
after iron feeding. Proc. Soc. Exptl. Biol. Med. 75: 124-127,
'950-
Gebauer-Fuelnegg, E , and C. A. Dragstedt. Studies
in anaphylaxis: IE The nature of a physiologically
active substance appearing during anaphylactic shock. 100.
Am. J. Physiol. 102: 520-526, 1932.
83. Geyer, G. F., S. M. Herbst, H. Thaler, and W. F.
Lever. The permeability of capillaries to scrum choles-
terol. J. Clin. Invest. 35: 281-284, '95°- IOO<
84. Gilman, T., and A. C. Ivy. A histological study of the
participation of the intestinal epithelium, the reticulo-
endothelial system and the lymphatics in iron absorption 101.
and transport. Gastroenterology 9: 162-169, '947-
85. Glenn, \V. W. L, J. Muus, and C. K. Drinker. Observa-
tions on the physiology and biochemistry of quantitative
burns. J. Clin. Invest. 22: 451-460, 1943.
81
82
Glenn, W. W. L., D. K. Peterson, and C. K. Drinker.
The flow of lymph from burned tissue, with particular
reference to the effects of fibrin formation upon lymph
drainage and composition. Surgery 12: 685-693, 1942.
Gley, E., and V. Pachon. Influence des variations de la
circulation lymphatique intra-hepatique. (Sur Taction
anticoagulante de la peptone) Arch. Physiologic, 5th Series.
7: 711-718, 1895.
Goodwin, W. E., and J. J. Kaufman. The renal lym-
phatics. I. Review of some of the pertinent literature.
Urol. Survey 6: 305-329, 1956.
Goodwin, W. E., and J. J. Kaufman. Renal lymphatics.
II. Preliminary experiments. J. Urol. 76: 702-707, 1956.
Gottschalk, C. W., and M. Mylle. Micropuncture
study of the mammalian urinary concentrating mecha-
nism: Evidence for the countercurrcnt hypothesis. Am. J.
Physiol. 196:927-936, 1959.
Green, H. D., G. A. Bergeron, J. Little, and J. E.
Hawkins, Jr. Evidence from cross transfusion experi-
ments, that no toxic factor is present in ischemic compres-
sion shock capable of inducing a shock state in normal
dogs. Am. J Physiol. 149: 1 12-122, 1947.
Grotte, G. Passage of dextran molecules across the
blood lymph barrier. Acta Chir. Scand. 211: 1-84, 1956.
Guyton, A. C., G. G. Armstrong, and J. W. Crowell.
Negative pressure in the interstitial spaces. Physiologist
3 (No. 3) : 70, i960.
Haddy, F. J., J. Scott, M. Fleishman, and D. Emanuel.
Effect of change in renal venous pressure upon renal
vascular resistance, urine and lymph flow rates. Am. J.
Physiol. 195:97-110, 1958.
Haynes, F. W. Factors which influence the flow and
protein content of subcutaneous lymph in the dog. II.
The effect of certain substances which alter the capillary
circulation. Am. J. Physiol. 101 : 612-620, 1932.
Heidenhain, R. Versuche und Fragen zur Lehre von der
Lymph Bilding. Pfiiigers Arch. ges. Physiol. 49: 209, 301,
1891.
Hellman, L., H. L. Bradlow, E. L. Frazell, and T. F.
Gallagher. Tracer studies of the absorption and fate of
steroid hormones in man. J. Clin. Invest. 35: 1033-1044,
'95°.
Hellman, L., E. L. Frazell, and R. S. Rosenfeld.
Direct measurement of cholesterol absorption via the
thoracic duct in man. J. Clin. Invest. 39: 1 288-1 294, i960.
Hewson, W. Experimental Inquiries: Part the Second. Con-
taining a Description of the Lymphatic System in the Human
Subject, and in Other Animals. Together with Observations on
the Lymph, and the Changes Which it Undetgoes in Some
Diseases. London: Johnson, No. 72, 1774.
Hollander, W., P. Reilly, and B. A. Burrows. Lym-
phatic flow in human subjects as indicated by the disap-
pearance of I131 labelled albumin from the subcutaneous
tissues. J. Clin. Invest. 35: 713, 1956.
i.Hudack, S., and P. D. McMaster. The permeability
of the wall of the lymphatic capillary. J. Exptl. Med. 56:
223-236, 1932.
Hunter, W. Two introductory lectures to his last course
of anatomical lectures at his theatre in Windmill Street.
London: pp. 58-59, 1884. (Quoted by C. K. Drinker,
Lane Medical Lectures. Stanford: Stanford LIniv. Press
I942-)
1070
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
102. Hyde, P. M., E. A. Doisy, Jr., W. H. Elliott, and E. A.
Doisy. Absorption of cnterally administered 1 y-a-
methyl-C14 testosterone and its metabolites. J. Biol. Chem.
290: 257-263, 1954.
103. Irisawa, A., and R. F. Rushmer. Relationship between
lymphatic and venous pressure in leg of dog. Am. J.
Physiol. 196:495-498, 1959.
104. Jaques, L. B., and E. B. Waters. The identity and origin
of the anticoagulant of anaphylactic shock in the dog. J.
Physiol., 99: 454-466, 1940-41 .
105. Kaiserling, H., and T. Soostmeyer. The importance of
the lymph system of the kidneys for kidney function.
Wien. klm. Wochenschr. 52 : 1 1 13-1 1 16, 1939.
106. Kampmeier, O. F. Further observations on the numerical
variability, position, function and fate of the valves in
the human thoracic duct. Anal. Record 38: 225-231, 1928.
107. Kaplan, A., M. Friedman, and H. E. Kruger. Observa-
tions concerning the origin of renal lymph. -4m. J. Physiol.
'38:553"556. !943-
108. Katz, Y. J. Some factors affecting renal lymphatic
pressure. Circulation Research 6: 452-455, 1958.
109. Katz, Y. J., and A. T. K. Cockett. Elevation of inferior
vena cava pressure and thoracic lymph and urine flow.
Circulation Research 7: 1 18-122, 1959.
1 10. Katzenstein, R., E. Mvlon, and M. C. Winternitz.
The toxicity of thoracic duct fluid after release of tourni-
quets applied to the hind legs of dogs for the production
of shock. Am. J. Physiol. 139: 307-312, 1943.
111. Kellnor, A. The lipid and protein content of tissue
fluid in normal and hyperlipemic rabbits. Symposium on
Atherosclerosis. Natl. Acad. Set. — Natl. Research Council
Publ. No. 338: 42-49, 1955.
112. Kinmouth, J. B. Lymphangiography in man. Clin. Sci.
11 : 13-20, 1952.
113. Klitgaard, H. M., and J. P. Toth, Jr. Lymphatic
transport of C14 thyroxine. Federation Proc. 14: 86, 1 955.
114. Klitgaard, H. M., J. P. Toth, Jr., P. A. Kot, and
R. A. Whaley. C14 thyroxine transport in thoracic lymph
in rats. Proc. Soc. Exptl. Biol. Med. 96: 122-124, 1957.
1 15. Koler, R. D., and J. D. Mann. Iron content of intestinal
lymph of rats. Proc. Soc. Exptl. Biol. Med. 76: 221-222,
I951-
116. Korner, P. I., B. Morris, and F. C. Courtice. An
analysis of factors affecting lymph flow and protein
composition during gastric absorption of food and fluids,
and during intravenous infusion. Australian J. Exptl. Biol.
Med. Sci. 32: 301-320, 1954.
117. Krogh, A. Anatomy and Physiology of Capillaries. New
Haven: Yale Univ. Press, 1922.
118. Landis, E. M. Capillary permeability and the factors
affecting the composition of capillary filtrate. Ann. N. Y.
Acad. Sci. 46: 713-731, 1946.
119. Landis, E. M., L. Jonas, M. Angevine, and W. Erb.
The passage of fluid and protein through the human
capillary wall during venous congestion. J. Clin. Invest.
'i : 7 '7-734. I932-
120. Langdell, D. R., L. \V. Bowersox, R. A. Weaver, and
W. A. Gebson. Coagulation properties of canine thoracic
duct lymph. Am. J. Physiol. 199: 626-628, i960.
121. LeBrie, S. J., and H. S. Mayerson. Composition of
renal lymph and its significance. Proc. Soc. Exptl. Biol.
Med. 100: 378-380, 1959.
122. LeBrie, S. J., and H. S. Mayerson. Influence of uranium
nitrate induced nephrosis on flow and composition of
renal lymph. Physiologist 3 (No. 3): 102, i960.
123. LeBrie, S. J., and H. S. Mayerson. Influence of elevated
venous pressure on flow and composition of renal lymph.
Am. J. Physiol. 198: 1 037-1 040, i960.
124. Lee, F. C. Some observations on lymph pressure. Am. J.
Physiol. 67:498-513, 1923-24.
1 24a. Lee, F. C. Permeability of lymph vessels and lymph
pressure. Arch. Surg. 48: 355-365, 1944.
125. Linder, E., and R. Blomstrand. Technic for collection
of thoracic duct lymph of man. Proc. Soc. Exptl. Biol. Med.
97:653-657, 1958.
126. Lindner, E., W. Marx, and H. E. Kruger. Absence in
lymph of capillary permeability factors in traumatic
shock. Proc. Soc. Exptl. Biol. Med. 55: 181, 1944.
127. Logan, G. B. Histamine in intestinal lymph of white rat
during anaphylactic shock. Proc. Soc. Exptl. Biol. Med.
104: 532"536. '96°-
128. Lowgren, E. Lymphuria as an explanation of the postural
proteinuria. Acta Med. Scand. 144: 245, 1952.
129. Ludwig, C, and T. Sawarykin. Die Lymphwurseln in
der Niere des Saugesteires. Sitz-Ber. Akad. Wiss. Hem.
44: 155, 1863.
130. McCandless, E. L., and D. B. Zilversmit. Distribution
and turnover of fatemulsion components in dogs. Am. J.
Physiol. 183:642, 1955.
131. McCandless, E. L., and D. B. Zilversmit. Disappearance
of I131 -labelled lymph triglycerides and phosphatides from
blood of dogs. Federation Proc. 16: 85, 1 957.
132. MacCallum, W. G. The relations between the lymphatics
and the connective tissue. Bull. Johns Hopkins Hosp.
14: 1-9, 1903.
1 32a. MacCallum, W. G. On the mechanism of absorption
of granular materials from the peritoneum. Bull. Johns
Hopkins Hosp. 14: 105-115, 1903.
133. Macallum, A. B. On the absorption of iron in the animal
body. J. Physwi. 16: 268-297, 1894.
134. McClure, C. F. W., and C. F. Silvester. A comparative
study of the lymphatic-venous communications in adult
mammals. Anat. Record 3: 534-551, I9°9-
135. McM aster, P. D. Lymphatic participation in cutaneous
phenomena. Harvey Lectures 37: 227-268, 1942.
136. McMaster, P. D. The lymphatic system. Ann. Rev. Physiol
5: 207-228, 1943.
137. McMaster, P. D. Conditions in skin influencing inter-
stitial fluid movement lymph formation, and lymph
flow. Ann. N. Y. Acad. Sci. 46: 743-787, 1946.
138. Mann, J. D., and G. M. Higgins, Lymphocytes in
thoracic duct, intestinal and hepatic lymph. Blood 5:
i77-'9°. 195°-
139. Mann, J. D., F. D. Mann, and J. L. Bollman. Hypo-
prothrombinemia due to loss of intestinal lymph. Am. J.
Physiol. 158:311-314, 1949.
140. Marble, A., M. E. Field, D. K. Drinker, and R. M.
Smith. The permeability of the blood capillaries to lipoids.
Am. J. Physiol. 109: 467-474, 1934.
141. Mayerson, H. S., C. G. Wolfram, H. H. Shirley, Jr.,
and K. Wasserman. Regional differences in capillary
permeability. Am. J. Physiol. 198: 155-160, i960.
142. Meng, H. C. Removal of intravenously injected fat from
PHYSIOLOGIC IMPORTANCE OF LYMPH
IO71
the circulation and its appearance in the thoracic duct
lymph. Am. J. Physiol. 168: 335-344, J95-J-
143. Miles, A. A., and D. L. Wilhelm. Distribution of globu-
lin permeability factor and its inhibitor in the tissue fluid
and lymph of the guinea pig. Nature 181 : 96-98, 1958.
144. Miles, A. A., and D. L. Wilhelm. The activation of
endogenous substances inducing pathological increases in
capillary permeability. In : The Biochemical Response to
Injun, edited by H. B. Stoner. Springfield, 111. : Thomas,
i960.
145. Miller, A. J., R. Pick, and L. N. Katz. Ventricular
endomyocardial pathology produced by chronic cardiac
lymphatic obstruction in the dog. Circulation Research 8 :
94'-947. '96°-
146. Miller, A. J., R. Pick, and L. N. Katz. Do lymphatic
vessels exist in the heart valves of the dog? Circulation 22:
789, i960.
147. Moore, C. V., W. R. Arrowsmith, J. Welch, and V.
Minnich. Studies in iron transportation and metabolism
IV. Observations on the absorption of iron from the
gastro-intestinal tract. J. Clin. Invest. 18: 553-580, 1939.
147a. Moore, D. H., and H. Ruska. The fine structure of
capillaries and small arteries. J. Bwphys. Biochem. Cytol.
3:457-462, 1957-
148. Morris, B. The interrelationships of the plasma and
lymph lipide fractions before and during fat absorption.
Australian J. Exptl. Biol. Med. Sci. 32: 763-782, 1954.
149. Morris, B. The hepatic and intestinal contributions to
the thoracic duct. Quart J. Exptl. Physiol. 41 : 318-325,
■956-
150. Morris, B. The exchange of protein between the plasma
and the liver and intestinal lymph. Quart. J. Exptl.
Phisiol. 41 : 326-340, 1956.
151. Morris, B., and F. C. Courtice. The origin of chylo-
microns in the cervical and hepatic lymph. Quart. J.
Exptl. Physiol. 41 : 341-348, 1956.
1 52. Mueller, J. H. The mechanism of cholesterol absorption.
J. Biol. Chem. 27: 463-480, 1 916.
153. Murphy, J. J., and M. K. Myint. The renal lymphatics
II. Effect of increasing pressure in the renal pelvis upon
absorption of substances of various molecular sizes.
Surg. Forum 7: 661-667, 1956.
154. Myint, M. K., and J. J. Murphy. The renal lymphatics
I. The effect of diuresis and acute ureteral obstruction
upon the rate of flow and composition of thoracic duct
lymph. Surg. Forum 7:656-660, 1956.
155. Nath anson, I. T, A. L. Nutt, A. Pope, P. C. Zamecnik,
J. C. Aub, A. M. Brues, and S. S. Kety. The toxic factors
in experimental traumatic shock. I. Physiologic effects of
muscle ligation in the dog. J. Clin. Invest. 24: 829-834,
'945-
156. Nisimaru, Y., and H. Irisawa. Lymph capillaries in the
frog's web. Federation Proc. 16: g4, 1957.
157. Nix, J. T., E. V. Flock, and J. L. Bollman. Influence of
cirrhosis on proteins of cisternal lymph. Am. J. Physiol.
164: 1 17-1 18, 1 951.
158. Nix, J. T., F. C. Mann, J. L. Bollman, J. H. Grindlay,
and E. V. Flock. Alterations of protein constituents of
lymph by specific injury to liver. Am. J. Physiol. 164:
119-122, 1951.
159. Nordmann, W., H. J. Loeblich, and W. Koch. The
pathology of lymphatic channels. Arch. Kreislaujforsch
19:38-58, 1953-
160. Nuck, A. Adenogtaphia curiosa et uteri foeminei anatome nova.
1692. [Quoted by Rusynak, Foldi, and Szabo (189).]
1 60a. Odor, D. L. Uptake and transfer of particulate matter
from the peritoneal cavity of the rat. J. Biophys. Biochem.
Cytol. 2: Suppl. 4, pt. 2, 105-107, 1956.
161. Page, I. The occurrence of a vasoconstrictor substance in
blood during shock induced by trauma, hemorrhage
and burns. Am. J. Physiol. 1 39 : 386-398, 1 943.
162. Page, I. H., L. A. Lewis, and G. Plahl. The lipoprotein
composition of dog lymph. Circulation Research 1 : 87-93,
1953-
163. Paine, R., H. R. Butcher, F. A. Howard, and J. R.
Smith. Observations on mechanisms of edema formation
in the lungs. J. Lab. Clin. Med. 34: 1544-1553, '949-
163a.PAL.vy, S. L., and L. S. Karlin. An electron micro-
scopic study of the intestinal villus. I. The fasting animal.
J. Biophys. Biochem. Cytol. 5: 363-372, 1959.
164. Palay, S. L., and L. J. Karlin. An electron microscopic
study of the intestinal villus. II. The pathway of fat
absorption. J. Biophys. Biochem. Cytol. 5: 373-383, 1959.
165. Papamiltiades, M. Sur la communication entre la
chambre anterieure et le reseau lymphatique de la
conjonctive de 1'oeil chez Fhomme. Ann. oculist., Paris
l89: 939-945. >956-
166. Pappenheimer, J. R. Passage of molecules through capil-
lary walls. Physiol. Rev. 33: 387-423, 1953.
167. Patek, P. R. The morphology of the lymphatics of the
mammalian heart. Am. J. Anat. 64: 203-249, 1939.
168. Patterson, R. M., C. L. Ballard, K. Wasserman, and
H. S. Mayerson. Lymphatic permeability to albumin.
Am. J. Physiol. 194: 120-124, '958-
169. Pecquet, J. Expeiimenta Nova Anatomica Quibus Incognitum
Hactenus Chyli Receptaculum, et ab eo per Thoracem in Ramos
Usque sub Clavios Vasa Lactea Detegunler. Paris: Cramoisy
and Cramoisy, 1951 .
170. Perlmann, G. E., W. W. L. Glenn, and D. Kaufman.
Changes in the electrophoretic pattern in lymph and
serum in experimental burns. J. Clin. Invest. 22 : 627-633,
1943-
171. Perry, T. T. Role of lymphatic vessels in the transmission
of lipase in disseminated pancreatic fat necrosis. A.M. A.
Arch. Pathol. 43: 456-465, 1947.
I7ia.PETERS, J. P. Body Water. Springfield, 111.: Thomas, 1935.
172. Petersen, VV. F., and T. P. Hughes. Inorganic altera-
tions of the lymph in canine anaphylactic shock. J. Biol.
Chem. 63: 179-196, 1925.
173. Petersen, W. F., and S. A. Levinson. Studies in endo-
thelial permeability. II. Role of the endothelium in
canine anaphylactic shock. J. Immunol. 8: 349-359, 1923.
174. Peterson, R. E., and J. D. Mann. Transport of radio-
active iron in intestinal lymph. Am. J. Physiol. 169:
763-766, 1952.
1 75. Pullinger, B. D., and H. W. Florey. Some observations
on the structure and functions of lymphatics: Their
behavior in local edema. Brit. J. Exptl. Pathol. 16: 49-61,
!935-
176. Rabin, E. R., and E. C. Meyer. Cardiopulmonary effects
of pulmonary venous hypertension with special reference
to pulmonary lymphatic flow. Circulation Research 8:
324-335. !96°-
177. Rampone, A. J. Experimental thoracic duct fistula for
conscious dogs. J. Appl. Physiol. 14: 150-152, 1959.
1072
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
I78.
179.
180.
181b
182.
183.
185.
186.
.87.
188.
189.
190.
192.
193-
194.
'95-
196.
197-
Rampone, A. J. Role of phospholipids in lymphatic trans-
port of dietary lipids in the dog. Am. J. Physiol. 199:
1015-1020, i960.
Rampone, A. J., and J. D. Sigurdson. Effect of bile
deprivation on absorption and lymphatic transport of
dietary soaps and triglycerides in the dog. Physiologist
3 (No. 3) : 128, i960.
Rapport, D., R. Guild, and A. Canzanelli. The
transmission by crossed circulation of a shock producing
factor. Am. J. Physiol. 143: 440-443, 1944.
Rawson, A. J. Distribution of the lymphatics of the
human kidney as shown in a case of carcinomatous
permeation. A. MA. Arch. Pathol. 47: 283-292, 1949.
Recklinghausen, F. T. von. Die Lymphgejdsse und Ihre
Bez'ehung zum Bindegcwebe. Berlin: Hirschwald, 1962.
.Recklinghausen, F. T. von. Zur Fettesorption. Arch,
pathol. Anat. 26: 172-208, 1862.
Reinhardt, W. O., M. C. Fishler, and I. L. Chaikofp.
The circulation of plasma phospholipids : Their transport
to thoracic duct lymph. J. Biol. Chem. 152: 79-82, 1944.
Reizenstein, P. G., E. P. Cronkite, L. M. Meyer, and
E. A. Usenik. Lymphatics in intestinal absorption of
vitamin B,2 and iron. Proc. Soc. Exptl. Biol. Med. 105:
233-236. I96°-
Ritchie, H. D , J. H. Grindlay, and J. L. Bollman.
Flow of lymph from the canine liver. Am. J. Physiol.
196: 105-109, 1959.
Rouviere, H., and G. Valette. Physiologic du Systeme
Lymphatique. Paris: Masson, 1937.
Rudbeck, O. Nova Exercilatio Anatomica, Exlubens Ductus
Hepaticos Aquosos, el Vasa Glandulorum Serosa. Upsala, 1653.
Rusznyak, I. New studies on the physiology and pathology
of the lymphatic circulation. Minerva med. 45: 1468-1473,
1954-
Rusznyak, I., M. Foldi, and G. Szab6. Lymphagio-
spasm. Acta Med. Scand. 137: 37-42, 1950.
Rusznyak, I., M. Foldi, and G. Szabo. Lymphatics and
Lymph Circulation. New York: Pergamon, i960.
Sabin, F. R. A critical study of the evidence presented in
several recent articles on the development of the lym-
phatic system. Anat. Record 5: 417-443, 191 1.
Sabin, F. The origin and development of the lymphatic
system. Bull. Johns Hopkins Hosp. 17: 347-440, 1916.
Sage, H. H., and B. V. Gozun. Methods for studying
lymphatic function in intact man utilizing Au198. Proc. Soc.
Exptl. Biol. Med. 97 : 895-896, 1 958.
Salter, W. T. Circulating thyroid hormone in blood
andlymph. Western J. Surg. Obstet. Gynecol. 55: 15-25, 1947.
Sappey, P. G. Anatomic, physiologic, pathologic des vaisseaux
lymphatiques consideres chez I'homme et les vertebres. Paris:
A. Delahaye, 1874.
Schmidt, C. F., and J. M. Hayman. A note upon lymph
formation in the dog's kidney and the effect of certain
diuretics upon it. Am. J. Physiol. 91 : 157-160, 1929.
Shafiroff, B. G. P., H. Doubilet, A. L. Preiss, and F.
CoTui. The effect of thoracic duct drainage and hemor-
rhage on the blood and lymph. Surg. Gynecol. Obstet. 76:
547-55°. '943-
Shirley, H. H., Jr., C. G. Wolfram, K. Wasserman,
and H. S. Mayerson. Capillary permeability to macro-
molecules: stretched pore phenomenon. Am. J. Physiol.
190: 189-193, 1957.
■99-
203.
204.
205.
206.
207.
208.
209.
213.
214.
2>5
216.
217.
Shrewsbury, M. M. Thoracic duct lymph in unanesthe-
tized mouse. Method of collection, rate of flow and cell
content. Proc. .Sue. Exptl. Biol. Med. 101 : 492-494, 1959.
Silk, M. H., and A. R. R. Mears. Withdrawal of periph-
eral lymph from the foot of the dog. J. Appl. Physiol. 1 4 :
212-214, ]959-
Silvester, C. F. On the presence of permanent communi-
cations between the lymphatic and the venous system at
the level of the renal veins in adult South American
monkeys. Am. ./. Anat. 12: 447-460, 1911-12.
Simmonds, W. J. The effect of fluid, electrolyte and food
intake on thoracic duct lymph flow in unanesthetized
rats. Australian J. Exptl. Biol. Med. Sci. 32: 285-299, 1954.
Smith, R. O. Lymphatic contractility — A possible intrinsic
mechanism of lymphatic vessels for the transport of
lymph. J. Exptl. Med. 90 : 497-509, 1 949.
Starling, E. H. The Fluids of the Body. Chicago: Keener,
1908.
Sugerman, J., M. Friedman, E. Barrett, and T. Addis.
The distribution, flow, protein and urea content of renal
lymph. Am. J. Physiol. 138: 108-112, 1942.
Swann, H. G., A. A. Ormsby, J. B. Delashaw, and W. W.
Tharp. Relation of lymph to distending fluids of the
kidney. Proc. Soc. Exptl. Biol. Med. 97: 517-522, 1958.
Swell, L., M. D. Law, H. Field, Jr., and C. R. Tread-
well. Composition of lymph cholesterol ester fatty acids
after feeding of cholesterol and oleic acid. Proc. Soc. Exptl.
Biol. Med. 104: 7-8, i960.
Swell, L., E. C. Trout, Jr., H. Field, Jr., and C. R.
Treadvvell. Labelling of intestinal and lymph cholesterol
after administration of tracer doses of cholesterol-4-C.
Proc. Soc. Exptl. Biol. Med. 101 : 519-521, 1959.
Tasker, R. R. The collection of intestinal lymph from
normally active rats. J. Physiol. 115: 292-295, 1951 .
Taylor, G. W., J. B. Kinmonth, E. Rollinson, J. Rot-
blat, and G. E. Francis. Lymphatic circulation studied
with radioactive plasma protein. Brit. Med. J. 1 : 133-137,
■957-
Uhley, H., S. E. Leeds, J. J. Sampson, and M. Fried-
man. Some observations on the role of the lymphatics in
experimental acute pulmonary edema. Circulation Research
9: 688-693, 1961.
Vahouny, G. V., I. Fawal, and C. R. Treadwell.
Factors facilitating cholesterol absorption from the
intestine via lymphatic pathways. Am. J. Physiol. 188:
342-346, 1957.
Vahouny, G. V., and C. R. Treadwell. Changes in
lipid composition of lymph during cholesterol absorption
in the rat. Am. J. Physiol. 191 : 179-184, 1957.
Vahouny, G. V., and C. R. Treadwell. Absorption of
cholesterol esters in the lymph-iistula rat. .4m. J. Physiol.
i95:5ID-52°, i958-
Von Kaulla, K. N., and E. B. Pratt. Influence of
intravenously administered heparin on clotting of lymph
in the dog. Am. J. Physiol. 187 : 89-93, '956-
Warren, M. F. The lymphatic system. Ann. Rev. Physiol.
2: 109-124, 1940.
Warren, M. F., and C. K. Drinker. The flow of lymph
from the lungs of the dog. Am. J. Physiol. 136: 207-221,
1 942.
Warren, M. F., D. K. Peterson, and C. K. Drinker.
The effects of heightened negative pressure in the chest,
PHYSIOLOGIC IMPORTANCE OF LYMPH
10/3
together with further experiments upon anoxia in increas-
ing the How of lung lymph. Am. J. Physiol. 137: 641-648,
[942.
218. Wasserman, K , J. D. Joseph, and H. S. Mayerson.
Kinetics of vascular and extravascular protein exchange
in unbled and bled dogs. Am. J. Physiol. 184: 175-182,
1956.
219. Wasserman, K., L. Loeb, and H. S. Mayerson. Capillary
permeability to macromolecules. Circulation Research 3:
594-603, 1955.
220. Wasserman, K., and H. S. Mayerson. Dynamics o
lymph and plasma protein exchange. Cardiologia 21:
296-307, 1952.
22!. Wasserman, K.., and H. S. Mayerson. Mechanism of
plasma protein changes following saline infusions. Am.
J. Physiol. 170: 1 -10, 1952.
222. Webb, R. C, and T. E. Starzl. The effect of blood
vessel pulsations on lymph pressure in large lymphatics.
Bull. Johns Hopkins Hosp. 93: 401-407, 1953.
223. Webb, R. L. The lymphatic system. Ann. Rev. Physiol.
14:315-327. !952-
224. Webb, R. L., and P. A. Nicoll. Behavior of lymphatic
vessels in the living rat. Anat. Record 88: 351-367, 1944.
225. Webb, R. L., and P. A. Nicoll. Persistence of active
vasomotion along blood and lymphatic vessels in bat's
wing after denervation. Anal. Record 109: 414, 1951.
226. Wessely, J. Lymph circulation of dogs in experimental
thermal, hemorrhagic and tourniquet shock. Acta Physiol.
Acad. Sci. Hung. 14: 327-351, 1958.
227. White, A. The lymphatic system. Ann. Rev. Physiol. 1 1 :
355"386. >949-
228. White, R. P., and P. H. Woodward. Studies on heparin
release in anaphylactic dogs. Federation Proc. 9: 134, 1950.
229. White, R. P., and P. H. Woodward. Relation of size
of shock dose of antigen to blood pressure fall and heparin
release in canine anaphylactic shock. Federation Proc. 1 1 :
171-172, 1952.
230. White, R. P., and P. H. Woodward. Heparin content
of thoracic duct lymph following shock in dogs. Am. J.
Physiol. 1 88 : 1 89- 1 92 , 1 957.
231. Wicksell, F. A simplified method for estimating the
histaminolytic activity of plasma in pregnancy. Acta
Physiol. Scand. 17: 359-369, 1949.
232. Wicksell, F. Observations on histamine and histaminoly-
sis in pregnancy. Acta Physiol. Scand. 17: 395-414. 1949.
233. Woo, C. H., and C. R. Treadwell. Lipide changes in
chylomicra and subnatant fractions of rat lymph during
cholesterol absorption. Proc. Soc. Expll. Biol. Med. 99:
709-712, 1958.
234. Yoffey, J. M., and F. C. Courtice. Lymphatics, Lymph
and Lymphoid Tissue. Cambridge: Harvard Univ. Press,
I956-
235. Zamecnik, P. C, J. C. Dub, A. M. Brues, S. S. Kety,
I. T. Nathanson, A. L. Nutt, and A. Pope. The toxic
factors in experimental traumatic shock. IV. Chemical
and enzymatic properties of muscle. J. Clin. Invest. 24:
850-855. !945-
CHAPTER 31
The peripheral venous system
ROBERT S. ALEXANDER
Department of Physiology, Albany Medical College,
Union University, Albany, New York
CHAPTER CONTENTS
Anatomical Considerations
Structure
Vasa Venarum
Innervation
Venous Valves
Venous Capacity
Physiological Characteristics of Veins
Principles of Venous Hemodynamics
Venous Distensibility
Nature of Venous Constriction
Assessment of Venomotor Activity
In Vitro Studies
Direct Observation
Inferences from Venous Pressure
Measurements of Pressure Gradients
Pressure Measurements in an Occluded Venous Segment
Pulse Methods
Venous Distensibility Patterns
Distensibility by Venous Increment
Summary of Venomotor Responses
insight which anyone may gain into the function of
the arterial system by the simple registration of arte-
rial blood pressure. As a consequence, very few
physiologists or students of physiology have had any
personal opportunity to make observations, other
than the experiments of Harvey, which could be
interpreted with confidence as manifestations of
venous function. A major effort of this presentation,
therefore, will be to stress the technical problems of
obtaining reliable information concerning venous
function and to review the degree to which presently-
imperfect methods have yielded interpretations that
are in substantial agreement. This will lead us to
some positive convictions about the functional role
of the venous system in spite of many unresolved prob-
lems of methodology.
ANATOMICAL CONSIDERATIONS
if one were to consult textbooks for information
on venous physiology, the impression would be gained
that knowledge of this subject has not progressed
since the classical observations of William Harvey.
The error of this misconception should have been
laid to rest by the excellent review of Gollwitzer-Meier
(36), in 1932 and the comprehensive monograph of
Franklin (32) published in 1937. The bibliography of
this monograph, containing well over 1 000 references,
is scarcely compatible with the ignorance of the subject
which is often reported. It is our impression that a
major deterrent to appreciation of our knowledge of
venous function stems from a failure to develop valid
techniques that can be applied to the venous system
with ease and technical accuracy, comparable to the
Structure
In general structural pattern, veins are composed of
the same elements as are the arteries, but with some
important quantitative differences. Surrounding the
endothelial lining of the lumen is a network of elastic
and collagenous fibers which form a clearly defined
intima only in the larger veins; in the smaller veins
there is very poor differentiation of the intimal layer.
Encircling these intimal fibers is the muscular media,
which remains essentially a layer of spirally arranged
smooth muscle fibers without any major contribution
of elastic fibers. This lack of a heavy elastic investment
of the media constitutes the major structural differ-
ence between veins and arteries. Externally, the vessel
is surrounded bv the meshwork of elastic and col-
!°75
1076
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
lagenous fibers constituting the adventitia. The
adventitial layer becomes the major component of
the wall of larger veins.
Another important difference between arteries and
veins is in the structural relationships adjacent to the
capillary bed. Whereas the arterial channels possess
significant muscle terminating in conspicuous pre-
capillary muscular elements at the arteriole-capillary
junction, minute venules are devoid of muscle. Con-
verging capillaries become surrounded with a col-
lagenous network to form small venules which mav
not acquire a continuous muscular media until
diameters of the order of 0.5 mm are reached. It must
be clearly recognized on a purely structural basis,
therefore, that there is no mechanism at the venous
end of the capillaries capable of throttling blood flow-
in the way that blood flow may be controlled at the
arteriolar end (85).
In describing these general structural features of
veins, reference is specifically being omitted to some
of the important variations which are found in the
adaptations of specific venous beds to local problems.
Bv way of illustration, suffice it to say that in the long
veins of the extremities there is the development of a
significant component of longitudinally oriented
muscle capable of counteracting the gravitational
stresses to which these vessels are subjected, while
within the cranium venules develop to considerable
size without the appearance of any muscular ele-
ments (61).
Vasa Yenarum
Crucial to an understanding of some aspects of
venous function is a recognition of the role of the vasa
venarum, which constitute the normal route through
which both nutrients and vasoactive substances reach
the vein wall. Older literature on this topic has been
reviewed by Ramsey (76). There is a dense network
of minute vessels in the adventitia of the larger blood
vessels which is particularly conspicuous in veins.
Although some techniques have failed to reveal a
penetration of the capillary plexus into the media,
adequate methods have succeeded in demonstrating a
profuse capillary bed extending almost to the intima
(67). In addition to the capillary plexus, there is clear
evidence of an accessory duct system, presumably
lymphatic in nature, which is distributed through the
adventitia and media.
It mast be emphasized, however, that the vasa
venarum do not penetrate the intimal layer and drain
through the local endothelium. Venous drainage from
the capillary plexus returns to venules running along
the superficial layer of the adventitia, and eventually
drains into either an entirely different vein or a
remote site of the same vein. O'Neill has pointed out
that this relationship assures that local obstruction in
a venous segment will not block the flow in the vasa
venarum, nor can local pockets of high intraluminal
pressure induce backflow in the vasa venarum of the
venous wall. A similar relationship exists in the
arteries.
Functional confirmation of the anatomical rela-
tionships described above has been provided by
O'Neill (67). Extensive damage to the intima fol-
lowed stripping the tissues surrounding the vein so as
to interrupt the vasa venarum, even though blood
flow was maintained through the lumen of the vein.
This indicates that oxygen and nutrients do not pass
in significant amounts from the lumen into the sur-
rounding tissue of the vein wall, and that the venous
wall is clearly dependent upon the vasa venarum.
Comparable evidence may be observed with drugs.
Minimal response to vasoactive agents can be demon-
strated when the drug flows through the lumen of the
veins, while very effective vascular responses result
when the drug is applied systemically or topically so
that it may reach the media from the adventitial side.
This anatomical arrangement seriously handicaps
the study of functional changes in the vasa venarum.
In the case of arteries, the vascular wall is supplied
not only by vessels penetrating from the adventitial
side, but also by some vasa vasorum interna which
penetrate the wall directly from the lumen. Smith
(83, 84) has taken advantage of this relationship to use
the amount of leakage, in response to internal pres-
sure changes, from the surface of an excised arterial
segment as a measure of vasa vasorum flow. A similar
technique would not be applicable to veins. In the
study of diseased veins, a further complication arises
in that the anatomical pattern changes qualitatively
as well as quantitatively. The early inflammatory
phase of vascular disease stimulates a dense invasion
of vascular elements into the wall of the vessel with the
creation of venous channels that penetrate the intima
directly into the lumen of the vessel, creating vasa
venarum which have no counterpart in normal veins.
Innervation
Veins are copiously supplied with nerves which
Thompson (86) demonstrated, in 1893, to be capable
of producing constriction of the vein. Bayliss & Star-
ling (8) confirmed the existence of neurogenic veno-
PERIPHERAL VENOUS SYSTEM
IO77
constrictor mechanisms and, on the basis of changes
in arterial and venous pressure following spinal
transection, inferred that the nervous system must be
of importance in maintaining venous tone. Donegan
(19) studied this innervation in greater detail and
established that it was sympathetic in nature. As with
other sympathetic pathways, localization is rather
gross, with a given vein segment responding to stim-
ulation from several adjacent spinal segments. There
is now clear evidence of a tonic constrictor activity
of this sympathetic innervation, since venous dilation
occurs with sympathectomy (9, 54) or with sym-
patholytic drugs. This adrenergic sympathetic influ-
ence appears to be purely constrictor without any
dilator component (62). Conversely, there appears
to be no evidence of parasympathetic innervation of
veins. Although pharmacological doses of cholinergic
drugs may influence venous musculature, neither
parasympathetic stimulation nor atropinization have
any effect on venous tone (31). Even in such a highly-
specialized vascular function as penile erection, para-
sympathetic control appears to be restricted to the
arterial side of the circulation, with the veins playing
a purely passive role (47).
Venous Valves
A unique feature of the venous system is the pres-
ence of venous valves. The dramatic simplicity with
which the nature of this valve action can be demon-
strated in the veins on the dorsum of the hand remains
one of the classical observations of physiology. Clini-
cally, the role of the venous valves in the lower ex-
tremities have received particular attention. The
superficial veins of the leg, lacking protection from
surrounding muscle, are often subjected to prolonged
hydrostatic loads. This excessive distension of the
vessels may eventuate in valvular incompetence. The
consequences of this valvular incompetence and its
relation to venous varicosities and varicose ulcers has
been analyzed extensively in the literature on periph-
eral vascular surgery. (See Burch, Chapter 36.)
This focusing of attention on the venous valves of
the extremities has distorted an appreciation of the
significance of valves in the venous system as a whole.
For example, a widely prevalent notion is typified by
the following statement from a leading textbook of
histology (46): "Valves are especially abundant in
the veins ot the extremities and they are generally
absent from the veins of the thorax and abdomen."
In actual fact, valves or valve-like structures have
been reported in most segments of the venous system,
although generalizations are difficult because of the
marked species variation which has been reported
(29). The distinguishing feature of the valves in the
extremities is not their presence but the degree of
competence which thev exhibit; venous valves in
areas not confronted with severe hydrostatic strains
are usually more rudimentary and therefore less
easily demonstrated.
An example of the latter type of valvular structure
in abdominal veins is illustrated in figure 1 from a
preparation made by Dr. Darrell Davis. This is a
photograph of a plastic cast of venous vessels obtained
by retrograde injection of a segment of dog intestine.
Numerous valve impressions are clearly indentifiable
on this preparation. Just before each point of junction,
most of the tributaries contain a valve. In every in-
stance the injection mass terminates with a bilobed
indentation which clearly represents a valve. Blood
traversing this venous bed must pass through a series
of valves before gaining access to the portal vein.
It should be appreciated that figure 1 also demon-
strates a relative incompetency of these valves, in that
the injection mass has readily passed beyond a num-
ber of the valves. Accordingly, in spite of the profusion
of valves in this bed, it is reasonably easy to reverse
the flow of blood in the intestine. If artery and vein of
a loop of dog intestine are sectioned and a circuit
re-established whereby the intestinal vein is connected
to an arterial supply and the artery led out to a route
of venous drainage, a substantial retrograde flow is
observed for several minutes, eventually becoming
reduced as massive edema develops from the abnormal
capillary pressure relationships. Measurement of
pressure gradients and flow demonstrates that retro-
grade flow encounters a resistance of three to ten
times the vascular resistance to forward flow during
fig. 1 . Plastic cast of mesenteric veins of a dog demonstrating
multiple valves. (Preparation made by Dr. Darrell Davis.)
1078
II WMS! 11 ik 11I I'll', Ml il 1 ". -i
CIRCULATION II
the early phase of the reversal. While a good part of
this resistance is undoubtedly attributable to the
valves, the situation contrasts with that observed
in healthy veins in the extremities where valves
present infinite resistance to retrograde flow until
very high pressures are reached.
The functional contribution of venous valves
should be clearly defined. In the idealized circulatory
scheme with continuous venous flow, the valves must
necessarily remain open and hence make no func-
tional contribution. With intermittency of flow, due,
for example, to intermittency of flow in the peripheral
bed, the valves would tend to close during the inter-
vals of flow cessation. Nevertheless, we must reject
the view that this "breaking up" of the venous
column into segments relieves the dependent parts
from the hydrostatic load of a continuous fluid
column. The hydrostatic gradient is inherent in the
hydraulics of the system; energetically the valves
cannot contribute to the need for adequate pressure
energy to overcome the hydrostatic barrier between
dependent parts and the heart. In the system as a
whole, intermittent flow must preserve the same mean
pressure gradient as is required of continuous flow.
Valves make their functional contribution by
translating extramurally applied forces into flow
energy. When an external force compresses a fluid-
filled vessel, local intramural pressure will rise and
tend to drive the blood in both directions from the
point of compression. The actual flow which will
occur in the two possible directions will be a function
of the pressure gradients and resistances in the alter-
nate directions. The resistance to retrograde flow
toward the capillary bed is far higher than the re-
sistance to forward flow toward the heart, and the
pressure gradient, which normally favors central
return, would very strongly favor central flow the
moment the retrograde flow combined with con-
tinuing capillary drainage to build up peripheral
venous pressure. Thus it should be appreciated that a
'"milking" action of intermittent venous compression
will effectively propel blood toward the heart even
in the complete absence of valves. Valves, however,
can greatly increase the efficiency of this process by
producing an almost immediate rise of retrograde
resistance to infinitv.
The importance of this process is most clearly
demonstrated by the dramatic relief from orthostatic
hypotension which is produced by movements of the
legs and their associated compressing forces on the
leg veins. Walking movements are so effective in pro-
pelling flow up the venous channels that they can
restore adequate venous return to the heart even
when vasomotor tone has been completely abolished
by sympatholytic drugs (71). Direct measurements
have demonstrated that the sequential compression
of venous segments during walking milks blood up the
legs efficiently enough to reduce the pressure in the
uncompressed veins of the ankle to less than one
quarter of the hydrostatic gradient from the ankle to
the heart (74). There is no question that such extra-
vascular forces constitute a significant "booster pump"
(49) for maintaining the circulation. The idea cham-
pioned by Henderson (48), that muscular activity
acting in conjunction with the venous valves was the
primary "venopressor" mechanism, should therefore
not be dismissed lightly, even though this mechanism
does not appear to play quite such a comprehensive
role in maintaining venous return as Henderson
claimed (90).
Venous Capacity
Because of the relatively large caliber of veins, and
also the conspicuous venous sinusoids which occur in
some organs, it is commonly supposed that the capaci-
tative function of the vascular bed resides dominantly
in the venous system. The full functional significance
of this concept will be developed in the next chapter.
Our concern at the moment is confined to examining
the evidence underlying this basic assumption.
Widely quoted data in support of this venous reser-
voir concept are those published by Green (37) calcu-
lated from an analysis of the intestinal vascular bed
of the dog reported by Mall. These data picture some
70 per cent of the vascular capacity to reside in the
venous system with 62 per cent in veins greater than
1 mm in diameter. Landis & Hortenstine (58) carried
out a very similar calculation based upon the in-
testinal data of Schleiser, which yielded a value of
75 per cent of the vascular volume within the venous
system, 50 per cent of the total being found in veins
greater than 1 mm in diameter. The Landis calcula-
tion did not include the venae cavae, inclusion of
which would have increased the percentage of
volume in the large veins.
It is distressing, however, when one realizes how
little direct evidence there is to support these esti-
mates. By measuring the mean transit time for dye
passage between the femoral vein and the right
atrium, Milnor & Bertrand (65) were able to calcu-
late a volume between these two sites which averaged
18 per cent of the total blood volume. Since the in-
ferior vena caval system is a notoriously poor mixing
PERIPHERAL VENOUS SYSTEM
1079
chamber, one would expect the true volume to exceed
that calculated by this method. This study therefore
represents some substantial support for the idea that
large veins represent a significant contribution to the
total vascular capacity. On the other hand, Knisely
and associates (57) have challenged the conventional
point of view with some data obtained from plastic
injections of whole rats. Using a plastic free of
particulate matter which flowed freely through the
circulation when initially injected, they obtained
casts of the entire vascular bed. When these casts were
fragmented, and the fragments sorted according to
caliber, over 80 per cent of the plastic was found to
be contained in vessels with a diameter of less than
200 n and only 12 per cent in vessels larger than 700 n.
This study suffers from the fact that veins are col-
lapsible, and it is not at all clear that their method
would have preserved a normal degree of filling of
the venous system. Nevertheless, such an extreme dis-
crepancy between the relative contribution of large
vessels and small vessels to total vascular capacity
clearly challenges the point of view that is usually
held. The burden of proof has been returned to the
proponents of the venous reservoir concept to offer
some more substantial documentation of their
hypothesis.
Apart from the question of total capacity, however,
there is much better support for the thesis that the
venous division, together with the lesser circulation, is
the most variable capacity of the vascular bed. One
very simple observation leading to such an inference
is the minimal change in pressure produced by an
injection into the venous system as compared to the
pressure change produced by an injection of an equal
volume at the same rate into the arterial system. More
direct evidence on this point was presented by Green-
field & Paterson (39), who compared volume changes
in the forearm produced by venous obstruction to
the volume changes produced by a negative pressure
applied to the whole arm. In the former instance, the
increase in transmural vascular pressure would be
essentially confined to the venous side; in the suction
experiment, the transmural pressure of all vessels
should be increased equally. Yet venous occlusion
provided 85 per cent of the volume increase observed
when suction was applied to the whole arm. A similar
experiment was reported by Capps (16). Such data,
together with venous distensibility characteristics to
be discussed later, justify reasonable confidence in the
hypothesis that the venous system plays an important
role in contributing a reservoir of variable capacity
to the vascular system.
PHYSIOLOGICAL CHARACTERISTICS OF VEINS
Principles of Venous Hemodynamics
The most frequent measurement made of the
venous system is the venous pressure. For the pur-
poses of our interests however, venous pressure is of
relatively little meaning. As competently reviewed by
Landis & Hortenstine (58), venous pressure can have
profound influence on capillary dynamics and the
transudation of fluid across the capillary endothelium.
Central venous pressure plays a key role in cardiac
filling and the control of cardiac output. For reasons
that will be developed shortly, however, venous pres-
sures tell very little about the venous system itself.
Indeed, it can be fairly stated that venous pressure
measurements in themselves are just as unimportant
to the physiologist interested in the venous system as
they are important to the physiologist interested in the
arterial system.
Before specifically considering the principles of
hemodynamics underlying this statement, a word of
emphasis should be given in reference to the implica-
tion of the studies of Pappenheimer (69, 70) on the
capillary bed. He has lucidly argued that, under
steady-state conditions, the mean capillary pressure
must be in equilibrium with the effective osmotic
pressure of the plasma proteins, excluding lymph
flow which at best is a small fraction of total blood
flow. Since central venous pressure shows relatively
small variations under most conditions, this indicates
a relatively constant pressure gradient from capillaries
to the central veins. Furthermore, large changes
in blood flow may occur without significant altera-
tions in either plasma protein concentration or central
venous pressure, yielding the apparent paradox of
blood flow that varies widely in spite of a fixed pres-
sure gradient. A corollary to this is that the venous
smooth musculature cannot effectively control blood
flow by imposing a variable resistance in the venous
portion of the circulation. The role of the venous
musculature must therefore be confined to producing
capacity changes in the system. These capacity
changes can indirectly influence blood flow only
insofar as more effective venous return increases
cardiac output, or higher mean circulatory pressures
increase capillary transudation.
If some of the preceding statements appear to be
in conflict with irrefutable principles of fluid dy-
namics, the reader must be reminded that the venous
system is a collapsible system and is therefore not
governed by the usual principles of fluid dynamics in
io8o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
cylindrical tubes. The phenomenon of venous col-
lapse was inherent in the classical observations of
Harvey. It has remained a commonplace observation
in the use of the height above the heart at which super-
ficial veins collapse as a clinical estimate of central
venous pressure. Yet the hemodynamic significance of
venous collapse has been all too rarely appreciated.
First, it is important to note that '"collapse" of
veins is not an all-or-none characteristic. Complete
collapse of the vein with obliteration of its lumen
represents an obstruction to blood flow which can
onlv exist on a transient basis. The collapse phenome-
non relates to the fact that the vein wall is not struc-
turally self-supporting. Energy is required to push
the vein walls out into a cylindrical configuration.
Any time that the intraluminal pressure becomes
equal to or less than the extravascular pressure, the
venous walls will tend to approximate each other in
an ellipsoidal cross section (73).
This is best visualized by examining veins above
heart level. Hydrostatic forces will act to drain blood
from these veins and create a negative intraluminal
pressure. In addition, finite tissue pressures always
produce some degree of positive extravascular com-
pression. In the absence of blood flow, such a vessel
would remain completely collapsed. To preserve
flow in such a segment, intraluminal pressure must
be raised until it slightly exceeds the extravascular
pressure so as to open the collapsed vein. Intraluminal
pressure must further be elevated enough above
extravascular pressure to provide the necessary pres-
sure head to produce forward flow against the re-
sistance it confronts. However, since a slightly positive
transmural pressure will widen the collapsed lumen
and produce a marked fall in resistance, very little
pressure gradient is required to produce flow. There-
fore, the pressure measured in veins that are above
heart level will be essentially the same as the extra-
vascular tissue pressure, as originally emphasized by
Holt (51, 52, 80). It follows that such pressures have
no hemodynamic significance in the usual sense of
gradients along the vascular circuit, and they are in
no way specifically related to constriction or dila-
tion of the veins.
A more rigorous statement of this relationship has
been clearly set forth in the exposition by Brecher
(11). The classical formulation of the Poiseuille law
for cylindrical tubes:
resistance oc radius *
must be modified for collapsible tubes to the more
complex expression:
Rex
2a3b3
in which R is resistance and a and b are the major
and minor axes of the ellipse. It should be noted that
in a cylinder where a = b, the second expression
reduces to the first. As an operational tool, this
formulation of resistance relationships is rarely of
practical value to the physiologist because the desired
dimensions are not accessible. Nevertheless, from a
theoretical standpoint it defines the fact that re-
sistance to flow will increase markedly as the vessel
progressively collapses to a flattened ellipse.
The full import of this collapsibility resides in the
consequences it has upon the significant variables
determining blood flow. In a system of cylindrical
tubes, as represented by the arterial system, pressure
is normally maintained at homeostatic levels in the
arterial reservoir and, for any given vascular bed,
blood flow is controlled by resistance changes through
the activity of the vascular smooth muscle in the
arterial supply to that bed. To emphasize this point,
one might consider the pressure as essentially con-
stant (Pa) under a given situation and the significant
variables of flow (Q) and resistance (Ra) expressed as:
6 = 4r x P„
R„
In contrast, in any local venous bed, the flow is
obligate since in a steady state the veins must trans-
port the volume of blood delivered by arterial inflow.
Flow may therefore be considered constant and in
any venous segment resistance is controlled by the
local pressure (P,,) which, as outlined above, must
represent a small increment over the extravascular
pressure :
Stated descriptively, in the venous system operating
under a state of partial collapse, local venous pressure
determines the cross section of the ellipse and thereby
adjusts resistance to accommodate the volume of
flow presented to the system.
Extending this analysis further, in the arterial
system an increase in the reference pressure (P„) will
immediately lead to an equivalent increase in the
flow (neglecting factors of vessel elasticity and auto-
regulation). In the venous system, an increase in the
reference flow (Q.) will tend to increase pressures
slightly along the venous route. This will widen the
PERIPHERAL VENOUS SYSTEM
I08l
0
z
1 *//
\ « 1 ///«
\ K|^
in
in
LxJ
LEV
(0
a
to
«
UJ
a.
+ \ /
-0 \ /'
^/ \l Resistance
/ 1 .
- 0 +
*
0
_i
u.
Flow -— — »^^
-O+i
1 1
ad
ca
CENTRAL VENOUS PRESSURE
fig. 2. Hemodynamic relationships in the venous bed,
apted from Holt (51) and Brecher (10). Solid lines indi-
te relationships without compensation; dashed lines indi-
cate relationships with compensation in arterial inflow resis-
tance. "O" pressure refers to the hydrostatic level of the
peripheral vein.
ellipse and lower resistance so as to support an in-
creased flow without change in the total pressure
gradient from capillaries to heart.
The key relationships in this pattern of venous
hemodynamics are illustrated in figure 2. For veins
above heart level, the hydrostatic column of blood
descending toward the heart creates a potentially
subatmospheric intraluminal pressure tending to suck
the walls of the vein in to cause collapse. Typically,
this phenomenon will be most manifest at the point
just before the venous channel enters the thoracic
cavity, since in this region the negative intrathoracic
pressure combines with the hydrostatic forces to
aspirate blood from the veins. At the left of figure 2,
therefore, central venous pressure is indicated as
below atmospheric pressure; yet peripheral venous
pressure is maintained at a definite positive value.
Since under these conditions there is a significant
pressure gradient, an appreciable resistance exists
between the peripheral and the central veins. This
resistance is created by the state of partial collapse
near the central end of the channel. As the central
venous pressure rises toward atmospheric pressure,
the aspiration effect causing collapse becomes pro-
gressively less, so that the resistance to flow progres-
sively lowers. Peripheral pressure, however, remains
unchanged. As the central venous pressure rises
above atmospheric pressure, there continues to be an
interval when the intravascular pressure remains
below the extravascular pressure because of the
existence of a positive tissue pressure in the area sur-
rounding the veins.
As the central venous pressure reaches the value of
the extravascular tissue pressure, a dramatic altera-
tion occurs. The intraluminal pressure will now be
sufficient to prevent collapse of the vein. As a conse-
quence, the venous channel is distended and the more
typical Poiseuille relationship pertains. Neglecting
the minor influence of elastic distension of the veins,
resistance between the peripheral and central veins
remains constant at a relatively low value, and corre-
spondingly a relatively constant small pressure differ-
ence exists between the peripheral and the central
veins. Peripheral venous pressure will therefore rise
almost parallel with central venous pressure.
It is to be emphasized that at all central pressures
below the level of tissue pressure, the peripheral
venous pressure remains at essentially the same level
as the tissue pressure. Assuming there are no changes
in arterial pressure or resistance factors, a constant
peripheral venous pressure dictates a constant
capillary blood flow. It is to be noted, therefore, that
flow remains constant in spite of the significant
changes in pressure gradient and resistance along the
venous route. Once the central venous pressure rises
above the tissue pressure, venous congestion occurs
with a rise in peripheral venous pressure and a corre-
sponding reduction in the arteriovenous pressure
gradient. This will have some influence in reducing
flow through the system unless compensated by other
changes. In actual fact, a large rise in peripheral
venous pressure reduces peripheral blood flow so
that vasodilator metabolites accumulate in the
tissues. The resulting compensatory dilation of the
arterial inflow channels will counteract the elevation
of peripheral venous pressure and maintain constant
flow, as illustrated in the dashed lines of figure 2.
Any factors leading to a change in the extra-
vascular tissue pressure will produce an equivalent
change in the peripheral venous pressure, a corre-
sponding alteration in resistance, and a shift of the
point of inflection of the curves. Flow, however, will
remain unchanged.
Students of the venous svstem must wrestle with
io8i
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
this concept until they come to recognize its profound
implications. While the application of Poiseuille's
law to collapsible tubes requires only minor modifica-
tions in arithmetic, the degree of determinacy of the
respective variables is radically different. For example,
there is a vast body of older literature which demon-
strates that venous pressure is not altered by a host
of physiological factors known to alter the regulation
of the cardiovascular system. The conclusion of these
authors, that the venous system therefore played no
part in cardiovascular regulation, may be a signifi-
cant reason why most textbooks are devoid of positive
statements in regard to venomotor mechanisms.
What these experiments actually proved was that
the various experimental maneuvers had no effect on
the extravascular tissue pressure which is the major
determinant of venous pressure; such measurements
are completely meaningless in reference to venous
tone.
It would be a serious error, moreover, to regard the
collapsibility of veins as a structural defect in the
system which serves no better purpose than to com-
plicate the understanding of venous hemodynamics.
Since the venous system operates in the same pressure
ranges as the gravitational forces and tissue pressures
to which it is exposed, drastic disturbances would
result if veins were rigid tubes. Consider, for example,
an individual turning a handspring. If veins were
rigid, there would be drastic surges in venous blood
flow and chaotic alterations in venous return to the
heart. More conventional running and jumping move-
ments would seriously tax the homeostatic adjustment
of a low pressure system of cylindrical tubes. The
collapse mechanism serves to check such hydrostatic
shifts of venous blood. As soon as pressure in the
veins becomes reduced to tissue pressure levels, col-
lapse occurs to throttle flow and maintain the periph-
eral bed at more nearly normal functional levels.
Duomarco and associates (22-24) have extended
this concept of venous hemodynamics to claim that
the design of a collapsible venous system guarantees
that extravascular factors capable of altering pressure
relationships can have no influence on venous flow.
Duomarco's enthusiasm for the teleological magnifi-
cence of such a scheme apparently exceeds the actual
facts. It must be appreciated that, with normal
blood volume, a significant fraction of the venous bed
is distended so that it does behave as a system of
cylindrical tubes. This will hold for most of the
extrathoracic veins which are below heart level and
which are not in regions subjected to significant
extravascular compression. Furthermore, the work
of Brecher (10) has established that phasic pressure
changes are capable of producing phasic changes in
flow during the intervals when geometric adjustments
in the degree of collapse are taking place. Such
phasic pressure changes are conspicuous in intra-
thoracic and intra-abdominal veins in association
with respiration, and also seem to be a characteristic
manifestation of venous vasomotion in the small
peripheral veins (44).
An additional word of caution should be appended
to emphasize that there are some important excep-
tions to generalizations as to the collapsibility of
veins. This is particularly true of venous structures
that are bound by connective tissue to rigid skeletal
elements which prevent their collapse, such as the
sinuses of the dura mater and the vertebral venous
sinuses. In these vessels, gravitational or respiratory
forces may lower the intraluminal pressure to values
significantly below the pressure existing on the out-
side of the vessel. A clinical consequence is the danger
of air aspiration into the vascular system if these
vessels are opened to the atmosphere during surgical
procedures or by accidental trauma. A similar prob-
lem exists to a lesser degree at the point where veins
enter the chest. The thyroid surgeon is well aware
that veins near the base of the neck have sufficient
connective tissue attachments so that traction may
pull open an incised vein that has not been securely
ligated, and aeroembolism result when inspiratory
pressure changes lower the central venous pressure
below atmospheric pressure.
To qualify generalizations about venous collapse,
however, should not obscure the importance of this
phenomenon in venous function as a whole. Any
approach to the venous circulation which neglects
the collapsibility of veins will lead to serious distor-
tions of the hemodynamic factors which control the
flow of venous blood.
Venous Distensibihty
In view of the collapsible nature of veins, a vein
segment will empty freely from cut ends, the vessel
will flatten and all blood will leave the lumen except
for a minute amount retained within the folds on
opposite sides of the vessel. If fluid is now added to
this collapsed vessel, two theoretically distinct proc-
esses will occur. The first phase will be '"filling,"
during which the geometry of the vessel wall is
restored to the cylindrical shape without increasing
in circumference. The succeeding phase will repre-
PERIPHERAL VENOUS SYSTEM
IO83
sent elastic distension of the vein segment through
increase in its circumference and length.
The inference sometimes encountered, however,
that the vein remains at "zero" pressure until it is
"filled" is quite unrealistic. Some finite pressure is
required to restore the wall to its cylindrical shape.
For veins in situ, filling the vein must also overcome
tissue pressure, and local tissue pressure will itself
be augmented by the swelling of the vein. Finally, as
fluid starts to fill the vein, hydrostatic pressures will
be created unless the vein is perfectly horizontal.
Consequently, if one records intraluminal pressure
accurately while fluid is progressively added to an
empty vein in vivo, pressure starts to rise almost
immediatelv and does not show any recognizable
inflection when the vein is "filled" (fig. 3).
The pressure-volume curves obtained from venous
segments, or the tension-length curves obtained from
strips or rings, are most commonly described as
curvilinear with considerable convexity toward the
length or volume axis. Clark (17) was the first to
recognize that these two types of measurement must
be related by the Laplacian relationship whereby
in a cylindrical tube, wall tension ( T) increases as a
function of pressure (P) and radius (R ) :
T = P X R
It follows that at small radii, relatively less wall
tension is created by a given pressure increment than
at large radii. The pressure-volume curve shows
correspondingly less curvature than the tension-
length curve.
Figure 3 illustrates the volume change due to radial
distension as contrasted with the volume increase due
to elongation. It is apparent that most of the volume
increment results from radial distension within the
physiological range of venous pressures. It is only
above pressures of 30 to 40 cm H20 that radial dis-
tension becomes restricted to the relatively low degree
of distensibilitv which is characteristic of longitudinal
distension. A further important characteristic of the
longitudinal distension of veins, that is not shown by
such data, is the spiral twist which veins exhibit when
they are subject to sufficient pressure to produce
significant longitudinal stretch. Presumably because
of the spiral structures within the vein wall, there is a
definite rotation of one end of the vein in respect to
the other end as the vein lengthens. Teleologically,
70
60
so-:
40 —
30-
20-
10
c
— 0
01
e
u
— ul
en
in 1
t/i I
_UJ J
a. f
a. 1
/ Elongation
Radiol
D istension
- J
'To » a 1
" VOLUME
1 1
— cc
1
1
fig. 3. Volume distensibility of a segment of a dog's jugular vein in vivo 88.8 mm in initial length,
prepared by double ligation, cannulation, and ligation of side branches through small skin incisions
with as little disturbance of the surrounding tissue as possible. Length measurements were obtained
directly with calipers; radial distension was calculated from the known length and volume. Fluid
added at the rate of 0.4 cc/min.
1084
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
it appears reasonable to suggest that this spiral twist
of a distended vein may be of importance in relieving
kinking of the veins when the tissue is mechanically
distorted.
Quite a different curve of vascular distensibility
was presented by MacWilliam (60) as shown in
figure 4. He took the precaution to collect fresh
tissues and observe their behavior carefully during
the postmortem period. Shortly after a segment of
living vessel was excised, it developed marked spasm.
This spasm persisted for many hours if the tissue was
kept cool. In this contracted condition a stretch curve,
such as that shown in the lower half of the figure, was
observed. When the state of contraction was elimi-
nated by warming the vessel, a more conventional
stretch curve was obtained as shown in the upper
portion of the figure. The marked sigmoid curve ob-
served originally has now been replaced by a simple
bow convex to the length axis. MacWilliam inter-
preted the lower curve as a manifestation of the
resistance to stretch of the smooth muscle, which
gradually gave way as tension increased until even-
tually stretch was restricted by the elastic and fibrous
tissue of the vessel wall. The upper curve lacked this
muscle component, and therefore revealed the simpler
manifestation of elastic tissue distension. This change
in the distensibility pattern was quite characteristic
of arteries; it was not so evident in the veins studied
bv MacWilliam.
BRARY *
Length
fig. 4. Stepwise loading of a ring of artery which in the
lower section was in a contracted state from preservation in the
cold. In the upper tracing the identical loading sequence was
repeated after the vessel had been dilated by warming. [From
MacWiUiam (60).]
Constricted
Total Venous Volume *•
Fic. 5. Distensibility patterns recorded from veins in vivo (4).
A majority of investigators have considered this
spasm of the excised vessel as a postmortem artifact,
and many describe techniques employed to remove
this state of spasm in the tissue before carrying out
studies of its elastic behavior. The potential signifi-
cance of this observation of MacWilliam therefore
lay dormant for many years, until Capps (16) ob-
served the same type of curves in plethysmographic
recordings obtained from the human arm. The sig-
moid type of distensibility curve was associated with
constricted veins, and the smooth bow was associated
with dilated veins.
More recently the significance of this distensibility
pattern of constricted vessels and of dilated vessels
has been analyzed by the author (2-4, 78). There is
now ample evidence that these distensibility patterns
are exhibited by living veins in vivo (fig. 5) and, as
will be discussed later, their analysis can serve as a
useful index to the state of contraction of the veins.
Another feature of vascular distensibility, which has
been recognized since the time of Roy (79), is the
marked time dependency in elastic behavior (1).
This has been variously identified as "elastic after-
action," "elastic hysteresis," "delayed compliance,"
or probably more properly by the physical phe-
nomena of "stress relaxation," in which pressure dissi-
pates following sudden distension to a constant
volume, and "creep," in which volume slowly in-
creases after sudden distension by a constant pressure.
With a continuous cycle of injection and withdrawal
of fluid, this characteristic manifests itself as a wide
loop of disparity between the pressure-volume rela-
tionships observed on injection and the pressure-
volume relationships found on withdrawal (figs. 6
and 7). To a degree, the width of this loop demon-
strates time dependency, in that it tends to become
PERIPHERAL VENOUS SYSTEM
I085
60
c
J
V)
/ 1
E
/ /
40
„ u
Id
tr
^3
yV
,/
to
**'/ *'
IO
20
or
-f^\/ y'
a.
~?*
,*'
^^
h-°--o-°
(X,
50
A
\/
/
0.g=6=s-
«=6=S-o-
25-
£
6
\
• Total
"Small Vessels |
* Arteries #—<
0 Veins /j
20-
1.5-
10-
0.5 J
Time in Minutes
fig. 9. Pressure gradients and resistance changes recorded
from the foreleg and paw of a dog that was artificially perfused
at a constant rate of flow, demonstrating the application of the
Haddy technique (45).
constriction of the large arteries and veins, and only
a minor change in the smaller vessel segment.
The fact that pressures in minute vessels show con-
siderable pulsatile variation (41, 55), presumably
associated with a type of venovasomotion (44), renders
such measurements of greater qualitative than quan-
titative value, since it is difficult to have faith that
the discrete vessel from which the peripheral pressure
was recorded is accurately representative of the mean
pressure in the peripheral venous bed as a whole.
This reservation is reinforced by the possibility that
the catheter might be wedged so that the recorded
pressure represents that of a collateral somewhat
remote from the catheter, and also the possibility
that the presence of the catheter itself might alter
pressure-flow dynamics. Nevertheless, if one keeps
these reservations in mind, and takes precautions to
be extremely critical of instrumental techniques and
record analysis so as to exclude the many possible
artifacts which may creep into a method of this type,
this appears to be one of the most valuable techniques
currently available for studying peripheral venous
function. It should be appreciated that this method
also has the merits of being applicable to human
studies (88).
Pressure Measurements in an
Occluded Venous Segment
Doupe and associates (20) appear to have been
the first to have employed the method of isolating a
segment of a superficial vein between a pair of com-
pressing wedges. If the wedges are placed so that the
intervening segment is free of branches, this creates
a blind cul-de-sac with an entrapped volume. If a
needle is then carefully introduced into this segment,
it is possible to record pressures in a system of fixed
volume, hence any change in pressure must reflect a
change in the muscle tone. In actual practice, it ap-
pears to be necessary to distend the vein with a con-
siderable volume increment so as to yield basal pres-
sures of the order of 40 to 60 cm H20 in order to
achieve significant pressure changes. This also serves
the purpose of elevating the recorded pressure out of
the range of confusion with extravascular pressure
effects.
A recording from such a preparation is illustrated
in figure 10. This illustrates one feature that is con-
spicuously demonstrated by this technique: the
marked influence of psychic stimulation on venomotor
response as the experimental subject witnesses the
preparation of some experimental maneuver which
he is to undergo. Since responses of the type illu-
strated in figure 10 disappear after sympatholytic
drugs, there can be little doubt as to their venomotor
origin. This technique has been exploited with con-
siderable success (14, 68) and is the most direct quali-
tative method for studying the venous reactions of
the human subject. Because it concerns itself with a
unique venous segment, however, it does not appear
feasible to standardize the quantitation of pressure
responses observed in such studies.
A variation of this technique is to introduce a small
cylindrical balloon into a vein segment and measure
pressure changes in this balloon. This obviates the
necessity of selecting a segment free of branches, and
gives greater confidence that the observations are
being obtained from a closed system. As with the
method previously described, it also seems essential
to distend the balloon to pressure of 30 cm H20 or
higher in order to observe significant changes. Salz-
man (81) has successfully applied this technique to
the study of the venomotor response to pressoreceptor
reflexes. Connolly & Wood (18), on the other hand.
iogo
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
RISE IN PRESSURE
^-^ mm Hg
/
EFFECT OF ANTICIPATION OF
< HAND IN ICE -
TEST
WATER
1
A
1
i
0
1
2
3
TIME IN MINUTES
PNEUMOGRAM
^.x*s/Wv/Wv-s->M^ •*?}
fig. io. Pressure changes recorded from a segment of a superficial vein of the forearm that had
been isolated between wedges and kept at constant volume. [From Duggan et al. (21).]
were unable to record temperature reactions in
superficial veins in man by this technique, even
though veins adjacent to that containing the balloon
exhibited obvious caliber changes. The author has
similarly been unsuccessful in attempting to get suffi-
cient response for accurate analysis in a variety of
applications of this technique to various dog prepa-
rations. Further studies are in order to determine the
full potentialities and limitations of this method.
Another type of method which is based on the
same principle was introduced by Hooker (54) and
has been adapted to human studies by Wallace (87,
88). By use of a pressure cuff on the arm, pressure
is first developed to occlude venous drainage and
produce venous congestion, and then further elevated
to stop all blood flow to the arm. Venous pressures
measured between 2 and 8 min after obstruction to
blood flow show a slow decline, presumably due to
capillary transudation. Yenomotor stimulation pro-
duces pressure changes superimposed on this slow
decline. Although this method has yielded clear quali-
tative evidence of venomotor reactions (64), and
possibly has some merits of simplicity, it introduces
a number of complicating features, such as prolonged
ischemia, which would vitiate precise quantitation
and therefore place it in an unfavorable position as
compared with other methods that are available.
Pulse Methods
Pulsatile changes in the volume contained within
a vascular segment will produce pulsatile pressure
changes, the magnitude and rate of transmission of
which are determined by the elastic properties of the
vessel wall. Since muscle tone is one of the factors
influencing elasticity of the venous wall, this suggests
another possible approach to an assessment of ven-
omotor tone. Unfortunately, the pulsations which
occur normally in the venous system are too small
in magnitude and too complex in etiology to be sus-
ceptible to this type of analysis.
Peterson (72, 73) has overcome this limitation by
generating pulses artificially with a high speed in-
jection system. There results a momentary peak of
pressure which increases in magnitude as venous tone
increases. As yet, limited applications of this type of
method have been reported. The author has had
extensive experience with a related phenomenon that
he has referred to as the ""acceleration transient"
which appears at the moment of initiating a constant
speed injection into a vein. It is clear that many de-
tails at the tip of the injection cannula can influence
the pressure peak produced. The exact dimensions
and orientation of the injection orifice in relation to
the vessel lumen are of critical importance in deter-
mining the exact pattern of pressure development,
and this problem can be gravely augmented by the
tendency for some veins to develop a segment of local
constriction in the area of cannulation. Extending
the orifice to a site somewhat remote from the point
of cannulation introduces problems of proper orien-
tation of the injection tip, and also requires pressure
recording through a separate channel in order to
prevent the flow resistance of an elongated injection
cannula from dominating the pressure recording.
Although the potentialities of such a pulse method
PERIPHERAL VENOUS SYSTEM
iogi
should not be overlooked, at present it has not been
developed to the point that it can be applied to the
accurate measurement of venous tone.
Venous Distensibility Patterns
As has been pointed out in reference to figure 5,
constriction of a vein alters its distensibility diagram.
The maximally dilated vein exhibits a smooth curve
convex toward the volume axis. As progressively
more constriction occurs, the distensibility curve is
transformed into a sigmoid form showing initially a
relatively rapid rise in pressure with initial volume
increments, a very much slower pressure rise as inter-
mediate volumes are added, and then a final steep
rise in pressure as still further volume is introduced.
If the rate of venous distension is carefully controlled
so as to prevent stress relaxation effects from dis-
torting the slopes of these curves, evidence of venocon-
striction should be obtainable from studies of the
shape of the distensibility curve.
The first application of this principle was presented
by Capps (16) using a plethysmographic method on
the human forearm. The use of the plethysmographic
technique for venous distensibility measurements
was a logical outgrowth of measurements of blood
flow by the Hewlett & Van Zwaluwenburg method
(50). In the latter method, while the distal portion
of the arm is enclosed in a plethysmograph to record
arm volume, a pressure cuff around a proximal por-
tion is suddenly inflated to a pressure slightly less
than the arterial diastolic pressure. This suddenly
blocks venous outflow from the arm without any
immediate interference with arterial inflow, and
hence the arm will increase in volume at a rate equal
to the rate of blood flow into the arm. After a short
interval, this blood flow will be reduced by the pro-
gressive congestion of the distal vascular bed. The
point is reached eventually where arm volume be-
comes relatively stable, and this must mean that
venous pressure has increased to equal cuff pressure
so that venous blood will be forced past the occluding
cuff at a rate equal to the reduced inflow rate. As
has been pointed out earlier, the most significant
factor in the change in the volume of the arm under
these conditions is the congestion of the venous bed
(17, 39). Therefore, using stepwise increments in
pressure in the occluding cuff yields a series of incre-
ments in arm volume which, as a reasonably good
first approximation, represents the increase in venous
volume at the corresponding occluding pressures.
With this method, Capps obtained clear evidence of
a sigmoid distensibility curve in veins constricted by
cold and other venoconstrictor stimuli, while dilated
veins exhibited a typical convex distensibility pattern.
Many subsequent authors have reported on veno-
motor reactions using the plethysmographic method.
A number of modifications in technique have been
introduced to minimize the artifact associated with
inflation of the pressure cuff, and to permit accurate
pressure reference levels. Errors in the pressure-volume
determinations due to unequal distribution of pres-
sure in the transitional zone at the margin of the
plethysmograph may be corrected by use of a double
plethysmograph. Both compartments are exposed to
equal pressures, but only the distal segment is used
for volume recording (93). Burch (12) has used
another method to correct for the occlusion artifact
in developing the plethysmographic technique for
use on the digit. By measuring the volume change
produced by the venous occlusion cuff during an
interval when all blood flow had been arrested by
arterial compression, he obtains an uncomplicated
record of the artifact alone, which can be subtracted
from the blood flow curves. When these corrected
curves are compared with unoccluded digital pulse
curves, he feels that he can analyze arterial inflow
and venous outflow dynamics with sufficient accuracy
to quantitate the phasic changes that occur during
each pulse wave.
Other plethysmographic devices, such as the mer-
cury-in-rubber resistance strain gauge (89) or the
impedance plethysmograph (75), may be adapted
for venous studies, although the simplicity of their
application should not encourage neglect of estab-
lishing their quantitative reliability.
There is one inherent difficulty in the plethysmo-
graphic method for distensibility determinations,
however, which in the opinion of the author has not
been adequately resolved. The sudden increment in
venous outflow pressure will have its most direct
effect in elevating venous pressure; to a lesser extent
it will elevate capillary pressure, and to a slight ex-
tent it will elevate the distal portions of the arterial
pressure gradient. This justifies the assumption that
most of the immediate volume change will occur on
the venous side, an assumption which has been rea-
sonably well substantiated. On the other hand, the
elevated capillary pressure promotes capillary transu-
dation, so that the recorded volume never reaches a
true plateau, but shows a slow increase persisting
after the initial major increase. Owing to the time-
dependent characteristics of vascular elasticity, more-
over, venous distension occurs rapidly at first and
1092
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
then distends more slowly toward its equilibrium
condition. As a consequence, volume increase due
to the delayed distension of the veins merges indis-
tinguishably with the volume increase due to capillary
leakage. One might attempt to control this problem
by using rigorously standardized intervals of pressure
exposure before reading the volume. Since the magni-
tude of the effect increases as higher pressures are
reached, however (cf fig. 6), virtually all reports of
the use of the plethysmography technique have de-
scribed vague and highly subjective criteria for select-
ing the end point at which to read the volume. There
results a random scatter of the determined points
which often obscures the actual form of the distensi-
bility curve. Therefore most recent authors have
abandoned any attempt to interpret the form of the
curve, and base their interpretations solely on the
total volume increment between two arbitrary pres-
sure levels. We will return to consider the significance
of this type of interpretation after first considering
other techniques which have focused on the form of
the distensibility curve.
The author has developed a method based upon
the change in distensibility pattern of intestinal veins
in the dog. After surgical isolation of the blood flow
through an intestinal loop, the circulation is momen-
tarily interrupted while blood is injected in a retro-
grade direction into the venous bed. By dividing
the volume change required to raise the venous pres-
sure from 10 to 20 cm saline into the volume change
required to raise the pressure from 20 to 30 cm saline,
it is possible to calculate a "venomotor index" which
expresses numerically the degree of sigmoid curvature.
This method has been standardized by use of a motor-
driven syringe and accurate timing of the injections
so that highly reproducible readings can be obtained,
permitting following changes in venomotor tone in
an animal over a period of several hours (5). Also,
if injection rates are adjusted so as to yield equivalent
rates of pressure rise in preparations of different sizes,
reasonably quantitative comparisons can be made
between different animals.
Unfortunately, this method is not without its limi-
tations. There remain some unanswered questions
as to the exact nature of such a retrograde injection.
The method was originally designed on the assump-
tion that there were no valves in this bed. Since India
ink injections invaded the minute vessels of the viscus,
and yet there was no suggestion of any rise in pressure
on the arterial side of the loop, there was an empiri-
cal basis for accepting with some confidence that
the distensibility measurements were representative
of the venous bed. Recognition that these veins are
in fact well supplied with valves of rather poor com-
petence (fig. 1) raises the question as to whether the
valves might be making some contribution to the
pattern of the distensibility curves. To admit this
possibility, however, would not alter the actual inter-
pretation of the data. The retrograde pressure re-
quired to force blood past a valve must be a function
of the muscular tone in the wall of the vessel. A dilated
vein should develop valvular incompetence quite
readily, while a constricted vein should be able to
withstand somewhat higher back pressures before
developing incompetence. Such valve action might
act to exaggerate the degree of sigmoid curvature
in the distensibility of a constricted venous bed, but
this would reinforce rather than detract from the
acceptance of the degree of sigmoid pattern as an
index to the degree of venoconstriction.
A more serious difficulty relates to the moderately
extensive surgical preparation required in the tech-
nique for setting up the loop, which taxes the com-
pensatory ability of the animal. Any further major
manipulation, such as an open-chest procedure, often
leads to deterioration of its circulatory status. This
method is also restricted to the analysis of pressure-
volume increments; a satisfactory quantitation of the
total venous volume has not yet proved feasible. This
negates some of the potential value of obtaining a
quantitative index, since it cannot be related in mean-
ingful terms to total circulatory function. Although
this is a defect which this technique shares in common
with most other methods for assessing the functional
activity of the peripheral venous bed, it is important
not to lose sight of the ultimate goal of being able to
evaluate quantitatively the contribution of the pe-
ripheral venous bed to over-all circulatory dynamics.
A further extension of this general method has
recently been introduced by Bartelstone (7), wfho
has studied the pattern of pressure development be-
hind a sudden obstruction of the vena cava. His re-
sults also afford evidence of the sigmoid distensibility
pattern which develops with venous constriction,
although in this method the quantitative interpreta-
tion is complicated by simultaneous changes in flow.
Distensibility by Volume Increment
As implied above, the plethysmographically re-
corded distensibility of the venous bed has been sub-
jected to two types of interpretation. Capps stressed
the pattern of the data, with the constricted veins
showing relatively less distensibility at low pressures
PERIPHERAL VENOUS SYSTEM
I093
and greater distensibility at intermediate pressures.
A majority of authors, however, have stressed the
increment in volume between two arbitrary pressures,
and interpreted a decrease in the volume change be-
tween these two pressures as an indication of venous
constriction.
To argue that a decreased volume increment is an
indication of venoconstriction might appear to repre-
sent a conflict with our previous discussion of dis-
tensibility patterns. The great distensibility of the
constricted vein over the intermediate range of the
sigmoid curve would seem to demand a great total
distensibility in the constricted vein. As can be ap-
preciated by reference back to figure 5, however, a
comparison of the volume increments in constricted
and dilated veins between any two arbitrary pressures
will yield different relationships at different pressure
levels. In the experiment shown in figure 7, for ex-
ample, the injection happens to have been stopped
at the point that the pressure-volume relationships
in the constricted and dilated veins were virtually
identical; a comparison of the total pressure-volume
increment would offer no suggestion of the significant
differences in distensibility above and below this
particular point.
The author is not aware of any data, obtained with
sufficient precision to identify the pattern of the dis-
tensibility curve, in which apparent discrepancies
in the interpretation of decreased venous distensi-
bilities cannot be resolved by reference to the pressure
level. To illustrate this point, the data in figure 1 1
were taken from a report of a reasonably well-stand-
ardized application of the plethysmographic method
and described by the authors as a "'typical" response.
fig. 1 1 . Volume increments in the human forearm recorded
by congesting the veins to successively higher pressure. Open
circles indicate control values; solid circles are the volumes
determined while infusing noradrenaline at the rate of 0.4
jug/min. [Redrawn from Glover el al. (34).]
The authors interpreted these data as demonstrating
that noradrenaline acts to decrease venous disten-
sibility. One cannot argue with such an interpretation
as a correct description of the data as far as they go.
Nonetheless, this interpretation ignores the different
form of the curves. Although the control data ex-
hibit a relationship convex to the volume axis, the
data recorded during noradrenaline infusion are
clearly concave to the volume axis. These data are
therefore completely compatible with the pattern
interpretation used by Capps and generalized in
figure 5. An extrapolation of the curves in figure 1 1
to higher pressures would clearly lead to a relative
increase in the total distensibility of the constricted
vessels and a relative decrease in the total disten-
sibility of the dilated vessels. Unfortunately, since
the plethysmographic method cannot be used effec-
tively when venous congesting pressures approach
arterial pressures, this technique does not appear
suitable for bringing out the full sigmoid distensibility
pattern in these veins. It should be recognized that
veins in the extremities are subjected to much greater
hydrostatic loads than are visceral veins, and there-
fore it would not be at all surprising to discover that
much higher pressures were required to achieve the
full sigmoid pattern in arm veins.
It is of interest to note that a similar argument has
arisen in reference to isolated vein preparations.
Leonard & Sarnoff (59) state unequivocally that a
constrictor drug always reduces venous distensibility.
Inspection of the Leonard and Sarnoff report,
however, reveals a definite alteration of the pattern
of their distensibility data which is quite compatible
with the suggestion that sufficient stretch will even-
tually reveal a very significant distensibility of the
constricted vein.
Nevertheless, there is another factor in these disten-
sibility characteristics which must be considered.
The original interpretation of the sigmoid curve
visualized that at high pressures the constricted
vessel was pulled out to the same dimension as the
dilated vessel, and thus demands that the total
distensibility must be greater in the constricted
vessel. As has been discussed earlier, the validity of
this interpretation is open to some question. It is
conceivable that a sigmoid distensibility pattern may
be compatible with some reduction in total disten-
sibility. If, in addition, dilated veins were in a state
of partial collapse at the point that the initial volume
was measured, this would augment the possibility of
observing a greater total distensibility- in the dilated
vein.
io94
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
One is therefore justified to interpret distensibility
data in whatever fashion appears compatible with the
responses observed with his particular method. Since
there is overwhelming evidence that adrenergic
drugs are potent venoconstrictors, a reproducible
response to such drugs may be used as a reference
standard to which responses evoked by other stimuli
may be related. It would nonetheless be desirable
to adopt accurately standardized methods for dis-
tending veins so that data will have greater validity.
In view of the significant stress relaxation and creep
phenomena demonstrated by veins, one can scarcely
hope to obtain any precise information by making
measurements "as soon as the volume seems rea-
sonably stable."
SUMMARY' OF VENOMOTOR RESPONSES
From the great variety of techniques that have been
employed for the assessment of venomotor activity, a
wealth of information has been obtained which
demonstrates at least qualitativelv the types of
stimuli which evoke venomotor responses. In table i
are listed responses which stand without controversy
as representative of the active responses of the
venous system. It should be understood that this
listing does not pretend to be comprehensive or
cover any significant fraction of the full literature on
the subject; in general it has proven convenient to
confine the citations in this table to reports which
have been referred to for other purposes in our pre-
vious discussion. This evidence of a broad spectrum
of reactivity suggests that the venous system must
play an important function in active regulation of
the circulation as a whole. For a discussion of this
important aspect of venomotor action, the reader
is referred to Chapter 32. The remainder of our
comments will be confined to response characteristics
which appear to be of some unique importance to
the venous system.
Most investigators of the venous system have been
impressed with the fact, first emphasized by Goll-
witzer-Meier (36), that the venous system appears
to act synergistically with the arterial system. This
is emphasized in the evidence presented in table 1 ;
with the exception of histamine, all responses listed
are direct counterparts of similar reactions known to
occur on the arterial side of the circulation. Beyond
this qualitative similarity, the interesting question
arises as to the relative sensitivity of arteriomotor
1 'enoconstrictor Responses
References
Adrenergic drugs
'2, 4, '4. >9. 25. 34. 4°. 43.
59. 61, 63,82, 88, 91, 93
5 -Hydroxy try ptamine
34. 43. 45
Histamine
25, 61
Carotid sinus hypotension
4, 5, 68, 81
Hypercapnia (central)
4. *'. 35. 54
Hypoxia
4
Cold
16, 20, 39, 42, 56, 68, 88, 93
Deep inspiration
14, 21 , 64
Intense sensory stimulation
4. 8> 54
Psychic stimulation
14, 21 , 68
Exercise
41 , 64, 68
Venodilator Responses
Nitrites
61 , 91, 92
Acetyl choline
*5
Adenylic acid
3
Carotid sinus hypertension
4, 68, 81
Neurogenic syncope
'4
Sleep
'4
and venomotor systems. The literature contains
definite suggestions that the venous system may have
a greater sensitivity (55, 92) and also make a greater
contribution to the total circulatory response (53),
although more information is needed to permit
sound generalizations.
There are a few instances, however, in which the
responses of the venous system appear to be unique.
One is the influence of 5-hydroxytryptamine pre-
viously discussed in reference to figure g. Haddy's
data would indicate that this compound is more
effective in producing venoconstriction than it is in
blocking pre-existing venous constriction resulting
from high sympathetic tone, while for the small
vessels on the arterial side of the circulation, the
svmpathetic blockade can dominate the direct con-
strictor action of the compound. Therefore, 5-
hydroxytryptamine shares with histamine the capacity
to produce both venoconstriction and arteriolar
dilation; as a consequence both of these compounds
have the capacity to induce edema formation. There
is an indication that local tissue acidity may also
have such an action (27, 32). An inverse type of
dissociation between arterial and venous effects has
been reported to occur in circulatory shock, where
venoconstrictor mechanisms fail at a point at which
arterial constrictor tone is still well maintained (5).
A particularly interesting dissociation also appears
to exist in reference to temperature effects. While cold
produces significant constriction of cutaneous
arterioles, it is even more effective as a venoconstric-
PERIPHERAL VENOUS SYSTEM
io95
tor. This cold venoconstriction has a significant
reflex basis as well as a local component (18). As a
consequence, capillary pressure should rise, account-
ing for the increased transudation of fluid and the
tendency for the hematocrit to rise in cold. Second-
arily, the fall in temperature so interferes with dissocia-
tion of oxyhemoglobin as to produce some tissue
anoxia. The anoxic metabolites tend to act as vaso-
dilators competing with the cold constriction of the
arterioles, but seem to have no capacity to cause
venodilation. As a consequence, the skin becomes
plethoric because of some arterial inflow in the face
of a continued resistance to venous outflow. Rewarm-
ing of the tissue will now cause arteriolar dilation.
There is considerable evidence, however, that the
veins do not dilate in response to heat (39, 42, 56,
93). Indeed, the extreme venoconstriction produced
by the previous cold stimulus seems to dissipate
slowly after the tissue is rewarmed, so that during
the initial phase there is significant arteriolar dilation
in spite of persisting venous constriction (42). This
accounts for the extreme degree of plethora and
tendency toward edema formation which is observed
in the rewarming phase.
Finally, attention should be called to fragmentary
information relating to the possibility of venovenous
reflexes. Apart from the general homeostatic regula-
tors of the circulation, such as the arterial presso-
receptor mechanism and the chemoreceptor responses,
are there mechanisms for adjusting venous capacity
as a function of the venous pressure? In view of the
importance of the central venous pressure in deter-
mining cardiac output, circulatory homeostasis
would be enhanced if there were such a mechanism
for adjusting venous capacity to central venous
pressure, so that the venous reservoir tended to
expand in response to an increase in venous pressure
and contract in response to a decrease in venous
pressure.
There are several suggestions that such a mecha-
nism exists. With acute hypotension produced by
vagal bradycardia, Fleisch (26) observed an initial
venous constriction, attributable to the carotid
sinus reflex, followed by a secondary venous dilation
which he felt was associated with the venous conges-
tion produced by the bradycardia. A very similar
observation was reported by Schretzenmayr (82) in
the response to adrenergic drugs. Although these
drugs usually produced a conspicuous venocon-
striction, in instances where the circulatory load
became so great as to momentarily embarrass the
heart and produce cardiac distension, there appeared
to be some mechanism that was counteracting the
venous constriction. More direct evidence of this
mechanism has been provided by the author (6),
who demonstrated venodilation to be produced
reflexly when venous congestion was produced by
inflating a balloon in the thoracic vena cava. It is
important to note, however, that to demonstrate
this effect it proved quite essential to prevent changes
in pressure on the arterial side of the circulation.
Unless such precautions are taken, the arterial
pressoreceptor system dominates the circulatory
responses, and the venovenous reflex mechanism
described above is completely overwhelmed.
This dominance of the arterial pressoreceptor
reflexes accounts for a number of observations which
otherwise would argue against the venovenous
reflex. Wood (93) has demonstrated reflex venous
constriction in man associated with venous con-
gestion produced by occluding cuffs on the extremi-
ties. Inasmuch as they observed a reflex tachycardia
as well as arterial constriction associated with what
was estimated to be a 15 per cent reduction in cir-
culating blood volume, this response would relate
primarily to arterial pressoreceptor mechanisms.
In confirmation of this, when the subjects were in a
supine position so as to minimize pooling of blood
with venous tourniquets, neither the venous reflex
nor other signs of compensation to arterial hypoten-
sion were observed. Similarly, Page and co-workers
(68) found venoconstriction to be produced by the
Valsalva maneuver, which would also be explainable
in terms of baroreceptors on the arterial side dominat-
ing the simultaneous effects of congestion on the
venous side.
There remains the problem of interpreting the
evidence presented by Burch (13) that patients in
congestive heart failure are characterized by an
augmented venomotor tone. If one assumes that the
only significant feature in this condition is venous
distention, this finding would not be compatible
with the postulation of a reflex venodilation in
response to venous distension. It may be of signifi-
cance that Burch found definite evidence of veno-
constriction only in those patients who were rather
severely decompensated and in whom numerous
other sources of reflex stimulation might therefore
have been operable. Our inability to define more
clearly the exact nature of the venomotor reactions
occurring in such an important clinical problem
should afford adequate stimulus to seek still further
clarification of the nature of venomotor control.
iog6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
REFERENCES
i. Alexander, R. S., W. S. Edwards, and J. L, Ankeney.
The distensibility characteristics of the portal vascular
bed. Circulation Research I: 271, 1953.
a. Alexander, R. S. The influence of constrictor drugs on the
distensibility of the splanchnic venous system, analyzed on
the basis of an aortic model- Circulation Research 2: 140,
IQ54-
3. Alexander, R. S. The source of delayed compliance in
the vascular bed. Circulation Research 2: 183, 1954.
4. Alexander, R. S. The participation of the venomotor
system in pressor reflexes. Circulation Research 2: 405, 1954.
5. Alexander, R. S. Venomotor tone in hemorrhage and
shock. Circulation Research 3. 181 , 1955.
fi. Alexander, R. S. Reflex alterations in venomotor tone
produced by venous congestion. Circulation Research 6: 49,
[956-
7. Bartelstone, H. J. Role of the veins in venous return.
Circulation Research 8: 1059, i960.
8. Bayliss, W. M., and E. H. Starling. Observations on
venous pressures and their relationship to capillary pres-
sures. J. Physiol. 16: 159, 1894.
9. Beaconsfield, P. Veins after sympathectomy. Surgery 36:
-71. 1954-
10. Brecher, G. A. Mechanism of venous flow under different
degrees of aspiration. Am. J. Physiol. 169: 423, 1952.
11. Brecher, G. A. Venous Return. New York: Grune &
Stratton, 1956.
12. Burch, G. E. A method for recording and a study of the
venous occlusive technique for measuring the time course
of the rate of inflow and the time course of the rate of
outflow in the finger tip of man during a single pulse cycle.
In : Peripheral Circulation in Man. Boston : Little, Brown, 1954.
13. Burch, G. E. A method for measuring venous tone in
digital veins of intact man. Evidence for increased digital
venous tone in congestive heart failure. A.M. A. Arch.
Internal Med. 94: 724, 1954.
14. Burch, G. E., and M. Murtadha. A study of the veno-
motor tone in a short intact venous segment of the forearm
of man. Am. Heart J. 51 : 807, 1956.
15. Burton, A. C. Relation of structure to function of the
tissues of the wall of blood vessels. Physiol. Revs. 34: 619,
■954-
16. Capps, R. B. A method for measuring tone and reflex
constriction of the capillaries, venules and veins of the
human hand with the results in normal and diseased states.
J. Clin. Invest. 15: 229, 1936.
17. Clark, J. H. The elasticity of s'eins. Am. J. Physiol. 105:
4'8> '933-
18. Connolly, D. G, and E. H. Wood. Distensibility of
peripheral veins in man determined by a miniature balloon
technique. J. Appl. Physiol. 7: 239, 1954.
ig. Donegan, J. F. The physiology of the veins. ./. Physiol. 55:
226, 1 92 1 .
20. Doupe, J., R. A. Krynauw, and S. R. Snodgrass. Some
factors influencing venous pressure in man. J. Physiol. 92:
383> !938-
21. Duggan, J. J, V. L. Love, and R. H. Lyons. A study of
reflex venomotor reactions in man. Circulation 7: 869, 1953.
22. Duomarco, J., P. Recarte, and R. Rimini. Influencia de
las presiones abdominal y toracica sobre el retorno venoso
en la cava inferior. Rev. arg. Cardiol. 1 1 : 286, 1 944.
23. Duomarco, J., R. Rimini, and F. N. Predari. Sobre el
estado de distension o colapso de las venas cavas. Rev. arg.
Cardiol. 12:333, '946-
24. Duomarco, J. I.., and R. Rimini. Energy and hydraulic
gradients along systemic veins. Am. J. Physiol. 178: 215,
'954-
25. Fleisch, A. Die wirkung von Histamin Acetylcholine und
Adrenalin auf die Venen. Pfliigers Arch. ges. Physiol. 228:
35'. '93'-
26 Fleisch, A. Venomotorzentrum und VenenreHexe. II
Mitteilung. Blutdruckzugler und Venereflexe. Pfliigers Arch.
ges. Physiol. 226: 393, 1 931 .
27. Fleishman, M., J. Scott, and F. J. Haddy. Effect of pH
change upon systemic large and small vessel resistance.
Circulation Research 5: 602, 1957.
28. Franklin, K. J. The pharmacology of the isolated vein
ring. J. Pharmacol. Exptl. Therap. 26: 215, 1925.
29. Franklin, K. J. Valves in veins: an historical survey. Proc.
Roy. So, Med 21 : 1 , 1927.
30. Franklin, K. J. The physiology and pharmacology of
veins. Physiol. Revs. 8: 346, 1928.
31. Franklin, K. J., and A. D. McLachlin. Further studies
upon reactions of the abdominal vena cava. J. Physiol. 87 :
87, 1936.
32. Franklin, K. ). .1 Monograph on Veins. Springfield, 111.:
Thomas, 1 937.
33. Franklin, K. J., and A. D. McLachlin. Dilation of veins
in response to tapping in man and in certain other animals.
J. Physiol. 88: 257, 1937.
34 Glover, W. E., A. D. M. Greenfield, B. S. L. Kidd,
and R. F. Whelan. The reactions of the capacity vessels
of the human hand and forearm to vasoactive substances
infused intra-arterially. J. Physiol. 140: 113, 1958.
35. Gollwitzer-Meier, K., and H. Bohn. Liber die venocon-
strictorische VVirking der Kohlensaure unde ihre Bedeutung
fur den Kreislauf. Klin. Wochschr. 9: 872, 1930.
36. Gollwitzer-Meier, K. Venesystem und Kreislaufregulie-
rung. Ergeb. Physiol. 34: 11 45, 1 93^-
37. Green, H. D. Circulation: physical principles. In: Medical
Physics, edited by O. Glasser. Chicago: Yr. Bk. Pub.,
1944, sol. 1, p. 208.
38. Green, H. D. (editor). Transactions of the Third Conference
on Shock and Circulatory Homeostasis. New York: Josiah
Macy, Jr., Foundation, 1953.
39. Greenfield, A. D. M., and C. G. Paterson. On the
capacity and distensibility of the blood vessels of the
human forearm. J. Physiol. 131 : 290, 1956.
40. Gunn, J. A., and F. B. Chavasse. The action of adrenin
on veins. Proc. Roy. Soc, London B 86: 192, 1913.
41. Haddy, F. J., A. G. Richards, J. L. Alden, and M. B.
Visscher. Small vein and artery pressures in normal and
edematous extremities of dogs under local and general
anesthesia. Am. J. Physiol. 176: 355, 1954.
42. Haddy, F. J., M. Fleishman, and J. B. Scott. Effect of
change in air temperature upon systemic small and large
vessel resistance. Circulation Research 5: 58, 1957.
4,3. Haddy, F. J., K. Fleishman, and D. A. Emanuel. Effect of
PERIPHERAL VENOUS SYSTEM
'097
51
52
epinephrine, norepinephrine, and serotonin upon systemic
small and large vessel resistance. Circulation Research 5: 247,
'957-
44. Haddy, F. J. Vasomotion in systemic arteries, small
vessels, and veins determined by direct resistance measure-
ments. Minn. Med. 41 : 162, 1958.
45. Haddy, F. J., P. Gordon, and D. A. Emanuel. The
influence of tone upon responses of small and large vessels
to serotonin. Circulation Research 7: 123, 1959.
46. Ham, A. W. Textbook of Histology (3rd ed.). Philadelphia:
Lippincott, 1957, p. 496.
47. Henderson, V. E., and M. H. Roepke. On the mechanism
of erection. Am. J. Physiol. 106: 441, 1933.
48. Henderson, Y. The veno-pressor mechanism. Am. J.
Physiol. 42: 589, 1917.
49. Henderson, Y. Tonus and the venopressor mechanism:
the clinical physiology of a major mode of death. Medicine
22: 223, 1943.
50. Hewlett, A. W., and J. G. Van Zwaluwenburg. The
rate of blood flow in the arm. Heart 1 : 87, 1909.
Holt, J. P. The collapse factor in the measurement of
venous pressure. Am. J. Physiol. 134: 292, 1941.
Holt, J. P. The effect of positive and negative intra-
thoracic pressure on peripheral venous pressure in man.
Am. J. Physiol. 139: 208, 1943.
53. Holt, J. P., W. J. Rashkind, R. Bernstein, and J. C.
Greisen. The regulation of arterial blood pressure. Am. J.
Physiol. 146:410, 1946.
54. Hooker, D. R. The veno-pressor mechanism. Am. J.
Physiol. 46: 591, 1918.
55. Kelly, W. D., and M. B. Visscher. Effect of sympathetic
nerve stimulation on cutaneous small vein and small
artery pressures, blood flow, and hindpaw volume in the
dog. Am. J. Physiol. 185: 453, 1956.
56. Kidd, B. S. L., and S. M. Lyons. The distensibility of the
blood vessels of the human calf determined by graded
venous congestion. ./. Physiol. 140: 122, 1958.
57. Knisely, W. H., M. S. Mahaley, and H. J. Harriman.
Approximation of "total vascular space" and its distri-
bution in three sizes of blood vessels in rats by plastic casts.
Circulation Research 6 : 20, 1 958.
58. Landis, E. M., and J. C. Hortenstine. Functional sig-
nificance of venous blood pressure. Physiol. Revs. 30 : 1 ,
>95°-
59. Leonard, E., and S.J. Sarnoff. Effect of aramine-induced
smooth muscle contraction on length-tension diagrams of
venous strips. Circulation Research 5: i6g, 1 957-
60. MacWilliam, J. A. On the properties of the arterial and
venous walls. Proc. Roy. Soc, London B 70: 109, 1902.
6t. Maloff, G. Pharmakologische versuche an isolierten Ve-
nen des Menschen. Pfliigers Arch, ges. Physiol. 229: 38, 1932.
62. Maynard, E. A., R. L. Schultz, and D. C. Pease.
Electron microscopy of the vascular bed of rat cerebral
cortex. Am. J Anat. 100: 409, 1 957.
63. Mellander, S. Comparative studies on the adrenergic
neuro-hormonal control of resistance and capacitance
blood vessels in the cat. Ada. Physiol. Scand. 50: Suppl. 176,
i960.
64. Merritt, F. L., and A. M. Weissler. Reflex venomotor
alterations during exercise and hyperventilation. Am.
Heart J. 58: 382, 1959.
65. Milnor, W. R., and C. A. Bertrand. Estimation of
venous blood volume in the dog by the indicator-dilution
method. Circulation Research 6: 55, 1958.
66. Ochsner, A., and M. DeBakey. Thrombophlebitis: the
role of vasospasm in the production of the clinical mani-
festations. J. Am. Med. Assoc. 114: 117, 1940.
67. O'Neill, J. F. The effects on venous endothelium of
alterations in blood flow through the vessels in vein walls,
and the possible relation to thrombosis. Ann. Surgery 126:
270, 1947.
68. Page, E. B., J. B. Hickam, H. O. Sieker, H. D. McIntosh,
and M. D. Pryor. Reflex venomotor activity in normal
persons and in patients with postural hypotension. Cirt il-
lation 1 1 : 262, 1955.
69. Pappenheimer, J. R., and J. P. Maes. A quantitative
measure of the vasomotor tone in the hindlimb muscles of
the dog. Am. J. Physiol. 137: 187, 1942.
70. Pappenheimer, J. R., and A. Soto-Rivera. Effective
osmotic pressure of the plasma proteins and other quantities
associated with the capillary circulation in the hindlimbs
of cats and dogs. Am. J. Physiol. 152: 471, 1948.
7 1 . Paton, YV. D. M. The paralysis of autonomic ganglia with
special reference to the therapeutic effects of ganglion
blocking agents. Brit. Med. .1 . 1 : 773, 1951.
72. Peterson, L. H. Participation of the veins in active regu-
lation of the circulation. Federation Proc. 10: 104, 1951.
73. Peterson, L. H. Certain aspects of reflex and mechanical
influences upon venous circulation. Federation Proc. 1 1 : 122,
'952-
74. Pollack, A. A., and E. H. Wood. Venous pressure in the
saphenous vein at the ankle in man during exercise and
changes in posture. J. Appl. Physiol. 1 : 649, 1 949.
75. Powers, S. R., C. Schaffer, A. Boba, and Y. Nakamura.
Physical and biological factors in impedance plethysmog-
raphy. Surgery 44: 53, 1958.
76. Ramsey, E. Nutrition of the blood vessel wall: review of
the literature. Yale J. Biol, and Med. 9:14, 1936.
77. Remington, J. W., and R. S. Alexander. Relation of
tissue extensibility to smooth muscle tone. Am. J. Physiol.
185: 302, 1956.
78. Remington, J. YV. (editor). Tissue Elasticity. YVashington :
Am. Physiol. Soc, 1957.
79. Roy, C. S. The elastic properties of the arterial wall. J.
Physiol. 3 : 1 25, 1881 .
80. Ryder, H. YV., W. E. Molle, and E. B. Ferris. The
influence of the collapsibility of veins on venous pressure,
including a new procedure for measuring tissue pressure.
J. Clin. Invest. 23: 333, 1944.
81. Salzman, E. YV. Reflex peripheral venoconstriction induced
by carotid occlusion. Circulation Research 5: 149, 1957.
82. Schretzenmayr, A. Die Motorik des intakten Venen-
systems. II. Nachweis und Bedeutung der Hohlvene-
motorik. Arch. Expll. Pathol. Pharmakol. 180: 295, 1936.
83. Smith, D. J. Constriction of isolated arteries and their
vasa vasorum produced by low temperatures. Am. J.
Physiol. 171:528, 1952.
84. Smith, D. J. Immediate sensitization of isolated swine
arteries and their vasa vasorum to epinephrine, acetyl-
choline, and histamine by thyroxine. Am. J. Physiol, \-j~j:
7, '954-
iog8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
85. Spalteholz, YV. Die Vcrtheilung der Blutgefasse in der
Haut. Arch. Anal. u. Physiol. Anat. 1 : 1893.
86. Thompson, W. H. Uber die Abhangigkeit der Gliedervenen
von motorischen Nerven. Arch. Anal. u. Physiol. 8: 102,
87. Wallace, J. M. Pressure relationships among arteries and
large and small veins. Circulation 14: 1013, 1956.
88. Wallace, J. M., and E. A. Stead. Spontaneous pressure
elevations in small veins and effects of norepinephrine and
cold. Circulation Research 5: 650, 1 957-
89. Whitney, R. J. The measurement of volume changes in
human limbs. J. Physiol. 121: I, 1953.
90. Wiggers, C. J. Peripheral circulation. Ann. Rev. Physiol. 9:
255. '947-
91. Wilkins, R. YV., F. W. Haynes, and S. YVeiss. The role of
the venous system in circulatory collapse induced by
sodium nitrite. J. Clin. Invest. 16: 85, 1937.
92. Wilkins, R. YV., S. YY'eiss, and F. W. Haynes. The effect
of circulatory collapse induced by sodium nitrite. J. Clin.
Invest. 17:41, 1938.
93. Wood, J. E., and J. YV. Eckstein. A tandem forearm
plethysmograph for study of acute responses of the pe-
ripheral veins of man: the effect of environmental and local
temperature change, and the effect of pooling blood in the
extremities. J. Clin. Invest. 37: 41, 1958.
CHAPTER 32
Venous return
ARTHUR C. GUYTON
Department of Physiology and Biophysics, University of Mississippi
School of Medicine, Jackson, Mississippi
CHAPTER CONTENTS
Introduction
Principles of Circuit Analysis as They Apply to Venous
Return
The problem
Solution to the problem
The Classical Analysis of Venous Return — Vis a Tergo
and Vis a Fronte
History of More Complete Circulatory Analyses
Simplified Graphical Analysis of Venous Return, Cardiac
Output, and Right Atrial Pressure
Cardiac Output Curves
Effectiveness of the heart as a pump
Alteration of the load against which the heart must
pump
Alterations in pressure on the outside of the heart
Summary of factors that affect the cardiac output
curves
Venous Return Curves
Venous return curves as complements to cardiac output
curves
Method for recording venous return curves
Effect of right atrial pressure on venous return — the
normal venous return curve
Effect of peripheral resistance on venous return
Effect of mean systemic pressure on venous return
Summary of factors that affect the venous return
curves
Equating the Venous Return and Cardiac Output Curves
Effect of sympathetic stimulation on venous return,
cardiac output, and right atrial pressure
Effect of muscular exercise on venous return, cardiac
output, and right atrial pressure
Effect of rapid transfusion on venous return, cardiac
output, and right atrial pressure
Effect of shock on venous return, cardiac output, and
right atrial pressure
Effect of opening the chest on venous return, cardiac
output, and right atrial pressure
Effect of myocardial damage on venous return, cardiac
output, and right atrial pressure
Analysis of decompensation and compensation in con-
gestive heart failure
Analysis of effects resulting from changes in vascular
resistance
A More Complex Graphical Analysis of Venous Return,
Ventricular Outputs, and Atrial Pressures
Balance of the two ventricular outputs with each other
Effect of acute left heart failure on cardiac output,
venous return, left and right atrial pressures, mean
systemic pressure, and mean pulmonary pressure
Effect of acute right heart failure
Effect of blood volume change
Summary of the complex analysis
Specific Factors that Affect Venous Return
Effect of the venous pump on venous return
Effect of venous collapse on venous return
Effect of central pulsation on venous return
Effect of local tissue activity on venous return — effect
of oxygen usage by the tissues
Venous Pressures
Effect of resistance to flow in the veins
Effect of venous flow on peripheral venous pressures
Effect of hydrostatic pressure on peripheral venous
pressures
Summary
INTRODUCTION
the term "venous return" means very simply the
flow of blood from the veins into the heart, and this,
obviously, can be divided into ''systemic venous re-
turn" and "pulmonary venous return." The function
of the veins and the importance of venous return have
been covered from different points of view in several
important monographs (31, 70) and reviews (72,
82-85, 1 33> 137)- Also, the basic characteristics of the
veins as blood vessels are reviewed by Alexander in
Chapter 31 . Therefore, the purpose of this chapter will
be to discuss especially the regulation of venous return
and secondarily the associated factor venous pressure.
1099
I IOO
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Though the venous return is normally exactly
equal to ventricular output, this may not be true for
short periods of time. However, when the venous
return is greater than the ventricular output, blood
will accumulate in the heart. During the ensuing few
heartbeats a new state of equilibrium will develop,
and venous return will again become equal to the
output. Yet, since there are times when venous return
and cardiac output are not equal, it is justified to use
the term ""venous return" separately from the term
"cardiac output."
Normally, in speaking of cardiac output, one thinks
principally of cardiac activity, whereas in speaking of
venous return, he thinks of all the functions of the
peripheral circulation that have to do with blood
flow into the heart. For this reason many circulatory
physiologists consider cardiac output to be regulated
principally by the heart and venous return to be regu-
lated principally by peripheral factors. By all means,
the reader must be cautioned at the outset against
this viewpoint, because except for instantaneous
periods of time, any factor that affects cardiac output
also affects venous return, and any factor that affects
venous return also affects cardiac output. This princi-
ple can be expressed in another way: The circulatory
system is a circuit, and the total flow of blood through
any one cross section of the circuit is exactly the same
as the total flow through any other cross section.
Principles of Circuit Analysis
800
400-
Mean systemic
pressure =
7 mm Hg
0 44
RIGHT ATRIAL PRESSURE (mm Hg)
fig. 5. The normal venous return curve.
-4
ATRIAL
+ 8
VENOUS RETURN
I IO7
The second method for establishing the normal
venous return curve has been to determine different
points along the curve intermittently by suddenly
elevating the right atrial pressure and making venous
return measurements within the next 5 to 7 sec
before circulatory reflexes can take place. Then the
circulation is returned to normal, and after a reason-
able control period another intermittent measure-
ment is made.
The venous return curves recorded by these two
different procedures have been identical. Further-
more, venous return curves have been recorded in
closed-chest animals in which a special occluding
system has been surgically placed around the pul-
monary artery so that the pulmonary artery could be
occluded to any desired degree (97). Then, using
especially the intermittent procedure, points along
the venous return curve were established. The results
agree with the measurements established when using
the above two procedures.
The venous return curve of figure 5 is the average
curve, recorded in approximately 100 separate dogs
anesthetized with sodium pentobarbital, and then
extrapolated on a weight basis to the 12-kg dog.
Several features of this curve deserve special comment.
First, when the right atrial pressure becomes more
negative than o to — 4 mm Hg, a further increase in
the negativity of the right atrial pressure does not
cause a further increase in venous return. In other
words, the venous return curve reaches a '"plateau."
The cause of this effect is the well-known collapse
factor in veins (88, 112). One can actually see the
veins entering the thoracic cavity begin to collapse
when the right atrial pressure becomes negative with
respect to atmospheric pressure. Furthermore,
measurements in the veins immediatelv bevond the
collapsed points show that these veins all have ap-
proximately o mm Hg pressure in them regardless of
how low the right atrial pressure falls. Thus, the
collapse factor effectively sets the venous pressure of
the blood leaving the systemic circulation almost
exactly at o.
The second important point in relation to the
venous return curve is that elevation of the right
atrial pressure above o causes a very rapid decrease
in return of blood from the systemic circulation
(98). On the average, for each mm Hg rise in pressure
above o, the venous return decreases 14 per cent, and
it reaches zero when the right atrial pressure has
risen to approximately +7 mm Hg in "areflex" dogs.
The third important point is that when venous
return reaches zero, the right atrial pressure at this
level is equal to the mean systemic pressure (98). The
mean systemic pressure is the pressure in the systemic
circulation that is measured if the root of the aorta
and the large systemic veins entering the heart are
suddenly occluded and all pressures in the systemic
circulation are brought instantaneously to equilib-
rium. That is, when blood flow ceases absolutelv
in the systemic circulation, the pressures in all its
segments become equal. Therefore, the right atrial
pressure becomes equal to the pressure everywhere
in the systemic vessels. This equilibrium pressure is
the mean systemic pressure.
The fourth point of major significance in relation
to the venous return curve is the almost complete
linearity of the venous return curve in the range
between o right atrial pressure and the mean systemic
pressure level. That is, the venous return is approxi-
mately proportional to the difference between mean
systemic pressure and right atrial pressure (Pms -
Pra). This difference is called the "pressure gradient
for venous return" (81), and it is an important
concept in establishing the forces that lead to the flow
of blood toward the heart. This will be seen below,
especially in relation to alterations in systemic
resistances, for when there is no pressure gradient for
venous return, there will be no venous return to the
heart regardless of the changes in systemic resistances.
EFFECT OF PERIPHERAL RESISTANCE ON VENOUS RETURN.
Figure 6 illustrates the effect on the venous return
2400
2000-
1600
CE
\-
UJ
CC
w
z>
o
z
UJ
>
1200
800
400-
+4 +8
RIGHT ATRIAL PRESSURE (mm Hg)
fig. 6. Effect on the venous return curve of changing the
peripheral resistance. Note that the mean systemic pressure
remains constant at approximately 7 mm Hg.
no8
HANDBOOK OF PHYSIOLOOV
CIRCULATION II
fig. 7. Effect on the venous return curse of suddenly open-
ing large bilateral femoral A-V fistulae.
curve of changing the systemic resistance from normal
(87). Note that the venous return is exactly zero in
the case of each of these three curves when the right
atrial pressure is equal to the mean systemic pressure.
That is, when there is no pressure gradient for venous
return, there is likewise no flow toward the heart.
Yet, when the right atrial pressure falls to some value
below the mean systemic pressure, then a pressure
gradient does exist for forcing blood toward the
heart, and the return of blood is inversely propor-
tional to the resistance. The greater the resistance,
the less is the return of blood to the heart, and the
less the resistance, the greater is the venous return.
Thus, figure 6 shows the normal venous return curve,
a venous return curve in which the resistances
throughout the systemic circulation are approximately
one-half normal, and a venous return curve in which
the resistances are approximately two times normal.
Figure 7 illustrates a typical experiment in which
peripheral resistance was suddenly changed while
all other conditions of the circulation were kept as
nearly constant as possible (103). In this instance two
large femoral A-V fistulae were suddenly opened so
that the total peripheral resistance was decreased to
approximately 60 per cent of the control value.
Circulatory pressures remained exactly constant.
Note that the study depicts precisely the same effects
as those illustrated in the previous figure but this
time showing a typical and actual experimental
study
It should not be supposed, however, that increasing
the resistance to blood flow in the arteries affects
venous return equally as much as increasing the
resistance in the veins. Indeed, for a given increase
in venous resistance, the venous return decreases
approximately eight times as much as when the
arterial resistance is increased the same amount.
This was illustrated by a comparative study in which
arterial resistance was increased by injecting micro-
spheres into the arterial system and venous resistance
was increased by progressive occlusion of all the large
veins emptying into the right atrium (87). Figure 8
illustrates the difference between these two effects,
the upper curve showing that the total peripheral
resistance could be increased by arterial embolization
to as much as 400 to 500 per cent of control values
before the venous return decreased a great amount.
On the other hand, increasing the total peripheral
resistance only 30 per cent by the method of venous
compression decreased the venous return to one-half
normal.
The cause of this difference between venous re-
sistance and arterial resistance is that the arterial
system proximal to the arterioles has very little
capacitance (DV/DP) in relation to the total capa-
citance of the s\stemic circulation proximal to the
venous constriction at the outflow of the veins into
the heart (89). Because of the small storage ability of
the arteries for blood, increasing the resistance at
the arterioles elevates the arterial pressure almost as
100
° 80-
60-
C40-
S20i
o
^9 °"er,ol resistance
100 200 300 400 500
TOTAL PERIPHERAL RESISTANCE (% of control value)
fig. 8. Effect on venous return of increasing the total
peripheral resistance when the resistance is increased in three
different ways : 1 ) by injecting microspheres into the arteries
to increase arterial resistance, 2) by constricting the inflow-
veins to the heart, and 3) by a combination of these two
procedures. [From Guyton et til. (87).]
VENOUS RETURN
I IOg
much as the resistance rises, and the arterial pressure
then simply forces the blood on past the resistance.
On the other hand, constricting the veins where they
empty into the heart causes the pressure in the veins
to rise only a few mm Hg because of the great storage
capacity of the veins. This small rise in venous
pressure is far too little to overcome the increasing
resistance, and, as a consequence, the venous return
becomes tremendously depressed. Therefore, venous
resistance affects venous return to the heart many
times as much as arteriolar or arterial resistance of
the same magnitude.
EFFECT OF MEAN SYSTEMIC PRESSURE ON VENOUS
return. Basically the mean systemic pressure is the
resultant of the ratio of a) the blood volume to b) the
ability of the circulatory system to hold blood. As
the blood volume increases, the mean systemic
pressure remains essentially zero until the blood
barely begins to distend the blood vessels. But, once
this point has passed, any further increase in blood
volume increases the mean systemic pressure directly
in proportion to the additional amount of blood
injected into the circulation, the mean systemic
pressure rising approximately 1 mm Hg for each 2
per cent increase in blood volume (86). Thus, it can
be seen that very small changes in blood volume can
cause relatively large changes in mean systemic
pressure and, as a consequence, can have a marked
effect on venous return unless other circulatory
compensations prevent this.
Only a few measurements of mean systemic pressure
have ever been made, for this requires instantaneous
stoppage of the circulation and then rapid equilibra-
tion of the pressures in all segments of the systemic
circulation before any blood can leave or before
circulatory reflexes or other factors can change the
vascular distensibility. In our laboratory, we have
measured this pressure a few times by suddenly
constricting the aorta and pulmonary arteries (138),
utilizing devices implanted several weeks previously
in the thoracic cavity to cause the constrictions. These
pressure measurements showed a normal mean
systemic pressure of almost exactly 7 mm Hg.
Far more measurements have been made of the
mean circulatory pressure (101, 105, 180, 182) than
of the mean systemic pressure; approximately 1000
such measurements have been made in our laboratory.
Since the mean systemic pressure is of such extreme
importance in determining venous return to the
heart, the few measurements of mean systemic
pressure have been compared with measurements of
mean circulatory pressure. In all instances, the mean
systemic pressure has been almost identical with the
mean circulatory pressure except in the case of
extreme engorgement of the pulmonary circulation,
and even here the difference has been only 1 mm Hg
or so. Therefore, insofar as venous return from the
systemic circulation is concerned, one can consider
the mean systemic pressure and mean circulatory-
pressure to be almost identical.
Measurements of mean circulatory pressure can be
made very easily by electrically fibrillating the heart
with 60-cycle current applied to needle electrodes in
the anterior chest wall at approximately 50 v. Studies
have demonstrated that all pressures of significance in
the measurement of mean circulatory pressure come
to equilibrium within only a few seconds after cardiac
fibrillation begins, except for the pressures of the sys-
temic arterial chamber and the systemic venous
chamber. Therefore, immediately after fibrillation of
the heart begins, arterial blood is pumped from a
catheter lying in the descending aorta and thence
through another catheter into the inferior vena cava.
After only 3 to 5 sec, the pressures in these two cham-
bers are brought to equilibrium and the mean circu-
latory pressure measured. During the next few seconds
the heart is defibrillated by passing 4 to 10 amperes
of 60-cycle alternating current at 440 v for i/[0 sec
directly through the chest anteroposteriorly (104).
After 2 to 3 min of recovery, the animal returns to
essentially normal circulatory conditions.
Figure 9 illustrates the typical effect on the venous
2400-
2000-
\
1600-
"
Nor
mal
1200-
800-
"2
X
normal ft^ \
400-
^""sf \
0
\ Pms ; \
\ ,35mm^
Pms =
\l, 7 mm
\ Pms ■
\yl4mn
1
— 1
1 1
1
1 1
-12 -8 -4 0 +4 +8 +12 +16
RIGHT ATRIAL PRESSURE (mm Hg)
fig. 9. Effect on the venous return curve caused by changes
mean systemic pressure.
I I 10
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
return curve caused by altering the mean systemic
pressure. Note that when the mean systemic pressure
is increased from the normal value of 7 mm Hg up
to 1 4 mm Hg the curve is shifted to the right and its
plateau becomes approximately twice as high as
normal. Conversely, when the mean systemic pressure
is decreased to 3.5 mm Hg, which is one-half normal,
the curve is shifted to the left, and the plateau becomes
reduced to one-half normal. Since the flow of blood
to the heart is proportional to the pressure gradient
for venous return, which in turn is equal to the mean
systemic pressure minus the right atrial pressure, one
can see that any increase in mean systemic pressure
causes a corresponding increase in venous return at
any given right atrial pressure. Likewise, any decrease
in the mean systemic pressure will cause a cor-
responding decrease in venous return at all right
atrial pressures.
Figure 10 depicts venous return curves in a series
of normal dogs and then again in the same dogs after
an average infusion of 200 ml of blood and also
after hemorrhage of 122 ml (99). Note that the
normal mean systemic pressure was approximately
7.7 mm Hg, that this rose to 1 1 .3 mm Hg in the
infused dog, and that it fell to 4.7 mm Hg in the
hemorrhaged dog. This experiment illustrates typical
shift of the venous return curves to the right as the
blood volume increases, thereby increasing the mean
systemic pressure.
Changes in the distensibility of the vascular system
or changes in the pressure on the outside of the
vessels can alter the mean circulatory pressure in
exactly the same manner as can alterations in blood
volume. These changes include a) increased vaso-
motor tone, caused either by sympathetic stimulation
or by infusion of sympathomimetic drugs; b) pressure
on the abdomen, which compresses large intra-
abdominal blood reservoirs; c) increased intrathoracic
pressure, which compresses the blood reservoirs of
the chest; and d) increased interstitial fluid volume,
which causes pressure on the outside of blood vessels
throughout the body. In normal circulatory ad-
justments the most important of these is the effect of
vasomotor tone on the mean systemic pressure.
Figure 1 1 illustrates the average results from 1 1
typical experiments in dogs in which the degree of
vasomotor tone was altered from the minimal level
up to almost the maximal level (95). This shows the
typical effects one would expect when the mean
systemic pressure is elevated, that is, progressive
shift of the venous return curves to the right as the
vasomotor tone is increased.
One might have expected the administration of a
sympathomimetic drug to cause increased resistance
to blood flow toward the heart as well as to increase
the mean circulatory pressure. This was not evident
from these studies, for the venous return curves did
not decrease in slope as the rate of epinephrine in-
jection was increased. On second thought, one can
understand why this was true. When vasomotor tone
is increased throughout the circulation while the blood
volume remains constant, pressures everywhere in
the circulation will tend to rise because of tightening
of the vessels around the blood. But, if any single
segment of the circulation constricts, some other
segment of the circulation must dilate. On the
average, then, for every constriction that occurs in
the systemic circulation following the injection of
fig. 10. Effect of increasing or decreasing the
blood volume on the venous return curves. Total
spinal anesthesia was instituted to prevent cardio-
vascular reflexes during the course of the experi-
ment. [From Guyton et a/, (qq).1
Z>
h-
LU
o
z
UJ
>
2400
2000
1600
1200
800
400 H
0
MEAN ART PRESS =184
TOTAL SPINAL ANESTHESIA
EPINEPHRINE
ELEVEN DOGS
MEAN WEIGHT = 14 kg
0.0035 mg /kg /minute
MEAN ART PRESS. = 120
0.0015 mg /kg /minute
MEAN ART PRESS. = 62
0.0005 mg /kg /minute
MEAN ART PRESS = 4
NO EPINEPHRINE
fig. II. Effect of different rates of epinephrine
injection on the venous return curves. Note that
the principal effect is to increase the mean systemic
pressure. [From Guyton el al. (95 I
■16 -12 -8 -4 0 + 4 +8 +12
RIGHT ATRIAL PRESSURE (mm Hg)
epinephrine, as depicted in the experiments of figure
1 1 , there had to be equal dilatation somewhere else.
Indeed, measurements have shown that, as the
arterioles constrict under these conditions, there is a
tendency for the veins to dilate even though the walls
of the veins do tighten to a very great extent (55).
This elevates the mean systemic pressure but does not
increase the resistance to blood flow from the systemic
vessels toward the heart. In essence, then, we can say
that an increase in vasomotor tone affects venous
return principally by increasing the mean systemic
pressure, and, usually, an increase in vasomotor
tone does not increase the average resistance that
opposes the return of blood to the heart.
SUMMARY OF FACTORS THAT AFFECT THE VENOUS
return curves. Basically, there are only two different
patterns of changes in venous return curves, those
that result from a) changes in resistance in the systemic
circulation, and h) changes in the mean systemic
pressure. Figure 6 depicts the pattern of venous
return curves that results from alteration of vascular
resistance, while figure 9 illustrates the curves that
result from alteration of mean systemic pressure.
Any factor that alters resistance, whether this be a
localized or generalized alteration, will correspond-
ingly alter the venous return curve. However, altera-
tion of the venous resistance affects venous return far
more drastically than alteration of the arterial re-
sistance.
The factors that affect the mean systemic pressure
can be divided into two main groups: /) those that
affect the blood volume, and 2) those that affect the
ability of the circulatory system to hold blood. The
two most important factors of all that affect mean
systemic pressure are blood volume itself and changes
in vasomotor tone.
The interrelationships of all these different factors
on the return of blood to the heart can be expressed
mathematically by the following formula (98) :
VR-
P
Pro
RvCv+(Rv+Ro)Ca
cv + ca
In this formula VR is venous return, Pms is mean
systemic pressure, Pra is right atrial pressure, C,. is
capacitance of the veins, C„ is capacitance of the
arterial tree, Rv is the average resistance to blood flow
from the veins to the heart, and R„ is the resistance
from the arterial tree to the venous tree. This formula
shows that venous return is approximately pro-
portional to the mean systemic pressure minus right
atrial pressure, which has been called the ''pressure
gradient for venous return," while, on the other hand,
venous return is inversely proportional to the re-
sistances in the systemic circulation. The capacitances
in the formula are constants for any given animal,
and they determine the relative importance of arterial
resistance versus venous resistance. In the normal
animal a given change in venous outflow resistance
affects venous return approximately eight times as
much as the same change in arterial resistance (87).
Equating the Venous Return and Cardiac Output Curves
If one understands the different factors that affect
venous return and cardiac output curves, he can
readily determine the approximate effects of any
given circulatory change on each of these two types
I I 12
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
2400,
f 2O0O-
-4 0 +4 +8 +12
RIGHT ATRIAL PRESSURE (mmHql
fig. i 2. Equating of a normal cardiac output curve and a
normal venous return curve for a 12-kg dog.
of curves. Then by plotting the two curves on the same
coordinates, as shown in figure 12, the equations
represented by the separate curves can be solved
(8i). Figure 12 illustrates the equating of a normal
venous return curve with a normal cardiac output
curve, as depicted for the 10-kg dog. Note that there
is only one single point at which the flows and the
pressures for the two curves are equal, and this point
is called the "equilibrium point." It represents the
solution to our graphical analysis, depicting in figure
12 that this particular dog at this particular time has
a cardiac output of 1200 ml per min, a venous return
also of 1200 ml per min, and a right atrial pressure
of — 2 mm Hg (referred to the level of the tricuspid
valve).
Looking once again at figure 1 2, let us assume that
an extra quantity of 25 ml of blood is suddenly
injected into the right atrium. This would raise the
right atrial pressure to approximately +4 mm Hg,
and, as depicted by the venous return curve, the
elevated right atrial pressure would decrease the
venous return to approximately 500 ml per min. On
the other hand, the high right atrial pressure would
cause the cardiac output, as depicted by the cardiac
output curve, to rise to approximately 2000 ml per
min. Thus, a disparity of 1500 ml per min develops
between venous return and cardiac output so that
far more blood is pumped out of the heart than
returns to it. Therefore, within the next three to six
heartbeats, the right atrial pressure falls back to the
level of —2, thus causing the venous return to rise up
to 1200 ml and the cardiac output to fall to 1200 ml.
In other words, within a few seconds, equilibrium
will be re-established whenever venous return and
cardiac output deviate from each other (35).
EFFECT OF SYMPATHETIC STIMULATION ON VENOUS
RETURN, CARDIAC OUTPUT, AND RIGHT ATRIAL PRES-
SURE. Using the same principles for equating venous
return and cardiac output curves as depicted in
figure 12, we can now show in figure 13 the effect of
strong sympathetic stimulation on venous return,
cardiac output, and right atrial pressure. The dashed
curves of the figure illustrate the normal curves. Then,
suddenly, sympathetic stimulation changes both the
venous return and cardiac output curves to the
respective solid curves (95). Note that the venous
return curve shifts far to the right and upward be-
cause of an increase in "mean systemic pressure,"
and the cardiac output curve shifts upward, as is
characteristic of a hypereffective heart. These two
curves equate at an entirely new point, the new
equilibrium point occurring at a right atrial pressure
of —3 mm Hg and a cardiac output and venous
return of 1800 ml per min.
This analysis of the effects of sympathetic stimula-
tion agrees with the typical experimental result
found when the sympathetics are stimulated through-
out the body, that is, a mild to moderate increase in
venous return and cardiac output and usually a
slight decrease in right atrial pressure (169). Almost
E 2800
-8 -4 0 +4 +8 412
RIGHT ATRIAL PRESSURE (mm Hg)
fig. 13. Effect of sympathetic stimulation on the venous
return and cardiac output curves, showing an increase in
cardiac output and venous return and a decrease in right
atrial pressure.
VENOUS RETURN
III3
exactly the opposite effects occur when vasomotor
tone is greatly reduced throughout the body by the
administration of nitrites (195) or ganglion-blocking
agents ( 1 89). Likewise, pooling of blood in the lower
part of the body when one stands (5, 76, 14Q-144,
175, 184, 196) or sequestration of blood in the limbs
by the application of tourniquets (59, 193) reduces
the venous return and cardiac output in a closely
similar manner.
EFFECT OF MUSCULAR EXERCISE ON VENOUS RETURN,
CARDIAC OUTPUT, AND RIGHT ATRIAL PRESSURE. Figure
14 illustrates an analysis of the circulation during
exercise, showing changes in both the cardiac output
and venous return curves. The cardiac output curve
is elevated as a result of a) sympathetic stimulation of
the heart, and b) inhibition of the vagi to the heart,
thus giving a cardiac output curve of a hypereffective
heart.
Three different factors cause the observed altera-
tions in the venous return curve. First, tightening of
the musculature throughout the body, particularly
4000
-8 "4 0 +4 +8 + 12 +16 +20
RIGHT ATRIAL PRESSURE (mm Hg)
pig. 14. Effect of exercise on the venous return and cardiac
output curves, showing that the new curves equate at greatly
elevated venous returns and cardiac outputs. Also, the right
atrial pressures are still close to zero mm Hg.
tightening of the abdominal musculature, causes an
instantaneous increase in mean systemic pressure of
several millimeters of mercury (100). Second, sympa-
thetic stimulation causes considerable increase in
mean systemic pressure (101). Third, the blood vessels
of the musculature are likely to become markedly
dilated, thus decreasing the resistance to blood flow
through the systemic circulation (12); this, in turn,
increases the slope of the venous return curve (87).
Thus, we find that in moderate exercise the cardiac
output and venous return may be increased to two or
more times normal, and the right atrial pressure will
still be only slightly elevated (14, 169, 174). On the
other hand, in severe exercise, the heart is then often
taxed to its limit so that the right atrial pressure rises
considerably as depicted by the highest equilibrium
point in the figure. In an animal or human being that
has been thoroughly trained for athletics, the cardiac
output curve of the heart can rise to one and one-half
to two times that depicted in figure 13, thus giving as
much as a five- to sevenfold increase in cardiac output
without an elevation of right atrial pressure above
zero.
Here, again, it is quite evident that simultaneous
analysis of the function of the heart and of peripheral
circulatory factors is needed to ascertain the inte-
grated effects of exercise on venous return, cardiac
output, and right atrial pressure (10, 39, 50, 169,
177, 188).
EFFECT OF RAPID TRANSFUSION ON VENOUS RETURN,
CARDIAC OUTPUT, AND RIGHT ATRIAL PRESSURE. Figure
1 5 depicts the effects of rapidly infusing an animal
with whole blood. The immediate effect stems mainly
from an increase in mean systemic pressure (99) — in
this instance from 7 mm Hg to 1 1.5 mm Hg. How-
ever, the increased blood volume also distends the
vessels of the systemic circulation, thus decreasing
the peripheral resistance and therefore increasing
the "slope" of the venous return curve. In the case
of the heart, circulatory reflexes, especially the
pressoreceptor reflex, actually weaken the heart
because the excess blood volume tends to elevate
arterial pressure, thereby initiating inverse reflexes.
As a consequence, the cardiac output curve decreases
very slightly. The net result, as depicted by the
equilibrium point, is a moderate increase in venous
return and cardiac output and a very marked rise in
right atrial pressure, which are effects that have been
observed many times by many different investigators
(66-69, M5> 1 83)- This figure and the previous one
illustrates that the relationship between right atrial
ii4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
O
<
a
o
<
Q
<
U
Q
Z
<
I-
Ui
cc
(/>
o
2000
1600
1200
800
400
"4 0 +4 +8 +12 +I6
RIGHT ATRIAL PRESSURE (mm Hg)
fig. 18. Effect of acute myocardial infarction and subse-
quent stages of recovery on the cardiac output and venous
return curves. The sequence of events is explained in the text.
1 1 16
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and venous return curves up to those depicted by the
dashes and dots. Therefore, within 2 min, the venous
return and cardiac output will have returned to ap-
proximately two-thirds normal, and the right atrial
pressure will have risen another 2 mm Hg up to +6
mm Hg. But, even this is an abnormally low cardiac
output which is still insufficient to supply all the tissues
of the body with adequate amounts of blood. Further-
more, there is intense sympathetic vasoconstriction
throughout the circulatory system during this period
of time as well as sometimes a decreased arterial pres-
sure; the intense sympathetic vasoconstriction (134)
and the decreased arterial pressure (135) both de-
crease renal output. Furthermore, the semishock
state that exists at this stage causes the adrenal glands
to secrete large quantities of aldosterone. This, in
turn, promotes rapid reabsorption of sodium from
the renal tubules, associated also with rapid reabsorp-
tion of water (48). The net effect on the kidneys,
therefore, is to reduce renal output greatly or, at
times, even to stop renal output completely. Over a
period of the next few days, fluid is progressively re-
tained in the circulatory system, thus shifting the
venous return curve in figure 1 8 further and further
to the right (99, 105). The solid venous return curve
depicts approximately that which will obtain after a
week or so of fluid retention.
Simultaneously with the changes that take place
in the venous return curve, the heart will also be
changing. If the infarction is an uncomplicated one
and recovery from the infarction begins immediately,
then the cardiac output curve of figure 18 will pro-
gressively rise. The solid cardiac output curve depicts
approximately that which one would expect after a
week of recovery. As illustrated by the point at which
the solid cardiac output curve equates with the solid
venous return curve, we find that the cardiac output
and venous return will have now returned almost
completely to normal but that the right atrial pressure
again will have risen another 2 mm Hg. This is charac-
teristic of the chronic stage of congestive heart failure,
that is, the cardiac output may be normal or slightly
below normal, but the venous pressures are essentially
always greatly elevated.
Another effect that occurs as the cardiac output
approaches normal is that the degree of sympathetic
activity throughout the body also gradually becomes
reduced toward normal. Furthermore, the body is
no longer in a shocklike state so that the output of
aldosterone also becomes reduced. As a consequence,
renal output once again returns toward normal, thus
preventing further retention of fluid. Therefore, the
circulatorv svstem has now reached a new steady
state, with the cardiac output and venous return
essentially normal, renal output once again essen-
tially normal, but the right atrial pressure consider-
ably elevated.
It should be noted again that the analysis illus-
trated in figure 1 8 is that for myocardial damage
affecting both ventricles approximately equally. The
course of events depicted by the dotted line is typical
of that normally observed following acute generalized
myocardial infarction (13, 71, 73, 75, 106-108, 121,
162, 163, 166). We shall see that the more complicated
graphical analysis presented later in the chapter is
much more satisfactory than the simple graphical
analysis when one side of the heart fails to a greater
extent than the other side.
ANALYSIS OF DECOMPENSATION AND COMPENSATION IN
congestive heart failure. Figure ig illustrates an
analysis of decompensation in severe cardiac failure.
This shows by the two dashed curves the analysis for
the normal circulatory system in a normal 10-kg dog.
Then, at the bottom of the graph, it shows the typical
cardiac output curve for a severely damaged heart
after all sympathetic reflexes and all recovery that
are possible have taken place (38). If we assume that
the cardiac output curve suddenly falls from the
normal down to this depressed curve, then we find
that the cardiac output immediately falls to point Ay
about two-fifths normal, with a right atrial pressure
of approximately +4 mm Hg. This cardiac output
is far too little to cause normal renal function, and,
for the same reasons discussed above, renal output
becomes severely depressed. As a result, fluids are
retained in the body, and the mean systemic pressure
progressively rises, shifting the venous return curves
to the right and progressively elevating their plateaus.
Thus, during the ensuing days, with the progressive
retention of fluid, the equilibrium points in figure 19
proceed to B, C, D, E, F, and G. It is especially in-
teresting that cardiac output curves of severely dam-
aged hearts do not rise to a plateau but, instead, rise
to a peak and then begin to descend (117). Therefore,
after the process of decompensation has proceeded
past the peak at point E, further retention of fluid
causes a reduction in cardiac output rather than an
increase.
The significant factor in decompensation is that
even at its greatest peak, the cardiac output curve
never reaches the necessary cardiac output level re-
quired to re-establish normal renal function. Con-
sequently, fluids continue to be retained indefinitely
until death of the animal.
VENOUS RETURN
I I I 7
fic. 19. Analysis of decompensated heart
disease, showing a greatly depressed cardiac
output curve and a progressive shift of the
venous return curves to the right until death
occurs, as explained in the text.
-4 0 + 4
RIGHT ATRIAL
+ 8 +12 »I6
PRESSURE (mm Hq)
Figure 20 depicts recompensation of the animal
that had been almost dead from decompensated
heart disease. The lower curve illustrates a cardiac
output curve of a decompensated heart, showing that
after a period of time the venous return curve had
already reached the far right curve with equilibrium
occurring at point A. Then, upon instituting appropri-
ate treatment, such as the administration of digitalis
(117), the heart becomes considerably stronger, and
the cardiac output curve rises to the upper curve. If
this rise is relatively rapid, the venous return curve
will not be immediately affected. Therefore, the new
equilibrium point becomes point B, which represents
a cardiac output greater than that required for normal
renal function. As a consequence, the output of urine
now becomes actually far greater than normal, which
is a well-known effect of digitalis when a decompen-
sated state is converted into a compensated state. The
output of urine causes a decrease in mean circulatory
pressure and a progressive shift in the venous return
curves toward the left. Thus, during the ensuing days,
the equilibrium points in figure 20 shift from point
B to C, to D, and finally E. At point E the venous
return curve becomes stable because now the cardiac
output has fallen back to a value that is just sufficient
to maintain a renal output equal to the daily intake
of fluid and salts. One can see that we now have three
different curves equating with each other, the cardiac
output curve, the venous return curve, and a straight
line which is a curve representing the level required
for normal renal function. It is where these three
£ 2800-
1
O
V
2400-
H
2
K
g 2000-
O
<
g 1600-
\ \ XfL.
\ \S^\
\£3
<
O
Level required for
normol renal function
§ I2O0-
j£T^_ -A
I — \
<
&/ \ \
z
£/ \ \
,\a
* 800-
v #t^\K
1-
*/ ^s^ \
E
f *r >
u
or
1-
z
Id
>
2400-
2000
1600-
W 1200
O
z
<
UJ 400-
3
O
800-
T
► 20
w +50 +40 +30 +20 +J0 0 -10
LEFT ATRIAL PRESSURE (mm Hg)
pig. 22. Transposition of the curves illustrated in fig. 21.
This transposition allows the left heart analysis to be correlated
with the right heart analysis in the following figures.
— 2800n
E
—2400-
K
3
0-
1-
3 2000-
«\ 5
tc
<
\ 3/
3 1600-
l\f
K
— \ 1
1-
Z
Equilibrated
output
* \l
> 1200-
O
//
**.
RN AN
00
8
*,
— 1 I
^1
0
V
s/s\
\-
*/
>v
UJ 400-
1
X*
*•/ ""1
cr
■c/
•&1
(0
*/
\S I
i 0-
1
11
-4 -2 0 »2 «-4 +6
RIGHT ATRIAL PRESSURE (mm Hg)
+ 50 +40 .30 +20 +10 0 -10
LEFT ATRIAL PRESSURE (mm 1+))
fig. 23. Simultaneous analysis of left and right heart
function, showing that in the steady state the venous returns
and outputs of the two sides of the heart are all in equilibrium
with each other.
Finally, we superimpose the analysis for the left
heart onto a simultaneous analysis for the right heart,
as shown in figure 23. In this superimposition, we
place the left atrial and right atrial pressure scales
so that the 7 mm Hg level of one coincides with the
7 mm Hg level of the other. The reason for this is
that our preliminary measurements of mean pulmo-
nary pressure show it to be almost identical in the
normal state with the mean systemic pressure, that
is, almost exactly 7 mm Hg. Now, we can explain
the composite analysis of the two sides of the heart.
Note in figure 23 that the right ventricular output
curve and the right ventricular venous return curve
equate at the 1200 ml per min level. Likewise, the left
ventricular output curve and the left ventricular
venous return curve also equate at this same level.
Therefore, under normal circumstances the two
venous returns and the two ventricular outputs are
all equal to each other, and the circulation is in a
steady state, without any momentary transference
of blood from one of the circulatory segments to
another.
BALANCE OF THE TWO VENTRICULAR OUTPUTS WITH
each other. Proceeding to figure 24, we see the
normal situation again depicted by the solid curves.
However, the dotted venous return curves represent
a situation in which excess blood has momentarily
been transferred from the pulmonary circulation to
the systemic circulation. Note especially that the two
venous return curves intersect the zero venous return
~ 2800h
1 f
1
j /
8
" 2400-
1 1
3
a.
1 /
3 2000-
<
-1
Va'I
l\
g 1600-
Normal right
1 \
z
venlrtclei /
\ 1 Normal lefl
1 U / ventricle
__//B
Equilibrated
output
> 1200-
0
z
<
800-
/* ~~~
\ TB<
z
K
Z>
t-
UJ 400-
\
N
\
On
1
l
N ■ i
CO
3
2 0-
\
-4 -2 0 t2 +4 +6
RIGHT ATRIAL PRESSURE (mm Hg)
+ 8
► 50 »40 +30 .20 .10 0 -10
LEFT ATRIAL PRESSURE (mmHg)
pig. 24. An analysis showing the manner in which the two
sides of the heart automatically balance their outputs. The
sequence of events is explained in the text.
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
level at the same point, and this point represents the
mean systemic pressure on the right atrial pressure
scale and the mean pulmonary pressure on the left
atrial pressure scale. Thus, the mean systemic pres-
sure has risen from 7 mm Hg to 7.5 mm Hg, and
this has caused the right ventricular output to rise to
point B, a value about 10 per cent above normal.
On the other hand, the shift of blood out of the lungs
has decreased the mean pulmonary pressure from
7 mm Hg to 3.5 mm Hg, thus shifting the left ven-
tricular venous return curve to the right and decreas-
ing the left ventricular output to point B' , an output
40 per cent below normal. This represents a disparity
of outputs between the two ventricles of 2 to 1 with
far greater amounts of blood being pumped by the
right heart than by the left heart. As a consequence,
a major shift of blood occurs from the systemic circu-
lation back to the pulmonarv circulation, increasing
the mean pulmonary pressure and decreasing the
mean systemic pressure. As a result, the outputs of
the two sides of the heart once again become equi-
librated.
Conversely, a sudden shift of blood from the sys-
temic circulation into the pulmonary circulation is
illustrated by the dashed-dot curves, showing a de-
crease in mean systemic pressure to 6 mm Hg and a
rise in mean pulmonary pressure to 14 mm Hg. The
net result is diminution of right ventricular output
by approximately 10 per cent and enhancement of
left ventricular output by approximately 40 per cent.
Here again there is almost 2 to 1 disparity between
the outputs of the two ventricles, thus resulting in a
rapid shift of blood out of the lungs into the systemic
circulation; this shift continues until the right ven-
tricular output rises to equal the falling left ventricular
output. In this manner, the outputs of the two ven-
tricles once again hecorre re-equilibrated, thus ex-
plaining the experimental findings of many different
investigators that the two sides of the heart always
automatically re-equilibrate with each other within
a few heartbeats (11, 18, 129, 159).
EFFECT OF ACUTE LEFT HEART FAILURE ON CARDIAC
OUTPUT, VENOUS RETURN, LEFT AND RIGHT ATRIAL
PRESSURES, MEAN SYSTEMIC PRESSURE, AND MEAN PUL-
MONARY pressure. Figure 25 illustrates the sequence
of events that occurs when the left ventricle suddenly
fails. Point A is the normal equilibrium point for the
right ventricle and point A' the normal equilibrium
point for the left ventricle. These two are in equi-
librium with each other. Then, suddenly, the left
ventricular output curve falls to less than one-half
normal as depicted by the lower solid curve. Instan-
taneously, this depressed left ventricular output curve
equilibrates with the normal left ventricular venous
return curve at point B which represents onlv 30 per
cent of normal output. Now a 70 per cent disparity-
exists between the momentary right ventricular out-
put and the momentary left ventricular output, this
causing blood to shift into the lungs from the systemic
circulation (138, 139). In a matter of a few heart-
beats the new venous return curves become the dashed
fig. 25. Effect of sudden reduction in pumping
effectiveness of the left ventricle. This shows a shift
of both venous return curves (as illustrated by the
dashed curves) to the left until right ventricular
output falls (point C) to equal the rising left ventric-
ular output (point C)
O
z
ID
>
-4 -2 0 + 2 + 4 +6
RIGHT ATRIAL PRESSURE (mm Hg)
+ 8
+ 50 +40 +30 +20 +10 0 -10
LEFT ATRIAL PRESSURE (mm Hj)
VENOUS RETURN
I 121
curves of the figure, with the systemic venous return
curve being governed by a new mean systemic pres-
sure of only 4 mm Hg and the pulmonary venous
return curve being governed by a very high pulmo-
nary pressure of +28 mm Hg. Blood continues to
shift from the systemic circulation into the pulmonary
circulation until the output of the left ventricle rises
to equal the falling output of the right ventricle. These
conditions are reached at equilibrium point C for
the right ventricle and equilibrium point C" for the
left ventricle. Since the mean pulmonary pressure
has risen to 28 mm Hg, the pulmonarv circulation
has become engorged with blood, and the pulmonary
capillary pressure will probably be above the critical
value of about 25 mm Hg, above which pulmonary
edema begins to appear (94).
This is only a partial analysis because within the
next 30 sec or so sympathetic reflexes will elevate at
least three of the curves, the right ventricular output
curve, the systemic venous return curve, and the left
ventricular output curve, thereby resulting in a fur-
ther elevated equilibrium level of cardiac output
but also further increase in atrial pressures.
EFFECT OF ACUTE RIGHT HEART FAILURE. Figure 26
illustrates the sequence of events when the right heart
fails acutely. Points A and A' represent normal con-
ditions, and point B represents the instantaneous
effect of the acute failure on right heart output, show-
ing that the right heart output is only about one-half
the output of the left ventricle at that point. Immedi-
ately, blood begins to shift from the pulmonary circu-
lation into the systemic circulation (138, 139), and
this shift continues until the left ventricular output
falls to equal the rising right ventricular output. The
new equilibrium points are C for the right heart and
C" for the left heart, both of which now have the same
ventricular outputs and venous returns of 700 ml per
min. During the re-equilibration of blood between
the systemic and pulmonary circulation, the mean
systemic pressure has risen from 7 to 7.4 mm Hg,
while the mean pulmonarv pressure has fallen from
7.0 to 4 mm Hg. This minute increase in mean sys-
temic pressure explains the failure of systemic vascu-
lar pressures to rise greatly in acute right heart failure
(122, 181). After another moment or so, sympathetic
reflexes tend to elevate the different curves as ex-
plained above, and the cardiac output can return
part way toward normal.
effect of blood volume change. Figure 27 analyzes
the effect of hemorrhage on the outputs of both ven-
tricles. Note that the primary effect of reduced blood
volume is to shift the scales for left and right atrial
pressures, moving the zero pressure points toward
each other. An increase in blood volume causes ex-
actly the opposite effect. Here again, since the
capacitance of the pulmonary circulation is only one-
seventh that of the systemic circulation, the left atrial
pressure scale is still one-seventh that of the right
atrial pressure scale. Thus, in figure 27, the mean
systemic pressure has fallen to 1.7 mm Hg and the
■| 2800-
fig. 26. Analysis of the effect of sudden right
ventricular weakness on cardiovascular dynamics
showing a decrease in right ventricular output to
point C and left ventricular output to point C".
-4 -2 0 +2 + 4 +6
RIGHT ATRIAL PRESSURE (mm Hg)
+ 50 +40 +30 +20 +10 0 -10
LEFT ATRIAL PRESSURE (mm Hg)
I 12 2
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
-4 -2 ~0~ +2 +4
RIGHT ATRIAL PRESSURE ImmHg)
jj! +50 +40 +30 +20 +10 0 -10
LEFT ATRIAL PRESSURE (mm Hg)
fig. 27. Effect of reduced blood volume on cardiovascular
dynamics. This figure shows that the mean systemic pressure
and mean pulmonary pressure are both greatly reduced, thus
causing corresponding decreases in the two venous return
curves.
mean pulmonary pressure to 2.5 mm Hg. Thus, the
net effect of a decrease in blood volume is simply to
reduce both the mean systemic and mean pulmonary
pressures. This does not affect, at least temporarily,
the output curves of either the left or the right ven-
tricles until sympathetic reflexes occur. Also, it does
not affect, at least temporarily, the slopes of the two
venous return curves. Therefore, the only significant
effect is a reduction in both the systemic and pul-
monary venous return curves because of the reduced
mean systemic and mean pulmonary pressures. As a
consequence, both the right ventricular and left
ventricular outputs are reduced, in this instance to
approximately 55 per cent of normal.
summary of the complex analysis. This more com-
plex analysis of the circulation has been presented
to illustrate a method for analyzing the effects of
unilateral excess load or unilateral alteration in pump-
inn effectiveness of the heart. It has particular im-
portance in analyzing abnormalities of the pulmonary
circulation. On the other hand, as one can readily
see from the last few figures, even when relativelv
large quantities of blood shift into or out of the pul-
monary circulation, rather small changes occur in
the dynamics of the svstemic circulation. Therefore,
from a practical point of view, when one is concerned
principally with systemic effects of the circulation,
the simplified analysis is usually quite adequate.
Obviously, only a few examples of the vast number
of uses of these two types of analysis have been given.
Because of the multitude of different quantitative
values that can be assumed by different venous re-
turn and different output curves, the analyses can
likewise assume literally thousands of different forms.
However, the various alterations in the individual
curves that can occur under manv different circu-
latory conditions obey rather simple principles.
Therefore, in almost any circulatory condition, one
can either establish the different curves experimentally
or can predict them very accurately, and from these
he can proceed with an analysis of the different effects
which will occur in the circulation, particularly as
they relate to venous return, cardiac output, left and
right atrial pressures, pulmonary blood volume, and
svstemic blood volume.
SPECIFIC FACTORS THAT AFFECT VENOUS RETURN
Thus far, we have considered only a general analy-
sis of venous return. Now we need to consider several
factors that at times play highly significant and spe-
cific roles in the local process of blood flow along
the veins. These include especially the effects of a)
the venous pump, b) the collapse factor, c) central
pressure pulsations, and d) local factors in the tissues
that help to govern venous return such as local tissue
activity and tissue utilization of oxygen.
EFFECT OF THE VENOUS PUMP ON VENOUS RETURN. Al-
most every student of physiologv is already familiar
with the function of the so-called "venous pump."
That is, all peripheral veins beyond the visceral cavi-
ties are supplied with valves oriented toward the
heart, and any factor that causes successive compres-
sions of the veins exerts a pumping action that propels
blood toward the heart. The different types of com-
pression that have been implicated in the venous
pump include a) compression incident to muscular
movement either as a result of direct muscular pres-
sure on the veins or indirectly as a result of movements
of the joints and tissues, and b) pulsatile compression
of the veins caused by arteries lying in the same
sheaths as the veins. The second of these has not
proved to be of any particular significance. Therefore,
the venous pump is also frequently called simply the
"muscle pump" (16, 17, 27, 49, 155, 156, 191).
VENOUS RETURN
I I 23
In quiet standing, blood from the legs returns to
the heart only with great difficulty, and the pressures
in the veins of the lower limb rise to values equal to
the weight of blood between the lower limbs and the
heart, that is, to as much as 90 mm Hg. However,
during walking, the venous return from the lower
limbs will be so satisfactory that venous pressures in
the feet may be as low as 20 to 25 mm Hg (156). In
the absence of an active venous pump, a person can
develop such high pressures in the lower part of the
body when he stands that he actually loses as much
as 15 to 20 per cent of his blood volume in less than
one-half hour, thereby in many instances provoking
fainting.
EFFECT OF VENOUS COLLAPSE ON VENOUS RETURN. The
phenomenon of "venous collapse" is based on the
simple fact that it is impossible to suck fluid through
a collapsible tube. Since the heart is located in the
thoracic cavity where the pressure is normally ap-
proximately — 5 mm Hg and since the right atrial
pressure often is also in the range of —2 to —3 mm
Hg, suction frequently is applied to the central veins.
This is particularly true of the veins entering from
above downward when a person is in an upright po-
sition, because, under these conditions, the negative
hydrostatic pressure of the blood flowing downward
toward the heart adds to the negative pressures al-
ready in the chest, thus causing essentially complete
collapse of the veins in the neck. However, this, too,
is a very old story known by almost every student of
physiology (34, 54-58, 112, 114, 170), and it can be
summarized by simply saying that any factor which
makes the right atrial pressure more negative than
normal does not cause a significant increase in venous
return. That is, the venous return will be as great
when the right atrial pressure is approximately —2
mm Hg as it will be should the right atrial pressure
fall to as low as — 1 5 mm Hg. For instance, when a
person breathes air from a chamber that is under
negative pressure, this negative pressure is trans-
mitted through his lungs to the chambers of his heart.
Yet, breathing against the negative pressure does not
increase the venous return to values above normal —
all because the veins collapse any time there is an
attempt to suck blood from the periphery.
Venous collapse also occurs whenever pressure is
applied to the outside of the veins. This very fre-
quently occurs in the case of elevated abdominal
pressure (24, 88). Figure 28 illustrates the effect of
intra-abdominal pressure on the pressure along the
inside of the inferior vena cava. In these studies, a
catheter was introduced upward from the femoral
vein until it entered the right atrium, showing that
the venous pressure all along the extent of the intra-
abdominal veins was always slightly greater than
36-
32: _
28
AP=35CMH20 AP=ABDOMINAL
■,"Q-e--®"*,-«-e-e..e-e-e-e..o-«..„ PRESSURE
£24
<
5
u.20
O
u
16
tr ia
>
uj 8 ■■
a
<0 4
O
>
-4
AP=30 CM
AP^2SCM
AP = 20CM
"—»__ AP=I5CM
AP=IOCM
._ AP=5CM
NORMAL AP = -0 7 CM H20
fig. 28. Effect of increased abdomi-
nal pressure (AP) on pressures meas-
ured from the tip of a catheter inserted
up the femoral vein and along the vena
cava until it entered the right atrium.
[From Guyton & Adkins (88).]
2 6 10 14 18 22 26 30 34 38 42 46 50
LENGTH OF CATHETER INTRODUCED IN CM
124
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the intra-abdominal pressure. In other words, for
blood to flow through a vein as it returns toward the
heart, the pressure inside the vein must be greater
than the pressure applied to the outside of the vein.
If the abdominal pressure is 25 mm Hg, then the
pressure in all the lower veins of the body that feed
blood through the abdominal cavity, including the
leg veins, must be greater than the 25 mm Hg intra-
abdominal pressure. Likewise, if a bone or some other
structure presses against a vein with a pressure of
10 mm Hg, the pressure in the vein beyond that point
must rise above 10 mm Hg to force blood past the
compression point. These are simple hydrodynamic
principles.
EFFECT OF CENTRAL PULSATION ON VENOUS RETURN.
Probably the most extensively studied factor that has
been considered to affect local venous flow is central
pulsation. There are two different types of central
pulsation which can affect blood flow to the heart.
These are /) increases and decreases in venous pres-
sure resulting from the contractions of the heart it-
self (6, 21, 30, 124, 185), and 2) increases and de-
creases in central venous pressure resulting from
respiration (1, 28, 32, 33, 62, 130, 149, 187). All studies
that have ever been reported on phasic blood flow
from the peripheral veins to the heart have demon-
strated that the flow of blood toward the heart in-
creases greatly during the negative phases of the
central pressure pulses. Then, during the positive
phases, blood flow becomes markedly reduced and
can even flow backward from the right atrium into
the veins. A very significant inflow of blood into the
right atrium occurs during inspiration for two differ-
ent reasons: First, movement of the diaphragm down-
ward decreases the intrathoracic and right atrial
pressures slightly, which helps to move blood toward
the heart. Second, and much more important, down-
ward depression of the diaphragm compresses the
veins of the abdomen, thus forcing large quantities
of blood toward the heart. Brecher and his colleagues
(31) have recently been foremost among a long line
of investigators, extending back a hundred years, in
pointing out the phasic flow of blood to the heart
caused by central pressure pulsations.
Still more important to our present discussion,
however, is not whether or not blood flows into the
heart in greater amount during the negative phase
than the positive phase but, instead, whether or not
central pulsation on the average aids venous return.
Different investigators in the past have gone so far
as to state that central pulsations are among the most
important of all the factors tending to return blood
to the heart, while others have gone so far as to state
that, if anything, central pulsations are harmful to
the venous return rather than beneficial. Brecher's
monograph on venous return presents very admirably
the first point of view (31 ). On the other hand, studies
from our own laboratories during the past year have
indicated that central pulsations on the average
(though not during the negative phases of the pulsa-
tions) cause considerable diminution of venous return
rather than enhancement (91 ). For this reason, it
would be impossible for the author to present any
arguments in favor of the importance of central pul-
sations in returning blood to the heart. Therefore,
the reader is referred to Brecher's thorough mono-
graph for this point of view.
The basis for our belief that central pulsations are
harmful rather than helpful, on the average, to venous
return is depicted in figure 29. This shows the typical
venous return curve, and it shews by means of the
horizontal sine waves the central pulsation excursions,
varying in this instance between the values —6 and
+ 2 mm Hg. The figure then shows by the vertical
pulsations the effects of these pressure changes on
venous return as would be predicted from the venous
return curve. Note that venous return is considerably
depressed during the positive phase of the pulsatile
cycle. On the other hand, venous return is only slightly
increased during the negative phase. Therefore, the
average venous return is decreased approximately
10 per cent as a result of the central venous pulsation.
To test this premise experimentally a cannula was
inserted in the wall of the right atrium, and varying
quantities of blood were injected and removed from
the right atrium at frequencies varying between 60
— I6OO-1
c
Mean
right atrial pressure
E
^-
u
-Average venous return
^t 1 ■*« — ^^— y^- -7^ *y
a 1200-
Nr^^tTTTr7vE
RETURN
1
r\iWt
CO
S 40O-
z
111
>
1
fed \
u-
8
1
-4
6 +4 +8 +12
RIG
-IT
ATRIAL PRESSURE (mm Hg)
fig. 29. Effect of central pulsations on venous return,
illustrating a rectification phenomenon that causes depressed
venous return when central pulsations occur.
VENOUS RETURN
1125
and 160 cycles per min and in volume between o and
64 ml per cycle. In over 200 successive records not a
single instance of increased venous return occurred.
On the contrary, even the minutest increase in right
atrial pulsation always reduced venous return, and
very intensive pulsations actually reduced venous
return (at any given mean right atrial pressure) to as
low as 50 per cent of normal. Thus, there is a "rectifi-
cation phenomenon" occurring in the venous return
to the heart. That is, on the negative pressure cycle
collapse of the veins prevents very much enhance-
ment of venous return, while on the positive pressure
cycle, no such event prevents the positive pressure
from reducing venous return (29, 36, 37, 44, 113, 118,
141, 151). The net effect, based on both theoretical
grounds and experimental grounds, and supported
by studies from other laboratories as well as from our
own (54), is that central pulsations are not of any
value in promoting venous return.
EFFECT OF LOCAL TISSUE ACTIVITY ON VENOUS RETURN —
EFFECT OF OXYGEN USAGE BY THE TISSUES. The best
known condition in which local tissue activity affects
venous return is muscular exercise, in which case the
venous return may be increased several fold. Earlier
it was pointed out that this is caused both by an in-
crease in mean circulatory pressure and by vascular
dilatation in the muscles. The problem still remains,
however, to explain the cause of the vascular dilata-
tion in the muscles which in turn leads to the greatly-
enhanced venous return. In recent years, much evi-
dence has accumulated that oxygen usage by the
tissues might well be the initiating factor that con-
trols vascular dilatation (9, 41, 45, 51, 64, 65, 77, 79,
119, 125, 136, 190, 199). Some research workers have
felt that relative oxygen lack in the tissues causes them
to form a humoral substance which then causes vaso-
dilatation (7, 8, 15, 153). Humoral substances that
have been suspected are carbon dioxide, hydrogen
ions, adenosine phosphate compounds, histamine,
and lactic acid. Thus far, however, none of these
substances has been isolated in sufficiently large quan-
tities from the blood to prove that it is truly acting
as a vasodilator substance.
Another concept is that the tissue cells compete
for the available oxygen in the arterial blood with
the vascular smooth muscle, perhaps with the smooth
muscle of the metarterioles and precapillary sphinc-
ters (45). If the tissues utilize excess oxygen, then
the blood vessels will be without adequate oxygen.
As a result, these vessels might dilate simply because
their smooth muscle walls cannot remain contracted
in the face of oxygen lack. This concept is supported
by the following experiment : venous blood was re-
moved from the right ventricle, and arterial blood
was removed from the aorta of the same dog at the
same time. These two bloods were then alternately
passed through an isolated hind limb of a dog in which
the input and outflow pressures were controlled and
in which the blood temperature was very exactly
controlled. The arterial blood always caused vaso-
constriction, while the venous blood always caused
vasodilatation (to 250% of the arterial value). Fur-
thermore, the degree of vasodilatation depended al-
most proportionately on the degree of unsaturation
of the blood entering the limb as shown in figure 30.
The onlv difference between the two bloods was that
one had passed through the lungs and the other had
not. Therefore, if any vasodilator substance were in
the venous blood, then it would have to have been
removed by the lungs. Since the lungs are not known
to have this ability to remove vasodilator substances
of any type, and since controlled breathing of carbon
dioxide illustrated that carbon dioxide had no sig-
nificant local effect on peripheral vascular flow, we
must presume that it is lack of oxygen that initiates
the vasodilatation in the limb and not some inter-
mediary vasodilator substance.
The reason for discussing this oxygen lack theory
of peripheral vasodilatation so completely is that, in
the final analvsis, it may be oxygen usage by the tis-
100
BLOOD FLOW (Per cent of control volue)
100
120
WO
160
180
200
220
240
fig. 30. Effect of reducing arterial oxygen saturation on
the blood flow through an isolated hind limb of the dog.
[From Crawford el al. (45).]
1 126
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
sues that is the primary factor which normally regu-
lates venous return and, therefore, also cardiac output.
That is, the degree of local dilatation of peripheral
vessels would increase with each increase in local
tissue activity; consequently, the return of blood to
the heart would be governed by tissue utilization of
oxygen. On summating all the individual flows
through all the individual tissues of the body we ob-
tain a summated value which equals venous return,
and, since this automatically equates with cardiac
output, the summated flows of the individual tissues
are also equal to the cardiac output. Therefore, if it
is true that oxygen lack in all individual tissues does
cause vasodilatation, then we find that in the final
analysis the rate of local oxygen utilization could be
the single most important controller of venous return
and cardiac output. Indeed, this is supported by
many isolated studies of the relationship between
oxygen utilization or oxygen lack and circulatory
blood flow, beginning with the study of Douglas
& Haldane (51) in 1922 in which it was shown that
oxygen lack increases the cardiac output to a con-
siderable extent, and extending through studies by
Gorlin and co-workers showing a greatly increased
cardiac output in severe oxygen lack (77), and a
more recent study by Huckabee (119) showing an
increase in cardiac output of as much as twofold
in animals poisoned with cyanide.
Besides the acute peripheral dilatation that re-
sults from oxygen lack, a very marked additional
increase in tissue blood flow occurs over a period of
several weeks if excessive oxygen usage or oxygen
deficiency persists for this long period of time (130a).
This, however, results from increased ''vascularity"
of the tissues, that is, increased numbers of blood
vessels. Nevertheless, this too, despite its slowness to
develop, represents a very important and very power-
ful regulatory mechanism for control of venous re-
turn in response to oxygen need by the tissues.
Aside from the experimental observations on the
control of venous return and cardiac output by oxy-
gen lack, there is one compelling theoretical reason
for believing that oxygen lack should be the main
controller of venous return and cardiac output, and
that is the following: Of all the essential substances
supplied to the tissues by the blood, oxygen is by far
the one most critically dependent upon an adequate
blood flow. For instance, blood flow can be de-
creased to as little as '20 normal, and adequate
quantities of glucose, fats, and proteins can still be
carried to the tissues. Also, if the depth of breathing
is increased, carbon dioxide can be carried away
from the tissues in adequate quantities even when
the cardiac output is decreased to as little as }/\§
normal. On the contrary, the tissues become severely
damaged from anoxia whenever cardiac output
remains only slightly below normal for a prolonged
period of time. Therefore, it is readilv obvious that
oxygen transport to the tissues is normally markedly
"flow limited," while the transport of no single
other essential substance to or from the tissues is
limited to a significant extent under normal or any-
where near normal conditions. For this reason, it is
especially reasonable that oxygen should be the major
regulator of venous return and cardiac output; this
would provide a closed loop regulatory system that
would help to maintain an adequate supply of oxygen
to all the tissues at all times.
VENOUS PRESSURES
The regulation of venous pressure is inextricably
related to the regulation of venous return and car-
diac output, as has already been pointed out in
both the simplified and more complex circuit analy-
ses presented earlier in this chapter. All the different
significant factors which affect right atrial pressure
have already been discussed. On the other hand,
the right atrial pressure is not the same as the more
peripheral venous pressures. Therefore, we need
now to conclude our discussion of the return of blood
to the heart by summarizing the different factors
that determine the peripheral venous pressures.
These include, first and paramount, the right atrial
pressure itself. In addition, they include a) resistance
to blood flow along the veins, b) rate of blood flow
in the veins, and c) hydrostatic pressure effects.
EFFECT OF RESISTANCE TO FLOW IN THE VEINS. Dilated
central veins are so large that they have almost no
resistance to blood flow, but semicollapsed veins,
on the other hand, have very high resistance. This
effect is particularly important at the different com-
pression points where the veins pass over the ribs
or lie against some relatively solid organ (52, 53).
In the ordinary circulation, therefore, the resistance
to venous flow is not negligible, principally because
of the compression points against the veins. On the
other hand, when the right atrial pressure rises to a
very high value, blood can dam up in the veins,
elevating the pressures in the veins to values equal
VENOUS RETURN
I 127
to or perhaps considerably greater than those on
the outside of the veins. In these instances the vein
become distended and the venous resistance becomes
automatically reduced. This turns out to be an im-
portant safety factor in venous return, for often an
elevated right atrial pressure results from a damaged
heart, in which case return of blood to the heart
would become inadequate if the venous resistance
should remain as high under these conditions as it
is in the normal circulation. Fortunately, however,
the reduced resistance of the veins allows the existing
pressure gradient from the periphery to the heart to
force blood toward the heart almost equally as well
as it occurs normally. For this reason, the peripheral
pressures ordinarily do not rise significantly until
the right atrial pressure has risen above approxi-
mately + 4 to +6 mm Hg (74). Above this point,
the veins by then will have become distended, and
any additional rise in right atrial pressure is there-
after reflected by a similar increase in peripheral
venous pressure (83).
parts of the body and creates negative pressure in
areas above the heart. The collapse factor and the
venous pump that modify these pressures were
described earlier in the chapter. Particularly impor-
tant is the fact that the veins of the neck collapse and
their resistances automatically become greatly ele-
vated. Therefore, venous pressure in the neck almost
never falls below atmospheric pressure unless un-
usual circumstances prevent the veins from collaps-
ing.
Because of the importance of the hydrostatic fac-
tor in all venous pressure measurements, two very-
similar methods have been suggested for determin-
ing a "physiological zero" pressure in the venous
system (93, 1 1 1 ). The second of these, which was
presented from our laboratory, depends on rotating
a dog about two different axes. It was found that
venous pressures referred to a point barely inside
the right ventricle at the tricuspid valve did not
vary a measurable amount regardless of the position
of the animal.
EFFECT OF VENOUS FLOW ON PERIPHERAL VENOUS
pressures. An increase in the volume of venous
blood flowing toward the heart theoretically would
cause essentially the same effects on peripheral
venous pressures as would an increase in venous
resistance. However, from a practical point of view
this is not true, because an increase in volume of flow
normally simply distends the collapsed veins to a
greater degree, thus reducing the resistance to flow.
The flow and decreased resistance ordinarily com-
pensate for each other so that increasing the flow-
has relatively minor effect in increasing the peripheral
venous pressures rather than a major effect as might
be expected (88). This has been demonstrated es-
pecially in the case of blood flowing from the periph-
eral limbs through the abdominal cavity when the
intra-abdominal pressure is elevated. For instance,
if the intra-abdominal pressure is +10 mm Hg,
whether the flow from the leg to the right heart
is 0.5 ml per min or 200 ml per min, the pressure
in the femoral vein leading into the abdominal cavity
still remains only 1 mm Hg or so greater than the
10 mm Hg intra-abdominal pressure.
EFFECT OF HYDROSTATIC PRESSURE ON PERIPHERAL
venous pressures. Finally, we have the well-known
effect of hydrostatic forces on peripheral venous
pressures. That is, the simple weight of the blood
increases the venous pressures in the dependent
SUMMARY
To summarize this entire chapter, its important
point has been that one cannot analyze venous re-
turn separately from a simultaneous analysis of
many other factors in the circulation. However,
relatively simplified analyses, based principally on
four major segments of the circulation, the right
heart, the pulmonary circulation, the left heart,
and the systemic circulation, can provide an almost
complete understanding of the interrelationships
between a) venous return, /;) cardiac output, c)
right atrial pressure, d) left atrial pressure, e) mean
systemic pressure, /) mean pulmonary pressure, g)
mean pulmonary volume, and /() mean systemic blood
volume.
If we should choose any single factor that might
be the primary regulator of venous return, and
hence also the primary regulator of cardiac output,
it might be the tissue utilization of oxygen. Cer-
tainly, in over half of the tissues of the body if not
in the entire body, local blood flow seems to be
controlled by the local utilization of oxygen, and
the summated value of all the local flows is the venous
return. Therefore, oxygen utilization by the tissues
might well be, in the final analysis, the primary
regulator of venous return.
I I2t
H \M)H( ic IK (H I'm SIIH.l ICY
CIRCULATION II
R I . F E R E N C E S
>3'
14.
15
17
Alexander, R. S. Influence of the diaphragm upon
portal blood flow and venous return. Am. J. Physiol. 167:
738, I95i- 2'-
Alexander, R. S. The participation of the venomotor
system in pressure reflexes. Circulation Research 2 : 405, 22.
'954-
Alexander, R. S. Venomotor tone in hemorrhage and
shock. Circulation Research 3: 181, 1955. 23.
Alexander, R. S. Reflex alterations in venomotor tone
produced by venous congestion. Circulation Research 4:
49, 1956. 24.
Allen, S. C, C. L. Taylor, and V. E. Hall. A study
of orthostatic insufficiency by the tiltboard method. Am.
J. Physiol. 143: 11, 1945. 25.
Altmann, R. Uber den entstehungsmechanismus des
systolischen kollapses der venenpulskurve. Z. Kreis-
laufforsch. 43: 728, 1954. 26.
Anrep, G. V., G. S. Barsoum, S. Salama, and Z. Souidan.
Liberation of histamine during reactive hyperemia and
muscle contraction in man. J. Physiol. 103: 297, 1944-
Anrep, G. V., and E. Saalfield. The blood flow through 27.
the skeletal muscle in relation to its contraction. J.
Physiol. 85: 375, 1935.
Asmussen, E., and M. Nielsen. The cardiac output in 28.
rest and work at low and high oxygen pressures. Acta
Physiol. Scand. 35: 73, 1955. 29.
Asmussen, E., and M. Nielsen. Cardiac output during
muscular work and its regulation. Physiol. Revs. 35: 778,
'955- 3°-
Barcroft, H. Cardiac output and blood distribution.
J. Physiol. 71 : 280, 1931.
Barger, A. C, V. Richards, J. Metcalfe, and B. 31.
Gunther. Regulation of the circulation during exercise;
cardiac output (direct Fick) and metabolic adjustments 32.
in the normal dog. Am. J. Physiol. 184: 613, 1956.
Barger, A. C, B. B. Roe, and G. S. Richardson. Rela-
tion of valvular lesions and of exercise to auricular 33.
pressure, work tolerance and to development of chronic
congestive failure in dogs. Am. J. Physiol. 169: 384, 1952.
Barratt-Boyes, G. B., and E. H. Wood. Hemodynamic 34.
response of healthy subjects to exercise in the supine
position while breathing oxygen. J. Appt. Physiol. 1 1 : 129,
1957- 35-
Barsoum, G. S., and F. H. Smirk. Observations on the
increase in the concentration of a histamine-like sub-
stance in human venous blood during a period of reactive
hyperemia. Clin. Sci. 2: 353, 1936. 36.
Beecher, H. K., M. E. Field, and A. Krogh. Method
of measuring venous pressure in human leg during
walking. Skand. Arch. Physiol. 73: 7, 1936. 37.
Beecher, H. K., M. E. Field, and A. Krogh. Effect
of walking on venous pressure at ankle. Skand. Arch.
Physiol. 73: 133, 1936. 38-
Berglund, E. Ventricular function. VI. Balance of
left and right ventricular output : relation between left
and right atrial pressures. Am. J. Physiol. 178: 381, 1 954.
Berglund, E. The function of the ventricles of the heart. 39.
Acta Physiol. Scand. 33: Suppl. 1 19, 1955.
Berglund, E., S.J. Sarnoff, and J. P. Isaacs. Ventricu-
lar function : Role of the pericardium in regulation of 40.
cardiovascular hemodynamics. Circulation Research 3 :
133. '955-
Blair, H. A., and A. M. Wedd. The action of cardiac
ejection on venous return. Am. J. Physiol. 145: 528, 1946.
Blalock, A. Exposure of heart to atmospheric pressure;
effects on cardiac output and blood pressure. Arch. Surg.
26:516, 1933.
Bolton, C. The experimental production of uncom-
pensated heart disease with especial reference to the
pathology of dropsy. J. Pathol. Bacteriol. 9: 67, 1903.
Booker, W. M., D. M. French, and P. A. Molano.
Further studies on the acute effects of intra-abdominal
pressure. Am. J. Physiol. 149: 292, 1947.
Boucek, R. J., J. H. Grindlay, and H. B. Burchell.
Experimental constrictive pericarditis : analysis of induced
circulatory failure. .4m. J. Physiol. 169: 434, 1952.
Boucek, R. J., J. H. Grindlay, and H. B. Burchell.
Experimental constriction of inflow tracts in the heart:
analysis of circulatory failure. Am. J. Physiol. 169: 442,
'952-
Bowers, E., E. J. M. Campbell, and C. H. P. Johnston.
Factors promoting venous return from arm in man.
Lancet I : 460, 1945.
Brecher, G. A. Mechanism of venous flow under differ-
ent degrees of aspiration. .4m. J. Physiol. 169: 423, 1952.
Brecher, G. A. Venous return during intermittent
positive-negative pressure respiration studied with a new
catheter flowmeter. Am. J. Physiol. 174: 299, 1953.
Brecher, G. A. Cardiac variations in venous return
studied with a new bristle flowmeter. Am. ./. Physiol. 176:
423. '954-
Brecher, G. A. Venous Return. New York: Grune & Strat-
ton, 1956.
Brecher, G. A., and G. Mixter, Jr. Augmentation of
venous return by respiratory efforts under normal and
abnormal conditions, ,4m. J. Physiol. 171 : 710, 1952.
Brecher, G. A., and G. Mixter, Jr. Effect of respiratory
movement on superior cava flow under normal and
abnormal conditions. .4m. J. Physiol. 172: 457, 1953.
Brecher, G. A., G. Mixter, Jr., and L. Share. Dynam-
ics of venous collapse in superior vena cava system. .4m.
J. Physiol. 171 : 194, 1952.
Buckley, N. M, E. Ogden, and D. S. Linton, Jr. The
effects of work load and heart rate on filling of the iso-
lated right ventricle of the dog heart. Circulation Research
3: 434. '955-
Candel, S., and D. E. Ehrlich. Venous blood flow dur-
ing the Valsalva experiment including some clinical
applications. Am. J. Med. 15: 307, 1953.
Carr, D. T., and H. E. Essex. Certain effects of positive
pressure respiration on circulatory and respiratory sys-
tems. Am. Heart J. 31 : 53, 1946.
Case, R B., E. Berglund, and S. J. Sarnoff. Ventric-
ular function. II. Quantitative relationship between
coronary flow and ventricular function with observations
on unilateral failure. Circulation Research 2: 319, 1954.
Chapman, C. B., and R. S. Fraser. Studies on the effect
of exercise on cardiovascular function. I. Cardiac output
and mean circulation time. Circulation 9: 57, 1954.
Charlier, R. Le role des regions sinusales et cardio-
VENOUS RETURN
I I 29
aortique dans la regulation reflexe du debit cardiaque.
Acta Cardiologica 3:1,1 948.
41. Chiodi, H., D. B. Dill, F. Consolazio, and S. M.
Horvath. Respiratory and circulatory responses to acute
carbon monoxide poisoning. Am. J. Physiol. 134: 683,
■94! •
42. Clowes, G. H. A., and L. R. Del Guercio. Circulatory
response to trauma of surgical operations. Metabolism 9 :
67, i960.
43. Cohen, S. M., O. G. Edholm, S. Howarth, J.
McMiciiael, and E. P. Sharpev-Schafer. Cardiac
output and peripheral blood How in arteriovenous
aneurysm. Clin. Sci. 7: 35, 1948.
44. Cournand, A., H. L. Motley, L. Werko, and D. W.
Richards, Jr. Physiologic studies of effects of intermittent
positive pressure breathing on cardiac output in man.
Am. J. Physiol. 153: 162, 1948.
45. Crawford, D. G., H. M. Fairchild, and A. C. Guyton.
Oxygen lack as a possible cause of reactive hyperemia.
Am. J. Physiol. 197:613, 1959.
46. Daly, I. de B. A closed circuit heart-lung prepara-
tion. I. Effects of alterations in blood volume. J.
Physiol. 60: 103, 1925.
47. Daly, I. de B., P. Egcleton, C. Hebb, J. L. Linzell,
and O. A. Trowell. Observations on the perfused
living animal (dog) using homologous and heterologous
blood. Quart. J. Exptl. Phys.ol. 39: 29, 1954.
48. Davis, J. O., B. Kliman, N. A. Yankopoulos, and R. E.
Peterson. Increased aldosterone secretion following
acute constriction of the inferior vena cava. J. Clin.
Invest. 37: 1783, 1958.
49. Desliens, L. Muscular contractions and blood circula-
tion; role of venous valves. Bull. Acad, med., Paris 130: 476,
1946.
50. Donald, K. W., J. M. Bishop, G. Gumming, and O. L.
Wade. The effect of exercise on the cardiac output and
circulatory dynamics of normal subjects. Clin. Set. 1 4 :
37, 1955-
51. Douglas, C. G., and J. S. Haldane. The regulation of
the general circulation rate in man. ./. Physiol. 56: 6g,
1922.
52. Duomarco, J., and R. Rimini. La pression veineuse des
membres chez l'homme normal et chcz Finsufficient
cardiaque. Compt. rend. Congr. Cardiologie 3:1,1 950.
53. Duomarco, J., and R. Rimini. La presion venosa en los
miembros superiores, en condiciones normales. Rev. arg.
Cardiol. 1 7 : 236, 1 950.
54. Duomarco, J., and R. Rimini. Energy and hydraulic
gradient along systemic veins. Am. J. Physiol. 178: 215,
■954-
55. Duomarco, J., R. Rimini, and F. N. Predari. Sobre
el estado de distension o colapso de las venas cavas. Rev.
arg. Cardiol. 1 2 : 333, 1 946.
56. Duomarco, J., R. Rimini, and P. Recarte. La presion de
los troncos venosos del torax. Rev. arg. Cardiol. 1 1 : 1 29,
■945-
57. Duomarco, J., R. Rimini, and J. P. Sapriza. Intento de
apreciacion de la presion venosa efectiva por medio de
la angiocardiograf fa. Rev. arg. Cardiol. 17: [5, 1950.
58. Duomarco, J., R. Rimini, J. P. Sapriza, and G. H.
Surraco. A proposito del colapso yuxtadiafragmatico
de la vena cava inferior estudio angiocardiograhco A'.
arg. Cardiol. 17: 220, 1950.
59. Ebert, R. V., and E. A. Stead, Jr. The effect of the
application of tourniquets on the hemodynamics of the
circulation. J. Clin. Invest. 19: 561, 1940.
60. Eckstein, R. W., D. Book, and D. E. Gregg. Blood
viscosity under different experimental conditions; effect
on blood flow. Am. J. Physiol. 135: 772, 1942.
61. Eckstein, R. VY\, G. R. Graham, I. M. Liebou, and
C. J. Wiggers. Comparison of changes in inferior cava
flow after hemorrhage and circulatory failure following
transfusion. Am. J. Physiol. 148: 745, 1947.
62. Eckstein, R. W.. C. J. Wiggers, and G. R. Graham.
Phasic changes in inferior cava flow of intravascular
origin. Am. J. Physiol. 148: 740, 1947.
63. Farber, S. J., J. D. Alex\nder, and D. P. Earle.
Shock produced by obstruction of venous return to the
heart in the dog. Am. J. Physiol. 176: 325, 1954.
64. Feinburg, H., A. Gerola, and L. N. Katz. Effect of
hypoxia on cardiac oxygen consumption and coronary
flow. Am. J. Physiol. 195: 593, 1958.
65. Feldman, M., Jr., S. Rodbard, and L. N. Katz. Rela-
tive distribution of cardiac output in acute hypoxemia.
Am. J. Physiol. 154: 391, 1948.
66. Fergusen, T. B., D. E. Gregg, and O. W. Shadle.
Effect of blood and saline infusion on cardiac performance
in normal dogs and dogs with arteriovenous fistulas.
Circulation Research 2 : 565, 1 954.
67. Fergusen, T. B., O. W. Shadle, and D. E. Gregg.
Effect of blood and saline infusion on ventricular end
diastolic pressure, stroke work, stroke volume and cardiac
output in the open and closed chest dog. Circulation Research
1 : 62, 1953.
68. Fleisch, A., and W. Temaszewski. L'influence de la
masse sanguine totale et de l'acide carbonique sur le
debit cardiaque. Arch, intern, physiol. 42: 367, 1936.
69. Fletcher, A. G., Jr., J. D. Hardy, C. Riegel, and
C. E. Koop. Effects of intravenous infusion of gelatin
on cardiac output and other aspects of circulation of
normal persons, of chronically ill patients, and of normal
volunteers subjected to large hemorrhage. J. Clin. Invest.
2i- 405. '945-
70. Franklin, K. J. A Monograph on Veins. Springfield, 111.:
Thomas, 1937.
71. Gammill, J. F., J. J. Applegarth, C. E. Reed, and A J
Antenucci. Hemodynamic changes following acute
myocardial infarction using the dye injection method for
cardiac output determination. Ann. Internal Med. 43: 100,
"955-
72. Gauer, O. H. Die wechselbenziehungen zwischen herz-
und venesystem. Verhandl. deut. Ges. Kreislaufforsch. 22:
61, 1956.
73. Gibbons, T. B. The behavior of the venous pressure
during various stages of chronic congestive heart failure.
Am. Heart J. 35: 553, 1948.
74. Gibert-Queralto, J., R. Nolla-Panades, and F.
Jove-Batalla. L'hemodynamie des veines caves et la
pression veineuse. Acta Med. Scand. 154: Suppl. 312, 673,
■95°-
75. Gilbert, R. P., M. Goldberg, and J. Griffin. Circu-
latory changes in acute myocardial infarction. Circulation
9-847. 1954-
II |0
HANDBOOK OF PHYSIOI.OCY
CIRCULATION II
76. Goldbloom, A. C, M. L. Kramer, and A. Lieberson.
Clinical studies in circulatory adjustments; physiologic
relation between posture and cardiac output. Arch.
Internal Med. 65: 175, 1940.
77. Gorlin, R., and B. M. Lewis. Circulatory adjustments
to hypoxia in dogs. J. Appl. Physiol. 7: 180, 1954.
78. Grodins, F. S. Integrative cardiovascular physiology: a
mathematical synthesis of cardiac and blood vessel hemo-
dynamics. Quart. Rev. Biol. 34: 93, 1959.
79. Grollman, A. Effect of high altitude on cardiac output
of man and its related functions; account of experiments
conducted on summit of Pike's Peak, Colorado. Am. ./.
Physiol. 93: 19, 1930.
80. Guntheroth, W. G. Function of liver and spleen as
venous reservoirs. Federation Proc. 1 7 : 63, 1 958.
81. Guyton, A. C. Determination of cardiac output by
equating venous return curves with cardiac response
curves. Physiol. Revs. 35: 123, 1955.
82. Guyton, A. C. Factors which determine the rate of
venous return to the heart. In: World //ends in Cardiology.
New York: Hoeber, 1956, p. 32.
83. Guyton, A. C. The venous system and its role in the
circulation. Modern Concepts Cardiovascular Disease 27 :
483, '958-
84. Guy'TON, A. C. La circulation veineuse. Symposia from the
Illrd World Congress of Cardiology. Brussels, 1958, p. 109.
85. Guyton, A. C. Cardiac output and venous return in
heart failure. In: Cardiology. New York: McGraw-Hill,
vol. 4, 1959, p. 18.
86. Guyton, A. C. Textbook of Medical Physiology (2nd ed.).
Philadelphia: Saunders, 1961, pp. 350 and 446.
87. Guyton, A. C, B. Abernathy, J. B. Langston, B. N.
Kaufmann, and H. M. Fairchild. Relative importance
of venous and arterial resistances in controlling venous
return and cardiac output. Am. ./. Physiol. 197: 1008,
'959-
88. Guy'ton, A. C, and L. H. Adkins. Quantitative aspects
of collapse factor in relation to venous return (relation
between intra-abdominal pressure and venous pressure).
Am. J. Physiol. 177: 523, 1954.
89. Guyton, A. C, G. C. Armstrong, and P. L. Chipley.
Pressure-volume curves of the entire arterial and venous
systems in the living animal. Am. ./. Physiol. 184: 253, 1956.
90. Guyton, A. C, Batson, H. M., Jr., and G. M. Smith,
Jr. Adjustments of the circulatory system following very
rapid transfusion or hemorrhage. Am. ./. Physiol. 164:
35'. '95'-
in. Guyton, A. C, and O. Carrier. Decrease in venous
return caused by venous pulsation. Federation Proc. 20:
1 20, 1 96 1 .
92 Guyton, A. C, and J. W. Crowill. Dynamics of the
heart in shock. Federation Proc. 20: 51, Suppl. 9, 1961.
93. Guyton, A. C, and F. P. Greganti. A physiologic
reference point for measuring circulatory pressures in the
dog — particularly venous pressure. Am. J. Physiol. 185:
'37. '956-
94. Guyton, A. C, and A. VV. Lindsey. Effect of elevated
left atrial pressure and decreased plasma protein concen-
tration on the development of pulmonary edema. Circula-
tion Research 7: 649, 1959.
95. Guyton, A. C, A. W. Lindsey', B. Abernathy, and
J. B. Langston. Mechanism of the increased venous
return and cardiac output caused by epinephrine Am.
J. Physiol. 192: 126, 1958.
96. Guyton, A. C, A. W. Lindsey, B. Abernathy, and T. Q.
Richardson. Venous return at various right atrial pres-
sures and the normal venous return curve. Am. J. Physiol.
189: 609, 1957.
97. Guyton, A. C, A. VV. Lindsey, and J. J. Gilluly. The
limits of right ventricular compensation following acute
increase in pulmonary circulatory resistance. Circulation
Research 2: 326, 1954.
98. Guyton, A. C, A. VV. Lindsey, and B. N. Kaufmann.
Effect of mean circulatory filling pressure and other
peripheral circulatory factors on cardiac output. Am. J.
Physiol. 1 80 : 463, 1 955.
99. Guyton, A. C, A. W. Lindsey, B. N. Kaufmann, and
J. B. Abernathy. Effect of blood transfusion and hemor-
rhage on cardiac output and on the venous return curve.
Am. J. Physiol. 194: 263, 1958.
100. Guyton, A. C, B. H. Douglas, J. B. Langston, and
T. Q. Richardson. Instantaneous increase in mean
circulatory pressure and cardiac output at onset of
muscular activity. Circulation Research. In press.
101. Guyton, A. C, D. Polizo, and G. G. Armstrong. Mean
circulatory filling pressure measured immediately after
cessation of heart pumping. Am. J. Physiol. 179: 261, 1954.
102. Guyton, A. C, and T. Q. Richardson. Effect of hema-
tocrit on venous return. Circulation Research 9: 157, 1961.
103. Guyton, A. C, and K. Sagawa. Compensations of
cardiac output and other circulatory functions in areflex
dogs with large A-V fistulae. Am. J. Physiol. 200: 1157,
1 961.
104. Guyton, A. C, and J. Satterfield. Factors concerned
in electrical defibrillation of the heart through the
unopened chest. Am. J. Physiol. 167: 81, 1951.
105. Guyton, A. C, J. H. Satterfield, and J. W. Harris.
Dynamics of central venous resistance with observations
on static blood pressure. Am. J. Physiol. 169: 691, 1952.
106. Harrison, T. R. Arterial and venous pressure factors
in circulatory failure. Physiol. Revs. 18: 86, 1938.
107. Harrison, T. R., B. Friedman, G. Clark, and G. Resnik.
Cardiac output in relation to cardiac failure. Arch.
Internal Med. 54: 239, 1934.
108. Harrison, T. R., VV. G. Harrison, J. A. Calhoun,
and J. P. Marsh. Congestive heart failure. Arch. Internal
Med. 50: 690, 1932.
109. Hatcher, J. D., F. A. Sunahara, O. G. Edholm, and
J. M. Woolner. The circulatory adjustments to post-
hemorrhagic anemia in dogs. Ciiculation Research 2 : 499,
'9r>4
1 10. Hochrein, M., and K. Matthes. Verschiedenheiten der
schlagvolumina und ungleichmassigkeiten der leistung
beicler ventrikel in ihrer auswirkung auf lungendepot
und herzdurchblutung. Pftiigers Arch. ges. Physiol. 231 :
207, 1932.
mi. Holt, J. P. The measurement of venous pressure in man
eliminating the hydrostatic factor. Am. J. Physiol. 130:
635> '94°-
112. Holt, J. P. The collapse factor in the measurement of
venous pressure. Am. ./. Physiol. 134: 292, 1 94 1 .
113. Holt, J. P. The effect of positive and negative intra-
thoracic pressure on peripheral venous pressure in man.
Am. J. Physiol. 139: 208, 1943.
VENOUS RETURN
I I;} I
114. Holt, J. P. Effect of positive and negative intrathoracic
pressure on cardiac output and venous pressure in dog.
Am. J. Physiol. 142: 594, 1944.
115. Holt, J. P., and P. K. Knoefel. Changes in plasma
volume and cardiac output following intravenous injec-
tion of gelatin, serum and physiologic saline solution. J.
Clin. Invest. 23: 657, 1944.
116. Holt, J. P., W. J. Rashkind, R. Bernstein, and J. C.
Greisen. The regulation of arterial blood pressure. Am.
J. Physiol. 146: 410, 1946.
117. Howarth, S., J. McMichael, and E. P. Sharpey-
Shafer. Effects of venesection in low output heart
failure. Clin. Sci. 6: 41, 1946.
118. Hubav, C. A., R. C. Waltz, G. A. Brecher, J. Praglin,
and R. A. Hingson. Circulatory dynamics of venous
return during positive-negative pressure respiration.
Anesthesiology 15:445, 1954.
119. Huckabee, VV. E. Circulatory response to cytochrome
oxidase inhibition in vivo. Federation Proc. 19: 119, i960.
120. Isaacs, J. P., E. Berglund, and S. J. Sarnoff. Ventric-
ular function: III. The pathologic physiology of acute
cardiac tamponade studied by means of ventricular
function curves. Am. Heart J. 48: 66, 1954.
121. Jouve, A. X. Exploration clinique dc la circulation de
retour au cours de l'insufnsance cardiaque. Paris med.
1 : 385. '938-
122. Kagan, A. Dynamic responses of the right ventricle
following extensise damage by cauterization. Circulation
5: 816, 1952.
123. Katz, L. N. Relation of initial volume and initial pres-
sure to dynamics of the ventricular contraction. Am. J.
Physiol. 87: 348, 1928.
124. Katz, L. N. The role played by the ventricular relaxation
process in filling the ventricle. Am. J. Physiol. 95: 542,
125. Katz, L. N., and H. Feinburg. The relation of cardiac
effort to myocardial oxygen consumption and coronary
flow. Circulation Research 6: 656, 1958.
126. Katz, L. N., W. Wise, and K. Jochim. The dynamics
of the isolated heart and heart-lung preparations of the
dog. Am. J. Physiol. 143: 463, 1945.
127. Katz, L. N., W. Wise and K. Jochim. The dynamics
of the non-failure period of the isolated heart and heart-
lung preparation. Am. J. Physiol. 143: 495, 1945.
128. Katz, L. N., W. Wise, and K. Jochim. The dynamic
alterations in heart failure in the isolated heart and
heart-lung preparation. Am. J. Physiol. 143: 507, 1945.
129. Kjellberg, S. R., U. Rudhe, and T. Sjostrand. The
relationship between the pulmonary blood content, the
heart volume and the filling rate of the left ventricle.
Acta Physiol. Scand. 24: 49, 1951.
130. Knebel, R., and D. Wick, Uber den Einfluss der atmung
auf den zentralen venendruck. /.. Kreislaufforsch 47: 623,
'958-
130a. Korner, P. I. Circulatory adaptations in hypoxia. Physiol.
Revs. 39 : 687, 1959.
131. Krayer, O. Uber die beziehung zwischen pulsfrequenz,
minutenvolumen und venendruck am isoliertcn
saugetierherzen. Arch, exptl. Pathol. Pharmakol. 157: go,
'93°
132. Landis, E. M., E. Brown, M. Fauteaux, and C. Wise.
Central venous pressure in relation to cardiac "com-
petence," blood volume and exercise. J. Clin. Invest. 25 :
-*37. '946-
133. Landis, E. M., and J. C. Hortenstine. Functional
significance of venous blood pressure. Physiol. Revs. 30:
', 195°-
134. Langston, J. B., and A. C. Guyton. Effect of epinephrine
on the rate of urine formation. Am. ■/. Physiol. 192: 131,
■958.
135. Langston, J. B., A. C. Guyton, and W. J. Gillespie.
Acute effect of changes in renal arterial pressure and
sympathetic blockade on kidney function. Am. J. Physiol.
■97: 595. '959-
136. Levy, S. E., and A. Blalock. Fractionation of output
of heart and of oxygen consumption of normal unanes-
thetized dogs. Am. J. Physiol. 118: 368, 1937.
137. Lindhard, J. L'eber die regulierung des kreislaufcs im
gesunden und kranken organismus. Cardwlogia 1 : 366,
■937-
138. Lindsey, A. W., B. F. Banahan, R. N. Cannon, and
A. C. Guyton. Pulmonary blood volume of the dog
and its changes in acute heart failure. Am. J. Physiol
■9°: 45. 1957-
139. Lindsey, A. W., and A. C. Guyton. Continuous record-
ing of pulmonary blood volume, and pulmonary pressure
and volume changes during acute right or left ventricular
failure. Am. J. Physiol. 197: 959, 1 959-
140. Loo, A. V., and E. C. Heringman. Circulatory changes
in the dog produced by acute arteriovenous fistula. Am.
J. Physiol. 158: 103, 1949.
141. Maloney, J. V., Jr., and S. W. Handford. Circulatory
responses to intermittent positive and alternating posi-
tive-negative pressure respirators. ./. Appl. Physiol. 6:
453. '954-
Mateeff, D. Der orthostatische kreislaufkollaps —
gravitationsshock (gravity shock) — beim menschen nach
korperlichcr. Arbeitsphysiolgie 8: 595, 1 935.
Mateeff, D., and C. Petroff. Gravitationsshock beim
menschen nach muskelarbcit. Z. ges. exptl. Med. 85: 115,
'93*-
144. Mayerson, H. S., and G. E. Burch. Relationships of
tissue (subcutaneous and intramuscular) and venous
pressures to syncope induced in man by gravity. Am. J.
Physiol. 128: 258, 1939.
145. Metcalfe, J., J. W. Woodbury, V. Richards, and
C. S. Burwell. Studies in experimental pericardial
tamponade; effects on intravascular pressures and
cardiac output. Circulation 5: 518, 1952.
146. Milnor, W. R., A. D.Jose, and C. J. McGaff. Pulmon-
ary vascular volume, resistance, and compliance in man.
Circulation 22: 130, i960.
147. Morhardt, P. E. Collapsus et syncopes par arret de la
circulation en retour. Vie med. 16: 109, 1935
148. Nickerson, J. L., F. W. Cooper, Jr., R. Robertson,
and J. V. Warren. Arterial, atrial and venous pressure
changes in the presence of an arteriovenous fistula.
Am. J. Physiol. 167: 426, 1951.
149. Opdyke, D. F., H. F. Van Noate, and G. A. Brecher.
Further evidence that inspiration increases right atrial
inflow. Am. J. Physiol. 162: 259, 1950.
150. Opdyke, D. F., and C. J. Wiggers. Studies of right and
left ventricular activity during hemorrhagic hypotension
and shock. Am. J. Physiol. 147: 270, 1946.
142.
'43-
1 132
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
151. Otis, A. B., H. Rahn, and W. O. Fenn. Venous pressure 168.
changes associated with positive intrapulmonary pres-
sures; their relationship to the distensibility of the lung.
Am. J. Physiol. 146: 307, 1946.
15 2. Page, E. B.,J. B. Hickam, H. O. Sieker, H. D. McIntosh, 169.
and W. W. Pryor. Reflex venomotor activity in normal
persons and in patients with postural hypotension. 1 70.
Circulation 1 1 : 262, 1955.
153. Patterson, G. C, and R. F. VVhalen. Reactive hy-
peremia in the human forearm. Clin. Sci. 14: 197, 1955.
154. Patterson, S. W., and E. H. Starling. On the me- 171.
chanical factors which determine the output of the ven-
tricles. J. Physiol. 48: 357, 1 91 4.
155. Pollack, A. A., B. E. Taylor, T. T. Myers, and E H.
Wood. The effect of exercise and body position on the 172.
venous pressure at the ankle in patients having venous
valvular defects. J. Clin. Invest. 28: 559, 1949.
156. Pollack, A. A., and E. H. Wood. Venous pressure in
the saphenous vein at the ankle in man during exercise 1 73.
and changes in posture. J. Appl. Physiol. 1 : 649, 1949
157. Post, R. S. Decrease of cardiac output by acute peri-
cardial effusion and its effect on renal hemodynamics and
electrolyte excretion. Am. J. Physiol. 165: 278, 1951. 174.
158. Rashkind, W. F., D. H. Lewis, J. B. Henderson, D. F.
Heiman, and R. B Dietrick. Venous return as affected 175.
by cardiac output and total peripheral resistance. Am.
J. Physiol. 175: 415, 1953. 176.
159. Reiss, R. A., and J. R. DiPalma. Right and left heart
failure: unilateral rises in right and left auricular pressure 177.
in hypervolemic cats following near lethal doses of
quinidine, auricular fibrillation and epinephrine. Am.
J. Physiol. 155:336, 1948. 178.
160. Remincton, J. W., W. F. Hamilton, G. H. Boyd, Jr.,
W. F. Hamilton, Jr., and H. M. Caddell. Role of
of vasoconstriction in the response of the dog to hemor-
rhage. Am. J. Physiol. 161 : 116, 1950. 179.
161. Remington, J. W., W. F. Hamilton, H. M. Caddell,
G. H. Boyd, Jr., and W. F. Hamilton, Jr. Some cir- 180.
dilatory responses to hemorrhage in the dog. Am. J.
Physiol. 161 : 106, 1950.
162. Richards, D. W., Jr., A. Cournand, R. G. Darling,
and W. H. Gillespie. Pressure in the right auricle of 181.
man, in normal subjects and in patients with congestive
heart failure. Trans. Assoc. Am. Physicians 56: 218, 1941.
163. Richards, D. W., A. Cournand, R. C. Darling, W. H.
Gillespie, and E. DeF. Baldwin. Pressure of blood in
the right auricle, in animals and in man: under normal 182.
conditions and in right heart failure. Am. J. Physiol.
'36: "5. >942-
164. Richardson, T. Q., and A. C. Guyton. Effects of
polycythemia and anemia on cardiac output and other
circulatory factors. Am. J. Physiol. 197: 1167, 1959- 183.
165. Richardson, T. Q_., J. O. Stallings, and A. C. Guyton.
Pressure-volume curves in live, intact dogs. Am. J.
Phisiol. 201 : 471, 1 96 1.
166. Roos, A., and J. R. Smith. Production of experimental
heart failure in dogs with intact circulation. Am. J. 184.
Physiol. 1 53 : 558, 1 948.
167. Root, W. S., W. W. Wolcott, and M. I. Gregersen. 185.
Effects of muscle trauma and of hemorrhage upon cardiac
output of dog. Am. J. Physiol. 151 : 34, 1947.
Rose, J. C, S. J. Cosimano, Jr., C. A. Hufnagel, and
E. A M assullo. The effects of exclusion of the right
ventricle from the circulation in dogs. J. Clin. Invest. 34:
■625, '955-
Rushmer, R. F., and D. A. Smith, Jr. Cardiac control.
Phisiol. Revs. 39: 41, 1959.
Ryder, H. W., W. E. Molle, and E. B. Ferris, Jr.
The influence of the collapsibility of veins on venous
pressure, including a new procedure for measuring tissue
pressure. J. Clin. Invest. 23: 333, 1944.
Sarnoff, S. J., and E. Berglund. Ventricular function.
I. Starling's law of the heart studied by means of simulta-
neous left and right ventricular function curves in the
dog. Circulation 9: 706, 1954.
Sarnoff, S. J., R. B. Case, E. Berglund, and L. C.
Sarnoff. Ventricular function. V. The circulatory
effects of aramine; mechanism of action of "vasopressor"
drugs in cardiogenic shock. Circulation 10: 84, 1954.
Schlesincer, E. G., and R. Hazen. The cardiovascular
effects of arteriovenous fistulae above and below the
heart. Trans. Am. Neurol. Assoc. 79th meeting, 1954,
p. 214.
Schneider, E. C, and R. Collins. Venous pressure
responses to exercise. Am. J. Physiol. 121: 574, 1938.
Scott, J. C. Cardiac output in standing position. Am. J.
Physiol. 115:268, 1936.
Sharpey-Schafer, E. P. Cardiac output in severe
anemia. Clin. Sci. 5: 125, 1944.
Sleator, W, Jr., J. O. Elam, W. N. Elam, and H. L.
White. Oximetric determinations of cardiac output
responses to light exercise. J. Appl. Physiol. 3: 649, 1 95 1.
Smith, E. L., R. A. Huggins, R. W. Randall, and G. A.
Jeffery. Hemodynamic changes resulting from insertion
of a rotameter in the venous circulation of a dog. Texas
Repts. Biol, and Med. 10: 674, 1952.
Starling, E. H. Some points in the pathology of heart
disease. Lancet 1 : 652, 1897.
Starr, I. Role of the "static blood pressure' in ab-
normal increments of venous pressure, especially in
heart failure. II. Clinical and experimental studies.
Am. J. Med. Sci. 199: 40, 1940.
Starr, I., W. A. Jeffers. and R. H. Meade. The
absence of conspicuous increments of venous pressure
after severe damage to the right ventricle of the dog, with
a discussion of the relation between clinical congestive
failure and heart disease. Am. Heart ./. 26: 291, 1943.
Starr, I., and A. J. Rawson. Role of the "static blood
pressure" in abnormal increments of venous pressure,
especially in heart failure. I. Theoretical studies on an
improved circulation schema whose pumps obey Star-
ling's law of the heart. Am. J. Med. Sci. 199: 27, 1940.
Sunahara, F. A., J. D. Hatcher, L. Beck, and C. W.
Govvdey. Cardiovascular responses in dogs to intravenous
infusions of whole blood plasma, and plasma followed
by packed erythrocytes. Can. ./. Biochem. and Phvsiol.
33: 349. '955-
Sweeney, H. M., and H. S. Mayerson. Effect of posture
on cardiac output. Am. J. Physiol. 120: 329, 1937.
Takashima, M. Experimental and clinical study of
venous return. I. Relationship between cardiac systole
and venous return. Biol. Abstr. 28: 18036, 1954.
VENOUS RETURN
1 1 33
1 86. Takashima, M. Clinical and experimental study on
venous return. II, III. Influence of pneumothoraces on
venous return. Biol. Abstr. 28: 23210, 1954.
187. Takashima, M. Clinical and experimental study on
venous return. IV, V. Influence of respiration on venous
return. Biol. Abstr. 28: 2321 1, 1954.
188. Tichv, V. L , and B. W Shaw. Augmentation of femoral
venous flow in dog by electrical stimulation of muscles.
Proc. Soc. Exptl. Biol. Med. 69: 368, 1948.
[89. Trapold, J. H. Role of venous return in the cardiovas-
cular response following injection of ganglion-blocking
agents. Circulation Research 5: 444, 1 957.
190. Ullrick, W. C, VV. V. Whitehorn, B. B. Brannan,
and J. G. Krone. Tissue respiration of rats acclimatized
to low barometric pressure. J. Appl. Physiol. 9: 49, 1956.
191. Walker, A. J., and C. J. Longland. Venous pressure
measurement in the foot in exercise as an aid to investiga-
tion of venous disease in the leg. Clin. Sci. 9: 101, 1950.
192. Warner, H. R. The use of an analog computer for
analysis of control mechanisms in the circulation. Proc.
IRE. 47: 1913, 1959.
193. Warren, J. V., E. S. Brannon, E. A. Stead, Jr., and
A. J. Merrill. Effect of venesection and pooling of blood
in extremities on atrial pressure and cardiac output in
normal subjects with observations on acute circulatory
collapse in three instances. J. CHn. Invest. 24: 337, 1945.
194. Weber, E. H. Ber sacks. Ges. (Afcad.) IViss., 196, 1850.
(Quoted by F. S. Grodins. Integrative cardiovascular
physiology: a mathematical synthesis of cardiac and
blood vessel hemodynamics. Quart. Rev. Biol. 34: 93,
■959)
195. Weiss, S., R. W. Wilkins, and F. W. Havnes. The
nature of circulatory collapse induced by sodium nitrite.
J. Clin. Invest. 16: 73, 1937.
196. Weissler, A. M., J. J. Leonard, and J. V. Warren.
Effects of posture and atropine on the cardiac output.
./. Clin. Invest. 36: 1656, 1957.
197. Wigcers, C. J. The failure of transfusions in irreversible
hemorrhagic shock. Am. J. Physiol. 144: gi, 1945.
198. Wiggers, C. J., and J. M. Werle. Cardiac and pe-
ripheral resistance factors as determinants of circulatory
failure in hemorrhagic shock. Am. J. Physiol. 136: 421,
■942-
199. Yonge, L. R., and W F. Hamilton. Oxygen consump-
tion in skeletal muscle during reactive hyperemia. Am.
J. Physiol. 197: 190, 1959.
CHAPTER 33
Effects of ions on vascular smooth muscle1
SYDNEY M . FRIEDMAN
CONSTANCE L. FRIEDMAN
Department of Anatomy, The University of
British Columbia, Vancouver, Canada
CHAPTER CONTENTS
Introduction
General Physical Chemistry of Ions
Osmotic equilibrium
Ion size
Ion activity
Ion mobility
Ion penetrability or membrane permeability
Special Properties of Ions in Biological Systems
Donnan equilibrium
Electrochemical gradients and membrane potentials
Special Properties of the Physiologically Important Ions
Classification and Critical Appraisal of Methods
Studies of Vascular Smooth Muscle Tension Associated
with Manipulation of the Milieu
Measurement of diastolic blood pressure
Measurement of peripheral resistance in regional vascular
beds
Measurement of tension in the vascular strip or ring or in
an analogous smooth muscle strip
Studies of the Milieu During Manipulation of Vascular or
Analogous Smooth Muscle Tension
Measurement of extracellular and/or intracellular ions
and water
Measurement of osmotic pressure and pH
Electrical measurements
Continuous monitoring of Na or K. ion activity
Role of Sodium and Potassium in Vascular Smooth Muscle
Tension
Evidence from Studies of Diastolic Blood Pressure or
Reactivity
General relation of Na and K to clinical and experimental
hypertension
Effects of varying Na and K. intake or loss on blood
pressure
Evidence from Studies of Tension or Reactivity of Vascular
or Analogous Tissue
Effects of manipulation of Na in the medium
Effects of manipulation of K in the medium
1 Submitted for publication December ->o, i960.
Evidence From Studies of Resistance in Regional Vascular
Beds
Effects of Na infusions
Effects of K infusions
Evidence from Measurement of Na and K in Relation to
Blood Pressure
Measurement of Na and K in chronic hypertension or
hypotension
Measurement of Na and K. in acute hypertension or
hypotension
Evidence from Studies of the Relation of Electrical Activity
to Tension in Vascular or Analogous Tissues
Role of Calcium and Magnesium in Vascular Smooth Muscle
Tension
Calcium
Magnesium
Role of H+ and OH" in Vascular Smooth Muscle Tension
Role of Anions in Vascular Smooth Muscle Tension
Theoretical Interpretations
Summary
INTRODUCTION
General Physical Chemistry of Ions
We are here concerned only with pointing out in
simplified form the basic physical-chemical features
of ions and the manner in which these affect their
role in biological processes. For detailed treatment of
this subject the reader is referred to standard chemistry
texts as well as to the basic articles of Conway (31,
32), Hodgkin (112) and Shanes (182, 183).
osmotic equilibrium. We may begin by considering
osmotic pressure to be the equivalent of the mechanical
pressure which must be applied to a solution to
prevent osmosis of the surrounding solvent into the
solution through the membrane. It is, therefore, not
"35
■ 36
HANDBOOK OF PHYSIOLOGY
( !IR( I'LATION II
primarily a characteristic of the membrane but is a
measure of some real difference between pure solvent
and solution. The membrane merely allows this
difference to show itself.
The osmotic pressure, IT, of any solution is propor-
tional to temperature and concentration:
II*kCT
Accordingly, in dealing with a cell, we must con-
sider the osmotic pressure not only of the solution
inside the cell but also that of the environment which
bathes it. The cell contains an amount of nondif-
fusible material in solution which is essential to its
metabolism. Clearly, the amount of this material
which can be retained without causing the cell to
swell will be sharply limited unless there is also a
counterbalancing nondiffusible material outside its
membrane. This is perhaps the major niche into
o 7
o 3
I
(No) +- 5-7
(X 100)
02
06
10
I -A
IB
pig. I. The dependence of sartorius weight on [Na+] of
the medium (24-hour immersion at 3 C). Closed circles : stepwise
NaCl reduction replaced by KC1. Crosses: stepwise NaCl reduc-
tion not replaced by K.C1. Open circles: stepwise NaCl reduction
in the presence of cyanide (2 X io-3 m). Volume control at
equilibrium depends on [Na+] even in the presence of cyanide
and is independent of [K+] above maintenance level (20
meq/liter). [From Conway (31).
which evolution has fitted the sodium ion. Cell
membranes are almost impermeable to this ion and,
since the cellular nondiffusible material is almost
constant, the amount of cellular water is controlled
In variations in extracellular sodium, Na0. Conway
(31) has demonstrated this point by showing, for
example, that K. concentration in the medium, K„,
can be varied over wide limits without influencing
the basic dependence of cell volume on Na<, after
equilibration (fig. 1 ). This essential point is often
overlooked in experiments dealing with alterations
of the medium.
The development of osmotic pressure is one of the
colligative properties of solutions. Ideally it is de-
termined by the number of particles in solution :
/7i/--/?/?r
Its expression, however, depends on whether or not
the membrane is permeable to the particle in question.
At equilibrium, no osmotic pressure is exerted by
a particle to which a membrane is freely permeable.
ion size. Ions may be defined as particles which have
gained or lost an electron on passing into solution.
The elements of Group 1 of the periodic table, the
alkali metals Li, Na, K, Rb, and Cs, are all charac-
terized by possessing a single electron in their outer
orbital shell. In solution this is lost so that the element
loses its electroneutralitv and remains positively
charged as Li+, Na+, etc. Other elements, like those
in Group 7, take up one or more electrons into their
outer shell and so become, in solution, negatively
weighted, e.g., Cl— , Br-, I , etc. The formation of
ions is not restricted to elements but also occurs with
more complex radicals which can collectively gain
or lose one or more electrons, e.g., OH-, XH4+,
etc.
Ions share the ordinary colligative properties of
substances in solution as, for example, freezing point
depression, osmotic pressure, etc. Additionally, they
possess a number of special properties based on the
fact that they are electrically charged.
We must distinguish here the size of the ion con-
sidered as a solid ball, so to speak, and defined by its
crystal radius, from its size when associated with
water molecules and defined by its hydrated radius.
The increase in size of the monoatomic crystals of
Group 1 falls naturally into the same order as their
periodic arrangement Li < Na < K < Rb < Cs.
For many years it has been the accepted practice
to emphasize the hydrated ion in biological systems.
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE
I 137
Crystal Radius,
A
Hydrated Radii, A
From Mobility
O.60
2.31
°-95
1.78
1 -33
I .22
1 .48
I ,|8
1.69
1. 16
..48
I .21
table I . Ionic Radii of Alkali Metal Ions
Li
Na
K
Rb
Cs
NH,+
From Ussing et a/. (202;.
The smaller the ionic crystal the more densely is
its electrical field packed and the greater its attraction
for water dipoles. Consequently, the size of the pack-
age, crystal plus water, moving in a solution, is
larger the smaller the crystal. Thus, the rank order
for the hydrated ion size is Cs < Rb < K < Na < Li
(table 1). This model should not be accepted un-
critically (160) although it does provide us with a
useful working framework.
ion activity. Ions in solution are subject to interionic
forces which limit their availability. In consequence,
the measured concentration of an ion may be greater
or less than its reactive concentration by some
measurable degree defined as the activity coefficient.
The concentration of the substance, corrected by the
activity coefficient, defines the activity of the ion in
the given solution. Since reactions in solutions and
their resultant equilibria are determined by activities
rather than by concentrations this point has special
importance. As figure 2 shows, each salt has its own
characteristic curve relating activity to concentra-
tion. Standard tables are available. For similar
reasons, the ionization constant of salts is also of
import, since not all salts dissociate with equal com-
pleteness into ions. Activity will be symbolically
written (Na+) and concentration [Na+]. Where such
precision is not necessary in a given context we shall
simplv write Na for sodium or Na+ for sodium ion.
ion mobility. In general, the mobility of an ion in
free solution varies inversely with its hydrated size.
In fact, the concept of ion hydration was in part
developed to explain the relative mobilities of ions.
Ion mobility is ordinarily measured as velocity in a
standard electrical field. Table 2 shows the relation
of increasing size to slower velocity.
ion penetrability or membrane permeability. Cell
membranes in general appear to have channels so
limited in size as to allow K to enter freely and just
fig. 2. Mean activity coefficients of various electrolytes at
25 C. [From Prutton & Maron (161 a).]
to exclude Na. Conway (32) has presented interesting
and basic data pertinent to this point (table 3).
Frog sartorius immersed in Ringer fluid, to which 100
iriM of a particular salt is added, at first loses weight
(osmotic withdrawal of water). Then, if the membrane
is freely permeable to the salt, the weight increases
back to its original base. Thus, at equilibrium, the
added salt has not upset the osmotic balance. The
time required to recover 50 per cent of the weight loss
is taken as a measure of the permeability of the mem-
brane for the particular ion and table 3 is so con-
structed. It is evident that taking KC1 as standard,
RbCl and CsCl enter readily, while the chlorides of
Na, Li, Ca, and Mg do not enter appreciably. Simi-
larly, in the anion series, bromide and nitrate enter
easily while phosphate, acetate, bicarbonate, and
sulfate are excluded.
On the right side of the table Conway compares
the diffusion coefficients of the ions rather than rela-
tive ion diameter with K and points out that the
correspondence is far from exact (cf Rb and Cs pene-
tration rates with diffusion rates).
The permeability of the cell membrane to ions is
not a fixed characteristic but must be expected to
vary physiologically and pathologically.
1 138
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
table 2. Relation 0} Ion Size to Mobility in an Electric Field
Velocities of Ions Under Gradient of i 17cm or 0.5 17cm for Divalent Ions
Relative Ion Diameters (Diameter of Potassium Ion = 1.00J
Cations
Anions
Cations
Anions
H
3>5-2
OH
'73-8
H
O.20
OH
°-37
Rb
67-5
Br
6/.3
Rb
O.96
Br
0.96
Cs
64.2
I
66.2
Cs
I .00
1
o-97
NH,
643
CI
65.2
NH,
I .00
CI
0.98
K
64.2
NO3
61 .6
K
I .OO
N03
1 .04
Na
43-2
CH3COO
35-°
Na
I.49
CH3COO
1 .84
Li
33 0
SO,
34°
Li
'•95
so.
..89
Ca
25-5
HPO,
28.0
Ca
2-5'
HPO,
2.29
Mg
22.5
Mg
2.84
From Conway (32).
table 3. Relative Entrance Rates of Ions into
Muscle (Left Column) Compared with Their
Diffusion Constants, D, Though Water Relative to A
Taken as 100 (Right Column).
Cation series
D for sinp
le ions, with K value =
= 100
KC1
100
K
100
RbCl
38
Rb
103
CsCl
8
Cs
104
NaCl
0
Na
67
LiCl
0
Li
52
CaCL
0
Ca
40
MgCl2
0
Mg
35
Anion series
KC1
100
CI
100
KBr
63
Br
105
KN03
'7
N03
96
K phosphate
4
H2PO,
5°
KOOCCH3
3
HPO,
39
KHCOs
1
CH3COO
54
K,SO,
0
HCO.,
67
so.
53
From Conway (31).
Special Properties of Ions in Biological Systems
donnan equilibrium. By thermodynamic principles
it can be demonstrated that the product of diffusible
cations and anions on the two sides of a membrane
must attain equality at equilibrium. Thus:
(cations), (anions), = (cations)<,(anions)e,
When the cell is nonpermeable to some charged
material on one side of its membrane, a sufficient
number of diffusible ions of opposite charge is re-
quired to balance this. In the cell this fixed material
is mainly negative in charge so that it can be shown
that at equilibrium the sum of the diffusible cellular
cations must be rather larger than the sum of dif-
fusible anions. In brief, this shows up in the low
intracellular and high extracellular CI-. It is im-
portant to realize that in determining an expected
Donnan equilibrium in a tissue only the diffusible
ions count. Thus, Na+ figures only to the small extent
that it penetrates the cell membrane.
electrochemical gradients and membrane po-
tentials. It is beyond the scope of this article to at-
tempt to deal with this fascinating subject in any sort
of detail. Since our discussion of ions and smooth
muscle must make frequent reference to bioelectric
potentials, however, a brief outline will be presented.
Two different dilutions of a substance A' and A"
have necessarily different activities and different
chemical potentials stored in them. This may be
expressed as the change in free energy required to
move one mole of A' from the lower to the higher
activity or, conversely, the amount of free energy
liberated when A" slides from higher to lower activity.
This relation follows the very general form
(A)'
(A)
AG can be expressed in electrical terms as volt-cou-
lombs and factored into a potential difference E and
nF faradays so that :
F_RT .(A)'
E — P ln 777"
nF (A)
At 25 C this simplifies to
E! * l>(Na*Jj * P
j 50
>
I
0 J
: 1.5 1
Brachial Artery
Small Artery
Small Vein
Cephalic Vein
Total
10 - Small VeBfit-l
s
i
pi
Arteries
Veins
MHmMWHH«M»l»B»tt«Mmwitr-»VWHJM»
1478
155.5 1685
mEij Na/L
fig. 4. Relation of serum [Na+] in venous outflow to % as-
cular pressures and resistances in the dog forelimb. Average
of 1 1 animals. Ten per cent NaCl infused into brachial artery
at 0.0, 1.0, 2.3, and 0.0 ml/min in that order. [From Haddy
(101).]
series of experiments and observed that even fairly
small additions of Na produced vasodilatation.
Marshall & Shepherd (148), in a beautiful series,
used the newly developed ultrasonic flowmeter to
monitor flow in the femoral artery of the dog and
studied changes in limb resistance in response to Na
salts. They found that the rapid injection of 2 ml of
10 to 20 per cent NaCl produced vasodilatation. The
same effect was obtained with a continuous infusion
of the same solutions at 2.3 ml per min. A series of
other salts (citrate, lactate, bicarbonate, acid and
alkaline phosphate) all produced the same effect.
Binet & Burstein (12) obtained similar results in the
dog limb isolated from the circulation of the rest of
the body, except as perfused under control from the
systemic circulation through a constant pressure
pump.
Haddy and his associates have used perhaps the
most detailed and elegant technique to study this
same problem. Basically, these workers perfuse the
forelimb of the dog with arterial blood rerouted from
the femoral artery. The connecting polvethylene
tubing uniting the femoral to the brachial artery
passes through a constant output pump which operates
by intermittent compression of the tubing. Pressure
1 148
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
lhiour*d ur«r rcspiRa
TION -W\
200 —
r
9^6
^ t E
o a. c
Jr o ~- 200
'W
o1--' 2 p/ ^ofc tea
info Ltff Femoral Arlfrr
^J ;
5 25 45 65
fig. 5. A: changes in renal vein flow, FRy, and venal artery pressure, PRA, with perfusate composi-
tion changes. TRV = temperature of renal vein outflow. [From Harvey (104).] B: effect of hypertonic
saline solution on blood flow through femoral artery measured with ultrasonic flowmeter. Note the
zero calibration check at the beginning of the record and again during the period of increase in How.
The fall in flow rate at time of injection is an artifact. [From Marshall & Shepherd (148).]
is monitored by means of fine cannulae passed into
the vascular tree at several points distal to the pump
so that a detailed description of the resistance of the
various segments of the circulation can be compiled.
Haddv (101), too, found that amounts of NaCl,
insufficient to alter systemic pressure, produced
arteriolar dilatation in the dog forelimb.
Unfortunately, it is not possible to elevate plasma
Na concentration in the perfusing blood without at
the same time raising its tonicity. Marshall &
Shepherd (148) first noted this as an experimental
defect and found that they could obtain a similar
degree of vasodilatation by infusing dextrose and
urea matched for tonicity to their sodium salts and
concluded that the mechanism of vascular relaxation
was uncertain. Muirhead et al. (154) had earlier
noted this effect of hypertonic infusions. This point
led Overbeck & Haddy (156) to restudy the problem.
Thev found that hypertonic solutions of NaCl,
Na->S04 and Na>HP04, which produced the same
final serum Na concentration, evoked decreases in
limb vascular resistance in parallel with their actual
tonicity. Equally hypertonic infusions of Na>S04,
and NaCl, irrespective of the amount of Na supplied,
evoked equal decreases in small vessel resistance.
They concluded that the addition of Na apparently
had little or no independent effect apart from that of
its tonicity. On the other hand, these same workers
(102) have obtained some evidence that a reduction
of Na in the perfusing medium is slightly vasocon-
strictive and decreases the caliber of the small vessels.
Harvey (104) and Read et al. (164) have studied the
effects of hypertonic solutions on the renal vascular
bed and have arrived at the same conclusion regarding
the relation of hypertonicity of the infusion and its
vasodilators- effect.
These findings are in substantial agreement with
those obtained with vascular and other smooth
muscle strips studied in vitro. It will be recalled that
following exposure to high Na„ such tissues may show
a temporary increase in tension, but very high levels
are required for this. On the other hand, after equili-
bration in such high Na media, tension is usually lower
than normal and responsiveness of these tissues is
reduced, and this obtains in vivo as well. Thus,
following perfusion of the rat with Na salts we (86)
found the blood pressure responses to norepinephrine
and to Pitressin sharply reduced. Haddy (ioi), in
more precise studies, has shown this to be due to a
reduction in the responsiveness of peripheral vessels
to both pressor and depressor agents. Since this effect
persists, it max not be directly related to the osmotic
effect of the hypertonic solution.
It should be emphasized that an osmotic effect
produced by infusing a hypertonic solution cannot,
in fact, be dissociated from an ionic effect. The
withdrawal of water from cells causes a proportionate
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE
"49
increase in both Na, and K.,. We shall deal with the
quantitative aspects of this type of shifi later.
effects of K. infusions. Attempts to link the effects
observed in vitro to in vivo responses of vascular
tissue have necessarily been limited in scope. This
again reflects the technical problems limiting such
exploration.
Mathison (149) long ago produced what is still one
of the best pieces of work on this subject. He injected
a few milliliters of M/7 (isotonic) KC1 into the arterial
circulation of the cat. A rise in blood pressure followed
at once; this was also obtained in the decerebrate,
spinal, or spinal-pithed animal. The effect was not
of cardiac origin and was only partly due to excitation
of vasomotor centers, since a considerable rise of
pressure was still obtained after ergotoxine. This
author paid careful attention to tonicity, anion
control, and pH. He pointed out also that peripheral
vascular relaxation rather than constriction followed
the injection of less concentrated solutions beginning
at mil Hoff et al. (115) obtained much the same
result.
Hazard & Quinquaud (108) carried out an ex-
ceedingly nice pharmacological study of the pressor
effect of intra-arterially injected K.C1. They showed
that a large part of the effect was due to adrenal
medullary discharge. Then, using a series of blocking
agents, they came to the conclusion that a significant
part of the vasoconstrictive action of K was directly
exerted on vascular smooth muscle.
McKeever and associates (146) perfused the left
coronary in the dog with blood from a donor animal,
interposing a constant output pump in the line
(cf 101). K.C1 was then added at constant rate to
raise plasma K from 2 to 20 meq per liter. Except at
the very lowest concentrations the infusions produced
a transient dilatation of large and small arterial
segments lasting for about 1 min, followed by a more
sustained constriction. Still higher concentrations were
entirely constrictive, but the degree to which an
adrenal discharge may have contributed was not
assessed.
Emanuel et al. (58) have carried out a careful
analysis of the changing pattern of resistance in the
different segments of the dog forelimb during infusion
of K. salts and have correlated their findings with
systemic and local measurements of serum Na and K.
Small vessel resistance decreased at all infusion rates.
By contrast, arterial resistance did not change at
lower rates and then, as the rate increased, gradually
began to show an increase. The net effect of these
changes was an over-all fall in resistance at low rates
and a rise at higher ones. The primary phenomenon
held for increases in serum K up to about 8 meq
per liter, at which point the secondary net constrictive
effect appeared. The arterial constrictive phase may
be in large part related to adrenal discharge. These
results were similar for the chloride, nitrate, lactate,
and acetate. They applied equally well to the renal
vascular bed (178). Even these moderate elevations
of K„ reduced the sensitivity of the peripheral vas-
culature to challenging doses of pressor and depressor
agents.
The work of this group satisfactorily explains the
phase of falling peripheral vascular resistance noted
by all authors who have infused small amounts of
KC1. It also shows that K does not produce smooth
muscle vasoconstriction in the physiological range
and agrees in this with in vitro studies. It leaves
unexplored, and correctly so, the effect of high,
unphysiological amounts of K which are vasocon-
strictive in vitro. This latter point has great theoreti-
cal importance if a change in membrane potential
is involved in peripheral vasoconstriction. Un-
fortunately, the experimenter cannot explore the
problem in vivo, for although high K infusions do not,
like Na, raise problems of osmotic pressure they do
produce adrenal, cardiac, and nervous effects which
presently defy rational interpretation.
Speaking critically, the perfusion of regional
vascular beds is not a totally satisfactory approach to
the problem. Technically, many of the procedures
give detailed information concerning the responses of
each segment of the vascular bed and for this are
most satisfactory. The problem resides not in this
facet of the approach but in the attempt to alter a
single variable in the medium while still perfusing
with whole blood. Such a situation cannot be fully
controlled. Using the Na and K electrodes we have
found many times that the solution we thought was
presented to the cells was not the same as the solution
the cells actually met. To interpret such perfusion
data fully requires information about the Na, K, and
Ca levels actually attained, together with an estimate
of pH.
Evidence from Meamrement of Xa and A in
Relation to Blood Pressure
measurement of Na and K in chronic hypertension
or hypotension. Deoxycorticosterone (DC A) hyper-
tension. Ledingham (137, 138) studied the relation of
Na and K. partition to blood pressure in a series of
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
DCA-treated rats using inulin to estimate water
distribution. Unfortunately, he analyzed only skeletal
and cardiac muscle and hence his report of an absence
of correlation must be treated cautiously. Despite his
own rather negative conclusion, if we assume that his
methods can only distinguish extremes and not
transitional stages, his tables show several striking
features. Blood pressure in the control-adrenal-
ectomized group averaged 86 mm Hg, while the
adrenalectomized saline-treated group reached 114
mm Hg. The two groups attaining the highest
pressure were likewise adrenalectomized and received
either DCA-saline or DCA-saline-cortisone and had
average blood pressures of 1 9 1 and 221 mm Hg,
respectively. Both Na» and Na, values were highest
in these groups and the Na0/Na, gradients lowest.
Again, K„ was lowest in both these groups, while K,
did not fit any pattern. The value for K„ is perhaps
particularly important, since it has also been reported
as a finding in the accelerated phase of essential
hypertension in man (111, 151). Low levels of serum
Na have also been noted at this stage of the disease
(46).
Woodbury & Koch (211) also noted in the rat that
DCA and aldosterone produce an increase in skeletal
muscle Na which, judging by measurements of
chloride space, is largely intracellular and represents
an increase in Na,. Potassium was little altered.
Unfortunately, blood pressure was not measured.
Ferrebee et al. (62), using more laborious techniques,
had long ago shown that DCA in the dog caused a
gain in Na, at the expense of K,. Cier and co-workers
(29) and Gross & Schmidt (99) also claimed that
DCA increases Na, in skeletal muscle.
Insofar as blood pressure regulation is concerned
it is more important to attempt to estimate the effect
of DCA on Na and K. in a representative of vascular
tissue. The aorta has so far been the only tissue
amenable to study and here, by and large, the
evidence points the same way. The classic and most
quoted paper in this field is that of Tobian & Binion
(196), who reported an increase in both Na and K
in the aorta of rats made hypertensive with DCA.
From estimates of the extracellular space (based on
chloride measurements) they believed that most of
the increase represented a true intracellular gain.
Tobian & Redleaf (199, 200) reaffirmed these
findings in later studies.
Daniel & Dawkins (40) claimed that aorta electro-
lyte changes demonstrable in early DCA hypertension
disappeared later in the disease. In early hypertension
they noted a tendency for a gain in Na, but, more
significantly, for K depletion in hypertensive rats
under treatment compared with normotensive also
under treatment. Most recently, Laszt (133) also
reported a gain in aorta Na following DCA treat-
ment.
By and large it would seem that DCA causes a gain
in Na in both skeletal muscle and in aorta in the rat;
a goodly part of this gain probably represents a true
intracellular increase, but with the methods so far
used it is difficult to assess just how much is actually
intracellular. The gain in Na is apparently not
accompanied by a parallel gain in water, so we are
fairly safe in inferring from all authors that intra-
cellular Na concentration, Na„ is elevated. All reports
uniformly fail to provide us, however, with simul-
taneous estimates of Na, of the aorta and Na„ of the
medium, which is a crucial piece of information. This
same lack of information tends to nullify Laszt's
claim (133) that blood pressure cannot be related to
the total Na content of the aorta. In our opinion, it
would be exceptional if blood pressure could indeed
be related to one such simple parameter as that.
Other hypertensive states. We have considered DCA
separately because of its obvious effects on electrolvte
metabolism which might perhaps be considered to
make it a special case. We turn now to a more general
review of the findings for Na and K analysis in a
divergent series of conditions united only by the fact
that a sustained hypertension is a common feature.
Ledingham (139) has recently reviewed this and
decided that the onlv common feature in these varied
states is the elevated blood pressure itself. His negative
view is based on his findings in hvpertension induced
by DCA, cortisone, renal arterv constriction, and
bilateral nephrectomy (136-138). As described above,
he may have overemphasized the negative aspects of
his data. It is unfortunate too that this conclusion
should have been arrived at without any attempt to
measure electrolytes and water in vascular tissue.
In their original report Tobian & Binion (196)
considered renal as well as DCA hypertension in the
rat. Aorta Na and K were both elevated following
renal constriction in rats developing hypertension
compared to animals remaining normotensive after
the same operation. Tobian (192) extended this work
by using a low sodium diet to control the blood
pressure rise of animals with renal constriction and
found the changes directly related to the presence or
absence of hypertension.
More recently Tobian & Redleaf (200) found that
rats with post-DCA sustained hypertension, with
adrenal regeneration hvpertension, and with the
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE
hypertension that may persist after excision of an
ischemic kidney all show an increase in aorta Na and
K.. This increase seemed to them to represent a true
gain in Na, and K,. There is a suggestion that these
studies may be pertinent to the problem of hyper-
tension in man, for Tobian & Binion ( 1 95 ) found an
increase of both Na and water in the renal arteries of
human hypertensives. Na was increased more than
water, but the technique used cannot distinguish
intracellular from extracellular locations. Not too
much weight should be given this type of stud\ ,
however, for as all workers in the field know, electro-
lytes may exchange rapidly across vessel walls both
immediately preceding and certainly after death.
In general, experiments in many laboratories
support the thesis that Na and K in tissues are altered
in hypertension, but the emphasis shifts now to the
one, now to the other. Thus, Eichelberger (53) long
ago measured an increase in Na and fall in K in the
skeletal muscle of dogs made hypertensive by renal
constriction. Assuming an extracellular position for
chloride there appeared to be a true rise in Na, and
fall in K,. Laramore & Grollman (132) found a
general rise in tissue Na and water and a fall in K
in the later stages of renal hypertension in the rat.
Later, however, Grollman (96) found the same
quantities unchanged in hypertension produced as a
late sequela of choline deficiency. More recently
Kolctsky et al. (i2g) analyzed the mesenteric arteries
of rats with acute renal hypertension and also found
a gain in Na, K, chloride, and water. These authors,
however, cautiously refrained from attempting to
partition the electrolytes on the basis of chloride
space.
In agreement with Tobian and associates, Freed
et al. (70) found an elevation in aorta K in renal
hypertensive rats but a less well-defined increase in
Na. The reduction of the hypertension by dietary K
deprivation was followed by a proportionate decrease
in aorta K, and the return to hypertensive levels on
refeeding K was accompanied by a return rise in
aorta K. On examining the data, however, it is
evident that the increase in aorta K occurs in rats
with renal constriction whether or not the pressure
goes up. This same inconsistency was noted by
Tobian & Binion (196).
Laszt (133) does not find an increase in aorta Na
at all consistent with the presence of hypertension in
rats, but claims the rise of K to be so.
Houck (119) pointed out that dogs maintained in
good balance for 5 to 1 1 1 days following bilateral
nephrectomy show a gain in tissue Na and fall in K
despite relatively normal extracellular values. This
apparently indicates an association of their sustained
hypertension with an elevation of Na, and fall in K.,.
Greene & Sapirstein (95) found an increase in total
body Na in rats made hypertensive by subtotal
nephrectomy. Haight & Weller (103) studied rats
made hypertensive by chronic high salt feeding;
they found hypernatremia and an increase in skeletal
and heart muscle Na and K. especially pronounced in
the hypertensive rats but no conspicuous differences
in aorta electrolytes.
Other sustained abnormal blood pressures. The regularly
observed fall in blood pressure in Addison's disease or
following adrenalectomy requires no comment.
Among other phenomena, it is associated with a
reduction of tissue Na from both intracellular and
extracellular compartments (30, 52).
Freed et al. (69) examined the relation of plasma
to aorta Na and K in rats made hypotensive by K
deprivation. Both Na and K declined in plasma as
well as in the aorta, although the authors stress only
the change in K. Tobian (191) found that rats on a
low Na diet lose a sizeable amount of aorta Na, while
serum Na may actually rise a little. Although blood
pressure was not measured it was assumed to tend
toward lower values.
Trauma may lead to a "posttraumatic sodium-
potassium shift" during which plasma Na falls and
K rises. This is associated with hypotension (186).
MEASUREMENT OF Na AND K IN ACUTE HYPERTENSION
or hypotension. Although investigation of the
association between acute changes in blood pressure
and electrolyte exchanges is comparatively recent,
the findings are more clear-cut than any we have so
far considered. The independent findings from
different laboratories all fit together nicely even
though interpretations vary. Since this approach
bears directly on the relation of ions to vascular
smooth muscle tension, the facts obtained must form
the basis of all theoretical discussion. Accordingly,
we shall present these facts in some detail.
Perhaps the earliest report of a relation between
ionic concentration in the serum and a pressor agent
was that of D'Silva (50) in 1934. He reported that,
in the cat, following the intravenous injection of 50
/ig of epinephrine or 10 units of Pitressin intra-
venously serum K. rose sharply as much as 3 meq per
liter within 1 min. Since these massive doses of the
order of 10 fig per kg epinephrine and 2 units of
Pitressin per kg also caused a rise in blood sugar, this
author related the K rise to the glycogenolytic effect
1152
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
IWJECTIOM
(.005 mg/>g EPIWEPHRIIIE)
CYCLOPROPANE
NEMBUTAL
UNANESTHETIZED
D— O ETHER
40 SO
imecTioi (b)
i iv I ^
VT
JOIVEIS.
CYCLOPROPANE 30 MINUTES
EPINEPHRINE
•— • ARANTHOL
a — □ NEOSYNEPHRIN
O— O EPHEORIBE
SECONDS
40 80
120
110
240
300
fig. 6. A: influence of anesthesia on plasma [K+j rise induced
by epinephrine in the dog. Arrhythmias and time of their
occurrence shown by horizontal lines at top of figure. VT
= ventricular tachycardia. B: effect of a series of sympatho-
mimetic amines on plasma [K+] in the dog. VT = ventricular
tachycardia, NVEXS — numerous ventricular extrasystoles.
[From O'Brien et al. (155).]
of the injections rather than to the blood pressure
effect. This explanation for the rise of K remained
unchallenged and uncritically accepted for almost two
decades.
O'Brien et al. (155) in 1953 obtained beautiful
curves showing the rise of plasma K after rather lower
doses of norepinephrine, 5 ng per kg, in dogs. Although
norepinephrine does not have any marked effect on
the mobilization of glucose and although these authors
pointed out that it is blocked by Dibenamine, which
would not block a glycogenolytic effect, the original
explanation still persisted. O'Brien and his associates
also pointed out that the choice of anesthetic modified
the effects, ether being the worst for blurring the
effect, cyclopropane affecting it least, Nembutal
almost as good as cyclopropane. These important
observations were unfortunately ignored by most
later workers including ourselves. (In our more
recent studies we have found the effects considerably
sharpened if a barbiturate mixture is used instead of
ether. )
Muirhead et al. (153) in the next year restudied the
problem. They gave norepinephrine by infusion in
total doses of 1 to 7 mg per kg in 50 to 180 ml of
saline over periods ranging from 20 to 50 min. Con-
cerning their results they wrote: "In many of the
experiments the sodium curve represents an approxi-
mate mirror image of the blood pressure curve. In
most of the experiments the changes of potassium
were not as pronounced as those of sodium. In
addition there seemed to be little if any correlation
between blood pressure and plasma potassium. The
latter is in contrast to the variations in sodium levels
which reflected even sudden changes in blood pres-
sure." No real change in radiosulfate space or in
chloride concentration was observed. These observa-
tions did not have the impact they ought to have had
for the doses used were very large.
Tobian & Fox (197) then approached the problem
directly and analyzed segments of dog femoral
artery before and after infusing norepinephrine
sufficient to maintain a blood pressure elevation for
30 min. A consistent fall in artery K and a less con-
sistent gain in Na was observed. These changes were
believed to represent a fall in K, and rise in Na„ but
the authors tended to underestimate the importance
of the Na change since it was less consistent. In our
view, the difficulties inherent in such tissue anah sis
and their basic range of error make it highly sig-
nificant that even this trend for a sodium increase was
observed; in fact, a sizeable Na gain was found in 9
of 12 dogs. In the light of later evidence, we must
conclude that a fall in K, and a real gain in Na„
probably with water, actually did occur.
Shortly thereafter, following up the possibility
that Pitressin might have an extrarenal action, we
observed that this agent caused a shift of water out
of the extracellular space in the bilaterally nephrecto-
mized rat. In a more detailed study, we found that
Na also left the extracellular space in association
with, but in excess of, water so that there was a
measurable fall in plasma Na concentration (82). The
relation of dose to response was presented at this time
and the correlation of the shifts with blood pressure
noted. In interpreting these exchanges of salt and
water, we took into account the changes which
other workers had previously observed to follow
norepinephrine administration and hypothesized thai
both sets of observations could be related by a general
rule that blood pressure regulation depends on the
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE
I 153
sodium transfer systems, broadly defined. The rise in
plasma K following Pitressin administration was
considered an integral part of the phenomenon, but
left aside from this first theoretical approximation for
later consideration.
In these first studio uc overestimated the absolute
magnitude of the water shift that follows Pitressin
administration. This was corrected in subsequent
studies. The effects of norepinephrine on Xa, K, and
inulin space in the nephrectomized rat were then com-
pared with those produced by Pitressin and again a
net loss of extracellular Na was observed in association
with the blood pressure rise (75). Pitressin in the large
doses used caused a measurable fall in plasma Na
indicating that Na moved in excess of water. With the
techniques at hand (flame photometry, arterial blood
sampling) no clear fall in plasma Na concentration
taken alone was observed with norepinephrine, but a
shift of Na and water was claimed on the basis of
replicate experiments in which extracellular sodium
was calculated. Angiotensin was then compared with
Pitressin in the nephrectomized rat (76). Both agents
given intravenously produced a measurable fall in
plasma Na concentration and a fall in extracellular
fluid volume (inulin) associated with the rapid rise
of blood pressure. While Pitressin produced a meas-
urable increase in extracellular K, angiotensin did
not.
It was clear at this time, at least for norepinephrine,
that the fall in Na and water and the rise in K which
we were measuring in the extracellular compartment
of the rat corresponded qualitatively to the gain in
Na and loss in K which Tobian & Fox (197) had
measured in the femoral arterv of the dog. Daniel et
al. (41 ) then injected a pressor dose of norepinephrine
within the physiological range, 1 /xg per kg, in the rat.
They found that the aorta was rapidly depleted of K
while Na tended to increase. This is surprisingly good
confirmation in view of the fact that the Na shift is
probably partly obscured by a movement of water
which these workers could not measure.
Subsequently, this group (42) studied the effects of
Pitressin and isoproterenol (isopropyl norepinephrine,
a peripheral vasodilator) on aorta electrolytes in the
rat. They concluded from the variations in aorta
sodium that during blood pressure changes, Na
moves into (rising blood pressure) and out of (falling
blood pressure) vascular muscle cells. Since they were
dealing with the aorta, an outward movement of K
occurred only with those drugs known to cause an
aorta strip to contract. They pointed out, however,
that the total amount of Na which we had reported to
leave the extracellular space could not possibly be
accommodated within the cells of the vascular tree.
This difficulty has now been satisfactorily resolved by
our observation that skeletal muscle also takes up
sodium under the influence of Pitressin (85).
In the rat, studies of changes in plasma Na, K, and
inulin space during changes in blood pressure are
technically difficult, since each step in drug, dose, or
time interval requires the use of separate groups of
animals. To circumvent this, as well as to extend the
observations to the dog, we studied the problem in the
bilaterally nephrectomized dog using norepinephrine,
isoproterenol, angiotensin, and Pitressin (73). We
found that the calculated extracellular Na (product of
inulin space and plasma Na) declined as pressure rose
and increased as it fell; the two measurements con-
sistently formed mirror images. Calculated extracellu-
lar K in general moved inversely to Na and hence in
parallel with the pressure except in the case of angio-
tensin where, as in the rat, no K shift was found.
In the case of norepinephrine, the simple measure-
ment of plasma Na was an inconsistent index of Na
movement, since the real decrease in this ion is
partially masked by a movement of water in the same
direction. For the same reason, K concentration is a
consistent but inaccurate estimate of K movement,
since the change is magnified by inverse movement of
water. In the case of Pitressin, although both Na and
water move out of the extracellular compartment, the
Na shift is well in excess of the water so that, if the
dose is adequate, a fall in plasma Na is readily ob-
served. These findings are remarkably similar to those
obtained in the rat.
Warren (205) has recently studied the effect of
Pitressin on Na, K, and inulin space in the trained,
conscious, intact dog. He observed similar exchanges
to those previously reported in the nephrectomized
dog even though he used considerably smaller doses
of Pitressin (30 mU kg as a single i.v. injection versus
200 mU/kg/min for 10 min by infusion).
Recently the Na and K. electrodes have been ap-
plied to this problem. In the first experiments we used
only a sodium electrode interposed into the femoral
artery of the dog (80). The aim was to determine
whether pressor and depressor agents actually shift Na
levels as blood sampling procedures indicated. The
result was unequivocal; the pressor response to nor-
epinephrine, epinephrine, and angiotensin was regu-
larly accompanied by a fall in electrode potential
indicating a fall in sodium concentration, or more
precisely, sodium activity. In terms of degree of
change, time course, and duration of effect, each agent
1 1 54
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
DOG 9.
ISOPROPYL
NOREPINEPHRINE
J Iy /Kq/min.
« 2 »
40 50
TIME — MINUTES
80
90
fig. 7. Changes in extracellular Na and K associated with blood pressure changes in the dog.
[From Friedman et al. (73).]
produced its own characteristic pattern. The depres-
sor response to acetylcholine, histamine, and isopro-
terenol was accompanied by oscillations in the tracing
which tended to be inverse to those observed with
pressor agents. Later, in a similar arrangement using
two electrodes, norepinephrine was shown to produce
a rise in (K+) inverse to the fall in (Na+) (81 ).
The technique for electrode monitoring of (Na+)
and (K+) in flowing blood was then modified after
Haddy so as to control flow rate through the electrode
as well as through the vasculature of the limb (122).
The femoral artery of the dog was interrupted by a
length of polyethylene fed through a Sigmamotor
pump. The femoral vein was similarly lengthened and
passed through a smaller division of the pump. The
venous outflow passed through Na and K cannula
electrodes in a shielded enclosure. This arrangement
ensured not only a constant limb inflow but also the
passing of a proportion of the venous outflow at
a constant rate past the electrodes. Quantitative
measurements could also be made, since calibrating
solutions could be injected into the venous tubing
proximal to the pump. One pressure transducer was
inserted on the arterial side between pump and limb
and another into a brachial artery. Small amounts of
vasoactive agents sufficient only to activate the limb
vasculature without producing any noticeable sys-
temic effects were used.
In general, limb vasoconstriction induced by nor-
epinephrine or epinephrine was associated with a fall
in blood (Na+) and often with a rise of (K+). Larger
doses tended to produce a biphasic response in (Na+),
that is, an initial transient rise preceding the fall.
Vasoconstriction produced by serotonin or angioten-
sin was associated with similar (Na+) change unac-
companied by any consistent (K+) deflection. Vaso-
constriction produced by Pitressin was associated with
a fall in (Na+) and consistent rise in (K+), both notice-
ably greater in degree and duration than with other
agents producing an equal degree of vasoconstriction.
Limb vasodilatation induced by isoproterenol, acetyl-
choline, or histamine was accompanied by a rise in
blood (Na+) without any consistent change in (K+).
A full analysis of rates and relations of ion and water
movement is clearly required. For the moment, we
mav conclude that the movements of Na and K.
associated with changes in blood pressure reflect
changes in tension in the peripheral blood vessels.
Onlv one report disturbs the general consistency
of this phase of the investigation. Headings et al. (109)
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE I 1 55
""" man
8.M9 NOREPI
14 kg cf
fig. 8. Changes in blood (Na+) and (K+J monitored with the Na electrode (Na/K = 250/1)
and K electrode (K/Na = 5/1) in the dog during A, systemic blood pressure rise induced with
norepinephrine and B, limb pressure rise induced with norepinephrine. [.4, from Friedman et at.
(81).]
found that dog carotid artery rings stimulated electri-
cally gained Na and lost K. Epinephrine, however, in
an amount sufficient to produce the same contractile
response did not produce these changes.
We may conclude that, in general, an acute in-
crease in tension of vascular smooth muscle is asso-
ciated with a gain in Na, and a gain in water. There
is a strong suggestion that in some instances, at
least, the gain in water may overshadow the gain in
Na and may also anticipate it. A loss in K from cells
to environment is almost always observed. Similar
ionic exchanges have been observed both in taenia
coli and uterus during activity (16, 124).
These experiments give no information regarding
the time relations connecting these phenomena nor do
they suggest which event is cause and which is effect.
Evidence from Studies of the Relation of Electrical
Activity to Tension in Vascular or Analogous Tissues
Bacq & Monnier (7) studied the relation of electri-
cal activity to tension in a variety of smooth muscles
obtained from the cat. They claimed that the laws
common to all excitable tissues apply to smooth
muscle as well. In their view the response to every
excitation, in this case contraction or increase in tone,
no matter how produced, is accompanied by a de-
crease in polarization. They considered the change in
polarization to be the cause of the change in tonus.
In accordance with theories current at that time de-
polarization was attributed to the exit of K+ from
cells.
Although Bozler (17) carried out the first basic
studies of electrical activity in smooth muscle using
modern techniques it remained for Bulbring and her
associates to carry out the difficult task of denning the
ionic basis of that activity. In 1954 data were pre-
sented for guinea pig taenia coli suggesting that
tension is inversely, and spike frequency directly, re-
lated to the membrane potential (19).
We can summarize these first experiments in a
simplified form. A resting membrane potential of 60
± 9 mv fell to 43 ± 10 and spike frequency increased
when the tissue was stretched (increased tension).
Histamine induced a fall in potential from 58 to 40 mv
while tension and spike frequency increased. Epi-
nephrine induced an increase in potential and a de-
crease in tension and spike frequency. Acetylcholine
caused a fall in potential and increase in spike fre-
quency and tension. Shortly thereafter Bulbring (20)
reported that the increase in rate of spike discharge
was proportional to the increase in tension. Then,
in 1 955 (2 1 ), fluctuations in membrane potential were
observed to be related to the spontaneous rhythm of
the taenia coli strip and periods of depolarization as-
sociated with increased tension and increased rate of
spike discharge alternated with periods of repolariza-
tion, reduced spike frequency, and lower tension.
From this basis Born & Bulbring (16) then pro-
ceeded to the still more difficult technical problem of
1 1 56
HANDBOOK OF PHYSIOLOGY
( IRCI1.ATION II
120"
o 100-
o
12 3 4 5 12 3 4 5 6
fig. g. Fluctuations of tension {broken line) and of radioactivity (K.12) appearing in washing solu-
tion (continuous line) of taenia coli during spontaneous activity .4, before. B, in the presence of atro-
pine 2 X io-6. [From Born & Biilbring (16).]
--110
--90 m
3
o
adding measurements of ionic exchange to the simul-
taneous monitoring of tension and electrical activity.
By restricting their attempt to the simpler problem
of measuring only K they succeeded quite elegantly
and in so doing proved beyond doubt that basic ionic
theory is generally valid, at least for the gut strip.
They used K42 as tracer and analyzed the medium
flowing past the tissue. Spontaneous activity was
characterized by the parallel rise and fall of K efflux
and tension so that as tension rose, K efflux increased
and as tension fell, K efflux decreased. Similarly,
histamine and acetylcholine produced contraction
associated with a parallel increase in K efflux. Epi-
nephrine, which causes relaxation of this particular
preparation, seemed to do so by increasing K influx.
The fall in membrane potential previously observed
to parallel the increase in tension was evidently asso-
ciated with an ionic shift here measured as K efflux.
Presumably, if it had been technically feasible to
measure, a primary sodium influx would have been
recorded.
Born (15) then turned to a study of some of the
metabolic problems concerned with contraction in
smooth muscle and for the first time we find a firm
separation of the relatively rapid changes in tension
from the maintained changes which we recognize as
tonus. The position is best stated by the author: "The
development of tension by smooth muscle involves
two mechanisms. One mechanism is responsible for
the immediate rise in tension which occurs when the
muscle is stimulated and this mechanism continues
to function in anoxia and in the presence of 2:4
dinitrophenol. The other mechanism is responsible for
the sustained tension which the muscle shows, both
spontaneously and following stimulation. This mecha-
nism is abolished when metabolism is interfered with,
e.g., by depriving the muscle of glucose or of oxygen,
or by exposing it to 2:4 dinitrophenol."
Later studies of electrical activity have rather
tended to cloud the picture. Biilbring & Ltillman (22)
demonstrated that spike frequency and tension could
be dissociated. Using dinitrophenol they showed that
spike frequency could be made to increase or decrease
without particular reference to tension changes. The
inverse relation of tension to membrane potential still
held under these circumstances, however, so that at
this point we might tend to disregard spike activity.
This idea is reinforced by Holman's observation (117)
that the addition of KC1 to the medium bathing a
taenia coli strip increases tension and decreases the
membrane potential. Further, at concentrations above
20 meq per liter, the relation of K„ to membrane
potential is linear with a 33 mv slope per log unit
change, an important fit with ionic theory although
the low slope remains to be accounted for.
Burnstock & Straub (26) using an improved pro-
cedure, the sucrose-gap technique, verified the fact
that K salts produced a membrane depolarization,
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE
I]57
showed the importance of penetrability of the ac-
companying anion, and improved the relation of E„,
to log K„ to yield a slope of 45 mv. This is still lower,
however, than the ideal 58 mv which one would
expect.
Holman (116) found more evidence to relate E„,
and tension. She reported that raising Na„ to two or
three times the normal level first increased tension and
spike rate while reducing Em. Later, as exposure was
prolonged, the spikes disappeared although tension
remained high and E„, low. These observations again
argue for dissociation of spikes and tension, but later,
Holman (118) took the position that spike frequency
was an important factor in the development of ten-
sion and Axelsson (5) sustained this view.
The question of whether or not tension changes
and electrical spikes are interdependent is of particu-
lar importance to our problem, since vascular tissue,
so far as it has been studied, shows no spike activity
whatever. Accordingly, tissues such as taenia coli are
relevant only insofar as their tension may correlate
with electrical activity other than that of a train of
action potentials constituting the spike activity. Burn-
stock's recent study and argument is thus particularly
important (24). He noted that smooth muscle of
guinea pig taenia coli is relaxed by epinephrine and
that this is associated with a rise in Em and decrease
of spike activity. The smooth muscle of the muscularis
mucosa of the dog, by contrast, is contracted by
epinephrine and in this case the epinephrine effect is
associated with a fall in E„, and an increase in spike
activity. It would therefore appear that at least for
these types of visceral muscle both the membrane
potential and the spike activity are correlated with
tension. Under certain circumstances it is possible to
dissociate the spike activity from the tension and the
membrane potential alone then remains inversely
correlated.
A similar situation arises in connection with studies
of the uterus. The observation of Woodbury & Mc-
Intyre (212) that oxytocin, which contracts the
pregnant uterus, reverses the membrane potentials of
the single muscle cell is quite relevant for it again
relates membrane potential to activity. On the other
hand, the claim (43) that tension of the uterus strip is
only correlated with spike activity suggests that this
tissue is not analogous to vascular smooth muscle.
Electrical studies of vascular smooth muscle were
almost nonexistent until very recently. Bozler (17)
considered this type of muscle to be distinctive in
being "multi-unit" unlike many other types which
behave like single units. Recent detailed and elegant
studies by Burnstock & Prosser (25) and Prosser et al.
(161) have placed this on a firmer footing. Vascular
smooth muscle is here shown to consist of widely sepa-
rated cells and its extracellular space calculated from
electron microscopy is about 40 per cent of the total
(see table 4). This contrasts strikingly with the other
types of smooth muscle in which the extracellular
space is estimated at less than 20 per cent. Vascular
smooth muscle shows no conducted electrical activity
and no spikes. Our theoretical discussion must reckon
with these distinctive features.2
ROLE OF CALCIUM AND MAGNESIUM IN
VASCULAR SMOOTH MUSCLE TENSION
Calcium
There is too little information concerning the de-
tailed effects of calcium and magnesium on vascular
tissue to permit any elaborate discussion. What little
evidence we do have is fortunately consistent. The
older literature has been reviewed by Evans (60). A
much larger literature deals with the general direct
involvement of calcium in the actomyosin system (51)
and in the metabolic cycle of cells (141). We shall
not develop this broad field of physiological chemistry,
which would lead us far from our immediate subject,
but we must note that calcium ions are evidently
necessary for the contractile machinery to work.
The physiological implications of this have been
demonstrated by Heilbrunn & Wiercinski (1 10). They
showed that Ca in high dilution injected directly into
the single skeletal muscle fiber caused an immediate
and pronounced shortening. This effect is not shared
by any other ion normally present in any quantity
in muscle, but it is also produced by Ba. These au-
thors, like others since (142), support the view that
Ca links the ionic processes at the membrane to the
contractile mechanism. This point obviously has as
much importance for contraction and tonus in vascu-
lar as in any other muscle tissue.
In studies of intestinal segments there seems to be
general agreement that the addition of Ca to the
medium increases tone (189, 208). More important
to our thesis is the demonstration that withdrawal of
2 Recent successful impalement of single smooth muscle cells
in turtle aorta and inferior vena cava segments has shown
specialized types of action potential in association with tension
changes. (Roddie, I. C. and S. Kirk. Transmembrane action
potentials from smooth muscle in turtle arteries and veins.
Science 134: 736, 1961.)
n58
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
' —
MAGNESIUM
SMALL VESSEL
SMALL
VESSEL
r
-VFNOUS
O O Jj? O O ro to O O
§600 666-0
INFUSION RATE ml/mm
fig. 10. Average effect of 10% CaCU or iofc MgS04
infused into the brachial artery on dog forelimb vascular
resistances. [Graph prepared from tabular data in Haddy
(101).]
Ca from the medium causes a dissociation of the con-
tractile mechanism from action potentials in the
guinea pig taenia coli. From this, Axelsson & Biil-
bring (6) have concluded that Ca is essential for the
activation of the contractile mechanism by action
potentials. Hurwitz et al. (121) have shown this in
another equally direct way. They observed that pro-
longed exposure of the guinea pig ileum to a calcium-
free environment divests the tissue of its ability to
contract in the presence of an appropriate chemical
stimulus. Further, the substitution of Mg for Ca in
the medium accelerates the loss of contractility. Even
so the membrane processes governing ionic exchanges
still function so that a stimulus which no longer causes
contraction will still cause K efflux
Zsoter & Szabo (216) have reported that the
feeding of a high calcium diet to rats for 10 to 15 weeks
causes an increased sensitivity of the mesoappendix to
the topical application of epinephrine. As we have
discussed earlier, however, this type of result is
difficult to interpret, since the variable (calcium
feeding) is so remote from the target. Much more re-
vealing is the observation of Haddy who showed that
the infusion of hypertonic calcium salts caused con-
striction of all segments of the peripheral vascular bed
of the dog forelimb under conditions of controlled
flow (101 ). It will be recalled that hypertonic solutions
in general produce vasodilatation so that the result
with calcium is particularly striking. In a later study
Overbeds & Haddy (156) reported that while the
infusion of isotonic KC1 produced peripheral vasodila-
tation, isotonic CaCl2 caused vasoconstriction.
Woolley (213) has suggested that serotonin acts
directly on the cell membrane to transfer calcium
from the exterior to the interior of the cell. His evi-
dence is quite incomplete, however, and a similar
argument could be developed with equal reason to
suggest that most if not all smooth muscle-contracting
agents act through some similar mechanism involving
calcium.
Magnesium
Haury (105) has clearly demonstrated that Mg
relaxes bronchial smooth muscle and opposes the
action of stimulating drugs. In well-controlled experi-
ments in the dog and frog he found that small
amounts of Mg given intravenously produced a blood
pressure fall which was in large part due to peripheral
vasodilatation (106). Schmid et al. (174) carried out a
careful hemodynamic study in conscious dogs and also
concluded that Mg salts produce peripheral vasodila-
tation as Hoff et al. (115) had earlier claimed. Stan-
bury (187) emphasized that the action of Mg is com-
plex, since it produces changes in the autonomic
nervous system and the heart as well as the peripheral
vasculature. Zadina & Kriz (215) claimed that Mg
had a direct relaxing effect on the isolated guinea pig
intestine and depressed the response to stimulating
agents.
Engbaek (59) reviewed the subject in 1952 and
concluded that although it seemed reasonably certain
that Mg ions acted to relax peripheral blood vessels
this had not yet been shown to be a direct effect.
Pending any evidence to the contrary it seems
reasonable to conclude, in summary, that Ca causes
peripheral vasoconstriction and Mg relaxation. In
general, these actions do not appear to be in any
way specific to vascular smooth muscle. The special
role of these bivalent metal ions in the chemistry of
contractile protein may be involved and both ions, or
at least Ca, may link membrane phenomena to the
contractile mechanism.
ROLE OF H+ AND OH- IN VASCULAR
SMOOTH MUSCLE TENSION
This subject is in a highly unsatisfactory state and
permits no real conclusions other than that the pH of
the medium is a most important variable, as we might
have guessed. At this stage in the investigation of ions
most workers are more concerned with maintaining
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE
• ' 59
pH as an invariant than with noting and interpreting
the effects of changes.
Schuler (177) measured the tension of mesenteric
and phrenic artery rings while shifting pH to either
side of normal. He found that tonus increased in both
cases. Tobian et al. (198), using the spiral aorta strip
of the rat, found that the contractile response to
norepinephrine was maximal at relatively higher pH
and minimal at lower. It seems to us that nowhere is
the duration of immersion or exposure to the altered
environment more important than in studies of the
effect of H ions. Reference to table 2 will remind us
of the extremely high mobility of H+ and the ease with
which it penetrates the membrane.
Rogers & Fenn (166) have shown that H+ added
to the medium exchanges rapidly with K+ and Na+
of cells. More recently, Saunders et al. (173) have also
shown a partial replacement of K,+ with H,+ during
dietary potassium depletion. At equilibrium, then,
an original alteration in medium pH is replaced by an
altered Na+ and K+ distribution. Duration of exposure
must then be a critical variable. With this in mind,
we can now examine the findings of workers using in
vivo preparations. Technically, all the procedures to
be quoted are beyond reproach as far as they go.
Burget & Visscher (23) showed a nice decrease in
epinephrine response of the pithed cat proportionate
to a stepwise fall in pH. This accords with Tobian et
al. (198). Fleisch et al. (63) used good techniques to
measure flow and pressure and found that a fall in
pH of as little as 0.05 caused generalized vasodilata-
tion.
More recent studies using the technique of con-
trolled perfusion of a vascular bed are conflicting. Deal
& Green (45), like Kester et al. (127) earlier, reported
that solutions on the acid or alkaline side of physio-
logical neutrality increased blood flow to limb mus-
cles, indicating peripheral vascular relaxation. Skin
vessels, however, showed a decrease in resistance as
pH fell and an increase as pH rose. Fleishman et al.
(64) showed that the picture is complicated by the
fact that small vessel segments constitute independent
resistances the magnitudes of which may actively
vary in opposite directions. The net effects were
dilatation of small vessels with an acute fall in pH and
constriction with a rise in pH. Emanuel et al. (57)
also reported that an acute rise in pH caused an in-
creased peripheral vascular resistance through the
renal vascular bed.
Clearly, we are in no position yet to draw any sort
of general conclusions except that H ion effects are
interwoven with those of Na+ and K+.
~
.
■r-
\
•
• TOTAL
-
!*
0 ARTERIES
18
~"
\/
a SMALL VESSELS
« VEINS
INTACT
NERVE BLOCK
1
1.6
•-/\
NERVE BLOCK
PLUS 1
. PHENTOL AMINE
( •- "
•- /
1 A
V
c
E
1 1.2
r
1 10
a
o
/
a
a'
/
a
\
0
\
•hB
•
/
•
\
••
•
0
z
<
g 8
u,
I ixx-«-rT "
»-J<>x-«~,
1 1
1 1
1
1 1 1 1
1 1 1 1 |l 1 l| 1 1 1 III
10
10
20
20 0 10 20 0
TIME IN MINUTES
fig. 1 1 . Average effects of pH change upon total and seg-
mental vascular resistances in the nerve intact, nerve blocked
and nerve blocked phentolamine dog forelimb. [From Fleish-
man et al. (64).]
ROLE OF ANIONS IN VASCULAR
SMOOTH MUSCLE TENSION
There is a great need for systematic study in this
field. So far, although anions have been considered
from time to time, they have been studied only to
underline the effect of their associated cations. No
approach to this problem will really make much sense,
however, until an acceptable basic model for the role
of cations in the regulation of vascular tension is pre-
sented. This model need not be the final one as long
as it provides a good rational framework. We hope to
present such an integrated view in the theoretical
discussion to follow.
THEORETICAL INTERPRETATIONS
A rational theoretical interpretation of the available
evidence is now quite possible and has been attempted
i i6o
HANDBOOK OF PHYSIOLOGY
cikcn.ATION II
by several workers Raab ( i • > _• i has suggested that the
amount of sodium in the smooth muscle cell deter-
mines its responsiveness to catecholamines which he
considers important in the pathogenesis of hyperten-
sive states Tobian & Redleaf (200) suggest that the
amount of both sodium and potassium increases in
the vascular smooth muscle cell in chronic hyperten-
sion and, by osmotic attraction, causes cell swelling
and water logging. Tobian has recently reviewed this
position (193, 194)- VVe have presented the theory
that the sodium transfer systems, broadly defined,
and expressed in the sodium gradient, determine
vascular tone (78). Raab (163) has recently revised his
position to incorporate the sodium gradient into his
basic thesis ol catecholamine sensitivity.
Insofar as the cell is concerned, neither the amount
nor the concentration of Na or K enclosed by its
membrane has any meaning apart from their relation
to the external environment as the gradients Na„/Na,
and K, K.„. An increase in Na,-, for example, attracts
water into the cell until osmotic equilibrium is estab-
lished only if Na, has increased relative to Na„. Again,
an increase in K„ will redistribute itself so as to pro-
duce no osmotic effect at equilibrium if Na„ Na, is
kept constant. Or again, insofar as membrane poten-
tials are concerned, an increase in K, hyperpolarizes
the cell only if K, K.„ is made steeper thereby.
We need not belabor the point implicit in the basic
principles of the introduction to this chapter but only
urge that a satisfactory theory must be based on con-
centration (or activity) gradients and not stress either
cell or environment alone in isolation. It is equally
apparent from the evidence presented that a satisfac-
tory theory must embrace both sodium and potassium.
We believe that the following theoretical interpreta-
tion will lit many of the presently known facts and
will perhaps serve to stimulate further thought. It will
be presented in the form of generalizations with some
supporting evidence. The remainder of the evidence
is contained in the body of this chapter.
/ ) Vascular smooth muscle tension is inversely pro-
portional to the membrane potential, that is, to the
sum of the equilibrium potentials of Na+ and K+
where, in the basal state, the permeability of the cell
lo K4 considerably exceeds its permeability to Na+.
Laborit & Huguenard (130) and Furchgott (87) have
already expressed this view.
The simple shift of water from cells to environment
which can be induced b\ increasing the external
tonicity will increase both Na, and K,, hyperpolarize
the membrane and relax the cell. This explains the
vasodilatation which consistently follows the infusion
of hyperosmotic solutions. Sustained exposure to hy-
perosmotic solutions containing particles other than
Na+ will not only lower Na0 but induce a flow of K+
from cells to medium so that at equilibrium the mem-
brane potential will be reduced and tension increased.
This may explain postnephrectomy hypertension (83).
2) Acute change in vascular smooth muscle tension
is ordinarily accomplished by agents which alter the
permeability of the membrane to Na+. An agent
which increases the permeability to Na+ will produce
an immediate depolarization and increase in tension
followed by a flow of sodium from environment to cells.
Such a flow of sodium has been consistently induced
in vivo by all vasoconstrictors.
If cell volume is to be maintained unchanged during
this process, potassium must leave the cell as sodium
enters. The expected increase in K„ does not occur
with all vasoconstrictors. In this case we must assume
that some cell swelling occurs. Fending further data
we recognize that real changes in cell volume may
also be involved in changes of tension in vascular
smooth muscle (65, 195).
■-;) Sustained change in vascular smooth muscle
tension may be accomplished by agents which ad-
just and sustain the membrane permeability to Na.
The equilibrium state fora given permeability is mani-
fest in the Na gradient. Since the entrance and exit
mechanisms for sodium are not necessarily the same
(see Goldman equation) the same result can be
achieved by varying either influx or efflux rate. The
sodium gradient falls, for example, if influx rate is
increased or efflux hindered. If Na, tends to accumu-
late in a sustained manner due to either of these
changes the cell can, within reason, compensate by
increasing its work of extrusion. Presumably the first
effort of the cell to compensate will be reflected by an
increase in the cell machinery involved in the work of
such Na extrusion. This capacity must, however, be
limited so that equilibrium will next be attained at a
lower gradient, that is, Na, increases until equilibrium
is re-established. The resultant accumulation of .Na,
must lead to the extrusion of K,, a new and lower
membrane potential and an increase of tension.
We have described the evidence that Na, is actually
increased in sustained hypertensive states. It is
equally clear that chronic sodium-depleting proce-
dures tend to re-establish the basic normal situation.
There is also good evidence that mineralocorticoids
regulate the permeability of cell membranes to
sodium (33, 74, 123). A control system which allows ,i
small trickle of sodium to enter the cell and then
EFFECTS OF IONS UN VASCULAR SMOOTH MUSCLE
I l6l
regulates the case with which it is extruded permit>
very fine control of the sodium gradient.
4) The role of cell volume remains to be assessed
both in acute and chronic changes of vascular smooth
muscle tension. This is self-evident. We are repeating
this point at this time to emphasize the fact that this
problem cannot be dealt with properly until such time
as water movements can be accurately measured (65,
195)-
SUMMARY
The detailed supporting evidence leading to our
final theoretical interpretation is contained in the body
of this chapter. In order to underline our intention
we have referred briefly to some essential evidence
which cannot be easily explained in any alternate
way. It is our opinion that most of the apparentlv
complex material presented can be temporarily but
usefully rationalized by reference to the theory pre-
sented. It is allied with general ionic theory as it
applies to other contractile elements (113) modified
to serve the special needs of this particular tissue.
We conclude that vascular smooth muscle tension
depends on ionic distributions and mobilities across
the cell membrane. The transmembrane equilibra-
tion of both sodium and potassium has been stressed
as has the possibility of a direct link with calcium.
REFERENCES
I Al.KKSANDROW, D., W. W'VSZNACKA, AND J. GaJEWSKI.
Studies on the mechanism of hypotensive action of
chlorothiazide. New Engl. J. Med. 260: 51, 1959.
2. Allen, F. M., and J. \V. Sherrill. The treatment of
arterial hypertension. J. Metabolic Research 2: 429, 1922.
3 Ambache, N. Interaction of drugs and the effect of
cooling on the isolated mammalian intestine. ./. Physiol.
104: 266, 1946.
4. Ambard, L., and E. Beaujard. Causes de l'hypertension
arterielle. Arch. gen. Med. 1 : 520, 1 904.
5. Axelsson, J. Further studies of the dissociation between
action potentials and the contractile mechanism in
smooth muscle. J. Physiol. 152: 16P, i960.
6. Axelsson. J., and E. Bulbring. Some means of abolishing
the tension response in smooth muscle during continued
electrical activity at the cell membrane. J. Physiol. 1 49 :
50P, 1959.
7. Bacq, Z. M., and A. M. Monnier. Recherches sur la
physiologie et la pharmacologic du systeme nerveux
autonome. XV. Variations de la polarisation des muscles
lisses sous l'influence du systdme nerveux autonome et de
ses mimetiques. Arch, intern, physiol. 40: 467, 1935.
8. Baer, J. E., H. F. Russo, and K. H. Beyer. Saluretic
activity of hydrochlorothiazide i6-chloro-7-sulfamyl-
3,4-dihydro-i ,2 ,4-benzothiadiazine-i , 1 -dioxide j in the
dog. Proc. Soc. Exptl. Biol. Med. 100: 442, 1959.
g. Barr, L. M. Distribution of ions in intestinal smooth
muscle. Proc. Soc. Exptl. Biol. Med. 101 : 283, 1959.
10. Barr, L. M., D. F. Bohr, and V. Headings. Recovery
of carotid artery strips from cold storage. Federation Proc.
19: 258, i960.
1 1 . Bevan, J. A. The use of the rabbit aorta strip in the
analysis of the mode of action of /-epinephrine on vascular
smooth muscle. J. Pharmacol. Exptl. Therap. 129: 417,
i960.
12. Binet, L., and M. Burstein. Action de quelques cations
sur le tonus des vaisseaux peripheriques. Compt. rend. soc.
biol. 142: 1363, 1948.
13. Bohr, D. F., D. C. Brodie, and D. H. Cheu. Effect
of electrolytes on arterial muscle contraction. Circulation
n- 746. 195°-
14. Bohr, D. F., and P. L. Goulet. A direct recording of
tension from isolated arteriolar smooth muscle. Physiologist
3 (No. 3) : 25, 1 960.
15. Born, G. V. R. The relation between the tension and
the high-energy phosphate content of smooth muscle.
J. Physiol. 131 : 704, 1956.
16. Born, G. V. R., and E. Bulbring. The movement of
potassium between smooth muscle and the surrounding
fluid. J. Physiol. 131: 6go, 1955.
1 7. Bozler, E. Conduction, automaticity and tonus of
visceral muscles. Experientia 4: 213, 1948.
18. Braun-Menendez, E. Water and electrolytes in experi-
mental hypertension. In: Ciba Foundation Symposium on
Hypertension. Boston: Little, Brown, 1954, p. 238.
19. Bulbring, E. Membrane potentials of smooth muscle
fibres of the taenia coli of the guinea-pig. J. Physiol. 125
302, 1954.
20. Bulbring, E. Correlation between membrane potential,
spike discharge and tension in smooth muscle. J. Physiol.
127: 9P, 1955.
21. Bulbring, E. Correlation between membrane potential,
spike discharge and tension in smooth muscle. J . Physiol.
128: 200, 1955
22. Bulbring, E., and H. Lullmann. The effect of metabolic
inhibitors on the electrical and mechanical activity of
the smooth muscle of the guinea-pig's taenia coli. .7.
Physiol. 136: 310, 1957.
23. Burget, G. E., and M. B. Visscher. Variations of the
pH of the blood and the response of the vascular system
to adrenalin. Am. J. Physiol. 81 : 113, 1927.
24. Burnstock, G. Membrane potential changes associated
with stimulation of smooth muscle by adrenalin. Xature
186: 727, i960.
25. Burnstock, G, and C. L. Prosser. Conduction in
smooth muscles : comparative electrical properties. Am. J.
Physiol. 199: 553, i960.
26. Burnstock, G, and R. W. Straub. A method for study-
ing the effects of ions and drugs on the resting and action
potentials in smooth muscle with external electrodes.
J. Physiol. 140: 156, 1958.
27. Cameron, D. R, D. M. Dunlop, R. Platt, M 1.
Rosenheim, and E. P. Sharpey-Schafer. The rice diet
I 162
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
in the treatment of hypertension. A report to the Medical 47.
Research Council. Lancet 2 : 509, 1950.
28. Cantoni, G. L., and G. Eastman. On the response of 48
the intestine to smooth muscle stimulants. J. Pharmacol.
Exptl. t'herap. 87: 392, 1946. 49.
29. Cier, J. F., R. Chambon, and P. Rigaud. La penetra-
tion intracellulaire du sodium dans l'hypertension par 50.
la deoxycorticosterone chez le rat. Compt. rend. soc. biol.
'53: '392> '959- 51-
30. Cole, D. F. Chemical changes in the tissues of the rat
after adrenalectomy. J. Endocrinol. 6: 245, 1950. 52.
31. Conway, E. J. Exchanges of K, Na and H ions between
the cell and its environment. Irish J. Med. Sci. 262 :
593, '947- 53-
32. Conway, E. J. Principles underlying the exchanges of K
and Na ions across cell membranes. ./. Gen. Physiol. 43:
17, 1960. 54
33. Conway, E. J., and D. Hingerty. The effects of corti-
sone, deoxycorticosterone and other steroids on the active
transport of sodium and potassium ions in yeast. Biochem. 55.
J- 55 : 455. '953-
34. Corcoran, A. C, R. D. Taylor, and I. H. Pace. Con- 56.
trolled observations on the effect of low sodium dieto-
therapy in essential hypertension. Circulation 3: 1, 1951.
35. Cottier, P. T., J. M. Weller, and S. W. Hoobler. 57.
Sodium chloride excretion following salt loading in
hypertensive subjects. Circulation 18: 196, 1958.
36. Dahl, L. K.. Salt intake, adrenocortical function and 58
hypertension. Nature 181 : 989, 1958.
37. Dahl, L. K., and R. A. Love. Evidence for relationship
between sodium (chloride) intake and human essential 59.
hypertension. A.M. A. Arch. Internal Med. 94: 525, 1954.
38. Dahl, L. K., and R. A. Love. Etiological role of sodium
chloride intake in essential hypertension in humans. 60.
J. Am. Med. Assoc. 164: 397, 1957.
39. Daniel, E. E., and B. N. Daniel. Effects of ovarian 61.
hormones on the content and distribution of cation in
intact and extracted rabbit and cat uterus. Can. J. Bio- 62,
chem. and Physiol. 35: 1205, 1957.
40. Daniel, E. E., and O. Dawkins. Aorta and smooth
muscle electrolytes during early and late hypertension.
Am. J. Physiol. 190: 71, 1957.
41. Daniel, E. E., O. Dawkins, and J. Hunt. Selective 63.
depletion of rat aorta potassium by small pressor doses
of norepinephrine. Am. J. Physiol. 190:67, >957-
42. Daniel, E. E., A. Dodd, and J. Hunt. Effects of pitressin
and isoproterenol on aorta electrolytes. Arch, intern. 64.
pharmacodynamic 119: 43, 1959.
43. Daniel, E. E., and H. Singh. The electrical properties
of the smooth muscle cell membrane. Can. J. Biochem. 65.
and Physiol. 36: 959, 1958.
44. Davf.y, D. A. Measurement of changes of tension in the
walls of perfused segments of blood vessels. J. Physiol.
132: 1 P, 1 956. 66.
45. Deal, C. P., Jr., and H. D. Green Effects of pH on
blood How and peripheral resistance in muscular and
cutaneous vascular beds in the hind limb of the pento- 67.
barbitalized dog. Circulation Research 2: 148, 1954.
46. De VVesselow, O. L. V. S., and W. A. R. Thomson. A
study of some serum electrolytes in hypertension. Qiiart. 68
./. Med. 8: 36., 1939.
Dickinson, C. J. Rapid contractile properties of isolated
mammalian arteries. Nature 185:620, i960.
Dodd, W. A., and E E. Daniel. Vascular muscle
reactivity. Circulation Research 8 : 446, 1 960.
Dodd, VV. A., and E. E. Daniel. Electrolytes and arterial
muscle contractility. Circulation Research 8: 451, i960.
D'Silva, J. L. The action of adrenaline on serum potas-
sium. J. Physiol. 82: 393, 1934.
Ebashi, S. Calcium binding and relaxation in the acto-
myosin system. J. Biochem., Tokyo 48: 150, i960.
Efron, D. H. The effect of adrenalectomy on the content
and turnover of sodium and potassium in various organs.
Acta Endocrinol. 26: 209, 1957.
Eichelberger, L. The distribution of water and elec-
trolytes between blood and skeletal muscle in experi-
mental hypertension. J. Exptl. Med. 77: 205, 1943.
Eichler, O. Die Pharmakologie anorganischer Anionen.
Handbuch der Expenmentellen Pharmakologie. Berlin : Springer,
1950, vol. 10.
I.isi nman, G, D. O. Rudin, and J. U. Casby. Glass
electrode for measuring sodium ion. Science 126: 831, 1957.
Eliakim, M., S. Z. Rosenberg, and K. Braun. Effect of
hypertonic saline on the pulmonary and systemic pres-
sures. Circulation Research 6: 357, 1958.
Emanuel, D. A., M. Fleishman, and F. J. Haddy.
Effect of pH change upon renal vascular resistance and
urine How. Circulation Research 5: 607, 1 957-
Emanuel, D. A., J. B. Scott, and F. J. Haddy. Effect
of potassium upon small and large blood vessels of the
dog forelimb. Am. J. Physiol. 197 : 637, 1959.
Engbaek, L. The pharmacological actions of magnesium
ions with particular reference to the neuromuscular and
the cardiovascular system. Pharmacol. Reus. 4: 396, 1952.
Evans, C. L. The physiology of plain muscle. Physiol.
Revs. 6: 358, 1926.
Fenn, VV. O. The role of potassium in physiological
processes. Physiol. Rei's. 20: 377, 1940.
Ferrebee, J. W., D. Parker, W. H. Carnes, M. K.
Geritv, D. VV. Atchlf.y, and R. F. Loeb. Certain effects
of desoxycorticosterone. The development of "diabetes
insipidus" and the replacement of muscle potassium by
sodium in normal dogs. Am. J. Physiol. 135: 230, 1 941 ■
Fleisch, A., I. Sibul, and V. Ponomarev. Uber nutritive
Kreislaufregulierung. I. Kohlensaure und Sauerstoff-
mangel als auslosende Reize. Pfliigers Arch. ges. Physiol.
230, 814, 1932.
Fleishman, M., J. Scott, and F. J. Haddy. Effect of
pH change upon systemic large and small vessel resist-
ance. Circulation Research 5: 602, 1 957-
Folkow, B., and B. Oberg. The effect of functionally
induced changes of wall/lumen ratio on the vasocon-
strictor response to standard amounts of vasoactive
agents Acta Physiol. Scand. 47: 131, 1959.
Freed, S. C, and M. Friedman. Hypotension in the
rat following limitation of potassium intake. Science 112:
788, 1950.
Freed, S. C, and M. Friedman. Depressor effect of
potassium restriction on blood pressure of the rat. Proc.
Soc. Exptl. Biol. Med. 78: 74, 1951.
Freed, S. C, R. H. Rosenman, and M. Friedman. The
relationship of potassium in the regulation of blood
EFFECTS OF IONS ON VASCULAR SMOOTH MI si I I
I 163
pressure with special attention to corticosteroid hyper- 87.
tension. Arm. A*. )'. Acad. Sci. 56: 637, 1953.
69. Freed, S. C, S. St. George, and R. H. Rosenhan. 88.
Aorta electrolytes of hypotensive potassium-deficient
rats. Am. J. Physiol. 1 95 ; 445, 1 958.
70. Freed, S. C, S. St. George, and R. H. Rosenman.
Arterial wall potassium in renal hypertensive rats. 89.
Circulation Research 7:219, 1 959.
71. Fregly, M. J. Production of hypertension in adrenalec-
tomized rats given hypertonic salt solution to drink. 90.
Endocrinology 66: 240, i960.
72. Friedman, M., S. C. Freed, and R 11. Rosenman.
Effect of potassium administration on ( 1 ) the peripheral
vascular reactivity and I21 blood pressure of the potas- 91.
sium-deficient rat. Circulation 5: 415, 1952.
73. Friedman, S. M., R M. Butt, and G. L. Friedman. 92.
Cation shifts and blood pressure regulation in the dog.
Am. J. Physiol. 190:507, 1957. 93.
74. Friedman, S. M., and C. L. Friedman. Effect of aldo-
sterone and hydrocortisone on sodium in red cells.
Expenentia 14: 452, 1958.
75. Friedman, S. M., C. L. Friedman, and M. Nakashima. 94.
Cationic shifts and blood pressure regulation. Circulation
Research 5: 261, 1957.
76. Friedman, S. M., C. L. Friedman, and M. Nakashima.
Effect of angiotonin on the distribution of sodium, 95.
potassium and water in the rat. Nature 180: 194, 1957.
77. Friedman, S. M., J. A. M. Hinke, and D. F. Hardvvtck.
Sodium tolerance in experimental hypertension. Circula- 96.
lion Research 3 : 297, 1 955.
78. Friedman, S. M., J. D. Jamieson, and C. L. Friedman. 97.
Sodium gradient, smooth muscle tone and blood pressure
regulation. Circulation Research 7 : 44, 1 959. 98.
79. Friedman, S. M., J. D. Jamieson, J. A. M. Hinke, and
C. L. Friedman. Use of glass electrode for measuring
sodium in biological systems. Proc. Soc. Exptl. Biol. Med.
99: 727. !958- 99-
80. Friedman, S. M., J. D. Jamieson, J. A. M. Hinke, and
C. L. Friedman. Drug-induced changes in blood pressure
and in blood sodium as measured by glass electrode. 100.
Am. ./. Physiol. 196: 1049, '959-
81. Friedman, S. M., J. D. Jamieson, M. Nakashima, and ioi.
C. L. Friedman. Sodium ion and smooth muscle con-
traction. Proc. Council for ///;'/? Blood Pressure Research 8:
57. '959- 102.
82. Friedman, S. M., M. Nakashima, and C. L. Friedman.
Extrarenal effects of intravenous pitressin in nephrecto-
mizcd rats. Circulation Research 4: 557, 1956. 103.
83. Friedman, S. M., M. Nakashima, and C. L. Friedman.
Relation of saluretic and hypotensive effects of hydro-
chlorothiazide in the rat. Am. J. Physiol. 198: 148, i960. 104.
84. Friedman, S. M., J. R. Pollev, and C L. Friedman.
The effect of desoxycorticosterone acetate on blood 105.
presstire, renal function and electrolyte pattern in the
intact rat. J. Exptl. Med. 87: 329, 1948.
85. Friedman, S. M., and F. A. Sreter. Effects of vasopressin
on sodium, potassium and water distribution in rat 106.
gastrocnemius muscle. Endocrinology 69: 386, ig6i.
86. Friedman, S. M., W. A. Webber, J. D. Jamieson, and
C. L. Friedman. Pressor responsiveness following acute toy
elevation of sodium in the rat. Can. J. Biockem. and Physiol.
35- 327. '957-
Furchgott, R. F. The pharmacology of vascular smooth
muscle. Pharmacol. Revs. 7: 189, 1955.
Furchgott, R. F., and S. Bhadrakom. Reactions of
strips of rabbit aorta to epinephrine, isopropylarterenol,
sodium nitrite and other drugs. J. Pharmacol. Exptl.
Therap. 108: 129, 1953.
Gaudino, M., and M. F. Levitt. Inulin space as a
measure of extracellular fluid. Am. .1 . Physiol. 157: 387,
'949-
Gellhorn, E. Beitrage zur allgemeinen Zellphysiologie.
V. Weiterc Untersuchungen iiber die YVirkung der
Kationen auf die glatte Muskulatur. P/lugers Arch. ges.
Physiol. 213: 789, 1926.
Goffart, M., and Z. M. Baco_. Les sensibilisateurs au
potassium. Ergeb. Physiol. 47: 555, 1952.
Goldman, D. E. Potential, impedance and rectification
in membranes. J. Gen. Physiol. 27: 37, 1944.
Green, D. M., F. M. Sturtevant, and C. G. Van
Arman. The temporal course of fluid intake and response
to fluid loads in perinephritic hypertension in rats.
(.111 illation Research 2: 73, 1954.
Green, D. M., H. G. Wedell, M. H. Wald, and B.
Learned. The relation of water and sodium excretion
to blood pressure in human subjects. Circulation 6: 919,
■952-
Greene, R. W., and L. A. Sapirstein. Total body sodium,
potassium and nitrogen in rats made hypertensive by
subtotal nephrectomy. Am. J. Physiol. 169: 343, 1952.
Grollman, A. The water and electrolyte content of the
tissues in hypertension. Circulation Research 2: 541, 1954.
Grollman, A. (editor). New diuretics and antihyper-
tensive agents. Ann. N.Y. Acad. Sci. 88: 771, i960.
Grollman, A.,T R. Harrison, M. F. Mason, J. Baxter,
J. Crampton, and F. Reichsman. Sodium restriction in
the diet for hypertension. J. Am. Med. Assoc. 129: 533,
'945-
Gross, F., and H. Schmidt. Natrium-und Kaliumgehalt
von Plasma und Geweben beim Cortexon-Hochdruck.
Arch. Exptl. Pathol. Pharmacol. 233: 311, 1958.
Guyton, A. C. Textbook of Medical Physiology. Philadelphia:
Saunders, 1956.
IIaddy, F. J. Local effects of sodium, calcium and
magnesium upon small and large blood vessels of the dog
forelimb. Circulation Research 8 57, 1960.
Haddy, F. J., and H. W. Overbeck. The effect of
hyper- and hypotonic solutions on small vessel resistance
in the dog forelimb. Physiologist 3 (No. 3): 71, 1 g6o.
Haight, A. S., and J. M. Weller. Tissue electrolytes
of rats given excess of sodium chloride. Federation Proc.
'9: 2.54. '96°-
Harvey, R. B. Vascular resistance changes produced
by hyperosmotic solutions. Am. J. Physiol. 199: 31, i960.
Haury, V. G. The broncho-dilator action of magnesium
and its antagonistic action (dilator action) against
pilocarpine, histamine and barium chloride. J. Pharmacol.
Exptl. Therap. 64 : 58, 1 938.
Haury, V. G. The effect of intravenous injections of
magnesium sulfate on the vascular system. J. Pharmacol.
Exptl. Therap. 65: 453, 1939.
Hazard, R., and A. Cornec. Action du potassium sur
1'intestin isole de rat et sur sa reactivite a ^acetylcholine.
Compt. rend. soc. biol. 146: 896, 1952.
I 164
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
108. Hazard, R., and A. Quinquaud. L'ion potassium vaso- 129.
stricteur. J. Physiol., Paris 44: 259, 1952.
109. Headings, V. E., D. F. Bohr, and P. A. Rondell.
Electrolytes in dog carotid in vitro following electrical 130.
and epinephrine stimulation. Federation Proc. 19: 104, i960.
1 10. Heilbrunn, L. V., and 1 [ Wiercinski. The action of
various cations on muscle protoplasm. J. Cellular Comp. 131.
Physiol. 29: 15, 1947.
111. Hilden, T., and A. R. Krogsgaard. Low serum potas-
sium level in severe hypertension. Am, J. Med. Sci. 236:
487, 1958. 13*.
112. Hodgkin, A. L. The ionic basis of electrical activity in
nerve and muscle. Biol. Rev. Cambridge Phil. Soc. 26:
339. '951- '33-
113. Hodgkin, A. L., and P. Horowicz Movements of Na
and K in single muscle fibres. J. Physiol. 145: 405, 1959.
114. Hoerr, N. L. Illumination of living organs for micro- 134.
scopic study. In: Medical Physics, edited by O. Glasser.
Chicago: Yr. Bk. Pub. 1944. p. 625. 135.
115. Hoff, H. E., P. K. Smith, and A. W. Winkler. The
relation of blood pressure and concentration in serum 1 36.
of potassium, calcium and magnesium. Am. J. Physiol.
127: 722, 1939.
116. Holman, M. E. The effect of changes in sodium chloride 137.
concentration on the smooth muscle of the guinea-pig's
taenia coli. J. Physiol. 136: 569, 1 957.
1 1 7. Holman, M. E The effect of changes in potassium chlor-
ide concentration on the membrane potential, electric 1 38.
activity and tension of intestinal smooth muscle. J.
Physiol. 137: 77 P, 1957.
118. Holman, M. E. Membrane potentials recorded with
high-resistance micro-electrodes; and the effects of 139.
changes in ionic environment on the electrical and
mechanical activity of the smooth muscle of the taenia
coli of the guinea-pig. J. Physiol. 141 : 464, 1958. 140.
119. Houck, C. R. Hypertension in the nephrectomized dog.
Trans. Am. Coll. Cardiol. 6: 144, 1956.
120. Hughes, F. B., R. J. S. McDowall, and A. A. I. Soli-
man. Sodium chloride and smooth muscle. J. Physiol. 141.
134: 257, 1956.
131. Hurwitz, L., B. Tinsley, and F. Battle. Dissociation 142
of contraction and potassium efflux in smooth muscle
Am. J. Physiol. 199: 107, i960.
122. Jamieson, J. D., and S. M. Friedman. Sodium and 143.
potassium shifts associated with peripheral resistance
changes in the dog. Circulation Research 9: 996, 1961.
123. Jones, E. S. Cellular electrolytes and adrenal steroids.
Nature 1 76 : 269, 1 955.
124. Kao, C. Y., F. Bronner, and D. Zakim. Evidence for 144
increased sodium permeability during activity in mam-
malian smooth muscle. Federation Proc. 19: 257, i960.
125. Katz, L. N., and E. Lindner The action of excess 145.
Na, Ca and K on the coronary vessels. Am. J. Physiol.
>24: '55. "938. 146-
126. Kempner, W. Treatment of hypertensive vascular disease
with rice diet. Am. J. Med. 4: 545, 1948.
127 Kester, N. C, A. W. Richardson, and H. D. Green. 147
The effect of controlled hydrogen-ion concentration on
peripheral vascular tone and blood flow in innervated
hind leg of the dog. Am. J. Physiol. 169: 678, 1952.
128. Koletsky, S., and A. M. Goodsitt. Natural history 148.
and pathogenesis of renal ablation hypertension. A.M. A
Arch. Pathol. 69: 654, i960.
Koletsky, S., H. Resnick, and D. Behrin. Mesenteric
artery electrolytes in experimental hypertension. Proc.
Soc. Exptl. Biol. Med. 102: 12, 1959.
Laborit, H, and P. Huguenard. Influence possible des
variations du potentiel de membrane sur la valeur de la
pression differentielle. J. Physiol , Paris 48 : 87 1 , 1 956.
Larach, J. H, S. Click, V. Januszewicz. Q. B. Deming,
\V. G. Kelly, and S. Lieberman. Aldosterone secretion
and primary and malignant hypertension. J. Clin. Invest.
39: 1091, i960.
Laramore, D. C, and A. Grollman. Water and elec-
trolyte content of tissues in normal and hypertensive
rats. Am. J. Physiol. 161 : 278, 1950.
Laszt, L. Correlation between the electrolyte and water
content of the organs and hypertension after adminis-
tration of corticosteroids. Nature 185: 695, 1960.
Laszt, L. Effect of potassium on muscle tension, especially
on that of vascular muscle. Nature 1 85 : 696, 1 960.
Laszt, L. Effect of the cations of the lyotropic series on
the tension of vascular muscle. Nature 187: 329, i960.
Ledingham, J. M. The distribution of water, sodium and
potassium in heart and skeletal muscle in experimental
renal hypertension in rats. Clin. Sci. 13: 337, 1953.
Ledingham, J. M. The distribution of fluid and elec-
trolytes in experimental hypertension. In : Ciba Founda-
tion Symposium on Hypertension. Boston: Little, Brown
■954. P 25°
Ledingham, J. M. Hypertension and disturbances of
tissue water, sodium and potassium distribution asso-
ciated with steroid administration in adrenalectomized
rats. Clin. Sci. 13: 543, 1954.
Ledincham, J. M. Disturbances in water and electrolyte
metabolism in experimental hypertension. Brit. Med.
Bull. 13: 33, 1957.
Leonard, E. Alteration of contractile response of artery
strips by a potassium-free solution, cardiac glycosides
and changes in stimulation frequency. Am. J. Physiol.
189: 185, 1957.
Lowenstein, J. M. Synergism of bivalent metal ions in
transphosphorylation. Nature 187: 570, i960.
Luttgau, H. C, and R. Niedergerke. The antagonism
between Ca and Na ions in the frog's heart. J. Physiol.
143: 486, 1958.
McCance, R. A., and A. B. Morrison. The effects of
equal and limited rations of water, and of 1 , 2 and 3
per cent solutions of sodium chloride on partially nephrec-
tomized and normal rats. Quart J. Exptl. Physiol. 41 : 365,
'956-
McDowall, R. J S., and A. A. I. Soliman. Sodium
chloride and the response of smooth muscle. J. Physiol.
122: 42P, 1953.
McDowall, R. J. S., and A. F. Zayat. Sodium chloride
and cardiac muscle. J. Physiol. 120: 13P, 1953.
McKeever, W. P., H. Braun, D. Coder, and J. Croft,
Jr. The local effect of potassium on different segments
of the coronary vascular bed. Clin. Research 3 : 1 88, 1 960.
Magee, H. E., and C. Reid. Studies on the movements
of the alimentary canal. I. The effects of electrolytes on
the rhythmical contractions of the isolated mammalian
intestine. J. Physiol. 63: 97, 1927.
Marshall, R. J., and J. T. Shepherd. Effect of injec-
tions of hypertonic solutions on blood flow through the
femoral artery of the dog. Am. J. Physiol. 197: 951, 1959.
EFFECTS OF IONS ON VASCULAR SMOOTH MUSCLE
I 165
149. Mathison, G. C. Potassium and peripheral vascular 168.
resistance. J. Physiol. 42: 471, igii.
150. Meneely, G. R., C. O. T. Ball, and J. B. Youmans.
Chronic sodium chloride toxicity. The protective effect 169.
of added potassium chloride. Ann. Internal Med. 47: 263,
'957-
151. Meneely, G. R., R. G. Tucker, W. J. Darby, and
S. H. Auerbach. Chronic sodium chloride toxicity: 170.
hypertension, renal and vascular lesions. Ann. Internal
Med. 39: 991, 1953.
152. Meneely, G. R., R. G. Tucker, W. J. Darby, and
S. H. Auerbach. Chronic sodium chloride toxicity in 171.
the albino rat. II. Occurrence of hypertension and a
syndrome of edema and renal failure. J. Exptl. Med.
98: 71. '953- '72-
153. Muirhead, E. E., A. Goth, and F. Jones. Sodium and
potassium exchanges associated with nor -epinephrine
infusions. Am. J. Physiol. 179: 1, 1954.
154. Muirhead, E. E., R W. Lackey, C. A. Bunde, and 173.
J. M. Hill. Transient hypotension following rapid
intravenous injections of hypertonic solutions. Am. J.
Physiol. 151 : 516, 1947. 174.
155. O'Brien, G. S., Q. R. Murphy, Jr., and \V. J. Meek.
The effect of sympathomimetic amines on arterial
plasma potassium and cardiac rhythm in anesthetized 175.
dogs. J. Pharmacol. Exptl. Therap. 109: 453, 1953.
156. Overbeck, H W., and F. J. Haddy. Acute effects
of Na+, K+ and Ca++ on vascular resistance in the dog 1 76.
forelimb. Physiologist 3 (No. 3): 122, i960.
157. Paton, W. D. M. The response of the guinea-pig ileum 177.
to electrical stimulation by coaxial electrodes. J. Physiol.
127: 40P, 1955.
158. Perera, G. A. Depressor effects of potassium-deficient
diets in hypertensive man. J. Clin. Invest. 32 : 633, 1 953. 1 78.
159. Pines, K. L., and G. A. Perera. Sodium chloride restric-
tion in hypertensive vascular disease. Med. Clin. North
Am. 33: 713, 1949. 179.
160. Podolsky, R. J. The structure of water and electrolyte
solutions. Circulation 21 : 818, i960.
161. Prosser, C. L., G. Burnstock, and J. Kahn. Conduction 180.
in smooth muscle: comparative structural properties.
Am. J. Physiol. 199: 545, i960.
i6ia.PRUTTON, C. F., AND S. H. Maron. Fundamental Prin-
ciples of Physical Chemistry. New York: Macmillan, 1951. 181.
162. Raab, W. The integrated role of catecholamines, min-
eralocorticoids and sodium in hyper and hypotension.
(A working hypothesis). J. Ml. Sinai Hosp. N.Y. 19: 182.
233. '952-
163. Raab, W. Transmembrane cationic gradient and blood
pressure regulation. Interaction of corticoids, catechol-
amines and electrolytes on vascular cells. Am. J. Cardiol. 183
4^ 752. '959-
164. Read, R. C, J. A. Johnson, J. A. Vick, and M. W.
Meyer. Vascular effects of hypertonic solutions. Circula- 184.
lion Research 8: 538, i960.
165. Robinson, J. R. Metabolism of intracellular water.
Physiol. Revs. 40: 112, i960. 185.
166. Rogers, T. A., and W. O. Fenn. Effect of extra -cellular
pH on muscle electrolytes. Federation Proc. 16: 109, 1957.
167. Rosenman, R. H.. S. C. Freed, and M. Friedman. 186.
Effect of variation of potassium intake on pressor activity
of desoxycorticosterone. Proc. Soc Exptl. Biol. Med. 78:
77. >95'-
Rosenman, K H ., S. C. Freed, and M. Friedman. The
peripheral vascular reactivity of potassium deficient rats.
Circulation 5: 412, 1952.
Rosenman, R. H., S. C. Freed, and M. Friedman.
Effect of desoxycorticosterone acetate upon the blood
pressure of rats fed varied dietary intakes of potassium
and sodium. J. Clin. Endocrinol. 14: 661, 1954.
Rosenman, R. H., S. C. Freed, S. St. Georce, and
M. K. Smith. The effect of varying dietary potassium
on the blood pressure of hypertensive rats. Am. J. Physiol.
'75- 386> '953-
Sapirstein, L. A. Sodium and water ratios in the patho-
genesis of hypertension. Proc. Council for High Blood Pus-
sure Research !> -'8, 1957.
Sapirstein, L. A., VV. L. Brandt, and D. R. Drury.
Production of hypertension in the rat by substituting
hypertonic sodium chloride solutions for drinking water.
Proc. Soc. Exptl. Biol. Med. 73: 82, 1950.
Saunders, S. J., R. O. H. Irvine, M. A. Crawford,
and M. D. Milne. Intracellular pH of potassium-defi-
cient voluntary muscle. Lancet 1 : 468, i960.
Schmid, E., M. v. Bubnoff, U. Wagenmann, and R.
Taugner. Zur Kreislaufwirkung der Magnesiumsalze.
Arch. Exptl. Pathol. Pharmacol. 224: 426, 1955.
Schroeder, H. A. Renal failure associated with low
extracellular sodium chloride. The Low Salt Syndrome.
./. Am. Med. Assoc. 141: 117, 1949.
Schroeder, H. A. Hypertensive Diseases. Philadelphia :
Lea & Febiger, 1953.
Schuler, \V. A. EinHuss der Wasserstoffionenkonzentra-
tion auf Tonus und Adrenalinreaktion von isolierten
Mesenterial-und Zwerchfellarterien. Pfliigers Arch. ges.
Physiol. 240: 393, 1938.
Scott, J., D. Emanuel, and F. J. Haddy. Effect of
potassium on renal vascular resistance and urine flow
rate. Am. J. Physiol. 197: 305, 1959.
Selye, H., C. E. Hall, and E. M. Rowley. Malignant
hypertension produced by treatment with DCA and
sodium chloride. Can. Med. Assoc. J. 49: 88, 1943.
Selye, H., J. Mintzberg, and E. M. Rowley. Effect
of various electrolytes upon the toxicity of desoxycorti-
costerone acetate. J. Pharmacol. Exptl. Therap. 85: 42,
'945-
Selye, H., H. Stone, P. S. Timiras, and C. Schaffen-
burg. Influence of sodium chloride upon the actions of
desoxycorticosterone acetate. Am. Heart J. 37 : 1 oog, 1 949.
Shanes, A. M. Electrochemical aspects of physiological
and pharmacological action in excitable cells. The resting
cell and its alteration by extrinsic factors. Pharmacol. Revs.
!°: 59. '958-
Shanes, A. M. Electrochemical aspects of physiological
and pharmacological action in excitable cells. The action
potential and excitation. Pharmacol. Revs. 10: 165, 1958.
Skelton, F. R. Development of hypertension and
cardiovascular-renal lesions during adrenal regeneration
in the rat. Proc. Soc. Exptl. Biol. Med. 90: 342, 1955.
Skelton, F. R. A study of the natural history of adrenal-
regeneration hypertension. Circulation Research 7: 107,
'959-
Smith, L L , J. T. Hamlin III, VV. F. Walker, and
F. D. Moore. Metabolic and endocrinologic changes in
acute and chronic hypotension in man Metabolism 8:
862, 1959.
1 1 66
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
187. Stanbury, J. B. The blocking action of magnesium ion
on sympathetic ganglia. ./. Pharmacol. Exptl. Therap. 93:
52. "948-
188. Streeten, D. H. P. The effects of sodium and chloride
lack on intestinal motility and their signilicance in
paralytic ileus. Surg. Gynecol. Obstet. 91 : 421, 1950.
189. Streeten, D. H. P., and E. M. Vaughan Williams.
Loss of cellular potassium as a cause of intestinal paralysis
in dogs. J. Physiol. 118: 149, 1952.
190. Tate, G., and A. J. Clark, The action of potassium and
calcium upon the isolated uterus. Arch intern, pharmaco-
dynamic 26: 103, 1922.
191. Tobian, L. Effect of a low sodium diet on electrolyte
composition of arterial wall. Am. J. Physiol. 181 : 599, 1955.
hi-' Tobian, L. The electrolytes of arterial wall in experi-
mental renal hypertension. Circulation Research 4: 671,
1956.
193. Tobian, L. Physiology of the juxtaglomerular cells.
Ann. Internal Med. 52: 395, i960.
194. Tobian, L. Interrelationship of electrolytes, juxtaglo-
merular cells and hypertension. Physiol. Revs. 40: 280,
i960.
195. Tobian, L., and J. T. Binion. Tissue cations and water
in arterial hypertension. Circulation 5: 754, 1952.
196. Tobian, L., and J. T. Binion. Artery wall electrolytes
in renal and DCA hypertension. ./. Clin. Invest. 33: 1407,
'954-
197. Tobian, L., and A. Fox. The effect of nor -epinephrine
on the electrolyte composition of arterial smooth muscle.
./. Clin. Incest. 35: 297, 1956.
198. Tobian, L., S. Martin, and W. Eilers. Effect of pH on
norepinephrine-induced contractions of isolated arterial
smooth muscle. Am. ./. Physiol. 1 96 : 998, 1 959.
199. Tobian, L., and P. D. Redi.eaf. Effect of hypertension
on arterial wall electrolytes during desoxycorticosterone
administration. Am. J. Physiol. 189: 451, 1 957.
200. Tobian, L. , and P. D. Redi.eaf. Ionic composition of
the aorta in renal and adrenal hypertension. Am. J.
Physiol. 192: 325, 1958.
201. Toussaint, C, R. Wolter, and P. Sibille. Hyperten-
sion et lesions arterielles provoquees chez le rat par
l'ingestion de quantites excessives de chlorure de sodium.
Rev. beige pal hoi. et rned. exptl. 23 : 83, 1 953.
202. USSING, H. H., P. K.RUH0FFER, J. H. THAYSEN, AND
N. A. Thorn. The alkali metal ions in biology. Handbuch
der Experimeniellen Pharmakologie. Berlin: Springer, i960,
vol. 13.
203. Vick, J., H. E. Ederstrom, and T. Vergeer. Epin-
ephrine sensitivity of blood vessel strips from salt-fed
and castrated rats. Proc. Soc. Exptl. Biol. Med. 93: 536,
>956-
204. Vogt, M The site of action of some drugs causing stim-
ulation of the circular coat of the rabbit's intestine J.
Physiol. 102 : 1 70, 1943.
205. Warren, J. D. Cation and Water Shifts in Response to Pressor
Agents in the Conscious Dog. (Thesis). Univ. British Colum-
bia, 1 96 1.
206. White, H. L., and D. Rolf. Whole body and tissue
inulin and sucrose spaces in the rat. Am. ./. Physiol. 188:
'5'. '957-
207. White, II. L., and D. Rolf. Comparison of various
procedures for determining sucrose and inulin space in
the dog. J. Clin. Invest. 37: 8, 1958.
208. Whitehead, R. W. Responses of excised intestines to
alterations of electrolyte concentrations (Na, Ca, K).
Am. J. Physiol. 89: 253, 1929
209. Williamson, A. W. R., and F. D. Moore. Norepinephrine
sensitivity of isolated rabbit aorta strips in solutions
of varying pH and electrolyte content. Am. ./. Physiol.
198: 1 157, i960.
210. Winter, H. A., H. E. Hoff. and L. Dso. Effects of
potassium deficiency upon gastrointestinal motility.
Federation Proc. 8: 169, 1949
211. Woodbury, D. M., and A. Koch. Effects of aldosterone
and desoxycorticosterone on tissue electrolytes. Proc. Soc.
Exptl. Biol. Med. 94: 720, 1957.
212. Woodbury, J. W., and D. M. McIntyre. Electrical
activity of single muscle cells of pregnant uteri studied
with intracellular ultramicroelectrodes. Am. J. Physiol.
■77: 355. '954-
213. Woolley, D. W. A probable mechanism of action of
serotonin. Proc. Null. Acad. Sci. 44: 197, 1958.
214. Yamabayasiii, II., and W. F. Hamilton. Effect of
sodium ion on contractility of the dog's aortic strip in
response to catecholamines. Am. J. Physiol. 197: 993,
■959-
215. Zadina, R., and V. Kriz. L'action du magnesium sur
la contraction de l'intestin isole. Compt. rend. soc. biol. 142:
1037, 1948.
216. Zsoter, T., and M. Szabo. Effect of sodium and calcium
on vascular reactivity Circulation Research 6: 476, 1958.
217. Zweifach, B. W. Microscopic observations of circulation
in rat meso-appendix and dog omentum: use in study of
vasotropic substances. In: Methods in Medical Physics
Chicago: Yr. Bk. Publ., 1948, vol. 1, p. 131.
CHAPTER 34
Lipid metabolism in relation to physiology
and pathology of atherosclerosis
SAMI A. HA SHIM
WILLIAM C . FELCH
THEODORE B. VAN ITALLIE
Department of Medicine, St. Luke's Hospital, and Institute of
Nutrition Sciences, Columbia University, New York City
CHAPTER CONTENTS
Pathology
Pathogenesis
Metabolic Consequences of Ingestion of Food
Diet
Fat Absorption and Digestion
Adipose Tissue
Hormonal Influences on Adipose Tissue
The Serum Lipids
Chylomicrons
The Lipoproteins
Free Fatty Acids
Role of the Liver
Cholesterol Disposal
Factors That Influence Serum Lipids
Stress
Sex
Dietary Fatty Acids
Essential Fatty Acid (EFA) "Deficiency"
Chain Length, Unsaturation, and Melting Point
Dietary Cholesterol
Practicable Diets
Mechanism of Cholesterol Lowering
Additional Influences on Serum Lipids
Blood Lipids and Atherosclerosis
Role of Blood Clotting and Thrombosis
the term "lipid" enables us to assemble under one
heading a number of organic substances which,
although variable in chemical structure, are closely
related in biological behavior. The physical and
chemical processes by which a living organism oper-
ates are summarized bv the term "metabolism."
Thus, lipid metabolism refers to the behavior in
living organisms of fatty acids, their esters, certain
hydrocarbons, phospholipids, and sterols. Recent
technical advances permitting better separation,
identification, and quantification of the various lipids
have resulted in a vast store of new information about
lipid metabolism. Much of this material still needs
to be organized and evaluated in terms of its relevance
to problems of human health.
''Atherosclerosis" (Gr. athero, mush) refers to a
lesion of the arterial wall characterized, inter alia,
by accumulation of lipid in the intima. The term was
first suggested in 1904 by Marchand (144). Today
atherosclerosis, by virtue of its deleterious effects on
the various arteries of the heart, brain, and other
important areas of the body, appears to be the major
public health problem of Western man. Thus, by
extension, lipid metabolism as it relates to the physi-
ology and pathology of blood vessel walls has become
a subject of vital importance.
The search for the etiology of atherosclerosis has
included consideration of all elements in the classic
epidemiologic triad — agent, host, and environment.
Environmental factors have received special attention
since evidence — epidemiologic, experimental, and
clinical — has accumulated suggesting that dietary
constituents and particularly dietary fats influence the
development of atherosclerosis. Such evidence as
applied to man necessarily has been indirect because
of the inaccessibility of atherosclerotic lesions during
life. It is now recognized that dietary constituents
can profoundly influence lipid metabolism. The role
1 167
I 1 68 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
of diet-induced changes in serum lipids in the de-
velopment of atherosclerosis has not been established.
In man, atherogenesis appears to be a chronic process,
requiring considerable time, perhaps years, to evolve.
The disease can culminate in an acute obstructive
event, frequently with disastrous consequences. At
such a time a disturbance in the coagulability of the
blood may occur resulting in the formation of an
arterial thrombus. Thus, attempts have been made
to correlate changes in serum lipids as influenced by
diet with the development not only of atherosclerosis
but also of a more acute change in coagulability of
the blood.
The clinical sequelae of atherosclerosis, ischemia,
and infarction of the heart, brain, and other tissues,
have been carefully documented for years. Also, the
gross pathologic processes, such as lumen encroach-
ment, fibrosis, ulceration, calcification, and throm-
bosis, which underlie these clinical events base been
understood by pathologists since the time of Virchow.
Moreover, a reasonable explanation for the patho-
genesis of the disease was advanced more than half a
century ago and has not been disproved. Yet from
the standpoint of etiology and intimate pathogenesis
the basic nature of the disease remains obscure and
debatable.
PATHOLOGY
The pathologic entity atherosclerosis must be dis-
tinguished from other blood vessel lesions, some ot
which have been previously lumped together with
atherosclerosis under the generic designation, arterio-
sclerosis. Monckeberg's medial sclerosis differs patho-
logically, pathogenetically, and clinically from ath-
erosclerosis. Various inflammatory lesions of blood
vessel walls also can be sharply separated, although
the generally held concept that thromboangiitis
obliterans is an entity different from peripheral ath-
erosclerosis recently has been questioned (209).
Another important consideration is that common
textbook descriptions of atherosclerosis may in fact
describe mostly complications or sequelae of an initial,
clinically silent process that may start in infancy.
Thus, lumen encroachment, fibrosis, calcification,
ulceration, hemorrhage, and thrombosis are all late
conditions.
What then is the initial lesion? What is the patho-
logic essence of the disease? The answer to these
questions necessarily involves consideration of patho-
genesis (to be discussed later) as well as descriptive
pathology. Fortunately, precise studies of early gross
and microscopic lesions from human and experi-
mental material are available (21, 55, 108, 109, 130,
162). The first gross lesion, often visible in infants,
is the fatty streak, a linear yellow elevation usually
found in the aorta. Microscopic examination of such
a streak reveals underneath the heaped-up intima
an accumulation of lipophages, cells which show a
foamy, reticular cytoplasm with ordinary stains con-
taining lipid solvents, but which are found, with
appropriate fat stains, to be packed with lipid. Lipid
is also found lying free between the lipophages.
Whether the lipid which makes up the fatty streak is
first intracellular or extracellular is as yet unknown.
With larger lesions, lipid is also found below the
internal elastic lamella in the media, but the smallest,
earliest, grossly invisible lesions consist of a few foam
cells lying directly under the endothelial surface of
the intima. Thus, the lipid-containing foam cell is
usuallv considered to be the earliest recognizable unit
of the atherosclerotic process.
However, careful microscopic studies show other
subtle anatomic changes (39, 126, 195) occurring
pari passu with the appearance of lipid in the blood
vessel wall. Elastic tissue stains reveal stretching and
fragmentation of elastic fibers in the intima as an
early feature. Other special stains show metachro-
matic changes in the ground substance of the arterial
wall, and chemical studies have demonstrated muco-
polysaccharide accumulations that occur along with,
or possibly before, the appearance of visible lipid. An
abundance of evidence, clinical and experimental,
has shown that preceding damage to the arterial wall,
toxic, infectious, chemical or physical, will accelerate
and will influence the site of the atherosclerotic
proce^. These findings have given rise to the theory
that subtle, perhaps submicroscopic, alterations in
the physicochemical state of the arterial wall may
actually precede the more gross lipid accumulations.
PATHOGENESIS
Since the earliest pathologic lesions are only adum-
brative, even with modern histochemical techniques,
it follows that the intimate pathogenesis of the
atherosclerotic lesion also remains obscure. It is
understandable that nineteenth century pathologists
considered the disease a degenerative one, an inevi-
table concomitant of the aging process and a simple
result of wear and tear on the arterial wall. Even
when atherosclerosis was separated from other arte-
LIPID METABOLISM
I l6g
rial lesions, it must still have appeared to pathologists
of that era to be another phenomenon of aging, found
along with cataracts, osteoarthritis and wrinkled skin,
and occurring with increasing frequency with advanc-
ing years.
The pendulum did not swing until Ignatowski (1 10)
in 1908 succeeded, by administering lipid-rich foods
to rabbits, in producing arterial lesions similar to
those occurring spontaneously in human subjects. A
few years later Anitschkow (10) demonstrated con-
vincingly that cholesterol in the diet was the athero-
genic factor in experimental rabbit atherosclerosis.
Since that time, a large body of evidence has led
away from the degenerative theory. Some of these
evidences are: the finding at autopsy of an occasional
octogenarian virtually free of arterial disease; the
contrary finding of fatal atherosclerosis in soldiers in
their twenties; the relative freedom from the disease
of premenopausal women; the increased incidence at
autopsy of atheromatous lesions in patients with
diseases involving lipid abnormalities such as diabetes,
hypothyroidism, and other processes associated with
hypercholesteremia and hyperlipidemia. Further
information resulted from the epidemiologic finding
of certain population groups relatively free of the
disease at postmortem examination. Pathologists
made further contributions to this change in concept
by their studies in experimental animals; it is now
recognized that, although man, certain other pri-
mates, birds, and swine are the only animals which
seem regularly to acquire atherosclerosis sponta-
neously, there is a wide variety of species which can
be caused to develop arterial lesions similar to those
found in human material, providing only that appro-
priate experimental manipulations involving lipid
metabolism are made.
All this pathologic evidence, along with a huge
volume of clinical, epidemiologic, and biochemical
studies, has led to the modern concept that athero-
sclerosis is potentially a preventable disease, a result
of metabolic disorder rather than a degenerative
process. This lipid concept of the pathogenesis of
atherosclerosis can be stated in simple terms: man
ingests an excess of lipid which overwhelms the mech-
anisms for its disposal; lipid then accumulates in the
circulating blood and is deposited in the arterial wall.
How does this rather simple concept fit with the
facts of the earliest recognizable pathologic lesion
described above? At first glance, the fit seems perfect.
An excess of lipoprotein material in the circulating
blood filters through the endothelium of the arterial
wall and is taken up there by tissue histiocytes to
form foam cells; the simple accumulation of these
lipophages results in the gross arterial atheroma and
sets off the chain of events leading to fibrosis, throm-
bosis, and the rest. Yet there are a number of ques-
tions which cast doubt on this simple hypothesis.
First, if the mechanism is merely one of filtration
through the endothelium to the arterial intima, why-
are the anatomically similar veins not more susceptible
to the atheromatous process? That intraluminal pres-
sure plays some role is shown by the increased inci-
dence of atherosclerosis in hypertensive patients, the
occurrence of pulmonary artery atherosclerosis in
individuals with pulmonary hypertension, and the
finding of phlebosclerosis adjacent to arteriovenous
fistulas.
Another disturbing question concerns the fact
that the lipid deposit is not a universal arterial finding,
coating the intima of the entire arterial tree, but
rather a spotty, localized one, involving certain seg-
ments of certain arteries. A number of possible ex-
planations for this finding have been offered. One
argument is that localized changes in filtration pres-
sure, occasioned by intraluminal physical forces such
as whirlpool and eddy formation, determine the site
at which lipid is deposited; the frequent occurrence
of atheromatous lesions at bifurcations, branches and
coarctations favors this theory (175). Another pro-
posal relates the clinical predilection for thrombosis
in atherosclerosis to its pathogenesis; the earliest
lesion, by this concept, is a chance fibrin deposit on
the endothelial surface, the spotty lipid lesion occur-
ring secondarily to fibrin deposition (56). Another
explanation depends on the clinical and experimental
evidence that preceding arterial wall damage, physi-
cal, chemical, mechanical, or bacterial, will foster
premature and extensive lipid deposits; by this theory,
occult damage to the elastic tissue or ground sub-
stance (or both) of the arterial wall, from degenera-
tive or extraneous cause, serves as the spotty focus for
lipid deposit. Still another theory explains the spotti-
ness on the basis of localized differences in various
areas of the arterial wall, in the mechanisms for
removal of lipid, either metabolic (enzyme overload),
scavenging (number of histiocytes present) or ana-
tomic (number of lymph channels present). Yet
another hypothesis explains localization by denying
filtration from the lumen; according to this concept,
atheromatous lesions are preceded by a localized
overproduction in the arterial wall of the lipids which
make up the lesion (105). Sensitive radioactive tracer
studies have indeed shown that arterial tissue can
synthesize lipids, but recent reports (214) have indi-
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
cated that if any atheroma lipid comes from local
synthesis, it is probably only the phospholipid com-
ponent.
One other major question relating to the intimate
pathogenesis of the atherosclerotic lesion remains
unanswered: What are the mechanisms for incorpora-
tion of lipid into cells to form lipophages? In each of
the hypotheses mentioned above for the pathologic
background of the early lesion, the common hall-
mark, whether primary or secondary in time and in
importance, is the lipophage or foam cell (156), the
very name of which suggests incorporation of lipid
(fig. 1). The knowledge of the exact derivation of the
foam cell is fundamental to a better understanding
of atherogenesis.
Is anabolic activity neccsary for the accumulation
of lipid in the cellular cytoplasm (as opposed to an
engulfing mechanism)? One clue may be that the
lipophage differs from the lipocyte, or adipose-tissue
cell, by the former's higher content of protein and
lipids other than neutral fat. Current attempts to
studv this problem bv in vitro tissue culture tech-
niques (180) may help to answer this and other
important questions about the role of the lipophage
in atherogenesis.
Do lipophages form simply because there is lipid
material available to be engulfed or phagocytized?
In favor of this concept is the observation that lipo-
phages are not peculiar to the atherosclerotic lesion;
they are found as part of the detritus in hemorrhage
into the various tissues; they are found as apparent
scavengers in lipoid pneumonia; they occur in
degenerating tumors; they are found experimentally
after the subcutaneous injection of cholesterol sus-
pensions; and they are found in the lipoidoses
(Niemann-Pick, etc.) in massive accumulations in-
volving the reticuloendothelial system. In most, if
not all, of these situations, it is reasonable to assume
that a scavenging attempt to rid the tissue of a local
excess of lipid is involved.
Yet, in the atheromatous lesion, there appears to
be an additional element — one of accumulation. A
single minute atheroma, invisible to the naked eye,
is made up of a tremendous number of lipophages,
packed together in the subintima in such volume as to
displace adjacent normal tissue and to project into
>*
fig. I. Photomicrograph of a
foam cell protruding from the
subendothelial space of a rat
aorta. X 8,400. (Courtesy of
Robert M. O'Neal, Baylor
University.)
LIPID METABOLISM
I 171
the lumen. What the stimulus is to cause this pro-
liferative (or accumulative) element is unknown; the
answer to this question is vital to a proper under-
standing of the intimate pathogenesis of athero-
sclerosis.
In summary, the major facts concerning the pa-
thology of atherosclerosis, particularly its grosser
aspects and its sequelae, are well documented. Debate
still exists, however, concerning the more subtle,
microscopic manifestations of the early atheroma,
particularly in regard to the primacy of lipid deposi-
tion.
After more than fifty years, the lipid theory, despite
some unanswered questions, seems to be standing the
test of time. It will not become a universally accepted
theory until certain difficulties are overcome. The
spotty localization of arterial lesions, the mechanism
of incorporation of lipid into tissue cells, and the
stimulus to cellular accumulation in atheromatous
lesions all are unsolved problems.
METABOLIC CONSEQUENCES OF INGESTION OF FOOD
rebuilt by the body into new protein. Some of the
amino acids can be converted to carbohydrate and
thence to fat. Their carbon skeletons also are available
for oxidation.
Following digestion, fatty acids passing through the
intestinal mucosa are incorporated into very low-
density lipoproteins (chylomicrons); these "mole-
cules" are distributed in the systemic circulation to
be disposed of by hydrolysis, oxidation, interconver-
sion (but not into carbohydrate), or storage in various
tissues.
Thus, carbohydrate and protein can be converted
to and stored in the body as fat.
Soon after a conventional meal has been consumed,
changes in concentration of glucose, amino acids,
and fat (chylomicrons) occur in the blood. These
"primary" changes induce "secondary" changes in
the metabolic state. Ingestion of fat is followed by a
postprandial lipemia, which may last for many
hours. Thus, to evaluate the serum lipids properly,
it is important to obtain blood samples from subjects
who are in the postabsorptive state.
Assimilation of foodstuffs is a condition of animal
life. Yet food is never deposited unchanged. For
absorption to take place, foodstuffs must be split,
and far-reaching chemical transformations follow the
absorption of digested food. The transformed food
may be oxidized for the immediate production of
energy or stored for short or long periods, depending
on the needs of the body. With the exception of cer-
tain essential nutrients, the body is able to synthesize,
interconvert, store, and mobilize its constituents.
When an individual ingests an assimilable carbo-
hydrate, practically all of it is absorbed from the
digestive tract and eventually reaches the liver as
hexose. Part of the hexose is converted to liver glyco-
gen; part is released into the circulation to be dis-
tributed to extrahepatic tissues; part enters muscle,
where it is either burned or stored as glycogen. Once
glucose enters muscle, becoming phosphorylated, it
can no longer leave as such. One of its breakdown
products, lactic acid, can diffuse out of muscle cells
and re-enter the circulation. The adipose cells trans-
form glucose into fatty acids, which are esterified with
a-glycerophosphate to form triglyceride and are
stored in this form. Glucose products, by a process of
transamination, can be converted into amino acids.
Protein must be hydrolyzed into amino acids prior
to absorption. Subsequently the amino acids can be
DIET
The average American diet, according to a sum-
mary of the 1 955 Household Food Consumption
Survey conducted by the United States Department
of Agriculture (65), derives 44 per cent of its caloric
content from fat, 1 3 per cent from protein, and 43
per cent from carbohydrate. The survey made no
deductions for food discarded. The breakdown of
calories derived from fat was: 18.3 per cent from satu-
rated fatty acids, 18.6 per cent from oleic acid, and
4.5 per cent from linoleic acid. As expected from such
an extensive survey, there were some regional differ-
ences in types and quantities of food consumed.
Knowledge of the chemical composition of natural
fats remains incomplete, although great strides for-
ward are being made. It is generally agreed that
most natural fats, whether animal or vegetable, con-
tain about 98 to 99 per cent triglycerides. The re-
maining 1 or 2 per cent includes diglycerides, mono-
glycerides, free fatty acids, phospholipids, and
unsaponifiable sterols. Fatty acids comprise over go
per cent of the triglycerides, with the remainder
being glycerol. The naturally occurring triglycerides
are mixtures varying widely in their patterns of fatty-
acids. The complexity of such glycerides is underlined
1 172
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
by the observation that at least 64 different fatty
acids have been identified in butter fat (101).
In general, the degree of unsaturation of the fat
depends upon the source of the fat. Fats of aquatic
origin contain a wide range of unsaturated C16, Cis,
C2o, and C22 acids. Fats from land animals contain
25 to 30 per cent C^, the remainder being mostly of
the Cis series (102). The so-called essential fatty acids
(mainly linoleic) apparently cannot be synthesized
by animals, and must be obtained from the diet.
The depot fat of certain animals, such as the pig, can
be varied markedly in its content of linoleic acid,
depending on the feed (60). Similarly, the adipose
tissue fatty acids of man eventually reflect the dietary
fatty acid pattern. This is true with respect to linoleic
acid (103); but the medium chain fatty acids (C12
and below) have not been identified in the fat depots.
Fats of vegetable origin vary tremendously in their
pattern of fatty acids, as well as in their degree of
unsaturation. For instance, coconut oil contains only
a small quantity of linoleic acid, while safflower oil
may contain 70 per cent or more. Dietary fats should
not be described merely as "animal or vegetable,"
"saturated or unsaturated," but the actual composi-
tion in terms of fatty acids should be identified. Thus,
when the effects of dietary fats on lipid metabolism
are being evaluated, the specific fatty acids involved,
their chain length, isomeric configuration, degree of
unsaturation, and relative proportion in the diet
must be considered.
Dietary phospholipids are found as complex mix-
tures in organ fats and certain raw vegetable fats,
rather than in depot fats. Egg yolk is a rich source of
phospholipid. As indicated by their name the phos-
pholipids area group of phosphorus-containing lipids;
in addition, they contain a nitrogenous base. The
lecithins, in which the base is choline, and the cepha-
lins, in which the base is ethanolamine or serine, are
classified as mono-amino-phosphatides. The com-
ponent fatty acids are usually both saturated and
unsaturated. The inositol phospholipid contains
ethanolamine and tartaric acid. It is found in soy-
bean phospholipids and in brain tissue. Other phos-
pholipids include sphingomyelin, which is a diamino-
phosphatide containing choline and sphingosine.
Plasmalogens contain higher fatty aldehydes and
ethanolamine (69). The complexity of dietary phos-
pholipids is illustrated by their occurrence in egg
yolk: 72.8 mols per cent phosphatidyl choline; 14.8
per cent phosphatidyl ethanolamine; 2.1 per cent
lysophosphatidyl ethanolamine; 5.8 per cent sphingo-
myelin; 0.9 per cent plasmalogen; 0.6 per cent inositol
phospholipid; and 0.2 per cent phosphatidyl amine
acids (169). One egg contains about 2 g of phos-
pholipid.
The unsaponifiable fraction of food fats consists of
sterols, including; cholesterol (absent from vegetable
fats), long-chain aliphatic alcohols, glycerol ethers,
pigments, etc. Finally the fat-soluble vitamins, A, D,
E, and K, may be found in this fraction.
FAT DIGESTION AND ABSORPTION
Generally speaking, lipids are not readily miscible
in water. To be able to absorb, transport, and utilize
fatty acids and other lipids, man has had to evolve
rather elaborate mechanisms for making these water-
immiscible or hydrophobic materials compatible with
a system whose basic medium is water. The mech-
anisms used to deal with the water-insoluble lipids
as they enter the body include hydrolysis, emulsifi-
cation, chemical combinations with substances con-
taining hydrophilic groups, and complex formation
with substances conferring greater water miscibility
and dispersibility, such as bile acids and proteins.
The mechanisms of digestion and absorption of
dietary fat have been subjects of controversy for many
decades. An early theory was proposed by Pfluger
(161) who described dietary fats as being emulsified
by bile salts in the small intestine. The triglycerides
were then completely hydrolyzed by pancreatic lipase
to fatty acids and soaps. Being water-soluble, these
products were readily absorbed. However, it soon
became apparent that intestinal pH is too low for
fatty acids to exist as soaps. It also became apparent
that the absorbed fat in lymph is mainly in triglyceride
form. Thus, glyceride resynthesis by the intestinal
mucosa was postulated. The modern concepts of fat
absorption arise from Frazer's work (67, 68). It is
now believed that hydrolysis of dietary glycerides
need not be complete in order for absorption to occur.
Frazer's original "partition theory" (66) postulated
that fatty acids passed directly into the portal circula-
tion while the partial and unchanged triglycerides
were somehow transported across the mucosa into
lymph as chylomicrons. This theory has failed to
survive in its original form as a result of more recent
work (16, 145) including Frazer's own (70). Portal
venous transport of fat is now known to occur only
with fatty acids of less than ten carbons, which com-
prise less than 5 per cent of dietary fats.
The digestion and absorption of long-chain fats
remain a subject of controversy. Frazer (70) has
LIPID METABOLISM
11 73
presented evidence that finely emulsified fat particles
of 0.5 fx or less in diameter can penetrate intact the
small spaces between mucosal "microvilli.'' More
recently, evidence has been presented that hydrolysis
of triglycerides in the intestinal lumen is extensive but
incomplete, and that approximately 65 per cent of
the fatty acids is absorbed in the free form while 35
per cent passes into the mucosa as glycerides ( 1 7, 29).
The mucosal cells resynthesize the free fatty acids,
the monoglycerides, and diglycerides into triglycer-
ides. Studies (48, 49, 127) have indicated that in-
testinal mucosal synthesis of triglycerides proceeds
along pathways similar to those defined by Kennedy
and his associates (206) for hepatic triglyceride syn-
thesis. In contrast to what may happen in the liver,
free glycerol does not appear to be a starting material,
although quite recently evidence for free glycerol
incorporation into triglyceride in the intestine in man
has been reported (107). Free fatty acids become
activated by linkage with coenzyme A. Two such
activated fatty acids combine with L-a-glycerophos-
phate to form diglyceride phosphate (phosphatidic
acid) which can then form diglyceride following
dephosphorylation by a suitable phosphatase. The
diglyceride reacts with a third activated fatty acid
to yield a triglyceride, or, like phosphatidic acid,
may also be converted to phospholipid. The major
pathway is that of triglyceride synthesis. A small
amount of phospholipid and cholesterol gets in-
corporated with the triglycerides into the "chylo-
microns" which range in diameter from 350 A to
0.5 /u. The chylomicrons are subsequently discharged
into the intestinal lacteals from which they drain into
the thoracic duct and ultimately the systemic circula-
tion. Studies on the composition of chylomicrons in
man, rat, and dog (98, 160, 182) have shown that
they consist of 85 to 93 per cent triglyceride, 8 to 1 1
per cent phospholipid, 1.5 to 4.5 per cent cholesterol
(free and ester) and 1.9 to 2.5 per cent protein (|3-
globulin). Some events that occur during fat digestion
and absorption are summarized in figure 2.
ADIPOSE TISSUE
The concept that adipose tissue is a dynamic
"organ" capable of participating in a number of
metabolic processes is now generally accepted. The
effect of caloric abundance or inadequacy on the
quantity of stored fat is well recognized. A variety of
stimuli are known to induce an increase or a decrease
in body fat. However, the mechanisms involved in
fat deposition (lipogenesis) and fat release (lipolysis)
are not completely understood. Most body fat acts as
a highly efficient caloric reservoir. However, it must
be remembered that this reservoir is composed of
myriads of living cells the function of which includes
svnthesis and mobilization of fat as well as storage.
The number of calories stored as fat is necessarily
a function of two variables, energy intake and energy
expenditure. A normal young adult man in caloric
balance may contain an average of 14 per cent pure
fat (116). Thus a young man weighing 70 kg may
carry approximately 10 kg of fat. This is almost two
and a half times the weight of his bone minerals.
The relative amount of fat in the body has been
shown to increase with age reaching, at age 55,
approximately 25 per cent of body weight in clini-
LUMEN
TRIGLYCERIDE
MUCOSA
FATTY ACID
CHOLESTEROL
CHOLESTEROL
Esters and Ereel
LYMPHATICS
-CHYLOMICRONS*
Triglyceride 89%
Phospholipid 7%
Cholesterol 2%
Protein 2%
fig. 2. Schema of lipid absorp-
tion (long -chain fats). * Propor-
tions of chylomicron constituents
vary with diet.
1 1 74
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
cally normal men (38). In obese individuals the
amount of body fat may reach 33 to 40 per cent of
body weight.
During the past twenty years there has been an
increasing awareness that adipose tissue is a dynamic
organ capable of responding to a variety of stimuli.
The experiments of Schoenheimer ( 1 84) and others
have shown that depot fat has a definite turnover. Its
half-life in the rat has been estimated to be 6 to 8
days. In man, turnover of depot fat is much slower,
with a half-life of many months, in the presence of
adequate caloric intake. When the availability of
carbohydrate is reduced, depot fat cells can quickly
mobilize free fatty acids which are then bound to
albumin and carried in the blood to muscle, liver,
and other tissues (52, 84).
The relationship between the availability of carbo-
hydrate and the rate at which fatty acids are mobilized
from depot fat is important. Breakdown of glucose
through the Embden-Myerhof pathway and the
hexosephosphate shunt may provide certain cofactors
necessary for fatty acid synthesis (lipogenesis). Insulin
promotes entry of glucose into the cell and thereby
provides a stimulus for further metabolism of this
hexose. Fatty acid synthesis appears to be somehow
dependent on the rate of glycolysis. In diabetes melli-
tus glycolysis is depressed because of insulin lack; in
addition, the rate of fatty acid synthesis is greatly
suppressed. At the same time, the rate of hydrolysis
of depot fat increases, and the resultant fatty acids
are carried into the circulation as free fatty acids.
For example, in uncontrolled diabetes, enormous
amounts of fat can be mobilized, leading to fatty
liver, hypertriglyceridemia, and hyperketonemia.
Administration of insulin corrects the situation in a
manner that has not been elucidated (203), but prob-
ably relates to the ability of insulin to reduce the rate
of fatty acid mobilization from adipose tissue.
Hormonal Influences on Adipose Tissue
In addition to insulin, several other hormones have
been found to influence lipid mobilization. It is em-
phasized that hormones do not initiate events within
cells, but merely regulate the rate at which some of
these events occur. Early studies (148) indicated
that when large doses of posterior pituitary extract
were injected into rats or rabbits there resulted an
accumulation of fat in the liver. Similarly, an increase
in liver fat and a decrease of carcass fat of the rat
were found following the injection of anterior pitui-
tary extract (19). There have been recent reports
that a posterior pituitary component, a relatively
small polypeptide, has a potent mobilizing effect on
omental and mesenteric fat in animals and man (185,
213). As yet there have been no reports on the influ-
ence of this posterior pituitary material on mobiliza-
tion of free fatty acids (95). In contrast, it was found
(177) that the injection into rabbits of crude extracts
of whole or anterior pituitary gland of hogs, sheep,
cattle, or man induced visible lipemia which was con-
siderably greater than that induced by recognized
anterior pituitary hormones. Thus it was suggested
that the lipemia-producing principle might be an
independent hormone, or that the lipemia was the
result of synergistic action of known hormones, al-
though the lipemia-producing anterior pituitary
component apparently contains negligible or un-
detectable amounts of eight known pituitary hormones
(178). Also, following injection of this material into
rabbits, there was a rapid and enormous (tenfold)
increase in free fatty acid levels. This was followed
within 12 hours by a twofold to fivefold increase in
serum total lipid concentration, including significant
increases in the serum levels of triglycerides, choles-
terol, and phospholipid. It is difficult to ascertain
from these studies whether or not the lipemia that
was produced was mediated through another gland,
since the recipient animals were neither hypophy-
sectomized nor adrenalectomized. The authors have
explained their results by postulating that the active
principle mobilizes free fatty acids from depot fat.
The increase in triglyceride, cholesterol, and phos-
pholipid in serum is believed due to increased rate of
formation of these substances in the liver, in response
to increased hepatic uptake of free fatty acids.
The possibility that there may be two new factors
capable of mobilizing depot fat, one from the anterior
and one from the posterior pituitary, deserves further
exploration and confirmation by other investigators.
Of the known pituitary hormones, growth hormone
has been shown to possess free fatty acid-mobilizing
properties in the intact organism (168). Growth hor-
mone action on depot fat (and other sites) appears
to depend upon the phylogenetic relationship of the
recipient animal to the donor source. For example,
beef growth hormone mobilizes fat in cattle and in
the rat but not in monkeys or human subjects. Knobil
(125) has shown that the administration of growth
hormone stimulates, while hypophysectomy inhibits,
free fatty acid release from the epidydimal fat body of
rats. The manner in which growth hormone promotes
free fatty acid release has not been fully elucidated.
Adrenocorticotropic hormone (ACTH) has been
found to be very active in inducing free fatty acid
release from adipose tissue in vitro (43, 61). However,
such an effect is not readily demonstrated in vivo.
The reason for this discrepancy is not clear, although
a difference in species response to ACTH with respect
to free fatty acid release has been offered as a possible
explanation (61 ).
The role of thyroid in the metabolism of depot fat
also is under study. The lipolytic response of adipose
tissue from thyroidectomized animals (which had
been suppressed) was made normal by restoration of
the euthyroid state, while hyperthyroidism accen-
tuated such a response (170). Moreover, it has been
shown that the treatment of hypothyroid patients
with thyroid restored to normal their free fatty acid
response to growth hormone (167). Thus the thyroid
appears to play a permissive role in fat mobilization,
potentiating the action of certain lipolytic agents.
The autonomic nervous system was long suspected
of playing an important part in the metabolism of
depot fat (24, 207). Early observations on the auto-
nomic innervation of adipose tissue have received
support from more recent studies on the action of epi-
nephrine and norepinephrine on depot fat. Subcu-
taneous administration or intravenous infusion of
epinephrine and norepinephrine induced significant
elevations of free fatty acids in the plasma of intact
animals and human subjects (52, 80, 84). Moreover,
adipose tissue in vitro has been found to be exquisitely
sensitive to epinephrine and norepinephrine in terms
of free fatty acid release (27, 61, 137, 210). It has
been shown that adipose tissue liberates glycerol in
response to epinephrine and norepinephrine (61, 137)
suggesting that the mode of action of these hormones
on adipose tissue involves hydrolysis of triglyceride.
An epinephrine-sensitive lipolytic system has been
reported in adipose tissue (172). The influence on
depot fat of chronic administration of epinephrine
and norepinephrine, as well as other sympathomi-
metic agents, remains unknown.
The various factors influencing lipogenesis and
mobilization of fat from adipose tissue are summarized
in figure 3.
In view of the importance of the subject of lipid
mobilization and lipogenesis in the scheme of knowl-
edge about metabolism, it is surprising that investiga-
tion in this area has lagged until recently. Conceiv-
ably, as knowledge of the subject increases, the
physician will be the beneficiary of valuable adjuncts
in the treatment of lipid disorders. Thus, adipose
tissue, once thought to be a relatively inert storehouse
of dense calories, has been found capable of partaking
Glucose
Insulin
LIPID METABOLISM
Fatty Acid - Albumin
'I 75
CO,
-ACETYL CoA
GLUCOSE 0 - P
x- Glycer
FATTY ACID -CoA
Glycerol
Fat
Droplet
Chylomicron
fig. 3. A scheme of lipogenesis and lipolysis in the adipose
cell. Free fatty acid release is promoted by: glucose lack,
starvation, insulin lack, epinephrine, glucagon (in vitro),
norepinephrine, growth hormone, ACTH (in vitro), thyroid
hormone (? permissive), and certain extracts from anterior
and posterior pituitary glands.
actively and rapidly in a number of metabolic proc-
esses and of responding to a variety of humoral and
autonomic stimuli.
THE SERUM LIPIDS
The serum lipids can be broadly classified under
two major headings: the lipoproteins and the free
(or nonesterified) fatty acids (FFA or NEFA). The
lipoproteins represent a whole spectrum of lipid mole-
cules containing varying proportions of phospholipid,
cholesterol (free and esterified), protein (polypeptide),
glyceride, and water. The lipoproteins have been
classified according to their behavior in the ultra-
centrifuge, by their migratory behavior during elec-
trophoresis, by their principal X-terminal amino
acid residue, by their solubility characteristics, and
other ways (155, 194). The density of the lipoprotein
molecule is largely a function of the proportion of
lipid contained within it; hence, the more "obese"
the molecule, the lower its density.
The lipoprotein species with the lowest density are
the chylomicrons; the remaining species can be further
divided according to their electrophoretic migration
with alpha globulins and beta globulins into two
groups, the lower density (beta) lipoproteins and the
higher density (alpha) lipoproteins.
The free fatty acids are present in a relatively low
concentration in plasma under basal conditions
(approximately 0.2 to 12.0 meq liter). However,
they appear to have a rapid turnover rate. They
travel in the circulation bound to albumin and per-
176
HANDBOOK OF PHYSIOLOOY
CIRCULATION II
haps to other substances, and consist principally of
fatty acids common in the diet, such as palmitate,
stearate, oleate, and linoleate.
Chylomicrons
The elaboration of chylomicrons by the intestinal
mucosa has been discussed. These small particles
enter the systemic circulation via the intestinal lac-
teals and thoracic duct. The chylomicrons in the
blood are responsible for the visible lipemia that
occurs after a meal containing an appreciable quan-
tity of fat. Similar (but not identical) particles manu-
factured by the liver seem to be responsible for the
lactescence that occurs in uncontrolled diabetes,
nephrosis, and carbohydrate-induced hyperlipemia.
From an analytical standpoint, chylomicrons have
been characterized as the material floating at the top
of a tube when chyle or serum is layered under sa-
line of density 1 006 and centrifuged for a few min-
utes at high speed. Varying speeds and time of cen-
trifugation have been suggested, but it is assumed
that the lower the speed and the shorter the time
of centrifugation, the purer will be the chylomicron
fraction (71). One procedure (135) uses 9500 g for
10 min. The actual density of chylomicrons is 0.94 g
per ml.
When chylomicrons are released into the systemic
circulation from the thoracic duct, they are removed
with considerable rapidity by the liver and extra-
hepatic tissues. The mechanisms of removal are in-
completely understood; however, these small fat
particles apparently are not hydrolyzed to any appre-
ciable degree in the circulating blood, although evi-
dence for intravascular hydrolysis has been pre-
sented (62). To some extent, fatty acids may be split
away from chylomicron triglyceride through the
intervention of the enzyme, lipoprotein lipase, but
this action probably occurs primarily at endothelial
surfaces and other cell membranes and not in the
main stream of the circulation (71).
More important mechanisms for chylomicron
removal may include direct diffusion into cells through
"pores," and phagocytosis by appropriate cells.
When the subject is in the postabsorptive state and
carbohydrate no longer is readily available, a larger
proportion of the chylomicrons from a test meal of
fat will be removed by liver and muscle; when excess
carbohydrate is available (that is, during hyper-
glycemia), the chylomicrons are shunted to a greater
extent to the fat depots. The clearing of visible
lipemia after ingestion of fat may be inhibited if a
previous fat load has been given and has recently
been cleared. Although a large amount of chylo-
micron fatty acid is directly oxidized, some of it may
recirculate in the form of free fatty acid. Thus, a
variable rise in free fatty acids occurs in blood during
the course of an alimentary lipemia (71, 72).
The Lipoproteins
In normal human plasma, the lipoproteins consti-
tute approximately 12 to 15 per cent of the total
protein (155). A fundamental difference between the
lipoproteins and the remainder of the plasma pro-
teins is that the former are lipid-laden molecules with
relatively low density. The plasma lipoproteins ex-
hibit densities from 0.9 to 1.2, in contrast to densities
of 1.26 to 1.38 for most other proteins. Thus, the
ultracentrifuge has become a useful tool in the sepa-
ration of the plasma proteins and lipoproteins, utiliz-
ing measurement of the differing sedimentation rates
of molecules in solvent systems of known density.
Gofman and his associates (78, 79, 135) have used
a sodium chloride solution with a density of 1 .063 to
differentiate lipoproteins. These workers have also
introduced the "Sf" nomenclature which is used to
describe the rates of flotation (varying Sf values) of
the various lipoproteins in sodium chloride of density
of 1.063. Such rates of flotation are measured in
Svedberg units (Sf unit, io~13 cm g sec-1 dyne-1).
Apparently the Sf value is dependent on the density,
shape and size of the lipoprotein molecules (155).
A plethora of terminology relating to the lipopro-
teins has evolved depending on methods of isolation
and identification (71, 78, 134). The subdivisions
sometimes have been rather arbitrary, yet certain
correlations have been made as, for example, be-
tween the ultracentrifugal and electrophoretic
behavior of the lipoproteins. It has been demon-
strated that there are two major groups of lipopro-
teins (three if chylomicrons are included) in human
plasma: /) high-density lipoproteins (density > 1.063),
or a-lipoprotein by virtue of their electrophoretic
mobility; 2) low-density (< 1.063) or /3-lipoproteins.
The latter (< 1.063) include the classes Sf 0-400 of
Gofman (78). The chylomicrons (density 0.94; see
above) have virtually no electrophoretic mobility,
but exhibit no definite line of demarcation from the
Sf 400 low-density lipoproteins, and their Sf value
may reach 40,000. Studies of the protein moiety of
the lipoproteins have yielded information which
may further help characterize the ultracentrifugally
separated fractions. Information is available (71)
LIPID METABOLISM
■77
with respect to molecular weight, end group analyses
for specific peptide chains, amount and type of N —
and C — terminal amino acids, etc. Such information
remains preliminary in nature. A schematic concep-
tion of the various human plasma lipoproteins is
shown in figure 4.
The lipid moieties of the various lipoproteins (other
than chylomicrons) may comprise from 40 to (per-
haps) 90 per cent of the molecule. These include a
small amount of free fatty acids, and virtually all the
esterified fatty acids as esters of glycerol or more
complex alcohols, and cholesterol. Free cholesterol
also is present. Normal human postabsorptive plasma
contains approximately 400 mg per 100 ml esterified
fatty acids. Approximately 70 per cent of such fatty
acids exist as triglycerides and phospholipids, and
the remainder as cholesterol esters (141). Small
amounts of esterified fatty acids may be found as
diglycerides and monoglycerides, cerebrosides and
acetals (71).
Despite the high proportion of lipid in the lipo-
proteins, they have the chemical and physical char-
acteristics of protein molecules. Such behavior sug-
gests that the protein moiety is on the surface of the
molecule. For instance, it has been estimated (35)
that a chylomicron of 0.5 ju may be covered com-
pletely by protein, assuming the protein to be all at
the surface and constituting 1.5 per cent of chylo-
micron. However, it has been pointed out that in the
larger ^-lipoprotein there is only enough protein to
cover about half the surface, assuming a thickness of
one peptide chain. On the basis of titration data,
Oncley el at. (154) have postulated a kind of mosaic
surface comprising both peptide and phospholipids,
the latter being oriented with their charged groups
at the surface.
Under ordinary circumstances, the alpha lipopro-
teins, or high-density lipoproteins, probably do not
transport triglyceride for oxidative purposes. Recent
studies (176) with labeled amino acids have suggested
that the plasma does not contribute a major portion of
the protein found in either the chylomicrons or the
high-density lipoproteins in thoracic lymph. Since the
cells of the intestinal mucosa incorporate amino acids
into proteins having the same electrophoretic mobility
as chylomicron protein, it was theorized that the in-
testine may be the source of the protein of both the
high-density (alpha) lipoproteins and the chylo-
microns.
The alpha lipoproteins contain approximately 40
per cent lipid; are not remarkably influenced by diet
or fasting and do not increase with age. In terms of
blood level, they are relatively stable.
In contrast, the low-density (beta) lipoproteins,
which contain 75 per cent or more lipid, are labile;
they are affected by diet, fasting, age and gonadal
DENSITY
40,000
LIPOPROTEIN SPECIES
PRINCIPAL
N-TERMINAL AMINO
ACID RESIDUE
fig. 4. Schematic conception of
human lipoproteins. Cross-hatching or
stippling represents polypeptide portions
of the molecules. [From Frederickson
& Gordon (71).]
1 1 78
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
hormones, and by a variety of other influences. In
fasting states and when carbohydrate utilization is de-
creased, serum levels of these low-density lipoproteins
tend to increase; however, appreciable rises in concen-
tration of such molecules occur only after the fast has
been sustained. Ordinarily, such a rise in circulating
beta lipoproteins is preceded by an increase in plasma
levels of free fatty acids.
There is some evidence that lipoprotein interconver-
sions occur and that, as fatty acids are split off a low-
density lipoprotein molecule, its density progressively
increases. When heparin is administered, thereby stim-
ulating lipoprotein lipase activity, the interconversion
process is greatly accelerated (62).
Free Fatty Acids
The free fatty acids comprise less than 10 per cent
of the total fatty acids found in plasma. Strictly speak-
ing, these acids are not "free" since they circulate
bound to albumin. Each molecule of albumin can
bind two or more molecules of long-chain fatty
acid (83).
The free fatty acids represent one important form in
which fatty acids are transported from sites of storage
(fat depots) to working cells (see above). They do not
appear to derive directly from dietary fat. However,
dietary fatty acids with chain lengths of C10 or less, a
very small fraction of fat in the diet, may enter the
circulation from the gut via the portal vein in the
"free" form (28), although their esterification with
glycerol by the intestinal mucosa may also occur (160).
Actually, the level of circulating free fatty acids falls
after a normal meal. On the other hand, during the
course of clearing of alimentary lipemia, a variable
fraction of the circulating free fatty acids may originate
from chylomicron triglyceride (71).
Studies of the distribution of C14-labeled palmitate
(bound to albumin) in rats 15 min after intravenous
injection have disclosed a general uptake of the label
by various organs and by muscle. Liver lipids were
particularly active, whereas adipose tissue had no
detectable activity (53). If such experiments can be
considered representative of the behavior of free fatty
acids under normal circumstances, it would seem that
the plasma free fatty acids are removed rapidly in
various parts of the body, with subsequent oxidation
or esterification, depending upon metabolic circum-
stances.
There is growing evidence that the free fatty acids
constitute an important source of the body's energy
in the fasting state; they are released in increasing
amounts by the adipose tissue and used very rapidly
at times when carbohydrate utilization is diminished
(52). Experimentally induced elevation of the blood
sugar or of blood amino acids (94) results in a recipro-
cal drop in plasma levels of free fatty acids.
Calculations suggest that the circulating free fatty
acids probably do not supply more than 50 per cent
of the energy in the fasting state; hence, it may be
suspected that another important source of energy
during fasting is esterified fatty acid (71). Isotope
studies have shown that the fatty acid turnover is much
more rapid in triglyceride than in cholesteryl ester
and phospholipid (13). Thus, the serum glycerides
may well function as a major vehicle for transport of
esterified fatty acid to sites of utilization.
Ordinarily, in the postabsorptive state, triglyceride
moieties travel in the blood as parts of low-density
lipoprotein molecules that are neither large nor nu-
merous enough to affect the gross clarity of the serum.
However, under conditions of metabolic stress, and in
certain disorders, lipoproteins of very low density
appear in the circulation in quantities sufficient to
render the plasma lactescent. It is unlikely that such
particles can be released directly from the fat depots
and evidence is accumulating that they originate from
the liver.
Role of the Liver
The liver plays a major role in the synthesis and
disposal of lipids and lipoproteins. With the exception
of the chylomicrons, it appears that lipoprotein
manufacture takes place principally, if not exclu-
sively, in the liver. The various factors that influence
hepatic synthesis of lipoproteins are not well under-
stood. The process whereby the intestinal mucosa
handles dietary fats and transforms them into chylo-
microns, some of which are removed by the liver, has
been discussed already. Adipose tissue releases fat in
the form of free fatty acids "bound" to albumin. These
acids also are extracted in appreciable quantities by
the liver. Carbohydrate and protein can be converted
into fat by the liver. In short, the liver is presented
with lipid from several sources, and can itself synthe-
size lipid, including cholesterol and phospholipid.
Thus, the lipoproteins that the liver manufactures
and sends out to the circulation are made from a
variety of building blocks and are subject to a variety
of metabolic influences including diet and hormones.
These concepts concerning the origin of serum lipids
are shown in schematic form in figure 5. In the case
of certain lipids, such as cholesterol, the liver is the
LIPID METABOLISM
'79
V, A T/ 0
FFA
CHYLOMICRONS
CARBOHYORATE
fig. 5. Origins of serum lipids.
[From Van Itallie & Felch (201).]
major organ of catabolism and excretion. Elaborate
mechanisms exist in the liver for disposal of the
steroid nucleus of cholesterol, since the body lacks
the mechanisms capable of opening the rings of
phenanthrene-like structure.
Cholesterol Disposal
As mentioned above, cholesterol represents a spe-
cial disposal problem. Fatty acids and glycerol are
readily metabolized, and phospholipids are freely
miscible with water and can be degraded rapidly. On
the other hand, although the isopropyl side chain of
the cholesterol molecule can be oxidized, with forma-
tion of bile acids and certain hormones, the steroid
nucleus itself is not degraded.
It is now established that bile acids constitute the
major catabolic end products of cholesterol metabo-
lism in man and in a variety of animal species (17, 18,
97). In man, conversion of cholesterol to bile acids oc-
curs in the liver. The biochemical details of this con-
version have not been completely worked out. It is
generally agreed that in man two '"primary" bile
acids (hydroxycholanic acids) result from catabolism
of cholesterol in the liver: cholic acid and chenodeoxy-
cholic acid (fig. 6). Five important biochemical
changes must occur in the cholesterol molecule, and
not necessarily in the order listed : a) isomerization of
the 3 /J-OH into 3 a-OH; b) saturation of the 5:6
double bond; c) hydroxylation at the 7 position
(chenodeoxycholic) d) hydroxylation at both 7 and
1 2 positions (cholic) ; and e) oxidation of the terminal
isopropyl group resulting in a C-24 acid (cholanic).
The resultant bile acids are secreted into the extra-
hepatic biliary system as micellar conjugated com-
pounds with either glycine or taurine. In man the con-
jugation process favors glycine by a factor of three
(18). In the intestine the bile acids may undergo fur-
ther chemical transformations attributed to intestinal
microorganisms, giving rise to "secondary" bile acids.
For example, cholic acid will lose its 7 a-OH group to
yield deoxycholic (3a, i2a-hydroxycholanic) acid, and
in a similar manner chenodeoxycholic acid will yield
lithocholic (3 a-hydrOxycholanic) acid. Thus the
7 a-dehydroxylation is a bacterial function. A number
of additional bacterial metabolites of hepatic bile acids
have been found in human feces, although their impor-
tance quantitatively has not been determined (in).
In addition, the bacteria split the conjugated com-
pounds into glycine, taurine and their corresponding
bile acids. Most of the bile acids are reabsorbed from
the intestine into the liver via the portal vein. A small
portion is excreted in the stool in unconjugated form.
In the normal gastrointestinal tract virtually no
cholic acid can be identified in the feces.
It has been estimated that the normal adult indi-
vidual synthesizes about 1.2 g of cholesterol per day.
Approximately 70 per cent of this amount (0.8 g) is
Il8o HANDBOOK OF PHYSIOLOGY — CIRCULATION II
HO \"^ ^"-""^ OH
CHENODEOXYCHOLIC ACID
I
INTESTINAL
BACTERIA
OH
CHOLIC ACID
I
INTESTINAL
BACTERIA
*
OH
COOH
LITHOCHOLIC ACID
I
DEOXYCHOLIC ACID
C-NH-CH2-C00H (GLYCINE CONJUGATES)
■NHCH,-CH,-SO,OH (TAURINE ' )
I CONJUGATION IN LIVER
fig. 6. Formation of "primary" and "secondary" bile acids
in man. (From Van Itallie & Hashim, M. Clin. North America.
In press.)
excreted as bile acids, and a portion of the remainder
is excreted in the feces as cholesterol, coprostanol,
cholestanol, and other nonacidic sterols. The newly
formed bile acids are excreted into the small intestine
and participate in an enterohepatic cycle through
which 20 to 30 g of bile acids circulate per day
(fig. 7). The net loss of bile acids in the feces nor-
mally corresponds to the amount converted from
cholesterol in the liver.
That bile acids are the major catabolic products of
cholesterol has been suspected for about a century.
However, this fact was clearly established in 1943 by
Bloch et al. (26) who demonstrated that, when choles-
terol labeled with deuterium was administered to
dogs, a minimum of two-thirds of the label appeared
in the excreted bile acids. These observations have
now been amply confirmed by studies involving use
of C'Mabeled cholesterol, notably by Siperstein et al.
(191, 192), Bergstrom et al. (17, 18), and others.
The behavior of the enterohepatic cycle of bile
acids profoundly affects the rate of cholesterol con-
version to bile acids. This rate in turn appears to
influence plasma cholesterol concentration. Rats with
approx 0.8gm/day
in feces
fig. 7. Enterohepatic cycle of bile acids. [Adapted from
Bergstrom (18).]
indwelling cannulae in the common bile duct exhibit
a 10-fold to 15-fold increase in bile acid output
through the externalized cannulae compared with
the fecal bile acid output of intact rats (18). This also
has been observed clinically in patients with common
duct intubation. Conversely, in animals with experi-
mental ligation of the common duct (41), and in
patients with cholestasis (202), the rate of conversion
of cholesterol to bile acids decreases. Moreover, in
such situations cholesterol synthesis actually may
increase, sometimes with marked elevation of serum
cholesterol.
FACTORS THAT INFLUENCE SERUM LIPIDS
Serum lipid levels result from a complex interaction
between host and environment (157). Genetic dis-
orders such as familial hypercholesteremia and essen-
tial hyperlipemia are associated with gross abnormali-
ties of serum lipids. Among the environmental factors,
diet and "stress" have attracted particular attention.
Certain diseases also have conditioning effects on
serum lipids. Included in this category are nephrosis,
hypothyroidism, biliary cirrhosis, diabetes mellitus,
pancreatitis, and others. In addition, such factors
as age, sex, race, culture, occupation, exercise, body
composition, and cigarette smoking have been
LIPID METABOLISM
1 1 81
regarded by investigators as capable of exerting a
significant influence on serum lipids.
It is much simpler to speak of ■"abnormalities" in
serum lipids than to define what is ''normal." The
difficulty has arisen because of a discrepancy between
what is statistically normal and what is probably
physiologically desirable. Many authorities now
believe that statistical means for serum total choles-
terol in a population (like that of the United States)
that is beset with cases of atherosclerosis, clinical and
subclinical, may be misleading if they are used as
criteria of biologic normality. For example, in popula-
tion groups in which clinically manifest athero-
sclerosis is very rare, the mean serum total cholesterol
is frequently at least 30 per cent lower than it is in
the United States (118).
Stress
Despite many attempts to characterize "stress,"
this phenomenon remains to be defined in generally
acceptable terms and its physiologic effects codified.
The "epidemiology" of stress is exceedingly complex.
The response of the individual to his environment
seems to be much more critical than the events
overtly taking place in the environment. Hence, it is
difficult to assess the degree of stress inherent in a
given occupation unless one also knows how the
individual is reacting to the "life situation" with
which he is confronted. For these and other reasons
the effect of stress on serum lipids remains contro-
versial (11, 179). It has been reported that students
displav a transient elevation of the serum cholesterol
level immediately prior to important examinations
(197) and that accountants exhibit similar elevations
when deluged with income tax returns (73). Even if
it is granted that such changes occur, it is not yet
known whether the lipid elevations result principally
from direct neurohumoral stimulation or because of
some associated change in the habits of the person
concerned. For example, it is well known that under
stress, activity rates may change and certain persons
may eat more or otherwise change their pattern of
living.
There is some evidence that a certain type of "per-
sonality profile" is associated with predisposition to
coronary heart disease and that patients with such a
profile secrete significantly more epinephrine and
norepinephrine than control subjects (74, 75, 147).
The coronary-prone individual is characterized as
exhibiting excessive, frankly competitive drive and an
enhanced sense of time urgency (42). In addition,
this type of individual may display a rapid, frequent,
forced, audible inspiration, tense facial and body
musculature, frequent fist clenching and a propensity
for hastening the pace of conversation. Such individ-
uals have been found to have higher serum cholesterol
and increased urinary excretion of vanillyl mandelic
acid than those exhibiting the converse of this be-
havior pattern. Vanillyl mandelic acid (VMA) con-
stitutes about 75 per cent of the metabolic end
products of norepinephrine and epinephrine. It has
already been mentioned that these catecholamines
stimulate mobilization of free fatty acids from adipose
tissue. Chronic administration of epinephrine (187)
has been found capable of inducing a rise in serum
cholesterol and phospholipids. Presumably this effect
is secondary to fatty acid mobilization from adipose
tissue. It is too early to draw any conclusions from
attempts to relate coronary proneness to behavior
pattern and catecholamine excretion rate.
Sex
Surveys have shown that American females be-
tween 20 and 50 years of age have significantly lower
levels of serum total cholesterol and low-density lipo-
proteins than age-matched American males (i2g).
It is obvious that the mode of life of females usually
differs from that of males in a given culture, and the
effect of such a differing pattern of activity upon serum
lipids and susceptibility to coronary heart disease is
difficult to assess. Nevertheless, a number of studies
have suggested that the endocrine differences between
male and female can adequately account for the fact
that during their reproductive period, women have
lower levels of certain serum lipids. In general, ad-
ministration of androgenic hormones to patients is
associated with a rise in concentration of /3-lipopro-
teins, while estrogenic hormones induce a fall in this
same lipoprotein fraction (58, 59).
Whether such differences in lipid levels can account
for the established disparity in susceptibility to coro-
nary-artery disease between men and women remains
to be proved. However, supporting evidence is to be
found in the fact that women who have undergone
oophorectomy have "male" serum lipids and an
increased incidence of coronary-artery disease (212),
and that men treated with estrogen for prostatic
carcinoma have "female" lipids and less than the
expected degree of atherosclerosis (173). A recent
report from Edinburgh (151) has revealed that, over
a five-year period, one hundred men who had recov-
ered from a myocardial infarction, and who were
1 1 82
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
treated with estrogen daily, showed appreciable lower-
ing of their serum cholesterol levels without a sig-
nificant effect on their death rate, compared with a
comparable untreated control group. The apparent
lack of protection may have been related to the choice
of subjects who already had experienced myocardial
infarction.
Dietary Fatty Acids
It is now established that diet can have a profound
influence on serum lipids; indeed, the relation of diet
to serum lipids has been under intensive study in
recent years. More information is available concern-
ing the influence of dietary fat on serum lipids than
about the effect on serum lipids of other dietary con-
stituents. Not long ago, it was believed that the total
quantity of fat in the diet was the major factor affect-
ing serum lipids (117). In particular, the American
diet providing 40 to 45 per cent of its calories as fat
was implicated as being responsible for the relatively
high serum cholesterol values observed in adults in
the United States. It is now clear that the "quality"
of the fat in the diet is of primary importance in
determining the response of the serum cholesterol
fraction, although the relative proportions of carbo-
hydrate and fat in the diet appear to influence the
serum triglyceride concentration. It is of interest to
review briefly some of the events that have led to
these conclusions.
In 1933 Schoenheimer (183) reported that feeding
a wholly vegetarian diet to a patient with hyper-
cholesteremia resulted in marked lowering of serum
cholesterol. During the ensuing two decades, rela-
tively little further information of this kind was gath-
ered. Extensive studies regarding the metabolism of
cholesterol and other lipids were undertaken during
this time and significant discoveries were made. It was
found that cholesterol is readily synthesized in the
body from small carbon fragments (171), and that
the major catabolic pathway for cholesterol involves
conversion to bile acids (26).
In 1952, Groen and associates (87) demonstrated
by means of well-controlled and prolonged experi-
ments that substitution of vegetable for animal fats
in the diet can lower serum cholesterol in man, even
if the total fat content remains unchanged. During
the same year Kinsell and associates (122) reported
that ingestion of diets containing relatively large
amounts of vegetable fat consistently resulted in a
significant fall in the level of serum cholesterol and
phospholipids in human subjects. These observations
have been amply confirmed (2, 3, 22, 36, 123, 140).
It soon became apparent that the fatty acid com-
position of dietary fat was of primary importance in
determining serum cholesterol response. The experi-
ments in which vegetable oils were used stimulated
interest in the possible role of the essential fatty acids.
Subsequently, Kinsell and associates (124) concluded
that the major cholesterol-lowering ingredient in
various vegetable fats was, in fact, linoleic acid. Then,
in a series of well-controlled comparative experiments
in man, utilizing formula diets deriving 40 per cent of
their calories from fat, Ahrens et at. (4) observed that
the effects on serum cholesterol of various edible fats
could be related to their iodine value. Thus, fats with
high iodine values such as safflower, corn, and cotton-
seed oils proved to be relatively hypocholesteremic,
while fats with low iodine values such as palm oil,
beef, butter, cocoa butter, and coconut oil tended to
raise the serum cholesterol level. Intermediate or
neutral effects on serum cholesterol were obtained
with fats with intermediate iodine values such as
peanut and olive oils. In a later study (5) the Rocke-
feller group found that menhaden oil, a fat virtually
free of linoleic acid and yet with an extremely high
iodine value (I, number 180) was at least as effective
as corn oil in lowering serum cholesterol. Keys and
associates (120) have proposed a formula designed to
predict the effect of a given pattern of dietary fatty
acids on serum cholesterol in man. The formula at-
tempts to take into account the different roles of the
saturated, monounsaturated, and polyunsaturated
fatty acids in the diet; however, it remains to be
demonstrated that such an analysis can be applied
successfully in a variety of dietary situations (4).
In any event, it is clear that serum levels of choles-
terol and low-density lipoproteins can be changed
significantly when the pattern of fatty acids in the
diet is rearranged. When the glycerides of a dietary
fat contain predominantly saturated long-chain fatty
acids, concentrations of serum total cholesterol and
certain low-density lipoproteins tend to rise. When
such dietary glycerides contain an appreciable pro-
portion of polyunsaturated fatty acids (whether essen-
tial or not), serum cholesterol, and low-density lipo-
proteins tend to decrease. The degree of change in
serum lipids seems to depend upon the magnitude
of change in the pattern of the fatty acids in the diet.
Thus, it may be necessary to double or triple the
polyunsaturated fatty acid content of the diet (with-
out change in the total fat intake) to induce an appre-
ciable lowering of serum cholesterol. However, even
LIPID METABOLISM
I l83
when such drastic changes are made in the diet,
variations in individual responses are great (113).
Essential Fatty Acid (EFA) "Deficiency"
The fact that serum lipids can be lowered, when
dietary fats rich in polyunsaturated fatty acids are
fed, has stimulated considerable interest in the bio-
chemistry of the essential fatty acids (linoleate,
arachidonate, etc.), their role in nutrition, and, in
particular, their possible role in the metabolism of
cholesterol. Excellent reviews and discussions of these
subjects are available (1, 51, 106, 146, 189). Holman
(106) has suggested that the term essential fatty acid
(EFA) include "only those substances which are
active both for growth and for maintenance of dermal
integrity, limiting the term to linoleic and arachidonic
acids and to such other acids as may be derived
metabolically from them." As has been pointed out
by Aaes-Jorgensen (1), this definition leaves out lino-
lenic acid and C22 polyenoic acids from brain phos-
phatides which have been shown by Thomasson (50,
198) to be active only as growth factors.
Despite numerous studies since 1929, when Burr
& Burr (40) first recognized EFA deficiencv in
young rats, the EFA requirement for human adults
has not been determined. In fact, EFA deficiency in
adult man has not been demonstrated. In 1958, the
Food and Nutrition Board of the National Research
Council (65) suggested that one per cent of calories
should be the minimum daily EFA allowance for
humans. In any reasonable variation of the American
diet, this quantity is certainlv present. In the human
infant, however, Hansen and associates (90) have
shown that linoleic acid is definitely required in
amounts as little as 1.3 per cent of daily dietary
calories to prevent or cure certain dermatoses. In-
fants fed low fat diets (EFA-deficient) exhibited low
serum values for dienoic and tetraenoic fatty acids,
and high serum values for trienoic acids. The reverse
serum picture was obtained following addition of
linoleic acid to the diet.
On the basis of evidence now available it seems
unlikely that Sinclair's (190) hypothesis attributing
"nutritional" hypercholesteremia and atherosclerosis
to essential fatty acid deficiency is correct. Patients
with clinically manifest atherosclerosis and elevated
levels of serum total cholesterol do not necessarily
exhibit a lack of linoleic acid in their serum or depots
(103, 112). Moreover, Ahrens and associates (5) have
shown that formula diets, containing as their source
of fatty acids the nonessential polyethenoid fatty
acids predominating in certain fish oils, lower serum
cholesterol as effectively as formula diets containing
oils exceedinglv rich in linoleic acid.
Chain Length, Unsatnration, and Melting Point
Serum lipid responses to dietary fats have been
correlated with certain characteristics other than
essential fatty acid content or iodine value. Two
major variables affecting the physical and biochemical
properties of fatty acids are degree of unsaturation
and chain length (table 1). For example, linoleic acid
(2 double bonds) and stearic acid (no double bonds)
have the same chain length, and yet the melting
point of linoleic acid is — 1 1 C, whereas that of stearic
acid is 69.4 C. Stearic acid (Cw) and capric acid
(C10) are both fully saturated acids; however, the
melting point of the shorter chained capric acid
is 31.5 C.
Since the melting point of a fat (as well as other
characteristics) depends upon the component fatty-
acids, it is possible to lower the melting point of a tri-
glyceride either by increasing the unsaturation or by
reducing the chain length of its fatty acids. Accord-
ingly, many of the physical characteristics of a fully-
saturated fat containing shorter chain saturated fatty
acids may resemble those of a highly unsaturated fat
containing predominantly long-chain monounsatu-
rated or polyunsaturated fatty acids.
As previously mentioned, when subjects are fed
diets containing as their fat source solid fats such as
butter and mutton tallow, their cholesterol levels tend
table 1 . Classification of Fatty Acids According to Chain
Length and Degree of Unsaturation
Category &
Number of
Number of
Typical
Typical Acid
Carbon Atoms*
(Always even numbered)
Double Bonds
Food Sources
1. Medium Chain
6-12
None
Milk fat
Saturated
(12)
Coconut oil
Uauric)
2. Long Chain
14 - 24
None
Practically oil
Saturated
(16)
onimal &
(Palmitic)
vegetable fats
3. Long Chain
14 - 22
One
Most fats
Monoun saturated
(18)
ond oils
(Oleic)
4, Polyunsaturated,
18 - 20
Two- Four
Seed fats
Essential
(18)
(Two)
(Organ fats)
(Linoleic)
5. Polyunsaturated,
18- 26
Three- Six
Fish oils
Non-essential
(22)
(Five)
(Clupanodonic)
1 184
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
to remain elevated or to rise. When highly unsatu-
rated liquid fats such as corn oil and menhaden oil
are substituted isocalorically for the "hard" fats in
the diet, serum cholesterol levels usually fall appre-
ciably. A reasonable correlation can be made between
the physical state of a fat and its effect on serum
cholesterol. It has been suggested that the cholesterol-
lowering effect of the liquid oils may be a function of
the constituent polyunsaturated fatty acids and that
these fatty acids have an effect by virtue of their
polyethenoid configuration.
Recently, it was reported that a synthetic medium-
chain triglyceride, a liquid with a melting point below
o C, made up entirely of saturated fatty acids of chain
lengths ranging from C6 to C12, can lower serum cho-
lesterol significantly when substituted for butter in a
formula diet (96). Such results are of interest since
medium-chain triglyceride (MCT) is devoid of poly-
ethenoid fatty acids and has an iodine value of less
than 1.0 (that of butter is approximately 40). Al-
though the preparation is highly saturated, its shorter
chain fatty acids confer upon it many of the physical
characteristics of the natural vegetable oils rich in
long-chain polyunsaturated fatty acids.
Since medium-chain fatty acids may be metab-
olized differently from longer chain fatty acids, the
results with medium-chain triglyceride may not help
one interpret the cholesterol-lowering effects of the
highly unsaturated oils. On the other hand, such
studies have again called attention to the possible
importance of the physical properties of fatty acids
apart from the number and location of their double
bonds.
There remains a great need for further character-
ization of dietary glycerides in order that their physi-
cochemical characteristics may be better related to
their effects on serum lipids. The mechanisms whereby-
dietary fats influence cholesterol metabolism still are
not well understood. The picture is further compli-
cated by the complex nature of the dietary glycerides.
As pointed out earlier, 64 different fatty acids have
been identified in butter. For a time it was believed
by some investigators that the hypercholesteremic
effect of butter was due to its relatively high content
of shorter chain saturated fatty acids. Studies em-
ploying synthetic glycerides of simplified fatty acid
composition were helpful in clarifying this problem
inasmuch as it was possible to study the effect on
serum lipids (96) of a synthetic medium-chain tri-
glyceride (MCT) containing predominantly the very
fatty acids thought to be hypercholesteremic. As a
result of such experiments it was shown that in rela-
tion to butter the MCT preparation was actually
hypocholesteremic.
Dietary Cholesterol
Since the early work of Ignatowski (110) and
Anitschkow (10) on experimental atherosclerosis in
rabbits, dietary cholesterol has been an essential
ingredient of diets used to induce hypercholesteremia
and atherosclerosis in a variety of animal species. Not
long ago, it was fashionable to prescribe diets low in
cholesterol for patients with an elevated serum choles-
terol. As cholesterol metabolism in man was studied
more intensively, it became evident that the liver
normally synthesizes about three times as much choles-
terol per day as is consumed in the average diet. It
was further learned that an increased intake of
cholesterol is likely to result in a proportionate inhibi-
tion of cholesterol manufacture. Subsequently, care-
fully controlled studies by Keys et al. (119) and others
have suggested that, within wide limits, variations in
the cholesterol content of an ordinary diet do not
affect serum cholesterol levels to any significant de-
gree, provided other elements in the diet are constant.
Such results have encouraged many physicians to
abandon use of diets specifically low in cholesterol
in treating patients with hypercholesteremia.
Despite such negative reports, the influence of
dietary cholesterol on serum cholesterol in man con-
tinues to be a subject of investigation and contro-
versy. Beveridge et al. (23) have recently reported
that the addition of relatively small amounts of choles-
terol to formula diets can raise serum cholesterol
levels, depending on the nature of the accompanying
fat. Indeed, these investigators attribute the hyper-
cholesteremic effect of butter in part to its content of
cholesterol. Keys (121) has reviewed the results of
Beveridge and associates and has questioned their
significance. Connor et al. (45) have studied the effect
of adding or subtracting moderate amounts of choles-
terol as egg yolk in diets equivalent in amounts and
composition of fat. In their studies, the addition of
475 to 1425 mg cholesterol (the amount present in
one to three large eggs) raised serum total cholesterol
by an average of 68 mg per 100 ml. On the other
hand, crystalline cholesterol added to the diet in
amounts ranging from 1 200 to 3600 mg per day
increased the mean cholesterol level by only 18 mg
per 100 ml.
It would seem that the effect of crystalline choles-
terol added to the diet cannot be equated with the
effect of the cholesterol that occurs naturally in food.
LIPID METABOLISM
ii 8=
It is possible that the differences in response to these
two forms of cholesterol relate to considerations of
solubility of this sterol in dietary fat. Recently, it has
been pointed out (37) that cholesterol is more soluble
in the saturated than in the polyunsaturated fats.
Practicable Diets
It is of considerable practical interest that palatable
diets can now be devised that are rich in polyunsatu-
rated fatty acid content and provide the same pro-
portion of fat to which Americans are accustomed. At
present, need exists for controlled studies in man to
determine the effects on serum lipids of "normal"
diets exhibiting a variety of fatty acid patterns. At-
tempts in this direction have begun (46, 93, 111).
Earlier experiments with semipurified formula-type
regimens suggested that when the ratio of poly-
unsaturated to saturated fatty acids (P:S ratio) in
the diet was increased, serum cholesterol usually
could be lowered.
Approximately 5 years ago, experiments were
begun to determine whether everyday diets could be
altered so as to reduce serum cholesterol Levels and
yet remain palatable and acceptable to most indi-
viduals. From progress reports of these studies, it is
now clear that manipulation of the fatty acid pattern
of the diet is effective in lowering serum cholesterol
in most subjects with cholesterol levels higher than
230 mg per 100 ml. The change in pattern is effected
principally by substituting one form of dietary fat for
another in order to increase the P:S ratio. In prac-
tice, this change involves a drastically decreased
consumption of butter fat and of certain margarines,
and a reduced intake of meats from ruminants, such
as bovine animals and sheep. At the same time, con-
sumption of poultry, fish, nuts, and plant seed oils
is materially increased. A typical diet designed to
lower cholesterol prescribes an increase in intake of
polyunsaturated fatty acids from approximately 1 5
to 42 per cent of total fat and a decrease in intake of
saturated fatty acids from approximately 42 to 1 5 per
cent. The intake of the monounsaturated fatty acid,
oleic acid, remains unchanged. Total dietary fat is
reduced from 44 to 36 per cent of calories, although
this is not an essential feature of the diet.
Such diets are acceptable and palatable. The effect
of a diet of similar fatty acid composition on serum
cholesterol in 97 men of normal weight, 50 to 59
years old, was determined by Jolliffe et a/. (113). This
study demonstrated the fall in cholesterol by tertiles
over a period of 6 months. The upper third, with
cholesterol levels of 270 mg per 100 ml and over,
dropped an average of 45 mg per 1 00 ml. The lower
third, with cholesterol values under 230 mg displayed
a decrease averaging 16 mg per 100 ml. Similar
studies performed on smaller groups of subjects have
yielded generally similar results (200).
The fact that it is indeed feasible to lower serum
cholesterol levels by dietary means has had and is con-
tinuing to have a tremendous impact on the public,
the medical and dietetic professions, and the food
industry. The public is being made increasingly aware
of the possibly ominous significance of an elevated
serum cholesterol level in terms of danger from
obstructive coronary artery disease. At the same time,
there appears to be decreasing use, per capita, of
butter and hydrogenated products, and increasing
use of liquid vegetable oils, such as corn, safflower,
and cottonseed oils.
A number of cookbooks on the subject of fatty acid
"control" are now appearing, and the demand for
them is great. Recently, several food companies have
come out with new margarines with an increased
content of cis-cis linoleic acid. There is increasing
interest in the development of cheeses and spreads
and commercial products resembling ice cream, all
containing appreciable quantities of linoleate. "Rea-
sonable substitution of polvunsaturated for saturated
fats, under medical supervision" has been recom-
mended by an ad hoc committee of the American
Heart Association (8). Whether or not diets of this
kind will have a clinically useful effect is one of the
most urgent questions facing medicine today.
Mechanism of Cholesterol Lowering
The exact mechanism whereby a diet rich in poly-
unsaturated fatty acids lowers serum total cholesterol
(and low-density lipoproteins) remains unknown. A
few studies have indicated that when such a diet is
fed more cholesterol and its end products (including
bile acids) are excreted in the feces. When a diet rich
in saturated fatty acids is fed, less sterols and bile
acids are excreted in the feces and the serum choles-
terol level rises (85, 99, 1 1 1 ).
There appears to be no evidence for direct inter-
ference by the polyunsaturated fatty acids with
cholesterol synthesis in the liver. However, the ability
of the liver to excrete cholesterol and to convert
cholesterol to bile acids may depend in part on certain
physicochemical characteristics of cholesterol esters
or of the lipoprotein molecules of which cholesterol
and its esters can constitute a significant portion.
1 1 86
HANDBOOK OF PHYSIOLOGY
CIRCULATION' II
Cholesterol esterifies readily with polyunsaturated
fatty acids such as linoleic acid (115). When the die-
tary fatty acids are predominantly saturated, esters
such as cholesteryl oleate and palmitate are likely to
occur in increasing amounts and, conceivably, may
be less available for excretion and conversion to bile
acids bv the liver.
Additional Influences on Serum Lipids
Other dietary manipulations also can influence
serum lipids (63, 153, 166). A drastic decrease in
dietary intake of protein is associated with lowering
of serum total cholesterol (and fi-lipoproteins) in
man (76, 152). High protein intakes above 10 per
cent of total daily calories have been found to be
effective in lowering serum cholesterol and /3-lipo-
proteins in animals but not in man (153). A sub-
stantial decrease in the proportion of fat in the diet
may be associated with a lowering of serum choles-
terol, but in certain individuals a considerable rise in
serum triglycerides may occur (4, 6). Such diets
usually contain a large quantity of carbohydrate,
much of which gets converted by the body into fat.
There is evidence from studies in laboratory animals
and human subjects that the type of carbohydrate in
the diet can affect serum lipids. Compared to sucrose,
starch promotes bile acid excretion and tends to
lower serum cholesterol in the rat (165). It has also
been reported that when the carbohydrates of
legumes are substituted isocalorically for sucrose in
the diets of human subjects, cholesterol levels are
reduced to a slight degree (9).
The effect of a given diet on caloric balance must
also be taken into account. Patients in negative
caloric balance can often have a transient decrease in
serum lipids on this basis; on the other hand, during
active weight gain, serum lipids tend to rise (142,
204). Weight reduction may induce a decline in
serum lipids in persons with hyperlipidemia. It is
not certain whether such a decrease occurs only while
weight actuallv is being lost or whether, in some in-
stances, the improvement in serum lipids will be sus-
tained for as long as weight is maintained at a reduced
level. However, as was mentioned earlier, during a
sustained fast the serum levels of the low-density
lipoproteins tend to increase; this contrasts with the
decline in serum lipids shown bv nonfasting patients
in negative caloric balance.
Pharmacologic approaches to lowering serum
cholesterol have included use of agents with the fol-
lowing mechanisms of action :
/) Inhibition of cholesterol biosynthesis. Under
this category are included triparanol (138, 193),
benzmalecene (15), and possibly nicotinic acid (7, 82,
158). Agents which inhibit cholesterol biosynthesis
also may interfere with other important synthetic
processes such as steroid biogenesis; hence, they are
potentially toxic for man.
_') Inhibition of cholesterol absorption from the
gastrointestinal tract. Plant sterols such as sitosterol
have been used for this purpose in man (20, 188). The
mechanism for inhibition of cholesterol absorption
remains to be demonstrated. Moreover, the effective-
ness of sitosterol in lowering serum cholesterol in man
has been questioned. Studies by Levere and his asso-
ciates (131) suggested that no decrease in serum
cholesterol could be attributed confidently to sitos-
terol administration, and that any apparent decrease
might be caused by great fluctuations in serum choles-
terol observed in such studies in man.
3) Promotion of cholesterol degradation. Reference
already has been made to the polyunsaturated fatty
acids and the possibility that they might act by pro-
moting the rate of cholesterol breakdown to bile
acids. Simple addition to the diet of relatively small
quantities of polyunsaturated fatty acids cannot be
relied upon to induce significant lowering of the
serum cholesterol level (159). Pharmaceutical prepa-
rations containing linoleic acid and sometimes sup-
plemented with small quantities of pyridoxine and
tocopherol offer no advantage over the various
linoleate-rich oils such as corn and cottonseed that
can be purchased inexpensively at the grocery.
Moreover, as mentioned earlier, the polyunsaturated
fatty acids, in order to be effective in lowering serum
cholesterol, must be consumed in relatively large
amounts and their intake integrated with an over-all
reduction in the consumption of saturated fatty acids
as part of a carefully adhered-to regimen.
An interesting group of substances has been found
capable of lowering serum cholesterol by binding bile
acids in the gut and thereby promoting their fecal
excretion. These are the anion exchange resins (14,
196, 202) that form nonabsorbable complexes with
bile acids. Appropriately, they have been termed
"bile acid sequestrants" and of these cholestyramine
has been studied extensively in man. Cholestyramine
is apparently innocuous since it does not seem to enter
the body. Long-term effects of bile acid sequestration
in man are at present unknown. The net effect of the
bile acid sequestrants appears to be similar to that of
the polyunsaturated fatty acids; namely, promotion
of cholesterol degradation.
Thyroid hormones and their analogues ([]^, 77, 151)
may lower cholesterol bv virtue of an effect on bile
LIPID METABOLISM
U87
acid metabolism (100); however, the exact mecha-
nism remains to be clarified. Studies of (/-thyroxine
in man revealed no reliable dose-response of serum
cholesterol over a 6-month period, nor was there a
dose which would be effective in lowering serum
cholesterol without provoking angina (151 ).
Neomycin has been reported to lower serum
cholesterol and simultaneously to increase fecal ex-
cretion of bile acids (81, 82, 181). The influence of
neomycin on bile acid metabolism may be related
to its profound effect on intestinal flora and the
possible damage it inflicts on the villi of the intestinal
mucosa.
4) Increased tissue removal of cholesterol. It is
suspected that the estrogenic hormones and their
congeners (12, 34) may lower serum cholesterol
by increasing the activity of the reticuloendothe-
lial system and thereby accelerating the removal
of cholesterol-rich lipoproteins from the plasma.
Whether cholesterol catabolism is further enhanced
within the reticuloendothelial system remains un-
known. The effect of estrogens on cholesterol bio-
synthesis also is not clear; however, there is evidence
that estrogen administration is not associated with
an increased rate of bile acid excretion. A variety of
teals
fig. 8. Mechanisms for lowering plasma cholesterol.
/ : Inhibition of cholesterol biosynthesis (triparanol). 2:
Inhibition of cholesterol absorption (sitosterols). 3: Promotion
of cholesterol degradation — a, polyunsaturated fatty acids;
b, bile acid sequestrants. 4: Increased tissue removal of
cholesterol (estrogens). (From Bergen & Van Itallie, Ann.
Internal Med. In press.)
other substances (63, 166) too numerous to discuss
in detail also are capable of inducing a reduction in
serum cholesterol. The four mechanisms for lowering
cholesterol that have been described are summarized
in figure 8.
BLOOD LIPIDS AND ATHEROSCLEROSIS
Virtually no information based on direct observa-
tion is available concerning the relationship in man
between circulating lipids and atherosclerosis. The
difficulty has been that the tools for diagnosing occult
atherosclerosis are inadequate. There is indirect evi-
dence— epidemiologic, experimental, clinical, and
pathologic — that sustained elevation of the plasma
low-density lipoproteins is associated with an in-
creased rate of atheroma formation. It is suspected
that plasma lipoprotein may be '"filtered" through
intimal cells under arterial pressure and that accumu-
lation of lipid in the subintimal area is accelerated by
continued traffic of plasma rich in "unstable" lipo-
protein through the arterial wall (153, 201).
Animal experiments have shown that when an
increased plasma concentration of ^-lipoproteins is
achieved by dietary or other means, atherosclerosis
usually results (64, 143, 211). The early vascular
changes observed in experimental atherosclerosis are
believed by many observers to have a close resem-
blance to early human lesions.
Evidence is also available from studies in man that
prolonged elevation of the serum cholesterol is asso-
ciated with an increased susceptibility to athero-
sclerosis and its clinical manifestations. In diseases
such as diabetes and hypothyroidism in which serum
lipids tend to be elevated, the incidence of athero-
sclerosis also is increased. Moreover, patients with
angina pectoris or a history of myocardial infarction
tend to have serum lipid levels higher than those of
apparently healthy control subjects. This difference
in lipid levels is most striking when groups below the
age of 45 are compared (201).
Epidemiologic studies of populations in various
parts of the world (18) involving correlation of serum
cholesterol levels with prevalence of clinically mani-
fest atherosclerosis and presence of the disease at
autopsy have given additional support to the pro-
posed relationship between hypercholesteremia and
atherosclerosis. Most epidemiologic studies attempt-
ing to relate coronary heart disease to levels of serum
cholesterol are retrospective in that the subjects
studied have already exhibited clinical manifesta-
tions of atherosclerotic heart disease. The Framing-
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
ham Heart Program (47, 114) is of interest because it
has used the prospective approach. In this study,
approximately 5,000 individuals originally free of
manifestations of overt coronary heart disease in the
town of Framingham, Massachusetts, have been fol-
lowed for 8 years, and the study continues. From the
Framingham data, it has been shown that an associa-
tion exists between a number of factors, other than
age and sex, and an increased risk of developing
coronary heart disease: these are obesity, hyperten-
sion, electrocardiographic evidence of left ventricular
hypertrophy, heavy smoking, and hypercholesteremia.
Increasing levels of serum cholesterol were found to
be associated with increasing risk of developing cor-
onary heart disease. Among the various lipid meas-
urements (excluding cholesterol esters and trigly-
cerides about which data were lacking) serum total
cholesterol was found to be the best measurement for
predicting the occurrence of overt coronary heart
disease. While the results of the Framingham study
have been interesting and provocative, the relatively
small number of subjects studied make it difficult to
arrive at firm conclusions about the relationship
between such variables as the serum cholesterol and
incidence of obstructive coronary disease. Until
larger samples can be obtained there is always a risk
of drawing too many unqualified conclusions from
insufficient data (132).
In addition to the increased risk of heart disease
associated with hypercholesteremia, the increased
incidence of coronary atherosclerosis in patients with
essential hyperlipemia (hypertriglyceridemia) (139)
should be mentioned, as well as the reports that
patients with coronary heart disease tend to display
elevated levels of serum triglycerides (6) and impair-
ment of rate of clearing of alimentary lipemia (186).
Such preliminary observations suggest that the lower-
density lipoproteins carrying an increased burden of
triglyceride may play a more important etiologic
role in coronary atherosclerosis than has been hitherto
appreciated.
Although it is common practice to use the term
"atherosclerosis" as though it described one disease,
evidence has been accumulating that in certain
countries where clinically manifest coronary-artery
disease is rare, atherosclerosis of the aorta may be
quite common (86). Similarly, American women
possess a relative immunity to coronary heart disease
during their reproductive years; yet they are not
equally immune from atherosclerosis at other ana-
tomic sites (174). It is worthy of comment that al-
though coronary heart disease is very common in
patients with familial hypercholesteremic xantho-
matosis (primary hypercholesteremia), this disease
does not seem to predispose to the development of
peripheral or cerebrovascular disease (89). Con-
versely, coronary heart disease is rare among the
South African Bantu, a group in whom serum choles-
terol levels tend to be very low; yet, cerebral catas-
trophes occur as frequently among Bantu as among
populations with substantially higher levels of choles-
terol (128, 205). Thus, coronary heart disease may
prove to be a special manifestation of atherosclerosis,
with its own epidemiology and, perhaps, its own bio-
chemical pathology.
Chemical studies of the atheroma have supported
the belief that a number of the fatty constituents of
the atheroma are derived from the plasma. Studies
have demonstrated that the distribution of lipids in
early atheroma is roughly similar to that in plasma
(104). As the atheromas of human aortas progress in
severity, they exhibit a parallel increase in content
of carotenoid and cholesterol (25). Since carotenoid
is derived entirely from the diet, such observations
also suggest that atheromatous lipid derives from the
circulation and does not originate de novo within
the arterial wall. A similar interpretation can be
made of the demonstration that linoleic acid, a sub-
stance which the body apparently cannot synthesize,
is a significant constituent of atheromas (187).
With the advent of gas-liquid chromatography and
other improved techniques for lipid separation and
identification, it has become possible to obtain more
precise information about the lipid constituents of
atheromas at various sites and at various stages of
evolution (30, 31, 54, 133, 136, 199, 208). It has been
reported that the saturated and monounsaturated
fatty acid moieties of cholesterol esters tend to accumu-
late preferentially in early atheromatous lesions.
However, linoleic acid also can be found in atheromas
in appreciable quantities. A report from Leiden (32)
has described the results of detailed analyses of the
lipids in aortas and coronary and cerebral arteries in
various stages of atherosclerosis. It was found that
as the aortic lesions became more advanced their
relative content of cholesterol and cholesterol esters
increased strikingly. A comparison of the fatty acid
composition of "early" and "late" atheromas with
uninvolved aorta showed an increase in the propor-
tion of the polyunsaturated fatty acids of cholesterol
esters in the older lesions. The phospholipids exhibited
a slight increase in their proportion of long-chain
saturated fatty acids. Generally similar results were
obtained for the circle of Willis, in which the content
LIPID METABOLISM
l89
•of cholesterol esters increased in more advanced
lesions while phospholipid content decreased. In the
coronary vessels, the content of triglycerides in rela-
tively healthy specimens was quite high; cholesterol
and its esters increased with advancing atherosclerosis
while glyceride content fell. With respect to fatty acid
patterns, the trends with increasing atherosclerosis
in coronary and cerebral samples were similar to
that shown by the aorta.
The studies reported to date on the lipid composi-
tion of atheromas appear to support the impression
that the cholesterol esters in the intima and media
start by being extremely saturated in comparison with
those circulating in the plasma; with increasing ather-
osclerosis of the wall the apparently healthy parts
contain more polyunsaturated cholesterol esters than
the adjacent lesions. The mechanism whereby even
apparently normal vascular tissue becomes infil-
trated by plasma lipids remains unknown. In any
event, it seems clear that the fatty acids of the arterial
wall, like those of the plasma, are responsive to dietary
influences. Hence, this variable (among others) must
be taken into account when data reported by various
investigators are being compared and evaluated.
It should not be forgotten that a number of en-
vironmental and host factors may play a role in the
pathogenesis of atherosclerosis. Groen (88) has listed
eight exogenous and six endogenous elements that
have been given major attention by various investi-
gators. Included in this list are such items as constitu-
tion, age, obesity, physical exercise, social class, emo-
tional influence and such concomitant diseases as
hypertension and diabetes. This very multiplicitv of
considerations makes human atherosclerosis an
exceedingly complex problem. However, a common
pathway still must be defined through which a given
influence can operate. In the present discussion the
early atheroma has been viewed as a phenomenon
secondary to an abnormality of the serum lipids;
thus, the lesion can be said to be biochemical first
and histologic second.
ROLE OF BLOOD CLOTTING AND THROMBOSIS
Commonly, the dramatic clinical manifestations of
atherosclerosis such as myocardial infarction and
"stroke" result from acute thrombotic obstruction of
an artery. It is generallv believed that the athero-
sclerotic lesion acts as a nidus for thrombus formation
which, in turn, occludes the vessel. This assumption
is not invariably correct. Arterial thrombosis can
occur at sites where atherosclerosis is minimal or
absent, and arterial occlusion by an atherosclerotic
lesion can occur without thrombosis. The events that
lead to thrombosis in vivo still are not clearly under-
stood.
Certain inconsistencies appear if arterial thrombosis
is regarded as a simple epiphenomenon of athero-
sclerosis. Epidemiologic studies have suggested an
increase in mortality rates from coronary artery dis-
ease in the Western World during the past 25 years
without a corresponding increase in atherosclerosis
(149). The discrepancy in certain countries between
clinical coronary artery disease (rare) and athero-
sclerosis of the aorta (common) already has been
mentioned. In certain experimental animals, in which
atherosclerosis has been produced readily by dietary
means, production of myocardial infarction has been
difficult [although achieved in rats fed high fat diets
supplemented with cholesterol, cholic acid, and thio-
uracil (91)]. Thus, any attempt to relate development
of occlusive arterial disease to dietary fat consumption
must take into account the role of thrombosis (150).
In this regard, Duguid (56) has reviewed and modified
a concept introduced by Rokitansky a century ago
relating atherogenesis to fibrin deposition on arterial
intimal surfaces. Duguid has shown that arterial
narrowing can be produced by organization of mural
thrombi with subsequent endothelialization and lipid
deposition. The end result is difficult if not impossible
to distinguish from "atherosclerosis.'' Subintimal
hemorrhage and other "mechanical" factors have
been considered as initiating the process (57). From
such a standpoint, lipid deposition in atherosclerosis
is thought to be secondary to thrombus formation.
However, most of the available evidence still favors
the view that lipid deposition in the arterial wall is
the primary event in atherogenesis.
The possible influence of dietary fat on blood coagu-
lation has been the subject of several reviews (92,
163). It is clear that the factors involved in main-
tenance of blood fluidity are complex and deserving
of further investigation. Methods for detecting in-
cipient thrombosis in the intact organism are inade-
quate. A distinction must be made between results
obtained by feeding fat on the various coagulation
tests in vitro and the influence of lipemia on coagula-
tion in vivo. There appears to be agreement concern-
ing the accelerating effect of certain dietary and
synthetic phospholipids, platelet lipid extracts, and
postprandial plasma on the Stypven time, a short-
ening of the clotting time of plasma in the presence
of Russell's viper venom. The relevance of these
ngo
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
findings to in vivo coagulation remains unknown.
Recently Poole (164) has emphasized the difference
between clotting and thrombosis and has pointed
out that factors important in clotting may be un-
important in thrombosis. Under the electron micro-
scope, clots and thrombi are different in structure.
The thrombus contains areas of closely packed plate-
lets while the clot contains predominantly red cells
and a few platelets distributed at random in a fibrin
network. It has been shown (44) that the coagulum
formed when blood is made to flow through a closed
circular loop of plastic tubing mounted on a rotator
resembles a natural thrombus. In this system the
unesterified long-chain saturated fatty acids accel-
erate thrombus formation, while the polyunsaturated
and shorter chain fatty acids are inactive. More
information is needed about the relationship between
clotting and thrombosis before a decision can be
made concerning the part played by dietary fat and
lipemia in the mechanism of thrombosis.
REFERENCES
1. Aaes-Joroensen, E. Essential fatty acids. Physiol. Revs.
41 : 1-51, 1961.
2. Ahrens, E. H., Jr., T. T. Tsaltas, J. Hirsch, and W. 15.
Insull, Jr. Effects of dietary fats on the serum lipides of
human subjects. J. Clin. Invest. 34: 918, 1955.
3. Ahrens, E. H., Jr., W. Insull, Jr., R. Blomstrand, J.
Hirsch, T. T. Tsaltas, and M. L. Peterson. The in- 16.
fluence of dietary fats on serum-lipid levels in man. Lancet
1 : 943-953. '957-
4. Ahrens, E. H, Jr., J. Hirsch, \V. Insull, Jr., and M. L.
Peterson. Dietary fats and human serum lipide levels. In:
Chemistry of Lipides as Related to Atherosclerosis, edited by I. 17.
H. Page. Springfield, 111.: Thomas, 1958, pp. 222-261.
5. Ahrens, E. H., Jr., VV. Insull, Jr., J. Hirsch, W. Stoffel, 18.
M. L. Peterson, J. W. F. Farquhar, T. Miller, and H.
J. Thomasson. Effect on human serum-lipids of a dietary 19.
fat, highly unsaturated, but poor in essential fatty acids.
Lancet 1: 115-ug, 1959. 20.
6. Albrink, M. J., and E. B. Man. Serum triglycerides in
coronary artery disease. Arc h. Internal Med. 103: 4-8, 1959.
7. Altschul, R., A. Hoffer, and J. D. Stephen. Influence 21.
of nicotinic acid on serum cholesterol in man. Arch. Bio-
chern. Biophys. 54: 588-559, 1955.
8. American Heart Association, Central Committee for 22.
Medical and Community Program Report. Dietary fat
and its relation to heart attacks and strokes. Circulation 23:
'33- '35. 1961- 23.
9. Anderson, J. T., F. Grande, and A. Keys. Effect of
carbohydrates of leguminous seeds, wheat and potatoes on
serum cholesterol in man. Federation Proc. 19: 18, i960.
10. Anitschkow, N. Uber Organvcrandcrungen bei Abla- 24.
geiiing von anisotropin Lipoiden. Ber. Ges. Russ.,Arzte
St. Petersburgh. Cited in: Experimental atherosclerosis in
animals. In : Atherosclerosis, edited by E. V. Cowdry. New 25.
York: Macmillan, 1933, p. 283.
1 1 . Arnott, E. M. Changing aetiology of heart disease. Brit.
Med. J. 2: 887-891, 1954.
12. Barr, D. P. Influence of sex and sex hormones upon de- 26.
velopment of atherosclerosis and upon lipoproteins of
plasma. J. Chronic Diseases 1 : 63-85, 1955.
13. Bates, M. W. Turnover rates of fatty acids of plasma 27.
triglyceride, cholesterol ester and phospholipid in post-
absorptive dog. Am. J. Physiol. 194: 446-452, 1958.
14. Bergen, S. S., Jr., T. B. Van Itallie, D. M. Tennent,
and W. H. Sebrell. Effect of an anion exchange resin on 28.
serum cholesterol in man. Proc. Soc. Exptl. Biol. Med. 102:
676-679, 1959.
Bergen, S. S., Jr., T. B. Van Itallie, and W. H.
Sebrell. Hypocholesteremic effect in man of benz-
malecene: inhibitor of cholesterol synthesis. Proc. Soc. Exptl.
Biol. Med. 103: 39-40, i960.
Bergstrom., S., and B. Borgstrom. Intestinal absorption
of fats. In : Progress in the Chemistry of Fats and Other
Lipides, edited by R. T. Holman, VV. O. Lundberg
and T. Malkin. London: Pergamon, 1955, vol. 3, pp.
352-393-
Bergstrom, S., and B. Borgstrom. Metabolism of lipides.
Ann. Rev. Bwchem. 25: 177-200, 1956.
Bergstrom, S. Metabolism of bile acids. Federation Proc.
20: Suppl. 7, 121-126, 1961.
Best, C. H., and J. Campbell. Anterior pituitary extracts
and liver fat. J. Physiol., London 86: 190-203, 1936.
Best, M. M., C. H. Duncan, E. J. Van Loon, and J. D.
Wathen. The effects of sitosterol on serum lipids. Am. J.
Med. 19:61-70, 1955.
Bevans, M., J D. Davidson, and L. L. Abell. Early
lesions of canine atherosclerosis. A.M. A. Arch. Pathol. 51 :
278-287, 1 95 1.
Beveridge, J. M. R., W. F. Connell, and G. A. Mayer.
Dietary factors affecting levels of plasma cholesterol in
humans. Can. J. Biochern. and Physiol. 34: 441-455, 1956.
Beveridge, J. M. R., W. F. Connell, H. L. Haust, and
G. A. Mayer. Dietary cholesterol and plasma cholesterol
levels in man. Can. J. Biochern. and Physiol. 37: 575-582,
■959-
Beznak, A., and Z. Hasch. Effect of sympathectomy on
fatty deposit in connective tissue. Quart. J. Exptl. Physiol.
27: '-'5. '937-
Blankenhorn, D. H , D. G. Freiman, and H. C
Knowles, Jr. Carotenoids in man. The distribution of
epiphasic carotenoids in atherosclerotic lesions. J. Clin.
Invest. 35: 1243-1247, 1956.
Bloch, K., B. N. Berg, and D. Rittenberg. Biological
conversion of cholesterol to cholic acid. J. Biol. Chem.
■49: 5i'-5'7. 1943-
Bodgonoff, M. D., A. M. Weissler, F. L. Merritt,
Jr., \V. R. Harlan, and E. H. Estes, Jr. The role of the
autonomic nervous system in human lipid metabolism. J.
Clin. Invest. 38: 989, 1959.
Borgstrom, B. Transport form of C14 decanoic acid in
LIPID METABOLISM
Iigi
porta and inferior vena cava blood during absorption in
the rat. Acta Physiol. Stand. 34: 71-74, 1955
29. Borgstrom, B., N. Tryding, and G. Westoo. On extent
of hydrolysis of triglyceride ester bonds in lumen of human
small intestine during digestion. Acta Physiol. Scand. 40:
241-247, 1957.
30. Bottcher, C. J. F., F. P. Woodford, C. Ch. Ter Haar
Romnev-Wachter, H. E. Boelsma-Van Houte, and C.
M. Van Gent. Composition of lipids isolated from aorta,
coronary arteries and circulus willisii of atherosclerotic
individuals. Nature 183:47-48, 1959.
31. Bottcher, C. J. F., F. P. Woodford, C. Ch. Ter Haar
Romnev-Wachter, H. E. Boelsma-Van Houte, and C.
M. Van Gent. Fatty-acid distribution in lipids of the
aortic wall. Lancet 1 : 1378-1383, i960.
32. Bottcher, C. J. F., and F. P. Woodford. Chemical
changes in the arterial wall associated with atherosclerosis.
Federation Proc. 21 : Suppl. 1 1, 15-19, 1962.
33. Boyd, G. S. Thyroid function, thyroxine analog, and cho-
lesterol metabolism in rats and rabbits. In : Hormones and
Atherosclerosis, edited by G. Pincus. New York : Acad. Press,
■959. PP- 49-°2-
34. Boyd, G. S. Effect of linoleate and estrogen on cholesterol
metabolism. Federation Proc. 21 : Suppl. 11, 86-92, 1962.
35. Bragdon, J. H. On the composition of chyle chylo-
microns. J. Lab. Clin. Med. 52: 565-570, 1958.
36. Bronte-Stewart, B., A. Antonis, L. Eales, and J. F.
Brock. Effects of feeding different fats on serum cholesterol
level. Lancet 1 : 521-527, 1956.
37. Bronte-Stewart, B. Lipids and atherosclerosis. Federa-
tion Pwc. 20: Suppl. 7, 127-134, 1961.
38. Brozek, J. Changes of body composition in man during
maturity and their nutritional implications. Federation Proc.
11 : 784-793. '952-
39. Bunting, C. H., and H. Bunting. Acid mucopolysac-
charides of aorta. A.M. A. Arch. Pathol. 55: 257-264, 1 953.
40. Burr, G. O, and M. M. Burr. New deficiency disease
produced by rigid exclusion of fat from the diet. J. Biol.
Chem. 82: 345-367, 1929.
41. Byers, S. O., M. Friedman, and F. Michaelis. Observa-
tions concerning production and excretion of cholesterol
in mammals; source of excess plasma cholesterol after liga-
tion of bile duct. J. Biol. Chem. 188: 637-641, 1951.
42. Byers, S. O., and M. Friedman. Excretion of 3-mcthoxy-
4-hydroxymandelic acid in men with behavior pattern as-
sociated with high incidence of coronary artery disease.
Federation Proc. 21 : Suppl. 1 1, 99-101, 1962.
43. Cahill, G. F., Jr., B. LeBoeuf, and A. E. Renold.
Factors concerned with the regulation of fatty acid metab-
olism by adipose tissue. Am. J. Nutrition 8: 733-739, i960.
44. Chandler, A. B. In vitro thrombotic coagulation of the
blood; a method for producing a thrombus. Lab. Invest. 7:
110-114, 1958.
45. Connor, W. E., R. E. Hodges, and R. Bleiler. Serum
lipids in men receiving high cholesterol and cholesterol-
free diets. Circulation 22: 735, 1960.
46. Davis, C. B., R. E. Clancy, B. E. Cooney, D. M. Heg-
sted, and J. Huett. Effect of mixed fat formula feeding
on serum cholesterol in man. II. Further study utilizing
a twenty per cent fat formula. Am. J. Clin. Nutrition 8:
808-811, i960.
47. Dawber, T. R., F. E. Moore, and G. V. Mann. Coro-
nary heart disease in the Framingham study. Am. J. Public
Health 47: 4-28, 1957 (Symposium).
48. Dawson, A. M., and K.J. Isselbacher. Esterification of
palmitate-i-G" by homogenates of intestinal mucosa. J.
Clin. Invest. 39: 150-160, ig6o.
49. Dawson, A. M., and K. J. Isselbacher. Fat absorption.
Arch. Internal Med. 107: 305-308, 1961.
50. De Iongh, H., and H. J. Thomasson. Essential fatty acid
activity of docosapolyenoic acids from brain glycerophos-
phatides. Nature 178: 1 051 -1052, 1956.
51. Deuel, H. J., Jr., and R. Reiser. The physiology and
biochemistry of the essential fatty acids. Vitamins and Hor-
mones 13: 29-70, 1955.
52. Dole, V. P. A relation between non-esterified fatty acids
in plasma and the metabolism of glucose. ./. Clin. Invest.
35- l5°-'54. '956-
53. Dole, V. P. Transport of non-esterified fatty acids in
plasma. In : Chemistry of Lipides As Related to Atherosclerosis,
edited by I. 11. Page. Springfield, 111.: Thomas, 1958, pp.
189-204.
54. Dole, V. P., A. T. James, J. P. W. Webb, M. A. Rizack,
and M. F. Sturman. The fatty acid patterns of plasma
lipids during alimentary lipemia. J. Clin. Invest. 38:
■544" 1 554. '959-
55. Duff, G. L., G. C. McMillan, and A. C. Ritchie.
The morphology of early atherosclerotic lesions of the
aorta demonstrated by the surface technique in rabbits
fed cholesterol; together with a description of the anatomy
of the intima of the rabbit's aorta and the spontaneous
lesions which occur in it. Am. J. Pathol. 33: 845-873, 1957.
56. Duguid, J. B. Thrombosis as factor in pathogenesis of
coronary atherosclerosis. J. Pathol. Bacteriol. 58: 207-212,
1946.
57. Duguid, J. B., and W. B. Robertson. Mechanical factors
in atherosclerosis. Lancet 1: 1 205-1 209, 1957.
58. Eder, H. A. The effects of hormones on human serum
lipoproteins. Recent Progr. Hoimone Research 14: 405-425,
1958.
59. Eilert, M. L. The effect of estrogens upon the partition
of the serum lipids in female patients. Am. Heart J. 38:
472-473. '949-
60. Ellis, N. R., and H. S. Isbell. Soft pork studies; effect of
food fat upon body fat, as shown by separation of indi-
vidual fatty acids of body fat. J. Biol. Chem. 69: 239-248,
1926.
61. Engel, F. L., and J. E. White. Some hormonal influences
on fat mobilization from adipose tissue. Am. J. Clin.
Nutrition 8: 691-704, i960.
62. Engelberg, H. Heparin lipemia clearing reaction and
fat transport in man. Summary of available knowledge.
Am. J. Clin. Nutrition 8: 21-33, i960.
63. Felch, W. C, L. Sinisterra, T. B. Van Itallie, and
F. J. Stare. Vitamins and other nutrients in cardio-
vascular disease. Vitamins and Hormones 16: 127-145, 1958.
64. Fillios, L. C, S. B. Andrus, G. V. Mann, and F. J.
Stare. Experimental production of gross atherosclerosis
in the rat. J. Exptl. Med. 104: 539-554, 1956.
65. Food and Nutrition Board Report. The role of dietary
fat in human health. Natl. Acad. Sci.—Natl. Research
Council. Publ. No. 575, 1958, p. 32.
66. Frazer, A. C. Differentiation in absorption of olive oil
IIy'2
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and oleic acid in rat. ./. Physiol., London 102: 306-312, 85.
■943-
67. Frazer, A. C. Absorption of triglyceride fat from in-
testine. Physiol. Revs. 26: 103-itg, 1946. 86.
68. Frazer, A. C. The mechanism of fat absorption. Bwchem.
Soc. Symposia, Cambridge, Engl. No. 9, 5-13, 1952.
69. Frazer, A. C. Lipid metabolism. In: Biochemistry and 87.
Physiology of Nutrition, edited by G. H. Bourne and G. W.
Kidder. New York: Acad. Press, 1953, pp. 212-264.
70. Frazer, A. C. Fat absorption and its disorders. Brit.
Med. Bull. 14: 212-220, 1958.
71. Fredrickson, D. S., and R. S. Gordon, Jr. Transport of 88.
fatty acids. Physiol. Revs. 38: 585-630, 1958.
72. French, J. E., B. Morris, and D. S. Robinson. Removal
of lipides from the blood stream. Brit. Med. Bull. 14: 89.
234-238, 1958.
73. Friedman, M., R. H. Rosenman, and V. Carol. 90.
Changes in the serum cholesterol and blood clotting time
in men subjected to cyclic variations of occupational
stress. Circulation 17: 852-861, 1958.
74. Friedman, M., and R. H. Rosenman. Association of 91.
specific overt behavior pattern with blood and cardio-
vascular findings: blood cholesterol level, blood clotting
time, incidence of arcus senilis and clinical coronary
artery disease. J. Am. Med. Assoc. 169: 1286-1296, 1959. 92.
75. Friedman, M., S. M. St. George, S. O. Byers, and
R. H. Rosenman. Excretion of epinephrine, norepineph- 93.
rine, and other hormones in men exhibiting behavior
pattern (A) associated with coronary artery disease.
Circulation 20: 698, 1959.
76. Furman, R. H., R. P. Howard, and L. N. Norcia. 94.
Modification of the effect of adrenal cortical steroids
and androgens on serum lipids and lipoproteins by caloric
supplementation and by isocaloric substitution of carbo- 95.
hydrate for dietary protein. In: Hormones and Athero-
sclerosis, edited by G. Pincus. New York : Acad. Press, 96.
'959. PP- 349-37°-
77. Gildea, E. F., E. B. Man, and J. P. Peters. Proteins in
hypothyroidism. J. Clin. Invest. 18: 739-755, 1939. 97-
78. Gofman, J. W., F. Lindgren, H. A. Elliott, W. Mantz,
J. Hewitt, B. Strisower, V. Herting, and T. P. Lyon. 98.
The role of lipids and lipoproteins in atherosclerosis.
Science III: 1 66-1 71, 1950.
79. Gofman, J. W., F. Glazie, A. Tamplin, B. Strisower,
and O. De Lalla. Lipoproteins, coronary heart disease, 99.
and atherosclerosis. Physiol. Revs. 34: 589-607, 1954.
80. Goldfien, A., and R. J. Havel. The effects of nor-
epinephrine and epinephrine on unesterified fatty acid
metabolism. J. Clin. Invest. 38: 1007, 1959. 100.
81. Goldsmith, G. A. Investigation of mechanisms by which
unsaturated fats, nicotinic acid and meomycin lower
serum lipid concentration: excretion of sterols and bile 101.
acids. Trans. Assoc. Am. Physicians 72: 207-217, 1959.
82. Goldsmith, G. A. Mechanisms by which certain pharma-
cologic agents lower serum cholesterol. Federation Proc.
21 : Suppl. 1 1, 81-85, '962.
83. Goodman, D. S. Interaction of human serum albumin 102.
with long -chain fatty acid anions. J. Am. Chem. Soc. 80:
3892-3898, 1958. 103.
84. Gordon, R. S., Jr., and A. Cherkes. Unesterified fatty
acid in human blood plasma. ./. Clin. Invest. 36: 206-212,
'956
Gordon, 1 1 , B. Lewis, L. Eales, and J. F. Brock. Dietary
fat and cholesterol metabolism. FecaJ elimination of bile
acids and other lipids. Lancet 2: 1 299-1 306, 1957.
Gore, I., A. E. Hirst, Jr., and Y. Koseki. Comparison
of aortic atherosclerosis in United States, Japan and
Guatemala. Am. ./. Clin. Nutrition 7: 50-54, 1959.
Groen, J., B. K. Tjiong, C. E. Kamminga, and A. F.
Willebrands. Influence of nutrition, individuality and
some other factors, including various forms of stress, on
serum cholesterol; experiment of nine months' duration
in 60 normal human volunteers. Voeding 13:556-587, 1952.
Groen, J. Present status of knowledge of the various fac-
tors in the etiology of atherosclerotic heart disease. Ned.
melt Zuiveltijdschr. 12: 282-338, 1958.
Guravich, J. L. Familial hypercholesteremic xanthoma-
tosis: a preliminary report. Am. J. Med. 26: 8-29, 1 959.
Hansen, A. E., M. E. Haggard, A. N. Boelsche, D.J. D.
Adam, and H. F. Wiese. Essential fatty acids in infant
nutrition. Ill: Clinical manifestations. J. Nutrition 66:
565-57°. '958-
Hartroft, W. S., and W. A. Thomas. Pathological
lesions related to disturbances of fat and cholesterol
metabolism in man. J. Am. Med. Assoc. 164: 1899- 1905,
'957-
Hashim, S. A., and R. E. Clancy. Dietary fats and blood
coagulation. New Engl. J. Med. 259: 1115-1123, 1958.
Hashim, S. A., R. E. Clancy, D. M. Hegsted, and F.
J. Stare. Effect of mixed fat formula feeding on serum
cholesterol level in man. Am. J. Clin. Nutrition 7: 30-34,
■959-
Hashim, S. A., and T. B. Van Itallie. Effect of intra-
venous amino acids on nonesterified fatty acids. Proc.
Soc. Exptl. Biol. Med. 100: 576-579, 1959.
Hashim, S. A. Endocrine factors in lipid mobilization.
Diabetes 9: 135-138, i960.
Hashim, S. A., A. Arteaga, and T. B. Van Itallie.
Effect of a saturated medium chain triglyceride on serum
lipids in man. Lancet 1: 1105-1108, 1960.
Haslewood, G. A. D. Recent developments in our knowl-
edge of bile salts. Physiol. Revs. 35: 178-196, 1955.
Havel, R. J., H. A. Eder, and J. H. Bragdon. The dis-
tribution and chemical composition of ultracentrifugally
separated lipoproteins in human serum. J. Clin. Invest.
34 : 1 345-' 353. '956-
Hellman, L., R. S. Rosenfeld, W. Insull, Jr., and
E. H. Ahrens, Jr. Intestinal excretion of cholesterol:
a mechanism for regulation of plasma levels. J. Clin.
Invest. 36: 898, 1957.
Hellstrom, K., and J. Sjovall. Conjugation of bile
acids in patients with hypothyroidism (bile acids and
steroids 105). J. Atherosclerosis Research 1: 205-210, 1961.
Herb, S. F., P. Magidman, F. E. Luddy, and R. W.
Riemenschneider. Fatty acids of cows' milk. B. Composi-
tion by gas-liquid chromatography aided by other meth-
ods of fractionation. J. Am. Oil Chemists Soc. 39: 142-146,
1962.
Hilditch, T. P. The Chemical Constitution of Natural Fats
(3rd ed.). New York: Wiley, 1956.
Hirsch, J., J. YV. Farquhar, E. H. Ahrens, Jr., M. L.
Peterson, and \V. Stoffel. Studies of adipose tissue in
man. A microtechnique for sampling and analysis. .4m. J.
Clin. Nutrition 8: 499-51 1, 1960.
LIPID METABOLISM
"93
104. Hirsch, E. F., and S. Weinhouse. The role of the lipids
in atherosclerosis. Physiol. Revs. 23: 185-202, 1943.
105. Holman, R. L., H. D. McGill, J. P. Strong, and J. C.
Geer. Filtration versus local formation of lipids in patho-
genesis of atherosclerosis. J. Am. Med. Assoc. 170: 416-420,
'959-
106. Holman, R. T. Essential fatty acids. Nutrition Revs. 16:
33-35. '95s-
107. Holt, P. Incorporation of G14 labeled glycerol into
urinary lipids in a patient with chyluria. Clin. Research
10: 228, 1962.
108. Hueper, W. C. Arteriosclerosis. A.M. A. Arch. Pathol.
38: 162-181; 245-285; 350-364, 1944.
109. Hueper, W. C. Arteriosclerosis. A.M. A. Arch. Pathol.
39:5" ^5; ii7-I3I! 187-216, 1945.
1 10. Ignatowski, A. S. Alterationen in den parenchymatosen
Organen und in der Aorta der Kaninchen unter dem
Einfiuss des tierischen Eiweiss. Isviest Imp. Voyenna. Med.
Acad. St. Petersburg. 16: 154, 1908.
ill. Intengen, C. L. Studies on Coconut Oil. I. Relation to
growth and serum cholesterol levels of rats. II. Relation to bile
acid excretion in man (Thesis). New York: Columbia Uni-
versity, 1 96 1 .
112. James, A. T., and J. E. Lovelock. Essential fatty acids
and human disease. Brit. Med. Bull. 14: 262-266, 1958.
113. Jolliffe, N., S. H. Rinzler, and M. Archer. The
anti-coronary club, including a discussion of the effects
of a prudent diet on the serum cholesterol level of middle-
aged men. Am. J. Clin. Nutrition 7: 451-462, 1959.
1 14. Kannel, W. B., T. R. Dawber, A. Kagan, N. Revotskie,
and J. Stokes III. Factors of risk in the development of
coronary heart disease — six year follow-up experience.
The Framingham study. Ann. Internal Med. 55: 33-50,
1961.
1 15. Kelsey, F. E., and H. E. Longenecker. Distribution and
characterization of beef plasma fatty acids. J. Biol. Chem.
>39: 727-740. >94i-
116. Keys, A., and J. Brozek. Body fat in adult man. Physiol.
Revs. 33: 245-325, 1953.
1 17. Keys, A. Atherosclerosis: Problem in newer public health.
J. Alt. Sinai Hasp. N. Y. 20: 1 18-139, 1953.
118. Keys, A. Diet and the epidemiology of coronary heart
disease. J. Am. Med. Assoc. 164: 1912-1919, 1957.
1 19 Keys, A., J. T. Anderson, O. Mickelsen, S. F. Adelson,
and F. Findanza. Diet and scrum cholesterol in man : Lack
of effect of dietary cholesterol. J. Nutrition 59: 39-56, 1956.
120. Keys, A., J. T. Anderson, and F. Grande. Prediction of
serum-cholesterol response of man to changes in fats in
the diet. Lancet 2: 959, 1957.
121. Keys, A. Effect of dietary cholesterol on serum cholesterol
in man. Am. J. Clin. Nutrition 9: 126, 1961.
122. Kinsell, L. W., J. W. Partridge, L. A. Boling, S.
Margen, and G. D. Michaels. Dietary modification of
serum cholesterol and phospholipid levels. J. Clin. Endo-
crinol. 12: 909-913, 1952.
123. Kinsell, L. W., and G. D. Michaels. Letter to the editor.
Am. J. Clin. Nutrition 3: 247-253, 1955.
124. Kinsell, L. W., G. D. Michaels, R. W. Friskey, and
S. Splitter. Essential fatty acids, lipid metabolism, and
atherosclerosis. Lancet 1 : 334, 1958.
125. Knobil, E. Direct evidence for fatty acid mobilization in
response to growth hormone administrations in rat.
Proc. Soc. Expll. Biol. Med. 101: 288-289, '959-
126. Lansing, A. I. The role of elastic tissue in the formation
of the arteriosclerotic lesion. Ann. Internal Med. 36: 39-49,
'952-
127. Laster, L., and F.J. Ingelfinger. Intestinal absorption
— aspects of structure, function and disease of the small-
intestine mucosa. New Engl. J. Med. 264: 1 192-1200;
1246-1253, 1961.
128. Laurie, W., and J. D. Woods. Atherosclerosis and its
cerebral complications in the South African Bantu.
Lancet 1 : 231-232, 1958.
129. Lawry, E. Y., G. V. Mann, A. Peterson, A. P. Wysocki,
R. O'Connell, and F. J. Stare. Cholesterol and beta
lipoproteins in the serums of Americans. Am. J. Med.
22:605-623, 1957.
130. Leary, T. Crystalline ester cholesterol and atherosclerosis.
A.M. A. Arch. Pathol. 47: 1-28, 1949.
131. Levere, A. H., R. C. Bozian, G. Graft, R. S.Jackson,
and C. F. Wilkinson. The "sitosterols": variability of
serum cholesterol levels and difficulty of evaluating de-
cholesterolizing agents. Metabolism 7: 338-348, 1958.
132. Lew, E. A. Biostatistical pitfalls in studies of athero-
sclerotic heart disease. Federation Proc. 21: Suppl. II,
62-70, 1962.
133. Lewis, B. Composition of plasma cholesterol ester: in
relation to coronary-artery disease. Lancet 2: 71-73, 1958.
134. Lewis, L. A., and I. H. Page. Electrophoretic and ultra-
centrifugal analysis of serum lipoproteins of normal,
nephrotic and hypertensive persons. Circulation 7 : 707-
7'7, '953-
135. Lindgren, F. T., H. A. Elliott, and J. W. Gofman.
Ultracentrifugal characterization and isolation of human
blood lipides and lipoproteins, with applications to the
study of atherosclerosis. J. Phys. and Colloid Chem. 55:
80-93, '95' ■
136. Luddy, F. E., R. A. Barford, R. W. Riemenschneider,
and J. D. Evans. Fatty acid composition of component
lipides from human plasma and atheromas. J. Biol. Chem.
232: 843-851, 1958.
137. Lynn, W. S., R. M. MacLeod, and R. H. Brown. Effects
of epinephrine, insulin, and corticotrophin on the metabo-
lism of rat adipose tissue. J. Biol. Chem. 235: 1 904-1 911,
i960.
138. MacKenzie, R. D., and T. R. Blohm. Effects of MER 29
on cholesterol biosynthesis. Federation Proc. 18: 417, 1959.
139. Malmros, H., B. Swahm, and E. Truedsson. Essential
hyperlipemia. Acta Med. Scand. 149: gi-108, 1954.
140. Malmros, H., and G. Wigand. The effect on serum-
cholesterol of diets containing different fats. Lancet 2 :
1-7. "957-
141. Man, E. B., and M.J. Albrink. Serum lipids in different
phases of carbohydrate metabolism. Yale J. Biol, and
Med. 29: 316-334, 1956.
142. Mann, G. V., and F.J. Stare. Nutrition and atherosclero-
sis. In : Symposium on Atherosclerosis. Natl. Acad. Sci.-Natl.
Research Council. Publ. No. 338, 1955, pp. 1 69-1 80.
143. Mann, G. V., and S. B. Andrus. Xanthomatosis and
atherosclerosis produced by diet in an adult rhesus mon-
key. J. Lab. Clin. Med. 48: 533-550, 1956.
144. Marchand, F. Uber Arteriosklerose (Athero-sklerose).
Verhandl. Kongr. Inn. Med. 21: 23, 1904.
i 1 1,4
HANDBOOK OF PHVSIOI .Ol ;"i
CIRCULATION II
145. Mead, J. F., and D. R. IIowton. Digestion and absorp- 167
tion. In: Radioisotope Studies of Fatty Acid Metabolism,
edited by J. F. Mead and D. R. Howton. New York: 168
Pergamon, i960, pp. 1-14.
146. Mead, J. F. The metabolism of the polyunsaturated
fatty acids. Am. J. Clin. Nutrition. 8: 55—6 1 , i960. 169
147. Miles, H. H. W., S. Waldfogel, E. L. Barrabee and
S. Cobb. Psychosomatic study of 46 young men with
coronary artery disease. Psyckosomat. Med. 16: 455-477, 170.
'954-
148. Mobilization of depot fat. Nutrition Revs. 13: 207-209,
1955- «7i'
I49_Morris, J. N. Recent history of coronary disease. Lancet
1: 1-7; 69-73, >95'-
150. Morris, J. N. Fats and disease. Lancet 1 : 687-689, 1956.
151. Oliver, M. F., and G. S. Boyd. Reduction of
serum-cholesterol by dextro-thyroxine in men with 172.
coronary heart-disease. Lancet 1: 783-785, 1961.
152. Olsen, R. E., J. W. Vester, D. Gursey, N. Davis, and 173.
D. Longman. Effect of low protein diets upon serum
cholesterol. Am. J. Clin. Nutrition 6: 310-324, 1958.
■53- Olsen, R. E., and J. W. Vester. Nutrition-endocrine
interrelationships in the control of fat transport in man. 1 74.
Physiol. Revs. 40: 677-733, i960.
1.54- Oncley, J. L., F. R. N. Gurd, and M. Melin. Prepara-
tion and properties of serum and plasma proteins. XXV.
Composition and properties of human serum /^-lipoprotein.
J. Am. Chem. Soc. 7 2 : 458-464, 1 950. I 75-
'55- Oncley, J. L. Plasma lipoproteins. In: Chemistry of Lipides
as Related to Atherosclerosis, edited by I. H. Page. Spring- ljS.
field, 111.: Thomas, 1958, pp. 1 14-133.
'5D- O'Neal, R. M., and W. J. S. Still. Pathogenesis of
atherosclerosis. Federation Proc. 21 : Suppl. II, 12-14, '9D2- '77-
'57- Osborne, R. H., D. Adlersberg, F. V. DeGeorge, and
C. Wang. Serum lipids, heredity and environment.
Am. J. Med. 26: 54-59, 1959. l7&-
158- Parsons, W. B. Studies of nicotinic acid use in hyper-
cholesteremia. Arch. Internal. Med. 107: 653-667, 1961.
159- Perkins, R., I. S. Wright, and B. W. Gatje. Effect of
safflower oil emulsion on serum cholesterol levels in young '79-
adult males. J. Am. Med. Assoc. 166: 2 132-2 135, 1958.
160. Peterson, M. L. The Transport of Fat in Man: A Study of
Chylomicrons (Thesis). New York: Rockefeller Institute,
i960.
161. Pfluger, E. F. W. Fortgesetzte Untersuchungen fiber die
Resorption der kiinstlich gefarbten Fette. Pflugers Arch. 180.
ges. Physiol. 85: 1-58, 1901.
162. Poole, J. C. F., and H. VV. Florey. The changes in the
endothelium of the aorta and the behavior of macrophages
in experimental atheroma of rabbits. J. Pathol. Bacteriol. lg[
75 : 245-252, 1958.
163. Poole, J. C. F. Fats and blood coagulation. Bnl. Med.
Bull. 14: 253-258, 1958. ig2
164. Poole, J. C. F. Effect of diet and lipemia on coagulation
and thrombosis. Federation /'roe. 21: Suppl. [I, 20-24,
1962.
165 Portman, O. W., E. Y. Lawry, and D. Bruno. Effect l83-
of dietary carbohydrate on experimentally induced
hypercholesteremia and hyperbetalipoproteinemia in
rats. Proc. Soc. Exptl. Biol. Med. 91 : 321-323, 1956. 184.
166 Portman, O. W., and F. J. Stare. Dietary regulation of
serum cholesterol levels. Physiol. Revs. 39. 407-442, 1959.
Potiiier, I.., and T. B. Van Itallie. Role of the thyroid
in lipid mobilization. Clin. Research 8: 377, 1960.
Raben, M. S., and C. H. Hollenberg. Effect of growth
hormone on plasma fatty acids. J. Clin. Invest. 38: 484-
488, 1959-
Rhodes, D. N., and C. H. Lea. Phospholipids. IV. On
the composition of hen's egg phospholipids. Biochem. J.
65: 526-533. '957-
Rich, C, E. L. Bierman, and I. L. Schwartz. Plasma
nonesterilied fatty acids in hyperthyroid states. J. Clin.
Invest. 38: 275-278, 1959.
Rittenberg, D., and R. Schoenheimer. Deuterium as
indicator in study of intermediary metabolism; further
studies on biological uptake of deuterium into organic
substances, with special reference to fat and cholesterol
formation. J. Biol. Chem. 121: 235-253, 1937.
Rizack, M. A. The effect of epinephrine on the lipolytic
activity of adipose tissue. Federation Proc. 19: 221, 1960.
Rivin, A. U., and S. P. Dimitroff. Incidence and severity
of atherosclerosis in estrogen-treated males, and in females
with hypoestrogenic or hyperestrogenic state. Circulation
9- 533-539. '954-
Roberts, J. C, Jr., R. H. Wilkins, and C. Moses.
Autopsy studies in atherosclerosis. II. Distribution and
severity of atherosclerosis in patients dying with morpho-
logic evidence of atherosclerotic catastrophe. Circulation
20: 520-526, 1959.
Rodbard, S. Physical forces and the vascular lining. Ann.
Inter mil Med. 50: 1 339-1 351, 1959.
Rodbell, M., D. S. Fredrickson, and K. Ono. Metabo-
lism of chylomicron proteins in dog. J. Biol. Chem. 234:
567-57i. 1959.
Rudman, D., and F. Seidman. Lipemia in the rabbit
following injection of pituitary extract. Proc. Soc. Exptl.
Biol. Med. 99: 146-150, 1958.
Rudman, D., M. DiGirolamo, F. Seidman, and M. B.
Reid. Purification and properties of a pituitary com-
ponent which produces lipemia in the rabbit. J. Clin.
Invest. 39: 1023, 1958.
Russek, H. I., and B. L. Zohman. Relative significance
of heredity, diet and occupational stress in coronary heart
disease of young adults: based on analysis of 100 patients
between ages of 25 and 40 years and similar group of 100
normal control subjects. .4m. J. Med. Sci. 235: 266-277,
1958.
Rutstein, D. D., E. F. Ingenito, J. M. Craig, and M.
Martinelli. Effects of linolenic and stearic acids on
cholesterol-induced lipoid deposition in human aortic
cells in tissue culture. Lancet 1 : 545-552, 1958.
Samuel, P., and A. Steiner. Effect of neomycin on serum
cholesterol level of man. Pioc. Soc. Exptl. Biol. Med. 100:
i93-'95. 1959-
Scanu, A., and I. H. Page. Separation and characteriza-
tion of human serum chylomicrons. J. Exptl. Med. 109:
239"256, '959-
Schoenheimer, R. Uber eine Storung der Cholesterin-
Ausscheidung. (Ein Beitrag zur Kenntnis der Hyper-
cholesterinamien.) Z. klin. Med. 123: 749-763, 1933.
Schoenheimer, R. The investigation of the intermediary
metabolism with the aid of heavy hydrogen. In: Harvey
Lectures. Baltimore: Williams & Wilkins, 1937, p. 122.
I.II'ID METABOLISM
I 195
185. Seifter, J., and D. H. Baeder. Lipid mobilize]- (LM) from
posterior pituitary of hogs. Proc. Soc. Exptl. Biol. Med.
95 : 3'8-32°. '957-
186. Seller, R. H., J. Braciifeld, H. Sandberg, and S.
Bellet. Use of I13l-labellcd fat in study of lipid handling
in patients with coronary artery disease. Am. J. Med.
27: 231-240, 1959.
187. Shafrir, E., K. E. Sussman, and D. Steinberg. The
nature of the epinephrinc-induced hyperlipidemia in
dogs and its modification by glucose. J. Lipid Research
1: 109-117, 1959.
188. Shipley, R. E. Symposium on sitosterol. 1. Effects of
sitosterol ingestion on serum cholesterol concentration.
Trans. N. Y. Acad. Set. 18: 111-118, 1955.
189. Sinclair, H. M. (Editor). Essential Fatly Acids. New
York: Acad. Press, 1958.
190. Sinclair, H. M. Deficiency of essential fatty acids and
atherosclerosis. Lancet 1: 381-383, 1956.
191. Siperstein, M. D., F. M. Harold, I. L. Chaikoff, and
W. G. Dauben. C'-cholesterol : bilary end-products of
cholesterol metabolism. J. Biol. Chem. 210: 181-191, 1954.
192. Siperstein, M. D., and A. W. Murray. Cholesterol
metabolism in man. J. Clin. Invest. 34: 1449-1453, 1955.
193. Steinberg, D., J. Avigan, and E. B. Feigelson. Effects
of triparanol (MER-29) on cholesterol biosynthesis and
on blood sterol levels in man. J. Clin. Invest. 40: 884-893,
1961.
194. Surgenor, D. M. Extracellular lipoproteins, [n: Symposium
on Atherosclerosis. Natl. Acad. Sci.-Natl. Research Council Publ.
No. 338, 1955.
195. Taylor, H. E. The role of mucopolysaccharides in the
pathogenesis of intimal fibrosis and atherosclerosis of the
human aorta. Am. J. Pathol. 29: 871-883, 1953.
196. Tennent, D. M., H. Siegel, M. E. Zanetti, G. W.
Kuron, W. H. Ott, and F. J. Wolf. Plasma cholesterol
lowering action of bile acid binding polymers in experi-
mental animals. J. Lipid Research. 1 : 469-473, i960.
ig7. Thomas, C. B., and E. A. Murphy. Further studies on
cholesterol levels in Johns Hopkins medical students :
effect of stress at examinations. J. Chronic Diseases 8:
661-668, 1958.
198. Thomasson, H. J. Biological standardization of essential
fatty acids. Intern. Rev. Vitamin Research 25: 62, 1953.
199. Tuna, N., L. Regkers, and I. D. Frantz. Fatty acids of
total lipids and cholesterol esters from normal plasma
and atheromatous plagues. ./. Clin. Invest. 37: 1153-1165,
1958-
203.
204.
205.
206.
207.
208.
209.
213.
214.
Van Itallie, T. B. Nutritional research in atherosclerosis;
a progress report. J. Am. Dietet. Assoc. 34: 248-253, 1958.
Van Itallie, T. B., and W. C. Felch. Reflections on the
pathologic physiology of atherosclerosis. New Engl. J.
Med. 263: 1179-1184; 1243-1246, i960.
Van Itallie, T. B., S. A. Hashim, R. S. Grampton, and
D. M. Tennent. The treatment of pruritus and hyper-
cholesteremia of primary biliary cirrhosis with cholestyra-
mine. New Engl. J. Med. 265: 469-474, 1961.
Vaughan, M. The metabolism of adipose tissue in vitro.
J. Lipid Research 2: 293-316, 1961.
Walker, W. J., E. Y. La wry, D. E. Love, G. V. Mann,
S. A. Levine, and F. J. Stare. Effect of weight reduction
and caloric balance on serum lipoproteins and cholesterol
levels. Am. J. Med. 14: 654-664, 1953.
Wai ki.r, A R. P., and H. Grusin. Coronary heart disease
and cerebral vascular disease in South African Bantu :
examination and discussion of crude and age specific
death rates. Am. J. Clin. Nutntion 7: 264-270, 1959.
Weiss, S. B., E. P. Kennedy, and J. Y. Kiyasu. The
enzymatic synthesis of triglycerides. J. Biol. Chem. 235:
40-44, i960.
Wertheimer, E., and B. Shapiro. The physiology of
adipose tissue. Physiol. Revs. 28: 451-464, 1948.
Werthessen, N. T., W. R. Nelson, A. T. James, and
R. L. Holman. Composition of fatty acids in cholesterol
esters derived from normal and abnormal intima. Cir-
culation 20: 972, 1959.
Wessler, S. Thromboangiitis obliterans: fact or fancy.
Editorial. Circulation 23: 165-167, 1961.
White, J. E., and F. L. Engel. A lipolytic action of
epinephrine and norepinephrine on rat adipose tissue.
Proc. Soc. Exptl. Biol. Med. 99: 375-378, 1958.
Wigand, G. Production of hypercholesteremia and
atherosclerosis in rabbits by feeding different fats without
supplementary cholesterol. Acta Med. Scand. Suppl. 351:
i-9". '959-
Wuest, J. H., T. J. Dry, and J. E. Edwards. Degree of
coronary atherosclerosis in bilaterally oophorectomized
women. Circulation 7: 801-809, '953-
Zarafonetis, C. J. D., G. M. Miller, J. Seifter, D.
Baeder, R. M. Myerson, and W. A. Steiger. Metabolic
studies in patients receiving lipid mobilizer hormone.
Am. J. Med. Sci. 234: 493-5°4. '957-
Zilversmit, D. B., E. L. McCandless, P. H. Jordan, Jr.,
W. S. Henly, and R. F. Ackerman. The synthesis of
phospholipids in human atheromatous lesions. Circulation
23: 370-375. ]961-
CHAPTER 35
The role of endocrines, stress, and
heredity on atherosclerosis1
L. N. KATZ
R. PICK2
Cardiovascular Institute, Michael Reese Hospital and Medical Center, Chicago, Illir,
CHAPTER CONTENTS
Hormones
Thyroid
Pancreatic Hormones
Chronic pancreatitis
Diabetes mellitus
Adrenal and Pituitary Hormones
Adrenal cortical hormones and ACTH
Adrenal medullary hormones
Anterior pituitary hormones
Conclusion
Sex Hormones
Heredity
Stress
Physical Activity
Emotional Stress
Summary
atherosclerosis, manifested in the lipid-containing
intimal lesions of small and large arteries, is the most
common pathological form of vascular disease and
the most detrimental in its effect on the blood and
oxygen supply to any given organ. It is one form of
arteriosclerosis, the most important one, leading to
widespread morbidity and mortality in man in our
Western civilization.
Several investigative approaches have led to the
1 Work of the institute mentioned in this communication
has been supported by grants from the National Institutes of
Health, National Heart Institute USPHS (H-2276, ^3031), the
Chicago Heart Association, the Albert and Mary Lasker
Foundation, and the Michael Reese Research Foundation.
2 Established Investigator of the American Heart Association.
conclusion, held by most but not by all workers in the
field, that it is a disease primarily due to disturbance
of the metabolism of lipid, lipoprotein, or cholesterol,
or all three (72). Whether atherosclerosis develops
into a major health problem within a population
depends to a large extent on the life-span pattern of
its diet. As early as 1934, Rosenthal (133) established
that in no population with a high intake of fat and
protein from animal sources is atherosclerosis absent,
while populations subsisting on a diet low in animal
fat and protein are uniformly free from the disease
anatomically and, therefore, from the sequelae which
produce morbidity and mortality. These findings
have been amply confirmed in recent years by world-
wide epidemiological studies (42, 74, 79, 107, 157).
A tangible concomitant of the ingestion of a diet
rich in saturated fats and cholesterol is a hypercholes-
terolemic tendency in a population. Thus, while
the mean serum cholesterol level of the atherosclero-
sis-free populations is 150 to 180 mg per cent, the
level for clinically healthy men of comparable age in
the United States is 220 mg per cent (74). It is also a
well-accepted fact today that serum cholesterol level
is the most closely related single factor determining
an individual's risk of developing clinical athero-
sclerotic coronary disease, i.e., the higher the serum
cholesterol, the greater is the risk (30).
According to our present knowledge, the mode of
action by which a diet rich in fats, particularly satu-
rated fats and cholesterol, acts to influence lipid
metabolism and to produce atherosclerosis can be
summarized as follows: cholesterol synthesis in the
1 197
"98
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
liver is finely attuned to the amount of ingested
cholesterol (61). This homeostatic mechanism is
disturbed or may even be exhausted by a high-fat,
high-cholesterol diet over the life span. This could
explain the slowly increasing serum cholesterol
levels with aging in our population. Recent findings
also suggest that cholesterol synthesis by extrahepatic
tissues, not regulated by dietary intake, may con-
tribute significantly to the development of hyper-
cholesterolemia and, therefore, atherosclerosis (i).
Animal experiments lead us to suspect that the daily
pattern of eating, the number of meals, for example,
may help to determine the metabolic fate of the
constituents of a potentially atherogenic diet (28).
In the case of a particular individual, the tendency
to develop this increase in blood cholesterol and to
acquire vascular disease will be subject to many
factors other than the nature of the diet (73, 114).
These other factors per se do not actually produce or
prevent atherosclerosis, but they are capable of
influencing it in the presence of a potentially athero-
genic diet. Those which will be considered in this
chapter are: a) the endocrines, b) heredity, and c)
stress. They are the most significant ancillary factors
so far known.
Because of discrepancies between the amount of
anatomical vascular disease and the occurence and
magnitude of organ involvement, doubt has recently
been expressed as to the relationship of atherosclerosis
and, for instance, coronary heart disease (129, 163).
However, no evidence is available that ischemic heart
disease and ischemic disease of the brain, the extrem-
ities, or other organs occur without vascular disease
(except on rare occasion). The major exception is one
in which, with only minimal atheroma formation, it
is possible to produce experimental coronary and
renal thrombosis, and myocardial and renal infarction
in rats (63), but even here vascular abnormalities
were produced only in the presence of a high
saturated-fat diet.
It is safe to state that without the basic arterial
process of atheroma no morbid consequences would
exist except as a rare phenomenon. Equally well
documented is the fact that even with moderate and
se\ ere atherosclerosis no such morbid or mortal
consequences need occur. These findings clearly point
out that in atherosclerosis research we have to deal
with two major questions: /) What produces the
basic vascular lesions? 2) What factor or factors may
lead to the complications — ulceration, thrombosis,
hemorrhage into a plaque— that will ultimately
determine the clinical fate of an individual? It is
possible that the same factors may determine both
aspects. For example, prolonged hyperlipemia and
hypercholesterolemia produce lipid deposition and
atheroma in the arteries, and these blood changes also
facilitate blood clotting, so that after an atheroma has
developed in this fashion the stimulus is there to give
rise to subsequent thrombus formation. Similarly,
different neurogenic or hormonal factors, or both,
may conceivably influence both processes. However,
their effect may be preferential upon one or the other
of these two stages. It must also be remembered that
the vascular wall as an organ is capable of synthesizing
cholesterol in small amounts and phospholipids in
larger quantities (170, 183). Furthermore, it has
been shown that species differences exist in the 02
uptake between normal and atherosclerotic aortas,
the ()•> uptake being higher in susceptible species
and in atherosclerotic specimens (173). Permeability
of the vascular endothelium is another factor that
may be influenced by metabolic alterations due to
hormonal, genetic, or stressful circumstances mediated
by hormone release. Electron microscopy has con-
firmed the concept that lipids are being deposited
in the intima by permeation from the blood stream
('59)-
Furthermore, differences in the responses of the
vessels in different vascular beds to hormonal and
other influences must not be overlooked. Whether
these differences are due to the particular metabolism
within the organ, to the nervous influences acting
upon it, or to anatomical differences — possibly due to
genetic factors — is not known at present. Some evi-
dence for each of these causes is available (83, 94,
138).
Whether a given duration and intensity of hyper-
lipemia and hypercholesterolemia will or will not
lead to the emergence of atherosclerotic disease,
either in the form of the anatomical substrate alone or
accompanied by the associated sequelae, is deter-
mined by the genetic make-up of the individual and
very likely also In the nature of the environmental
conditions under which he lives out his existence.
Emotional factors, dependent in part upon genetic
make-up and in part upon external environment,
have recently been implicated in aberrations of lipid
metabolism, in the genesis of atherosclerosis, and in
the transformation of a silent vascular disease into a
clinically overt one. Whether emotional states
operate through hormones, or by way of the auto-
nomic nervous system, or both, is not known at
present. It is possible that hormonal and nervous
factors themselves produce the emotional upsets as a
ENDOCRINES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
I!99
side effect independent of their direct actions upon
atherosclerosis and ischemic disease. It is more likely,
however, that emotional upsets induce hormonal and
nervous factors which lead to ischemic disease.
It is thus apparent that any attempt to understand
the pathogenesis of the multifaceted process of
atherosclerosis requires that many factors be con-
sidered. In considering them, the effects of each on
lipid metabolism, on the vascular wall, on blood
coagulation, and on fibrinolysis have to be studied
separately, and after that all the facts must be inte-
grated to reconstruct the whole complex process. At
present there are many gaps in our knowledge which
are difficult to bridge. While the influences of hor-
mones and heredity on lipid metabolism and on the
vascular wall have been studied to some extent, their
influences on blood clotting and fibrinolysis are poorly
understood at present, since such studies are still in
their infancy. Equally scanty is our knowledge of the
effect of emotional factors on atherosclerosis and its
sequelae.
HORMONES
That hormones influence lipid metabolism and
atherosclerosis has long been suspected from clinical
findings. Several diseases of endocrine organs show
alterations in serum lipid levels and are associated
with significant deviations in the incidence and
severity of atherosclerosis. Several hormones are also
known to affect the morphologic characteristics of the
ground substance of the vascular wall and its cell
membrane permeability. This would indicate that
hormones may influence atherosclerosis either by a
direct action on the vascular wall or through their
influence on lipid metabolism (synthesis, absorption,
transport, storage, excretion, and destruction), or
both. The recognition of these factors stimulated
extensive research into the mechanism of these
actions. Only the action of the following hormones
will be considered here: a) thyroid, b) pancreas, c)
adrenal and pituitary, and d) sex. The part played
by others is too poorly understood and too unimpor-
tant to warrant discussion.
Thyroid
The effect of thyroid hormone on lipid metabolism,
particularly cholesterol metabolism, has been studied
extensively in man and in various species of experi-
mental animals. Endogenous thyroid hypersecretion,
as occurs in thyrotoxicosis, and the exogenous
administration of the hormone have identical effects
and will be discussed together. Different forms of
hypothyroidism, whether primary or secondarily
induced in man and animals by surgical thyroid-
ectomy or by I131 administration, also show similar
effects. The suppression of thyroid hormone secretion
by thiourea drugs shows — in rats at least — a greater
effect on cholestrol metabolism than that produced
by the other methods of inducing hypothyroidism
(22, 62).
Hyperthyroidism decreases serum cholesterol levels
in man and animals. Recent tracer studies indicate
that thyroid hormone increases synthesis of cholesterol
in the liver, particularly of the free cholesterol
fraction, and also increases catabolism and fecal
excretion of this sterol (cf 22). Boyd (22) found in
rats that neither exogenous thyroid hormone nor
active thyroid hormone analogues lower normal serum
cholesterol levels appreciably; however, if animals are
made slightly hypercholesterolemic by dietary means,
then these hormone preparations depress the dietary
hypercholesterolemia. It has been demonstrated that
this action of thyroxin or thyroid hormone is not due
to the increase in basal metabolic rate per se, as
several thyroxin analogues show the cholesterol
depressant action without increase in basal metab-
olism (21). Furthermore, in chicks it was shown that
dinitrophenol, a drug that increases basal metab-
olism, has no effect on serum cholesterol levels
(148). Thyroid hormones also reduce /^-lipoprotein
concentration and that of certain classes of high
density a-lipoproteins (70).
Hypothyroidism in man and animals produces a
decreased synthesis of cholesterol while the biological
half-life of serum cholesterol is increased and fecal
excretion is reduced (22). The effect of hypothy-
roidism on lipoproteins is the direct opposite of the
effect of exogenous thyroid hormone administration,
i.e., it causes an increase in /3-lipoproteins.
Pituitary thyroid stimulating hormone is without
direct effect on cholesterol metabolism and athero-
sclerosis.
Thyroid hormone also affects the vascular wall.
This has long been established in the older literature.
Large doses of thyroxin or dessicated thyroid cause
damage to the vascular media. They produce necrosis
and calcification, similar to the changes produced by
catecholamines (10). These are arteriosclerotic
changes, not atherosclerosis. Smaller doses of thyroid
hormone preparations have been shown to reduce
cholesterol-oil-induced hypercholesterolemia and
1200 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
atherosclerosis in rabbits (164). In chicks, larger
doses of these hormones always depress diet-induced
hypercholesterolemia; however, the effect on the
vascular lesions is inconstant and inconsistent (152,
154). This may be due to the second action of the
thyroid in producing vascular damage, providing a
favorable site for lipid deposition and atherosclerosis.
In contrast to the inconsistent effects of excess
thyroid hormones, the deficiency of the hormone
(hypothyroidism) always produces increased athero-
genesis in animals on a potentially atherogenic diet.
This has been shown in chicks (154), rats (178),
rabbits (165), and monkeys (146). Dogs develop
atherosclerotic lesions only when a high-fat, high-
cholesterol diet is combined with hypothvroidism
(158).
No information is available at present indicating
an effect of thyroid hormones on blood coagulation or
fibrinolysis.
It can be concluded from all available data that
thyroid hormone has a significant and important
effect on cholesterol metabolism. The direct effect on
atherosclerosis is undetermined and questionable.
The continuing effort to separate calorigenic from
hypocholesterolemic effects in thyroid analogues
ultimately may alter the utility of thyroid preparations
as antiatherogenic substances.
Pancreatic Hormones
Studies in man and in experimental animals
indicate that two hormonal systems in the pancreas
are actively involved in lipid metabolism and,
therefore, in the control of the circulating serum
lipids. These can operate independently or, more
often, in an interrelated manner.
Knowledge of the two hormonal systems of the
pancreas in man has been derived from the study of
pancreatitis and diabetes mellitus. In addition,
pancreatic enzyme systems are known to influence
absorption from the upper digestive tract. Elastase,
presumably a pancreatic enzyme, by influencing the
elastic tissue in the media of the blood vessels, can
change wall permeability and thus modify calcium
and lipid deposition in the intima (82).
chronic pancreatitis. Chronic pancreatitis in man
with hyperlipemia and xanthomatosis, without
diabetes, was first described by Wiesel (175) in 1905.
Binet & Brocq (15) in 1929 reported a transient
hyperlipemia and hypercholesterolemia in dogs with
experimental pancreatitis. An antifatty liver sub-
stance high in bound choline was prepared from the
pancreas of dogs by Drae;stedt (34, 35). Adlersberg
and co-workers carried out the most recent studies on
experimental pancreatitis in dogs and rabbits, and
also studied chronic pancreatitis in man (7, 167).
He described the serum changes as consisting of a
two- to three-fold increase in cholesterol and phos-
pholipids with a four- to ten-fold increase of total
lipids, the triglycerides, rendering the serum
lactescent. The mechanism producing these serum
lipid changes has not been clarified. An action by way
of the enzyme system affecting lipid absorption has
to be considered. Also destruction of the a-cells of
the islets of Langerhans and their glucagon content
may be involved (26). The elevation of triglycerides
is considered the primary change leading secondarily
to hypercholesterolemia and hyperlipemia by others
(47). The significance of these findings in the patho-
genesis of atherosclerosis needs further study.
diabetes mellitus. The grossly and significantly
increased incidence of atherosclerosis in individuals
with diabetes mellitus has led to numerous clinical
and experimental studies on the influence of the
hormones of the islets of Langerhans, particularly
insulin, on carbohydrate and lipid metabolism and on
atherosclerosis.
The morphology of the arterial lesions in the
diabetic does not differ from that in the nondiabetic.
The difference between the two, then, is quantitative.
However, in diabetes mellitus a characteristic capil-
lary lesion in the retina and the kidney is found,
consisting of capillary microaneurysms. Changes in
serum lipids and complex carbohydrates are usually
found when capillary lesions are present.
Severely atherosclerotic diabetic patients frequentlv
show distinct disturbances of lipid and lipoprotein
metabolism, including hyperlipemia, hypercho-
lesterolemia, hyper-/j-lipoproteinemia, and a marked
elevation of esterified fatty acids (5, 16). They also
have increased levels of serum polysaccharides. In
diabetic acidosis and ketosis marked hyperlipemia and
hvpercholesterolemia are present, in addition to
hyperglycemia. Insulin treatment results in bringing
all three abnormalities toward normal. However,
insulin given to normal individuals has no cho-
lesterol-lowering effect (20).
Experimental studies on the effect of diabetes
mellitus and of insulin on lipid metabolism and
atherosclerosis have been carried out on numerous
animal species, including dogs, rabbits, rats, and
chicks. In all animals tested, diabetes produced b\
ENDOCRINES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
I 20 I
alloxan injections or by pancreatectomy failed to
cause atherosclerosis. When experimental diabetes
was combined with cholesterol-fat feeding, athero-
sclerosis incidence was not higher than in normal
animals on the same diet. Of particular interest are
the findings by Duff et al. (38), who showed a de-
creased atherogenesis in diabetic rabbits on a cho-
lesterol-oil diet. This trend was reversed when the
animals were treated with insulin. These workers
attributed this effect to the particular serum lipid
picture developing under these circumstances.
Alloxan diabetes produced a marked hyperphos-
pholipemia with hypercholesterolemia and an in-
crease in serum neutral fat. These animals, therefore,
had a low ratio of total cholesterol to phospholipids
(C/P ratio) in the hypercholesterolemic state. This
particular lipid picture is usually accompanied by a
low incidence and severity of vascular lesions. Insulin
given to these rabbits caused the lipid picture to
change so as to resemble the usual pattern obtained
by cholesterol-fat feeding alone, namely a marked
hypercholesterolemia with a mild hyperphospho-
lipemia. This, in turn, resulted in an elevated C/P
ratio and the attendant increased incidence and
severity of atherosclerosis.
Interesting results were also obtained in studies of
the pancreas and atherogenesis in chicks (154, 156)
Pancreatectomized birds show no overt signs of
disturbances of lipid or glucose metabolism. However,
latent disturbances can be detected when these birds
are given a high-cholesterol, high-fat diet or when
adrenal steroids are administered. On this diet they
show enhanced hypercholesterolemia and athero-
sclerosis, as well as retarded healing of lesions. With
glucocorticoids they show a definite hyperglycemic
response which is much greater than occurs in normal
animals given these steroids. Cholesterol-fed, steroid-
diabetic chicks do not show increased atherogenesis.
Insulin, in hypoglycemic doses, when given to normal
chicks does not increase the atherogenic potential of
a high-cholesterol, high-fat diet. However, in these
same doses insulin prevents regression of coronary
artery lesions when it is given to chicks which are
first made atherosclerotic and then placed on a
plain, nonatherogenic diet — a diet which by itself
normally leads to rapid regression of these early
coronary lesions. The mechanism by which insulin
prevents regression, while at the same time appearing
to be without effect during the induction phase, is
not clear. Large doses of insulin were used in these
experiments, and this did cause marked hypoglycemia
which in some way acted in a detrimental manner.
Also, the insulin probably increased the secretion of
catecholamines and corticoids, as evidenced by the
occurrence of periods of reactive hyperglycemia.
Furthermore, some recent observations indicate that
chronic insulin administration may produce pro-
longed hyperglycemia after the drug administration
is discontinued, indicating some profound hormonal
derangement. Local effects within an atheroma also
cannot be excluded.
How much of the effect of diabetes mellitus or
insulin on atherosclerosis is due to the changes in
lipid metabolism and how much to factors influencing
the vascular wall is not clearly established. Further-
more, several authors (32, 84) have suggested that in
diabetics, and even in nondiabetic members of their
families, the ground substance of the vascular wall is
subtly changed, making it particularly prone to
atherosclerosis. In addition, blood coagulation is
changed in uncontrolled diabetes as in other hyper-
lipemia states.
From all this it is apparent that the increased
tendency of the diabetic to develop atherosclerosis
must depend on a number of factors.
Adrenal and Pituitary Hormones
Adrenal cortical and medullary hormones have
been shown to influence lipid metabolism and the
vascular wall. However, the lipid metabolic responses
to these hormones differ among the several animal
species studied, including man. Also, their acute and
chronic effects on circulating serum lipids differ.
The mechanism of their action has not been satis-
factorily elucidated.
adrenal cortical hormones and acth. Hyper-
activity of the adrenal cortex in Cushing's disease is
frequently associated with hypercholesterolemia and
hyperlipemia and a tendency to severe premature
atherosclerosis (64, 68, 177). In contrast, bilateral
destruction of the adrenals in Addison's disease is
accompanied by low serum cholesterol levels (142).
Furthermore, the adrenal cortex has a high chol-
esterol content and it can synthesize and discharge
cholesterol and steroid hormones readily (147, 1 74)-
Adrenalectomized dogs maintained on desoxycorti-
costerone acetate (DCA) show a marked decrease of
serum cholesterol and phospholipid levels (31, 182).
When cortisone is substituted for DCA, a marked
rise in these lipids occurs. Combined DCA and corti-
sone administration showed no further increase over
cortisone alone. It was concluded from these studies
I 202 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
that the primary effect on circulating lipids is due to
cortisone (31, 182). It has to be borne in mind,
however, that data derived on adrenalectomized
animals are complicated by the fact that simul-
taneously with the depletion of cortical hormones
there is also a lack of medullary catecholamines, and
these, too, have an effect on lipid metabolism.
Aldosterone, the adrenal hormone affecting elec-
trolyte metabolism, apparentlv has no effect on the
circulating lipids, but may affect the vascular wall
according to some recent data (80).
The pituitarv adrenocorticotropic hormone
(ACTH) has an effect similar to, but less marked than,
cortisone. In man, the administration of either corti-
sone or ACTH produces an initial depression of
serum cholesterol levels with a subsequent rise on
continued administration (144). In the dog, the
response of the serum lipids to corticosteroids,
particularly' cortisone, is relatively mild. In the rabbit
and the rat, the effects of ACTH and corticosteroids
are qualitatively similar but much more pronounced.
The serum cholesterol elevation resulting from corti-
sone treatment is especially marked in the free
cholesterol fraction. Phospholipids are elevated
concomitantly, resulting in a normal C/P ratio,
despite elevated cholesterol levels. Triglycerides are
also increased, rendering the serum lactescent (7).
In the chick, the active steroid is 1 7-hydroxycorti-
costerone (compound F). It has lipid effects similar
to those described in the rabbit for cortisone.
Atherosclerosis has not been induced in dogs,
rabbits, or chicks by the administration of adrenal
cortical hormones despite the lipid changes they
produce. If these steroids are given in the presence of
an atherogenic diet, the effect on the circulating
lipids is variable depending upon the species, but
the effect on atherogenesis is similar — corticoids
depress cholesterol-induced atherogenesis.
It has been postulated that the atherosclerosis-
depressing action of these steroids is due to their
decreasing the permeability of the vascular endo-
thelium. Adlersberg's group has shown that when
hyaluronidase — a substance which increases cell
permeability — is given simultaneously with cortisone
the atherosclerosis-inhibiting action of the corticos-
teroids is overcome and atherosclerosis proceeds as
in the controls (166).
Some authors produced increased arteriosclerosis
and secondarily atherosclerosis by the administration
of ACTH in rats (171, 172) and dogs (100). It is
possible that these effects are due to the action of the
hormone on the vascular ground substance (muco-
polysaccharides) and fibroblasts (102).
No specific data are available implicating the
adrenal cortical hormones in blood coagulation or
clot lysis.
adrenal medullary hormones. /-Epinephrine is
the adrenal medullary hormone most extensively
studied. It influences lipid metabolism and produces
damage of the vascular wall in the form of medial
necrosis and calcification. The other catecholamines
probably act in a similar manner. Both of these
actions of /-epinephrine may produce arterio- and
atherosclerosis. In addition, /-epinephrine, being a
pressor agent, may further increase atherogenesis due
to the arterial hvpertension which ensues on chronic
endogenous overproduction or by protracted exog-
enous administration of the hormone (126).
In the older literature disparate data on the
circulating lipids after catecholamine administration
have been described. Some observers noted a transient
hyperlipemia, probably due to an action on mobiliza-
tion and transport (33, 45, 60, 71). Others observed a
decrease in serum cholesterol, phospholipid, and total
lipids (39, 71). Recently, Shafrir el a!. (143) clarified
some of these discrepancies. They showed that a single
subcutaneous injection of /-epinephrine in dogs
produces a prompt, transient elevation of serum-free
fatty acids and a delayed elevation of /i-lipoproteins.
Prolonged daily /-epinephrine administration, how-
ever, produced a marked increase in cholesterol
levels, with a smaller concomitant rise in phos-
pholipids. This epinephrine reaction was abolished by
adrenalectomy and restored by cortisone treatment.
anterior pituitary hormones. Pituitary growth
hormone (somatotropin) influences lipid mobilization
and transport as well as the distribution of lipid
between the liver and fat depots (85). Information
on the influence of this hormone on circulating
serum lipids is scant and the effects are variable in
different species. However, this may be due, in part
at least, to the fact that there are differences, both
physiological and chemical, in the nature of growth
hormone preparations obtained from different animal
species (19). Some stimulation of fibroblast growth
with this hormone has been described (102). No data
are available indicating any possible effect of this
hormone on atherogenesis.
Recently, Rudman et al. (134) demonstrated the
existence of a separate and distinct hyperlipemia-
producing hormone of the anterior pituitary.
ENDOCRINES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
1203
conclusion. All these studies would suggest that the
adrenal and pituitary glands have a significant
influence on lipid metabolism, but the exact mecha-
nism of these effects is still poorly understood. They
also affect the metabolism of the vascular wall and
may, therefore, be intimately related to athero-
sclerosis. Whether they have a direct effect on blood
coagulation and clot lysis has not yet been explored.
Some effects attributable to nervous and emotional
factors may actually be related to the release of
adrenal hormones under these circumstances. Further
studies of these hormones should be conducted.
Sex Hormones
Numerous clinical and experimental studies
indicate a profound influence of male and female
gonadal hormones on lipid and lipoprotein metab-
olism and atherosclerosis. Also, it has been reported
that these hormones exert a marked influence on
ground substance and connective tissue elements as
well as a slight, less well understood, effect on the
clotting mechanism. The influences of female sex
hormones are more pronounced than those of the
male hormone, and the effects of the male and female
hormones are in general opposite and antagonistic.
The levels of circulating lipids and lipoproteins in
normal males and females of all ages have been
extensively studied (6, 96, 135). Up to the age of 20
years, total cholesterol and phospholipid levels are
similar in the two sexes. Both these lipid fractions rise
significantly in men up to the age of 33, and then
remain stable up to age 60. In women, on the con-
trary, they stay constant up to the age of 32, and
from then on a steady rise occurs until 58 years of
age. The 1 :3 ratio of free to esterified cholesterol is
fairly constant in both sexes at all ages.
Serum lipoprotein patterns show a distinctly sex-
linked difference, as does the cholesterol content in
the different lipoprotein fractions. Young women
have more a-lipoproteins and a-lipoprotein cho-
lesterol than men of all ages, postmenopausal women,
or castrated women. Oliver & Boyd (no, 111)
studied lipids and lipoproteins during the menstrual
cycle and during pregnancy and found a depression of
^-lipoproteins and of the cholesterol phospholipid
(C/P) ratio coincident with the peak of estrogen
secretion at ovulation. During the third trimester of
pregnancy /3-lipoproteins and the C P ratio increase
despite large estrogen secretion. These contradictory
findings need clarification.
Sex differences in serum cholesterol levels are also
observed in animals. Female rats have higher serum
cholesterol levels than males (22). Egg-laying hens and
pigeons have elevated serum cholesterol levels with
very high phospholipid levels and, therefore, signifi-
cantly depressed C P ratios.
The normal lipoprotein pattern of the chick differs
from that of man and most mammals in that the main
component is a-lipoprotein (151). Furthermore,
giving estrogens to the cockerel elevates ^-lipoprotein,
instead of a-lipoprotein as in man (151). Incidentally,
the effect of diet on lipoprotein levels in chicks is also
opposite to that seen in man.
In man, androgens increase /^-lipoproteins and
serum cholesterol. Eunuchs have lower cholesterol
and /^-lipoprotein le\-els than normal men (52).
Estrogen administration to men or postmenopausal
women changes the serum lipoprotein pattern to the
young female type and this pattern remains as long
as therapy is continued, even over several years (151,
155). The effect on serum cholesterol is not so uni-
form. Several authors described a fall (7, 110), while
others found no change (151, 155). However, there is
uniform agreement that the phospholipid level
rises and therefore that the C/P ratio falls.
Androgens even in small doses, given to men
concomitantly with estrogens, counteract the estro-
genic serum lipid effect without counteracting the
feminizing effect on the secondary sex characteristics.
The latter action represents one instance where the
action of the two hormones is not antagonistic, at
least in man.
The mechanism by which the gonadal hormones
influence lipid metabolism is not yet entirely clarified.
Boyd (22) carried out tracer studies with C14-labeled
acetate in rats and found that estrogens slightly
depress plasma cholesterol synthesis and significantly
reduce the biological half-life of cholesterol. Ovariec-
tomy in female rats had the opposite effect (22, 44).
Furman et al. (53) have shown an interrelationship
of methyltestosterone and dietary protein intake on
serum lipoproteins in men. On a low-protein or
protein-free formula diet both a- and ^-lipoproteins
were significantly depressed, beyond the depression of
the protein-free diet alone. These findings have been
confirmed by Olson & Vester (1 15).
A very definite action of the sex hormones, particu-
larly the estrogens, on atherosclerosis has also been
established. Data from clinical medicine are sugges-
tive, experimental data on animals are indicative.
Premenopausal women have less gross coronary
atherosclerosis than men or castrated women (127,
181) and a markedly lower incidence of myocardial
[204
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
infarction (113, 130). After the menopause the
incidence of myocardial infarction in women slowly
rises to become almost equal to that in men by the
eighth decade (110, 113). Furthermore, Marmorston
et al. (99) have shown that postmenopausal women
with coronary artery disease have lower urinary
estrogen levels than healthy women of the same age,
and also have lower levels of protein-bound iodine
indicating a decreased thyroid function. Bersohn &
Oelofse (14) made similar observations in man. Aortic
atherosclerosis shows no significant sex difference (43,
128, 176). The protection of the female from coronary
atherosclerosis is lost in the presence of familial
hypercholesterolemia and in diabetes mellitus.
Cockerels on a high-fat, high-cholesterol (athero-
genic) diet are protected against coronary athero-
sclerosis when given estrogens (1 19). Also, previously
induced atherosclerosis can be completely reversed by
the hormone (120). Aortic atherosclerosis is not
influenced. Sexually mature, estrogen-secreting hens
fed the atherogenic diet develop aortic atherosclerosis
but no coronary atherosclerosis (150). Castration of
these hens makes them susceptible to coronary lesions
(121). Estrogens given to chicks on a normal non-
estrogenic diet induce aortic atherosclerosis, but not
coronary atherosclerosis (27, 67, 86). These different
effects of estrogens on coronary and aortic athero-
sclerosis are a good example of the previously stated
observation that local anatomic or metabolic factors
are of importance in atherogenesis. This makes it
imperative for the investigator to study the several
vascular beds separately, and not to draw the con-
clusion that observations made in one vascular bed
necessarily apply to other parts of the arterial tree.
Freedom from coronary lesions in chickens is
accompanied by the previously described char-
acteristic serum lipid changes resulting in a normal
C V ratio in the presence of hypercholesterolemia.
The same effect on lesions and lipids was obtained
in rats (108). In male rabbits neither serum lipid
changes nor coronary protection can be achieved by
estrogen administration (151). Ludden et al. (91)
observed that both androgens and estrogens protect
intact female rabbits from cholesterol-induced aortic
atheroscerosis. Neither hormone was effective in
males or in castrated females.
Another exception to this sex phenomenon is
atherosclerosis in a susceptible strain of pigeons
(vide infra). Old, egg-laying pigeons show coronary
atherosclerosis despite the usual low C/P ratio
characteristic for female birds (66, 88-90).
Androgens in large doses depress diet-induced
hypercholesterolemia without influencing athero-
sclerosis (123).
Studies in chicks revealed that estrogen protection
is preserved even if estrogen administration is com-
bined with androgen administration or with ad-
ministration of DCA or compound F, or is used after
pancreatectomy (149, 154)- The only clear-cut
reversal of the estrogen effect was obtained when
chicks were made hypothyroid by the administration
of thiouracil (122). A slight decrease of estrogen
reversal of previously induced lesions was observed
when insulin was administered concomitantly with
estrogens during the period when the lesions were
regressing (154). Recently,3 we have noted that
blocking the reticulo-endothelial system also pre-
vented the estrogen effect.
Estrogens have also been shown to stimulate
growth of ground substance, particularly collagen and
fibroblasts. They also stimulate the reticuloendothelial
system (18, 25). Furthermore, there is some indication
that estrogens influence fibrin content and fibrinolytic
activity of the blood (11, 56, 57). It has also been
reported that intravenous injection of estrogens,
particularly conjugated equine estrogens, has a
hemostatic effect (58).
An indication that the local influence of estrogens
on the vascular wall may be related to athero-
sclerosis was recently obtained in chicks. It was shown
that atherosclerotic abdominal aorta and coronary
lesions, produced by a high-fat, high-cholesterol, low-
protein diet, can ulcerate if large doses of estrogens are
given (76). In the chick this was shown to occur as a
stage in the healing process of these lesions. This is
the first suggestion that estrogens may also influence
the vascular wall of the aorta.
The action of estrogens on lipid metabolism and
atherogenesis stimulated several long-term research
projects in man using different female sex-hormone
preparations in the therapy of patients with proven
ischemic heart disease (101, 112, 131, 151, 155). The
results show a possible life-prolonging action only
when a natural estrogen preparation (conjugated
equine estrogens) is being used. It is not known wh\
this difference exists between natural and synthetic
compounds. The therapeutic value of this regimen is,
however, limited by the accompanying feminizing
action of the hormone. Several nonfeminizing estrogen
3 Pick, R., L. N. Katz, P. J. Johnson, and D. E. Century. The
role of the reticulo-endothelial system and estrogens on coro-
nary atherogenesis in cholesterol-fed cockerels. Circulation. In
press. I October 1962.)
ENDOCRINES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
I 21)-
derivatives are being explored, so far without worth-
while results.
From all these data it is evident that gonadal
hormones have a significant influence upon lipid
metabolism and atherogenesis.
HEREDITY
That genetic or hereditary factors may influence
the development of atherosclerosis is suggested by
animal experimental and human studies. It has to be
emphasized, however, that these tendencies become
evident only in the presence of a potentially athero-
genic diet. If the environment is favorable, genetic
tendencies may not become evident. As in many
other diseases, the interplay between host and en-
vironment is of the utmost importance.
Animal experiments have indicated species differ-
ences in the susceptibility of atherosclerosis. Man and
several species of birds develop atherosclerosis spon-
taneously (51). Also old dogs, kept as pets, have been
found to exhibit aortic and coronary atherosclerosis
(87). Most animals, however, living in their natural
environment do not exhibit vascular lesions, with the
possible exception of the baboon (55, 93) which
shows fatty streaks in the aorta.
Species differences are also found in the response to
high-level cholesterol-fat feeding. Chicks and rabbits
respond to this regimen with severe hypercholestero-
lemia and atherosclerosis in a short period of time
(10, 72). It is more difficult to produce similar effects
in ducks, guinea pigs, and hamsters (8, 9, 59). In the
dog, cholesterol feeding has to be combined with
supression of the thyroid activity in order to produce
both lipid and vascular changes. In the monkey,
cholesterol feeding has to be combined with a de-
ficiency in sulfur-containing amino acids in the diet to
produce lesions (98). Recently, however, athero-
sclerosis was induced in rhesus monkeys by a high-
saturated fat, high-cholesterol, nondeficient diet alone
(29). In the rat, the species most resistant to the
induction of atherosclerosis, this disease has been
produced by a combination of multiple dietary and
hormonal manipulations, i.e., cholesterol, cholic acid,
and saturated fat in the diet, plus hypothyroidism and
unilateral nephrectomy (63).
The cause of these species differences has not been
entirely clarified. Recent tracer studies, however,
indicate species differences in cholesterol synthesis,
turnover, and degradation rates, and in the handling
of dietary cholesterol (61, 62). Other studies indicate
differences in the number of vasa vasorum in the
aorta — richest in resistant species and poorest in the
very susceptible (139). Whether or not these species
differences are the actual cause of the varied suscepti-
bility to hypercholesterolemia and atherosclerosis is
not known. Nor is it known how they are inherited.
More significant perhaps than species differences,
are strain differences which occur within a single
species. These have been demonstrated in rabbits
(145). In chicks they have been described by Opdyke
& Ott (116) and others (46). They have been indi-
cated in dogs (100). The most recent and thorough
investigations into strain differences was carried out
by Lofland & Clarkson (88, 89, 90) who have studied
several breeds of pigeons, in particular: the White
Carneau, the White Racer, and the Autosexing King.
The first shows severe aortic and some coronary
atherosclerosis in old birds of both sexes kept on a
low-cholesterol, low-fat commercial diet. The second
strain does not show any lesions on the same diet.
The third, genetically a cross-breeding between the
first two, has intermediate incidence and severity of
lesions. The onset of atherosclerosis in response to
high cholesterol feeding of the three breeds parallels
the severity of the spontaneous lesions. All three
breeds spontaneously have high serum cholesterol
levels, around 400 mg per cent, with marked seasonal
variations (66), and resemble one another in many of
the biochemical aspects studied by these authors.
Wherein lies the definitely genetic difference in the
production of lesions is unexplained.
In man, it has often been suggested that genetic
and hereditary factors may play a role in lipid
metabolism and coronary atherosclerosis. The first
indication of such a relationship was described in
1930 (65). A vast literature on statistical and genetic
investigations has since appeared proving the familial
occurrence of a xanthomatous tendency, i.e., a
tendencv for hypercholesterolemia and atherosclerosis
to appear in families. This was reviewed recently by
McKusick (94). The most extensive clinical studies
were carried out by Adlersberg et al. (2-4) and by
Thomas & Cohen (161). Some limited studies are
available on the incidence of these disorders in
identical and fraternal twins, living together or
separately, which may aid in determining the respec-
tive roles of heredity and the environment (118).
Adlersberg and others (3, 17, 40, 137, 161) con-
sider hypercholesterolemia an inborn error of metab-
olism, probably inherited as an "incomplete"
dominant trait. Familial hypercholesterolemic xantho-
matosis, a disorder of lipid metabolism characterized
I 2( 'I '
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
by the triad: hypercholesterolemia, cutaneous or
tendon xanthomata, and severe premature athero-
sclerosis (sometimes occurring even in childhood),
is the most severe stage of this inherited disorder. It
is probably homozygotic. A hypercholesterolemic
tendency without xanthomata also occurs; this
milder form is heterozygotic (3). Recently, Epstein
rt a/. (41 ), restudying the families originally published
l>\ Adlersberg, re-emphasized the interplay between
genetic tendency and the environment. C. B. Thomas,
in her study of the families of healthy medical stu-
dents, showed a definite trend for the offspring of
parents with hypertension and or coronary artery
disease to have more hypertension and coronary
disease than children of parents not so afflicted. If one
parent had either of these diseases the occurrence
among the offspring was intermediate. Also, she
reported a fourfold greater frequency of occurrence
of coronary artery disease among the siblings of the
afflicted parents than among siblings of parents not
so afflicted. She concluded that the gradation of the
disorder rates were consistent with the Mendelian
law of inheritance. However, she could not exclude a
multiplicity of genetic factors and associated modify-
ing environmental agents.
Whether genetic and hereditary factors influence
atherosclerosis also by determining the anatomical
pattern of the circulatory tree, particularly that of
the coronary circulation, is difficult to evaluate — but
the possibility does exist (94).
In addition, the studies of Gertler & White (54)
on young coronary patients have yielded information
regarding body build. Even though no particular
'"coronar\- habitus" could be established, young
patients with coronary disease as a group belonged
predominantly to the mesomorph body build. Re-
cently, attention has also been focused on personality
and character traits, as well as the responsiveness of
the autonomic nervous system, partially genetically
determined, and their possible relationship to
coronary disease proneness. But these interrelation-
ships need further clarification (136).
No data are as yet available on the familial tend-
ency to accelerated blood clotting and thrombus
formation, other than the tendency of hyperlipemic
serum to shorten coagulation time. However, it is not
inconceivable that such genetic traits may be un-
c 1 1\ cred.
From the evidence presented it can be concluded
that genetic and hereditary traits may be an important
predisposing factor in an individual's response to
dietary and environmental factors leading to athero-
sclerosis, anatomically and clinically. But this is still,
for practical purposes, an uncultivated field of
systematic study of great importance.
STRESS
In recent years, interest has grown concerning the
possible influence of physical activity and psycho-
logical or emotional stress on the development of
atherosclerosis. It has also been suggested that both
of these types of "stress" may be involved in precipitat-
ing thrombosis or sudden occlusion of a blood vessel
in which pre-existing but clinically occult disease is
present. Further, such stress may act as the trigger
mechanism in aggravating the ischemia of an organ,
particularly of the heart and the brain, which already
has a deficient blood supply because of an athero-
sclerotic process. The presence of atherosclerotic
disease per se limits the ability of the circulation of an
organ to adjust to augmented demands placed upon it.
As has been pointed out for other facets of the
problem in previous sections of this chapter, "stress,"
too, exerts its role only in the presence of a life-span
pattern of diet high in cholesterol and fat, particularly
saturated fat. When this potentially atherogenic diet
is absent, differences attributable to stress and other
factors fail to appear. Therefore, differences in the
incidence of clinical coronary disease according to
occupation for instance, are found only in those
populations in which the over-all incidence of this
disease is high, presumably because of the dietary
factor.
Physical Activity
Results from the animal laboratory with regard to
the influence of enforced physical activity in the
presence of an atherogenic diet are contradictors.
Brown et a/. (24) found no differences in rabbits.
Kobernick & Niwayama (81 ), working with chol-
esterol-fed rabbits which were forced to exercise
adequately by combining a mechanical treadmill with
conditioning to electric shocks, found significantly less
atherosclerosis in the exercised rabbits as compared to
the sedentary controls — although the degree of
hypercholesterolemia was similar in both groups.
Brainard (23), working with rabbits exercised on a
treadmill, found no differences in the amount of
aortic cholesterol between the active and the seden-
tary group. Myasnikov (109) obtained positive
results in the rabbit in favor of a protection of the
ENDOCRINES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
exercised group. However, he also found an increased
incidence of myocardial infarction in the exercised
group despite the decrease in gross aortic and coronary
atherosclerosis. Serum cholesterol levels were found
to be significantly lower in rats forced to swim than
in sedentary controls and in pair-gained sedentary
controls (69). Orma (1 17), Warnock et al. (168), and
Wong et al. (180) reported that exercise decreased
hypercholesterolemia and atherogenesis in cholesterol-
fed cockerels. McAllister et al. (92), on the contrary,
found more severe atherosclerosis in exercised,
cholesterol-fed, hypothyroid dogs as compared to
sedentary ones. Their findings are complicated by
the fact that the exercised dogs were ingesting their
rations as meals while the sedentary hypothyroid
animals, with the usual depression of appetite, ate
their food slowly over the entire 24-hour period.
Such differences in feeding pattern in chicks have
been shown to influence the atherogenicity of the
diet per se (28).
Data in man relating physical activity to anatomical
atherosclerosis or clinical coronary disease are even
more difficult to evaluate. Several investigators (78,
97, 160) observed that increasing the caloric intake
did not produce the expected increase in serum
lipoprotein and cholesterol levels when the subjects
were exercised intensely enough to prevent weight
gain. They concluded that only a positive caloric
balance over a long-time period could elevate serum
lipid levels.
Pomeroy & White (125) reviewed the life history
of former football players and found fewer deaths
from cardiovascular disease among those who con-
tinued a program of regular exercise into the middle
years than among those who stopped physical activity
after their school years.
The most indicative data relating the amount of
physical activity at work with a decreased incidence
of death from atherosclerotic vascular disease,
particularly ischemic heart disease, come from the
studies of Morris in Great Britain (104, 106). His
data were obtained from a relatively homogenous
population of a similar socio-economic group : by a
comparison of sedentary bus drivers with physically
active conductors, by a comparison of sedentary
telephone operators with active postmen, and by
other comparisons of a similar nature. His findings
indicate that the incidence of ischemic heart disease
in middle age tends to be lower in the groups habit-
ually engaged in a greater amount of physical
activity. These investigations, although indicative,
are not to be taken as final proof, because, in the bus
workers at least, there was a difference in obesity —
the drivers were more obese from the start than the
conductors, as judged by the size of the uniforms
(103). This leaves open the question of whether the
difference between jobs was fortuitous, dependent
upon self-selection, which in turn was dependent on
temperament and body build of the individual
worker. Studies from other countries, i.e., Sweden,
Finland, and Italy (74), with a generally high
morbidity and mortality rate from atherosclerotic
heart disease, are not so clear cut as the British
studies. Studies from the United States show no
difference between active and sedentary groups in an
urban population (153, 157); however, farmers have
less atherosclerotic heart disease than city dwellers.
Some authors suggest that continued physical activity
through middle age may be of possible benefit in the
prevention of atherosclerotic disease ( 1 79).
One fact clearly emerges from these studies: that
no difference between physically active and inactive
groups can be observed in populations with a low
incidence of atherosclerotic heart disease and low
mean serum lipid levels. In populations with a high
incidence, however, there is a difference in some
but not in all countries. Furthermore, even where a
difference has been well documented, as in Great
Britain, this is only relative; the absolute incidence of
this disease in the physically active is still high com-
pared to all groups in a country with a low incidence.
Therefore, physical activity must play a minor role
compared to other factors such as diet.
The mechanism by which physical activity might
influence atherosclerosis is not clear. The data
regarding serum cholesterol and lipoprotein levels
suggest an influence via metabolism. Other data also
indicate that the factors preventing blood coagulation
and aiding fibrinolysis are favorably influenced by
heavy physical work (11, 12, 77). This was pointed
out in human studies, and Warnock et al. (168) report
the same effect in chicks. These latter effects may be
important, particularly since Morris' work points to
a decrease of coronary thrombosis and major occlusion
in active middle-aged men, without any noticeable
decrease in vascular atheroma and diffuse, nonfatal
myocardial fibrosis (104).
Furthermore, physical work may have another
effect. There are several studies indicating a stimula-
tion of the production of intercoronary anastomoses by
physical work (13, 105, 184). Nor must it be over-
looked that physical activity is a form of training
which permits the body to adjust more readily to
periods of stress.
I208
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
It is apparent that there is room for further studies
of this important aspect of atherosclerosis.
Emotional Stress
A number of investigators have shown that emo-
tionally stressful life situations transiently elevate
serum cholesterol levels and shorten the blood-
clotting time. This has been noted in medical students
at the time of examination (36, 162), and Friedman
et al. (48) observed it in accountants when they were
under professional peak loads. Several other workers
have published data linking the acute episode of
coronary occlusion to immediately preceding stressful
life situations (37, 136, 169). The proponents of the
hypothesis that emotional stress influences athero-
genesis and may precipitate clinical episodes of
occlusion implicate the stresses of our modern
mechanized civilization in particular. Emotional
stress also produces elevation of blood pressure, and
this in turn may have a deleterious effect on the
vascular wall. In this and other ways hypertension
favors atheroma formation.
The study of emotional factors in relation to
cardiovascular disease is in its infancy. The main
reason for the difficulties in the evaluation of this
factor is the lack of an effective measure of emotional
stress and of the various personality profiles (75). The
mechanism by which psychological stress influences
body homeostasis is also difficult to assess. It may
operate: a) by disturbing endocrine balance, e.g., via
pituitary-adrenal stimulation and catecholamine
release (50, 95, 132, 140, 162), thereby influencing
blood pressure, cholesterol metabolism, coagulation,
and fibrinolytic activity; or b) by other, as yet un-
known, mechanisms including an action via the
nervous system. Numerous acute psychological
episodes of this type over the life span may lead to the
establishment of chronic changes, such as those which
may operate in hypertension.
Experience from the animal laboratory is equally
fragmentary. Cold stress was shown to produce
coronary artery changes in rats (141). Friedman &
Uhley (49) have shown that rats kept tense in antic-
ipation of electric shocks showed significantly
shortened blood-clotting time. They found no
difference in coronary atherosclerosis between aggres-
sive and passive chicks on an atherogenic diet. Other
results confirm this observation. Recent studies in
our laboratorv indicate that isolation of chicks in a
quiet, undisturbed room increases the atherogenic
response to a high-fat, high-cholesterol diet (124).
Also, isolation during the healing phase of athero-
sclerosis prevents regression of lesions. Such isolated
cockerels, in addition, showed retarded sexual growth
as evidenced by decreased testis weight and comb
size, as well as a decreased food efficiency expressed by
smaller weight gain on a food intake similar to that of
the controls.
It cannot yet be determined from these results
whether the observed effects of isolation of the birds
constituted a response to a severe stress, specifically
the unnatural isolation, or were, on the contrary, the
result of the lack of normal stresses. However, all
these data indicate that certain environmental
influences, mediated via the central nervous system
and involving the nervous and hormonal regulation
of body functions, including that of the vessels them-
selves, can influence the vascular response to a
potentially atherogenic diet. These findings may have
far reaching implications, particularly should "lack
of normal stress" be the underlying cause.
The evidence relating emotional or psychological
stress to atherosclerosis, clinical or experimental, is
at best fragmentary. It is much too early to extrap-
olate the findings and to make any major generaliza-
tion. Many further well-controlled and systematic
studies are required to help our understanding of the
complicated mechanisms which operate in this elusive
area.
SUMMARY
From the data presented in this chapter the
following conclusions can be drawn:
Whether or not atherosclerosis emerges as a major
health problem in a population is largely and mainly
determined by the life-span pattern of the diet. In
the case of any individual member of a population
group which is habitually ingesting a potentially
atherogenic diet, several other factors will determine
the degree and extent of atherosclerosis and clinical
atherosclerotic disease — this is not always mirrored
by the blood cholesterol or other lipid levels. There is
a complex interplay between diet and these other
factors which operate to accelerate or retard athero-
genesis. The most important of these accessory factors
determining the individual's fate with regard to
atherosclerosis are hormones, heredity, and stress.
It is hopefully felt that by taking all these factors
into account it will become possible to single out
persons particularly prone to develop atherosclerotic
disease at a relatively early age and to suggest dietary
ENDOCRINES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
[20g
or other measures for them, with the expectation that
the development of atherosclerosis can be retarded,
its sequelae avoided or delayed, and a clinical
catastrophe prevented or put off for a variable time.
Much work is still needed to reach this goal and many
aspects of the problem are poorly understood at the
present time. But progress is being made year by
year in the multidisciplinary attack aiming to under-
stand the pathogenesis of atherosclerosis, and seeking
to further its prevention and to improve its manage-
ment once it develops.
The work of this department, which is included in this review,
was accomplished over the years in collaboration with the fol-
lowing former associates: Drs. D. V. Dauber (deceased), L.
Horlick, S. Rodbard, and, especially, J. Stamler.
Several dedicated research assistants made these complicated
investigations possible. They were: Mrs. C. Bolene-VVilliams,
Mrs. D. Century, and Mr. P. Johnson. Also, the cooperation
of many research technicians is gratefully acknowledged.
REFERENCES
1. Abbuhl, R., C. B. Taylor, D. Patton, and G. Cox.
Comparative quantitation of the sources of plasma cho-
lesterol in dog and man. Circulation 20: 966, 1959. 18.
2. Adlersberg, D., A. D. Parets, and E. P. Boas. Genetics
of atherosclerosis. J. Am. Med. Assoc. 141 : 246, 1949.
3. Adlersberg, D. Hypercholesterolemia with predisposition 19.
to atherosclerosis: An inborn error of lipid metabolism.
Am. J. Med. 1 1 : 600, 1951.
4. Adlersberg, D. Inborn errors of lipid metabolism: 20.
clinical, genetic and chemical aspects. A.M. A. Arch.
Pathol. 60:481, 1955. SI.
5. Adlersberg, D., C. I. Wang, H. Rifkin, J. Brekman, G.
Ross, and C. Weinstein. Serum lipids and polysaccharides
in diabetes mellitus. Diabetes 5 : 1 1 6, 1956. 22.
6. Adlersberg, D., L. F. Schaeper, and A. Steinberg. Age,
sex, serum lipids and coronary atherosclerosis. J. Am. 23.
Med. Assoc. 162:619, '95^-
7. Adlersberg, D. Hormonal influences on the serum lipids.
Am. J. Med. 23: 769, 1957. 24.
8. Altschul, R. Experimental cholesterol arteriosclerosis.
II. Changes produced in golden hamsters and in guinea
pigs. Am. Heart J. 40: 401, 1950. 25.
9. Altschul, R. Selected Studies on Arteriosclerosis. Springfield,
111. : Thomas, 1950. 26.
10. Anitschkow, N. Experimental arteriosclerosis in animals.
In: E. V. Cowdry, Arteriosclerosis. New York: Macmillan,
■933. P-27'- 27-
1 1 . Astrup, T. Role of blood coagulation and fibrinolysis in
the pathogenesis of arteriosclerosis. In: I. Page, Connective
Tissue, Thrombosis and Atherosclerosis. New York: Acad.
Press, 1959, p. 223. 28.
12. Astrup, T. The biological significance of fibrinolysis.
Lancet 2: 565, 1956.
13. Baroldi, G, O. Mantero, and G. Scomazzoni. The col- 29.
laterals of the coronary arteries in normal and pathologic
hearts. Circulation Research 4: 223, 1956.
14. Bersohn, I., and P. J. Oelofse. Urinary oestrogen levels
in myocardial infarction. S. African Med. J. 32: 979, 30.
I958.
15. Binet, L., and P. Brocq. Le lactesence du serum sanguin
au course de la pancreatite hemorrhagique (etude experi- 31.
mentale). Paris med. 1 : 489, 1929.
16. Bloor, W. R. The lipoids ("fat") of the blood in diabetes.
J. Biol. Chern. 26: 417, 191 6.
17. Boas, E. P., A. D. Parets, and D. Adlersberg. Heredi- 32.
tary disturbance of cholesterol metabolism: A factor in the
genesis of arteriosclerosis. Am. Heart J. 35: 61 1, 1948.
Boucek, R. J., N. L. Noble, and J. F. Woessner. Proper-
ties of fibroblasts. In : I. Page, Connective Tissue, Thrombosis
and Atherosclerosis. New York : Acad. Press, 1959, p. 193.
Boyd, G. S., and M. F. Oliver. The physiology of the
circulating cholesterol and lipoproteins. In: R. P. Cook,
Cholesterol. New York: Acad. Press, 1958, p. 181.
Boyd, G S., and M. F. Oliver. Hormonal control of the
circulating lipids. Brit. Med. Bull. 14: 239, 1958.
Boyd, G. S., and M. F. Oliver. The effect of certain
thyroxine analogues on the serum lipids in human sub-
jects. J. Endocrinol. 21 : 33, 1960.
Boyd, G. S. Endocrines in lipid metabolism. Federation
Proc. 20: Part 3, 152, 1961.
Brainard, J. B. Effect of prolonged exercise on athero-
genesis in the rabbit. Proc. Soc. Exptl. Biol. Med. 1 00 : 244,
1959-
Brown, C. E., T. C. Huang, E. L. Bortz, and C. M.
McCay. Observations on blood vessels and exercise. J.
Gerontol. 1 1 : 292, 1956.
Burrows, H. Biological Actions of Sex Hormones. London :
Cambridge Univ. Press, 1949, pp. 454, 466.
Caren, R., and L. Carbo. Pancreatic alpha-cell function
in relation to cholesterol metabolism. J. Clin. Endocrinol.
l6:5°7. '956-
Chaikoff, I. L., S. Lindsay, F. VV. Lorenz, and C.
Entenman. Production of atheromatosis in the aorta of the
bird by the administration of diethylstilbesterol. J. Exptl.
Med. 88:373, 1948.
Cohn, C, R. Pick, and L. N. Katz. Effect of meal eating
compared to nibbling upon atherosclerosis in chickens.
Circulation Research 9: 139, 1961.
Cox, G. E., C. B. Taylor, L. G. Cox, and M. A. Counts.
Atherosclerosis in rhesus monkeys. I. Hypercholesterolemia
induced by dietary fat and cholesterol. A.M. A. Arch.
Pathol. 66: 32, 1958.
Dawber, T. R., F. E. Moore, and G. V. Mann. Coro-
nary heart disease in the Framingham study. Am. J.
Public Health 47, Part 2 : 4, 1957.
DiLuzio, N. R., M. L. Shore, and D. B. Zilversmit.
Effect of cortisone and desoxycorticosterone acetate on
plasma lipids of adrenalectomized dogs. Metabolism 3 : 424,
■954-
Ditzel, J., P. White, and J. Duckers. Changes in the
HANDBOOK OF PHYSIOLOGY -" CIRCULATION II
pattern of the smaller blood vessels in the bulbar con-
junctiva in children of diabetic mothers. A preliminary re-
port. Diabetes 3 : 99, 1 954.
33. Dole, V. P. Relation between non-esterihed fatty acids in
plasma and the metabolism of glucose. J. Clin. Invest. 35 :
■5°. I95b-
34. Dragstedt, L. R. The role of the pancreas in arterioscle-
rosis. Biol. Symposia 11:118, 1945.
35. Dragstedt, L. R., J. S. Clarke, G. R. Hlavacek, and
P. V. Harper, Jr. Relation of the pancreas to the regu-
lation of blood lipids. Am. J. Physiol. 179: 439, 1954.
36. Drevfuss, F., and J. W. Czaczkes. Blood cholesterol and
uric acid of healthy medical students under the stress of
an examination. A. MA. Arch. Internal Med. 103: 708,
'959-
37. Drevfuss, F. Role of emotional stress preceding coronary
occlusion. Am. J. Cardiol. 3: 590, 1959.
38. Duff, G. L., D. J. H. Brechim, and W. E. Finkelstein.
Effect of alloxan diabetes on experimental cholesterol
atherosclerosis in the rabbit. IV. Effect of insulin therapy
on inhibition of atherosclerosis in the alloxan-diabetic rab-
bit. J. Exptl. Med. 100:371, 1954.
39. Durv, A. Effects of epinephrine on lipid partition and
metabolism in the rabbit. Cireulation Research 5: 47, 1957.
40. Epstein, F. H., E. P. Boas, and R. Simpson. The epidemi-
ology of atherosclerosis among a random sample of clothing
workers of different ethnic origins in New York City.
I. Prevalence of atherosclerosis and some associated charac-
teristics. II. Associations between manifest atherosclerosis,
serum lipid levels, blood pressure, overweight and some
other variables. J. Chronic Diseases 5: 300, 329, 1957.
41. Epstein, F. H., W. D. Block, E. A. Hand, and T.
Francis, Jr. Familial hypercholesterolemia, xanthomatosis
and coronary heart disease. Am. J. Med. 26: 39, 1959.
42. Epstein, F. H. Epidemiology of coronary heart disease.
In: A. M. Jones, Modern Trends in Cardiology. New York:
Hoeber-Harper, i960, p. 155.
43. Faber, M., and F. Lund. The human aorta. Influence of
obesity on the development of arteriosclerosis in the
human aorta. A.M.A. Arch. Pathol. 48: 351, 1949.
44. Fillios, L. C, R. Kaplan, R. S. Martin, and F. J.
Stare. Some aspects of the gonadal regulation of cho-
lesterol metabolism. .4m. J. Physiol. 193: 47, 1958.
45. Frederickson, D. S., and R. S. Gordon, Jr. Transport of
fatty acids. Physiol. Revs. 38: 585, 1958.
46. Friedman, D., P. Johnson, R. Pick, J. Stamler, and L.
N. Katz. Aorta atherogenesis in different strains of
hybrid cockerels. Circulation 14:498, 1956.
47. Friedman, M., and S. Byers. Role of hyperlipemia in the
genesis of hypercholesterolemia. Proc. Soc. Exptl. Biol. Med.
90:496, 1955.
48. Friedman, M., R. H. Rosenman, and V. Carroll.
Changes in the serum cholesterol and blood clotting time
in men subjected to cyclic variation of occupational stress.
Circulation 17: 852, 1958.
49. Friedman, M., and 11. L'iilev. Experimental stress, blood
lipids and atherosclerosis. In: G. Pincus, Hormones and
Atherosclerosis. New York: Acad. Press, 1959, p. 205.
50. Friedman, M., and R. H. Rosenman. Association of
specific overt behavior pattern with blood and cardio-
vascular findings: Blood cholesterol level, blood clotting
time, incidence of arcus senilis and clinical coronary
artery disease. J. Am. Med. Assoc. 169: 1286, 1959.
51. Fox, H. Arteriosclerosis in lower mammals and birds: Its
relation to the disease in man. In: E. V. Cowdry, Ar-
iel losclerosis. New York: Macmillan, 1Q33, p. 153.
52. Furman, R. H., and R. P. Howard. The influence of
gonadal hormones on serum lipids and lipoproteins :
studies in normal and hypogonadal subjects. Ann. Internal
Med. 47:969, 1957.
53. Furman, R. H , R. P. Howard, and L. N. Norcia.
Modification of the effects of adrenal cortical steroids and
androgens on serum lipids and lipoproteins by caloric
supplementation and by isocaloric substitution of carbo-
hydrate for dietary protein. In: G. Pincus, Hormones and
Atherosclerosis. New York : Acad. Press, 1949, p. 349.
54. Gertler, M. M., and P. D. White. Coronary Heart Disease
in Young Adults: A Multidisciplinary Study. Cambridge:
Harvard Univ. Press, 1954.
55. Gillman, J., and C. Gilbert. Atherosis in the Baboon
(Papio ursinus). Exptl. Med. Surg. 15: 181, 1957.
56. Gillman, T., and S. S. Naidoo. Gonadal influences on
plasma fibrin and librinolytic activity : A possible basis for
the further analysis of some forms of coronary thrombosis.
Endocrinology 62: 92, 1958.
57. Gillman, T., S. S. Naidoo, and M. Hathorn. Sex differ-
ences in plasma fibrin, fibrinolytic capacity and lipids as
influenced by ingested fat, gonadectomy and hormonal
implants. Clin. Sci. 17:393, 1958.
58. Gitman, L., and I. J. Greenblatt. Effect of intra-
venously administered estrogen in cardiovascular disease.
Angiology 4: 502, 1953.
59. Goldman, J., and O. J. Pollak. The hamster as experi-
mental animal for the study of atheromatosis. Am. Heart J.
38: 474, 1949.
60. Gordon, R. S., Jr., and A. Cherkes. Unesterified fatty
acids in human blood plasma. J. Clin. Invest. 35 : 206
■956-
61. Gould, R. G, and R. P. Cook. The metabolism of cho-
lesterol and other sterols in the animal organism. In: R. P.
Cook, Cholesteiol. New York: Acad. Press, 1958, p. 237.
62. Gould, R. G. The relationship between thyroid hormones
and cholesterol biosynthesis and turnover. In: G. Pincus.
Hormones and Atherosclerosis. New York : Acad. Press, 1 959,
P- 75-
63. Hartroft, W. S., and W. A. Thomas. Production of
coronary thromboses and myocardial infarcts in rats by
dietary means. Circulation 16: 481, 1 957-
64. Heinbecker, P., and M. Pfeiffenberger, Jr. Further
clinical and experimental studies on the pathogenesis of
Cushing's syndrome. Am. ./. Med. 9: 3, 1950.
65. Herapath, C. E. K, and C. B. Perry. The coronary
arteries in a case of familial liability to sudden death.
Brit. Med. ./. 1 : 685, 1930.
66. Hoffman, R. A. Observations in serum and gonad cho-
lesterol in pigeons. Endocrinology 67: 31 1, i960.
67. Horlick, L., and L. N. Katz. The effect of diethylstil-
besterol on blood lipids and the development of athero-
sclerosis in chickens on a normal or low fat diet. J. Lab.
Clin. Med. 33: 733, 1948.
68. Hueper, W. C. Arteriosclerosis. A.M.A. Arch. Pathol. 38:
162, 245, 350, 1944 and 39: 51, 117, 187, 1945.
69. Jones, E. M., P. B. Johnson, H. J. Montoye, and E. D.
Van Huss. Comparative effects of exercise and food re-
striction on bodv composition and blood serum cholesterol
ENDOCRINES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
rats. Federation Proc. 20, Part
207,
concentration
1961.
70. Jones, R. J., L. Cohen, and H. Corbus. The serum lipid
pattern in hyperthyroidism, hypothyroidism and coro-
nary atherosclerosis. Am. J. Med. 19: 71, 1955.
71. Kaplan, A., S. Jacques, and M. Gant. Effect of long-
lasting epinephrine on serum lipid levels. Am. J. Physiol.
'9' :8> '957-
72. Katz, L. N., and J. Stamler. Experimental Atherosclerosis.
Springfield, 111. : Thomas, 1953.
73. Katz, L. N, J. Stamler, and R. Pick. The role of the
hormones in atherosclerosis. Natl. Acad. Sci.—Natl. Re-
search Council Publ. No. 338, 1954, p. 236.
74. Katz, L. N., J. Stamler, and R. Pick. Nutrition and
Atherosclerosis. Philadelphia: Lea & Febiger, 1958.
75. Katz, L. N., J. Stamler, and R. Pick. Approaches to the
problem of the relation of emotions to hormonal function
and atherosclerosis. In: G. Pincus, Hormones and Atheroscle-
rosis. New York: Acad. Press, 1959, p. 377.
76. Katz, L. N., and R. Pick. Morphological aspects of
atherosclerosis in the chick. Conn. Slate Med. J. 25: 84,
1 96 1.
77. Keys, A., and R. Buzina. Blood coagulability. Effects of
meals and differences between populations. Circulation 14:
479- '956-
78. Kevs, A., J. T. Anderson, and O. Mickelsen. Serum
cholesterol in men in basal and nonbasal states deports
and letters). Science 123: 29, 1956.
79. Keys, A., and P. D. White. World trends in cardiology:
I. Cardiovascular epidemiology. Selected Papers from Second
World Congress and Twenty-Seventh Annual Scientific Sessions
of the American Heart Association. New York: Hoeber, 1956.
80. KlTTINGER, G. W., B. C. WEXLER, AND B. F. MlLLER.
Abnormal adrenal function in arteriosclerotic rats. Feder-
ation Proc. 19: 16, i960.
81. Kobernick, S. D., and G. Niwayama. Physical activity in
experimental cholesterol atherosclerosis of rabbits. Am. J.
Pathol. 36: 393, i960.
82. Lansing, A. I. Elastic tissue in atherosclerosis. In: I. H.
Page, Connective Tissue, Thrombosis and Atherosclerosis. New
York: Acad. Press, 1959, p. 167.
83. Laurie, \\ '., and J. D. Woods. Anastomosis in the coro-
nary circulation. Lancet 2: 812, 1958.
84. LeCompte, P. M. Vascular lesions in diabetes mellitus.
J. Chronic Diseases 2: 178, 1955.
85. Levin, L., and R. K. Farber. Hormonal factors which
regulate the mobilization of depot fat to the liver. Recent
Progr. Hormone Research 7: 399, 1952.
86 Lindsay, S., and I. L. Chaikoff. Coronary arterioscle-
rosis of birds. A.M. A. Arch. Pathol. 49: 434, 1950.
87. Lindsay, S., I. L. Chaikoff, and J. W. Gilmore. Arterio-
sclerosis in the dog. A.M. A. Arch. Pathol. 53: 281, 1952.
88. Lofland, H. B., T. B. Clarkson, R. W. Prichard, and
H. G. Netsky. Further studies on spontaneous athero-
sclerosis in pigeons. Circulation 20: 973, 1959.
89. Lofland, H. B., and T. B. Clarkson. A biochemical
study of spontaneous atherosclerosis in pigeons. Circulation
Research 7: 234, 1959.
90. Lofland, H. B., and T. B. Clarkson. Serum lipoproteins
in atherosclerosis susceptible and resistant pigeons. Proc.
Soc. Exptl. Biol. Med. 103: 238, i960.
91. Ludden, J. B., M. Bruger, and I. S. Wright. Experi-
mental atherosclerosis IV. Effect of testosterone
92.
93
94-
95-
96-
97-
99-
propionate and estradiol dipropionate on experimental
atherosclerosis in rabbits. A.M. A. Arch. Pathol. 33: 58,
'942-
McAllister, F. F , R. Bertsch, J. Jacobson, and G.
D'Alessio. Accelerating effect of muscular exercise on
experimental atherosclerosis. A.M. A. Arch. Surg. 80: 54,
i960.
McGill, H. C, Jr., J. P. Strong, R. L. Holman, and
N. T. Werthessen. Arterial lesions in the Kenya baboon.
Circulation Research 8: 670, i960.
McKusick, V. A. Genetic factors in cardiovascular dis-
eases: I. The four major types of cardiovascular disease.
II. Disorders of primarily genetic etiology. Modern Con-
cepts Cardiovascular Disease 28: 535, 547, 1959.
Macfarlane, R. G, and R. Biggs. Fibrinolysis: Its
mechanism and significance. Blood 3: 1167, 1948.
Man, E. B., and J. P. Peters. Variations of serum lipids
with age. J. Lab. Clin. Med. 41 : 738, 1953.
Mann, G. V, and H. S. White. The influence of stress on
plasma cholesterol levels. Metabolism 2: 47, 1953.
Mann, G. V., and S. B. Andrus. Xanthomatosis and
atherosclerosis produced by diet in an adult rhesus mon-
key. J. Lab. Clin. Med. 48: 533, 1956.
Marmorston, J., O. Hoffman, H. Sobel, and P. Starr.
Urinary estrogen and serum protein-bound iodine levels
in a group of post-menopausal women with and without
myocardial infarction. In: A. Keys, Arteriosclerosis. Minne-
apolis: Univ. Minnesota Press, 1955, p. 70.
100. Marmorston, J., S. Rosenfeld, and J. Mehl. Experi-
mental atherosclerosis in dogs. In: G. Pincus, Hormones
and Atherosclerosis. New York: Acad. Press, 1959, p. 213.
101. Marmorston, J., O. Magdison, O. Kuzma, and F. J.
Moore. Estrogen therapy in men with myocardial in-
farction. J. Am. Med. Assoc, 174: 241, i960.
102. Moon, H. D. Connective tissue reactions in the develop-
ment of arteriosclerosis. In: I. H. Page, Connective Tissue,
Thrombosis and Atherosclerosis. New York: Acad. Press
'959. P 33
103. Morris, J. N, J. A. Heady, and P. A. B. Raffle.
Physique of London busmen : Epidemiology of uniforms.
Lancet 2: 569, 1956.
104. Morris, J. N., and M. D. Crawford. Coronary heart
disease and physical activity of work. Brit. Med. J. 2 :
■485, i958-
105. Morris, J. N. Epidemiology and coronary heart disease.
Circulation 17:321. 1958.
106. Morris, J. N. Occupation and coronary heart disease.
A. ALA. Arch. Internal. Med. 104: 903, 1959.
107. Morris, J. N. Epidemiology and cardiovascular disease
of middle age. Modern Concepts Cardiovascular Disease 29:
625, i960 and 30: 633, 1 96 1.
108. Moskowitz, M. S., A. A. Moskowitz, W. L. Bradford,
Jr., and R. W. Wissler. Changes in the serum lipids
and coronary arteries of the rat in response to estrogens.
A.M. A. Arch. Pathol. 61: 245, 1956.
log. Myasnikov, A. L. Influence of some factors on develop-
ment of experimental cholesterol atherosclerosis. Circula-
tion 17: 99, 1958.
1 10. Oliver, M. F., and G. S. Boyd. Coronary atherogenesis —
an endocrine problem? In : A. Keys, Arteriosclerosis. Minne-
apolis: Univ. Minnesota Press, 1955, p. 64.
ill. Oliver, M. F., and G. S. Boyd. Plasma lipid and serum
HANDBOOK OI PHYSIOLOGY
CIRCULATION II
I'3
114.
"5
116.
117.
"9-
123.
124.
'25-
126.
127.
128.
129.
130.
lipoprotein patterns during pregnancy and puerperium.
Clin. Sri. 14: 15, 1955.
Oliver, M. F., and G. S. Boyd. The influence of the sex
hormones on the circulating lipids and lipoproteins in
coronary sclerosis. Circulation 13: 82, 1956.
Oliver, M. F., and G. S. Boyd. Effects of bilateral
ovariectomy on coronary artery disease and serum-lipid
levels. Lancet 2: 690, 1959.
Oliver, M I Metabolic factors in the aetiology of
coronary heart disease. In: A. M. Jones, Modem Fiends
m Cardiology. London: Butterworth, i960, p. 172.
Olson, R. E., and J. W. Vester. Nutrition-endocrine
interrelationships in the control of fat transport in man.
Physiol. Rets. 40: 677, i960.
Opdyke, D. F., and W. H. Ott. Influence of source of
cholesterol, grade of cottonseed oil, and breed on experi-
mental avian atherosclerosis. Proc. Soc. Exptl. Biol. Med.
85: 4'4> '954-
Orma, E. J. Effect of physical activity on atherogenesis;
An experimental study in cockerels. Acta Physiol. Scand. 41 :
Suppl. 142, 1, 1957.
Osborne, R. H., and D. Adlerseerg. Serum lipids in
adult twins. Science 127: 1294, 1958.
Pick, R., J. Stamler, S. Rodbard, and L. N. Katz.
The inhibition of coronary atherosclerosis by estrogens
in cholesterol-fed chicks. Circulation 6: 276, 1952.
Pick, R., J. Stamler, S. Rodbard, and L. N. Katz.
Estrogen-induced regression of coronary atherosclerosis
in cholesterol-fed chicks. Circulation 6, 868, 1952.
Pick, R., J. Stamler, and L. N. Katz. Susceptibility of
the ovariectomized hen to cholesterol-induced coronary
atherogenesis. Circulation Research 5: 515, 1957.
Pick, R., J. Stamler, and L. N. Katz. Effects of hypo-
thyroidism on estrogen-induced inhibition of coronary
atherogenesis in cholesterol-fed cockerels. Circulation
Research 5: 510, 1957.
Pick, R., J. Stamler, S. Rodbard, and L. N. Katz.
Effects of testosterone and castration on cholesterolemia
and atherogenesis in chicks on high-fat, high-cholesterol
diets. Circulation Research 7: 202, 1959.
Pick, R., and L. N. Katz. Social milieu and atherosclero-
sis in cockerels. Federation Proc. 20: Part 1, 93, 1961.
Pomeroy, W. C., and P. D. White. Coronary heart
disease in former football players. J. Am. Med. Assoc. 167:
711, 1958.
Raab, W. Neurohormonal atherogenesis. Am. J. Cardiol.
' : 113. '95°-
Rivin, A. U., and S. P. Dimitroff. The incidence and
severity of atherosclerosis in estrogen-treated males and
in females with a hypoestrogenic or hypercstrogenic state.
Circulation 9: 533, 1954.
Roberts, J. C, Jr., C. Moses, and R. H. Wilkins.
Autopsy studies in atherosclerosis: I. Distribution and
severity of atherosclerosis in patients dying without
morphologic evidence of atherosclerotic catastrophe. II.
Distribution and severity of atherosclerosis in patients
dying with morphologic evidence of atherosclerotic
catastrophe. Circulation 20: 511, 520, 1959.
Robertson, W. B. Atherosclerosis and ischaemic heart
disease. Lancet 1 : 444, 1959.
Robinson, R. W., N. Higano, and W. D. Cohen. In-
creased incidence of coronary heart disease in prematurely
castrated women. Circulation 18: 771, 1958.
131 . Robinson, R. \\\, W. D. Cohen, and N. Higano. Estrogen
replacement therapy in women with coronary athero-
sclerosis. Ann. Internal Med. 48: 95, 1958.
132. Rosenman, R. H., and M. Friedman. The possible rela-
tionship of the emotions to clinical coronary heart disease.
In: Ci. Pim us. Hormones and Atherosclerosis. New York:
Acad. Press, 1959, p. 283.
133. Rosenthal, S. R. Studies in atherosclerosis: chemical, ex-
perimental and morphologic. AM .A.Auh. Pathol. 18:473,
660, 1934.
134. Rudman, D., F. Seidman, and M. B. Reid. Lipemia pro-
ducing activity of pituitary gland: Separation of lipemia-
producing component from other pituitary hormones.
Proc. Soc. Exptl. Biol. Med. 103: 315, 1960.
135. Russ, E. M ., H. A Eder, and D. P. Barr. Protcin-lipid
relationships in human plasma. I. In normal individuals.
Am. J. Med. II: 468, 195 1.
136. Russek, H. I., and B. L. Zohman. Relative significance
of heredity, diet and occupational stress in coronary heart
disease of young adults: Based on an analysis of 100 pa-
tients between the ages of 25 and 40 years and a similar
group of 100 normal control subjects. Am. J. Med. Sa.
235: 266, 1958.
137. Schaefer, L. E., D. Adlersberg, and A. G. Steinberg.
Heredity, environment and serum cholesterol. Circulation
17: 537. '958.
138. Schlesincer, M. J. Relation of anatomic pattern to
pathologic conditions of the coronary arteries. A.M. A.
Arch. Pathol. 30: 403, 1940.
139. Schlichter, J. G., and R. Harris. The vascularization
of the aorta. II. A comparative study of the aortic vascu-
larization of several species in health and disease. Am. J.
Med. Sci. 218: 610, 1949.
140. Seifter, J., D. Baeder, C. Zarafonetis, and J. Kalas.
Effect of adrenals, pituitary, liver and mucopoly-
saccharides on blood lipids. In: G. Pincus, Hormones and
Atherosclerosis. New York: Acad. Press, 1959, p. 265.
141 Sellers, E. A., and R. \V. You. Deposition of fat in
coronary arteries after exposure to cold. Brit. Med. J. 1 :
815. !956-
142. Selye, H. Textbook of Endocrinology. Montreal : Acta, 1949.
143. Shafrir, E., K. E. Sussman, and D. Steinberg. The
nature of the epinephrine-induced hyperiipidemia in
dogs and its modification by glucose. J. Lipid Research
'■ '09. '959-
144. Skanse, B., W. Von Studnitz, and N. Skooc. The effect
of corticotrophin and cortisone on serum lipids and lipo-
proteins. Acta Endocrinol. 31 : 442, 1959.
145. Smith, D. H., and E. Gaman. Breed susceptibility in
rabbits to hypercholesterolemia and atherosclerosis.
Circulation 20: 973, 1959.
146. Sperry, W. M., J W. Jailer, and E. T. Engle. The
influence of diet on the cholesterol concentration of the
blood serum in normal, spayed, and hypothyroid monkeys.
Endocrinology 35 : 38, 1 944.
147. Srere, P. A., I. L. Chaikoff, and W. G. Dauben. The
in vitro synthesis of cholesterol from acetate by surviving
adrenal cortical tissue. J. Biol. Chem. 176: 829, 1948.
148. Stamler, J., E. N. Silber, A. J. Miller, L. Akman,
C. Bolene, and L. N. Katz. The effect of thyroid and
of dinitrophenol-induced hypermetabolism on plasma
and tissue lipids and atherosclerosis in the cholesterol -fed
chick. J. Lab. Clin. Med. 35: 351, 1950.
ENDOCR1NES, STRESS, AND HEREDITY ON ATHEROSCLEROSIS
1213
149. Stamler. J., R. Pick, and L. N. Katz. Estrogen prophy- 167.
laxis of cholesterol-induced coronary atherogenesis in
chicks given adrenal corticoids or ACTH. Circulation 10:
247, 1954. 168.
150. Stamler, J., R. Pick, and L. N. Katz. Inhibition of
cholesterol-induced coronary atherogenesis in the egg-
producing hen. Circulation 10: 251, 1954.
151. Stamler, J., R. Pick, and L. N. Katz. Experiences in 169.
assessing estrogen antiatherogenesis in the chick, the
rabbit and man. Ann. -V. }'. Acad. Sci. 64: 596, 1956.
152. Stamler, J., R. Pick, and L. N. Katz. Further observa-
tions on the effects of thyroid hormone preparations on 1 70.
cholesterolemia and atherogenesis in cholesterol-fed
cockerels. Circulation Research 6: 825, 1958.
153. Stamler, J. The epidemiology of atherosclerotic coronary
heart disease. Postgrad. Med. 25: 610, 685, 1959. 171.
154. Stamler, J., R. Pick, and L. N. Katz. Influences of
thyroid, pancreatic and adrenal hormones on lipid me-
tabolism and atherosclerosis in experimental animals. 172.
In: G. Pincus, Hormones and Atherosclerosis. New York:
Acad. Press, 1 959, p. 1 73.
155. Stamler, J., R. Pick, L. N. Katz, A. Pick, and B. M.
Kaplan. Interim report on clinical experiences with long-
term estrogen administration to middle-aged men with 173.
coronary heart disease. In : G. Pincus, Hormones and
Atherosclerosis. New York: Acad. Press, 1959, p. 423. 174.
156. Stamler, J., R. Pick, and L. N. Katz. Effect of insulin
in the induction and regression of atherosclerosis in the
chick. Circulation Research 8: 572, i960. 175.
157. Stamler, J., M. Kjelseerg, Y. Hall, and N. Scotch.
Epidemiologic studies on cardiovascular-renal diseases.
I. Analysis of mortality by age-race-sex-occupation. J. 176.
Chronic Diseases 1 2 : 440, 1 960.
158. Steiner, A., and F. E. Kendall. Atherosclerosis and
arteriosclerosis in dogs following ingestion of cholesterol
and thiouracil. A.M. A. Arch. Pathol. 42: 433, 1946.
159. Still, W. J. S., and R. M. O'Neal. Experimental athero- 1 77.
sclerosis in the rat: The pathogenesis of the early lesion.
Federation Proc. 20: Part 1, 94, 1961. 178.
160. Taylor, H. L., J. T. Anderson, and A. Keys. Diet,
physical activity and serum cholesterol in man. Circulation
16: 516, 1957.
161. Thomas, C. B., and B. H. Cohen. The familial occurrence 179.
of hypertension and coronary artery disease with ob-
servations concerning obesity and diabetes. Ann. Internal
Med. 42: 90, 1955. 180.
162. Thomas, C. B., and E. A. Murphy. Further studies on
cholesterol levels in the Johns Hopkins medical students:
The effect of stress at examination. J. Chronic Diseases 8:
661, 1958. 181.
163. Thomas, A. J., A. L. Cochran, and I. T. Higgins. Meas-
urement of the prevalence of ischaemic heart disease.
Lancet. 2:540, 1958. 182.
164. Turner, K. B. Studies on the prevention of cholesterol-
induced atherosclerosis in rabbits. I. The effects of whole
thyroid and potassium iodide. J. Exptl. Med. 58: 115, 1933. 183.
165. Turner, K. B., C. H. Present, and E. H Bidwell. The
role of the thyroid in the regulation of the blood choles-
terol of rabbits. J. Exptl. Med. 67 : 1 1 1 , 1938. 184.
166. Wang, C. I., L. E. Schaefer, and D. Adlersberg.
Tissue permeability — A factor in atherogenesis. Circula-
tion Research 3: 293, 1955.
Wang, C. I., F. Paronetto, and D. Adlersberg. Hyper-
lipemia and pancreatitis: In man and in experimental
animals. Clin. Research Proc. 5: 197, 1957.
Warnock, N. H., T. B. Clarkson, and R. Stevenson.
Effect of exercise on blood coagulation time and athero-
sclerosis of cholesterol-fed cockerels. Circulation Research 5:
478. '957-
Weiss, E., B. Dolin, H. R. Rollin, H. K. Fischer, and
C R. Bepler. Emotional factors in coronary occlusion. I.
Introduction and general summary. A.M. A. Arch. Internal
Med. 99:628, 1957.
Werthessen, N. T. Control of aortal lipid metabolism
and lipid movement by hormones and vitamins. In:
G. Pincus, Hormones and Atherosclerosis. New York: Acad.
Press, 1959, p. 131.
Wexler, B. C, and B. F. Miller. Coronary arterio-
sclerosis and thrombosis in the rat. Proc. Sue. Exptl. Biol.
Med. 100: 573, 1959.
W 1 xi er, B C, T. E. Brown, and B. F. Miller. Athero-
sclerosis in rats induced by repeated breedings, ACTH,
and unilateral nephrectomy — acid mucopolysaccharides,
fibroplasia, elastosis and other changes in early lesions.
Circulation Research 8 : 278, 1 960.
Whereat, A. F. Oxygen consumption of normal and
atherosclerotic intima. Circulation Research 9: 571, 1961.
White, A. Integration of the effects of adrenal cortical,
thyroid and growth hormones in fasting metabolism.
Recent Progr. Hormone Research 4: 153, 1949.
Wiesel, J. Uber Leberveranderungen bei multipler ab-
domineller Fettgeivebsnekrose und Pankreatitis haemor-
rhagica. Mitt. Grenzg. Med. Chir. 14: 487, 1905.
Whkins, R. H., J. C. Roberts, Jr., and C. Moses.
Autopsy studies in atherosclerosis. III. Distribution and
severity of atherosclerosis in the presence of obesity,
hypertension, nephrosclerosis and rheumatic heart disease.
Circulation 20: 527, 1959.
Williams, R. H. Hyperadrenocorticism. Am. J. Med. 10:
612, 1 951.
Wissler, R. W., M. L. Eilert, M. A. Schroeder, and
L. Cohen. Production of lipomatous and atheromatous
arterial lesions in the albino rat. A.M.A. Arch. Pathol. 57:
333. '954-
W'olffe, J B Continued vigorous physical activity as a
possible factor in the prevention of atherosclerosis. Cir-
culation 16: 517, 1957.
Wong, H. Y. C, R. L. Simmons, and E. W. Hawthorne.
Effects of controlled exercise on experimental athero-
sclerosis in androgen -treated chicks. Federation Proc. 15:
203, 1956.
Wuest, J., T. J. Dry, and J. E. Edwards. The degree of
coronary atherosclerosis in bilaterally oophorectomized
women. Circulation 7: 801, 1 953-
Zilversmit, D. B., T. N. Stern, and R. R. Overman.
Effects of adrenal hormones on blood phospholipids.
Am. J. Physiol. 164: 31, 1951.
Zilversmit, D. B. Phospholipid turnover in atheromatous
lesions. In : G. Pincus, Hormones and Atherosclerosis. New
York: Acad. Press, 1959, p. 145.
Zoll, P. M., S. Wessler, and M. J. Schlesinger. Inter-
arterial coronary anastomoses in the human heart, with
particular reference to anemia and relative cardiac anoxia.
Circulation 4: 797, 1 951.
CHAPTER 36
Peripheral vascular diseases-
diseases other than atherosclerosis1
GEORGE E. BURCH
JOHN PHILLIPS
Department of Medicine, Tulane University School of Medicine,
and Charity Hospital of Louisiana, New Orleans, Louisiana
CHAPTER CONTENTS
Vascular Malfunction in General
Approach of the Clinician and Clinical Physiologist to the
Study of Patients with Diseases of the Peripheral Circula-
tion
The History
Symptoms of arterial disease
Symptoms of venous disease
Manifestations of capillary and lymphatic disease
The Physical Examination and Simple Clinical Tests of
Vascular Function
Determination of the adequacy of the cutaneous circula-
tion
Evaluation of the status of the main arteries
Evaluation of the status of the venous system
Evaluation of the status of capillary and lymphatic vessels
Special Laboratory Procedures for Examining the Peripheral
Circulation
Effects of Circulatory Arrest
Classification of Peripheral Vascular Disease
Mechanisms in Peripheral Vascular Disease
Vasoconstrictor Disease Syndromes
Raynaud's syndrome or phenomenon
Acrocyanosis
Livedo reticularis
Causalgia and related syndromes
Miscellaneous states
Vasodilative Syndromes
Erythromelalgia (erythermalgia)
Mechanisms in Other Vascular Diseases
Appendix
one of the great interests in the vascular system
is its reaction and response to disease. Although most
1 Work supported by grants from the L'. S. Public Health
Service.
physiologic studies have been concerned with the
normal state, the pathophysiology of vascular disease
is of considerable importance to the physiologist and
the clinician. Attempts will be made in this chapter
to correlate the interactions of vascular malfunction
with the pathologic lesions and their clinical mani-
festations. The presentation will be limited primarily
to diseases of and observations on man.
In the discussions to follow the term "peripheral
vascular disease" will refer to disease affecting largely
the circulation to the limbs. This obviously excludes
a discussion of disease in other circulatory beds;
notable among these sites are the pulmonary, portal,
renal, and cerebral vessels. Further, the discussions
will be limited in large part to "primary" vascular
disease, or disease states in which alterations in blood
vessels and their function are the basic cause for the
disease manifestations. Vascular changes associated
with or secondary to primary disease in other organs
are important but they are beyond the scope of this
presentation. Examples of these secondary vascular
disturbances are the spider angiomata and palmar
erythema occurring in liver disease, aging, rheumatoid
arthritis, and pregnancy; the pale avascular skin of
castrate and eunuchoid men; the reddish flushes of
the menopausal states; the cyanotic flushes of seroto-
nin-producing carcinoids; the pallor of the nephrotic
syndrome, hypothyroidism, and pituitary insuffi-
ciency; the vasoconstriction of pheochromocvtomas;
the vasodilatation of thyrotoxicosis; the vascular
changes of acute exanthema, scarlet fever, and other
infectious diseases; and the digital clubbing and
cyanosis of cardiac and pulmonary disease.
1215
12 I 6
HANDBOOK OF PHYSIOLOGY " CIRCULATION II
\ VSCULAR MALFUNCTION IN GENERAL
Vascular malfunction might be denned as that
temporar\ or permanent condition which exists
when the circulation fails to meet its intended func-
tions of a) temperature regulation, b) tissue nutrition,
and c I repair.
Malfunction might arise through active vasomotor
or passive structural (anatomic) mechanisms. Vaso-
motor (not limited only to neuromuscular) mecha-
nisms include a) increase in vessel tone ("vasotighten-
ing"), decrease in luminal cross-sectional area, or a
combination of these (vasoconstriction), or b) decrease
in vessel tone ("vasoloosening"), increase in luminal
cross-sectional area, or a combination ol these
(vasodilatation). Vasomotor changes might be
induced through neural mechanisms, humoral
mechanisms, primary muscular action, physical
factors affecting any of the vessel wall coats, other
unknown factors, or any combination of these. These
reactions imply a degree of reversibility.
Structural or anatomic mechanisms by which
vascular malfunction might occur are a) structural
obstruction (occlusive), b) structural dilatation
(aneurysm or varix), or c) abnormal vascular com-
munications. These changes imply a degree of
irreversibility.
Somewhat difficult to define as either active vaso-
motor or passive structural changes are vascular
distention and vascular collapse (the latter not refer-
ring to the "shock syndrome"). Vascular distention
or congestion implies a relative increase in vessel
tone (as opposed to vasodilatation) but with an
increase in luminal cross-sectional area. Vascular
collapse implies a relative decrease in vessel tone (as
opposed to vasoconstriction) but with a decrease in
luminal cross-sectional area. These are, of course,
potentially reversible states.
Structural diseases have been termed "organic,"
whereas the vasomotor diseases have been called
"functional."' Division of peripheral vascular diseases
into organic and functional categories, although
convenient, is purely arbitrary. Certainly, altered
phvsiology has its structural counterpart. Means to
resolve this artificial dichotomy then are dependent
simply on the sensitiveness of methods for morpho-
logic observations. In the past, the division of diseases
from the anatomic standpoint has been dependent
largely on light microscopy. Under existing classi-
fications, for example, early and mild Raynaud's
disease is a functional disorder. However, with the
use of more sensitive methods such as electron
microscopy the same stage of the disease might be
shown to be associated with structural defects whether
it be in the vasculature itself or in the nervous system
or in both. Thus, by present "policy" the disease is
now considered both organic and functional.
Regardless of the above criticisms, it is still con-
venient for clinical purposes to classify vascular
disease within organic or functional categories. This
focuses attention on the more observable underlying
mechanisms in the characteristic manifestations of
the disease. With this in mind, more emphasis shall
be placed here on the disorders in which altered phy-
siology is the most readily detectable underlying
mechanism; these diseases include predominately,
but certainly not exclusively, the functional disorders.
It should be remembered that none are purely
organic or purely functional and that all vascular
diseases have elements of each.
Changes which influence the circulation and its
functions, although not directly arising from the
vessel wall itself, pertain to such factors as blood
volume, cardiac output, pulse rate, blood viscosity,
sludging, blood gases, neurogenic and psychogenic
disorders, endocrine and humoral factors, and many-
others. Many of these factors operate simultaneously
to various degrees and with temporal variations. These
topics are covered in ether chapters of this volume.
APPROACH OF THE CLINICIAN AND CLINICAL
PHYSIOLOGIST TO THE STUDY OF PATIENTS WITH
DISEASES OF THE PERIPHERAL CIRCULATION
The clinical peripheral vascular physiologist has
a difficult and complex task. He must observe the
symptoms and clinical and laboratory signs in his
subject which suggest the possibility cf pathological
alteration in the peripheral circulation. He must
attempt to discover the underlying pathologic anat-
omy of the clinical manifestation. Most difficult of
all, he must attempt to explain the observed changes
in terms of pathophysiologic mechanisms, establish
a diagnosis and then introduce corrective therapeutic
measures based upon established pharmacodynamic
and physiologic principles. This is done with an aim
to modify the altered pathophysiology in order to
establish as near normal vascular function as possible
for as long a period of time as possible. This objective
requires a satisfactory understanding of the normal
and abnormal functions of the interrelated organ
systems which may influence the diseased state. The
clinician must attempt to estimate properly the
PERIPHERAL VASCULAR DISEASES
12 I 7
relative quantitative, qualitative, and temporal roles
of the many contributing factors such as the numerous
effects of disease of other organ systems. All of these
factors must be carefully integrated in order to decide
upon the type, amount, and time of administering
various therapeutic measures.
Progress in the basic understanding of peripheral
vascular disease has been slow. Because of the enor-
mous number of variables and the complex nature
of these diseases, the clinical physiologist has had
great difficulty in elucidating underlying patho-
physiologic mechanisms. No small hindrance to this
progress has been the nature of the experimental
animal himself, namely, man. Nowhere else in
physiology does nondestructive observation place its
strictest limitations. Available counterparts of spon-
taneously acquired human peripheral vascular disease
are rare indeed in lower animals.
The History
The detection, clarification, and interpretation of
a patient's own experience with his altered circula-
tion (symptoms) may be just as important to the
clinician and clinical physiologist as graphic record-
ings of circulatory parameters may be to the basic
physiologist. For this reason, it seems worthwhile
to discuss briefly the important aspects of this means
of investigation. Where known, the physiologic
mechanisms underlying these symptoms will be noted.
Several general aspects of the history are note-
worthy. Age, sex, and race are important; e.g., arterio-
sclerosis is more common in the aged; Raynaud's
disease is much more frequent in females; and
Buerger's disease is extremely rare in Negroes and
women (74). Because of their predisposing influences
on subsequent vascular disease, a past history of poly-
cythemia, frostbite, thrombophlebitis, diabetes mel-
litus, and many other disease states is important.
Occupational factors should be explored, e.g., the
predisposition to Raynaud's phenomenon seen in
truck drivers and pneumatic hammer operators.
Evaluation of environmental influences such as
temperature, humidity, and body position is also
important. The effects of drugs may be important,
e.g., ergot, nicotine, and sympathicomimetic agents.
Evaluation of emotional and other psychic factors
are also of considerable importance.
More specific aspects of the history are as follows:
symptoms of arterial disease. Among the common
symptoms associated with reduced arterial flow are
pain, tenderness, fatigue, paresthesia, altered sensa-
tions ranging from hyperesthesia to anesthesia, muscle
cramps, and sensitivity to thermal change. Pain may-
be divided into three main groups: intermittent
claudication, rest pain due to ischemic neuritis, and
rest pain associated with trophic changes.
Intermittent claudication characteristically is pain
produced by exercise and relieved with rest. It may
appear in any muscle group and is usually due to
organic arterial obstruction. It is related to the degree
and or rate of work (in the physiologic sense) over
a certain time interval performed by a particular
muscle group with its compromised circulation.
Increasing the amount or rate of work produces a
more rapid onset, a more severe degree of pain, or
both. With reference to the lower extremities, clini-
cians attempt to quantitate claudication in terms of
onset of pain after walking a certain distance (claudi-
cation distance) or after walking a certain period of
time (claudication time) at a prescribed pace. This
symptom usually starts as a sense of "fatigue" then
progresses to a "cramping" pain. A major charac-
teristic of claudication is relief with rest. When post-
exercise relief does not ensue within 5 to 10 min,
another cause for the pain is suspected.
"Vasospastic claudication" is a term used to
describe a syndrome in which peripheral arterial
pulsations are normally present at rest but disappear
during exercise. The affected limb may then become
pale and typical claudicatory distress occur (43).
From clinical studies it is considered that the majority
of this comparatively small group of patients have
partial segmental occlusion of large arteries proximal
to the site of claudication (36) and that superimposed
arterial spasm is responsible for the ischemic mani-
festations on exercise.
The exact mechanism of intermittent claudication
is not clear. That claudication is not clue to muscle
cramps has been repeatedly stressed, since the muscles
are flaccid during the attack, and it is not due to
vascular spasm of small vessels because the vessels to
the muscles are dilated rather than constricted with
exercise. Claudication indicates insufficient blood
supply to the painful muscles to meet the increased
metabolic needs of the muscle during exercise.
It appears that there are at least two basic require-
ments for the production of intermittent claudication :
/) oxygen lack and 2) muscular contraction in the
presence of this anoxia (36). In this regard it should
be noted that claudicatory pain has been produced
in severely anemic patients with patent arteries (78)
and that it has been produced by exercising normal
I2l8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
people while breathing air with low oxygen concen-
trations (40). It is noteworthy that intermittent
claudication may be produced in a normal limb by
exercise after artificial arrest of the blood supply.
In the normal limb following circulatory arrest, the
pain disappears quickly (usually within 3 sec) after
restoration of the circulation, but if arrest is main-
tained the pain persists, presumably because the
agents responsible for the pain are not inactivated in
the presence of an inadequate circulation.
Lewis has shown that it is not oxygen lack itself
that causes the claudicatory pain but rather the
stimulation of sensory nerves by metabolic products
of muscular activity which are ordinarily inactivated
in the presence of an adequate blood supply with
sufficient oxygenation. The mediating agent or
agents from ischemic muscles to the pain-sensitive
nerve endings has been called "factor P" or "pain
factor." Apparently, it is a metabolic product of
muscle and is rather stable, acid, and nonvolatile.
Whether or not it is produced in increased quantities
or is inadequately neutralized, inhibited, or dis-
persed in the face of ischemia is not known. Few
definitive studies in this area are available. Among
the more important are those of Lewis et al. (47,
51 ) and Katz et al. (37).
One observation that requires further exploration
is the relief of deep pain by the application of ethyl
chloride spray to the skin surface. Travell and
associates (98) have reported pronounced ameliora-
tion of claudicatory distress by this means.
Ischemic neuritis is considered to be one of the
mechanisms of rest pain in arterial disease. This
type of pain, severe lancinating sensations and
paresthesias, is characteristic of stimulation of neural
elements. It is usually more troublesome at night
when the patient is in bed. The pain of ischemic
neuritis is found most frequently in vascular dis-
turbances associated with diabetes mellitus and
Buerger's disease.
The pathogenesis of the symptoms due to ischemic
neuritis appears to be related to neural degeneration
secondary to impairment of blood flow through
the nutrient vessels of the nerves. In support ol
nerve degeneration is the associated reduction in
vibratory sense perception and pinprick sensation.
Trophic changes may be responsible for pain occur-
ring in the resting state. These painful sites are areas
of ulceration and pre-ulceration which probably
cause sensory nerve irritation through inflammation
and ischemia. This type of pain is usually continuous
in nature. It is more common in Buerger's disease
and in diabetic neuropathy probably because of
the involvement of nerves in the inflammatory
processes.
The pain associated with occlusive arterial disease
is frequently accentuated by elevation and diminished
by dependency of the involved part. Excessive local
heat may be harmful and often increases the pain
because of increased metabolism with restricted
blood supply. Excessive cold may induce vascular
spasm and also accentuate pain and tissue damage.
symptoms of venous disease. Symptoms due to
disease of the veins may include pain, muscle fatigue,
muscle cramps, and paresthesias. Many of the
symptoms of venous insufficiency (deep-vein throm-
bosis and obstruction, valvular incompetence, and
varicose veins) are due to congestion and edema of
the involved parts and therefore are affected by
gravity. These symptoms are accentuated by depend-
ency and diminished by elevation. Additional man-
ifestations of venous disease and their pathogenesis
will be found in subsequent sections.
manifestations of capillary- and lymphatic
disease. These diseases are discussed in other parts
of this volume.
Physical Examination and Simple Clinical
Tests of Vascular Function
In the sense that the basic physiologist must define
the conditioning influences under which his labora-
tory experiments are conducted, so must the clinician
define the conditions under which he makes his
observations of disease processes. In this respect,
in the examination of his patient for peripheral
vascular disease, the clinician must make every
attempt to control influencing variables in the
environment. LTnder ideal conditions then, the
changes observed during a "steady state" estab-
lished by proper conditions for examination may be
assumed to be due to the disease itself. To this end,
standardization of the condition and technique of
the examination is necessary.
The subject should rest supine in bed in a com-
fortable position with no constricting garments.
The environmental temperature and humidit)
should be in a comfortable range. The parts should
be dry and free from exposure to drafts. Blankets
may be applied, if necessary, but when employed
should cover all parts symmetrically. Local heat or
cold should be avoided. Other influencing factors
PERIPHERAL VASCULAR DISEASES
12 19
such as recent use of tobacco, alcohol, or certain
drugs should be controlled.
Because of the great number of variables in disease
and because of the wide range of normal variation,
the clinician must take advantage of "built-in"
controls. To this end, he should carefully and con-
tinually examine and compare symmetrical parts
of the body.
DETERMINATION OF THE ADEQUACY OF THE CUTANEOUS
circulation. The presence and location of cutane-
ous ulcerations should be noted. In arterial disease
these tend to be at the tips of the digits and over
pressure areas, whereas in venous disease they tend
to be located over the medial lower one-third of the
leg. The skin should be examined for texture and
consistency. The skin tends to be thin and shiny in
arterial disease and thick and brawny in long-
standing venous and lymphatic disease. Tissue
swelling and edema tend to be absent in arterial
disease, unless there has been considerable capillary-
injury, but they are frequent findings in venous and
lymphatic insufficiency. Changes in the growth rate
and appearance of the nails may be clues to impaired
cutaneous circulation. The nails tend to be thickened,
ridged, deformed, brittle, and pigmented. In vaso-
spastic states there may be thinning of the proximal
nail fold with merging into the cuticle (pterygium).
Hair growth may be impaired. The degree of sweating
is important. Absence of sweating may indicate
complete ischemic destruction of sympathetic nerve
fibers or ischemic impairment of sweat gland func-
tion. Excessive sweating, in the absence of a demand
of this function for thermal regulation, usually
indicates increased sympathetic activity with intact
nerves, frequently due to psychogenic disturbances.
Other vascular manifestations of increased sympathe-
tic activity with respect to temperature and color of
the skin are usually present.
Temperature and color changes are of such
importance in the evaluation of the cutaneous
circulation that they demand special comment. The
observations of Lewis (49) are still authoritative.
Under standardized conditions, the amount of heat
brought to the skin may be considered a gross reflec-
tion of the rate of local blood flow. It should be noted
that the temperature of a part cannot decline more
than 1 C to 2 C below room temperature and then
only if the part is moist and the circulation com-
pletely arrested. The temperature rarely decreases
below 20 C in a cool room and rarely exceeds 34 C
in a warm room. Temperature differences of sym-
metric areas, similarly exposed, should arouse sus-
picion of circulatory disorder. When exposed to
cold, the part with the better circulation will remain
warm longer and on rewarming its temperature will
increase faster.
Except for modification by skin pigments, the
color of the skin is due mostly to blood in the venules
of the subpapillary venous plexus and to a lesser
34
24.
_^.^_ REFLEX VASODILATATION
(BODY HEATED)
POSTERIOR TIBIAL
NERVE BLOCK
MINUTES
"I I I I I I I I I T"
5 10
"~ 1 1 I I I I I I I I I I I I I 1 1 I I 1 1 1 1 1 1 1 r~
15 20 25 30 35 40
45
fig. I. Thermocouple recording of thermal change in the right big toe demonstrating the effects
of reflex vasodilatation and posterior tibial nerve block.
i 220
HANDBOOK OF PHYSIOLOG1
CIRCULATION- II
O
Control Stott
Subj JM.J5WM
Tr*nchfool
2RT J
_ IOTIOO en
S olo t
E
I
= 4RT
° 10 _ 100
«W ^am)
FIG. 2. Rheoplethysmographic recordings showing the simultaneous curves of volumes and rates
of digital inflow, outflow, and the difference between inflow and outflow for the tip of the right
second toe (2RT) during a single pulse cycle. A represents the curves for the subject resting supine
in a comfortable environmental atmosphere, B following heating of the trunk, and C following pro-
caine block of the posterior tibial nerve. Iy and IK represent the time courses of the volume and rate,
respectively, of inflow; Oy and Or, volume and rate, respectively, of outflow ; Dy and DR difference
between the volume and rate, respectively, of inflow and outflow. The reader should refer to the
iterature (10, 11) for a discussion of rheoplethysmography. See following pages for 2/? and iC.
degree to blood in die cutaneous capillaries. When
the velocity of blood is slow, more oxygen is removed
by the tissues, the concentration of reduced hemo-
globin increases, and the color of the skin darkens
and becomes bluer.
The integration of skin color and temperature
has been aptly stated by Lewis (49). These charac-
teristics, of considerable physiologic and clinical
significance, are:
"Warm pale skin: This is a skin through which
blood flows rapidly for many minutes. It is warm
because flow is fast, pink because of the abundant
supply of fully oxygenated blood, and pale because
the skin is well nourished and minute vessel tone is
therefore high.
"Warm deeply coloured red skin: Such skin has been
irritated, by heat or otherwise, it is in a state of
inflammation, or it is skin in which arterial vasodila-
tation has recently been brought about through
nervous channels or by means of drugs such as
amyl nitrite.
"Cold pale cyanosed skin: This is skin to which the
blood-flow is very slow or absent. If the tint of the
cold skin is violaceous or if the skin is blanched, the
circulation to it is absent and has been arrested in it
for many minutes. Minor grades of cyanosis are, as
previously stated, of much less significance.
"Cold deeply coloured cyanosed skin: This is skin in
which the circulation is very slow, and in which
blood-flow has been failing for a long time or in
which there is a process of low-grade inflammation.
"Cold deeply coloured red skin: If skin is sufficiently
cold, 10° C (50° F) or less, the blood will not part
with its oxygen, but the minute vessels are damaged
PERIPHERAL VASCULAR DISEASES
I 221
H«t*fl of Tntn»
Svt.] JM, S9 WM
TftftcMOOt
FIG. 2S
y
/ .
/
**•
0»
Ov
and dilate, and thus the skin becomes bright red in
colour although the blood-flow through it may be
small."
Some generalizations might be made from these
correlations. If the temperature of a part is warm,
there is a large volume flow; if it is cool, the volume
flow is small. If the depth of skin color is pale, the
cross-sectional area of the minute vessels is decreased;
whereas if a deep color is present, the cross-sectional
area is increased (vessels open). If the color is pink,
the velocity of blood flow is fast and oxygen satura-
tion high; if red, the velocity of flow is intermediate
and the oxygen saturation is intermediate; if blue, the
blood flow velocity is slow and the oxygen saturation
low.
A crude estimation of the cutaneous circulatory
status may be obtained from the subpapillary venous
plexus filling time. Digital pressure on the skin for
several seconds results in displacement of blood into
adjacent and deeper lying areas. On sudden release
of this pressure, the normal skin shows a change
from the pallor to a normal color within i or i sec.
Gravitational effects on the cutaneous circulation
may be employed in diagnosis and estimation of the
state of the circulation by means of elevation and
dependency tests. Here, the patient lies on his back
with the legs flexed to go degrees at the hip. He holds
this position for a certain length of time, usually i
min. If he will tolerate it, the patient may be requested
to flex and extend the ankle during this period. With
impairment of the circulation, the skin assumes a
white, waxy color during this maneuver. The patient
is then instructed to sit on the edge of the examining
table with his legs hanging dependent. Normally,
gravity and reactive hyperemia cause a return of
flushed color to the skin within a short interval of
time, usually 1 5 sec. A delay in return of color is
roughly proportional to the degree of circulatory
insufficiency.
The test of venous filling may be performed simul-
taneously with the above gravitational test. The dura-
tion of time from the moment the patient sits until
the superficial veins of the legs are filled is a gross
indication of the circulatory status in the legs.
1222
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Procolnt Block ©t PotHnor Tlt-lol N*rvt
Sub>JM,35 WM
FIG. iC
Normally, venous filling starts within 30 sec. For this
maneuver to be valid as a test for arterial sufficiency,
the veins must be relatively normal with competent
valves. Venous valvular insufficiency is evidenced by
abrupt venous filling.
Tests of vascular dilatability may be useful in the
evaluation of the cutaneous circulation. Reactive
hyperemia may be employed as such a test. The
techniques and mechanisms for this reaction are
discussed in detail elsewhere in this volume, but brief
comments may be in order. As originally recom-
mended (49, 77), the part to be tested is placed in
water at 35 C to 40 C for 10 min then removed and
raised above body level until pale. The purposes of
this procedure are to ensure that vessels in spasm are
relaxed and to empty the minute vessels of blood. A
sphygmomanometer cuff is then placed about the
part to be investigated and inflated to a pressure
above systolic blood pressure. The limb is then
returned to the water bath and maintained there for
5, 10, or even 15 min with the circulation arrested, if
tolerated by the patient. The limb is then lifted out
of the bath, dried, and its circulation released. In
normal limbs, the reactive hyperemic flush reaches
the tips of the digits within 2 to 5 sec, becomes
maximal in 1 5 sec, then quickly fades. When organic
arterial occlusion is present, the flush spreads slowly,
is patchv in distribution, and may be delayed up to
a minute or so in reaching the tips of the fingers or
toes. When disease is present and the onset is delayed,
the flush lasts for much longer periods of time. More
quantitative methods such as plethysmography or
thermometry may be used to measure the reaction.
Vascular dilatability may be tested by methods
which decrease lympathetic tone. For example, local
nerve block (figs. 1, 2A, B, and C) or paraverte-
bral and stellate sympathetic ganglionic nerve block
mav be employed with the responses of the circula-
tion being measured graphically by temperature
recordings, plethysmographic recordings, or other
means. Vasospastic states due to sympathetic nervous
activitv become evident from the recordings. In
normal subjects or in patients with functional vaso-
spasm, interruption of sympathetic activity results in
a rise in digital cutaneous temperature to approxi-
mately 32 to 35 C when the subject is at rest in a
comfortable environment. For obvious reasons, the
total rise above control levels is much greater in
PERIPHERAL VASCULAR DISEASES
1223
patients with vasospasm than in normal subjects. In
patients with obliterative arterial disease (organic),
interruption of sympathetic innervation results in
little change or only a moderate increase in digital
cutaneous temperature, depending upon the degree
of obstructive disease (usually only to about 28 C).
A decline in cutaneous temperature from control
levels following sympathetic inhibition indicates
severe impairment of arterial circulation. This
response may be due to a pre-existing lack or dys-
function of sympathetic innervation in the diseased
area caused by ischemic degeneration of sympathetic
nerves. Rigidity of the diseased blood vessel walls
may be another factor. After induced sympathetic
inhibition, blood is apparently shunted away to
more healthy areas where the resulting decrease in
vascular resistance would be proportionately greater.
It should be remembered that tests dependent on
inhibition of sympathetic innervation are of value
clinically only in evaluation of the cutaneous circula-
tion and are of little or no value in the investigation
of the circulation to muscle.
The above tests, however, are of a special or some-
what complex nature. More simply, sympathetic
inhibition may be induced by means of reflex vasodila-
tation. Reflex vasodilatation is produced by heating
a part of the body other than that which is being
tested. Application of a radiant heat tent over the
trunk may be used with the temperatures main-
tained at approximately 50 to 60 C. Responses can
be determined graphically by temperature or plethys-
mographic recordings (figs. 1, 2 A, B, and C) or other
suitable means. Normally, the unheated tested limb
may reach approximately 32 C when other parts of
the body are warmed. Reflex vasodilatation is
caused in large part by indirect sympathetic vaso-
constrictor inhibition resulting from the action of
warmed blood on central sympathetic temperature-
regulating centers (30), but in part also by increased
activity of the vasodilator fibers (100).
Another simple clinical procedure is the histamine
wheal test (48, 93). The wheal formation of the
triple response produced by intradermal injection of
histamine is dependent upon the rate of local blood
flow and capillary pressure. The rate at which the
wheal forms is a rough indication of the status of
cutaneous circulation. Briefly, the test is performed
by slightly puncturing the skin several times with a
sharp needle through a drop of 1:1000 solution of
histamine acid phosphate. The subsequent reaction
is then observed. It has been stated that if a wheal
fails to appear within 3 to 5 min, ischemia of tissue
is severe; if a wheal does not develop at all, gangrene
is imminent. Normally, the wheal develops in 3 min.
The mechanisms of the triple response have been
discussed in detail by Lewis (48, 49).
EVALUATION OF THE STATUS OF THE MAIN ARTERIES.
In the evaluation of the arterial circulation, one of
the most important and informative procedures is
a careful and methodical palpation of the main
arteries, including all the major branches from the
aorta to the digital ones. The abdominal portion of
the aorta is readily palpated. The thoracic aorta may
be palpated in advanced aneurysm formation. The
digital arteries of normal people are usually palpable.
When arterial spasm is present and the pulses are
weak, sublingual administration of 0.4 mg of nitro-
glycerin (24) or the inhalation of amyl nitrite may
release the vascular tone sufficiently to cause marked
accentuation of arterial pulses. Obviously, arteries
which have been totally and organically occluded
will remain impalpable after this procedure.
For further information concerning functional
states of the main arterial circulation, one often
must employ more specialized procedures. With
the foregoing diagnostic procedures, however, this
is usually not necessary.
EVALUATION OF THE STATUS OF THE VENOUS SYSTEM.
The common signs of venous disease should be
observed and evaluated. One of the most frequent is
edema, which tends to be compressible or ""pitting"
in acute or subacute stages of formation but firm and
less compressible when of long duration. Abnormally
dilated and distended veins and venous varicosities
are frequently present. Ulcers are common in long-
standing venous insufficiency and tend to be located
over the medial lower one-third of the leg. Also
located in this area and frequently found in chronic
disease is the so-called ""stasis dermatitis," an atrophic,
pigmented, chronic, low-grade inflammatory area
of skin.
Several simple clinical tests are available for
determining the competency of the venous valves
and the patency of the deep veins. They are usually
employed in the evaluation of varicose veins. Infor-
mation pertaining to the clinical tests may be found
in several publications (9, 54, 65, 67, 72, 7g, 80, 99).
The tests are well known and will not be repeated
in detail here, but briefly some are as follows:
Brodie- Trendelenburg test (gg)- This test is designed
to test the valvular competence of the saphenous
and communicating venous system. It involves two
1224
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
parts. In the first part the leg is elevated above the
body, the veins are emptied, and a tourniquet is
applied over the upper thigh tightly enough to occlude
the superficial but not the deep veins. The patient
then stands. If the superficial veins fill quickly (in
less than 30 sec) then the valves of the deep to super-
ficial communicating veins are incompetent. The
second part of the test is similar except that the
tourniquet is removed at the moment the patient
stands. If the superficial veins fill immediately in a
retrograde fashion, then the saphenous valve system
is incompetent.
Perthes test (72). This test is designed to evaluate the
competency of the saphenous and communicating
valves and to test for deep vein patency. A tourniquet
is applied to the thigh and the patient is asked to
walk for 5 min. If the superficial veins collapse during
this walk, there is an indication that the communicat-
ing valves are competent and the deep veins are
patent. If no change is observed, the indication is
that the communicating valves are incompetent. If
the superficial veins become more prominent and
pain is produced, there is an indication that the
deep veins are obstructed and the communicating
valves may be incompetent. This test utilizes the
pumping effect of exercise on venous flow.
The Perthes test has been modified by Mahorner
& Ochsner (54) to determine the location of the
incompetent communicating veins. The modifica-
tion involves the application of tourniquets to the
upper thigh, above the knee and below the knee
before walking. Observations are made on the veins
below the tourniquet before and after each walking
period, and if they collapse then the communicating
valves below the tourniquet are competent.
The mechanisms involved in the above and similar
tests should be obvious and will not be pursued
further.
Determinations of venous pressure at rest and
during exercise are of help not only in clinical evalua-
tion but also in understanding the pathophysiology
of many manifestations of venous insufficiency (valvu-
lar incompetence or venous obstruction) (g, 79).
With significant venous obstruction or valvular incom-
petency, venous pressure in the leg fails to drop
during walking as it does in the normal. Venous
blood flow also fails to increase during walking. Thus,
high venous pressure and sluggish venous flow persist
during the normal daily activity of these patients.
When venous obstruction is marked, venous pressure
is even higher and the venous flow more sluggish
than normal even in a recumbent position. The
increase in venous pressure is transmitted back to
the venules and capillaries. Thus, the many factors
of venous hypertension, stagnant flow, lowered blood
oxygen tension, compression of tissue by dilated
veins, delayed removal of waste products of metab-
olism, previous and present inflammatory reactions,
edema, infection, and everyday trauma combine
to produce the secondary manifestations of venous
disease such as stasis dermatitis and dermal ulcers.
In advanced stages, thickening of the walls, endo-
thelial proliferation and degenerative changes may
be found in arterioles as well as venules (3).
Edema formation secondary to venous obstruction
or valvular incompetence is a topic of its own and will
not be discussed here. Some aspects, however, are
discussed in a subsequent section of this presentation.
EVALUATION OF THE STATUS OF CAPILLARY AND LYM-
PHATIC vessels. This evaluation is available in other
portions of this volume.
Special Laboratory Procedures for Examining
the Peripheral Circulation
Many of the specialized laboratory procedures
for investigating the peripheral circulation are dis-
cussed elsewhere in this volume and are beyond the
scope of the present discussion. Such methods include
plethysmography, thermometry, calorimetry, intra-
vascular and tissue pressure recordings, circulation
times, blood-gas analyses, rate of radioisotopic
clearance, arteriography, venography, lymphangio-
graphy, sweat studies, nailbed and scleral capil-
larioscopy, infrared photography, and oscillometry.
EFFECTS OF CIRCULATORY ARREST
Complete circulatory arrest lies at the farthest end
of the spectrum from the normal state. Before pro-
gressing to disease, the effects of which may lie any-
where between these two extremes, it might be
beneficial to review briefly the effects of complete
arrest of the circulation. Much of these data are
found in the publications of Lewis (48, 49).
With complete circulatory arrest the temperature
of the part decreases to room temperature. The rate
of decline is dependent upon environmental tem-
perature, relative humidity, and air currents. The
greater the mass of the part, the slower the decline.
Pallor is the first change in color as the blood drains
out of the minute vessels in the first 30 to 60 sec.
PERIPHERAL VASCULAR DIS1
[22 =
During the next few minutes the part becomes bluer
and then definitely cyanotic.
Lewis (48) made interesting observations on the
blood vessel reactions in the skin of the arm after
obstruction of the brachial artery by a pneumatic
cuff. After a few minutes of obstruction, reel and
white spots (Bier's spots) form in the cyanotic back-
ground of the surrounding skin. The red spots are
due to leakage of blood into the area from collateral
circulation through bone. The white spots are due to
intense contraction of the vessels responsible for
skin color, in particular, the subpapillary plexus of
veins. Lewis excluded cold temperature and central
and local nervous factors as primary causes of the
white spots. They are formed in previously dener-
vated areas. Through a careful series of experiments,
Lewis showed that in cutaneous areas in which the
circulation has been sufficiently reduced, vaso-
constrictor as well as vasodilator substances are
formed. The vasoconstrictor substances are released
locally in the tissue spaces and are not derived from
the blood. These substances act against potent
vasodilator factors known to be released when the
circulation is arrested. As noted by Lewis (48), these
white spots enlarge and coalesce progressively as the
skin is deprived of its circulation. He noted that at
death the skin is initially congested but shortly after-
ward the white spots appear, spread, and coalesce
until all but dependent parts of the skin are white or
universally blanched. Responses comparable to Bier's
spots have been described for organs other than the
skin (86).
After its induction, if circulatory arrest to a limb
continues, nervous changes develop, distally at first
and then progress proximally up the extremity.
Numbness occurs within 15 min and is followed by
hypesthesia, first for pain and cold and last for
warmth. With respect to the arm, motor changes occur
in approximately 20 min, appearing first in the
thenar eminence. Within 25 min motor paralysis is
usually present in the thenar muscles and within
30 min in the interossi and extensor muscles of the
wrist.
If the circulation is re-established within 30 to 60
min after its arrest, complete recovery usually occurs.
When arrest is prolonged, severe changes occur.
Within 6 to 1 2 hours there is muscle death and
whealing and blistering of the skin. After circulatory-
arrest for 12 to 20 hours there is nerve destruction
and after 24 to 48 hours, necrosis of the skin.
Recent studies (33, 88) of pathologic changes from
acute ischemia in man have shown that after a few
hours of circulatory arrest a muscle contracture similar
to rigor mortis develops. This does not progress
inevitably to fibrotic (Yolkmann's) contracture.
Although the latter is a frequent occurrence, the
initial contracture can be reversible. It was noted
that the early changes can be accompanied by little
or no obvious histologic alteration but somewhat
paradoxically, restoration of the circulation at this
stage often leads to sudden increase in the apparent
severity of muscle damage. It is thought that the
previous "normal" histology might be merely that
of dead or dying muscle preserved in a cool environ-
ment and that subsequent circulation of warm blood
results in the demonstrable vascular engorgement,
swelling, exudation, and focal hemorrhage with
release of myoglobin and consequent muscle pallor.
Depending on severity and duration, the involved
muscle can recover completely or suffer any degree
of damage (with subsequent fibrotic contracture) up
to complete necrosis. It was observed that skin is
more resistant to ischemic damage than is muscle,
and muscle can be irreversibly damaged even though
the skin remains viable. From these studies it was
difficult to place a definite length of time for circula-
tory arrest to produce irreversible change, but a gross
estimate was 1 2 hours or less.
With respect to the clinical implications of the
above studies in the management of acute arterial
occlusion by thrombo-embolectomy, it is worthwhile
noting that blood in an artery distal to an occlusion
usually remains fluid for 8 to 12 hours (4). After-
ward the tendency to thrombosis and progressive
distal arterial occlusion increases rapidly.
CLASSIFICATION OF PERIPHERAL VASCULAR DISEASE
The classification of peripheral vascular disease
included in the Appendix to this chapter although
not all-inclusive is fairly complete. It represents a
modification of the classification suggested by the
Criteria Committee of the New York Heart Associa-
tion (15). It serves to emphasize the enormous prob-
lem and types of peripheral vascular disease. Each
entity in the classification represents a separate
complex experiment in nature. An adequate dis-
cussion of each would be impossible. For more elabo-
rate descriptions of these diseases and for references,
the reader may consult monographs on the subject
(1,3, 87, 104) and two recent symposia on peripheral
vascular diseases (95, 96). The diseases selected for
discussion here are the more common ones as well
I _•_>(>
HANDBOOK OK PHYSIOLOGY
CIRCULATION II
as tliose primarily oriented best from the standpoint
of discussion of mechanisms in peripheral vascular
disease. Particular emphasis is placed on vasocon-
strictor and vasodilator disease states.
MECHANISMS IN PERIPHERAL VASCULAR DISEASE
I asoconstrictoi Disease Syndromes
Raynaud's syndrome or phenomenon. This syndrome
or phenomenon had been known for years before the
time of Raynaud. Ragnetta, Huguier, Virchow,
Zambaco and others commented on the syndrome,
but Raynaud's thesis published in 1862 (84) first
brought wide attention to the syndrome as a distinct
entity. Since that time Lewis' work has been out-
standing (46, 48, 49).
For purposes of classification and diagnosis one
may refer to a) primary (idiopathic) Raynaud's
syndrome or phenomenon, and b) secondary Ray-
naud's syndrome or phenomenon. When Raynaud's
svndrome occurs as a primary manifestation and
without any obvious underlying or predisposing
cause, it is termed "primary." When the syndrome
occurs as a result of, or in association with, some
other disease which is known to be of significance in
predisposition to or production of the syndrome, it is
termed "secondary." Obviously, the classification
of the secondary type is somewhat crude, since it is
based upon empiric observation of an association of
the syndrome with some other disease process with
.1 frequency not expected in otherwise normal people.
Certainly the primary syndrome must be secondary
to its cause. Nevertheless, for diagnostic, prognostic,
and therapeutic reasons, this classification is helpful.
Raynaud's disease is the term applied when the
typical phenomena have been present for 2 years
without detection of any obvious cause. Although
this terminology is arbitrary, it is clinically valid
since most diseases in which Raynaud's syndrome is
a secondary manifestation are usually diagnosed
within a 2-year period (22).
Raynaud's syndrome characteristically consists
of transient episodes of digital pallor, cyanosis, and
erythema. The typical progression would be from
pallor, to cyanosis, to erythema, but this is not always
true. Erythema is not invariably noted and its
presence is not a requirement for diagnosis. Although
a pale blue-gray reaction usually precedes the stage
of pallor (or cyanosis when pallor is absent), it
frequently escapes notice. In order to diagnose
Raynaud's syndrome confidently there should be at
least intermittent attacks or crises of either digital
pallor (syncope) or digital cyanosis. Both may be
present and either or both may be associated with
subsequent erythema.
Primary Raynaud's phenomenon and Raynaud's disease.
Raynaud's disease usually, but not invariably, appears
before the age of forty and is much more frequent in
lemales. Typically the vasomotor episodes are precipi-
tated by exposure to cold and occasionally by emo-
tional stress. In diagnosis, blanching can often be
produced by submerging the hands or feet in water
at an optimum temperature (49) of approximately
15 C (range, 12 to 18 C) for 10 to 15 min, but failure
to produce the characteristic manifestations of the
attack does not exclude the diagnosis (1, 49, 60).
However, failure to produce blanching by this
means plus additional preliminary or simultaneous
general body cooling (e.g., a cold shower) is reliable
evidence that Raynaud's syndrome does not exist.
Water of icy coldness tends to produce a red reaction
even in patients with Raynaud's syndrome. It is
important to differentiate Raynaud's phenomenon
from cold allergy which produces an erythematous
pruritic edema but not true blanching (60).
The vascular reactions and color changes of Ray-
naud's syndrome tend to occur segmentally and
bilaterally in the digits, generally terminating at the
interphalangeal or metacarpophalangeal articula-
tions. Although there is a distinct tendency for the
syndrome to occur bilaterally and symmetrically
some asymmetry in degree of involvement of either
hand is not uncommon. Involvement of an extremity
characteristically does not extend proximal to the
metacarpo(tarso)phalangeal joint. The feet may be
involved, but the hands are involved much more
frequently. Involvement of other parts of the body
is occasionally seen, especially such acral parts as
the ear lobes, cheeks, tongue, and the tip of the nose.
The localization of the vascular lesions to the
hands and feet is of interest. When only a single
phalanx is involved, it is the distal one; when two
phalanges are involved, they are the distal two. All
three phalanges of a digit or several digits of either
hand may be involved. In a single digit the direction
of progression of changes during an attack is from
distal to proximal. The second or fifth or both digits
are involved most often. When only the very tip of
the distal phalanx is involved, this suggests changes
in vessels smaller than the digital arteries. Localiza-
tion of Raynaud's reaction, which rarely occurs in
parts of the body other than the digits, is apparently
PERIPHERAL VASCt LAR DISEASES
[227
due to vascular changes in the reticular-perpendicular
arterioles in the skin of these sites.
Manifestations other than the typical color changes
may be present. During the pallid or cyanotic crises,
digital paresthesias may be present. During the
erythematous phase there may be increased warmth
and a painful throbbing sensation in the affected
digits. After long-standing or severe disease, ulcera-
tion, necrosis, edema, or subungual and paronychial
infection may develop (fig. 3). Ulcers usually occur
on the digital tips and these may be quite painful.
When they heal, they typically leave small pitted
"stellate" scars. Another change that occasionally
develops in long-standing or severe disease is sclero-
dactylia. The digits show a tight, tough, inelastic,
fibrotic, and contracted skin with areas of hyper-
pigmentation and hypopigmentation. This change
must be differentiated from a) acrosclerosis, in which
similar changes involve not onlv the digits but also
the face and neck; and b) scleroderma, in which
fibrotic changes are generalized, even involving
multiple visceral organ systems.
Practically nothing is known of the earliest patho-
logic changes in Raynaud's disease. This is due in
large part to the lack of biopsy specimens obtained
during the early stages of the disease. Physicians
and investigators have been reluctant to obtain
biopsies in these patients. Simple digital biopsy
methods and other means are now available whereby
early changes may be observed, both by light and
electron microscopy not only in Raynaud's syndrome
but also in other peripheral vascular disorders (32,
73, 75). Based on light microscopy, it has been
assumed that pathologic changes are absent in early
Raynaud's disease. More sensitive methods, such as
electron microscopy, may alter this impression. In
FIG. 3. Primary Raynaud's disease with trophic changes
and early sclerodactylia.
the later stages of the disease, intimal thickening of
the digital arteries is almost always present. In still
more advanced stages the internal elastic membranes
split and there is endarteritis obliterans with thrombi
in various stages of recanalization. The latter changes
are particularly frequent in association with
ulceration.
Briefly, and in general, the mechanisms for Ray-
naud's phenomenon are as follows: Pallor is due to
digital arterial constrictive crises to the point of, or
almost to the point of, complete occlusion with
resultant absence or near absence of digital arterial
blood flow. Capillary pressure drops to about 5 to 10
mm Hg (60). Cyanosis occurs when digital arterial
constrictive crises are slightly less severe, allowing
some blood to flow. In this situation the slow rate
of flow fosters an increased dissociation of ox\ gen
from hemoglobin with resulting local cyanosis.
During the recovery phase from the vasoconstriction,
erythema frequently ensues due to reactive hyperemia.
These physiologic changes are fairly well accepted
as the vascular reactions responsible for the typical
digital color changes. Controversy still exists, how-
ever, regarding the location and nature of the under-
lying factors responsible for initiating vascular reac-
tions. Raynaud felt that the primary factor was a
derangement of the nervous system (84). Adson &
Brown (2) also considered the basic fault in early
Raynaud's disease to exist entirely in the vasomotor
nerves, since complete relief of Raynaud's reaction
occurred in many patients following sympathetic
ganglionectomy.
Lewis (48, 49), however, maintained that the
basic fault is in the digital arteries themselves and
that the defect consists of an abnormal sensitivity
of the arteries to direct stimuli, particularly to cold.
In Lewis' own words (49), "The central fact is tran-
sient loss of circulation to the digits occurring on
exposure to cold. I have shown that this spasmodic
loss of circulation is due to closure of the digital
arteries, and that, irrespective of its nature, the fault
lies in these vessels; the closure does not involve
arteries of much larger size, neither does it include
arterioles or veins. But since the attack is induced by
exposure to cold, to which all vessels normally
respond, a general reduction of their size happens.
In most of the vessels the degree of closure can be
regarded as no more than normal. In the small
arteries only is the response to cold manifestly abnor-
mal; these are in a state rendering them particularly
liable to shut on direct exposure to cold. In sensitive
cases, the blood-flow to a single finger can be arrested
I22t
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
at will by cooling this finger alone, or even by cooling
a short stretch of it; for the digital arteries in their
whole length possess this liability to closure. The state
of closure once established can be released by warm-
ins; the hands; and this can also be affected in the
arteries of separate fingers, or even in the arteries
at the base of a finger, by warming the finger or its
base separately.
"It has been indicated already that, when a normal
subject is exposed to cold, arteries like the digital
narrow under two influences; they constrict as a
direct reaction to cold, and because vasomotor
nervous tone increases. These same two factors
operate in the fingers of the cases we are discussing,
under conditions of general cooling. But, because
in these cases there is an abnormality, the vessels
do what they will not do in normal subjects, they
close to obliteration. The evidence proves that the
abnormal element is local, and not, as formerly
thought, in the response of the nervous system. Thus,
if vasomotor tone is deliberately reduced by warming
the subject's body, immersion of the hand of a sus-
ceptible subject in cold water will still induce the
attack; but if the hand is kept warm, an increase of
vasomotor tone, induced by cooling the body, will
not provoke the attack. Again, if the circulation to
the fingers of such a patient has become arrested by
general exposure to cold, local destruction of vaso-
motor tone by nerve anaesthetisation does not bring
instant release of blood-flow, which would happen
inevitably if vasomotor tone were alone responsible:
it brings delayed release, or the release fails. Likewise,
as experience has shown, destruction of the sympa-
thetic nerve supply to the limb by surgical inter-
vention does not cure the malady: for it frequently
happens that patients so treated continue subse-
quently to display attacks on exposure to cold; and
after sympathectomy the local susceptibility can
always be demonstrated by special tests in sensitive
cases and this is so even when the sympathectomy is
preganglionic. The local abnormality is the reason
for this, for it remains unchanged.
"Although the facts show that the fault is not in
the nervous system, that is not to say that the nervous
system plays no part in the attacks. If under the
direct influence of cold the arterial channels of a
hand become unusually narrowed, but not quite
obliterated, then subsequent cooling of the trunk,
or an emotional disturbance, or a painful stimulus,
by normally increasing vasomotor tone, will cause
the vessels to close completely and thus determine
an attack. It is this kind of event that has been
misinterpreted in the past, and has given support to
the wrong idea that the vasomotor nervous system is
primarily at fault. Further it will be apparent that
anything reducing or abolishing vasomotor tone will
on occasion bring an attack to an end, and continuing
as an influence will tend to prevent the recurrence of
attacks. This is the basis upon which the modern
treatment by sympathectomy rests; its successful
results are due, not to interference with the passage
of abnormal nervous impulses, but to the destruction
of normal vasomotor tone."
Lewis emphasized that if one finger of a subject
with the disease is immersed in cold water, the attack
is frequently confined to this finger. He felt that such
a sharply localized response could not be explained
on the basis of a nervous reflex.
More recent publications (8, 52, 62) presented evi-
dence in support of Lewis' theory of a "local fault" in
the blood vessels. By plethysmography and thermom-
etry it was demonstrated that patients with Ray-
naud's phenomenon have an increased sensitivity
to cold as compared to normal subjects and that this
state persists even after successful sympathetic dener-
vation. These studies were not meant to imply that
sympathectomy is of no benefit to the patients. When
the patient's peripheral vessels are maximally dilated
by the procedure and heated by the warm blood
flowing through them, a decrease of the vessel tem-
perature to the critical level is not as easily produced.
Further, cooling of the vessels by vasoconstriction
can no longer be induced reflexly from emotional
disturbances, pain, or body chilling.
Whether or not ether vessels besides the digital
arteries participate actively in the Raynaud's reac-
tion has been debated. Naide & Sayen (66) con-
sidered that arterial spasm alone cannot explain the
entire clinical picture. They presented evidence,
though not conclusive, that spasm of the digital veins,
as well as the digital arteries, exists. It was based
largely on observations that in some patients with
Raynaud's disease the digits began to appear puffed
and cyanotic before blanching occurred. The authors
considered this to indicate venoconstriction prior
to arterial constriction.
Capillarioscopy has been of value in detecting
vascular change in the various reactions of Raynaud's
disease (3, 18). During the stage of pallor, no blood
enters the capillaries of the involved digits. During
the stage of cyanosis, more than the usual number of
of capillaries are engorged with blood, and many
are greatly dilated. They are filled with stagnant
blood. Venules may also be dilated during this stage,
PERIPHERAL VASCULAR DISEASES
1229
and there may be reflux of blood from the venules
into the capillaries. The transient cyanotic reaction
in Raynaud's disease is similar to the more nearly
permanent cyanosis of acrocyanosis in which second-
ary dilatation of capillaries and venules results from
arteriolar spasm (3). During the erythematous phase,
capillary pulsations may be detected. It should be
recalled that in Raynaud's disease the vessels supply-
ing the hand (radial and ulnar) are normal and
continue to pulsate normally during the crises.
The main controversy remains concerning the
initial basic defect in Raynaud's disease, that is,
whether or not it is in the nervous system or in the
vessel wall itself. Certainly in advanced stages the
easily demonstrable intimal thickening and throm-
bosis of the vessels contribute significantly in reducing
blood flow. It is probable that a vicious cycle is
induced whereby repeated vasospastic attacks cause
increasing injury and structural changes in the
digital vessels which then become more vulnerable
to vasospastic influences (60). Nevertheless, the
important pathogenic factors in the early stages are
still not known and very little recent definitive
research in this area has been reported.
Some recent studies, however, are of interest.
Using chromatography and biologic assay, Peacock
(70) determined the concentrations of epinephrine
and norepinephrine in the peripheral venous blood
collected at the wrist in a group of normal subjects
and in a group of patients with primary Raynaud's
disease. He found that under warm resting environ-
mental conditions, the Raynaud's patients showed a
significantly higher blood level of these amines than
did the normal subjects. Following sympathetic
nervous stimulation by cold, the patients with
Raynaud's disease had an increase primarily of the
norepinephrine fraction. This increase varied directly
with the clinical severity of the disease. Peacock con-
sidered the high concentrations of these amines to
be due to an abnormality in metabolism of these
substances. It was noted that in Raynaud's disease
the average digital cutaneous temperature in a room
temperature of 20 C ± 5 was 22.3 C compared with
30.2 C for the normal subjects. Similar differences
had been reported for environmental temperatures
as high as 25 C. It was reasoned that over this range
of temperature, due to precooling of blood by counter-
current flow mechanisms, the intraluminal temper-
ature of the blood in the digital arteries of patients
with Raynaud's disease was probably considerably
lower than that seen in normal control subjects. Thus,
it was concluded that the lower temperature inhibited
enzyme systems which inactivate epinephrine and
norepinephrine and that this was responsible for the
higher concentrations of the vasoconstrictor sub-
stances and the intense peripheral vasoconstriction.
In this respect monoamine oxidase activity of digital
arteries of two patients with Raynaud's disease was
absent, whereas that of two normal subjects was found
to be 552 ii\ 02 per g per hour.
These studies are interesting but many questions
remain unanswered. For example, the effect of
reduced blood flow per se on the concentrations of
amines in the venous blood draining these areas is
not known. Concentrations may be greater but the
total amount may be the same or less. Further, with
respect to these amines, the relative contributions of
a) increased amount in stores, b) increased release
from stores, c) decreased destruction, d) impeded
physical removal, and e ) increased vascular sensitivity-
are also unknown. Furthermore, the role played by
reduced formation of vasodilator metabolites from
cooled tissues [as proposed by Freeman (27, 28) and
later by Perkins et al. (71)] in the pathophysiology of
Raynaud's disease needs investigation.
Recently, Mendlowitz & Naftchi (61) have reported
observations on digital blood pressure (Gaertner
capsule) and digital blood flow (calorimetry) in 20
patients with primary or idiopathic Raynaud's
disease. The patients were studied at rest, under
standardized conditions, before and after vasodilata-
tion (reflex vasodilatation) and after vasoconstriction
produced by infused /-norepinephrine. After appro-
priate calculations, they noted that their patients
fell into two groups: /) those with digital vascular
obstruction and normal vasomotor tone, and 2)
those without obstruction but with heightened vaso-
motor tone. Neither group showed increased sensitiv-
ity to norepinephrine. Thus, the authors suggested
that the digital vasospastic crises in Raynaud's
disease could be produced either by vascular obstruc-
tion acted upon by normal vasomotor tone, or by
heightened vasomotor tone produced by increased
sympathetic neural discharge acting on otherwise
normal vessels.
The relationship of these latter findings to those
of Lewis (49) and the more recent ones of Peacock
(70) is not clear. With reference to the grouping
offered by Mendlowitz and Naftchi, it is possible,
but not proved, that Lewis was studying patients in
the group with vascular obstruction and normal
vasomotor tone, whereas Peacock was studying
patients in the group without obstruction but with
230
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
heightened tone. Thus, there are many problems in
Raynaud's disease which await clarification.
Secondary Raynaud's syndrome or phenomenon. A glance
at the classification of peripheral vascular disease
(see Appendix to this chapter) reveals that Ray-
naud's phenomenon occurs as a secondary mani-
festation in a number of disease states. The proposed
mechanisms invoked in these entities are complex
and a description of these is beyond the scope of the
present discussion. Further, little definitive work has
been done to determine the true physiologic mecha-
nisms responsible for the relationship between these
diseases and secondary Raynaud's syndrome. In
general, what has been said regarding primary
Raynaud's phenomenon may be applied to the
secondary phenomenon. There are a lew known
differences. In the secondary state the associated
manifestations of the primary predisposing disease
are evident. With obliterative arteriosclerotic endarte-
ritis, Buerger's disease, or other obliterative arterial
states, the degree of gangrene associated with second-
ary Raynaud's phenomenon may be considerably
more than in primarv Raynaud's because of the
underlying obliterative disease. In the secondary
state, exposure to cold or emotional stress may or
may not precipitate Raynaud's phenomenon. Lastly,
secondary Raynaud's phenomenon is frequently
neither bilateral nor symmetrical.
It is often suggested that the observation of Ray-
naud's phenomenon occurring in diseases of the
nervous system refutes Lewis' idea of a local arterial
defect and that primary Raynaud's disease is a
disease of neural origin. Even though this problem
is unsettled, it must be remembered that differences
between primary and secondary Raynaud's pheno-
menon do exist. Further, Raynaud's phenomenon
from neural disease still might be due to excessive
sympathetic nervous activity in the presence of
"normal" digital vessels, whereas primary Raynaud's
disease might be due to a local vessel defect in the
presence of normal sympathetic nerve activity.
acrocyanosis. Acrocyanosis is a disorder character-
ized bv a persistent cyanotic rubor to the skin of the
hands and feet and other acral portions of the bodv
associated with a reduced skin temperature. The term
"acrocyanosis" was first applied by Crocq (16) in
iHi)ti, and Cassirer's description (12) in 191 2 helped
clarify this disorder as a distinct clinical entity. Lewis
& Landis (50) initiated basic investigations into the
pathophysiology of this disease, but only a few con-
tributions to its mechanisms have been available
since.
The etiology of acrocyanosis is unknown. In his
large series of several hundred patients, Stern (94)
was unable to detect any constant precursor or accom-
paniment of the disorder other than cold, with fre-
quent moderate cooling of affected parts, and possibly
inactivity, the latter because of the occurrence in
lethargic types of mental disorders. It is much more
frequent in females and usually present in young or
middle-aged individuals. There is frequently a family
history of the disorder. It has been described as being
rare in the general population but rather common
among the inmates of mental institutions.
The patient usually visits the doctor for cosmetic
reasons, complaining of almost constant coldness and
bluish discoloration ol the fingers, hands, nose,
cheeks, chin, and pinna of several years' duration
(fig. 4). The toes and feet may be involved, but usually
to a lesser degree than the hands. The changes,
though present during the summer, are usually more
marked in the winter. The affected parts are usually
deeply cyanotic when cold, and bright red when they
are very cold or when they are warm. Frequently
they present a mixture of the two colors, red and
bluish-purple. The deep reddish color (as opposed to
cyanosis) produced by a very cold temperature (less
than 10 C) is due to arteriolar injury and dilatation
n
*
fig. 4. Acrocyanosis. [Reprinted with the permission of
H. K. Lewis & Company, Ltd., London (94).]
PERIPHERAL VASCULAR DISEASES
1231
and inhibition of oxygen dissociation from hemo-
globin.
The palms are often sweaty. The hands are usually
much colder than normal during exposure to a com-
fortable temperature but warm readily in a hot room.
The disease varies considerably in degree from very
mild to severe.
In the past, acrocyanosis has been confused with
Raynaud's disease but many differentiating features
are apparent. In acrocyanosis the color changes are
persistent rather than episodic. The changes are not
limited to the digits but include the entire hand and
foot, though they rarely extend proximally to the
wrist or ankle areas. There may be associated livedo
reticularis or pernio involving more proximal areas
of the extremities. There are usually no episodes of
blanching, sclerodactylia does not develop, and areas
of ulceration and gangrene are generally absent.
Swelling however may occur, particularly in cold
weather, and occasionally localized areas may become
tender or painful. Although spontaneous ulceration is
extremely rare, traumatic lesions in affected areas
may become infected and heal slowly. Palmar clammi-
ness is a well-known feature of acrocyanosis and
differs from the dry skin of Raynaud's disease which
appears when the local circulation ceases (94). In
true acrocyanosis, examination of peripheral arteries
reveals no evidence of occlusive organic arterial dis-
ease. The dependent cyanosis frequently present in
the feet of patients with occlusive arterial disease
should not be classified as acrocyanosis.
As in early Raynaud's disease, very little is known
of the pathology of acrocyanosis. Stern (94) studied
sections from the dorsal skin of the hands and feet of
of 12 patients with acrocyanosis. It was found that
the medial coats of nearlv all arterioles were thick-
ened. Local edema and dermal fibrosis frequently
were present in association with considerable dilata-
tion of the superficial capillaries with formation of
new capillaries. Others have described distention of
the venules and venous limb of the capillaries and
have noted large capillary loops occurring in increased
numbers in the nail bed (3, 6, 41). In fact, this tend-
ency to dilatation of the venous side of the circulation
with a marked decrease in venomotor tone is a char-
acteristic feature of acrocyanosis.
Little work is available on the vascular mechanisms
responsible for acrocyanosis. Most evidence points to
excessive arteriolar constriction which occurs at
ordinary environmental temperatures and which is
increased by cooling. The arteriovenous anastomoses
are also probably constricted (60). This constriction
is followed by secondary dilatation of capillaries and
venules with stasis in the minute vessels of this skin.
There is loss of capillary and venular tone (6, 60),
thought to be due in large part to anoxia. Stasis allows
increased formation of reduced hemoglobin and the
associated deep cyanosis, the blue color being due to
increased amounts of reduced hemoglobin and the
deep character of the color being due to the engorge-
ment of the vessels. That venous obstruction is not a
significant factor has been pointed out by Lewis &
Landis (50) from the simple observation that cyanosis
is abolished by venous drainage produced by eleva-
tion of the involved part. The color of the cyanosed
'.kin is not uniform, since it frequently contains
bright red areas (cinnabar red spots) (94) and occa-
sionally changing reticular areas of pink color due
probably to temporary relatively normal rate of blood
through the perpendicular-arteriole reticuiar-capil-
lary network (22). It has been noted that acrocyanosis
is less pronounced in the presence of hypertension,
since the latter tends to produce more normal circu-
lation in spite of the dermal arteriolar spasm (60).
In acrocyanosis, the white spot produced by ex-
ternal pressure on the cyanosed part disappears spon-
taneously in a very characteristic fashion (94). The
color returns from the periphery and not from below
as in normal skin.
Comments by Lewis concerning the mechanism of
acrocyanosis are worth quoting (49) :
"The minute vessels of the skin are verv dilated, as
is evident equally Irom macroscopic and microscopic
examination. But the temperature of the hands and
other tests show the blood-flow to the skin to be
reduced greatly. The veins though contracted l>\ cold
are not occluded. The pulses in the main arteries are
normal. The constriction is in the small arteries or
arterioles of the skin. If the hand at the time is
cyanosed and a small part of it is warmed, the latter
soon becomes sharply defined as a bright red area;
similarly if a little histamine (1 in 3000) is pricked
into the skin, the skin reddens locally and its tempera-
ture rises. This is in contrast to the cyanosed skin in the
attack of Raynaud's disease, where the obstruction
lies in the main digital arteries, and in which redden-
ing of the skin does not occur in similar tests, but only
after these arteries open. In acrocyanosis all the
arteries and arterioles are capable of opening widely;
gross structural impediment is not present in any of
their channels."
Again, as in Raynaud's disease, whether the basic
underlying disorder lies in the sympathetic nervous
system or in the vessels themselves is unknown. As
1232
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
FIG. 5. Livedo reticularis (lower portions of both
legs of patient lying supine).
Lewis (49) noted, however, the arterioles in affected
areas are in an unusually high state of tone. He was of
the opinion that this is due to a fault in the vessels
themselves, since anesthetizing the appropriate
nerves does not result in prompt relief of the arteriolar
spasm, as it would if the mechanism were neural in
origin. In contrast, observations by Day & Klingman
(17) were interpreted as showing predominant sym-
pathetic nervous influences as the basic mechanism.
They noted that during sleep the cyanosis and cold
skin are relieved and replaced by warm, red skin.
No definitive studies are available in acrocyanosis
concerning the significance of tissue catecholamines
and other vasoactive substances. Nevertheless, com-
ments made above on Raynaud's disease in this
regard might be equally applicable here.
Since the course of acrocyanosis is relatively benign
and complications are few, sympathectomy has rarely
been indicated in its treatment. Because of this, con-
trolled studies on the effects of sympathectomy are
unavailable. In severe cases, however, sympathectomy
may be of value, especially when there is an associated
hyperhidrosis. The usual protection from cold or
sudden and marked decline in temperature is indi-
cated. The patient should keep warm and dry with
serious attention being given to his general state of
health.
livedo reticularis. Livedo reticularis is character-
ized by a prominent mottled, reticular, or blotchy
reddish-blue discoloration of the skin of the extremi-
ties (fig. 5). Between the reticular discolorations, the
skin presents a more normal but pale appearance.
Kaposi was probably the first to use the term
"livedo reticularis" (3). The etiology of this disease
is unknown. That it may represent a congenital
anomaly of blood vessels has been suggested by
some (85, 103). Some are of the opinion that there
is some inherent vascular instability in the back-
ground of most patients (3). In one series, 30 per
cent of the patients had associated hypertension and
50 per cent demonstrated marked nervous instability
(3, 5). Livedo reticularis is more frequent in females
and usually appears before the age of 40.
The disorder usually involves the skin of the legs
and feet in greatest severity, but it frequently also
involves the arms and hands. Occasionally, the thighs
and the lower part of the trunk may be affected.
There is a distinct tendency for the disease to occur
bilaterally and symmetrically.
The characteristic color changes are usually in-
tensified on exposure to cold and tend to be alleviated
on exposure to a warm environment. Patients may
complain of numbness, tingling, coldness, or aching
over the involved legs and feet. Ulcerations in livedo
reticularis are not frequent but they do occur. Ulcers
usually begin as an intensification of change in
areas of marked cyanosis, usually over the medial
lower one-third of the leg. These lesions may be very
painful and slow to heal. Ulcers in some patients
seem to be precipitated by cold weather, whereas in
others warm weather seems to be important in their
formation (23).
The pathophysiology and clinical findings in
livedo reticularis have been the subject of several
reports (20, 48, 85, 103); the most recent one of sig-
nificance is by Feldaker el al. (23). The latter authors,
following the suggestion of Williams & Goodman
(103), preferred to classify livedo reticularis into
three groups: /) cutis marmorata, 2) idiopathic
PERIPHERAL VASCULAR DISEASES
I233
livedo reticularis (primary livedo reticularis), and
3) symptomatic livedo reticularis (secondary livedo
reticularis).
Cutis marmorata refers to a state characterized
by transient reticular discoloration producing a
marble (hence the term "marmorata") pattern to
the skin which appears on exposure to cold but, un-
like the other types of livedo, it is not permanent and
disappears with warmth. It is considered that in this
state there is no organic pathologic alteration in the
peripheral circulation but rather that the disturb-
ance is a vasomotor phenomenon. It has been noted
to be frequent in infants and young girls and may
disappear as they grow older.
In idiopathic livedo reticularis (primary livedo
reticularis) the reticular discoloration is relatively
permanent and persists to some degree regardless
of temperature changes. As noted before, however,
the degree of discoloration is accentuated by exposure
to cold. There may be minimal to no organic changes
in the vessels except increased number and dilatation
of capillaries in the livid areas. Feldaker et al. (23)
also noted in these areas varying degrees of endar-
teritis and endophlebitis of the smaller vessels. At
times there is occlusion, periarteritis and periphlebitis
and occasionally, thickening of the walls of arterioles
in the dermis and subcutaneous tissue.
Symptomatic livedo reticularis (secondary livedo
reticularis) is the form of the disorder associated
with or secondary to other dermal, vascular, or
systemic diseases. These have been outlined in the
accompanying classification of peripheral vascular
diseases (see Appendix to this chapter).
The large arteries such as the dorsalis pedis, pos-
terior tibial, and popliteal are not involved by occlu-
sive disease and likewise venous insufficiency is not a
factor in livedo reticularis. Digital blood flow after
interruption of sympathetic nerve supply is usually
normal (60). Feldaker el al. (23) have recently
summarized the probable pathophysiology of this
disease. The perpendicular arterioles, supplying the
skin from below, and the central zone capillary
arborizations have a slightly greater tone and faster
linear rate of blood flow than the peripheral capil-
laries. Either because of organic change (as described
earlier) or vasospasm of arteries and arterioles of the
skin or both, capillary atony and slowing of blood in
peripheral capillaries are further increased, resulting
in a livedo reticularis pattern in annular rings about
central paler areas. Cold causes an increased vaso-
constriction of the arteries and arterioles, resulting
in an intensification of the livedo. When the periph-
eral capillaries are only temporarily atonic and di-
lated, and the arteriolar supply is only temporarily
reduced, the transient cutis marmorata results;
but if the changes are more or less permanent, then
true livedo reticularis is produced. On elevation of
the affected parts, the livedo decreases if the venules
draining the areas are patent and can drain the
stagnant blood from the capillaries. Warmth and
sympathectomy reduce the spasm of the arteries
and arterioles and thus reduce the degree of dis-
coloration.
These observations and interpretations are at-
tractive. As in Raynaud's disease and acrocyanosis,
however, the basic factors underlying the vascular
disturbances and manifestations are unknown.
Whether or not the defect is primarily one of local
vessel fault or one of sympathetic nerve disturbance,
and whether or not localization of the disorder to
these sites is determined by congenital or acquired
mechanisms, are not known. Relief of livedo reticu-
laris and return to normal color has been reported
following sympathectomy and also following the
administration of acetyl-beta-methylcholine (22).
The problems surrounding supposition of a sympa-
thetic nerve disease as the basic disturbance are
essentially as discussed for Raynaud's disease and
acrocyanosis.
CAUSALGIA AND RELATED SYNDROMES. This is one of
the most confused areas of all in peripheral vascular
disease today. Definitions are poor; criteria for classi-
fication and diagnosis, variable; and terminology,
diffuse. The unifying characteristic of this group is
the development of a bizarre symptom complex
following some type of injury to an extremity. This
posttraumatic syndrome consists in general of pain,
paresthesia, trophic changes, edema, and evidence
of autonomic nervous system dysfunction. In this
group of diseases are included major causalgia,
minor causalgia, traumatic vasospasm, acute atrophy
of bone, Sudeck's atrophy, reflex nervous dystrophy,
traumatic angiospasm, posttraumatic painful osteo-
porosis, neurovasospastic phenomenon, chronic post-
traumatic edema, posttraumatic reflex dystrophy,
sympathetic dystrophy, neurovascular reflex dys-
trophy, atypical causalgia, posttraumatic spreading
neuralgia, reflex nervous atrophy, irritative nerve
lesions, sympathalgia, posttraumatic pain syndrome,
peripheral acute trophoneurosis, postamputation
syndrome, and traumatic neuralgia. All these terms
have been employed in reports in the literature, and
undoubtedly others have been used. Each term has
[234
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
served to focus not only on the outstanding manifes-
tations in each particular patient studied but also on
the particular interests, specializations, and orien-
tations of the various investigators. This type of
terminology usually implies gross confusion, and
such is the case. Although these syndromes have
elements of both vasodilator and vasoconstrictor
mechanisms, they are classified under the latter,
since these manifestations are the most classic ol the
disorders.
To discuss the manifestations of each of the dis-
orders listed above would be beyond the scope of this
presentation. Adequate descriptions and reference
sources may be found in other publications (i, 3,
19. 34. 44, 49, 53, 58, 82, 87, 89, 90, 104).
There is great overlap of the manifestations in the
syndromes listed. In general they may follow any
type of injury to an extremity whether minor or
severe. Some investigators feel that there is a con-
stitutional predisposition to the development of
these syndromes in certain individuals. Whether or
not psychogenic factors and a previous history of
vasomotor instability are important has been de-
bated.
The provoking injury usually, but not invariably,
involves nerves (especially the median or sciatic)
or tissues around joints (particularly the wrist or
ankle). Pain and vasomotor disturbances may occur
almost immediately after the injury, or be delayed
and develop gradually over the next several weeks.
Pain from the original injury with its associated
accentuation on movement, with resulting disuse,
may be important factors in the pathogenesis.
The outstanding characteristic of the causalgia
syndrome is burning superficial pain. Pain is usually-
referred distal to the site of original injury and
frequently involves the digits and the volar surfaces
of the hands and feet. Hyperesthesia is a common
associated complaint which may be localized to a
sensory nerve, but is frequently incomplete and
neither segmental nor somatic in distribution. Be-
cause of this, these patients are frequently considered
to be malingerers or "neurotic." Patients at times
go to extremes to protect their hyperesthetic extrem-
ities, avoiding many direct as well as indirect stimuli,
even loud noises. Frequently they obtain relief
from the application of moist cloths. Obviously,
since the disease is posttraumatic the symptoms are
usually unilateral. In later stages, however, vaso-
motor disturbances may spread to the opposite
extremity.
The vasomotor changes are of particular interest.
Initially, the affected extremity is usually somewhat
edematous, erythematous, dry, and warmer than its
unaffected counterpart. The blood vessels are dilated,
the rate of blood flow is increased, and local tempera-
ture and oscillometrically recorded pulsations are
increased. Later, the vasodilatation subsides and
vasospastic phenomena usually become prominent
and remain so during the chronic stage of the disease.
In this chronic stage, the skin is usually cold, hyper-
hidrotic, cyanotic, and atrophic. The limb may then
be especially sensitive to cold and secondary Ray-
naud's phenomenon may be observed.
Early X-ray study reveals a spotted, often cyst-
like, decalcification of the bones in the involved
part. This is especially true for the ankle and wrist
and the bones of the hands and feet (Sudeck's at-
rophy). It is considered that these changes in bone
occur much too early to be explained simply as
atrophy from disuse. Later in the disease, however,
osteoporosis may become diffuse and difficult to
differentiate from osteoporosis of disuse. There is
emphasis by the patient on immobilization and
disuse of the part because of pain and, therefore,
disuse may be a contributing factor.
The tendency among most investigators has been
to divide the syndrome into at least two subgroups:
namely, major causalgia and minor causalgia. In major
causalgia there is usually a history of a penetrating
wound in the region of a major nerve trunk of the
limb and the subsequent characteristic symptom is
that of severe burning superficial pain. In minor
causalgia the provoking trauma is frequently minor
in type and major peripheral nerve trunks are not
involved. Although there are evidences of vasomotor
dysfunction and trophic changes in both the major
and minor varieties, spotty osteoporosis and edema
have been much more frequently seen and severe
burning superficial pain less frequently seen in the
latter.
Successful treatment with marked or complete
relief of symptoms has been reported to follow intra-
arterial or orally administered sympatholytic drugs,
paravertebral sympathetic nerve blocks, and sym-
pathectomy (82). In fact, success with these measures
may be strong ancillary factors in substantiating
the diagnosis.
The pathogenesis of the causalgia syndromes is
unknown and it would be profitless to discuss the
many proposed theories. These are available in other
publications (1,3,19, 34.49. 53. 58, 82> 87, 89, 90, 104).
They are interesting and thought-provoking but
largely unfounded. The whole field is complex and
PERIPHERAL VASCULAR DISEASES
235
confusing, but it might he helpful to indicate inter-
esting factors which have pathogenic relationships
to the disease.
Though sometimes very minor, tissue trauma is a
regularly associated factor. Fractured bones with
injury to adjacent nerves, surgery, tight bandages
and dressings, automobile accidents, falls and the
like seem to produce the syndrome.
Afferent neural conducted impulses (possibly ab-
normally integrated, distributed or modulated) must
certainly be factors. Pain is perceived by the patient
and is frequently the outstanding symptom. That
this is mediated through regular sensory-type nerves
is probable, but debatable.
Efferent neural conducted impulses apparently
play a part. Most probably a large proportion is
mediated through the sympathetic nervous system.
As in the afferently conducted impulses, these may
be integrated, distributed, or modulated abnormally.
Vasomotor disturbances are paramount and relief
with sympathetic block or sympathectomy is frequent.
Afferent neural conduction and efferent neural
conduction imply, but do not prove, since they may
be dissociated phenomena, that some reflex arc is
involved in the disturbance. The level in the nervous
system at which this occurs and how it functions is
not clear. It could be an axon reflex, a short-circuiting
in a peripheral nerve trunk, or an arc in the spinal
cord at segmental or higher levels or even in the
vasomotor centers or higher. Furthermore, the in-
tegration, distribution, and control, as well as the
modulation of the frequency, intensity, and time
course of the action potential of the impulses may be
abnormal in causalgia. This phase of the patho-
physiology has been neglected and needs investigation.
Blood vessel reactions are apparent from the pre-
ceding discussions but exactly how they are induced
is unknown.
The keynote of the causalgia syndromes is that the
magnitude of the resulting physiologic and anatomic
manifestations are out of proportion to the magni-
tude of the provoking injury. This implies altered
responsiveness on the part of the body to trauma.
To integrate all the observed or apparent mani-
festations of the causalgia syndromes into one unified
theory is difficult at the present time. Lewis (49) had
interesting ideas concerning the mechanisms of pain
and vasodilatation in the syndrome. He referred to
evidence from Tinel showing that section of the nerve
distal to the responsible lesion may relieve the
causalgic pain when section proximal to the lesion
had alreadv failed to do so. He further noted that
when a normal cutaneous nerve or the posterior
nerve root is cut and its distal end excited electrically,
that the corresponding area of skin reddens and
becomes hotter than previously ("antidromic effect")
and a burning itching pain is produced. Lewis be-
lieved that this resulting vasodilatation in the skin
is produced by a local release of a histamine-like
substance. He thought that the substance released
affects overlapping nerve endings in the area. Thus,
analogous pain impulses in causalgia might be con-
veyed back along these intact paths as well as along
the injured nerve. Lewis thus concluded that the
erythema and heat stage of causalgia was an anti-
dromic effect produced by distal stimulation of the
injured nerve and that this was in accordance with
the observations of Tinel. Lewis' theory does not
explain all the findings, however, such as the vaso-
constrictor phenomena and the relief with inter-
ruption of the sympathetic nerve supply.
One theory which has recently been attractive to
many investigators serves in part to explain the pain
and its relief with sympathectomy (19). In general,
it might be conceived as follows: In a zone of nerve
injury, the insulating factors that normally keep
one nerve fiber from interfering with its neighbor are
defective. Thus, efferent impulses might cross-
stimulate afferent fibers resulting in sensory dis-
turbances and pain. Regarding the efferent im-
pulses, the autonomic fibers logically would be the
most offensive in the damaged nerve since these have
continuous vasotonic activity. Therefore, during
periods of increased vasomotor activity there would
be more cross stimulation in the injured nerve and
thus more pain. Increased pain might thereby result
in increased sympathetic nervous activity and thus
propagate a vicious cycle.
The preceding theory may adequately explain
changes occurring unilaterally in the injured limb,
but would be inadequate to explain the extension of
vasomotor dysfunction into the contralateral mem-
ber. Thus, some higher source of nervous dysfunction
might well be involved. Further, it is accepted by
some that chronic neural irritation, especially if
excessive, is capable in some way of changing the
normal behavior of the neurons within the central
nervous system and of eventually modifying the
pattern of excitation registered in the conscious
levels (87). This may be an expression of disturbance
in integration, distribution, and modulation of the
action potentials of the nerve impulses within the
central nervous system. There are occasional patients
with causalgia unrelieved by peripheral nerve sec-
[Q 36
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
tion, section of the posterior roots or sympathectomy.
It has been assumed that these patients represent
examples of "thalamic dysfunction."
Other theories involving higher centers are of
interest, but will not be expanded. For example, it
has been suggested that an irritative focus in the
extremity produces afferent impulses over sensory-
nerves to the spinal cord which results in continual
impulse discharges and nervous disturbance in this
zone. This results in stimulation of the elements of
the lateral and anterior horns, which then produce
the characteristic peripheral signs (53). A theory-
proposed by Leriche (44) involved reflex overactivity
of the central vasomotor center.
Some theories of note have suggested involvement
of afferent sympathetic nerve fibers. Of interest in
this regard are studies by Kuntz (42) demonstrating
afferent spinal nerve fibers which traverse the sym-
pathetic trunk and communicating rami. Stimulation
of appropriate nerves results in conduction of pain
impulses by these afferent fibers which appear to be
distributed chiefly in relation to blood vessels rather
than to the skin and muscles of the extremity. Whether
or not these phenomena function in the causalgia
svndromes is unknown, but relief of pain by sym-
pathectomy might be explained by such mecha-
miscellaneous states. Spasm of major arteries. Spasm
of a large artery is initiated by some type of trauma
in or near the artery. The initiating trauma is usually
a severe penetrating injury such as a gunshot wound,
but it may be provoked by contusing or crushing
injuries even though the artery itself is not directly-
involved in the injury. The spasm may be sufficient
to occlude the lumen completely. The spasm may be
limited to a small isolated discrete segment of the
artery or it may involve a long segmental length
including the orifices of many collateral arteries (3).
The exact mechanism of this type of spasm is not
clear but the ''myogenic" factor or the inherent
property of smooth muscle to contract when directly
traumatized appears to be paramount. It has been
shown experimentally that local segmental spasm
in large arteries can be produced by mechanical
trauma irrespective of the presence of the adventitia
or nerve supply (38). These facts are in accordance
with observations that this type of spasm usually
cannot be released with periarterial injections of
local anesthetics, sympathetic nerve interruption,
periarterial nerve stripping, or even by amputation
of the involved extremity above the site of the arterial
spasm (3, 13, 26). It has been shown, however, both
experimentally and clinically, that direct applica-
tion of a 2.5 per cent solution of papaverine to an
artery will relieve traumatic spasm in the majority
of instances (39). The mechanism of this response is
unknown.
Vasoconstrictor mechanisms in acute arterial occlusion.
The changes which occur after acute circulatory
arrest have been presented in a preceding section. It
is not the purpose of this discussion to present the
clinical signs and symptoms nor the pathogenesis of
events leading to acute circulatory arrest, which
include occlusive arterial disease, thrombosis, and
embolism. Arteriosclerosis, intravascular clotting,
and embolism are discussed in other chapters of this
Handbook. The purpose of the present discussion is to
indicate, briefly, concepts concerning possible vaso-
constrictor mechanisms which operate in acute
arterial occlusion.
The obvious factor in acute arterial occlusion is
acute impairment of blood flow through the arterial
lumen. Studies suggest, however, that this is not
necessarily the major cause of the resulting profound
ischemia associated with acute arterial occlusion,
since only mild to moderate degrees of ischemia may-
be produced when a comparable peripheral artery is
ligated. The implication then is that a superimposed
functional disturbance must be operative in patho-
logic occlusion.
It has been considered that reflex vasospasm of the
distal portion of the artery and the collateral arteries
is mediated in the efferent arc through the sym-
pathetic nervous system (3, 87). This has been the
basis for recommendation of prompt sympathetic
interruption in patients with acute arterial occlusion
(87). Experimental work has shown, however, that
the superimposed diminution in blood supply af-
fected by spasm and inadequate dilatation of col-
lateral arteries is temporary and spontaneously
disappears in a few hours (31, 64). Several investi-
gators (3) have postulated that the spasm in col-
lateral arteries, if prolonged, produces degenerative
changes in the intima of distal arteries and veins
which in turn provokes widespread vascular thrombo-
sis with the resultant organic obstruction to flow even
after the spasm disappears. This would account for
progression to complete irreversible circulatory arrest
in some patients. However, when spasm has not
been severe or prolonged, a satisfactory collateral
circulation may be established permitting the limb
to survive (3).
Some experimental studies on the mesenteric
PERIPHERAL VASCULAR DISEASES
I237
vascular responses in young dogs are of interest (56).
After acute arterial occlusion the artery developed
severe spasm whereas the vein exhibited a mild
degree of spasm. On release of the occlusion there
was a period of residual spasm in artery and vein.
Sympathectomy abolished both the arterial and
venous spasm during occlusion as well as after re-
lease. Papaverine administered during the arterial
occlusion had no effect on the arterial spasm, but
after release of the occlusion the residual spasm was
abolished.
Recent studies, however, have cast considerable
doubt on the significance of diffusely distributed
vasospastic phenomena in response to acute arterial
occlusion in man. Compensatory mechanisms in
response to sudden arterial occlusion have been the
subject of a recent report on clinical, pathologic,
and experimental observations by Wessler et al. (102).
They noted that three major important compensa-
tory phenomena follow sudden arterial occlusion,
namely, clot fragmentation, clot lysis, and function
of preformed inter-arterial collateral anastomoses.
The authors considered that clot fragmentation and
clot lysis, although not disproving the role of "spasm,"
provide an alternative to the concept of release of
spasm as an explanation for the occasionally witnessed
sudden relief of arterial insufficiency in some pa-
tients. The gradual enlargement of anastomotic
channels, bypassing complete obstructions, accounts
for the delayed and gradual improvement (even
with return of distal pulsatile flow) observed in some
patients weeks to months after the initial occlusion.
The authors further stated that unlike embolectomy,
blockage of autonomic nervous supply for the relief
of ischemia, secondary to arterial occlusion, has
neither a sound physiologic rationale nor satisfactory
clinical documentation of its efficacy (102). Based
on their own (102) and other studies (76) they found
little evidence that vessels in the ischemic zone are
in spasm in organic arterial insufficiency.
More recent observations by Hardy & Tibbs (33)
have further minimized the role of diffuse ''spasm" in
acute arterial thrombosis. These authors emphasized
that a healthy artery is normally in a state of con-
siderable elastic distention and that when occluded
the vessels distal to the occlusion become narrow from
"elastic recoil." Apparently this recoil has been the
basis for the erroneous diagnosis of diffuse arterial
"spasm." Patients are described (33) in whom the
distal arteries remained contracted and pulseless after
embolectomy, and in whom a residual "consecutive"
clot was found. When this residual clot was removed
completely by retrograde irrigation, the "spasm"
disappeared and pulsation returned.
Regardless of the above and other arguments, it
is impossible to state dogmatically whether or not
arterial "spasm" is significant in the pathophysiology
of the circulation in acute arterial occlusion. The
suggestion that a powerful vasoconstrictor substance
(possibly serotonin) is released from a fresh thrombus
and that it causes spasm of the affected vessel and
adjacent collaterals (25, 83, 101) needs further
study.
Vasoconstrictor mechanisms in chronic arterial occlusion.
This topic has caused considerable discussion, espe-
cially among surgeons who advocate sympathectomy
in the treatment of chronic arterial occlusive disease.
One major basis for this suggestion has been the
thesis that even if superimposed arterial spasm is not
of pathogenic importance, sympathectomy is of
benefit because it reduces normal arterial "tone"
causing arterial dilatation and fostering collateral
circulation. Particularly with respect to muscle
circulation, neither experimental nor clinical evidence
in man justifies pursuing this topic further.
Vasoconstrictor mechanisms in collagen diseases and
diseases of the fine blood vessels. These diseases and dis-
eases of "immune" mechanisms arc on the forefront
of medicine today. Much progress has been made in
understanding these conditions. In general, the
"collagen diseases" include lupus erythematosus,
scleroderma, dermatomyositis, periarteritis nodosa,
rheumatic fever, and rheumatoid arthritis. Also in-
cluded among these diseases are thrombotic thrombo-
cytopenic purpura, multiple forms of "vasculitis"
(see Appendix) and several other disease states. The
present day concept of the collagen diseases is that
they represent diseases which primarily involve con-
nective tissue structures. Since connective tissue is
ubiquitous the manifestations of these diseases are
protean. Regardless of terminology, there is no
reason to assume that the collagen fiber is the only
structure involved in these processes, but rather that
the disease is generalized including all connective
tissue constituents such as reticulum fibers, elastic
fibers, ground substance, and all related cells such
as fibroblasts, histiocytes, lymphoid elements, plasma
cells, and mast cells.
Even to attempt to discuss briefly the generalities
of this group of diseases would be beyond the scope
of this presentation. Numerous sources are available
in the literature. A brief review with particular
emphasis on cardiovascular manifestations has been
published (97). The blood vessels certainly are the
I238
HANDBOOK OF PHYSIOL! )( ;Y
CIRCULATION II
major shock organs of these
diseases. It is likely that the
vasculitis is responsible for a
great portion o f the mani-
festations of the various dis-
ease entities. Any type of
vessel may be involved, but
the fine blood vessels are
usually major participants.
Pathological changes in-
clude subendothelial fibrin-
oid degeneration, fibro-
blastic proliferation, intimal
thickening, varied inflam-
matory responses, and
thromboses.
From the pathophysio-
logic standpoint the vascular
manifestations probably re-
sult in greatest part from
organic structural change
and occlusion. That a func-
tional vasospastic compo-
nent may be superimposed,
however, has been proposed. It has been stated that
angiitis of the fine acral vessels is particularly apt
to give rise to vasoconstriction both reflexly and
by direct stimulation of the vessel network (21).
The organic changes plus the "spastic" factors lead
to agglutination of the cellular elements of blood in
the fine vessels in the distal reaches of the circulation,
which is an effective precursor of tissue necrosis. It
is impossible to quantitate the degree to which func-
tional vasospasm contributes to the pathogenesis of
these diseases. Vasospasm probably exists to some
degree as suggested (but not proved) from the fre-
quent association of secondary Raynaud's phenom-
enon (as high as 25% in some series).
Scleroderma {progressive systemic sclerosis). The terms
sclerodactylia, acrosclerosis, and scleroderma (pro-
gressive systemic sclerosis) have been introduced in
the section on Raynaud's phenomenon; the collagen
diseases in general have been discussed above. Be-
cause of its importance, scleroderma or progressive
systemic sclerosis is discussed further.
Scleroderma is a systemic disorder which involves
connective tissue of skin, muscles, tendons, fascia, and
all internal organs. Its outstanding manifestation is
a generalized increase in collagen fibers (97).
As with other collagen diseases, the etiology of
scleroderma is unknown. It affects both the white
and Negro races (74), is more common in females,
fig. 6. Severe scleroderma.
and usually occurs during early adult life and middle
age.
Many organs of the bodv may be involved. Thicken-
ing of the skin with tightening, increased rigidity,
and reduced distensibility, involving the face, ex-
tremities, and trunk produce a characteristic ap-
pearance (fig. 6). In the early stages of the disease
the skin may be edematous, but later it characteris-
tically becomes firm and nondistensible, with areas
of hyperpigmentation and hypopigmentation, and
the joints become stiff and contracted. Calcification
of tissues, absorption of the terminal phalanges,
atrophy of the fingertips, deformed nails, and cutane-
ous ulcers may occur. The face may offer a striking
appearance being tight, expressionless, and masklike
without wrinkles. The features are pinched, the nose
is pointed, and there is difficulty in smiling and open-
ing the mouth. Acral vasomotor disturbances such
as color changes, coldness, hyperhidrosis, and Ray-
naud's phenomenon are not infrequent.
Histologic sections through affected skin show that
the Malpighian layer is atrophic. The deep layers of
the skin show increased fibrosis which extends into
the subcutaneous tissues. It may extend into muscles
of the limbs and may bind the skin of the fingers to
bone. Blood vessels are entrapped in the dense fibrotic
change. The feet may be involved but the hands
show by far the most extensive change.
PERIPHERAL VASCULAR DISEASES
12 39
Involvement of the gastrointestinal tract is common,
especially the esophagus, frequently the small bowel
and occasionally the colon. This may produce
obvious functional changes in these organs. Joint
involvement may mimic rheumatoid arthritis and
skeletal muscle involvement with atrophy and fibrosis
may resemble dermatomyositis. Pulmonary involve-
ment is very common with peribronchial and inter-
stitial fibrosis and destruction of alveolar walls.
The vascular manifestations of the disease may be
widespread. There is thickening of vessel walls, peri-
vascular fibrosis, intimal proliferation, obliterative
vasculitis, thrombosis, fibrinoid necrosis, and cellular
infiltrations with polymorphonuclear and round
cells.
Usually smaller vessels are predominantly affected,
but lesions may be encountered in any vessel of the
body. The coronary, pulmonary, dermal, and renal
vascular beds are notable participants in this vascular
disease. Primary changes in the myocardium, apart
from the involvement of mvocardial vessels, are
frequent. There is interstitial fibrosis, myocardial
degeneration, endocardial, epicardial and peri-
cardial fibrosis, and ventricular dilatation and hyper-
trophy.
Obviously, clinical manifestations may be multiple
and varied. Renal involvement may cause albumi-
nuria, hematuria, and hypertension. Myocardial
disease may present any of the findings seen in
congestive heart failure. Pulmonary involvement
may be expressed as respiratory alveolar-capillary
block syndrome, fibrosis and emphysema, obstructive
and restrictive ventilatory dysfunction, hypoxia,
carbon dioxide retention, pulmonary hypertension,
polycythemia, cor pulmonale, and the like. Thoracic
involvement with the tight constricting skin may
produce hypoventilation and circulatory dysfunc-
tion.
All the above and other changes in scleroderma
are obviously of importance to a discussion of the
peripheral circulation because, potentially, they may
all contribute to disordered peripheral vascular
physiology. To quantitate their effects, however,
would be a difficult or impossible task. In addition
to these general effects, more direct factors in sclero-
derma may influence vascular function. Comments
made in the preceding section on vasculitis and
collagen disease, in general, are applicable here. As
noted, obliterative vascular change is probably the
major factor in vascular dysfunction, but a super-
imposed functional vasospastic element has been
suggested.
The frequent occurrence of Raynaud's phenome-
non in scleroderma is of interest. It may precede,
accompany, or follow the clinical disease onset. The
true relationship between the two is unknown. It
was long held by many that the vasomotor ab-
normality was of etiologic importance in the patho-
genesis of scleroderma. The majority of current
opinion, however, is that associated Raynaud's
phenomenon is a secondary manifestation of sclero-
derma due to the primary disturbance in blood
vessels and connective tissue. In this regard it is
analogous to secondary Raynaud's phenomenon
occurring in other collagen diseases.
Of the collagen diseases, however, scleroderma
presents an additional unique factor in that the
vessels are entrapped in a fibrotic ever-contracting,
poorly distensible environment. Thus, in addition to
intravascular occlusion there may be, in effect,
extravascular constriction or strangulation (59, 81,
92). Greatly increased pressures have been found
in the subcutaneous tissue in patients with sclero-
derma. Studies have shown tissue pressure values
up to 320 mm of water in patients with scleroderma,
whereas normal persons do not exceed 54 mm (92).
These added factors may be of significance in ex-
plaining the altered peripheral vascular function in
scleroderma. In this disease, reduced skin tempera-
ture and decreased digital pulsations are common.
That these changes are largely structural or organic
in origin is suggested by failure of these parameters
to return to normal after use of sympatholytic drugs
or inhibition of sympathetic tone through nerve
block or sympathectomy.
Vasoconstrictor mechanisms in acute thrombophlebitis.
The clinical manifestations of thrombophlebitis have
been adequately described in a number of publica-
tions (1, 3, 87, 104) and will not be repeated here.
Mechanisms of intravascular clotting are discussed
elsewhere in this volume.
Perhaps a few words regarding terminology are in
order. It has been common practice in the past for
clinicians to use the terms "thrombophlebitis" and
"phlebothrombosis" to represent two different and
distinct clinical syndromes (69). Thrombophlebitis
was considered to represent a rather intense inflam-
matory reaction in the involved vein with a more
firmly attached thrombus. Although it produces a
more dramatic local reaction, it was considered to be
less dangerous, since there was less likelihood for
emboli to break from the thrombus. In contrast,
phlebothrombosis, although bland with respect to
local reactions and manifestations, was considered to
1240
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
be the more lethal, since the associated loosely at-
tached thrombus was more predisposed to break
into emboli. Recent experimental and clinical evi-
dence suggests, however, that phlebothrombosis is
merely the silent forerunner of thrombophlebitis and
that the two diseases are stages of the same process
(H. 29, 55)-
Other terminology is dependent upon whether or
not superficial or deep veins are involved for which
the terms superficial and deep thrombophlebitis are
applied. The process of course is named according
to the particular vein or veins involved. When in-
fection is a predominant accompaniment, the term
"septic" or "suppurative" thrombophlebitis is ap-
plied.
An interesting rare variant of thrombophlebitis,
the Trousseau syndrome, or "migratory thrombo-
phlebitis," should perhaps be mentioned. This dis-
ease may involve either superficial or deep veins, in
one or more sites, either concurrently or separated by
considerable lengths of time. The importance of this
svndrome is related to the frequency with which
underlying serious disease is present, especially,
thromboangiitis obliterans, polycythemia vera, occult
carcinoma (usually of the stomach, pancreas, or
lung), or collagen disease (14).
The outstanding finding in typical acute superficial
thrombophlebitis is pain and tenderness over the
involved area, but embolic phenomena may occur.
Deep thrombophlebitis especially predisposes to
embolism. In deep thrombophlebitis the main
physiologic disturbance is obstruction to venous
blood flow. Pain of various types may be a feature
and is helpful in diagnosis, but edema is the most
objectively demonstrable physiologic alteration. The
degree of this disturbance is obviously dependent on
the size and location of the involved vein, the extent
of the thrombus, and the adequacy of collateral
circulation. When a large trunk such as the iliofemoral
or axillary vein is involved with a long thrombus also
compromising collateral flow, considerable obstruc-
tion to venous flow may occur and the increase in
venous pressure may be marked. This is in contrast
to simple ligation of a venous trunk in which col-
lateral circulation is not adversely affected.
Edema formation, secondary to venous occlusion
in thrombophlebitis, is much more than a simple
process of increased venous pressure with resultant
increase in mechanical transudation of fluid into
the tissue space, but this factor seems to be important.
The importance of associated lymphatic obstruction
in thrombophlebitis in the production of the edema
is debated and not yet clarified. Certainly, fibrotic
reactions in long-standing edema with a high protein
content of the extracellular fluid impairs lvmph flow.
Appropriate to the present discussion are studies
concerning possible vasomotor or sympathetic factors
in the pathogenesis of the manifestations in thrombo-
phlebitis. That arterial spasm may occur in some
patients with deep vein thrombophlebitis is not
denied, but whether or not it is a significant factor
in most instances is debatable. In some patients
during the acute stages of the disease spasm may he
so severe that pulsations in the large arteries disappear
for several hours. Several patients with actual ischemic
gangrene have been reported. The terms "phlegmasis
alba dolens" and "phlegmasia cerulea dolens" have
been used in some of these patients to describe the
associated color changes thought to be due to ac-
companying arterial spasm.
Some studies (68) suggest that vasoconstrictor
impulses are initiated by the thrombosed segment
of vein which produces spasm of both arterioles and
venules in the distant portions of the limb. Experi-
mental and clinical evidence has been presented in
favor of the idea that the thrombosed venous segment
initiates a detrimental spinal reflex arc with the
sympathetic nerves as its efferent arm. The induced
arterial, arteriolar, venous, and venular spasm was
said to propagate edema formation by increased
venous pressure with augmentation of filtration pres-
sure, by relative anoxia of capillary endothelium
with increased fluid transudation and by retarded
lymph flow secondary to reduced "pumping action"
from the arterial and arteriolar vessels in spasm.
Rather dramatic clinical improvement in patients
following paravertebral sympathetic blocks, both in
subsidence of pain as well as edema, has been re-
ported.
Subsequent experimental studies on mesenteric
vessels of young dogs support some of these concepts
(56). It was found that after acute occlusion of the
main stem vein the artery reflexly underwent spasm,
whereas the vein became moderately dilated. On
release of the occlusion there was a period of residual
arterial constriction, whereas the vein returned to its
preocclusion caliber. After sympathectomy, however,
it was noted that all reflex arterial constriction, as
well as the residual arterial constriction that followed
release of the occlusion, were abolished. It was noted
that during occlusion the vein became dilated to a
diameter exceeding that of the control. Although
PERIPHERAL VASCULAR DISEASES
I 24I
these experiments tend to lend support in part to
the concept of reflex arterial constriction, the latter
observation regarding venous dilatation is not in
accordance with the concept of venous or venular
constriction. This factor in thrombophlebitis had
been difficult to accept in the light of the intense
congestion and obvious distention of the small veins
(3)-
The theory that associated vasospasm in throm-
bophlebitis is a frequent and important pathophysio-
logic factor is intriguing but more definitive ex-
perimental work is required for confirmation and
general acceptance in clinical medicine.
Vasodilative Syndromes
Vasodilatation as an important vascular response
is seen in a number of physiologic states such as
thermoregulation, tissue inflammatory response, reac-
tive hyperemia, early stages of causalgia, and the like,
but the cardinal condition which concerns us here
is erythromelalgia.
erythromelalgia (erythermalgia). Early con-
tributors to the literature of this disease state were
Graves in 1834 (63), Mitchell in 1872 (63), Cassirer
in 1 91 2 (12), and May & Hillemand in 1924 (57).
Significant contributions after that time include the
work of Brown in 1932 (7), Lewis in 1933 (45), and
Smith & Allen in 1938 (91 ). Since then, little definitive
work has been done and published on erythro-
melalgia.
Erythromelalgia is a vasodilative syndrome char-
acterized by episodes of erythema, increased heat
and pain involving the hands and especially the feet.
It has been placed into primary and secondary
categories. "Primary erythromelalgia" occurs in
otherwise healthy individuals who manifest no de-
tectable evidence of organic disease of the nervous
or vascular systems. Analogous to Raynaud's phe-
nomenon, "secondary erythromelalgia" occurs in
association with or as a secondary symptom complex
of some other primary disease, such as hypertension,
occlusive organic vascular disease or polycythemia.
Gout, organic neurologic disease, frostbite, immersion
foot, trenchfoot, infectious diseases, and heavy metal
poisoning are also incriminated.
The mechanism of erythromelalgia is unknown and
the pathology has not been clarified. Symptoms
usually start in middle age or later and may affect
either sex. It is apparently rare in the Negro (74).
The clinical picture may be quite dramatic. The main
complaint is usually burning pain in the extremities,
especially in the feet and frequently in the hands.
Occasionally the disturbance may extend as high as
the knees or thighs. The patient usually complains
that the distress affects primarily the balls of the feet
and tips of the toes or corresponding parts of the hand.
It may last from a few minutes to several hours.
Usually the patient relates aggravation by dependency
of the part, by warmth, accentuation during summer
months, relief by cold and elevation of the part, and
lessened symptoms during winter. Attacks may be
precipitated by exercise which increases the warmth
of the skin. For unknown reasons, dry heat seems to
be more provocative than wet heat at the same tem-
perature. The discomfort may start as a "pricking"
paresthetic feeling then progress to a more typical
severe burning pain. During the subsidence of the
episode the pricking stage may again be noticed. In
the primary disease, neurological examination is
negative and examination of the peripheral vascula-
ture shows no evidence of occlusive arterial disease.
Trophic changes, ulceration, and gangrene are quite
rare, though some swelling may be evident in the
involved extremities.
What is known of the pathophysiology of this
syndrome is interesting. The most important part of
this syndrome is its intimate relationship with the
temperature of the skin. Lewis (45) has designated a
"critical point" in skin temperature at which this
syndrome may be produced in susceptible individuals.
It usually is around 32 C (range, 31 to 36 C). With
temperatures higher than this critical point, the
distress continues and with temperatures lower than
this point the distress disappears (3, 91). The tem-
perature at which the syndrome may be produced
varies with different patients and also to some degree
in different parts of the extremity in the same person,
but in the same person the range is usually within
±1 C.
Vasodilation per se seems to be an important
vascular factor in the production of the erythro-
melalgic crises. Increased blood flow is only an indirect
accompaniment. Thus, the syndrome may be pro-
duced by warming the extremity to the critical level,
and the symptoms continued even though blood
flow is arrested by an inflated constricting blood
pressure cuff. This is so, provided the skin tempera-
ture is maintained at levels equal to or greater than
the critical point.
As Lewis has pointed out, however, although
1242
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
vasodilation may be the essential vascular reaction,
it alone is not enough to explain the clinical state,
since an equivalent degree of vasodilatation may
occur in asymptomatic normal subjects in response
to warmth or exercise. Thus, he concluded that the
essential abnormality was a hypersensitive state of
the cutaneous pain fibers to heat or tension (by
dilated and engorged vessels), i.e., a "susceptible
state" of the skin. Thus, he suggested that a chemical
substance liberated into the skin served as the im-
mediate stimulus to the nerve endings and supported
this by the observation that the pain of the erythro-
melalgic skin was prolonged or intensified by arresting
the circulation to the part.
The essential vessels involved in the vasodilatation
are not definite, but it seems that all small vessels
participate. During a typical attack, the accompany-
ing features of vasodilation may be observed. In
addition to increased temperature of the skin, there
may be increased amplitude of arterial pulsation,
throbbing sensations, increased elimination of heat,
and increased content of oxygen in the returning
venous blood. The affected part assumes a deep,
dusky red color. The dusky color of the skin which
indicates a low oxygen content of small vessel blood
is of interest in light of the high oxygen content of
returning venous blood. An explanation offered for
this is arteriolar-venous shunting of some of the
peripheral flow through open arteriovenous anas-
tomoses (91 ).
Other observations are of interest (3, gi ). If the
skin temperature is slightly lower than the critical
point, distress may be induced by artificially in-
creasing the venous pressure by a proximal constrict-
ing blood pressure cuff inflated to a pressure less than
arterial pressure or by holding the part below heart
level to produce venous congestion. Similarly,
symptoms may be lessened if an extremity is elevated
even though the skin temperature remains unchanged.
In addition, direct pressure on the skin of the in-
volved area may cause relief.
A vasoconstrictor factor has been suspected in
some patients during intervals when they are free
of the burning distress. This is manifested as local
coldness and cyanosis or pallor of the skin during
these pain-free periods. Some patients have been
reported to suffer from Raynaud's phenomenon when
cold and erythromelalgia when warm.
In diagnosis of erythromelalgia, one must exclude
the burning sensations in the extremities of patients
who are suffering with peripheral neuritis, occlusive
arterial disease, and other states, but who do not
have erythromelalgia either primary or secondary.
In these patients the skin temperature is frequently
low (especially in organic vascular disease) or normal,
and the intimate relationship of distress to a critical
thermal level is not apparent. Further, it should be
noted that in organic vascular disease elevation fre-
quently accentuates symptoms and causes the in-
volved part to assume a pale and waxy color, whereas
in erythromelalgia color largely persists on elevation
and the symptoms may lie somewhat alleviated. In
establishing a diagnosis of erythromelalgia it is es-
sential to demonstrate that skin temperature and the
distress are related. For this purpose, the patient's
reaction and skin temperature are observed while
the temperature is raised either by reflex vasodilata-
tion or by direct application of heat.
One other interesting fact bears comment. Acetyl-
salicylic acid, in an oral dose of as little as 0.65 g,
may produce marked and persistent relief in erythro-
melalgia for as long as several days. The mechanism
of this response is unexplained, but it may be related
to effects on the local release of bradykinin.
In the pathophysiology of erythromelalgia, al-
though the vascular responses to temperature are in
part well established, the mechanisms that induce
these responses are unknown. Whether the basic
defect is in the nervous system or in the blood vessels
themselves is not clear. Regardless of the site, the
mediators involved need study. Furthermore, it is not
known whether or not the disturbance is congenital
or acquired. The possible contributions of vasoactive
humoral agents and the vasodilator nerves and the
mechanisms of their actions are also unknown. It may
be worthwhile, however, to direct attention to the
renewed interest in vasodepressor polypeptides, in
particular bradykinin. Depressor polypeptides have
been the subject of a recent review (35) and their
possible physiologic functions are covered elsewhere
in this volume. Further study might well incriminate
bradykinin as an important factor in the patho-
physiology of peripheral vascular disease, not only in
erythromelalgia, but in numerous other vasodilator
reactions.
Mechanisms in Othc
n ular Diseases
In the preceding discussions it was not possible
to survey the pathophysiologic mechanisms of many
other diseases of the peripheral vascular system. The
reader may obtain insight into the scope of the prob-
lem by referring to the appendix of this chapter.
PERIPHERAL VASCULAR DISEASES
I243
A P P E N D 1 X
CLASSIFICATION OF PERIPHERAL VASCULAR DISEASE
Diseases Affecting Primarily the Arteries and Arterioles
I. FUNCTIONAL CONDITIONS
A. Vasoconstrictor
1 . Raynaud's syndrome (primary Raynaud's disease)
2. Raynaud's syndrome (secondary)
a. Traumatic vasospastic syndrome
b. Neurovascular mechanisms
(1) Shoulder girdle syndromes
(a) Scalenous amicus
(b) Cervical rib
1 c 1 Costoclavicular
(d) Hyperabduction
(e) Thoracic outlet
(f) Malposition
(g) Pectoralis minor
1 2 I Spondylitis
; 1 Neuritis
c. Secondary to organic vascular disease
(1) Arteriosclerosis
1 j 1 Syphilitic arteritis
(3) Thromboangiitis obliterans
(4) Thrombotic or embolic occlusion
(5) Other occlusive arterial disease
d. Secondary to intoxications
( 1 ) Arsenic
j 1 Ergot
(3) Lead
(4) Nicotine
(5) Tobacco
e. Scleroderma and acrosclerosis
f. Miscellaneous mechanisms (e.g., rheumatoid
arthritis, lupus erythematosus, cold injury, and
other factors listed in Category 5 below)
3. Acrocyanosis
4. Livedo reticularis
a. Idiopathica
b. Svmptomatic livedo reticularis
(1) Questionable factors
Rickets, mongolism, various endocrin-
opathies, malnutrition, varicose veins,
other peripheral vascular diseases, infec-
tious diseases, intoxications, congenital
vascular defects, ectodermal abnormalities,
cirrhosis of the liver and other unusual
diseases, and neural disorders
(2) Probable factors
1 a ) Hypertension
(b) Nervousness and emotional instability
(c) Arsenic or lead poisoning (?)
(3) Purported causes
(a) Tuberculosis
(b) Syphilis
(c) Periarteritis nodosa and allergic cu-
taneous vasculitis
c. Cutis marmorata
5. Vasospasm, secondary to
a. Lesions of peripheral nerves
b. Lesions of brain and spinal cord including polio-
myelitis, prolapsed nucleus pulposus, hemiplegia,
tumors, multiple sclerosis, epileptic equivalent,
spinal bifida, spinal arthritis, lesions of midbrain
and internal capsule, etc.
c. Thrombophlebitis
d. Embolism
e. Thrombosis
f. Trauma
(1) Posttraumatic reflex sympathetic dystrophy
(2) Major causalgia
(3) Minor causalgia
(4) Sudeck's atrophy
(5) Posttraumatic osteoporosis
(6) "Vibrating-machine disease"
(7) Shoulder-hand syndrome
(8) Crutch arteritis
(9) Metabolic, adynamic and hormonal, includ-
ing rheumatoid arthritis, malnutrition and
asthenia, terminal rheumatic heart disease,
hypothyroidism, castration, menopause, hy-
poglycemia, Addison's disease, polycythemia,
cold hemagglutination and cryoglobulins,
leprosy, etc.
B. Vasodilator
1 . Erythromelalgia, primary
2. Erythromelalgia, secondary to
a. Polycythemia vera
b. Arteriosclerosis
c. Thromboangiitis obliterans
d. Hypertension
e. Miscellaneous factors: trauma, gout, frostbite,
immersion foot, trenchfoot, infectious diseases,
heavy metal poisoning, etc.
II. ORGANIC CONDITIONS (STRUCTURAL)
A. Occlusive [organic)
1. Arteriosclerosis
a. Atherosis
b. Atherosclerosis
c. Atherosclerosis obliterans
d. Medial ( Monckeberg's) arteriosclerosis
e. Arteriolosclerosis and hypertensive ischemia
f. Combined
2. Thromboangiitis obliterans
3. Arteritis (inflammatory diseases) and arteriolitis
a. Disseminated arteritis
b. Erythema induratum
c. Erythema nodosum
d. Nodular panniculitis
e. Nodular vasculitis
f. Temporal arteritis
g. Syndromes of necrotizing and/or allergic vas-
culitis
(1) General syndromes
(a) Purpura rheumatica
(b) Schonlein-Henoch syndrome
(c) Allergic angiitis
(d) Anaphylactoid purpura
(e) Necrotizing vasculitis
(f) Periarteritis nodosa of hypersensitivity
i244
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
(g) Wegener's granulomatosis and lethal
midline granuloma of the face
(2) Cutaneous syndromes
1 a ) Acute parapsoriasis
(b) Nodular allergid of Gougerot
(c) Allergic granulomatosis
d \llcrgic microbid
(e) Erythema elevatum diutinum
(f) Malignant atropic papulosus of Degos
4. Ergotism
5. Collagen diseases
a. Dermatomyositis
b. Disseminated lupus erythematosus
c. Periarteritis nodosa
d. Scleroderma
6. Hypertensive vascular disease
7. Arterial thrombosis
a. Associated with infectious diseases
b. Associated with blood dyscrasias
c. Secondary to trauma or compression (Volk-
mann's contracture, shoulder girdle syndrome)
d. Secondary to surgery
e. Associated with parturition
f. Associated with cardiac insufficiency
g. Associated with slowed blood stream
h. Associated with exposure to radiation
i. Idiopathic
8. Abscess of wall of artery
q. Cold injuries
a. Chilblains (pernio)
b. Cold urticaria
c. Frostbite
d. Immersion foot
e. Trenchfoot
10. Livedo reticularis
1 1 . Arterial embolism
a. Thrombus
b. Fat
c. Air
d. Bacterial
e. Neoplastic
f. Fungus
g. Inorganic substances
12. Ainhum
13. Blood agglutination
a. Dyscrasias (polycythemia, leukemia, thrombotic
thrombocytopenic purpura)
b. Cold reactions (possible cold agglutinins, cryo-
globulinemia)
c. Idiopathic (necrotizing acrocyanosis)
d. Massive venous thrombosis
B. Nonocclusive (organic)
1 . Aneurysm
a. Congenital
b. Syphilitic
c. Arteriosclerotic
d. Mycotic
e. Traumatic
f. Embolic
g. Idiopathic
2. Arteriovenous anastomosis (fistula)
a. Congenital
b. Traumatic
c. Secondary to malignancy
d. Secondary to bacterial infections
e. Secondary to fungus infections
3. Congenital anomalies of artery
4. Trauma of artery
5. Shoulder girdle syndromes
6. Rupture of artery
7 I.ffects of exposure to radiation
8. Nonocclusive arteritis
Diseases Primarily Affecting the Veins
I. FUNCTIONAL CONDITIONS
A. Spasms
II. ORGANIC CONDITIONS (STRUCTURAL)
A . Occlusive
1. Thrombophlebitis and venous thrombosis (phlebo-
thrombosis)
a. Primary
(1) Thromboangiitis obliterans
(2) Migratory thrombophlebitis
(3) Essential or idiopathic, local
b. Secondary to
( 1 ) Mechanical injury
(2) Muscular effort or strain
(3) Chemical injury
(4) Inflammatory or suppurative lesions (etio-
logic agent to be indicated)
(5) Infectious diseases
(6) Severe ischemia
( 7 ) Varices
(8) Blood dyscrasias
(a) Polycythemia
lb) Myelogenous leukemia
(c) Lymphatic leukemia
id) Pernicious anemia
(e) Disturbances of blood clotting mech-
anism
ill Other blood dyscrasias
(9) Cardiac insufficiency
(10) Carcinoma
2. Neoplastic invasion of vein
3. Venous compression by
a. Gravid uterus
b. Neoplasm
c. Aneurysm
d. Scar tissue
e. Scalenus amicus syndrome
f. Hyperabduction syndrome
g. Fractures
h. Dislocations
i. Increased intra-abdominal pressure (ascites)
4. Postphlebitic syndrome
B. Nonocclusive
1 . Varicose veins
a. Primary
b. Secondary to
( 1 ) Posture
(2) Occupation
(3) Clothing
PERIPHERAL VASCULAR DISEASES
I245
(4) Proximal obstructive lesions or pressure
(see II, A, 3)
(5) Thrombophlebitis
(6) Arteriovenous anastomosis
( 7 ) Hemangioma
(8) Congenital anomalies of veins
Arteriovenous anastomosis (fistula)
a. Congenital
b. Traumatic
c. Secondary to malignant lesions
d. Secondary to bacterial infections
e. Secondary to fungus infections
Aberrant position of vein
Hypoplasia of vein
Phlebectasia
Periphlebitis
Phlebosclerosis
Rupture of vein
Neoplasms of Blood Vessels
BENIG!>
r
A.
Hemangioma
1 . Cavernous
2.
3-
Capillary
Plexiform
4-
5-
6.
Sclerosing
Sturge-Parkes VVeber-Dimitr
Von Hippel-Lindau disease
syndrome
B.
7. Maffucci's syndrome
8. Multiple hemangiomas and
syndrome)
Glomus
chrondromata
(Kast's
C.
Telangiectasia
1 . Hereditary hemorrhagic
2.
Senile
3-
4-
5-
Simple
Spider
Papillary varices
II.
MALIGNANT
A. Ewing's sarcoma
B. Hemangioendothelioma
C. Hemangiosarcoma
D. Kaposi's sarcoma
(3) Fungus
(4) Erysipelis
b. Mechanical, chemical and physical
( 1 ) Abrasions
(2) Burns
(3) Chemical irritation
(4) Lacerations
(5) Trauma
(6) X-ray
c. Granulomata
(1) Lymphogranuloma
(2) Syphilis
(3) Tuberculosis
d. Postphlebitic
e. Surgery
(1) Removal of lymph nodes
(2) Removal of lymph vessels
f. Neoplastic invasion of lymph nodes
(1) Endothelioma
(2) Hodgkin's disease
(3) Leukemia
(4) Lymphangioma
(5) Lymphocytoma
(6) Lymphoma
(7) Lymphosarcoma
(8) Obstruction of thoracic duct
(9) Sarcoma of lymph nodes
(10) Reticular cell sarcoma
g. Dependency edema
II. LYMPHANGITIS
A. Primary (idiopathic)
B. Secondary
1 . Infection
2. Infestation
3. Trauma
III. NEOPLASMS OF LYMPH VESSELS
A. Benign
1. Lymphangiectasia
2. Lymphangioma
a. Simple
b. Cavernous
c. Cystic
B. Malignant
I . Lymphangiosarcoma
Diseases Primarily Affecting the Lymphatics
I. LYMPHEDEMA
A. Primary (idiopathic)
1. Congenital and hereditary (Milroy's disease)
2. Congenital but not hereditary
a. Without constricting bands
b. With constricting bands
3. Praecox
B. Secondary
1. Lymphangitis and lymphadenitis
a. Infection and infestations
( 1 ) Filariasis
(2) Pyogenic
Diseases Affecting Primarily the Minute Vessels
I. INCREASED FRAGILITY (PURPURA) OF VESSELS
A. Thrombocytopenic purpura
1. Primary: idiopathic (Werlhoff's) disease
2. Secondary
a. Vascular defects
(1) Thrombotic thrombocytopenic
(2) Blood dyscrasias
(a) Acquired hemolytic anemia
(b) Hodgkin's disease
(c) Leukemia
(d) Malignant lymphoma
(e) Myeloma
(f) Pernicious anemia
purpura
1246
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
(3) Infections
(a) Bacterial
(b) Rickettsial
(c) Viral
(4) Splenomegaly
(a) Band s syndrome
(b) Felty's syndrome
(c) Gaucher's disease
(d) Hodgkin's disease
(e) Leukemia
(5) Malignancy
(6) Drugs and chemical agents
(a) Allyl-isopropyl-acetyl-carbamide
(Sedormid)
(b) Arsenic
(c) Benzene
(d) Bismuth
(e) Certain foods, such as orris root
(f) Chloramphenicol
(g) Dichloro-diphenyl-trichloro-ethane
(DDT)
(h) Digitoxin
(j) Dinitrophenol
(j) Ergot
(k) Gold
(1) Hair dyes
(m) Iodine
(n) Methylphenylethyl hydantoin
( Mesantoin)
(o) Nitrogen mustard
(p) Pertussis vaccine
(qj Phenol
(r) Phenolphthalein
(s) Phosphorus
(t) Quinidine
(u) Quinine
(v) Snake venom
(w) Streptomycin
(x) Sulfa drugs
(y) Triethylenemelamine
(z) Trimethadione
(7) Physical factors
(a) Heat stroke
(b) Radiation
(c) Burns
B. Nonthrombocytopenic purpura
1 . Primary
a. Senile
b. Purpura simplex
c. Hereditary hemorrhagic diathesis
2. Secondary
a. Stasis — increased venous pressure
b. Traumatic or mechanical
c. Allergic or anaphylactoid
( 1 ) Schoenlein-Henoch purpura
(2) Purpura fulminans
(3) Other
d. Skin diseases
e. Chemical agents
(1) Acetophenetidin (Phenacetin)
1 2 ) Atropine
(3) Belladonna
(4) Bismuth
(5) Chloral hydrate
(6) Iodine
(71 Mercury
(8) Penicillin
(9) Quinine
( 10) Salicylic acid
lii) Various anticoagulants
f. Systemic diseases and infections
( 1 ) Nephritis
1 2 ) Purpura fulminans
(3) Septicemia
(4) Enanthema
(5) Scarlet fever
(6) Other bacterial diseases
(7) Rickettsial diseases
(8) Viral diseases
g. Avitaminosis
( 1 ) Scurvy
(2) Vitamin P deficiency
(3) Vitamin K deficiency
14) Other vitamin deficiency
h. Cryoglobulinemia
II. INCREASED PERMEABILITY OF VESSELS
A. Allergic urticaria
B. Angioneurotic edema
C. Inflammation
D. Physical irritants
1. Trauma
2. Cold
3. Heat
E. Scrum sickness
REFERENCES
1 . Abramson, D. Diagnosis and Treatment of Peripheral Vascular
Disorders. New York: Hoeber-Harper, 1956.
2. Adson, A., and G Brown. The treatment of Raynaud's
disease by resection of the upper thoracic and lumbar
sympathetic ganglia and trunks. Surg. Gynecol. Obstet.
48:577. '929-
3. Allen, E., N. Barker, and E. Hines. Peripheral Vascular
Diseases (2nd ed.). Philadelphia: Saunders, 1956.
4. Allen, J., P. Moulder, D. Emerson, C. Basinger, J.
Landv, and D. Glotzer. Physiology of intravascular
coagulation in health and disease. Surg. Clin. North
Am. 37: 1473, 1957-
5. Barker, N, E. Hines, and W. Craig. Livedo reticularis.
A peripheral arteriolar disease. Am. Heart J. 21 : 592, 1 941 .
6. Boas, E. Capillaries of extremities in acrocyanosis. J. Am.
Med. Assoc. 79: 1404, 1922.
PERIPHERAL VASCULAR DISEASES
H7
'3'
14.
>5'
16.
'7-
'9-
23-
24-
25-
26.
27-
28.
29-
Brown, G. Erythromelalgia and other disturbances of
extremities accompanied by vasodilatation and burning.
Am. J. Med. Sci. 183: 468, 1932.
Buchanan, J., J. Cranlev, and R. Linton. Observations
on direct effect of cold on blood vessels in human ex-
tremity and its relation to peripheral vascular disease.
Surgery 31 : 62, 1952.
Burch, G. A Primer of Venous Pressure. Philadelphia : Lea &
Febiger, 1950.
Burch, G. Digital Plethysmography. New York : Grune &
Stratton, 1954.
Burch, G. George E. Brown Memorial Lecture: Digital
rheoplethysmography. Circulation 13:641, 1956.
Cassirer, R. Die Vasomotor ischtropischen Neurosen. Berlin :
Karger, 191 2.
Cohen, S. Traumatic arterial spasm. Lancet 1:1, 1944.
Coon, W., and P. Willis. Deep venous thrombosis and
pulmonary embolism; prediction, prevention and treat-
ment. Am. J. Cardiol. 4: 61 1, 1959.
Criteria Committee of the New York Heart Association,
Inc. Nomenclature and Criteria for Diagnosis of Diseases of the
Heart and Blood Vessels (5th ed.). New York: N. Y. Heart
Assoc, 1953.
Crocq, C. De l'acrocyanose. Semaine med. 16: 298, 1896.
Day, R., and W. Klingman. Effect of sleep on skin tem-
perature reactions in case of acrocyanosis. J. Clin. Invest.
18: 271, 1939.
Deutsch, F., O. Ehrentheil, and O. Peirson. Capillary
studies in Raynaud's disease. J. Lab. Clin. Med. 26: 1729,
1 941.
Doupe, J., C. Cullen, and C. Chance. Post-traumatic
pain and causalgia syndrome. J. Neurol. Neurosurg.
Psychiat. 7: 33, 1944.
Ebert, M. Livedo reticularis. Arch. Dermatol, and Syphilol.
16: 426, 1927.
Edwards, E. Varieties of digital ischemia and their man-
agement. New Engl. J. Med. 250: 709, 1954
Estes, J. Vasoconstrictor and vasodilative syndromes of
the extremities. Mod. Concepts Cardiovas. Dis. 25 : 355,
!956-
Feldaker, M., E. Hines, and R. Kierland. Livedo
reticularis with ulcerations. Circulation 13:1 96, 1 956.
Foley, W., E. McDevitt, J. Tulloch, M. Tunis, and
I. Wright. Studies of vasospasm. I. Use of glyceryl
trinitrate as a diagnostic test of peripheral pulses. Circula-
tion 7:847, 1953.
Foley, W., and I. Wright. Color Atlas and Management
of Vascular Disease. New York: Appleton, 1959.
Freeman, N. Acute arterial injuries. J. Am. Med. Assoc.
>39: "25. '949-
Freeman, N. Effect of temperature on rate of blood flow
in normal and in sympathectomized hand. -4m. J. Physiol.
1 '3: 384, 1935-
Freeman, N., and J. Zeller. Effect of temperature on
volume flow of blood through sympathectomized paw of
dog with observations on oxygen content and capacity,
carbon dioxide content and pH of arterial and venous
blood. .4m. J. Physiol. 120: 475, 1937.
Fuller, C, C. Robertson, and R. Smithwick. Manage-
ment of thromboembolic disease. New Engl. J. Med.
263:983, i960.
30. Goetz, R., and F. Ames. Reflex vasodilatation by body
heating in diagnosis of peripheral vascular disorders.
A.M. A. Arch. Internal. Med. 84: 396, 1949.
31. Gosset, A., I. Bertrand, and J. Patel. Sur la physio-
pathologie, des embolies arterielles des membres (re-
cherches expermentales). Ann. anat. pathol. 9: 841, 1932.
32. Hale, A., and G. Burch. Arteriovenous anastomoses and
blood vessels of human finger; morphological and func-
tional aspects. Medicine 39: 191, i960.
33. Hardy, E., and D. Tibbs. Acute ischaemia in limb in-
juries. Brit. Med. J. 1 : 1001, i960.
34. Homans, J. Minor causalgia, a hyperesthetic neurovascu-
lar syndrome. New Engl. J. Med. 222: 870, 1940.
35. Huggins, C, and E. Walaszek. Depressor polypeptides.
Am. Heart J. 60: 976, i960.
36. Juergens, J. Intermittent claudication. Med. Clin. North
Am. 42 : 981, 1958.
37. Katz, L., E. Linder, and H. Landt. On nature of sub-
stance(s) producing pain in contracting skeletal muscle:
its bearing on problem of angina pectoris and intermittent
claudication. J. Clin. Invest. 14: 807, 1935.
38. Kinmonth, J., F. Simeone, and V. Perlow. Factors
affecting diameter of large arteries with particular refer-
ence to traumatic spasm. Surgery 26: 452, 1949.
39. Kinmonth, J., G. Hadfield, J. Connolly, R. Lee, and
E. Amoroso. Traumatic arterial spasm: its relief in man
and in monkeys. Brit. J. Surg. 44: 164, 1956.
40. Kissin, M. Production of pain in exercising skeletal muscle
during induced anoxia. J. Clin. Invest. 13: 37, 1934.
41. Kistiakovsky, E. Erythrocyanosis cutis symmetrica,
angioneurosis endocrinopathica polyglandularis. Arch.
Dermatol, and Syphilol. 20 : 780, 1 929.
42. Kuntz, A. Afferent innervation of peripheral blood
vessels through sympathetic trunks; its clinical implica-
tions. Southern Med. J. 44: 673, 1951.
43. Leary', W., and E. Allen. Intermittent claudication as a
result of arterial spasm induced by walking. .4m. Heart J.
22: 719, 1941.
44. Leriche, R. De l'elongation et de la section des nerfs
perivasculaires dans certain syndromes douloureux d'ori-
gine arterielle et dans quelques troubles trophiques.
Lyon chir. 1 : 378, 1 91 3.
45. Lewis, T. Clinical observations and experiments relating
to burning pain in extremities, and to so-called "eryth-
romelalgia" in particular. Clin. Sci. 1 : 175, 1933.
46. Lewis, T. Experiments relating to peripheral mechanisms
involved in spasmodic arrest of circulation in fingers, a
variety of Raynaud's disease. Heart 15: 7, 1929.
47. Lewis, T. Pain in muscular ischemia. A. MA Arch. Inter-
nal Med. 49:713, 1932.
48. Lewis, T. The Blood Vessels of the Hitman Skin and Their
Responses. London: Shaw, 1927.
49. Lewis, T. Vascular Disorders of the Limbs (2nd ed.) London :
Macmillan, 1949.
50. Lewis, T., and E. Landis. Observations on vascular
mechanisms in acrocyanosis. Heart 15: 229, 1930.
51. Lewis, T., G. Pickering, and P. Rothchild. Observa-
tions upon muscular pain in intermittent claudication.
Heait 15: 359, 1 93 1.
52. Linton, R. Peripheral vascular diseases. New Engl. J.
Med. 260: 322, 1959.
1248
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
53. Livingston, W. K. Pain Mechanisms. A Physiological Inter-
pretation of Causalgia and Its Related States. New York :
Macmillan, 1943.
54. Mahorner, H., and A. Ochsner. A new test for evaluat-
ing circulation in venous system of lower extremity
affected by varicosities. Aich. Surg. 33: 479, 1936.
55. Marin, H., and M. Stefanini. Experimental production
of phlebothrombosis. Surg. Gynecol. Obstet. 110: 263,
i960.
56. Martin, W., H. Laufman, and S. Tuell. Rationale of
therapy in acute vascular occlusions based upon micro-
metric observations. Ann. Surg. 129: 476, 1949.
57. May, E., and P. Hillemand. Erythromelalgie; etude de
la pathologie du sympathique. Ann. med., Paris 16: 51,
I924-
58. Mayfield, F. Causalgia. Springfield, 111.: Thomas, 1951.
59. Mayo, W., and A. Adson. Raynaud's disease, thrombo-
angiitis obliterans and scleroderma : Selection of cases for
and results of sympathetic ganglionectomy and trunk
resection. Ann. Surg. 96: 771, 1932.
60. Mendlowitz, M. The Digital Circulation. New York:
Grune & Stratton, 1954.
61. Mendlowitz, M., and N. Naftchi. The digital circula-
tion in Raynaud's disease. Am. J. Cardiol. 4: 580, 1959.
62. Menendez, C, and R. Linton. Peripheral vascular dis-
eases. New Engl. J. Med. 251 : 382, 432, 1954.
63. Mitchell, S. Clinical lecture on certain painful affections
of the feet. Philadelphia Med. Times 3: 81, 1872.
64. Mulvihill, D., and S. Harvey. Studies on collateral
circulation. I. Thermic changes after arterial ligation and
ganglionectomy. J. Clin. Invest. 10: 423, 1931.
65. Myers, T., and J. Cooley. Varicose vein surgery in
management of postphlebitic limb. Surg. Gynec. Obstet.
99:733. 1954-
66. Naide, M., and A. Sayen. Venospasm: Its part in pro-
ducing the clinical picture of Raynaud's disease. A.M. A.
Arch. Internal Med. 77: 16, 1946.
67. Ochsner, A., and H. Mahorner. Varicose Veins. St. Louis:
Mosby, 1 939.
68. Ochsner, A., and M. DeBakey. Therapy of phlebo-
thrombosis and thrombophlebitis. Arch. Surg. 40 : 208,
1940.
69. Ochsner, A., and M. DeBakey. Thrombophlebitis and
phlebothrombosis. Southern Surgeon 8: 269, 1939.
70. Peacock, J. Peripheral venous blood concentrations of
epinephrine and norepinephrine in primary Raynaud's
disease. Circulation Research 7: 821, 1959.
71. Perkins, J., M. Li, F. Hoffman, and E. Hoffman,
Sudden vasoconstriction in denervatcd or sympathec-
tomized paws exposed to cold. Am. ./. Physiol. 155: 165,
1948.
72. Perthes, G. Ueber die operation der unterschenkel-
varizen nach Trendelenberg. Devi. med. Wochschr. 1 :
253. i895-
73 Phillips, J., and G. Burch. Digital biopsy in man: An
adjunct to the study of peripheral circulation. Am. J
Med. Sci. 235: 6, 1958.
74. Phillips, J., and G. Burch. Review of cardiovascular
diseases in white and Negro races. Medicine 39: 241, i960.
75. Phillips, J., G. Burch, and R. Hibbs. Applications of
digital biopsy to peripheral vascular investigations in
man, with special considerations to dermal chromaffin
cells. Am. J. Med. 27: 320, 1959.
76. Pickering, G. Vascular spasm. Lancet 2: 845, 1951.
77. Pickering, G. On clinical recognition of structural disease
of peripheral vessels. Brit. Med. J. 2: 1106, 1933.
78. Pickering, G., and E. Wayne. Observations on angina
pectoris and intermittent claudication in anaemia. Clin.
Sci. 1 : 305, 1934.
79. Pollack, A., B Taylor, T. Myers, and E. Wood.
Effect of exercise and body position on venous pressure
at ankle in patien's having venous valvular defects. J. Clin.
Invest. 28:559, 1949.
80. Pratt, G. Cardiovascular Surgery. Philadelphia: Lea and
Febiger, 1954.
81. Prinzmetal, M. Studies on mechanism of circulatory
insufficiency in Raynaud's disease in association with
sclerodactylia. Arch. Internal Med. 58: 309, 1936.
82. Owens, J. Causalgia. Am. Surgeon 23: 636, 1957.
83. Rapport, M., A. Green, and I. Page. Serum vasocon-
strictor (serotonin); IV. Isolation and characterization.
J. Biol. Chem. 176: 1243, 1948.
84. Raynaud, M. De lasphyxie locale el de la gangrene symitrique
des extremites. Paris: Rignoux, 1862.
85. Rothman, S. Physiology and Biochemistry of the Skin. Chicago :
Univ. Chicago Press, 1954.
86. Rous, P., and H. Gilding. Meaning of Bier's spots.
Proe. Soc. Exptl. Biol. Med. 26: 497, 1929.
87. Samuels, S. Diagnosis and Treatment of Vascular Disorders.
Baltimore : Williams & VVilkins, 1 956.
88. Scully, R., and C. Hughes. Pathology of ischemia of
skeletal muscle in man. Am. ./. Pathol. 32: 805, 1956.
89. Shumacker, H., and D. Abramson. Post-traumatic
vasomotor disorders; with particular reference to late
manifestations and treatment. Surg. Gynecol. Obstet. 88:
417, 1949.
90. Shumacker, H., I. Spiecel, and F. Upjohn. Causalgia.
I. The role of sympathetic interruption in treatment.
Surg. Gynecol. Obstet. 86: 76, 1948.
91. Smith, L., and E. Allen. Erythermalgia (erythro-
melalgia) of extremities; A syndrome characterized by
redness, heat and pain. Am. Heart. J. 16: 175, 1938.
92. Sodeman, \V., and G. Burch. Tissue pressure: An objec-
tive method of following skin changes in scleroderma.
Am. Heart J. 17:21, 1939.
93. Starr, I., Jr. Change in reaction of skin to histamine.
./. .4m. Med. Assoc, go: 2092, 1928.
94. Stern, E. The aetiology and pathology of acrocyanosis.
Brit. J. Dermatol. Syphilis 49: 100, 1937.
95. Symposium on peripheral vascular diseases. Am. J.
Cardiol. 4: 565, 1959.
96. Symposium on peripheral vascular diseases. Am. J. Med.
23:673, "957-
97. Taubenhaus, M., B. Eisenstein, and A. Pick. Cardio-
vascular manifestations of collagen diseases. Circulation
12:903- "955-
98. Travell, J., S. Baker, B. Hirsch, and S. Rinzler.
Myofascial component of intermittent claudication.
Federation Proc. II: 164, 1952.
99. Trendelenburg, F. Ueber die unterbindung der saphena
magna vein. Beitr. klm. Chir. 7: 195, 1891.
PERIPHERAL VASCULAR DISEASES
[249
ioo. Uvnas, B. Vasodilator nerves. Am. Heart J. 62: 277,
1 961.
101. Werner, M., and S. Udenpriend. Relationship of
platelet serotonin to disturbances of clotting and hemo-
stasis. Circulation 15: 353, 1957.
102. Wessler, S., S. Sheps, M. Gilbert, and M. Sheps.
Studies in peripheral arterial occlusive disease; acute
arterial occlusion. Circulation 17: 512, 1958.
103. Williams, C, and H. Goodman. Livedo reticularis.
J. Am. Med. Assoc. 85: 955, 1925.
104. Winsor, T. Peripheral Vascula) Diseases. Springfield, 111.:
Thomas, 1959.
CHAPTER 37
Situations which lead to changes
in vascular patterns
AVE RILL A. LIEBOW
Department of Pathology, Yale University School of Medicine,
New Haven, Connecticut
CHAPTER CONTENTS
Normally Occurring Arteriovenous Communications
Structure
Distribution and Size
Development and Fate
Function
Role in Bodily Economy
Abnormal Arterial Communications
Traumatic or Surgically Induced Arteriovenous Connections
Collateral Circulation
Some Aspects of Angiogenesis in General
Types of Collaterals
Forces Affecting the Development of Collateral Circulation
Mechanical factors
Neural factors
Chemical factors
Rate of Development
Regression of Collaterals
Arterial Versus Venous Collaterals
Some Effects of Collateral Circulation
Structure of Collateral Vessels
Measurement of Collateral Circulation
Some Outstanding Problems
AFTER WILLIAM HARVEY HAD DISCOVERED the circu-
lation of the blood, there remained the mystery of its
transfer from arteries to veins. A solution was provided
by Marcello Malpighi in 1 66 1 when he first saw the
capillaries in the transparent lung of the frog (146).
It took another half century before direct connections
between an artery and vein (the spermatic), in this
instance probably anomalous, were reported by Lealis
Leali (43). Precapillary arteriovenous anastomoses
are now known to exist normally in many organs and
tissues. Masson (117) credits Berres (14) with their
discovery in 1832 in erectile tissue where they were
later described in considerable detail by Johannes
Miiller (123). The transparent wing of the bat pro-
vided an opportunity for observing the vessels in the
living subject and here Paget (130) saw large arterio-
venous anastomoses. Hyrtl (88) noted that when these
structures were open there was pulsation of veins and
arterialization of the blood within them. Sucquet (172)
soon found precapillary arteriovenous connections to
be widely distributed in man, but his results were dis-
credited by such observers as Hover (80) and Berliner-
blau (13) for the reason that they were based on
injection of fluids of low viscosity. Arnold (4, 7) recog-
nized the "coccygeal gland," which had been dis-
covered by Luschka in 1859, to be analogous to the
glomeruli caudales of animals and to represent in
reality vascular complexes replete with arteriovenous
anastomoses. He remarked on the muscular nature of
some of the vessels. The first detailed histological
description of the specialized transitional segment was
by Hover (80) and this was elaborated by Grosser (65)
in 1902. Max Clara (29, 30), author of the most
extensive monographs on these structures, considered
Schumacher (159) to be the discoverer of the epi-
thelioid cells. The relationship of arteriovenous an-
astomoses to nerves was definitely demonstrated by
Masson (1 16-1 18), and was subsequently investigated
by Brown (25), and, with special reference to tumors,
by Popoff (134).
Both normal and abnormal arteriovenous shunts
can exert physiological effects, but these have been
explored only in part. The observations of Grant (61)
1 251
1252
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
on the intact rabbit's ear have been supplemented by
a whole series of investigations by the Clarks (34, 36)
who used a chamber technique. Recently, some at-
tention has been paid to a possible secretory function
of these structures.
Arteriovenous communications can be classified as:
A. "Normal" arteriovenous connections
1. Simple
2. Complex
B. -'Abnormal"
1. Congenital
a. Familial
b. Isolated
2. Progressive acquired
a. Hemangiomatous
b. Within neoplasms
c. Associated with disease, e.g., cutaneous
spiders
C. Traumatic
D. Surgically induced
NORMALLY OCCURRING ARTERIOVENOUS CONNECTIONS
Structure
Connections between arteries and veins can range
from simple bridges only slightly larger than capil-
laries (fig. 1) to complex channels with specialized
cells in their walls (figs. 2 and 3). The former have
been well described as components of the microcircu-
lation by Zweifach (200). Both extremes can be en-
countered, for example in the rabbit's ear (148).
Spanner, in particular, has emphasized the existence
of transitional forms (165).
The more complex of these structures can be de-
rived from a larger artery at a bifurcation, one division
of which may be distributed to capillaries in the usual
fashion. The other, or both, can become remarkably
contorted and characteristically differentiated before
joining a vein. In the intermediate or intercalated
segment, called variously the Sucquet-Hoyer or
Hover-Grosser canal, the wall becomes thickened by
the presence of a broad layer of ''epithelioid" cells
which abut upon or partly replace the endothelium
and appear to be differentiated from smooth muscle
cells. At the beginning of the intercalated segment the
epithelioid and muscle cells may be intermingled.
When fully formed, however, the former approach a
spherical shape and contain few or no myofibrils.
Their cytoplasm is hyaline, or somewhat vacuolated,
and gives no reaction for glycogen, fat, or mucin. A
fig. i. A relatively direct arteriovenous anastomosis from the
human ear. Specimen injected with Berlin blue, stained with
hematoxylin and cleared. The arrow points to the terminal
portion of the intercalated segment. There is a slight fusiform
thickening nearer the arterial end of the latter, suggesting
accumulation of muscle or epithelioid cells. [From Prichard &
Daniel (135).]
circular layer of muscle fibers may or may not be
preserved externally to the epithelioid cells. The elastic
laminae usually disappear in the intermediate seg-
ment. The adventitia is a delicate collagenous re-
ticulum supporting a very rich plexus of both medul-
lated and nonmedullated nerves. The latter were well
described by Masson (117, 118), and also in some of
their finest details in the tongue of the dog by Brown
(25). The latter noted thin unmyelinated fibers to
terminate in the media and thick myelinated fibers
(afferents?) with termination in the adventitia (fig. 3).
Groups of such complicated arteriovenous anastomoses
may be closely associated to form a "glomus" which
may be enclosed within a dense connective tissue
capsule.
Less complex arteriovenous anastomoses exist in
which the intercalated segment is not tortuous. In
some there is simply a well-developed inner layer of
longitudinal muscle fibers without special epithelioid
characteristics.
Distribution and Size
The distribution of arteriovenous shunts is now
known to be almost universal. Aside from the glomus
coccygeum, some of the largest and most complex
glomera in man have been described in the skin and
subcutaneous tissue on the flexor surfaces of the fingers
and toes and in the nail beds. Their numbers have
CHANGES IN VASCULAR PATTERNS
1253
V e n e
Vine <<^
■-~*lK Vent
Vene
Venen
Arterie
fig. 2. Graphic reconstructions of arteriovenous anastomoses; relatively simple (left) and complex
(rightj communications. The accumulations of epithelioid cells are indicated. [From Staubesand &
Genschovv (168).]
been variously stated. For example, Grant & Bland
(62) found 593 per cm2 in the nail bed of the toe, and
293 per cm2 on the plantar side, but Popoff (134)
counted only 24 per cm2 in the nail bed, and 18 on
the ventral aspect of the same extremity. The latter
considered only the more complicated glomera. Their
size also varies: Grosser (65) found the external di-
ameter to be between 55 and 85 y. in the nail bed,
between 90 and 150 y, in the finger pad, and the
internal diameter to be 18 to 22 and 10 to 30 y, re-
spectively. In the wings of bats the intermediate seg-
ment had an external diameter of from go to 280 y,
and an internal diameter of from 60 to 150 y. The
length of the junctional segment as measured in the
tongue of the dog by Prichard & Daniel (136) was
between 100 and 500 y, usually between 200 and 300 y.
Other locations where arteriovenous anastomoses
have received detailed study include: erectile tissue,
the ears in man (135); the nose, including skin,
septum, and turbinates; and the gastrointestinal tract
(8). Their existence in the kidney has been denied by
Trueta (178), and by Staubesand & Hammarsen
(169), although Spanner (165) described them in the
region of the sinus renalis, and Simkin et al. (162)
found that spheres as large as 90 to 440 y would pass
from renal arteries to veins. In the lung, Weibel (186)
could find no precapillary connections between pul-
monary arteries and veins, but Prinzmetal et al. (137)
found that glass spherules as large as 150 y would
pass from the former to the latter, and Parker et al.
(131) observed that spheres of 75 to 80 y would
traverse the capillaries but those of 300 y would not.
Tobin & Zariquiey (176) and Rahn et al. (140) have
also concluded that pulmonary arteriovenous com-
munications must exist normally. In perfused lobes
Niden & Aviado (126) observed glass beads as large
as 420 y on the venous side in a perfusate introduced
intra-arterially. Bostroem & Piiper (23), however,
found that spheres of 28 to 36 y. would pass only
exceptionally, and criticized the high pressures used
by Tobin and his associates. Gordon et al. (60) also
concluded from their own work, based on an appli-
'254
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 3. An arteriovenous anastomosis
from the tongue of the dog. The richest
nerve supply is to the intercalated seg-
ment. The filaments end in all levels of
the wall, in which large rounded epi-
thelioid cells predominate. Portions of a
thick sensory fiber and sensory termi-
nations are seen in the adventitia.
[From Brown (25).]
1
cation of the principles of surface tension, that no
connections larger than 25 /* were present in the lungs,
intestines, or kidneys of rats and rabbits, in contrast
with the extremities of these animals. Fritts and co-
workers (55) stated on the basis of recently developed
methods utilizing simultaneous T-1824 and radio-
active krypton injections that if such shunts are
functional they could account for not more than 1 per
cent of total left cardiac output in normal human
subjects. These problems have been judiciously re-
viewed by de Burgh Daly (41, 42) with special
reference to possible influences of the nervous system.
Connections between bronchial and pulmonary
arteries are mentioned in the discussion of collateral
circulation in this chapter.
Development and Fate
Arteriovenous shunts, at least those with a differ-
entiated intercalated segment and complex "orga-
noid" structure, do not exist in embryos. Popoff (134)
could not find them in the extremities from 4.5
months of intrauterine life to term, although Clara
(30) stated that they may be present in the newborn.
After the age of 60 the complex cutaneous arterio-
venous shunts tend to undergo atrophy and sclerosis.
Clark & Clark (36) observed the new formation of
arteriovenous anastomoses in transparent chambers of
the rabbit's ear where the tissue was induced to grow
into an originally vacant space. Here the anastomoses
were relatively straight, but were characterized by the
addition of an extraendothelial layer of differentiated
cells. Stimuli leading to increase in blood flow seemed
to increase the formation of these structures. Most of
these shunts were temporary, and disappeared early
or late, but some were permanent and had the
property of contractility. This seemed to be associated
with the development of nerves. Newly formed arterio-
venous shunts were also found within 2 weeks after
resecting a marginal segment of the rabbit's ear (148).
Function
With the disclosure of arteriovenous anastomoses in
erectile tissue it became obvious that their functional
state must vary from time to time. The ability of these
structures to close was established in the living trans-
illuminated rabbit's ear as early as 1930 by Grant
(61 ). The use of the rabbit ear chambers with "pre-
formed tissue" provided a clearer view in the hands
of the Chirks (34), and they were able to make
quantitative observations on the number, size, and
rate of contraction of the anastomoses over intervals
of many months. The specialized intercalated seg-
ments with their greater thickness and complex mural
arrangements and rich nerve supplv showed a faster
and more complete contraction than the arteries, and
the rhythm was independent of that of the latter. No
explanation was apparent for the extremely variable
responses of various arteriovenous anastomoses (fig.
4). The mechanism of closure has come under dis-
cussion and has been considered to be contraction ol
CHANGES IN VASCILAR PATTERNS
1255
y.'!:.'}!'.['."n>i>nnM»/ll/,li- ,. ■■■.
<>' I 2"
1 ;\_WV"\ /WWWLA.
V~AA^
FIG. 4. A series of four arteriovenous communications from
the rabbit's ear. The curves represent the rapid and uncoordin-
ated rate of contraction or dilatation of each of these structures.
[From Clark & Clark (34) as modified by Clara (29).]
an external layer of circular muscle in some instances,
the accumulated epithelioid cells simply acting as a
cushion that partly restricts the lumen even when the
intercalated segment is open. In the absence of the
circular muscle however it has been thought that
closure is the result of swelling of the epithelioid cells,
in consequence of a still unknown process. Bcnning-
hoff (12) was the first to suggest this idea, and Havlicek
(71) called these cells "Quellzellen." Since these are
approximately spherical, shortening, as with ordinary
muscle fibers, is not possible. This mechanism has
been extensively considered by Mark (112, 1 13).
Even direct observations have their limitations,
since the functions of these structures may be multiple,
and are not necessarily the same at all times, nor in
all vascular beds in the same animal, nor in different
species. There is now good evidence that they can re-
spond both to neural and chemical stimuli, but there
are numerous contradictions in details (119). It is of
interest that so large a glomeroid structure as the
coccygeal body can be removed, as in resection of the
coccyx, without known physiological effects (160).
Local mechanical stimulation, such as rubbing, re-
sults in opening of the intercalated segments (34, 61).
The effects of temperature appear to be determined
in degree as well as direction by quantitative factors.
Upon warming the whole animal both Grant and the
Clarks found in rabbits a widening of the anastomoses.
Moderate cooling was accompanied by their closure,
and this was observed also in the paw of the dog by
Bostroem & Schoedel (24). Sonomoto (163) also
found that the arteries and a majority of the arterio-
venous anastomoses were constricted in the rabbit's
ear during the winter. In the fingers of man, however,
the anastomoses were closed by warming to an ex-
ternal temperature of 33 to 37 C, while cooling pro-
duced the opposite result. In the rabbit's ear Grant
(61) found that cooling below 15 C produced an
opening of the anastomoses, and that with greater or
more prolonged reduction in temperature the arteries
also became dilated, whereupon there was a rapid
flow of blood through the anastomoses. He explained
that this phenomenon kept the extremities from
getting too cold. In human skin and in the feet of
birds Grant & Bland (62) further established this
function of the arteriovenous shunts by temperature
measurements.
The consequences of anoxia were examined by
Schroeder et al. (158) with plethysmography methods,
under the assumptions that the volume of an ex-
tremity kept at an initial pressure of 35 mm Hg will
reflect changes in capillary pressure, and that of an
extremity compressed at 15 mm Hg will reflect
changes in venous pressure. They found that when the
dog was breathing an atmosphere containing 8 to 9
per cent 02, the capillaries were wide without alter-
ation in the functional state of the arteriovenous
anastomoses, but when the oxygen concentration was
between 6 per cent and 8 per cent the anastomoses be-
came narrow without change in capillaries. In an
atmosphere of between 5 per cent and 6 per cent 02,
perfusion was slowed in consequence of constriction of
the nutrient bed as well as of the shunts.
Stimulation of the cervical sympathetic was noted
by Grant (61) to constrict the anastomoses as well as
the small arteries in the rabbit's ear. The denervation
of an extremity in the dog resulted in dilatation of the
shunts (24). Folkow (54) stated that the cutaneous
arteriovenous anastomoses become maximally dilated
as soon as their constrictor fibers are cut, provided
that there is no significant increase in hormone output
of the adrenal medulla. Claude Bernard's classical
observation that, when the peripheral end of the
chorda tympani is stimulated, the rate of blood flow
from the submaxillary vein becomes greater and the
blood becomes bright red, has been interpreted to
indicate the shunting of blood through the arterio-
venous anastomoses (71). After vagotomy Curri et al.
(40) reported that the arteriovenous anastomoses be-
came widely open but lost reactivity to various
stimuli.
The injection studies of Vastarini-Cresi (179) had
suggested that, in general, vasoconstrictor substances
1256
HANDBOOK OF PHYSIOLOGY -" CIRCULATION II
decreased and vasodilators increased the size of the
arteriovenous connections. The vasoconstrictor effect
of adrenaline was evident in direct observations of
living vessels in rabbits' ears and it was found in the
same preparations that histamine and acetylcholine
dilated these shunts (61). This has been confirmed.
According to Curri et al. (40) serotonin introduced
intravenously in a dose of 8 mg resulted in a cessation
of rhythmic activity of the intercalated segments.
Role in Bodily Economy
Surely one function of the arteriovenous shunts as a
component of the microcirculation is concerned with
regulation of regional blood flow, as exemplified in
erectile tissue. In general, when the shunts are open
the capillary bed may be largely or entirely bypassed,
and total blood flow traversing the part may be
maximally increased. This phenomenon, according to
Grant (61, 62), helps to maintain the temperature of
the extremity when exposed to extreme cold. Also it
has been known since 1840, from the observation of
Julius Robert Mayer, surgeon to the threemaster
"Java," that venous blood tends to become "arteri-
alized" in the tropics, indicating a dilated state of the
arteriovenous connections (71). Thus, a thermoregu-
latory function has been suggested for these structures.
Many of the arteriovenous anastomoses are however
deeply situated, for example, in the periosteum or
even within parenchymatous organs, and must have
other than thermoregulatory functions. A third
physiological role which has been considered, but
which has not been truly demonstrated or tested
experimentally, is in the regulation of blood pressure.
It seems logical that if sufficient numbers of the direct
arteriovenous anastomoses are widely open, systemic
arterial blood pressure might fall.
It was suggested by Schumacher (160), largely on
theoretical grounds, that the specialized epithelioid
cells might have a secretory function — more specifi-
cally that they could secrete acetylcholine. Luckner &
Staubesand (no) found in extracts of the coccygeal
body a substance with the biological properties of
acetylcholine in concentrations of 9000 jug per g.
Indeed Schumacher (160) conceived that the pulsa-
tion of the arteriovenous anastomoses was a mecha-
nism to maintain a level of the short-lived acetyl-
choline in the blood. This concept is of interest in that
the epithelioid cells are rather widely distributed in
small groups within the walls of arteries, for example
at the vascular pole of the glomerulus, where they
had been described first by Ruyter (149) and later by
Goormaghtigh (59).
Also rather theoretical is the idea that the connect-
ing segments may be pressoreceptors, i.e., that the
metabolism of the cells could be altered by variations
in pressure, and that this effect could somehow be
transmitted to the associated extensive neural plexuses.
Schumacher (160) thought that cells of the non-
chromaffin paraganglia were analogous to the epi-
thelioid cells of the intercalated segments, but there
is no evidence that the carotid body is related to
arteriovenous anastomoses, although some of these
structures exist in its connective tissue capsule (1).
Clearly, there is much to be learned in the domain
of function of the arteriovenous anastomoses. Further
study doubtless will be highly rewarding.
ABNORMAL ARTERIOVENOUS COMMUNICATIONS
Arteriovenous connections of unusual size or loca-
tion can occur as single or multiple lesions and,
especially when multiple, can be familial (Osler-
Weber-Rendu disease). The lesions vary from insig-
nificant blue or purple spots on the skin or mucous
membranes to complex cirsoid masses with the
arrangement of hemangiomas. These are important
chiefly because they can bleed, as for example into
the gastrointestinal tract.
In the lung, the pulmonary arteries and veins can
come into free anastomosis with a right to left shunt.
When of sufficient size, there are the expected conse-
quences of desaturation of systemic arterial blood,
cyanosis, polycythemia, clubbing, and at times
thrombotic complications. Cardiac failure does not
occur unless immense numbers of the arteriovenous
fistulas are present (66). These can be of such small
size as to be undetectable by angiography.
It was known to Virchow that acquired hemangio-
mas with a cavernous component also are the seat of
arteriovenous communications, as indicated by the
bright color of the effluent blood. Fistulation also can
occur within certain neoplasms, especially when they
become necrotic or hemorrhagic as in the case of
chorionepithelioma. A bruit may then become audi-
ble over the lesion.
The "cutaneous arterial spider" has been recog-
nized to consist in part of arteriovenous connections.
The arterial component has in its walls specialized
"glomus cells" (epithelioid cells), like other arterio-
venous shunts. Such structures develop commonly in
CHANGES IN VASCULAR PATTERNS
-01
association with severe chronic liver disease, in preg-
nancy, in persons with deficiency of the vitamin B
complex, in the carcinoid syndrome, and also in
certain apparently healthy individuals. The subject
has been well reviewed by Bean (10).
TRAUMATIC OR SURGICALLY INDUCED
ARTERIOVENOUS CONNECTIONS
The establishment of a connection of sufficient
magnitude between an artery and vein may have
major or even catastrophic consequences. These
have been carefully worked out by experiment (76,
77> 79, i53)-
The immediate effects upon opening the fistula are
a fall in blood pressure, an increase in the heart rate
and venous filling, and consequently a greater cardiac
output. The regional veins become engorged. A
thrill becomes palpable and a murmur audible over
the fistula, and these can be abolished by exerting
pressure over the vein proximal to the fistula. With a
sufficiently large shunt, the total blood volume in-
creases in course of time, and the blood becomes more
dilute. Although initially the size of the heart and of
the artery on both sides of the fistula becomes re-
duced, there is gradually a dilatation of the arteries
and veins proximal to the fistula. The heart also
becomes enlarged, chiefly because of dilatation. In
late stages, the proximal artery may even become
aneurysmally dilated (144).
Blood flow is toward the fistula, even from the
distal artery. A large flow depends on a fistula which
exceeds in size that of the proximal artery. Lewis (98)
stated that the blood supply to the distal parts of the
limb is at first diminished, but that with passage of
time blood flow tends to become restored and may
even exceed that to the normal limb. This results
from the development of an extensive collateral circu-
lation as will be discussed (fig. 5). Evidence for in-
crease in the flow through the fistula is that cardiac
dilatation and decompensation can occur late after
the arteriovenous fistula is established. For the flow to
increase progressively the distal artery must be dis-
tensible. Excessive scarring can interfere with this
distensibility (79). Schenk el al. (153) made quantita-
tive observations on the regional blood flow in all
limbs of experimental arteriovenous fistulas using a
square wave electromagnetic flowmeter. They found
that in the femoral fistulas in dogs, flow tended to
increase and at the end of approximately 1 year had
*
fig. 5. Traumatic arteriovenous fistula. The trauma oc-
curred accidentally during attempted biopsy of lymph nodes
from the anterior scalene region in a 72-year-old man. Several
weeks later the patient noted pain and a pulsatile swelling in
the region of the wound over which a systolic bruit was audible
The specimen is a vinylite cast of the arteriovenous fistula
showing numerous tortuous arterial channels related to the
fistula.
not yet reached stability. In contrast with the femoral
fistulas, flow through the proximal artery of carotid-
jugular fistulas tended to diminish with passage of
time, but in both types of fistulas there was a marked
increase in flow through both the arterial and venous
distal limbs.
COLLATERAL CIRCULATION
Collateral circulation may be defined as blood flow
that pursues a channel or system of vessels which is
alternative to or develops in substitution for a major
vascular pathway. To understand collateral circula-
tion would require not only a complete knowledge of
the mechanisms of angiogenesis, and therefore of all
growth, but also of the anatomical and functional
responses of blood vessels in general. At the present
writing only limited answers can be supplied to such
questions as: What starts the growth of collateral
vessels and what controls the rate of their increase;
what stops them from expanding indefinitely; when
newly formed, what guides them to their proper
place; what determines the structure of their walls.
Some two hundred years ago the great John Hunter
was amazed to find that not only did the growth of
the stag's antler proceed uninterrupted when its
1 2 -,8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
major nutrient artery was ligated, but that in time
there was a prodigious growth of new vessels (146).
He was not the first to observe this providential
activity of nature, since Antyllus, a pioneer in the
surgical treatment of aneurysms fifteen centuries
before him, had noted that interruption of the artery
to a limb does not necessarily result in its loss (97).
Morgagni also had anticipated Hunter in observing
collateral vessels. After numerous experiments Hunter
could conclude only that "vessels go where they are
needed'' ( 124).
In attempting in this chapter a precis of some
major advances in the understanding of this complex
subject a sense of wonder and frustration still remains.
Scholarly reviews have been published by Mulvihill
& Harvey (124), Quiring (139), Longland (108),
Learmonth (96) and Rati ]
had become intrigued by Leriche's studies of the
periarterial sympathetic nerves. Mulvihill & Harvey
(125) found that with sympathectomy the temperature
of an extremity did not fall after ligature of the
external iliac artery (fig. 10). Theis (173) and later
Longland (108) confirmed this observation by various
methods. Both indicated the persistence of the effect
over several months. Injection with alcohol of the
main artery beyond a ligature in the anterior ex-
tremity of the dog, a procedure that was presumed
to produce a destruction of the sympathetic nerves
within it, seemed to result in a better development
of collateral circulation than in the leg of the control
side (91, 92).
Turning to the "microcirculation," Fulton et al.
(57) provide a description of several orders of nerve
plexuses related to the small vessels in the cheek
pouch of the hamster. The networks are sufficiently
rich to innervate all the smooth muscle cells of the
vessels. The development of a complement of non-
medullated nerve fibers in newly formed arterioles
in chamber preparation of the rabbit's ear has been
demonstrated by the Clarks (37). They have estab-
lished that only vessels supplied with such fibers are
capable of spontaneous contraction. The bearing of
these observations on the development of collateral
circulation remains for further exploration.
In 1958, North & Sanders (127) reported that the
innervation of the mouse ear seemed to have no
effect on the growth of collateral vessels.
The existence of a "basal vascular tone"' of local
muscular rather than neural origin has been con-
sidered by Folkow (53, 54). Evidence for an appar-
ently nonneural dilator response, probably trans-
mitted by the musculature of the vessel itself, has been
adduced by Hilton (75). He found that cocainization
of a femoral artery feeding actively contracting
muscle abolished its dilatation, while cutting the
nerve to the extremity and curarizing the animal
did not. This dilator response traveled up the artery
at a slow rate, of the order of 10 cm per sec. Such
phenomena may have a bearing on the total problem
of the reaction of collaterals.
When the influence of the nervous system, and
possibly also of intrinsic myogenic influences, is con-
sidered, it is clear that the important effect relative
to collateral circulation is the lysis of vascular tone.
This is expressed essentially in the alteration of
mechanical forces. Possible influences on the growth
of vessels are as yet unknown.
chemical factors. There are at least four ways in
which chemical substance could affect collateral
circulation: /) By regulating vasodilatation. 2) By
controlling the proliferation of new vessels. 5) By-
stimulating and inhibiting the growth of vessels.
(Growth itself is obviously a chemical process,
although it could be initiated by mechanical or
chemical factors.) ./) By guiding vessels to specific
destinations.
1264
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 10. Effect of left sympa-
thectomy upon temperature of an
extremity after bilateral ligation
of iliac artery. On the side of
sympathectomy the temperature
is maintained. On the opposite
side it falls gradually to room
temperature but rises with great
rapidity at about the 8th hour
following the ligation. [From
Mulvihill & Harvey (125).]
The idea that metabolites formed locally in anoxic
tissue can produce vasodilatation received early ex-
perimental support. Marey (115) had observed that
hyperemia occurred in the arm after a sufficient
degree of compression either by mercury (150-200
mm Hg) or in a pneumatic chamber. That this did
not require participation of the central nervous
system and that it was not the result of squeezing the
nerve itself was demonstrated by Bier (17, 18), who
noted reactive hyperemia after clamping and then
releasing the artery in the limb of a pig severed from
all connections with the body except for the major
vessels; compression of the nerve in another prepara-
tion produced no such effect. John & Warren (90)
showed that an increased flow was associated with
reactive hyperemia. Possible sources of error in
interpretation are that nerves intimately associated
with the arteries, or axone reflexes, might be in-
volved, or that a local myogenic conducting mecha-
nism might exist. Bier explained enlargement of the
major arteries on the basis of decrease in peripheral
resistance of the capillary bed.
Controversy regarding chemical mechanisms in
collateral circulation was further stimulated by
Thomas Lewis's (98) dictum " — and we are brought
to ask if arterial growth is not directly controlled by
a stimulant, a chemical stimulant arising locally as a
product of tissue need and acting locally." If this
had no other good effect, it at least inspired a vigorous
investigation of the mechanisms of collateral circula-
tion, especially those related to arteriovenous fistula.
The primacy of mechanical over chemical phe-
nomena, under some circumstances when both might
be operative, was suggested in the case of the ex-
tremity after ligature of a major artery by Winblad
et a!. (192), and by John & Warren (90), among
others, and in the case of arteriovenous fistula by
Holman (77) in his amputation experiment.
Certain factors, such as anemia, exercise, and de-
creased arterial pO-2, can increase interarterial
anastomoses at least in the coronary system (45, 46,
197, 198). Just how this is brought about is not
known.
The Clarks (33) had observed that the same
growth conditions which favored the formation of
new blood vessels also stimulated the growth of other
tissues in the same region. They were led to suggest:
"As for the chemical substance or substances which
may stimulate the formation of new capillaries, they
should be sought in embryonic tissue and in inflam-
matory exudates, since it is in such environments that
active vascular formation takes place." In autografts
of connective tissue in transparent chambers Williams
(190) concluded that hypoxia of a certain degree is a
stimulus for growth of vascular endothelium. No
specific data are given, however, to support this
statement.
For the growth of vessels, Nothnagel (129) offered
a deceptively simple explanation: "'Anemia of
peripheral parts results in an increased flow through
collaterals, whereupon there is an augmented
nourishment of the walls of these vessels, by the mate-
rials with which they become increasingly perfused."
To substantiate this idea it is necessary to demonstrate
CHANGES IN VASCULAR PATTERNS
1265
that substances transported through the lumen of a
vessel serve to nourish it, or that after the flow be-
comes greater both nourishment and growth increase.
Compelling evidence does exist that chemical
factors affect the growth of vessels in general, and the
development of collateral circulation in particular.
This becomes especially clear in organs with a
double blood supply such as the lung. For example,
there is general agreement that the actively metaboliz-
ing cells of primary malignant pulmonary neoplasms
receive their blood supply from systemic vessels (39,
99, 194, 195). This would indicate the effect of
chemical rather than mechanical factors, since the
vast majority of the pulmonary capillaries represent
a bed supplied by the pulmonary artery. In fact,
with the growth of tumors the pulmonary arteries and
veins tend to be obstructed or peripherally displaced.
This may have some practical importance, since at-
tempts have been made to subject pulmonary neo-
plasms to high concentrations of cytotoxic agents by
injecting them into the pulmonary arteries leading to
the involved segments. There is disagreement on the
blood supply of metastatic tumors, since some (99)
have found them likewise to be vascularized from
the aorta, while others (39) concluded that the
bronchial arteries did not nourish the metastatic
tumors. Possibly the disagreement indicates variation
in the blood supply. In the case of the liver, according
to Hales (personal communication), both primary
and metastatic tumors are supplied by the hepatic
arteries.
Other newly formed tissues in the lung also are
nourished by systemic arteries. Again this implies
that chemical factors must be at work. Thus, in or-
ganizing pulmonary disease, as in bronchiectasis, the
granulation tissue is derived at least in large part
from the bronchial arteries (104, 194). Ultimately
precapillary anastomoses are formed with branches
of the pulmonary arteries. That this is not necessarily
the outcome of enlargement of existing precapillary
anastomoses, is indicated by the observation that
vessels penetrating into the lung from intercostal
arteries via adhesions anastomose with pulmonary-
arteries in similar fashion (102).
The peculiar "tropism" exhibited by newly formed
collaterals in the lung must also have a chemical
explanation. It is well established that ligature of the
pulmonary artery induces expansion only of arterial
collaterals (102). After interruption of the pulmonary
veins, only the venous collaterals expand (86). Pro-
liferating branches of intercostal vessels that enter the
lung through the pleura are of the same type, arterial
or venous, depending on the stimulus, and the ar-
teries form precapillary connections only with ar-
teries, and the veins only with veins. When both
major limbs of the pulmonary circulation are inter-
rupted, the expansion of both collateral systems is
induced, and both types of collaterals penetrate in-
ward through pleural adhesions (183). These col-
laterals must be newly formed, from capillaries in
granulation tissue which did not exist previously but
which comes to obliterate the pleural space after the
operative manipulation. Remarkably, again the
transpleural branches of the intercostal arteries
establish precapillary connections only with the pul-
monary arteries, and the intercostal veins connect
likewise only with pulmonary veins. If these were
mechanically induced there should be "short cir-
cuits'' between the intercostal arteries and veins,
since they share a common capillary bed in the granu-
lation tissue as it is first formed, and the greatest
pressure gradient under the circumstances is obvi-
ously from intercostal arteries to intercostal veins.
Yet such short circuits have rarely if ever been ob-
served. Rather, the blood pursues the longest course,
from the intercostal arteries through precapillary
anastomoses into the pulmonary arteries, to the pul-
monary capillaries, to pulmonary veins, and finally
through precapillary connections into the collateral
veins. The chemical factors that must be responsible
for this are still unknown.
There is some suggestive, but as yet imperfect,
evidence that hormones can exert an influence on the
development of collateral circulation. The collateral
circulation that develops within a few months after
ligature of a pulmonary artery in puppies less than
48 hours old seems immensely greater by gross in-
spection than that appearing after a comparable
interval of time in adult animals after the same pro-
cedure (105). The results of a study of the effect of
hormones on collateral circulation after interruption
of the iliac artery in rats were, however, equivocal
(147). Cortisone seemed to inhibit the collateral circu-
lation, just as it did the connective tissue proliferation
in the region of the lower abdominal incision where,
as a result, hernias appeared. The weight increase of
these animals, however, stopped when cortisone was
administered. It might be expected at the end of the
experiment that the smaller animals would have
smaller vessels. This problem, however, should be
reinvestigated.
Studies of the earliest phases of development of the
1266
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
M
t 1 ■
1
fig. 12. The calculated regression lines for rates of increase
of collateral blood flow are plotted when both pulmonary
arteries and veins are ligated (solid line) and when the pulmon-
ary artery alone is interrupted (broken line). These are re-
markably congruent; this indicates that the collateral veins can
expand at least as rapidly as the arteries. The data for "artery
alone" have been recalculated in ml/kg/min from a paper by
Bloomer el al. (20). An adjustment for nitrogen shift in the
bronchospirometry has also been made and this has made pos-
sible the construction of the graph, using also data previously
published by Vidone & Liebow (183).
In certain other situations progressive increase in
expansion of collateral beds can take place for at least
1 year and possibly longer. As discussed previously,
this has been observed in the arterial collateral circu-
lation related to an arteriovenous fistula.
Regression of Collaterals
That preformed collaterals can open rapidly and
disappear as quickly when the stimulus to their
formation is abrogated has been shown angiographi-
cally by Winblad et al. (192) and John & Warren (90).
As a collateral bed develops, certain of its com-
ponents tend to enlarge and to persist as major
channels, while others regress. This was noted in
successive angiograms after ligating the femoral
artery in the rabbit (108). North & Sanders (127)
found in the ear of the mouse that when continuity
of an interrupted vascular channel was regained cer-
tain minor collaterals regressed.
Even collaterals of long standing remain only so
long as the stimuli that led to their expansion are
maintained. Bosher et al. (22) observed regression of
collateral circulation associated with a peripheral
arteriovenous fistula by comparing angiograms im-
mediately and again 6 weeks after fistulectomy; some
regression was apparent as early as the fourth or fifth
day. Anatomically the collaterals in regression were
described as showing marked subendothelial pro-
liferation. The results after fistulectomy were similar
to those after ligature of the major participating ves-
sels in the fistula, and this was considered further
evidence against the "tissue need" theory. Winblad
et al. (192) Schoop (156) and Hasse & Schoop (70)
noted the regression of collaterals after adequate
thrombo-intimectomy or bypass grafting in major
systemic arteries. Similar phenomena were described
by Jacobson & McAllister (89).
Arterial Versus Venous Collaterals
"Nature has been more prodigal in the provision of
alternative venous and lymphatic routes than she
has been in arranging for arterial collaterals" [Lear-
month (96)].
In the lung the stimuli to the development of
arterial and of venous collaterals are independent.
This is true not only where mechanical forces seem
dominant as in the expansion of pre-existing col-
laterals, but also where chemical influences appear to
be pre-eminent as in the case of newly formed trans-
pleural vessels.
When both arteries and veins are compromised
under appropriate conditions, both arterial and
venous limbs of the collateral circulation will expand.
This has already been discussed for the lung. In seg-
ments of small intestine transplanted to the sub-
cutaneous tissue by the Florey- Harding method
(128) both arterial and venous collaterals appeared
when the original mesenteric vascular pedicle was
severed.
It is of interest that in these experiments the venous
collaterals seemed to develop to a larger size more
quickly than the arterial. Similar observations had
been reported by North & Sanders (127) in the ear
of the mouse. The veins seemed to expand within 24
hours, while it took 4 to 5 days for visible expansion
of arteries to take place. Quantitative data are avail-
able for the lung. When both the pulmonary arteries
and veins are interrupted, the collateral blood flow is
approximately the same as when arteries alone are
ligated (fig. 12). This means that expansion of the
venous collateral can at least keep pace with that
of the arterial.
In these experiments arteries became joined to
arteries, and veins to veins, but there are circum-
1268
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
stances where arteriovenous connections develop in
newly formed circulations, as in the rabbit ear
chamber (36). It is of interest that such arteriovenous
shunts are normally present in the rabbit's ear (34).
Some Effects of Collateral Circulation
Blood supply arriving of necessity by way of col-
lateral routes is usually not so efficient as the original
in maintaining full function. Rarely it may exceed
the needs of the tissue, as in association with large
peripheral arteriovenous fistulas, and in the lungs, as
has been described. In the latter, however, it never-
theless falls short of normal pulmonary perfusion
that is carried on in the service of the body as a
whole.
Certain special effects of collateral circulation
occur under pathological circumstances, but these
can only be mentioned in passing. Thus, in the lung,
the seat of severe fibrosing disease such as bronchi-
ectasis, where large precapillary anastomoses are
formed between bronchial and pulmonary arteries,
the higher pressure in the former tends to shunt the
pulmonary arterial blood into normal tissue where
oxygenation can occur. Consequently, there may be
no peripheral arterial desaturation. When suffi-
ciently numerous, these connections may contribute
to an increased resistance to the output of the right
ventricle. Reverse flow in pulmonary arteries from
the periphery via these anastomoses has also been
demonstrated, by analysis of gases in blood drawn
from catheters placed within such arteries, and by
angiography (103), and most convincingly by
aortography (3). In the last mentioned a radiopaque
substance introduced above the origins of the bron-
chial vessels has been shown to fill the pulmonary
arteries retrogradely.
The hepatic circulation bears certain analogies to
the pulmonary. Enlargement of the hepatic arteries
has been demonstrated in cirrhosis, and the suggestion
has been made that they may contribute to portal
hypertension as a consequence of more direct connec-
tions with the portal veins (68).
Bronchial veins so enlarged that their valves become
incompetent have been demonstrated in pulmonary
emphysema, and the possibility of reverse flow of
blood, i.e., from the azygos system into the pul-
monary veins has been inferred (100, 114). Such
shunting has also been considered as an explanation
of the cyanosis sometimes encountered in fibrosis of
the liver (27). Enlargement of bronchial veins,
probably as a result of high pressure in the azygos
system which may carry a large volume of blood
bypassing the liver, has been demonstrated by injec-
tion.
Structure of Collateral Vessels
It is now well established that vessels reflect in their
structure the mechanical conditions to which they
are subjected. As early as 1883 von Recklinghausen
(142) stated in his textbook that as collaterals carry
more blood, they become thicker and more tortuous.
That there is both hypertrophy and hyperplasia of
smooth muscle in the larger collaterals was described
and illustrated by Nothnagel (fig. 9). Fischer &
Schmieden (51) provided an experimental demon-
stration of adaptive changes in larger vessels sub-
jected to altered circumstances of pressure and flow.
When a segment of external jugular vein was in-
serted into the course of the carotid artery of a dog,
it became reduced in caliber, firmer, and as much as
two or three times thicker. Histologically, the media
was shown to contain much more muscle and con-
nective tissue (fig. 13). The medial elastic fibers were
thought to be reduced, but this was not convincingly
demonstrated. The adventitia also was seen to con-
tain denser connective tissue. The intima generally
remained unchanged. The trunk of the pulmonary
artery, when subjected to a sufficiently increased
pressure, becomes markedly thickened with an
increase both in elastic tissue and smooth muscle
(105, 145).
With the enlargement of small arteries as they
become able to carry more blood there often appear
remarkable aggregates of longitudinal smooth muscle
fibers that dissect or even replace the internal elastic
lamina, and that may lead to the subtotal or even
complete obliteration of the lumen. Such vessels
have been most extensively studied in the lung and
in manv tvpes of chronic pulmonary disease where
bronchial collateral circulation is characteristically
increased (109, 187). Some have called these "Sperr-
arterien" (72-74, 93) and have thought them to
possess a regulatory function in relation to their
anastomoses with pulmonary arteries. In more
general terms it may be said that longitudinal muscle
tends to increase in other small muscular arteries with
an augmentation in the blood that they carry, as in
the bases of the cardiac valves in rheumatic fever, in
the vasa vasorum of the aorta in syphilis, and in
the intercostal vessels as they traverse adhesions to
enter the lung. Probably the hypertrophy and hyper-
plasia of muscle is in fact a response to increased
CHANGES IN VASCULAR PATTERNS
1269
a
I k %
m
fic. 13. Adaptive changes in a seg-
ment of jugular vein which had been
inserted into the course of the carotid
artery for 86 days. At left is shown the
appearance of the vessel before its
exposure to the higher pressure. Key:
i — intima; m — media. [From Fischer and
Schmieden (51).]
tension which, in general, appears to increase the
tone of muscle (9, 16, 171). This has been suggested
for the increased muscle characteristic of bullae in
pulmonary emphysema, the walls of which are under
stretch consequent to air trapping (106). A brilliant
experimental demonstration of this mechanism in
small vessels has been provided by Weibel (187).
In his experiment, increased tension in mesenteric
vessels was produced by stretching the mesentery
slightly and attaching it to the diaphragm. The inner
longitudinal muscle then did increase to a remarkable
degree.
Gaps in the internal elastic lamina of the large
arteries serving as collaterals have been noted by
several observers (102, 108). Some newly formed
collaterals mav possess relatively little or no elastica
(16.).
Newly formed collateral vessels that develop from
capillaries, as for example in the adhesions between
visceral and parietal pleura, ultimately acquire a
structure appropriate to their function as arteries
or veins at the size which they ultimately attain
(86).
The growth of muscle in the walls of vessels func-
tioning as collaterals in the mesentery of the rat has
been well described by Weyrauch & De Garis (189).
They considered the stimulus to be increased blood
volume. The tortuosity of the vessels was said to be
the result of the fact that the muscle fibers do not
grow in a single plane and this may be one factor to
account for the tortuosity of collaterals in general.
They also described the appearance of muscle in
vessels which they thought were newly formed.
Less well understood than the structural changes
are the forces that bring them about. They are prob-
ably similar to those that govern the differentiation
of arteries and veins from the retiform capillary
anlagen of the early embryo, as previously discussed.
Measurement of Collateral Circulation
Attempts have been made to estimate the extent of
collateral circulation by both anatomical and physio-
logical methods. The former offers only a general
and not necessarily reliable guide to the latter, in the
sense that the size of a bridge cannot always provide
a clue to the magnitude, nor even to the direction of
traffic.
The early observers, such as Morgagni and Porta,
made many excellent observations with the naked
eye. Direct visual observation continues to be of
value and details of the formation of smaller col-
lateral vessels can be observed microscopically at
intervals in the process of their formation, for example,
in the ear of the living mouse (127). John Hunter
(83) early used injection methods in his famous
studies of the new blood supply to the stag's antler.
Some of his casts of the vessels are preserved to this
day in the Museum of the Royal College of Surgeons
in London. Attempts have been made to quantitate
12 70
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the results of the injections. This has been accom-
plished by some who have standardized an injection
mass that does not penetrate through vessels of less
than a known diameter. Important information has
been \ ielded by such radiopaque materials as Schle-
singer's (154, 155). Colored masses in gelatin or simi-
lar materials have also been used to inject tissues
that have subsequently been cleared, for example by
the method of Spalteholz (164). Materials that
harden to provide casts of the vessels and which
resist subsequent corrosion of the tissues have been
useful in many applications. In the study of such
casts, or other injections, certain critical points can
be established, and the magnitude of the collateral
can be estimated by whether or not the material has
penetrated into particular segments of the system
(147). Casts can be measured or weighed. Glass or
plastic beads of graded sizes have been used to es-
tablish the size of vascular communications (2, 3,
126, 131, 137, 138, 162). With these materials over-
pressure must be avoided and the possibility of con-
tamination must also be considered. Angiography,
microangiography (11), and more recently cineange-
ography have become increasingly important with
improvements of technique, especially since they
offer a way of investigating the collateral circulation
in intact animals. Quantitation can be achieved by
such procedures as that of Longland (108), who
counted the number of vessels in his angiograms that
exceeded a stated size at a selected level. In the lung,
a degree of refinement can be obtained by relating
the vessels to the diameters of the bronchi which they
accompany (188). The same can be done in any
organ with an appropriate reference structure.
To measure collateral blood flow, direct and more
or less indirect methods have been applied. The
simplest perhaps is the collection of blood from the
veins draining the part. This procedure is useful only
if the tissue is supplied exclusively from a collateral
source during the period of measurement, and if the
veins carry away all or a known proportion of blood.
These conditions cannot often be met. In man,
plethvsmography has been employed in the study of
collateral circulation (193).
The collection of "backflow" from a vessel beyond
a point of occlusion has been used on the presumption
that it will increase if the vessels circumventing an
occlusion come to carry an increased volume of
blood. This principle has been extensively applied in
the study of the coronary circulation (63). Upon
opening the vessel beyond the obstruction the periph-
eral resistance confronting the blood in the collateral
vessels is, of course, reduced, and the backflow can in
no sense be considered an absolute measure of col-
lateral blood supply. As a relative measure the prin-
ciple is valid if no uncontrollable change, such as
spasm, has occurred in the diameter of the vessel
beyond the point of occlusion. Backflow then would
reflect the pressure in the vascular bed in the distal
arterial segment, which also is related to the extent
of the collateral connections. Pressures as well as
flows have been measured for this purpose.
Other "'direct" measurements have been made by
introducing such devices as the bubble flowmeter
into the major feeders of the collateral bed (26).
In such structures as the lung, attempts have been
made to perfuse separately the greater and lesser
circulations (151). Both procedures require extensive
surgery, with denervation and possibly other dis-
turbing factors.
In the lungs bronchospirometry and blood gas
analyses, with temporary balloon blockade of a pul-
monary artery and application of "mixing formulas"
where indicated, can provide data on "'effective"
collateral arterial flow, i.e., blood arriving by syste-
mic arteries that becomes oxygenated in the lungs
(20, 52).
The fact that the temperature of a tissue bears a
relationship to the quantity of arterial blood perfusing
it in a unit of time has been used to measure col-
lateral circulation (50). One source of error lies in the
fact that blood flow is not necessarily distributed in a
uniform manner through all tissues of a part, nor
through all portions of a tissue.
The distribution of such dyes as Evans blue or Fox
green or of radioactive materials, or "labeled"
ervthrocytes (138) in various vascular compartments
has been used for qualitative detection of shunts, but
under specific conditions. Isotonic solutions differing
in temperature or conductivity from blood can be
employed instead of dyes, and records similar to dye
concentration curves can be obtained with appro-
priate sensing, amplifying, and recording devices.
Under special circumstances such methods can
also be applied in a quantitative fashion. In the lung
where an extensive bronchial collateral circulation
represents a left-to-left shunt, originating as it does in
the left ventricle and aorta, and returning from the
lungs via the pulmonary veins to the left heart, intro-
duction of an indicator material such as T-1824,
Fox green, or radioactive iodinated serum albumin
into the circulation produces characteristic altera-
CHANGES IN VASCULAR PATTERNS
tions from the usual in the arterial dye curve: A more
rapid reversal of the downward limb of the first wave
and a double-humped camel rather than dromedary
recirculation curve. A known quantity of the indi-
cator can be injected rapidly into a systemic vein,
and concentration curves can be obtained simul-
taneously from the pulmonary artery and aorta by
appropriate methods such as cuvette densitometry.
The "left cardiac output" measured from the latter
should exceed the "right cardiac output" calculated
from the former by the volume of the collateral blood
supply to the lung. This principle has been applied
by a number of workers (38, 55, 56), but very rapid
left-to-left recirculation introduces problems that
may make this procedure inapplicable for quantitative
use.
To measure collateral blood flow from extra-
coronary sources to the heart by means of vessels in
anastomosis with the coronary arteries, the dye has
been introduced into the aorta above the orifices of
the presumed collaterals and well below the origins
of the coronary arteries in the sinuses of Valsalva.
If collaterals exist, dyed blood will reach the coronary
sinus by the collateral route before recirculation can
take place. Quantitation has been attempted by com-
paring the peak concentrations in the aortic blood
with that of the coronary sinus peak, or better, the
areas beneath appropriate segments of the two
curves (182).
Some Outstanding Problems
The problems of collateral circulation are in-
separable from those of angiogenesis, "histodynamic->"
in Thoma's sense and hemodynamics in general.
Methods for study have advanced notably, but new
developments can be expected to accelerate progress.
Catheters, as used currently for measuring pressures
and in obtaining dye concentration curves, carry
inherent artifacts. Accurate sensing units sufficiently
small so as not to interfere significantly with blood
flow are needed for both purposes
It is clear that many of the basic mechanisms must
be essentially physicochemical. These must underlie
the molding influence of mechanical forces on the
structures of vessels. They must also be responsible for
what is now vaguely recognized as "tropism." None
of the essential chemical information is yet available
to explain how, in a newly formed collateral bed,
arteries are joined directly to arteries and veins to
veins, with no arteriovenous connections, while the
latter are constantly present normally in certain
other parts.
John Hunter's (185) remark of 1785 still well
defines the present state of knowledge: "All the uses
arising from the anastomosing of the vessels are, per-
haps, not yet perfectly understood; general reasons
can, I think, be assigned for them, but these will not
apply to all cases; it is something, therefore, more
than we are yet acquainted with."
REFERENCES
1 . Adams, W. E. The Comparative Morphology of the Carotid
Body and Carotid Sinus. Springfield, 111.: Thomas, 1958.
2. Aeby, C. Der Bau des Menschlichei. Korpers. Leipzig: Vogel,
1 87 1.
3. Alley, R. D., A. Stranahan, H. Kausel, P. Formel,
and L. H. S. van Mierop. Demonstration of bronchial-
pulmonary artery reverse flow in suppurative pulmonary
disease. Clin. Research 6: 41, 1958.
4. Arnold, J. Ein Beitrag zu der Structur der sogenannten
Steissdriise. Virchow's Arch, pathol. Anat. 32: 293, 1865.
5. Arnold, J. Experimentelle Untersuchungen iiber die
Entwickelung der Blutcapillaren. Virchow's Arch, pathol.
Anat. 53: 70, 1 87 1.
6. Arnold, J. I. Experimentelle Untersuchungen iiber die
Entwickelung der Blutcapillaren. II. Die Entwickelung
der Capillaren bei der Keratitis vasculosa. Virchow's
Arch, pathol. Anat. 54: I, 1872.
7. Arnold, J. Ueber die Glomeruli caudales der Sauge-
thiere. Virchow's Arch, pathol. Anat. 39: 497, 1867.
8. Barclay, A. E., and F. Bentley. The vascularization
of the human stomach. A preliminary note on the shunt-
ing effect of trauma. Brit. J. Radiol. 22: 62, 1949.
g. Bayliss, W. M. On the local reactions of the arterial wall
to changes of internal pressure. J. Physiol. 28: 220, 1902.
10. Bean, W. B. The cutaneous arterial spider: A survey.
Medicine 24: 243, 1945.
11. Bellman, S., H. A. Frank, P. B. Lambert, and A. J.
Roy. Studies of collateral vascular responses. I. Effects
of selective occlusions of major trunks within an exten-
sively anastomosing arterial system. Angio/ogy 10: 214,
'959-
12. Benninghoff, A. Blutgefasse und Herz. Arteriovenose
Anastomosen, Glomus coccygeum und Polsterarterien.
In : Handbuch der Mikroskopiscken Analomie des Alenschen.
Berlin: Springer, 1930, vol. 6, pt. 1, pp. 107-112.
13. Berlinerblau, F. Ueber den directen L'ebergang von
Arterien in Venen. Arch. Anat. Physiol. 117: 1875.
14. Berres, J. Analomie der mikroskopischen Gebildc des mcnch-
lichen Korpers. Vienna: Ceroid, 1837.
15. Berry, J. L., and I. de B. Daly. The relation between
1272
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the pulmonary and bronchial vascular systems. Proc.
Roy. Soc. London, B. 1 09 : 3 1 g, 1 93 1 .
16. Biedermann, W. Beitrage zur allgemeinen Nerven- und
Muskel-physiologie. Sitzungsberichte der Wiener Akademie
89: .9, 1884.
17. Bier, A. Die Entstehung des Collateralkreislaufs, Theil
I. Der arterielle Collateralkreislauf. Virchow's Arch, palhol.
Anal. 147: 256, 1897.
18. Bier, A. Die Entstehung des Collateralkreislaufs, Thiel
II. Der Ruckfluss des Blutes aus ischamischen Korper-
theilen. Virchow's Arch, pathol. Anat. 153: 306, 1898.
ig. Bing, R. J., L. D. Vandam, and F. D. Gray, Jr. Physio-
logical studies in congenital heart disease. II. Results of
preoperative studies in patients with tetralogy of Fallot.
Bull. John Hopkins Hasp. 80: 121, 1947.
20. Bloomer, W. E., W. Harrison, G. E. Lindskog, and
A. A. Liebow. Respiratory function and blood flow in
the bronchial artery after ligation of the pulmonary
artery. Am. J. Physiol. 157: 317, 1949.
21. Blumgart, H. L. Anatomy and functional importance
of intercoronary arterial anastomoses. Circulation 20: 816
■959-
22. Bosher, L. H., F. Harper, and I. A. Bicger. A study of
the collateral circulation after excision of arteriovenous
fistulas. Surgery 26: 918, 1949.
23. Bostroem, B., and J. PiipER. Uber arterio-venose Anasto-
mosen und Kurzschlussdurchblutung in der Lunge.
Pfliigers Arch. ges. Physiol. 261 : 165, 1955.
24. Bostroem, B., and W. Schoedel. Uber die Durchblutu-
tung der arteriovenosen Anastomosen in der hinteren
Extremitat des Hundes. Pfliigers Arch. ges. Physiol. 256 :
371. 1953-
25. Brown, M. E. The occurrence of arterio-venous anasto-
moses in the tongue of the dog. Anat. Record 6g: 287, 1937.
26. Bruner, H. D., and C. F. Schmidt. Blood flow in the
bronchial artery of the anesthetized dog. Am. J. Physiol.
148: 648, 1947.
27. Calabresi, P., and W. H. Abelmann. Porto-caval and
porto-pulmonary anastomoses in Laennec's cirrhosis
and in heart failure. J. Clin. Invest. 36: 1257, 1957.
28. Chapman, W. B. The effect of the heart-beat upon the
development of the vascular system in the chick. Am. J.
Anat. 23: 175, 1918.
29. Clara, M. Die Arteno-Vcnosen Anastomosen. Vienna:
Springer, 1956.
30. Clara, M. Die arterio-venosen Anastomosen der Vogel
und Saugetiere. Ergeb. Anat. Entwicklungsgeschichte. 27:
246, 1927.
31. Clark, E. R. Studies on the growth of blood-vessels in
the tail of the frog larva — by observation and experi-
ment on the living animal. Am. J. Anat. 23: 37, 1918.
32. Clark, E. R., and E. L. Clark. Caliber changes in
minute blood-vessels observed in the living mammal. Am.
J. Anat. 73: 215, 1943.
33. Clark, E. R., and E. L. Clark. Microscopic observa-
tions on the growth of blood capillaries in the living
mammal. Am. J. Anat. 64: 251, 1939.
34. Clark, E. R., and E. L. Clark. Observations on living
arterio-venous anastomoses as seen in transparent cham-
bers introduced into the rabbit's ear. Am. J. Anat. 54:
229, 1934.
35. Clark, E. R., and E. L. Clark. Observations on living
preformed vessels as seen in a transparent chamber
inserted in the rabbit's ear. Am. J. Anat. 49: 441, 1932.
36. Clark, E. R., and E. L. Clark. The new formation of
arterio-venous anastomoses in the rabbit's ear. Am. J.
Anat. 55: 407, 1934.
37. Clark, E. R, E. L. Clark, and R. G. Williams. Micro-
scopic observations in the living rabbit of the new growth
of nerves and the establishment of nerve-controlled
contractions of newly formed arterioles. Am. J. Anat.
55: 47. '934-
38. Cudkowicz, L., W. H. Abelmann, G. E. Levinson,
G. Katznelson, and R. M Jreissatv. Bronchial arterial
blood flow. Clin. Sci. 19. 1, i960.
3g. Cudkowicz, L., and J. B. Armstrong. The blood supply
of malignant pulmonary neoplasms. Thorax 8: 152, 1953.
40. Curri, S. B., F. Tischendorf, and C. C. Maggi. Experi-
mentelle Untersuchungen zur Histophysiologie und
Pathologic der arteriovenosen Anastomosen (nach
Lebendbeobachtungen am Kaninchenohr). Acta neuro-
veget. (Vienna) 14: 149, 1956.
41. Daly, I. de B. Intrinsic mechanisms of the Lung. Quart.
J. Exptl. Physiol. 43: 2, 1958.
42. Daly, I. de B. Reactions of the pulmonary and bronchial
blood vessels. Physiol. Rev. 13: 149, 1933.
43. Del Guerra, G. The first description of arteriovenous
anastomosis. J. Cardiovascular Surg. 1:218, 1 960.
44. Deterling, R. A , H. E. Essex, and J. M. W'augh.
Arteriovenous fistula: Experimental study of influence
of sympathetic nervous system on development of col-
lateral circulation. Surg. Gynecol. Obstet. 84: 629, 1947.
45. Eckstein, R. W. Development of interarterial coronary
anastomoses by chronic anemia. Disappearance following
correction of anemia. Circulation Research 3: 306, 1955.
46. Eckstein, R. W. Effect of exercise and coronary artery
narrowing on coronary collateral circulation. Circulation
Research 5: 230, 1957.
47. Eckstein, R. W., D. E. Gregg, and W. H. Pritchard.
The magnitude and time of development of the collateral
circulation in occluded femoral, carotid and coronary
arteries. Am. J. Physiol. 132: 351, 1941.
48. Evans, H. M. On the development of the aortae, cardinal
and umbilical veins, and the other blood vessels of
vertebrate embryos from capillaries. Anat. Record 3: 498,
1909.
49. Evans, H. M. On the earliest blood-vessels in the anterior
limb buds of birds and their relation to the primary
subclavian artery. Am. J. Anat. 9: 281, 1909.
50. Ferris, H. W., and S. C. Harvey. A physiological study
of the development of the collateral circulation in the leg
of the dog. Proc. Soc. Exptl. Biol. Med. 22: 383, 1 924-1 925.
51. Fischer, B., and V. Schmieden. Experimentelle Unter-
suchungen liber die funktionelle Anpassung der Gefass-
wand. Histologic transplantierter Gefasse. Frankfurt.
Z. Pathol. 3:8,1 909.
52. Fishman, A. P., G. M. Turino, M. Brandfonbrener,
and A. Himmelstein. The "effective" pulmonary col-
lateral blood flow in man. J. Clin. Invest. 37: 1071, 1958.
53. Folkow, B. Intravascular pressure as a factor regulating
the tone of the small vessels. Acta Physiol. Scand. 1 7 : 289,
■949-
54. Folkow, B. Role of the nervous system in the control of
vascular tone. Circulation 21 : 760, i960.
CHANGES IN VASCULAR PATTERNS
I273
55. Fritts, H. W., Jr., A. Hardewic, D. F. Rochester, 77.
J. Durand, and A. Cournand. Estimation of pulmonary
arteriovenous shunt-flow using intravenous injections of
T-1824 dye and Kr85. J. Clin. Invest. 39: 1841, i960.
56. Fritts, H. W\, Jr., P. Harris, C. A. Chidsev, III, 78.
K H. Clauss, and A. Cournand. Estimation of flow
through bronchial-pulmonary vascular anastomoses with
use of T-1824 dye. Circulation 23: 390, 1961
57. Fulton, G. P., B. R. Lutz, and A. B. Callahan. Inner- 79.
vation as a factor in control of microcirculation. Physiol.
Revs. 40: 57, i960.
58. Golubew, A. Bcitrage zur Kenntniss des Baues und der 80.
Entwicklungsgeschichte der Capillargefasse des Frosches.
Arch, mikroskop. Anal. 5: 49, 1869.
59. Goormaghtigh, N. Les segments neuro-myo-arteriels 81.
juxta-glomerulaires du rein. Arch. biol. 43: 575, 1932.
60. Gordon, D. B., J. Flasher, and D. R. Drury. Size of
the largest arteriovenous vessels in various organs. Am. 82.
J. Physiol. 173: 275, 1953.
61. Grant, R. T. Observations on direct communications
between arteries and veins in the rabbit's ear. Heart 83.
15: 281, 1929-1931.
62. Grant, R. T., and E. F. Bland. Observations on arterio- 84.
venous anastomoses in human skin and in the bird's foot
with special reference to the reaction to cold. Heart 15:
385, 1 929-' 93 '■ 85-
63. Gregg, D. E. Coronary Circulation in Health and Disease.
Philadelphia: Lea & Febiger, 1950.
64. Gross, L. The Blood Supply to the Heart in its Anatomical 86.
and Clinical Aspects. New York: Hoeber, 1921.
65. Grosser, O. Ueber arterio-venose Anastomosen an den
Extremitatenenden beim Menschen und den kral-
lentragenden Saugethieren. Arch, mikroskop. Anat. 60: 87.
igi, 1902.
66. Hales, M. R. Multiple small arteriovenous fistulae of
the lung. Am. J. Pathol. 32: 927, 1956.
68. Hales, M. R., J. S. Allan, and E. M. Hall. Injection 88.
corrosion studies of normal and cirrhotic livers. Am. J.
Pathol. 35: 909, 1959. 89.
69. Halsted, W. S. A striking elevation of the temperature
of the hand and forearm following the excision of a sub-
clavian aneurysm and ligations of the left subclavian 90.
and axillary arteries. Bull. Johns Hopkins Hasp. 31 : 219,
1920. 91.
70. Hasse, H. M., and W. Schoop. Der Kollateralkreislauf
vor und nach operativer Wiederlr-rstellung der Strom-
bahn bei Arterienverschlussen. Z. Kreislaujforsch. 50: 242,
1 961.
71. Havlicek, H. Vasa privata und vasa publica. Neue 92.
Kreislaufprobleme. Hippokrates 2: 105, 1929.
72. Hayek, H. v. Die Menschluhe Lunge. Berlin : Springer, 1 953.
73. Hayek, H. v. Uber einen Kurzschlusskreislauf (arterio-
venose Anastomosen) in der menschlichen Lunge. Z. 93.
Anal. Entwicklungsgeschichte 110: 412, 1940.
74. Hayek, H. v. Uber verschlussfahige Arterien in der
menschlichen Lunge. Anat. Ariz. 89: 216, 1 939-1 940.
75. Hilton, S. M. A peripheral arterial conducting mecha-
nism underlying dilatation of the femoral artery and 94.
concerned in functional vasodilatation in skeletal muscle.
J. Physiol. 149: 93, 1956.
76. Holman, E. Arteriovenous Aneurysm. New York : Macmillan, 95.
■937-
Holman, E. Problems in the dynamics of blood flow.
I. Conditions controlling collateral circulation in the
presence of an arteriovenous fistula, following the ligation
of an artery. Surgery 26 : 889, 1 949.
Holman, E., and M. E. Edwards. A new principle in
the surgery of the large vessels. Ligation of vein proximal
to site of ligation of the artery : An experimental study
J. Am. Med. Assoc. 88: 909, 1927.
Holman, E., and G. Taylor. Problems in the dynamics
of blood flow. II. Pressure relationships at site of an
arteriovenous fistula. Angiology 3: 415, 1952.
Hoyer, H. Lleber unmittelbare Einmiindung kleinster
Arterien in Gefassaste venosen Charakters. Arch, mikroskop.
Anat. 13: 603, 1877.
Hughes, A. F. W. Studies on the area vasculosa of the
embryo chick. I. The first differentiation of the vitelline
artery. J. Anat. 70: 76, 1 935-1 936.
Hughes, A. F. W. Studies on the area vasculosa of the
embryo chick. II. The influence of the circulation on
the diameter of vessels. J. Anat. 72: 1, 1937- 1938.
Hunter, J. Essays and Observations, edited by R. Owen.
London: Van Voorst, 1861, vol. 1, p. 126.
Hunter, W. The history of an aneurysm of the aorta,
with some remarks on aneurysms in general. Med. Obs.
& Inquiries by a Society of Physicians in London 1 : 323 : I 756.
Hunter, W. Further observations on a particular species
of aneurysms. Med. Obs. (3 Inquiries by a Society of Physicians
in London 2:390, I 76 1 .
Hurwitz, A., M. Calabresi, R. W. Cooke, and A. A.
Liebow. An experimental study of the venous collateral
circulation of the lung. I. Anatomical observations. Am.
J. Pathol. 30: 1085, 1954.
Hurwitz, A., M. Calabresi, R. W. Cooke, and A. A.
Liebow. An experimental study of the venous collateral
circulation of the lung. II. Functional observations. J.
Thoracic Surg. 28: 241, 1954.
Hyrtl. Anatomical Notes. 8. On the radial artery in the
cheiroptera. Natural History Rev. 2: 99, 1862.
Jacobson, J. H., II, and F. F. McAllister. The harmful
effect of arterial grafting on existing collateral circulation.
Surgery 42: 148, 1957.
John, H. T., and R. Warren. The stimulus to collateral
circulation. Surgery 49: 14, 1961.
Kolesnikow, V. Die VVirkung der Desympathisierung
von Arterien mit Alkohol nach Rasumowsky auf die
Entwicklung von Kollateralen. (Anatomisch-experi-
mentelle Untersuchung). Z. Anat. Entwicklungsgeschichte
(1 Abt.) 89: 405, 1929.
Kolesnikow, V. Uber einige Eigenschaften der Kol-
lateralen der vorderen Extremitaten beim Hunde.
(Anatomisch experimen telle Untersuchung). Z. Anat.
Entwicklungsgeschichte (lAbt.) 89:412, 1929.
Lapp, H. Liber die Sperrarterien der Lunge und die
Anastomosen zwischen A. bronchialis und A. pulmonalis,
uber ihre Bedeutung, insbesondere fur die Entstehung
des hamorrhagischen Infarktes. Frankfurt. Z. Pathol. 62:
537, '95'-
Latschenberger, J., and A. Deahna. Beitrage zur Lehre
von der reflectorischen Erregung der Gefassmuskeln.
Arch. Physiol. 12: 157, 1876.
Laurie, VV., and J. D. Woods. Anastomosis of the
coronary circulation. Lancet 2: 812, 1958.
1274
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
96. Learmonth, J. Collateral circulation, natural and
artificial. Surg. Gynecol. Obstet. 90: 385, 1950.
97. Leonardo, R. A. History of Surgery. New York: Froben
Press, 1943.
98. Lewis, T. The adjustment of blood flow to the affected
limb in arteriosenous fistula. Clin. Sci. 4: 277, 1 939-1942.
99. Liebow, A. A. Tumors of the lower respiratory tract.
Fascicle 17, "Alias of Tumor Pathology." Washington,
D. C. : Armed Forces Institute of Pathology, 1952.
1 10. Liebow, A. A. The bronchopulmonary venous collateral
circulation with special reference to emphysema. Am.
J. Pathol. 29: 251, 1953.
1 01. Liebow, A. A., M. R. Hales, and W. E. Bloomer.
Relation of bronchial to pulmonary vascular tree. In :
Pulmonary Circulation, edited by W. R. Adams, and I. Veith.
New York: Grune & Stratton, 1959.
102. Liebow, A. A., M. R. Hales, W. E. Bloomer, VV.
Harrison, and G. E. Lindskog. Studies on the lung
after ligation of the pulmonary artery. II. Anatomical
changes. Am. J. Pathol. 26: 177. 1950.
103. Liebow, A. A., M. R. Hales, VV. Harrison, W. Bloomer,
and G. E. Lindskog. The genesis and functional implica-
tions of collateral circulation of the lungs. Yale J. Biol,
and Med. 22:637, '95°-
104. Liebow, A. A., M. R. Hales, and G. E. Lindskog.
Enlargement of the bronchial arteries, and their anasto-
moses with the pulmonary arteries in bronchiectasis.
Am. J. Pathol. 25: 211, 1949.
105. Liebow, A. A., VV. Harrison, and M R. Hales. Experi-
mental pulmonic stenosis. Bull. Intern. Assoc. Med Museums
31: 1, 1950.
106. Liebow, A. A., VV. E. Loring, and VV. L. Felton, II.
The musculature of the lungs in chronic pulmonary
disease. Am. J. Pathol. 29: 885, 1953.
107. Loeb, J. Ueber die Entwicklung von Fischembryonen
ohne Kreislauf. Pfliigers Arch. ges. Physiol. 54:525, 1893.
108. Longland, C. J. The collateral circulation of the limb.
Ann. Roy. Coll. Surg. Engl. 13: 161, 1953.
109. Loring, VV. E., and A. A. Liebow. Effects of bronchial
collateral circulation on heart and blood volume. Lab.
Invest. 3: 175, 1954.
1 10. Luckner. H., and J. Staubesand. Die inkretorische
Funktion des Glomus coccygicum. Z. ges. exptl. Med. 117:
96, 1951.
ill. Makins, G. Gunshot Injuries of the Blood Vessels (8th Am.
ed). Philadelphia: Wood, 1909.
112. Mark, VV. Arterio-venose Anastomosen in Lippen und
Nase der Saugetiere. Z. mikroskop-anat. Forsch. 52: 1, 1942.
113. Mark, VV. Uber arterio- venose Anastomosen, Gefas-
sperren und Gefasse mit epitheloiden Zellen beim
Menschen. Z. mikroskop-anat. Forsch. 50: 392, 1 94 1 .
114. Marchand, P., J. C. Gilrov, and V. A. Wilson. An
anatomical study of the bronchial vascular system and
its variations in disease. Thorax 5: 207, 1950.
115. Marey, E. J. La Circulation du Sang. Paris: Masson, 1881.
116. Masson, P. Innervation des glomus cutanes de l'homme.
Tr. Roy. Soc. Can. V. 30: 31. [936
117. Masson, P. Le glomus neuro-myo-arteriel des regions
tactiles et ses tumeurs. Lyon chir. 21 : 257, 1924.
118. Masson, P. Les Glomus Neuro-Vasculaires. Paris: Hermann,
IQ37-
119. Mendlowitz, M. Cardiovascular shunts (editorial).
Am. J. Med. 22: 1, 1957.
120. Merwin, R. M., and G. H. Algire. The role of graft
and host vessels in the vascularization of grafts of normal
and neoplastic tissue. J. Nat. Cancer Inst. 17: 23, 1956.
121. Miller, W. S. The Lung (2nd ed. 1. Springfield, 111.:
Thomas, 1961.
122. Moore, R. L. Adaptation of the transparent chamber
technique to the ear of the dog. Anat. Record 64: 387, 1936.
123. Muller, J. Entdeckung der bei der Erektion des mann-
lichen Gliedes wirksamen Arterien bei den Menschen
und den Thieren. Arch. Anat. Physiol, wiss. Med. 202, 1833.
124. Mulvihill, D. A., and S. C. Harvey. The mechanism
of the development of collateral circulation. New Engl.
J. Med. 204: 1032, 1931.
125. Mulvihill, D. A., and S. C. Harvey. Studies on col-
lateral circulation. I. Thermic changes after arterial
ligation and ganglionectomy. J. Clin. Invest. 10: 423, 1931.
126. Niden, A. H., and D. M. Aviado, Jr. Effects of pul-
monary embolism on the pulmonary circulation with
special reference to arteriosenous shunts in the lung.
Circulation Research 4: 67, 1956.
127 North, K A K . and A. G. Sanders. The development
of collateral circulation in the mouse's ear. Circulation
Research 6: 721, 1958.
128. North, K. A. K. , A. G. Sanders, and H. VV. Florey.
The development of an anastomotic circulation to trans-
planted tissue. Brit. J. Exptl. Pathol. 41 : 520, i960.
129. Nothnagel, H. L'eber Anpassungen und Ausgleichungen
bei pathologischen Zustanden. Ill Abhandlung. Die
Entstehung des Collateralkreislaufs. Z. klin. Mid. 15:
42, 1889.
130. Paget, J. Lectures on Inflammation. Lecture I. London
Med. Gaz- 10: 965, 1850.
131. Parker, B. M., D. C. Andresen, and J. R. Smith.
Observations on arteriosenous communications in lungs
of dogs. Proc. Soc. Exptl. Biol. Med. 98 : 306, 1 958.
132. Parker, B. M., and J. R. Smith. Studies of experimental
pulmonary embolism and infarction and the deselopment
of collateral circulation in the affected lung lobe. J. Lab.
Clin. Med. 49: 850, 1957.
133. Pepler, VV, J . and B. J. Meyer. Interarterial coronary-
anastomoses and coronary arterial pattern. A comparative
study of South African Bantu and European hearts.
Circulation 22: 14, i960.
134. Popoff, N. VV. The digital vascular system with reference
to the state of glomus in inflammation, arteriosclerotic
gangrene, diabetic gangrene, thrombo-angiitis obliterans
and supernumerary digits in man. A.M. A. Arch. Pathol.
l8: 295. '934-
135. Prichard, M. M. L., and P. M. Daniel. Artcrio-venous
anastomoses in the human external ear. J. Anat. 90:
3°9. l9$6-
136. Prichard, M. M. L., and P. M. Daniel. Arterio-senous
anastomoses in the tongue of the dog. J. Anat. 87: 66,
■953-
137. Prinzmetal, M., E. M. Ornitz, B. Simkin, and 11. C.
Bergman. Arteriosenous anastomoses in liser, spleen,
and lungs. Am. J. Physiol. 152: 48, 1948.
138 Prinzmetal, M , B. Simkin, H. C. Bergman, and H. E.
Kruger. Studies on the coronary circulation. II. The
collateral circulation of the normal human heart by
CHANGES IN VASCULAR PATTERNS
1 -75
coronary perfusion with racioactive erythrocytes and
glass spheres. Am. Heart J. 33: 420, 1947.
139. Quiring, D. P. Collateral Circulation. Philadelphia: Lea
& Febiger, 1 949.
140. Rahn, H., R. Stroud, and C. E. Tobin. Visualization
of arteriovenous shunts by cinefluorography in the lungs
of normal dogs. Proc. Soc. Exptl. Biol. Med. 80: 239, 1952.
141. Rau, G., and W. Schoop. Entwicklung des Kollaternal-
kreislaufes. Arzneimittcl-Forsch. 14: 192, i960.
142. Recklinghausen, F. v. Handbuih der allgememen Pathologic
des Kreistaufs, und der Erndhrung. Stuttgart: Enke, 1883,
PP 35~52-
143. Reichert, F. L. An experimental study of the anasto-
motic circulation in the dog. Bull. Johns Hopkins Hosp. 35 :
385> '924-
144. Reid, M. R. Abnormal arteriovenous communications,
acquired and congenital. III. The effects of abnormal
arteriovenous communications on the heart, blood vessels
and other structures. Arch. Surg. 1 1 : 25, 1925.
145. Reid, M. R. Partial occlusion of the pulmonary aorta
and inferior vena cava with the metallic band. Observa-
tions on changes in the vessel wall and in the heart. J.
Exptl. Med. 40: 289, 1924.
146. Robinson, V. Pathfinders in Medicine. New York : Medical
Life Press, 1929.
147. Rosenberg, M. Z., and A. A. Liebow. Effects of age,
growth hormone, cortisone, and other factors on collateral
circulation. A.M. A. Arch. Pathol. 57: 8g, 1954.
148. Rossatti, B. Observations on the blood supply of the
rabbit's ear and on the experimental new formation of
arterio-venous anastomoses. J. Anat. 90: 318, 1956.
149. Ruvter, J. H. C. Uber einen merkwiirdigen Abschnitt
der Vasa afferentia in der Mauseniere. Z. Zellforsch. 2 :
242, 1925.
150. Sabin, F. R. Origin and development of the primitive
vessels of the chick and of the pig. Carnegie Inst. Wash.
Publ. No. 226 18: 61-124, 1917-
151 . Salisbury, P. F., P. Weil, and D. State. Factors influenc-
ing collateral blood flow to the dog's lung. Circulation
Research 5: 303, 1957.
152. Sandison, J. C. A new method for the microscopic study
of living growing tissues by the introduction of a trans-
parent chamber in the rabbit's ear. Anat. Record 28: 281,
'924-
153. Schenk, W. G., Jr., J. W. Martin, M. B. Leslie, and
B. A Portin. The regional hemodynamics of chronic
experimental arteriovenous fistulas. Surg. Gynecol. Obstet.
1 10: 44, i960.
154. Schlesinger, M. J. New radioopaque mass for vascular
injection. Lab. Invest. 6:1, 1957.
155. Schlesinger, M. J. The relation of anatomic patterns
to pathological conditions of the coronary arteries.
A.M. A. Arch. Pathol. 30: 403, 1940.
156. Schoop, W. Die Entwicklungsbedingungen des arteriellen
KoUateralkreislaufes. Arzneimittel U'ochnschr. 15: 45, igfio.
157. Schoop, W., and W. Jahn. Entwicklungsstadicn arteriel-
ler Kollateralen und ihre begriffliche Definition. Z.
Kreislaufforsch. 50: 249, 196 1.
158. Schroeder, W\, W. Schoop, and E. Stein. Die Durch-
blutung der Extremitat im akuten Sauerstoffmangel
unter besonderer Berucksichtigung der Funktion der
arterio-venosen Anastomosen. Pfliigers Arch, ges Physiol.
259: I24, 1954-
159. Schumacher, S. v. Uber das Glomus coccygeum des
Menschen und die Glomeruli caudales der Saugetiere.
Arch, mikroskop. Anat. 71 : 58, 1908.
160. Schumacher, S. Uber die Bedeutung der arteriovenosen
Anastomosen und der epitheloiden Muskelzellen (Quell-
zellen) Z. mikroskop-anat. Forsch. 43: 107, 1938.
161. Sewell, W. H , and D. R. Koth. A basic observation
on the ability of newly formed capillaries to develop into
collateral arteries. Surg. Forum 9: 227, 1958.
162. Simkin, B., H. C. Bergman, H. Silver, and M.
Prinzmetal. Renal arteriovenous anastomoses in rabbits,
dogs and human subjects. A.M. A. Arch. Internal Med. 81 :
"5> '95°-
163. Sonomoto, A. Studies on the structure and function of
arteriovenous anastomoses in the rabbit's ear. Kyushu
Mem. Med. Sci. 4: 175, 1953.
164. Spalteholz, W. Die Arterien der Hcr-wand. Leipzig:
Hirzel, 1924.
165. Spanner, R. Zur Anatomie der arterio-venosen Anasto-
mosen. Verhandl. deut. Ges. Kreislaufforsch. 18-19:257, 1952.
166. State, D., P. F. Salisbury, and P. Weil. A study of the
bronchial artery How in the dog. Surg. Forum 7: 214, 1957.
167. State, D., P. F. Salisbury, and P. Weil. Physiologic
and pharmacologic studies of collateral pulmonary flow.
J. Thoracic Surg. 34:599, 1957.
168. Staubesand, J., and C. Genschow. Die arterio-venosen
Anastomosen im Loffel des Kaninchens nach graphischen
Rekonstruktionen. Z. Anal. Entwicklungsgeschichte 116:
446, I952-
169. Staubesand, J., and F. Hammersen. Zur Problematik
des Nachweises arterio-venoser Anastomosen im Injek-
tionspraparat. Z. Anat. Entwicklungsgeschichte iig: 365,
1 955" ' 95"-
1 70. Stefani, A. Delia influenza del sistema nervoso sulla
circolazione collaterale. Spenmentale 58: 225, 1886.
171. Straub, W. Zur Muskelphysiologie des Regenwurms.
Pfliigers Arch. ges. Physiol. 79: 379, 1 900.
172. Sucquet, J. P. De La Circulation du Sang dans les Membres
el dans la Tete chez V Homme. Paris : Bailliere, 1 860.
173. Theis, F. V. Effect of sympathetic neurectomy on the
collateral arteriole circulation of the extremities. Experi-
mental study. Surg. Gynecol. Obstet. 57: 737, 1933.
174. Thoma, R. I 'ntersuchungen uber die Histogenese und Histo-
mechamk des Gej'dsssy stems. Stuttgart: Enke, 1893.
1 75. Tobin, C. E. The bronchial arteries and their connec-
tions with other vessels in the human lung. Surg. Gynecol.
Obstet. 95: 741, 1952.
176. Tobin, C. E., and M. O. Zariquiey. Arteriovenous
shunts in the human lung. Proc. Soc. Exptl. Biol. Med.
75 : 827. !95°-
177. Tondury, G., and E. Weibel. Anatomie der Lungenge-
fasse. Ergeb. ges. Tuberk-Forsch. 14:61, 1958.
178. Trueta, J. Studies of the Renal Circulation. Oxford: Black-
well, 1947.
179. Vastarini-Cresi, G. Comunicazioni dirette tra le arterie
e le vene (anastomosi artero-venose ) . Mordt. zool. ital.
13-14: 136, 1902-1903.
180. Verloop, M. C. On the arteriae bronchiales and their
anastomosing with the arteria pulmonalis in some
rodents: A micro-anatomical study. Acta anat. 7:1, 1949.
1276
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
181. Verloop, M. C. The arteriae bronchiales and their
anastomoses with the arteria pulmonalis in the human
lung: A micro-anatomical study. Acta anat. 5: 171, 1948.
1 82. Vidone, R. A., J. L. Kline, M. Pitel, and A. A. Liebow.
The application of an induced bronchial collateral
circulation to the coronary arteries by cardiopneumono-
pexy. II. Hemodynamics and the measurement of col-
lateral flow to the myocardium. Am. J. Pathol. 32 : 897,
1956.
183. Vidone, R. A., and A. A. Liebow. Anatomical and
functional studies of the lung deprived of pulmonary
arteries and veins, with an application in the therapy of
transposition of the great vessels. Am. J. Pathol. 33 : 539,
'957-
184. Volpel, W. Uber die Entstehungsbedingungen des
arteriellen Kollateralskreislaufes. Acta Biol, el Med. Ger.
3- 557. 1959-
185. Wakeley, C. John Hunter and experimental surgery.
Hunterian oration, 1955. Ann. Roy. Coll. Surg. Engl.
16: 69, 1955.
186. VVeibel, E. Die Blutgefassanastomosen in der mensch-
lichen Lunge. Z. Zelljarsch. 50: 653, 1959.
187. Weibel, E. Early stages in the development of collateral
circulation to the lung in the rat. Circulation Research 8:
353. >96°-
188. VVeibel, E. Die Entstehung der Langsmuskulatur in den
Asten der A. bronchialis. Z. Zelljarsch. 47: 440, 1958.
189. Weyrauch, H. B., and C. F. De Garis. Normal and
interrupted vascular patterns in the intestinal mesentery
of the rat. An experimental study of collateral circulation.
Am. J. Anat. 61 : 343, 1937.
190. Williams, R. G. Experiments on the growth of blood
vessels in thin tissue and in small autografts. Anat. Record
133: 465. '959-
191. Williams, R. G. The fate of minute blood vessels in
omentum transplanted as autografts to the rabbit's ear.
Anat. Record 116: 495, 1953.
ig2. Winblad, J. N., K. Reemtsma, J. L. Vernhet, L. P.
Laville, and O. Creech, Jr. Etiologic mechanisms
in the development of collateral circulation. Surgery 45:
io5> '959-
193. Winsor, T., J. H. Payne, N. Rudy, and J. O. Beatty.
Collateral circulation in health and disease. A.M. A.
Arch. Surg. 74: 20, 1957.
194. Wood, D. A., and M. Miller. The role of the dual
pulmonary circulation in various pathologic conditions
of the lungs. ./. Thoracic Surg. 7: 649, 1938.
195. Wright, R. D. The blood supply of abnormal tissues in
the lung. J. Pathol. Bactenol. 47: 489, 1938.
196. Ziegler, E. Experimentelle Untersuchungen uber die Hcrkunjt
der Tuberkelelemrnte mil besonderer Berucksichtigung der
Histogenese der Riesen-ellen. Wiirzburg : Staubinger, 1875.
197. Zoll, P. M., and L. R. Norman. Effect of vasomotor
drugs and of anemia upon interarterial coronary anasto-
moses. Circulation 6: 832, 1952.
198. Zoll, P. M., S. Wessler, and M. J. Schlesinger.
Interarterial coronary anastomoses in the human heart,
with particular reference to anemia and relative cardiac
anoxia. Circulation 4: 797, 1 951.
199. Zuckerkandl, E. Uber die Anastomosen der Venae
pulmonales mit den Bronchialvenen und mit dem
mediastinalcn Venennetze. Sitzber. Akad. Win. Wien,
Malh.-naturw. Kl. 84, Abt. 3: 1 10, 1882.
200. Zweifach, B. W. Basic mechanisms in peripheral vas-
cular homeostasis. In : Factors Regulating Blood Pressure.
Transactions of the Third Conference, May 5-6, 1949,
New York: Macy, 1950, pp. 13-52.
CHAPTER 3!
Methods of measuring blood flow
KURT KRAMER
WILHELM LOCHNER
E. WETTERER
Physiologisches Inslitut der Universitdt, Gbttingen, Germany
Physiologisches Institut der Medizinischen Akademie, Diisseldorf, Germany
Physiologisches Inslitut der Universitdt Miinchen, Miinchen, Germany
CHAPTER CONTENTS
Varied Methods and Instruments for Flow Measurement
Admixing Methods for Measurement of Regional Blood Flow
Flowmeters: Their Theory, Construction, and Operation
Perhaps no other field of physiological methodology encom-
passes such a variety of physical and chemical principles as
that of flow measurement. Principles of measurement may be
and have been developed from almost every topic in physics
textbooks: mechanics {solid, liquid, and gas), sound, elec-
tricity, magnetism, optics, thermodynamics, and atomic
physics. For this reason we have divided the duties of this
section and each author has taken the field of his choice; or,
more correctly, two authors have chosen and one (K. K.),
like Cinderella, has made do with the remainder.
We have set ourselves the task first to review these various
principles, or at least to sketch their historical development,
and second to acquaint the reader with the manner in which
each method fits the special purposes of the investigator.
We have attempted to give a more detailed description of
modern techniques or older ones which are still in use today;
in this we have tried to present not so much an account of
technical details of a piece of apparatus as special suggestions
which will facilitate its use, permit judgment of its reliability,
and guard against sources of error. What we take for reality
sometimes changes so that it is often difficult to distinguish
that which is true only for the moment from that which will
endure.
If older methods no longer in use today are mentioned, it
is to point out particular disadvantages which caused them
to be abandoned. In this way we hope to guide the young
traveler who might otherwise take these fruitless paths again.
K. KRAMER
W. LOCHNER
E. WETTERER
I. Varied methods and instruments for flow measurement
KURT KRAMER
CONTENTS
Outflow Measurements
Venous Outflow Collection
Drop Recording
Methods Based on Ludwig's Principle
Bubble Flowmeter
Venous-Occlusion Methods
Pulse Plethysmography
Photoelectric Plethysmography
Thermal Methods
Thermostromuhr
Skin Blood Flow Measurement Based on Thermal Con-
ductance Measurement
Flowmeters Based on the Measurement of Thermal Con-
ductivity
1277
I278 HANDBOOK OF PHYSIOLOGY --" CIRCULATION II
OUTFLOW MEASUREMENTS
Venous Outflow Collection
the simplest and most reliable way to measure mean
blood flow of an organ consists in the collection of
blood from an opened vein into a graduated cylinder
over a measured period of time. Several venous out-
flow recorders with intermittent indication of flow
rate have been designed (44). In Gaddum's model
(34) blood from the opened vein runs into a cylinder,
the bottom of which is automatically opened after
known periods of time. The collected volume in the
cylinder may be recorded making use of Brodie
bellows, strain gauges, or other devices for measuring
volume or pressure. The Gaddum principle is in fact
a continuous recording of graduated cylinder and
stop watch readings. The dimensions of the apparatus
do not allow measurements lower than 10 ml per
min. Readings every 2 sec furnish reliable results.
The diameter of the cylinder must be adapted to the
amount of blood expected to leave the vein per unit
time. The reliability depends mainly on the rapidity
of emptying the cylinder between collection periods.
Drop Recording
Measurements of flow rates lower than 2 ml per
min can be obtained by recording every drop of blood
leaving the blood vessel. In most devices the drop
closes an electric circuit thereby giving an electro-
magnetic signal. Enumeration of drop signals, how-
ever, is inaccurate and troublesome. Therefore con-
struction of an instrument that records time elapsing
between two drops was a great improvement in the
method. In 1935 Fleisch (29) described an apparatus
in which a motor-driven lever is moved up on a
smoked drum until the drop falls. Closure of the
electric circuit by the drop initiates the interruption
of a coupling link between motor and lever so that the
lever returns to its original level. The next period of
measurement always begins after 0.12 sec regardless
of the height of the lever. It is obvious that such a
recorder is more complex than the simple marking
apparatus and its construction involves a high degree
of precision work.
With the development of electronics a principle
was applied in which the time measurement was
performed by measuring the increase of voltage on a
condenser during the time between drops. The drop
initiates a sudden breakdown of the condenser charge.
PERSPEX
OR POLY-
ETHYLENE
BLOOD
SILICONE
WATER
fig. i. Schematic drawing of the drop chamber according
to Lindgren. A concentric water jacket maintains constant
temperature of the blood. [From Lindgren (62).]
The voltmeter records deflections which are pro-
portional to the time between two drops (63a).
Drop recording has been used mainly to measure
venous outflow. The drawback of all outflow measure-
ments is loss of blood and the necessity for prompt
reinfusion. A definite improvement therefore was the
introduction of a drop chamber that can be used in a
closed circulatory system (58). The blood from a vein
entering the drop chamber falls in drops between
electrodes to the bottom and returns to the distal part
of the dissected vessel. The air cushion does not seem
to introduce any disadvantage in the return of blood
to the vein.
Since electrolysis at the electrode contacts and their
coating with coagulated blood often makes readings
unreliable, a photoelectric drop recording device has
been constructed. A combination of both improve-
ments— the enclosed drop chamber and the photo-
electric recording of drops — seems to be the best of the
fairly simple methods (fig. 1). Lingren"s device (62)
uses a drop chamber filled with silicone instead of
air, thereby avoiding elastic effects especially impor-
tant in arterial blood flow measurements. In recording
of pulsatile arterial flow, one should consider also
that the device may impair the transmission of pulse
waves to the peripheral arterial bed, thereby diminish-
ing original mean flow rate.
METHODS OF MEASURING BLOOD FLOW
12 79
O
Water
^m^v,
'IHMMMl
Rotating
plate
A-\sB
fig. 2. Schematic drawings of
direct recording flowmeters de-
rived from Volkmann's and
Ludwig's principles. [From
Dawes et al. (22).] a: Volkmann
(1850). Open 7"i, close 7\, and
time movement of blood through
U-tube. b: Ludwig (Dogiel,
1867). Time movement of blood
through one chamber and then
reverse chambers by hand, c:
Pavlov (1887). Time movement
of blood through one chamber
and reverse direction of flow
automatically by opening elec-
tromagnetic tap Ti and closing
T-.. d: Dawes et al. (22) Close T
and time movement of blood be-
tween electrodes A and B, restore
blood levels by opening T.
METHODS BASED ON LUDWIG S PRINCIPLE
Volkmann, (81 ) in 1850, was the first to measure
blood flow per unit time in arteries (fig. 2a). His device
consisted of a U-tube inserted in an artery. The U-
tube could be bypassed by two 3-way stopcocks. When
measurements were taken, the U-tube was filled with
saline. After turning both stopcocks simultaneously
the blood flowed through the U-tube, the time was
measured between the moments when the blood
entered and left the U-tube.
This method obviously did not allow continuous
measurement of blood flow. Another drawback was
the repeated infusions of saline with each measure-
ment of flow. Therefore Ludwig and colleagues
(see 24) modified Volkmann's device (fig. 2b). In their
version, the upstream limb of the U-tube was filled
with oil and the downstream one with blood. The
tube itself could be turned by hand through 1800,
to connect the two limbs alternately to the distal and
proximal ends of the artery. The blood was allowed to
enter the upstream limb and the tube had to be turned
when the oil content reached the entrance to the
distal arterial connection. Each turn was marked on
a smoked drum, thereby recording blood flow con-
tinuously. Apparatus based on Ludwig's principle
have been constructed with many modifications and
are still in use (6, 12, 69). The directness of the meas-
urement of flow can be regarded as the main reason
for its popularity. In Ludwig's laboratory Pavlov
(70) developed in 1887, a self-recording flowmeter
based on the same principle. To avoid manipulations
for reversing the direction of flow, he designed his
meter so that blood could be made to enter alternately
either limb of the U-tube by opening and closing
electromagnetic taps, (see fig. 2c). The taps were
automatically operated by means of floats in both
limbs moving with the direction of flow and closing
contacts in the electromagnetic circuit when the rising
float reached the top of its limb.
The Pavlov type flowmeter has been used in numer-
ous modifications. The U-tube can be made very
small for low flow rates. To avoid electrical contacts
within the blood stream the electromagnetic taps
can be controlled by photoelectric relays (63).
An ingenious device based on Ludwig's principle
has recently been described by Dawes et al. (22) (fig.
2d). The upper part of the U-tube is filled with silicone
oil, the lower part of both limbs with blood. The by-
pass can be closed by an electromagnetic tap. In the
inflow limb of the U-tube, two electrodes which oper-
ate a relay for opening and closing the electromag-
netic tap of the bypass are inserted with a distance
between them, such that about 1.5 ml of fluid is
enough to cover both contacts. When blood enters
the proximal limb it touches one electrode. After
1.5 ml more have entered, the other electrode is con-
nected with the first, thereby closing an electric cir-
cuit and setting a relay causing the tap to open. This
allows the blood which entered the proximal limb
12 80
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Return
chamber};
chamber
I P 1
5cm
fig. 3. Sectional view of the Dawes' flowmeter. The cham-
bers are made of Perspex, the washers of Portex sheet, the con-
necting tubes of rubber, and the electrodes of silver wire. The
bag is molded of rubber solution on a form (see text). [From
Dawes et al. (22).]
to return to the arterial stream driven by a pressure
difference between the two limbs. This pressure dif-
ference is supplied by the density difference between
blood and silicone oil. Therefore the authors have
called their apparatus "density flowmeter." The tech-
nical details are more involved than the description
of the principle indicates. As can be seen from figure 3,
in the actual device blood is not allowed to enter the
electrode chamber. A rubber bag filled with saline
placed in the lower part of the proximal limb is com-
pressed by the inflowing blood, emptying its contents
into the part of the limb containing the electrode.
The apparatus of the dimensions given in figure 3
can measure blood flow at rates as high as 45 ml per
min with an absolute accuracy of ±4 per cent. The
pressure drop does not exceed 3 to 4 mm Hg at maxi-
mum rates. The dead space to be filled with blood
amounts to about 4 to 5 ml. The range of flow may be
extended by using larger measuring chambers.
For measurement of time intervals any kind of
ordinate writer (29) can be used. Gaddum's drop-
timer (1938) was used by the authors (44).
A --till simpler sell-recording flowmeter making use
of a single electromagnetic tap in the bypass was first
described by Dawes et al. and has been constructed
recently by Wretlind (88). When the bypass is closed
the blood enters the U-tube, consequently bulging
a membrane in proportion to the volume flow. The
displacement of the membrane is recorded by a lever
on a smoked drum. The tap is automatically opened
every 2 sec so that blood in the U-tube returns to the
artery, allowing the membrane to return to its original
position. Then the cycle begins again with the closing
of the tap. Since the lever indicates blood flow per 2
sec, the record gives direct readings of flow rate.
BUBBLE FLOWMETER
The bubble flowmeter developed by Soskin et al.
(75) consists of a glass tube of known caliber and
Amwiii
fig. 4. Schematic drawing of a bubble flowmeter :B, bubble
reservoir; E, entrance of the bubble into the flowmeter; Mi Aft,
measuring points (platinum electrodes) for timing the passage
of the bubble; S, rubber tubing; V, magnetic tap used as auto-
matic bubble injector. [From Rockemann (72).]
METHODS OF MEASURING BLOOD FLOW
I28l
length which is inserted into the blood stream from
an artery. Near the proximal end of the tube an air
bubble of such a size as to completely fill a short sec-
tion of the tube is injected and time required for the
bubble to pass the length of the tube is recorded. Near
the distal end of the tube the bubble is caught in a
trap. The rate of flow is calculated from the ratio of
volume and time, as is done for all the foregoing re-
corders.
To utilize this principle for continuous recording
of blood flow, an automatic injector for air bubbles,
automatic removal of the bubbles after they pass the
tube, and a recorder of time required for passage of
each bubble are necessary. Several solutions of the
problem have been proposed (13, 15, 33, 57, 64, 65,
87).
A recorder for the passage time of the bubble which
uses photoelectric signals caused by the bubble when
passing a light source and phototube was introduced
by Selkurt (73). Baumgartner et al. (11) added an
automatic bubble injector, the operation of which is
timed by the passage of the bubble past the photocell
detector. A schematic drawing of a recent model (72)
is seen in figure 4. They also studied the over-all prop-
erties of the principle. They could not confirm the
assumption that blood and bubble velocity are equal.
Rather, they found that at low flows the bubble ve-
locity is less and at more rapid flows it is greater than
the blood velocity. The reasons for these deviations
are complex. At high flows the bubble seems to lose
contact with the wall and to move in the faster axial
stream of the blood. Viscosity influences the bubble
velocity somewhat but not seriously. Maximal devia-
tions are not greater than ±5 per cent. However, if
an accuracy within 1 to 2 per cent is desired, they
suggest calibration of the apparatus with blood of the
animal. Pulsation is without influence on the calibra-
tion curve. They used a tube 3.5 mm in diameter
and 35 cm in length. The resistance to flow in such
a tube is low in comparison to that of the peripheral
vascular beds. The maximal flow they studied
amounted to 300 ml per min. At this rate the pressure
gradient was not more than 4 cm of H.O. This value
is comparable with other methods used on opened
vessels. The diameter of 3.5 mm cannot be much in-
creased because at diameters of more than 4.5 mm
the air bubble will not fill the flowmeter tube. Length-
ening the tube increases the sensitivity of the meas-
urement, but also increases resistance to flow. This
fact is to be considered mainly in measurements of
venous flow.
VENOUS-OCCLUSION METHODS
The principle of the venous-occlusion method con-
sists in temporarily blocking the venous outflow from
an organ which is enclosed in a plethysmograph. The
blood that enters the organ via the artery is thereby
retained and indicated as a volume increase by the
plethysmograph.
In this way the method is an almost direct volume
measurement per unit time and thus comparable in
principle to those described in the foregoing para-
graphs. Brodie & Russell (14), who first described the
venous-occlusion principle, were aware of the main
conditions to be fulfilled: ". . . It is obviously essential
that the blockage of the vein must not be maintained
so long as to impede the flow through the capillaries.
Under all ordinary conditions the veins are never
completely filled, so that it is possible to store up in
them a small extra quantity of blood without checking
the inflow into them from the capillaries." As long
as the volume recorder indicates a uniform increase,
the inflow is not impeded. Brodie's method was
adapted by Hewlett & von Zwaluwenburg (56) to
measure blood flow in extremities in man. A plethys-
mograph similar in construction to that of Mosso
was used : a glass cylinder wide enough to enclose
the hand and forearm, from which a rubber tube of
small dimensions leads to the recorder. The whole
system is filled with water to avoid volume errors due
to temperature changes. The veins are blocked by
applying pressure of 50 mm Hg into a pneumatic cuff"
placed on the upper arm.
Several modifications (79) of the original device
have been described (8-io, 46). H. Barcroft's as-
sembly is now most commonlv in use (fig. 5). Mosso's
glass cylinder is replaced by a conic metal tube. The
hand is covered with a large surgical rubber glove
which is fixed outside the plethysmograph to avoid
leakage of water. Since any movement of the forearm
will change the volume of water inside the plethys-
mograph, the circumference of the glove is stiffened
by a diaphragm J/4-inch thick. The diaphragm is
bolted to a 2-inch-wide flange on the end of the ple-
thysmograph by means of metal plates and wing
nuts. (For further details see the original paper.) The
pneumatic cuff is connected through a three-way tap
with a reservoir of compressed air at 60 to 70 mm Hg.
The three-way tap allows inflation of the cuff from
the reservoir and deflation when it is opened to room
air. Two or even four measurements can be taken in
1 min, if blood flow is high. At this rate of measure-
ment the cuff is inflated for only 5 sec. It is found that
1282
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
TO FLOAT RECORDER
THERMOMETER
COLLECTING
CUFF
HIGGINSON S SYRINGE
fig. 5. Plethysmograph for the hand according to H.
Barcroft. The hand is enclosed in a loose-fitting surgical rubber
glove. [From Barcroft & Swan (10).]
during this time the volume of the hand increases
uniformly, indicating that the venous reservoir is not
filled to an extent which would impair capillary flow.
The recording system used in Barcroft's experiments
consists of a small spirometer writing on a smoked
drum. The rubber tubing connection between the
plethvsmograph and spirometer is filled with air. The
use of a spirometer makes it necessary to have air in
the rubber tubing connections between spirometer
and plethysmograph. A small cylinder on top of the
plethysmograph allows control of the water level of
the apparatus. An electrically recorded tracing of
spirometer movements is used in our laboratory mak-
ing use of the electromagnetic principles applied in
the rotameter recording technique (see below).
The use of small rubber cuffs as plethysmography
lias been recommended recently by Dohm. Models
suitable for measurements on forearm and calf, with
which it is possible to secure good venous-occlusion
records and to measure blood flow during muscular
exercise, are especially useful on moving subjects (41).
The plethysmographic cuffs are made of thin-walled
rubber 5 cm wide. The filling pressure can best be
about +40 mm H.,0 and increase with 1 per cent
volume changes of the extremity segment up to about
50 mm HaO. The pressure was measured by a con-
denser manometer.
Avoiding any plethysmographic devices, Whitney
(86) proposes the use of a strain gauge mounted di-
rectly on the limb. It records the changes of tension
due to changes in blood volume. This occlusion tech-
nique furnishes results not remarkably different from
those obtained by using water or air plethysmography.
Assuming that the limb is distended only in the dia-
metrical direction, changes of circumference can be
converted directly into volume changes. However,
corrections for compression of the limb by increases
in blood volume are deemed to be necessary.
A serious objection to the venous-occlusion method
is discussed by Gaskell & Burton (35), who observed
a decrease of blood flow in the dependent leg. These
authors believe in a venovasomotor reflex elicited
by distension of veins. Since the venous-occlusion
method relies on the fact that blood entering the
region of measurement is collected in the veins,
thereby distending them, it is important to the valid-
ity of the method to study the influence of venous
distension upon vascular reflexes.
Greenfield & Patterson (45) showed in experiments
on the forearm at different states of venous distention
that the blood flow, as measured with their venous-
Phase 1
Phase 2
Phase 3
fig. 6. Events during venous occlusion plethysmography.
Actual inflow = actual outflow + apparent inflow. Each divi-
sion on vertical scale for inflow and outflow represents 1 ml/ 100
ml of forearm per min. Total duration of collection: 130 sec.
[From Greenfield & Patterson (45).]
METHODS OF MEASURING BLOOD FLOW
!283
occlusion technique, did not change. Even in states of
venous congestion leading to 2 per cent increase of
the limb volume, the blood flow was almost unaltered.
Less than 1 per cent increase of limb volume is usually
necessary in the application of the venous-occlusion
method. Considering all this, they offer several ex-
planations of Gaskell and Burton's findings.
Greenfield and Patterson give an instructive dia-
gram of events during venous occlusion (fig. 6). In
the first phase of occlusion the plethysmographic
record shows a straight line increase indicating a
constant inflow of blood into the extremity. In the
second phase, the volume increase of the extremity
declines asymptotically, indicating that the inflow
of blood progressively decreases. This can be explained
by the decreasing arteriovenous pressure difference.
In a third phase the venous pressure reaches the oc-
clusion pressure. A new equilibrium obtains in which
there probably is a much lower blood flow through
the extremity. The volume increase in the occluded
region levels off.
The "afterdrop" (a decrease in venous pressure
and limb volume which occurs on release of cuff pres-
sure if the veins are distended) can be considered as a
vasomotor phenomenon. It can also be explained on
mechanical grounds. The release of the pneumatic
cuff opens up an area of compressed veins thereby
acting like a muscle pump on the underlying veins
(2). [See also (1) and (82).]
Pulse Plethysmog raphy
According to Fick's suggestion it is generally ac-
cepted that the first differential quotient of volume
change in an extremity occurring during the course
of the arterial pulse equals the change in the rate of
arterial inflow, if the outflow is constant. Von Kries
(60) and later Frank (31) used tachographs and ple-
thysmographs on the forearm and measured changes
of volume and of the rate of arterial inflow during the
arterial pulse. A combination of pulse plethysmog-
raphy and venous-occlusion technique was used by
Burton (19, 27), Burch (16, 18), and others in order
to obtain absolute values for flow rates during the
time course of the arterial pulse in fingers and hands.
The plethysmographic devices (cylinders, cuffs, re-
cording systems) are adapted to the size of the ex-
tremities in question. The recording systems consist
of capsules covered with thin membranes, the bulging
of which corresponds to volume displacements and
are recorded optically. [For details see (17, 66).]
Photoelectric Plethysmography
Measurements of transparence and reflectance of
infrared light in skin areas furnish almost the same
values for blood volume changes as do the direct
mechanical methods (52-55). The calibration of such
instruments involving calorimetric or venous-occlu-
sion techniques cannot claim great accuracy. How-
ever, the simplicity of the experimental procedure
allows the use of instruments adapted to special pur-
poses not only in various skin areas but also on the
surfaces of organs such as the brain or kidney. The
latter, especially with its high blood content (about
23%), has been the object of blood flow studies uti-
lizing the light absorption properties of Hb in the red
and infrared regions. Procedures have been elabo-
rated (59) that allow measurement of blood content
in cortical and medullary areas of the kidneys, as well
as total blood flow using dye dilution and oxymetric
principles.
THERMAL METHODS
Thermostromuhr
Thermal methods of measuring blood flow are
based on the principles of measurement of heat con-
duction. It is assumed that any condition leading to
loss or gain of heat in the blood stream would depend
among other variables on its volume flow. Gesell &
Bronk (37) cannulated the blood vessel and let the
blood pass through a tube surrounded by a concentric
water jacket which was flushed by a constant flow
of water at room temperature. The loss of heat from
the blood measured by the temperature increase in
the outflowing water was found to be inversely pro-
portional to the volume flow of blood. Corrections
were of course made for different blood temperatures.
The response to changes in flow is slow — of the order
of 1 min.
A few years later H. Rein (71) constructed his
thermostromuhr, which was made for use on un-
opened blood vessels. This method was regarded as a
great improvement both as to lag time and conven-
ience.
The original conception of the thermostromuhr
was based on the assumption that an alternating cur-
rent of high frequency applied to a blood vessel would
heat the blood radially. The temperature rise (A 7")
of this disc of blood would then be proportional to
the product of square of the current (T2) and electrical
1284
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
resistance (/?) and inversely proportional to the blood
flow (V) and specific heat (c). Measurements of blood
flow using a device with two thermojunctions placed
on each side of a pair of heating electrodes seemed
to justify the above assumptions, and permit using
the following equation :
1/-C-"
AT
0.239
(I)
According to this equation, the calibration curve is
hyperbolic. This type of curve has actually been
found in all thermostromuhr devices. However, quan-
titative measurements of AT (3) show values about
ten times higher than expected. This finding indicates
that the assumption of a uniformly heated cross sec-
tion of the blood vessel is not valid. The error in de-
termining AT is found to result from heating the
vessel wall much more than the blood. Due to the
complicated arrangement of electrical resistances to
high-frequency current in the wall, the liberation of
heat in the blood column amounts to only 10 per cent
in arteries and 20 to 40 per cent in veins (83). The
original assumption, therefore, must be revised: the
radial heat gradient is directed from outside to inside
the vessel and not, as suggested by Rein, from inside
to outside. The basic principle by which the stromuhr
measures flow is the change in T in the vessel wall
with blood flow, because of cooling it by the blood
stream. Findings based on this assumption are in good
agreement with the earlier results obtained with the
direct current method, showing that there is no basic
difference between the methods (4, 5, 74, 77).
Further studies (3) on heat dissipation in the wall
of the vessel and in the blood stream have revealed
a temperature profile of complex nature.
The temperature gradients are directed from out-
side to inside the vessel radially and also along the
length of the wall both upstream and downstream
with highest temperature underneath the heating
electrodes.
This temperature profile, however, is not sym-
metrical for two reasons: first, since heating electrodes
are attached to a segment of the wall, the tempera-
tures measured in the plane of the heating electrodes
are higher than in a plane at an angle to it; second,
since the blood stream cools the upstream wall sec-
tion more than the downstream section, the tempera-
ture profile is lengthened in the downstream direction.
The temperature profile changes with blood flow.
The asymmetry of temperature distribution along
the wall of the vessel increases with decreasing blood
flow. The temperature of the upstream section changes
less than that of the downstream section. It is this
fact which makes the device a flowmeter.
From Gregg's investigations (43, 74) on direct cur-
rent stromuhrs it was expected that pulsations of the
blood stream should distort the temperature profile
in an unpredictable manner. Wever & Aschoff (84),
working with a stream having large pulsations, found
that thermojunctions arranged at an angle of 90°
to the heating electrodes yield false readings which
are opposite to those obtained at an angle of 0°. The
practical application of these studies has led to the
construction of a device using ring electrodes, by
which temperatures of the complex profile are aver-
aged, and errors due to pulsation are avoided. These
electrodes also compensate for errors resulting from
nonlinearity of the calibration curve (fig. 7). Since
the highest temperature exists on the outside of the
vessel wall, any uncontrolled heat dissipation to the
outside of the unit would lead to an undetectable
error of measurement. In the new models (3), a
double wall including air for thermoisolation is in-
troduced.
Where backflow occurs, the deviation in the meas-
urements is always in the direction of increasing flow.
The effect of backflow can be diminished by means
of asymmetrical placement of the thermojunctions
(77). When the downstream thermojunction is placed
close to the heating electrodes and the upstream junc-
tion is farther off, the backflowing blood heated during
its passage through the hot vessel wall will reach the
upper junction later and will have less influence on
the measurement.
Although methods based on the thermostromuhr
principle have been abandoned during the last dec-
ades because of inherent inaccuracies (7, 25, 26, 43),
the new analysis given by Aschoff and Wever has
revived interest in the matter.
Skin Blood Flow Measurement Based on Thermal
Conductance Measurement (20, 32, 48-51, 85)
Since the heat produced in animals and humans is
transported mainly by blood flow, the heat flow of a
defined area of the skin is related to blood flow through
it. However, it is obvious that any change of tempera-
ture gradient, such as that induced by changes of the
surrounding temperature, will influence the heat flow
and therefore invalidate the measurement of blood
flow. The best values are obtained with devices which
measure heat flow and temperature gradient simul-
taneously.
The following equation gives a measure of blood
METHODS OF MEASURING BLOOD FLOW
1285
flow from the relationship of these two variables in
the form of a thermal conductance coefficient :
a (2)
Tc
Tsk
where Q equals heat flow in (cal cm2 sec), Tc = core
temperature and T,k = skin temperature in °C. The
dimension of A is calories per square centimeter second
°C. The Tc as measured does not always represent the
temperature of the arterial blood in the region under
study. Heat may be lost during the passage of blood
irom the core to the site of measurement. Hensel (50)
points out that, among other things, the special
geometry of the skin area, insulation, local metabo-
lism, and countercurrent heat exchange between
arteries and veins may modify k without changes of
blood flow.
If it is possible to keep these variables constant,
relative changes in blood flow in skin areas can be
estimated by measuring k. Several methods are pro-
posed. The measuring device should avoid the "re-
action-error'' which would occur if calorimeter de-
vices are used with large heat capacities and
temperatures different from those of the skin (48).
However, it is necessary that the heat resistance of
the device be made much lower than that of the skin,
the resistance of which is determined by the blood
flow.
A device (85) that fulfills the above conditions con-
sists of a cork plate 1 mm thick covered with two silver
plates with two thermojunctions. The unit is fixed
tightly on the skin. The temperature gradient meas-
ured between these plates is proportional to Q, the
heat flow from the skin. The temperature difference
Tc — T,k is measured by connecting the skin thermo-
junction with a third junction placed in the mouth
or rectum. The quotient Q/(TC — Tsk) is measured
by a bridge circuit or by a ratiometer. Q can only be
measured if the thermal conductivity of the cork
plate is known. This value must be determined ex-
perimentally. Synchronous measurement of k and
blood flow of the finger with the venous occlusion
technique furnish a fairly good proportionality. This
was found at different room temperatures (15-30 C)
as well as at different skin and rectal temperatures.
Also, insulation of the arm did not influence the meas-
urements. It seems therefore that, according to Aschoff
and Wever's results, blood flow is the main factor
determining k.
Vein
Artery
7
50
100 150 200
fig. 7. Measurements with original Rein elements and
Aschoff and Wever ring-element. Figure shows effect on flow
readings when heating electrodes are placed at 900 and o° to
the thermojunctions. The compensating effect of a ring unit in
which the thermojunctions are fixed on silver rings surrounding
the vessel is shown to be effective for pulsations up to 1 20% of
mean flow. Abscissa = oscillations in percentage of mean flow.
Ordinate = thermostromuhr readings. [From Wever & Aschoff
(84)-]
Flowmeters Based on the Measurement
of Thermal Conductivity
In 1 92 1 the mathematician Carlslaw showed that
when a special source of heat is surrounded by an
infinitely extended mass of material a steady state is
approached in which the relation between heat pro-
duction, such as that generated electrically, and heat
loss is described by the equation:
IZR
4vrATX
(3)
where / = electric current heating a filament with
the resistance R, r = radius of the sphere, T = tem-
perature of the sphere, and X = the thermal con-
ductivity constant.
In application to our problem we have to consider
that X, because of the complexity of the tissue, is not
a simple constant but depends on several parameters
of the tissue under study (80), and most importantly
on blood flow. This dependence on flow provides the
basic principle for measurement with this type of
flowmeter. Experimental data (42) provided by meas-
urements of AT on living organs have shown that a
1286
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
linear relationship exists between blood flow and the
apparent increment of X. Direct readings of X may
be recorded (42) by keeping T constant through
variation of 7. It should, however, be borne in mind
that only relative values for blood flow can be ob-
tained. The principle can be applied either to surfaces
or to inner regions of organs. For these various ap-
plications several types of instruments have been
developed. As an example, for measurements of blood
flow through deep layers, heat source and temperature
measuring units are contained in a needle (38, 39,
42, 47, 68).
Hensel's modification shown in figure 8 contains
both thermojunctions within the needle, one at the
tip together with the heating wire, and the other at
the middle of the needle. According to Graf & Rosell
(40), reliable measurements are obtained only when
the tip of the needle is placed in close proximity to a
vessel, either an artery or vein. This condition is
checked by comparison of X-values obtained in the
ischemic and normal state which should be in the
order of 1 X io~4 cal per sec cm C. Lower values
Connections
to healing sou/re
nj^—to galvanometer
Cross
section
-Cannula
-Heating wire
rhermojunction
(hot)
fig. 8. Schematic drawing of longitudinal and cross sections
of Hensel's needle for measuring thermal conductivity in tissue.
[From Hensel el at. (47)-]
T her moj unction
( cold J
□ V2A steel
f~1 Copper
^ Iron
■I Constantan
r~"l Solder metal
I I Glass
■
350J
A
■ ■
-
■
4
O
250-
»*
■
■
■
■
•
■
•
•
• •
150-
a
'•a
1954-
17. Burch, G. E. Digital Plethysmography. New York: Grune &
Stratton, 1954.
18. Burch, G. E. Recording the time course of digital rate of
flow. J. Appl. Physiol. 7: 95-104, 1954.
ig. Burton, A. C. The range and variability of the blood
flow in the human fingers and the vasomotor regulation of
body temperature. Am. J. Physiol. 127: 437, 1939.
20. Burton, A. C. The direct measurement of the thermal
conductance of the skin as an index of peripheral blood
flow. Am. J. Physiol. 129: 326, 1940.
21. Burton, A. C. The thermal insulation of the tissues of
the body. In: Man m a Cold Environment, edited by A. C.
Burton and O. G. Edholm. London : Arnold, 1 955, p. 73.
22. Dawes, G. S., J. C. Mott, and J. R. Vane. The density
flowmeter, a direct method for the measurement of the
rate of blood flow. J. Physiol., London 121 : 72, 1953.
23. Delaunois, A. L., and L. A. Rovati. A new method for
continuous measurement of cardiac output. Arch, intern
pharmacodynamic 116:228-236, 1958.
24. Dogiel, J. Die Ausmessung der stromenden Blutvolumina.
Arb. physiol. (Leipzig: Anstalt), p. 196, 1867.
25. Dorner, J. Fehlermoglichkeiten bei der Durchblutungs-
messung mit der Diathermie-Thermostromuhr nach H.
Rein. Arch, exptl. Pathol. Pharmacol. 220: 490, 1953.
26. Dorner, J. Beitrag zur Frage einer quantitativen Stro-
mungsmessung mit der Thermostromuhr nach H. Rein.
Arch, exptl. Pathol. Pharmacol. 221 : 312-322, 1954.
27. Edwards, M., and A. C. Burton. Correlation of heat
output and blood flow in the finger, especially in cold-
induced vasodilatation. J. Appl. Physiol. 15: 201-208, i960.
28. Felix, E. Erganzende Bemerkungen zur Blutstrommessung
mit Thermistoren. Z. Biol. 1 08 : 121, 1 956.
29. Fleisch, A. Die Registrierung zeitlicher Intervalle direkt
als Ordinate mit dem Pulszeitschreiber. In : Abderhalden,
Handbuch der biologischen Arheitsmethoden. Wien: 1935, vol.
5, sect. 8, p. 905.
30. Fleming, D. G. Precautions in the physiological applica-
tion of thermistors J. Appl. Physiol. 13: 529, 1958.
31. Frank, O. Konstruktion und Theorie eines neuen Tacho-
graphen. Z. Biol. 32: 303, 1908.
32. Franke, E. K. Uber den Zusammenhang der kapillaren
Durchblutung mit der Warmeleitfahigkeit der Haut. Pflii-
gers Arch. ges. Physiol. 270: 657-659, i960.
33. Friedburg, H., U. E. Schafer, and R. Taugner. Verbes-
serungen am Bubble-Flowmeter mit automatischer
Registrierung. Arch, exptl. Pathol. Pharmacol. 233: 567-568,
1858.
34. Gaddum, J. H. An outflow recorder. J. Physiol. London,
67: 16 P, 1929.
35. Gaskell, P., and A. C. Burton. Local postural vasomotor
reflexes arising from the limb veins. Circulation Research
1:27. '953-
36. Gersmever, E. F., H. Weyland, and H. Spithbarth.
Zur Messung der Blutstromgeschwindigkeit mit Ther-
mistoren in grossen Gefassen des Menschen. Klin. U'ochschr.
36:872, 1958.
37. Gesell, R., and D. W. Bronk. A continuous thermo-
electric method of recording the volume-flow of blood.
Am. J. Physiol. 79:61, 1926-27.
38 Gibbs, F. A. A thermoelectric blood flow recorder in the
form of a needle. Proc. Soc. Exptl. Biol. Med. 31 : 141-146,
■933-
39. Gibbs, F. A., E. L. Gibbs, and W. G. Lennox. The cere-
bral blood flow in man as influenced by adrenalin, caffein,
amyl nitrite and histamine. Am. Heart J. 10: 916-924, 1935
40. Graf, K., and S. Rosell. Untersuchungen zur fort-
laufenden Durchblutungsregistrierung mit Warmeleitson-
den, Beobachtungen an der Skeletmuskulatur der Katze.
Acta Physiol. Scand. 42 : 5 1 , 1 958.
41. Graf, K., and A. Westersten. Untersuchungen uber
Eigenschaften und Verwendungsmoglichkeiten eines
flexiblen Extremitatenplethysmographen. Acta. Physiol.
Scand. 46: 1 -1 8, 1959.
42. Grayson, J. Internal calorimetry in the determination of
thermal conductivity and blood flow. J. Physiol., London
118: 54, 1952.
43. Gregg, D. E., W. H. Pritchard, R. W. Eckstein, R E.
Shipley, A. Rotta, J. Dingle, T. W. Steege, and J. T.
Wearn. Observations on the accuracy of the thermo-
stromuhr. Am. J. Physiol. 136: 250, 1942.
METHODS OF MEASURING BLOOD FLOW
1289
46
47
44. Green, H. D., Venous drainage recorders. In: Methods in 63a
Medical Research. Chicago: Yr. Bk. Pub., 1948, vol. 1, p. 68.
45. Greenfield, A. D. M., and G. C. Patterson. The effect 64.
of small degrees of venous distension on the apparent rate
of blood inflow to the forearm. J. Physiol., London 125:
525> IQ54-
Greenfield, A. D. M. A simple water-filled plethysmo- 65.
graph for the hand or forearm with temperature control.
J. Physiol. , London 123: 62, 1954.
Hensel, H., J. Ruef, and K. Golenhofen. Fortlaufende
Registrierung der Muskeldurchblutung am Menschen mit 66.
einer Kalorimetersonde. Pfliigers Arch. ges. Physiol. 259:
a67. !954-
48. Hensel, H. Ein neues Verfahren zur peripheren Durch- 67.
blutungsregistrierung an beliebigen Korperstellen. Z.
Kreislaufforsch. 41 : 252, 1952.
49. Hensel, H., and F. Bender Fortlaufende Bestimmung der 68.
Hautdurchblutung am Menschen mit einem elektrischen
Warmeleitmesser. Pfliigers Arch. ges. Physiol. 263: 603, 1956.
50. Hensel, H. Kritische Betrachtungen zur Messung der 69.
Hautdurchblutung mit thcrmischen Methoden. Klin.
Wochschr. 34: 1273, 1956.
51. Hensel, H. Mefikopf zur Durchblutungsregistrierung an 70.
Oberflachen. Pfliigers Arch. ges. Physiol. 268: 604, 1959.
52. Hertzmann, A. B. The blood supply of various skin areas
as estimated by the photoelectric plethysmograph. Am. J. 71.
Physiol. 124: 328, 1938.
53. Hertzmann, A. B., W. C. Randall, and K. E. Jochim.
The estimation of the cutaneous blood flow with the photo- 72.
electric plethysmograph. -4m. J. Physiol. 145: 716, 1946.
54. Hertzmann, A. B., W. C. Randall, and K. E. Jochim.
Relations between cutaneous blood flow and blood content 73-
in the finger pad, forearm and forehead. .4m. J. Physiol.
150: 122, 1947.
55. Hertzmann, A. B. Photoelectric plethysmography of the 74.
fingers and toes in man. Proc. Soc. Exptl. Biol. Med. 37 :
529> '937-
56. Hewlett, A. W., and J. van Zwaluwenburg. Method
for estimating the blood flow in the arm. Heart 1 : 87, 1909. 75.
57. Hierholzer, K., K. Frohner, and S. Schleer. Ein neuer
Blasengeber fur das Bubble-flowmeter. Pfliigers Arch. ges.
Physiol. 264: 94, 1957. 76.
58. Hilton, S. M. A perspex drop chamber. ./. Physiol.,
London 117: 48 p. 1952. 77.
58a.K.ANZow, E. Quantitative fortlaufende Messung von
Durchblutungsanderungcn in der Hirnrinde. Pfliigers Arch,
ges. Physiol. 273: 199, 1961. 78.
59. Kramer, K., K. Thurau, and P. Deetjan. Hamodynamik
des Nierenmarks. Pfliigers Arch. ges. Physiol. 270: 251, i960.
60. Kries, J. von. Uber ein neues Verfahren zur Beobachtung 79.
der Wellenbewegung des Blutes. Arch. Anat. u Physiol.
Anal. Abt. (Physiol. Abt.) P, 254, 1887.
61. Katsura, S., R. Weiss, D. Baker, and R. F. Rushmer. 80.
Isothermal blood flow velocity probe. IRE Trans, on Med.
Electronics. Me-6: 283, 1959.
62. Lindgren, P. An improved method for drop recording of 81.
arterial or venous blood flow. Acta Physiol. Scand. 42 : 5,
1958. 82.
63. Lu, F. C. , and K. I. Melville. A new apparatus and
procedure for continuous registration of changes in coronary
flow concurrently with changes in heart contractions.
J. Pharmacol. Exptl. Therap. 99: 277, 1950. 83.
Lullies, H. Ein Zeitordinatenschreiber auf elektrischer
Grundlage. Pfliigers Arch. g«r. Physiol. 241: 354, 1938.
Lutz, J. Bubble-flowmeter mit unmittelbarer Anzeige der
DurchflufigrolSe und elektrischer Registrierung auf einem
Direktschreiber. Arch, exptl. Pathol. Pharmacol. 238: 228,
i960.
Lutz, J. Blasen-Stromuhr (Bubble-flowmeter) mit Rohr-
elektroden und einem Mel5wertumformer zur linearen
Registrierung auf Direktschreibern. Arch, exptl. Pathol.
Pharmacol. 240:341, 1961.
Mead, J., and R. C. Schoenfeld. Character of blood
flow in the vasodilated finger. J. Appl. Physiol. 2 : 680,
1950.
Mellander, S., and R. F. Rushmer. Venous blood flow
recorded with an isothermal flowmeter. Acta Physiol. Scand.
48: 13, i960.
Mowbray, J. F. Measurement of tissue blood flow using
small heated thermocouple needles. J. Appl. Physiol. 14:
647. '959-
Olerud, S. Experimental studies on portal circulation at
increased intra-abdominal pressure. Acta Physiol. Scand.
30: Suppl. 109, 1953.
Pavlov, I. P. Uber den Einfluss des Vagus auf die Arbeit
der linken Herzkammer. Arch. Anat. Physio'.. (Physiol.
Abt.) 1887, p. 452.
Rein, H. Uber Durchblutungsmessungen an Organen in
situ, insbesondere mit der Thermostromuhr. Ergeb.
Physiol, exptl. Pharmacol. 45: 514, 1944.
Rockemann, W. Ein Bubble-Flowmeter mit elektrischer
Blasenregistrierung und vereinfachtem Blasengeber. Pflii-
gers Arch. ges. Physiol. 272: 393, 1961.
Selkurt, E. E. An optically recording bubble flowmeter
adapted for measurement of renal blood flow. J. Lab. Clin.
Med. 34 : 1 46, 1 949.
Shipley, R. E., D. E. Gregg, and S. T. Wearn. Opera-
tive mechanism of some errors in the application of the
thermostromuhrs method to the measurement of blood
flow. Am. J. Physiol. 136: 263, 1942.
Soskin, S., W. S. Priost, and W. J. Schultz. Influence
of epinephrine upon exchange of sugar between blood and
muscle. Am. J. Physiol. 108: 107, 1934.
Suckling, E. E., and A. Vogel. Thermistor bridge for
blood flow measurement. J. Appl. Physiol. 15: 966, i960.
Schmidt, C. F. , and A. M. Walker. A thermostromuhr
operating on storage -battery current. Proc. Soc. Exptl.
Biol. Med. 33:346, 1935-
Schmidt, L., and R. Engelhorn. Die Abhangigkeit der
Coronardurchblutung vom arteriellen Blutdruck. Arch,
exptl. Pathol. Pharmacol. 218: 115, 1953.
Stead, E. A., Jr., and P. Kunkel. A plethysmographic
method for the quantitative measurement of the blood
flow in the foot. J. Clin. Invest. 17:711, 1 938.
Vendrik, A. J. H., and J. J. Vos. A method for the
measurement of the thermal conductivity of human skin.
J. Appl. Physiol. 11:211-215, 1957.
Volkmann, A. W. Die Haemodynamik. Leipzig: Breitkopf
und Hartel, 1850.
Wallace, W. F. M. Does the hydrostatic pressure of the
water in a venous occlusion plethysmograph affect the
apparent rate of blood flow to the forearm? J. Physiol.,
London 143: 380, 1958.
Wever, R. Die Verteilung des Diathermie-stromes im
1290
HANDBOOK OK PHYSIOLOGY
CIRCULATION II
Blutgefass bei der Thermostromuhr-Messung. Pflugers
Arch. ges. Physiol. 262: 1, 1955.
84. Wever, R., and J. Aschoff. Durchfluflmessung mit der
Diathermie-Thermostromuhr bei pulsierender Stromung.
Pflugers Arch. ges. Physiol. 262: 152, 1956.
85. Wever, R., and J. Aschoff. Die Warmcdurchgangszahl
als DurchblutungsmalS am Menschen. Pflugers Arch. ges.
Physiol. 264: 272, 1957.
86. Whitney, R. J. The measurement of volume changes in
human limbs. ./. Physiol., London 121 : I, 1953.
87. Winder, C. V., J. Wax, and R. W. Thomas. Stable
precision in a readily assembled, continuously recording
bubble-flowmeter. J. Lab. Clin. Med. 42: 766, 1953.
88. Wretlind, A. Recorder for blood flow determination.
Acta Physiol. Scand. 40: 196, 1957.
89. Zijlstra, W. G., J. R. Brunsting, and L. B. Slikke,
Intravascular and intracardiac blood velocity patterns
recorded by means of NTC resistors. Xature 184: Suppl
13. 99>. '959-
II. Admixing methods for measurement of regional blood flow
WILHELM LOCHNER
CONTENTS
Blood Tissue Exchange Methods
Nitrous Oxide Method
Measurement of cerebral blood flow
Measurement of coronary blood flow
Other Test Substances
Test Substance Dilution Methods
Measurement of Coronary Blood Flow
Measurement of Cerebral Blood Flow
Measurement of Flow in Other Organs
Measurement of Flow in a Blood Vessel Without Interposing
an Organ
BLOOD TISSUE EXCHANGE METHODS
X 1I1 mis Oxide Method
MEASUREMENT OF CEREBRAL BLOOD FLOW. The nitrOUS
oxide method for determination of cerebral blood
flow was developed by Kety & Schmidt (20, 22) in
1945. Since then it has become a standard method
for determinations in man of both cerebral and cor-
onary blood flow, especially because extensive opera-
tive procedures can be avoided. The nitrous oxide
method makes use of Fick's principle of blood flow
estimation. The test substance, nitrous oxide, is an
easily diffusible, inert gas which diffuses into the tis-
sues fast enough to allow equilibrium between gas
tensions in tissue and venous capillaries. With a known
partition coefficient of the gas, and under the assump-
tion that equilibrium between tissue and blood is
reached, the amount of test substance taken up by
100 g of tissue can be calculated (21). Simultaneous
measurement of arteriovenous nitrous oxide difference
then permits calculation of the blood flow per minute
per 100 g of tissue. Applying Fick's principle, the
formula (22) is:
CBF-
IOOVu-S
wherein
.1 = arterial N-iO — concentration
I" = venous N;0 — concentration
S = partition coefficient for NoO between blood and
tissue
['„ = venous N20 concentration after equilibrium reached
in tissue during time u
CBF = cerebral blood flow per 100 g brain tissue per min
The procedure of measurement is as follows: The
patient breathes a gas mixture of oxygen, nitrogen
and 1 5 per cent nitrous oxide over a period of 10 min
(time u). During this time, five consecutive blood
samples are taken simultaneously from the internal
jugular vein and from a peripheral artery. The sam-
ples must be collected under anaerobic conditions.
They are analyzed for NoO according to the method
of Orcut & Waters (33). [See also Kety (23).] Figure 1
shows arterial and venous time-concentration curves
of nitrous oxide in a typical determination. As can
be seen from the figure, ten blood samples have to be
analyzed, an undesirable feature of the method. A
modification of this method has been proposed by
Scheinberg & Stead (38) and by Bernsmeier &
Siemons (3). Intermittent sampling is replaced by
continuous sampling of only two probes, one arterial
METHODS OF MEASURING BLOOD FLOW
1291
4 -
3 -
0 2
fig. i. Typical arterial (A) and internal jugular (V) curves
of N20 concentration during a 10-min period of inhalation of
[5% Nl.O. [From Kety & Schmidt (22).]
and one venous, during the measuring period of 10
min. The advantages of this modification are seen
in the fact that a) only one person is needed for taking
the samples, b) less blood is taken from the patient,
and c) the number of gas analyses is reduced from
ten to three (25). The modified method yields results
identical with those obtained with the original
method. However, Kety (24) prefers his primary pro-
cedure of intermittent sampling since he believes that
the course of arterial and venous time concentration
curves allows an estimation of the volume of blood
from extracranial vessels which has been intermixed
with cerebral blood flow. This point is of general im-
portance and will be briefly discussed here. The main
premise is that representative mixed venous blood
of the brain is obtained for measurement of gas con-
centrations. It is not necessary for the total blood
flow of the organ to pass through one vein. However,
it is required that the concentration of test substance
be equal in all veins. Samplings from the bulbus
cranialis of the internal jugular vein have proved to
be fairly free of extracranial blood. Shenkin et al. (40)
estimate a maximal admixture of 2 to 3 per cent. For
proper use of the partition coefficient of nitrous oxide
in blood and tissue — its value for cerebral tissue is
about one — it is important that concentration equilib-
rium between tissue and blood has been reached.
Even small differences between arterial and venous
concentration lead to errors, as Sapirstein & Ogden
(37) have shown. This fact would seem to make the
original intermittent sampling technique the method
of choice. Simultaneous measurements of cerebral
blood flow in monkeys by the nitrous oxide method
and by bubble flowmeter show good agreement (22).
MEASUREMENT OF CORONARY BLOOD FLOW. Soon after
the introduction of the nitrous oxide method for the
determination of cerebral blood flow, the method was
used to measure coronary blood flow (6, 7, 12). The
blood of the coronary sinus is representative of the
left ventricular coronary flow, and since the improve-
ment of catheterization technique has made it pos-
sible to sample blood from the coronary sinus, the
nitrous oxide method can be applied successfully in
man (4). In coronary blood flow experiments on dogs,
Gregg and co-workers (14) found good agreement
between values obtained by use of the rotameter and
the nitrous oxide method. A series of investigations
has been undertaken using the desaturation course
of nitrous oxide. The results were similar to those
obtained by the method of saturation [Goodale &
Hakel (1 1 ) and Bargeron et al. (2)].
Other Test Substances
Radioactive krypton 85 has been proposed by
Lassen & Munck (26, 29) for use in the determination
of cerebral blood flow. The procedure is very similar
to that of the nitrous oxide method. The application
of krypton 85, although it allows greater accuracy,
has the disadvantage of requiring special instrumenta-
tion and the risk to the patient of radiation exposure.
Munck & Lassen have recommended that internal
jugular blood should be sampled bilaterally because
concentration of the test substance may differ in the
two veins. Blood flow is then calculated twice and
the mean is taken. Since gaseous test substances re-
quire special care in sampling and storing of blood,
and since the analyses are time consuming and diffi-
cult under conditions of gaseous anesthesia, Huckabee
(17) proposed the use of 4-aminoantipyrine. This is
a nonvolatile, biologically inert substance, which
diffuses rapidly from blood into tissue fluid, and is
relatively easy to measure. [See also (14), (19), (18).]
TEST-SUBSTANCE DILUTION METHODS
These also employ the Fick principle. Blood flow
through an organ is determined from the ratio of the
amount of injected test substance to its concentration
in the effluent blood. In case there is, for any reason,
some of the indicator substance in the blood at the
1292
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
time of injection, the arteriovenous difference in con-
centration of the substance may be substituted. A
rapid injection of the substance into the blood stream
is used. This procedure was originally developed to
determine cardiac output (15). Description of the
method and its theoretical implications have been
given by Hamilton (14a) and Zierler (41). For ap-
plication of the method to blood flow of any organ,
the general formula holds:
F-
f0°°c(t)dt
wherein
F = blood flow
C = concentration of test substance in the effluent blood
I = time elapsing during passage of test substance
m = amount of test substance injected.
Measurement oj Coronary Blood Flow
Hirche & Lochner (16) have adapted the method
to determine coronary blood flow in anesthetized
dogs. A main branch of the left coronary artery (de-
scending branch or circumflex branch) and the cor-
onary sinus are catheterized. Cardiogreen (5) is used
as the test substance. For continuous measurement of
dye concentration, the mixed venous blood of the
heart muscle is drawn by a pump from the sinus cath-
eter through a cuvette photometer (27). When the
dye is injected either into the descending or circum-
flex branch, time-concentration curves of equal area
are obtained. The sinus catheter should be placed
close to the outflow. Hirche and Lochner have con-
cluded that the method gives values of the sinus out-
flow only. Since about 10 per cent of left coronary
artery blood may not be returned by the coronary
sinus, and therefore a proportionate amount of dye
does not appear with this venous outflow, 10 per cent
of the calculated blood flow must be subtracted for
quantitative measurement. Values obtained with this
method are in agreement with those measured by
other methods reported in the literature. The meas-
urements can be repeated in intervals of 1 to 2 min
and in practically unlimited number.
Measurement oj Cerebral Blood
Flow (10, 22, 30-32, 39)
The human brain receives nearly all of its blood
through two vertebral and two internal carotid ar-
teries. The blood leaves the brain through two main
veins, the two internal jugulars. It would be theoreti-
cally justified to apply the test-substance injection
method if, following injection into one arterial branch,
one could obtain identical time-concentration curves
in all veins. This would indicate that the mixing of
the test substance in all brain vessels was complete.
However, the results described below show that such
mixing is not obtained. Injection of test substance into
one internal carotid vields three distinct types of time-
concentration curves in the separate internal jugular
veins.
/) The test substance may appear only on one
side. This finding would allow the conclusion that
blood flow of only one hemisphere is measured. 2) The
concentration curves may be identical in both in-
ternal jugular veins. This would correspond to the
ideal case where blood flow through the brain as a
whole is measured. 3) Most frequently, however, it
happens that although the dye appears in both in-
ternal jugular veins, the concentrations differ to a
high degree.
It is proposed to average the blood flow values
obtained from both time-concentration curves. How-
ever, it seems questionable whether this procedure
yields accurate quantitative values for brain blood
flow. Using the test-substance injection method,
Shenkin et al. (40) have studied the "dynamic anat-
omy of the brain," mainly to test the validity of the
nitrous oxide method. The advantage of the latter
lies in the fact that all cerebral arteries show the same
concentrations of nitrous oxide at any period of time.
In spite of this, the measurements made from the in-
ternal jugular veins sometimes do not give identical
values (29). When the method of sudden and short
injection of the test substance is applied, the inter-
pretation is much more complicated. Even in the
simple case of two identical time-concentration curves
cerebral blood flow cannot be correctly estimated,
since not all the test substance appears at the internal
jugular measuring points. About 22 per cent of the
blood flow through the external jugular vein is derived
from internal carotid blood, as Shenkin has shown.
One cannot assume that the test substance leaving
the brain via the external jugular vein has mixed
thoroughly with all the blood passing the brain. The
difference between two time-concentration curves
from the internal jugular veins speaks against it. This
means that accurate measurement of cerebral blood
flow cannot be obtained with the test-substance in-
jection method, even when concentration curves of
both internal jugular veins are recorded.
METHODS OF MEASURING BLOOD FLOW
l293
Measurement of Flow in Other Organs
The blood flow of the extremities has been measured
in man using the test-substance dilution method
[Andres et al. (i)] and in the isolated kidney by Loch-
ner & Ochwadt (28). Piiper (35, 36) determined the
site of main resistance to flow in the vascular bed of
the lungs by injection technique, and also the site of
capillaries in the vascular volume of the isolated lungs.
Measurement of Flow in a Blood Vessel
Without Interposing an Organ
In the procedures described above the perfusion of
an organ was measured. The organ served as a mixing
chamber, and a reliable time-concentration curve of
the test substance was obtained as it left the organ.
Peterson et al. (34), on the other hand, have developed
a method to measure the outflow of the left ventricle
which does not involve mixing of the blood and in-
dicator substance within the heart. The indicator is
injected into the root of the aorta and its concentra-
tion is measured in an artery. In the same way, Grace
et al. (13) have measured flow in the thoracic aorta.
Whereas the above-mentioned two groups used dyes,
Frank et al. (8) used cold solutions. Fronek & Ganz
(9), using cold injections, were able to measure blood
flow in individual small vessels by placing the in-
jection and recording sites in close proximity. Since
laminary flow is dominant in the blood vessels, special
care must be taken to achieve complete mixing of the
test substance with the blood. This can be done by
choosing a proper diameter and arrangement of holes
at the tip of the injection catheter or needle.
REFERENCES
1. Andres, R., K. L. Zierler, H. M. Anderson, W. N.
Stainsbv, G. Cader, A. S. Ghrayyib, and J. L. Lilien-
thal, Jr. Measurement of blood flow and volume in the
forearm of men; with notes on the theory of indicator-
dilution and on production of turbulence, hemolysis and
vasodilation by intra-vascular-injection. J. Clin. Invest. 33:
482, 1954.
2. Bargeron, L. M., D. Ehmke, F. Gonlubol, A. Castel-
lanos, A. Siegel, and R. J. Bing. Effect of cigarette
smoking on coronary blood flow and myocardial metabo-
lism. Circulation 15: 251, 1957.
3. Bernsmeier, A., and K. Siemons. Die Messung der Hirn-
durchblutung mit der Stickoxydulmethode Pfliigers
Arch. ges. Physiol. 258: 149, 1953.
4. Bing, R. J., M. M. Hammond, J. C. Handelsman, S. R.
Powers, F. C. Spencer, J. E. Eckenhoff, W. T. Goodale,
J. H. Hafkenschiel, and S. S. Kety. The measurement
of coronary blood flow, oxygen consumption and efficiency
of the left ventricle in man. Am. Heart J. 38: 1, 1949.
5. Cherrik, G. R., S. W. Stein, C. M. Leevy and Ch. S.
Davidson. Indocyanine green : Observations on its physical
properties, plasma decay and hepatic extraction. J. Clin.
Invest. 39:592, i960.
6. Eckenhoff, J. E., J. H. Hafkenschiel, C. M. Landmesser,
and M. H. Harmel. Cardiac oxygen metabolism and
control of the coronary circulation. Am. J. Physiol. 149:
634. '947-
7. Eckenhoff, J. E., J. H. Hafkenschiel, M. H. Harmel,
W. T. Goodale, M. Lubin, R. J. Bing, and S. S. Kety.
Measurement of coronary blood flow by the nitrous oxide
method. Am. J. Physiol. 152: 356, 1948.
8. Frank, A., H. J. Bretschneider, E. Kanzow, and V.
Bernard. Uber die Wirkungen von Lacarnol, Oxyaethyl-
theophyllin, Dioxypropyltheophyllin und von Kombina-
tionen dieser Stoffe auf Coronardurchblutung und Herz-
stoffwechsel. Z. ges. exptl. Med. 128: 520, 1957.
9. Fronek, A., and V. Ganz. Measurement of flow in single
blood vessels including cardiac output by local thermo-
dilution. Circulation Research 8: 175, i960.
10. Gibbs, F. A., H. Maxwell, and E. L. Gibbs. Volume flow
of blood through the human brain. A.M. A. Arch. Neurol.
Psychiat. 57: 137, 1947.
11. Goodale, W. T., and D. B. Hackel. Measurement of
coronary blood flow in dogs and man from rate of myo-
cardial nitrous oxide desaturation. Circulation Research 1 :
502, 1953-
12. Goodale, W. T., M. Lubin, J. E. Eckenhoff, J. H.
Hafkenschiel, and W. G. Banfield. Coronary sinus
catheterization for studying coronary blood flow and
myocardial metabolism. Am. J. Physiol. 152: 340, 1948.
13. Grace, J. B., I. J. Fox, W. P. Crowley, and E. H. Wood.
Thoracic-aorta flow in man. J. Appl. Physiol. 1 1 : 405, 1957.
14. Gregg, D. E., F. H. Longino, P. A. Green, and L. J.
Czerwonka. A comparison of coronary flow determination
by the nitrous oxide method and by a direct method
using a rotameter. Circulation 3 : 89, 1 95 1 .
1 4a. Hamilton, W. F. Measurement of the cardiac output.
In : Handbook of Physiology. Washington, D. C. : Am.
Physiol. Soc, 1962, Sect. 2, vol. II, p. 551.
15. Hamilton, W. F., J. W. Moore, J. M. Kinsman, and
R. G. Spurling. Simultaneous determination of the pul-
monary and systemic circulation times in man and of a
figure related to cardiac output. Am. J. Physiol. 84: 338,
1928.
16. Hirche, H., andW. Lochner. Messung der Durchblutung
und der Blutfullung des coronaren Gefalibettes mit der
Teststoffmjektionsmethode am narkotisierten Hund bei
geschlossenem Thorax. Pfliigers Arch. ges. Physiol. 274:
624, 1962.
17. Huckabee, W. E. Use of 4-aminoantipyrine for determining
volume of body water available for solute dilution. J. Appl.
Physiol. 9: 157, 1956.
1294
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
10
23'
24-
25-
26.
27
29
Huckabee, \V. E., and G. Walcott. Determination of 30.
organ blood flow using 4-aminoantipyrine. J. Appl.
Physiol. 15: 1 139, i960.
Huckabee, W. E., and D. H. Barron. Factors affecting 31.
the determination of uterine blood How in vivo. Circulation
Research 9:312, 1 961.
Kety, S. S., and C. F. Schmidt. The determination of 32.
cerebral blood How in man by the use of nitrous oxide in
low concentration. Am. J. Physiol. 143: 54, 1945.
Kety, S. S., M. H. Harmel, H T. Broomell, and C. B.
Rhode. The solubility of nitrous oxide in brain and blood. 33.
J. Biol. Chan. 173: 487, 1948.
Kety. S. S. , and C. F. Schmidt. The nitrous oxide method
for the quantitative determination of cerebral blood How
in man : theory, procedure and normal values. J. Clin. 34.
In. est. 27: 476, 1948.
Kety, S. S. Quantitative determination of cerebral blood
flow in man. In: Methods in Medical Research. Chicago: Vr. 35.
Bk Pub. 1948, vol. I, p. 204.
Kety, S. S. Comment on continuous, constant-rate, 36.
sampling modification of nitrous oxide method for cerebral
blood flow in man. In : Methods in Medical Research. Chicago :
Yr. Bk. Pub. 1961, vol. 8, p. 268. 37.
Lambertsen, C. J., and S. G. Owen. Continuous, con-
stant-rate sampling modification of nitrous oxide method
for cerebral blood flow in man. In: Methods in Medical Re- 38.
search. Chicago: Yr. Bk. Pub. i960, vol. 8, p. 262.
Lassen, N. A., and O. Munck. Cerebral blood flow in man
determined by the use of radioactive krypton. Acta Physiol.
Scand. 33:30, 1955.
Lochner, \V., and H. Hirche. Ein Photometer zur
fortlaufenden Messung von Farbstoffkonzentrationskurven 39.
im stromenden Blut bei 805 m^. Klin. Wochschr. 39: 1142,
1 96 1.
Lochner, VV., and B. Ochwadt. Uber die Beziehung 40.
zwischen arteriellem Druck, Durchblutung, Durchflulizeit
und Blutfiillung an der isolierten Hundeniere. Pfliigers
Arch. ges. Physiol. 258: 275, 1954. 41.
Munck, O., and N. A. Lassen. Bilateral cerebral blood
flow and oxygen consumption in man by use of krypton -85.
Circulation Research 5: 163, 1957.
Nylin, G, and H. Blomer. Studien uber die cerebrale
Zirkulation mit radioaktiven Isotopen. Z. Kreislaufforsch.
44 ':»■ '955-
Nyi in, G., and H. Blomer. Studies on distribution of cer-
ebral blood How with thorium-B-labeled erythrocytes. Cir-
culation Research 3: 79, 1955.
Nylin, G., H. Blomer, H. Jones, S. Hedlu.nd, and C. G.
Ryi \nder. Further studies on the cerebral blood flow
estimated with thorium-B-labeled erythrocytes. Brit.
Hart J. 18: 385, 1956.
Orcutt, F. S., and R. M. Waters. Method for deter-
mination of cyclopropane, ethylene and nitrous oxide in
blood with Van Slyke-Neill manometric apparatus. J.
Biol. Chetn. 117:509, 1937.
Peterson, I. H., M. Helrich, L. Greene, C. Taylor,
and G. Coquette. Measurement of left ventricular output.
./. Appl. Physiol. 7: 258, 1954.
Piiper, J. Eine Methode zur Lokalisierung des Stromungs-
widerstandes. Pfliigers Arch. ges. Physiol. 266 : 1 99, 1 958.
Piiper, J. Uber die Lage der Capillaren im Gefalibett der
isolierten Hundelunge. Pfliigers Arch. ges. Physiol. 267 : 1 ,
'958.
Sapirstein, L. A., and E. Ogden. Theoretical limitations
of the nitrous oxide method for the determination of
regional blood flow. Circulation Research 4: 245, 1956.
Scheinberg, P., and E. A. Stead. Cerebral blood flow in
male subjects as measured by the nitrous oxide technique:
normal values for blood flow, oxygen utilization, glucose
utilization, and peripheral resistance, with observations on
the effect of tilting and anxiety. J. Clin. Invest. 28: 1163,
■949
Schimmler, W. Zur Messung der Gehirndurchblutung
mit T-1824 (Evans-blue) am Menschen. Z. Kreislaufforsch.
45:47> i956-
Shenkin, H. A.. M. H. Harmal, and S. S. Kety. Dynamic
anatomy of the cerebral circulation. A.M. A. Arch. Neurol.
Psychiat. 60: 240, 1948.
Zierler, K. L. Circulation times and the theory of indi-
cator-dilution methods for determining blood flow and
volume. In: Handbook of Physiology. Washington. D. C. :
Am. Physiol. Soc, 1962, Sect. 2, vol. II, p 585.
III. Flowmeters: their theory, construction, and operation
E WETTERER
Ultrasonic Flowmeters
Traveling Markers
Miscellaneous Methods
contents
Flowmeters Based on the Registration of Pressure Differences
The Rotameter
The Electroturbinometer
Bristle and Pendulum Flowmeters
Methods Based on the Electromagnetic-Induction Principle
the purpose of most registrations of blood flow is the
recording of the fluid volume passing the cross section
per unit of time. The flowmeter used will therefore
be calibrated in terms of rate of volume flow. In ad-
METHODS OF MEASURING BLOOD FLOW
I295
diameter must also be considered :
fig I. Parabolic velocity profile according to Poiseuille's
law in steady laminar flow. For explanation see text.
dition, the fluid velocity at particular points within
the cross section may be of interest, especially in hy-
drodynamic studies. In these cases, the flowmeter
is calibrated in terms of fluid velocity.
Since different flow types occur in the circulation,
and even in the same blood vessel, any calibration in
terms of flow rate presupposes an examination of the
dependence of flowmeter response on the velocity
distribution over the cross section.
In case of steady laminar flow, the velocity distri-
bution is in the form of a paraboloid, the profile of
which is represented in figure i . If v is the velocity
at the distance r from the axis and R is the radius of
the tube, then we have, according to Poiseuille's law:
v = K(Rz-rz)
(I)
where A' = (AP/Ax)-(i 4m); AP Ay = pressure
gradient in axial direction; /x = viscosity of the fluid.
The maximum velocitv is at the axis where r = o:
v *KR'
ax
(2)
while the lamina adhering to the wall (r = R) is at
rest. When equation I is integrated over the cross-
sectional area, the flow rate Q, is obtained :
Q = 2irfvrdr= -jKirR
(3)
The average velocity taken over the cross-sectional
area is vA:
v. ' —*— * —KR
(4)
It follows from equations i and 4 that the fluid lamina
moving at the velocity f., is at a distance of R/y/2
from the axis.
With respect to the performance of some flow-
meters, the velocity vK averaged over the radius or
;VD;
o
From equations 2, 4 and 5 we obtain the ratios:
(5)
ax
= 2:1
vK = 4:3
A
(6)
(7)
If the critical Reynolds number is exceeded, the
flow becomes turbulent; the profile of the net forward
velocities is then flattened and approaches, with
increasing turbulence, complete flatness. Other condi-
tions, which will be mentioned below, may also give
rise to a flattening of this profile. In the extreme case
of complete flatness, all fluid particles are moving
uniformly at the net forward velocity vA.
Now we may consider how different flowmeter
types will behave when the velocity profile changes
from the parabolic shape to complete flatness. A
flowmeter which responds to the axial flow only has
a relatively high sensitivity when the flow profile is
parabolic, since, according to equation 6, Vox'-Va =
2:1. When the profile is completely flat, the axial
velocity will be as high as the velocity at any other
point so that the sensitivity in terms of flow rate is now
reduced by 50 per cent from the case of a parabolic
profile. In theory, this loss in sensitivity could be
avoided by placing the flow-sensing element at a dis-
tance of R/y/o. from the axis where, in the parabolic
profile, the local velocity equals vA as discussed above.
If the response of a flowmeter is determined by the
sum of the velocities at all points covering the diameter
or the radius, this response is proportional to vR. The
sensitivity in terms of flow rate will then decrease by
25 per cent when the profile changes from the para-
bolic shape to complete flatness since vr'.va = 4:3.
It is obvious that the sensitivity of those flowmeters
which respond primarily to the velocity vA averaged
over the cross-sectional area is independent of the
velocity profile. The conditions are more complicated
if the flowmeter's response to the fluid velocity is not
linear, as is the case with most devices based on hydro-
dynamic principles.
Particular conditions are given in the inlet section
of a tube into which fluid is driven from a larger
reservoir as is the case in the trunks of the aorta and
pulmonary artery. At the entrance of such a tube the
velocity profile is flat except in a small marginal zone
where a thin boundary layer showing a steep radial
velocity gradient exists. When the site of observation
I296
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
is shifted along the tube, the boundary layer is iound
to increase in thickness and will finally occupy the
whole cross section forming the parabolic profile of
laminar flow, provided that the Reynolds number
is below the critical value. The so-called inlet length,
i.e., the distance between the beginning of the tube
and the site where a parabolic profile is just estab-
lished, can be calculated [see McDonald (93)]. If the
Reynolds number is above the critical value, the inlet
length is the distance between the beginning of the
tube and the site where turbulence is fully developed;
this inlet length, too, is calculable and will be much
shorter than for laminar flow. Under both conditions,
however, the velocity profiles in the trunks of the
aorta and pulmonary artery are almost flat so that
the use of flowmeters involves no essential difficulties
regarding the velocity profile, unless the flow type is
altered by abnormalities such as valvular stenosis.
Some flattening of the velocity profile also occurs
when the fluid is streaming from a wider into a nar-
rower tube segment through a conical intermediate
section. This effect may be utilized to improve the
performance of some flowmeters regarding the de-
pendence on the velocity profile.
The pulsatile flow in peripheral arteries is charac-
terized by phase differences between the layers oscil-
lating at various distances from the axis. Generally,
the oscillation of the layers near the axis shows a
phase lag in relation to the more marginal zones.
While at low frequencies of the flow oscillations the
phase lag increases continuously from the margin
toward the axis, the inner zones will swing closer in
phase to each other when the frequency is raised. At
high frequencies, a wide central column of fluid will
oscillate uniformly, and the profile of oscillation will
approach flatness (93). It is obvious that flowmeters
which respond to v^ or to vR are showing, in case of
pulsatile flow in peripheral arteries, errors not only in
amplitude but also in phase, as will be discussed below
with special reference to the pendulum and bristle
flowmeters. Only flowmeters responding to vA will
deliver records free from such distortions.
Another point of view is the consideration of the
frequency characteristics which a flowmeter must
possess to obtain adequate recordings of pulsatile
flow. The highest frequencies occurring in the central
flow pulse of the dog under physiological conditions
amount to 50 to 100 cycles per sec (cps). For record-
ing the main features of the central pulse a frequency
response up to about 50 cps is sufficient (34, 35).
In larger animals and in man the upper frequency
limit may be somewhat lower, but it is remarkablv
higher in small animals. In case of mechanical pickup
systems capable of vibrating, the natural frequency
should be at least double the highest frequency to be
recorded when the system is critically damped. In an
electrical system based on a carrier-frequency proce-
dure, the carrier frequency must be high enough to
reach an adequate band width. Further details will
be discussed in the description of the various flow-
meters.
The application of flowmeters to the circulation
usually involves a local alteration of the flow condi-
tions resulting from insertion of a cannula, from
placing an obstacle to flow within the streaming
fluid, from constriction of the blood vessel from out-
side, or at least from surrounding the vessel with a
rigid sleeve. A slight constriction extended over a
short length generally will not give rise to objection-
able changes of the flow conditions. The factional
drop of the mean pressure is often used as a measure
of the impediment to flow caused by the flowmeter.
It is obvious that the pressure drop should be small
as compared to the absolute pressure level. This
criterion alone, however, is not sufficient, since an
arterial segment which contains an obstacle or is
made rigid by an inserted cannula or a surrounding
sleeve can change the hemodynamic conditions by
causing pulse-wave reflections even if there is no
remarkable drop of the mean pressure. For this
reason, the length of a rigid segment should not exceed
1 cm (93).
FLOWMETERS BASED ON THE REGISTRATION
OF PRESSURE DIFFERENCES
When a liquid flows through a tube, a pressure
differerfce between two points along the tube may be
generated by friction and by mass inertia. Whereas
the influence of friction results in a pressure difference
proportional to the flow velocity, the effect of inertia
is causally connected with flow acceleration. Two
kinds of acceleration, convective and local, have to be
considered. Convective acceleration (dv/dx = change
in velocity along the axial direction) occurs in steady
as well as in pulsatile flow as a result of variation in
cross-sectional area of the tube or bv an arrangement
which causes a locally circumscribed stagnation of the
fluid or a change of the flow direction. According to
Bernoulli's theorem, pressure differences due to con-
vective acceleration are proportional to the square of
flow velocity. Local acceleration (do dt = velocity
change in time, observable at a single point, i.e.,
"locally") takes place only when the flow rate is
changing in time. Pressure differences between two
points which are caused by local acceleration are
proportional to the differential quotient of the flow
velocity, to the fluid's density, and to the distance
between both points. Thus, we have, according to
Frank (39), the following equation:
(8)
•p**F,v
* ?*v.
n
dv
* ?"»
m
where Pi and P2 = instantaneous pressures at two
different points; v = instantaneous flow velocity;
t = time; Cu C2, and C3 = coefficients; I = frictional
term; II = inertia term due to convective accelera-
tion (Bernoulli); III = inertia term due to local
acceleration. The pressure difference (Pi — Pj) is
recorded by a suitable differential manometer (see
below). In most experimental cases, the coefficients
are determined by practical calibration although they
are, under certain conditions, calculable from the
tube dimensions and from density and viscosity,
respectively, of the fluid. If, instead of the linear
velocity v, the rate of volume flow is used in equation
8, then d is equivalent to the so-called effective mass
M' (Frank):
M -k
<°L
(9)
where p = density of fluid, L and A = length and
cross-sectional area, respectively, of the fluid column
contained in the tube between both points and k =
correction factor for velocity distribution within this
column; k = 1.0 if the velocity profile is flat [see
(46, 48)]. As Ranke (107) pointed out, equation 8
must be regarded as an approximation, since the
coefficients change with the Reynolds number for the
flow and further terms may have to be taken into
account.
The properties of most differential-pressure flow-
meters which respond to pulsatile flow are dependent
upon the three terms of equation 8, although, for
certain models, one or two terms may play a dominat-
ing role. If only mean flow is recorded, term III can
be ignored; nevertheless, a great effective mass
(coefficient C3) should be avoided because it alters
hemodynamic conditions in the case of pulsatile
flow [cf McDonald (93)]. If the tube diameter is
large, as in great central vessels, term I usually has
little significance as compared to term II.
If the pressure difference is generated mainly by
friction as in figure 2, the device must be constructed
in such a way that the resulting pressure drop will
METHODS OF MEASURING BLOOD FLOW 1 297
UPO DPO
fig. 2. Friction device. UPO, DPO = lateral openings
upstream and downstream from the constriction for connection
with differential manometer. [From Green (50).]
not be so great as to disturb the physiological condi-
tions. A friction device consisting of a long, narrow
plastic tube inserted into a blood vessel was applied
by Ueno & Takenata (129) for recording the mean
flow; the pressure drop was measured by a rolling
manometer. It seems likely that it interferes with
normal blood flow.
An older method may be mentioned here. In 1935,
Green et al. (53) tried to estimate the systolic and
diastolic coronary-artery flow from the pressure
difference between the aorta and a peripheral coro-
nary branch. There is, however, no simple relation-
ship between these magnitudes, because waves travel-
ing in elastic tubes are concerned. Therefore, the
method was abandoned [cf Gregg's criticism (54) and
Chapter 7, vol. I, this Handbook].
The principle of the Venturi meters is based on the
generation of convective acceleration by a variation
in the cross-sectional area of a tube (Venturi 1797;
Herschel 1887). As shown in figure 3, the fluid has to
move from a wider into a narrower tube segment.
According to the continuity law, equal quantities of
an incompressible fluid must pass each cross section
of a rigid tube during the same time interval. The
fluid's linear velocity is therefore augmented in the
narrow segment so that here the kinetic energy is
increased and the lateral pressure is decreased. This
results in a pressure difference between UPO and
DPO in figure 3, which is proportional to the square
of the average flow velocity (term II in equation 8).
If the tube widens again downstream from DPO to
the same cross-sectional area as before, the former
pressure is restored. The additional influence of fric-
tion will augment the pressure difference between
both points (term I); this part of the pressure drop,
of course, is not reversible by rewidening of the tube.
When the rate of volume flow is changing in time, a
third kind of pressure difference corresponding to
term III appears which should be kept minimal
because it distorts the records. Devisers of such flow-
meters often failed to take this source of error into
consideration. Lauber's Venturi cannula (87), for
instance, was criticized by Frank (42) because its
[2g8 HANDBOOK OF PHYSIOLOGY <-" CIRCULATION II
UPO
fig. 3. Venturimeter of original type. [From Green (50).]
manometer connections were very distant from each
other. The aortic flow records obtained with this
cannula therefore represented acceleration curves
rather than velocity curves. A detailed polemic was
carried out on this point by Frank (41, 42) on the
one side and by Broemser (16) and Ranke (107) on
the other. To minimize the distortions, the distance
between both manometer connections (length L in
equation 9) must be as small as possible. In contrast
to a widely held opinion, the distorting effect of term
III cannot be detected by comparing the directly
measured mean flow with mean pulsatile flow deter-
mined by planimetering the recorded curves. This is
true because when areas are determined by the
planimeter, the distortions generated during flow-
acceleration may be compensated for by opposite
distortions generated during flow deceleration. An
analytical correction of the records would be feasible,
but very difficult. The best way, therefore, is to keep
Cz minimal by appropriate construction of the flow-
meter. Similar considerations apply also to Pitot
tubes (see below).
Venturi tubes like those of figure 3 are used rela-
tivelv seldom. The cannulae of de Burgh Daly (20)
and of Lauber (87) may be mentioned here. Lawson
& Holt (88) modified Daly's method.
The Venturi principle is applicable also to other
designs. Figure 4 shows the effect of an inflection of
the tube wall on the streamlines. In case of such an
inflection of small length, the point at which the
streamlines run closest to each other is not situated
at the tip of the inflection, but somewhat downstream
from it. This means that the fluid's linear velocity
is higher and the lateral pressure is lower on the
downstream side than on the upstream side of an
inflection or constriction. Thus a pressure difference
corresponding to term II is generated between two
points situated upstream and downstream from a
nearby constriction even if the tube's cross sections
are equal at both points.
Broemser (15) and Reissinger (109) in 1928, making
use of this effect, constructed an instrument which
proved appropriate for recording pulsatile flow in the
ascending aorta (fig. 5). The advantages of this can-
nula consist in the very short distance between the
fig. 4. Deviation of streamlines caused by an inflection of
the wall. [From Reissinger (109).]
fig. 5. Cannula of Broemser
and Reissinger. [From Reissinger
(109).]
lateral openings and in their symmetrical arrangement
which provides equal sensitivity to forward and back-
ward flow. The optimal inflection angle between tube
axis and wall was found to be 7 to 8°; by using this
angle, sufficient sensitivity is achieved and no eddies
occur even at the highest physiological flow velocities.
Since the planes of the lateral openings are not par-
allel with the vessel axis, an additional Pitot effect
(see below) may be involved.
Nilsson & Kramer (97) in 1954 developed a Venturi
meter according to the aforementioned principles for
the registration of the pulsatile flow in the intra-
thoracic vena cava. Steady and oscillatory flow
calibrations showed that, for this device, the terms
I and III are of subordinate significance.
The orifice flowmeter of Gregg & Green (55)
[cf Green (50) and Gregg (54)] is also based on the
Venturi principle. The pressure difference is generated
by an opening ( = orifice) in a thin disk placed across
the stream (see fig. 6). Lateral manometer connec-
tions are arranged upstream and downstream from
the disk at distances equal to the tube radius. As seen
from figure 6, the streamlines converge downstream
from the orifice so that there is a point at which the
pressure reaches a minimum as described above. Due
to its symmetrical arrangement, this device has equal
sensitivities to forward and backward flow. The size
of the orifice can be adjusted during the experiments
either by substituting disks of different orifice diam-
eters through a slot in the cannula wall or, according
to a modification devised by Shipley et al. (120), it
can be altered from the outside by means of a stud
screw, the rounded end of which protrudes into the
METHODS OF MEASURING BLOOD FLOW
^99
-P0«
OP
fig. 6. Principle of orifice flowmeter of Gregg and Green.
PO, connections to differential manometer; OP, orifice plate.
[From Green (50).]
cannula lumen. Details of construction, as well as
the connection of the orifice-meter cannula to the
differential manometer, are shown in figure 7. The
rubber membrane of the manometer, to which the
mirror is attached, bulges under the action of pressure
differences between its two sides, while it is insensitive
to the absolute pressure. The steady-flow calibration
curve is virtually quadratic if saline solution is used.
When blood is used, the effect of term I is noticeable.
The natural frequency of the differential manometer
amounts to 50 to 70 cps or more (54). It is difficult to
judge the effect of term III on the records of pulsatile
flow; the coefficients of equation 8, including C3,
vary with the cannula and orifice diameter. Arterial
flow curves recorded with the orifice meter, particu-
larly those of the femoral, axillary, and carotid
arteries (120), might suggest that the contour of the
svstolic flow peak and the registered backflow phase
could be markedly affected by term III, although
arterial flow patterns are widely variable for physio-
logical reasons, as McDonald (93) discusses in detail.
The applicability of the orifice meter to veins is
limited because of its frictional pressure drop.
Schroeder's differential - pressure flowmeter
("Druckdifferentialstromuhr") (119) may be re-
garded as a further developmental stage, obtained by
new technical means, of Broemser's differential sphyg-
mograph (see below). Schroeder designed his instru-
ment, which is shown in figure 8, for application on
unopened carotid and femoral loops of conscious dogs.
In a special compartment (C), the artery (.4) is com-
pressed by a screw device (S) so that its wall is relaxed.
Two rubber diaphragms (Di , Z)2 ), arranged at the
bottom of the compartment, are in direct contact with
the skin surrounding the vessel, one diaphragm is
placed at an upstream vessel segment, the other at a
downstream segment. Each diaphragm covers a
water-filled chamber [Chi , Ch ), and can transmit
the pressures from both vessel segments into these
chambers. Due to the compression of the vessel, its
skin and wall tissues are deformed so that the vessel
lumen is narrowed toward the middle of the compart-
ment; the slight constriction gives rise to a flow-related
pressure difference between the two vessel segments,
which is detected by a thin metal membrane (.\/i )
interposed between the chambers and is transferred
by a lever (L) to an air-pressure nozzle amplifier
(A.h ) for optical manometer registration. By two
other membranes (Mo , A/3 ) connected to similar air-
pressure amplifiers (NA2 , NAz ), the pressure of each
chamber is picked up, and the sum of both pressures
is optically recorded by an adding manometer. In
this way, the difference between the pressures (related
to the blood flow) and their sum (related to the blood
pressure) are recorded simultaneously. The steady-
flow calibration curve is almost quadratic; it is cor-
rected automatically by an optical linearizing device.
Distortions due to term III of equation 8 are reduced
by attaching an air chamber to the differential-
pressure air-transmission system. Although some me-
chanical functions of this design may require further
theoretical clarification, its records of pulsatile ar-
terial flow and pressure resemble to a surprisingly
high degree those obtained by other well-recognized
instruments.
Two other devices mav be mentioned here for re-
OP O PD
fig. 7. Construction of orifice flowmeter and differential
manometer. OP, orifice plate; 0, orifice; PU and PD, upstream
and downstream connections to manometer. .?, shell; B, base;
C, cap of manometer. L, lens, carried by a ball. M, mirror
attached to rubber diaphragm of manometer [From Green
(50).]
1300
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
flow
5 i>
mmnri wmisrffiM/WiIE/lWflM WMMh
WimUfiMM&lJBjIPWIIlL
jgiMMBlinniMtBiaik
wnnnk
fig. 8. Differential-pressure flowmeter of Schroeder (119)
for application on skin-coated arterial loops. For description
see text.
cording mean flow. A Yenturi cannula in connection
with a differential water manometer, for application
to abdominal arteries, was used by Wagoner &
Livingston (131). YVretlind (138), modifying a plan
by Stephenson (1948), carefully designed a meter for
mean flow in the ascending aorta of the cat. As figure
9 shows, the blood streaming from A to B passes a
constriction (£>) of small length which causes a pres-
sure drop of a few mm Hg. Pulsations in the differen-
tial manometer (G) records are eliminated by expand-
ing chambers (C,E) at the upstream and downstream
side of the constriction; within these chambers,
pulsating blood columns rise to different mean levels
corresponding to the pressure difference which is
necessary to drive the mean blood flow through the
constriction. The tops of the two chambers are con-
nected to each other by an air-filled tube (F) which
acts as an elastic bypass transmitting a part of flow
pulsations from the central to the peripheral end of
the meter.
The principle of Pitot flowmeters (1 728-1732)
consists in the measurement of the hydrodynamic
increment in pressure which is generated by the
locally circumscribed stagnation of a small part of
the streaming fluid. For this purpose, a thin tube, the
opening of which faces upstream, is placed in the
fluid. The difference between the pressure exerted on
that opening ("end" or "total" pressure) and the
"lateral" (or •'static") pressure is indicated by a
differential manometer. The opening which picks up
the lateral pressure may be placed in the wall (fig.
10) or near the opening facing upstream (fig. 1 1). In
other devices, two thin tubes are inserted, with one
fig. 9. Flowmeter of YVretlind
for ascending aorta of cat. For
description see text. [From
YVretlind (138).]
opening upstream and the other downstream (fig. 12);
the pressure difference is greater with this design
because suction is effected at the downstream opening
by eddy formation. This arrangement also offers the
advantage that almost equal physical conditions can
be provided to measure forward and reverse flow. If
equation 8 is applied to Pitot meters, v of term II is
not the average velocity of the fluid, but the velocity
of that small bundle of streamlines which hits the
opening facing upstream. This is an advantage be-
cause it offers the possibility of using Pitot meters like
those illustrated in figures 10 to 12 as probes which
can be shifted along the tube radius in order to
measure, point by point, the hydrodynamic pressure
distribution between the axis and the wall. Thus, the
velocity profile is determinable for hydraulic investi-
gations of theoretical and practical interest [Miiller
(95)]. If, on the other hand, the average flow velocity
is to be detected by Pitot meters, errors resulting from
changes of the velocity profile have to be taken into
account. If the flow is pulsatile, term III of equation 8
requires special consideration.
Prandtl's tube (fig. 1 1 ) is a modification of the Pitot
meter; the openings lie at the surface of a probe which
is placed in the streaming fluid. This device minimizes
eddy formation.
The construction of most Pitot meters applied in
cardiovascular physiology can be deduced from one of
the types shown in figure 10 to 12. Aortic flow was re-
corded in 1899 with Frank's (36) double-lumen cathe-
ter which was introduced through the carotid artery.
Other Pitot devices were used by Baxter & Pearce (4)
and by Jameson (67) for recording the pulmonary
artery flow, by Johnson & Wiggers (70) for recording
the coronary sinus outflow, and by Eckstein et al. (26)
for recording the vena cava flow.
"Torpedo"-shaped Pitot meters offering low re-
sistance to flow were built in 1953 by Brecher (8) and
METHODS OF MEASURING BLOOD FLOW 130I
j Fl G lg I
I FIG. II I
ML
jm—_
> . 1
- K
fig. 10. Pitot meter with asymmetrical pressure taps, one facing upstream, the other arranged to
measure lateral pressure. [From Miiller (95).]
fig. 11. Prandtl's tube, based on the Pitot principle. 1, upstream facing pressure tap. 2, lateral
pressure tap. L, distance between 1 and 2. [From Hardung (57).]
fig. 12. Pitot meter with symmetrical pressure taps facing upstream and downstream. [From
Miiller (95).]
within a wide range by changes of viscosity (water-
blood), by changes of the flow type (laminar-turbulent),
and even by changes of the velocity profile caused by
flow pulsations, provided the flow remains unidirec-
tional. Using his modification of the Pitot meter, he
constructed a catheter-tip cannula for recording the
coronary-sinus outflow as well as another cannula for
measuring the pump output in extracorporeal-circu-
lation devices. It seems possible to combine these
advantages with the arrangement described by Har-
dung (see below) which avoids distortions caused by
local acceleration.
Besides the aforementioned Pitot types, there is an
older modification used by Cybulski (19). As shown in
figure 1 4, the sharp angle in the tube causes a sudden
change in direction of flow creating a hydrodynamic
by Mixter (94). Brecher's device (fig. 13), designed for
introduction into the superior vena cava from the
jugular vein, consists of a rigid three-tube catheter at
the tip of which a streamlined lead "torpedo" is
attached. It contains the upstream and downstream
facing ends of the differential-manometer tubes U and
D while the longer tube A, serving to detect the
pressure in the right atrium, is connected to a separate
manometer. Tube A can be moved along U and D by
working an outside handle; in this way, springs are
expanded to form a basket (dashed lines in fig. 13)
around the torpedo in order to keep it centered in a
vessel of constant diameter. Mixter's torpedo is placed
in a metal tube for direct insertion into a large vein.
A promising attempt to improve the performance
of the Pitot meter was recently made by Bretschneider
(13), who realized the difficulties involved in the
estimation of the average velocity from the velocity of
a small bundle of streamlines. He placed the opening
facing upstream at a point where the local fluid veloc-
ity equals the average velocity for both laminar and
turbulent flow. Theoretically, this point is situated at a
distance of about 0.7 R from the tube axis, or 0.3 R
from the wall, i.e., R/y/2. This arrangement, how-
ever, is impracticable if the lumen diameter is 6 mm
or less. In such cases a greater relative distance lrom
the wall (0.4-0.6 R) has to be chosen. Errors due to
such malposition are substantially reduced by using a
flow cannula with a conical inlet section which flattens
the velocity profile. Thus Bretschneider obtained an
average flow calibration curve which is unaffected
DIFFERENTIAL
PRESSURE
MANOMETER
CATHETER
FLOWMETER HEAD
*;e
S3r
'— .D'
RIGHT
ATRIUM
-> SUP. CAVAL FLOW >
ATRIAL
PRESSURE
MANOMETER
fig. 13. Pitot "torpedo" of Brecher for recording superior
vena cava flow. For description see text. [From Brecher (8).]
1302
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fiu. 14. Cybulski's modification of the Pitot meter. [From
Muller (95).]
pressure elevation which acts upon the adjacent limb of
the differential water manometer.
Broemser's differential sphygmograph (14) was
built for application on unopened arteries. The instru-
ment (fig. 15) consists of a double sphygmograph
capsule, the two lower openings of which are covered
with thin rubber diaphragms. The planes of the dia-
phragms form an obtuse angle to each other. The air-
filled capsules are connected to an optical differential
manometer as well as to a simple optical manometer.
When the lower end of the instrument is pressed
against an artery so that one diaphragm is directed
upstream, the other downstream, a wedge-shaped
inflection of the vessel wall is produced, and the blood
pressure bulges the diaphragms into the capsules. Due
to the pressure difference effected by the blood flow,
the upstream diaphragm will bulge more than the
downstream one; thus a flow-related deflection of the
differential manometer takes place, while the deflec-
tion of the simple manometer is proportional to the
blood pressure. The function of this device may be
derived partly from the Venturi and partly from the
Pitot principle in that the wall inflection is typical of
the Venturi meters while the inclination of the dia-
phragms to the vessel axis results in a Pitot effect. The
calibration curve determined by perfusion of excised
arteries or of elastic tubes is almost quadratic. Al-
though the records obtained with this device from the
ascending and abdominal aorta of rabbit, cat, and
dog show the typical contours known from other
flowmeter registrations, the instrument did not find
frequent application. This may be due to the fact that
its exact positioning and its calibration in situ are
difficult.
Besides historical notes, Muller (95) published a
theoretically and experimentally based criticism of
Pitot meters. He stated that the arguments raised by
previous investigators against these meters are, on the
whole, not justified. Pieper & Yogel (101) calculated,
for the device shown in figure 1 1 , the distortions due
to term III of equation 8 at various distances L. Al-
though, on the one hand, L should be kept as small as
possible so as to minimize C'3 , the flow sensitivity of
the device (term II) is, on the other hand, also dimin-
ished when opening 2 is placed too near to 1 . The
optimal L must therefore be found by compromise.
Hardung (57) came to the conclusion that, for instru-
ments such as shown in figure 10, term III can theo-
retically be eliminated by placing the upstream-facing
opening at a certain optimal distance from the long
axis of the lateral tube.
A new and interesting catheter-tip method for re-
cording the blood velocity in great central arteries was
developed by Fry et al. (46-48). The tip of the double-
lumen catheter used has two openings, both facing in
lateral direction and placed several centimeters apart.
Here the difference of the pressures acting on the
openings ("axial pressure gradient") is due neither to
a Venturi nor to a Pitot effect so that only the terms
I and III of equation 8 are involved. While, in the
aforementioned instruments, term III is a very un-
desired source of distortions, this very term plays the
main role in Fry's method. For this reason, the co-
efficient Ci is purposely made very large by choosing a
great distance between the openings. It is obvious
that the time course of the pressure difference itself,
which is picked up by an electrical differential manom-
fic. 15. Differential sphygmograph
of Broemser for application to unopened
arteries. Thin lines at the lower ends =
rubber diaphragms. Upper ends con-
nected to differential and adding
manometers. [From Broemser (14)]
METHODS OF MEASURING BLOOD FLOW
!303
eter, in no way represents the time course of the
flow velocity v. The magnitude v, however, is con-
tained in the linear differential equation which re-
mains when term II is removed from equation 8. In
order to solve this equation for v continuously, the
electrical signal delivered by the differential manom-
eter is fed into an analogue computer which has
been adjusted according to the magnitudes C\ and C3 .
The output signal of the computer will then follow the
actual flow course, provided the coefficients C\ ("veloc-
ity resistance") and C3 ("velocity inductance") are
known with sufficient accuracy and the pressure
difference is not affected by other physical influences.
Difficulties arising from these conditions are discussed
by Fry (46) and by McDonald (93). Flow records
from the ascending aorta of dog and man demonstrate
this method to be promising, while records of the
pulmonary artery flow seem to require further
clarification. For simplified catheter-tip approaches,
consult the papers of Evans (29), Jones et al. (71), and
their discussion by McDonald (93).
Many types of differential manometers have been
used to record the pressure differences delivered by
the instruments described above. It is obvious that
low frequency manometers, particularly water manom-
eters, are far from able to follow the rapid fluctua-
tions in differential pressure which occur when arterial
or central venous flow is recorded. Therefore mem-
brane manometers with adequate frequency response
are required for recording pulsatile flow. Their
sensitivity must be considerably higher than that of
common blood pressure recorders because the flow-
related pressure differences are relatively small.
The difficulty of combining high sensitivity and
high natural frequency in the same instrument is
manifest in the discussions of physical principles by
Frank and the models designed by Frank (37, 39),
Gregg (54) and Green (50, 51 ).
The difficult task of recording the very small
flow-conditioned pressure differences is made possible
by the amplification of electrical signals from rela-
tively stiff manometers of high natural frequency.
Most important of these are capacitance or inductance
manometers as well as resistance manometers of the
strain gauge type (6, 50, 51).
It has been emphasized that, due to term II of
equation 8, the calibration curve of most differential-
pressure flowmeters is not linear. As in the case of
bristle flowmeters, an attempt has been made to avoid
the cumbersome graphical correction of records by
using linearizing or, especially, square-root extracting
devices, some of which a) are employed after registra-
tion, for evaluation of the records, while others b) are
designed to deliver already linearized registrations.
For a, Frank (39) proposed a photographic projec-
tion method, Broemser (15) and Ranke (108) com-
bined the linearizing element with a planimeter. For
b, Schroeder (119), as described above, uses optical
linearization which works during the registration;
Baxter & Pearce (4) connected a linearizing circuit to
the electrical differential manometer. See also Green
(50).
Finally, the so-called constant-pressure flowmeters
or air-expansion systems may be mentioned although
they are not differential-pressure meters in the proper
sense [see Gregg (54), Green (50)]. If a blood reservoir
is connected to a large air chamber, any inflow or
outflow of blood will change the air pressure. The
rate of volume flow of the blood entering or leaving
the reservoir can therefore be determined from the
slope of the change in air pressure which is recorded by
a sensitive manometer. If the pressure variations are
very small as compared to the absolute pressure level,
the system delivers a virtually constant pressure for
perfusing a vascular bed and acts, at the same time, as
a flowmeter. Such systems have been preferentially
employed for studies on the coronary circulation
[Wiggers & Cotton (137); Green & Gregg (52); Eck-
stein et al. (25)].
THE ROTAMETER
The rotameter used in physiological experiments is
a device for measuring mean blood flow in cannulated
vessels. The prototype instrument was designed for
measuring gas flow. In a vertical conic tube a float
moves up and down in proportion to the rate of flow,
stabilizing its position by fast rotation which results
from spiral grooves around the float body. This rota-
tion has given the instrument its name. If fluid instead
of gas is used, the float does not rotate: stabilization is
achieved by other means. Devices in use are based on
designs by Gregg et al. (56). The instrument of
Shipley & Wilson (121, 122) is shown in figure 16.
The conic tube is made of Lucite or plexiglass with
various flow capacities from 0.1 to 3.0 liters per min.
The movement of the brass float is detected electro-
magnetically. An iron rod pierces the middle of the
float vertically and is fixed in such a position that it
protrudes equally above and below the body of the
float. The lower part of the rod is guided by a ring,
whereas the upper part enters into the lumen of an
electromagnetic coil fed by alternating current. As the
1304
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
I —
fig 16. Rotameter of Shipley and Wilson. .4, contact holder;
B, one of the contacts to induction and compensating coils,
C, detecting assembly; D, rubber cap for removing air bubbles;
E, coils; F, protecting sleeve; G, outflow spout; H, soft -iron
float wire; /, brass float; J, metering chamber; A', conical
metering portion; L, support; M, float rest at zero flow; N,
float guide; 0, inflow spout. [From Shipley & Wilson (121).]
rod moves into it, the inductance of the coil increases,
and this is recorded continuously by means of a
bridge circuit, rectifier, and galvanometer. Since the
lift of the float is proportional to the flow of blood,
the record can be calibrated in terms of flow rate.
The theoretical basis of the rotameter may chiefly be
derived from mass inertia of the streaming fluid ac-
cording to the Bernoulli effect. The fluid streaming
upward is accelerated in the ring slot around the
float. The fluid reaches its maximal velocity not in
the plane of the slot, but at a somewhat higher level.
This velocity difference causes a pressure difference to
develop between the levels below and above the float.
Other effects, such as eddy formation, may play an
additional role. The pressure difference is augmented
by friction due to viscosity of the fluid. By the action
of the total pressure difference on the cross-sectional
area of the float, a force is brought about which lifts
the float, a force is brought about which lifts the float.
In steady states the lifting force must be in balance
with the float's weight diminished by its buoyancy.
Since the weight minus buoyancy remains constant,
the force must be constant also. This is achieved by
the fact that the ring slot area increases with flow rate
by elevation of the float to a higher level where the
tube diameter is larger.
Since the rotameter should be independent of
viscosity (due to changes in temperature or hemato-
crit), the frictional force must be kept minimal as
compared to the inertia force. This can be achieved
either by increasing the inertia force by special
shaping of the float (121) or by diminishing the fric-
tional force by using large ring slot areas and floats of
light weight (60).
The relationship of volume flow to galvanometer
deflection can be made linear if the electrical settings
are adjusted. Since the response to changes of flow is
slow, only mean flows are recorded. However, large
pulsations, such as occur in arteries, are not averaged
correctly, particularly by units which use heavy brass
floats. This results in the recording of lower mean
values than are actually present (60), and it may be
useful in such cases to diminish the amplitudes of the
flow pulsations by an air chamber arranged upstream
from the rotameter (122).
THE ELECTROTURBINOMETER
The Potter electroturbinometer, originally built for
technical purposes, has been applied by Sarnoff et al.
116, 117) for the registration of aortic blood flow in
dogs. It consists of a stainless-steel turbine which is
driven by the blood stream. The turbine is suspended
within a Lucite tube by spring clips. The necessity of
using thrust bearings is avoided by shaping the rotor
in such a way that the stream generates, in addition to
rotation, a hydrodynamic force which acts in an up-
stream direction. The rotor contains a permanent
magnet which induces, by its rotation, an alternating
voltage in a pickup coil outside the tube. The fre-
quency of that voltage is proportional to the rotational
speed; by means of a counting and integrating
electronic system, an output signal is obtained,
the strength of which is a measure of the number of
turbine revolutions per time unit. Two models of
different sizes are described, the smaller of which
responds to flow from about 0.5 to more than 4.5
liters per min. The calibration curve shows a slight
bend and is, in the case of blood, independent of
temperature from 22 C to 40 C and of the hematocrit
down to 22 per cent. Although the instrument is un-
able to follow the instantaneous changes of the aortic
METHODS OF MEASURING BLOOD FLOW
I305
0
flow-
©
0
©
diaphragm
0\y
©
^■diaphragm
©
fig. 1 7. Schematic diagram of different
arrangements of bristle and pendulum flow-
meters, a: Bristle in the proper sense, b: Stiff
needle with flexible origin (short flat spring or
similar), c: Stiff needle the fulcrum of which is
formed by a diaphragm, d: Similar to a or bt
with a body at the tip. e: Similar to c, with a
body at the tip. /.• Coaxial compact cylinder,
held by one or two springs. The types a, b, c
are now commonly called "bristles," the
types d, e, f: "pendulums." [Redrawn and
modified from Taylor (127).]
flow pulses, it is said to indicate mean flow faithfully,
irrespective of whether the flow is steady or pulsatile,
provided no phases of essential backflow occur. Its
resistance to blood flow is relatively high. The small
model causes pressure drops of about 5, 1 3, 30, and 50
mm Hg at flow rates of 1, 2, 3, and 4 liters per min,
respectively. The pressure drops caused by the large
model are lower. Heparinization of the dogs is, of
course, necessary. Some alteration of the pulse wave
and hemolysis might be caused by the instrument.
BRISTLE AND PENDULUM FLOWMETERS
When a body is immersed in a streaming fluid, it
represents an obstacle to the flow. The force exerted
on the body by the streaming fluid is due partly to
friction and partly to mass inertia of the fluid. Ac-
cording to Frank (40), this force F is given approxi-
mately by the formula:
F'V + Cf* ('0)
where v = velocity of the fluid acting on the body; C\
and C2 = coefficients which depend on the viscosity
and density, respectively, of the fluid, on the size and
shape of the body, and on the local distribution of
velocities; C\ v = frictional term; C« v2 = inertia term.
Sometimes, another approximation is used:
F -" Cv*
(II)
where, in case of blood, the exponent k is found to be
between 1.2 and 2.0. As will be discussed below, theo-
retical estimation of the coefficients and of the expo-
nent is impossible except under very simple conditions,
so that an empirical determination is usually neces-
sary.
If the body is held in its position by an elastic device,
it will undergo some displacement due to the force F,
and thus the degree of displacement can be taken as a
measure of that force. Since the force is related to the
fluid velocity, according to equation 10 or 11, the
registration of the displacements by mechanical,
optical, electrical, or other means represents a con-
tinuous recording of the flow. For the construction
of such devices, the following requirements should be
taken into account: a) The resistance to flow produced
by the obstacle must be so small that the flow is not
significantly influenced, b) Where pulsatile flow is to
be recorded, the natural frequency of the elastically
suspended body must be much greater than the
highest significant frequency. If this condition is ful-
filled, the displacements of the body will be very small.
c) As far as possible a fixed relationship should be ob-
tained between force and displacement on the one
hand and average flow velocity on the other, inde-
pendently of the velocity profile. This will be discussed
below.
The body itself can be in the form of a rod or
needle set perpendicularly to the direction of flow.
This needle is usually attached at its origin to the end
of a side tube while its tip remains free and reaches the
axis of the main tube as shown in figure 1 73o8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
flowmeter, the deflections of which are detected either
by two coils acting as two arms of an inductance
bridge or by four coils forming a differential trans-
former. The coils are wound around the flow cannula
upstream and downstream from the pendulum. This
pendulum consists of a flexible ferromagnetic paddle
or of a spring-suspended ferromagnetic disk fixed on
the inner side of the tube wall. The authors empha-
sized improvement in stability achieved by using the
differential transformer. They implanted such devices
into the abdominal aorta of dogs under anesthesia
and obtained flow records some days later, the animals
being conscious.
The most recent model of a pendulum flowmeter
with electromagnetic transmission was built by Pieper
(ioo). The flow-sensing element containing a differ-
ential transformer is arranged at the tip of a catheter
and can be introduced from a peripheral vessel into
an unopened central vessel, e.g., from the carotid
into the ascending aorta. The transformer coils are
wound around a longitudinal iron core. A ferromagne-
tic cylinder, surrounding the coils at a small distance
and covering about three fourths of their length, is
suspended by elastic springs so that it can be shifted
to and fro in its longitudinal direction. On its circum-
ference, the cylinder carries a small ring-shaped disk
which faces the blood stream. The force exerted on
the disk by the flow will shift the cylinder and thus
change the mutual inductances of the two transformer
parts in opposite directions. The natural frequency of
the elastically suspended cylinder is 120 cps per sec.
The frequencv response was found to be flat up to 20
cps. The probe is held centered in the vessel axis by an
umbrella-like arrangement. Rods surrounding the
probe are folded when the catheter is introduced, and
are spread by means of a mechanism actuated from
outside when the tip has reached its final position.
The RCA 5734 transducer tube represents a new
and very useful means of electrical transmission in
pendulum and bristle flowmeters. This tube, which
was originally built for physical purposes, was em-
ployed in physiology for the measurement of small
forces, such as in the manometers and sphygmographs.
The essential characteristic of the 5734 triode is the
movable element consisting of the internal tube plate
(anode) and the external plate shaft (fig. 20). This
movable element extends through a thin and flexible
metal diaphragm, the center of which acts as a pivot
permitting small angular deflections of the plate
shaft so that the distance between the plate and the
fixed grid will be changed. This results in changes of
the plate current. Under the triode-operating condi-
fig. 20. Mechano-electric transducer tube no. 5734 of the
Radio Corporation of America. Schematic sectional view.
[Redrawn from Muller (96).] K, cathode; Hi, H-_, heater
(filament) connections; 6', grid; .4, internal plate (anode); S,
external plate shaft; M, flexible metal diaphragm. .4, M, and S
are electrically connected to the tube's metal shell. Terminal
leads in clockwise order. Bottom view: heater, grid, heater,
cathode. Tube dimensions as indicated by RCA : maximal total
length (excluding leads), 1.3". Maximal diameter, 0.328".
Tube weight, 1.75 g. Rotational compliance of the diaphragm,
0.075 degree /g crn- Resonance frequency of plate shaft, 12,000
cps. The connection of the plate to the electric circuit is pro-
vided by the supporting clamp attached to the metal shell of
the tube. If there exists contact between tube shell and blood or
tissues, the electric circuit has to be designed in such a way
that the tube plate is grounded. The plane of deflection of the
plate shaft must coincide with the plane through terminal
lead of grid and tube axis.
tions indicated by RCA (plate-supply voltage, 300
volts; grid voltage, o volts; load resistance, 75,000
ohms), the deflection sensitivity, i.e., the ratio of
change in output voltage to angular deflection of the
plate shaft, amounts to 40 volts per degree. Deflections
of more than ± 0.5 degree from the normal position
of the shaft may damage the diaphragm and the tube
electrodes. By virtue of its small weight and dimensions,
its high sensitivity and the low inertia of its moving
part, the 5734 is very appropriate to use in con-
structing a pendulum or bristle flowmeter. A further
advantage is its commercial availability.
In 1953, the 5734 was first used for blood flow
measurement independently by Brecher and his co-
workers, and by Scher el a!., and in 1954 by Muller.
Fundamentally, all these designs were based on the
principles shown in figure 1 yc and e. The transducer
tube is placed in the side branch of a T-cannula, and
a needle or pendulum which protrudes into the
streaming blood is attached to the plate shaft.
In the model of Scher et al. (1 18), the T-cannula is
made of stainless steel. Two types of obstacle to flow-
are used, the first being a flat paddle placed across
the stream, the second a streamlined rod or tube of
plastic. As can be expected, the sensitivity of the
paddle type is very high, and the deflection is ap-
proximately proportional to the square of flow-
velocity so that in such cases the second term of
METHODS OF MEASURING BLOOD FLOW
'3°9
FLOW
Y-<
w <
fig. 21. Diagram of the standard transducer-tube bristle
flowmeter of Brecher. a: Flowmeter cannula similar to that of
Brecher & Praglin (12), with improved socket, b: The same
model with "zeroing" cylinder. Total length, 50 mm. For
further description see text. [From Brecher (9, 10).]
equation 10 is preponderant. The streamlined ob-
stacle, however, offers much lower resistance to the
flow; its sensitivity is therefore smaller, and its calibra-
tion curve approaches linearity in that the deflection
is proportional to vl-\ thus indicating that it is chiefly
due to viscous drag. It may be noteworthy that the
authors found the output voltage of the transducer
tube to respond to cooling by the streaming blood so
that the base line of the records was not sufficiently
stable.
The most extended use of the 5734 tube-bristle
flowmeter in arteries and veins was made by Brecher
et al. Their "standard model" was published by
Brecher & Praglin (12) and is shown in figure 21 with
two additional improvements described by Brecher
(10). The system consists of four parts A', Y, W, '/.,
which are screwed together. The transducer tube is
held by the parts X and W which form a metal screw
socket pressing the two lead washers R\ and R<
against the lower rim and shoulder of the tube's
metal shell. In this way, the tube is fixed, and perm-
anent electrical contact is established between tube
plate and grounded cannula. Also seepage of fluid
from part Z to the lead wires M is prevented. Part
Z, as "head" of the assembly, is ligated into the blood
vessel and connected with A' by part Y. Heads of
various dimensions are available for use in blood
vessels of different diameters. The small bristle B is
20 to 35 mm long; it can be made of glass, nylon, or
metal. The natural frequency of the transducer
system with the bristle amounts to about 200 cps.
The side tube A' is used for the removal of air bubbles
and for attaching a manometer for pressure record-
ing. Temperature changes of the transducer tube
caused by the streaming blood are prevented by the
long stationary fluid column in the parts Y and Z.
When the meter is inserted into the blood vessel, zero
flow can be determined at any time by means of the
"zeroing" cylinder G (fig. 216) without stopping the
blood flow. The cylinder position is controlled by the
handle (?; it can be moved forward beyond the
bristle tip in order to protect the bristle from de-
flection by the flow.
For blood-flow recordings in large arteries,
especially in the trunk of the pulmonary artery,
Brecher & Hubay (11) developed a modification
of the standard bristle flowmeter which can be in-
serted without clamping the vessel (fig. 22). The
device consists of three main parts A, B, C. The
transducer tube is fixed in C by cement. The most
characteristic part of the device is the lip D which is
introduced into the vessel through a "buttonhole"
opening with little loss of blood. The vessel wall is
then firmly held between D and plate E, the latter
being pressed by screw nut F. In order to protect
the bristle against damage during the insertion of lip
D into the vessel, the tip of the bristle U is withdrawn
behind lip D by screwing part C backward. The
fig. 22. Brecher's bristle flowmeter modified by Brecher &
Hubay (11) for use in large arteries, especially in pulmonary
artery trunk. Total length, 75 mm. a: Diagram of longitudinal
section, b: Application to pulmonary artery. For description
see text. [From Brecher (10).]
I3IO
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 23. Schematic diagram of
the transducer-tube flowmeter of
Miiller. T, 5734; E, micrometer
screw. A', resistance body. [From
Mailer (96).]
artery's diameter is kept constant by the metal band
N placed around the vessel wall. As in the standard
model, the "zeroing" cylinder G is used for the
determination of the base line of flow during the
records, and the side tube A' for the removal of air
bubbles and for manometer connection.
As to the additional electrical equipment, Brecher
uses a load resistor of 500,000 ohms for the transducer
tube which gives higher sensitivity and better d-c
stability than the 75,000-ohm resistor recommended
by RCA. The load resistor is connected to the cathode
while the plate is grounded in all cases. The changes
in plate current due to the deflections of the plate
shaft cause proportional variations of the cathode
potential which are amplified either by battery-
operated or main-fed d-c amplifiers. Both types were
designed by Praglin and are described by Brecher
(g). It should be noted that the greatest plate-shaft
deflection ever observed in Brecher's blood-flow
experiments is 3 min of arc. This deflection results in
a potential change of 4 volts at the tube's cathode,
corresponding to a sensitivity of 80 volts per degree.
Thus the deflections caused by the blood flow remain
within the mechanically safe range, which extends to
30 min of arc on either side.
The transducer-tube flowmeter developed by
Miiller (96) is shown in figure 23. It was built for
flow measurement in blood vessels as well as for
investigations of more general hydrodynamic interest,
particularly for the study of the forces exerted by
streaming fluids of various Reynolds numbers on
resistance bodies of different shapes. An accurate
calibration in terms of force is therefore needed and
can be performed by the micrometer screw E which
causes a small spring to press on the tube's plate shaft.
The transducer tube is arranged in a bridge circuit,
the adjacent limb of which contains a second triode
of similar properties which compensates for fluctua-
tions of operating voltage, etc. The bridge output is
connected to a push-pull d-c amplifier. A special
model for coronary-artery flow recording was de-
signed by Laszt & Miiller (86) (fig. 24). The hori-
IJ_
l-P
M
fig. ->4- Transducer-tube flowmeter of Laszt and Miiller.
^r' 5734; A, plate shaft; M, manometer; A\ vertical limb of
T-cannula; S, extension rod of bristle. Si, movable cylinder
around R, Wk, cylindrical resistance body at the bristle tip; A',
horizontal limb of T-cannula. [From Laszt & Miiller (86).]
zontal cannula A' is inserted into the vessel without
interruption of the flow and without using ligatures
around the vessel. The vessel wall is pierced by the
sharp edge, and the cannula is tilted and moved
slightly until A slips into the vessel. The cylinder St
is then pressed downward to immobilize the vessel
wall around the incision.
Critical remarks on the bristle flowmeter technique
are based on experimental data and on theoretical
considerations. The main practical advantages
enumerated by Brecher (9) are as follows: negligible
resistance to and interference with the flow; equal and
opposite response to forward and backward flow;
high sensitivity and frequency response. The main
practical disadvantages are: necessity for opening the
vessel and using anticoagulants; gravitational effects
on the bristle when the position of the cannula is
altered; and the nonlinearity of the calibration
curve. This latter drawback can be overcome by
electrical linearization (9, 103). Since, for this purpose,
the amplification of low flow signals is made much
greater than that of high ones, the device becomes
very sensitive to minute shifts of the base line so that
an exceedingly stable base line is required. The
METHODS OF MEASURING BLOOD FLOW
1311
linearizing circuit has to be adjustable to correct the
calibration curves in the range from about v1'2 to v2-0.
In case of large flow pulsations, the recording of the
mean flow by integrating circuits must be preceded
by linearization. Obviously, some of these properties
concern other flowmeters as well (particularly differ-
ential-pressure flowmeters). As to flowmeters which
are equipped with the transducer tube, the temper-
ature of material surrounding the tube should be kept
constant, and the heater current should be stabilized.
Furthermore, some 5734 tubes show "pressure
artifacts," i.e., changes in plate current when the
pressure exerted on the tube's diaphragm is altered.
According to Brecher, very few new factory-delivered
tubes respond to pressure; however, careless handling
of a tube, especially anything causing deflections of
the plate shaft beyond 30 min of arc, can effect
permanent distortion of the diaphragm which will
give rise to such artifacts.
Muller (96) showed theoretically that at present,
exact mathematical calculation of the forces exerted
on a body immersed in the streaming fluid is im-
possible even in the case of steady flow. Only in the
range of very small Reynolds numbers will forces
be fully calculable as the sum of a term proportional
to v and another term proportional to ;»'2 (cf equation
10). Also, in the range of high Reynolds numbers,
friction cannot be ignored. For the force exerted on
the body is due to a thin boundary layer of fluid
around its surface (Prandtl's theory). Within this
layer, the velocity gradient perpendicular to the body
surface is very high. The lower the fluid's velocity,
the thinner will be the boundary layer and the greater
the velocity gradient. Muller's experimental data
show that the boundary layer around a streamlined
bristle is stable up to Reynolds numbers of about 900;
above this, disturbances of the layer and hence
irregularities of the force exerted on the bristle are
observed, even if the flow is laminar. As to pulsatile
flow in blood vessels, the conditions are still more
complicated as the blood is nonhomogeneous and
a very large range of Reynolds numbers (from 0 up to
several thousand) may occur within one pulse cycle.
According to Muller's experimental findings as well
as to Womersley's theory, the fluid laminae moving
at various distances from the vessel axis are oscillating
out of phase with each other. Thus, one must consent
to Muller's conclusion that, from a theoretical point
of view, this type of flowmeter type is far from having
a clear theoretical and mathematical basis.
Taylor (127) simplified some of the physical
presumptions and presented a valuable theoretical
study of the recording properties of bristle and
pendulum flowmeters. Considering the velocity
profile at various frequencies (fundamental and higher
Fourier harmonics) of oscillatory flow according to
Womersley's theory, he found that simple bristles
(see fig. 1 7a, b and c) give relatively true records of
the oscillatory flow. Compared with their response to
steady laminar flow, these instruments progressively
underestimate the average velocity as the flow
oscillations increase in frequency. The error in
amplitude approaches 25 per cent at higher fre-
quencies, and the maximum phase lag is about 70.
Taylor also compared an actual femoral-artery flow
curve [recorded by McDonald (g2) with gas-bubble
high-speed cinematography] to the record which a
bristle would give according to his calculation. He
showed that the errors which are mainly due to the
higher harmonic components have no great distorting
effect because of their small amplitude. In Taylor's
words, "the final "recording' is a quite acceptable
reproduction of the flow." Attaching a paddle to the
bristle (fig. 1 yd and e) gives rise to greater errors in
amplitude and phase while the recording by a
coaxial cylinder (fig. 1 7/) shows small errors in
amplitude, but an enormous phase lead at higher
frequencies. In case of an almost flat velocity profile,
as in the great arteries near the heart, bristles and
even paddle-mounted pendulums will give still more
satisfactory results.
METHODS BASED ON THE
ELECTROMAGNETIC-INDUCTION PRINCIPLE
This type of flow measurement is notable in several
respects. It furnishes direct transformation of the
mechanical magnitude into an electrical signal. Its
interference with the blood flow is so small that it
can be completely neglected. It delivers strictly linear
calibration curves and equal sensitivities with opposite
signal directions to forward and backward flow so
that the assessment of mean flow by integrating
circuits can be easily achieved. Its calibration in
terms of average velocity or flow rate is independent
of the velocity profile and of the density, viscosity,
and temperature of the fluid. Its range of frequency
response is theoretically unlimited and depends in
practice on the electrical equipment used. It is
applicable to all fluids having electrical conductivity
equal to or higher than that of tap water, e.g., saline
solutions, blood, mercury.
In addition to these physically inherent character-
I 3 1 2
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
istics which render the method almost ideal, there are
other favorable properties of great practical value.
Most important, the method is applicable to unopened
blood vessels and therefore requires neither damaging
the vessel wall nor using anticoagulants. For this
reason, the method can be applied to the anesthetized
animal and man under surgery and by implanting
electromagnetic probes, measurements can be made
on conscious and freely moving animals.
The electrical flow signal is produced by the motion
of the fluid across the lines of a magnetic field. For
explanation, a simple physical experiment is shown
in figure 25. A metal strip is moving at a velocity ;'
in the direction indicated by the arrows. This direc-
tion is at right angles to the lines of magnetic force
present between the magnet poles N and S so that a
voltage (potential difference) is generated in the metal
strip according to Faraday's induction law. The
induced voltage, which is directed perpendicularly to
the lines of magnetic force and to v, is picked up by
sliding contacts ("electrodes") et and c2, and measured
by the voltmeter I'. Assuming that the magnetic field
permeating the metal strip between the electrodes is
homogeneous and that the lines of force, the velocity
v, and the line extended between the electrodes are
directed mutually at right angles to each other, the
induced voltage Eind is:
~ind
BDvIO volts
(12)
where B = density of magnetic flux (gauss); D =
width of the metal strip, which is also the distance
between the electrodes (cm); v = instantaneous
velocity of the metal strip (cm/ sec). Reversal either
of the direction of the magnetic field or of the motion
will reverse the polarity of the induced voltage. Small
deviations from the assumed right-angle arrangement
between the directions of B and v, say by ±10 per
cent, have little effect on £,„,,.
Now suppose that the moving metal strip in figure
25 is replaced by a conductive fluid streaming through
a tube (fig. 26). The tube wall may consist of insu-
lating material, and the electrodes et and e2 which are
inserted into the wall may be in contact with the
fluid. In this case, also, equation 12 is generally valid
if D is the diameter of the fluid column and ;■ the
instantaneous fluid velocity. The velocity, however,
will usually not be uniform within the space between
the electrodes (as is the case for the solid strip of fig.
25). The induced voltage must therefore be calcu-
lated from the total sum of all differentials v.dr
fig. 25. So-called unipolar induction in a metal strip
moving across the lines of magnetic force. For explanation see
text.
fig. 26. Basic arrangement for electromagnetic flow meas-
urement. Metal strip of fig. 25 is replaced by conductive fluid
streaming through a tube. For explanation see text.
existing along D, i.e., from
fvdr - D7
(12a)
where vR = velocity averaged over the diameter D
or radius R. Thus, this velocity has to be used in
equation 1 2 instead of v, and the induced voltage is
obviously dependent on the velocity profile. This
would be an essential drawback of the method, if
there were not an additional compensating effect
which is of the greatest importance. Let us assume
that, as in case of steady laminar flow, the fluid near
the axis runs much faster than that near the wall. The
outer fluid layers, in which smaller voltages per unit
length are induced than in the inner layers, act as a
sort of load resistor to the latter, and circular electric
currents take place within the fluid thus bringing
about a change in the originally induced voltage
distribution. Making a valuable theoretical and
experimental contribution to the achievements of
earlier workers (see historical notes), Thurlemann
METHODS OF MEASURING BLOOD FLOW
I3!3
(128) found that, due to this effect, the resulting
voltage picked up by the electrodes is, in case of a
parabolic velocity profile, just as high as £,„,, would
be if all the fluid layers were to move uniformly at the
velocity vA averaged over the cross-sectional area.
This means that the method delivers a flow-signal
voltage which is linearly proportional to the in-
stantaneous velocity vA or to the instantaneous flow
rate. The only conditions required are that the
magnetic field be homogeneous, that the fluid be
homogeneous with respect to its electrical con-
ductivity, and that the velocity distribution be sym-
metrical in relation to the tube axis (79). It follows
that the flow-signal voltage Ef which is picked up by
the electrodes differs from the originally induced
voltage Eind (except the extreme case of a completely
flat velocity profile where both are equal) so that
for fluid flow, equation 12, has to be modified to:
E. = BDv.-IO'8 volts,
f
(13)
As Kolin (78-80) stated, equation 13 is valid also
in case of any other velocity profile and even in the
idealized case of a central and coaxial fluid jet which
is moving through a tube while the annular fluid
cylinder around this jet is quiescent. From this Kolin
concluded that the electrically conducting vessel
wall may be regarded as representing such a quiescent
fluid cylinder surrounding the streaming blood and
that, therefore, no error is caused if the voltage Es
is picked up by electrodes placed at the outer surface
of the vessel wall. Thus the application of the method
on unopened blood vessels, which had been carried
out earlier as an experimentallv proved procedure by
Kolin and other workers, was also justified on a
theoretical basis. Any inaccuracy due to the differ-
ence in specific conductivities of blood and wall
tissue is of minor significance (80, 82).
If, instead of v,,, the instantaneous flow rate Q,
(cm3/sec) is used in equation 1 3, we get with D =
2RandvA = Q,/(/x2tt):
Ef
2BQ
-8
■ 10 volts
(14)
It is obvious that, according to the aforementioned
considerations, R is the vessel radius including the
wall thickness. Equation 14 shows that the sensitivity
£//Q, is inversely proportional to R or to the distance
between the electrode tips and is independent of the
wall thickness (80, 82), provided that B is fixed and
that the vessel is surrounded by insulating material.
The history of electromagnetic flow measurement
[see notes in (78, 84, 123)] shows that several authors
found the principle independently of each other.
Faraday demonstrated electromagnetic induction in
solid as well as in liquid conductors. But he did not
conceive the idea of measuring fluid flow which in-
volves recognition of velocity distribution. His ex-
periment at the Waterloo Bridge, in which he tried to
detect an induced electromotive force (emf) in the
River Thames due to the water's motion through the
earth's magnetic field, simply represented his search
for an induction phenomenon on a terrestrial scale.
This experiment was unsuccessful, probably due to
electrode-polarization difficulties. Young et al., in
1920, were able to record such an emf. Williams, in
1 930, performed the first electromagnetic measure-
ments of the velocity distribution in copper sulfate
solutions, but made no measurement of flow rate. In
1932, Fabre (30) suggested, in a short note, electro-
magnetic recording of variations in blood flow in can-
nulated vessels [see also (84)]. Kolin [1936 (5)] is to be
regarded as the real founder of electromagnetic blood-
flow measurement. He was the first to recognize the
applicability of the method to unopened vessels and to
obtain successful records from dogs. In the following
years and decades, he also made the major contribu-
tions to further development of the procedure, espe-
cially by introducing and refining the a-c modification
instead of the d-c type which was employed earlier.
The d-c method was also described by Wetterer [1937
(133)] and was particularly used for recording flow in
the unopened ascending aorta. Valuable contributions
were further made by Einhorn (28) concerning the a-c
method and by Thurlemann (128) whose findings
have already been mentioned. In 1953, the square-
wave modification was initiated by Denison (see 125)
and, since then, has been undergoing considerable
development by the work of Denison and Spencer.
It combines, at least theoretically, the advantages of
the d-c type with those of the a-c sine-wave type.
The d-c procedure (68, 69, 72, 75, 84, 133) is the
simplest approach to the electromagnetic flowmeter
technique (see fig. 26). A constant magnetic field is
used from either an electromagnet or a permanent
magnet. The pole pieces of the magnet should be con-
structed in such a way that the gap can be adapted to
the vessel size and the pole faces are large enough to
insure a uniform magnetic field across the entire
vessel segment. The field strength should be as high as
possible, e.g., 1000 to more than 10,000 gauss. In case
of 10,000 gauss and a vessel diameter of 0.5 cm, a flow
i3*4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
rate of i ml per sec will generate a flow-signal voltage
of about 0.25- io~3 volts (equation 14). In the ascend-
ing aorta or pulmonary artery trunk, signals of several
millivolts can be recorded at flow peaks. Non-
polarizable electrodes are indispensable. Zn-ZnS04
electrodes are useful; calomel half cells give still more
satisfactory results [Katz & Jochim (72); Jochim
(68)]. The electrodes are connected to the vessel wall
by wicks soaked in saline-agar solution or by saline-
agar filled glass tubes. Also Ag-AgCl electrodes are
recommended (33). Furthermore, the vessel's di-
ameter and cross-sectional area must be kept constant
throughout the measurements. The best way is to use
a rigid sleeve of insulating material (76). The size of
the sleeve should be carefully chosen so that the
vessel is narrowed down to that diameter which would
be reached if the blood pressure fell to the lowest
level expected during the experiment. This moderate
constriction will not essentially interfere with the
hemodvnamic conditions, nor will the rigid sleeve
give rise to pulse-wave reflections if it is not longer
than about 1 cm. The tips of the saline connections to
the electrodes are contained in two small holes placed
in the sleeve wall at right angles to the vessel axis and
to the lines of magnetic force. The sleeve also assures
a fixed position of the vessel relative to the magnet and
protects the exposed vessel from drying as well as from
undesired contact with neighboring tissues. To permit
introducing the vessel, either the sleeve has a small
longitudinal slot, or is composed of two halves which
are joined together around the vessel. If the flow-
signal voltage picked up by the electrodes is high
enough, it can be directly recorded by a string gal-
vanometer (75, 133). However, d-c amplifiers are
generally employed (33, 64, 68, 110, 128, 134), or
the input voltage is converted into alternating current
by a mechanical chopper (72, 76), and a capacitance-
coupled amplifier may then be used. The over-all fre-
quencv response of the amplifier system and recording
galvanometer should be uniform up to at least 50 cps.
The base line is assessed during the experiment
either bv clamping the vessel distal to the site of
measurement or by de-energizing the magnet. See
also (32). Calibration is performed by perfusing the
excised vessel or the vessel in situ with blood or saline
solution at known flow rates. Because of the strictly
linear calibration curve, it is sufficient to determine,
in addition to the zero point, only one point corre-
sponding to a flow rate near the upper limit of the
range under investigation. The calibration can also
be done using a nonsteady flow: a known quantity of
fluid is injected into the vessel by means of a syringe,
and the course of the corresponding flow signal is re-
corded. Thus the mean flow rate and the mean flow
signal can be calculated from the known injected vol-
ume and the time and deflection as recorded on the
tracing (73).
In spite of its theoretical simplicity, the d-c pro-
cedure has been widely abandoned because of several
practical drawbacks. The magnet and most types of
nonpolarizable electrodes are rather bulky. In the
case of flow measurements on small vessels, the flow-
signal voltage is very low so that high-gain d-c ampli-
fication with its inherent difficulties is required and
changes of the electrode potential will cause drift of
the base line. The results of Richards & Williams (1 10)
and of Inouye el al. (65) show that, in spite of utmost
care, such difficulties are present even in the applica-
tion of the d-c procedure to the dog's carotid and
femoral arteries. By improving the electrodes and
using modern stabilized d-c amplifiers, however, satis-
factory short-time recordings of the flow in the de-
scending aorta of the dog have been made possible
[Feder & Bay (33)]. As to vessels close to the heart,
cardiac action potentials may be picked up by the
electrodes in addition to the flow signal.
The a-c modification [Kolin (76, 77)] is character-
ized by the use of an alternating magnetic field which
is generated by energizing the coils of the electro-
magnet with sinusoidal alternating current:
B = BQsin
40 cps
pulse
\\\z
fig. 30. Block diagram of the square-wave flowmeter circuits required to energize the magnet,
eliminate spurious emfs, amplify the flow signal, and convert it to direct current. [From Spencer &
Denison (126).]
TYPE U
TYPE C
C 3 2 I A B
8 cm
H
r-
5 cm
H
fig. 31. Three types of magnet-sleeve units used in the square -wave flowmeter technique: Type U
(horseshoe) mainly employed in surgical measurements, type C for implantation about large vessels,
type I for small vessels. .4, B, C, magnet-coil terminals; I, 2, 3, electrode leads; 2, 3, split lead. The
units are imbedded in plastic cast. [From Spencer & Denison (126).]
TYPE I
375cm
ULTRASONIC FLOWMETERS
The measurement of blood velocity by recording
sound-transit times upstream and downstream within
a vessel segment of small length offers, in principle,
several important advantages. The device placed
around the vessel is very lightweight and simple in
construction. The vessel remains intact, and there is
no interference with the blood flow or pulse wave
except that effected by a short rigid sleeve causing a
moderate constriction. The calibration curve can be
made to be a straight line passing through the zero
point with equal slopes for forward and backward
flow. The signals obtained can follow the most rapid
changes of the instaneous blood velocity occurring in
the circulation. However, in contrast to the simple
device applied to the vessel itself, very involved
electronic equipment is required to detect and evalu-
ate the extremely small effects exerted by the flow on
the sound transit times. Since sound velocity in blood
is about 1.5- 1 o5 cm per sec, the time required for
traveling over a distance of 1 cm is about 7 n sec. If
the blood is moving at the velocity v along the direc-
tion of sound propagation, the apparent sound veloc-
ity measured between two quiescent points is (c — v)
or (c + v) for upstream or downstream sound direc-
tion, respectively. A flow velocity of 1 cm per sec will
therefore change the sound transit time (7 fisec/cm)
by about ±5'io~n sec. Thus utmost precision is
necessary if differences of such a minute order of
magnitude are to be measured with sufficient accu-
racy, and the admirable advances made in this field
to date are based on very difficult and detailed work.
The development of ultrasonic blood flowmeters has
been carried out mainly by two groups using different
approaches.
Haugen et al. (58) and Herrick & Anderson (59),
modifying the design of Kalmus, developed a phase-
METHODS OF MEASURING BLOOD FLOW
>3>9
difference procedure. Cylindric ceramic "trans-
ducers" are placed about i inch apart around the
vessel wall. They are arranged to transmit and receive
ultrasound (/ = 400 kc sec) alternately upstream and
downstream at a rate of 75 per sec. The phase differ-
ences between the signals received upstream and
downstream are detected by phase meters and used as
a measure of the differences (A/) between the upstream
and downstream sound transit times:
*"*■&
c+v
l_) ~ 2Lv
(18)
where L = distance between the transducers. Since
the phase angle A = 2irf-At, we get:
A
4irfLv
radians.
(19)
In the present design (31, 59) where/ = .l-io5 cps
and L = 2.5 cm, a blood velocity of 1 cm per sec will
cause a phase angle of about 5- io~4 radians or 0.03°.
The output signal of the apparatus is proportional
to A4>. Due to undesired phase differences, assessment
of the base line remains a difficult problem. An im-
provement was achieved by introducing an automatic
phase-shift control. The authors succeeded in con-
structing a reliable recorder of extracorporeal blood
flow. Preliminary findings indicate that satisfactory
results may ultimately be attained on vessels in vivo.
For this purpose, the switching rate of the transducers
has to be increased, and the time constants of some
of the circuits have to be reduced (59).
Franklin et al. (43-45) made use of the pulse tech-
nique in detecting and evaluating the differences in
upstream and downstream sound-transit times. As
seen in figure 32, their flow-sensing element consists
of a short (1-3 cm) Lucite cylinder which is split
longitudinally and mounted snugly about the vessel.
The sound is transmitted and received by two barium
titanate crystals placed on the vessel wall diagonally
from each other across the vessel lumen. The crystals
are set to function alternately 800 times per sec as
transmitter and receiver. The respective transmitter
crystal is pulse-excited at a repetition rate of 12,000
per sec so that it will, during each switching period of
1/800 sec, give off a train of ultrasound bursts at
its resonant frequency of 3 mc per sec. These waves
travel through the adjacent vessel wall, the blood, and
the opposite vessel wall to reach the receiver crystal.
In the next switching period, the functions of both
crystals are exchanged so that in every 1 400th sec a
train of upstream and a train of downstream transits
are available for determination of At. Equation 18 is
applicable to this device if L is replaced by d-cos 9
where d = length of diagonal between the crystals
and 9 = angle between diagonal and vessel axis.
The transit-time voltage converter generates a ramp
voltage showing a strictly constant slope of 40 volts
per ^sec. This ramp voltage is started at the beginning
of every sound transmission, and its ascent is abruptly
stopped when the respective receiver crystal begins to
be excited by the sound, so that the amplitude of the
ramp voltage is proportional to the sound-transit
time. It is obvious that, due to the blood flow, the up-
stream ramp-voltage amplitude is greater than the
downstream one. This difference amounts to 4 mv per
io-10 sec and is detected by the voltage comparator
which delivers a 400 cps square-wave voltage with an
amplitude proportional to the difference between the
upstream and downstream ramp-voltage amplitudes.
Finally, a synchronized detector converts the square
wave into a d-c voltage, which indicates the instanta-
neous magnitude and direction of the blood velocity.
The device possesses satisfactory sensitivity to flow
and high stability of the base line. The stability is
achieved mainly by using whenever possible, only one
functional unit or channel for detecting differences in
time or voltage of consecutive events. Due to the
carrier frequency of 400 cps, the apparatus is capable
of an excellent frequency response to pulsatile blood
flow. As far as seen from the tracings published in
reduced scale, the flow patterns recorded on blood
vessels of different sizes in vivo are very similar to
those obtained by the electromagnetic method. In
addition, the simultaneous application of several or
many ultrasonic meters is possible without any
mutual interaction (44). One is inclined to predict
that this kind of versatile flowmeter is on its way
toward becoming a favorite instrument in cardio-
vascular research. The same may happen regarding
the application of ultrasound to the recording of
instantaneous dimensional changes of organs [Keidel
(74); Edler & Hertz (27); Rushmer et al. (1 15)].
However, it seems that the possible dependence of
the calibration of the ultrasonic flowmeters on the
velocity profile has not yet been duly considered. The
sound passes from the transmitter to the receiver
crystal on a diagonal path which crosses the stream-
lines of moving fluid at the angle 9 (see fig. 32). It
may be assumed that only the streamlines crossing
this diagonal will cause flow-related changes of the
sound-transit times. Furthermore, the relative velocity
distribution taken over the diagonal may be con-
sidered to equal that taken over the vessel's diameter
or radius. This means that the device will indicate the
[320
HANDBOOK OF PHYSIOLOGY -^ CIRCULATION II
fig. 32. Simplified diagram of the
pulsed ultrasonic flowmeter. For de-
scription see text. [From Franklin el al.
(44)-]
UPSTREAM- DOWNSTREAM
VOLTAGE COMPARATOR
1-
SYNCHRONOUS
DETECTOR
RECORDER
AORTIC FLOW
JUUjUX
H-lsec*
velocity vR averaged over the vessel's diameter or
radius, and that therefore the calibration in terms of
the flow rate varies with the velocity profile. On the
other hand, corrections might be brought about by
additional effects, such as some flattening of the
velocity profile by the slight constriction of the vessel
caused by the sleeve, and the fact the ultrasound
beam reaching the receiver is not in the form of a line,
but of a band. Rushmer describes the calibration of
his ultrasonic flowmeter as being independent of the
velocity profile within ±5 per cent (44). With respect
to the growing importance of this flowmeter type, the
problem should be reconsidered both theoretically
and practically.
TRAVELING MARKERS
Estimations of blood velocity can be made by
observing, continuously photographing, or filming
the movement of any substance which acts as a dis-
tinguishable marker traveling with the blood stream.
In most cases, such a procedure will allow only single
short-time recordings which can be repeated at
intervals. The marker may be represented by dye, by
a drop of fluid nonmiscible with the blood [see (54,
p. 60) (50, p. 116)], by radiopaque material for
cineradiography (2, 7, 24, 90, 106), or by a gas bubble.
Even the blood's own corpuscles can be used as
markers (63), and the progress of blood columns
differing in oxygen saturation may be assessed photo-
electrically (85) (cf Chapter 18, vol. I, of this Hand-
book). Foreign substances are usually injected into a
side branch and then observed through the wall of
the vessel under investigation. In contradistinction
to flowmeters in the proper sense, the use of traveling
markers does not give the volume flow at a fixed site;
it rather gives a function of time and space since the
mark changes position during the measurement. In
case of relatively small displacements, the change in
the site of measurement may be neglected. Dyes are
particularly useful for the study of the flow course in
small vessels. Valuable results were obtained with
China ink and high-speed cinematography on pulmo-
nary capillaries (130) and on very small arteries of
the rabbit ear (136).
In a carefully elaborated procedure, McDonald (91-
93) studied the flow pulse in the rabbit aorta and in
peripheral arteries of the dog by filming the move-
ment of injected gas bubbles through the translucent
vessel wall. Gas embolism was avoided by using pure
oxygen instead of air. An injected gas bubble travels
at a velocity quite near to the average blood velocity
vA if the bubble is spherical and just fills the lumen
completely. Smaller spherical or larger cylindrical
bubbles will run faster. High-speed cinematography
METHODS OF MEASURING BLOOD FLOW
1321
at about 1000 frames per sec is used, the exposure of
each frame being 200 /usee. For evaluation, the
distance-time relation is plotted from the projected
film, and the time course of the velocity is obtained
by graphic differentiation. McDonald's work is of
particular significance regarding hemodynamics be-
cause his simultaneous recordings of the pressure
gradient make possible hydrodynamic flow calcula-
tions and comparison of the calculated flow pattern
with that determined by cinematography.
MISCELLANEOUS METHODS
Rockemann (114) tried to measure the blood
velocity by means of electrolytic polarization taking
place at electrode surfaces which are in contact with
the streaming blood. A stable calibration, however, is
not obtainable for this method.
The application of nuclear magnetic resonance to
blood flow measurements was described almost simul-
taneously by Buchman and by Singer in 1959. Buch-
man's device (17) passes protons, the spin axes of
which have been aligned, through a varying magnetic
field. Energy is required to bring them into resonance.
Thus the absorbed energy is a measure of the number
of protons passing per time unit, and hence, is propor-
tional to the flow rate. Singer (124) uses several
methods based on nuclear magnetic resonance. In
one of the procedures, the nuclear relaxation time of
the protons in the water of streaming blood is meas-
ured and compared with the relaxation time of
those in stopped blood; only single determinations of
relative values are obtained. By another procedure
described by Singer, absolute flow velocities can be
recorded at short time intervals. The nuclei are per-
turbed by the 60 mc per sec field of a transmitter coil,
and the time required by the nuclei to reach a second
detecting coil is measured. Singer also considers
nuclear or electron magnetic resonance as a tracer
detection system. It seems worthwhile to carry on the
development of these methods since they are appli-
cable to unopened vessels, even from outside through
the intact skin.
ADDENDUM
Since completion of the manuscript, several papers have
been published which should be referred to:
Elliott, S. E., J. I. E. Hoffman, and A. Guz. An electro-
magnetic flowmeter for simultaneous measurement of ventric-
ular ejection in the conscious animal. Digest of 4th Intern. Conf.
Med. Electronics, New York, 1961, p. 150. Coreless electro-
magnetic flowmeter units were implanted around the ascending
aorta and pulmonary artery of the dog. A-c sine-wave type,
400 cps. Some magnetic interference between both units was
observed.
Yanof, H. M. A New Trapezoidal-wave Electromagnetic
Blood Flowmeter and Its Application to the Study of Blood
Flow in the Dog (Ph.D. thesis). Berkeley: Univ. of California,
iq6o. Description of circuitry. 1000 cps. Adjustment of mini-
mum transformer emf by an additional ferrite slug.
Wyatt, D. G. Problems in the measurement of blood flow
by magnetic induction. Phys. in Med. Biol. 5: 289, 1961. Thor-
ough examination of performance, and possible error sources
in the application, of electromagnetic flowmeters.
Zarnstorff, W. C, and C. A. Castillo. An ultrasonic flow-
meter. Digest 0/ 4th Intern. Conf. on Med. Electronics, New York,
1 96 1, p. 86. Stable phase-difference device appropriate for
recording of blood How in unopened arteries.
Franklin, D. L., D. W. Baker, and R. F. Rushmer. Pulsed
ultrasonic transit time flowmeter. IRE Trans, on Bio-Med.
Electronics BME-g, 44, 1962. Diagrams of electronic circuitry.
Higasi, K. (ed.). Platinum blood flowmeter. Research Inst.
Appl. Elec, Hokkaido Univ., Monograph Ser. No. 10, 1962. Con-
tains several papers by M. Mochizuki and co-workers on the
relation between polarographic current for oxygen and the flow
velocity. Application to flow measurement in arteries. Catheter-
tip method. Calibration curve concave to flow abscissa at low
velocities and linear at higher velocities.
REFERENCES
1 . Abel, F. L. Chopper -operated electromagnetic flowmeter.
IRE Trans, on Med. Electronics ME-6: 216, 1 959.
2. Anschutz, F., and F. Heuck. Uber die durch Aorten-
sklerose verursachten Veranderungen der arteriellen
Blutstrbmung. Z. Kreislauforsch. 49 : 1 20, 1 960.
3. Barnes, C. W. A new method for obtaining How signals
from the electromagnetic flowmeter. Naturwissenschaften
47: 56, i960.
4. Baxter, I. G, and J. W. Pearce. Simultaneous measure-
ment of pulmonary artery flow and pressure using
condenser manometers. J. Physiol., London 115: 410, 1 951 -
5. Bergmann, G. Die "Stromborste", ein elektrischer Ge-
schwindigkeitsmesser fur Fliissigkeiten. (2. Mitteil.) Z.
Biol. 98: 536, 1938.
6. Betticher, A., J. Maillard, and A. Muller. Un mano-
metre differentiel a transmission electrique entierement
alimente sur le reseau alternatif, pour mesurer la vitesse
d'ecoulement dans des tuyeaux et des vaisseaux sanguins.
Helv. Physiol, et Pharmacol. Acta 12: 112, 1954.
7. Bohme, W. Uber den aktiven Anteil des Herzens an der
Forderung des Venenblutes. Ergeb. Physiol. 38: 251, 1936.
Eortschr. Rontg. Sir. 57: 59, 1938.
[322
HANDBOOK OF PHYSIOLOGY ~ CIRCULATION II
■4-
'5-
it).
'7-
iq.
23-
24-
26
27-
28.
=9'
3°
Brecher, G. A. Venous return during intermittent 31.
positive-negative pressure respiration studied with a new
catheter flowmeter. Am. J. Physiol. 174: 299, 1953.
Brecher, G. A. Critical review of bristle flowmeter 32.
techniques. IRE Trans, on Med. Electronics ME-6: 294,
'959-
Brecher, G. A. Bristle flowmeter. In: Methods in Medical 33.
Research. Chicago: Yr. Bk. Pub., i960, vol. 8, p. 307.
Brecher, G. A., and C. A. Hubay. A new method for
direct recording of cardiac output. Proc. Soc. Expll. Biol.
Med. 86: 464, '954- 34-
Brecher, G. A., and J. Praglin. A modified bristle
flowmeter for measuring phasic blood flow. Proc. Soc.
Exptl. Biol. Med. 83: 155, 1953. 35-
Bretschneider, H.J. Verhandl. deut. Ges. Chirurgie, 1961.
Broemser, P. Der Differentialsphygmograph. Z. Biol.
88: 264, 1928. 36.
Broemser, P. Untersuchungen iiber die Messung der
Stromstarke in BlutgefalJen. (3. Mitteil.) Z. Biol. 88:
296, 1928. 37-
Broemser, P., and O. F. Ranke. Beitrag zur Re-
gistrierung der Kurve der Stromungsgeschwindigkeit 38.
pulsierender Strome, zugleich eine Erwiderung an Otto
Frank. Z. Biol. 91 : 267, 1931.
Buchman, P. Nuclear Magnetic Resonance Blood Flow- 39.
meter. (Thesis). Seattle : Univ. of Washington, 1959.
Clark, J. W., and J. E. Randall. An electromagnetic 40.
blood How meter. Rer. Sci. Instr. 29: 951, 1949.
Cvbulski, N. Die Bestimmung der Stromgeschwindigkeit 41.
des Blutes in den Gefassen mit dem neuen Apparat-
Photohamotachometer. Pfliigers Arch. ges. Physiol. 37 :
382, 1885. 42-
Daly, I. de Burgh. A blood velocity recorder. J. Physiol.,
London 61 : 21P, 1926.
Daly, I. de Burgh. The resistance of the pulmonary 43.
vascular bed. J. Physiol., London 69: 238, 1930.
Denison, A. B., M. P. Spencer, and H. D. Green. A 44.
square-wave electromagnetic flowmeter for application
to intact blood vessels. Circulation Research 3 : 39, 1 955.
Denison, A. B., and M. P. Spencer. Magnetic flowmeters. 45.
In: Medical Physics. Chicago: Yr. Bk. Pub., i960, vol. 3,
p. 178. 4D-
Dotter, C. T., and L. H. Frische. Radiologic technic
for qualitative and quantitative study of blood flow.
Circulation 18:961, 1958. 47.
Eckstein, R. W., M. Stroud, C. V. Dowling, R. Eckel,
and W. H. Pritchard. Response of coronary blood flow
following stimulation of cardiac accelerator nerves. 48.
Federation Proc. 8 : 38, 1 949.
Eckstein, R. W., C. J. Wiggers, and G. R. Graham.
Phasic changes in inferior vena cava flow of intravascular 49.
origin. Am. J. Physiol. 148: 740, 1947.
Edler, J., and C. H. Hertz. Kgl. Fysiograf Sdllskap. Lund 50.
Fb'rh. 24: 5, 1954. (Quoted from Effert et at. Z. Kreis-
laufforsch. 48 : 230, 1 959.)
ElNHORN, H. D. Electromagnetic induction in water. 51.
Trans. Roy. Soc. S. Africa 28: 143, 1940.
Evans, R. L. Cardiac output and central pressure data. 52
Nature 181 : 1471, 1958.
Fabre, P. Utilisation des forces electromotrices d'induc-
tion pour renregistrement des variations de vitesse des 53
liquides conducteurs : un nouvel hemodromographe sans
palette dans le sang. Compl. rend. Acad. Sci. 194: 1097, 1932.
Farrall, \V. R. Design considerations for ultrasonic
flowmeters. IRE Trans, on Med. Electronics ME-6: 198,
'959-
Feder, W. Resume of dc electromagnetic flowmeter
group discussion. IRE Trans, on Med. Electronics ME-6:
250, 1959-
Feder, \V. , and E. B. Bay. The dc electromagnetic
flowmeter and its application to blood flow measurement
in unopened vessels. IRE Trans, on Med. Electronics ME-6:
240, 1959.
Ferguson, D. J., and H. S. Wells. Frequencies in pulsa-
tile flow and response of magnetic meter. Circulation
Research 7: 336, 1959.
Ferguson, D. J., and H. S. Wells. Harmonic analysis of
frequencies in pulsatile blood flow. IRE Trans, on Med.
Electronics ME-6: 291, 1959.
Frank, O. Die Benutzung des Prinzips der Pitot'schen
Rohrchcn zur Bestimmung der Blutgeschwindigkeit.
Z. Biol. 37: 1, 1899.
Frank, O. Kritik der elastischen Manometer. Z. Biol.
44 : 445. I9°3-
Frank, O. "Hamodynamik." In: Handbuch der physiolo-
gischen Methodik, edited by R. Tigerstedt. Leipzig: Hirzel,
1908, vol. 2.
Frank, O. Der Ablauf der Stromungsgeschwindigkeit in
den Gefassen. Z. Biol. 88: 249, 1928.
Frank, O. Theorie und Konstruktion eines optischen
Strompendels. Z. Biol. 89: 83, 1929.
Frank, O. Kurze Bemerkungen iiber die Bestimmungen
der Blutgeschwindigkeit. Sitz-Ber. Ges. Morphol. Physiol.
Mitn, hen 39: 19, 1929.
Frank, O. Bemerkungen zu der Abhandlung von Otto
Ranke: Uber die Registrierung der Stromungsge-
schwindigkeit usw. Z. Biol. 90: 181, 1930.
Franklin, D. L., and R. M. Ellis. A pulsed ultrasonic
flowmeter. Federation Proc. 1 7 : 48, 1 958.
Franklin, D. L., D. W. Baker, R. M. Ellis, and R. F.
Rushmer. A pulsed ultrasonic flowmeter. IRE Trans, on
Med. Electronics ME-6: 204, 1959.
Franklin, D. L., R. M. Ellis, and R. F. Rushmer. Aortic
blood flow in dogs. J. Appl. Physiol. 14: 809, 1959.
Fry, D. L. The measurement of pulsatile blood flow by
the computed pressure gradient technique. IRE 'Trans, on
Med. Electronics ME-6: 259, 1959.
Fry, D. L. Methods of flow estimation by pressure sensing
techniques. IRE Tunis, on Med. Electronics ME-6: 264,
1959-
Fry, D. L., A. J. Mallos, and A. G. T. Casper. A
catheter tip method for measurement of the instantaneous
aortic blood velocity. Circulation Research 4: 627, 1956.
Gauer, O. H., and E. Gienapp. A miniature pressure-
recording device. Science 112: 404, 1950.
Green, II. D. Differential pressure flow meters. In:
Methods in Medical Research. Chicago : Yr. Bk. Pub., 1948,
vol. 1 .
Green, H. D. Circulatory system: methods. In: Medical
Physics. Chicago: Yr. Bk. Pub., 1950, vol. 2.
Green, H. D. and D. E. Gregg. The relationship
between differential pressure and blood flow in a coronary
artery. Am. J. Physiol. 130: 97, 1940.
Green, H. D., D. A. Gregg, and C. J. Wiggers. The
phasic changes in coronary flow established by differential
pressure curves. Am. J. Physiol. 112: 627, 1935.
METHODS OF MEASURING BLOOD FLOW
I323
54. Gregg, D. E. Coronary Circulation in Health awl Disease.
Philadelphia: Lea & Febiger, 1950.
55. Gregg, D. E., and H. D. Green. Registration and inter-
pretation of normal phasic inflow into a left coronary-
artery by an improved differential manometric method.
Am. J. Physiol. 1 30 : 114, 1 940.
56. Gregg, D. E., R. E. Shipley, R. W. Eckstein, A. Rotta,
and J. T. Wearn. Measurement of mean blood flow in
arteries and veins by means of the rotameter. Proc. Soc.
Exptl. Biol. Med. 49: 267, 1942.
57. Hardung, V. Zum Gebrauch des Pitot-Rohres bei
nichtstationarer Stromung. Arch. Kreislaufforsch. 26: 337,
'957-
58. Haugen, M. G., W. R. Farrall, J. F. Herrick, and
E. J. Baldes. An ultrasonic flowmeter. Proc. Natl. Elec-
tronics Conf. 1 1 : 464, 1955.
59. Herrick, J. F., and J. A. Anderson. Ultrasonic flow-
meter. In: Medical Physics. Chicago: Yr. Bk. Pub., i960,
vol. 3, p. 181.
60. Hilger, H. H., and H. Brechtelsbauer. Erfahrungen
liber Stromungsmessungcn mit verschiedenen Typen
elektrisch registrierender Rotameter. Pfliigers Arch. ges.
Physiol. 263:615, 1957.
61. Holzlohner, E. Die "Stromborste," ein elektrischcr
Geschwindigkeitsmesser fur Flussigkcitcn. (1. Mitteil.)
Z. Biol. 98:533, 1938.
62. Holzlohner, E., and B. Schonerstedt. Der Strompuls
der Vena jugularis. Z. Biol. 100: 51, 1940.
63. Hurthle, K. Eine Methode zur Registrierung der
Geschwindigkeit des Blutstroms in den kapillaren Ge-
fassen. Pfliigers Arch. ges. Physiol. 162: 422, 1915.
64. Inouve, A., and H. Kuga. On the applicability of the
electromagnetic flowmeter for the measurement of
blood flow rate. Japan. J . Physiol. 4: 205, 1954.
65. Inouye, A., H. Kuga, and G. Usui. A new method for
recording pressure-flow diagram applicable to peripheral
blood vessels of animals and its application. IE Japan. J.
Physiol. 5 : 236, 1 955.
66. James, W. G. An induction flowmeter design suitable for
radioactive liquids. Rev. Sri. Instr. 22: 989, 1951.
67. Jameson, A. G. Instantaneous linear velocity of flow in
pulmonary artery measured by a catheter tip method.
Science 128:592, 1958.
68. Jochim, K. E. Electromagnetic flowmeter. In: Methods
in Medical Research. Chicago: Yr. Bk. Pub., 1948, vol. 1,
p. 108.
6g. Jochim, K E. Electromagnetic flowmeter. In: Medical
Physics. Chicago: Yr. Bk. Pub., 1950, vol. 2, p. 224.
70. Johnson, J. R., and C. J. Wiggers. Alleged validity of
coronary sinus outflow as criterion of coronary reactions.
Am. J. Physiol. 118: 38, 1937.
71. Jones, W. B., L. L. Hefner, J. R. Bancroft, and \V.
Klip. Velocity of blood flow and stroke volume obtained
from the pressure pulse. J. Clin. Invest. 38: 2087, 1959.
Katz, L. N., and K. E. Jochim. Electromagnetic flow-
meter. In: Medical Physics. Chicago: Yr. Bk. Pub., 1947,
vol. 1, p. 377.
Katz, L. N., and A. Kolin. The flow of blood in the
carotid artery of the dog under various circumstances as
determined with the electromagnetic flowmeter. .4m. J.
Physiol. 122: 788, 1938.
74. Keidel, W. D. Uber eine neue Methode zur Registrierung
75-
76.
77-
78.
79-
80.
81.
82.
83-
84.
85-
86.
87
88.
72
73
90.
9'-
92.
93-
94-
95-
96.
der Volumenanderungen des Herzens am Mensehen.
Z. Kreislaufforsch. 39: 257, 1950.
Kolin, A. An electromagnetic flowmeter. Principles of the
method and its application to blood flow measurements.
Proc. Soc. Exptl. Biol. Med. 35: 53, 1936.
Kolin, A. Electromagnetic rhcometry and its application
to blood flow measuremenLs. Am. J. Physiol. 122: 797, 1 938.
Kolin, A. An a.c. induction flowmeter for measurement of
blood flow in intact blood vessels. Proc. Soc. Exptl. Biol.
Med. 46: 235, 1 94 1.
Kolin, A. Electromagnetic selometry. I. A method for
the determination of fluid velocity in space and time.
J. Appl. Physiol. 15: 150, 1944.
Kolin, A. An alternating held induction flowmeter of
high sensitivity. Rev. Sci. Instr. 16: 109, 1945.
Kolin, A. Improved apparatus and technique for electro-
magnetic determination of blood flow. Rev. Sci. Instr. 23 :
235. '952-
Kolin, A. Electromagnetic blood flow meters. Science 1 30 :
1088, 1959.
Kolin, A. Blood flow determination by electromagnetic
method. In: Medical Physics. Chicago: Yr. Bk. Pub., i960,
vol. 3, p. 141.
Kolin, A., and R. T. Kado. Miniaturization of electro-
magnetic flowmeter. Proc. Acad. Sci. 45: 131 2, 1959.
Kolin, A., and L. N. Katz. Observation de la vitesse
instantanec du sang a l'aide du rheometre electromag-
netique. Ann. Physiol. 13: 1022-1029, 1937.
Kramer, K. Uber die Messung der Stromungsge-
schwindigkeit des Blutes in unerrjffneten Arterien. Ein un-
blutiges Kontrollverfahren zur Reinschen Thermostrom-
uhr. Pfliigers Arch. ges. Physiol. 238: 91, 1936.
Laszt, L., and A. Muller. Uber Druck-und Geschwindig-
keitsverhaltnisse im Coronarkreislauf des Hundes.
Helv. Physiol, et Pharmacol. Acta 1 5 : 38, 1 957.
Lauber, H. Untersuchungen iiber die Messung der Strom-
starke in Blutgefassen. (1. Mitteil.). /. Biol. 88: 277, 1928.
Lawson, H., and J. P. Holt. A differential manometer
method for the measurement of blood flow. J. Lab. Clin.
Med. 24:639, 1939.
Lutz, J., O. Harth, W. Ohler, and W. Kreienberg.
Durchblutungsmessung mit einem technischen Durch-
flussmesser nach dem Induktionsprinzip. Pfliigers Anh. ges.
Physiol. 270: 540, i960.
Lynch, P. R., B. L. Carter, J. Gimenez, and R. Krisch.
Venae cavae flow pattern in cats: as studied with high-
speed cineradiography. .4m. ./. Physiol. 199: 1135, i960.
McDonald, D. A. The velocity of blood flow in the rabbit
aorta studied with high-speed cinematography. J. Physiol.,
London 118:328. 1952.
McDonald, D. A. The relation of pulsatile pressure to
flow in arteries. J. Physiol., London 1 27 : 533, 1 955.
McDonald, D. A. Blood Plow in Arteries. London : Arnold,
i960.
Mixter, G. Respiratory augmentation of inferior vena
cava flow demonstrated by a low-resistance phasic flow-
meter. .4m. J. Physiol. 172:446, 1953.
Muller, A. Uber die Verwendung des Pitot-Rohres zur
Geschwindigkeitsmessung. Helv. Physiol, el Pharmacol.
Acta 12: 98, 1954.
Muller, A. Uber die Verwendung des Castelli-Prinzips
zur Geschwindigkeitsmessung. Helv. Physiol, et Pharmacol.
Acta 12 : 300 1954.
1324
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
97. Nilsson, N. J., and K. Kramer. Stromvolumpulse der 118.
herznahen Venen bei verschiedenen Kreislaufzustanden.
Z. Biol. 1 06 : 386, 1 954.
98. Olmstead, F. Measurement of cardiac output in unre-
strained dogs by an implanted electromagnetic meter. 119.
IRE Trans, on Med. Electronics AIE-6 : 2 1 o, 1 959.
99. Olmstead, F., and F. D. Aldrich. Improved electro-
magnetic flowmeter; phase detection, a new principle. J.
Appl. Physiol. 16: 197, 1961. 120.
100. Pieper, H. Registration of phasic changes of blood flow by
means of a catheter-type flowmeter. Rev. Sci. Instr. 29: 965,
1958. 121.
1 01. PrEPER, H., and W. Vogel. Zur Messung der Stromungs-
geschwindigkeit des Blutes mittels katheterformiger 122.
Differenzdruckmanometer. Z. Biol. 109:62, 1956.
102. Pieper, H., and E. Wetterer. Strompendel fur elek-
trische Registrierung der Blutstromungsgeschwindigkeit. 123.
Z. Biol. 105:214, 1952.
103. Pieper, H., and E. Wetterer. Elektrische Registrierung
der Blutstromungsgeschwindigkeit mit neuartigen Strom- 124.
pendeln. Verhandl deut. Ges. Kretslaujforsch. 19:264, 1953.
104. Pieper, H., and E. Wetterer. Die Beziehungen zwischen 125.
Blutdruck und direkt gemessener diastolischer Strom-
starke einzelner arterieller Gebiete bei kiinstlich herbeige-
fiihrten periodischen Druckanderungen. Verhandl deut. Ges. 1 26.
Kreislaufforsch. 2 1 : 439, 1 955.
105. Potter Engineering Co., 87 Academy St., Newark, N. J.
106. Prec, O., L. N. Katz, L. Sennett, R. H. Roseman, A. P.
Fishman and W. Hwang. Determination of kinetic energy
of the heart in man. .4m. ./. Physiol. 159: 483, 1949. 127.
107. Ranke, O. F. Uber die Registrierung der Kurve der
Stromungsgeschwindigkeit bei ungleichmaliiger Stro-
mung. Z. Biol. 90: 167, 1930.
108. Ranke, O. F. Das Entzerrungsgerat. Z. Biol. 93: 227, 128.
!932-
109. Reissinger, H. Untersuchungen iiber die Messung der
Stromstarke in Blutgefalten. Z. Biol. 88: 286, 1928. 129.
1 10. Richards, T. G., and T. D. Williams. Velocity changes in
the carotid and femoral arteries of dogs during the cardiac 1 30.
cycle. J. Physiol., London 120: 257, 1953.
111. Richardson, A. W. A simplified electromagnetic flow-
meter with high fidelity recording. J. Appl. Physiol. 14: 131.
658> '959-
1 12. Richardson, A. W., A. B. Denison, and H. D. Green. A
newly modified electromagnetic blood flowmeter capable 132.
of high fidelity flow registration. Circulation 5: 430, 1952.
113. Richardson, A. W., J. E. Randall, and H. M. Hines. A 133.
newly developed electromagnetic flow meter. J. Lab.
Clin. Med. 34: 1 706, 1949.
114. Rockemann, W. Versuche zur Messung der Blutge- 134.
schwindigkeit mit Hilfe der elektrischen Polarisation. Z.
ges. exptl. Med. 120: 375, 1953. 135.
115. Rushmer, R. F., D. L. Franklin, and R. M. Ellis. Left
ventricular dimensions recorded by sonocardiometry. 136.
Circulation Research 4: 684, 1956.
116. Sarnoff, S. J., and E. Berglund. The Potter electro- 137.
turbinometer: An instrument for recording total systemic
blood flow in the dog. IRE Trans on Med. Electronics ME-6:
270. '959-
117. Sarnoff, S. J., E. Berglund, and P. E. Waithe. The 138.
measurement of systemic blood flow. Proc. Soc. Exptl. Biol.
Med. 79: 414, 1952.
Scher, A. M., T. H. Weigert, and A. C. Young. Com-
pact flowmeters for the use in the unanesthetized animal,
an electronic version of Chauveau's hemodrometer. Science
118:82, 1953.
Schroeder, W. Druckdifferentialstromuhr zur Messung
der Stromungsgeschwindigkeit des Blutes in Arterien-
schlingen des wachen Hundes. Pfliigers Arch. ges. Physiol.
261 :5o7, 1955.
Shipley, R. E., D. E. Gregg, and E. F. Schroeder. An
experimental study of flow patterns in various peripheral
arteries. Am. J. Physiol. 138: 718, 1943.
Shipley, R. E., and C. Wilson. An improved recording
rotameter. Proc. Soc. Exptl. Biol. Med. 78: 724, 1951.
Shipley, R. E., and C. Wilson. A simplified recording
rotameter. In: Methods in Medical Research Chicago: Yr.
Bk. Pub., i960, vol. 8: p. 346.
Shirer, H. W., R. B. Shackelford, and K. E. Jochim.
A magnetic flowmeter for recording cardiac output. Proc.
WE, 1959. >901-
Sincer, J. R. Blood flow rates by nuclear magnetic
resonance measurements. Science 130: 1652, 1959.
Spencer, M. P. Differential pressure measurement :
Paired transducer system. In: Methods in Medical Research,
Chicago: Yr. Bk. Pub., i960, vol. 8: 341.
Spencer, M. P., and A. B. Denison. Square-wave electro-
magnetic flowmeter for surgical and experimental appli-
cation. In : Methods in Medical Research. Chicago : Yr. Bk.
Pub., i960, vol. 8: 321. (See also IRE Trans, on Med.
Electronics ME-6: 220, 1959.)
Taylor, M. G. The discrepancy between steady- and
oscillatory-flow calibration of flowmeters of the "bristle"
and "pendulum" types: A theoretical study. Phys. Med.
Biol. 2 : 324, 1 958.
Thurlemann, B. Methode zur elektrischen Geschwindig-
keitsmessung von Flussigkeiten. Helv. Physica Acta 14:
383> "941-
Ueno, A., and F. Taken ata. A new measurement of
blood flow. Japan. J. Pharmacol. 4:98, 1955.
Vogel, H. Die Geschwindigkeit des Blutes in den Lun-
genkapillaren. He/vet. Physiol, el Pharmacol. Acta 5: 105,
■947-
Wagoner, G. W., and A. E. Livingston. Application of
the Venturi meter to measurement of blood flow in vessels.
J. Pharmacol. Exptl. Therap. 32:171, 1 928.
Westersten, A., G. Herrold, and N. S. Assali. A gated
sine wave blood flowmeter. J. Appl. Physiol. 15: 533, i960.
Wetterer, E. Eine neue Methode zur Registrierung der
Blutstromungsgeschwindigkeit am uneroffneten GefalS.
Z. Biol. 98: 26, 1937.
Wetterer, E. Der Induktionstachograph. Z. Biol. 99:
158, 1938.
Wetterer, E. Eine neue manometrische Sonde mit elek-
trischer Transmission. Z. Biol. 101 : 332, 1943.
VVidmer, L. K. Zur Stromungsgeschwindigkeit in klein-
sten peripheren Arterien. Arch. Kreislaufforsch. 27 : 54, 1 957-
Wiggers, C. J., and F. W. Cotton. Studies on the coro-
nary circulation. II. The systolic and diastolic flow
through the coronary vessels. Am. J. Physiol. 106: 597,
■933-
Wretlind, A. Apparatus for the determination of the
mean blood flow in the ascending aorta of the cat. Ada
Physiol. Scand. 46: 291, 1959.
CHAPTER 39
The circulation through the skin
A. D. M. GREENFIELD
Department of Physiology, The Queen's University of Be/fast,
Belfast, Northern Ireland
CHAPTER CONTENTS
Introduction
Arrangement of the Blood Vessels of the Skin
Measurement of the Flow of Blood Through the Skin
Total cutaneous blood flow
Color of the Skin
Temperature of the Skin
Responses of Skin Blood Vessels to Physical Disturbances
Response of the Circulation Following Periods of Arrest or
Insufficiency
Reactive hyperemia
Hyperemia after prolonged insufficiency of the circulation
Responses of Skin Vessels to Changes in Transmural Pressure
Effect of Local Temperature on the Skin Circulation
Local temperatures in the range 15 C to 45 C
Local temperature in the range o C to 15 C: Cold vaso-
dilatation
Prolonged exposure to cold: Trench foot and immersion
foot
Exposure to severe cold : Frostbite
Reactions to Injury
Mechanical injury
Ultraviolet light
Arterial gas embolism
Nervous Control of Skin Blood Vessels
Vasomotor Nerves
Vasoconstrictor sympathetic nerves
Vasodilator sympathetic nerves
Vasodilatation caused by antidromic stimulation of
dorsal root sensory nerves. The axon reflex pathway
Innervation of the Blood Vessels of the Skin in Different
Areas
The human hand and fingers
The human forearm
Other areas of the human body
The skin of animals
Late effects of sympathetic denervation
Late effects of total denervation
Reflex Control of Blood Vessels of the Skin
Body temperature regulation
Emotion
Fainting
General sensory stimuli
Response to a deep inspiration
Response to distention of the bladder
Hypoglycemia
Posture
Responses to baroreceptor stimulation
Action of Humoral Agents on the Blood Vessels of the Skin
Adrenaline and Noradrenaline
Histamine
Acetylcholine
5-Hydroxytryptamine (Serotonin)
Adenosine Triphosphate
Bradykinin
Carbon Dioxide
Vasopressin
Oxytocin
INTRODUCTION
this section deals principally with the circulation
through human skin, since this has been so frequently
and carefully studied, mostly in unanesthetized sub-
jects. In selecting references no attention has been
in general paid to priority of discovery. Papers have
been chosen for the completeness of the information
they contain, for the value of their bibliography,
and very often because the work is personally known
to the present author. A wider bibliography will be
found in several excellent monographs and reviews
(1, 19, 22, 44, 115, 119, 139, 148, 152, 155, 165,
176, 187, 193).
The great bulk of observations relates to the cir-
culation through the skin of the extremities, and par-
ticularly of the digits. Here the striking features are,
firstly, the very great variability of the blood flow
under different circumstances, greatest in the tips
'3*5
[326
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. I. Projection drawing (X 95) of
a group of anastomoses in the nail bed
of the toe. The artery crosses the center
of the drawing and gives rise to 1 2
thick-walled anastomotic branches. The
thin-walled venous terminations ol
several are shown. [From Grant &
Bland (99).]
of the fingers where the maximum flow is probably
between 100 and 200 times the minimum (42),
and secondly, the fact that in normal persons the
blood flow to the skin is greatly in excess of its meta-
bolic requirements, being chiefly determined by the
need to maintain thermal balance (82). This lavish
circulation is, no doubt, valuable in the repair of
trauma and wounds to which the skin is especially
exposed. Perhaps because of methodological diffi-
culties, there is very little information about the
circulation through other areas of skin, but it is
almost certainly much less reactive than that through
the extremities.
Arrangement of the Blood I 'esseli of the Skin
The skin is supplied with a profuse system of capil-
lary loops which rise in the papillae of the corium
and return to enter a subpapillarv venous plexus.
The vessels of the latter are large and have thin walls,
and it is probable that when distended they contain a
very large proportion of all the blood in the skin.
There are rich capillary networks around the sweat
glands, at the base of hair follicles, around the seba-
ceous glands, and in the nail bed and nail fold.
In the skin of the extremities a special and promi-
nent feature is the large number of arteriovenous
anastomoses (50, 147). These are coiled channels
(fig. 1 ) with thick muscular walls and a lumen which
in the dilated state is between 20 and 70 n in diam-
eter, the average being 35 /x- They are abundantly
supplied with nerve endings, and a high concentra-
tion of cholinesterase has been reported around them
(28, 124, 151). They directly connect arterioles and
venules in the dermis at the level of, or a little super-
ficial to, the sweat glands. They are most numerous
in the nail bed, numerous at the tips of the digits,
less numerous on the palmar surface of the phalanges,
and almost absent from the dorsum of the phalanges.
They are fairly numerous in the palm of the hand
and sole of the foot, but are absent from the areas
of the forearm and calf which have been examined.
CIRCULATION THROUGH THE SKIN
I327
table I. Xumber of Anastomoses per
Square Centimeter of Surface Area
Hand
Index finger
Nail bed 501
Tip 236
Palm, 3rd phalanx 150
2nd phalanx 20
1st phalanx 93
Palm
Metacarpo-phalangeal joint 3rd finger 31
Thenar eminence 113
Hypothenar eminence 96
Foot
2nd toe
Nail bed 593
Pad 293
Sole, near heel 197
These are nil for dorsal surfaces of fingers, toes, hand, and
foot; flexor surfaces of lower forearm and lower calf of leg;
lower half of ear. [From Grant & Bland (99).]
table 2. Percentage Composition by Volume
of Parts of Human Limbs
Hand
Foot
Forearm
Skin
30.2
'7
8.6
'3-4
Subcutaneous
tissue
24
8.0
Fat
54-3
2
28.0
Bone
43
'3-7
Tendon
6.1
Muscle
'5-5
14
63.6
58.6
REFERENCES
Hand Average of 3 hands (2).
Foot Average of 2 feet (12).
Forearm A Average of 5 forearms (56).
Forearm B Average of 3 forearms (2).
Table 1, from Grant & Bland (99), summarizes the
distribution in the human. They have since been
found in the human ear (164). Some observers (151,
163) have reported rather smaller numbers (20-25/
cm2) than did Grant and Bland in the finger pad.
They are numerous in the external ear of the rabbit,
where their reactions have been carefully studied
(98), in the ear of the cat and dog and in the feet of
webfooted birds. Grant's (98) summary of the func-
tions of the anastomoses is still valid, and applies to
the human extremities as well as to the rabbit's
ear: "The anastomoses serve two functions (a) local,
and (b) general, (a) It is mainly through their agency
that the temperature of the ears is maintained when
they are exposed to cold, (b) They are important
factors in regulating of body temperature, aiding the
dispersal of heat by allowing an enormous blood
flow through the ears." (See also Chapters 27 and
37-)
Measurement of the Flow of Blood Through the Skin
The fingers and toes are composed largely of skin.
Of the total flow through them, the greater part
normally passes through skin, and total digital blood
flow is often used as an index of digital skin blood
flow (table 2). Digital flow may be directly measured
by venous occlusion plethysmography (40, 95), a
method which permits variation in the rate to be
followed from one heart beat to the next, and even
during a single beat. Flow may be estimated by
calorimetry (148), a method which, because of the
thermal capacity of the tissues, is incapable of follow-
ing rapid changes in flow, but which conveniently
integrates flow over a period of time. Calorimetry
finds its most successful application in the digits,
and the method has been rendered more versatile by
the use of copper-tellurium heat flow discs (47).
In the more proximal parts of the limbs, the total
blood flow depends a great deal on the circulation
through tissues, especially muscle, deep to the skin.
The flow through the skin can be deduced by com-
paring the total flow in a pair of segments in one of
which the circulation through the skin has been
suppressed by iontophoresis of adrenaline (71).
More often, indirect indices of skin blood flow have
been employed. If venous blood can be obtained from
vessels exclusively draining skin, and if the oxygen
usage of the skin is assumed to remain constant,
changes in flow can be inferred from changes in the
oxygen content of the blood (170). Measurements
of the temperature of the skin have provided useful
qualitative information in the proximal as well as
in the distal parts of the limbs (100), but this tempera-
ture, as explained later, depends on so many other
factors that it is a very imperfect index of skin cir-
culation. It is incapable of following accurately
rapid fluctuations in flow; if the circulation is com-
pletely arrested the temperature of the skin falls
very slowly. A more sensitive index of blood flow is
the change in thermal conductivity of the skin, which
can be conveniently measured by a surface applicator
containing two small plates, one of which is elec-
trically warmed while the temperature difference
1328
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
between the plates is recorded (1 14). The instrument
does not, of course, distinguish between the effect
of blood circulating through the local blood vessels
of the skin, and blood flowing through nearby veins
draining distal regions. The results cannot be quan-
titatively translated into measurements of blood
flow, but the method retains its sensitivity over the
wide ranges of flow for which measurement of skin
temperature is of little help.
The rate of clearance of radiosodium from an in-
jection site (130) probably depends on the rate of
blood flow through those vesse's which nourish the
tissues, and is probably little affected by the rate of
blood flow through, for example, arteriovenous
anastomoses.
The capillary loops of the nail fold are among the
most easily visualized in the living body, and they
have been much observed ( 1 75).
total cutaneous blood flow. This quantity has
not been measured with precision, but in a warm
subject it is a considerable fraction of the cardiac
output. Hardy & Soderstrom (112) by a study of deep
and superficial temperature and heat exchange ar-
rived at a blood flow through the skin of 278 ml per
m2 of body surface per min in a nude subject at rest
at an environmental temperature of 35 C. Behnke &
Willmon (29) measuring helium absorption through
the skin under similar conditions arrived at a figure
of 230 ml per m2 per min. During generalized maxi-
mum cutaneous vasodilatation the total blood flow-
is presumably very much greater. Assuming, for
example, a mean thickness of 1.2 mm, and a maxi-
mum flow of 180 ml per 100 ml skin per min, which
has been reported in digits and inferred in the fore-
arm, the flow would be 1200 ml per m'2 per min.
Another estimate, based on skin conductance, is
2000 ml per m2 per min (1 15).
Color of the Skin
The color of the skin due to tissue pigment is re-
vealed by expelling the blood by local pressure. The
additional color, due to circulating pigment, de-
pends on the quantity, quality, and distribution of
this pigment in the skin and subcutaneous vessels.
It is not dependent on the rate of blood flow (139).
Although it often happens that the skin contains more
blood when the flow is fast than when it is slow, the
amount of blood contained in the tissue and the rate
at which blood flows through the tissue by no means
run parallel to each other (54). Thus, the intensity
of the color indicates the amount of pigment present,
and how near the surface are the vessels containing
it. The hue is determined by the proportions of the
various hemoglobin derivatives (oxy-, reduced, met-,
carboxy- etc.) present.
Temperature of the Skin
The temperature of the skin in air depends partly
on the rate of blood flow through it. It depends on
the temperature at which the arterial blood arrives;
that of the blood in the radial artery may be as low
as 21.5 C in a subject who is not feeling unduly
cold (25). It depends also on the rate of blood flow
through both distal and subjacent tissues, on the
activity of nearby muscle (101), on the rate of evap-
oration of sweat, on the temperature, humidity,
motion, and pressure of the surrounding air, and on
the exchange of radiant heat with the environment.
It is clear, therefore, that there can be no simple rela-
tionship between the temperature of the skin and the
rate of blood flow through it. The simplest relation-
ship between the two quantities is probably found in
the digits, examined in still air at a comfortable
temperature. If the circulation is arrested, the fingers
cool until their temperature settles near that of the
air. With the circulation fully opened up, the tem-
perature of the skin of the fingers comes to within
about 1 C and that of the toes to within about 3 C of
the temperature of the mouth. Between these ex-
tremes the relationship between flow and temperature
is by no means linear. For example, in a room at
22 C, the temperature of the fingers may rise to 34 C
when the blood flow is one quarter of the maximum,
and to 36 C when the maximum is attained (55). In
a room at 20.5 C, a skin temperature of 24 C corre-
sponded with a blood flow through the toes of 3 ml of
blood per 100 ml of toe per min; 29 C, with 10 ml,
and 32 C, with more than 30 ml. Even flows of 70 ml
do not cause the temperature to reach 34 C (75).
This does not mean that the higher ranges of blood
flow are always wastefully employed by the body,
for in colder air, or in moving air, the difference in
temperature to which the skin is raised by, and the
difference in heat dissipation at, one quarter of the
maximum flow and the maximum flow may be very
considerable.
By far the greatest variations in skin temperature
are found in the extremities, particularly the hands
and feet in man, and the ears in the rabbit.
The temperature at the surface of the skin in
thoroughly stirred water is essentially that of the
CIRCULATION THROUGH THE SKIN
1329
water. This follows because stirred water can convey
heat to or from the surface at a rate which is very
great indeed compared with the rate at which it can
be conveyed to or from the surface by even the most
profuse flow of blood through the tissues. If an in-
sulating layer is formed, by allowing the water to
stagnate, or by covering the skin with fabric, the
skin becomes a point on the temperature gradient
from the body core to the water. It assumes a tempera-
ture which depends on the ratio of the thermal
insulation between the body core and the skin, and
between the skin and the water. The thermal insula-
tion between the body core and the skin is highly-
dependent on the state of the circulation. The circu-
lation of the blood is the main means of transfer of
heat between the body core and the periphery. The
thermal conductivity of the skin is also highly de-
pendent on the rate of blood flow through it (44).
RESPONSES OF SKIN BLOOD VESSELS TO
PHYSICAL DISTURBANCES
Response of the Circulation Following
Periods of Arrest or Insufficiency
reactive hyperemia. The circulation through the
skin is very frequently arrested by local pressure; it is,
for example, arrested in the sole of the foot while
standing and in the parts of the hand supporting a
heavy object. The skin is better able than most other
tissues to survive fairly prolonged arrest of the circu-
lation without permanent damage. It shows, con-
spicuously, the phenomenon of reactive hyperemia,
by which is meant the bright red flushing (51) and
increase in blood flow above the resting level (139)
when the circulation is released following obstruction.
This is a local change, and clearly depends upon a
local dilatation of the blood vessels responsible for
resistance to flow. The size and duration of the reac-
tive hyperemia are related to the duration of previous
arrest. Although some observations have indicated
that the extra blood flowing during the period of
hyperemia is closely similar to the amount that would
normally have flowed during the period of arrest
[debt and repayment hypothesis (142)] the corre-
spondence is by no means always exact (83), the debt
being frequently underpaid (157). Indeed, it is
possible in the forearm, by gradually releasing the
main vessel, to restore the circulation without any
repayment of debt (35), the blood flow never exceed-
ing the resting level.
Reactive hyperemia is most readily demonstrated
when a limb is warm; it was found by Lewis & Grant
(142) to be much reduced in a cooled part. Thus
following arrest of the circulation for 5 min, Catch-
pole & Jepson (47) found average peak flows of
3.05 ml per 100 ml per min while the hand was
immersed in water at 15 C, 6.8 ml at 20 C, g.8 ml at
25 C, and 19.8 ml at 30 C.
Bier (33, 34) demonstrated in 1897 that reactive
hyperemia is independent of nervous connections
with the central nervous system. While amputating
limbs he first divided the nerves and flesh, leaving
the main artery and vein intact. Occlusion of the
artery was followed, on release, by the usual hypere-
mia. Lewis & Grant (141) observed that skin which
had long been anesthetized, as a result of old standing
lesions of the main nerves, flushes uniformly with the
adjacent skin still possessing normal innervation. In a
chronically denervated and wasted forearm, the peak
blood flow during reactive hyperemia was found by
Eichna & Wilkins (73) to be 26 per cent greater, in
relation to the volume of the part, than in the nor-
mally innervated arm. Similar observations on four
other cases have been made by Duff & Shepherd
(70). The height and the duration of the reactive
hyperemia were found by Freeman (83) to be similar
in the normal and the chronically sympathectomized
hand. The reaction seems, therefore, to be inde-
pendent of all nervous elements which degenerate
following section of peripheral, somatic, and auto-
nomic nerves.
The commonly observed rough correspondence of
debt and repayment has suggested that a chemical
substance may accumulate during circulatory arrest,
and act as a vasodilator. Histamine has been found
in the venous blood following arrest of the circulation
(14), but in the forearm antihistamine substances do
not influence the hyperemia following brief arrest,
though they somewhat reduce that following more
prolonged arrest of the circulation (6g).
There is some evidence that the lowered pressure
in the resistance vessels during circulatory arrest
may lead to a relaxation of their muscular tissue,
perhaps by a local mechanism. Thus Wood et al.
(194) and Patterson (158) have found that reactive
hyperemia in the forearm is reduced if the blood
vessels are packed with blood, thus maintaining a
high transmural pressure during the period of arrest
of the circulation.
Present evidence suggests that reactive hyperemia
depends on local chemical and physical changes,
133° HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
which may contribute in varying proportion accord-
ing to the circumstances.
HYPEREMIA AFTER PROLONGED INSUFFICIENCY OF THE
circulation. A hyperemia, with the blood flow
several times the normal level and lasting for some
weeks, is seen in the feet of some patients after the
relief of chronic arterial obstruction by an arterial
graft (86). The mechanism of this hyperemia and its
relationship, if any, to reactive hyperemia is not
yet known.
Responses oj Skin I 'essels to Changes
in Transmural Pressure
Measurements of the circulation through the finger
have led Burton (43) and co-workers to conclude that
in the resistance vessels there is an unstable equi-
librium between the tension in the wall and the
transmural pressure. If the transmural pressure falls
below the ''critical closing pressure,"' the value of
which depends on the state of activity of sympathetic
vasomotor nerves, the vessels close completely and
arrest the flow of blood. This behavior of the vessels
has been independently confirmed in the finger tip
by Roddie & Shepherd (167).
Calorimetric measurements on the hand (53) and
toes (52) indicate that when the transmural pressure
is progressively increased beyond the normal value
(as by local exposure to subatmospheric pressure) the
resistance vessels at first are passively dilated. At
somewhat higher pressures they react by active con-
traction of their walls and may become narrower
than normal; this is a form of autoregulation of the
skin circulation, the purpose of which may be to
assist the antigravity defenses of the body rather than
to maintain constancy of the skin blood flow.
Effect of Local Temperature on the Skin Circulation
The effects of local temperature on the skin circula-
tion are of great importance, because the skin is
normally exposed to a greater range of temperatures
than any other part of the body except perhaps the
upper end of the alimentary canal. In the latter,
exposure to extremes of temperature is brief, but in
the skin it may be prolonged.
A great many observations have shown that the
circulation through the skin is greatly influenced by
local temperature. The exposure of any part of the
body to a change of temperature probably causes
some alteration to the circulation in all other parts,
40
30
20
IO
A WARM
• COMFORTABLE
O COLD
15 25 35
LOCAL TEMPERATURE C
45
fig. 2. The blood How through the hand measured by venous
occlusion plethysmography, in warm, comfortable, and cold
subjects, and with the hand immersed in water at various local
temperatures. [Data from: /) Abramson el at. (4), 2) Catchpole
& Jepson (47), 3) Killian & Oclassen (132), 4) Kunkel & Stead
(136), 5) Kunkel el at. (137), 6") Peacock (159), 7) Roddie &
Shepherd (166), 8) Speaman (180).]
partly by nervous reflexes and partly by alteration in
temperature of the blood returning from the part to
the heat-regulating center. However, the effects now
to be described are predominantly local ones. When,
for example, the temperature of the water around
one finger or hand is altered, the changes in the
circulation through it are very much greater than
those simultaneously observed in the opposite member
immersed in water at a constant temperature (55,
166).
LOCAL TEMPERATURES IN THE RANGE 1 3 C TO 45 C.
Figure 2 summarizes some representative observa-
tions on the effect of immersion in water at tempera-
tures in the range 15 C to 45 C on the rate of blood
flow through the hand. Between the observations
there are differences of age, sex, and number of
subjects, of present and previous environmental
temperature, in the length of exposure to the local
temperature, and in the details of the venous oc-
clusion plethysmography technique. In general,
however, it may be said that the blood flow through
the hand is at its lowest value at about 15 C, when it
may be as little as 0.3 ml per 100 ml of hand per min
in a cold subject, and 0.9 ml in a warm one. From
CIRCULATION THROUGH THE SKIN 1 33 1
15 C to 29 C there is a modest rise, and from 29 C
to 35 C a faster rise in flow with temperature (180);
35 C to 37 C is the highest temperature to which the
hand is normally warmed by the body's own heat.
At local temperatures in the range 25 C to 35 C, the
level of blood flow is greatly influenced by the heat-
regulating mechanism of the body, and the observed
values are distributed over a wide range. With further
rise in local temperature from 35 C to 45 C, there is
a steep increase in flow, to a maximum of about
35 ml per 100 ml per min (4); Peacock (159) found
in 12 women an average of 36.0, and a range of 30.8
to 41.0; Kunkel & Stead (136) in 18 subjects at 43 C
found an average of 32 and a range of 18.7 to 54.4.
Even at high local temperatures the heat-regulating
mechanism still exerts an influence, for if the subject
is generally warmed the blood flow through the hand
at 44 C increases to about 56 ml per 100 ml per min,
individual observations of over 70 ml per 100 ml per
min having been recorded (166). Most people find
immersion in stirred water hotter than 45 C to be
painful or intolerable.
In the foot, the effect of immersion in water at
various temperatures is very similar to that in the
hand, but the blood flow per unit volume of tissue is
generally about 50 per cent, and per unit of surface
area about 75 per cent (136), of that in the hand.
Allwood & Burry (12) report average blood flows in
four subjects ranging from 0.2 ml per 100 ml per
min at 15 C to 16.5 ml at 44 C, and these seem typi-
cal. Thus in the range 43 C to 45 C flows have been
reported of 14.8 in one subject (132); 16.3 with a
range 11.1 to 20.9 in 33 male feet; and 18.7 with a
range 13.4 to 25.9 in 15 female feet at 43 C, go per
cent of the observations falling between 13 and 20
(136); 15.2 in one subject (4) and 20.5 in 33 subjects
(191). The high blood flow with local heating prob-
ably has a useful protective effect. By conducting
heat away from the tissues it reduces the temperature
below the surface, and the likelihood of thermal
damage. A hand immersed in stirred water at 45 C
becomes painful if the circulation is arrested.
It takes time for the blood flow through an ex-
tremity to settle after a change of the temperature
of the water in which it is immersed. Figure 3 shows
blood flows after immersing the feet in water at
various temperatures. The delay may be partly
explained by the time needed for the internal tissues
to reach a new equilibrium temperature. Once estab-
lished the blood flow through the hand and fingers
is well maintained after immersion for as long as 2
hours at 41 C (8).
40 60
Time (min)
fig. 3. Foot blood flow plotted against time of immersion
during experiments at seven different temperatures. Each
point represents the average blood flow over 5 min. [From
Allwood & Burry (12).]
The local effect of temperature is usually very-
similar to normal in chronically sympathectomized
hands (83), but an anomalous response has been
reported in one case with a reduction in the blood
flow through a sympathectomized hand on raising
the temperature to 41 C (7). The response after
chronic total denervation also appears to be similar
to normal at local temperatures above 18 C (62).
The vessels supposedly respond directly, but some
recent evidence suggests that a local nervous pathway
may assist. Irradiating the proximal half of the fore-
arm with infrared rays causes a vasodilatation which
spreads to the nonirradiated distal half; the spread
is prevented by a cutaneous nerve block at the
junction of the two halves of the forearm, and it is
unaffected by sympathectomy, or by nerve block at
the elbow (59).
LOCAL TEMPERATURE IN THE RANGE O C TO 15 C:
cold vasodilatation. Lewis (140) observed that
following exposure to low temperature the tempera-
ture of the skin rose above its former resting level.
For example, following cooling for 15 min at 7 C,
the temperature of the skin of the index finger rose to
above 28 C, while that of the nonimmersed third
finger remained at 19 C, the subject being in a room
at 17.8 C to 19. 1 C. The temperature of the index
finger was at its maximum 1 1 min after cooling
ended, and was raised for about 50 min.
Further observations showed that the vasodilata-
tion started while the finger was exposed to cold.
Figure 4 shows Lewis's experiment in which the Ri
'332
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
J 4
3 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 78 80 84
o
e
34
32
30
28
M
NUTE,
'
^■J?*
, r*
!==S
L
t
?'
'ti.
"*^"
.«' >
^S« >
-^
R3
.-"s..
26
24
22
20
18
16
14
**
iJT
**•
t'
of* o— -»-o^_-. .
R2
/
/;
L2
I i
If
12
IQ
8
6
4
2
n
■
g
,'
li
b'
/ 1
^V,,.
b
"\
^ ■
J-
»***«
'.. \
/
f
A *
J
_i
+ 4
*»,
■5s
V>.
"
R.T.
19 3°0
fig. 4. Skin temperature measurements with a thermoelectric junction covered by adhesive plaster.
Fingers i?3 and Z.3 in air throughout. Finger In in crushed ice from 11 to 71 min, finger R2 from
16 to 71 min. The curves of temperature rise during immersion are in this case at first discordant,
but become concordant. [From Lewis (140).]
and L-2 fingers were immersed in a mixture of crushed
ice and water, the immersion of R-2 being delayed for
5 min. The 7?3 and L3 fingers remained in air and
served as controls. The temperature of all digits
was measured by thermoelectric junctions, covered
by adhesive plaster; the thermal insulation of the
plaster enabled the junction to assume a temperature
different from that of the ice water with which it
would otherwise have been in direct contact. On
immersing the fingers, the temperature fell at first
abruptly and then more slowly to about 3 C. About
10 min from the start the temperature of both im-
mersed fingers began to rise. The rise was, in other
experiments, prevented by arrest of the circulation
and it clearly indicated vasodilatation. The tem-
perature thereafter fluctuated slowly (the so called
"hunting reaction") and in the case shown the
fluctuations in the two fingers became synchronous,
although in other experiments initially synchronous
fluctuations sometimes became discordant. Following
removal from the ice water the immersed fingers
became warmer than the control fingers.
A similar cold vasodilatation is strongly manifested
in the toes, the lobe of the ear, and the tip of the nose;
it is difficult to detect in the skin of the forearm, calf
of the leg, and on the dorsum of the hand and foot
(99, 140). Strong reactions are seen in the rabbit's
ear (98) and in the foot of the domestic fowl and
CIRCULATION THROUGH THE SKIN
'333
fic. 5. Reaction of arteriovenous
anastomoses in the rabbit's ear to local
cooling, 0, before and b, during cooling.
.4, artery, !', vein; AV, arteriovenous
anastomosis, closed in a and open in b.
[From Grant (98).]
Mkk Ski Oil
LIBRA
Mi M Hi
60 o
MINUTES
60 O
60
fig. 6. The heat loss in cal/100 ml/min from the R and L index fingers to water in the range
0-6 C with intervals of o min (left), 5 min (middle), and 10 min (right) between their insertion
into the calorimeters. The full width of the lower frames is 60 min. The clear areas represent heat
derived from the tissues of the finger in cooling to calorimeter temperature during the first 6 min
of insertion. Pain is represented on a roughly quantitative scale by marks at the top of the frames.
The full height of the frame corresponds to a blood flow of not less (1 10) than 80 ml/100 ml of
finger/min. [From Greenfield el al. (109).]
cluck. The arteriovenous anastomoses in the rabbit's
ear (fig. 5) were directly seen by Grant (98) to dilate
to cold. In the hands and feet, and particularly in
the digits, the intensity of the cold vasodilatation was
found by Grant & Bland (99) to parallel closely the
density of the arteriovenous anastomoses, and it seems
likely that the dilatation of the latter is mainly re-
sponsible for the increased blood flow.
Subsequent calorimetric observations (16, 105)
have shown that for the first 5 to 10 min of immersion
in ice cold water there is a constriction of the vessels
with almost complete arrest of blood flow (fig. 6).
At this time there is a considerable degree of pain.
The vessels then rapidly dilate, the pain goes and the
finger feels warm and comfortable. In a warm sub-
ject, the blood flow may rise to a value which is
probably as high as is attained by any other type of
vasodilatation (105).
With continued immersion, the dilatation is ir-
regularly interrupted by periods of constriction
lasting a few minutes. These may be abrupt in onset
and termination, and may cause almost complete
arrest of blood flow (108). The pattern and timing of
these periods of constriction differ in different digits
simultaneously observed, and appears to be locally
determined (109). During continued immersion the
general level of the peaks of vasodilatation often
tends to decline, but if the subject is kept warm,
alternation of dilatation and constriction may con-
tinue for several hours (37). On removal of a finger
from the cold water the dilatation persists, and for
about half an hour the finger may be warmer than
its nonimmerscd neighbors (192). The vasodilator
response is conspicuous on immersion at temperatures
near o C, but it is detectable at temperatures as high
as 12 C or 15 C.
The vasodilator response is present after inter-
ruption of the sympathetic outflow from the central
nervous system by local anesthetic block or by chronic
section. It is, however, influenced by sympathetic
activity. Among chilled individuals there are con-
siderable differences in the response, but the vaso-
!334
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
15 C
2000 •
L
IOOO -
O
29 C
2000
IOOO
mi
42 C
Hi in
60 O
INUTES 6° °
60
fig. ~. Heat loss in cal/ioo ml of finger/min to water in the range o-6 C from the L (anesthetic)
and R (normal) 5th fingers, the observations being made between the 30th and 40th days after di-
vision of the L ulnar nerve. In the denervated finger, cold vasodilatation is extremely small or absent
after preliminary immersion in water at 1 5 C, small but definite after 2Q C, and considerable after
42 C. [From Greenfield et al. (109).]
dilatation is delayed, sometimes for as much as 90
min from the time of immersion, and it is reduced
in size sometimes to less than one-tenth (127). The
response continues in the early days after interruption
of the somatic nerves, but becomes difficult to elicit
later when the separated distal parts of the nerves
have degenerated. This led Lewis (140) to conclude
that the response depends on a local axon reflex from
cutaneous receptors to the blood vessels. If, however,
the chronically denervated limb is warmed for some
time before the digits are immersed in cold water
(109), a reduced (to 20-90 Tr) but definite vasodilator
response is seen (fig. 7). It is always difficult to be
certain that denervation is complete, but the observa-
tions probably indicate that the axon reflex pathway-
is not essential for the response. This view is strength-
ened by the finding of a vasodilator response in a
finger tip locally injected with anesthetic solution
(110). Other vascular responses which appear to be
depressed in the chronically denervated limb are
improved if the limb is first warmed for an hour or
two.
A chemical stimulus to cold vasodilatation has not
been identified. Acetylcholine and histamine in-
jected intra-arterially or introduced by electrophoresis
during the first few minutes of immersion of a finger
in ice water do not provoke an earlier vasodilatation,
but it is possible that at this time of intense vasocon-
striction they do not reach the blood vessels. Cold
vasodilatation, however, is not reduced by atropine
nor by antihistamine (188). The mechanism of the
response remains obscure.
The effect of the dilatation is to raise the tempera-
ture of the exposed extremities at the expense of a
considerable loss of heat from the bodv. Even while
immersed in stirred water near the freezing point,
the average internal temperature of a finger may be
raised to as much as 30 C, and in the central parts
must presumably be only slightly below the tempera-
ture of the body core (107). In a warm person the
heat loss per minute from a whole hand and from the
distal half of a foot may be 800 and 407 cal per min,
respectively (104). These figures are similar to total
resting heat production, and the loss of heat from one
hand can cause a fall in esophageal temperature
of 0.6 C in 9 min (106). When plenty of heat is
available, the reaction keeps fingers exposed to
reasonable cold sufficiently warm to preserve move-
ment and sensation. Dwellers in cold climates nor-
mally wear clothing which provides their body with
a warm microclimate, and are able to afford some
heat loss. When there is a need to conserve heat,
the reaction is greatly diminished and the fingers are
only slightly warmed by it. For example, Australian
aborigines are able, by restricting the peripheral
circulation, to retain sufficient heat to sleep naked
through the night in a temperature which may fall
to o C (177).
Several observations suggest that the extremities
can become acclimatized to cold (44). Local pain,
and the reflex increases in arterial pressure and pulse
rate are reduced after repeated immersion in water
at 4 C (92). Exposure of the fingers to severe cold
causes less numbness in persons habitually exposed
than in others. Such observations suggest that there
may be a local adaptation of the circulation in the
exposed parts but the evidence for this is not strong.
Repeated exposure of the hands to cold, as in Nor-
wegian and Lapp fishermen, leads to a more rapid
onset of cold vasodilatation (135) but to no increase
CIRCULATION THROUGH THE SKIN'
J335
in the level of blood flow at the height of the vaso-
dilatation (113, 135)- In such subjects, kept warm to
release sympathetic vasoconstrictor tone, the reactive
hyperemic blood flow with the hands at 40 C, and
the resting blood flow with the hands at 40 C, 20 C,
and 10 C is no different from that in normal controls
(135). The improved circulation in the hands of cold-
habituated persons reported by other observers may
depend more on general adaptation of the circula-
tion than on a local adaptation in the periphery.
PROLONGED EXPOSURE TO COLD: TRENCH FOOT AND
immersion foot. Prolonged local cooling to tempera-
tures above the freezing point is capable of causing
serious injury. Although in many recorded cases the
parts have been wet as well as cold, the main factor
is the cooling of the extremities in a chilled subject
(183). The feet are particularly liable to injury, and
most cases have been seen after exposure for many
hours or days in war time.
The four stages of the condition have been well
described by Ungley (183).
/) During exposure, the limb is numb, power is
reduced, and movement is clumsy. Pain is unusual.
Swelling is common, the limb often looks bright red,
and there may be periods of warmth, presumably due
to cold vasodilatation, but the chilling of the subject
reduces this to small proportions.
2) Immediately after rescue and return to warmth
and shelter there is a prehyperemic stage, which may
last for 2 to 5 hours. The limb is cold and either
pale with cyanotic patches or cyanosed. The arterial
pulsations cannot be felt. There is a partial or com-
plete '"stocking" sensory loss.
3) A hyperemic stage follows, the part becoming
red, swollen, painful, and sometimes blistered. When
the arterial pulses return, they are very strong, and
the temperature of the skin is as high as that of the
axilla or groin. The hyperemia is judged clinically
to be at least as great as that following sympathec-
tomy, and it is often much more persistent, lasting as
long as 14 weeks. There is partial anesthesia, and
vasomotor and sudomotor paralysis, indicating nerve
damage. In addition there is direct vascular damage.
4) In mild cases there is a return from the hy-
peremic state to normal, but in severe cases a post-
hyperemic state follows. The circulation decreases
greatly, and although vasomotor reflexes to heating
and cooling the rest of the body return, the response is
slow and incomplete. There is often an increased
sensitivity to cold, reduction of the blood flow for
many hours sometimes following immersion in water
at a temperature as high as 24 C. Once constricted
or dilated, the vessels tend to remain so for a long
time. The cause of this altered vascular reactivity is
not known.
Although the vascular damage may not be an es-
sential feature (183) it may sometimes be severe (84)
with dilatation and engorgement of vessels, rupture,
and thrombus formation. Exposure for many days
to water at as high a temperature as 21 C has been
sufficient to cause the feet to become swollen, hy-
peremic, and painful (186). Of nine volunteers
living for 5 days in a covered raft in arctic waters,
seven developed hyperemic swollen feet, a condition
which in two cases persisted for several weeks; the
lowest toe temperature recorded during exposure
was 1 1 C, and the temperatures were usually 13 C to
15 C. (44). The vascular changes during exposure
have not been followed in man. It is presumed that
cold vasodilatation subsides after a time, perhaps
because the subject becomes generally chilled, and
that there is an extremely low blood flow for a long
time.
exposure to severe cold, frostbite. Exposure to a
temperature sufficiently low to cause freezing of the
tissues may cause frostbite, which is commonly
followed by gangrene and loss of tissues. During
exposure there is arterial spasm and capillary stasis.
On rewarming, there is an intense hyperemia, and
the capillary permeability is greatly increased, lead-
ing to edema and to blockage of the vessels with blood
cells. There is frequently thrombosis in some vessels
and this may lead to a permanent reduction in blood
flow (149).
The freezing point of living fingers is about —0.6 C,
but supercooling is usual so that fingers immersed in
brine at —1.9 C, the freezing point of sea water, do
not always freeze (128). Supercooling to —1.9 C
does not cause the tenderness, redness, and warmth
which persist for several days after freezing at that
temperature. The damage on freezing the tissues is
probably caused partly by the formation of ice
crystals, and partly by the concentration of the dis-
solved substances in the liquid water that remains
(i5°)-
In dogs, after immersion of the hind leg in an
alcohol and dry ice mixture at —25 C for 30 min,
or — 4 C to — 8 C for 2 1 o min, the blood flow, on
rewarming the limb, is increased for several hours
to several times the level in the contralateral control
limb (125) and this vasodilatation appears to depend
on the integrity of sympathetic outflow (126).
'336
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Reactions to Injury
mechanical injury. The reactions to mechanical
injury were very completely studied by Lewis ( 1 39).
The while reaction. When warm skin is lightly stroked
with a blunt point there is a temporary blanching as
blood is expressed from and then returns to the super-
ficial vessels. About 1 5 sec later, the line of the stroke
becomes pale again, the pallor reaching its maximum
about 30 sec after the stroke, and fading in about 3
to 5 min. The white line is sharply localized to the
area stroked. Its development is unaffected by the
temporary arrest of the circulation, and it was there-
fore taken by Lewis to indicate active contraction of
the vessels responsible for the color of the skin, and
not merely deprivation of these vessels by contrac-
tion of the arterioles that supply them. The vessels
responsible for color are able to sustain their con-
traction against a distending pressure of 80 to 100
mm Hg produced by venous congestion.
The triple response. When the stroke is much or very
much firmer, the white reaction is replaced by a
different response which, when fully developed, has
three components, the red line, flare, and wheal, a)
The most constant component is a sharply de-
marcated red line which develops along the line of
the stroke with a latency of 3 to 15 sec, and the in-
tensity and duration of which increase with the
strength of the stimulus. Like the white reaction, the
red line develops even when the circulation is tem-
porarily arrested. It was considered by Lewis (139)
to indicate active dilatation of the vessels responsible
for the color of the skin, b) In susceptible skins, and
with strong or repeated stimuli, an irregular red
flare develops about 15 to 30 sec after the red line,
and gradually extends for 2 to 3 cm on each side of
the line of the stroke. The flare remains a bright
scarlet color, unlike the red line which becomes
progressively dusky. As the flare fades, it becomes
mottled. A white reaction can be developed across
the flare by light stroking, but not across the red
line. The flare was considered by Lewis (139) to
indicate arteriolar dilatation, c) In sensitive skins,
or in others following a strong stimulus such as the
lash of a whip, a raised wheal usually begins to appear
along the line of stroke in 1 to 3 min, reaching full
development in 3 to 5 min. It overlies the red line
and the line becomes pale, presumably because of
the pressure exerted by the transuding fluid upon
the minute vessels.
The triple response is unaffected when the sensory
nerves are freshly interrupted by section or local
anesthesia. The red line and the wheal continue in
chronically denervated skin, but the flare is lost after
about the sixth or seventh day when the sensory
nerves degenerate. This led Lewis (139) to conclude
that the red line and wheal are independent of nerves,
but that the flare depends on a local axon reflex (49).
The nerve impulse arises in a receptor in the skin
and, after ascending a sensory nerve for some distance,
returns antidromically along a branch to arrive at an
arteriole and cause it to dilate.
The triple response is the standard reaction of the
skin to a great variety of injurious stimuli. The re-
sponse to mechanical trauma can be exactly repro-
duced by pricking histamine into the skin. Further,
if trauma or a histamine prick is applied while the
circulation is arrested the development of the flare
is delayed until the circulation is released. This and
other evidence led Lewis (139) to postulate that the
flare depends on the activation of the skin receptors
by an H-substance, which may be histamine, rather
than directly by the mechanical trauma.
ultraviolet light. Irradiation with ultraviolet
light, which penetrates to a very small depth in the
fig. 8. Two experiments. Hand blood How in ml/ 100 ml/
min. Solid circles: injected arm; open circles: control arm. The
heights of the vertical columns indicate the percentage satura-
tion of venous blood with oxygen. Intra-arterial injection of 5
ml of nitrous oxide is indicated by the black rectangle starting
at o min. [From Duff el at. (67).]
CIRCULATION THROUGH THE SKIN
'337
skin, causes a delayed erythema sharply confined to
the exposed area. It is probable that a chemical
agent is concerned, but that this is not histamine
(•56).
arterial gas embolism. After injection into the
brachial artery of 1 to 10 ml of gas there is usually
an immediate reduction in blood flow through the
hand lasting for a few minutes, followed by a pro-
longed increase, to several times the normal resting
rate, which does not entirely subside for many hours
(66). The increase in the oxygen saturation of the
venous blood parallels the increase in flow (fig. 8).
All of the several gases tested are effective, provided
they are given as bubbles and not in solution. The
response is present in both sympathectomized and
chronically denervated limbs, and is unaltered in the
presence of amounts of antihistamine substances
which prevent the action of histamine (67). The
vessels of muscle as well as of skin are affected. The
mechanism is not understood, but the reaction ap-
pears to result from some trauma to the tissues caused
by the bubbles, and a peripheral arterial conducting
mechanism may be involved (120).
nervous control of skin blood vessels
Vasomotor Nerves
vasoconstrictor sympathetic nerves. Claude Ber-
nard (30, 31) showed that division of the cervical
sympathetic chain in the rabbit caused the ear on
the same side to become flushed and warm. Stimula-
tion of the trunk had the reverse effect (32). Similar
observations were made by Brown-Sequard (39).
These observations indicate that the sympathetic
nerves contain vasoconstrictor fibers, and that under
ordinary conditions the activity in these fibers keeps
the vessels in a partially constricted state. The warm-
ing of, and increased circulation through, the human
feet and hands following lumbar and thoracic sym-
pathectomy was first described by Adson & Brown
(5, 6) and this established that these areas were
similarly under sympathetic vasoconstrictor control.
Walker et al. (184) made quantitative measurements
of the effects of sympathectomy in patients with ap-
parently normal blood vessels in whom the operation
had been carried out as a treatment for excessive
sweating. The blood flow in milliliters per 100 ml of
hand per min was increased from an average value
of 5.2 before the operation to peak values in the range
22.7 to 59.2 after the operation, and in the feet was
increased from an average value of 2.1 before the
operation to peak values in the range 20.8 to 28.0
after the operation. The averaged results on five
hands and six feet are shown in figure g.
In vasoconstrictor sympathetic nerves low rates of
discharge have a powerful effect; in the cat's paw,
stimulation at the rate of 1 per sec increases the re-
sistance to flow about 10 times, and stimulation at
10 per sec increases it about 100 times (48).
vasodilator sympathetic nerves. The evidence for
vasodilator nerves to the skin rests at present on
experiments of the type employed by Grant &
Holling (100), that is to say, on the simultaneous
observation under suitable conditions of reflex
stimulation of a greater blood flow in normally
innervated skin than in a corresponding area of skin
acutely deprived of its vasomotor innervation. Since
both areas are perfused with blood of identical com-
position and at the same pressure, the resistance
vessels may be presumed to be more widely dilated
in the innervated skin, and this dilatation to result
from nervous activity. It must be noted that chroni-
cally denervated skin is not a satisfactory tissue for
this comparison, because of the decline in blood flow
due to contraction of the blood vessels.
Grant & Holling (100) found that blocking the
cutaneous nerves, which convey sympathetic fibers,
to parts of the skin of the forearm not only failed to
cause a flushing and rise of temperature but pre-
vented the vasodilatation and also the sweating
normally seen in the forearm during body warming.
Evidently, in the forearm vasomotor nerves actively
bring about vasodilatation. Whether they do so by a
direct action on the vessels or as a consequence of
increased sweat gland activity, or by both means
was, and is, uncertain. If the fibers are called "vaso-
dilator," it must be remembered that vasodilatation
may be only a consequential and not a direct effect.
It is, of course, important for heat to be brought
to the skin if sweat is to be evaporated. At least 60 ml
of blood are required to transport from the body
core the heat required to evaporate 1 g of sweat when
the skin is 10° below the temperature of the core. At
least 600 ml of blood are required when the skin is
1 C below the temperature of the core. Sweating can
therefore be effective only when accompanied by
vasodilatation.
I338 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
fig. 9. Abscissae: time in days, the vertical broken line indicating the day on which sympathectomy
was performed. Ordinatcs: blood flow in ml/100 ml tissue/min. The averaged results of experiments
on five hands and six feet show the effect of sympathectomy on the blood flow in the hand and
foot, and on the skin temperatures of the fingers and toes. The denervations were performed to
prevent excessive sweating; hence the responses are those of normal human blood vessels. Note the
transient increase in blood flow after operation and the subsequent decline as intrinsic tone returns
to the vessels. [From Barcroft (18).]
VASODILATATION CAUSED BY ANTIDROMIC STIMULATION
OF DORSAL ROOT SENSORY NERVES: THE AXON REFLEX
pathway. Stimulation of the peripheral end of a
cut sensory nerve often causes vasodilatation in the
area of skin supplied by the nerve, and Bayliss (24)
showed that this was due to impulses traveling toward
the periphery along neurons of the dorsal root system.
The vasodilator effect of such antidromic nervous
impulses is considerable, but evidence has not been
forthcoming for the use of this route from the central
nervous system to the periphery for any reflex ad-
justments of the circulation. It seems probable that
the artificially provoked antidromic impulses travel
to the vascular nerve endings of the peripheral axon
reflex pathway. The usual source of impulses ar-
riving here is from nearby sensory- nerve endings,
probably subserving pain sensation. The transmitter
substance for the vasodilatation produced in the
chronically sympathectomized ear of the rabbit by
stimulation of the sensory great auricular nerve is
neither histamine nor acetylcholine (123) but is
probably adenosine triphosphate (122), the presence
of which has been demonstrated in nerve roots (121).
Innervation of the Blood Vessels of the Skin
in Different Areas
The problem is to define, for different areas of
skin, the existence and range of action of vasocon-
strictor and vasodilator fibers. This has been most
fully investigated in the human limbs, particularly
in the upper limbs, and the innervation here will be
first described.
the human hand and fingers. In the cold subject,
the blood flow through the hand is very small, often
less than 1 ml per 100 ml per min. When the subject
is warmed, either in a hot cabinet (143) or by im-
mersing the legs in stirred water at 44 C (89), the
blood flow increases to about 30 ml per too ml per
min, this being part of the general response by which
the body attempts to lose heat. Although abundant
cholinesterase is found in the arteriovenous anastomo-
ses of the fingers, and although Lewis & Pickering
(143) obtained evidence in cases of Raynaud's dis-
ease for the activity of vasodilator nerves, the increase
in blood flow in normal persons can be entirely ac-
counted for by a reduction in the activity of vaso-
constrictor nerves; several careful investigations have
CIRCULATION THROUGH THE SKIN
l339
failed to detect any contributions from vasodilator
nerves. Thus Pickering (161) during body heating
found equal rates of heat elimination from the two
hands, the ulnar nerve conveying part of the sym-
pathetic supply to one hand having been blocked.
In a more sensitive test, Arnott & Macfie (15) meas-
ured the heat elimination from the fifth fingers during
body heating. The sympathetic supply to one was
entirely interrupted by ulnar nerve block, but the
rates of heat elimination were equal. Warren et al.
(185) found that paravertebral block of the sym-
pathetic outflow increased rather than decreased the
blood flow through the hand of a heated subject.
Gaskell (87) compared the rates of blood flow through
the two hands by venous occlusion plethysmog-
raphy, which is probably the most accurate method.
He heated the subject, and then blocked on one
side near the elbow the radial, ulnar, and median
nerves which probably convey the great majority
of sympathetic fibers to the hand. This caused no
alteration in the rate of blood flow. Roddie et al.
(172) found no difference between the rates of blood
flow through the two hands in similar experiments
in which the nerve block on one side preceded the
body heating; this eliminated the possibility that in
GaskelTs experiments (87) a stable chemical vaso-
dilator substance was released by sympathetic nerves
before they were blocked. The most probable ex-
planation of these observations is that in the ade-
quately heated subject there is a complete cessation
of activity in the vasoconstrictor nerves to the hand,
and no activity in vasodilator nerves. The less prob-
able alternative is that in the hand of a heated subject
there is a balance of vasoconstrictor and vasodilator
activity, and that the vessels are unaffected when both
activities are abolished by nerve block.
Although there is no evidence for the participation
of vasodilator nerves in the response to body heating
in normal persons, a vasodilatation dependent on an
intact sympathetic nerve supply may accompany
the sweating in the hand which is provoked by emo-
tional stress. The direct and immediate effect of
emotion is to reduce the blood flow through the hand
(2) by increasing the activity in vasoconstrictor nerves.
If, however, the emotional stress is continued, as by
mental arithmetic, the vasodilatation consequent on
sweating may outweigh the constriction even in
normal persons, and in persons suffering from ex-
cessive sweating the vasodilatation may be very
large indeed (10), the flow rising from 5 to over
30 ml per 100 ml per min (fig. 10).
B.P. (mm
118/76
Hg)
Arithmetic
120/76
134/82
I
ml /mm)
1
\
0
0
1 / *4
_1
1
S
0
T3
O
O
-co
L
\ Hand
Forearm
30
20 --5.
10
0 Minutes 10
fig. 10. Results showing the marked increase in hand blood
flow (•) during mental arithmetic in a hyperhidrotic subject.
There was little change in the forearm blood flow (O) or
arterial blood pressure. Plethysmograph temperature 36 C.
[From All wood et al. (10).]
the human forearm. Both vasoconstrictor nerves,
and nerves which directly or indirectly cause vaso-
dilatation (vasodilator nerves) regulate the circula-
tion through the skin of the forearm. Of these, the
vasodilator nerves, first described by Grant & Holling
(100), are by far the more important. The role of the
two sets of nerves is clearly displayed during the
response of total forearm blood flow to general body
heating.
This response has recently been shown, by several
methods, to be confined to the skin, the muscle cir-
culation remaining unchanged. Thus Edholm et al.
(71) found that intensive iontophoresis of adrenaline,
sufficient to arrest the circulation in the skin of the
forearm, prevented the normal increase in total
forearm blood flow with body heating. Barcroft et al.
(20) found that when a person is heated the total
blood flow through the calf of the leg, measured
plethysmographically, increases, but that through
the muscle, measured by a heated thermocouple
method, does not; the increase must have been in the
skin. The rate of clearance of radioactive sodium
from muscle is unchanged or reduced (154). Roddie
et al. (170) found that during general body heating
there was a gradual increase, from an initial 40 to
72 per cent to a final 85 to 99 per cent, in the oxygen
saturation of the blood in the superficial veins of the
forearm predominantly draining skin, but no change
in the deep veins mainly draining muscle (fig. 11);
the changes in the superficial blood closely paralleled
the increase in total forearm blood flow in the op-
posite arm.
i 34o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 1 1 . The effect of body heating and of change of posture
on the oxygen saturation of deep and superficial forearm venous
blood. The black rectangle indicates the period of general body-
heating. The intervals between the dotted lines represent the
periods during which the subject's legs were passively raised.
O, Left forearm blood How, ■ oxygen saturation of superficial
venous blood in right forearm; •, oxygen saturation of deep
venous blood in right forearm. [From Roddie el al. (170).]
On warming a rather cold subject, the blood flow
through the forearm, and hence through the skin of
the forearm, increases in two steps (171). The first
increase, from about 2 to about 4 ml per 100 ml per
min, is of the same order of size as the increase which
follows block of the superficial nerves to the forearm,
and may be assumed to be due to withdrawal of
vasoconstrictor activity. The subject is by now com-
fortably warm. If body heating is continued, there
is a further increase in forearm flow from 4 to 10 to
1 5 ml per 1 00 ml per min. This is accompanied by
sweating. The total forearm blood flow and the oxygen
saturation of the blood from superficial veins now far
exceed the levels seen after cutaneous nerve block.
Blocking the cutaneous nerves at this stage causes the
forearm blood flow to fall to about the level seen in
an unheated subject (72). Without block, the blood
flow through the skin of the forearm is now very large
indeed; Edholm el al. (71 ) give a figure of 165 ml per
100 ml of forearm skin per min, but do not claim
that this is more than an approximate figure. The
sweating can be prevented and the vasodilatation
delayed by the injection of atropine into the brachial
artery before the heating starts. Fox & Hilton (81)
have found that during sweating there is a fivefold
increase in the bradykinin-like activity in the per-
fusate of the subcutaneous tissue of the forearm, and
that a bradykinin-formintj enzyme is present in
sweat. Bradykinin is a very powerful vasodilator
substance. Injected into the human brachial artery
it is more powerful, per molecule, than acetylcholine
or histamine, or indeed any other known substance
(80). It is suggested (81) that the vasodilatation in
the skin of the human forearm is produced in the main
by bradykinin resulting from sweat gland activity,
itself provoked by cholinergic sympathetic nerves.
In the region of the wrist there must be a transi-
tion from the vasoconstrictor control of the hand
vessels to the predominantly vasodilator control of
the forearm vessels. The site, sharpness, and varia-
bility or constancy of the demarcation have not been
defined.
other areas of the human body. The innervation
of the foot has been less completely examined than
that of the hand, but as far as is known, the pattern
is similar. Elsewhere, our knowledge is fragmentary
and incomplete. In the upper arm, calf of leg, and
thigh (36) the pattern of vasomotor innervation is
like that of the forearm, there being a weak vasocon-
strictor innervation operating when the subject is
cold, and a more powerful vasodilator innervation,
probably associated with sweat gland control, which
operates when the subject is hot. Vasodilator control
is also dominant in the forehead and chin, and cutane-
ous vasodilatation accompanies sweating in these
areas; vasoconstrictor control is important in the
glabrous portion of the lips, and in the skin of the
nose (78, 79).
the skin of animals. Apart from the paws (138), the
skin of the limbs of cats and dogs lacks eccrine sweat
glands. In these species, heat vasodilatation results
from the reduction in the activity of vasoconstrictor
nerves, and there is no evidence for vasodilator
nerves (77, 102).
late effects of sympathetic denervation. Goltz
& Freusberg (97) noted that the freshly denervated
leg of the dog was warmer than its fellow, but that
the difference does not persist. It has since been
shown by several groups of workers (18, 23, 101, 145,
184, 190) that the blood flow in a limb several weeks
after sympathectomy differs little from the preopera-
tive value. The blood flow in both hands and feet
reaches its highest value about the second day after
the operation, and then declines steeply during the
next few days (fig. 9). In the hand, the decline in flow-
is equally rapid whether a preganglionic section or a
postganglionic section with ganglionectomy has been
performed (184). In the forearm (63) the maximum
CIRCULATION THROUGH THE SKIN
1341
flow is seen on the day of the operation, and the
decline is faster than in the hand; the extent to which
the vessels of muscle and skin, respectively, contribute
to these changes has not been defined.
The cause of the change in the vessels which leads
to the return of blood flow to near the normal level is
not known. The denervated vessels develop an in-
creased sensitivity to adrenaline and other vasoac-
tive agents (46, 74) and this develops at a rate which
closely parallels the decline in blood flow (22). In-
creased sensitivity of chronically sympathectomized
vessels has been demonstrated in the finger to ad-
renaline injected intravenously (178), and in the
hand to both adrenaline and noradrenaline injected
intra-arterially (fig. 14). Whether the return of tone
is due to an increased sensitivity of the vessels to
unknown circulating pressor substances, or to an
intrinsic change in the muscle of the vessel wall, or
to an effect of surviving accessory sympathetic fibers
is not decided.
LATE EFFECTS OF TOTAL DENERVATION. "While loss
of sympathetic supply causes the corresponding fingers
to be in general warmer than they otherwise would
be, loss of all nerve supply causes the corresponding
fingers to be in general colder than they otherwise
would be. And, since with combined loss of both
motor and sympathetic supply the digits remain
warm, it seems that sensory nerve loss must be an
important factor in determining the persistent cold-
ness in cases of mixed nerve lesions" (Lewis &
Pickering, 144). The extent to which the coldness of
denervated fingers depends on the loss of sensory as
opposed to sympathetic innervation has, however,
been questioned (62), and limbs normal except for
muscular paralysis are colder than normal (144). The
most conspicuous abnormality in the behavior of the
circulation in denervated digits is the great reduction,
under normal circumstances, in the vasodilator
response to cold.
Reflex Control of Blood Vessels of the Skin
The blood flow through the digits can be varied
through a very wide range by the activity of the
sympathetic vasoconstrictor nerves. At the upper and
lower extremes of the range the blood flow is normally
fairly steady from minute to minute. At intermediate
levels, such as are normally found in comfortably
warm subjects, the flow usually fluctuates, rising and
falling by 20 per cent or more several times a minute.
The fluctuations are abolished bv division of the
sympathetic nerves, occur simultaneously in the
digits of all limbs, and are often associated with
simultaneous changes in heart rate (41, 42, 45). The
frequency of the constrictions is greater when the
flow is near the lower than when it is near the upper
end of its range, and the size and pattern of the varia-
tions van,- considerably in different individuals. The
functional significance of the fluctuations is not
known; their occurrence makes desirable the use of
repeated rather than single observations in estimates
of the skin circulation in the extremities. They are
not found in the skin of the forehead (117).
Very little is known about the reflex responses of
skin other than that in the extremities, but such evi-
dence as is available indicates that the responses,
if present, are comparatively small (85, 116).
body temperature regulation. The skin is a main
route for loss of heat from the body, and by far the
most important route capable of adjustment by the
temperature-regulating center. The heat lost from the
surface, whether by conduction, convection, radiation,
or the evaporation of sweat must be transported to
the skin, and because of the low thermal conductivity
of body tissue, the transport is mainly in the circu-
lating blood.
Gibbon & Landis (8g) found that if one arm was
immersed in water at 42.5 C to 44.6 C the temperature
of the fingers of the opposite hand started to rise in 5
to 10 min, and reached 32 C in g to 16 min. If, how-
ever, (fig. 12) under similar conditions, the circula-
tion in the immersed arm was arrested (with brief
releases) by a pneumatic cuff for the first 35 min of
immersion, the rise in temperature of the fingers of
the opposite hand was delayed until 7 to 10 min
after the final release of the cuff (that is 42-45 min
from the start of immersion), and the fingers reached
32 C 11 to 16 min after the release. The response in
the fingers evidently depends on the return to the
body of hot blood from the immersed part, rather
than on the stimulation of peripheral receptor organs.
This conclusion is confirmed by the finding that
rapid intravenous infusions of hot saline are able to
provoke vasodilatation in the hand by a mechanism
independent of any surface heating ( 1 7g). The central
receptor mechanism is sensitive to the addition of as
little as 1 to 2 Calories of heat to the body, or to an
amount of heating sufficient to raise the sublingual
temperature by 0.15 C (88).
The temperature of surface receptors is, however,
of some importance in the reflex regulation of the skin
circulation. Kerslake & Cooper (129) found that
'34-1
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 12. The right arm, with the
circulation off, was immersed in hot
water at 42.8-44.3 C at the 38th min.
There was no dilatation in the left
second (Z.2 ) or left fourth (Z.4) finger
until the circulation was released at the
76th min. [From Gibbon & Landis
(89)-]
18
16
10
CIRCULATION OFF
RT ARM IN BATH 428- 443 C
30
50
MINUTES
70
90
heating the trunk or both legs with radiant heat
caused a substantial (2-fold or 3-fold) increase in blood
flow through the hand, with a latency of only 10 to
15 sec; this was too short a time for a mechanism
depending on the return of hot blood to the heat-
regulating center. Furthermore, on heating the legs
the response was unaffected when cuffs around the
thighs were inflated to 200 mm Hg to arrest the circu-
lation (fig. 13). In this case the vasodilatation there-
fore appears to depend on afferent information
conveyed by nerves from the heated skin. The response
to heating the legs disappears after bilateral lumbar
sympathectomy, and in persons with unilateral
sympathectomy it is obtained on heating the normal
but not the sympathectomized leg (57). It is not yet
certain whether the afferent nerves traverse the
sympathetic ganglia, or whether sympathectomy
modifies the response by altering the conditions at
somatic nerve thermoreceptors. Stimulating the
intact lumbar sympathetic chain or its cut central
end causes vasoconstriction in the hand, but it is not
known whether the stimulated afferent fibers are
from the skin or from the viscera (58).
As mentioned earlier, the effector side of the tem-
perature regulation reflex is mediated in the hand by
adjusting vasoconstrictor activity, and in the forearm
mainly by adjusting vasodilator activity. The reduc-
tion in vasoconstrictor activity is not simultaneous
in all areas; the individual fingers often dilate asyn-
chronously, and the foot often dilates many minutes
after and less completely than the hand (162). The
vessels in warm skin dilate sooner than those in corre-
sponding areas of cold skin.
Exposure of part of the body to cold causes changes
in the circulation in other areas by two mechanisms.
There is a rapid transient reflex vasoconstriction, due
to stimulation of afferent nerves (160), and a longer
lasting vasoconstriction due to cooled blood returning
to the heat-regulating center (160, 181). The cooling
effect on the temperature of the body core of exposure
of limbs to cold is usually restricted by local vasocon-
striction. This limits the quantity of cooled blood
returning to the core. Furthermore the efficient ar-
rangements for exchange of heat between arteries and
veins in the limbs (25, 26) reduce the cooling effect
of the blood. On the other hand, the temperature of
the blood that does return may, at the start of its
journey, be 30 C or more below that of the core,
while from a heated region the blood can hardly
start its return at a temperature more than 7 to 8°
above that of the core. Further, cold may sometimes
be sufficiently severe to cause cold vasodilatation,
CIRCULATION THROUGH THE SKIN
'343
•
OCIRCULATION OCCLUDED- NO HEAT
©HEAT-CIRCULATION OCCLUDED
• HEAT- NO OCCLUSION
60 ^.---....e 120
SECONDS
fig. 13. Changes in hand blood flow during heating of the
front of the legs. Each curve is the mean of three runs. [From
Kerslake & Cooper (129).]
and while this is in progress the heat loss may be
several times greater than the heat gain during
exposure to heat.
When a nude person is chilled, the temperature of
the skin and the loss of heat from it fall to much lower
levels over the limbs, and particularly the hands and
feet, than they do over the head, neck, and trunk.
This difference is partly accounted for by the more
vigorous vasoconstriction in the skin of the limbs, and
partly by the more favorable opportunities in the
limbs for economizing heat loss by exchange of heat
between arteries and veins. Thus there are poor
defenses against heat loss from the head and trunk,
and these regions are particularly dependent on
clothing for insulation. The loss of heat from the
uncovered head may be very large (85), amounting
to about one-half the resting heat production of the
body when the ambient temperature is — 4 C.
emotion. In the middle ranges of flow, the circulation
through the hand is often very sensitive to slight emo-
tional stimuli; it may suffer a considerable transient
reduction, lasting a minute or more, when a person
enters the room, if a remark is made or a question in
mental arithmetic posed (2), or if there is an unex-
pected noise. For this reason it is important to reduce
disturbance to a minimum in experiments in which
the blood flow to the hand (or foot) is measured. A
more prolonged emotional stimulus, such as is pro-
vided by mental arithmetic for 10 min under trying
conditions, causes in persons with hyperhidrosis and
in some normal persons an increase in the blood flow
through the hand; this is associated with emotional
sweating (10). All these responses are prevented by
division of the sympathetic nervous outflow. It is,
however, possible that during more severe emotional
upsets, the circulation through the skin may be
affected by adrenal gland activity. In the forearm,
the circulation through the skin is little affected by
the emotional stress of mental arithmetic (76).
fainting. In posthemorrhagic fainting the blood flow
through the hand was found by Barcroft & Edholm
(21) to be more reduced than would be expected
from the fall in arterial blood pressure. This indicated
vasoconstriction in the hand. Other observations have
indicated little change or vasodilatation. It seems very
probable that the response is a variable one, de-
pending perhaps on the degree of associated emo-
tional sweating. Little is known about the precise
changes in blood flow in other areas of skin.
general sensory stimuli. Transient reduction in
hand or finger blood flow has been described in re-
sponse to a great range of mildly unpleasant stimuli
such as immersing another part of the body in cold
water (160), pinching (2), or inflating a pneumatic
cuff around the arm. On the other hand, Lynn &
Simeone (146) were unable to provoke reflex vaso-
constriction in anesthetized dogs by electrical stimu-
lation or by distention of the femoral vein.
response to a deep inspiration. After a deep in-
spiration there is a transient decrease in finger volume
(38, 94). The size of the arterial pulsations diminishes
and the rate of blood flow falls sometimes to a very
low level (116, 189). The blood flow can be seen to
slow in the capillary loops of the nail bed (153). The
blood flow is similarly transiently decreased in the
hands and feet (138) but not in the more proximal
parts of the limbs. The response is lost after nerve
block or sympathectomy.
Gilliatt (90) found that the vasoconstrictor response
in the finger could be elicited by a sufficiently fast
and deep expansion of the lungs, whether brought
about by passive inflation or voluntary inspiration.
It did not follow obstructed inspiratory or expiratory
efforts, nor deep expiration. The response has been
observed in the fingers and toes of persons with a
complete break in the functional continuity of the
■:544
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
spinal cord above the level of the sympathetic outflow
to the hands (91), and in these cases the response
appears to be a purely spinal reflex. At least some of
the afferent fibers must enter the cord below the
second thoracic roots. The receptors and the precise
nature of the effective stimulus have not been identi-
fied. The reflex may be responsible for the reduction
in the blood flow through the hands in hyperventila-
tion (2), but its functional significance is unknown.
RESPONSE TO DISTENTION OF THE BLADDER. Distention
of the bladder causes a constriction of the blood
vessels of the skin, and elsewhere, by a spinal reflex.
The response was first described by Guttman &
Whitteridge (111) in patients with complete trans-
verse section of the spinal cord and in whom the
isolated cord was undamaged. With sections above
the level of the second lumbar outflow there was
constriction in the skin of the feet and legs; with high
section there was also constriction in the hands. The
response may be sufficient to raise the arterial pres-
sure, and to lead to consequent reflex adjustments in
that part of the circulation innervated by the brain
and upper part of the spinal cord. It is important
when making observations on the skin circulation
to start with the subject's bladder empty, and to re-
empty it before it becomes uncomfortably full.
hypoglycemia. The blood flow is increased through
the hand and the forearm in insulin hypoglycemia
(3). The increase in the forearm is partly in the skin,
and mediated by an active vasodilator mechanism
(11), probably associated with sweating. Injection of
an adequate dose of atropine into the brachial artery
reduces the blood flow through the forearm, but does
not affect that through the hand (13).
posture. Changes in body posture cause complex
changes and reactions in the circulation, but if the
inclination to the horizontal of the observed limb re-
mains unchanged the net effect is that the circulation
through the hands (27) and fingers and toes (131 ) is
little changed. By contrast, as a single arm or leg is
raised above the horizontal, the posture of all other
parts remaining unchanged, the rate of blood flow
through the digits is progressively diminished (174,
191) and in all dependent positions the blood flow is
slightly increased (174). The digital pulse volume is
greater in the raised limb, and less in the dependent
limb, than it is in a horizontal limb (96). This is a strik-
ing example of the way in which, under some circum-
stances, blood flow and pulse volume may even
change in opposite directions, although in other
circumstances their changes may correspond closelv
(42).
RESPONSES TO BARORECEPTOR STIMULATION. Present
evidence suggests that the blood vessels of the skin
are largely, and perhaps entirely, excused from par-
ticipation in baroreceptor reflexes.
Low pressure baroreceptor s. Unidentified low pressure
baroreceptors within the thorax can be stimulated
by raising the legs of a recumbent subject and allow-
ing part of the blood they contain to flow into the
central venous pool, or by breathing through a narrow
tube which restricts air flow and causes intrathoracic
pressure transients of +30 to — 20 mm Hg to be set
up. Such stimulation causes a reflex dilatation,
brought about by reduction in vasoconstrictor nerve
activity, in the blood vessels of the voluntary muscle
of the forearm, but no change in the resistance to
flow through the hand or through the skin of the
forearm (173).
Systemic arterial baroreceptors. The decisive animal
experiments, in which observations have been made
on a perfused isolated innervated limb, while the
baroreceptors have been stimulated in various ways
(sec 119), have dealt either with whole limbs or
skinned limbs. There do not appear to have been
anv decisive experiments dealing with the skin as
such. In the human, in experiments in which bilateral
arterial compression caused increases in heart rate
and arterial blood pressure (and was thereby shown
to affect the baroreceptors), the vascular resistance
through the hand remained unchanged (168).
ACTION OF HUMORAL AGENTS ON THE
BLOOD VESSELS OF THE SKIN
The action of drugs on the skin has been recently
reviewed by Herxheimer (118). Only substances of
ph\ siological importance will be considered here.
Their direct local action is best tested by a steady
intraarterial infusion. The dose is adjusted to the
volume of tissue to which it will be distributed, and
it is usually so small that any returning to the general
circulation causes a negligible disturbance of arterial
pressure and of the blood vessels elsewhere. By measur-
ing the blood flow in the contralateral limb as well
.is in the infused limb general disturbances of the
circulation can be detected and can be allowed for
since these normally affect the limbs symmetrically.
The effect of humoral agents released into the
CIRCULATION THROUGH THE SKIN
1345
40 -S
20
Adrenaline
Noradrenaline /
/
c
0
/
/
/
/
/
^7
0
0
>
tn
c
- V ^
^
C
<-3
/ p ^"""^
51
<7
1/
MCj/min
1 1
pg/min
1 1 1
64
16
_1_
64
16
fig. 14. The mean percentage
reduction in the rate of blood flow in 1 3
hands, tested before and after sym-
pathectomy, in response to infusions of
adrenaline and noradrenaline at various
rates into the brachial artery. The
effects of general disturbance of the
circulation have been eliminated by
referring the blood flow to the simul-
taneously measured blood flow in the
opposite hand. Noradrenaline has a
greater constrictor effect than adrena-
line in normally innervated hands. After
sympathectomy the response to both
substances is augmented. [From Duff
(64)0
general circulation is tested by intravenous infusion.
The rate should ideally follow the normal pattern of
secretion but, in the general absence of exact knowl-
edge of this, steady infusion is best employed (g) and
is much superior to a sudden injection. The effect
on the skin or any other part of the peripheral circu-
lation depends on a combination of direct local
action and of general circulatory disturbance involv-
ing changes in arterial pressure and vasomotor
control, and probably changes in the concentration
of other humoral agents in the arterial blood.
Adrenaline and Noradrenaline
subcutaneously both substances cause
1 pallor, and iontophoresis of adrenaline
virtually to arrest the circulation through
. Injected into the brachial artery in very
both substances (22) cause a reduction in
through the hand (fig. 14). There is
of the low pressure capacity vessels,
veins, as well as of the resistance vessels
Injected
intense loca
can be used
the skin ( 7 1
small doses,
blood flow
constriction
presumably
(93)-
During the infusion of either substance intra-
venously at 20 fig per min there is a severe reduction,
and sometimes nearly complete arrest, of the blood
flow through the hands. After adrenaline injection,
the flow usually increases for a time to above the
resting level (182) and there is often a flushing of the
face (17, 103). No such increase is seen in sym-
pathectomized hands, or after intravenous nor-
adrenaline, or either substance given intra-arterially
(182).
Histamine
Pricked into the skin histamine causes a local wheal
and a reddening of the skin or flare extending for a
radius of 3 to 4 cm (139). The temperature of the skin
is only slightly increased in the region of the flare
(61, 139) and, although the content of blood is greatly
increased, the increase in the flow is modest. Infused
into the brachial artery, histamine has a dilator
effect at all doses tested (fig. 15) and the skin becomes
deeply flushed. The flush does not always cover all
parts of the hand, and this illustrates a general
difficulty with intra-arterial infusions. The injected
material may not become thoroughly mixed with the
arterial blood at the site of injection, and the artery
may not be the exclusive supply to the area of tissue
examined. Further, the pattern of distribution may
varv with changes in the circulation.
Acetylcholine
This acts as a powerful dilator to the blood vessels
of the hand (fig. 15), but so rapid is its destruction in
the blood stream that for equal effect on the blood
flow through the hand the dose into the brachial
artery must be about one thousand times as great
as that into the radial artery (65).
5-Hydroxytryptamine ( Serotonin )
When infused into the brachial artery at the rate
of 1 fig per min or more, this causes a reduction in
the rate of flow of blood through the hand, but the
volume of the part increases because of edema forma-
1346 HANDBOOK OF PHYSIOLOGY -^ CIRCULATION II
fig. 15. The effect on forearm and
hand blood flow of various doses of
acetylcholine and histamine injected
over i-min periods into the brachial
artery of four normal subjects. Con-
tinuous line: injected side; dotted lines:
control side. Doses in jig. [From Duff
et al. (65).]
40
E
§
ACh, hand
AD.
16
256
4000
-
.,-. -.••■.■••
R.B.C.
5 M- 10
Minutes
Minutes
tion, and the skin becomes flushed and petechial
hemorrhages appear; thus the vessels controlling
flow are constricted, and those responsible for color
are dilated (169). On the other hand, the low pressure
capacity vessels as a whole are rendered less disten-
sible (93), so the reaction of those responsible for color
appears not to be typical of the low pressure vessels
as a whole.
Adenosine Triphosphate
Injected intra-arterially in man, magnesium
adenosine triphosphate causes a great increase in the
blood flow through the hand, and is nearly as ef-
fective as an equal weight of histamine (68). A dose
of 1 mg per min into the brachial artery raises the
blood flow in the hand to an average value of 34 ml
per 100 ml of hand per min. Adenosine triphosphate
may be the transmitter substance released from
sensory nerve endings causing antidromic vaso-
dilatation (122).
Bradykinin
Detailed information is not yet published but it
appears that, per molecule of injected substance,
bradvkinin has a more potent vasodilator effect than
any other tested substance (80).
Carbon Dioxide
Breathing mixtures containing high concentrations
of carbon dioxide has a complex effect on the circu-
lation, with great disturbance of vasomotor regulation.
The local effect of carbon dioxide, as seen when a
hand is immersed in a saturated solution of the gas
(60) or when carbon dioxide mixtures are injected
subcutaneously (61), is entirely vasodilator. The
effect has not, however, been quantitatively defined
in terms of the response to various tensions of the
gas in the tissues.
Vasopressin
Infused intravenously this causes an initial vaso-
constriction in the hands, which diminishes as the
infusion continues ( 1 33)-
Oxytocin
Injected intravenously in man this may cause
flushing. Injections of 500 units intravenously or
50 units into the brachial artery cause the blood
flow through the hand to double for a few minutes.
With repeated doses, by either route, the response
diminishes (134). The vasodilator effect of oxytocin
is balanced by one-twentieth of the number of units
of vasopressin.
CIRCULATION THROUGH THE SKIN
'347
REFERENCES
'3'
1 6.
17'
'9'
Abramson, D. I. Vascular Responses in the Extremities of Man
in Health and Disease. Chicago: Univ. Chicago Press, 1944.
Abramson, D. I., and E. B. Ferris. Responses of blood
vessels in the resting hand and forearm to various stimuli. 21.
Am. Heart J. 19: 541, 1940.
Abramson, D. I., M. Schkloven, M. N. Margolis, and
I. A. Mirskv. Influence of massive doses of insulin on 22.
peripheral blood flow in man. Am. J. Physiol. 128: 124,
■939- 23-
Abramson, D. I., H. Zazeela, and J. Marrus. Plethysmo-
graphy studies of peripheral blood flow in man. II.
Physiologic factors affecting resting blood flow in the 24.
extremities. Am. Heart J. 17: 206, 1939.
Adson, A. W., and G. E. Brown. Calorimetric studies of
the extremities following sympathetic ramisectomy and 25.
ganglionectomy. Am. J. Med. Sci. 170: 232, 1925.
Adson, A. W., and G. E. Brown. The treatment of
Raynaud's disease by resection of the upper thoracic and
lumbar sympathetic ganglia and trunks. Surg. Gynecol. 26.
Obstet. 48: 577, 1929.
Ahmad, A. Paradoxical responses to changes of local
temperature in the hands of a recently sympathectomized
hyperhidrotic subject. Clin. Sci. 13: 351, 1954. 27.
Ahmad, A. Response of the blood vessels of the upper
extremity to prolonged local heat. Clin. Sci. 15: 609, 1956.
Allen, W. J., H. Barcroft, and O. G. Edholm. On the 28.
action of adrenaline on the blood vessels in human
skeletal muscle. J. Physiol., London 105: 255, 1946.
Allwood, M. J., H. Barcroft, J. P. L. A. Hayes, and
E. A. Hirsjarvi. The effect of mental arithmetic on the 29.
blood flow through normal, sympathectomised and
hyperhidrotic hands. J. Physiol., London 148: 108, 1959.
Allwood, M. J., I. Birchall, and J. S. Staffurth.
Circulatory changes in the forearm during insulin hypo- 30.
glycaemia studied by regional 24Na clearance and by
plethysmography. J. Physiol., London 143: 332, 1958.
Allwood, M. J., and H. S. Burrv. The effect of local 31.
temperature on blood flow in the human foot. J. Physiol.,
London 124: 345, 1954.
Allwood, M. J., and J. Ginsburg. The effect of intra- 32.
arterial atropine on blood flow in the hand and forearm
during insulin hypoglycaemia J. Physiol., London 149:
486, 1959- 33-
Anrep, G V., G. S. Barsoum, S. Salama, and Z. Souidan.
Liberation of histamine during reactive hyperaemia and
muscle contraction in man. J. Physiol., London 103: 297, 34.
'944-
Arnott, \V. M., and J. M. Macfie. Effect of ulnar nerve
block on blood flow in the reflexly vasodilated digit. J. 35.
Physiol., London 107: 233, 1948.
Aschoff, J. Uber die Kaltedilatation der Extremitat des
Menschen in Eiswasser. Pfliigers Arch. ges. Physiol. 248:
183, 1944. 36.
Barclay, J. A., W. T. Cooke, and R. A. Kenney.
Observations on the effects of adrenaline on renal function
and circulation in man. Am. J. Physiol. 151 : 621, 1947. 37.
Barcroft, H. Problems of sympathetic innervation and
denervation. Brit. Med. Bull. 8: 363, 1952.
Barcroft, H. Sympathetic control of vessels in the hand 38.
and forearm skin. Physiol. Revs. 40: 81, i960.
Barcroft, H., K. D. Bock, H. Hensel, and A. H.
Kitchin. Die Muskeldurchblutung des Menschen bei
Indirektar Erwarmung und Abkiihiung. Pfliigers Arch. ges.
Physiol. 261: 199, 1955.
Barcroft, H, and O. G. Edholm. On the vasodilatation
in human skeletal muscle during post-haemorrhagic
fainting. J. Physiol., London 104: 161, 1945.
Barcroft, H., and H. J. C. Swan. Sympathetic Control of
Human Blood Vessels. London: Arnold, 1953.
Barcroft, H., and A. J. Walker. Return of tone in
blood vessels of the upper limb after sympathectomy.
Lancet 1 : 1035, 1949.
Bayliss, W. M. On the origin from the spinal cord of the
vaso-dilator fibres of the hind-limb, and on the nature of
these fibres. J. Physiol., London 26: 173, 1901.
Bazett, H. C, L. Love, M. Newton, L. Eisenburg,
R. Day, and R. Forster. Temperature changes in blood
flowing in arteries and veins in man. J. Appl. Physiol. 1 :
3, '948-
Bazett, H. G, E. S. Mendelson, L. E. Love, and B.
Libet. Precooling of blood in the arteries, effective heat
capacity and evaporative cooling as factors modifying
cooling of the extremities. J. Appl. Physiol. 1: 169, 1948.
Beaconsfield, P., and J. Ginsburg. The effect of body
posture on hand blood flow. J. Physiol., London 130: 467,
!955-
Beckett, E. B., G. H. Bourne, and W. Montagna.
Histology and cytochemistry of human skin. The dis-
tribution of cholinesterase in the finger of the embryo and
the adult. J. Physiol., London 134: 202, 1956.
Behnke, A. R., and T. L. Willmon. Cutaneous diffusion
of helium inhalation to peripheral blood flow and absorp-
tion of atmospheric nitrogen through the skin. Am. J.
Physiol. 131:627, 1941.
Bernard, C. Influence du grand sympathique sur la
sensibilite et sur la calorification. Compt. Rend. Soc. Biol.
3: 163, 1851.
Bernard, C. Sur les effets de la section de la portion
cephalique du grand sympathique. Compt. Rend. Soc. Biol.
4: 168, 1852, quoted by Monro (152).
Bernard, C. Sur les variations de couleur dans le sang
veineux des organes glandularies suivant leur 6tat de
fonction ou de repos. J. Physiol., Paris 1 : 233, 1858.
Bier, A. Die Enstehung des Collateralkreislaufs. Theil I.
Die arterielle Collateralkrcislauf. Arch. Pathol. Anal.
Physiol. 147:256, 1897.
Bier, A. Die Enstehung des Collateralkreislaufs. Theil II.
De Riickfluss des Blutes aus ischamischen Korpertheilen.
Arch. Pathol. Ana/. Physiol. 153: 306, 1898.
Blair, D. A., VV. E. Glover, and I. C. Roddie. The
abolition of reactive and post-exercise hyperaemia in the
forearm by temporary restriction of arterial inflow. J.
Physiol., London 148: 648, 1959.
Blair, D. A., W. E. Glover, and I. C. Roddie. Vaso-
motor fibres to skin in the upper arm, calf and thigh.
J. Physiol., London 153: 232, i960.
Blaisdell, R. K. Cold Induced Vasodilatation. Office of
Q. M. General, U. S. Army, Environment Protection
Section Rept. No. 177: 1, 1951.
Bolton, B., E. A. Carmichael, and G. Sturup. Vaso-
constriction following deep inspiration. J. Physiol., London
86:83, '936.
■348
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
39. Brown-Sequard, C.-E. Rccherches sur l'influencc dn
systeme nerveux sur les fonctions de la vie organique.
Med. Exam. Phila. 486, 1852.
40. Burch, G. E. Digital Plethysmography. New York: Grune
& Stratton, 1954.
41. Burch, G. E., A. E. Cohn, and C. Neumann. Spon-
taneous variations in volume of the finger tip, toe tip, and
postero-superior portion of the pinna of resting normal
white adults. Am. J. Physiol. 136: 433, 1942.
42. Burton, A. C. The range and variability of the blood flow
in the human fingers and the vasomotor regulation of body
temperature. Am. J. Physiol. 127: 437, 1 939.
43. Burton, A. C. On the physical equilibrium of small
blood vessels. Am. J. Physiol. 164: 319, 1951.
44. Burton, A. G, and O. G. Edholm. Man in a Cold Environ-
ment. London: Arnold, 1955.
45. Burton, A. G, and R. M. Taylor. A study of the
adjustment of peripheral vascular tone to the require-
ments of the regulation of body temperature. Am. J.
Physiol. 129: 565, 1940.
46. Cannon, W. B., and A. Rosenblueth. The Supersensitivity
of Denervated Structures. A Law of Denervation. New York :
Macmillan, 1949.
47. Catchpole, B. N, and R. P. Jepson. Hand and fingei
blood flow. Clin. Sci. 14: 109, 1955.
48. Celander, O. The range of control exercised by the
sympathico-adrenal system. Acta Physiol. Scand. 32 :
Suppl. 1954.
49. Celander, O., and B. Folkow. The nature and the
distribution of afferent fibres provided with the axon
reflex arrangement. Acta Physiol. Scand. 29 : 359, 1 953-
50. Clark, E. R. Arterio-venous anastomoses. Physiol. Revs.
18: 229, 1938.
51. Cohnheim, J. Gesammelte Abhandlungen Von Julius Cohnheim.
Berlin: Hirschwald, 1872, p. 301.
52. Coles, D. R. Heat elimination from the toes during the
exposure of the foot to subatmospheric pressures. J.
Physiol., London 135: 171, 1957.
53. Coles, D. R., and A. D. M. Greenfield. The reactions
of the blood vessels of the hand during increases in
transmural pressure. J. Physiol., London 131 : 277, 1956.
54. Coles, D. R., and G. C. Patterson. The capacity and
distensibility of the blood vessels of the human hand. J.
Physiol., London 135: 163, 1957.
55. Cooper, K. E., K. W. Cross, A. D. M. Greenfield,
D. McK. Hamilton, and H. Scarborough. A comparison
of methods for gauging the blood flow through the hand.
Clin. Sci. 8: 217, 1949.
56. Cooper, K. E., O. G. Edholm, and R. F. Mottram.
The blood flow in skin and muscle of the human forearm.
J. Physiol., London 128: 258, 1955.
57. Cooper, K. E., and D. McK. Kerslake. Abolition of
nervous reflex vasodilatation by sympathectomy of the
heated area. J. Physiol., London 119: 18, 1953.
58. Cooper, K. E., and D. McK. Kerslake. Vasoconstric-
tion in the hand during electrical stimulation of the
lumbar sympathetic chain in man. J. Physiol., London
127: 134. '955-
59. Crockford, G. VV., and R. F. Hellon. Vascular re-
sponses of human skin to infra-red radiation. J. Physiol.,
London 149: 424, 1959.
60. Diji, A. Local vasodilator action of carbon dioxide on
blood vessels of the hand. ./. Afipl. Physiol. 14: 414, 1959.
61. Diji, A., and A. D. M. Greenfield. The local effect of
carbon dioxide on human blood vessels. Am. Heart J.
60: 907, ig6o.
62. Doupe, J. Studies in denervation. B. The circulation in
denervated digits. J. Neurol. Psychiat. 6: 97, 1943.
63. Duff, R. S. Circulatory changes in the forearm following
sympathectomy. Clin. Sci. 10: 529, 1 951.
64. Duff, R. S. Effect of adrenaline and noradrenaline on
blood vessels of the hand before and after sympathectomy.
J. Physiol., London 129: 53, 1955.
65. Duff, F., A. D. M. Greenfield, J. T. Shepherd, and
I. D. Thompson. A quantitative study of the response to
acetylcholine and histamine of the blood vessels of the
human hand and forearm. J. Physiol., London 120: 160,
"953-
66. Duff, F., A. D. M. Greenfield, and R. F. VVhelan.
Vasodilatation produced by experimental arterial gas
embolism in man. Lancet 2: 230, 1953.
67. Duff, F., A. D. M. Greenfield, and R. F. Whelan.
Observations on the mechanism of the vasodilatation
following arterial gas embolism. Clin. Sci. 13: 365, 1954.
68. Duff, F., G. C. Patterson, and J. T. Shepherd. A
quantitative study of the response to adenosine triphos-
phate of the blood vessels of the human hand and forearm.
./. Physiol., London 125: 581, 1954.
6g. Duff, F., G. C. Patterson, and R. F. Whelan. The
effect of intra-arterial antihistamines on the hyperaemia
following temporary arrest of the circulation in the human
forearm. Clin. Sci. 14: 267, 1955.
70. Duff, F., and J. T. Shepherd. The circulation in the
chronically denervated forearm. Clin. Sci. 12: 407, 1953.
71. Edholm, O. G, R. H. Fox, and R. K. Macpherson.
Effect of body heating on the circulation in skin and
muscle. ./. Physiol., London 134: 612, 1956.
72. Edholm, O. G., R. H. Fox, and R. K. Macpherson.
Vasomotor control of the cutaneous blood vessels in the
human forearm. ./. Physiol., London 139: 455, 1957.
73. Eichna, L. VV., and R. Wilkins. Blood flow to the forearm
and calf. II. Reactive hyperaemia: Factors influencing the
blood flow during the vasodilatation following ischaemia.
Bull. Johns Hopkins Hosp. 68: 450, 1941.
74. Essex, H. E., J. F. Hfrrick, E. J. Baldes, and F. C.
Mann. Observations on the circulation in the hind limbs
of a dog ten years following left lumbar sympathetic
ganglionectomy. Am. J. Physiol. 139: 351, 1943.
75. Felder, D., E. Russ, H. Montgomery, and O. Horwitz.
Relationship in the toe of skin surface temperature to
mean blood flow measured with a plethysmograph.
Clin. Sci. 13: 251, 1954.
76. Fencl, V., Z. Hejl, J. Jirka, J. Madlafousek, and
J. Brod. Changes of blood flow in forearm muscle and
skin during an acute emotional stress (mental arithmetic).
Clin. Sci. 18: 491, 1959.
77. Folkow, B., J. Frost, K. Haeger, and B. Uvnas. The
sympathetic vasomotor innervation of the skin of the dog.
Acta Physiol. Scand. 17: 195, 1949.
78. Fox, R. H., R. Goldsmith, and D. J. Kidd. Cutaneous
vasomotor nerves in the human ear and forehead. J.
Physiol. , London 150: 12P, ig6o.
79. Fox, R. H., R. Goldsmith, and D. J. Kidd. The cuta-
neous vasomotor control in the human nose, lip and chin.
./. Physiol., London 150: 22P, i960.
80. Fox, R. H., R. Goldsmith, D. J. Kidd, and G. P. Lewis.
CIRCULATION THROUGH THE SKIN
'349
Bradykinin as a vasodilator in man J. Physiol., London
154: 16P, i960.
81. Fox, R. H., and S. M. Hilton. Bradykinin formation in
human skin as a factor in heat vasodilatation. J. Physiol.,
London 142: 219, 1958.
82. Fredericq, L. Sur la regulation de la temperature chez
les animaux a sang chaud. Arch, bwl., Liege 3: 687, 1882.
83. Freeman, N. E. Effect of temperature on rate of blood
How in normal and in sympathectomized hand. Am. J.
Physiol. 113:384, 1938.
84. Friedman, N. B. The pathology of trench foot. Am. ./.
Pathol. 21 : 387, 1945-
85. Froese, G., and A. C. Burton. Heat losses from the
human head. J. Appl. Phynol. 10: 235, 1957.
86. Gaskell, P. The rate of blood How in the foot and calf
before and after reconstruction by arterial grafting of an
occluded main artery to the lower limb. Clin. Sci. 1 5 :
-*59. '956-
87. Gaskell, P. Are these sympathetic vasodilator nerves
to the vessels of the hands? J. Physiol., London 131 : 647,
1956.
88. Gerbrandv, J., E. S. Snell, and \V. I. Cranston. Oral
rectal and oesophageal temperatures in relation to central
temperature control in man. Clin. Sci. 13: 615, 1954.
89. Gibbon, J. H. H., and E. M. Landis. Vasodilatation in
the lower extremities in response to immersing the
forearms in warm water. J. Clin. Invest. 11: 1019, 1932.
90. Gilliatt, R. W. Vaso-constriction in the linger after deep
inspiration. J. Physiol., London 107: 76, 1948.
91. Gilliatt, R. W., L. Guttman, and D. Whitteridge.
Inspiratory vasoconstriction in patients after spinal
injuries. J. Physiol., London 107: 67, 1948.
92. Glaser, E. M., and G C. Whittow. Retention in a
warm environment of adaptation to localised cooling. J.
Physiol., London 136: 98, 1957.
93. Glover, \V. E., A. D M. Greenfield, B. S. L. Kidd,
and R. F. Whelan. The reactions of the capacity blood
vessels of the human hand and forearm to vaso-active
substances infused intra-arterially. J. Physiol., London 140:
113, 1958.
94. Goetz, R. H. Der Fingerplethysmograph als Mittel zur
Untersuchung der Regulationsmechanismen in peri-
pheren Gefassgebieten. Pjlugers Arch. ges. Physiol. 235:
271. 1935
95. Goetz, R. H. Rate of control of blood flow through the
skin of lower extremities. .4m. Heart J. 31 : 146, 1946.
96. Goetz, R. H. Effect of changes in posture on peripheral
circulation with special reference to skin temperature
readings and the plethysmogram. Circulation 1 : 56, 1 950.
97. Goltz, F., and A. Freusberg. Uber gefasseriveiternde
Nerven. Pjlugers Arch. ges. Physiol. 9: 174, 1874.
98. Grant, R. T. Observations on direct communications
between arteries and veins in the rabbit's ear. Heart 15:
281, 1930.
99. Grant, R. T., and E. F. Bland. Observations on arterio-
venous anastomoses in human skin and in the bird's foot
with special reference to the reaction to cold. Heart 15:
385. '93'
100. Grant, R. T., and H. E. Holling. Further observations
on the vascular responses of the human limb to body
warming; evidence for sympathetic vasodilator nerves in
the normal subject. Clin. Sci. 3: 273, 1938.
101. Grant, R. T., and R. S. B. Pearson. The blood circula-
te^.
104.
105.
107
108.
log.
tion in the human limb, observations on the differences
between the proximal and distal parts and remarks on
regulation of body temperature. Clin. Sci. 3: 119, 1938.
Green, H. D., W. B. Howard, and L. F. Kenan. Auto-
nomic control of blood flow in hind paw of the dog. Am. J.
Physiol. 187: 469, 1956.
Green, D. M., A. D. Johnson, A. Lobb, and G. Cusick.
The effects of adrenaline in normal and hypertensive
patients in relation to the mechanism of sustained pressure
elevations. J. Lab. Clin. Med. 33: 332, 1948.
Greenfield, A. D. M., G A. Kernohan, R. J.
Marshall, J. T. Shepherd, and R. F. Whelan. Heat
loss from toes and fore-feet during immersion in cold water.
J. Appl. Physiol. 4: 37, 1 95 1.
Greenfield, A. D. M., and J. T. Shepherd. A quantita-
tive study of the response to cold of the circulation through
the fingers of normal subjects. Clin. Sci. 9: 323, 1950.
106. Greenfield, A. D. M., J. T. Shepherd, and R. F.
Whelan. The loss of heat from the hands and from the
fingers immersed in cold water. J. Physiol., London 112:
459. '95°.
Greenfield, A. D. M., J. T. Shepherd, and R. F.
Whelan. The average internal temperature of fingers
immersed in cold water. Clin. Set. g: 349, 1950.
Greenfield, A. D. M., J. T. Shepherd, and R. F.
Whelan. Cold vasoconstriction and vasodilatation.
Irish J. Med. Sci. 309: 415, 1 95 1.
Greenfield, A. D. M., J. T. Shepherd, and R. F.
Whelan. The part played by the nervous system in the
response to cold of the circulation through the finger tip.
Clin. Sci. 10: 347, 1 95 1 .
Greenfield, A. D. M., J. T. Shepherd, and R. F.
Whelan. Circulatory response to cold in fingers infiltrated
with anesthetic solution. J. Appl. Physiol. 4: 785, 1952.
Guttman, L., and D. Whitteridge. Effects of bladder
distension on autonomic mechanisms after spinal cord
injuries. Brain 70: 361, 1 947.
Hardy, J. D., and G F. Soderstrom. Heat loss from the
nude body and peripheral blood Mow at temperatures of
22°C to 35 °C J. Nutrition 16: 493, 1938.
Hellstrom, B., and K. L. Andersen. Heat output in the
cold from hands of Arctic fishermen. J. Appl. Physiol.
15: 771, i960.
Hensel, H., and F. Bender. Fortlaufende Bestimmung der
Hautdurchblutung am Menschen mit einem elektrischen
Warmeleitmesscr. Pjlugers Arch. ges. Physiol. 263: 603,
!95°-
Hertzman, A. B. Vasomotor regulation of cutaneous
circulation. Physiol. Revs. 39: 280, igsg.
Hertzman, A. B., and J. B. Dillon. Selective vascular
reaction patterns in the nasal septum and skin of the
extremities and head. Am. J. Physiol. 127: 671, I93g.
Hertzman, A. B., and L. W. Roth. The absence of
vasoconstrictor reflexes in the forehead circulation.
Effects of cold. Am. J. Physiol. 136: 692, 1942.
Herxheimer, A. The action of drugs on the skin. Ann.
Rev. Pharmacol. 1 : 351 , 1 96 1 .
Heymans, C, and E. Neil. Refle xogenic Areas of the Cardio-
vascular System. London : Churchill, 1 958.
Hilton, S. M. A peripheral arterial conducting mecha-
nism underlying dilatation of the femoral artery and
concerned in functional vasodilatation in skeletal muscle.
J. Physiol., London 149: 93, ig5g.
"3
114
"5
116.
117.
"35°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
123.
124.
125-
126.
127.
128.
129.
130.
I31
132.
• 33-
134-
■35-
136.
137-
138.
■39-
140.
141.
Holton, F. A., and P. Holton. The capillary dilator
substance in dry powders of spinal roots; a possible role of
adenosine triphosphate in chemical transmission from
nerve endings. J. Physiol., London 126: 124, 1954. 142
Holton, P. The liberation of adenosine triphosphate on
antidromic stimulation of sensory nerves. J. Physiol., 143
London 145: 494, 1959.
Holton, P., and W. L. M. Perry. On the transmitter
responsible for antidromic vasodilatation in the rabbit's
ear. J. Physiol., London 114: 240, 1951. 144
Hurley, H. J., and H. Mescon. Cholinergic innervation
of the digital arterio-venous anastomoses of human skin.
A histochemical localisation of cholinesterase. J. Appl.
Physiol. 9: 82, 1956. 145,
Imig, C. J., W. J. Roberson, M. Gault, and H. M.
Hines. Blood flow in the hind legs of dogs after exposure
to cold. Am. J. Physiol. 181 : 395, 1955. 146.
Imig, C. J., W. J. Roberson, and H. M. Hines. Compari-
son of blood flow in normally innervated and in sym-
pathcctomized legs of dogs after exposure to cold. Am. J.
Physiol. 186: 35, 1956. 147.
Keatinge, W. R. Effect of general chilling on the vaso-
dilator response to cold. J. Physiol., London 139: 497, 1957. 148.
Keatinge, W. R., and P. Cannon. Freezing point of
human skin. Lancet 1: 11, i960. 149.
Kerslake, D. McK., and K. E. Cooper. Vasodilatation
in the hand in response to heating the skin elsewhere.
Clin. Sci. 9: 31, 1950.
Kety, S. S. Measurement of regional circulation by the 150.
local clearance of radio-active sodium. Am. Heart J.
38: 321, 1949. 151.
Kidd, B. S. L., and R. V. McCready. Effect of change in
posture on the blood flow through the fingers and toes.
J. Appl. Physiol. 12: 121, 1958. 152.
Killian, J. A., and C. A. Oclassen. Comparative effects
of water baths and mustard baths at varying temperatures
on the rate of peripheral blood flow in man. Am. Heart J. 153.
i5:4->5. I938-
Kitchin, A. H. The effect of pitressin on hand and
forearm blood flow. Clin. Sci. 16: 639, 1957. 154.
Kitchin, A. H., S. M. Lloyd, and M. Pickford. Some
actions of oxytocin on the cardiovascular system in man.
Clin. Sci. 18: 399, 1959. 155.
Krog, J., B. Folkovv, R. H. Fox, and K. L. Andersen.
Hand circulation in the cold of Lapp and north Nor- 156.
wegian fishermen. J. Appl. Physiol. 15: 654, i960.
Kunkel, P., and E. A. Stead. Blood flow and vasomotor 157.
reactions in the foot in health, in arteriosclerosis and in
thrombo-angiitis obliterans. J. Clin. Invest. 17: 715, 1938. 158.
Kunkel, P., E. A. Stead, and S. Weiss. Blood flow and
vasomotor reactions in the hand, forearm, foot and
calf in response to physical and chemical stimuli. J. Clin. 159.
Invest. 18: 225, 1 939.
Langley, J. N., and K. Uyeno. The secretion of sweat.
Part II. The effect of vasoconstriction and of adrenaline. ,60
J. Physiol., London 56: 207, 1922.
Lewis, T. The Blood Vessels of the Human Skin and Their
Responses. London: Shaw, 1927.
Lewis, T. Observations upon the reactions of the vessels
of the human skin to cold. Heart 15: 177, 1930.
Lewis, T., and R. T. Grant. Vascular reactions of the '"2-
skin to injury. Part II. The liberation of a histaminelike
substance in injured skin; the underlying cause of factitious
urticaria and of wheals produced by burning; and
observations upon the nervous control of certain skin
reactions. Heart 1 1 : 209, 1924.
Lewis, T., and R. T. Grant. Observations upon reactive
hyperaemia in man. Heart 12: 73, 1925.
Lewis, T., and G. W. Pickering. Vasodilatation in the
limbs in response to the warming of the body, with
evidence for sympathetic vasodilator nerves in man.
Heart 16; 33, 1 931.
Lewis, T., and G. W. Pickering. Circulatory changes in
the fingers in some diseases of the nervous system, with
special reference to the digital atrophy of peripheral nerve
lesions. Clin. Sci. 2 : 1 49, 1 936.
Lynn, R. B., and H. Barcroft. Circulatory changes in
the foot after lumbar sympathectomy. Lancet 1: 1105,
1950.
Lynn, R. B., and F. A. Simeone. Observations of reflex
vascular responses to stimulation of blood vessels and
perivascular tissues in the dog. Am. J. Physiol. 169: 471,
I952-
Masson, P. Les Glomus Neuro-vasculaires. Actualites Scientifi-
ques et Industnelles. Paris: Hermann, 1937, p. 453.
Mendlowitz, M. The Digital Circulation. New York:
Grune & Stratton, 1954.
Mendlowitz, M., and H. A. Abel. The quantitative
blood flow measured calorimetrically in the human toe in
normal subjects and in patients with residua of trench foot
and frost bite. Am. Heart J. 39: 92, 1950.
Meryman, H. T. Tissue freezing and local cold injury.
Physiol. Revs. 37: 233, 1957.
Mescon, H., H. J. Hurley, Jr., and G. Moretti.
Anatomy and histochemistry of the arteriovenous anasto-
mosis in digital skin. J. Invest. Dermatol. 27: 133, 1956.
Monro, P. A. G. Sympathectomy. An Anatomical and Physio-
logical Study with Clinical Applications. London : Oxford
Univ. Press, 1 959.
Mulinos, M. G., and I. Shulman. Vasoconstriction in the
hand from a deep inspiration. Am. J. Physiol. 125: 310,
'939-
McGirr, E. M. Rate of removal of radioactive sodium
following its injection into muscle and skin. Clin. Sci. 1 1 :
91. '952-
Newburgh, L. H. Physiology of Heat Regulation. Philadel-
phia: Saunders, 1949.
Partington, M. W. The vascular response of the skin to
ultra-violet light. Clin. Sci. 13: 425, 1954.
Patel, D. J., and A. C. Burton. Reactive hyperaemia in
the human finger. Circulation Research 4: 710, 1956.
Patterson, G. C. The role of intravascular pressure in the
causation of reactive hyperaemia in the human forearm.
Clin. Sci. 15: 17, 1956.
Peacock, J. H. Vasodilatation in the human hand.
Observations on primary Raynaud's disease and acro-
cyanosis of the upper extremities. Clin. Sci. 17: 575, 1958.
Pickerinc, G. W. The vasomotor regulation of heat loss
from the human skin in relation to external temperature.
Heart 16: 115, 1933.
Pickering, G. W. The peripheral resistance in persistent
arterial hypertension. Clin. Sci. 2: 209, 1936.
Pickering, G. W., and W. Hess. Vasodilatation in the
hands and feet in response to warming the body. Clin.
Sci. 1: 213, 1933.
CIRCULATION THROUGH THE SKIN
'35'
163. Popoff, N. W. The digital vascular system. .1.1/.. I.
Arch. Pathol. 18: 295, 1 934.
164. Prichard, M. M. L., and P. M. Daniel. Arteriovenous
anastomoses in the human external ear. ./. Anal, go: 309,
195°.
165. Richards, R. L. The Peripheral Circulation in Health and
Disease. Baltimore: Williams & Wilkins, 1946.
166. Roddie, I. C, and J. T. Shepherd. The blood flow
through the hand during local heating, release of sym-
pathetic vasomotor tone by indirect heating, and a
combination of both. J. Physiol., London 131 : 657, 1956.
167. Roddie, 1. C, and J. T. Shepherd. Evidence for critical
closure of digital resistance vessels with reduced transmural
pressure and passive dilatation with increased venous
pressure. J. Physiol., London 136: 498, 1957.
168. Roddie, I. C, and J. T. Shepherd. The effects of carotid
artery compression in man with special reference to
changes in vascular resistance in the limbs. J. Physiol.,
London 139: 377, 1957.
169. Roddie, I. C, J. T. Shepherd, and R. F. VVhelan. The
action of 5-hydroxytryptaminc on the blood vessels of
the human hand and forearm. Brit. J. Pharmacol. 10:
445. '955
170. Roddie, I. O, J. T. Shepherd, and R. F. VVhelan.
Evidence from venous oxygen saturation measurements
that the increase in forearm blood flow during body
heating is confined to the skin. J. Physiol., London 134:
444, 1956.
171. Roddie, I. C, J. T. Shepherd, and R. F. VVhelan. The
contribution of constrictor and dilator nerves to the skin
vasodilatation during body heating. J. Physiol., London
136: 489, 1957.
172. Roddie, I. C, J. T. Shepherd, and R. F. Whelan. A
comparison of the heat elimination from the normal and
nerve-blocked finger during body heating. J. Physiol.,
London 138: 445, 1957.
173. Roddie, I. C, J. T. Shepherd, and R. F. Whelan.
Reflex changes in human skeletal muscle blood flow
associated with intrathoracic pressure changes. Circulation
Research 6: 232, 1958.
174. Roddie, R. A. Effect of arm position on circulation
through the fingers. J. Appl. Physiol. 8: 67, 1955.
175. Roth, G. M. In: Peripheral Vascular Diseases, edited by
E. V. Allen, N. W. Barker, and E. A. T. Hines. Philadel-
phia: Saunders, 1946.
1 76. Rothm an , S. Physiology and Biochemistry of the Skin. Chicago :
Chicago Univ. Press, 1954, p. 60.
177. Scholander, P. F., H. T. Hammel, J. S. Hart, D. H.
Lemessurier, and J. Steen. Cold adaptation in Australian
aborigines. J. Appl. Physiol. 13:211, 1 958.
178. Simeone, F. A., and D. A. Felder. Supersensitivity of
denervated blood vessels in man. Surgery 30: 218, 1 95 1 .
1 ~g. Snell, E. S. The relationship between the vasomotor
response in the hand and heat changes in the body
induced by intravenous infusions of hot or cold saline.
J. Physiol. , London 125: 361, 1954.
180. Spealman, G. R. Effect of ambient air temperature and
of hand temperature on blood flow in the hands. Am. J.
Physiol. 145:218, 1945.
181. Sturup, G., B. Bolton, D. J. Williams, and E. A.
Carmichael. Vasomotor responses in hemiplegic patients.
Brain 58 : 456, 1 935.
182. Swann, H. J. C Observations on a central dilator action
of adrenaline in man. J. Physiol., London 112: 426, 1951.
183. Ungley, C. C. The immersion foot syndrome. Advances in
Surg. 1 : 269, 1949.
184. Walker, A. J., R. B. Lynn, and H. Barcroft. On the
circulatory changes in the hand and foot after sympathec-
tomy. St. Thomas's Hosp. Rept. 6: 18, 1950.
185. Warren, J. V., C. W. Walker, J. Romano, and E. A.
Stead. Blood flow in the hand and forearm after para-
vertebral block of the sympathetic ganglia. Evidence
against sympathetic vasodilator nerves in extremities of
man. J. Clin. Invest. 21 : 665, 1942.
186. White, J. C. In : Rehabilitation 0/ the War Injured, edited by
W. B. Doherty and D. C. Runes. London : Chapman &
Hall, 1943.
187. White, J. G, R. H. Smithwick, and F. A. Simeone. The
Autonomic Nervous System. New York: Macmillan, 1952.
188. Whittow, G. C. Effect of antihistamine substances on
cold vasodilatation in the finger. Nature 176: 51 1, ig55-
189. Wilkins, R. W., J. Doupe, and H. W. Newman. The
rate of blood flow in normal fingers. Clin. Sci. 3: 403, 1938.
190. Wilkins, R. W., and L. W. Eichna. Blood flow to the
forearm and calf. 1 . Vasomotor reactions : Role of the
sympathetic nervous system. Bull. Johns Hopkins Hosp. 68 :
425* '94>-
191. Wilson, G. M. The blood flow to the lower limbs in
peripheral arterial disease and coarctation of the aorta.
Edinburgh Med. J. 58: 125, 1951.
192. Wolff, H. H., and E. E. Pochin. Vasodilatation after-
reaction in recently cooled fingers. Clin. Sci. 8: 145, ig4g.
193. Wolstenholme, G. E. W., J. C. Freeman, and J. Ether-
ington. Peripheral Circulation in Man. London: Churchill,
'954-
194. Wood, J. E., J. Litter, and R. W. Wilkins. Mechanism
of limb segment reactive hyperaemia in man. Circulation
Research 3: 581, 1955.
CHAPTER 40
Circulation in skeletal muscle
HENRY BAR CROFT
Sherrington School of Physiology, St. Thomas Hospital
Medical School, London, England
CHAPTER CONTENTS
Basal Tone
Automaticity
Automaticity in Human Muscle Vessels
Pressure-Flow Relations in Muscle Vessels Deprived of
Automaticity
Critical Closing Pressure
Local Temperature
The Problem of Structure and Function
Nervous Control
Sympathetic Vasoconstrictor Nerves
Effect of sympathetic vasoconstrictors upon resistance,
blood volume, and capillary filtration in skeletal muscle
vessels
Chemical transmission at sympathetic vasoconstrictor
nerve endings in skeletal muscle
Effect of stimulation of the arterial baroreceptors on
skeletal muscle vessels in the dog
Effect of stimulation of the arterial baroreceptors on the
circulation in human skeletal muscle
Effect of receptors in a low pressure area in the cardio-
pulmonary system on the sympathetic vasoconstrictor
tone in human skeletal muscle
Impulse frequency in sympathetic vasoconstrictor fibers
Sympathetic Vasodilator Nerves
Chemical transmission at sympathetic vasodilator nerve
endings in skeletal muscle
Activation of sympathetic vasodilator fibers to skeletal
muscle by hypothalamic stimulation
Sympathetic vasodilator fibers to human skeletal muscles
Do Posterior Root Fibers Affect Muscle Blood Flow?
Effect of the Temperature-Regulating Center on the Circula-
tion in Muscle
Role of Sympathetic Fibers to Muscle in Exercise
Action of Sympathomimetic Substances
Noradrenaline
Adrenaline
Effect of Adrenaline on the Circulation in Skeletal Muscle
During Exercise
Reactive Hyperemia
Exercise Hyperemia
few will deny that analytical study of the physiology
of the circulation in skeletal muscle began in the
Institute of Physiology at Leipzig. The paper bears
the name of Gaskell (108), but it was Carl Ludwig
who suggested the problem and who probably did
many of the experiments. In Gaskell's Obituary
Notice written by Langley (137) we read — "At this
time Ludwig's laboratory was much the most im-
portant school of physiological research in Germany
or elsewhere. It attracted students from all parts of
the world. All the work was planned by Ludwig,
who had an almost unerring sense of the lines of
work which would yield profitable results. To this
the success of the school was mainly due. Its popu-
larity was increased by the method of procedure
adopted by Ludwig. This has been described by T.
Lauder Brunton who was with Ludwig in 1869-70.
The experiments were carried out by Ludwig with
the pupil as an assistant, Ludwig wrote the paper
and then published it, occasionally as a conjoint
work, but usually in the name of his pupil. As I have
heard from Gaskell the method was the same in his
time."
Be that as it may, let us turn to the experiments
themselves — "On the changes of the blood stream in
muscles through stimulation of their nerves." By a
simple graphical method venous outflow was recorded
from the extensor group of muscles of an unanesthe-
tized dog. The changes in outflow were determined
during and after tetanic stimulation of the crural
nerve. From a typical record, such as that seen in
figure 2 (top), six phases could be discerned during
sustained contraction : a) an initial spurt due to
squeezing of the veins by the muscles; b) decrease in
flow caused to some extent by mechanical compres-
sion of the vessels by the contracted muscle; c) in-
'353
■354
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
FIG. I. Walter Holbrook Gaskell, 1847-1914.
crease in flow; then, following contraction; d) a
check in the rate of the stream while the veins re-
filled; e) a further large increase in flow; and finally
/) gradual restoration of the flow to the resting rate.
It is interesting to compare Gaskell's record with
that seen in figure 2 (bottom), which was made by
Kramer & Quensel (131) 50 years later. They deter-
mined the venous outflow of the dog's gastrocnemius
with a hot-wire anemometer. Kramer and his
colleagues recognized the following changes in out-
flow during maximal tetanic stimulation of the motor
nerve: a) an initial peak due to expression of blood;
b) decrease in flow due to mechanical compression;
c) increase in flow; then, after relaxation, d) transitory
decrease while the vessels refilled; e) hyperemia
reaching the maximum; /) restoration of the flow
to the resting rate. The agreement between Gaskell's
and Kramer's records is remarkable, the main dif-
ference being that the postexercise flow was greater
in Gaskell's experiment. Presumably in his experi-
ment the muscle had contracted more powerfully
during stimulation.
So much for tetanic contraction. Figure 22 shows
the changes during rhythmic contraction. The record
is from another experiment of Kramer's (132). The
motor nerve to the gastrocnemius was stimulated for
1 sec every alternate sec for 5 min. Venous outflow-
increased rapidly during the first minute to reach a
steady level. Further increase in outflow occurred
immediately after the exercise because the stream
was no longer checked repetitively by mechanical
compression. Then after a few seconds it subsided to
the resting rate.
In man the changes in flow in the forearm muscles
during strong sustained contraction were determined
by Grant (113), who recorded the rate of the blood
flow by venous occlusion plethysmography. An ex-
cellent description of the method has been published
by Greenfield (114). Grant's subject gripped an iron
bar as hard as possible for 1 min. There was a small
increase in flow during the exercise and a large one
afterward. The vasodilatation during contraction was
not conspicuous because of compression of the vessels
I iv the contracted muscle. As soon as the muscle
relaxed, compression ceased and then blood flowed
rapidly into the veins.
During strong contraction of the human gastroc-
nemius soleus the effect of mechanical compression
of the muscle vessels may stop the flow. For example,
when one is standing tiptoe on the ball of one foot,
supporting the whole weight of the body by contrac-
tion of the calf muscles, the blood flow in these
muscles is probably arrested. This was inferred from
records of the changes in temperature in these muscles
made while the subject was standing on tiptoe (28).
The length of time one can stand tiptoe on one leg
MINUTES
fig. a. Top: Changes in venous outflow from the extensor
group of muscles of the dog*s leg, during (R-R) and after
tetanic stimulation of the crural nerve. [From Gaskell (108).]
Bottom: Changes in the venous outflow from the gastrocnemius
muscle of the dog recorded during (R-R) and after tetanic
stimulation of the sciatic nerve. [From Kramer & Quensel
O3O.]
CIRCULATION IN SKELETAL MUSCLE
'355
4-0 -
20
RHYTHMIC EXERCISE
1
1
1
-c j
iiiiMiHi
^ !
ill ' 1
— 1
E |
1 1 ' \
O '
\
— 1
1 1 \
1
E 1
1
II \
1 i
I 1 '■ 1 1 ' ' ! CALF^
I
1 1 1 FLOW
1
Minutes
10
IE
fig. 3. Diagrammatic representation of changes in blood flow
in the calf muscles of the human leg during strong rhythmic
contraction. [From Barcroft & Dornhorst (19).]
resting value. At speeds of from 4 to 8 mph there was
no further increase in the size of the peak post-
exercise flow, but the flow returned to the resting rate
more and more slowly as the speed increased.
These opening paragraphs recall the circulatory
changes in muscle that take place during the per-
formance of its most important function — namely,
contraction. The mechanism of the hyperemia of
exercise is not yet understood and it is the most
important problem in this field. Besides dealing with
the hyperemia of exercise this article must refer to
many other matters. For example, we shall have to
deal with the basal tone of the vessels, and with their
nervous regulation, their responses to adrenaline, and
so forth. It will be convenient to refer first to these
general matters, and afterward, with such knowledge
as a background, to return to the central problem of
the hyperemia of exercise.
is not curtailed by previous arrest of the circulation
in the thigh. That is to say, the circulation in the calf
is of no functional significance in tiptoe standing, an
observation in accord with the fact that the gastroc-
nemius soleus arrests its own circulation when stand-
ing on tiptoe (28). During this exercise intramuscular
pressure in the calf does not exceed about 50 mm Hg
(1 19), so that it seems likely that the blood supply to
the muscle is stopped by nipping of its vessels.
The circulation in the calf muscles behaves quite
differently during weak sustained contraction. Then
there is marked hyperemia. The effect of the vaso-
dilatation predominates (28). The behavior of the
circulation during the sustained contraction of other
human muscles also depends upon the force of their
contraction and the extent to which vasodilatation
overcomes the effect of mechanical compression (67,
141, 142, 145).
When human muscles contract rhythmically, each
strong contraction checks the hyperemia (19). This
is shown in figure 3. In running, blood flow through
the calf must be intermittent; free flow through
widely dilated vessels when the muscles are relaxed
must alternate with partial or perhaps complete
arrest of the circulation during contraction.
Black (36) has investigated the effect on the post-
exercise blood flow of walking at different speeds
from 1 to 8 mph. The subject wore a light celluloid
plethysmograph on his calf. The distance covered was
130 yards. Up to 4 mph the size of the immediate
peak postexercise blood flow was directly propor-
tional to the speed. The flow returned rapidly to its
BASAL TONE
Skeletal muscle vessels exhibit a very pronounced
basal tone. In this respect they differ from the vessels
of the skin, or at any rate from the A-V anastomoses
in the skin. Lofving & Mellander (143) found that
the resistance to flow in acutely denervated cat
muscles can be decreased by 80 to 85 per cent by the
close arterial injection of supramaximal amounts of
acetylcholine or ATP; the resistance in the denervated
paw can only be decreased by 20 to 50 per cent.
The action of a circulating vasoconstrictor sub-
stance has often been invoked to explain the strong
basal tone in muscle vessels. If this were so then con-
strictor substances such as noradrenaline, adrenaline,
serotonin, angiotonin, and vasopressin should act
more powerfully in muscle, where basal tone is strong,
than in the skin where basal tone is weak. However
Lofving & Mellander (143) have shown that many
constrictor substances act more powerfully on the
skin vessels of the paw than on skeletal muscle vessels.
They concluded that the basal tone in muscle vessels
cannot be due to the action of adrenaline, noradren-
aline, serotonin, angiotonin, or vasopressin since
muscle vessels did not respond more sensitively to
any of these agents.
Human muscle vessels, too, exhibit strong basal
tone. Vascular resistance in the normal forearm is
about the same as that in the chronically sym-
pathectomized forearm (73) and in both it decreases
to about one-tenth in severe exercise (113). If the
smooth muscle coats of these vessels were to stop
1356
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
contracting spontaneously peripheral resistance and
arterial blood pressure might well fall to a danger-
ously low level. Our very lives must depend upon the
maintenance of basal tone in the vessels of the skeletal
muscular system.
Automaticity
This may be illustrated by experiments on the cat.
Folkow & Lofving (97) recorded the effect of lowering
and then raising the arterial pressure upon the blood
flow through the muscles of the leg and the following
results have been calculated, approximately, from
one of their experiments:
ABP
120
50
50
120
120
F
7
2.6
3-8
15
7
PRU
17
'9
13
8
'7
When the arterial pressure was suddenly lowered
from 1 20 to 50 mm Hg the resistance to flow in-
creased slightly from 1 7 to 19 units, probably because
of elastic recoil of the vessels. Now over the next
few minutes, arterial pressure being still 50 mm Hg,
vascular resistance gradually fell from 19 to 13 units
indicating a gradual reduction in smooth muscle tone.
When the arterial pressure was suddenly restored to
120 mm the resistance fell from 13 to 8 units, due to
stretching of the relaxed vessels. In the course of the
next few minutes the resistance rose again from 8 to
1 7 units, its initial value, indicating a gradual restora-
tion of smooth muscle tone. In short, lowering the
arterial pressure was soon followed by decrease in
basal tone and vice versa.
An even more striking example of automaticity is
shown in another of Folkow's (89) experiments.
Clamping the carotid arteries was followed by a rise
in arterial pressure from 100 to 150 mm Hg; blood
flow in the denervated muscular portions of the hind
parts rose initially but soon returned to its initial
level. In spite of the rise in arterial pressure, the lumen
of the muscle vessels must have decreased.
The explanation of automaticity is not yet com-
plete. Plain muscle responds to stretch by increased
contraction. Bayliss (35) pointed out the significance
of this. In a well-known experiment he recorded
volume changes in the dog's hind leg before, during,
and after splanchnic nerve stimulation. He notes:
"As the arterial pressure rises the limb is distended
passively, but instead of merely returning to its
original volume when the blood pressure has come
down again it constricts much below its previous
level and only gradually returns." He thought this
was probably because the plain muscle of the arterial
walls had responded to stretch by contraction. How-
ever, other factors may be involved. It will be re-
membered that when in one of Folkow's experiments
described above the arterial blood pressure was
raised the flow remained constant and the lumen of
the vessels became smaller. In that case the stimulus
cannot have been simply stretching the vessel walls
if by that is meant a maintained elongation of the
smooth muscle fibers. Nor can the vasoconstriction
have been due to the lowering of metabolite concen-
tration due to more rapid flow — in this experiment
the flow did not increase. Perhaps some of the capil-
lary bed shut down so that the same total quantity
of blood flowed faster through a restricted area.
Further work is needed on the fundamental signifi-
cance of automaticity.
Pressure-flow relations in muscle depend a good
deal on the condition of the animal. As this dete-
riorates in the course of an experiment, automaticity
declines and the effect of alteration in arterial pres-
sure upon muscle blood flow becomes more pro-
nounced.
Automaticity in Hit
Muscle lessels
Experiments by Greenfield & Patterson (115) show
that human vessels constrict when they are stretched.
The forearm was enclosed in a plethysmograph, for
measuring the rate of flow, modified so that pressures
of —50 and — 150 mm Hg could be applied for 30 sec
to the enclosed limb segment. Immediately after the
release of the negative pressure forearm blood flow-
was decreased; the vessels must therefore have con-
stricted. The vessels in the calf respond in the same
way to stretching (60). Blair and others (38) re-
corded the oxygen saturation changes in blood from
the skin and from the muscle. When suction was
applied oxygen saturation rose at once in the blood
from the skin and muscle, due to distention of the
vessels. This can be seen in figure 4. However, by the
end of the first few minutes of suction, oxygen satura-
tion of the blood from both skin and muscle had re-
turned to its initial value or was even less; thus the
vessels had contracted to their initial size or even
smaller. So much for the facts. Since the circum-
ferential size of the vessels was not increased, and
may have been decreased, the authors thought that
the response could not be explained simply by stretch-
ing.
Stretching the vessels of normal, sympathectomized,
and chronicallv denervated forearms bv venous con-
CIRCULATION IN SKELETAL MUSCLE
[357
gestion instead of by suction is also followed by con-
striction of the resistance vessels (157).
Pressure-Flow Relations in Muscle Vessels
Deprived of Automaticity
Folkow & Lofving (97) investigated pressure-flow
relations in maximally dilated muscle vessels in
which automaticity had been abolished by perfusing
100
fig. 4. Results showing that stretching of the forearm
vessels causes contraction. Oxygen saturations of blood samples
from a superficial (O) and a deep (•) forearm vein after general
body heating. During the time represented by the black
rectangle the subject's legs were passively raised. During the
period between the vertical lines the forearm was exposed to a
pressure 50 mm Hg below atmospheric. [From Blair el at.
(39)-]
them with dextran-Tyrode solution. Their results are
summarized in figure 5. Figure 5A (continuous curve)
shows the pressure flow relations when the arterial
pressure was increased stepwise, the venous pressure
being maintained at zero. The curve is convex to
the pressure axis indicating vascular distention as
the pressure increased, until further distention is
prevented by the connective tissue and the develop-
ment of edema. In Figure 5A (broken line) are seen
the results when the mean intravascular pressure
was kept constant at 50 mm Hg, and the perfusion
pressure was increased by increasing the arterial
pressure above and decreasing the venous pressure
below the mean value. The relation between the
perfusion pressure and the flow is then linear. This
is as would be expected since the distending force,
and hence the resistance to flow remain constant at
all values of the perfusion pressure. Finally in Figure
5B is seen the effect of raising arterial and venous
pressures together by equal increments so that the
perfusion pressure remains constant while mean
pressure increases. It will be seen that the greater
the mean pressure the larger is the flow correspond-
ing to a given difference in the perfusion pressure.
This follows because an increase in the mean pres-
sure distends the vessels and decreases intravascular
resistance. When the mean pressure is high the ves-
sels are almost maximally distended and resist further
distention like rigid tubes.
Critical Closing Pressure
When the arterial supply to the cat's muscles was
occluded in an animal in good condition the arterial
pressure did not level out at a value specific for the
FLOW ML/MIN
% INCREASE OF FLOW
-|=90
*580 FjSO
P.30
f=60
P,20
fig. 5. Perfusion of the calf of the cat's
leg with dextran-Tyrode solution. Vessels
maximally dilated. Arterial pressure, Pa;
venous pressure, Pv; perfusion pressure,
p p D Pa + Pv
Pa — Pv = Pn; mean pressure, ' =
2
Pmi. For further explanation see text.
[After Folkow & Lofving (96).]
1358
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
prevailing vascular tone (97), that is to say, not when
the muscle vessels exhibited automaticity. On the
contrary, the arterial pressure always fell to within
a few mm Hg of zero. During the occlusion basal
tone must have decreased, as was manifested after-
wards by reactive hyperemia. Observations in man
confirm this. After arrest of the circulation in the
upper arm, intrabrachial arterial pressure and pres-
sure in the antecubital vein fall progressively till
eventually intra-arterial sinks below intravenous
pressure. This is explained by progressive loss of
intravascular tone, without however the reflux of
venous blood, which is prevented by the action of the
venous valves (48).
Using a pressure plethysmograph, Burton &
Yamada (48) found that the vessels of a segment of
the forearm did close critically after reduction of their
transmural pressure. About half the tissue in their
forearm plethysmograph must have been muscle.
Further work is needed on critical closing pressure in
healthy muscle.
Local Temperature
Not enough is known about the effect of local
temperature on the blood flow through muscle.
Blood flow appears to decrease progressively when the
cat's hind limb with paw tied off is cooled from 40 C
to 25 C. Further cooling is accompanied by increase
in flow which at 10 C generally exceeds that at 40 C
(155). This is not so in the limb that has been treated
with cyanide: blood flow diminishes as the tempera-
ture falls, owing to diminution in the fluidity of the
blood (155).
In man the average forearm blood flow when the
limb was in water at 45 C was 1 7.6 ml per 100 ml per
min. When the water was 13 C forearm flow was
0.5 ml. However, these experiments tell us little
about the effect of local temperature on human
muscle blood flow (20). Forearm blood flow in-
creases as the temperature of the surrounding water
is lowered from 18 C to o C; the dilatation is mainly
in the muscles as it takes place after the circulation
in the skin has been arrested by adrenaline electro-
phoresis. Further work in this lield is needed.
THE PROBLEM OF STRUCTURE AND FUNCTION
The arrangement of blood vessels in striated
muscle was studied by Spalteholtz (173) and de-
scribed as follows by Krogh (134). "The arteries
supplying a muscle branch freely, and between the
branches there are very numerous anastomoses
forming a primary network. Into the meshes of this
net small arteries are given off at regular intervals,
and these again anastomose freely, forming a second-
ary cubical net of great regularity. From the threads
of this network the arterioles branch off, generally
at right angles to the muscle fibers and at very regular
intervals (of about 1 mm in the warm-blooded ani-
mal), and these arterioles finally split up into a large
number of capillaries running along the muscle fibers
and in the main parallel to them but with numerous
anastomoses, forming long narrow meshes about the
fibers. The capillaries unite into venules intercalated
regularly between the arterioles, and the whole
system of veins reproduces and follows almost ex-
actly that of the arteries. All the veins down to the
smallest branches are provided with valves allowing
the blood to flow in the direction of the heart only."
The number of capillaries per square millimeter
transverse section of striated muscle is related to the
metabolic activity of the animal. Krogh found 400
per mm2 of muscle in the cod, 1350 in the horse, 2630
in the dog, and the number in the smallest mammal
he thought would be more than 4000. Assuming a
figure of 2000 for the number of capillaries per square
millimeter of human muscle he calculated that the
total length of all the capillaries in all the skeletal
muscles of a man would be equal to a distance of two
and a half times round the earth, and he estimated
that when all these capillaries were open their surface
area would be 6300 m-\
Certain experimental findings are difficult to
explain on the basis of the classical description of
skeletal muscle vessels. For example, stimulation of
the vasoconstrictor nerves to the dog's hind legs is
accompanied by decrease in muscle blood flow, by
decrease in oxygen consumption of the muscle and
surprisingly by a rise in oxygen saturation of the
venous blood (154). Pappenheimer thought that the
blood must have been directed through A-V shunts
whose surface area available for Oa exchange was
small. Issekutz (127-129) came to the same conclu-
sion. Then, again, increase in muscle blood flow is
accompanied by decrease in oxygen consumption
during sympathetic vasodilator nerve stimulation
(126), but by increase in Oo consumption during
inhibition of sympathetic vasoconstrictor tone (169).
As figure 6 shows, hypothalamic stimulation is accom-
panied by increase in venous outflow from muscle
but the clearance of Xal131 from muscle does not
alter. To explain such results it has been suggested
CIRCULATION IN SKELETAL MUSCLE
!359
2CO
/
ISO
/
% BLOOO FLOW
o
SO
SO
O
CLEARANCE %
200
30
fig. 6. Radio-iodide clearance and blood flow in the
gastrocnemius: Nal131 injected intra-arterially. Results show
that hypothalamic stimulation increases muscle blood flow but
does not alter Nal131 clearance. O — Stimulation of the hypo-
thalamic vasodilator pathway; 5 trials, 4 cats. [From Hyman
el al. (126).]
the stimulation of the vasodilators shifts the flow
from nutritional to nonnutritional (A-V shunt)
channels.
Intravenous infusions of adrenaline increase the
rate of the blood flow through muscle without affect-
ing the rate of Na24 clearance (151). This too has
been attributed to the opening of A-V shunts (18).
A-V anastomoses have been invoked to explain the
circulatory changes in muscle during hypothermia
(61) and to account for the very small A-V 02 dif-
ference in resting muscle (31).
Zweifach (181 J claims that he has seen blood
short-circuitin g through "thoroughfare vessels" in
skeletal musc'e, and Redish et al. (161) have pub-
lished photomicrographs of A-Y shunts in human
skeletal muscle. However, most anatomists deny the
existence of A-V anastomoses in skeletal muscle
(43). The perfusion of skeletal muscles with fluids
containing minute plastic spheres shows, moreover,
that no sphere of 30 /* diameter or over traverses the
denervated gastrocnemius of the dog, though one-
fifth of the stream goes through vessels of 20 n diam-
eter. These vessels may be large capillaries (66, 159).
Barlow et al. (32, 33), having failed to find A-V
shunts in muscle, have suggested another explanation
of the action of adrenaline on the muscle circulation.
They found that muscle contains two entirely sepa-
rate circulations, one to the skeletal muscle fibers,
the other to the connective tissue. According to these
authors adrenaline increases the rate of flow through
the nutritional vessels as is shown plethysmographi-
cally. However, it has little effect upon the rate of
flow in the connective tissue where, in most experi-
ments, the Xa24 is located, so that the rate of Na24
clearance is scarcely altered.
Folkow (93) and Mellander (149) are using the
following scheme for the muscle circulation. After
large "windkessel vessels," which transform pulsatile
into fairly steady flow, come "resistance vessels,"
consisting of two variable sets — precapillary (pre-
dominantly the arterioles) and postcapillary (mainly
the small veins). These vessels determine the resist-
ance to flow and also affect the hydrostatic capillary
pressure and therefore the filtration rate. The "sphinc-
ter" vessels are a specialized section of the smallest
precapillary resistance vessels. These vessels can
cause intermittent closing of the capillaries and they
regulate the size of the capillary surface area ex-
posed to the blood flow and available for blood-
tissue fluid exchange. Then there are "capacitance
vessels" (mainly the veins) in which minor changes
in tone, too small to affect the resistance signifi-
cantly, will have a large effect upon the circulating
blood volume available for the heart. Lastly there
are, of course, the "exchange vessels" or true capil-
laries for the direct exchange of substances between
the blood and tissue fluids; they are devoid of smooth
muscle cells.
Further work is necessary to reconcile the function
of the vascular bed in muscle with its structural ar-
rangement.
NERVOUS CONTROL
Skeletal muscle vessels exhibit strong intrinsic
basal tone and correspondingly weak nervous control.
Their smooth muscle is supplied by sympathetic
vasoconstrictor and vasodilator fibers, though the
belief that these act reciprocally is no longer tenable.
Nor is there at present any convincing evidence that
muscle's sensory innervation has any effect on its
vessels, either by antidromic impulses or by axon
reflexes.
Sympathetic Vasoconstrictor Xerves
These have been found in the cat, dog, hare, mon-
key, and in man, in fact in all mammals so far in-
vestigated [for literature see 29, 92]. The evidence
for their existence in animals is conclusive. It may be
of some interest to refer briefly to the proof of their
presence in man (21, 29). It is as follows. The rate
of the blood flow in the upper muscular parts of both
forearms was measured plcthysmographically and
>36°
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
found to be equal. Radial, median, and ulnar nerve
blocks were performed in one arm, above the elbow,
and the rate of the blood flow doubled in that limb.
Nerve blocks did not affect the rate of the blood flow
in the sympathectomized forearm. Therefore the
doubling of the flow in the normal forearm must
have been due to blocking sympathetic vasocon-
strictor fibers. Whereabouts were the vessels supplied
by these fibers? Nerve block doubled forearm blood
flow after the circulation in the skin had been arrested
by adrenaline electrophoresis. Therefore the release
of vasoconstrictor tone was deep to the skin, probably
in the skeletal muscles. Other authors have confirmed
this (166). It is interesting to note that brachial
plexus block is followed by an even greater increase
in forearm flow (23). Release of vasoconstrictor tone
in all muscles would increase the circulation through
the skeletal muscular system by 1.5 liter per min — the
increase in severe exercise is far greater, about 20
liter per min (21). Sympathetic vasoconstrictor tone
is also present in the resting muscle of the cat (13),
dog (10), and man (21), and probably in the muscle
vessels of all other mammals.
EFFECT OF SYMPATHETIC VASOCONSTRICTORS UPON
RESISTANCE, BLOOD VOLUME, AND CAPILLARY FILTRA-
TION in skeletal muscle vessels. So far we have seen
that the vasoconstrictor fibers in muscle can increase
the resistance to flow. They can also reduce capillary
filtration and increase venous constrictor tone. A
beautiful preparation shown in figure 7 has been
developed by Mellander (149) for simultaneous re-
cording of these effects. These studies will be briefly
described. They concern the effects of stimulation
of the abdominal sympathetic chain upon the cir-
culation in the hind parts of the cat — almost the whole
of the cat distal to the fifth lumbar vertebra. This
part of the cat consisted of skin, muscle, and bone in
the proportions of 1:4:1. The circulation through
bone could be neglected and the hind parts could be
regarded as a "combined skin-muscle region," a
few experiments with rather similar results were
made on skinned hind parts or "muscle regions."
Figure 8 shows a typical tracing. The arterial in-
flow pressure was maintained constant, by means
of a screw clip, at 120 mm Hg. Atropine was given
to exclude the action of the vasodilator fibers. The
abdominal sympathetic was stimulated for periods of
1 min, indicated by the signal marker, at the dif-
ferent frequencies shown on the tracing. The tracing
also shows the corresponding changes in the volume
of the hind parts, which were enclosed in a plethysmo-
graph communicating with a piston recorder. During
stimulation the hind parts shrank rapidly at first
and then more slowly. The initial rapid shrinking,
shown bv an almost vertical downstroke of the lever,
KYMOGRAPH
VOLUME PISTON RECORDER
GADDUM RECORDER
PLETHYSMOGRAPH
SYMP. CHAINS
TO INF. MESENT ART
TO INF. CAVAL VEIN
FROM CAVAL VEIN
HEATING PAD
fig. 7. Preparation used by Mellander to investigate the effect of sympathetic vasoconstrictor nerve
stimulations upon vascular resistance, blood volume and capillary filtration rate in skeletal muscle
vessels. [After Mellander (149).]
CIRCULATION IN SKELETAL MUSCLE 1 36 1
• •
*
Art. blood pressure
•
Frequency of
sJirri/sec
Change of volume g,
ml
-2
-4
-6-
..- *
0.25 0.5 1
2
'4
■»
;z
6
8 16
^7^ 3^
—yr-~
-v - V \(
====2
-8
-10
Colcul.decreose
of capill. pressure
mmHg
Blood flow «oC I
ml/min 0tl
Signal
Time 20 sec ,
2 2 4
*
5 •
3t=
6
Ui
■
1 1
10 15
1
™m
fig. 8. Typical tracing obtained by Mellander using the preparation shown in fig. 7. For further
explanation see text. [After Mellander (149).]
was a measure of the decrease in the volume of blood
in the hind limbs, mainly due to contraction of the
venules. The subsequent slower shrinkage, indicated
by the dotted sloping lines, was a measure of the
rate of loss of tissue fluid. Below the record of the
volume changes are shown the corresponding changes
in capillary pressure; these were calculated after the
experiment and will be referred to later. Below this
again we see the changes in the venous outflow from
the hind parts, the rate of flow is proportional to the
height of the record. At the beginning of the experi-
ment the venous pressure was adjusted, by raising
or lowering the venous outflow cannula, so that the
volume of the hind parts remained constant. It
will only be necessary to consider the change in the
circulation produced by sympathetic nerve stimula-
tion at 2, 8, and 16 impulses per sec. Typical results
are shown in table i . The data obtained from the
tracing are shown by the figures in italic, namely
the arterial blood pressure (line i); the capacitance
changes (initial rapid shrinkage, line 8); the rate of
loss of tissue fluid (slow continuous shrinkage, line
12) and the rate of the venous outflow (line 2).
It will be convenient to consider first the changes in
resistance due to the effect of the vasoconstrictors in
the precapillary and postcapillary vessels, that is
mainly on the arterioles, and to a smaller extent on
the venules. Maximal vasoconstriction was produced
by stimulation at frequencies of 16 per sec or more.
The outflow decreased from 60 to 10 ml per min
(line 2), a reduction of 50 ml (line 3) which can be
regarded as 100 per cent maximal (line 4). This
corresponds to a 6-fold increase in resistance (line 5),
an increase of from 2 to 12 P.R. units (line 6).
However the maximum possible physiological
impulse frequency, as we shall see, is probably not
more than 6 to 10 per sec. The changes recorded
using a frequency of 8 per sec are therefore of par-
ticular interest. The blood flow from the hind parts
was reduced from 60 to about 13 ml (line 2), a reduc-
tion of about 47 ml (line 3), a response corresponding
to g4 per cent of the maximal (line 4) and to a four-
and-a-half-fold increase in resistance (line 5). These
figures correspond to a reduction in the rate of flow
in the hind parts from 8.5 ml per 100 ml hind part
per min to 1.85 ml per 100 ml per min (hind-part
volume 700 ml).
We must now refer to the effects of stimulating the
vasoconstrictor nerves upon the volume of blood
in the vessels of the hind parts. The maximal effect
was obtained at a frequency of 8 per sec. At this
frequency, 6.25 ml were expelled by venous contrac-
tion (table 1 : line 8). According to this the amount of
blood that could be expelled by venous contraction
from the whole of the skin and the entire skeletal
muscular system, tissues weighing half as much as
the whole body, would be only 4.5 per cent of the
animal's blood volume. But of course it must be
1362
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
table i . Changes in the Circulation and Tissue Fluid
Volume in the Hind Limbs of the Cat During
Stimulation of the Sympathetic Vasoconstrictor Nerves
Before
Stimu-
lation
After 1 min Stimulation at
the Following Frequences
2
8
Max
Physiol
16
Max
1 ABP, mm Hg
120
1 20
120
120
2 Blood flow from hind parts,
60
40
'3
IO
ml/min
3 Decrease in blood flow,
20
47
5°
ml/min
4 Decrease in flow as % max
40
94
IOO
decrease
5 Increase in resistance
«x
4]2
6
6 Resistance, PRU
■2
3
9
12
7 Percentage shortening of
35
plain muscle in resis-
tance vessels
8 Decrease in blood volume
O
5
ea
e\i
of hind parts, ml
9 Decrease in blood volume
80
IOO
100
of hind parts as % of max
10 Blood loss as % of total
25
33
33
blood in hind parts
1 1 Percentage shortening of
20
plain muscle in venules
and veins
12 Slow decrease in volume of
O
2
5
6V2
hind parts, ml
13 Decrease in capillary BP,
5
10
15
mm Hg
14 Approximate capillary BP,
24
]9
14
9
mm Hg
remembered that skin and resting muscle are rela-
tively avascular tissues and only contain 14 per cent
of the blood volume. Measurements of the volume
of blood in the hind parts, by a radioisotopic method,
showed that they contained about 20 ml of blood.
Of this, 6.25 ml, that is 33 per cent, was expelled by
sympathetic activity. If the sympathetic could expel
33 per cent, the same proportion, from the splanchnic
area which contains a relatively enormous amount
of blood, it is clear that this mechanism would be of
great importance.
But let us return to table 1 and to Mellander's
results. It will be seen that at 2 per sec hind-part
blood volume is reduced by 80 per cent of the maxi-
mal (line 9) and that resistance is not reduced by 80
per cent until the impulse frequency has been in-
creased to 8 per sec (line 4).
To account for the changes in resistance and hind-
part blood volume found during maximal electrical
stimulation Mellander has calculated that the inter-
nal circumferences of an "average arteriole" and an
"average venule" would have to decrease by 35 per
cent and 20 per cent, respectively. This would happen
if the smooth muscle coat in both arterioles and ven-
ules shortened by 20 per cent. In the case of the
arteriole, owing to the protrusion inwards of the
inner wall layers (99), this would reduce the internal
circumference not by 20 per cent but by about 35
per cent.
Table 1, line 12, shows the effect of sympathetic
chain stimulation on transcapillary fluid movement.
During stimulation tissue fluid entered the capillaries
and drained away, the amount being related to the
impulse frequency. The greater the impulse frequency
the more must capillary pressure have fallen. The
precapillary vessels must have constricted both
absolutely and relatively more than the postcapillary
vessels. The discrepancy must have increased as the
frequency increased. The falls in capillary pressure
corresponding to the different impulse frequencies
were determined as follows. In a control experiment
venous pressure was decreased by a known amount
by lowering the venous cannula, and the rate at
which fluid drained from the tissues out of the hind
parts was recorded. From this the rate at which fluid
entered the capillaries per 1 mm drop in capillary
pressure was calculated. This was the absorption
coefficient. Knowing both this and the rate of entry of
fluid into the capillaries recorded during the stimula-
tion at the different impulse frequencies, the cor-
responding falls in capillary pressure could be cal-
culated. These are shown in lines 13 and 14 and in
the tracing in figure 8.
Another interesting point is that at the end of 2-min
stimulation the absorption of tissue fluid ceases.
Nevertheless during maximal physiological sympa-
thetic stimulation for 2 min the volume of tissue fluid
draining out of the hind parts is almost as much as
that expressed from the capacitance vessels (lines 8
and 12).
Folkow & Mellander (98) have developed a tech-
nique for investigating the effect of a procedure upon
the capillary surface area. Maximal stimulation of
the sympathetic vasoconstrictors, they find, closes
many precapillary sphincters and reduces the capil-
lary surface area to about one-third under conditions
where blood flow is decreased to about one-sixth.
CHEMICAL TRANSMISSION AT SYMPATHETIC VASOCON-
STRICTOR NERVE ENDINGS IN SKELETAL MUSCLE. Folkow
CIRCULATION IN SKELETAL MUSCLE
'363
I - SYMPATHETIC CHAIN
DIVIDINO PLANE
fig. 9. Preparation used for investi-
gating the effect of stimulation of the
baroreceptors on muscle blood flow.
[After Folkow et al. (101).]
& Uvnas (ioq, 104) have shown that the transmitter
is probably noradrenaline. It is not adrenaline. Proof
of this was obtained in cats given Dibenamine. Other
procedures excluded the action of the vasodilator
fibers. After giving Dibenamine, stimulation of the
vasoconstrictors caused only weak contraction or
none at all. Injections of noradrenaline likewise
caused weak contraction or had no effect. On the
other hand, injections of adrenaline caused marked
vasodilatation in the muscle. From such results
Folkow and Uvnas concluded that the vasocon-
strictor nerve endings in the muscles of the cat (102)
and dog (104) might have released noradrenaline
but they had not released adrenaline. For a proper
account of these beautiful experiments and for the
literature, their papers should be consulted. Nor-
adrenaline has not yet been positively identified in
the venous effluent collected from a muscle vein
during vasoconstrictor nerve stimulation.
EFFECT OF STIMULATION OF THE ARTERIAL BARO-
RECEPTORS ON SKELETAL MUSCLE VESSELS IN THE DOG.
Folkow et al. (101) have shown that the sympathetic
vasoconstrictor fibers are solely responsible for medi-
ating the baroreceptor reflex. Their proof is as follows.
Figure 9 shows the preparation of the hind parts of
one dog (the recipient) which were perfused from
another dog (the donor); changes in blood pressure
and in hormone concentration in the upper part of
the recipient's body could not affect the circulation
in its hind legs. The venous outflow from the hind
legs, mainly from the muscles, was recorded, as
was that from an area of the hind-leg skin. The re-
sults are seen in figure 10. Reduction of the blood
pressure in the recipient's carotid sinuses, by carotid
occlusion, caused vasoconstriction in both muscle
and skin. After section of the abdominal sympathetic
nerves neither carotid occlusion nor stimulation of
the carotid sinus .terve had any effect whatsoever.
Dorsal root fibers could not have been implicated.
They could have mediated vasodilatation, as acetyl-
choline injections did. And they were still in good
physiological condition because vasodilatation was
recorded in the skin when the dorsal roots were stimu-
lated (101). In other experiments the vasoconstrictor
action of the abdominal sympathetic chains was
blocked by Dibenamine. Clamping the carotids no
longer caused vasoconstriction in the legs. Although
the sympathetic vasodilator pathway remained in-
tact there was no sign of reciprocal innervation. On
the other hand, vasoconstriction in the legs following
carotid occlusion was normal after the dilator fibers
had been blocked by atropine (104). Folkow and his
colleagues concluded that the effect of stimulation of
the arterial baroreceptors on the blood flow in muscle
must be mediated solely by inhibition of activity in
the sympathetic vasoconstrictor fibers.
EFFECT OF STIMULATION OF THE ARTERIAL BARORE-
CEPTORS ON THE CIRCULATION IN HUMAN SKELETAL
1364
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
mmHg.
FIG. 10. Results obtained with the
preparation shown in fig. 9. Sympathec-
tomy abolished the action of the carotid
sinuses upon the skin and muscle of the
hind limb. [After Folkow et al. (101).]
Note that an increase in saphenal blood
flow records downward.
DONOR
RECIPIENT
SAPHENAL
BLOOD FLOW
CAVAL OUTFLOW
SIGNAL
TIME 6OSEC
1.5: OCCLUSION COMMON CAROTIDS
2.4.6.8: STIMULATION RIGHT SINUS NERVE
3.7.9: INTRA-ARTERIAL INJECTION 0-2>ig ACh
muscle. Folkow and others (54) recorded the effect
of stimulation of the carotid sinus nerve in five pa-
tients during block dissections of the neck performed
for the treatment of cancer. Stimulation at 40 per
sec elicited maximal effects. Mean blood pressure
and pulse amplitude fell promptly, there was a slight
increase in forearm flow, implying considerable
vasodilatation, which was probably of nervous origin.
Nevertheless it is unlikely that arterial pressure
changes in the carotid sinuses in man have much
effect on the sympathetic tone in human muscle
vessels for the following reasons. Stretching the ca-
rotid sinuses by applying subatmospheric pressure to
the outside of the neck causes bradycardia and fall
in arterial pressure — signs of stimulation of the
baroreceptors — but vascular resistance in the fore-
arm is unaltered (81). Compression of the carotid
arteries, followed by fall of the carotid arterial pres-
sure to 20 mm Hg causes tachycardia, hyperpnea,
and rise in brachial arterial pressure — signs of de-
creased baroreceptor activity — but forearm vascular
resistance is unaltered (163). Although strong stimula-
tion of the carotid sinus nerve causes reflex vaso-
dilatation in human muscle quite large changes in
the transmural pressure in the carotid sinuses do not
seem to have any effect on the vascular resistance
in human muscle.
EFFECT OF RECEPTORS IN A LOW PRESSURE AREA IN
THE CARDIO-PULMONARY SYSTEM ON THE SYMPATHETIC
VASOCONSTRICTOR TONE IN HUMAN SKELETAL MUSCLE.
The evidence for this important reflex is as follows
(39, 164, 167, 168). Raising the legs of a recumbent
subject increases the forearm blood flow. It does not
have this effect in the sympathectomized forearm.
The dilatation is reflex. Raising the legs after the cir-
culation in them has been arrested has no effect
upon forearm blood flow. The reflex is elicited by a
shift of blood from the legs into the trunk. Raising
the legs has scarcely any effect on arterial blood
pressure We have already seen that in man neither
stretching the carotid sinuses (81) nor reducing the
blood pressure in them (163) affects the tone of blood
vessels of the forearm — so it seems very unlikely
that their discharge frequency would be affected
by the very small change in arterial pressure which
follows raising the legs. On the other hand, raising
the legs increases the central venous pressure. It
seems then reasonable that rise in pressure on the
venous side stretches structures in the cardiopul-
monary system and so stimulates low pressure re-
ceptors which reflexly increase forearm blood flow.
This conclusion is supported by the finding that blood
flow in the normally innervated forearm increases
when the thoracic vessels are stretched by negative
pressure breathing (39). When the thoracic contents
are repetitively stretched by rapid alternating positive
and negative intrathoracic pressure changes forearm
blood flow is trebled or quadrupled (163).
The low pressure receptors act reflexly by altering
sympathetic vasoconstrictor tone in the muscles.
This has been deduced from the following observa-
CIRCULATION IN SKELETAL MUSCLE
'365
uons. Raising the legs increases forearm blood flow
but it has no effect on blood flow in the hand, which
has very little muscle. The increase in forearm flow
is accompanied by increase in oxygen saturation of
the blood draining from the forearm muscles but
there is no change in the oxygen saturation of the
blood draining from the forearm skin. Therefore
the vasodilatation in the forearm must be in the mus-
cles. As mentioned previously, the reflex is mediated
by the sympathetic nerves as it is absent in sympa-
thectomized forearms and after deep nerve block.
Atropinization of the forearm does not weaken it.
Therefore vasodilator fibers do not seem to be impli-
cated, and it is probably due to decrease in the dis-
charge frequency in the vasoconstrictor fibers (167).
This low pressure receptor reflex in skeletal muscle
vessels may function so as to reduce the effect of
alterations in venous pressure upon the arterial
blood pressure. The effect of an increase in venous
pressure and cardiac output on arterial pressure
would be reduced because of reflex vasodilatation in
the muscles. Conversely reflex constriction in skeletal
muscles would tend to maintain arterial blood pres-
sure after a fall in venous pressure and cardiac output.
After major operations forearm blood flow is de-
creased for several days (1 10). This may be due to a
low pressure receptor reflex induced by hemorrhage
and decreased venous pressure. The importance of
low pressure receptor reflexes in man may be related
to the upright posture.
IMPULSE FREQUENCY IN SYMPATHETIC VASOCONSTRIC-
TOR fibers. Folkow (90) has also investigated the
impulse frequency in the vasoconstrictor fibers to
skeletal muscle — postganglionic C-fibcrs. He con-
cluded that whereas in somatic fibers the maximum
discharge frequency may reach 50 per sec the maxi-
mum frequency in the vasoconstrictors to muscle
hardly ever exceeds 6 to 10 per sec, and normal
sympathetic tone is maintained at a discharge fre-
quency of only about 1 per sec. The experiments
forming the basis of this statement are most elegant.
The preparation is seen in figure 1 1 and a typical
result in figure 12. Atropine was given to block the
action of the vasodilator fibers. Venous outflow from
the isolated muscles of one cat's leg was recorded
before, during, and after i-min periods of stimulation
of the abdominal sympathetic chain at frequencies
increasing stepwise from 0.5 to 20 per sec. As the
frequency was increased the reductions of the blood
flow became greater and were maximal at a fre-
quency of 16 per sec. After stimulations at frequencies
BLOOD REINFUSED
STIMULATION
OF
SYMPATHECTIC
VENOUS DRAINAGE
OF MUSCLES
OCCLUDE CAROTIDS
BLOOD PRESSURE
PARTIAL OCCLUSION
OF AORTA TO KEEP
PERFUSION PRESSURE
CONSTANT
BLOOD PRESSURE
fig. 1 1 . Preparation used by Folkow to investigate the
impulse frequency in sympathetic vasoconstrictor fibers. [After
Folkow (90).]
below 7 or so per sec the blood flow returned rapidly
to its initial value, but its restoration became more
and more delayed after stimulations at progressively
higher frequencies. Folkow then turned his attention
to the venous outflow from the leg muscles of the
opposite side of the cat, the side on which the abdomi-
nal sympathetic chain was still intact; maximal
physiological stimulation of the vasoconstrictor fibers
(only) was induced by clamping both carotid and
vertebral arteries. The vagi had been cut. The possi-
bility of adrenaline secretion had been eliminated
and arterial inflow pressure into the leg was kept
constant by tightening a screw clip on the abdominal
aorta. The reduction in flow during carotid and
vertebral occlusion and the rate of its restoration
afterward were then compared with the reductions
in flow and subsequent rates of restoration that had
been obtained during and after electrical stimulation
of the peripheral end of the cut abdominal sympa-
thetic chain. The comparison showed that maximal
physiological stimulation of the vasoconstrictors (by
occlusion of the arterial supply to the head) caused
changes in muscle blood flow closely resembling those
recorded during and after electrical nerve stimulation
at 6 to 8 per sec. Blood flow during stimulation at 6
to 8 per sec was reduced by 80 per cent of the maxi-
mal reduction recorded during maximum electrical
stimulation at 16 per sec.
[ ;M. HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
BLOOD PRESSURE
fig. 12. Results obtained with the
preparation seen in fig. i i showing that
maximum physiological frequency in vaso-
constrictor fibers to the leg muscle is about
6-8 per sec. [After Folkow (90).] I, 2,
carotid occlusion; ;j, sympathectomy. 4.
0.5 stim./sec; 5, ! stim./sec; 6, 2 stim./
sec; 7, 4 stim./sec; 8, 6 stim./sec; 9, 10
stim. sec; 10, 15 stim./sec; II, 20 stim./
sec. For further details see text.
mmHg
190
150
SIGNAL
TIME 30SEC
BLOOD
FLOW
mmHg
180
140
Sympathetic Vasodilator Nerves
Folkow & Uvnas (102) noticed that after Dibena-
mine, stimulation of the abdominal sympathetic
chain caused marked vasodilatation in the cat's
hind limbs. The reason was sought in a later paper
(103). What kind of fibers had they been stimulating?
Were they true sympathetic fibers? To decide this
they extirpated dorsal root ganglia Lj to L5 on one
side and allowed 10 to 14 days for the sensory fibers
to degenerate. However, this did not diminish the
vasodilator response in the legs during stimulation
of the abdominal sympathetic chain. From this and
other elegant experiments they concluded that the
increase in blood flow must be due to stimulation of
none other than sympathetic vasodilator fibers.
That being so they then investigated the where-
abouts of this vasodilatation in the hind parts.
Stimulation of the abdominal sympathetics did not
increase flow in the saphenous vein draining the skin,
on the other hand the flow from the vena cava drain-
ing skinned hind parts, with the paws tied off, did
increase very greatly. They concluded that the in-
crease must have been in the muscles. Other investi-
gators had shown that the skeletal muscles of the dog
are supplied by sympathetic vasodilator fibers (47).
CHEMICAL TRANSMISSION AT SYMPATHETIC VASODILATOR
nerve endings in skeletal muscle. These are cho-
linergic. Folkow et al. (96) showed this in cats given
Dibenamine to block the action of sympathetic vaso-
constrictor fibers. They noticed that the vasodilator
response to stimulation of the abdominal sympa-
thetic chain was much reduced by atropine. Atropine
did not reduce the vasodilator action of adrenaline
or of histamine. If the abdominal sympathetic vaso-
dilator fibers were cholinergic, they argued, the
vasodilatation should be potentiated after inactiva-
tion of cholinesterase by eserine. It was difficult to
test this because the combination of Dibenamine and
eserine caused almost maximal vasodilatation in the
hind legs. Positive results were obtained in only a
few experiments. Nor could they test the effect of
the venous effluent on the eserinized leech muscle
because there were no leeches in Sweden. However
the results of tests made with extracts of the venous
effluent upon the cat's blood pressure and upon the
frog's rectus muscle showed beyond any doubt that
sympathetic vasodilator nerve stimulation did re-
lease acetylcholine. Folkow & Uvnas (105) could
find no evidence for the existence of adrenergic
vasodilators to muscle vessels in the cat. In the dog
too these vasodilator fibers are cholinergic (46).
ACTIVATION OF SYMPATHETIC VASODILATOR FIBERS
TO SKELETAL MUSCLE BY HYPOTHALAMIC STIMULATION.
Eliasson et al. (78) were the first to show that stimula-
tion of the hypothalamus activated the sympathetic
vasodilator fibers to muscle blood vessels. These
fibers must have been solely responsible as the re-
sponse was abolished by minute doses of atropine or
by section of the abdominal sympathetic chains.
Figures 1 ; and 14 show the preparation and a typ-
ical result. As hypothalamic stimulation caused con-
striction in the skin and intestines, tachycardia,
constriction of the spleen, and dilatation of the pupils,
they thought that activation of the vasodilator fibers
to the skeletal muscles must be part of the reaction of a
state of emergency in which a sudden increase in
muscle blood flow is often needed for muscular activ-
ity. Further studies have since been made on the
central connections of these fibers (2-4, 79, 139, 140).
SYMPATHETIC VASODILATOR FIBERS TO HUMAN SKELETAL
muscles. Observations on man suggest that these
CIRCULATION IN SKELETAL MUSCLE
1367
LIGATURE
fig. 13. Preparation used for investigating the effect of
hypothalamic stimulation in the skin and skeletal muscle of the
dog's hind limb. [After Eliasson el at. (78).]
fibers exist and are activated during fainting and
emotional stress. During fainting, the vasovagal
syndrome, induced experimentally by hemorrhage,
the arterial blood pressure falls precipitously but
blood flow in the forearm increases. There must be
marked vasodilatation in the forearm (24). This
vasodilatation is absent in sympathectomized fore-
arms and is mediated by sympathetic fibers. It is
probably in the skeletal muscles, although this has
not yet been examined with the Hensel needle, or by-
observations of the changes in oxygen saturation
of the blood draining from the deep forearm veins,
or by inducing faints in subjects after arresting most
of the circulation in the forearm skin by adrenaline
electrophoresis. Is the vasodilatation due to inhibition
of sympathetic vasoconstrictor tone or to activation
of sympathetic vasodilator fibers? It is difficult to
devise a satisfactory experiment to decide which is
responsible. During the faint the average blood
flow in six nerve-blocked forearms was less than that
in six normally innervated forearms. Therefore it
seems likely that vasodilator fibers were activated
(22). However, in the cat simple inhibition of vaso-
constrictor tone causes fall in arterial blood pressure
accompanied by increase in muscle blood flow (Fol-
kow, personal communication).
It is worth noting that the vasodilatation in muscle
in fainting is probably large enough to be mainly
responsible for the fall in blood pressure and so for
loss of consciousness (24).
Vasodilators to human muscle are probably acti-
vated in emotional stress. Wilkins and Eichna found
that calf blood flow increased when a subject was
given a mental arithmetic problem which took him
about 15 sec to solve. They thought that this vaso-
dilatation was mediated both by the sympathetic
nerves and by adrenaline secretion (179). Others
have studied the effect on forearm blood flow of
harassing subjects with mental arithmetic problems
for several minutes. They have shown with the Hen-
sel needle that the vasodilatation is in the forearm
muscles (44,83, 111), and that the response is re-
duced, though not abolished by atropine, so that
it is probably mediated to some extent by activity
in vasodilator fibers (16, 42).
Blair et al. (37) frightened subjects by telling them
BLOOD PRESSURE
160
mmHg 140
I20
MUSCULAR
BLOOD FLOW
LEFT HIND LIMB
SIGNAL
TIME 6OSEC
CUTANEOUS
BLOOD FLOW
RIGHT HIND LIMB
SIGNAL
TIME 60 SEC
fig. 14. Results obtained with the preparation shown in fig.
13. Stimulation of part of the hypothalamus caused vaso-
dilatation in the skeletal muscles and vasoconstriction in skin of
the hind limb. Section of the lumbar spinal cord did not abolish
these effects which were mediated by the sympathetic chains.
i368
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Minutes
fig. 15. Results showing that active cholinergic vasodilator
nerves to human muscle contribute to the vasodilatation in the
forearm muscles during stress. Open circles: hand blood flow.
Solid circles : forearm blood flow. During the time represented by
the rectangle it was suggested to the subject that he was suffering
from severe blood loss. [Blair et al. (39).]
that they were suffering from severe blood loss.
In one experiment, the result of which is illustrated
in figure 15, forearm blood flow rose from 8 to 50
ml per min, while hand blood flow was not affected.
In another subject oxygen saturation of blood drain-
ing from muscle rose from 20 to 65 per cent. In six
subjects the vasodilator responses to a wide variety
of stimuli were found to be reduced by atropinization
of the forearm. They concluded that activation of
cholinergic vasodilator nerves to human muscle con-
tributed to the vasodilatation in the forearm muscles
during stress.
Do Posterior Root Fibers Affect Muscle Blood Flow?
There is no important evidence of any efferent path-
way via the posterior roots to animal or human muscle
vessels. These fibers certainly play no part in the
arterial baroreceptor reflex which is mediated solely
by sympathetic vasoconstrictor fibers (101), and
they play no part in hypothalamic vasodilatation
which is mediated solely by sympathetic vasodilator
fibers (78). In man, too, sympathectomy of the limbs
completely abolishes all known vascular responses
of central origin.
There remains the question of whether or not
axon reflexes from sensory endings in muscle influ-
ence the vessels. If so, then stimulation of the posterior
roots should cause "antidromic" vasodilatation in
muscle. Celander & Folkow (56) investigated the
effect on paw flow and muscle flow of stimulation
of the peripheral cut ends of L5-S2. There was
marked vasodilatation in the paw but no effect on
the circulation in muscle. Nor was there any change
in the flow through muscle when the small C-fibers
were selectively stimulated by heating the sciatic
nerve. They concluded that axon reflexes in muscle
were of very little significance.
Effect of the Temperature-Regulating Center
on the Circulation in Muscle
It is well known that rise in body temperature re-
leases sympathetic vasoconstrictor tone in the paws.
However the temperature-regulating center has
very little influence on the circulation in muscle.
Folkow et al. (100) heated the cat's hypothalamus by
diathermy and recorded marked cutaneous vaso-
dilatation, but there was no change in the venous
return from the skinned hind parts. In man, Edholm
et al. (75) recorded marked increase in flow in the
forearm during body heating, but this was absent
in the opposite forearm in which the cutaneous cir-
culation had previously been arrested by adrenaline
electrophoresis. This has been confirmed by observa-
tions of muscle flow made with the Hensel needle
(15) and by measurements of the oxygen saturation
of blood obtained from veins draining muscle (165).
Body heating which causes sweating and rise in
mouth temperature does not increase blood flow in
skeletal muscle.
Role of Sympathetic Fibers to Muscle in Exercise
Gaskell (108, iog) at first thought that vasodilator
nerves were responsible for the vasodilatation In
muscle in exercise, but later he realized that the ac-
tion of metabolites was more important (107). There
is strong evidence that the hyperemia of exercise is
due to the action of a local mechanism which is
triggered by the process of contraction. For example,
Hilton (120) and others showed that the muscular
contractions and vasodilatation elicited by motor
nerve stimulation are both completely abolished by
CIRCULATION IN SKELETAL MUSCLE
[369
20
FOREARMS
• • Normal
c
E
I]
"e
O
O
V
I 10
- \
5
0
_J
Li.
\
D
O
O
_J
m
O
1
10
Minutes after exercise
fig. 16. Results showing vasodilatation in normal and
sympathectomized muscle after exercise. Ordinate gives
increase in blood flow above normal level. [After Grant (113).]
curare. Curare does not paralyze vasodilator nerve
endings so that the vasodilatation must have been
due to the contractile process. Muscles given atro-
pine, and then stimulated, vasodilate quite normally
in spite of cholinergic vasodilator nerve block (11,
106). Dogs are normally active after extirpation of
both sympathetic chains (53) and human beings with
sympathectomized limbs walk and cycle and take
all forms of normal exercise. Before lumbar sympa-
thectomy a policeman ran 380 yards in 65 and 61
sec; 99 days after sympathectomy he did it in 60
and 61.5 sec (H. Barcroft and J. S. Paddle, un-
published observation). Figure 16 shows that in a
sympathectomized forearm the blood flow rose 19
ml per 100 ml per min after clenching a bar hard for 1
min, in a normal forearm the blood flow rose only
18 ml (113). Such findings show that the sympathetic
vasodilators were not responsible for muscle vaso-
dilatation in any of these activities.
However, there is no doubt that in exercise sympa-
thetic impulse discharge to skeletal muscular system
may alter. Blair el al. (41) recorded blood flow in
both forearms and harassed a subject to do his best
to exercise one of them in which voluntary movement
had been paralyzed by a curare-like substance.
The subject's strenuous efforts were accompanied
by vasodilatation in both his forearms. As this was
equal in the two sides, they concluded that the spe-
cific vasodilator fibers to a specific muscle group are
not activated during activation of the motor nerves
to the group in question.
In other experiments the circulatory changes in
the forearm were recorded while subjects, who were
recumbent, performed bicycling exercises with their
legs. During these exercises arterial blood pressure
rose and forearm blood flow decreased so that vascu-
lar resistance in the forearm increased. This vaso-
constriction was still present when the cutaneous
nerves were blocked but it was absent after deep
nerve block. It was mediated by the sympathetic
fibers to muscle vessels. Since it was not affected by
previous atropinization of the forearm it must have
been due to activation of the vasoconstrictor fibers.
There then is a paradox. The vasodilators were
activated when the subject tried hard to exercise his
paralyzed forearm, but it was the vasoconstrictors
that were activated during the bicycling experiments.
Can this be explained as follows? In the bicycling
experiment the vasoconstriction was probably a
manifestation of generalized vasoconstriction of the
resistance and capacitance vessels, involving the
splanchnic area too, and providing blood for the
large increase in output necessary to supply the active
legs. In this exercise the effect of activation of the
vasodilator fibers may have been overpowered by
much stronger activation of the constrictors. On the
other hand, when an emotionally stressed subject
begins exercise the combined actions of the vaso-
dilator nerves and the local factor would be expected
to cause more than usually rapid vasodilatation in
his active muscles.
ACTION OF SYMPATHOMIMETIC SUBSTANCES
Noradrenaline
Given intra-arterially, in animals or man, nor-
adrenaline constricts muscle vessels in all effective
doses (27, 29, 55, 59, 94, 117, 178). Given intrave-
nously in animals its constrictor action may be over-
come by the rise in blood pressure; if this is prevented
(59) or obviated (55) the muscle vessels constrict.
In man, at the beginning of an intravenous noradren-
aline infusion, there may be a transient vasodilatation,
and after this the flow settles down at about the initial
rate for the rest of the infusion period; reflex vaso-
dilatation of sympathetic nervous origin usually
masks noradrenaline's direct constrictor action
(25, 178)-
Adrenaline
The literature contains numerous references to the
effect of adrenaline on the circulation in skeletal
'37°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
muscle (see 1 16, 147). Many of the results are difficult
to interpret. Artificial perfusion pumps, often used in
these studies, must have damaged the blood and,
because of the release of vasodilator substances,
basal tone in the muscle vessels may be weakened
(91). About the use of artificial pumps Folkow (91)
says "The present experiments indicate that slight
interference with the blood supply may damage the
blood cells with release of substances that consider-
ably depress the tone of the vascular smooth muscles
and their reactions to different types of stimuli.
The mere passage of normal arterial blood through a
pump device of the type generally used in the per-
fusion experiments releases these substances in con-
centrations big enough to depress the vascular tone.
The rougher the handling of the blood the bigger
will their effect* be .... The substance (or substances)
is contained in the blood cells, probably in the eryth-
rocytes, and is rapidly destroyed while passing the
lungs, even when present in big concentrations ....
It must be a very potent vasodilator agent, as the
erythrocytes of only 0.5 mm3 blood contain amounts
enough to elicit a well-defined vasodilatation ....
All these characteristics are typical also for ATP.
"It should be stressed, that most blood pump
devices are very unsuitable for a study of the reactions
of the blood vessels, as their vascular smooth muscles
rapidly loose their tone and reactivity to most kinds
of influences due to the fact that big amounts of
vasodepressor agents are then released from the
formed elements of the blood. . . ." These remarks
apply to the cat (Folkow, personal communication).
Besides being observed during pump perfusion,
the action of adrenaline was usually studied after
the hormone had been given rapidly by single
injection, so that there was not enough time for the
resulting action on the vessels to reach a steady
state.
Dale & Richards (63, 64), in two classic papers,
showed that small doses of adrenaline cause vaso-
dilatation in the denervated muscles of the cat's
hind limb.
Clarke (57, 58) gave adrenaline by intra-arterial
infusion. His records of the venous outflow from the
skinned limb of the cat show that it had a biphasic
effect — vasodilatation followed by vasoconstriction.
The vasodilator effect has been attributed to libera-
tion of acetylcholine (174, 175), but this has been
denied. Celander (55) recorded the changes in
venous outflow from the denervated muscle of the
cat's hind leg. Close intra-arterial infusion of adren-
aline at 0.04 fig per kg per min had no effect (fig. 1 7 A) ;
0.07 fig per kg per min caused a large transient
dilatation accompanied by a small sustained one
lasting till the end of the infusion (fig. 17C); 0.13
fig per kg per min caused the initial transient vaso-
dilatation followed, in this case, not by sustained
vasodilatation but by sustained constriction (fig.
175). Whether the initial transient vasodilatation
was followed by small sustained vasodilatation or by
sustained vasoconstriction was a matter of dosage. As
to the explanation of this paradox Celander says:
"It is hard to conceive that the direct effect of /-adren-
aline on the smooth muscle cells of the muscular blood
vessels at a low dosage should be relaxation while the
same substance on the same substrate at a higher dos-
age would bring about a constriction. It seems more
reasonable to assume that the dilator action of
/-adrenaline is an "indirect" one and that its disap-
pearance at a higher dosage of /-adrenaline is related
to the 'direct' constrictor action of /-adrenaline. In
that case the dilatation would be due to the mobiliza-
tion of a vasodilator factor released by /-adrenaline
in the surrounding; skeletal muscle cells with a sec-
ondary influence on the smooth muscles of the blood
vessels." The author thought that the sustained vaso-
dilatation was a phenomenon which should be looked
upon more as a "metabolic" action of adrenaline
than as a direct "motor" action. He also thought
BLOOD
PRESSURE
MUSCULAR
BLOOD FLOW
SIGNAL
TIME-30SEC
fig. 17. Effects on muscular blood flow of /-adrenaline given
intra-arterially. Perfusing blood pressure 120 mm Hg. Body
weight 2.5 kg. A: I -A infusion /-adrenaline 0.04 jig/kg/min.
B. I -A infusion /-adrenaline 0.13 jig/kg/min. C: I-A infusion
/-adrenaline 0.07 jig/kg/min. For further details see text.
[From Celander (55).]
CIRCULATION IN SKELETAL MUSCLE
■37'
that the "after-dilatation" seen after the larger infu-
sions (fig. 1 -]B) might well be due to the action of
some carbohydrate metabolite diffusing slowly from
the skeletal muscle cells. This substance was perhaps
lactic acid as Lundholm (144) had suggested.
Celander (55) also recorded the changes in venous
outflow during intravenous adrenaline infusions.
Arterial pressure in the cat's legs was kept constant by
adjusting a screw clip on the lower abdominal
aorta. The general picture was the same — initial
transient vasodilatation followed by smaller sus-
tained vasodilatation or by constriction according
to the infusion rate. There was one important dif-
ference. Far greater amounts of adrenaline — about
five times as much — had to enter the leg before the
sustained vasodilatation gave place to constriction.
In the case of intravenous infusions the local con-
strictor action of adrenaline was believed to have
been opposed by the vasodilator action of a substance
liberated into the general circulation. Celander
thought this substance was perhaps lactic acid from
the other muscles.
Celander's (55) investigation also included the
changes in venous outflow caused by unilateral
splanchnic nerve stimulation. Here, too, arterial
pressure in the leg was kept from rising by tightening
a screw clip placed proximally. Stimulation at
frequencies of 1 to 6 per sec (corresponding to bi-
lateral splanchnic stimulation at 0.5-3/sec) caused
sustained vasodilatation in the skeletal muscles.
Further increase in the frequency was accompanied
by progressively less vasodilatation. The results may-
have been complicated by liberation of substances
from the liver; but they are interesting because it is
via the splanchnic nerves that the suprarenal gland
receives its natural stimulus.
The human experiments on the biphasic and other
actions of adrenaline on muscle vessels are of par-
ticular interest. To quote Lewis (138), "It is perhaps
impossible to measure the relevant quantities so
precisely in man as in animals that are reduced by
anaesthesia to perfect stillness and control. The
disadvantage is offset, however, in other directions.
It is the reaction in man himself of which we par-
ticularly require knowledge. Moreover, in human
experiments the nutrient fluids bathing the limb are
those natural to the limb and to the reaction, and
this has not always been the case in animal experi-
ments. Our observations are undertaken upon the
unanaesthetised subject, the body as a whole is
healthy and undisturbed, the general circulation is
perfect, conditions rarely, if ever, realised in animal
experiment, and yet probably essential to an eluci-
dation of the full truth where such a delicate reaction
is concerned."
The subject is given a continuous infusion of saline
into the brachial or femoral artery throughout the
experiment. When appropriate the syringe contain-
ing the saline is replaced by another containing the
same saline solution to which adrenaline has been
added. The subject does not know whether syringes
contain adrenaline or not. Thus when adrenaline
is given, changes in forearm or calf blood flow can
safely be attributed to the adrenaline itself; neither
the saline nor emotional stress can be responsible
(29). Soon after the beginning of an intra-arterial
adrenaline infusion blood flow in the muscular part
of the limb increases abruptly, reaching a peak in
about 1 min. From the peak the flow subsides abruptly
to a little above the initial level at which it remains
for the rest of the infusion period. The vessels, as
it were, "yawn" — they open wide and close. This
initial transient vasodilatation occurs at the begin-
ning of infusions at rates varying from about 0.00 1
to 2.0 Mg per min. The biphasic pattern of the re-
sponse is very striking (177). When the rate of the
infusion is increased stepwise, each increase in rate
is accompanied by its own transient initial vaso-
dilatation (29). Intra-arterial infusion of very large
amounts of adrenaline, far above the physiological
range, causes sustained vasoconstriction.
As figure 18 shows, the initial biphasic transient
dilatation is also the first response of the muscle
vessels of man to infusions of adrenaline given by the
intravenous route. The subsequent residual sus-
tained vasodilatation is larger than that recorded
during intra-arterial infusions (29,74, J 77)- Thus
in the forearm an intravenous infusion at 10 jug per
min is accompanied by an initial fivefold increase
in flow after which the rate subsides to about double
the initial value for the remainder of the infusion
period (5,29, 178). That both the initial transient
and the subsequent smaller sustained vasodilatation
take place in the skeletal muscle has been shown by
records taken with a Hensel needle implanted in the
calf muscles (14, 26). This is seen in figure 19.
There is then a close resemblance between the
action of adrenaline on the vessels of the skeletal
muscle of man and of animals. The mechanism of
the initial transient vasodilatation and of the later
sustained one is plainly of great fundamental signifi-
cance. It will be convenient to consider first the na-
ture of the initial biphasic effect which is such a
1372
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
SALINE
ADRENALINE 1 V.
SALINE
IO uq/min
W.J. A
E
Wf
I'5-
jl
E
•1
BLOOD FLOW
5
(
»
2
5 5-
UJ
a.
O
u.
/%^
<
>
v^
V
(
3 :
Tl
WF
IO 1
IN MINUTES
5 20
fig. 1 8. Results show the effect of an intravenous infusion of
adrenaline in man on forearm blood flow. The initial large
transient and later smaller sustained vasodilatation are due to
the action of adrenaline on the blood vessels in the skeletal
muscles.
constant and conspicuous response of the vessels
in the muscles of the calf and forearm.
During the initial vasodilatation blood flow in-
creases about fivefold. The main resistance vessels,
the arterioles and precapillary sphincters must be
widely dilated. This is in accordance with micro-
scopic observations made by Hartman & Walker
( 1 1 8) on the tibialis anticus of the cat. Small doses of
adrenaline dilated arterioles, capillaries, and venules.
In man this dilatation is not dependent on nervous
connections. It occurs after nerve block, after sympa-
thectomy, and in completely denervated limbs (29,
177). It must be due to the local action of adrenaline
in the skeletal muscles. This might be a direct action
on the plain muscle of the arterioles or a metabolic
action due to products of carbohydrate metabolism
released by the action of adrenaline on the skeletal
muscle fibers. That the action of adrenaline is direct
rather than metabolic is shown by the following ob-
servations. In the first place, the initial vasodilatation
is not accompanied by any rise in venous blood
lactate, which may even fall, so it is not likely that
there is a rise in the concentration of anv other
SALINE
ADRENALINE
SALINE
PLETH
NEEDLE
SALINE
ADRENALINE
SALINE
PLETH
NEEDLE
O-
O
IO
A^
SALINE
ADRENALINE
SALINE
PLETH
NEEDLE
A—
5
Sz
8>
CDq
_l -I
O c
UJ
> if) 10
PLETH
NEEDLE
— -— — —
f-
PLETH
NEEDLE
V-
PLETH
NEEDLE
0 <
u
IB
MINUTES
fig. 19. Simultaneous records of the changes in blood flow in calf of the leg (plethysmograph) and
in the muscles of the calf of the leg iHensel needle) recorded in six experiments before, during, and
after the intravenous infusion of adrenaline. [After Barcroft el al. (26).]
CIRCULATION IN SKELETAL MUSCLE
'373
metabolite from the skeletal muscle fibers (17).
Secondly, a perfectly normal initial vasodilatation
has been recorded in a patient whose muscles, owing
to an inborn error of metabolism, contained no
phosphorylase, so that the vasodilatation was prob-
ably not due to any product of glycolysis (146, 170;
also H. Barcroft and B. McArdle, unpublished ob-
servation). It seems reasonable to conclude that the
initial vasodilatation is not due to the action of
adrenaline on carbohydrate metabolism in the
skeletal muscle fibers, and that it probably is due to
the direct action of adrenaline on the plain muscle
of the arteriolar walls.
Now we must consider the second part of the bi-
phasic initial transient vasodilatation — the rapid
return of the flow from the peak toward the resting
level. Allwood & Ginsburg (7) and de la Lande &
Whelan (136) have shown that this is due to a direct
constrictor action of adrenaline on the muscle vessels.
It can be partially or completely prevented by adre-
nergic blocking agents, as is shown in figure 20. That is
to say adrenaline causes first vasodilatation in muscle
almost immediately followed by vasoconstriction,
which can be prevented by a blocking agent. The
question now arises as to whether the vasodilator and
vasoconstrictor phases of the initial transient vaso-
dilatation take place in the same vascular bed, or is
the opening of one set of vessels soon followed by the
closing of another set in parallel? A little considera-
tion of the extent of the changes in flow shows that
the initial vasodilatation and the ensuing vasocon-
striction must both take place in the same set of
vessels. Suppose, as often happens, the flow before
the beginning of the infusion was 3 ml per min.
Then constriction in one bed could not reduce the
flow by more than the preinfusion rate of 3 ml. In
fact typical flows before, at, and after the initial
transient vasodilatation may be 3, 15, 6 ml, respec-
tively. That is, constriction may reduce the flow by
9 ml, i.e., by three times the total preinfusion rate.
Plainly this could only happen if the vessels had first
dilated so that they could be constricted to this
extent.
It is well known that the action of adrenaline on a
piece of smooth muscle can be biphasic (45). It is
highly probable that the vasodilator and constrictor
phases of the initial transient vasodilatation are
both due to a direct biphasic action of adrenaline on
the smooth muscle coat of the arterioles of the skeletal
muscle vessels. This would be in accordance with
the fact that vasodilatation invariably comes before
constriction and that in any given infusion the sizes
M.G.
"E \
i * /im y\ 1
»
§ 10-
y^^^V
\
1
r
0
I
N*.
1 5"
till a *
4W^
*A*
earm
^*^^*+wi
Adrenaline l.V.
For
1 r
►0] 10 //g/min
r 1 r '
0-
1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1
Minutes
fig. 20. Results showing that the rapid return of the forearm
blood flow from the initial peak towards the resting rate during
adrenaline infusions is due to a direct vasoconstrictor action on
the muscle blood vessels. Open circles: before chlorpromazine.
Solid circles: after chlorpromazine. [From de la Lande & Whelan
(136).]
of the vasodilatation and vasoconstriction are usually
equal. If the inhibitory and excitatory actions of
adrenaline were to take place in different parts of
the same vascular bed, i.e., in the arterioles and
venules, it seems less likely vasodilatation would so
closely relate to constriction in both time and extent.
It is very difficult to imagine that the main resistance
could shift from the arterioles to the venules. It is
generally believed that adrenaline does constrict
arterioles.
We must now turn to the smaller sustained vaso-
dilatation that follows the large initial transient one.
Similar sustained vasodilatation in muscle would be
expected to accompany a continuous release of adren-
aline from the adrenal glands. It will be recalled that
Celander (55) observed sustained vasodilatation
during intra-arterial adrenaline infusions and at-
tributed it to the action of lactic acid liberated by the
metabolic action of adrenaline in the skeletal muscle
fibers. Intra-arterial infusions of adrenaline in man
are also accompanied by local release of lactic acid
(6) and the small sustained vasodilatation in such
infusions may be due to an indirect metabolic action
of adrenaline on the skeletal muscle fibers. The evi-
dence so far available shows that the sustained vaso-
dilatation in man is rather larger during intravenous
than during intra-arterial infusions (29, 74, 177).
Celander (55) found the same in animals and thought
that it was because during intravenous infusions the
1374
HANDBOOK OF I'lIVSK ,\
CIRCULATION II
muscle vessels were dilated not only by locally
liberated lactic acid but also by lactic acid liberated
from other muscles into the general circulation.
Many experiments have been done on man to try
to explain why the sustained vasodilatation is larger
during intravenous infusions than it is during intra-
arterial ones. It is not due to rise in arterial blood
pressure, nervous reflexes (29, 177), histamine forma-
tion (153), or secretion of the pituitary (135). It is
probably explained by the fact that during intrave-
nous infusions the concentration of adrenaline re-
mains constant independently of the rate of flow,
whereas the concentration varies reciprocally with the
rate of flow during intra-arterial infusions. The large
sustained vasodilatation typical of an intravenous
adrenaline infusion can be obtained too during an
intra-arterial infusion, if the rate of infusion of the
hormone is gradually increased so as to keep its
concentration constant (143a).
Effect of Adrenaline on the emulation in
Skeletal Muscle During Exercisi
It is often stated that adrenaline dilates muscle
blood vessels in exercise. So far as I am aware there
is no evidence that it actually does so. The idea is
often accepted as part of Cannon's Emergency
Theory of the function of the autonomic nervous
system. So far as I can find, Cannon himself never
suggested it (49-52). On the other hand, why are
the plain muscle coats of the arteries in the skeletal
muscles specialized so that they are rapidly dilated
by adrenaline? Is this of any teleological value in the
cat or man? Or was it of value in some extinct an-
cestor? Or is it just a coincidence?
A few minutes after the beginning of long-lasting
repetitive stimulation of the motor nerve to the dog's
gastrocnemius, the oxygen saturation of the venous
blood draining from the muscle sinks to its lowest
point to rise again later. This is because of delay in
the rate of opening of the vessels. If in the exercising
animal adrenaline secretion helped to open the ves-
sels, the provision of oxygen and disposal of waste
products would be facilitated.
When exercise begins the blood pressure rises,
sympathetic impulse discharge increases, and metab-
olite concentration in the muscle mounts up. It
would be very difficult to devise an experiment in the
cat, dog, or human to determine the extent to which
the initial dilatation of the muscle vessels was in
fact due to the action of adrenaline.
In man, emotional stress at the beginning of exer-
cise may be accompanied by adrenaline secretion
(16). This would be expected to cause an initial
transient vasodilatation throughout the entire skele-
tal muscular system. In active muscles the vessels
would be rapidly dilated and the constrictor phase
of the initial biphasic response might well be blocked
b) the action of the mounting concentration of metab-
olites. In animal experiments the constrictor action
of adrenaline is blocked if muscles are active (150).
In other muscles which were not contracting the
constrictor part of the biphasic response would mani-
festly be of use as it would prevent useless and waste-
ful hyperemia. To quote from August Krogh (134),
"Speculations such as these, though admittedly-
loose, are sometimes very useful. Sooner or later an
opportunity offers of putting them to the test. It is,
of course, very gratifying to find them confirmed, but
generally they are even more useful when they turn
out to be wrong, because, in that case, they serve to
discover at what point the reasoning went astray and
to guide it back into a channel which may possibly
lead it onward. The problems of physiology are so
complicated that, to put it tersely, one cannot expect
to be able to reason correctly from the facts for more
than 5 min at a stretch."
Apart from the beginning of exercise is the question
of the action of adrenaline on the vessels later on
(sustained vasodilator action). It is known that adren-
aline continues to be secreted in severe exercise in
man (82), but because of the very strong action of
metabolites its effect on the vessels would be ex-
pected to be negligible. This is in accordance with
the results of experiments. In dogs, Cannon et al.
(49) found that the amounts of work that dogs
could do to exhaustion on a treadmill was neither
prolonged by previous injection of adrenaline nor
shortened by previous adrenalectomy. In man Dorn-
horst & Whelan (68) found that the postexercise
"blood debt" was not diminished when the exercise
was performed during an infusion of adrenaline.
REACTIVE HYPEREMIA
Reactive hyperemia can be induced in skeletal
muscle vessels. Hilton ( 1 20) recorded it in the cat's
isolated gastrocnemius. Following temporary arrest
of the circulation through this muscle, achieved by
clamping the artery for 30 sec, the increase in blood
flow was as great as that recorded after 30 sec of
maximal tetanic contraction, but after ischemia the
flow subsided more quickly. Folkow & Lofving (97)
CIRCULATION IN SKELETAL MUSCLE
'375
found that temporary occlusion of the circulation
through the cat's hind limb (paw tied off) was
followed by reactive hyperemia. As the condition
of the animal deteriorated, basal tone diminished and
reactivity of the vessels decreased. Reactive hyperemia
in skeletal muscles is soon lost when they are per-
fused with saline.
It is generally agreed that reactive hyperemia
takes place independently of nervous connections.
Bayliss (35) thought that the relaxation of the vessels
during reactive hyperemia was due to lengthening of
the plain muscle because, during the period of ar-
rested circulation, the fibers were no longer subjected
to the stimulus of stretch. Lewis ( 1 38) denied this
because reactive hyperemia followed circulatory
arrest by venous occlusion, during which the smooth
muscle of the arterial walls was still distended by
the arterial blood pressure. He did numerous ex-
periments leading him to the conclusion that the
response was due to the action of a histamine-like
vasodilator substance the concentration of which,
in the tissue fluids, increased during the ischemic
period. Folkow et al. (95) and Emmelin & Emmelin
(80) found that reactive hyperemia occurred quite
normally in the limbs of animals in which the action
of injected histamine had been completely blocked by
antihistaminics. Therefore, they concluded that the
response was probably not due to the action of a
histamine-like substance. Guyton et al. (62) showed
that reactive hyperemia cannot be due to the action
of accumulated CO 2. Ventilating dogs with 20 per
cent CO 2 was not accompanied by any vasodilatation
in the legs. On the other hand, reduction of the
oxygen saturation of the blood to 30 per cent doubled
the rate of the blood flow. Guyton et al. thought that
oxygen deficiency might well be one of the causes of
reactive hyperemia.
Most studies of reactive hyperemia in muscle in
man have been made in the forearm or calf using
venous occlusion plethysmography. However, one
must alwavs remember that the blood flows recorded
by this method are not those in the skeletal muscle
only, but include also the blood flow through the
skin.
In man as in animals (120), the longer the period
of arrest lasts the greater is the subsequent hyperemia
in the forearm ( 1 58) ; the increase is mainly in the
duration of high flows, the peak value being relatively
little increased. Reduction of the arterial pressure
during the period of arrest and of the stimulus of
stretch may be partlv responsible for the loss of
vascular tone (156). Exposing the forearm to sub-
atmospheric pressure and thus '"packing" it with
blood before arresting the circulation lessens the fall
of intravascular pressure during the period of oc-
clusion. As in animals, so in the human forearm,
antihistamines, such as tripelennamine, mepyramine,
and antazoline when introduced into the brachial
artery do not diminish the reactive hyperemia that
follows 3 min of circulatory arrest, although they
completely abolish the increase in flow brought
about by the intra-arterial injection of histamine
(70-
Apart from histamine, various chemical causes,
such as anoxia, have been suggested to explain reac-
tivity in man. Lewis (138) says, "'It is manifest that
neither deficiency of oxygen nor an accumulation
of carbon dioxide or other weak acid in the blood
that is within the vessels can possibly form the direct
stimulus; were that so the reaction would always be
fleeting, the blood being at once replaced by the flood
of the reactive hyperaemia." Certainly the vessels
would be filled with fresh blood almost instantly, but
does it follow that the reaction would be fleeting?
After sudden removal of the stimulus how fast in
fact would the vessels contract? Some kinds of plain
muscle respond to stimulation rather slowly. Further
experiments are needed on this important point.
McNeill (148) showed that during the second minute
of a reactive hyperemia in the forearm the oxygen
saturation of the venous blood in the antecubital
vein may rise transiently to well above the resting
value. The effect is seen in figure 21. The reason for
this transient rise in venous oxygen saturation, as
McNeill showed, was that oxygen consumption re-
turned to the resting level more promptly than did
the blood flow; this corresponded to a transient de-
crease in utilization. Return of the blood flow to the
pre-occlusion level may have lagged behind restora-
tion of the circulation a) because of the presence of a
nonoxidizable metabolite, or b) because the vessels
simply could not contract fast enought to keep pace
with the rapid fall of the concentration of some
oxidizable vasodilator metabolite. Further work is
necessary on this topic.
Dornhorst & Whelan (68) showed that, after a
short period of arrest of the circulation in the calf,
the rate at which the blood flow during the subse-
quent reactive hyperemia returns to initial levels is
exponential; i.e., a straight line is obtained when log
flow is plotted against time (compare fig. 24Z?). The
significance of this fact is not clear. In other experi-
ments, using a pressure plethysmograph, they studied
the effect on reactive hyperemia of reduction of the
'376
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
100
1 2 3
Minutes
fig. 2i. Results showing that after circulatory arrest
(vertical lines) the oxygen saturation in the venous blood rises to
above the initial value. Oxygen consumption (open circles)
returns to the resting level more quickly than the blood flow
(filled circles). [From McNeil (148).]
arterial pressure. The resting rate of flow, post-
ishemic peak and "area under curve" or "blood
debt" were all much reduced, but this was not so
for the exponential rate of restoration of the flow.
They concluded that the flow could not depend on
the local concentration of some metabolite, the re-
moval of which depended critically on the rate of the
blood flow. They say "the facts were compatible with
the concept of a metabolite diffusing out of the tissues
with a concentration gradient which effectively
limits its rate of removal when the blood flow is
above some small value, or of a metabolite oxidized
at a rate dependent on its concentration and inde-
pendent of the local oxygen tension when this ex-
ceeds some low figure." The significance of this
finding is not yet known. Another relevant experi-
ment was performed by Blair et al. (41). They studied
the blood flow in the forearm after 5 min circulatory
arrest. Coincident with the release of the circulation
the brachial artery was compressed digitally for 5
min to prevent the blood flow from rising above the
resting level. Release of the artery was not followed
by any reactive hyperemia. They concluded that it
was not necessary to have an increase in blood flow
alter circulatory arrest to "repay" the "debt" in-
curred during this procedure. It would be interesting
to compare the oxygen debt incurred during the
circulatory arrest with the subsequent oxygen repay-
ment. So far as I am aware this has not been done,
because of technical difficulty. [However, see (180).]
It is certainly worth recalling that reactive hy-
peremia in totally denervated forearms is just as
great as in normal ones (72). Indeed, owing to the
withering of the muscles the response per 100 ml is
far greater in the denervated forearm. Can reactive
hyperemia be due to ischemia of the plain muscle of
the arterial tree?
EXERCISE HYPEREMIA
Active muscles must get oxygen from the air. To
this end total ventilation increases, heart rate is
speeded up, and muscle blood vessels dilate. We
still do not know for certain the mechanism of any
of these responses. The hyperpnea that we can see
and the tachycardia that we can feel and record have
attracted more attention than the deeply hidden
dilatation of the muscle vessels. We still seem a long
way from understanding the cause of the hyperemia
of exercise. I recall some sentences of my father's.
"Let us then jot down such information as is forth-
coming in the hope that the points at issue may be
taken up one by one by future workers, and that one
day systematic work may be done on the subject.
I say 'jot down' rather than 'put together' because
to make any sort of story from such unsatisfactory
material would be quite unwarrantable" (30).
Let us first "jot down" some points about the
hypothesis that exercise hyperemia is caused by
anoxia of the vascular tree. It is certainlv worth
noting that the rate at which muscle blood vessels
open may be as fast after simple arrest of the circu-
lation as during exercise. Eichna & Wilkins (77)
found that the peak forearm flow after 5 min of simple
circulatory arrest was as large as that after 5 min
combined circulatory arrest and rhythmic exercise.
Dornhorst & Whelan (68) recorded a peak flow of
about 20 ml per 100 ml calf per min after 2 min
ischemia; after 2 min rhythmic exercise the same post-
exercise peak flow was recorded. Hilton (120) noted
that the flow from the cat's gastrocnemius after 30
sec ischemia was the same as that after 30 sec exercise.
True, the mechanisms of ischemia and exercise are
not really comparable, but further work is necessary
to see if ischemia opens the vessels as fast alone as
when combined with exercise. If so, we must ask
CIRCULATION IN SKELETAL MUSCLE
'377
whether the smooth muscles of the arterial tree
relax simply because their oxygen is taken away
by the active skeletal muscle fibers.
The experiments of Kramer and his colleagues
(131-133, 160) have helped a great deal to establish
the effect of exercise on the circulation and metab-
olism of muscle. The dog's gastrocnemius muscle was
stimulated indirectly via its nerve. The rate of its
venous outflow was recorded continuously by an
optical method in milliliters per minute as also were
the arterial blood pressure and the oxygen saturations
of both arterial and venous blood; in some experi-
ments blood lactate was estimated. From these, the
relations between work done, blood flow, oxygen
consumption, and lactate output were calculated.
Figure 22 is from an experiment of Kramer's in
which the sciatic nerve was stimulated maximally
at 310 impulses per sec for 1 sec every alternate sec.
The findings are relevant:
/) At the beginning of exercise venous blood
oxygen tension fell abruptly to reach a "low" after
about 1 min. Blood flow rose exponentially to reach
a steady value in about 1 min. If the resting vessels
were opened wide by acetylcholine then oxygen
usage jumped up to the steady state as soon as the
exercise began.
These facts may be interpreted as follows. As soon
as exercise begins there is an immediate demand for
oxvgen, the supply of blood being quite inadequate,
tissue oxygen tension falls to a very low level. This is
reflected in the low oxygen saturation of the venous
blood, and possibly also in the gradual relaxation
of the plain muscle of the arterial tree. As vasodilata-
tion proceeds and oxygen supply improves venous
oxygen saturation rises somewhat.
2) During the steady state, blood flow, work done,
and rate of oxygen consumption are linearly related.
This is seen in figure 23. During submaximal exercises
the muscle gets all the oxygen it wants (or almost).
Opening the vessels still more with acetylcholine
does not increase the oxygen consumption. It is not
clear from these exeriments whether the blood flow
was linearly related to the decrease in venous blood
oxygen tension. It seems very significant indeed that
the rate of the blood flow is linearly related to the
rate of oxygen consumption. Further work is needed
to see how it is related to tissue oxygen tension.
3) Immediately after moderate exercise extra
oxygen usage stops in a muscle the vessels of which
are opened maximally with acetylcholine. When the
circulation is normal, oxygen consumption and blood
flow rapidly subside exponentially, oxygen consump-
z
o
cr
3
%, ARTERIAL
VENOUS
-10
ML/MIN
5
Q
□
O
o
1
1
ML
'MIN
1
1
1
ill
1
1
1
1 J
'
1 1
!
1 1
1 1
1 V
1
/
1
_
MINUTES
fig. 22. Results obtained by Kramer and his colleagues.
Top: records from which blood flow, oxygen saturation, and
oxygen consumption were obtained. Rhythmic stimulation
between the vertical lines. Note the remarkable rise in the
venous oxygen saturation after the end of exercise. Bottom: re-
sults calculated from records like those shown above. [After
Kramer et at. (132).]
tion falling a little more rapidly than blood flow.
The behavior of the venous O2 saturation is interest-
ing. Immediately after exercise stops, it rises tran-
siently to a peak and then subsides again to a low
level from which it recovers only very slowly. The
reason for the immediate postexercise peak may be
as follows. The demand for oxygen being soon satis-
fied, the arterial tree is no longer anoxic and its
plain muscle starts to contract. But the rate of con-
traction cannot keep pace with the fall off in demand
for O-i. Hence O2 saturation rises. It is not clear
why, later on, the postexercise venous blood oxygen
saturation subsides from the peak to a level almost
as low as in exercise, and from which recovery to the
pre-exercise level takes place only very gradually.
Since at this time the rate of the blood flow is de-
creasing, it is plain that blood flow cannot be in-
versely related to venous blood oxygen tension. To
be able to explain this odd finding would be to gain
much insight into the mechanism of exercise hyper-
emia.
i3?8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
140
ML/MIN
I20
IOO
5
o
80
8 6°
40
20
IO
ML/MIN
8
z
O
E 6
io 4
Z
o
1
N
X
••• J x«
o
• ^X
IN
y^''
0
s%x
o
^
•
% ^^
O 25 SO 75 IOO%
WORK PERFORMED AS °/c MAXIMUM
fig. 23. Results showing that blood flow and oxygen con-
sumption in the dog's gastrocnemius muscle were proportional
to the work done during rhythmic contraction. [From Kramer
el al. (132).]
So much for Kramer's experiments. It is worth
recalling that Barger el al. (31) in their experiments
on the regulation of the circulation in exercise in
the normal clog also concluded, though from less
direct evidence, that muscle blood flow and oxygen
consumption were directly related.
It is worth noting that in hyperthyroid subjects
the postexercise blood flow following a standard
exercise is much increased (1). Here again an ex-
planation might take us to the very center of the
mechanism of exercise hyperemia.
But now we must "jot down" some results which
make it difficult to picture a relationship between
tissue oxygen tension and blood flow. Let us begin
with experiments by Dornhorst & Whelan (68).
They studied the effect of reduction of femoral arterial
blood pressure on postexercise hyperemia in the calf
muscles. The effective arterial pressure could be
halved by means of a pressure plethysmograph.
Figure 24 shows a typical result. After the arterial
pressure was reduced by one-half, the postexercise
hyperemia flows and "blood debt" were lowered but
the hyperemia was not prolonged. The changes in
the peripheral vascular resistance of the calf were
exactly similar to those found in the limb with normal
blood pressure. Thus, if postexercise hyperemia
depended on the local concentration of a metabolite,
its removal or destruction could not depend on the
rate of blood flow nor could its oxidation depend upon
local oxygen tension. They concluded that for post-
exercise hyperemia, as for reactive hyperemia (see
above), the vasodilatation could be due to an intra-
cellular metabolite the removal of which from the
tissue of the vessel wall is limited more criticallv by
its diffusion gradient than by the rate of blood flow
through the vessel lumen. Such a metabolite, they
suggested, would be deactivated primarily by intra-
cellular oxidation.
Holling & Verel (125) studied the effect of lowering
the effective arterial pressure on the forearm blood
flow. Arterial pressure was lowered by elevating the
forearm. There was a linear relation between pressure
and flow. Peripheral vascular resistance did not
change. Compensation for reduction of perfusion
pressure by vasodilatation did not occur. Oxygen
tension in the elevated forearm was reduced as
shown by the polarograph, but oxygen uptake was
unaltered because of increased utilization. Their
findings did not support the concept that metabolism
of resting muscle played a prominent role in the
regulation of its blood flow. In line with this, Blair
et al. (40) showed that compression of the brachial
arterv for 5 min beginning at the end of 1 min
rhythmic exercise of the forearm muscles altogether
abolished postcontraction hyperemia. They concluded
that it was not necessary to have an increase in blood
flow after exercise to "repay" the "debt" incurred in
exercise. This agrees with their observations following
postocclusive (reactive) hyperemia (see above).
However, by supplying the tissues with an excess of
blood, postexercise hyperemia does return the tissues
to the resting state more quickly than does the resting
flow.
If oxygen lack of the arterial tree opens the vessels
in exercise, then asphyxiating skeletal muscle should
cause hyperemia. Bayliss (34), however, found that
the blood flow through the denervated hind limb
did not alter when the arterial blood was made
asphyxial. Yerzar (176) studied the effect of ven-
tilating the cat with 8 to 1 o per cent oxygen on the
gaseous metabolism of the gastrocnemius muscle.
Partial asphyxia of the muscle greatly reduced its
rate of oxygen consumption but had no effect on the
rate of the blood flow. The venous blood seldom
became less than 30 per cent saturated with oxygen,
nevertheless tissue oxvgen tension in regions farthest
CIRCULATION IN SKELETAL MUSCLE
'379
60 1 20 O 60
TIME IN SECONDS
I20
fig. 24. Results showing that reduction of the arterial blood
pressure did not affect the resistance changes in the muscle
vessels of the calf of the leg after exercise. Solid circles: average
of 1 2 runs in six subjects with pressure in plethysmograph
raised 67 cm H;0. Open circles: average of 12 control runs on
the same subjects. A: simple scale; B: semilogarithmic scale.
[From Dornhorst & Whelan (68). j
from capillaries could not have been far from zero.
Others have found that venous blood from exercising
muscles seldom contains less than 6 vol per cent
oxygen (12, 31), and that muscle oxygen tension
may be very low. For example, during tetanic
stimulation of the cat's soleus oxygen saturation of the
myoglobin falls from 90 to 50 per cent (152). Rough-
ton has kindly told me that this corresponds to a fall
in muscle oxygen tension of from about 40 to 5 mm
Hg (F. J. W. Roughton, personal communication).
Ehrlich (76) too, who measured the reduction of
alizarin blue, found that muscles reduce very ac-
tively.
But what about the effect of really low arterial
blood oxygen tensions on muscle flow? The trouble
is that these kill the heart. It is true that Fleisch el a/.
(86), who perfused hind legs with blood artificially
ventilated with 3 per cent oxygen, did not find much
dilator effect. But their preparation had lost basal
tone and was probably widely dilated before the
effect of rarefied oxygen was tested. Because of the
damage done to the blood of the cat by pumps (91)
the experimenter who wishes to test the effect on
muscle blood flow of complete reduction of the blood
is faced with a very awkward technical problem. In
dogs Guyton ct al. (62) recorded some vasodilatation
in the hind legs when the oxygen tension of the blood
perfusing them was reduced to 30 per cent. We do
badly need a study of the effect on muscle blood flow
and gaseous metabolism of progressive reduction
in 02 tension of the arterial inflow right down as
far as zero. It will be recalled that Hilton & Eicholtz
(124) found that ventilating the animal with N2
was accompanied by a large increase in coronary
blood flow. Results have not yet appeared on the
effect of ventilating the animal with N2, on the hy-
peremia of exercise, nor on the effect on the hypere-
mia of exercise of reducing arterial oxygen tension
down to zero.
So much for oxygen lack. It was Gaskell's (107)
idea that muscle blood vessels were opened by vaso-
dilator metabolites liberated from the skeletal muscle
fibers. It was he who first painted the arteries of the
frog's mylohyoid with 1:10,000 lactic acid and ob-
served the vasodilatation. However, lactic acid is
probably not responsible for exercise hyperemia.
Exercise is accompanied by the usual hyperemia in
muscles that have been poisoned with moniodoacetic
acid to prevent the formation of any lactic acid
(112, 162). Exercise is accompanied by hyperemia
in patients whose muscles, owing to congenital
absence of phosphorylase, are unable to form lactic
acid (146, 170). And there are awkward differences
between blood flow and lactic acid time relations
(i33)-
Fleisch & Sibul (85) found that neutral lactate
had no dilator effect. That of injected lactic acid
they thought to be due to its pH. Other substances
which caused vasodilatation in concentrations of
Mo to *300 ml Per m' blood were methylglyoxal,
Na-pyroracemic acid, acetaldehyde, acetates, sodium
acetoacetate, salts of fatty acids, and adenosine
phosphate. Their actions were additive. Fleisch &
Weger (88) investigated the action of fructose
1 ,6-diphosphate, dihydroxyacetaone phosphate, phos-
phoglyceric acid, phosphopyruvic acid, phospho-
glycerol, and creatine phosphate on the blood
vessels of the cat's hind leg. All were inactive or
weakly dilator. However, their results were not of
much quantitative value because they used pump
perfusion and worked with a preparation that had
lost basal tone.
Gaskell (107) had suggested that CO 2 might be a
factor and while Bayliss (35) showed that it had a
vasodilator action Krogh (134), Fleisch (86) and
others have shown that its effect is too weak to be of
much significance.
Fleisch & Sibul (85) thought that the additive
effect of carbohydrate metabolites might be con-
siderable because of their reduction of the pH. But
Gollwitzer-Meier's (112) determinations of the pH
changes in the venous effluent of the exercising
gastrocnemius of the dog did not support this hypoth-
1380
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
esis. Moreover, as we have seen, exercise hyperemia
occurs in muscles poisoned with moniodoacetic acid
although their pH probably increases (112).
All workers are agreed as to the great vasodilator
power of ATP and its related compounds in animals
(88, 162, 171) and man (70). And several have sug-
gested that it is concerned with exercise hyperemia.
In fact the amount of ATP in muscle is probably-
reduced in exercise because of conversion to ADP.
As we have seen the rate of the blood flow in muscle
closely parallels the rate of its oxygen usage. But so
far as I know there is no relation between muscle
02 consumption and the rate at which ATP or ADP
leaks out of the muscle fiber.
Dawes (65) suggested that potassium might be
partly responsible for the hyperemia of exercise. This
was based on the finding that intra-arterial injection
of 5000 ^g of potassium into the pump-perfused
muscles of the dog's hind legs caused some vasodilata-
tion. Once more attention is being focused on potas-
sium. Kjellmar (130) found that postcontraction
hyperemia in the cat's gastrocnemius muscle was
accompanied by a rise in the potassium concentra-
tion in the venous effluent. Intra-arterial infusions of
small amounts of potassium caused vasodilatation.
But, and this is important, infusions of amounts far
bigger than those found during exercise cause con-
striction. Tetanus of the cat's muscle during the
potassium-induced constrictor phase was no longer
followed by hyperemia, although injected vaso-
dilator agents induced vasodilator responses. This
suggests that the hyperemia of exercise may be at
least partly due to the action of potassium ions. I do
not know whether the rate at which K leaves the
skeletal muscle fibers is related to their rate of oxygen
consumption. We have seen how closely blood flow
is related to oxygen consumption and to the work
done.
Anrep and others (8, 9, 1 1 ) found that blood from
active muscles contained histamine and that during
contraction the amount of histamine in muscle di-
minished. Fleisch & Weger (87) repeated their
experiments and concluded that the loss of histamine
was due to the fact that the condition of the animals
had deteriorated. In man, active muscles do not
release a vasodilator substance, or if they do it does
not survive a single passage through the lungs. The
performance of leg exercise is not accompanied by
any change in the vascular resistance of the nerve-
blocked forearm (42).
Bradykinin has recently come very much to the
lore. Hilton (123) has shown that it is not implicated
in the mechanism of exercise hyperemia.
In exercise muscle blood flow may increase ten-
fold. Peripheral vascular resistance in the muscle
must have fallen to one-tenth of its normal value and
the principal resistance vessels, usually considered to
be the arterioles, must be widely dilated. This dilata-
tion results from the action of metabolites produced
either in the arterial tree itself or in the surrounding
skeletal muscle fibers or in both. The question arises,
if metabolites from the skeletal muscle fibers are
involved, by what mechanism do they cause relaxa-
tion of the multilayered arteriolar plain muscle
coat? The following points seem relevant:
1) It is not hard to imagine that vasodilator
metabolites from the tissue fluids could quickly
diffuse through the arteriolar walls. A good example
of diffusion through thick tissues is that of a dental
anesthetic, which in a short time seeps from the sub-
cutaneous tissue of the gum through the maxillary
bone into the tooth socket. Diffusion of metabolites
through minute arterioles might well be very rapid
indeed.
2) Schretzenmayr (172) made the curious dis-
covery that contractions of the skeletal muscles in the
lower part of the cat's leg are followed by increase
in the diameter of the femoral artery in the inguinal
region. Figure 25 illustrates this. Since this increase
in diameter was not abolished by denervation, but
was by painting the vessels with phenol, he thought
that it must be an axon reflex from the active muscles
to the arterial walls. Fleisch (84) confirmed this in
the dog and showed that intra-arterial injections of
acetic acid, of various intermediary products of
30 SEC
fig. 25. Exercise of the leg muscles by stimulation of the
sciatic nerve (A, between the arrows) caused dilatation of the
femoral artery proximal to the muscle. After curare neither
stimulation of the sciatic {B) nor of the muscle (C) had any
effect on the diameter of the femoral artery. [From Hilton
(122)-]
CIRCULATION IN SKELETAL MUSCLE
I38.
muscle metabolism, of histamine and especially of
acetylcholine all were followed by widening of the
femoral artery.
Hilton (121, 122) compared the effects of a variety
of drua;s and procedures on postcontraction hyperemia
and on the postcontraction dilatation of the femoral
artery. The actions of drugs on these two processes
resembled each so closely as to suggest strongly that
they had a common mechanism. He noted also that
the dilator response traveled slowly along the wall
of the artery at about 10 cm per sec. Intra-arterial
injections into the muscle of acetylcholine, histamine,
bradykinin, and nicotine were all followed by intra-
and extramuscular arterial vasodilatation. However
intra-arterial injection of ATP causes only intra-
muscular vasodilatation; he did not think that ATP
could be concerned with the dilator response which
accompanies muscular contraction (125).
5) D'Silva & Fouche (69) found that shunting
the blood from the artery to the vein causes widening
of the artery proximally. They think that the dilata-
tion of the artery in exercise may not be due to
metabolites but to a change in the rate of flow.
4) It seems important to bear in mind that muscle
blood flow, work done, and oxygen consumption are
closely related, though we do not understand the
nature of the underlying mechanism.
REFERENCES
1. Abramson, D. I., and S. M. Fierst. Peripheral vascular
responses to exercise in the hyperthyroid state. J. Clin.
Invest. 20: 517, 1941.
2. Abrahams, V. C, and S. M. Hilton. Active muscle
vasodilatation and its relation to 'flight and fright'
reactions in the conscious animal. J. Physiol.. London
140: 16P, 1958.
3. Abrahams, V. C, S. M. Hilton, and J. L. Malcolm.
Sensory input to the hypothalamic and mesencephalic
regions subserving the defence reaction. J. Physiol., London
149: 45P, 1959.
4. Abrahams, V. C, S. M. Hilton, and A. Zbrozvna.
Active muscle vasodilatation elicited by mesencephalic
stimulation. Its relation to the defence reaction. J.
Physiol., London 148: 32 P, 1959.
5. Allen, W. J., H. Barcroft, and O. G. Edholm. The
action of adrenaline on the blood vessels in human
skeletal muscle. J. Physiol., London 105: 255, 1946.
6. Allwood, M. J., and A. F. Cobbold. Lactic acid release
by intra-arterial adrenaline infusions before and after
dibenyline, and its relationship to blood flow changes in
the human forearm. J. Physiol., London 157: 328, 1961.
7. Allwood, M. J., and J. Ginsburg. The effect of dibeny-
line on the vascular response to the sympathomimetic
amines in the forearm. J. Physiol., London 147: 57P, 1959-
8. Anrep, G. V., and G. S. Barsoum. Appearance of
histamine in the venous blood during muscular con-
traction. J. Physiol., London 85: 409, 1935.
9. Anrep, G V., G. S. Barsoum, M. Talaat, and E.
Wieninger. Further observations on the release of
histamine by skeletal muscles. J. Physiol., London 96:
->4°. '939-
10. Anrep, G. V., A. Blalock, and A. Samaan. Effect of
muscular contraction upon blood flow in skeletal muscle.
Proc. Roy. Soc., London, B. 114: 223, 1934.
11. Anrep, G. V., and E. von Saalfeld. The blood flow
through skeletal muscle in relation to its contraction. J.
Physiol., London 85: 375, 1935.
12. Assmussen, E. and M. Nielsen. Cardiac output during
muscular work and its regulation. Physiol. Revs. 35: 778,
'955-
13. Baetjer, A. M. The relation of the sympathetic nervous
system to the contractions and fatigue of skeletal muscle
in mammals. Am. J. Physiol. 93: 41, 1930.
14. Barcroft, H. Action of epinephrine in man. Trans.,
Fourth Conf. on Shock and Circulatory Homeostasis. New York :
Josiah Macy Jr. Foundation. 1954, 9.
1 5. Barcroft, H., K. D. Bock, H. Hensel, and A. H. Kitchin.
Die Muskeldurchblutung des Menschen bei indirekter
Erwarmung und Abkiihlung. Pfliigers Arch. ges. Physiol.
261: 199, 1955.
16. Barcroft, H., J. Brod, Z. Heil, E. A. Hirsjarvi, and
A. H. Kitchin. The mechanism of the vasodilatation in
the forearm during stress (mental arithmetic). Clin. Sci.
19 : 577. i960.
17. Barcroft, H., and A. F. Cobbold. The action of adrena-
line on muscle blood flow and blood lactate in man. J.
Physiol., London 132: 372, 1956.
18. Barcroft, H., and A. C. Dornhorst. Blood flow re-
sponses to temperature and other factors. Ciba Found.
Symp., Peripheral Circulation Man. 1954.
19. Barcroft, H., and A. C. Dornhorst. Blood flow through
the human calf during rhythmic exercise. J. Physiol.,
London 109: 402, 1949.
20. Barcroft, H., and O. G. Edholm. The effect of tem-
perature on blood flow and deep temperature in the
forearm. J. Physiol., London 102: 5, 1943.
21. Barcroft, H., and O. G. Edholm. On sympathetic
vasoconstrictor tone in human muscle. J. Physiol., London
102: 21, 1943.
22. Barcroft, H., and O. G. Edholm. On the vasodilatation
in human skeletal muscle during post-haemorrhagic
fainting. J. Physiol., London 104: 161, 1945.
23. Barcroft, H., O. G. Edholm, C. A. Foster, R. H. Fox,
and R. K. Macpherson. The effect of nerve block on
forearm blood flow. J. Physiol., London 132: 16P, 1956.
24. Barcroft, H., O. G. Edholm, J. McMichael, and E. P.
Sharpey-Schafer. Post-haemorrhagic fainting. Lancet
1:489, 1944.
I382
HANDBIinK OF PHYSIOLOGY
CIRCULATION II
25. Barcroft, H., P. Gaskell, J. T. Shepherd, and R. F.
Whelan. The effect of noradrenaline infusions on the
blood tlow through the human forearm. J. Physiol., 45.
London 123: 443, 1954.
26. Barcroft, H., H. Hensel, and A. H. Kitchin. Compari-
son of plethysmography thermoelectric needle records of 46.
calf blood flow during intravenous adrenaline infusions. J.
Physiol. , London 127: 7 P, 1955.
27. Barcroft, H., and H. Konzett. On the actions of 47.
noradrenaline, adrenaline and isopropylnoradrenaline on
the arterial blood pressure, heart rate and muscle blood
flow in man. J. Physiol., London 110: 194, 1949.
28. Barcroft, H., and J. L. E. Millen. The blood flow 48.
through muscle during sustained contraction. J. Physiol.,
London 97: 17, 1939.
29. Barcroft, H., and H. J. C. Swan. Sympathetic Control of 49.
Human Blood Vessels. Physiological Society Monographs Series
No. 1. London: Arnold, 1953.
30. Barcroft, J. Respiratory Function of the Blood. I. Lessons from 50.
High Altitude. London : Cambridge Univ. Press, 1 925.
31. Barger, A. C, V. Richards, J. Metcalfe, and B.
Gunther. Regulation of the circulation during exercise. 51.
Am. J. Physiol. 184: 613, 1956.
32. Barlow, T. E., A. L. Haigh, and D. N. Walder. Dual 52.
Circulation in skeletal muscle. J. Physiol., London 149:
18P, 1959.
33. Barlow, T. E., A. L. Haigh, and D. N. Walder. 53.
Evidence for two vascular pathways in skeletal muscle.
Clin. Sci. 20: 367, 1 96 1.
34. Bayliss, W. M. The action of carbon dioxide on blood
vessels. J. Physiol., London 26: 32P, 1901. 54.
35. Bayliss, W. M. On the local reaction of the arterial wall
to change of internal pressure. J. Physiol., London 28:
220, 1902.
36. Black, J. E. Blood flow requirements of the human calf 55.
after walking and running. Clin. Sci. 18: 89, 1959.
37. Blair, D. A., E. W. Glover, A. D. M. Greenfield, and
I. C. Roddie. Excitation of cholinergic vasodilator nerves 56.
to human skeletal muscles during emotional stress. J.
Physiol., London 148: 633, 1959.
38. Blair, D. A., W. E. Glover, A. D. M. Greenfield, and 57.
I. C. Roddie. The increase in tone in forearm resistance
blood vessels exposed to increased transmural pressure. 58.
J. Physiol., London 149: 614, 1959.
39. Blair, D. A., W. E. Glover, and B. S. L. Kidd. The 59.
effect of continuous positive and negative pressure
breathing upon the resistance and capacity blood vessels
of the human forearm and hand. Clin. Sci. 18:9, 1 959.
40. Blair, D. A., VV. E. Glover, and I. C. Roddie. The 60.
abolition of reactive and post-exercise hyperaemia in the
forearm by temporary restriction of arterial inflow. J.
Physiol., London 148: 648, 1959. 61.
41. Blair, D. A., W. E. Glover, and I. C. Roddie. Vaso-
motor responses in the human arm during leg exercise.
Circulation Ren; mil i| 264, 1961.
42. Blair, D. A., K. Golenhofen, and W. Seidel. Muscle
blood flow during emotional stress. J. Physiol., London
63.
149: 61 P, 1959.
43. Boyd, J. D. General survey of visceral vascular structure.
Ciba Found. Sym/.., Visceral Circulation. 1952, p. 3. 5,
44. Brod, J., V. Fencl, Z. Heil, and J. Jirka. Circulatory
changes underlying blood pressure elevation during acute
emotional stress (mental arithmetic) in normotensive and
hypertensive subjects. Clin, Sci. 18: 269, 1959.
Bulbring, E. Biophysical changes produced by adrenaline
and noradrenaline. In: Adrenergic Mechanisms. London:
Churchill, i960, p. 275.
Bulbring, E., and J. H. Burn. The sympathetic dilator
fibres to the muscles of the cat and dog. J. Physiol., London
si 483. !935-
Burn, J. H. On vasodilator fibres in the sympathetic,
and on the effect of circulating adrenaline in augmenting
the vascular response to sympathetic stimulation. J.
Physiol., London 75: 144, 1932.
Burton, A. G, and S. Yamada. Relation between blood
pressure and How in the human forearm. J. Appl. Physiol.
4: 329. 1951 ■
Campos, F. A. de M., W. B. Cannon, H. Lundin, and
T. T. Walker. Some conditions affecting the capacity for
prolonged muscular work. Am. J. Physiol. 87: 680, 1927.
Cannon, W. B. The emergency function of the adrenal
medulla in pain and major emotions. Am. J. Physiol.
33 : 356, '9H-
Cannon, W. B. Bodily Changes in Pain, Hunger, Fear and
Rage. New York: Appleton, 1929.
Cannon, W. B. The sympathetic division of the autonomic
system in relation to homeostasis. Proc. Assoc. Research
Nervous Mental Disease 9: 181, 1930.
Cannon, W. B., H. F. Newton, E. M. Bright, V.
Menkin, and R. M. Moore. Some aspects of the phys-
iology of animals surviving complete exclusion of sym-
pathetic nerve impulses. Am. J. Physiol. 89: 84, 1929.
Carlsten, A., B. Folkow, G. Grimby, C A. Hamburger,
and O. Thesuleus. Cardiovascular effects of direct stimula-
tion of the carotid sinus nerve in man. Acta Physiol. Scand.
44 : >38, '958-
Celander, O. The range of control exercised by the
sympathico-adrenal system. Acta Physiol. Scand. 32 :
Suppl. 1 16, 1954.
Celander, O., and B. Folkow. The nature and dis-
tribution of afferent fibres provided with the axon reflex
arrangement. Acta Physiol. Scand. 29: 359, 1953.
Clark, G. A. The vaso-dilator action of adrenaline. J.
Physiol., London 80: 429, 1933.
Clark, G. A. Adrenaline vaso-dilatation in voluntary
muscle. J. Physiol., London 84: 344, 1935.
Cobbold, A. F., and C. C N. Vass. Responses of muscle
blood vessels to intraarterially and intravenously ad-
ministered noradrenaline. J. Physiol., London 120: 105,
'953-
Coles, D. R., B. S. L. Kidd, and G. C. Patterson.
Reactions of blood vessels of the human calf to increase
in transmural pressure. J. Physiol., London 134: 665, 1956.
Corovino, B. G , W. R. Beavers, and D. W. Rennie.
Hindlimb flow during immersion hypothermia. Am. J.
Physiol. 1 87 : 593, 1 956.
Crawford, D. G, H. M. Fairchild, and A. C. Guyton.
Oxygen lack as a possible cause of reactive hyperaemia.
Am. J. Physiol. 197: 613, 1959.
Dale, H. H., and A. N. Richards. The vasodilator
action of histamine and of some other substances. J.
Physiol., London 52: 110, 191 8.
Dale, H. H., and A. N. Richards. The depressor
(vasodilator) action of adrenaline. J. Physiol., London 63:
201, 1927.
CIRCULATION IN SKELETAL MUSCLE
1383
65-
66.
67.
68.
69.
70.
71-
72.
73-
74-
75-
76.
77-
78.
79-
80.
81.
82.
83-
84.
85-
Dawes, G. S. The vaso-dilator action of potassium. J.
Physiol., London 99: 224, 1941.
Dieter, E. Uber das Vorkoramrn arteriovenoser Anasto-
mosen im Skeletmuskel. Pfliigers Arch. ges. Physiol. 258:
47°. '954-
Dolgin, P. and G. Lehmann. Ein Beitrag zur Physiologie
der statischen Arbeit. Arbeitsphysiologie 2: 248, 1930.
Dornhorst, A. C., and R. F. Whelan. The blood flow in
muscle following exercise and circulatory arrest: the
influence of reduction in effective local blood pressure, of
arterial hypoxia and of adrenaline. Clin. Sei. 12: 33, 1953.
D'Silva, J., and R. F. Fouche. The effect of changes in
flow on the calibre of large arteries. J. Physiol., London
150: 23P, i960.
Duff, F., G. C. Patterson, and J. T. Shepherd. A
quantitative study of the response to adenosine triphos-
phate of the blood vessels of the human hand and forearm.
J. Physiol., London 125: 581, 1954.
Duff, F., G. C. Patterson, and R. F. Whelan. The
effect of intra-arterial antihistamines on the hyperaemia
following temporary arrest of the circulation in the hu-
man forearm. Clin. Sci. 14: 267, 1 955.
Duff, F., and J. T. Shepherd. The circulation in the
chronically denervated forearm. Clin. Sci. 12: 407, 1953.
Duff, R. S. Circulatory changes in the forearm following
sympathectomy. Clin. Sci. 10: 529, 1951.
Duff, R. S., and H. J. C. Swan. Further observations
on the effect of adrenaline on the blood flow through
human skeletal muscle. J. Physiol., London 1 14: 41, 1 951 .
Edholm, O. G, R. H. Fox, and R. F. Macpherson. The
effect of body heating on the circulation in skin and mus-
cle. J. Physiol., London 134: 612, 1956.
Ehrlich, F. Das SauerstoJJbediirfniss des Organismus. Berlin:
Hirschwald, 1885.
Eichna, L. W., and R. W. Wilkins. II. Reactive hyper-
aemia: Factors influencing the blood flow during the
vasodilatation following ischaemia. Bull. Johns Hopkins
Hasp. 68: 450, 1941.
Eliasson, S., B. Folkow, B. Lindgren, and B. Uvnas.
Activation of sympathetic vasodilator nerves to the skeletal
muscles in the cat by hypothalamic stimulation. Ada
Physiol. Scand. 23: 333, 1951.
Eliasson, S., B. Lindgren, and B. Uvnas. Representation
of the hypothalamus and the motor cortex in the dog of
the sympathetic vasodilator outflow to the skeletal mus-
cles. Acta Physiol. Scand. 27: 18, 1952.
Emmelin, K., and N. Emmelin. Histamine and reactive
hyperaemia. Acta Physiol. Scand. 14: 16, 1 947.
Ernsting J., and D. J. Parry. Some observations on the
effects of stimulating the stretch receptors in the carotid
artery of man. J. Physiol., London 137: 45P, 1957.
Euler, U. S. von, and S. Hellner. Excretion of nor-
adrenaline and adrenaline in muscular work. Acta Physiol.
Scand. 26: 183, 1952.
Fencl, V., Z. Heil, J. Jirka, J. Madlafousek, and J.
Brod. Changes of blood flow in forearm muscle and skin
during an acute emotional stress (mental arithmetic).
Clin. Sci. 18: 491, 1959.
Fleisch, A. Les reflexes nutritifs ascendents producteur
de dilatation arterielle. Arch, intern, physiol. 41 : 141, 1935.
Fleisch, A., and I. Sibul. Uber nutritive Kreislaufre-
gulierung. IE Die Wirkung von pH, intermidiaren Stoff-
86.
87,
9°-
9'-
92-
93-
94.
95-
96.
97-
99
103
104.
wechselprodukten und andern biochemischen Verbin-
dungen. Pfliigers Arch. ges. Physiol. 231: 787, 1933.
Fleisch, A., I. Sibul, and V. Ponomarev. Uber nutritive
Kreislaufregulierung. I. Kohlensaurc und Sauerstoff-
mangel als auslosende Reize. Pfliigers Arch. ges. Physiol.
230: 814, 1932.
Fleisch, A., and P. Weger. Uber das Auftreten von
gefasserweiternden Substanzcn im Venosen Blut. Pfliigers
Arch. ges. Physiol. 239: 354, 1937.
Fleisch, A., and P. Weger. Die gefasserweiternde Wirk-
ung der phosphorglierten Stoffwechselprodukte. Pfliigers
Arch. ges. Physiol. 239: 362, 1937.
Folkow, B. Intravascular pressure as a factor regulating
the tone of small blood vessels. Acta Physiol. Scand. 17: 289,
1949-
Folkow, B. Impulse frequency in sympathetic motor
fibres correlated to the release and elimination of a trans-
mitter. Acta Physiol. Scand. 25: 49, 1952.
Folkow, B. A critical study of some methods used in in-
vestigations on the blood circulation. Acta Physiol. Scand.
27: 10, 1952.
Folkow, B. Nervous control of blood vessels. Physiol
Revs. 35: 927, 1955.
Folkow, B. The efferent innervation of the cardiovascular
system. Verhandl. deut. Ges. Kreislaufforsch. 25: 84, 1959.
Folkow, B., J. Frost, and B. Uvnas. Action of adren-
aline, noradrenaline and some other sympathomimetic
drugs on muscular cutaneous and splanchnic vessels of
cat. Acta Physiol. Scand. 15: 412, 1948.
Folkow, B., H. Haeger, and G. Kahlson. Observations
on reactive hyperaemia as related to histamine, on drugs
antagonizing vasodilatation induced by histamine, and
on the vasodilator properties of adenosine triphosphate.
Acta Physiol. Scand. 15: 264, 1948.
Folkow, B., K. Haeger, and B. Uvnas. Cholinergic
vasodilator nerves in the sympathetic outflow to the mus-
cles of the hind limbs of the cat. Acta Physiol. Scand. 1 5 :
401, 1948.
Folkow, B., and B. Lofving. The distensibility of systemic
resistance vessels. Acta Physiol. Scand. 38: 37, 1956.
Folkow, B., and S. Mellander. Aspects of the nervous
control of the precapillary sphincters with regard to the
capillary exchange. Acta Physiol. Scand. 75: Suppl. 50, 52,
i960.
Folkow, B., and B. Oberg. The effect of functionally
induced changes of wall/lumen ration on the vasocon-
strictor responses to standard amounts of vasoactive
agents. Acta Physiol. Scand. 47: 131, 1959.
Folkow, B., G Strom, and B. Uvnas. Cutaneous vaso-
dilatation elicited by local heating of the anterior hypo-
thalamus in cats and dogs. Acta Physiol. Scand. 17: 317,
1949-
Folkow, B., G. Strom, and B. Uvnas. Do dorsal root
fibres convey centrally induced vasodilator impulses?
Acta Physiol. Scand. 21:1 45, 1 950.
Folkow, B., and B. LIvnas. The chemical transmission
of vasoconstrictor nerve impulses to the hind limbs and
splanchnic region of the cat. Acta Physiol. Scand. 15: 365,
1948.
Folkow, B., and B. Uvnas. The distribution and func-
tional significance of sympathetic vasodilators to the
hindlimbs of the cat. Acta Physiol. Scand. 15: 389, 1948.
Folkow, B., and B. Uvnas. The chemical transmission
.384
HANDBOOK OF PHYSIOLOGY
I IKI I I \ 1 l( l\ II
"3-
114.
"5-
116.
117.
of nerve impulses to the hind limbs of the dog. Acta Physiol.
Scand. 17: 191, 1949.
105. Folkow, B., and B. Uvnas. Do adrenergic vasodilator
nerves exist. Acta Physiol. Scand. 20: 329, 1950.
106. Ganter, G. Uber die Vorgange im Kreislauf bei der
Arbeit. Arch, exptl. Pathol. Pkarmakol. 138: 276, 1928.
107. Gaskell, W. H. On the tonicity of the heart and blood
vessels. J. Physiol., London 3: 48, 1880.
108. Gaskell, W. H. On the changes of the blood stream in
muscle through stimulation of their nerves. J. Anat. 1 1 :
360, 1877.
109. Gaskell, W. H. On the vasomotor nerves of striated
muscles. J. Anat. 11: 720, 1877.
1 10. Ginsburg, J. The Ejects oj Certain Stimuli on the Peripheral
Circulation m Healthy and Diseased Subjects (Thesis). Oxford
University, 1958.
111. Golenhofen, K., and G. Hildebrandt. Psychische
Einfliisse auf die Muskeldurchblutung. PJlugers Arch. ges.
Physiol. 263: 637, 1957.
112. Gollwitzer-Meier, K. Blood pH and blood flow during
muscular activity. Lancet 1 : 38 1 , 1 950.
Grant, R. T. Observations on the blood circulation in
voluntary muscle in man. Clin. Sci. 3: 157, 1938.
Greenfield, A. D. M. Venous occlusion plethysmog-
raphy. Methods in Med. Research 8: 293, i960.
Greenfield, A. D. M., and G. C. Patterson. Reactions
of the blood vessels of the human forearm to increase in
transmural pressure. J. Physiol., London 125: 508, 1954.
Griffiths, F. R. Jr. Fact and theory regarding the calo-
rigenic action of adrenaline. Physiol. Revs. 31 : 151, 1951.
Gross, F. Periphere Gefasswirkung von Adrenalin und
Noradrenalin. Helve! . Physiol, el Pharmacol. Acta 7.C: 43,
•949-
118. Hartman, F. A., and H. G. Walker. The action of epi-
nephrine upon the capillaries and fibres of skeletal muscle.
Am. J. Physiol. 85: 91, 1928.
119. Henderson, Y., A. W. Oughterson, L. A. Greenberg,
and C. P. Searle. Muscle tonus, intramuscular pressure
and the venopressor mechanism. .4m. J. Physiol. 114: 261,
■936-
120. Hilton, S. M. Experiments on the post-contraction
hyperaemia of skeletal muscle. J. Physiol., London 120:
230. !953-
121. Hilton, S. M. The Mechanism of the Hyperaemia Accompany-
ing Activity m Skeletal Muscle (Thesis). Cambridge Univ.,
1956.
122. Hilton, S. M. A peripheral arterial conducting mecha-
nism underlying dilatation of the femoral artery and
concerned in functional vasodilatation in skeletal muscle.
J. Physiol., London 149: 93, 1959.
123. Hilton, S. M. Plasma kinin and blood flow. Polypeptides
Which Affect Smooth Muscles and Blood Vessels. London:
Pergamon, i960, 260.
124. Hilton, R., and F. Eicholtz. The influence of chemical
factors on the coronary circulation. ./. Physiol., London
59 ; 4i3. i924-25-
125. Holling, H. E., and D. Verel. Circulation in the
elevated forearm. Clin. Sci. 16: 197, 1957.
126. Hvman, C, S. Rosell, A. Rosen, R. R. Sonnenschein,
and B. Uvnas. Effects of alterations of total muscular
blood flow on local tissue clearance of radio-iodide in the
cat. Acta Physiol. Scand. 46: 358, 1959.
127. Issekutz, B. v. Die Wirkung von Gefassmitteln auf den
lokalen Stoffwechsel des Muskels. Arch, exptl. Pathol.
Pharmakol. 197: 313, 1941.
128. Issekutz, B. V. Uber die Wirkung der Gefassmitteln auf
den Kreislauf der Extremitat. Arch, exptl. Pathol. Phar-
makol. 199: 233, 1942.
129. Issekutz, B. v., and M. Harangozo-Oroszy. Die
Wirkung der Sympathikomimetica auf den Gasstoff-
wechsel. Arch, exptl. Pathol. Pharmakol. 201 : 346, 1942.
130. Kjellmar, I. Some aspects of work hyperaemia in
skeletal muscles. Acta Physiol. Scand. 1 75 : Suppl. 50, 85,
i960.
131. Kramer, K, and W. Quensel. Untersuchungen iiber den
Muskelstoffwechsel des Warmebliiters. I. Mitteilung. Der
Verlauf der Muskeldurchblutung wahrend tetanischen
Kontraktion. PHiigers Arch. ges. Physiol. 239: 621, 1937.
132. Kramer, K., F. Obal, and W. Quensel. Untersuchungen
iiber den Muskelstoffwechsel des Warmebliiters. III.
Mitteilung. Die Sauresoffaufnahme des Muskels wahrend
rhythmischer Tatigkeit. PJlugers Arch. ges. Physiol. 241 :
717, 1939-
133. Kramer, K., W. Quensel, and K. E. Schafer. Unter-
suchungen iiber den Muskelstoffwechsel des Warme-
bliiters. IV. Mitteilung. Beziehungen zwichen Saure-
stoffaufhahme und Milchsaureabgabe des Muskels
wahrend der Tatigkeit. PJlugers Arch. ges. Physiol. 241
73°, '939-
134. Krogh, A. The Anatomy and Physiology oj the Capillaries.
New Haven: Yale Univ. Press, 226, 1922.
135. Kitchin, A. H. Observations on the Circulation in Human
Skeletal Muscles (Thesis). London Univ., 1954.
136. Lande, I. S. de la, and R. F. Whelan. The effect of
antagonists on the response of the forearm vessels to
adrenaline. J. Physiol., London 148: 548, 1959.
137. Langley, J. N. Obituary notice of W. H. Gaskell. Proc.
Roy. Soc, London, B. 88: xxvii, 191 4.
138. Lewis, T. The Blood Vessels of the Human Skin and Their
Responses. London: Shaw, 1927.
139. Lindgren, P., and B. Uvnas. Vasodilator responses in
skeletal muscles of the dog to electrical stimulation in the
medulla oblongata. Acta Physiol. Scand. 29: 137, 1953.
140. Lindgren, P., and B. Uvnas. Activation of sympathetic
vasodilator and vasoconstrictor neurones by electric
stimulation in the medulla of the dog and cat. Circulation
Research 1: 479, 1953.
141. Lindhard, J. Untersuchungen iiber statische Muskel-
arbeit. Pt. I. Skand. Arch. Physiol. 40: 145, 1920.
142. Lindhard, J. Untersuchungen iiber statische Muskel
Arbeit. Pt. II. Skand. Arch. Physiol. 40: ig6, 1920.
143. Lofving, B., and S. Mellander. Some aspects of the
basal tone of the blood vessels. Acta Physiol. Scand. 37:
135. 1956-
1 43a. Lowe, R. D., and B. F. Robinson. J. Physiol., London.
In press.
144. Lundholm, E. M. The mechanism of the relaxing effect
of adrenaline on smooth muscle. Acta Physiol. Scand. 2g:
Suppl. 108, 1953.
145. Marschak, M. Eine Untersuchung iiber den Gaswechsel
und iiber Milchsaure und Alkalireserve im Blut bei
statischer Arbeit. Arbeitsphysiologie 4: 1, 1931.
146. McArdle, B. Myopathy due to a defect in muscle
nlvcogen breakdown. Clin. Sci. 10: 13, 1951.
CIRCULATION IN SKELETAL MUSCLE
•385
F. G. Vablecasas.
im rahenden und
ges. Physiol. 237 : 454,
The measurement of
147. McDowall, R. J. S. The Control of the Circulation of the
Blood. London: Longmans, Green, 1938.
148. McNeill, T. A. Venous oxygen saturation and blood
flow during reactive hyperaemia in the human forearm.
J. Physiol., London 134: 195, 1956.
149. Mellander, S. Comparative studies on the adrenergic
neuro-humorai control of resistance and capacitance
blood vessels in the cat. Acta Physiol. Scand. 50: Suppl. 176,
i960.
150. Mertens, O., H. Rein, and
Gefasswirkung des Adrenalins
arbeitenden Muskel. Pfliigers Arch.
■936-
151. Miller, H., and G. M. Wilson
blood flow by the local clearance of radioactive sodium.
Brit. Heart J. 13: 227, 1951.
152. Millikan, G. A. Experiments on muscle haemoglobin in
vivo; the instantaneous measurement of muscle meta-
bolism. Proc. Roy. Soc, London, B. 123: 218, 1937.
153. Mongar, J. L., and R. F. Whelan. Histamine release by
adrenaline and D-tubocurarine in the human subject.
J. Physiol., London 120: 146, 1953.
154. Pappenheimer, J. R. Vasoconstrictor nerves and oxygen
consumption in the isolated perfused hind-limb muscles of
the dog. J. Physiol., London 99: 182, 1940.
155. Pappenheimer, J. R., S. L. Eversole, and A. Soto-
Rivera. Vascular responses to temperature of the isolated
perfused hind limb of the cat. Am. J. Physiol. 155: 458,
1948.
156. Patterson, G. C. The role of intra-vascular pressure in
the causation of reactive hyperaemia in the human fore-
arm. Clin. Sci. 15: 17, 1 956.
157. Patterson, G. C, and J. T. Shepherd. The blood flow
in the human forearm following venous congestion. J.
Physiol., London 125: 501, 1954.
158. Patterson, G. C., and R. F. Whelan. Reactive hyper-
aemia in the human forearm. Clin. Sci, 14: 197, 1955.
159. Piiper, J., P-W. Schneider, and W. Schoedel. Kurz-
schlussdurchblutung. Klin. Wochschr. 540, 1954.
160. Quensel, W., and K. Kramer. Untersuchungen iiber
den Muskelstoffwechsel des Warmebliiters. II. Mitteilung.
Die Saurestoffaufnahme des Muskels wahrend der tetani-
schen kontraktion. Pfliigers Arch. ges. Physiol. 241 : 698,
'939-
161. Redish, W., F. F. Tangco, and K. L. de C. H. Saun-
ders. Peripheral Circulation in Health and Disease. New York :
Grune & Stratton, 1957, 132.
162. Rigler, R. Uber die Ursache der vermerhten Durch-
blutung des Muskels wahrend der Arbeit. Arch, exptl.
Pathol. Pharmakol. 167: 54, 1932.
163. Roddie, I. C, and J. T. Shepherd. The effect of carotid
artery compression in man with special reference to
changes in vascular resistance in the limbs. J. Physiol.,
London 139: 377, 1957.
164. Roddie, I. C, and J. T. Shepherd. Receptors in the
high pressure and low pressure vascular systems. Lancet
I: 493. '958-
165. Roddie, I. C., J. T. Shepherd, and R. F. Whelan.
Evidence from venous oxygen saturation measurements
that the increase in forearm blood flow during body
heating is confined to the skin. J. Physiol., London 134:
444, I956-
166. Roddie, I. C, J. T. Shepherd, and R. F. Whelan.
The vasomotor nerve supply to the skin and muscle of
the human forearm. Clin. Sci. 16: 67-74, '957'
167. Roddie, I. C, J. T. Shepherd, and R. F. Whelan.
Reflex changes in vasoconstrictor tone in human skeletal
muscle in response to stimulation of receptors in a low
pressure area of the intrathoracic vascular bed. J. Physiol.,
London 139: 369, 1957.
168. Roddie, I. G, J. T. Shepherd, and R. F. Whelan.
Reflex changes in human skeletal muscle blood flow
associated with increased intrathoracic pressure change.
Circulation Research 6: 232, 1958.
169. Rosell, S., and B. Uvnas. Vasomotor control of oxygen
consumption in skeletal muscle. Acta Physiol. Scand. 175:
Suppl. 50, 129, i960.
170. Schmid, R., and R. Mahler. Chronic progressive myop-
athy with myoglobinuria; demonstration of a glycogen-
olytic defect in the muscle. J. Clin. Invest. 38: 2044, 1959.
171. Schoedel, W. Die Wirkung der Muskel-Adenylsaure
und chemisch venvandter Stoffe, auf die Durchblutung
des Skeletmuskels. Pfliigers Arch. ges. Physiol. 236: 93, 1935.
172. Schretzenmayr, A. Uber Kreislaufregulatorische Vor-
gange an den grossen Arterien bei der Muskelarbeit
Pfliigers Arch. ges. Physiol. 236: 190, 1933.
173. Spalteholtz, W. Die Verteilung der Blutgefasse im
Muskel. Abhandl. Ges. Wiss. Gbttmgen Math.-physik. Kl.
14: 509 (2), 1888.
174. Stjczs, E., E. Hetenyi, and I. Went. Analyse der bi-
phasischen Wirkung von Adrenalin an Kiinstlich durch-
stromter hinterer Extremitat des Hundes. Acta Physiol.
Acad. Sci. Hung. II: 317, 1 957.
175. Suczs, E., E. Hetenyi, and I. Went. Untersuchungen
auf Adrenalinwirkung primar auftretenden Vasodilata-
tion an denervierten, Strukturen. Acta Physiol. Acad. Sci.
Hung. 11 : 327, 1957.
176. Verzar, F. The influence of lack of oxygen on tissue
respiration. J. Physiol., London 45: 39, 1912.
177. Whelan, R. F. Vasodilatation in human skeletal muscle
during adrenaline infusions. J. Physiol., London 119: 575,
'95*-
178. Whelan, R. F. The effect of adrenaline and noradrenaline
on the blood flow through human muscle. Ciba Found.
Symp., Peripheral Circulation Man. 1954-
179. Wilkins, R. W., and L. W. Eichna. Blood flow to the
forearm and calf. I. Vasomotor reactions: role of the
sympathetic nervous system. Bull. Johns Hopkins Hosp.
68: 425. '94' •
180. Yonce, L. R., and W. F. Hamilton. Oxygen consump-
tion in skeletal muscle during reactive hyperemia. Am.
J. Physiol. 197: 190, 1959.
181. Zweifach, B. Basic mechanisms in peripheral vascular
homeostasis. Trans., Third Conf. on Factors Regulating
Blood Pressure. New York : Josiah Macy Jr. Foundation.
■949. P- >3-
CHAPTER 41
The hepatic circulation1
STANLEY E. BRADLEY
Department of Medicine, Columbia University College of Physicians
and Surgeons, and Presbyterian Hospital, New York City
CHAPTER CONTENTS
Anatomy
Methodology
Direct Methods
I lepatic and splanchnic blood Hows
Hepatic and splanchnic blood volumes
Hepatic and splanchnic blood pressures
Indirect Methods
Hepatic blood flow
Splanchnic blood volume and transit time
Normal Parameters of the Hepatic Circulation
Hepatic Blood Flow
Splanchnic Vascular Pressures and Resistances
Splanchnic Blood Volume
Primary Determinants of Hepatic Blood Flow and Volume
Cross Section
Path Length and Distributional Pattern
Viscosity
Volume and Distensibility
Secondary Determinants of Hepatic Hemodynamic Adjust-
ments
Neural Determinants
Neurohumoral Determinants
Epinephrine and norepinephrine
Acetylcholine
Autonomic blockade
Local Biochemical Determinants
Oxygen
Carbon dioxide
Histamine
Physical Determinants
Intra-abdominal pressure
Gravity
Respiration
Exercise
1 The preparation of this report was aided by a grant from
The Heart and Lung Foundation, New York City. It is sub-
mitted in honor of Chester S. Keefer and the Golden Anni-
versary of the Evans Memorial Department of Clinical Re-
search, Boston Massachusetts.
Hepatic Circulatory Integration and Dysfunction
Hepatosystemic Interrelationships
Hepatosplanchnic Interrelationships
a voluminous literature testifies convincingly,
and sometimes eloquently, to the importance of the
hepatic circulation in the body economy of verte-
brates. The volume and composition of the blood
perfusing the liver are undoubtedly major deter-
minants of hepatocellular function. The maintenance
of the hepatic parenchymal "milieu interieur" with
essential nutrients and the delivery of raw materials
from the gut and other parts of the body to the liver
for processing depends directly upon the blood
supply. In even the lowest vertebrates the liver lies
in the path of all the vessels draining the splanchnic
viscera, thus potentially controlling the total splanch-
nic venous outflow (85). The splanchnic vasculature
as a whole must be considered therefore an integral
part of the hepatic circulatory system. The liver is
influential in affecting general cellular metabolism
and homeostasis only to the extent to which it can
modify the chemical structure of the blood coming to
it. A copious flow of blood is required for this purpose
and the resultant anatomical arrangements and
size of the hepatic vasculature appear to confer upon
the liver an important place in cardiovascular dy-
namics.
Quantitative evaluation of the circulatory physiol-
ogy of the liver and the other splanchnic viscera has
proved extremely difficult owing to the inadequacies
of the methods available, to uncertainties arising
from species differences, and to the lack of data ob-
tained simultaneously to provide information regard-
ing the behavior of the remainder of the circulatory
1387
i388
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
system. Measurements of cardiac output and arterial
blood pressure are required to determine whether
changes in hepatic hemodynamics are produced by
local vasomotor activity or by passive changes in
response to alterations in the perfusing pressure.
Data on the correlated behavior of other vascular
beds aid also in delineating the mechanism of in-
tegration and in placing the role of the splanchnic
bed in proper perspective. Responses seem to differ
between the various vertebrate species either because
effective drug dosage levels and the intensities of the
various stimuli used are not comparable or because
the physiological mechanisms are fundamentally
dissimilar. More data are needed to determine which
of these alternatives is responsible for many phenom-
ena. Meanwhile, interpretation of the behavior of
the hepatic circulation in one species (in man for
instance), on the basis of the known behavior in
another (such as dog) must be made with caution.
Methodology is a major stumbling block. Regardless
of species, the hepatic circulation is difficult to ap-
proach and surgical procedures of some kind are
usually necessary. As a result the method of measure-
ment may modify or interfere with the response
under study. Coniinuous observations over any ex-
tended period or repeated examinations at long
intervals may be impossible owing to deterioration
of the preparation or to the ultimate irreversible
damaging effects of mensuration itself. All these dif-
ficulties may be laid to the inaccessibility and com-
plex arrangement of the hepatic vasculature.
ANATOMY
Recent investigations have contributed impor-
tantly in characterizing the structural patterns of
the hepatic vascular inflow and outflow systems.
A variety of techniques has been employed. The
injection of plastic masses and colored materials of
various kinds into the hepatic artery, portal vein,
and hepatic veins has been used with increasing skill
and efficacy (97, 120, 148, 205, 214). Careful re-
constructions by the wax plate method or by photo-
graphic procedures have resulted in a new appraisal
of the arrangement of minute hepatic vessels relative
to the parenchymal cells. Modern methods of micro-
dissection have been less frequently used, but direct
observation of the quartz-rod transilluminated liver
in living animals has played an important part in
providing information on the anatomy and behavior
of the sinusoids (185, 270, 299). A large number of
careful gross dissections of the splanchnic vascular
bed has resulted in more reliable statistical data on
the various types of arrangements of the hepatic
artery and portal vein (120, 145, 214). Although
anatomical facts are of vital importance in the inter-
pretation of functional data, it must be emphasized
that a priori inference regarding functional signifi-
cances on the basis of structure alone may be very
hazardous.
The character of the venous and arterial inflow-
tracts is particularly susceptible to misinterpretation.
The cross section of the hepatic artery is much smaller
than that of the portal vein in a ratio of approxi-
mately one to five — suggesting that arterial inflow
is roughly one-fifth of the portal venous inflow. Since
this conclusion has found some justification in the
measurement of blood flows, it has served to encour-
age further speculation. Cross-sectional area alone is
not a good indication of relative flows in the absence
of data on pressures and resistances, and it is not
surprising, therefore, to find on further study that
the only generalization regarding the relationship
between arterial and venous inflow, which seems per-
missible at present, is that they tend to show a degree
of reciprocity. Anatomically the two systems differ
greatly.
The portal venous system drains the vascular beds
of the spleen, pancreas, stomach, large and small
intestines, and the mesenteries. Each of these beds
presents certain unique features that cannot be dis-
missed simply because they are not immediately
concerned with the hepatic circulation. The volume
of blood flowing into the portal vein and the pressure
maintained upon the blood in the portal vein are
determined to a large extent by the resistances to
arterial perfusion within each of these portal units.
The dynamics of portal hepatic inflow are therefore
bound up intimately with the behavior of extra-
hepatic splanchnic circulation.
The arteries giving rise to the extrahepatic splanch-
nic vasculature include a large array of major
branches that spring directly from the aorta or from
the celiac axis in a rather bewildering variety of
patterns recently described in detail by several
anatomists. [See (214) for survey.] In general
there is an abundance of collateral anastomoses out-
side the organs supplied, but exceptions to this tend-
ency abound and surgeons must proceed warily in
ligating any large branch without prior demon-
stration of the area of supply.
The terminal vessels are equally diverse, ranging
from the well-muscled end arterioles (penicilli) in
I HE Hl'.l'AHC CIRCULATION
'3^9
the spleen to the thin freely anastomosing mucosal
arterioles in the gastrointestinal tract (22, 39, 184,
227, 317)- It seems probable that the major point
of splanchnic vascular resistance lies in these vessels,
but arteriovenous (A-V) anastomoses between mu-
cosal arteries and veins appear to be numerous.
There is evidence (22) that blood may be diverted
through these channels principally as a result of
changes in capillary resistance rather than active
changes in A-V cross section.
The capillary nets that drain into the various tribu-
taries of the portal vein are also highly variable. In
the gastric and mesenteric beds (22, 317) thorough-
fare channels (A-V) may provide direct routing of
blood from the arterioles to the venules, the degree
of capillary filling outside the A-V capillary depending
upon selective "sphincteric" action. Similar vessels
have been described in the spleen. Here the capillary
system is made more than usually complex by the
presence of a venous sinusoidal system which has
been the cause for much disagreement (184, 227).
The presence of capillary sphincters and A-V channels
elsewhere also remains disputed, since it is possible
that the phenomena described may be artifactitious.
Muscular tissue is not obviously present in the capil-
laries or at the sites of the so-called "sphincters."
Capillary vasomotion and the closure of sphincters
may therefore be attributable to changes in intra-
luminal pressure secondary to arteriolar activity
rather than local contractions. Capillary nets could
contribute importantly to frictional resistance through
mechanisms such as these, but further anatomical
investigation is necessary. In addition, it must be
shown more satisfactorily that the manipulation of
tissues prior to or during microscopic examination
is not responsible for the changes observed.
The portal vein enters the hilum of the liver in
close relationship to the hepatic artery and the
emerging common bile duct. It is a rather weakly
muscled vessel, most of the muscle fibers being ar-
ranged longitudinally with a sparse coat of circular
muscle (65). The structure of the portal vein sug-
gests limited distensibility and easy collapsibility.
Numerous communications between the portal
vein and the systemic veins have been demonstrated
by a variety of techniques. Edwards (119) has shown
(by roentgenography and dissection after injection
of a barium sulfate suspension into the femoral veins
of three cadavers) that the most important connec-
tions are to be found in man at the retroperitoneal
surfaces of the abdominal viscera, in the pelvis,
and in the mediastinum. Even in normal persons,
the portal system fills with radiopaque material
introduced in this way. The anastomoses are rela-
tively small and are probably of no importance in
determining portal venous pressure. They play a
more prominent role when portal venous inflow is
blocked. In dogs and other vertebrates in which
mesenteries are better developed than in man such
retroperitoneal links appear to be lacking (85).
At its entrance into the porta hepatis, the portal
vein displays a relatively uniform and constant ar-
rangement which contrasts sharply with the dis-
orderly configuration of the hepatic arterial supply.
According to Gilfillan (145) the portal vein is nearly
always formed behind the head of the pancreas at the
level of the second lumbar vertebra by the union of
the splenic and superior mesenteric veins to course
directly in the hepatoduodenal ligament to the hilum.
The hepatic artery (or arteries), on the other hand,
is highly variable in its origin, course, anastomoses
within the gastrohepatic ligament, and relation to
the portal vein. Within the liver, the venous and
arterial inflow tracts assure a fairly regular pattern
of distribution and relationships. At the hilum the
blood vessels and bile ducts do not penetrate the
capsule of the liver but are carried in a sheath of
connective tissue derived from it which accompanies
them in all their ramifications as the portal tract.
Recent studies of vascular casts have demonstrated
that the external fissuring or lobation of the liver is
not precisely followed by the vascular and ductal
system (173, 214). Instead the vessels are distributed
to the right or left lobar segments which are sepa-
rated by an avascular sagittal cleavage plane inter-
secting the visceral surface of the liver along a line
drawn from the fossa of the inferior vena cava through
the gallbladder. These "vascular lobes" are divided
into segments which are distinct and easily separable
in plastic corrosion casts. Communications between
the right and left hepatic arteries or between their
branches may be found within the liver (148) but
appear to be uncommon, at least in man (84, 205).
The portal venous system appears to have a similar
and parallel arrangement with even fewer intra-
hepatic anastomoses.
The finer branches of the portal vein and hepatic
artery communicate indirectly through a capillary
network which appears to furnish the blood supply
to the bile duct and other tissues in the portal tract
(97, 120, 148, 185, 205, 214, 270, 299). The small
hepatic artery lies close to the portal vein in all the
vertebrate species studied. Arterioportal anastomoses
resembling rungs of a ladder have been described by
'39°
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
some workers (185, 270, 299) — and denied by others
(84) — in the frog, rat, mouse, rabbit, and guinea
pig (not in the cat or man) close to the terminal
arborizations in the portal tract. Three-dimensional
reconstruction studies on the basis of injections of
dyes or colloids by a large number of workers now
seem to warrant the view that terminal branches of
the hepatic arteries as well as of the portal vein
give rise to the sinusoids in various mammals and
amphibia (84, 97, 102, 120, 148, 185, 205, 214, 270,
299). It is possible that these findings apply in general
to most vertebrates but additional studies of the
comparative anatomy of the finer hepatic vessels are
needed. The axial vein in the portal space appears to
give rise to smaller radicles which then course paral-
lel to the parent vessel, usually in the same direction,
giving off branches that penetrate a so-called "'limit-
ing plate" of parenchymal cells to enter the sinusoids.
Small branches leading directly into the sinusoid
may also spring from the axial portal vein. The capil-
lary network fed by the hepatic arterioles gives rise
to the sinusoids and by its link to the portal vein
provides an arterioportal anastomosis through which
the portal venous distribution may be supplied by
the arterial inflow; or the reverse could occur. It is
difficult to be certain in the welter of conflicting
claims, but it does seem likely that the sinusoids con-
stitute the most important region of terminal ar-
teriolar distribution rather than the portal tract and
supporting tissues. The hepatic arterioles may join
the terminal portal veins where they enter the sinus-
oids, or they may enter the sinusoids directly at any
point between the portal tract and the central veins
The junction of the arterial and portal venous streams,
therefore, apparently occurs chiefly (if not entirely)
under most circumstances at or within the sinusoid
and perhaps to some extent within the capillary net-
work in the portal tracts. At this level the vessels
appear to have very thin walls containing little if
any muscle tissue. Sphincters are described at the
point of entry into the sinusoid because closure of
the vessels in a manner suggesting sphincteric action
has been observed microscopically in transilluminated
livers of living animals. Apparently muscular sphinc-
ters have not been detected by histological techniques.
The point of emergence of a small capillary from a
larger muscular arteriole in the portal space has
been construed by Elias (120) as a sphincter. Certainly
distinct muscular sphincters do not seem to be demon-
strable within the sinusoidal system proper.
The structure of the hepatic lobule and the rela-
tionship between the parenchyma and the capillaries
or sinusoids has long been the subject of spirited
discussion, and disagreement that is not yet settled.
In recent years, Elias and his associates have taken
issue with the view that the liver is a complex tubular
gland modified by extensive coalescence and reorga-
nization to form a tightly packed mesh of cells bathed
on all sides by the blood in the sinusoids. They have
called attention to the dominance of long rows of
cells, one cell thick, in sections of mammalian livers
and the paucity of cylindrical cross sections such as
one might expect in a tubular organ. Careful three-
dimensional reconstructions indicate that the liver
may be considered a cell mass penetrated by a net-
work of tubular sinusoids separated by interconnect-
ing cellular sheets or plates, one cell thick in mammals
and certain birds, usually two cells thick in all other
vertebrates. A tubular layer of parenchymal cells
encloses the portal tracts as a limiting plate which is
pierced by the terminal branches of the portal vein
and hepatic artery. The limiting plate can be traced
along the tract to the surface of the liver where it
passes out to lie under the capsule. There may be
several such subcapsular (seemingly concentric)
limiting plates or none. The openings into the central
vein are so numerous that a clear-cut limiting plate is
not demonstrable but a similar lamina does appear
about the sublobular vein and layer branches of the
hepatic veins. These studies have not definitely ruled
out the possibility that the liver is basically a closely
conglutinated tubular structure in lower forms, with
flattening and realignment of the cells into single
cell layers in the mammals. Indeed, a tubular con-
struction is demonstrable when increased sinusoidal
pressure increases the spaces between the laminae
and seems to fragment them (121). Rappaport
(237) and his associates have attributed the usual
microscopic picture to the character of the basic
hepatic unit which they believe to be an "irregular
berry-like parenchymal mass situated around the
trio of terminal branches of portal vein, hepatic
artery, and bile duct, growing out from a small portal
triad and mainly running perpendicularly to the
central vein. The hepatic unit occupies adjacent
parts of neighboring hexagonal fields and extends
from the central vein of one hexagon to the central
vein of another." All sections of such a structure
would tend to be tangential and would, they claim,
yield a preponderance of longitudinal sections.
Regardless of the ultimate outcome of this argu-
ment it is evident that sinusoids are cylindrical or
saccular vessels closely encased in a kind of highly
flexible plastic sheathing that must operate to in-
THE HEPATIC CIRCULATION
[39'
flucnce their behavior. They radiate from the portal
tracts and converge upon the central veins, produc-
ing the appearance in section of hexagonal "lobules"
that are centered upon the central vein. This appear-
ance is apparently attributable to the degree of filling
of the peripheral vessels between the portal tracts,
since it has been shown that elevation of hepatic
venous pressure or reduction in portal venous pres-
sure changes the configuration of the lobule to one
centered upon the portal tracts as a result of relative
distension of the vessels running between the central
veins (121). The walls of the sinusoids are composed
of thin endothelial cells, possibly [as Knisely et al.
(185) claim] all capable of active phagocystosis,
though this question is not yet settled. There is no
evidence of muscular tissue and the endothelium is
usually closely attached to the parenchymal cells. A
narrow perisinusoidal fluid-filled (plasma?) space
(Disse) observed on many occasions by light micros-
copy has been clearly delineated by the electron
microscope (30, 126, i6g). Numerous relatively
large fenestrations in the sinusoidal endothelium
may permit the plasma to come into direct contact
with the hepatic cells. Both luminal and the canalic-
ular surfaces of the parenchymal cells are markedly
increased by folds and microvilli. The space between
the endothelium and the polygonal cells is apparently
no greater than 0.5/1 and it may be filled with an
amorphous material resembling basement mem-
brane rather than plasma. Since the extravascular
space is so narrow, the sinusoidal closure must in-
volve displacement and apposition of the surrounding
cell plates.
The hepatic venous drainage system begins in the
colander-like thin-walled central veins that empty
into the muscular sublobular veins. In certain re-
spects the central veins appear to be passive sumps
not strictly separable from the parenchyma and not
unlike a large receiving sinusoid. Opening and
closure of sinusoids at the point of entry into the
central vein have been described by workers using
transillumination techniques (186), but definite
structural evidence of muscular sphincters seems to be
lacking. In contrast the muscle coats of the "sub-
lobular" and other hepatic veins appear to be en-
tirely adequate for this purpose. Gibson (143) finds
that sinusoids empty only into the central veins al-
though Deysach (107) has claimed that sinusoids
may occasionally enter the larger hepatic venules
directly as "sluice channels" which may bypass the
central venous sumps. Gibson believes these vessels
are really central veins and he agrees with Deysach
in viewing the point of passage through the thick
muscular wall as a site at which contraction could
interfere with flow. In the dog, contraction of the
musculature can throw the large as well as the small
hepatic veins into corrugated folds that could con-
ceivably block outflow completely (287). The extent
of this musculature appears to vary greatly in dif-
ferent species, but it appears to be weak and rela-
tively unimportant in man (132, 143). The dis-
tribution of the hepatic veins and their tributaries
results in an intimate interdigitation with the system
of portal tracts. The finer radicles course at acute
angles or perpendicular to the portal tracts from
which they are derived. There is no evidence of seg-
mentation or lobation as there is in the arrangement
of the portal tracts. Except in animals with deeply
fissured and lobated livers, the hepatic veins freely
cross the "avascular plane" separating the hepatic
segments to bind the liver into a single vascular mass.
The system empties into the inferior vena cava by
three or more terminal branches just below the dia-
phragm or within the "caval tunnel" where the in-
ferior vena cava is closely applied to or incorporated
in the posterior surface of the liver. In certain species
(notably the dog and diving mammals) the muscular
coat of the hepatic vein becomes more prominent
and forms a sphincteric ring at the orifice into the
vena cava. The preponderance of the inferior portion
of the ring may serve as a "sling" to pull up the lower
lip of the opening into a valvelike ridge.
A lymphatic drainage system runs parallel to the
vascular inflow and outflow tracts, to communicate
at the hilum and at the junction of the hepatic veins
and inferior vena cava with larger trunks that ul-
timately carry the lymph through the local lymph
nodes to the cisterna chyli or thoracic duct (85). A
rich network of lymph vessels lies about and within
the walls of the draining vessels and beneath the
capsule of the liver, but there is little evidence for
lymphatic capillaries within the parenchyma. It is
possible that the perisinusoidal space is the terminus a
quo of the lymphatics. A more definite point lies in
fluid-filled spaces, the so-called spaces of Mall,
found in close proximity to the portal tracts, between
the limiting plate and the connective tissue making
up the bulk of the tract. Arrangements of cells and
channels resembling very small lymph nodes are
described or pictured at many points within these
nets and a rich lymphopoietic layer is found beneath
the capsule of the liver in some of the lower verte-
brates. Erythropoietic tissue may also occur in adult
1 392
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
forms, though it is usually more prominent in fetal
livers (120).
The liver is a remarkably malleable organ, the
adjoining organs molding its surface and determining
its shape. Changes in the filling of the stomach in the
dog, for example, induce considerable alteration in
the configuration of the portion of the liver lying in
contact with it. The influence of these deformations
on the local hepatic circulation does not seem to
have been studied. It is likely that only the more
superficial parts of the liver are involved and that
resistance to flow is increased in the compressed
regions. The liver appears to be equally plastic
when interference with outflow, as in congestive
heart failure, results in an elevation in vascular dis-
tending pressures. Presumably the marked hepatic
enlargement involves stretching of lamellae, canalic-
uli, and the connective tissue framework (121).
To what extent the distortion affects the degree and
distribution of resistance to blood flow remains un-
determined. Certainly, if the distortion is long main-
tained, persistent change in architecture occurs and
fibrosis develops.
METHODOLOGY
The structure and location of the hepatic vascular
bed and its tributaries indicate at once the physical
difficulties of quantitative evaluation of hepatic
hemodynamics and the variety of measurements
required. Among the latter the following appear to
be particularly important: a) determination of the
minute volumes of blood flowing into and out of the
liver, including flows through the various compo-
nents of the splanchnic bed; b) measurement of the
volume of blood within the hepatic vascular bed and
the contributory vessels of the splanchnic viscera;
and c) measurement of the blood pressure in the
arteries, the hepatic veins, and at the points of junc-
tion between the different streams in the portal vein
and the sinusoids. Given these data, a complete
analysis of the local determinants of flow and integra-
tion is possible. Until recently, most hepatic hemo-
dynamic parameters have been measurable only by a
direct approach invoking considerable traumatic
manipulation and interference with normal function.
Various indirect methods are now under study in
many laboratories which appear to provide a means of
measuring local blood pressure and hepatoportal
blood flows and volumes in intact animals and man
without operation or anesthesia.
Dirrct Methods
HEPATIC AND SPLANCHNIC BLOOD FLOWS. Blood flow
through the hepatic artery, the portal vein, and he-
patic veins has been measured directly in experimental
animals for many years by a number of devices. The
Ludwig stromuhr has been replaced by the thermo-
stromuhr and more recently by the rotameter and
other types of flowmeter (67). These methods re-
quire isolation of the artery or vein for insertion or
application of the measuring device. Additional
surgery is required to obtain a value for hepatic
venous outflow by difference between the flows
through the inferior vena cava above and below the
entry of the hepatic veins or as the retrograde flow
through the inferior vena cava (above the renal
veins) after ligation at the level of the diaphragm.
Trauma, anesthesia, manipulation, hemorrhage, and
loss through collateral channels all contribute to
the errors inherent in these procedures. Nevertheless,
they possess the great advantage of the direct ap-
proach.
The so-called "collection methods" are equally
direct but somewhat easier to use and more accurate,
at least with respect to the measurement of hepatic
venous outflow and portal venous inflow. Here the
outflow is collected, rapidly measured, and then
returned to the systemic circulation. Thus, portal
venous inflow may be measured as the outflow from
the severed splenic vein following splenectomy and
temporary occlusion of the portal vein close to the
liver (204). Blalock & Mason (35) introduced under
local anesthesia a blind brass cannula with lateral
openings via the right external jugular vein of the
dog, the superior vena cava, and right atrium into
the inferior vena cava where balloons affixed to the
cannula could be inflated temporarily at points above
and below the entry of the hepatic veins during
withdrawal of the total venous outflow. More re-
cently, Selkurt (264) has measured hepatic venous
outflow in dogs by a similar technique after shunting
blood from the hind portions of the animal via an
external circuit from the femoral veins to an external
jugular vein, with ligation of the inferior vena cava
below the hepatic veins, and collection of hepatic
venous blood from above by a special cannula.
Although this method requires general anesthesia
and abdominal surgery, inclusion of blood from the
lower portion of the vena cava is avoided and a period
of complete obstruction of flow from the hind por-
tions, with attendant circulatory disturbances, is
circumvented.
THE HEPATIC CIRCULATION
'393
Changes in local blood flow and velocity may be
detected by instruments recently developed upon the
principle of the "thermostromuhr." Grayson (155)
and his associates have used a tiny copper-constantan
thermocouple and heating wire implanted in the
liver for this purpose. The measured loss of heat to
the tissues appears to be a linear function of tissue
thermal conductivity and blood flow. The first of
these variables may be determined as a constant for
each liver after cessation of circulation; changes in the
second can then be computed in percentage terms
from changes in conductivity. The unit may be left
in place indefinitely and measurements made as
desired after healing of the wound through which the
leads emerge. Movement of blood can be assessed
only in a collar of tissue approximately 0.5 cm long
and 0.3 cm in diameter within the immediate vicin-
itv of the embedded thermocouple. Although total
flow cannot be measured, the instruments provide a
satisfactory means of following either acute or chronic
adjustments in small animals. A somewhat similar
device has been introduced by Grabner & Neumayr
(151 ) for the purpose of estimating blood flow through
a hepatic vein. A tiny thermistor affixed to the tip
of a Cournand catheter is heated several degrees
above the temperature of the blood after insertion
into an hepatic vein. Any change in temperature of
the element is directly related to a change in the
velocity of the blood flow in the immediate vicinity
of the ''pickup," or to the actual volume of flow if
the calibre of the vein is constant. Movements of
the catheter tip with respiration, reversal in the
direction of flow, scar formation, proximity of large
vessels, and changes in hepatic blood temperatures
may jeopardize the validity and usefulness of the
method, but it does possess the advantage of detecting
rapid and transient alterations.
The transillumination method (185, 270, 299),
discussed above in connection with the delineation
of the finer anatomy of the liver, has proved valuable
in defining the character of blood flow through the
terminal radicles of the hepatic artery and portal
vein. The technique invokes careful exposure of the
liver in anesthetized or pithed animals with as little
trauma and blood loss as possible. A quartz rod,
provided with a conduit through which warm Ring-
er's solution may bathe the tissue examined, must
be inserted under the edge of the liver. The liver
edge transilluminated by light conducted through
the quartz rod may then be examined microscopi-
cally at high magnification. Respiratory movements
of the liver can be prevented in the anesthetized ani-
mals by the introduction of 100 per cent oxygen
through a catheter placed in the trachea. Fluores-
cence microscopy and transillumination with ultra-
violet light following injection of fluorescent materials
permit somewhat better visualization of the blood
stream within the sinusoids and of the movement of
the materials from the blood into the parenchymal
cells and bile canaliculi (153). Dyes and particulate
substances have been used similarly to follow flow in
visible light. The conditions under which observations
must be made are obviously unphysiologic and limit
the extent to which generalizations may be adduced.
Local extraneous factors such as changes in tissue
tension, the direct effects of immobilizing and han-
dling the liver, the influence of foreign materials and
fluids within the abdominal cavity, as well as the
effects of anesthesia, prolonged immobilization of
the body in an abnormal position, and the limited
area available to study combine to make interpreta-
tions most uncertain. When considered in relation
to information obtained by other methods, however,
studies of the transilluminated liver may be most
helpful and revealing.
Studies of the perfusability of the isolated liver
have also contributed to the knowledge of the he-
patic circulation, though here again interpretation in
terms of the intact animal and the circulatory system
as a whole must be made with caution. The prepara-
tion of the liver in investigations of this kind has
varied widely. At one extreme the liver is handled
with greatest care to avoid prolonged interruption
of blood flow either by perfusing the liver in situ or
by rapid transfer from the living animal to the per-
fusion apparatus where it is maintained under con-
ditions as closely as possible approximating those
in situ (10, 25, 61, 83, 277). The contributions of the
arterial and venous inflow to total outflow, the
character of intrahepatic adjustments, and the re-
sponse to an array of controlled pressure-flow states
impossible to impose in the intact animal may be
precisely evaluated in the isolated perfused liver.
Techniques have steadily improved with the de-
velopment of more effective anticoagulation, oxy-
genation, and surgery. At the other extreme, the
liver is removed at a varying time after death and
perfused with different foreign substances, ranging
from saline to kerosene (1 1 1 ). Intrahepatic resistances
and the interplay of the inflow systems at different
pressures have been evaluated by this means. Kero-
sene oil has been used by Dock ( 1 1 1 ) because it is
confined to the vascular channels and does not dif-
fuse, as saline solutions do, into extravascular tissues
'394
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
to interfere with perfusability. As noted above, the
injection of colored plastic semisolid substances aids
in defining functional relationships between vascular
structures as well as their anatomic arrangements.
The extent to which streams of different color inter-
mingle or fill a given portion of the vascular bed
points to dynamic relations that may be important
in life.
Radiopaque injection masses have been used in
both living and "dead-' livers in order to visualize
the vascular tree by X ray (22, 84, 148). Daniel &
Prichard (102) have used microangiography to
study portal venous flow in rats, cats, guinea pigs,
rabbits, and goats. Contrast substance is injected
rapidly into an omental or intestinal vein and serial
radiographs taken thereafter at a rate of one or two
per sec over a 9- to 1 2-sec period or motion pictures by
high-speed cinefluorography (144). The dispersion
of the radiopaque material in the blood stream, the
distribution of portal inflow to the hepatic segments,
and the time of blood movement may be determined
graphically in this manner. Although there are ob-
vious drawbacks (anesthesia, immobilization, the
presence of a foreign material in very high concen-
tration, and manipulation of the gut), certain hemo-
dynamic effects can be examined only by this method.
Radiographic methods of studying the portal
venous system have also proved of value diagnosti-
cally. Roentgenograms taken at the operating table
immediately after injection of a concentrated solution
of Diodrast (85) (70% — 12 to 40 ml, depending on
the size of the patient) or Urokon (70% — in similar
dosage) into a tributary of the portal vein have been
helpful in determining the extent of collateral cir-
culation or the point of venous obstruction in patients
with portal venous hypertension. Percutaneous
splenoportal venography (15, 314) permits visualiza-
tion of the splenic and portal veins in anesthetized
patients and, when rapid serial radiography is em-
ployed, the character of blood flow and vascular
filling can be made out. Diodrast or Urokon may be
injected directly into the spleen through a long 17-
or 18-gauge needle that is inserted through the skin
under local anesthesia. In most patients, subjected
to this procedure, the spleen is palpable and the
needle may be placed obliquely into the body of the
spleen, or it may be introduced through the ninth
intercostal space at the midaxillary or posterior
axillary line. The contrast substance leaves the spleen
almost at once and may be detected radiographically
within one or two seconds in the portal vein and its
branches. The procedure is somewhat hazardous,
since intraperitoneal bleeding often occurs and
splenic infarcts may develop. Severe hemorrhage has
been reported.
Of even greater potential danger is a new variant
of the technique of splenoportal venography de-
scribed by Bierman and his associates (32), who
introduce a needle through the liver into the portal
vein. However, they report that no serious complica-
tions developed following or in the course of 1 44
transhepatic portal venipunctures in 73 seriously
ill patients. Under local anesthesia, while the patient
holds his breath, they insert a special large-gauge
styletted needle at a point 1 cm below the xiphoid
process and 1 cm to the right of the midline to a depth
of 12 cm. The obturator is then removed and the
needle is slowly withdrawn during application of
gentle suction until there is free flow of blood, indi-
cating that the laterally placed orifice lies in a vessel.
A small ureteral catheter or polyethylene tubing
may be threaded through the needle into the vein
and left in place for a prolonged period after the
needle has been withdrawn. A contrast medium
such as Diodrast, Urokon, or Neo-Iopax (sodium
acetrizoate, iodopyracet, or sodium iodomethamate)
may be injected through the needle or catheter.
In a number of instances, the hepatic vein, inferior
vena cava, or hepatic artery have been visualized.
Zeid ft al. (314) have had a similar experience. A
more recent development which employs the costal
intra-osseous route appears to be considerably safer
(262). The injection of contrast material directly
into the medullary cavity of a lower rib results in
visualization of veins in the vertebral, intercostal,
azygos, and hemiazygos drainage system. In con-
tradistinction to splenoportal venography, which
reveals portal collateral channels in the presence of
portal hypertension, intraosseous venography permits
detection of systemic venous collaterals.
HEPATIC AND SPLANCHNIC BLOOD VOLUMES. The vol-
ume of blood in the liver and the splanchnic bed
may also be estimated by radiographic and injection
techniques. The relative mass of the hepatic vascula-
ture has been evaluated qualitatively from veno-
grams and arteriograms, and from the volume of
plastic casts of vascular tree. Measurement of the
liver opacified by contrast medium or delineated
after inflation of the stomach or colon with gas is also
theoretically possible (300). Changes in the size of
the spleen have been followed radiographically
(20) and interpreted in terms of displacement or
filling with blood. Unfortunately, the extent of vascu-
THE HEPATIC CIRCULATION
'395
lar filling by contrast substance or injection mass is
most uncertain, and the assumptions required in esti-
mating volume from X-ray shadows are of dubious
validity. Nevertheless, further exploration in this
direction may prove fruitful.
Changes in the volume of spleen and liver may be
more accurately measured plethysmographically in
animals, but the fixation of the organ and the sur-
gical handling required seriously impair the validity
of the values obtained (132). These devices permit a
rough estimation of engorgement or disgorgement of
the liver during vascular adjustments, but they
provide no information on the absolute volume of
blood in the liver. The same difficulties are encoun-
tered in studies of the volume or weight of the iso-
lated liver (10, 25, 61, 203) or spleen (163, 287).
Measurement of the volume of blood retained in or
expelled from the liver or splanchnic bed as a whole
may also be made on the basis of the difference in
blood inflow and outflow during a period of shifting
volume.
It has proved extremely difficult, also, to deter-
mine the absolute volume of blood in the liver or
spleen and their tributaries by the direct approach.
With excision, blood runs off into the systemic veins
and is lost. Surgery in living animals, with care to
block inflow and outflow tracts simultaneously and
to avoid trauma that might induce physiologic re-
distribution of blood, is required to obtain reliable
values. The quantity of blood may then be evaluated
by extraction of hemoglobin and calculation of blood
volume from the hematocrit of arterial blood. A
serious difficulty arises at this point because the hema-
tocrit in the capillaries and sinusoids may differ
greatly from that in large vessels. Radioisotope label-
ing of plasma (I131-labeled human serum albumin,
Cr51 tagging of plasma proteins, T-1824 bound to
plasma proteins) and of red cells (P3'2, Cr51) has proved
helpful in surmounting this obstacle. Recovery of
the isotope is relatively easy and blood volume can
be calculated on the basis of the radioactivity per
unit volume of arterial plasma and red cells. Allow-
ance must be made for the possible uptake of radio-
isotope by the liver cells or lymph. Though these
jmethods (106, 175, 181) provide approximate values
for hepatic blood volume in steady states, changes in
the same animals cannot be obtained.
HEPATIC AND SPLANCHNIC BLOOD PRESSURES. Blood pres-
sure has been measured directly in the intra-abdominal
veins after laparotomy in animals and man (85).
Opening the abdomen may bring about changes in
pressure gradients independently of the effects of
anesthesia and surgery, but on the whole these
measurements are acceptable and revealing, par-
ticularly when analyzed in terms of simultaneous
measurements of arterial pressure and blood flow.
Pressure measurements may be made in unanes-
thetized subjects by percutaneous splenic or hepatic
puncture (15, 17, 32). Atkinson & Sherlock (17)
found a statistically linear correlation between intra-
splenic pressure and portal venous pressure over a
wide range in 24 patients. With transhepatic punc-
ture of the portal vein (32) reliable records may be
obtained that possess an advantage over intrasplenic
pressure tracings in showing phasic or respiratory
fluctuations that are damped out in the splenic
pulp spaces. Portal venous pressure has also been
estimated on the basis of pressure in large readily
accessible collateral veins in the abdominal wall of
patients with portal venous obstruction. Though
this approach may yield valid figures for the subject
under study, it is not feasible in the normal and
does not yield values of general application. Care
must be taken to refer all values to the same reference
plane, preferably at the level of the right atrium
determined radiologically. Many workers use a level
5 cm posterior to the angle of Louis as the reference
plane in human subjects; and in general this is quite
satisfactory, though it appears to be a less dependable
guide to the level of the right atrium than the plane
10 cm anterior to the back (246). Changes in pressure
are usually of particular interest and the importance
of the zero reference plane is not often stressed, since
the accuracy of pressure differences is not affected
by it. When absolute values obtained by different
groups of workers are compared, however, apparent
discrepancies are encountered that may be due to
inexact definition of the reference point.
The development of venous catheterization tech-
niques by Cournand, Richards, and their associates
(98, 246) has opened up a new approach to the study
of intravascular pressures. The insertion of a long
radiopaque ureteral catheter deep into the venous
system under fluoroscopic control is atraumatic, rela-
tively simple, and safe. The catheter is introduced
under local anesthesia into a vein in the antecubital
fossa (preferably lying at the medial aspect) in human
subjects and into a jugular vein in dogs. It is then
threaded into the right atrium and inferior vena cava.
A curved tip makes manipulation and control of
direction possible but at the same time interferes
with passage, since it may cause the catheter to move
in unintended directions or to catch at valvelike
■396
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
shelves, such as the Eustachian "valve" in the right
atrium at the point of entry of the inferior vena cava.
With experience it is usually possible to maneuver
the catheter tip past these obstacles and into the de-
sired vessel. Human subjects can assist in this opera-
tion by making voluntary movements of the arms,
shoulders, neck, and trunk or by deep breath holding,
thus shortening, lengthening, or straightening the
venous channels in accommodating the passage of
the catheters. Untrained dogs must be anesthetized
and external movement of the body employed as an
aid. Rappaport (77) has described a device for
"guided catheterization" which can be used to bend
the tip of the catheter after it is placed close to the
orifices of the hepatic veins. The angulation obtain-
able permits catheterization of an hepatic vein from
below, an approach which is impossible with the
Cournand catheter owing to the relatively fixed
obtuse angle of its tip. In at least 1000 catheterizations
of the hepatic veins in man in several laboratories
not a single fatality has occurred despite the fact
that many subjects were seriously ill. This good
record is undoubtedly attributable to the fact that
the right heart is not entered and thus a dangerous
source of arrhythmia is avoided. The use of a rela-
tively soft catheter is advisable even though it makes
manipulation more difficult. Excessive buckling or
coiling should be guarded against, since knots can
be tied in the catheter. Indeed, a knot in a catheter
which included the chorda tendinae of the tricuspid
valve has been observed in the dog. The procedure is
therefore not without hazard and should always be
used cautiously with the catheter under direct ob-
servation throughout. In man, a vein draining the
right lobe is easiest to enter; in the dog, a vein drain-
ing the left lobe. Since these lobes are the largest
hepatic lobes in man and dog, the catheter can be
inserted to a depth that permits reliable measure-
ments of intrahepatic venous pressure.
Taylor & Myers (286) have shown that thrusting
the catheter deep into the liver and obstructing the
hepatic vein provides a means of measuring portal
venous pressure. Occlusion is assured by introducing
the catheter until it buckles slightly within the he-
patic vein. The '"occluded hepatic venous pressure"
theoretically reflects the pressure transmitted from
the portal vein or venules through a stationary column
of blood extending from the tip of the catheter. The
outflow tract obstructed is probably quite large,
presumably consisting of a "wedge" of convergent
hepatic venules, sublobular veins, and sinusoids into
which both hepatic arterioles and portal venules
empty. The hepatic venous "end pressures" therefore
may mirror the mean pressure attained when flows
into and out of the obstructed area have reached
equilibrium and may more closely approximate
sinusoidal pressures than portal venous pressure.
The small gradient of pressure between the portal
vein and the sinusoids probably accounts for the good
agreement with the portal venous and intrasplenic
pressures reported by a number of workers (76, 135,
244)-
Indirect Methods
Hepatic venous catheterization has also proved of
major importance in the development of indirect
methods for the appraisal of the hepatic circulation.
Accurate measurement of changes in the blood as it
passes through the liver makes it possible to apply
the Fick principle in the estimation of flow, to study
hepatic clearances, and to follow the dilution of
isotopes within the splanchnic blood volume. Since
hepatic blood flow and hepatic arteriovenous dif-
ferences can be determined simultaneously, hepatic
removal of various substances from the blood can
be subjected to analysis. Under appropriate condi-
tions maximal hepatocellular activity can be em-
ployed as a means of approximating the mass of
tissue perfused by blood in order to permit more
precise definition of ischemia, hyperemia, and re-
distribution. On the basis of such analyses, more
sophisticated clearance techniques have been de-
veloped that may circumvent venous catheterization.
The blood volume in the splanchnic bed and the
distribution of flow and volume have been adduced
from studies of the time required for the movement of
tracers such as I131 human serum albumin across the
splanchnic bed, in relation to flow. These approaches
have been opened up in the past 15 years and are
already yielding a rich harvest of new information
regarding hepatic hemodynamics.
hepatic blood flow. The hepatic blood flow can be
estimated indirectly by three somewhat different
methods. In one the total quantity of some substance
removed from or added to the blood each minute
by the liver is determined and divided by the hepatic
arteriovenous concentration difference, i.e., the
amount removed from or added to each milliliter
of blood perfusing the liver. In a second procedure,
the percentile disappearance of some substance
more or less completely cleared from the blood per-
fusing the liver is measured and hepatic blood flow
THE HEPATIC CIRCULATION
1397
computed as that percentage of the blood volume.
And, finally, flow may be estimated from the extent
of dilution of a known quantity of some tracer by
total outflow during; a measured period of time.
Obviously, the validity of all measurements depends
upon a number of assumptions which are difficult
to verify. Nevertheless, suitable test substances have
been found and adequate evidence of reliability has
been forthcoming to warrant qualified acceptance
of much of the data set out in the literature.
Clearance and extraction techniques. A variety of dye-
stuffs has been employed in the development of con-
stant infusion clearance and extraction techniques
beginning with bromosulfophthalein (BSP or Brom-
sulfalein) in 1945 (49), and more recently employing
tetrachlor-tetraiodo-fluorescein (rose bengal) (258),
and a tricarbocyanine dye, indocyanine green (238,
304). The hepatic removal of these substances can be
estimated with reasonable accuracy from the rate of
infusion if it may be assumed that a) disappearance
from the blood depends solely upon hepatic ex-
traction, and b) changes in plasma concentration
can be taken into account simply by multiplying
the change in concentration (AP) by the plasma vol-
ume (PV). Subtracting (rising level) or adding
(falling level) this product (in milligrams per min-
ute) and the infusion rate yields a value for hepatic
removal. The hepatic arteriovenous difference is
measured as the difference between concentrations
in samples of blood obtained from a peripheral artery
and an hepatic vein at the same time. As a rule these
values are derived by interpolation at the midpoint
between successive samples in order to allow for
simultaneous correction for changing levels. A
number of additional assumptions must be made in
accepting this procedure including c) that a sample
of blood from any hepatic vein is representative of
the total mixed hepatic venous drainage; and d) that
the presence of the catheter does not affect represent-
ative sampling. Numerous thoroughgoing investi-
gations have elucidated each step and in doing so
have contributed importantly to knowledge of hepa-
tocellular function.
All these agents are apparently transferred from
blood to bile by fundamentally similar mechanisms.
The character of BSP removal has been the most in-
tensively studied but the data available on rose bengal
and indocyanine green suggest that they move by the
same pathways, since indocyanine green interferes
with hepatic uptake of BSP (304) and BSP with
uptake of rose bengal (177, 212). Considerable evi-
dence (45, 59, 75, 177, 212, 305, 306, 312) supports
the view that BSP is removed from the blood by a
dual mechanism that involves a) accumulation or
"storage" of the dye within the polygonal cells in a
higher concentration than in the plasma, and b)
transfer by a limited transport system from plasma to
bile. Analysis of the biochemical mechanisms of
further subsidiary processes and of the physiological
concomitants is far from completion, but it seems not
unlikely that both storage and transfer require energy
expenditure and depend upon enzymatic catalysis.
Uptake into storage apparently proceeds only when
the plasma concentration is rising and for a period
after stabilization until equilibration is complete.
Whether BSP moves into the bile only from the so-
called "storage space," directly from the blood to the
bile canaliculi, or by both routes remains unsettled.
Rose bengal appears to be handled in much the same
fashion and may indeed be visualized by fluorescent
microscopy as it accumulates in high concentration
within the parenchymal cells. Intercellular accumula-
tion of indocyanine green has not been proved but
seems probable in view of the rapidity with which it
disappears from the blood relative to its output in
the bile (304). The limit imposed upon removal by
the transfer maximum or Tm- results in reduced
extraction by the liver as arterial plasma concentra-
tions rise and makes it preferable to maintain levels
close to 1 or 2 mg per cent in order to assure suffi-
ciently large differences between peripheral and
hepatic venous concentration for accurate measure-
ments. Even at higher levels, however, hepatic
removal accounts almost exclusively for the disap-
pearance of these substances from the blood.
Considerable confusion has resulted from failure
to use such words as "removal," "extraction," and
"recovery" with precision. "Removal" may be
defined either as the amount of dye removed from
the blood each minute (the usage employed in this
2 Unfortunately the term Lm has been applied by Mason
et al. (208) to maximal hepatic removal of BSP. Maximal
transfer is determined by liver mass and the abbreviation,
Lm, is, therefore, justified to some extent. Nevertheless, trans-
fer is a functional phenomenon that may be affected without
change in liver mass by substances competing for the same
system, by various inhibitors (such as deoxycholic acid or
Benemid in the case of BSP), and by fever or hepatic disorders
(74, 306). For this reason use of "Tm" to refer to the "transfer
maximum" seems preferable and in keeping with usage in
other fields (280). Of even greater importance is the fact that
Lm as determined by Mason (208) and others 6285, 295)
includes movements of dye into storage as well as transfer
from blood to bile. Hence in referring to a more discrete (albeit
complex and multifarious) activity Tm appears to be the more
suitable term.
i398
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
paper) or as the total amount which has appeared
in the bile over a period of time — usually several
hours and usually expressed as a percentage of the
total dose administered. The latter is also often
referred to as "recovery" but may be confused with
"extraction," a term correctly applied to the per-
centage of dye removed from the blood perfusing
liver and calculated as the ratio between the arterial-
hepatic venous concentration difference and the
peripheral arterial or venous concentration. Hepatic
extraction must be computed on the assumption that
the arterial concentration is a measure of the concen-
tration in the blood perfusing the liver by way of both
the hepatic artery and portal vein from which ex-
traction has occurred to account for the concentra-
tion found in the hepatic venous blood. For this
reason due allowance must be made for splanchnic
circulation time when rapid changes are occurring.
Perhaps the most serious confusion has arisen in
discussions of extrahepatic removal of BSP. When the
plasma level was maintained at a constant level in
the dog at about i mg per cent, the extraction of BSP
averaged 34 ± (sd) 12 per cent (282), in association
with removal rates of from 0.57 to i.g8 mg per min.
No more and usually much less than 10 per cent of
the amount removed per minute could be ascribed
to extrahepatic loss, when direct measurements of
hepatic uptake of BSP were made in the anesthetized
dog (41, 305). Following hepatectomy, however, the
plasma level may fall by 25 or 35 per cent from a level
of 1 mg per cent 1 hour after a single intravenous
dose, an observation which has been claimed (87,
88, 302) to indicate a proportionately large extra-
hepatic contribution to removal. The confusion here
stems from comparing two fundamentally different
removals; one, the percentage of the total removal
rate per minute attributable to extrahepatic tissues;
the other, the percentage change in plasma concen-
tration over the course of 1 hour. If the dog's circu-
lating plasma volume following hepatectomy can be
taken as 800 ml, then a 30 per cent fall in BSP from a
concentration of 1.0 mg per cent to 0.7 mg per cent
in the course of 1 hour would entail a total loss of
2.4 mg or 0.04 mg of BSP per min, approximately 5
per cent of the expected removal of 1 .0 mg per min.
This figure does not differ greatly from those ob-
tained by direct measurement and it may be con-
cluded that extrahepatic loss is negligible under most
circumstances. The failure to escape from the vascu-
lature may be attributed to the fact that all three
dyes are almost completely bound by the plasma
proteins (49, 258, 304). Neither rose bengal nor
indocyanine green enters the urine (258, 304),
whereas Bromsulfalein is excreted by the kidney in
amounts equaling 0.06 to 2.0 per cent of the total
dose (49, 60, 88, 220, 232, 275, 311). Since disap-
pearance from the blood depends, therefore, almost
exclusively upon the liver, hepatic removal per minute
may be computed from the rate of infusion (plus or
minus, respectively, the amount removed from or
added to the plasma volume, i.e., AP X PV). This
conclusion is not vitiated by failure to "recover"
more than 60 to 80 per cent of a dose of BSP from
the bile nor is indocyanine green necessarily prefer-
able because 97.7 per cent of a single dose appears
in the bile. The total recovery is a measure of the
extent to which other excretory pathways are ac-
cessible and of the time allowed for collection. It
does not throw light upon the movement into other
tissues.
The incomplete recovery of BSP does suggest,
however, that BSP may undergo alteration in the
body and that, as a consequence, calculation of
hepatic removal may be erroneous. Brauer and his
associates (60, 188) have shown that BSP from the bile
of the cat, rat, sheep, and chicken can be separated
into four fractions by column chromatography having
the same absorption spectra. Recent work (45, 93,
105, 165, 178, 211) indicates that the chromato-
graphic fractions are conjugates of BSP formed in the
liver by combination with glutathione at the sulfhy-
dryl group (GSH), with the release of bromine. Since
GSH is confined to the cells, conjugation must occur
intracellularly. In addition to various isomers of
BSP-GSH, BSP-cysteinyl-glycine conjugates are
formed, presumably by enzymatic hydrolysis of
BSP-GSH, since free glutamic acid appears simul-
taneously in the bile. There is at present no evidence
that conjugation is essential to transfer or storage of
BSP (178). Indeed, free BSP appears in the bile and
both rose bengal (189) and indocyanine (304) are
excreted without any evidence of conjugation. All
the biliary BSP conjugates have been found in the
blood of man and dog indicating escape from cells.
This movement into the blood occurs chiefly within
the liver and does not interfere with the calculation
of estimated hepatic blood flow (EHBF) since it
affects the computation of both hepatic extraction
and removal to the same extent and thus cancels out.
Enterohepatic circulation (198, 199, 224) of dye is
similarly of no concern provided hepatic venous con-
centration does not exceed the arterial concentration
and provided portal venous blood does not bypass
the liver via collateral pathways. In any case, in-
THE HEPATIC CIRCULATION
'399
testinal absorption of BSP, though it does occur, does
not seem to result in a significant difference between
portal venous and arterial concentrations. Whether
derived from intestinal contents or from hepatic
cells, BSP conjugate displays the same spectral
properties as standard BSP, but its extinction coeffi-
cient appears to be slightly different ( 1 88). The error
so produced also tends to cancel out. A larger error
may result from conjugation that occurs elsewhere
in the body, as in the kidney ('252). However, BSP
conjugate from this source seems to contribute in-
significantly to the blood level, even in hepatec-
tomized dogs with prolonged maintenance of very
high plasma BSP concentrations. BSP conjugate ac-
counted at most for 10 per cent (1.0 mg per cent) of
the plasma BSP concentration 45 min after adminis-
tration, with plasma levels falling from 17 to 10 mg
per cent, in two dogs studied by Rosenau and his
associates (252).
The most important drawback in the use of BSP,
rose bengal, and indocyanine green for the measure-
ment of hepatic blood flow lies in the impossibility
of sampling a mixture of all the venous blood draining
from the liver. The liver is a large organ in which
nonuniform perfusion, inequalities in tissue activity,
and heterogeneity of bile formation may be induced
at any time by a large number of extraneous factors.
Nevertheless, many workers (77, 146, 232, 275) have
failed to find any significant difference between con-
centrations of BSP in blood taken from different
hepatic veins in the same animal, provided peripheral
plasma levels are kept constant and comparable.
Differences observed by others (49, 118) may be
ascribed to changing concentrations or to sampling
difficulties. Careful control is especially important
during withdrawal of blood through the catheter in
an hepatic vein (62, 118, 146, 257). Diaphragmatic
movements result in displacement of the tip of the
catheter by pressing the liver down and in doing so
predispose to retrograde suction of blood from the
inferior vena cava. In the dog, contraction of the
hepatic venous musculature seems occasionally to
block venous outflow from the liver without inter-
ference with reflux. Since this phenomenon occurs
infrequently and erratically, it is extremely difficult
to appraise quantitatively. Edwards' (118) failure to
observe it in three experiments is therefore not sur-
prising. As he notes, hasty sampling may result in
dilution by residual "washout" saline infusion
trapped in the catheter and veins. Care must be
taken to avoid wedging the catheter deep in the
hepatic vein in order to avoid any stimulus to hepatic
venous contraction or interference with outflow.
Of special importance is the fact that obstruction by
the wedged catheter may affect portal venous inflow
preponderantly so that the sample obtained consists
largely of blood originating in the hepatic artery.
Sapirstein & Reininger (257) have reported values
for sodium />-aminohippurate (PAH) concentration
in "wedged" hepatic venous samples during mesen-
teric venous infusion of PAH that suggest such a
possibility. Although their results may be explained
by nonuniform distribution of PAH attributable to
"streamlining," a recent paper by Brauer et al. (62)
brings forward new evidence supporting the idea of
interference with portal venous inflow by the catheter.
These workers have injected S'i5-labeled BSP into
the portal vein or hepatic artery as a means of dif-
ferentiating arterial and venous components in the
outflow. With the former, radioactivity remained
much lower in the hepatic vein than in the femoral
artery, whereas radioactivity rose promptly in the
hepatic vein and remained higher than in the femoral
artery when BS:i5P was injected into the hepatic
artery. This phenomenon would not affect determina-
tion of BSP extraction if BSP were removed to the
same extent from hepatic arterial and portal venous
inflows. Andrews and his associates (14) have claimed
that extraction is in fact more complete when BSP
is infused into the hepatic artery than when it is given
by the portal vein in perfused canine livers. Other
workers (62, 83) have failed to confirm this observa-
tion, however, and in a variety of critical studies have
found little difference in efficiency of extraction be-
tween the two routes. Nevertheless, the uncertainties
inherent in hepatic venous sampling call for caution
in interpretation and should be acknowledged by
referring to the measure as "estimated hepatic blood
flow" or EHBF.
The best evidence that clearance and extraction
techniques with constant infusion yield valid estimates
of hepatic blood flow has been obtained from simul-
taneous measurements by direct methods. Selkurt
(264) found that the BSP method overestimated
flow by 7.3 per cent on the average when total
hepatic venous outflow was measured by collection
and reinfused in 274 comparisons in 14 experiments.
Similar results have been obtained by Shoemaker
(275) and by Drapanas and his associates (114) using
other direct methods. Changes in blood flow follow-
ing hemorrhage or transfusion were accurately re-
flected in values for EHBF. In view of unavoidable
trauma and blood loss that would enhance extra-
1400
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
hepatic escape of the dye during these procedures,
the agreement is remarkably good. Although these
comparisons have been limited to Bromsulfalein
there is no reason to suppose that rose bengal and
indocyanine green would not prove equally reliable.
Of all the clearance materials at hand, BSP appears
at present to be clearly superior. Rose bengal as ob-
tained commercially is a mixture of several chro-
matographically separable components some of which
appear to be less readily cleared than others. Though
this defect is not important so long as hepatic extrac-
tion is determined, it may be troublesome. Composi-
tion varies from lot to lot with a resultant unpre-
dictable irregularity in extraction and removal rate.
The availability of I13I-labeled rose bengal (91, 213)
simplifies analysis but does not compensate for the
other difficulties. Indocyanine green is most attrac-
tive for many reasons. It is easily and accurately
measured in the plasma; it is not conjugated, and it
does not enter the urine nor move perceptibly from
the plasma into any tissue other than the liver. Un-
fortunately, it is unstable on standing in aqueous
solutions, and may prove unsuitable, therefore, for
constant infusion. The chemical determination of
Bromsulfalein offers certain difficulties, since it is
difficult to remove the dye from the plasma proteins
and to eliminate interfering materials present in the
blank. It is possibly this factor that accounts for
Sherlock's (273) finding [which others (48, 77, 78)
have failed to confirm] that values for EHBF tended
to be excessively high when plasma BSP concentra-
tions were less than 1 mg per cent. Even a small
error in the determination of arterial and hepatic
venous concentrations may produce a large error in
the A-V difference. In any case, interference by
substances in the ""blank" can be avoided for all
practical purposes by appropriate dilution and use of
the Beckman DU spectrophotometer.
A variety of other agents has been employed for
determining EHBF but none has won wide accept-
ance. Galactose has been found to be metabolized
by the liver alone with sufficient rapidity to permit
accurate measurement of extraction and computation
of hepatic removal at levels too low for significant
urinary loss (293). Metabolic changes may interfere
importantly, however. Alcohol has been suggested
for use in the same manner but more recent work
(191) has shown that it may be removed actively by
tissues other than the liver. Finally the role of the
liver as the major site of urea formation has been
exploited in the measurement of hepatic blood flow.
Urinary urea excretion has been taken as equal to the
rate of hepatic synthesis and divided by the amount
of urea added to each milliliter of blood perfusing the
liver (the hepatic venous-arterial urea concentration
difference) to yield values for EHBF that agree with
those obtained by the BSP method (217). The diffi-
culties of analysis, correction for urinary delay, and
maintenance of a steady state, militate against its
routine use. Bromsulfalein appears to be relatively
innocuous. Anaphylactic reactions are exceedingly
rare (283) and occasional febrile responses appear
to be due to contamination during preparation.
Intense local inflammation follows extravasation of
BSP into the tissues.
Single injection techniques. The hepatic ''clearance"
of any substance removed exclusively by the liver may
be computed from the change in plasma concentra-
tion with time, following intravenous administration
of a single dose. Here the word "'clearance" has been
used in a somewhat different sense than that out-
lined above. In renal physiology the term applies
to the amount of any material excreted in the urine
per minute relative to its concentration in each
milliliter of plasma. This ratio has the dimensions of
volume and is equivalent to the volume of plasma
that would have been " "cleared" completely of the
substance in question, if it had been completely ex-
tracted from each milliliter. But the clearance (195)
may also be computed from the falling plasma con-
centration following intravenous administration of a
single dose provided a) the disappearance follows a
simple exponential decay function :
(when Co is the initial concentration, C\ the concen-
tration at time t, and k is the disappearance constant),
and b) the plasma volume of distribution (V) is
known. In this case clearance is equal to the product
of V and k, since /. is equal to the fraction of the
volume that is completely cleared. The constant k
is also often referred to, rather confusingly, as the
"fractional clearance." In the estimation of hepatic
blood flow by "single injection" it is necessary to
find substances that are removed by the liver alone
with almost 100 per cent efficiency, that are dis-
tributed within a determinable volume of distribu-
tion and that may be used for repeated determina-
tions. If a radioisotope could be employed as such,
or as a label for the ideal test material, changes in
plasma radioactivity might be followed by external
monitoring (over the thigh, for example), thus
eliminating the objectionable features of the "con-
THE HEPATIC CIRCULATION
I 40 I
stant infusion" technique which includes repeated
blood sampling, hepatic venous catheterization, and
prolonged intravenous infusion.
The remarkable phagocytic activity of reticuloen-
dothelial cells situated in the liver and splanchnic
bed early suggested the possibility that particulate
substances in colloidal suspension might be removed
with sufficient efficiency to permit the development
of single injection techniques dog, 272). Various
agents in colloidal suspension including carbon, iron,
gold, chromium phosphate, polyvinylpyrrolidine, and
denatured plasma protein have been studied inten-
sively by many workers (33, 109, 1 10, 272, 296, 316).
In general it appears that phagocytic removal by
the R-E cells within the liver and spleen depends
upon particle size, "saturation," splanchnic blood
flow, body temperature, and the obscure factors that
determine preferential removal (55, 56, 316). Of these,
particle size seems to be critical though difficult to
define. Large particles (100 A or larger) are taken
up more actively than small ones; addition of plasma
to the suspension prior to administration appears to
enhance removal, possibly by increasing the bulk of
small particles with a protein coating like that ob-
served directly by Knisely et al. (185) prior to phago-
cytosis by KupfTer cells. The majority of studies in
intact animals and man have involved the administra-
tion of P32-labeled chromium phosphate, radioactive
gold (Au198) and heat-denatured plasma albumin
labeled with I131 (33, 231, 242, 296). From 80 to 100
per cent of a single dose of each of these agents has
been shown to accumulate in the liver and spleen
and each yields a disappearance curve that can be
resolved into two or more simple exponential func-
tions. The values in plasma radioactivity do not
usually fall to zero but tend to flatten into a straight
line on semilogarithmic paper. This phenomenon has
been attributed to the very slow removal of small
particles which represent only a minute fraction of
the dose. In practice the values obtained by extrap-
olating the "tail" of the curve back to zero time are
subtracted from the initial figures to obtain a disap-
pearance curve which is usually a single exponential
that can be evaluated simply in terms of the disap-
pearance half-time (t\ =):
k . 2.303 log C0/.5C0 _ .693
'2 y2
taking any value on the curve as C0 and lv«. as the
time required thereafter for the concentration,
plotted semilogarithmically, to fall to half C0. The
value for k, i.e., the fractional clearance, is usually
multiplied by the total plasma volume to yield a
figure for EHBF. Colloidal chromium phosphate is
difficult to prepare with a suitable range of particle
size and is now little used. Radioactive gold and
iodine are 7-emitters and their disappearance may be
followed by external counting.
Perhaps the most serious difficulty with the single
injection techniques lies in choosing the "volume of
distribution,' (Y) to which the hepatic fractional
clearance may be referred (108). Although the plasma
volume is usually employed, a substantial fraction of
the plasma volume within the splanchnic bed from
which clearance has occurred cannot be included.
The "volume" concerned is presumably one, too, in
which admixture is instantaneous and throughout
which the same concentration prevails at any mo-
ment. Attempts have been made to correct for "mix-
ing time" by including a nondiffusible dye, like
T-1824, with the test close, but the corrections have
proved relatively insignificant and have been
deemed unnecessary. Attempts to compute V have
proved less successful, especially following trauma or
blood loss. Many workers use the volume of dilution
calculated from the intercept at zero time obtained
by extrapolation of the disappearance curve; others
simply report values for k, or use the plasma volume
less a fraction held in the splanchnic vessels. Another
problem arises from the assumption that hepatic ex-
traction is nearly complete. In early studies (296),
values of 70 to 90 per cent were reported for col-
loidal Au198, but more recently reported figures (231)
range from 30 to 70 per cent, accounting perhaps for
lower values for EHBF. Changes in the composition
and properties of gold colloids commercially avail-
able may be responsible for this phenomenon. Al-
though heat-denatured serum proteins labeled with
I131 appear to be extracted very efficiently and, as an
added advantage, are ultimately eliminated by
normal metabolic processes, they have found little
use in large part because I131 rose bengal and indo-
cyanine green have proved more attractive (19, 183,
213). Both I131 rose bengal and indocyanine green
disappear rapidly and exponentially from the blood
and, since neither is lost in the urine nor taken up in
significant quantity by extrahepatic tissues, both
may be used to measure EHBF by the single injection
technique. Nevertheless, uncertainty remains re-
garding the character of the volume of distribution
from which the dyes disappear, the constancy and
magnitude of extraction under all circumstances, and
the part played by different mechanisms in deter-
I J.I l_>
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
mining the fractional clearance. Studies of BSP
disappearance and of I131 rose bengal accumulation
in the liver indicate the possibility of reflux, of inter-
play between coupled reservoirs and transfer systems,
and of secondary derangements (e.g., saturation and
competition) that may lead to error (45, 122, 177,
212, 312).
Dilution techniques. A third indirect approach to
estimation of hepatic (or splanchnic) blood flow
depends upon measurement of the dilution of a known
quantity of some tracer within the hepatic circulation
over an accurately measured time period. In essence,
these procedures are adaptations of the Hamilton-
Stewart method for the measurement of cardiac
output and the Kety-Schmidt method for cerebral
blood flow. For the first, which has been developed
by Reichman and his associates (239), I131-labeled
human serum albumin (HSA) is injected into the
spleen and the concentration curve followed either
a) over the liver by external counting with approxi-
mate correction for background, or b) in hepatic
venous outflow collected continuously at a constant
rate with sampling at regular intervals. Analysis of
the hepatic venous radioactivity curve (as in the
analysis of pulmonary arterial concentrations for
determination of cardiac output by the "dye method")
yields a value for the average hepatic venous activity-
resulting from dilution of the injectate by splanchnic
blood flow during the time chosen. The tracer in-
jected into the spleen appears to travel as a compact
"bolus" in the splenic venous blood though a frac-
tion (significant in 20 % of human subjects) may be
left behind in the subcapsular tissues. Delayed entry
into the splanchnic bed with "trailing" may result
from slow uneven injection. The amount actually
injected and diluted within the hepatic blood flow
can be computed as the product of the radioactivity
in the peripheral blood at equilibrium (taken at 10
min after injection) and the total blood volume deter-
mined separately. This quantity divided by the cal-
culated "average hepatic venous radioactivity"
yields a value for the total splanchnic outflow during
the period of analysis. Uncertainties arising from re-
circulation, nonuniform mixing, determination of the
quantity of Im-HSA injected, and possible pooling,
together with the difficulties involved in intrasplenic
injection, limit the usefulness of this method. A similar
procedure has yielded satisfactory results in the dog
with injection of iodinated albumin and Cr5J (labeled
erythrocytes) into the portal vein (278).
Application of the Kety-Schmidt technique has
been suggested by a number of students (176, 288).
The average arterial-hepatic venous concentration
difference during equilibration following the intra-
venous administration of substances freely diffusible
throughout the liver and splanchnic bed, such as
radioactive krypton, water labeled with deuterium or
tritium, or 4-amino antipyrine, may be divided into
the average hepatic venous concentration at equilib-
rium to obtain a value for splanchnic blood flow per
unit mass of splanchnic tissue. Sapirstein (256) claims
that the distribution within the body of such uni-
formly diffusible tracers shortly after injection is
determined by the distribution of cardiac output and
thus indicative of local flow as a fraction of output.
According to this view, if radioactive potassium
chloride is given to an experimental animal and
allowed sufficient time to pass through the heart and
lungs to the tissues of the body, and if the animal is
killed before appreciable venous drainage and re-
circulation have occurred, the K.4- content of the
various organs can be used to evaluate the pattern
of flow distribution. Periods of time ranging from 5 to
60 sec before death in the rat or 20 to 120 sec in the
dog do not appear to affect the results (except for the
brain), presumably because venous K4- content is
much smaller than the arterial levels during these
periods and because recirculation does not begin to
contribute for about 30 sec. Although the drawbacks
of such a procedure are obvious, interesting and help-
ful information may be obtainable by this means
alone.
SPLANCHNIC BLOOD VOLUME AND TRANSIT TIME. The
volume of blood within the splanchnic bed and the
mean splanchnic circulation time may also be meas-
ured by an adaptation of the dilution methods (50,
94). Comparison and careful timing of the moment-
to-moment changes in arterial and hepatic venous
concentrations during the period of equilibration
following injection of some substance which is con-
fined to the vascular bed, such as I131 HSA, affords a
measure of both the total quantity of tracer distributed
within the splanchnic bed at equilibrium and the
time required for passage from artery to the point of
venous sampling. Thus the amount of tracer entering
the splanchnic bed between its first appearance in the
arterial blood (sampled from a brachial or femoral
artery) and the point of equilibrium (defined as
agreement between arterial and venous concentra-
tions within the limits of analytical error over a
period of 30 sec or longer) is equal to the product of
the total splanchnic blood flow and the average
arterial radioactivity (.7) during that period (/,,, in
THE HEPATIC CIRCULATION
i4°3
sec). The amount leaving is the product of the blood
flow and the average hepatic venous radioactivity
(F) at the same time. Since blood flow can be meas-
ured by the BSP method and average arterial and
venous radioactivities can be obtained by "integrated
sampling" the amount of tracer distributed within the
splanchnic bed is easily determined following intra-
venous injection of I131 HSA and divided by the
arterial concentration at equilibrium (Aeq) to yield a
value for splanchnic blood volume:
SBV =
(A-V)x EHBF x te
(A.a)
where EHBF is expressed in milliliters of blood flow
per second. Since splanchnic blood volume is equal to
the product of the hepatic blood flow per second and
the mean splanchnic circulation time (MCT) in
seconds,
it follows,
SBV -" EHBF x MCT
MCT'^^L-
(A.a)
It will be recognized that the Hamilton-Stewart and
Kety-Schmidt methods alluded to above are applica-
tions of the same principle which has been treated at
greater length mathematically by Stephenson and
others (284, 315). [See also Chapters 18 and 19 of
this Handbook^ Radioactive phosphate or chromium-
labeled erythrocytes have been used to determine
splanchnic red cell mass and erythrocyte circulation
time (94). Any other relatively nondiffusible sub-
stance should yield equally reliable results, provided
the major assumptions upon which the method is
based are valid.
All the difficulties implicit in the measurement of
hepatic blood flow pertain with equal force to the
determination of the splanchnic blood volume. Of
added importance is the assumption of "representa-
tive hepatic venous sampling," because the distribu-
tion of tracer within splanchnic blood flow varies
from time to time during equilibration, appearing
first in the hepatic arterial inflow and later in other
parts of the bed. Thus the various splanchnic pathways
must be represented within each outflow tract to an
equivalent degree. In view of the anatomical arrange-
ments and the data yielded by study of circulation
time (see below) this assumption seems to be valid
in normal man and animals. Local changes within
the liver will certainly interfere and the effects of
streamlining (to be dealt with later) may also in-
troduce inequalities by predisposing to predominance
of splenic and gastrointestinal vascular routes within
the left and right hepatic venous outflows, respectively.
The fact that similar values are obtained with
sampling from right and left lobes suggests that this
possibility is not important, but further work is
necessary to settle the matter. Uniform and diffused
admixture of tracer throughout all the blood filling
the splanchnic blood vessels must have been com-
pleted by the "equilibrium time." Since equilibra-
tion appears to be attained within 3 min or less, it
seems most unlikely that volumes of blood held rela-
tively motionless, in contact with but not actively a
part of the circulating blood, are included in the
final value. Tracer undoubtedly must find its way
into the splenic pulp, but largely by diffusion rather
than by active mixing, thus probably accounting for
the lack of change in splanchnic blood volume (SBY)
noted following splenectomy. For this reason the
value should be referred to as the "circulating
splanchnic blood volume." Although the term specif-
ically indicates the volume of whole blood, the tracer
usually employed (I131 HSA) is actually distributed
within the plasma. Blood volume must therefore be
computed from the estimated plasma volume and
the arterial hematocrit. But the latter is not strictly
applicable because the phenomena of "lamination"
and "plasma skimming" result in a lower hematocrit
in blood flowing through the capillaries than in
arterial blood. The resulting error may be relatively
large and must be borne in mind in interpreting
shifts, particularly in association with a changing
hematocrit. Simultaneous determinations of red cell
mass and plasma volume should yield a more accurate
estimate of the total volume, though measurement of
the red cell mass is undoubtedly more seriously limited
by the difficulty of complete admixture. Blood flow
and volume must remain relatively constant during
the period of determination — at least 10 min — to
permit estimation of hepatic blood flow. Owing to
the limits of accuracy imposed by the analytical
procedures, blood flow must not be so large relative to
volume as to minimize critically the difference be-
tween the mean values for arterial and hepatic
venous radioactivities. Experiments with model sys-
tems in which the blue dye, T-1824, has been used
have demonstrated the validity of the method pro-
vided flow per minute is not greater than three times
the volume (63). A higher ratio may be compatible
with sufficiently accurate measurement of the arterio-
H°4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
venous difference when radioactive tracers are used
(234). For the splanchnic bed the actual ratio is
much lower, approximating unity under most cir-
cumstances. A more serious problem is loss of tracer
en route either into the interstitial fluid or into col-
lateral channels that bypass the liver. The first
possibility is usually not very important. The second
does not affect the measurement in normal subjects,
but with cirrhosis and other conditions leading to the
development of a collateral circulation measurement
may be impossible. This consideration applies with
equal force to evaluation of the "mean" circulation
time.
An accurate analysis of the time required for blood
to move through the splanchnic bed requires de-
termination of tracer levels in artery and hepatic
vein at i-sec or 2-sec intervals owing to the rapidity
of the change which must be followed (303). The
development of satisfactory means of doing this by-
Wheeler (303), Tornvall (290), and their associates
has made it possible to apply and to extend analyses
of transit times worked out in the course of a study
of urine formation (51). Since blood must be drawn
through a catheter there is distortion of the concentra-
tion curve by the velocity differential produced by
laminar flow (215, 274). This factor may be allowed
for by sampling arterial and hepatic venous blood
at the same rate through catheters having the same
dimensions. Wheeler's collection technique involves
the use of a 30-foot length of polyethylene tubing
into which the blood is drawn together with droplets
of mercury to break up the column of blood and to
prevent streamlining. Since the tubing has a uniform
calibre and since withdrawal is carefully timed, seg-
ments of tubing containing blood collected during
successive i-sec intervals can be heat sealed and cut
off as separate "timed" segments for determination
of I131 activity, after removal of the mercury droplets
by centrifugation. Tornvall's device consists of a
magazine of 50 U-shaped channels, each with a
capacity of 3 ml, arranged in a carrier that auto-
matically fills each channel in succession, at i-sec to
2-sec intervals, as blood is withdrawn at a constant
rate. It is possible to apply values so obtained in the
construction of a frequency distribution of arterial-
hepatic venous transit times. The effluent from a
system of tubes draining a reservoir describes a
frequency distribution of transit times from reservoir
to sampling site when the reservoir tracer concentra-
tion is suddenly set at some arbitrary level (taken as
1 00 7c) at zero time. Changes in reservoir (or arterial)
concentration are reflected in distortions in the effluent
(or venous) concentration curve which may be taken
into account by sequential comparison and graphic
integration. Although hepatic arterial and mesen-
teric vasculatures are undoubtedly characterized by
markedly different mean circulation times, there is
so much dispersion and overlap between these and
other splanchnic beds that separation of specific
populations has proved impossible. Nevertheless, the
method affords a more precise indirect approach to
an understanding of the intrasplanchnic distribution
of flow and volume than any other now available.
NORMAL PARAMETERS OF THE HEPATIC CIRCULATION
Although methodology is now far-advanced, a
reliable quantitative description of the hepatic and
splanchnic circulation at rest in man and experi-
mental animals is still a major desideratum. Un-
certainty results from all the technical difficulties
already noted. In addition, the control or "resting
state" is extremely difficult to define and is perhaps,
like the "normal," a relatively meaningless concept.
The splanchnic circulation (hence, the hepatic out-
flow) serves at one and the same time the demands of
viscera engaged in a diversity of metabolic activities
and the needs of the cardiovascular system as a
whole. The establishment of a steady state referable
to each of these factors would be almost impossible
and, in any case, of limited applicability. For this
reason it has seemed preferable to abstract suitable
"control" approximations from the literature and to
consider these as the basis for a reasonable appraisal
of what may be characterized as the "reference
state."
Hepatic Blood Flow
Of necessity, data obtained in studies of man and
dog must dominate the picture. Although the hepatic
circulation has been investigated extensively in the
cat, rat, mouse, rabbit, and other species, the in-
formation obtained has been largely qualitative; of
considerable importance in elucidating physiologic
and pathologic adjustments, of but inferential value
quantitatively. Systematic exploration of the field of
comparative physiology with the methods at hand
would be most rewarding. In both man and dog,
the figures for hepatic blood flow, portal venous and
sinusoidal pressures, and splanchnic blood volume
range widely. In 91 apparently normal fasting human
subjects, studied resting in recumbency, the BSP
THE HEPATIC CIRCULATION
1405
method yielded a mean value of 1530 ± (sd) 300 ml
per min (48), which appears to be fairly representa-
tive [and certainly not differing significantly from
the figures published by other workers using the
same or other methods (33, 78, 231, 242, 273, 296)].
The wide range observed suggests a considerable
variation in flow that is also evident (though by no
means to the same extent) during the course of a
single study in the same subject. For the dog, the
values obtained by different workers differ much more
significantly (37, 90, 129, 232, 275, 282). To a large
extent the disagreement may be ascribed to dif-
ferences in preparation, anesthesia, and surgical
manipulation. Anesthesia appears to be particularly
difficult to control, since it may be associated with a
varying degree of hypercapnia with resultant
splanchnic vasoconstriction. Light barbiturate anes-
thesia appears to produce no change in splanchnic
hemodynamics in man so long as the plasma carbon
dioxide tension is kept constant (123). Artificial
respiration with various mechanical devices pre-
disposes to hypercapnia in man and it may be assumed
that this is also true of the dog. Hence, it seems reason-
able to accept the mean values for EHBF obtained
in unanesthetized dogs by Pratt (232), Bollman (37),
Fisher (129), and their co-workers of 43.6 ml, 42.5
ml, and 45 ml per kg body wt per min, respectively,
as the best available estimates. As in man, variance is
relatively large (in Fisher's series, for example, the
standard deviation was ±9.3 ml/ kg body wt/min)
and a similar variation is observed during the course
of a single study. The values for EHBF are not cor-
related with body size in man as they are in the dog,
presumably because the range of variation in body
size in man is so much less than in the dog and be-
cause a correlation may be obscured by other factors
responsible for variance in "resting" EHBF. Dobson
& Jones (1 10) have reported mean values for hepatic
blood flow (chromic phosphate method) in the un-
anesthetized rabbit, rat, mouse (0.74, 1.2, and 1.4
ml/ml liver/ min, respectively) in rough agreement
with those for man and dog. Similar values have been
reported also, in terms of body weight, for sheep and
cattle (138, 260).
Splanchnic J 'oscular Pressures and Resistances
The figures available for arterial and venous pres-
sures and for pressure differentials in different species
also indicate close similarities though a definitive and
systematic investigation remains to be done. In
every series the values range so widely that inter-
species differences are apparently insignificant (17, ;_\
76, 135, 244, 286, 314). This variation may be ex-
plained largely by the technical difficulty of establish-
ing strictly comparable "zero reference planes,"
states of "resting normality," and laboratory condi-
tions. Nevertheless, mean arterial pressure may be
taken as approximately 100 mm Hg in both man and
dog, portal venous pressure as 10 mm Hg, and central
(or atrial) venous pressure as o. Since portal and
wedged hepatic venous pressure in dog and man
behave in the same way and attain the same levels,
the value for sinusoidal pressure of 8.5 mm Hg com-
puted for the dog by Friedman & Weiner (135) as
the midpoint between wedged hepatic and wedged
portal venous pressures may be accepted also for
man. The pressure gradients therefore are 90 mm
Hg between artery and portal vein, 91.5 between
hepatic artery and sinusoids, 1.5 between portal
vein and sinusoids, and 8.5 between sinusoids and the
right heart.
The resistances that determine these drops in
pressure between the arteries and veins can be
evaluated only when the distribution of blood flow
is known. Exact figures are not available but most
workers (though not all) tend to accept the view that
hepatic arterial inflow is approximately one-half
portal venous inflow (132). If this is the case and if
resistances may be computed as the ratio between
pressure drop (given above) and flow per second
(1530 ml in man and 550 ml in the dog) multiplied
by a factor — 1332 — to obtain figures in absolute
units (dynes cm-5 sec), the following values for re-
sistances within the splanchnic bed would obtain in
man and dog (10 kg):
Arterial sinusoidal (hepatic arterio-
lar— R,)
Arterial portal venous (splanchnic
arteriolar — R2)
Portal venous sinusoidal (portal ven-
ular— R3)
Sinusoidal inferior vena cava (post-
sinusoidal — Rt)
Man Dog
14,630 48,750
7200 24,000
1 20 400
450 1510
The interrelationship between resistances is com-
plicated by the fact that the splanchnic circulation
consists of a combination of resistances both in
series and in parallel (fig. 1). The computation of
any component requires a precise information re-
garding the distribution of total blood flow as well
as pressure gradients. Any attempt to infer behavior
of a given resistance from values for the pressure
1406
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. I. The hepatic and splanchnic cir-
cuits. The vascular resistances in the splanch-
nic bed (A) are shown here in diagrammatic
form (B). The resistances indicated are the
determinants of sinusoidal and portal venous
pressures and of flows through the portal
vein and hepatic artery. In addition to the
hepatic arteriolar resistance {R\), colic,
mesenteric, pancreatic, gastric, and splenic
arteriolar (/?;), portal venular (/?,) and post-
sinusoidal (R,) resistances, a fifth resistance
lying in direct communication between the
portal vein and inferior vena cava (the col-
lateral resistance) is shown as a dotted line.
It may be seen that the resistance pattern
resembles that of the Wheatstone bridge,
though the electrical analogy must not be
taken too literally. The total splanchnic
resistance (RT) may be expressed (44) in
terms of its constituent resistances as follows
(omitting the collateral resistance) :
RIRZ
R,R3
+ R.
R,+R2+R3
[Reprinted from (43) with permission of the
publishers.]
Inf V.C.
*2
Splanchnic
arter: 'i-
PORTAL V resist
AORTA
gradient alone is fraught with the danger of serious
error. If the values above are correct, it is evident
that the resistance to outflow from the liver and the
portal venous bed is a fraction of inflow resistance.
It may be inferred therefrom that relatively small
absolute changes in Rf and R* would influence portal
venous and sinusoidal pressure markedly, and in
doing so, affect the volume of blood distending the
portal and hepatic vasculature.
Splanchnic Blood Volume
The balance between input and outlet resistances
and the "capacity" of the vasculature together
presumably determine splanchnic blood volume and
the pressures under which the vessels are distended.
The relative contribution of each is difficult to assess
not only during change but also in the basal reference
state. The total circulating splanchnic blood volume
(regional dilution method) in both dog and man at
rest amounts to approximately 20 per cent of the
total blood volume, within a wide range attributable
both to technical and physiologic factors (43, 44, 50,
106, 175, 181). Of this, the bulk appears to be held
within the large veins (for details see below), though
an important moiety is lodged within the sinusoids of
the liver and the spleen. The intrinsic capacity of
this variegated system at any pressure thus depends
upon the elasticity of muscular veins and the counter-
forces operating to compress or distend the intra-
abdominal viscera and their vasculatures.
Muscular contraction may quickly modify the
former, whereas the introduction of food, water, and
air into the gastrointestinal tract and the movement
of fluid across the cell walls may change the latter
very slowly. It is difficult, under the circumstances,
to establish satisfactory reproducible control values.
Moreover, the pressures acting in the different parts
of the bed are effective in proportion to diameter,
in accord with Laplace's law (70) so that a much
greater pressure rise is necessary to increase sinusoidal
volume than to produce the same increment in
venous volume. Insufficient data are available to
permit quantitative evaluation of this factor in dif-
ferent regions and to give proper importance to
venous and capillary pressure levels.
The cross-sectional distribution of the vessels,
containing the blood, figures importantly not only
in determining the average distensibility but also in
fixing the average hematocrit and the composition
of the splanchnic blood volume. Lamination of
flowing blood results in a lower hematocrit in
capillaries than in large vessels, owing to the relatively
large volume of plasma in the layer immediately
adjacent to the vessel wall. Sequestration of blood
with sluggish turnover may lead to accumulation of
red cells, however, and to a higher hematocrit than
in the large vessels. The hematocrit of the circulating
THE HEPATIC CIRCULATION
1407
splanchnic blood volume of the dog has been proved
to average 79.4 =b 8.9 per cent of the simultaneously
determined arterial hematocrit (94). Splenectomy
does not significantly affect the value, presumably
because the tracer is not dispersed throughout the
spleen. Since circulating SBV is computed on the
basis of the arterial hematocrit it is evident that the
value is overestimated by the extent to which the
splanchnic hematocrit differs from the arterial and
underestimated by failure to include stagnant splenic
blood. Nevertheless, the magnitude of the value
indicates at once that the splanchnic reservoir can
contribute significantly in systemic circulatory homeo-
stasis by mobilizing a large volume of blood to repair
deficits in the peripheral circulating volume or by
expanding to accommodate an excess that might
threaten cardiac stability.
At present, methodology undoubtedly figures most
prominently as a cause for contemporary figures
denoting hepatic circulatory variance. Active vascular
adjustments must also play an important role in
producing the variability observed in measurements
of flows, pressures, and volumes at "rest" in view of
the abundant evidence of muscular tissue and mus-
cular activity in influencing flow and volume. The
same fundamental mechanisms are involved in the
circulatory changes observed during "acute" re-
sponses to various stimuli and stresses. Alterations in
vascular dimensions and elastic properties and in
hemodynamic patterns arise primarily from the
varied interplay of vasoconstriction, closure, or col-
lapse of vessels, and rearrangement of vascular path-
ways, but numerous additional extraneous factors
exert a vital modifying, integrating, and directive
influence. Among the latter it is necessary to consider
neural mechanisms, humoral agents, and external
physical forces that are imposed by abdominal
muscular contraction, tissue tension, gravity, respira-
tory movements, and the like.
PRIMARY DETERMINANTS OF HEPATIC BLOOD
FLOW AND VOLUME
Cross Section
Since blood flow and volume are functions not
only of the driving and distending pressures but also
of the dimensions of a vasculature, splanchnic vascular
anatomy may be considered an immediate de-
terminant of hepatic hemodynamics. Structure, as
such, however, is not constant in its physiologic
implications nor is it particularly helpful in indicating
the control reference state because death and dissec-
tion result in disarrangement of the delicate balances
that depend upon tissue turgor and muscle contrac-
tion. Nevertheless, anatomic data may help in sug-
gesting the points at which resistance to flow should
be most marked. Mall's (206) careful measurements
ol the dimensions and numbers of vessels within the
liver and splanchnic bed can still be used, more than
half a century after their publication, as a basis for
computing sites of resistance. Assuming that each
successive category of vessels gives rise to a new sys-
tem of resisting conduits in parallel, the cross section
of each conduit progressively diminishing to the level
of the capillaries; the frictional resistances to flow
at each level may be computed from Poiseuille's
law of fluid flow through capillary vessels in parallel
as follows:
-L+-L+J-
R, R2 R3
where RT is the total resistance imposed by any
category of parallel branches and Ru R2, R3, ■ ■ ■ are
the resistances imposed by each constituent branch.
Since resistance in each branch varies inversely as
the fourth power of its radius (r) and directly as its
length (7) and the viscosity (jj) of the perfusate:
and if the average radius (r) is used :
RT vl or Rr-JF*
where n is the number of vessels in each category.
Changes in the values for viscosity and length con-
tribute negligibly to the change in total resistance as
the vessels narrow and increase in number. The
values presented by Mall for the number of branches
and for the average cross section at each level indicate
that (i/nr ) reaches a maximum in the smallest
arterial branches (or arterioles) in the liver, spleen,
stomach, and intestines. It may be inferred, therefore,
that arteriolar resistance plays a preponderant role
in determining splanchnic and hepatic inflow. Be-
yond this point in both the portal venous and hepatic
venous systems, values for (i/nr ) fall to very low
figures though slight increases do occur at the level
of the smallest portal venular branches and the
sinusoids. No evidence of a significant postsinusoidal
resistance may be adduced from these data, though
1408
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the wedged hepatic venous-central venous pressure
differential gives proof to its operation while the
thick throttling musculature of the small hepatic
veins (in the dog, at least) suggests a mechanism for
its production. The musculature of the arterioles
also indicates an apparently adequate basis for
variation noted in intrahepatic or splanchnic re-
sistances and blood flows. Nevertheless, other factors
enter the equation and under certain circumstances
contribute effectively in changing the total vascular
cross section independently of vasomotor activity.
Collapse of vessels secondary to changes in trans-
mural pressure may result in the redistribution of
resistances and in the reduction of the number and
diameter of the units perfused at any level. Much
work in recent years (149, 162, 226, 307) indicates
that perfusing pressure and blood flow are linearly
correlated only above 20 to 40 mm Hg in the maxi-
mally dilated vascular beds of the isolated hind limb
of the dog. At lower pressures the pressure-flow curve
is sigmoid with a positive intercept on the pressure
axis. Green and his associates (162) have suggested
that vascular compliance may produce the convexity
to the base and the positive pressure intercept at zero
flow, the perfused vessels decreasing in cross section
and number as the distending pressures are lowered
with increasing resistance in consequence. This phe-
nomenon has been extensively studied by Burton
and his associates (70-72) who attribute it to an
inherent vascular instability that develops as the
product of the intraluminal pressure and the radius
falls below the "tension" in the wall. At this point,
that of "critical closing pressure," collapse occurs and
the vessel "shuts down." The mural tension is at-
tributed to the interplay of elastic tension or tissue
resistance to stretch, active tension generated by
muscular contraction, and interfacial tension arising
from the surface forces between blood and the "un-
wettable" intima. A significant correlation noted
between critical closing pressure and resistance to
flow in various preparations (including the perfused
ear of the rabbit, intact dogs with extracorporeal
circulation, the human hand during changes in
transmural pressure) has been interpreted as evidence
that the arterioles are chiefly concerned. The linear
relationship between high pressures and flows ob-
served by Whittaker & Winton (307), Pappenheimer
& Maes (226), and others (70, 71) indicates non-
distensibility of the resistance vessels and appears to
be a result of maximal dilatation in their experiments.
Levy (194), Folkow & Lofving (131), and others
(166) have shown that the resistance vessels are freely
distensible over a wide range of pressures produced by-
equal increments in both arterial and venous pressures,
less so when arterial pressure alone is raised and more
obviously with a rise in venous pressure alone; pro-
vided the bed is denervated and before local adjust-
ments obtrude. Thus resistance to flow through an
extremity may be diminished by raising the in-
travascular pressure and increased by lowering it.
Studies by Brauer el al. (57), Trapold (292), and
Selkurt it al. (269) indicate that critical closing pres-
sures may also be defined for the vasculatures of the
liver and intestines. All have used isolated denervated
tissues perfused in vitro over a wide range of pressures.
Brauer el al. (57) obtained a sigmoid relationship
between perfusion pressure and flow through the
isolated rat liver perfused via the portal vein alone.
They found an increment in resistance below pressures
of 5 to 10 mm Hg apparently attributable to closure
of a significant proportion of vessels that occurred in
association with impairment in bile formation. Pres-
sure-flow relationships were evaluated by Trapold
(292) and Selkurt et al. (269) in the vessels of isolated
loops of small intestine. Both observed linearity
between 60 and 1 50 mm Hg, convexity to the pressure
axis at about 60 mm Hg or lower, and a tendency for
flattening at higher pressures. The zero flow in-
tercept on the pressure axis was 16 mm Hg. They
interpreted these findings as evidence of "critical
closing" at low pressures and of distension with
diminishing resistance as pressure was increased. All
three groups noted the changes in critical closing
pressure (rising with vasoconstriction and falling with
dilatation) observed by Burton and his co-workers
(70-72) during vasomotor activity produced by
drugs and anoxia, and attributed by them to a
change in "active tension." In normally innervated
beds or in carefully prepared tissues, however, the
correlation between critical closing pressure and the
level of vasomotor tone did not appear to be readily
demonstrable and free distensibility was not ap-
parent (250). Indeed, additional evidence suggests
that stretch of the vessel wall by a rise in intravascular
or transmural pressure may actually elicit a reactive
contraction of the smooth muscle, that prevents dis-
tension and that may even reduce cross section.
The possibility that intraluminal tension might
determine vascular tone in this manner was raised
by W. M. Bayliss in 1902 as an explanation for his
observations that transient occlusion of the femoral
artery was followed immediately in the denervated
limb by hyperemia, and that a sharp rise in intra-
luminal pressure elicited an increase in "tone" of
THE HEPATIC CIRCULATION
1409
isolated arterial segments. Owing to the questionable
application of these observations to the situation in
intact animals and the uncertain role of local vaso-
active materials, the "myogenic theory of tone"
was not readily accepted, but recent work by Folkow
and others appears to have put it upon a sounder
basis. Folkow & Lofving (131) worked with the
denervated hind limb, skin, and mesenteric arterial
bed in anesthetized cats, dogs, and rabbits. They
found that lowering the arterial pressure for as short a
period as 3 to 5 sec by arterial or aortic occlusion
produced dilation, whereas raising the pressure by
bilateral carotid occlusion for 20 to 60 sec elicited a
vasoconstriction. Neither anoxia nor hypercapnia
altered the response. Denervation apparently elim-
inated neither vascular tone nor responsiveness to
vasoactive drugs such as acetylcholine, epinephrine,
norepinephrine, serotonin, vasopressin, and angioten-
sin. From these results and from direct study of
isolated arteries and minute blood vessels Folkow
concluded that ''vascular tone" is created by a
rhythmic unsynchronized activity of the smooth
muscle of the resistance vessels. Conclusive demon-
stration of myogenic autoregulation within the mesen-
teric vasculature has proved somewhat difficult,
though Johnson (180) has been successful in finding
it in 2 1 of 26 experiments. The response, he observed,
was not eliminated by infusion of enough procaine
to block a possible local autonomic reflex arc and
did not appear to depend upon a change in tissue
fluid content, oxygen consumption, or lactic acid
production. The use of a suitable perfusion system
and enough time to permit recovery from surgery,
venous cannulation, and denervation may have been
important in Johnson's success in demonstrating the
phenomenon. Study of pressure-flow relationships
in the portal venous drainage tract has been less
clear cut in showing evidence of myogenic mainten-
ance of tone. Although the data obtained by Riecker
(250) with perfusion of the canine liver via the portal
vein in situ are not marred by the effects of the trauma
and disorganization, inevitable during excision and
study in vitro, they exhibit considerable variance;
opposing, on the one hand, the view that the porto-
hepatic vasculature is a simple, passive elastic svstem
and failing to support, on the other hand, intrinsic
control of vascular cross section.
Although the data indicate that closure by col-
lapse may occur in the hepatic and splanchnic vascula-
ture, the role of a definite critical closing pressure
remains uncertain. Confusion arises particularly in
connection with the character of closure. According
to the myogenic theory, the "unstretched radius"
is reached after the complete contraction of elastic
recoil and is therefore zero. In this view, closure
consists in a concentric constriction, but it may also
be regarded as collapse to form a closed slit from
some finite value for the unstretched radius. Since
critical closing has apparently escaped direct observa-
tion, it is impossible to say which, if either, state
obtains. The fact (70) that the pressure at which
closure occurs does not differ from that at which the
vessels re-expand (critical opening pressure) favors
the first, at least so far as the resistance vessels are
concerned. There is ample evidence that the critical
opening pressure for the large veins greatly exceeds
their critical closing pressure. A response similar to
that of the large veins — and slit formation on closure —
seems more likely also at the level of the tenuous
venular channels and capillaries and at arteriovenous
communications. However, the liver plates must
move with expansion or deflation of the sinusoids to
impose special plastic properties quite unlike those
characteristic of other capillary nets. Critical closing
pressures in the depth of a lobe probably differ
markedly from those characteristics of sinusoids close
to the surface not only because deformation must
affect the periphery more easily but also because the
distance from the afferent vessels is shorter in the
central regions. No matter what the mechanism of
closure may be, it effectively changes resistance to
flow by reducing the vascular cross section. In addi-
tion, the distribution of collapse may affect the re-
sistance by altering the mean length of the resisting
circuits.
Path Length and Distributional Pattern
Resistance is directly related to path length and
though the length of the conduit contributes much
less to frictional loss of the energy head at any level
than does the radius, it figures importantly in the
total resistance from artery to vein. Arteriovenous or
veno-venous shunting is the most obvious means of
shortening the vascular bed. Arteriovenous anas-
tomoses (A-V) occur prominently in the wall of the
stomach (22, 39) and may be operative elsewhere in
the gastrointestinal tract, but there is little evidence
that they are significant hemodynamically. Even
when the capillaries of the perfused stomach are
completely blocked with starch granules, no more
than 5 per cent of the total flow passes through the
A-V anastomosis. Few or none are demonstrable in
the liver though Prinzmetal et al. (233) have re-
1410
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
ported that glass spheres up to 180 n in diameter
may be recovered from hepatic venous effluent fol-
lowing injection into the portal vein. In contrast,
Gordon et al. (150) obtained much smaller values with
a method based upon the established relationship
between perfusate surface tension, minimal perfusing
pressures, and the largest radius in a system of tubes.
They found that portal to hepatic vein "anastomoses"
did not exceed 24 n in diameter and hepatic A-V
anastomosis ranged from 18 to 26 ii in the rat and
rabbit. The former are probably the hepatic sinusoids
proper; the latter, the A-V anastomosis reported by
Wakim & Mann (299) and Seneviratne (270). With
respect to the data obtained with glass beads, they
note that the method "yields remarkable results in
that every organ investigated by this means has been
shown to have very large A-V anastomoses." What-
ever the merits of this dispute, vascular anastomoses
do not seem to play a large part in determining
hepatic hemodynamics. Possibly they operate in
establishing distributional patterns of flow but even
here cross section and path length per se appear to
be more important.
The characteristic patterns of the pathways by
which blood travels through the capillary beds within
the hepatic and splanchnic vasculature are imperfectly-
understood. There is now fairly general agreement
that capillaries themselves possess no intrinsic capacity
for contractility or for autoregulation of flow and
volume within them. Chambers & Zweifach (82)
claim that capillary nets are characterized by con-
tinuously active and well-marked "thoroughfare chan-
nels" or "A-V capillaries," which pass more or less
directly from the arterioles to the draining veins and
from which the bulk of the capillaries take origin.
According to these workers, the proximal portion of
the central channel is encircled by muscle cells and
is to be regarded as a junctional arteriole or "met-
arteriole" which gives rise to even less well-muscled
"precapillary" vessels or "precapillary sphincters"
controlling inflow cross section. This arrangement
has been described with what seems complete validity
in the mesenteries (82, 317) but does not seem to be
typical of capillary nets in other tissues (154) [see
also Chapter 27 of this volume]. Active, more or
less rhythmic, alternating dilatation and contraction
or so-called "vasomotion" has also been observed in
the terminal arterioles by time-lapse photography.
Xot all workers have been successful in convincing
themselves of the validity of vasomotion but all
seem in agreement regarding the phenomenon of
"intermittency" or transient nonperfusion of a frac-
tion of any given capillary bed. Flow ceases or capil-
laries empty completely and remain so for a time,
then flow resumes without apparent cause. During
hyperemia nearly every capillary visualized will be
active. Ischemia seems to reduce the number of
active capillaries as well as to diminish flow through
those remaining in function. This phenomenon has
been repeatedly observed (185, 225, 270, 299, 317)
in the hepatic sinusoidal system as well as in the
capillary beds of the mesenteric distribution, the
pancreas, and the spleen. Of course, the capillaries
accessible to direct visualization are an infinitesimal
fraction of the total and probably not a representative
or random sampling. Intermittency probably occurs
during normal life and may be involved in altering
actively the total resistance to flow, but it seems not
unlikely that it is an expression of capillary instability
resulting from a critical reduction in distending pres-
sures by "path-length resistance."
Innumerable routes of various lengths may be
followed by the blood from the aorta to the hepatic
vein. On the arterial side, the gastric, mesenteric, and
colic vessels are particularly long and variable with
interconnection by arcades that may serve to equalize
input pressures and flows. The hepatic and splenic
arteries are shorter and more direct but path length
varies nonetheless because hilar entry results in short
routes in the more central regions and longer ones
by way of the parenchymal tissues situated at the
periphery. The same configuration applies to the
hepatic portal inflow tract but here the low pressure
head and the minimal cross-sectional resistance ap-
pear to confer greater importance upon path length
as a determinant of energy loss. Daniel & Prichard
(10 1, 102) claim that portal blood does not always
perfuse the entire liver for this reason. They used
rapid serial angiography as a means of assessing
distribution of flow following injection of Thorotrast
into a mesenteric vein in cats, rabbits, guinea pigs,
pigs, and goats. The contrast medium was usually
found to move freely into the portal vein and its
branches, then into the sinusoids, opacifying the
organ diffusely with sharp definition of its profile,
and finally into the hepatic veins and inferior vena
cava. In a few rats and kittens, a "restricted intra-
hepatic circulation" was demonstrable with failure of
the Thorotrast to fill the outermost ramifications of
the portal vein, with an irregular and patchy opacifi-
cation limited to the central tissues and with filling
of only those segments of the hepatic veins which lie
relatively near the hilum. This phenomenon could be
induced by stimulation of the hepatic nerves and by
THE HEPATIC CIRCULATION
I4I
partial hepatectomy suggesting the possibility that
neurovascular mechanism is invoked. Although the
contrast substance appeared to move more rapidly
into the hepatic veins with the "restricted" pattern,
no evidence of veno-venous shunting could be ad-
duced. Circulation time estimated in this way is not a
reliable guide to the actual velocity of the blood. Any
reduction in the volume of blood held in the vessels
would reduce the transit time without necessarily
affecting flow. The rise and fall of Thorotrast concen-
tration in the entering blood must also be taken into
account as well as the extent of dilution by arteriolar
inflow. Blood flow was not measured and it is impos-
sible to say whether the alteration in portal inflow was
associated with a compensatory change in hepatic
arteriolar resistance. The phenomenon suggests that
the peripheral sinusoids which can be examined
directly and in which intermittency has been observed
may be peculiarly susceptible to shifts in the pattern
of perfusion. It is probable that the same considera-
tions are applicable to intermittency in mesenteric
and splenic vessels. Determination of the distribution
of circulation (or transit) times across the splanchnic
bed in the dog (303) indicates that separate popula-
tions of path lengths (e.g., hepatic arterial, splenic,
and mesenteric channels) overlap markedly, each
possessing very short and very long routes. Hence the
effect of special distribution patterns within any one
circuit (such as the hepatic) would have little detect-
able influence upon the composition of draining blood.
The fact that BSP transfer remains relatively constant
over a wide range of flow suggests that parenchymal
cells are uniformly perfused under most circumstances
(305)-
I 'iscosity
The equation of vascular resistance with the num-
ber, cross section, and length of the arterioles alone
implies that blood flows freely without turbulence as
an ideal Newtonian fluid in accord with Poiseuille's
law. In reality, of course, blood is a highly complex
and heterogenous suspension of red cells in a colloidal
solution of proteins. Much evidence indicates that its
viscosity is altered by the character of the conduit, by
perfusing pressure, and by flow (202). Although there
is little reason to believe that critical velocities are
frequently exceeded in any portion of the splanchnic
vasculature, turbulence may be induced by respira-
tory and body movements which check the flow of
blood and give rise to transient vortices and eddies.
Turbulence may also arise during arterial pulsation
with a tendency for the blood to move backward, even
in the capillaries, during diastole. Laminar flow re-
sults in inward movement of red cells and accumula-
tion about the axis presumably owing to nonuniform
distribution of the shearing force across the lumen of
the vessel and to shear rate dependence of plasma
viscosity (202, 301). However, this process appears
to be inconsistent, so that turbulence of a sort always
occurs and produces a "mixed flow." Both the cell-
free zone of plasma and the high velocity differential
next to the vessel walls permit "slippage" and result
in a lower than expected viscosity in vessels of small
diameters where the volume of "plasma-lining" is
proportionately larger. From this layer is derived the
plasma which enters capillaries by the process of
plasma-skimming observed in the hepatic sinusoids
and mesenteric capillaries by Knisely (185) and
others (225, 270, 299). Consequently, the blood
perfusing capillaries may vary widely in viscosity as
well as hematocrit with resultant irregularities that
are not readily resolved in hemodynamic analysis.
The development of turbulence under various cir-
cumstances and the effects of anomalous viscosity
complicate matters still more.
With turbulence, a more complete admixture of
blood results. Thus the blood entering the arteries
from the heart has undergone a thorough stirring and
may be regarded as having a relatively uniform com-
position. Within the large veins, lamination results
in an unequal mixing of converging streams of vary-
ing composition so that "representative sampling"
from the inferior vena cava, for example, may be
difficult or impossible. Similarly, "layering" may oc-
cur in the portal vein and give rise to nonuniform
distribution within the liver, of blood coming from
the gastrointestinal, pancreatic, and splenic veins.
This possibility, first broached by Glenard in 1890,
was given experimental support by studies of Serege,
who found that India ink injected into the splenic
vein of dogs was carried preferentially to the left lobe
of the liver. Later, Bartlett et al. (24) found that
absorption of copper sulfate from the stomach and
duodenum of dogs resulted in deposition preponder-
antly in the left lobe and that absorption from the
ileum led to deposition in the right. Gopher & Dick
(95) observed "stream lines" in the canine portal vein
directly with transillumination following injection of
trypan blue into various portal tributaries. Perhaps
the most convincing evidence of "bilaterality of
portal flow" was reported in 1945 by Hahn et al. (168).
These workers injected radioactive phosphorus as
orthophosphate into the splenic vein, mesenteric
I 41 2
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
vein, or jugular vein of dogs after placing loose liga-
tures around the portal vein, hepatic artery, and
inferior vena cava. After approximately 3 sec the
ligatures were tied and the liver removed immedi-
ately, transverse sections cut and "wet ashed," and
radioactivity determined on the measured aliquots.
After injection into the jugular vein, the radioactivity
was found uniformly distributed through the liver
from right to left. Three-quarters of the radioactivity
appeared in the left side of the liver after injection into
the splenic vein, and approximately the same pro-
portion appeared on the opposite side after injection
into the mesenteric vein. Each observation was made
but once, under anesthesia, with the abdomen open
and after extensive manipulation of the viscera.
Hence, application of these data to the situation in
man is most uncertain. Nonetheless, there has been a
tendency among clinicians to explain the distribution
of pathology in human hepatic disease on this basis.
Barnett & Cochrane (23) have recently pointed
out the importance of species and individual anatomic
peculiarities in altering lamination. Experiments on a
model system proved helpful in evaluating these
effects of branching and convergence of tributaries
upon the distribution of streamlines at varying rates
of flow with different perfusates. For branches like
those at the hilum of the liver, they found that the
chance that particles in the major trunk "remote
from the branch would pass into it are greater with
increasing viscosity and decreasing width of the
branch," the angle of outflow having little significance
at the rates of flow usually prevailing. Thus, for a
fluid like blood, small branches would seem likely to
be perfused by a fairly representative sample of the
total inflow. "Moreover, the manner of formation of
the portal vein is important. Where it is formed by a
tributary joining a straight main vein (Y) particles
are more likely to pass across the portal vein when the
rate of flow in the tributary is large and its diameter
small. The converse is true where the portal vein is
formed by the symmetrical union (Y) of a major and
minor tributary, unless the rate of flow in the minor
tributary is very high. In both types of junction more
crossing-over of the streams occurs when the angle of
union is larger than when it is acute."
Although these considerations are of importance it
is probable that local movements of vessels are of
even greater significance in determining the extent
to which the blood streams commingle in the portal
vein in the intact animal and man. In the anesthetized
dog secured in the dorsirecumbent position, with the
abdomen open and respiratory movements mini-
mized, it is not surprising that lamination may be
detected in the inferior vena cava (132). And even
under these circumstances Cole and his associates
(89) found that I131-labeled rose bengal was uniformly
distributed in the liver following injection into four of
the different divisions of the portal vein draining the
spleen, small intestine, cecum, and colon of the dog. In
intact man and dog it has proved a much more elu-
sive phenomenon. Portal venography by intrasplenic
injection of contrast substance has usually failed to
show much evidence of "physiological bilaterality."
Streamlining or a filling defect in the shadow of the
portal vein at the point of entry of the superior mes-
enteric vein attributable to lateral filling by radio-
lucent blood from the mesenteric vein has been
reported (15, 116) but it is by no means a constant, or
even a frequent observation. Indeed, Patrassi and his
colleagues (229) claim that injection of contrast sub-
stances into the spleen tends rather to make the right
lobe more opaque than the left. In addition, they
found no significant difference between the transit
times from spleen to each of the two lobes when small
amounts of sodium para-aminohippurate or red blood
cells labeled with radioactive phosphorus were in-
jected into the spleen of human subjects. Similar
studies in dogs yielded the same results. Incomplete
portal venous admixture may therefore be regarded
as a potential but unlikely result of the viscous
properties ot blood. Hemodynamically it is important
chiefly with respect to the movement of red cells and
variation of hematocrit within the splanchnic and
hepatic vessels.
Volume and Distensibility
The potential volume of the vasculature which
houses the "circulating blood" of the splanchnic bed
may be analyzed dimensionally with the data pub-
lished by Mall (206) which have already been
employed in determining the major points of vascular
resistance. Using data for the length of vessels in the
mesenteric circuit from the work of Schleier (261),
and estimates of vascular lengths in the liver, the
total volume of each vascular category may be com-
puted as for cylinders. The internal volume of hepatic
and mesenteric arterial inflow tract in a dog of
"medium size" (liver weight — ca. 175 g) was found
by this means to amount to 4.1 ml; the mesenteric
and portal venous systems, 42.6 ml; the sinusoids,
32.3 ml; and the hepatic venous outflow tract,
4 1. 1 ml. Thus the arteries accounted for some 3.3 per
cent of the total, the sinusoids and mesenteric capil-
THE HEPATIC CIRCULATION
1413
laries foi-27.5 per cent, and the veins for the remainder
or 69.2 per cent. Assuming that the gastric, colic,
and pancreatic vessels hold no more than twice the
amount in the mesenteric vessels, and omitting the
spleen, the splanchnic bed in Mall's "dog" held a
total of 153 ml or 28.5 per cent of the blood volume
of an animal weighing 7 kg (taking the liver weight
and blood volume as 2.5 per cent and 7.7 per cent,
respectively, of body weight). Since the total values
compare favorably with those yielded by other
methods, the figures indicating distribution of volumes
may be regarded as equally valid in pointing to a
predominance of the veins in determining the volume
of blood contained within the liver and the splanchnic
bed at rest. As noted above, the veins of the splanchnic
bed are generously supplied with muscle and it may
be surmised that their capacity is subject to change
by venomotor activity.
The obvious constriction or dilatation of veins —
including those of the splanchnic vasculatures — in
response to chilling, tapping, warming, or various
injurious manipulations indicates clearly the ability
of the venous musculature to alter the calibre and
length of the veins (132). The mechanisms by which
venous smooth muscle effects these changes, the
integration and function of circular, spiral, and longi-
tudinal fibers in different veins, and the patterns of
contraction and relaxation are most obscure. Zweifach
(317) and others (132) have reported spontaneous
"intermittent activity" not only in the arteries and
arterioles of the mesenteries but also in the small
venules, with cycles of alternate filling and emptying
that appear to be irregular, unpredictable, and
independent of the innervation. Similar fluctuations
have been observed by Knisely and his associates
(185) at the level of the central veins and sinusoids in
the liver, presumably secondary to activity of the
well-muscled sublobular veins. The mass, configura-
tion, and extent of the hepatic venous musculature
ranges widely among species and is apparently
capable of a corresponding range of constrictive
action, from complete sphincteric throttling at in-
numerable points throughout the total drainage net
to a modest reduction in capacity. Unfortunately,
quantitative data are lacking and even the qualita-
tive studies are so incomplete and fragmentary that
it is impossible at present to assess the pattern and
extent of change at different levels in a variety of
species.
Spontaneous vasomotion appears to be randomly
distributed, involved in strictly local shifts in volume
but not in sweeping changes that move blood between
major units of the cardiovascular system. Studies
based upon measurements of circulating splanchnic
blood volume (regional dilution of I '"-labeled HSA)
indicate that large changes in SBV may occur in the
course of normal circulatory adjustments. In man, for
example, both tilting into the upright position and
exercise in recumbency have been found to induce
splanchnic vasoconstriction with a fall in hepatic
blood flow and splanchnic blood volume (42). It is
not yet clear if a fall in distending pressure secondary
to a more marked increase in the gastrointestinal and
splenic inflow resistance than in hepatic venous out-
flow resistance, or if an active reduction in venous
capacity is responsible. The fact that splanchnic
denervation interferes with the response to tilting
suggests that venoconstriction may be essential. Even
if capacity is affected by venomotor activity, however,
the extent of filling still depends upon the level of
distending pressure and upon the manner in which
distensibility is altered by "stretch" itself.
The arrangement of collagenous tissue in the ad-
ventitia, of muscle in the media, and of elastic tissue
in the inner layers of vessels appears to result in an
elastic behavior resembling that of three springs in
parallel, the weakest representing the elastic tissue;
the intermediate, muscle; and the stiffest, collagen.
Interconnections and viscous changes in muscle and
elastic tissue complicate the effort to devise a truly
representative model (31, 241) [see also Chapters
24 and 26 of this volume]. The stretch or volume-
pressure response curve yielded by isolated vessels
proves to be concave to the pressure axis at low pres-
sures, linear over an intermediate range, and finally
convex at high pressures, suggesting that vascular
distensibility is dominated initially by muscle, then,
by elastic tissue and, finally, by collagen and fibrous
tissue, as stretching occurs. Such a sigmoid curve has
been obtained for canine splanchnic veins under a
variety of conditions in situ (4, 7). The inflections
occur at quite different pressures than they do in the
aorta in conformity with the differences in structure.
The concavity to the pressure axis and flattening
occur at a much lower pressure (at about 40 cm saline
as opposed to 140 mm Hg for the aorta) indicating
dominance of fibrous tissue in the vein. A more
marked increase in splanchnic venous distensibility
is evident at physiologic pressures during the vaso-
constrictive action of catecholamines, presumably
because smooth muscle contributes more importantly
under these circumstances. At lower pressures (below
15 cm saline), or when constriction results in a very
low venous cross section, distensibility seems to de-
1 4i4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
crease because of the operation of Laplace's law.
After a period of time at high pressures the stretched
vein does not return at once to the control volume
when distending pressures are lowered to the control
but assumes temporarily a new larger "zero volume."
The viscous element which is responsible for this
phenomenon is particularly difficult to explain. To
some extent it may be referred to architectural ad-
justments such as slippage or uncoiling of intertwined
elements, but all tissues in the wall, under sufficiently
prolonged stress, are subject to a kind of viscous flow.
Alexander (4, 5, 7) has encountered the same prob-
lem in a somewhat different form in the course of a
study of splanchnic venous distensibility in anesthe-
tized dogs. Volumes of blood were injected by a motor-
driven syringe at constant rates (10 — 250 ml min)
into the vein draining a loop of ileum isolated with
all collateral vessels and nerves ligated and cut.
Simultaneous arterial and venous pressure measure-
ments permitted direct determination of pressure-
volume relationships. By this means two types of
distensibility were evident; /) rapid elastic expansion
yielding the expected sigmoid pressure-volume curve
during the injection of relatively small volumes of
blood and, 2) an additional, more slowly developing
distention increasingly apparent at slower injection
speeds. Alexander attributes the latter phenomenon
to viscous creep and refers to it as "delayed com-
pliance." Apparently, it is a continuously changing
factor, possibly arising from the operation of multiple
viscoelastic units arranged in series with the other
components. Whatever the mechanism, delayed
compliance may be of major importance in deter-
mining splanchnic vascular pooling at any distending
portal venous pressure.
SECONDARY DETERMINANTS OF HEPATIC
HEMODYNAMIC ADJUSTMENTS
Although the dimensions and the physical proper-
ties of all parts of the splanchnic vascular bed are
primarily responsible for its over-all hemodynamic
character, both circulatory stability and rearrange-
ment are mediated by essential secondary mecha-
nisms. Neural, humoral, and physical agents are
demonstrably involved in the maintenance of the
"reference" state and in the production of appropriate
patterns of response. Change in any one of these
factors elicits adjustments in all the others that must
also be taken into account. A rich innervation assures
integration and spread of vascular adjustments.
Whether neural activity is also responsible for the
tone of vascular smooth muscle remains uncertain.
Xo matter how defined, tone is not clearly dependent
upon a continuous release of neural impulses. A very
slow and indetectable discharge rate might be in-
volved, but local factors, chemical and physical, still
seem to take precedence over and replace neural
regulation under certain circumstances. Neurohu-
moral transmitters, such as epinephrine, norepineph-
rine, and acetylcholine may also have considerable
importance, contributing by local release in the
maintenance of tone observed following denervation,
for example, or by release into the circulation, in
systemic integrations. Other local biochemical factors
that must be considered to participate include oxygen,
carbon dioxide, hydrogen ion, and metabolites like
histamine or serotonin. Among the physical deter-
minants are to be numbered intra-abdominal pres-
sure, gravity, intestinal motility, and the changes
associated with respiration and body movements.
Neural Determinants
The nerves of the liver, gall bladder, and bile ducts
form a plexiform structure made up of numerous
small ganglia with a) the anterior hepatic plexus
(derived from the left portion of celiac plexus and the
right abdominal branch of the left vagus) immeshing
the hepatic artery, and b) the posterior hepatic plexus
(derived from the right portion of the celiac plexus
and the branches of the right vagus that traverse the
celiac plexus) investing the portal vein and bile duct
(8, 190, 253). Ganglia required for parasympathetic
synapses are not present. The spleen receives its supply
almost entirely from the celiac plexus possibly with
some contribution by the left phrenic nerve. Like the
liver, the spleen receives no parasympathetic com-
ponent (53, 294). Throughout the splanchnic bed
bundles of nerves accompany blood vessels in their
distribution to the tissues. Within the walls of the
larger arteries subsidiary plexuses are arranged in a
more or less orderly manner. An outer plexus in the
adventitia, a deeper plexus between adventitia and
media, and a plexus within the muscular media have
been recognized. The complexity of these networks
becomes progressively less marked in the vessel walls
as caliber diminishes until at the capillars' level it i^
difficult or impossible to find any evidence of specific
innervation. The close association of vagal and sym-
pathetic fibers in many regions does not imply an
association in controlling vascular smooth muscle.
Indeed the reverse seems to be true for vagal fibers
THE HEPATIC CIRCULATION
1415
clearly innervate the smooth muscle of gastrointestinal
tract, biliary tract, and pancreatic ducts (including
secretory cells), but none has been traced to the blood
vessels (8, 190, 249, 253). The innervation of the blood
vessels within the splanchnic viscera appears to be
derived exclusively from the sympathetic venous
system. Moreover, all the sympathetic efferent path-
ways are now believed to be vasoconstrictive in ac-
tivity.
Recent work (80) strongly supports the view that
neither sympathetic nor dorsal root vasodilator fibers
run to the splanchnic vasculature. The appearance of
vasodilation evident in a rise in splanchnic blood flow
with no blood pressure change, or in the face of a
reduction in blood pressure, is therefore to be referred
to ''diminished vasoconstrictive tone." Although this
conclusion is not universally acceptable, it must be
admitted that a great weight of evidence gives it
strong support. Direct stimulation of the splanchnic
nerves, the hepatic plexus, or splenic nerve by a
tetanizing current induces only a vasoconstrictive
response in vivo or in situ which may be expressed by
diminished blood flow, by a tendency for the liver
and spleen to contract, by diminished cross section of
intrahepatic vessels under direct observation, and by
peripheral ischemia of the liver evident in micro-
radiographic studies (11, 20, 25, 102, 104, 132, 161,
204, 270, 299). Variation in the extent of this response
appears to be referable to differences in species studied
and in the techniques employed, but the general
agreement upon its qualitative features is unmistak-
able. In contrast, stimulation of the vagus produces
little or no obvious change in intrahepatic or splanch-
nic resistance under similar circumstances (11, 104,
161, 270). Richins (248) has claimed that vasodilation
may occur in the pancreas during stimulation of the
celiac plexus after cutting the splanchnic nerves in
the cat, because "quick-freezing" the pancreas during
this period and careful preparation of microscopic
sections of the tissue reveal larger cross sections of the
arterioles and veins. This method is obviously open to
question because it requires the assumption that
fixation and preparation of the tissues for study do
not affect the state of the vessels which obtains at the
moment of freezing. Somewhat stronger support for
active cholinergic hepatic vasodilation under special
conditions has been put forward by Grayson and his
associates (147, 157). These workers have attributed
increments in hepatic blood flow, measured by inter-
nal calorimetry in intact unanesthetized rats and
rabbits, during increments in arterial pressure pro-
duced by infusion of epinephrine or by transfusion of
rat blood, to reflex vasodilation because the response
could be blocked by section of the right vagus, celiac
neurectomy, atropine, and hexamethonium. They
could not compute the changes in hepatic vascular
resistance, however, and it seems more likely that the
failure for blood flow to change after neural blockade
during the rise in blood pressure is the result of active
vasoconstriction. In this view, an intact innervation
maintains flow by minimizing or blocking the funda-
mental vasoconstrictive response rather than by
inducing vasodilation. In fact, Grayson and Ginsburg,
like manv other workers, have found that stimulation
of the cut distal end of the vagus has no effect upon the
hepatic circulation.
The undeniable fact of abdominal pain clearly
indicates the presence of afferent pathways mediating
visceral perception. Within and about the vessels of
the splanchnic bed, myelinated and nonmyelinated
fibers may be found ending freely in a fine meshwork
or in Pacinian corpuscles. The first seem to accompany
the vessels closely, branching dichotomously at each
bifurcation and losing their myelin sheaths distal to
the last branching. A filmy plexus of nonmyelinated
nerves about the vessels extends into the avascular
portions of the mesenterv, onto the visceral perito-
neum covering the intestines and bladder, and into the
substance of the liver and kidney. Sheehan (271)
found that small nerve ganglion cells appear in this
network at wide intervals and concluded that single
fibers "branch and anastomose in a true network
arrangement." Fine twigs may be seen occasionally
issuing from the plexus to end freely among the en-
dothelial cells. Since they are demonstrable after
removal of the splanchnic sympathetic chain and the
vagi, they are presumably somatic in origin and pos-
sibly responsible for visceral sensations. Perhaps the
most prominent and clearly definable afferent nerve
endings associated with the splanchnic vasculature
are the Pacinian corpuscles. These structures vary
considerably in size and shape, ranging from easily
visible ovoid bulbs (1.0 by 0.6 mm in diameter) to
end bulbs measuring only 8 by 4 n. Typically the
Pacinian corpuscle is composed of a relatively thick,
laminated capsule with a central core through which
the main afferent nerve runs to its termination. It
lies embedded in the vessel walls particularly in the
pancreas, lymph nodes, and mesenteries or in the
surrounding connective tissue and fat, usually ar-
ranged with the long axis parallel to the vessel. The
number of corpuscles varies from species to species,
widely distributed and common in the cat, but almost
absent in the mesentery of the dog, rabbit, mouse.
! 4 I<)
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and man (251). Electrophysiologic studies (139, 141)
indicate that Pacinian corpuscles in the skin and
mesenteries are sensitive to pressure changes. The
spontaneous outflow of impulses in large afferent
fibers in the splanchnic nerve seem to derive in the
main from the Pacinian corpuscles. Gammon &
Bronk (139), recording impulses from the peripheral
end of the splanchnic nerve and its branches in cats,
found group discharges synchronous with systolic ar-
terial pressure peaks. During constant perfusion a
sustained discharge was observed. Sarnoff & Yamada
(259) have suggested that Pacinian corpuscles may
mediate the changes in arterial pressure noted during
manipulation of the splanchnic vessels, but recent
work (40) indicates that extrasplanchnic baroreceptors
are involved and that mesenteric pressure receptors do
not contribute significantly, at least in species other
than the cat. Nevertheless, afferent impulses originat-
ing in these areas may be implicated in local and
central reflex arcs of importance to hepatic circulatory
adjustments.
Impulses passing from the splanchnic bed by all
these afferent routes evidently pass directly to the
central venous system. Central representation of
afferent fibers from the abdomen has been explored
by a variety of methods. Bain et al. (18) found that
stimulation of the central end of the divided splanch-
nic causes pupillary dilatation that is due to inhibi-
tion of the oculomotor nucleus and not to change in
blood pressure, release of adrenaline, or activity of
somatic afferent fibers, since it occurs after transection
of the spinal cord between the fifth and sixth thoracic
roots. Using this method as a means of detecting
splanchnic afferent impulses they found that splanch-
nic afferents enter the cord from the sympathetic
chain via the rami and the dorsal roots. No synaptic
junctions seem to occur in the dorsal roots or in the
lateral sympathetic ganglia which resemble those for
sympathetic afferents in the sympathetic ganglia.
According to Downman (113), stimulation of these
fibers, in both cats and dogs, evokes detectable changes
in cerebral action potentials within the trunk areas of
somatic sensory representation, viz, contralateral
area I and both contralateral and ipsilateral area II.
The distribution and latency of the responses elicited
by centripetal stimulation of the splanchnic nerve do
not differ from those elicited by stimulation of a body-
wall nerve. Amassian (9) reported similar cortical
representation of visceral afferent impulses in the
rabbit, monkey, dog, and cat with maximal primary
cortical responses in the trunk region of contralateral
areas I and II, with ipsilateral representation in area
II for the cat alone. The lack of correlation between
the number of receptors and the intensity of responses
suggests that splanchnic afferent projection may be
but partially derived from Pacinian bodies. In addi-
tion, the projection to somatovisceral areas in the
cortex raises additional doubt whether Pacinian
corpuscles are primarily concerned in vascular read-
justments rather than visceral sensation. The path-
ways through the cord have been mapped out to some
extent by Aidar et al. (2) who found that action
potentials were detectable in the cat to levels as high
as the thalamus. Faster impulses course through the
ipsilateral fasciculus and nucleus gracilis, internal
arcuate fibers, and contralateral medial lemniscus to
reach the thalamus. Slower impulses ascend in the
lateral spinothalamic tracts. These findings have been
confirmed by Gardner et al. (140) in studies of cortical
projections of fast visceral afferent impulses in the
cat and monkey. Since section of the dorsal funiculi
does not always abolish cortical potentials evoked by
stimulation of the splanchnic nerve they suggest that
additional pathways are followed. The anatomical
basis for reflex regulation of the hepatic and splanch-
nic circulation is clearly evident in these studies. The
status of a controlling "vasomotor center" for the
splanchnic bed is most obscure and it cannot be said
with certainty that discrete splanchnic vasomotor
representation is detectable within the cortex. Never-
theless, the cortical representation of visceral sensory
and motor functions that may involve vascular smooth
muscle seems to imply, on the one hand, a measure of
influence upon splanchnic vascular changes by cortical
activity directly, or, on the other, a reflection of
visceral circulatory adjustments in cortical function.
Reflex responses almost certainly occur within the
splanchnic and hepatic vasculature, although they
are extremely difficult to characterize. Axon reflexes,
involving afferent nerves like those responsible for the
vasodilation of the "flare" in the skin during the
"triple response," are not demonstrable (80). The
phenomenon of "autoregulation" of hepatic blood
flow is possibly an exception but, as noted above, a
local stretch-and-response myogenic balance may be
responsible (180). Expansion of the portal venous
chamber may also elicit what Yamada & Burton
(313) have referred to as a "veni-vasomotor reflex"
characterized by arteriolar vasoconstriction proximal
to the site of venous distention. Mesenteric arteriolar
constriction observed during an elevation in portal
venous pressure (268) may be explained on this basis
but, here again, retrograde elevation of pressure to the
level of the arterioles with the slowing of flow cannot
THE HEPATIC CIRCULATION
[4I7
be eliminated as a cause for myogenic activation (180).
The possibility that elevated hepatic venous pressure
may induce reflex constriction in the hepatic arterioles
and portal venules should be investigated. Similarly,
conclusive demonstration of reflex changes mediated
through spinal or corticothalamic centers via visceral
afferent and efferent arcs is needed. There is little
doubt that disturbance of splanchnic and hepatic
vessels or stimulation of the central ends of the cut
splanchnic nerves can give rise to marked changes in
systemic hemodynamics. In both man and experi-
mental animals traction on the mesenteric vessels is
associated with a striking fall in arterial blood pressure
provided the nerve supply is intact (230, 291), whereas
splanchnic nerve stimulation results in transient
arterial hypertension (141). The mechanisms of these
reactions have not been subjected to detailed analysis
and it is impossible at present to evaluate them in
terms of venous return, peripheral vascular resistances,
and local splanchnic and hepatic hemodynamics.
The splanchnic vasculature also undergoes changes
that appear to arise reflexly from other parts of the
cardiovascular system, such as the carotid sinuses and
great veins, but the usual difficulties in interpretation
arise in connection with widespread and simultaneous
circulatory adjustments in the remainder of the body.
Perhaps the venoconstriction with increased carotid
sinus tension and the venodilation during a fall in
carotid sinus tension, or during distension of the in-
ferior vena cava noted by Alexander (4, 5) in isolated
innervated segments of the mesenteric veins, may be
regarded as reasonably clear-cut evidence of reflex
action, but even here uncertainty must remain re-
garding myogenic and humoral factors. Since the
venodilator response is abolished by section of both
vagi, these reservations seem ill-founded and a reflex
with an afferent pathway via the vagal trunks may be
postulated. Certainly afferent pathways from other
parts of the body seem to be capable of activating
visceral neural outflow and visceral vascular responses
to peripheral stimulation. Heating or chilling the
skin, a rise of pressure in the carotid sinus, distention
of the inferior vena cava, and stimulation of the
central end of the severed sciatic nerve have all been
shown to elicit changes in the splanchnic vasculature
(4, 5, 132). The blushing or blanching of the gastric
or rectal mucosa during emotional responses (156,
192) also suggests that cortical activity, mediated by
pathways that begin in the cortical motor projection
of the splanchnic autonomic system, may affect the
splanchnic and hepatic circulation.
Neurohumoral Determinants
Confusion regarding the neural determinants of
flows and volumes within the splanchnic vessels is
inevitable not only because responses are so complex
but also because so little is known about the structure
and function of the neurovascular units. Several types
of nerve endings and receptors in vascular smooth
muscle have been postulated to account for the varied
responses to neural stimulation, to neurohumoral
transmitters, and to blocking agents, but too little is
known with assurance to permit the formulation of a
fully satisfactory explanation. Recent work (161)
suggests that norepinephrine alone is released from
the nerve endings in the splanchnic vessels, and that
receptors (J3) responsive to circulatory epinephrine
are present in the mesenteric, splenic, and hepatic
vessels. Reserpine appears to be capable of releasing
norepinephrine from splanchnic nerves, whereas the
circulating amine, released by the adrenal medulla or
introduced extraneously, can replenish the depleted
store (69). Dopamine, the immediate precursor to
norepinephrine, accounts for more than 95 per cent of
the catecholamine demonstrable in the liver, jejunum,
and colon where it appears to occur to a large extent
in nonneural tissue (263). In the spleen and the
pancreas, norepinephrine and epinephrine are found
in approximately equal amounts as in adrenergic
nerves. The large local supply of dopamine may imply-
that it has an action of its own. Acetylcholine has been
found in large amounts in the spleens of some species
(100) in accord with some evidence for cholinergic
vasodilator receptors in the splenic vessels. The inter-
play of all these factors in any single neurovascular
reaction is extremely difficult to follow, especially in
view of differences in responsiveness, independent
myogenic reactivity, and innervation within what
Folkow (130) has referred to as the "series-coupled"
and "parallel-coupled circuits" of the hepatic and
splanchnic vasculature. Study of the pattern of re-
sponse to individual chemical agents may ultimately
clarify the mechanism of these responses and throw-
light upon the local and systemic role of the hormones
themselves.
EPINEPHRINE AND NOREPINEPHRINE. To what extent
the circulating catecholamines participate in vaso-
motor adjustments remains uncertain. Neural activity
appears to exert a profound and selective action,
effectively controlling splanchnic vasomotor function
without need for an adjuvant (79, 130). The total
range of control by direct sympathetic innervation is
[418
HANDBOOK OF I'llVMOI.DCV
CIRCULATION II
also much more impressive than that of the adrenal
medulla. Stimulation of the constrictor nerve fibers to
i In- spleen, for example, causes marked contraction at
rates as low as one impulse every other second,
whereas large doses (5 /ug/kg/min) of the medullary
amines fail to cause more than 40 per cent contraction
of the denervated spleen and an even smaller maximal
response is produced by stimulation of the adrenal
medullae. Celander (79) has concluded therefore that
motor control of smooth muscle in blood vessels is
dominated by the neural component. A corresponding
predominance of the adrenal medullary hormones
might apply to the sympathetic control of various
metabolic processes. In emergency situations, more-
over, it is possible that circulating hormones may have
greater importance in determining vascular responses.
Both epinephrine and norepinephrine are clearly
vasoconstrictor at all dosage levels in the perfused
liver (12, 13, 25, 81, 132). In the early work adren-
aline, known to be a mixture of /-epinephrine and
/-norepinephrine, was used. Fortunately, the effect of
epinephrine appears to dominate the vascular re-
sponse to the mixture and the findings of the earlier
studies do not differ substantially from those carried
out more recently with /-epinephrine alone. The
published data are often difficult to evaluate owing to
the rapidity with which a succession of shifts occurs
after introduction of the drugs. In part, the changes
may be attributed to the rearrangements involved in
passing from one state to another. Thus hepatic
venous outflow may increase transiently as the liver
shrinks with a reduction in intrahepatic blood volume,
although inflow may fall and remain depressed.
Differences in dosage are also obviously responsible
for certain variants and may indeed give rise to ir-
regularities in response pattern as the plasma con-
centration of the amine rises abruptly and then tails
off following injection, reaching some parts of the
vasculature early in high concentration, others later
after dilution within the vessels. Finally, the physio-
logic state of the organ, whether liver, spleen, or in-
testine, is especially important. Congestion and in-
creased resistance to perfusion arising from various
causes, chiefly on deterioration with time, may greatly
modify the response. With due allowance for all these
considerations, however, both drugs appear to in-
crease resistance to flow through the perfused hepatic
(12, 13, 25, 81, 132), mesenteric (132, 269), and
splenic (132) arterioles and to diminish the vascular
capacity by venoconstriction and splenic contraction
in all species. The facts are consistent with the pres-
ence of a-receptors mediated by norepinephrine (161).
The moderate vasodilation that may occur in the
perfused mesenteric circuit, but not in the hepatic
vessels, following the vasoconstrictive response to
epinephrine may be regarded as evidence that [1-
receptors occur in the former and not in the latter,
though the role of secondary changes with gut activity
or of balanced shifts within the liver is not easily
determined. Little information regarding the dis-
tribution of resistance and volume changes and the
relative intensities of smooth muscle contraction be-
tween the different beds or even within the same one
may be gleaned from these data.
In the intact animal, the responses are even more
varied and complex, but the interplay of local circuits,
pressures, and over-all cardiovascular adjustments
may be made out more readily. The early work (132)
on mammals yielded data generally consistent with
the conclusion that epinephrine gives rise to an eleva-
tion in arterial and portal venous pressure in asso-
tion with a reduction in splenic and hepatic volume,
and diminished hepatic venous outflow. All these
changes are in accord with those observed in the
isolated systems and suggest, furthermore, that a more
marked increase may develop in the hepatic vascular
resistances than in mesenteric or splenic to account
for the rise in portal venous pressure (1, 29, 47, 68,
104, 125, 142, 157, 159, 160, 163, 203, 2io, 223, 281).
Recent studies (1, 29, 47, 68, 125, 142, 157, 159, 160,
223, 281) indicate that norepinephrine may behave
similarly. With greater detail and precision, however,
interpretation has become somewhat more dubious.
In the first place, it is now clear that epinephrine is
essentially vasodilator in its total systemic effect,
physiologic doses producing no change or even a fall
in arterial mean pressure. Changes in flow must be
equated with mean pressures and cannot be taken
alone as evidence for vasodilation or vasoconstriction.
Furthermore, much of the published material relates
to the pattern of response observed after a single dose
of the drug that induces a succession of conflicting
local and reflex adjustments in which the assignment
of cause and effect may be quite impossible. With
constant infusion of epinephrine in unoperated,
unanesthetized man (47), dog (142), and rat (157),
hepatic venous outflow has been found to increase.
Since the increment in flow exceeded or was out of
phase with the increment in arterial mean pressure
in the studies of man (0.10 jug epinephrine kg min
for 30 min — BSP method) and dog (0.25 /ug/kg/min
for 1 min — blood flow velocity measured by implanted
"thermistorsonde") it may be concluded that over-all
splanchnic vascular resistance decreased during these
THE HEPATIC: CIRCULATION
1419
experiments. In the rat, however, the rise in hepatic
blood flow appeared to be less than the rise in arterial
pressure indicating a local vasoconstriction less than
that elsewhere in the body. Perhaps the differences
in results are attributable to differences in the
dosage — the lower doses inducing vasodilation of the
mesenteric arterioles by stimulation of /3-receptors.
Certainly if portal venous pressure rises together with
hepatic blood flow, as it seems to in man (187), it is
possible that mesenteric vasodilation dominates the
circulatory pattern, masking a moderate degree of
intrahepatic vasoconstriction. Whether a reflex me-
diated through the central nervous system also con-
tributes (21) remains uncertain. Norepinephrine, in
contrast, is clearly vasoconstrictor but as with epineph-
rine, nothing is known about its effect upon the
individual components in the intact hepatic and
splanchnic bed. The data available are too frag-
mentary and unreliable to permit even a tentative
synthesis.
Direct visual examination of the vessels within the
splanchnic bed gives further support to the view that
medullary amine usually tends to evoke a complex
vasoconstrictive response. Seneviratne (270) and
Wakim (299) agree in reporting that epinephrine
applied directly to the surface of both the mammalian
and amphibian liver produces contraction of the
sinusoids, presumably as a result of constriction of
hepatic arterioles and portal venules. No obvious
effect upon the visible portal and hepatic veins was
evident, however. A similar response was observed
when the drug was injected into the portal vein.
Interestingly, when the drug was given via the vena
cava, the response was delayed and then replaced by
"overactivity of the circulation in the whole liver,
both as to number of active sinusoids and as to en-
gorgement and rate of flow of blood in them." Similar
responses have been reported in the gastrointestinal
tract (132). Serial angiography also yields evidence of
intrahepatic vasoconstriction in rat, rabbit, cat, dog,
and the monkey (101). Intraportal injection of ad-
renalin (10-20 ,ug) resulted in changes like those
produced by stimulation of the splanchnic nerves —
i.e., a reduction in the number of fine vessels demon-
strable by the circulating contrast medium together
with an inconsistent diminution in calibre of the larger
portal vessels. Daniel & Pritchard (101) noted further
that "the rapid transhepatic passage of the portal flow
which was observed after administration of adrenaline
was associated in some but not all experiments with a
change in the distribution of the contrast medium
within the liver. . . . Frequently there was evidence
of an unequal distribution of the blood flow within
the liver, illustrating that a differential use was being
made by the portal venous blood of the various path-
ways through the liver." Much more work is required
to sort out the data available and to evaluate by more
precise methods the pattern of flow and pressure
redistribution throughout the splanchnic bed during
the action of epinephrine and norepinephrine.
acetylcholine. Little consistent change appears in
the isolated perfused splanchnic circulation following
the administration of acetylcholine regardless of the
method employed in its evaluation. Bauer et al. (25)
were unable to obtain a definite response in the
perfused liver of the dog, cat, or goat. If the "arterial
tone" was "high" an intra-arterial injection of ace-
tylcholine appeared to produce a relaxation com-
parable to that induced by histamine. Relatively
small doses intra-arterially, however, acted upon the
perfused goat's liver somewhat like adrenaline, that
is, "arterial and portal tone were increased, liver
volume diminished, and outflow practically
unchanged. . . . this effect of acetylcholine was of the
parasympathetic type, in that it was completely
abolished by atropine, which left those of adrenaline
and histamine unchanged." Chakravarti & Tripod
(81) found that acetylcholine produced an easily
detectable effect upon the circulation through the
perfused liver if adrenaline were added to the per-
fusate in order to provide a "vasoconstricted state"
in which a vasodilator action could be more easily
elicited. Unlike adrenaline or histamine, acetylcholine
had no action when injected into the portal vein.
Andrews et al. (13) reported results similar to these.
In most of their studies of the perfused canine liver,
acetylcholine in doses ranging from 0.35 to 15.0 ng
injected into the hepatic artery produced no change in
arterial inflow, and a fall in both portal inflow and
hepatic outflow with a rise in volume; but when adren-
aline (0.1 fig per ml) was added to the perfusate
acetylcholine raised the arterial inflow and slightly
decreased portal venous inflow with a rise in both
total outflow and volume. When injected into the
portal vein, acetylcholine produced the same response
as that produced by intra-arterial injection but 10 to
1 5 times the amount was required for an equivalent
effect. Administration of eserine equalized the re-
sponse to arterial and venous administration indicat-
ing that the difference might be due to greater de-
struction of acetylcholine in the portal vessels. They
interpreted their results as indicating a weak vaso-
constrictive response in the portal vein and hepatic
1420
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
vein with little or no action upon the arterial circuit.
Acetyl-/3-methylcholine chloride and carbaminoyl
chloride had a similar but longer action and did not
produce different responses depending upon the route
of administration, possibly because neither is readily
destroyed by hepatic esterases. In subsequent studies
of the effect of acetylcholine on the perfused hepatic
vasculature of the monkey, cat, and rabbit, the same
group (12) was impressed by the difficulty of obtaining
reproducible responses for "the vascular responses
showed considerable variation not only from species
to species but also from time to time in the same
animal." Similar equivocal results have been obtained
in studies of isolated gut (27). Indeed, Bean & Sidky
(27) have adduced evidence that the increase ob-
served in blood flow through a perfused intestinal
loop is attributable to the release of vasoactive ma-
terials locally rather than to a direct effect upon the
vasculature. In the case of the spleen, acetylcholine
does seem to have a direct effect at least insofar as
strips of splenic muscle are concerned (53).
In his initial studies in 191 8, Reid Hunt found that
the liver of the intact dog shrank following intravenous
administration of small doses but expanded with
larger ones. A similar response was noted by
McMichael (210) in the cat; the reduction in volume
coincided with a fall in portal venous pressure and
arterial pressure. In a similar study of the dog, Katz
& Rodbard (182) also noted that acetylcholine caused
an isolated fall in portal venous pressure that occurred
at approximately the same time as a fall in arterial
pressure and a rise in peripheral venous pressure.
Portal venous flow (Ludwig stromuhr) tended to
follow the portal venous pressure but did not fall to
the same extent as arterial pressure, suggesting the
development of mesenteric arteriolar dilatation. In
more recent studies, the results also suggest that
vasodilatation may occur in the liver though a change
inflow has not been seen on direct observation (299).
Using implanted devices to measure flow, Ginsburg
& Grayson (147) and Gersmeyer & Gersmeyer (142)
report an increase in hepatic venous outflow in rats
and dogs during intravenous infusion of acetylcholine.
Since arterial pressure fell at the same time, intra-
hepatic resistance must have diminished. An increase
in the volume of the gut and in mesenteric and gastric
outflow has been noted by other workers (132, 219).
The increment in hepatic blood flow cannot be ex-
plained entirely on this basis, however, since Ginsburg
& Grayson (147) found that hepatic blood flow in-
creased in the rat even when portal drainage was
diverted from the liver. The spleen in the intact dog
apparently changes little in volume with a slight
augmentation in arterial inflow despite arterial hy-
potension (164, 223).
Although the reactions, outlined above, to acetyl-
choline in the isolated and intact hepatic vasculature
are obviously equivocal, a change of some kind does
seem to occur. This fact is a little difficult to square
with the absence of cholinergic innervation of the
splanchnic and hepatic vessels and with the lack of
any response to vagal neurectomy and stimulation.
Cholinergic activity is definitely important in the
function of the gastrointestinal musculature and of
the glands of the biliary tract and pancreas. Hence,
it is possible, as Bean & Sidky (27) have suggested,
that secondary release of substances acting locally
upon the blood vessels may be implicated. Brandon
& Rand (53) and Burn & Rand (6g) have recently-
brought forward a more attractive hypothesis that
acetylcholine may release norepinephrine from stores
in the tissues (in the nerves or chromaffin cells) which
may be depleted by reserpine or neural degeneration
and replaced by an infusion of norepinephrine. The
variability of the results reported by many workers
may well be explained, in part at least, by variation
in the stores initially available and in the depletion of
the neurotransmitter during the preparation of the
tissues for study. Stimulation of the adrenal medulla
and a general systemic response must also be taken
into account in the interpretation of the effect of
acetylcholine upon the hepatic circulation in the
intact animal.
autonomic blockade. Autonomic denervation and
chemical interference with autonomic activity have
proved extremely helpful in the study of neurovascular
function. As knowledge has accumulated, however,
the complexities of the problem have become in-
creasingly apparent. Surgical denervation is neces-
sarily limited by the inaccessibility, diversity, and
versatility of the nerve supply. A remarkable array of
autonomic blocking agents is now available and grow-
ing in number and variety with each year. Unfortu-
nately, the supply has outrun the laborious and
gradual gathering of information regarding mode and
site of action. Most of these substances interfere with
both sympathetic and parasympathetic function,
most produce confusing side effects unrelated to the
splanchnic adrenergic blockade. Few have been
studied with special attention to the effects upon
hepatic inflow and outflow tracts. Since this is not the
place to embark upon a detailed examination of the
mechanisms of adrenergic and cholinergic blockade,
THE HEPATIC CIRCULATION
I 42 I
these matters will not be considered here. Interfer-
ence with autonomic transmission, no matter how
produced — whether by competition for receptor sites,
by depletion of stored transmitters, by changes in
membrane permeability or polarization, by change in
synthesis or degradation, or even by surgery — makes
possible a more meaningful appraisal of the relative
importance of neural, humoral, and local circulatory
determinants.
Denervation by any method does not seem to pro-
duce significant change in the blood flow in any part
of the splanchnic and hepatic vascular bed. Mesen-
teric vasodilation, increased hepatic blood flow, and
partial splenic contraction have been reported shortly
after section of the splanchnic nerves, splenic nerve,
and lumbodorsal sympathectomy, but these reactions
seem to be temporary (132). Wilkins and his associates
(310) have found that the EHBF was definitely higher
and splanchnic vascular resistance "significantly"
lower in 13 patients with hypertensive vascular disease
2 weeks after the Smithwick procedure. In six patients
studied 4 to 6 months later, EHBF had returned to
control values. Circulating splanchnic blood volume
has also been found to increase after sympathectomy
(42). High spinal anesthesia induces a fall in EHBF
that may be accounted for by the coincidental fall in
blood pressure in the absence of change in resting
splanchnic resistance (201). Section of the vagus has
no obvious effect (147).
Hexamethonium appears to be capable of blocking
cholinergic transmission within autonomic ganglia
though it does not block the response to sympathetic
nerve stimulation (92, 133, 245). During ganglionic
blockage, blood flow through the splanchnic bed
decreased in dog and man only to the extent to which
arterial pressure was reduced. The circulating
splanchnic blood volume, in contrast, was found to
expand significantly in the dog in the absence of
detectable change in portal venous or sinusoidal
pressures (92). These data are consistent with the view
that basal arteriolar cross section in the splanchnic
and hepatic beds at rest does not depend exclusively,
if at all, upon the integrity of the autonomic nerve
supply. Such a statement does not imply that no
change occurred in the vascular smooth muscle.
Whatever change in tone that may have occurred was
evidently not sufficient to result in dilatation of the
arterioles. It has been suggested that the contrasting
arteriolar and venous reactions can be explained on
the basis of Laplace's law, the difference in radii
accounting chiefly for the relative effectiveness of the
distending pressures at each level.
After administration of adrenergic blocking drugs
(ergotamine, Dibenzyline, Dibenamine, Ilidar), epi-
nephrine, norepinephrine, and neural stimulation
fail to induce the usual arteriolar vasoconstriction in
the liver, gastrointestinal tract, or spleen (159-161,
223). Norepinephrine appears to elicit little or no
change of any kind under these circumstances,
whereas epinephrine now causes vasodilation in the
mesenteric and splenic vessels. The effect of neural
stimulation is also reversed to some extent in the gut
and spleen. Both are without any demonstrable effect
upon the liver. These findings have been interpreted
as evidence for vasodilator /3-receptors in the mesen-
teric and splenic arterioles. Since study of the behavior
of the veins, with special reference to their volume
capacity, has not been made during adrenergic block-
ade, it is impossible to say whether these findings
apply in general to the other vascular levels in each
coupled system. It is also difficult to be certain whether
dilation is an active or passive process or whether it
involves the participation of other substances pro-
duced locally in response to epinephrine. There is no
doubt that general blockade at any level is associated
with a much more marked interference with vascular
responses than is evident in the reversal of responses to
epinephrine. Although the hemodynamic pattern at
rest is not materially affected, any shift in position,
imposition of stress, or environmental change unmasks
a serious loss of capacity to make corrective adjust-
ments. Tilting into the upright position, for example,
results in a sharp drop in arterial pressure without
eliciting the normal compensatory change in splanch-
nic vascular resistance. Pooling of blood within the
splanchnic veins actually enhances the tendency to
circulatory collapse (42). Neural and neurohumoral
mechanisms may not be essential to maintenance of
the resting state but they are clearly necessary for
coordination in systemic responses.
Local Biochemical Determinants
Changes in blood flow result in corresponding
changes in the delivery of oxygen and essential nu-
trients to, and in the removal of metabolites from, the
tissues. Moreover, the associated activity of smooth
muscle, the distention or deflation of capillaries, the
alterations in interstitial pressure and in lymphatic
drainage all impinge directly upon the cells. Meta-
bolic processes are certainly affected by the secondary
shifts in exchange and in the milieu interieur and by
the neurohumoral agents themselves concerned in
these responses. In consequence, vasoactive materials
I4'22
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
oi various kinds, for the most part still uncharacter-
ized, are released locally to play a more or less essen-
tial role in determining the pattern of circulatory
adjustments. Since these materials rarely enter the
blood in amounts sufficient for detection it is extremely
difficult to evaluate their contribution. Tissue gas
exchange is an exception because oxygen and carbon
dioxide content of the blood entering and leaving the
tissues is easily measured, but the influence of other
agents must be studied indirectly.
oxygen. Even assessing the role of the blood gases is
dillicult because it cannot be said with any certainty
how oxygen and carbon dioxide concentrations are
distributed within the capillary network. The abun-
dance of intercommunications makes it likely that a
relatively uniform admixture of blood occurs. Never-
theless, the gradient between artery and vein must
be reflected in a similar tissue differential so that
vessels like those on the periphery of the hepatic
lobules contain more highly oxygenated blood than
those entering the central veins. Hypoxia, per se, seems
to have relatively little effect upon the hepatic circula-
tion. Hypoxia does result in the rapid deterioration of
the perfused liver with the development of obvious
swelling and decreased perfusibility, but whether
these changes are to be ascribed to active vascular
changes or to cellular swelling alone does not seem to
have been subjected to systematic study (132).
Torrance (291) found no evidence of any change in
intrahepatic resistance to flow (internal calorimetry)
in anesthetized rabbits after complete occlusion of the
arterial and venous inflow tracts for 2 min. More
prolonged (2 hours) ischemia of the liver in anesthe-
tized, splenectomized dogs by occlusion of the hepatic
artery and diversion of the portal inflow via an ex-
ternal shunt to the jugular vein produces a complex
splanchnic vascular response, according to Selkurt
(265, 266). He found that arterial blood flow dropped
to 66 per cent of control on restoration of hepatic
perfusion in association with a rise in intrahepatic and
a fall in mesenteric resistances that together resulted
in a marked increase in portal venous pressure. This
work indicates, as does that of Torrance (291), that
"reactive hvperemia" does not develop in the liver
and it agrees with the more recent findings of Fischer
el al. ( 1 28) in showing hepatic arteriolar constriction.
Seneviratne (270) has noted sinusoidal dilatation
after 1 hour of airway obstruction in mice and rats
but in this instance carbon dioxide retention cannot be
eliminated as the cause. Of course, prolonged hypoxia
also elicits widespread compensatory adjustments
in the systemic circulation in which the hepatic and
splanchnic bed might be expected to participate and
which may produce changes opposed to those resulting
from its action locally. Thus, perfusion of the mesen-
teric vasculature in an isolated innervated segment of
a dog's intestine with hypoxic blood resulted in vaso-
dilation and increased flow (26), in line with Selkurt's
observations following protracted anoxia. When the
animal was allowed to breath a low oxygen mixture,
however, reflex mesenteric vasoconstriction developed
to a degree commensurate with the arterial oxygen
content. Mesenteric and portal venous distensibility
also decreases under these circumstances and it may
be presumed that reflex venoconstriction contributes
to the development of portal venous hypertension and
results in a reduction in splanchnic blood volume
during hypoxia.
carbon dioxide. The effect of hypercapnia is par-
ticularly dillicult to follow owing to interference by
an array of striking concomitant adjustments that
include hyperventilation, peripheral vasodilation,
hypertension, tachycardia, and an increase in cardiac
output (247). To circumvent these obstacles, the
splanchnic hemodynamic effects of elevated arterial
carbon dioxide tensions were studied by Epstein and
his associates (123) in normal human subjects during
light general anesthesia with thiopental and nitrous
oxide. Arterial carbon dioxide tension could be main-
tained at a constant level (Paco, = 56 mm Hg on the
average) by mechanically controlled respiration with
an appropriate gas mixture following neuromuscular
blockade with succinylcholine. Under these circum-
stances, interference with the response by an increase
in ventilation could be eliminated. Nevertheless, the
over-all response appeared to be inconsistent and
erratic; mean arterial pressure rising in six subjects,
falling in two, and not changing in five in association
with a fall in EHBF in nine but with little or no change
in EHBF in four. Splanchnic vascular resistance
always increased, however, EHBF changing in accord
with the balance between the perfusing pressure and
resistance. Since blood flow tended to fall, the splanch-
nic vasoconstriction appeared usually to be somewhat
in excess of that elsewhere in the body. Circulatine;
splanchnic blood volume also decreased significantly
in ten subjects and, since hypercapnia has been found
to increase portal venous pressure, the change may be
interpreted as evidence of constriction of splanchnic
veins. The vasoconstrictive response is probably
mediated through the central nervous system, since
several investigators (132, 216) have found that in-
THE HEPATIC CIRCULATION
1423
creased carbon dioxide tension in the blood perfusing
splanchnic blood vessels including those of the liver,
intestines, and spleen, elicits an arteriolar and venous
constrictive response only if the splanchnic nerve
supply is intact. After denervation, an elevation in
PaCo, regularly induces vasodilation. Evidently this
reflex pathway is relatively unaffected by cholinergic
blockade and by light general anesthesia. Whether
sufficiently high concentrations of carbon dioxide
could overcome the opposing reflex activity and dilate
the vessels directly remains unsettled, although Brick-
ner et al. (66) have shown that mesenteric vasodilation
may occur in the intact animal breathing gas mix-
tures that contain more than 8 per cent carbon di-
oxide. Splenic contraction induced by hypercapnia
appears to involve activity of both splanchnic nerves
and adrenals (235). Further work is required to
elucidate the role of adrenergic mediators, the re-
distribution of blood within the splanchnic bed, and
changes in the partition of hepatic inflow during
hypercapnia. In addition, the effect of carbon dioxide
released locally by metabolizing tissues, in amounts
too small to affect the vasomotor centers, deserves
investigation.
histamine. There is no doubt that carbon dioxide is
added to the blood perfusing the tissues and that it is
present, therefore, in varying concentrations which
are certainly in excess of those in the arterial blood at
the capillary level. Various other vasoactive materials
also appear at approximately the same site, though
relatively little is known regarding the mechanisms
and character of their release. Electrolyte shifts and
the production of hydrogen ion may be particularly
important though very little is known about the part
they play in the regulation of the microcirculation.
Another vasoactive substance which appears in high
concentration in all the tissues served by the splanch-
nic circulation is histamine (127). Although it has
excited intense interest and extensive study for more
than a half century, the function of histamine remains
puzzling and controversial. Its action upon the hepatic
and splanchnic circulations has received special atten-
tion. As early as 1899, it was discovered that a striking
engorgement of the liver occurs in the dog during
anaphylactic shock, and not long afterward the same
phenomenon was observed following introduction of
histamine into the portal vein (132). Since hepatic
venous outflow decreased during engorgement of the
liver, it was suggested that contraction of the venous
musculature might act as a throttle mechanism to
occlude the hepatic veins in the dog. Dale (25) and
others (12, 132) have confirmed these results, Dale
showing that the response could be eliminated by-
slitting the hepatic veins of the perfused liver. More-
over, hepatic swelling proved to be less marked or
absent altogether in animals like the cat, goat, or
monkey that have thinly muscled hepatic veins. A
diminution in portal inflow as well as in hepatic
venous outflow also points to the development of in-
creased resistance to perfusion. All these findings,
together with the fact that histamine causes contrac-
tion of the spleen in most animals (132), point to
vasoconstriction as the predominant effect. Else-
where in the body, however, histamine causes striking
arteriolar vasodilation, and in intact animals the net
effect appears to be vasodilator also in the splanchnic
bed. An increase in EHBF in the face of arterial hy-
potension has been observed in man after intramus-
cular injection of histamine phosphate (47). In tin-
anesthetized dogs, moreover, Gersmeyer & Gersmeyer
(142) have found that the velocity (thermistor) of
portal venous blood flow increased sharply, although
portal venous pressure rises very little as arterial and
inferior vena caval pressures fall. Since it seems likely
from what has been said above that outflow resistance
is increased, the increments in portal pressure [which
are quite marked in other studies (308)] and blood
flow are most reasonably explained on the basis of
mesenteric vasodilation in excess of augmented intra-
hepatic resistance. It is equally possible that hepatic
arteriolar dilatation occurs. More precise and detailed
information obtained under properly controlled
conditions is needed to evaluate the simultaneous
changes in splanchnic resistances and in venous
capacity. The data at hand are in accord with growing
evidence that "histamine may actively dilate arterioles
at the same time that it actively constricts veins"
(167) not only as a result of a direct action upon the
vessels but also as a secondary result of adrenal medul-
lary discharge (134, 228, 297). It is also possible that
the action of many other substances is mediated
through histamine release (228).
Still another substance released locally is serotonin,
or 5-hydroxytryptamine, which has been identified
and extensively studied in recent years. Selkurt (269)
reports that it behaves like norepinephrine in causing
vasoconstriction within the isolated mesenteric vessels
of the dog but little is known about its effect upon the
intrahepatic resistances or upon the volume of blood
held within the total splanchnic bed, liver, or spleen.
A steadily growing list of similar activating agents,
including a miscellany of amines and peptides pro-
duced during tissue injury, is making manifest the
1 424
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
inadequacy of current concepts of cardiovascular
regulation and stimulating numerous new investiga-
tions of the hepatic circulation.
Physical Determinants
Neural, neurohumoral, and local chemical influ-
ences operate for the most part reactively to alter
vascular perfusion and content by changing arteriolar
cross section and venous distensibility. As such they
mediate adjustments but do not actually produce
them. In contrast, external physical forces to which
the abdominal viscera are exposed affect the hepatic
and splanchnic vessels directly, frequently eliciting
corrective or compensatory responses in which all the
mechanisms discussed above are called into play.
Movements and gaseous distention of the gastroin-
testinal tract compress, stretch, and variously deform
the mesenteric vasculature with resultant change in
flow and volume that must affect the delivery of blood
to the liver by the portal vein. To what extent shifts
within the mesenteric circuit, repartitioning of the
total hepatic inflow, and redistribution of the splanch-
nic blood volume depend upon activity of the mus-
culature of the gut remains uncertain. These problems
are dealt with at length in Chapter 42 but must be
touched upon here in order to indicate the potential
importance of extravascular forces upon the hepatic
blood supply. In this instance a reciprocal relationship
obtains in that digestion, absorption, and the other
functions of the gastrointestinal tract depend in turn
upon the integrity of the mesenteric vasculature.
Moreover, the integrated function of the liver and
intestine hinges upon movement of absorbed material
through the portal system to the liver. Exercise and
respiration also impinge directly upon the hepatic
circulation by raising intra-abdominal pressure and
often by changing the position of the body. Both
activities are associated with widespread cardiovas-
cular changes affecting all parts of the body. The
splanchnic vasculature participates in these reactions
but it may actually be more markedly influenced
directly by the associated physical effects.
intra-abdominal pressure. The roughly cylindrical
container holding the abdominal viscera and their
vasculatures is almost completely muscular and
capable of actively increasing the pressure within the
peritoneal cavity. Moreover, the gastrointestinal
system is periodically filled with fluids, solids, and gas,
which occupy space and raise the pressure. As long
as the pressures distending the vessels are in excess
of the circumambient pressures, there is little effect
upon the resistance to flow or upon the forces con-
ducive to blood flow. When the external pressure
equals or exceeds the intraluminal pressure, by an
amount determined by the elastic properties of the
vessel, instability develops and collapse occurs at
some critical pressure, provided the contents can be
displaced. Any pressure increment is uniformly dis-
tributed throughout the abdomen, and blood, not
being compressible, must move from an area of high
pressure to one of low pressure to permit collapse of
those vessels in which the intraluminal pressure is
less than the external pressure. This occurs most
readily at the diaphragm, across which the usual
pressure drop is augmented. It may be inferred that
with a rise in intra-abdominal pressure, blood in the
veins close to the diaphragm is expressed and the
vessels collapsed, thus producing a dam at the point
of outlet. Since the arterial perfusing pressure is little
affected, blood continues to pour into the system,
but with the cessation of outflow, pressures gradually
rise and inflow begins to diminish. When the local
pressure at the point of collapse exceeds the critical
opening pressure, outflow is restored and a new equi-
librium established. Whether the expected increase
in resistance so induced would tend to result in pool-
ing of blood within the splanchnic circuit in the face
of the forces operating to prevent distention must
depend upon the relationship between the distribu-
tion of resistances and pressure gradients within the
bed. Little information is available on which one may-
base further speculation. Increased intra-abdominal
pressure has been found to reduce hepatic blood
flow in man and experimental animals (47, 221),
but precise localization of the sites of increased re-
sistance and the character of volume capacity changes
have not been satisfactorily settled.
gravity. Gravitational force is also constantly opera-
tive in affecting intravascular pressures within the
abdomen. Change of position results in a change in
the hydrostatic pressure at every point in the vascula-
ture by means of an addition or subtraction of a
column of blood, the height of which depends upon
the elastic properties of the vascular system. In com-
puting the hydrostatic head at any point in the vessels,
it is customary to take the center of the right atrium
as the zero base line in recumbency. All pressures
above that level in the supine position are negative
with reference to it, and all below are positive. With
a shift in position, the levels at which pressure in the
arteries and veins remains unchanged mav be taken
THE HEPATIC CIRCULATION
1425
as the zero reference planes for the respective sys-
tems. In man, the arterial zero reference plane lies
at the level of the diaphragm immediately after tilt-
ing into the upright position, with moderate change
thereafter during vasomotor adjustments (309).
Thus the elastic properties of the vasculature are
such as to maintain the arterial pressure relatively
constant at the level of the diaphragm; pressures at
points above, falling, and below, rising, solely by the
weight of the column of blood lying between each
point and the zero reference level. The imposed blood
column does not reach to the level of the uppermost
body surface, e.g., the top of the head, because the
closed elastic container exerts an attractive force in
supporting that portion of the blood above the refer-
ence plane. A similar elastic buffering of hydrostatic
shifts occurs in the veins. In both dog (86) and man
(309) the venous side of the circulation is divided
dynamically by cardiac activity into two separate
hydrostatic compartments each with a separate
zero reference plane. The immediate hydraulic
changes with change of position are minimized on
both the arterial and venous side and little immediate
change in hemodynamics occurs in dog or man.
Within a few seconds after assumption of the head-up
position, however, the blood pressure does tend to
fall and the heart rate to speed. Widespread vaso-
constriction quickly checks the decline in arterial
pressure. The splanchnic vasculature partakes in this
general response, since it has been observed that
splanchnic blood flow (EHBF) decreases significantly
in man during orthostasis (99). This change is im-
paired in hypertensive patients following lumbodorsal
sympathectomy (310). Circulating splanchnic blood
volume is also reduced in the upright position pre-
sumably as a result of reflex alteration in the splanch-
nic vascular capacity (42). Further work is needed
to define these changes in experimental animals.
respiration. The intra-abdominal pressure figures
prominently in a more active sense as one of the forces
involved in determining changes in splanchnic
blood flow during respiration. Evaluation of the
changes in pressure gradients and flows during the
respiratory cycle is complicated by the difficultv of
defining precisely what takes place in general terms.
The number of factors involved and the variety of
combinations possible under different circumstances
makes generalization extremely hazardous. Even the
dynamics of quiet breathing in recumbency in man
can vary from time to time and from person to person,
depending upon the individual pattern of abdominal
and thoracic muscular interplay, fatigue, extent of
gastrointestinal and bladder filling, apprehension
and the state of consciousness, muscular development,
pulmonary or cardiac dysfunction, blood volume,
and many other variables. It is to this irregularity
that a remarkable diversity of opinion must be at-
tributed (64, 132).
The fact that pressure within the thorax tends to
be lower than atmospheric pressure during inspira-
tion and somewhat higher during expiration is self-
evident; the question that is difficult to answer is how
and to what extent this phasic change in pressure is
transmitted to the splanchnic vasculature. With
descent of the diaphragm during inspiration the vis-
cera are forced into the abdominal cavity, the vascu-
lar bed subjected to shortening and buckling, and the
liver and spleen compressed to some extent as they
are thrust out of the thorax. During expiration shifts
in the opposite direction must occur. In association
with these changes, opposing changes in intra-abdom-
inal and thoracic pressure probably occur in such a
way as to increase the pressure gradient between
abdomen and thorax during inspiration and to re-
duce it during expiration. In the main, these in-
ferences find support in the experimental record but
they must be modified under many conditions.
For example, it is possible for contraction of the
abdominal muscles during expiration and relaxation
during inspiration to counter the usual effect easily
and to produce a reversed pattern. Indeed with
maximal or laborious breathing in man, this seems
to be the rule (73). In the upright position the super-
imposed hydrostatic pressure changes also have an
effect. As noted above, the intra-abdominal pressure
under these circumstances is governed to some extent
— modified, of course, by muscular tone and activity —
by the hydrostatic forces and the shift in arterial
and venous zero reference planes. The weight of the
upper abdominal viscera may be supported in part
by the retractive forces of the thorax and its contents
and, with impeded diaphragmatic excursion, an
abdominal component in the respiratory pattern
may figure more prominently. The magnitude and
direction of thoracico-abdominal pressure shifts
may have direct bearing upon blood flow into or
out of the splanchnic vasculature or upon the quan-
tity of blood held within it at any moment. This is
so because collapse of the large draining venous
channels may increase resistance to flow more than
the rise in the pressure gradient tends to enhance it.
This point has been much debated and still remains
unsettled. On the one hand, Holt (174) and Duo-
1426
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
marco & Rimini (117) believe that reduction in
venous cross section at the level of the diaphragm
and within the superior vena cava will impede return
of blood to the heart during inspiration. Thus, radio-
micrographic studies in the rat indicate that hepatic
outflow may be more rapid during expiration (58).
This thesis has been vigorously disputed, on the other
hand, by the group with Wiggers in Cleveland which
has included Opdyke (222), Alexander (3), and
Brecher (64). These workers have shown that,
though venous collapse may increase resistance with
an increased pressure drop at the diaphragm, there
is nonetheless an inspiratory increase in return,
provided the tendency to venous collapse is not ex-
aggerated by marked changes in the gradient, and
provided alteration in the timing of phasic changes
does not predispose to "depletion" of the venous
chamber with resultant collapse. Both groups are
probably correct, however, in view of the manner
in which the response may be altered by extraneous
factors, such as respiratory rate, position, and body
size. During inspiration under ordinary circumstances
the blood held in the hepatic venous tree and the
inferior caval system flows out somewhat more
rapidly than inflow so that venous return is initially
augmented — and splanchnic outflow increased. The
accompanying compression of outflow channels and
the rise in intra-abdominal pressure would operate
to reduce inflow so that net flow might change very
little, rise or fall, depending upon the duration and
frequency of the inspiration phase. Direct observation
as well as measurement of flow bears out this con-
clusion, at least for the dog. The increased outflow
resistance may actually conduce to portal venous
pooling (3), again depending upon the interplay of
all the other factors concerned. It may actually be
rather difficult to define splanchnic flow and volume
under these circumstances because the '"depleting"
phase of inspiration may be followed by filling during
expiration not only from the arterial side but also by
retrograde flow from the right atrium into the hepatic
venous chamber as angiographic studies have shown.
The effect of anatomic and dimensional differences
requires further study.
exercise. The effect of exercise must be determined
to a large extent by the manner in which it affects
respiratory activity, intra-abdominal pressure, and
gas exchange as well as by release of various vaso-
active agents. In quietly resting human subjects in
recumbency, exercise (alternate leg raising) induces a
significant reduction in both hepatic blood flow and
splanchnic blood volume presumably as a result of
vasoconstrictive activity (298). Reallocation of
splanchnic blood volume seems to occur quickly
and may indeed play a role in the maintenance of
cardiac output prior to the establishment of a new
equilibrium. When blood pressure rises, the fall
in splanchnic blood flow may not occur in spite of
vasoconstriction. This phenomenon has been ob-
served in dogs (superior mesenteric arterial flow-
measured by thermostromuhr) exercised on a tread-
mill (172). Possibly there was a similar absence of
change in EHBF, in the face of an increment in he-
patic temperature, in three normal human subjects
studied by Graf (152). The Bromsulfalein clearance is
of questionable value in evaluating the effect of exer-
cise, since BSP extraction tends to increase in associa-
tion with a rise in hepatic arteriovenous oxygen
difference (34, 200). Barcroft and his associates (20)
found that exercise (running) caused a significant
splenic contraction in dogs and cats, which tended to
persist in proportion to the duration and severity of
exertion. A definite pattern of response of splanchnic
arteriolar and venous constriction can be made out
but, in view of the varied and opposing forces that
are brought into play during exertion, a diversity
of responses is probably the rule in normal life.
HEPATIC CIRCULATORY INTEGRATION
AND DYSFUNCTION
Hepatosystemic Interrelationships
The participation of the hepatic and splanchnic
circulation in general systemic reactions is usually
diflicult to detect and to delineate. The changes ob-
served during exercise, assumption of the upright
position, and respiration have been noted above
because they entail a direct effect upon the intra-
abdominal vasculature. In addition, any tendency
for cardiac output or arterial pressure to fall or to
rise is associated with concomitant changes in hepatic
blood flow and splanchnic blood volume. In the
main, these adjustments appear to provide for con-
tinued perfusion of the liver without undue inter-
ference with corrective responses elsewhere in the
body to restore the status quo ante. Owing to the com-
plexities of the splanchnic circuitry, however, the
precise mechanisms of local adjustments are usually
obscure. Little or no information is available regard-
ing minor shifts. More is known about adjustments in
such extreme disorders as circulatory collapse and
congestive heart failure. Unfortunately, the need for
THE HEPATIC CIRCULATION
[427
anesthesia in the experimental study of shock in-
troduces an additional variable. General anesthesia
with nitrous oxide, thiopental, or pentobarbital does
not seem to affect EHBF and SBV provided gas
exchange is carefully controlled (123, 124). Reactions
may differ with the agent employed [splenic volume
is increased, for example, by barbiturates and de-
creased by ether (132)] and more carefully controlled
studies are necessary to evaluate the effects of dif-
ferent dosage levels and anesthetic planes. Anesthetic
drugs also act like autonomic "blockers" to diminish
responsiveness so that data collected in experimental
studies may not be strictly relevant to the clinical
state. Studies in man have been helpful for this reason,
though complete control is impossible. The hepatic
circulatory changes of congestive heart failure have
been investigated only in man and, though the data
are of great value, reliable evaluation must wait upon
definitive studies of the condition produced experi-
mentally in laboratory animals.
The effects of hemorrhage have been most ex-
tensively explored. In the anesthetized (pentobarbital)
dog and rat, splanchnic blood flow and volume de-
crease during and following blood loss (136, 137,
179, 243, 267). Blood flow appears to diminish in
proportion to the fall in arterial pressure. A transient
vasoconstriction may occur but the general tendency
appears to be rather in the direction of moderate
vasodilation, particularlv affecting the hepatic ar-
terioles. Mesenteric arteriolar constriction may de-
velop after prolonged hemorrhagic hypotension (96,
136) and may even be enhanced by the administra-
tion of norepinephrine, but there is no evidence that
it is effective in sustaining arterial pressure (193).
Portal venous pressure usually falls, too, presumably
as a result of both the shift in the balance between
input and output resistances (central venous pressure
also falls) and the drop in arterial pressure. Splanch-
nic blood volume decreases more than the total
blood volume (137, 243). The reduction in distend-
ing pressure is probably a major determinant of the
shift, but venoconstriction, demonstrable in isolated
preparations (6), must play a part also and un-
doubtedly accounts for the continued reduction in
splanchnic blood volume following restoration of
blood volume at a time when portal venous pressure
tends to rise. Certainly, splenic contraction occurs
in most species (132). The splanchnic venous reservoir
evidently participates as a whole in homeostatic
compensations by actively transferring blood into
the central veins and sustaining the "circulating blood
volume." This response continues to be detectable
even in the terminal irreversible phase (137, 243).
How and to what extent anesthesia, adrenal medul-
lary discharge and neural activity contribute to or
modify hepatic vasomotor adjustments remains un-
settled. Studies of changes during hemorrhage in
human volunteers suggest that splanchnic vaso-
constriction may be more prominent in the absence
of anesthesia (28).
A definite vasoconstrictive pattern is clearlv charac-
teristic of congestive heart failure in patients with
various cardiac diseases (236). In this situation,
EHBF has been found to be reduced to the same ex-
tent as cardiac output despite maintenance or even
elevation of the arterial blood pressure, indicating
vasoconstriction no greater than that elsewhere in
the body and certainly much less than that occurring
in the kidney. A uniform contraction of hepatic and
splanchnic arterioles probably occurs without much,
if any, change in postsinusoidal or portal venular
resistance, since wedged hepatic venous pressure did
not differ from the free hepatic venous pressure by
more than 2 mm Hg according to Rapaport and his
associates (236). Circulating splanchnic blood volume
is disproportionately increased by cardiac failure,
at least in those patients in whom atrial and wedged
hepatic venous pressure is elevated. There is no evi-
dence at present of either splanchnic venoconstriction
or venodilation and it is necessary to conclude, for
the moment, that the distention is passive. The se-
questration of a larger portion of the blood volume
in the splanchnic bed may effectively reduce the load
imposed upon the heart and, in so doing, serve as a
compensatory device. The large volume of blood
within the splanchnic veins is an ever-present hazard,
however, for exertion, violent respiratory movements,
or increased intra-abdominal pressure may displace a
large volume of blood from the abdomen and throw
an additional, and perhaps an overwhelming, burden
upon the heart at almost any time. More data are
needed to evaluate this possibility and to assess the
role of the hepatic vasculature in the pathogenesis
and therapy of congestive heart failure and other
cardiovascular disorders.
Hepatosplatulmic Inter) elationships
The interdependence of the liver and the gastro-
intestinal tract is self-evident. Digestion and absorp-
tion depend upon normal biliary secretion, while the
enterohepatic circulation of bile salts and the release of
secretin, in turn, determine bile flow and composition.
Water, electrolytes, and various organic compounds
[428
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
move rapidly from the gut into the portal venous
blood and are carried directly to the liver. Since the
bulk of the blood bathing the parenchymal cells
comes from the portal vein, and since the sinusoidal
walls are completely permeable to large molecules
(209), the extracellular tissue fluid of the hepatic
parenchyma must vary much more widely in com-
position, osmolarity, and acidity than interstitial
fluid elsewhere in the body. Cellular function is
undoubtedly influenced by the milieu interieur as a
whole as well as by certain active ingredients in it.
The gastrointestinal tract may be said to control the
chemical environment of liver cells not only by its
absorptive activity but also by its oxygen consump-
tion, for the liver must be content with the leavings of
the gut. Owing to the difficulty of sampling portal
venous blood, little is known about the fluctuations
in the chemical composition of portal venous blood
and their impact upon hepatocellular function. Much
more, but still too little, is known about the manner
in which portal venous pressures may be affected by
the interplay between hepatic and splanchnic re-
sistances.
The behavior of portal venous blood flow and pres-
sure under normal conditions has been discussed at
length above and the pattern of partition will be
covered in another chapter. Since there is a lack of
data on the distribution of flow and pressures through-
out the total hepatosplanchnic vasculature, the rela-
tive importance of pre- and postportal vein resistances
cannot be clearly defined. The fact that portal venous
inflow usually accounts for some two-thirds to three-
quarters of the hepatic venous outflow points to domi-
nance by the preportal resistances (35, 37, 115,
132, 204, 255). The rise in portal venous pressure
during the action of epinephrine may be ascribed
therefore largely to mesenteric (subsuming under
this term the total preportal bed) vasodilation, the
fall with norepinephrine to mesenteric vasoconstric-
tion. Vasopressin (103, 142, 308) has also been
found to be a most effective agent in lowering portal
venous pressure in man and dogs by contraction of
the mesenteric arterioles in association with a reduc-
tion in hepatic blood flow. The resulting rearrange-
ment of pressure gradients may result in a fall in
sinusoidal pressure, so that the much weaker hepatic
arteriolar constriction may be effectively countered
by the rise in arteriosinusoidal pressure difference
and hepatic arterial inflow actually increased (171).
There is relatively little evidence of an actively
maintained balance between arterial and portal
venous flow to the liver, although the arrangement of
resistances noted above does lead to an apparent
reciprocity when one or the other inflow is predomi-
nantly affected (36, 158). Ligation of the portal vein
is thus immediately followed by an increase in hepatic
arterial inflow, up to 100 per cent above control, and
hepatic arterial ligation has a similar effect upon
portal venous flow, but the increment fails to restore
total flow to the control level (158, 254). Perhaps
subsequent changes in tissue function set in train
delayed corrective adjustments that assure adequate
perfusion, but local mechanisms to provide immedi-
ately for reciprocity seem to be lacking. Indeed,
portal venous pressure may be persistently reduced
after hepatic arterial ligation so that an increment in
portal inflow fails to make up the deficit in perfusion.
A similar phenomenon has been encountered in
patients with cirrhosis (52) where parenchymal
cellular damage and extensive fibrosis have grossly
deformed the architecture of the liver and its vascula-
ture. The blood flow through the cirrhotic liver is
significantly reduced by attenuation of the total
vascular bed and by compression and distortion of
the hepatic venous outflow tract. The resultant ele-
vation in sinusoidal and portal venous pressure is
often combined with a fall in plasma albumin con-
centration. Portal venous hypertension appears to
be the primary event responsible for increased move-
ment of fluid across capillary and sinusoidal walls
and for the formation of ascites. Secondary cir-
culatory, humoral, and renal changes are also essen-
tial features [see (16) and (196) for a recent examina-
tion of this problem]. The portal venous pressure
may be markedly diminished by portacaval anasto-
mosis, in association with a significant fall in hepatic
blood flow that tends to persist without any evi-
dence of hepatic arteriolar dilatation. Hepatic venous
oxygen concentration is well below normal in cir-
rhotic patients and it falls still lower after establishment
of a portacaval shunt, indicating further that hepatic
arterial and portal venous inflows are not neces-
sarily correlated. The relative independence of the two
vascular supplies may indeed contribute in the patho-
genesis and perpetuation of cirrhosis (289).
Under normal resting conditions the gastroin-
testinal vasculature could conceivably determine
hepatic function through its domination of sub-
strate supply, but, in fact, the hepatic blood supply
appears to be adjusted to the metabolic requirements
of the body as a whole. Digestion and absorption,
as such, do not affect hepatic blood flow. Ingestion
THE HEPATIC CIRCULATION
I429
of protein, possibly carbohydrates, but not fat, is
followed by the development of hepatic hyperemia
in man (54, 152, 200, 240). Reininger & Sapirstein
(240) have found that hepatic blood flow increases
in rats after a protein meal, in proportion to the rise
in cardiac output and blood flow to other tissues
that occurs at the same time. Similar changes in
systemic and hepatic circulation have been de-
tected in man during febrile reactions to pyrogenic
agents (46, 152, 170) in association with increased
total oxygen consumption. Liver temperature rises
after protein feeding and during fever (152), pre-
sumably as a result of augmented hepatocellular
metabolism. When hepatic oxygen consumption is
increased by thyrotoxicosis, EHBF does not change
appreciably (218). According to Bondy and others
(38) uncontrolled human diabetes is not associated
with a significant change in EHBF, although Lips-
comb & Crandall ( 1 97) have observed high values
in diabetic dogs. A definite increment in EHBF has
been observed in dogs also during hyperglycemia
produced by glucagon administration and hypogly-
cemia produced by insulin (112, 276, 279). Epi-
nephrine release may be involved in the latter and
must, indeed, be weighed in the evaluation of hepatic
hyperemia, whatever its cause. Sympathoadrenal
activation is an unlikely participant, however, in the
action of /-hydrazinophthalazine which has been
found (207) to elicit a pattern of circulatory and
metabolic adjustments, in every respect like that
produced by pyrogen, except that body temperature
does not rise. If the expansion in splanchnic blood
volume during the action of hydralazine in dogs is
typical of the hepatic hyperemic reaction in general,
it may be concluded that vascular smooth muscle is
uniformly affected with simultaneous arteriolar and
venous dilatation. Moreover, the tendency for ar-
terial pressure to fall to low levels in these conditions
may be the result of interference with venous return
by "splanchnic pooling." Hyperemic responses de-
serve careful study not only for the light that may be
cast upon normal hepatosplanchnic interrelation-
ships but also for better understanding of derange-
ments in integration that may be involved in the
production of hepatic disorders.
Throughout the foregoing discussion attention
has been directed chiefly to the local and systemic
factors that may determine hepatic blood flow in
health and disease. Admittedly the account is sketchy
owing to inadequacies of the author and the space
available. An effort has been made to cover in some
detail the elements of hepatic and splanchnic hemo-
dynamics. The role of the preportal systems in the
spleen, gastrointestinal tract, and pancreas have
been alluded to frequently but it has been difficult to
give these factors the weight they deserve, chiefly
because the evidence available is so fragmentary and
questionable. As noted at the outset, methodology-
must take first rank as a cause for uncertainty. Dif-
ficulty in generalization arises also from dependence
upon data drawn from but one, or too few, experi-
mental animal species; from acute responses tram-
meled by unphysiologic conditions of anesthesia,
restraint, and surgery; and, finally, from portions,
rather than the totality of any reaction. Emphasis
has been placed upon correlation of structure and
function. For this reason, among others, the arteriolar
resistances have been stressed as determinants of
flow and pressure gradients. The volume of blood
contained within the vasculature has been assigned
chiefly to the large veins and translocations or re-
arrangement of content ascribed, therefore, to
alterations in venous smooth muscle function and
inlet-outlet balance. Neither of these inferences is
invalidated by the possibility touched upon at several
points, that extravascular influences may dominate
more slowly developing changes. Tissue turgor, fibro-
sis, extravascular cellular or fluid infiltration, and
distortion by compression or traction may affect
vascular path lengths and numbers, as well as cross
sections, with a corresponding effect upon resistance
and volume capacity. Much remains to be learned
about the dynamics of flow in sinusoidal capillaries
and it is possible that conventional explanations
will ultimately prove inadequate. The hepatic vascu-
lature and the splanchnic reservoir proximal to it
participate in systemic circulatory reactions, but the
evidence suggests that maintenance of hepatocellular
function has priority and that homeostatic adjust-
ments operate solely to produce a state that does not
actively impair over-all compensation, without adding
much to it. Perhaps the shift of blood from the
splanchnic bed is helpful but the data cannot be con-
strued to favor conclusively an active rather than a
passive role. The liver is undoubtedly essential to
metabolic activity, but the role of the hepatic
circulation in metabolic homeostasis remains to be
elucidated. In this direction, hepatic circulatory
physiology may stand upon the threshold to signifi-
cant discoveries.
'43«
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
REFERENCES
i. Ahlquist, R. P., J. P. Taylor, C. VV. Rawson, Jr., and
V. L. Sydow. Comparative effects of epinephrine and
levarterenol in intact anesthetized dog. J. Pharmacol.
Exptl. Therap. 110:352, 1954.
2. Aidar, O., W. A. Geohegan, and L. H. Ungewitter.
Splanchnic afferent pathways in the central nervous
system. J. Neurophysiol. 15: 131, 1952.
3. Alexander, R. S. Influence of the diaphragm upon
portal blood flow and venous return. Am. J. Physiol.
16/: 738, I951-
4. Alexander, R. S. The influence of constrictor drugs on
the distensibility of the splanchnic venous system,
analyzed on the basis of an aortic model. Circulation
Research 2: 140-147, 1954.
5. Alexander, R. S. The participation of the venomotor sys-
tem in pressor reflexes. Circulation Research 2 : 405-409,
'954'
6. Alexander, R. S. Venomotor tone in hemorrhage and
shock. Circulation Research 3: 181 -190, 1955.
7. Alexander, R. S., W. S. Edwards, and J. L. Ankenev.
The distensibility characteristics of the portal vascular
bed. Circulation Research 1: 271-277, 1953.
8. Alexander, VV. F. The innervation of the biliary system.
J. Comp. Neurol. 72: 357-370, 1940.
9. Amassian, V. E. Fiber groups and spinal pathways of
cortically represented visceral afferents. J. Neuro-
physiol. 14: 445^400, 1951.
10. Andrews, W. H. H. A technique for perfusion of the
canine liver. Ann. Trap. Med. Parasitol. 47: 146-155,
■953-
11. Andrews, W. H. H., R. Hecker, and B. G. Maegraith.
Observations on the innervation of the hepatic blood
vessels. Ann. Trop. Med. Parasitol. 52: 500-507, 1958.
12. Andrews, W. H. H., R. Hecker, and B. G. Maegraith.
The action of adrenaline, noradrenaline, acetylcholine
and histamine on the perfused liver of the monkey,
cat, and rabbit. J. Physiol., London 132: 509-521, 1956.
[3. Andrews, VV. H. H., R. Hecker, B. G. Maegraith,
and H. D. Ritchie. The action of adrenaline, L-nor-
adrenaline, acetylcholine and other substances on the
blood vessels of the perfused canine liver. J. Physiol.,
London 128:413-434, 1955.
14. Andrews, VV. H. H., B. G. Maegraith, and T. G.
Richards. The effect upon Bromsulphalein extraction
of the rate and distribution of blood flow in the per-
fused canine liver. J. Physiol., London 131 : 669-677, 1956.
1 5. Atkinson, M., E. Barnett, S. Sherlock, and R. E.
Steiner. The clinical investigation of the portal circu-
lation, with special reference to portal venography.
Ojiart. J. Med. 24: 77-94, 1955.
16. Atkinson, M., and M. S. Losowsky. The mechanism of
ascites formation in chronic liver disease. Quart. J.
Med. 30: 153-166, 1 961 .
17. Atkinson, M., and S. Sherlock. Intrasplenic pressure as
index of portal venous pressure. Lancet 1: 1 325-1 327,
'954-
18. Bain, VV. A., J. T. Irving, and B. H. McSwiney. The
afferent fibers from the abdomen in the splanchnic
nerves. J. Physiol., London 84: 323-333, 1935.
19. Banaszak, E. F., VV. J. Stekiel, R. A. Grace, and J. J.
24-
25-
Smith. Estimation of hepatic blood flow using a single
injection dye clearance method. Am. J. Physiol. 198:
877-880, i960.
20. Barcroft, J., H. A. Harris, D. Orahovats, and R.
Weiss. A contribution to the physiology of the spleen.
J. Physiol., London 60: 443-456, 1925.
21. Barcroft, H., and H. J. C. Swan. Sympathetic Control of
Human Blood Vessels. London: Arnold, 1953.
22. Barlow, T. E., F. H. Bentlev, and D. N. Walder.
Arteries, veins, and arteriovenous anastomoses in the
human stomach. Surg. Gynecol. Obstet. 93: 657-671, 1951.
23. Barnett, C H., and VV. Cochrane. Flow of viscous
liquids in branched tubes — with reference to the hepatic
portal vein. Nature 177: 740-742, 1956.
Bartlett, F. K., H. J. Corper, and E. R. Long. The
independence of the lobes of the liver. Am. J . Physiol.
35 : 3°-5°. '9'4-
Bauer, VV., H. H. Dale, L. T. Poulsson, and D. VV.
Richards. The control of circulation through the liver.
J. Physiol., London 74: 343-375, 1932.
26. Bean, J. W., and M. M. Sidky. Effects of low O2 on
intestinal blood flow, tonus and motility. Am. J. Physiol.
'89: 54'-547, 1957-
27. Bean, J. \V., and M. M. Sidky. Intestinal blood flow
as influenced by vascular and motor reactions to acetyl-
choline and carbon dioxide. Am. J. Physiol. 194: 512-
518, 1958.
28. Bearn, A. G., B. Billing, O. G. Edholm, and S. Sher-
lock. Hepatic blood flow and carbohydrate changes in
man during fainting. J. Physiol., London 115: 442-455,
■951-
Bearn, A. G., B. Billing, and S. Sherlock. The effect
of adrenaline and nor-adrenaline on hepatic blood
flow and splanchnic carbohydrate metabolism in man.
J. Physiol. , London 115: 430-44 1 , 1 95 1 .
30. Bennett, H. S., J. H. Luft, and J. C. Hampton. Mor-
phological classifications of vertebrate blood capillaries.
Am. J. Physiol. 196: 381-390, 1959.
31. Bergel, D. H. The static elastic properties of the arterial
wall. ./. Physiol., London 156: 445-457, 1961.
32. Bierman, H. R., K. H. Kelly, L. P. White, A. Coblentz,
and A. Fisher. Transhepatic venous catheterization and
venography. J. Am. Med. Assoc. 158: 1 331 -1334, 1955.
33. Biozzi, G., B. Benacerraf, B. N. Halpern, C Stiffel,
and B. Hillemand. Exploration of the phagocytic func-
tion of the reticuloendothelial system with heat denatured
human serum albumin labeled with I131 and application
to the measurement of liver blood flow, in normal man
and in some pathologic conditions. J. Lab. Clin. Med. 51 :
230-239, 1958.
34. Bishop, J. M., K. VV. Donald, S. H. Taylor, and P. N.
Wormald. Changes in arterial-hepatic venous oxygen
content difference during and after supine leg exercise.
J. Physiol., London 137:309-317, 1957.
35. Blalock, A., and M. F. Mason. Observations on the
blood flow and gaseous metabolism of the liver of unan-
esthetized dogs. Am. J. Physiol. 117: 328-334, 1936.
36. Bollman, J. L., and J. H. Grindlay. Hepatic function
modified by alteration of hepatic blood flow. Gastroen-
terology 25: 532-539, 1953.
29-
THE HEPATIC CIRCULATION
'431
37. Bollman, J. L., M. Khattab, R. Thors, and J. H. 56.
Grindlav. Experimentally produced alternations of
hepatic blood Mow. A.M. A. Arch. Surg. 66: 562-569, 1 953.
38. Bondv, P. K, \V. L. Bloom, V S. Whither, and B. W.
Farrar. Studies of the role of the liver in human carbo-
hydrate metabolism by the venous catheter technic. II. 57.
Patients with diabetic ketosis, before and after the admin-
istration of insulin. J. Clin. Invest. 28: 11 26-1 133, 1949.
39. Boulter, P. S., and A. G. Parks. Submucosal vascular
patterns of the alimentary tract and their significance. 58.
Brit. J. Surg. 47: 546-550, i960.
40. Bover, G. O., and A. M. Scher. Significance of mesen-
teric arterial receptors in the reflex regulation of systemic 59.
blood pressure. Circulation Research 1 3 : 845-848, 1 960.
41. Bradley, S. E. Clinical aspects of hepatic vascular
physiology. Josiah Macy Jr. Con/, on Liver Injury Trans. 60.
■95°. 7'-9°-
42. Bradley, S. E. Integration of the splanchnic circulation
in systemic hemodynamic adjustments. Proc. Ann. Meet-
ing, Council for High Blood Pressure Res. Am. Heart Assoc. 61.
4: 11-24, >955-
43. Bradley, S. E. Methods for evaluation of the splanchnic
circulation. Circulation. Proceedings of the Harvey Tercen-
tenary Congress, edited by J. McMichael. 1958, 255-265. 62.
44. Bradley', S. E. Structural and functional parameters of
the normal splanchnic circulation. Proc. Third World
Congress Cardiology. Symposia 1958, pp. 239-248.
45. Bradley, S. E. The excretory function of the liver. Harvey 63.
Lectures 54: 131 -155, 1959.
46. Bradley', S. E. Variations in hepatic blood flow in man
during health and disease. New Engl. J. Med. 240: 456-
461, 1949. 64.
47. Bradley, S. E., F. J. Ingelfinger, and G. P. Bradley.
Determinants of hepatic haemodynamics. Ciba Foundation 65.
Symposium, Vist nal Circulation. 1953, pp. 219-232.
48. Bradley, S. E., F. J. Ingelfinger, and G. P. Bradley. 66.
Hepatic circulation in cirrhosis of the liver. Circulation 5 :
4 '9-429. '952-
49. Bradley, S. E., F. J. Ingelfinger, G. P. Bradley', and 67.
J.J. Curry'. The estimation of hepatic blood flow in man.
J. Clin. Invest. 24: 890-897, 1945. 68.
50. Bradley', S. E., P. A. Marks, P. C. Reynell, and J.
Meltzer. The circulating splanchnic blood volume in 69.
dog and man. Trans. Assoc. Am. Physicians 66: 294-302,
'953-
51. Bradley, S. E., J. F. Nickel, and E. Leifer. The distri- 70.
bution of nephron function in man. Trans. Assoc. Am.
Physicians 65: 147-158, 1952.
52. Bradley, S. E., C. McC. Smythe, H. F. Fitzpatrick, 71.
and A. H. Blakemore. The effect of a portacaval shunt
on estimated hepatic blood flow and oxygen uptake in
cirrhosis. J. Clin. Invest. 32: 526-537, 1953. 72.
53. Brandon, K. W., and M. J. Rand. Acetylcholine and the
sympathetic innervation of the spleen. J. Physiol., London
157: 18-32, 1961. 73.
54. Brandt, J. L., L. Castleman, H. D. Ruskin, J. Green-
wald, and J. Kelly, Jr. The effect of oral protein and
glucose feeding on splanchnic blood flow and oxygen
utilization in normal and cirrhotic subjects. J. Clin. Invest. 74.
34: 1 01 7-1025, 1955.
55. Brauer, R. W., R. J. Holloway, and G. F. Leong.
Temperature effects on radiocolloid uptake by the isolated
rat liver. Am. J. Physiol. 189: 24-30, 1957. 75.
Brauer, R. W., G. F. Leong, R. F. McElroy, and R. J.
Holloway. Circulatory pathways in the rat liver as re-
vealed by P32 chromic phosphate colloid uptake in the iso-
lated perfused liver preparation. .4m. J. Physiol. 184:
593-598. '956-
Brauer, R. W., G. F. Leong, R. F. McElroy, Jr., and
R. J. Holloway. Hemodynamics of the vascular tree of
the isolated rat liver preparation. Am. J. Physiol. 186: 537—
542, '956-
Brauer, R. W., R. F. McElroy-, Jr., and G. F. Leong.
Blood flow in the hepatic veins of the rat (motion picture).
J. Physiol., London 3: 28, i960.
Brauer, R. W., and R. L. Pessotti. Hepatic uptake and
biliary excretion of bromsulphthalein in the dog. .4m. J.
Physiol. 162:565-574, 1950.
Brauer, R. W., R. L. Pessotti, and J. S. Krebs. The
distribution and excretion of S35 -labeled sulfobromophthal-
ein-sodium administered to dogs by continuous infusion.
J. Clin. Invest. 34: 35-43, 1955.
Brauer, R. W., R. L. Pessotti, and P. Pizzolato. Iso-
lated rat liver preparation. Bile production and other
basic properties. Proc. Soc. Exptl. Biol. Med. 78: 1 74-1 81,
■951-
Brauer, R. \V., O. S. Shill, and J. S. Krebs. Studies
concerning functional differences between liver regions
supplied by the hepatic artery and by the portal vein. J.
Clin. Invest. 38: 2202-22 14, 1959.
Braunwald, E., A. P. Fishman, and A. Cournand. Es-
timation of volume of a circulatory model by the Hamilton
and the Bradley methods at varying flow volume ratios.
J. Appl. Physiol. 12:445-447, 1958.
Brecher, G. A. Venous Return. New York: Grune & Strat-
ton, 1956, 148 pp.
Brendle, E. Uber den Bau der Menschlichen Pfortader
und ihrer Wurzeln. Acta Anal. 10: 108-129, 1950.
Brickner, E. W., E. G. Dowds, B. Willitts, and E. E.
Selkurt. Mesenteric blood flow as influenced by pro-
gressive hypercapnia. .4m. J. Physiol. 184: 275-281, 1956.
Bruner, H. D. (editor in chief). Peripheral blood flow
measurement. Methods in Medical Research 8 : 302-35 1 , 1 960.
Bltrn, J. H., and D. E. Hutcheon. The action of nor-
adrenaline. Brit. J. Pharmacol. 4: 373-380, 1949.
Burn, J. H., and M. J. Rand. New observations on the
sympathetic postganglionic mechanism. Am. J. Med. 29:
1 002-1 007, i960.
Burton, A. C. Laws of physics and flow in blood vessels.
Ciba Foundation Symposium, Visceral Circulation. 1953, PP-
70-86.
Burton, A. C. Relation of structure to function of the
tissues of the wall of blood vessels. Physiol. Revs. 34: 619-
642, 1954-
Burton, A. G, and R. H. Stimson. The measurement of
tension in vascular smooth muscle. J. Physiol., London
153:290-305, i960.
Campbell, E. J. M., and J. H. Green. The variations in
intra-abdominal pressure and the activity of the abdomi-
nal muscles during breathing; a study in man. J. Physiol.,
London 122: 282-290, 1953.
Cantarow, A., and C. W. Wirts, Jr. The effect of dog's
bile, certain bile acids and India ink on bilirubinemia and
the excretion of Bromsulfalein. Am. J. Digest. Diseases 10:
261-266, 1943.
Cantarow, A., C. W. Wirts, W. J. Snape, and L. L.
>432
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Miller. Excretion of bilirubin and Bromsulfalein in bile.
Am. J. Physiol. 154: 211-219, 1948.
76. Canter, J. W., W. S. Rosenthal, and I. D. Baronofskv.
The interrelationship of wedged hepatic vein pressure,
intrasplenic pressure, and intra-abdominal pressure. J.
Lab. Clin. Med. 54: 756-762, 1959.
77. Casselman, W. G. B., and A. M. Rappaport. "Guided"
catheterization of hepatic veins and estimation of hepatic
blood flow by the Bromsulphalein method in normal dogs.
J. Physiol., London 123: 173-182, 1954.
78. Castenfors, H., H. Eliasch, and E. Hultman. The
splanchnic blood flow and oxygen consumption estimated
in man by the Bromsulphalein method with special refer-
ence to the influence of the peripheral dye level. Scand. J.
Clin. & Lab. Invest. 12: 1 58-171, i960.
79. Celander, O. The range of control exercised by the
'sympathico-adrenal system.' Ada Physiol. Scand. 32:
Suppl. 116, 1-132, 1954.
80. Celander, O., and B. Folkow. The nature and the dis-
tribution of afferent fibres provided with the axon reflex
arrangements. Acta Physiol. Scand. 29: 359-370, 1953.
81. Chakravarti, M., and J. Tripod. The action in the
perfused liver of acetylcholine, sympathomimetic sub-
stances and local anaesthetics. J. Physiol., London 97: 316-
329. '94°-
82. Chambers, R., and B. W. Zweifach. Topography and
function of the mesenteric capillary circulation. Am. J.
Anat. 75: 173-205, 1944.
83. Chapman, N. D., P. D. Goldsworthy, L. M. Nyhus,
W. Volwiler, and H. N. Harkins. Studies in isolated
organ physiology : Bromsulphalein clearance in the iso-
lated perfused bovine liver. Surgery 48: 111-118, i960.
84. Chenderovitch, J. La microangioradiogi aphie du foie et de
la rate (M.D. thesis). Vichy: Wallon, 1956, 92 pp.
85. Child, C. G. III. The Hepatic Circulation and Portal Hyper-
tension. Philadelphia: Saunders, 1954, 444 pp.
86. Clark, J. H., D. R. Hooker, and L. H. Weed. The
hydrostatic factor in venous pressure measurements. Am.
J. Physiol. 109: 166-177, 1934.
87. Cohn, C, R. Levine, and M. Kolinsky. Hepatic and
peripheral removal rates in the dog, for intravenously
injected Bromsulphalein. Am. J. Physiol. 155: 286-289,
1948.
88. Cohn, C, R. Levine, and D. Streicher. The rate of
removal of intravenously injected Bromsulphalein by the
liver and extrahepatic tissues of the dog. Am. J. Physiol.
150:299-303, 1947.
89. Cole, J. W., J. Krohmer, F. J. Bonte, and W. Schatten.
An experimental study of intrahepatic distribution of
portal blood. Surg. Gynecol. Obstet. 102: 543-544, 1956.
90. Coleridge, J. C G., and A. Hemingway. Partition of the
venous return to the heart. J. Physiol., London 142: 366-
38", I958-
91. Combes, B. Estimation of hepatic blood flow in man and
in dogs by I131-labeled rose bengal. J. Lab. Clin. Med.
56: 537-543. '9oo.
92. Combes, B., J. R. K Preedy, H. O. Wheeler, R. M.
Hays, and S. E. Bradley. The hemodynamic effects of
hexamethonium bromide in the dog, with special refer-
ence to "splanchnic pooling." J. Clin. Invest. 36: 860-865,
■957-
93. Combes, B., and G. S. Stakelum. Conjugation of sulfo-
bromophthalein sodium with glutathione in thioether
linkage by the rat. J. Clin. Invest. 39: 1214-1222, 1960.
94. Cominsky, B., J. R. K. Preedy, R. Hays, and H. O.
Wheeler. The distribution of circulating blood within
the splanchnic vasculature. J. Clin. Invest. 34: 927, 1955.
95. Gopher, G. H., and B. M. Dick. "Stream Line" phe-
nomena in the portal vein and the selective distribution of
portal blood in the liver. A.M. A. Arch. Surg. 17: 408-419,
■9*8.
96. Corday-, E., and J. H. Williams, Jr. Effect of shock and
of vasopressor drugs on the regional circulation of the
brain, heart, kidney, and liver. Am. J. Med. 29: 228-241,
i960.
97. Couinaud, C, and C. Nogueira. Les veines sus-hepa-
tiques chez l'homme. Acta Anat. 34: 84-110, 1958.
98. Cournand, A., and H. A. Ranges. Catheterization of the
right auricle in man. Proc. Soc. Exptl. Biol. Med. 46: 462-
466, 1941.
99. Culbertson, J. W., R. W. Wilkins, F. J. Ingelfinger,
and S. E. Bradley. The effect of the upright posture upon
hepatic blood flow in normotensive and hypertensive
subjects. J. Clin. Invest. 30: 305-311, 1 95 1 .
100. Dale, H. H., and H. W. Dudley. The presence of his-
tamine and acetylcholine in the spleen of the ox and the
horse. J. Physiol., London 68: 97-123, 1929.
101. Daniel, P. M., and M. M. L. Prichard. Effects of stimu-
lation of the hepatic nerves and of adrenaline upon the
circulation of the portal venous blood within the liver.
J. Physiol., London 114: 538-548, 1 951 .
102. Daniel, P. M., and M. M. L. Prichard. Variations in
the circulation of the portal venous blood within the
liver. J. Physiol., London 114: 521-537, 1 95 1 .
103. Davis, W. D., Jr., R. Gorlin, S. Reichman, and J. P.
Storaasli. Effect of pituitrin in reducing portal pressure
in the human being. AW Engl. J. Med. 256: 108-111,
'957-
104. Deal, C. P., Jr., and H. D. Green. Comparison of
changes in mesenteric resistance following splanchnic
nerve stimulation with responses to epinephrine and
norepinephrine. Circulation Research 4: 38-44, 1956.
105. DeFraiture, W. H., H. Heemstra, J. J. M. Vegter.
and E. Mandema. Chromatographic separation of differ-
ent bromsulphalein metabolites in urine and bile. Acta
Med. Scand. 165: 153-156, 1959.
106. Delorme, E. J., A. I. S. Macpherson, S. R. Mukherjee,
and S. Rowlands. Measurement of the visceral blood
volume in dogs. Quart. J. Exptl. Physiol. 36: 219-231, 1951.
107. Deysach, L. The nature and location of the "sphincter
mechanism" in the liver as determined by drug actions
and vascular infections. Am. J. Physiol. 132: 713-724,
1 941.
108. Dobson, E. L. The measurement of liver blood flow. A
comparison of the parameters measured. In : Liver Func-
tion, edited by R. W. Brauer. Washington, D. C. : Am.
Inst. Biol. Sci. 1958, pp. 75-80.
109. Dobson, E. L, J. W. Gofman, H. B. Jones, L. S. Kelly,
and L. A. Walker. Studies with colloids containing radio-
isotopes of yttrium, zirconium, columbium and lantha-
num. II. The controlled selective localization of radio-
isotopes of yttrium, zirconium and columbium in the
bone marrow, liver and spleen. J. Lab. Clin. Med. 34:
305-312, 1949.
THE HEPATIC CIRCULATION
'433
1 10. Dobson, E. L., and H. B. Jones. The behaviour of intra-
venously injected particulate material. Acta Med. Scand.
144: Suppl. 273, 1-71, 1952. 130.
111. Dock, W. Role of increased hepatic arterial flow in portal
hypertension of cirrhosis. Tram. Assoc. Am. Physicians 57:
302-306, 194a. 131.
112. Dosekun, F. O., J. Grayson, and D. Mendel. The
measurement of metabolic and vascular responses in liver
and muscle with observations on their responses to insulin 132.
and glucose. J. Physiol., London 150: 581-606, i960.
113. Downman, C. B. B. Cerebral destination of splanchnic 133.
afferent impulses. J. Physiol., London 113: 434-441, 1951.
114. Drapanas, T, D. N. Kluge, and VV. G. Schenk, Jr.
Measurement of hepatic blood flow by bromsulphalein
and by the electromagnetic flowmeter. Surgery 48: 1017—
1021, ig6o. 134.
115. Drapanas, T., W. G. Schenk, Jr., E. L. Pollack, and
J. D. Stewart. Hepatic hemodynamics in experimental
ascites. Ann. Surg. 152: 705-716, i960. 135.
116. Dreyer, B. Streamlining in the portal vein. Quart. J.
Exptl. Physiol. 39: 305-307, 1954.
117. Duomarco, J. L., and R. Rimini. Energy and hydraulic
gradients along systemic veins. Am. J. Physiol. 178: 215- 136.
220, 1954.
1 18. Edwards, A. W. T. Sampling of hepatic venous blood in 137.
the dog. J. Appl. Physiol. 10: 305-313, 1957.
119. Edwards, E. A. Functional anatomy of the porta-sys- 138.
temic communications. Arch. Internal Med. 88: 137-154,
'95'-
120. Eli as, H. Liver morphology. Biol. Revs. Cambridge Phil. Soc.
30: 263-310, 1955. 139.
121. Elias, H., and A. Sokol. Dependence of the lobular
architecture of the liver on the porto-hepatic blood pres-
sure gradient. Anat. Record 115: 71-86, 1953.
122. Englert, E., B. A. Burrows, and F. J. Ingelfinger. 140.
Differential analysis of the stages of hepatic excretory
function with gamma emitting isotopes. II. Attempts to
alter rate phenomena. J. Lab. Clin. Med. 56: 193-206, 141.
i960.
123. Epstein, R. M., H. O. Wheeler, M. J. Frumin, D. V.
Habif, E. M. Papper, and S. E. Bradley. The effect of 142.
hypercapnia on estimated hepatic blood flow, circulating
splanchnic blood volume and hepatic sulfobromophthal-
ein clearance during general anesthesia in man. J. Clin. 143.
Invest. 40: 592-598, 1 96 1.
124. Evringham, A., E. M. Brenneman, and S. M. Horvath. 144.
Influence of sodium pentobarbital on splanchnic blood
flow and related function. Am. J. Physiol. 197: 624-626,
■959- '45-
125. Farrand, E. A., R. Larsen, and S. M. Horvath. Effects
of /-epinephrine and /-norepinephrine on the splanchnic 146.
bed of intact dogs. Am. J. Physiol. 189: 576-579, 1957.
126. Fawcett, D. W. Observations on the cytology and elec-
tron microscopy of hepatic cells. J. Natl. Cancer Inst. 15: 147.
1 475-1503, 1955.
127. Feldberg, VV. Distribution of histamine in the body.
Ciba Foundation Symposium, Histamine. 1956, pp. 4-13. 148.
128. Fischer, A., L. Takacs, and G. Molnar. Hepatic cir-
culation in arterial hypoxia. Acta Med. Acad. Sci. Hung. 149.
16: 61-74, i960.
129. Fisher, B., C. Russ, R. G. Selker, and E. J. Fedor.
Observations on liver blood flow. Its relationship to 150.
cardiac output in anesthetized and unanesthetized ani-
mals. A.M. A. Arch. Surg. 72: 600-611, 1956.
Folkow, B. Range of control of the cardiovascular system
by the central nervous system. Physiol. Revs. 40: Suppl. 4,
93-99. '960.
Folkow, B., and B. Lofving. The distensibility of the
systemic resistance blood vessels. Acta Physiol. Scand. 38:
37-52, I956-
Franklin, K. J. A Monograph on Veins. Springfield, 111.:
Thomas, 1937, 410 pp.
Freis, E. D, J. C. Rose, E. A. Partenope, T. F. Higgins,
R. T. Kelley-, H. W. Schnaper, and R. L. Johnson.
The hemodynamic effects of hypotensive drugs in man.
III. Hexamethonium. J. Clin. Invest. 32: 1285-1298,
'953-
Trendelenburg, V. The action of histamine on the sym-
pathetic nervous system. Ciba Foundation Symposium, His-
tamine. 1956, pp. 278-279.
Friedman, E. W., and R. S. Weiner. Estimation of
hepatic sinusoid pressure by means of venous catheters
and estimation of portal pressure by hepatic vein cathe-
terization. Am. J. Physiol. 165: 527-531, 1 95 1 .
Friedman, J. J. Mesenteric circulation in hemorrhagic
shock. Circulation Research 9: 561-565, 1961.
Friedman, J. J. Splanchnic blood volume in traumatic
shock. Am. J. Physiol. 200: 614-618, 1961.
Fries, G. F., and G. H. Conner. Studies on bovine portal
blood. II. Blood flow determinations with observations
on hemodilution in the portal vein. .4m. J. Vet. Research
22: 487-491, 1961.
Gammon, G. D., and D. VV. Bronk. The discharge of
impulses from Pacinian corpuscles in the mesentery and
its relation to vascular changes. Am. J. Physiol. 114:
77-84. "935-
Gardner, E., L. M. Thomas, and F. Morin. Cortical
projections of fast visceral afferents in the cat and mon-
key. Am. J. Physiol. 183: 438-444, 1955.
Gernandt, B., and Y. Zotterman. Intestinal pain: an
electrophysiological investigation on mesenteric nerves.
Acta Physiol. Scand. 12: 56-72, 1946.
Gersmeyer, E. F., and G. Gersmeyer. Stromungsge-
schwindigkeits- und Druckmessungen in der Pfortader des
wacken Hundes. Arch. Krieslaufforsch. 27: 206-219, 1957.
Gibson, J. B. The hepatic veins in man and their sphincter
mechanisms. J. Anat. 93: 368-379, 1959.
Gidlund, A. Development of apparatus and methods for
Roentgen studies in haemodynamics. Acta Radiol. Suppl.
130: 1-70, 1956.
Gilfillan, R. S. Anatomic study of the portal vein and
its main branches. A.M. A. Arch. Surg. 61 : 449-461, 1950.
Gilmore, J. P. Effect of anesthesia and hepatic sampling
site upon hepatic blood flow. Am. J. Physiol. 195: 465-
468, 1958.
Ginsburg, M., and J. Grayson. Factors controlling liver
blood How in the rat. J. Physiol., London 123: 574-602,
■954-
Glauser, F. Studies on intrahepatic arterial circulation.
Surgery 33: 333-341, 1953.
Gomez, D. M. Hhnodynamique el Angiocinetique ; Etude Ra-
tionnelle des Lois Regissant les Phenomenes Cardio-vasculaires.
Paris: Hermann, 1941, 731 pp.
Gordon, D. B., J. Flasher, and D. R. Drury'. Size of
'434
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the largest arteriovenous vessels in various organs. Am. J.
Physiol. 173: 275-281, 1953.
151. Grabner, G., and A. Neumayr. A continuous recording
method for the estimation of liver blood flow in man.
Circulation. Proc. Harvey Tercentenary Congress, edited by
J. McMichael. 1958, pp. 386-392.
152. Graf, W. Patterns of human liver temperature. Acta
Physiol. Scand. 46: Suppl. 160, 1959.
153. Grafflin, A. L. The excretion of fluorescein by the liver
under normal and abnormal conditions observed in vivo
with the fluorescence microscope. .4m. J. Anat. 81 : 63-1 16,
1947-
154. Grafflin, A. L., and E. H. Baglev. Studies of peripheral
blood vascular beds. Johns Hopkins Hasp. Bull. 92 : 47-
73. 1953-
155. Grayson, J. The application of internal calorimetry to
the measurement of liver blood flow responses. In: Liver
Function, edited by R. W. Brauer. Washington, D. (.'.. :
Am. Inst. Biol. Sci. 1958, pp. 106-112.
156. Grayson, J. Vascular reactions in the human intestine.
J. Physiol., London 109: 439-447, 1949.
157. Grayson, J., and D. H. Johnson. The effect of adren-
aline and noradrenaline on the liver blood flow. J. Physiol.,
London 120: 73-94, 1953.
158. Grayson, J., and D. Mendel. Observation on the intra-
hepatic flow interactions of the hepatic artery and portal
vein. J. Physiol., London 139: 167-177, 1957.
159. Green, H. D., C. P. Deal, Jr., S. Bardhanabaedya,
and A. B. Denison, Jr. The effects of adrenergic sub-
stances and ischemia on the blood flow and peripheral
resistance of the canine mesenteric vascular bed before
and during adrenergic blockade. J. Pharmacol. Exptl.
Therap. 113: 1 15-123, 1955.
160. Green, H. D., L. S. Hall, J. Sexton, and C. P. Deal.
Autonomic vasomotor responses in the canine hepatic
arterial and venous beds. Am. ./. Physiol. 196: 196-202,
'959-
161. Green, H. D., and J. H. Kepchar. Control of peripheral
resistance in major systemic vascular beds. Physiol. Revs.
39: 617-686, 1959.
162. Green, H. D., R. N. Lewis, N. D. Nickerson, and A. L.
Heller. Blood flow, peripheral resistance and vascular
tonus, with observations on the relationship between
blood flow and cutaneous temperature. Am. J. Physiol.
141: 5l8~536. '944-
163. Green, H. D., K. Ottis, and T. Kitchen. Autonomic
stimulation and blockade on canine splenic inflow, out-
flow and weight. .4m. J. Physiol. ig8: 424-428, i960.
164. Grindlay, J. H., J. F. Herrick, and F. C. Mann.
Measurement of the blood flow of the spleen. .4m. J.
Physiol. [ 27 : 1 06- 118, 1 939.
165. Grodsky, G M., J. V. Carbone, and R. Fanska.
Identification of metabolites of sulfobromophthalein.
./. Clin. Invest. 38: 1981-1988, 1959.
166. Guyton, A. C, A. \V. Lindsey, and G G. Armstrong.
Relationship of total peripheral resistance to the pressure
gradient from the arteries to the veins. -4m. ./. Physiol.
186: 294-298, 1956.
167. Haddy, F. J. Effect of histamine on small and large vessel
pressures in the dog foreleg. .4m. J. Physiol. 198: 161 -168,
.|tn 1
168. Hahn, P. F., VV. D. Donald, and R. C. Grier, Jr. The
170.
171.
174.
'75-
physiological bilaterality of the portal circulation. Am. ./.
Physiol. 143: 105, 1945.
169. Hampton, J. C. A re-evaluation of submicroscopic struc-
ture of the liver. Texas Repts. Biol, and Med. 18: 602-611,
i960.
Hamrick, L. \V., Jr., and J. D. Myers The effect of
subfebrile doses of bacterial pyrogens on splanchnic metab-
olism and cardiac output. J. Lab. Clin. Med. 45: 568-572,
'955-
Harris, P. D., and S. I. Schwartz. Polarographic eval-
uation of the effects of Pitressin on hepatic oxygen tension.
Surgery 49 : 51 4-5 19, 1 96 1 .
172. Herrick, J. F., J. H. Grindlay, E. J. Baldes, and F. C.
Mann. Effect of exercise on blood flow in superior mesen-
teric, renal and common iliac arteries. .4m. J. Physiol.
128: 338-344, 1940.
173. Hjortsjo, C. H. The topography of the intrahepatic duct
systems. Acta Anat. 11: 599-615, 1951.
Holt, J. P. The collapse factor in the measurement of
venous pressure: the flow of fluid through collapsible
tubes. Am. J. Physiol. 134: 292-299, 1941.
Horvath, S. M., T. Kelly, G. E. Folk, Jr., and B. K.
Hutt. Measurement of blood volumes in the splanchnic
bed of the dog. .4m. ./. Physiol. 189: 573-575, '957-
176. Huckabee, VV. E., and G Walcott. Determination of
organ blood flow using 4-aminoantipynne. J. Appl.
Physiol. 15: 1 1 39- 1 143, i960.
177. Ingelfinger, F. J., S. E. Bradley, A. I. Mendeloff,
and P. Kramer. Studies with Bromsulphalein. 1. Its
disappearance from the blood after a single intravenous
injection. Gastroenterology 11: 646-657, 1948.
178. Javitt, N. B., H. O. Wheeler, K. J. Baker, O. L.
Ramos, and S. E. Bradley. The intrahepatic conjugation
of sulfobromophthalein and glutathione in the dog. J.
Clin. Invest. 39: 1570-1577, i960.
179. Johnson, D. H. The effect of haemorrhage and hypoten-
sion on the liver blood flow. J. Physiol., London 126: 413-
433- '954-
180. Johnson, P. C. Autoregulation of intestinal blood flow.
Am. J. Physiol. 199: 311-318, i960.
181. Johnstone, F. R. C. Measurement of splanchnic blood
volume in dogs. Am. J. Physiol. 185: 450-452, 1956.
182. Katz, L. N., and S. Rodbard. The integration of the
vasomotor responses in the liver with those in other sys-
temic vessels. J. Pharmacol. Exptl. Therap. 67: 407-421,
!939-
183. Ketterer, S. G., B. D. Wiegand, and E. Rapaport.
Hepatic uptake and biliary excretion of indocyanine
green and its use in estimation of hepatic blood flow in
dogs. .4m. J. Physiol. 199: 481-484, i960.
184. Knisely, M. H. Spleen studies. 1. Microscopic observa-
tions of the circulatory system of living unstimulated
mammalian spleens. Anat. Record 65: 23-50, 1936.
185. Knisely, M. H., E. H. Bloch, and L. Warner. Selective
phagocytosis. I. Kgl. Danske Videnskab Selskab Biol.
Skrifter 4: 1-93, 1948.
186. Knisely, M. H., F. Harding, and H. Debacker. Hepatic
sphincters: brief summary of present-day knowledge.
Science 125: 1 023-1 026, 1 957.
Koiin, P. M., B. L. Charms, and B. L. Brofman. Effect
of epinephrine and posterior pituitary extract on the
wedged-hepatic-vein pressure in normal patients and in
187.
THE HEPATIC CIRCULATION
•435
those with liver disease. New Engl. J. Med. 261 : 323-327,
'959-
188. Krebs, J. S., and R. VV. Brauer. Metabolism of sulfo-
bromophthalein sodium (BSPj in the rat. Am. J. Physiol. 208.
■94: 37-43, 1958-
189. Rubin, R. H., G. M. Grodskv, and J. V. Garbone.
Investigation of Rose Bengal conjugation. Proc. Soc.
Exptl. Biol. Med. 104: 650-653, i960. 209.
igo. Kuntz, A. The Autonomic Nervous System (4th ed.). Phil-
adelphia: Lea & Febiger, 1953, 605 pp.
191. Larsen, J. A., N. Tygstrup, and K. Winkler. The sig- 210.
niiicance of the extrahepatic elimination of ethanol in
determination of hepatic blood flow by means of ethanol. 211.
Scand. J. Clin. & Lab. Invest. 13: 116— 121, 1961 .
192. Lee, R. E. Vasomotor reactions in the mesenteric and
serosal capillary bed during fright and violent muscular 212.
activity. Proc. Soc. Exptl. Biol. Med. 71 : 607-609, 1949.
193. Levy, M. N. Influence of levarterenol on portal venous
flow in acute hemorrhage. Circulation Research 6: 587-591,
1958. 213.
194. Levy, M. N. Relative influence of variations in arterial
and venous pressures on resistance to flow. Am. J. Physiol.
192: 164-170, 1958. 214.
195. Lewis, A. E. Investigation of hepatic function by clear-
ance techniques. Am. J. Physiol. 163: 54-61, 1950. 215.
196. Liebowitz, H. R. Bleeding, Esophageal Varices, Portal Hyper-
tension. Springfield, III.: Thomas, 1959, 986 pp.
197. Lipscomb, A., and L. A. Crandall, Jr. Hepatic blood 216.
flow and glucose output in normal unanesthetized dogs.
Am. J. Physiol. 148: 302-311, 1947.
198. Lorber, S. H., M. J. Oppenheimer, H. Shay, P. Lynch,
and H. Siplet. Enterohepatic circulation of Bromsulphal- 217.
ein: intraduodenal, intraportal and intravenous dye
administration in dogs. Am. J. Physiol. 173: 259-264,
'953-
199. Lorber, S. H., and H. Shay. Entero-hepatic circulation
of bromsulphalein. I. Studies in man with special refer- 218.
ence to the clinical BSP test. Gastroenterology 20: 262-271,
I952-
200. LOWENTHAL, M., K. HaRPUDER, AND S. D. BlATT.
Peripheral and visceral vascular effects of exercise and 219.
postprandial state in supine position. J. Appl. Physiol.
4: 689-694, 1952.
201. Lynn, R. B., S. M. Sancetta, F. A. Simeone, and R. W.
Scott. Observations on the circulation in high spinal 220.
anesthesia. Surgery 32: 195-213, 1952.
202. MacDonald, D. A. Blood Plow in Arteries. Baltimore:
Williams & Wilkins, i960, 328 pp. 221.
203. MacLean, L. D., E. L. Brackney, and M. B. Visscher.
Effects of epinephrine, norepinephrine and histamine on
canine intestine and liver weight continuously recorded 222.
in vivo. J. Appl. Physiol. 9: 237-240, 1956.
204. MacLeod, J. J. R., and R. G. Pearce. The outflow of
blood from the liver as affected by variations in the condi- 223.
tion of the portal vein and hepatic artery. Am. J. Physiol.
35: 87-105, 19.4.
205. Maegraith, B. Sinusoids and sinusoidal flow. In: Liver
/•'unction, edited by R. W. Brauer. Washington, D. C. : 224.
Am. Inst. Biol. Sci. 1958, pp. 135-319.
206. Mall, F. P. A study of the structural unit of the liver.
Am. J. Anal. 5: 227-308, 1906. 225.
207. Marks, P. A., P. C. Reynell, and S. E. Bradley.
Hemodynamic effects of L-hydrazinophthalazine in the
dog, with special reference to circulating splanchnic
blood volume. Am. J. Physiol. 183: 144-148, 1955.
Mason, M. F., G. Hawley, and A. Smith. Application
of the saturation principle to the estimation of functional
hepatic mass in normal dogs. Am. J. Physiol. 152: 42-47,
1948.
Mayerson, H. S., C. G. Wolfram, H. H. Shirley, Jr.,
and K. Wasserman. Regional differences in capillary
permeability. Am. J. Physiol. 198: 155-160, i960.
McMichael, J. The portal circulation. II. The action of
acetylcholine. J. Physiol., London 77: 399-421, 1933.
Meltzer, J. I., H. O. Wheeler, and W. I. Cranston.
Metabolism of sulfobromophthalein sodium (BSPj in dog
and man. Proc. Soc. Exptl. Biol. Med. 100: 174-179, 1959.
Mendeloff, A. I., P. Kramer, F. J. Ingelfinger, and
S. E. Bradley. Studies with Bromsulfalein. II. Factors
altering its disappearance from the blood after a single
intravenous injection. Gastroenterology 13: 222-234, 1949.
Meurman, L. On the distribution and kinetics of injected
J131 rose bengal. Acta Med. Scand. 167: Suppl. 354, 1-85,
i960.
Michels, N. A. Blood Supply and Anatomy of the Upper
Abdominal Organs. Philadelphia: Lippincott, 1955, 581 pp.
Milnor, W. R., and A. D. Jose. Distortion of indicator-
dilution curves by sampling systems. J. Appl. Physiol.
15: 177-180, i960.
Mohamed, S., and J. W. Bean. Local and general altera-
tions of blood COs and influence of intestinal motility
in regulation of intestinal blood flow. Am. J. Physiol.
167: 413-425, 1951.
Myers, J. D. The hepatic blood flow and splanchnic oxy-
gen consumption of man — their estimation from urea
production or bromsulphalein excretion during cathe-
terization of the hepatic veins. J. Clin. Invest. 26: 1130-
"37. '947-
Myers, J. D., E. S. Brannon, and B. C. Holland. A
correlative study of the cardiac output and the hepatic
circulation in hyperthyroidism. J. Clin. Invest. 29: 1069-
1077, 1950.
Necheles, H., R. Frank, W. Kaye, and E. Rosenman.
Effect of acetylcholine on the blood flow through the
stomach and legs of the rat. Am. J. Physiol. 114: 695-699,
'935-
Norcross, J. W., R. M. White, and R. F. Bradley, Jr.
Bromsulfalein liver function test with special reference to
renal excretion. Am. J. Med. Sci. 221 : 137-139, 1 95 1 .
Olerud, S. Experimental studies on portal circulation at
increased intra-abdominal pressure. Acta Physiol. Scand.
30: Suppl. 109, 1-95, 1953.
Opdyke, D. F., H. F. Van Noate, and G. A. Brecher.
Further evidence that inspiration increases right atrial
inflow. Am. J. Physiol. 162: 259-265, 1950.
Ottis, K., J. E. Davis, Jr., and H. D. Green. Effects of
adrenergic and cholinergic drugs on splenic inflow and
outflow before and during adrenergic blockade. Am. J.
Physiol. 189: 599-604, 1957.
Owen, C. A., Jr. The effect of enterohepatic circulation
on the Bromsulfalein test of hepatic function. J. Lab. Clin.
Med. 38: 583-584, 1 95 1.
Palmer, A. A. A study of blood flow in minute vessels of
the pancreatic region of the rat with reference to inter-
>436
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
mittent corpuscular flow in individual capillaries. Quart.
J. Exptl. Physiol. 44: 149-159, 1959.
226. Pappenheimer, J. R., and J. P. Maes. A quantitative
measure of the vasomotor tone in the hindlimb muscles
of the dog. Am. J. Physiol. 137: 187-199, 1942.
227. Parpart, A. K., A. O. Whipple, and J. J. Chang. The
microcirculation of the spleen of the mouse. Angiology 6:
350-362, 1955.
228. Paton, W. D. M. The mechanism of histamine release.
Ciba Foundation Symposium, Histamine. 1956, pp. 59-78.
229. Patrassi, G., B. D'Agnolo, C. DalPalu, and A. Ruol.
II ciriolo epatoportalc alia luce delle moderne techniche.
Acta Med. Patavma. Suppl. 3: 1-86, 1957.
230. Peterson, L. H. Some characteristics of certain reflexes
which modify the circulation in man. Circulation 2: 351 —
362, 1950.
231. Playoust, M. R., J. McRae, and R. W. Boden. Ineffi-
cient hepatic extraction of colloidal gold: resulting in-
accuracies in determination of hepatic blood flow. J. Lab.
Clin. Med. 54: 728-738, 1959.
232. Pratt, E. B., F. D. Burdick, and J. H. Holmes. Measure-
ment of liver blood flow in unanesthetized dog using BSP
dye method. Am. J. Physiol. 71: 471-478, 1952.
233. Prinzmetal, M., E. M. Ornitz, Jr., B. Simkin, and
H. C. Bergman. Arteriovenous anastomoses in liver,
spleen and lungs. Am. J. Physiol. 152: 48-52, 1948.
234. Rabinovvitz, M., and E. Rapaport. Determination of
circulating pulmonary blood volume in dogs by an arterio-
venous dye equilibration method. Circulation Research 2:
525-536> '954-
235. Ramlo, J. H., and E. B. Brown, Jr. Mechanism of splenic
contraction produced by severe hypercapnia. Am. J.
Physiol. 197: 1079-1082, 1959.
236. Rapaport, E., M. H. Weisbart, and M. Levine. The
splanchnic blood volume in congestive heart failure.
Circulation 18: 581-587, 1958.
237. Rappaport, A. M. The structural and functional unit
in the human liver (liver acinus). Anat. Record 130: 673-
689, 1958.
238. Reemtsma, K., G. C. Hottinger, A. C. DeGraff, Jr.,
and O. Creech, Jr. The estimation of hepatic blood flow-
using indocyanine green. Surg. Gynecol. Obstet. 110: 353-
356, i960.
239. Reichman, S., W. D. Davis, J. P. Storaasli, and R.
Gorlin. Measurement of hepatic blood flow by indicator
dilution techniques. J. Clin. Invest. 37: 1848-1856, 1958.
240. Reininger, E. J., and L. A. Sapirstein. Effect of diges-
tion on distribution of blood flow in the rat. Science 126:
■ 176, 1957-
241. Remington, J. W. Extensibility behavior and hysteresis
phenomenon in smooth muscle tissues. In : Tissue Elas-
ticity. Washington, D. C. : Am. Physiol. Soc, 1957, pp.
'38-153-
242. Restrepo, J. E , W. D. Warren, S. P. Nolan, and
W. H. Muller, Jr. Radioactive gold technique for the
estimation of liver blood flow : normal values and tech-
nical considerations. Surgery 48: 748-757, i960.
243. Reynell, P. C, P. A. Marks, C. Chidsey, and S. E.
Bradley. Changes in splanchnic blood volume and
splanchnic blood flow in dogs after haemorrhage. Clin.
Set. 14: 407-419, 1955.
244. Reynolds, T. B., D. C. Balfour, Jr., D. C. Levinson,
W. P. Mikkelsen, and A. C. Pattison. Comparison of
wedged hepatic vein pressure with portal vein pressure in
human subjects with cirrhosis. J. Clin. Invest. 34: 213-218,
'955-
245. Reynolds, T. B., A. Paton, M. Freeman, F. Howard,
and S. Sherlock. The effect of hexamethonium bromide
on splanchnic blood flow, oxygen consumption and glu-
cose output in man. J. Clin. Invest. 32 : 793-800, 1953.
246. Richards, D. W., Jr., A. Cournand, R. C. Darling,
W H. Gillespie, and E. Baldwin. Pressure of blood in
the right auricle, in animals and in man: under normal
conditions and in right heart failure. Am. J. Physiol. 136:
i'5-'23. '94'-l
247. Richardson, D. W., A.J. Wasserman, and J. L. Patter-
son, Jr. General and regional circulatory responses to
change in blood pH and carbon dioxide tension. J. Clin.
Invest. 40: 31-43, 1 96 1.
248. Richins, C. A. The effect of sympathetic nerve stimulation
on blood flow through the pancreas. Anat. Record 1 06:
237"-!38, 195°-
249. Richins, C. A. The innervation of the pancreas. J. Comp.
Neurol. 83: 223-236, 1945.
250. Riecker, G Uber die Beziehung zwischen Druck und
Stromstarke der portalen Lebergefasse. Pflugers Arch. ges.
Physiol. 262: 37-50, 1955.
251. Roberts, W. H. Lamellated corpuscles (Pacinian) in
relation to the larger human limb vessels and a compara-
tive study of their distribution in the mesentery. Anat.
Record 133: 593-604, 1959.
252. Rosenau, W., J. V. Carbone, and G M. Grodsky.
Metabolism of sulfobromophthalein in hepatectomized
and hepatectomized-nephrectomized dog. Proc. Soc.
Exptl. Biol. Med. 102: 131-133, 1959-
253. Russu, I. G, A. Vaida, D. Dumitrascu, and O. Lucaciu.
Beitrage zur Innervation der leber. Die nervenbahnen der
venae hepaticae beim Menschen. Acta. Anat. 44: 70—79,
1961.
254. Sancetta, S. M. Dynamic and neurogenic factors deter-
mining the hepatic arterial flow after portal occlusion.
Circulation Research 1 : 414-418, 1953.
255. Sapirstein, L. A. Indicator dilution methods in the
measurement of the splanchnic blood flow of normal dogs.
In: Liver Function, edited by R. W. Brauer. Washington,
D. C. : Am. Inst. Biol. Sci. 1958, pp. 93-105.
256. Sapirstein, L. A. Regional blood flow by fractional dis-
tribution of indicators. .4m. J. Physiol. 193: 161-168,1958.
257. Sapirstein, L. A., and E.J. Reininger. Catheter induced
error in hepatic venous sampling. Circulation Research 4:
493-498. '956-
258. Sapirstein, L. A., and A. M. Simpson. Plasma clearance
of rose bengal (tetraiodotetrabromiluorescein). Am. J.
Physiol. 182: 337-346, 1955.
259. Sarnoff, S. J., and S. I. Yamada. Abdominal presso-
receptors: the pancreas and abdominal Pacinian system.
Proc. World Congr. Cardiology 3: 54-55, 1958.
260. Schambve, P. Experimental estimation of the portal vein
blood flow in sheep. I. Examination of an infusion method
and results from acute experiments. Nord. Veterindr. med.
7: 961. II. Chronic experiments in cannulated sheep
applying infusion and injection methods. Nord. Veterindr.
med. 7: 1001-1016, 1955.
THE HEPATIC CIRCULATION
'437
261. Schleier, J. Der Energieverbrauch in der Blutbahn. 280.
Pflugers Arch. ges. Physiol. 173: 172-204, 1918.
262. Schobinger, R. Inlra-osseous Venography. New York: 281.
Grune & Stratton, i960.
263. Schumann, H. J. Formation of adrenergic transmitters.
Ciba Symposium, Adrenergic Mechanisms, edited by G. E. W.
Wolstenholme and R. M. O'Connor, i960, pp. 6-16. 282.
264. Selkurt, E. E. Comparison of the Bromsulphalein method
with simultaneous direct hepatic blood flow. Circulation
Research 2: 155-159, 1954.
265. Selkurt, E. E. Effect of acute hepatic ischemia on 283.
splanchnic hemodynamics and on BSP removal by liver.
Proc. Soc. Expll. Biol. Med. 87: 307-312, 1954. 284.
266. Selkurt, E. E. Splanchnic hemodynamics as influenced
by hepatic ischemia. Proc. Soc. Exptl. Biol. Med. 90: 427-
43'. '955- 285.
267. Selkurt, E. E., and G. A. Brecher. Splanchnic hemo-
dynamics and oxygen utilization during hemorrhagic
shock in the dog. Circulation Research 4: 693-704 1956. 286.
268. Selkurt, E. E., and P. C. Johnson. Effect of acute eleva-
tion of portal venous pressure on mesenteric blood volume,
interstitial fluid volume and hemodynamics. Circulation
Research 6: 592-599, 1958. 287.
269. Selkurt, E. E., M. P. Scibetta, and T. E. Cull. Hemo-
dynamics of intestinal circulation. Circulation Research 6:
92-99. '958. 288.
270. Seneviratne, R. D. Physiological and pathological re-
sponses in the blood vessels of the liver. Quart. J. Exptl.
Physiol. 35: 77-110, 1949.
271. Sheehan, D. The afferent nerve supply of the mesentery 289.
and its significance in the causation of abdominal pain.
J. Anal. 67: 233-249, 1933.
272. Sheppard, C. W., E. B. Wells, P. F. Hahn, and J. P. B.
Goodell. Studies of the distribution of intravenously
administered colloidal sols of manganese dioxide and gold 290.
in human beings and dogs using radioactive isotopes.
J. Lab. Clin. Med. 32: 274-286, 1947.
273. Sherlock, S. A. G. Bearn, B. H. Billing, and J. C. S.
Paterson. Splanchnic blood flow in man by the Brom- 291.
sulfalein method: the relation of peripheral plasma brom-
sulfalein level to the calculated flow. J. Lab. Clin. Med.
35: 923-932. 1950. 292.
274. Sherman, H., R. C. Schlant, W. L. Kraus, and C. B.
Moore. A figure of merit for catheter sampling systems.
Circulation Research 7: 303-313, 1959. 293.
275. Shoemaker, W. C. Measurement of hepatic blood flow in
the unanesthetized dog by a modified Bromsulphalein
method. J. Appl. Physiol. 15: 473-478, i960. 294.
276. Shoemaker, W. C, R. Mahler, J. Ashmore, and D. E.
Pugh. Effect of insulin on hepatic blood flow in the un- 295.
anesthetized dog. Am. J. Physiol. 196: 1250-1252, 1959.
277. Shoemaker, W. C, F. G. Panico, W. F. Walker, and
D. H. Elwvn. Perfusion of canine liver in vivo. J. Appl. 296.
Physiol. 15: 687-690, i960.
278. Shoemaker, W. C, R. W. Steenburg, L. L. Smith, and
F. D. Moore. Experimental evaluation of an indicator-
dilution technique for estimation of hepatic blood flow. 297.
■J. Lab. Clin. Med. 57: 661-670, 1961.
279. Shoemaker, W. C, T. B. Van Itallie, and W. F. 298.
Walker. Measurement of hepatic glucose output and
hepatic blood flow in response to glucagon. Am. J. Physiol.
'96: 3'5-3'8. '959-
Smith, H. W. The Kidney: Structure and Function in Health
and Disease. New York: Oxford Univ. Press, 1951, 1049 pp.
SMVTHE, C. McG, J. P. GlLMORE, AND S. W. H.ANDFORD.
The effect of levarterevol (L-norepinephrine) on hepatic
blood flow in the normal, anesthetized dog. ./. Pharmacol.
Exptl. Therap. 1 10: 398-402, 1954.
Smythe, C. McC, H. O. Heinemann, and S. E. Bradley.
Estimated hepatic blood flow in the dog. Effect of ethyl
alcohol on it, renal blood flow, cardiac output and
arterial pressure. Am. J. Physiol. 172: 737-742, 1953.
Stecher, J. L. Fatal reaction to sulfobromophthalein.
New Engl. J. Med. 261: 963, 1959.
Stephenson, J. L. Theory of the measurement of blood
flow by the dilution of an indicator. Bull. Math. Biophys.
10: 1 17-121, 1948.
Taleisnik, S. Liver mass determination by Bromsulfalein
in partially hepatectomized rabbits. Gastroenterology 29:
64-70, 1955-
Taylor, W. J., and J. D. Myers. Occlusive hepatic
venous catheterization in the normal liver, cirrhosis of
the liver and noncirrhotic portal hypertension. Circulation
i3:368-38°. '956-
Thomas, W. D., and H. E. Essex. Observations on the
hepatic venous circulation with special reference to the
sphincteric mechanism. Am. J. Physiol. 158:303-310, 1949.
Thompson, A. M., H. M. Cavert, N. Lifson, and R. L.
Evans. Regional tissue uptake of D2O in perfused organs :
rat liver, dog heart and gastrocnemius. Am. J. Physiol.
!97: 897-902, 1959-
Tisdale, W. A., G. Klatskin, and W. W. Glenn. Portal
hypertension and bleeding esophageal varices; their occur-
rence in the absence of both intrahepatic and extrahepatic
obstruction of the portal vein. New Engl. J. Med. 261 : 209-
218, 1959.
Tornvall, G., and L. Johansson. Liver circulation in
man as studied by means of dilution curves. A method
using catheterisation technique. Acta Med. Scand. 1 54 ■
491-500, 1956.
Torrance, H. B. Liver blood flow during operations on
the upper abdomen. J. Roy. Coll. Surgeons, Edinburgh 2:
216-228, 1957.
Trapold, J. H. Effect of ganglionic blocking agents upon
blood flow and resistance in the superior mesenteric
artery of the dog. Circulation Research 4: 718-723, 1956.
Tygstrup, N., and K. Winkler. Galactose blood clear-
ance as a measure of hepatic blood flow. Clin. Sci. 17
1-9, 1958.
Utterback, R. A. The innervation of the spleen. J. Comp.
Neurol. 81 : 55-68, 1944.
Verschure, J. C. M. Clinical use of measurements of
clearance and maximum capacity of the liver. Acta Med.
Scand. 142: 409-419, 1952.
Vetter, H., G Grabner, R. Hofer, A. Neumayr, and
O. Parzer. Comparison of liver blood flow values esti-
mated by the Bromsulphalein and by the radiogold
method. J. Clin. Invest. 35: 825-830, 1956.
Von Euler, U. S. Histamine and nerves. Ciba Foundation
Symposium, Histamine. 1956, pp. 235-241.
Wade, O. L., B. Combes, A. W. Childs, H. O. Wheeler,
A. Cournand, and S. E. Bradley. The effect of exercise
on the splanchnic blood flow and splanchnic blood volume
in normal man. Clin. Sci. 15: 457-463, 1956.
H38
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
299. Wakim, K. G., and F. C. Mann. The intrahepatic circu-
lation of blood. Anal. Record 82: 233-253, 1942.
300. Walk, L. Roentgenologic determination of the liver
volume. Acta Radiol. 55: 49-56, 1961.
301. Wells, R. E., Jr., and E. W. Merrill. Shear rate de-
pendence of the viscosity of whole blood and plasma.
Science 133: 763"764. '961-
302. Werner, A. Y., and S. M. Horvath. Measurement of
hepatic blood flow in the dog by the Bromsulphalein
method. J. Clin. Invest. 31: 433-439, 1952.
303. Wheeler, H. O., B. Combes, and A. W. Childs. The
splanchnic circulation time. Trans. Assoc. Am. Physicians
68: 177-184, 1955.
304. Wheeler, H. O., W. I. Cranston, and J. I. Meltzer.
Hepatic uptake and biliary excretion of Indocyanine
Green in the dog. Proc. Soc. Exptl. Biol. Med. 99: 11 -14,
1958.
305. Wheeler, H. O., R. M. Epstein, R. R. Robinson, and
E. S. Snell. Hepatic storage and excretion of sulfobro-
mophthalein sodium in the dog. J. Clin. Invest. 39: 236-
247, i960.
306. Wheeler, H. O., J. I. Meltzer, and S. E. Bradley.
Biliary transport and hepatic storage of sulfobromo-
phthalein sodium in the unanesthetized dog, in normal
man, and in patients with hepatic disease. J. Clin.
Invest. 39 : 1 1 3 1 - 1 1 44, 1 960.
307. Whittaker, S. R. F., and F. R. Winton. The apparent
viscosity of blood flowing in the isolated hindlimb of the
dog, and its variation with corpuscular concentration.
J. Physiol., London 78: 339-369, 1933.
308. Wiggins, C. J., D. F. Opdvke, and J. R. Johnson. Portal
pressure gradients under experimental conditions, includ-
ing hemorrhagic shock. .4m. J. Physiol. 146: 192-206,
1946.
309. Wilkins, R. W., S. E. Bradley, and C. K. Friedland.
The acute circulatory effects of the head-down position
(negative G ) in normal man, with a note on some meas-
ures designed to relieve cranial congestion in this position.
J. Clin. Invest. 29: 940-949, 1 950.
310. Wilkins, R. W\, J. W. Culbertson, and A. A. Rymut.
The hepatic blood flow in resting hypertensive patients
before and after splanchnicectomy. J. Clin. Invest. 31:
529-53 1. ]952-
311. Winkler, K\. Urinary elimination of Bromsulfalein in
man. Scand. J. Clin. & Lab. Invest. 13: 44-49, 1961.
312. Winkler, K., and G Gram. Models for description of the
bromsulfalein elimination curves in man after single
intravenous injections. Acta Med. Scand. 169: 263-272,
1 96 1.
313. Yamada, S., and A. C. Burton. Effect of reduced tissue
pressure on blood flow of the fingers; the veni-vasomotor
reflex. J. Appl. Physiol. 6: 501-505, 1954.
314. Zeid, S. S., B. Felson, and L. Schiff. Percutaneous sple-
noportal venography, with additional comments on trans-
hepatic venography. Ann. internal Med. 52: 782-805, i960.
315. Zierler, K. L. A simplified explanation of the theory of
indicator dilution for measurement of fluid flow and vol-
ume and other distributive phenomena. Bull. Johns
Hopkins Hasp. 103: 199-217, 1958.
316. Zilversmit, D. B., G. A. Boyd, and M. Brucer. The
effect of particle size on blood clearance and tissue dis-
tribution of radioactive gold colloids. J. Lab. Clin. Med.
40: 255-260, 1952.
317. Zweifach, B. W. Direct observation of the mesenteric
circulation in experimental animals. Anal. Record 120:
277-291, 1954.
&^r^A
CHAPTER 42
The flow of blood in the mesenteric vessels'
EUGENE GRIM Department oj Physiology, University of Minnesota, Minneapolis, Minnesota
CHAPTER CONTENTS
Magnitude of Total Mesenteric Blood Flow
Partition of Total Blood Flow
Major Organs
Individual Tissues
Vessels of Different Sizes
Mesenteric Blood Volume
Factors Affecting the Blood Flow and Its Distribution
Stomach
Intestine
Pancreas
Spleen
Mesenteric Circulation as a Whole
Relation of Blood Flow to Function of the Mesenteric Organs
the mesenteric circulation is usually considered to
be that part of the systemic circulation which sup-
plies the stomach, small intestine, large intestine,
pancreas, and spleen. These organs receive blood
from all the branches of the celiac (except the hepatic
proper), the superior mesenteric, and the inferior
mesenteric arteries. They are not drained directly
into the venous system as are most organs, but into
the portal vein from which the blood passes through
a second set of capillaries in the liver before entering
the inferior vena cava.
Because of the peculiar anatomy of this venous
drainage system, the flow of blood in the mesenteric
vessels may be altered profoundly by factors which
do not act directly on these vessels but rather change
the resistance of the hepatic vasculature. This poses a
problem for the investigator who uses the intact
animal as the most "physiological" subject for study.
1 This chapter was written during the tenure of a U. S.
Public Health Service Senior Research Fellowship (SF-161).
Great care must be exercised in the interpretation of
the results of such studies, particularly when they
disagree with those from investigations on the iso-
lated mesenteric organs.
The mesenteric circulation as such has not pre-
viously been the subject of a comprehensive review,
although it has been considered in a subsidiary
fashion in reviews on the total splanchnic blood flow
by such authors as La Croix (92) and Bradley (21).
Even the standard textbooks of physiology, in which
can be found sections devoted to the circulation
through the heart, brain, lungs, kidneys, liver, and
skeletal musculature, contain few statements con-
cerning the circulation through the mesenteric
organs.
In part, the cause of this is the relative scarcity of
quantitative information on the subject and the
many instances in which different investigators have
published contradictory results. For the same reasons
many of the statements that follow should be taken
as tentative. This review might be better viewed as
indicating guide lines for future research rather than
as a definitive dissertation.
MAGNITUDE OF TOTAL MESENTERIC BLOOD FLOW
The total flow of blood through the mesenteric
system can be most directly determined by measuring
the flow through the portal vein. Since there has
been no suitable method for this measurement in the
human, all the available quantitative information
has been obtained in experimental animals, especially
the dog.
One of the earliest measurements of portal venous
flow was made by Burton-Opitz (30) as a part of
1439
1440
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
what probably remains to this day the most complete
study of the flow of blood through the mesenteric
organs. He placed a stromuhr in the portal veins of
dogs anesthetized with ether and obtained a mean
blood flow divided by the mean body weight of 14.3
kg of 19 ml per min per kg. His animals had relatively
low arterial blood pressures, the average being about
100 mm Hg, and a normal mean portal vein pressure
of 1 1 mm Hg.
In several subsequent studies made with thermo-
stromuhrs, values between 16 and 20 ml per min per
kg were obtained. These include Grab et al. (59),
Soskin et al. (128), Grodins et al. (71), and Grindlay
et al. (70). Some of the dogs used were unanesthetized;
others were anesthetized with such agents as ether,
chloralose, or sodium pentobarbital. Arterial blood
pressure was given only in the report of Grab et al.,
the mean being 100 mm Hg.
These investigations were performed with in-
struments which have since been severely criticized,
the stromuhr because it introduces a flow resistance
into the vessel in which it is placed and the thermo-
stromuhr for a variety of reasons [see, for example,
Gregg (68)].
MacLeod & Pearce (97) cannulated the thoracic
vena cava of ether-anesthetized dogs, occluding it
above and below the entrance of the hepatic veins
with balloons, and collected the outflow before and
after portal vein ligation. The mean total liver outflow
in animals with an arterial blood pressure of about
140 mm Hg was 44 ml per min per kg. This was
reduced by 60 per cent upon portal vein occlusion,
indicating that the usual flow through the latter was
about 26 ml per min per kg. Blalock & Mason (17)
used a somewhat similar technique in unanesthetized
dogs to measure the hepatic venous outflow im-
mediately after hepatic arterial ligation and ob-
tained a mean value of 24 ml per min per kg.
Electromagnetic flowmeters have been placed on
the portal vein by several groups of investigators.
Stewart et al. (129) and Drapanas et al. (41) found
the mean portal flow to be 25 ml per min per kg at
arterial pressures of about 140 mm Hg. Green et al.
(66) obtained a lower value, 1 7 ml per min per kg,
but the mean arterial pressure of their animals was
only slightly above 100 mm Hg.
Direct measurements of portal venous flow by
cannulation and collection of the blood was made
by Heimburger et al. (75). They obtained a mean
value of 30 ml per min per kg; however, since they
collected the blood by gravity thus producing an
unphysiological, negative pressure in the portal vein,
it seems likely that this value is too high.
The highest value for portal venous flow has been
reported by Sapirstein (114). He injected radio-
potassium and radiorubidium into both rats and
dogs and observed that the concentration of these
isotopes in all organs except the brain remained
nearly constant for a period of approximately 10 to
60 sec after injection. He concluded that the ex-
traction ratios for these substances must necessarily
be the same for all the organs, and hence that the
fraction of injected isotope found in any organ was
equal to the fraction of the cardiac output passing
through the organ. By adding the isotope contents
of the organs drained by the portal vein, Sapirstein
found that 20 per cent of the cardiac output passed
through them, a value which agrees well with the
findings of the previous workers. However, when
Sapirstein converted this value to units of flow per
unit body weight, he obtained 34 ml per min per kg.
The discrepancy arises from the fact that his dogs
were small (6-8 kg) and had a mean cardiac output
by the K42-dilution technique of 170 ml per min per
kg. In the larger animals used by most investigators
(12-20 kg) the cardiac output is usually about 125
ml per min per kg, 20 per cent of which is 25 ml per
min per kg.
The discrepancies among portal venous flows ob-
tained by various workers would seem in large part
to have been due to differences in arterial pressures
rather than to the measurement procedures. The
workers who found values of about 25 ml per min
per kg studied dogs having arterial pressures of
about 130 mm Hg, while those who observed 18 or
19 ml per min per kg used animals with pressures
of about 100 mm Hg. It would appear that the portal
venous blood flow in dogs of 10 to 20 kg body wt
having a "normar' arterial pressure of 130 mm Hg
is about 25 ml per min per kg. This is equivalent to
about 20 per cent of the cardiac output or approx-
imately 350 to 450 ml per min in a 15-kg animal. It
does not seem that this value is too much affected by
anesthesia with a variety of agents.
A very few measurements of total portal flow in
other species can be found in the literature. Sapirstein
and co-workers (109, 113, 133) in three separate
studies found the portal flow in rats anesthetized with
sodium pentobarbital to be 14, 16, and 20 per cent
of the cardiac output. Fegler & Hill (44) using a
thermodilution technique in sheep obtained a very-
high portal flow of 31 per cent of the cardiac output;
however, as they pointed out, members of this
FLOW OF BLOOD IN MESENTERIC VESSELS
I44I
species are exceedingly sensitive to abdominal trauma.
In the human, portal flow is usually estimated on the
assumption (from studies with the dog) that two-
thirds to three-fourths of the total hepatic flow as
determined by the Bromsulfalein technique (800-
850 ml/min m2) is derived from the portal vein. On
this basis, the portal venous flow is 530 to 640 ml per
min per m2, somewhat less than 20 per cent of the
cardiac output.
PARTITION OF TOTAL BLOOD FLOW
Major Organs
Two investigators, Burton-Opitz and Sapirstein,
have measured the blood flow through all the major
mesenteric organs. The former measured gastric
flow (29) by placing a stromuhr in the gastrosplenic
vein, ligating the pancreatic and splenic branches.
Ligation of the gastroduodenal and pyloric veins
presumably forced all the gastric venous drainage
through anastomotic channels into the stromuhr.
He obtained a mean flow of 0.25 ml per min per g
of stomach in dogs with a mean arterial pressure of
85 mm Hg. In another group of animals (27) he
placed the stromuhr in the common mesenteric vein
thus obtaining the blood flow through all the intestine
except the duodenum which is drained by the pan-
creaticoduodenal vein. At a mean arterial pressure
of slightly less than 100 mm Hg, the mean flow was
0.31 ml per min per g of intestine. His measurements
of pancreatic blood flow (31) were more difficult to
make as this organ is supplied and drained by num-
erous vessels. He placed the stromuhr in the gastro-
duodenal artery, ligated the right gastroepiploic
artery, and so obtained the flow through the superior
pancreaticoduodenal artery. This vessel supplies the
body of the pancreas and a portion of the duodenum.
The head of the pancreas receives arterial blood by
way of the inferior pancreaticoduodenal, a branch of
the cranial (superior) mesenteric, and the tail of the
organ by way of branches of the splenic artery. In
two animals he was able to separate the body of the
pancreas from the duodenum and so obtained the
pancreatic flow alone. The mean flow was 0.8 ml
per min per g at 1 10 mm Hg. To measure the splenic
blood flow, Burton-Opitz placed the stromuhr in the
splenic vein (28). In 10 animals, he obtained a
mean flow of 0.58 ml per min per g at an arterial
pressure of 98 mm Hg.
From Burton-Opitz' data, the weights of the
stomach, intestine, pancreas, and spleen in a 15-kg
dog can be estimated as 250, 500, 50, and 70 g,
respectively. The total blood flows through the same
organs would be 60, 155, 40, and 40 ml per min,
respectively, and the partition of the total mesenteric
flow about 20, 55, 13, and 13 per cent. In Sapirstein's
study (114) the partition of blood flow was determined
directly. He obtained values of 13, 72, 8, and 7 per
cent for the same organs.
The discrepancies in these two sets of data may
be due to one or more of several factors. As stated
earlier, Sapirstein's dogs were much smaller than
those of Burton-Opitz. In the former the weight of
the intestines was greater in relation to the weights
of the other mesenteric organs than in the latter.
Further, Sapirstein's dogs were anesthetized with
sodium pentobarbital and presumably had arterial
pressures 30 to 40 mm Hg higher. Finally, the meas-
urement techniques used may have resulted in er-
roneous values for one or more organs in the study.
It is possible that the resistance offered by the stro-
muhr to the intestinal venous outflow may have
caused Burton-Opitz to underestimate the propor-
tion of the total flow that passed through the gut.
On the other hand, Sapirstein's method may result
in either an underestimate or an overestimate of
flow through an organ. His assumption that con-
stancy of isotope content with time in all organs im-
plies identical extraction ratios is not wholly justified.
It is probably not too much in error as the extraction
ratio for radiopotassium in the first few seconds after
injection is nearly one for all organs. However, isotope
constancy can be observed in the presence of different
extraction ratios if the potassium ion concentrations
of the organs differ, as they do.
In order to compare the data of Burton-Opitz
and Sapirstein for any one organ with data obtained
by other investigators, it is desirable to express the
flows per unit weight of organ. Also, an attempt must
be made to normalize the values to some kind of
average animal. For Burton-Opitz' results this can
be done by correcting to an arterial pressure of
1 30 to 1 40 mm Hg, assuming that the flow increases
linearly with arterial-venous pressure difference.
The following values are thus obtained : stomach,
0.4; intestine, 0.4; pancreas, 1.0; and spleen, 0.8
ml per min per g of tissue. Sapirstein's values for the
same organs are 0.4, 0.7, 1.0, and 0.6, respectively.
These values were obtained by taking the cardiac
output to be 170 ml per min per kg. If Sapirstein's
distribution is applied to 1 5- or 20-kg animals with
a cardiac output of 125 ml per min per kg, his values
1442
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
would all be reduced by some 25 per cent; that is,
with the exception of the intestine, they would be
lower than those of Burton-Opitz.
Of the other workers who have measured gastric
blood flow, Boenheim (18) collected the venous
drainage directly and obtained a mean flow of 0.26
ml per min per g at the very low arterial pressure
of 60 mm Hg. Lim et al. (94) perfused with a donor
dog an isolated surviving stomach and collected the
venous outflow to find a mean value of 0.34 ml per
min per g at a perfusion pressure of 100 mm Hg.
Recently, Salmon et al. (112) used a method similar
to that of Boenheim's in dogs with blood pressures of
130 to 150 mm Hg and obtained a mean flow of
0.37 ml per min per g.
The literature contains widely varying values for
the blood flow through the intestines. Selkurt et al.
(119) measured mesenteric venous outflow in dogs
anesthetized with sodium pentobarbital and having
pressures of 1 30 mm Hg or more. They found a mean
flow of 8.7 ml per min per kg body wt which is
equivalent to about 0.2 ml per min per g organ.
In a later study with a rotameter, Selkurt (120)
obtained flows 50 per cent or higher, but no body-
weights were given so direct comparison cannot be
made.
A large number of measurements of venous outflow
from segments of small intestine have been made in
the writer's laboratory in the past few years. These
were innervated and denervated segments, in situ,
in dogs weighing 1 2 to 20 kg, anesthetized with
sodium pentobarbital and having arterial pressures
of 1 20 to 1 50 mm Hg. Although there was a large
variation in values from segment to segment, the
mean flows were about 0.6 ml per min per g, being
slightly higher in the upper jejunum than in the
lower ileum. These values were obtained with a
venous pressure of zero. When the venous pressure
was elevated to 10 mm Hg, the flow was generally
reduced by 10 to 15 per cent.
Brodie and co-workers (23, 24) measured the
blood flow through small intestinal segments
plethysmographically and obtained a mean value of
0.4 ml per min per g. Neely & Turner (103) used a
somewhat similar technique, measuring weight
changes following venous occlusion to find 0.28 ml
per min per g. Results of such studies as these two
must be considered in light of the prompt rise in
intestinal vascular resistance which follows an acute
rise in venous pressure [see, for example, Selkurt &
Johnson (122) and Johnson (85)].
Selkurt et al. (121) artificially perfused segments of
ileum and obtained flows with an arterial-venous
pressure difference of 130 mm Hg of about 0.25 ml
per min per g. This is much lower than flows ob-
tained with similar preparations in this writer's
laboratory. In our early work, very low flows were
frequently obtained; however, more normal flows of
0.5 to 0.6 ml per min per g were usual in later ex-
periments. The cause of the vasoconstriction in
intestinal segments which follows arterial cannulation
is not known to this writer, but it seems to be generally
prevented by topical application of procaine at the
site of the cannulation. This vasoconstriction can at
times be so intense as to reduce blood flow to less
than 0.05 ml per min per g.
Geber (54) has recently placed an electromagnetic
flowmeter on cannulae placed in the arterial circuit
of segments of dog's intestine and obtained very
high flows. His values were duodenum, 1 .38; jejunum,
0.98; ileum, 0.82; and colon, 0.73 ml per min per g.
It is difficult to believe that these values are not falsely
high. If correct, the intestinal venous outflow would
be equal to or greater than the total portal venous
flow as measured by most workers. It is possible that
Geber trimmed the mesentery from the intestinal
segments before weighing them. In some animals,
this would reduce the segment weight by 25 to 50
per cent, and thus result in high estimates of the
perfusion rates.
Several investigators have attempted to measure
the blood flow through the cranial mesenteric artery
of the dog. Trapold (132) using a Shipley rotameter
and Deal & Green (38) using an electromagnetic
flowmeter found flows in the range of 10 to 60 ml per
min, the average being less than 2 ml per min per kg
body wt. This is a surprisingly low value; with the
exception of the flow through the relatively small
caudal mesenteric artery, the cranial artery supplies
the same tissues as are drained by the common mesen-
teric vein. It is possible that manipulation of the
mesenteric artery may result in vasoconstriction just
as does cannulation of the intestinal arteries. Cull
et al. (36) obtained higher flows in the cranial mesen-
teric artery (120 ml/min in dogs of unspecified
weights), but these are still significantly lower than
would be expected. Grodins et al. (71) used a thermo-
stromuhr to obtain mesenteric artery flows of a
more expected value of 1 2 ml per min per kg. Meyer
(100) ligated the gastroduodenal and caudal mesen-
teric arteries and collected venous outflow from that
part of the gut supplied by the cranial artery (je-
junum, ileum, and proximal portion of the colon).
He obtained flows of the same order as those of
FLOW OF BLOOD IN MESENTERIC VESSELS
H43
Grodins, 180 to 190 ml per min from tissue averaging
370 g in weight (0.5 ml/ min g).
The studies on pancreatic blood flow have not
generally supplied sufficient information to permit
calculation of the flows per unit organ weight;
hence comparison with the values of Burton-Opitz
and Sapirstein are difficult. Babkin & Starling (6)
perfused the superior pancreaticoduodenal artery
from a heart-lung preparation and collected the
venous outflow in dogs under morphine and chloralose
anesthesia. They did not separate the pancreas from
the duodenum. In one experiment they observed a
control flow of 100 ml per min and in another 30
to 40 ml per min. No animal or organ weights were
given. Gayet & Guillaume (52, 53) measured the
outflow from the superior pancreatic vein in dogs and
obtained resting values of 20 to 25 ml per min in
animals of unspecified weight. Bennett & Still (14)
placed a stromuhr in the superior pancreatico-
duodenal vein of dogs of 8 to 9-kg body wt. Various
anesthetic agents were used: sodium barbital, sodium
amytal, and chloralose. They separated the pancreas
from the duodenum and estimated that they were
measuring the drainage of about one-half of the
organ. The mean flow was 6.2 ml per min. This is of
the order of 0.5 to 0.6 ml per min per g of tissue.
Grindley et al. (69) used a thermostromuhr to
measure splenic blood flow in unanesthetized dogs.
They obtained a mean value of 95 ml per min which
is 50 per cent higher than that observed by Burton-
Opitz. Ottis et al. (104) employed electromagnetic
flowmeters in dogs anesthetized with sodium pento-
barbital and observed flows of about 30 ml per min
in dogs of approximately the same weight. While
this is much lower than observed by Burton-Opitz,
it is in better agreement with the findings of Sapir-
stein. It is important to note that in expressing splenic
blood flows per unit weight of organ, care must be
exercised in the definition of the organ weight.
Burton-Opitz found the spleen of his etherized dogs
to weigh about 5 g per kg body wt. postmortem. As
is well known, the spleen in animals anesthetized
with pentobarbital weigh three to four times this,
and Burton-Opitz' value of 0.8 ml per min per g
would accordingly be reduced to 0.2 or 0.25 ml per
min per g.
An attempt has been made to summarize all this
data and to choose mean values which it is hoped
approximate the situation in the intact dog; these
are shown in table 1 .
table 1. Blood Flow Through Mesenteric Organs of a
15-kg Dog Having an Arterial Blood Pressure
of Approximately 130 mm Hg
Weight, g
Blond Flow
Organ
ml min, g organ
mJ/min
animal
% of total
Stomach
Intestine
Pancreas
Spleen
250
500
5°
70* (25ot)
035
°-5
0.8
0.7* (0.2 f)
90
250
40
50
20
60
10
IO
* Weight and flow in ether anesthesia. f Same in pento-
barbital anesthesia.
Individual Tissues
Since the arterial supply and venous drainage of
any tissue of a complex organ is by way of thousands
of microscopic vessels, the usual direct methods
cannot be employed to measure tissue blood flow.
Instead, one must resort to methods based on the
Fick principle.
Shore et al. (125), in a study of the secretion of
basic drugs by canine gastric mucosa, found that
the clearance rates of drugs having a pK of 5 or
greater were equal and maximal for all the drugs
tested. They concluded that this maximal value was
equal to the mucosal blood flow. By measurement of
the concentration of drug in the gastropyloric venous
blood, they further discovered that the maximal
clearance was about two-thirds of the total gastric
blood flow; that is, approximately two-thirds of the
total flow passed through the secreting portion of the
mucosa. Schanker et al. (116) determined the clear-
ance of one of the same drugs, aniline, by the rat
stomach to be 75 ml per hour. On the basis of Sapir-
stein's findings, the total gastric blood flow in rats
of the same size is 140 to 150 ml per hour. Thus, a
minimum of one-half the total flow passes through
the secreting mucosa.
Two different techniques have been employed by
workers in this writer's laboratory to obtain a reason-
ably complete analysis of the distribution of blood
flow through the tissues of the canine small intestine.
Lindseth (95) employed glass microspheres of 12 /u
in diameter labeled with Na24 in dogs anesthetized
with sodium pentobarbital. After the injection of a
small quantity of the spheres into an intestinal artery,
the segment supplied by that artery was removed
and separated into its component tissues: mucosa,
submucosa, muscularis, and mesentery. The frac-
tion of the injected spheres in each tissue was de-
i444
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
termined by counting the isotope. Since the spheres
were too small to lodge in any arteriovenous vessels
except the capillaries, this fraction should represent
the proportion of the total blood flow which passed
through the capillaries of that tissue. For fasted ileal
segments the proportions were mucosa, 38 per cent;
submucosa, 8 per cent; muscularis, 22 per cent; and
mesentery, 15 per cent. The remaining 17 per cent
passed through vessels larger than 12 /x in diameter.
The flows in milliliters per minute per gram of tissue
were mucosa, 0.42; submucosa, 0.34; muscularis,
0.48; and mesentery, 0.69. For fasted jejunal seg-
ments, the distribution and flows per gram of tissue
were not significantly different, except that the total
blood flow was somewhat higher, the difference
passing through arteriovenous channels larger than
12 xx.
The other method was employed by Rayner et al.
(107) in dogs under morphine analgesia and by
Weiner (137) in dogs under pentobarbital anesthesia.
Segments of fasted ileum were artifically perfused
with blood containing deuterium oxide for short
periods of time. On the assumption that the kinetics
of distribution of isotopic water is blood-flow limited,
the perfusion rate for each tissue was calculated from
its isotopic content at the end of the perfusion period.
In milliliters per minute per gram of tissue, Rayner
et al. obtained the following flows: mucosa, 0.38;
submucosa, 0.56; muscularis, 0.66; and mesentery,
0.23. Weiner found 0.42, 0.50, 0.51, and 0.16 for
the same tissues.
Comparison of the results obtained with the
microsphere and with the D20 methods in pento-
barbital-anesthetized animals shows the major
discrepancy to be in the estimation of flow through
the mesentery. It seems most likely that the D20
technique underestimates flow through this fatty
tissue. The higher submucosal flow obtained with
the latter technique may indicate that some D>0
exchange occurs across the walls of arteriovenous
bridges which abound in that tissue.
Vessels of Different Sizes
Many histological studies have demonstrated that
the arterioles and venules of the mesenteric organs
are connected by vessels varying from true capillaries
with a diameter of 10 ju or less to arteriovenous bridges
or throughfares with diameters in the range 10 to
20 xx to true arteriovenous anastomoses having
diameters in excess of 20 xt. The distribution of organ
blood flow to these different-sized channels has been
investigated in only a very few instances.
Walder (135) estimated the flow through arterio-
venous anastomoses in the artifically perfused human
stomach by measuring the perfusion rate before and
after presumably blocking all the capillaries with
starch granules. On this basis, he concluded that the
proportion of the total flow passing through arterio-
venous anastomoses was about 5 per cent.
Lindseth (95) measured the blood flow through
arteriovenous vessels of different sizes by injecting
small, known numbers of radioactive spheres of
different diameters into the arterial supply of the
canine intestine. He employed spheres with mean
diameters of 12, 20, and 44 /x. Measurement of the
number of such spheres which passed through the
organ into the venous blood permitted him to calcu-
late the partition of the total blood flow among the
arteriovenous channels of the three sizes. He found
that essentially no 44 xx spheres passed through in
either the jejunum or ileum; 3 to 4 per cent of the
blood flowed through vessels having diameters be-
tween 20 and 44 /x; and 24 per cent in the jejunum
and 14 per cent in the ileum passed through vessels
1 2 to 20 xx in diameter. Thus, in the fasting canine
intestine, the fraction of the total blood flow which
passes through true arteriovenous anastomoses
(vessels greater than 20 /x in diameter) is very small.
However, a quite significant fraction may flow
through arteriovenous bridges and hence bypass the
capillaries.
The finding that such a small fraction of the blood
passes through channels larger than 20 xx is in
agreement with results of investigations by Gordon
et al. (58). These workers estimated the size of the
largest arteriovenous channels in the intestine of rats
and rabbits anesthetized with sodium pentobarbital
by determining the minimum pressure required to
force mercury, air, or kerosene through the vascula-
ture of these organs. They concluded that the intes-
tines contain no vessels larger than 25 xx in diameter.
Several investigators have injected large quantities
of glass microspheres into the arteries supplying one
of the mesenteric organs and determined the max-
imum size of the spheres which passed through.
Sherman & Newman (124) did this in the stomach
and duodenum of the dog and recovered some
spheres as large as 100 to 180 it. Prinzmetal (106)
reported spheres of 160 to 37011 in the splenic vein
of dogs studied in the same way. Walder (134) made
a more complete study and used much smaller
quantities of spheres in the artificially perfused human
stomach. He found the diameter of the largest
spheres to pass through were 1 40 ll and of the mean
FLOW OF BLOOD IN MESENTERIC VESSELS
'445
iooju. The fraction of injected spheres which passed
through was not determined in these investigations.
MESENTERIC BLOOD VOLUME
Attempts to measure the volume of blood in the
mesenteric organs of the dog are complicated by the
presence of the spleen, the quantity of blood in this
organ varying greatly with the nature of the anes-
thetic agent. Further, since its hematocrit is much
higher than that of the body as a whole, estimates of
mesenteric blood volume from the distribution of
labeled red cells can be expected to be too high and
those made with labeled plasma constituents too low.
A correct value can be obtained only if both red cell
and plasma volume are measured simultaneously.
Although such a direct study has not been made,
it is possible to approximate the mesenteric blood
volume by combining the results of several different
investigations. One of the most pertinent of these is
Johnstone's (87). He placed ligatures around the
esophageal-gastric junction and the rectum of dogs
anesthetized with sodium pentobarbital, and in-
jected P'i2-labeled red cells. After a 5-min mixing
period, he clamped the celiac axis, mesenteric arteries,
and portal vein, simultaneously. By analyzing these
organs for P32, he found that they contained 22 per
cent of the injected red cells.
To calculate the blood volume from this observa-
tion, the hematocrit of the mesenteric organs, es-
pecially the spleen, must be known. Allen & Reeve
(2) determined both the red cell and plasma volume
of spleens from pentobarbital-anesthetized dogs.
They found the blood volume to be 4 to 10 per cent
of the total body blood volume and the hematocrit
1.7 times that of the large vessels. The ratio of the
large vessel hematocrit to that of the whole body is a
variable quantity as pointed out by Baker & Reming-
ton (7); however, in dogs anesthetized with pento-
barbital like those of Allen and Reeve, Reeve et al.
(108) found the ratio to be about 0.9; that is, the
splenic hematocrit would be about 50 per cent greater
than that of the whole body. The dogs of Allen and
Reeve were only lightly anesthetized and other
studies have shown that the spleen in more deeply
anesthetized animals may contain more than 10
per cent of the total blood volume. A reasonable
estimate, though, would be that the spleens of dogs
anesthetized with pentobarbital have 10 per cent of
the total blood volume and 15 per cent of the total
red cell mass. Combining this with Johnstone's
observations, the other mesenteric organs would
contain 7 per cent of the body's red cells and, as-
suming their hematocrit to be about the same as the
body's, 7 per cent of the total blood volume. Thus,
the mesenteric organs would hold 1 7 per cent of the
total blood volume, or about 15 ml per kg body wt
in dogs under pentobarbital anesthesia. Under ether
anesthesia, with the spleen essentially empty of
blood, this value would drop to nearly 7 per cent
(6 ml/kg). In unanesthetized animals, the volume
should be between these two extremes. Friedman
(49) has shown that the spleens of unanesthetized
mice contains about one-half as much blood as those
of animals under pentobarbital anesthesia.
Horvath et al. (82) used the "exclusion technique"
of Delorme and co-workers (39) to determine the
volume of blood in the mesenteric organs plus the
liver. They found this to be 21 per cent of the total
blood volume, 6 per cent in the hepatic, 6 per cent
in the splenic, and 9 per cent in the mesenteric
artery beds. Most of their experiments were with
I131-labeled albumin and hence probably gave under-
estimates of the blood volume, particularly of the
splenic artery distribution. Their findings do not,
therefore, disagree significantly with the estimate
given above.
Measurement of the volume of blood contained in
the minute vessels of some of the mesenteric organs
was made by Gibson et al. (55) in dogs under light
morphine narcosis. These workers determined both
the red cell and plasma content of the drained organs
and found the stomach and intestine to contain 0.04
ml blood per g tissue and the spleen 0.5 ml per g.
The blood volume of organs of the rat was determined
by Everett et al. (43) with Fe59-labeled cells and
I131-labeled plasma in quick-frozen animals. For the
small intestine they obtained a blood content of
0.034 ml per g and for the spleen, 0.17 ml per g
Rieke & Everett (in) made similar measurements
with rats under pentobarbital anesthesia and found
0.047 ml per g of intestine and 0.32 ml per g of
spleen.
If the minute vessels of the stomach and intestine
contain about 0.04 ml per g of blood, these organs
in a 1 5-kg dog would contain about 30 ml of blood,
2-3 in the intestine. Thus, a dog not too deeply anesthe-
tized with pentobarbital would have a total mesen-
teric blood volume of some 200 to 250 ml, 60 per
cent of which would be in the spleen, 10 to 15 per
cent in the minute vessels of the other organs, and
the remaining 25 to 30 per cent in the large gastric
and intestinal vessels. This partition as well as the
I446 HANDBOOK OF PHYSIOLOGY -~ CIRCULATION II
total volume can, of course, vary greatly in both
physiological and pathological states.
FACTORS AFFECTING THE BLOOD FLOW
AND ITS DISTRIBUTION
Stomach
Stimulation of the splanchnic nerves decreases the
blood flow through the gastric vessels. This has been
demonstrated by Burton-Opitz (29) in the ether-
anesthetized dog, by Lim et al. (94) in the blood-
perfused canine stomach, by Thompson & Vane
(131) in cats anesthetized with chloralose, and by
Walder (134) in Ringer-perfused human stomachs.
Friesen & Hemingway (51), using a calorimetric
method, showed that the mucosal flow decreased
during sympathetic stimulation in unanesthetized
dogs. In the rat, Schnitzlein (117) observed blanching
of the gastric mucosa during splanchnic stimulation
and Arabehety et al. (5) found engorgement following
block or section of the same nerves. The latter ob-
servations, it should be emphasized, are of the
mucosal blood volume and do not necessarily demon-
strate that the blood flow through this tissue is de-
creased by sympathetic stimulation.
Care should be exercised in the interpretation of
the many observations of changes in mucosal color,
labeled red cell content, India ink density, etc.
produced by nervous stimulation or drug administra-
tion. Blanching may well occur without significant
change in the blood flow or even in face of an in-
creased blood flow. Engorgement may accompany an
increase in flow resistance, especially if that occurs as
a consequence of venular constriction. Such observa-
tions can properly be taken as indicating changes in
blood volume only.
Many of the investigations cited above (18, 94,
131, 134) have shown that the influence of epineph-
rine on the gastric circulation is quite similar to that
of splanchnic stimulation. In addition, Henning
et al. (76) using an acetylene clearance method
observed an apparent reduction in human mucosal
blood flow in response to administration of sympatho-
mimetic drugs. Peters & Womack (105) found that
epinephrine produced mucosal blanching in the dog.
They also injected glass microspheres into the arterial
supply and, finding more large spheres in the venous
outflow than in control studies, concluded that adren-
aline dilated arteriovenous anastomoses. This latter is
not in agreement with the findings of Walder (135),
who concluded that the increase in arteriovenous
anastomotic flow was due only to increased resistance
in the capillary system rather than anastomotic
dilation. Miller & Haszczyc (101 ) found that epineph-
rine reduced the number of blood-filled capillaries
in biopsy specimens from human gastrostomies.
Dolcini et al. (40) made similar observations in the
rat. Schnitzlein (117) observed mucosal engorgement
in rats given ergotoxine to block adrenergic influences,
although the same drug in Walder's (134) exper-
iments did not alter the perfusion rate significantly
from control values.
Burton-Opitz (29), Lim et al. (94), Boenheim (18),
Friesen & Hemingway (51) all found little or no
effect of vagal stimulation on gastric blood flow
unless peristaltic activity appeared, in which case
blood flow declined. Schnitzlein (117) did observe
mucosal engorgement in the rat with vagal stimula-
tion. He also found that the application of acetyl-
choline to the gastric muscularis produced contrac-
tions and mucosal blanching. Necheles et al. (102)
found that acetylcholine usually produced vasocon-
striction in Ringer-perfused rat stomachs. It was
stated that this was not the consequence of increased
motor activity, although the latter was not recorded.
Walder (134) reported that acetylcholine in some
cases reduced and in others increased the rate of
perfusion through human stomachs. He made no
comments concerning motor activity.
In the studies already referred to by Lim et al.,
Thompson and Vane, and Walder, histamine caused
a vasodilation in the stomach. Cutting et al. (37)
also observed increased gastric blood flow in cats
with this compound. Richards et al. (no), using a
calorimetric method, found that histamine increased
mucosal flow in the human stomach. In contra-
distinction, Necheles et al. (102) could observe no
effect of histamine in their Ringer-perfused rat
stomachs; and Boenheim (18) reported a decrease in
etherized dogs, although the arterial pressures of his
animals were very low. Peters & Womack (105)
observed a marked increase in the mucosal content
of arterially injected starch granules and India ink
during histamine administration in the dog. Miller
& Haszczyc (101) also saw an increase in filled capil-
laries in human mucosa as a consequence of the
drug. Kimbel et al. (90), on the other hand, found a
marked decrease in the P32-labeled red cell content
of the gastric mucosa of polycythemic patients given
histamine.
The influence of several other chemicals on gastric
blood flow has also been studied. Cutting et al. (37)
FLOW OF BLOOD IN MESENTERIC VESSELS
1447
found that pilocarpine increased flow in the cat.
Bishton (16) made a similar observation in guinea
pigs when the pilocarpine was administered topically.
Schnitzlein (117) saw mucosal engorgement in the
rat under the influence of this drug. Lim et al. (94)
observed little effect of Pitressin in their perfused
preparation, but both Cutting et al. and Boenheim
(18) found a decrease in flow under the influence of
this hormonal preparation. Lim et al. found that
sodium nitrite and Cutting's group that erythrol
tetranitrate increased flow. Dolcini and co-workers
(40) observed an increase in the gastric mucosal
content of arterially administered India ink in the
rat given serotonin or 5-hydrox\ tryptophan.
Salmon et al. (112) demonstrated that cooling the
dog stomach to 15 C reduced the blood flow to 30 to
40 per cent of control. Heating was shown by Cut-
ting's group to have the opposite effect.
Richards et al. (1 10) found that a variety of emo-
tional states, anxiety, tension, resentment, all in-
creased flow in the human gastric mucosa as evi-
denced by the increase of heat uptake by the luminal
surface. Wolf & Wolff (138) made an extensive study
of color changes (i.e., blood volume changes) in a
human gastrostomy. There was an increase in redness
following administration of histamine, alcohol, beef
juice, acetylmethylcholine, exposure to local warm-
ing, during discussion of food and coincident with
evidence of hostility. Blanching occurred during
fear, sadness, discouragement, exposure to cold, and
after administration of epinephrine, ergotamine, or
Pitressin.
Intestine
It is generally agreed that stimulation of the
splanchnic nerves causes vasoconstriction in the
intestine. As early as 1899, Bayliss & Starling (11)
demonstrated that such stimulation decreased the
volume of intestinal segments. Burton-Opitz (27)
observed a reduction in mesenteric venous flow in
etherized dogs without a significant increase in portal
vein pressure, thus showing that mesenteric resistance
was increased. Deal & Green (38) measured flow in
the cranial mesenteric artery and the appropriate
pressures in dogs anesthetized with pentobarbital
in order to determine intestinal vascular resistance.
Although their control flows were abnormally low
(less than 0.1 ml 'min g of tissue), they found an
increase of 50 per cent in resistance during splanchnic
stimulation. Both Celander (34) and Kock (91)
determined venous outflow from jejunal loops in
vagotomized cats under pentobarbital or chloralose-
urethan anesthesia and also found that splanchnic
stimulation reduced the flow.
Bulbring & Burn (25) observed a reduction in
intestinal volume in plethvsmographic studies with
etherized, adrenalectomized dogs and cats during
stimulation. After administration of ergotoxine, the
same procedure produced an increase in volume.
Atropine did not block the dilation phase, and they
concluded that the splanchnic nerves contained some
noncholinergic vasodilator fibers as well as the vaso-
constrictor elements. Deal and Green also found that
the sympatholytic agent, Ilidar, sometimes reversed
the constrictor effect of splanchnic stimulation and
that atropine had no influence on the reversal.
Folkow et al. (47) in their studies on cats concluded
that the vasodilator fibers could not be adrenergic
either and hence that there were probably no splanch-
nic vasodilators. They thought that the vasodilation
seen during splanchnic stimulation after ergotamine
or Dibenamine was probably due to relaxation of the
intestinal smooth muscle.
The primary effect of both epinephrine and nor-
epinephrine on the intestinal vasculature seems to
be the same as that of splanchnic stimulation.
Schwiegk (118) found epinephrine to decrease both
arterial and venous flow in dogs anesthetized with
chloralose. In cats, also anesthetized with chloralose,
Clark (35) found that epinephrine in all concentra-
tions reduced intestinal venous outflow. Folkow et
al. (47) observed vasoconstriction in the cat with
both epinephrine and norepinephrine, as also did
Kock (91). Grayson's group (61-64), using a calori-
metric method, demonstrated that both compounds
produced vasoconstriction in the mucosa and muscle
of human ileostomies and colostomies. Binit et al.
(15) observed an increase in the resistance of the
mesenteric arterial bed upon intra-arterial injection
of epinephrine in the dog under chloralose anesthesia.
Green and co-workers (38, 65) in their studies on
mesenteric artery flow in dogs anesthetized with
pentobarbital found that both compounds caused a
several hundred per cent increase in resistance.
Selkurt et al. (121) observed a reduced flow through
artificially perfused, denervated ileal segments under
the influence of both substances. Bohr et al. (19)
used the Zweifach preparation of the rat mesoap-
pendix to show that epinephrine and norepinephrine
were both constrictors whether administered intra-
venously or topically. Although these workers found
epinephrine the more potent compound, all the
other investigators (47, 65, 91) who compared the
'44-8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
two substances in the dog and the cat found nor-
epinephrine to be the more potent (see below).
Two groups of workers (65, 118) have observed a
secondary increase in blood flow in the dog intestine
following the primary vasoconstriction due to epineph-
rine administration, although in the cat, Folkow et
al. (47) could find the constrictor effect only. Several
studies (35, 47, 65) have shown that only vasodilation
occurs with epinephrine if given after such substances
as ergotoxine, Dibenamine, or Ilidar. These same
substances are capable of blocking the constrictor
effect of norepinephrine but do not reverse it.
Not all the results of investigations of volume or
weight changes of the intestine under the influence of
epinephrine and norepinephrine agree with the
findings on blood flow. Woods et al. (139), Goetz
(57) and MacLean et al. (96) all found that epineph-
rine caused a primary reduction in volume followed
by a secondary increase. The latter observed a de-
crease in weight with norepinephrine. Opposed to
these are the observations of Biilbring & Burn (25),
Goetz (57), with small doses, and Burn & Hutcheon
(26) that epinephrine increased the volume oi in-
testinal segments of dogs and cats. As Folkow (48)
has stated, these discrepancies might well be ex-
plained by the possibility that the smooth muscle
relaxing effects of epinephrine result in a decreased
transmural pressure in the intestinal vessels which for
small doses of the drug overbalance its usual con-
strictor effects. Aside from this possibility, it remains
highly questionable whether or not intestinal vascular
resistance changes can be deduced from variations in
the volume of the organ.
The influence of the parasympathetic portion of
the autonomic nervous system on the intestinal
vasculature is not completely clear. Celander &
Folkow (32) observed an increase in intestinal blood
flow in the cat as a part of the depressor response to
sinus nerve stimulation. Since this disappeared after
ergotamine, they concluded that there were no
dilator fibers involved and that the increase in flow
was the consequence of a reduction in constrictor
tone. The parasympathetic mediator, acetylcholine,
does seem to cause vasodilation, as shown by the
just-mentioned workers, as well as by Binit et al.
(15) and by Bean & Sidky (13). The latter took great
care to separate the effects of the compound on the
vasculature and the visceral smooth muscle and
showed that the increase in blood flow appeared
before the augmentation of motor activity. The
increase in blood flow was abolished or reversed by
vigorous segmental contractions. Howe\ er, one note
of caution should be made regarding all three of
these studies. All were performed with perfused
preparations and it may be that the control blood
flows were abnormally low as is so frequently the
case with perfused intestinal segments. Although
adequate control data are not given in any instance,
estimations from Bean and Sidky's results indicate
that their preparations may have had an abnormally
high constrictor tone. Care should be exercised,
therefore, in concluding that acetylcholine has a
dilator effect in the intact normally perfused in-
testine.
It seems reasonable to conclude that splanchnic
nerve stimulation, and administration of the adrener-
gic substances, epinephrine and norepinephrine,
produce vasoconstriction in the intestine, presumably
by excitation of alpha adrenergic constrictor re-
ceptors as reviewed by Green & Kepchar (66a).
The usual secondary dilation observed after epineph-
rine injection and especially the primary dilation
seen when epinephrine is administered following
Ilidar or ergotamine blockade indicates the presence
of beta adrenergic dilator receptors as well. This
offers an explanation of the lesser constrictor potency
of epinephrine since this compound, unlike nor-
epinephrine, stimulates both the constrictor and
dilator receptors. The small increase in flow which
occurs during splanchnic stimulation after admin-
istration of Ilidar or ergotamine seems more likely
to be explained by intestinal smooth muscle relaxa-
tion, or by mechanical distention of blood vessels
due to a rise in blood pressure, than by stimulation of
the beta dilator receptors. It is highly questionable
whether the vagus has any influence on the intestinal
circulation other than that secondary to augmenta-
tion of motor activity. Acetylcholine probably has a
dilator effect but unequivocal proof of this in the
intact intestine is not available.
The influence of other chemical compounds on the
intestine may be summarized as follows. Biilbring &
Burn (25) found histamine to produce a slight vaso-
dilation, as did Binit et al. (15). In this writer's
laboratory, on the other hand, this compound has
been found to produce constriction fairly consistently
in artificially perfused intestinal segments. Both
Selkurt's group (121) and Bohr and his colleagues
(19) found serotonin to be an intestinal vasocon-
strictor. The latter also found Pitressin to be a con-
strictor of intestinal surface vessels.
Vasodilation has been produced by isopropylnor-
epinephrine in the hands of Green et al. (65), by
curare in a study by Elwell & Bean (42), by adenosine
FLOW OF BLOOD IN MESENTERIC VESSELS
1449
triphosphate in the investigations of Selkurt et al.
(121) and of Binit et al. (15), and by topically applied
procaine in Grayson's research (61) on human mu-
cosal blood flow. Grayson also observed that cooling
a limb caused dilation in the colostomy mucosa
whereas heating the body produced constriction, the
direction of the changes being opposite to those in
the skin. Trapold (132) found that several ganglionic
blocking agents caused a small decrease in resistance
to flow in the mesenteric artery bed, although this
must be interpreted in light of the fact that his control
flows were abnormally low.
Sidky & Bean (12, 126) used their isolated in-
testinal segment preparation to investigate the effects
of variations in the concentration of the respiratory-
gases in the perfusion fluid. They found that hyper-
capnia and hypoxia resulted in an increase in blood
flow; hypocapnia resulted in vasoconstriction. Brick-
ner et al. (22) determined the total mesenteric flow
less that through the spleen in dogs breathing gas
mixtures containing various percentages of C02.
With less than 5 per cent C02, the circulatory changes
were minor; at levels of 5 to 16 per cent, there was a
significant decrease in mesenteric resistance.
Intestinal blood flow is profoundly influenced by
motor activity. Anrep et al. (4) perfused loops of dog
intestine and observed a decrease in venous outflow
during muscular contractions. Sidky & Bean (127)
in their studies of artificially perfused intestinal seg-
ments found that early in a contraction arterial inflow
decreased and venous outflow increased, with venous
pressure sometimes exceeding arterial pressure. If
the contractions were rhythmic and of short dura-
tion, they could augment the flow. If the duration of
a contraction was longer, the flow through the seg-
ment would decrease as a consequence of the fall in
arterial inflow. As expected, these effects were more
pronounced the stronger the contractions.
Lawson & Chumley (93) showed that increases in
intraluminal pressure to values below 30 mm Hg
caused a temporary decrease in blood flow followed
by recovery to control values. At higher pressures
only a partial recovery' was noted. Recovery was not
observed in segments placed in plaster casts or treated
with procaine, and denervation was without in-
fluence. They concluded that the stretching of the
gut wall initiated a vasodilation mediated through
intrinsic nerve networks.
Selkurt et al. (121) have investigated the relation
between blood flow through an artificially perfused
denervated ileal segment and the arterial-venous
pressure difference. They found the relationship to
be slightly curvilinear, convex toward the pressure
axis, with a positive intercept on that axis of about
15 mm Hg. Since, as already pointed out, their
observed flows at normal arterial-venous pressure
differences were quite low, some caution must be
exercised in applying their results to the normal
situation. However, Johnson et al. (84) found a
similar intestinal pressure-flow relationship in the
totally perfused dog, although with higher flows for
any given pressure.
Selkurt & Johnson (122) and Johnson (85) ob-
served that the effect of increasing intestinal venous
pressure was to produce a rise in vascular resistance
in the mesenteric bed. They concluded that the re-
sistance changes were not dependent on nervous
mechanisms but suggested that the elevation of
venous pressure induced a myogenic response in the
resistance vessels.
Johnson (86) also investigated the influence on
flow resistance of partial occlusion of an intestinal
artery. In 70 per cent of the cases the resistance de-
creased with arterial pressure reduction. He con-
cluded that this autoregulation of intestinal blood
flow was not clue to a local reflex but rather was the
consequence of a myogenic response of the vascular
smooth muscle.
Occlusion of the mesenteric artery also has an
effect on the systemic circulation, causing a rise in
arterial blood pressure. Sarnoff & Yamada (115)
observed large increases in blood pressure in the cat
and concluded that this effect was dependent upon
reflexes initiated by receptors in the abdominal
organs, particularly in the pancreas. In this species,
they considered such reflexes more important than
those originating in the carotid sinus and aortic
arch. Boyer & Scher (20) observed smaller pressure
changes in the same animal and concluded that
there was no evidence for the presence of baroceptors
in the mesenteric artery, and that the rise in systemic
arterial pressure was due only to mechanical diver-
sion of the blood away from the abdominal viscera.
Heymans et al. (79) performed similar studies with
the dog and decided that the general blood pressure
rise was a purely hemodynamic effect due to the
exclusion of an important arterial vascular area and
did not indicate the existence of abdominal baro-
ceptors. Selkurt & Rothe (123) performed similar
studies in both dogs and cats. The results with cats
agreed with the findings of Sarnoff and Yamada.
Those obtained from dogs led the authors to con-
clude, in agreement with Heymans, that splanchnic
baroceptor activity in that species is slight.
1 45°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Herrick et al. (78) measured blood flow through
the cranial mesenteric artery of unanesthetized dogs
with a thermostromuhr during treadmill exercise and
found that the flow was essentially unchanged despite
an increase in arterial blood pressure, indicating an
intestinal vasoconstriction. Barcroft & Florey (9)
observed exteriorized preparations of colonic mucosa
of dogs during exercise. Early in the period mucosal
pallor was evident but, as the exercise continued,
the color returned to normal.
Several workers (50, 60) have attempted to study
the influence of various emotional states such as
depression, anxiety, fear, etc., on human intestinal
blood flow by making inferences from observations on
the degree of mucosal engorgement in colostomies.
The possible errors inherent in such inferences have
already been alluded to.
Pancreas
The information available on the influence of
nervous stimulation and of drugs on pancreatic
circulation is scanty. Both Anrep (3) and Gayet &
Guillaume (52) showed a reduction in venous outflow
as a consequence of splanchnic stimulation. The
effect of vagal stimulation is not quite so clear. Anrep
concluded that the vagus carried neither constrictor
nor dilator fibers to the pancreas. Gayet and Guil-
laume consistently found a marked increase in blood
flow during vagal stimulation.
Gayet & Guillaume (52), Maltesos & Watson
(98), Jones (88), and Bennett & Still (14) all ob-
served an increase in blood flow when secretin was
administered, in contradistinction to Weaver (136),
who could find no change in venous outflow. J ones
found that the rise in flow was a function of splanchnic
vasomotor tone and could be quite small when the
tone was high. Because Bennett and Still observed a
secretin-induced rise in blood flow only when the
pancreatic duct pressure increased, they proposed
that the apparent vasodilator action of the hormone
was not due to a direct effect on the vasculature but
was the consequence of a reflex initiated by the rise
in ductal pressure during secretion. They concluded
that a truly "vasodilation-free" secretin may be
prepared.
Recently, Holton & Jones (80) used a photoelectric
technique to measure blood content changes in the
pancreas and found that acetylcholine, histamine,
secretin, and pancreozymin all produce vasodilation,
whether or not secondary to a rise in ductal pressure
is not clear.
Spleen
In most of the investigations on the splenic circula-
tion, attention has been directed toward changes in
the volume of the organ rather than the blood flow
through it. Adrenergic stimulation causes a marked
decrease in volume in dogs and cats. As shown by
Celander (33) in cats under chloralose anesthesia,
sympathetic stimulation is more potent in this regard
than epinephrine, which in turn is several times more
effective than norepinephrine. Others, such as Ahl-
quist et al. (1) and Holtz et al. (81) have demon-
strated that epinephrine is also more effective than
norepinephrine in the dog. Many other compounds
produce splenic contraction; ephedrine, pituitrin,
histamine, acetylcholine, and amyl nitrite. Anesthetic
agents also exert a profound influence; as shown by
Hausner and co-workers (74), ether causes a reduc-
tion in size and various barbiturates a marked en-
largement over that of the waking animal. Hahn
et al. (72) reported that spleens taken from dogs
anesthetized with pentobarbital weighed four times
those from etherized animals. Almost any change in
the environment which can produce a sympathetic
discharge in the animal causes splenic contraction.
Thus, Hargis & Mann (73) and Barcroft and co-
workers (8, 10) observed this in waking dogs sub-
jected to a loud noise, tail pinching, hemorrhage,
exercise, or exposure to cold. The first mentioned
workers thought that most of these responses were
reflex, since they occurred so rapidly and were not
observed after denervation. Barcroft and Elliott,
however, did find contraction of the denervated spleen
after a loud noise, although it was delayed and
progressed slowly. One of the few maneuvers which
increases splenic volume is feeding.
A number of investigators, for example, Glaser
et al. (56), have concluded that the spleen is not an
important blood storage organ in the human body,
and hence does not change volume as markedly as
in the dog or cat.
With regard to factors influencing the splenic blood
flow, Burton-Opitz (28) found that stimulation of the
splanchnic nerve or any of the fibers of the splenic
plexus caused a reduction in blood flow through the
splenic vein. Green and co-workers (67, 104) studied
this in more detail and found that splanchnic stimula-
tion of short duration decreased arterial inflow but
temporarily increased venous outflow, thus accounting
for the reduction in volume of the organ. They also
FLOW OF BLOOD IN MESENTERIC VESSELS
1451
showed that epinephrine and norepinephrine had a
similar effect, with the former being more potent.
Phenoxybenzamine reversed the inflow reduction, and
reduced the increase in venous outflow and the volume
change. Acetylcholine and methacholine increased
arterial and venous flow and slightly increased organ
volume, these effects being blocked by atropine.
This observation is in agreement with that of Hunt
(83), but at variance with those of Ferguson et al.
(45) and of Fleming & Parpart (46) who observed
arteriolar constriction in the mouse spleen with
topical application of acetylcholine as well as epineph-
rine, norepinephrine, and histamine. An extensive
investigation by Grindlay and co-workers (6g) with
thermostromuhrs in unanesthctized dogs showed that
a loud noise resulted in a temporary increase in
venous outflow while having no effect on arterial
inflow, thus accounting for the usual volume reduc-
tion of the organ. They also found that both arterial
and venous flow rose after feeding and fell after
hemorrhage in agreement with volume changes.
1 Hiring exercise both flows increased. Since splenic
volume decreases during exercise, this provides a
good example of the danger inherent in assuming
that the direction of volume change of an organ
indicates the direction of flow change.
Mesenteric Circulation as a Whole
The influence of nervous stimulation or drug
administration on the mesenteric circulation as a
whole must for the most part be inferred from a
synthesis of the effects of these factors on the separate
organs. Most studies in the intact animal have been
on the total splanchnic flow with no separation of
this into its hepatic arterial and portal venous com-
ponents. Even where the portal flow is determined
separately care must be exercised in the interpreta-
tion of the results, since the factor under study may
alter the portal flow by affecting hepatic resistance
and have no effect on mesenteric resistance. Only
when measurement of the portal flow is accompanied
by determination of the mesenteric arterial-venous
pressure difference is it possible to infer the effects
of the factor on the mesenteric circulation, and even
then the effect may not be direct; for example, a
passive dilation of mesenteric vessels due to a rise in
portal venous pressure as a consequence of a hepatic
resistance increase or the contrary myogenic vaso-
constriction studied by Selkurt and Johnson. One
such study in which pressures were recorded, al-
though mesenteric resistances were not calculated, is
that of Katz & Rodbard (89). Another pertinent
investigation is that of McMichael (99). The results
of these workers are considered below with a sum-
mary of what seems to be the best evidence to the
present time on the factors affecting blood flow
through the separate mesenteric organs.
There is general agreement that splanchnic stim-
ulation increases the resistance to blood flow through
the mesenteric circuit. Most results indicate that the
effect of norepinephrine and the primary effect of
epinephrine are similar, with norepinephrine the
more potent of the two except in the spleen. In gen-
eral, epinephrine has a secondary dilator effect which
is the only consequence of its administration following
treatment with various sympathetic blocking agents.
Katz and Rodbard, and McMichael found that
epinephrine first increased then decreased mesenteric
resistance. It might be noted that adrenergic stim-
ulation may result in a temporary increase in portal
venous flow despite the primary rise in resistance,
because such stimulation evokes splenic contraction
and the discharge of its stored blood.
Vagal stimulation probably has little if any sig-
nificant influence on the mesenteric blood flow,
except insofar as flow is changed secondary to an
increase in motility in the stomach and gut. Because
the results of studies with acetylcholine are contra-
dictory with all organs except the intestine, and there
their validity may be questioned, much the same
conclusion must be drawn for this factor for the
present.
Pitressin seems to have a constrictor effect in most
of the mesenteric organs. Again the data of Katz and
Rodbard, and of McMichael confirm this for the
mesenteric circuit as a whole. Since serotonin seems
to be a constrictor in the intestine, and since the
major part of the mesenteric flow passes through this
organ, the effect of this hormone on the circulation
as a whole is probably the same. The influence of
histamine on the gastric circulation seems to be
dilatory; however, its effect on the other organs is not
so clearly established. Katz and Rodbard's data
indicate little change in the over-all flow resistance in
the mesenteric organs under the influence of this
compound; the dilation in the stomach may be
balanced by constriction elsewhere. Finally, one
physiological maneuver, exercise, seems to cause
vasoconstriction in all the mesenteric circulation
except in the spleen.
H52
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
RELATION OF BLOOD FLOW TO FUNCTION
OF THE MESENTERIC ORGANS
The influence of the blood flow on the function of
the mesenteric organs seems to be clearly established
in only one respect; namely, that a certain minimum
flow is essential for the maintenance of the integrity
of the cells. Whether or not the alimentary activities,
secretion, absorption, and motility, of the stomach,
intestine, and pancreas require an augmentation of
the blood supply above the basal level is the subject
of conflicting evidence, although most of the admit-
tedly scanty evidence indicates that, necessary or not,
there is an increase in blood flow through these or-
gans after the ingestion of a meal.
Herrick et al. (77), in thermostromuhr studies in
the unanesthetized dog, observed that 1 to 2 hours
after taking a meal the cranial mesenteric artery flow
was increased to 50 to 60 per cent above control.
Since there were increases of similar magnitude in
flows through the femoral and carotid arteries at the
same time, they concluded that digestion caused a
general increase in cardiac output rather than a shift
of blood supply from other regions of the body to the
abdominal viscera. Reininger & Sapirstein (109)
used their K42 method to demonstrate that there was
a similarly uniform increase of about 30 per cent in
blood flow to all parts of the body of the rat after
feeding.
Brodie and co-workers (23, 24) measured the
oxygen uptake by segments of canine small intestine
and observed a 30 per cent increase during absorp-
tion of dilute salt solutions and a 60 per cent increase
during absorption of protein solutions. They reported
similar rises in blood flow, but this must be interpreted
in light of their use of the plethysmograph to make
the measurements. Lindseth (95) determined both
total intestinal segment flow and its partition among
the individual tissues in anesthetized fasted and fed
dogs. In upper jejunal segments, feeding produced no
significant change in total venous outflow but caused
a diversion of the flow through the mesentery to the
capillaries of the mucosa and submucosa. In the
ileum, there was a 25 to 30 per cent increase in total
flow, almost all of which went to the muscle, there
being essentially no change in that through the
absorbing mucosa. He also found no significant altera-
tion in the fraction of the total flow which passed
through arteriovenous anastomoses.
Numerous attempts have been made to determine
the relation between blood flow and secretion by the
gastric mucosa. Thompson & Vane (131) in their
studies with the perfused cat stomach observed
parallel changes in secretory rate and blood flow as
a consequence of sympathetic stimulation, epineph-
rine administration, and celiac arterial infusion of
histamine, and concluded that secretion could be
directly influenced by changing blood flow. Lim
et al. (94), on the other hand, in similar studies with
the dog found that histamine-induced secretion could
occur in face of a falling blood flow. Further, sodium
nitrite increased the blood flow without initiating
secretion. Cutting et al. (37) observed increases in
both parameters in the cat when given histamine or
pilocarpine. Pituitrin decreased blood flow and
volume secretion but had little if any effect on the
amount of acid produced. Warming the stomach or
the administration of erythrol tetranitrate increased
flow without stimulating secretion. Finally it has
been noted that vagal stimulation which induces
secretion has little or no effect of the total blood flow
through the stomach although it does cause mucosal
engorgement. There seems little doubt that the parie-
tal cells must require an increase in oxygen supply
during secretion. There are, however, a number of
ways by which this can occur without alteration in
the total gastric blood flow. Oxygen extraction can
rise, although the work of Peters & Womack (105)
indicates that such is not the case in the dog in re-
sponse to histamine injection or vagal stimulation.
Other possibilities include shifts of flow from arterio-
venous anastomotic channels, from other tissues,
or from other regions of the stomach to the fundic
mucosa. Experiments to measure the distribution of
blood flow to the different tissue and arteriovenous
channels in the basal and in the secretory states are
needed. It is the mucosal flow that is of real signifi-
cance and it may not vary in the same manner as
the total flow. On the basis of the evidence presently
available reasonable conclusions seem to be that
agents which decrease gastric blood flow below the
basal level prevent or at least markedly reduce
secretion, that increased blood flow does not of itself
initiate or augment secretion, and that whether or
not secretion is necessarily accompanied by an in-
crease in total gastric blood flow cannot be answered
definitely.
The investigations on the relation of pancreatic
secretion to blood flow have been critically reviewed
recently by Tankel & Hollander (130). They pointed
out the contradictory nature of the evidence presently
FLOW OF BLOOD IN MESENTERIC VESSELS
'453
available and stated that it does not warrant the
conclusion that pancreatic secretion is dependent
on the blood supply, except that a minimum flow is
required to maintain cellular activity and provide
fluid for secretion.
The relation between motor activity and blood
flow in the stomach and intestine have been referred
to earlier. Vigorous contractions, such as are pro-
duced by vagal stimulation, cause a reduction in
blood flow. On the other hand, a reduction in blood
flow may, as suggested by Celander (34), be re-
sponsible for the usually observed inhibition of
motility during sympathetic stimulation.
The best general conclusion seems to be that there
is as yet no clearly established demonstration that
the mesenteric organs need an augmentation of their
basal blood supply to perform their alimentary func-
tion. It seems clear that these organs receive their
proportionate share of the general rise in cardiac
output which follows feeding, but whether this is
coincidental or to satisfy an essential requirement is
debatable.
REFERENCES
1. Ahlquist, R., J. Taylor, C. Rawson, and V. Sydow.
Comparative effects of epinephrine and levarterenol in
the intact anesthetized dog. J. Pharmacol. Expll. Therap.
110:352, 1954.
2. Allen, T., and E. Reeve. Distribution of "extra plasma"
in the blood of some tissues in the dog as measured with
P32 and T-1824. Am. J. Physiol. 175: 218, 1953.
3. Anrep, G. The influence of the vagus on pancreatic
secretion. J. Physiol., London 50: 421, 1916.
4. Anrep, G., S. Cerqua, and A. Samaan. The effect of
muscular contraction upon the blood flow in the skeletal
muscle, in the diaphragm and in the small intestine.
Proc. Roy. Soc, London, B 114: 245, 1934.
5. Arabehety, J., H. Dolcini, and S. Gray-. Sympathetic
influences on circulation of the gastric mucosa of the rat.
Am. J. Physiol. 197: 915, 1959.
6. Babkin, B., and E. Starling. A method for the study of
the perfused pancreas. J. Physiol., London 61 : 245, 1926.
7. Baker, G, and J. Remington. Role of the spleen in
determining total body hematocrit. Am. J. Physiol. 198:
906, i960.
8. Barcroft, J., and J. Stephens. Observations on the size
of the spleen. J. Physiol., London 64: 1, 1927.
g. Barcroft, J., and H. Florey. The effects of exercise on
the vascular conditions in the spleen and the colon. J.
Physiol., London 68: 181, 1929.
10. Barcroft, J., and R. Elliott. Some observations on the
denervated spleen. J. Physiol., London 87: 189, 1936.
11. Bayliss, W., and E. Starling. The movements and
innervation of the small intestine. J. Physiol., London 24:
99. '899-
12. Bean, J., and M. Sidky. Effects of low 02 on intestinal
blood flow, tonus and motility. Am. J. Physiol. 189: 541,
■957-
13. Bean, J., and M. Sidky. Intestinal blood flow as in-
fluenced by vascular and motor reactions to acetylcholine
and carbon dioxide. Am. J. Physiol. 194: 512, 1958.
14. Bennett, A., and E. Still. A study of the relation of
pancreatic duct pressure to the rate of blood flow through
the pancreas. Am. J. Physiol. 106: 454, 1933.
15. Binit, L., M. Burstein, and D. Coullaud. Sur les
reactions vasomotrices au niveau de l'intestin grele
Compt. rend. soc. biol. 148: 1954, 1954.
16. Bishton, R. The effect of pilocarpine on gastric blood
flow. J. Physiol., London 124: 26P, 1954.
17. Blalock, A., and M. Mason. Observations on the blood
flow and gaseous metabolism of the liver of unanesthetized
dogs. Am. J. Physiol. 117: 328, 1936.
18. Boenheim, F. Uber das Minutenvolumen des Magens
und seine Beeinflussung durch Blutdruck, durch Vagus-
reizung, durch Histamin und durch Organextrakte.
Z. ges. expll. Med. 71 : 88, 1930.
■ 9- Bohr, D., M. Wolf, and P. Rondell. Comparison of
intravenous and topical effectiveness of various vaso-
constrictors on the terminal vascular bed of the rat
mesoappendix. Am. J. Physiol. 182: 311, 1955.
20. Boy-er, F., and A. Scher. Significance of mesenteric
arterial receptors in the reflex regulation of systemic blood
pressure. Circulation Research 8: 845, i960.
21. Bradley, S. Methods for the evaluation of the splanchnic
circulation. Proc. Harney Tercentenary Congress. 1958, p. 355.
22. Brickner, E., E. Dowds, B. Willits, and E. Selkurt.
Mesenteric blood flow as influenced by progressive
hypercapnia. Am. J. Physiol. 184: 275, 1956.
23- Brodie, T., and H. Vogt. The gaseous metabolism of the
small intestine. Part I. The gaseous exchanges during the
absorption of water and dilute salt solutions. J. Physiol.,
London 40: 135, 1910.
24- Brodie, T., W. Cullis, and W. Halliburton. The
gaseous metabolism of the small intestine. Part II. The
gaseous exchanges during the absorption of Witte's
peptone. J. Physiol., London 40: 173, 1 910.
25- Bulbring, E., and J. Burn. Sympathetic vasodilatation
in the skin and the intestine of the dog. J. Physiol., London
87: 254, 1936.
26. Burn, J., and D. Hutcheon. The action of noradrenaline.
Brit. J. Pharmacol. 4: 373, 1949.
27. Burton-Opitz, R. Uber die Stromung des Blutes in dem
Gebiete der Pfortader. I. Das Stromvolum der Vena
Mesenterica. Pfliigers Arch. ges. Physiol. 124: 469, 1908.
28. Burton-Opitz, R. fiber die Stromung des Blutes in dem
Gebiete der Pfortader. II. Das Stromvolum der Vena
lienalis. Pfliigers Arch. ges. Physiol. 129: 189, 1909.
29. Burton-Opitz, R. fiber die Stromung des Blutes in dem
Gebiete der Pfortader. III. Das Stromvolum der Vena
gastrica. Pfliigers Arch. ges. Physiol. 135: 205, 1910
i 454
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
30. Burton-Opitz, R. The vascularity of the liver. IV. The
magnitude of the portal inflow. Quart J. Exptl. Physiol.
4: 113, 1911.
31. Burton-Opitz, R. Uber die Stromung des Blutes in dem
Gebiete der Pfortader. V. Die Blutversorgung des Pfortners
und Pankreas. Pflugers Arch. ges. Physiol. 146: 344, 19 12.
Celander, O., and B. Folkovv. Are parasympathetic
vasodilator fibers involved in depressor reflexes elicited
from the baroceptor regions? Ada Physiol. Scand. 23 : 64,
■951-
Celander, O. The range of control exercised by the
sympathoadrenal system. Acta Physiol. Scand. 32 : Suppl.
1 16, 1954.
Celander, O. Are there any centrally controlled sympa-
thetic inhibitory fibers to the musculature of the intestine.
Acta Physiol. Scand. 47: 299, 1959.
Clark, G. The vaso-dilator action of adrenaline. J.
Physiol., London 80: 429, 1934.
Cull, T., M. Scibetta, and E. Selkurt. Arterial inflow
into the mesenteric and hepatic vascular circuits during
hemorrhagic shock. Am. ./. Physiol. 185: 365, 1956.
Cutting, W., E. Dodds, R. Noble, and P. Williams.
Effect of alterations in blood flow on gastric secretion.
Proc. Roy. Soc, London, B 123: 29, 1937.
Deal, C, and H. Green. Comparison of changes in
mesenteric resistance following splanchnic nerve stimula-
tion with responses to epinephrine and norepinephrine.
emulation Research 4: 38, 1956.
Delorme, E., A. MacPherson, S. Mukherjee, and S.
Rowlands. Measurement of the visceral blood volume in
dogs. Quart. J. Exptl. Physiol. 36: 219, 1 951.
Dolcini, H., I. Zaidman, and S. Gray. Hormonal and
pharmacologic influences on microcirculation in the rat
stomach. Am. J. Physiol. 199: 1 157, i960.
Drapanas, T., D. Kluge, and VV. Schenk. Measurement
of hepatic blood flow by bromsulphalein and by the
electromagnetic flowmeter. Surgery 48: 1017, i960.
Elwell, L., and J. Bean. Intestinal blood flow in curari-
zation. Am. J. Physiol. 174: 185, 1953.
Everett, N., B. Simmons, and E. Lasher. Distribution of
blood (Fe59) and plasma (I131) volumes of rats determined
by liquid nitrogen freezing. Circulation Research 4: 419,
IQ56
Fegler, G., and K. Hill. Measurement of blood flow
and heat production in the splanchnic region of the
anaesthetized sheep. Quart. ./. Exptl. Physiol. 43: 189, 1958.
Ferguson, J., A. Ivy, and H. Greengard. Observations
on the response of the spleen to the intravenous injection
of certain secretin preparations, acetylcholine and
histamine. Am. J. Physiol. 117: 701, 1936.
Fleming, W., and A. Parpart. Effects of topically-
applied epinephrine, norepinephrine, acetylcholine and
histamine on the intermediate circulation of the mouse
spleen. Angiology 9: 294, 1958.
F01 kcjw, B., J. Frost, and B. Uvnas. Action of adrena-
line, noradrenaline and some other sympathomimetic
drugs on the muscular, cutaneous and splanchnic vessels
of the cat. Acta Physiol. Scand. 15: 412, 1948.
48. Folkow, B. The nervous control of the blood vessels. In:
The Control of the Circulation 0/ tin Blood. London: Dawson,
I956.
49. Friedman, J. Effect of Nembutal on circulating and
3-'-
33-
34-
35-
36.
37-
38.
39-
40.
41-
42.
43-
44-
45-
46.
47-
tissue blood volumes and hematocrits of intact and
splenectomized mice. ,4m. J. Physiol. 197: 399, 1959.
50. Friedman, M., and W. Snape. Color changes in the
mucosa of the colon in children as affected by food and
psychic stimuli. Federation Proc. 5: 30, 1946.
51. Friesf.n, S., and A. Hemingway. The vascular response
01 the stomach to experimental alterations in the auto-
nomic nervous system of the dog. Am. Surgeon 18: 195,
'952-
52. Gayet, R., and M. Guillaume. Les reactions vasomo-
trices du pancreas etudiees par la mesure des debits
sanguins. Compl. rend. soc. biol. 103: 1106, 1930.
53. Gayet, R., and M. Guillaume. Les relations quanti-
tatives reciproques de la secretion du sue pancreatique
et du debit sanguin. Compt. rend. soc. biol. 103: I 2 16, 1930.
54. Geber, W. Quantitative measurement of blood flow in
various areas of small and large intestine. Am. J. Physiol.
198: 985, i960.
55. Gibson, J., A. Selicman, W. Peacock, J. Aub, J. Fine,
and R. Evans. The distribution of red cells and plasma
in large and minute vessels of the normal dog, determined
by radioactive isotopes of iron and iodine. J. Clin. Invest.
25 : 949- '946-
56. Glaser, E., D. McPherson, K. Prior, and E. Charles.
Radiological investigation of the effects of hemorrhage
on the lungs, liver and spleen with special reference to
the storage of blood in man. Clin. Sci. 13: 461, 1954.
57. Goetz, R. The control of the blood-tlow through the
intestine as studied by the effect of adrenaline. Quart. J.
Exptl. Physiol. 29 : 32 1 , 1 939.
58. Gordon, D., J. Flasher, and D. Drury. Size of the
largest arterio-venous vessels in various organs. .4m. ./.
Physiol. 173: 275, 1953.
5Q. Grab, VV., S. Janssen, and H. Rein. Uber die Grosse
der Leberdurchblutung. Z. Biol. 89: 324, 1929.
60. Grace, W., S. Wolf, and H. Wolff. The Human Colon.
New York: Hoeber, 1951.
61. Grayson, J. Observations on blood flow in human
intestine. Brit. Med. J. 2: 1465, 1950.
62. Grayson, J , «d H. Swan. Action of adrenaline, nor-
adrenaline and dihydrocrgocornine on colonic circulation.
Lancet I : 488, 1950.
63. Grayson, J., and H. Swan. The reactions of the colonic
circulation in man to adrenaline and noradrenaline.
J. Physiol., London ill: 14P, 1 950.
64. Grayson, J. The measurement of intestinal blood flow-
in man. J. Physiol., London 114: 419, 1951.
65. Green, H., C Deal, S. Bardhanabaedya, and A.
Denison. The effects of adrenergic substances and is-
chemia on the blood flow and peripheral resistance of the
canine mesenteric vascular bed before and during adren-
ergic blockade. J. Pharmacol. Exptl. Therap. 113: 115. 1955.
66. Green, H., L. Hall, J. Sexton, and C. Deal. Autonomic
vasomotor responses in the canine hepatic arterial and
venous beds. Am. J. Physiol. 196: 196, 1959
66a. Green, H., and J. Kepchar. Control of peripheral
resistance in major systemic vascular beds. Physiol. A' 1
39:6l7. '959-
67. Green, H., K. Ottis, and T. Kitchen. Autonomic
stimulation and blockade on canine splenic inflow,
outflow and weight. Am. J. Physiol. 198: 424, i960.
FLOW OF BLOOD IN MESENTERIC VESSELS
'455
68. Gregg, D. Thermostromuhr. In : Methods in Medical
Research. Chicago: Yr. Bk. Pub., 1948, p. 89.
69. Grindlay, J., J. Herrick, and F. Mann. Measurement
of the blood flow of the spleen. Am. J. Physiol. 127: 106,
■939-
70. Grindlay, J., J. Herrick, and F. Mann. Measurement
of the blood flow of the liver. Am. J. Physiol. 132: 489,
1 941.
71. Grodins, F., S. Osborne, A. Ivy, and L. Goldman. The
effect of bile acids on hepatic blood flow. Am. J. Physiol.
'32: 375. "941-
72. Hahn, P., W. Bale, and J. Bonner. Removal of red cells
from the active circulation by sodium pentobarbital.
Am. J. Physiol. 138: 415, 1943.
73. Hargis, E., and F. Mann. A plethysmography study of
the changes in the volume of the spleen in the intact
animal. Am. J. Physiol. 75: 180. 1925.
74. Hausner, E., H. Essex, and F. Mann. Roentgenologic
observations of the spleen of the dog under ether, sodium
amytal, pentobarbital sodium and pentothal sodium
anesthesia. Am. J. Physiol. 121 : 387, 1938.
75. Heimburger, I., S. Teramoto, and H. Shumacker.
Influence of general hypothermia and local gastric cooling
on portal blood flow. Surgery 47 : 534, 1 960.
76. Henning, N., L. Demling, and R. Gromotka. Con-
servative methods for the determination of blood flow of
the digestive organs. Am. J. Digest. Diseases 5: 655, i960.
77. Herrick, J , H. Essex, F. Mann, and E. Baldes. The
effect of digestion on the blood flow in certain blood vessels
of the dog. Am. J. Physiol. 108:621, 1934.
78. Herrick, J., J. Grindlay, E. Baldes, and F. Mann.
Effect of exercise on the blood flow in the superior mesen-
teric, renal and common iliac arteries. Am. J. Physiol.
128: 338, 1939.
79. Heymans, C, A. De Schaepdryver, and G. De Vle-
eschhouwer. Abdominal baro- and chemosensitivity
in dogs. Circulation Research 8: 347, i960.
80. Holton, P., and M. Jones. Some observations on changes
in the blood content of the cat's pancreas during activity.
J. Physiol., London 150: 479, i960.
81. Holtz, P., F. Bachmann, A. Engelhardt, and K.
Greeff. Die Milzwirkung des Adrenalins und Arter-
enols. Pfliigers Arch. ges. Physiol. 255: 232, 1952.
82. Horvath, S., T. Kelly, G. Folk, and B. Hutt. Measure-
ment of blood volumes in the splanchnic bed of the dog.
Am. J. Physiol. 189: 573, 1957.
83. Hunt, R. Vasodilator reactions I. Am. J. Physiol. 45:
197, 1918.
84. Johnson, J., V. Gott, and F. Welland. Perfusion rates
of brain, intestine and heart under conditions of total
body perfusion. Am. J. Physiol. 200: 551, 1961.
85. Johnson, P. Myogenic nature of increase in intestinal
vascular resistance with venous pressure elevation.
Circulation Research 7: 992, 1959.
86. Johnson, P. Autoregulation of intestinal blood flow.
Am. J. Physiol. 199: 31 1, i960.
87. Johnstone, F. Measurement of splanchnic blood volume
in dogs. Am J. Physiol. 185: 450, 1956.
88. Jones, M. The effect of secretin on pancreatic blood flow.
J. Physiol., London 151 : 49P, i960.
89. Katz, L., and S. Rodbard. The integration of the vaso-
motor responses in the liver with those in other systemic
vessels. J. Pharmacol. Exptl . Therap. 67: 407, 1939.
90. Kimbel, K., H. Kinzlmeier, and N. Henning. Unter-
suchungen zur Magendurchblutung. I. Mitterlung: Ver-
suche mit radioaktiven Phosphor. Gastroenterologia 82: 317.
■954-
91. Koch, N. An experimental analysis of mechanisms engaged
in reflex inhibition of intestinal motility. Acta Physiol. Scand.
47:Suppl. 164, 1959.
92. LaCroix, E. Splanchnic circulation. Ada gastro-enterol.
belg- 23:534. I96°-
93. Lawson, H., and J. Chumley. The effect of distention on
blood flow through the intestine. Am. J. Physiol. 131 : 368,
1940.
94. Lim, R., H. Necheles, and T. Ni. The vasomotor reactions
of the (vivi-perfused) stomach. Chinese J. Physiol 1 : 381,
1927.
95. Lindseth, E. Vascular Flow Patterns in the Tissues of the Dog
Intestine (Ph.D. Thesis). Minneapolis: Univ. of Minnesota,
i960.
96. MacLean, L., E. Brackney, and M. Visscher Effects of
epinephrine, norepinephrine and histamine on canine
intestine and liver weight continuously recorded in vivo.
J. Appl. Physiol. 9: 237, 1956.
97. MacLeod, J., and R. Pearce. The outflow of blood from
the liver as affected by variations in the condition of the
portal vein and hepatic artery. Am. J. Physiol. 35: 87,
]9'4-
98. Maltesos, C, and R. Watson. Durchblutung und
Sekretion des Pankreas bei humoraler Anregung. Pfliigers
Arch ges. Physiol. 241 : 516, 1939.
99. McMichael, J. The portal circulation. I. Action of
adrenaline and pituitary pressor extract. J. Physiol.,
London 75 : 241 , 1932.
100. Meyer, M. Hemodynamic Studies of Endotoxin Shock in the
Dog (Ph.D. Thesis). Minneapolis: Univ. of Minnesota,
1 961.
101 . Miller, E., and V. Haszczyc. Gastric mucosal capillaries
in the human. A.M. A. Arch. Surg. 73: 465, 1956.
102. Necheles, H., R. Frank, W. Kaye, and E. Rosenman.
Effect of acetylcholine on the blood flow through the
stomach and legs of the rat. Am. J. Physiol. 114: 695, 1936.
103. Neely, W., and M. Turner. Measurement of blood flow
in kidney and isolated segments of intestine. J. Appl.
Physiol. 14:37, 1959.
104. Ottis, K, J. Davis, and H. Green. Effects of adrenergic
and cholinergic drugs on splenic inflow and outflow
before and during adrenergic blockade. Am. J. Physiol.
l89:599> >957-
105. Peters, R., and N. Womack. Hemodynamics of gastric
secretion. Ann. Surg. 148: 537, 1958.
106. Prinzmetal, M. Arterio-venous anastomoses in the liver,
spleen and lungs. Am. J. Physiol. 152 : 48, 1948.
107. Rayner, R., L. MacLean, and E. Grim. Intestinal tissue
blood flow in shock due to endotoxin. Circulation Research
8: 121 2, i960.
108. Reeve, E., M. Gregersen, T. Allen, and H. Sear
Distribution of cells and plasma in the normal and
splenectomized dog and its influence on blood volume
estimates with P32 and T-1824. Am. J. Physiol. 175: 195,
'953-
109. Reininger, E., and L. Sapirstein. Effect of digestion on
1456
HANDBOOK OF PHYSIOLOGY-
CIRCULATION II
distribution of blood flow in the rat. Science 1 26 : 1 1 76,
'957-
Richards, C, S. Wolf, and H. Wolff. The measure-
ment and recording of gastroduodenal blood flow in man
by means of a thermal gradientometer. J. Clin. Invest. 21 :
55'. '94'2-
Rieke, W., and N. Everett. Effect of pentobarbital
anesthesia on the blood values of rat organs and tissues.
Am. .1. Physiol. 188: 403, 1957.
Salmon, P., W. Griffin, and O. Wangensteen. Effect of
intragastric temperature changes upon gastric blood flow.
Proc. Soc. Exptl. Biol. Med. 101 : 442, 1959.
Sapirstein, L. Fractionation of the cardiac output of rats
with isotopic potassium. Circulation Research 4: 689, 1956.
Sapirstein, L. Regional blood flow by fractional distribu-
tion of indicators. Am. J. Physiol. 193- 161, 1958.
Sarnoff, S., and S. Yamada. Evidence for reflex control
of arterial pressure from abdominal receptors with special
reference to the pancreas. Circulation Research 7 : 325, 1 959.
Schanker, L., P. Shore, B. Brodie, and C. Hogben.
Absorption of drugs from the stomach. I. The rat. J.
Pharmacol. Exptl. Therap. 120:528, 1957.
Schnitzlein, H. Regulation of blood flow through the
stomach of the rat. Anat. Record 127: 735, 1957.
Schwiegk, H. Untersuchungen iiber die Leberdurchblu-
tung und den Pfortaderkreislauf. Arch, exptl. Pathol.
Pharmakol. 168:693, 1 932.
Selkurt, E., R. Alexander, and M. Patterson. Role of
mesenteric circulation in the irreversibility of hemorrhagic
shock. Am. J. Physiol. 149: 732, 1947.
120. Selkurt, E. Splanchnic hemodynamics as influenced by
hepatic ischemia. Proc. Soc. Exptl. Biol. Med. 90: 427, 1955.
121. Selkurt, E., M. Scibetta, and T. Cull. Hemodynamics
of intestinal circulation. Circulation Research 6: 92, 1958.
122. Selkurt, E., and P. Johnson. Effect of acute elevation of
portal venous pressure on mesenteric blood volume, inter-
stitial fluid volume and hemodynamics. Circulation Re-
search 6: 592, 1958.
123. Selkurt, E., and C. Rothe. Splanchnic baroceptors in
the dog. Am. J. Physiol. 199: 335, i960.
124. Sherman, J., and S. Newman. Functioning arteriovenous
anastomoses in the stomach and duodenum Am. J.
Physiol. 179: 279. 1954.
125. Shore, P., B Brodie, and C. Hogben. The gastric secre-
"3-
114.
"5-
116.
117.
118.
"9
126.
127.
128.
129.
130.
i3»-
133-
134
'35-
■ 36.
137
138
139
tion of drugs: A pH partition hypothesis. J. Pharmacol.
Exptl. Therap. 119:361, 1 957.
Sidkv, M., and J. Bean. Local and general alterations of
blood CO. and influence of intestinal motility in regula-
tion of intestinal blood flow. Am. J. Physiol. 167: 413,
'95'-
Sidky, M , and J. Bean. Influence of rhythmic and tonic
contraction of intestinal muscle on blood flow and blood
reservoir capacity in dog intestine. Am. J. Physiol. 193
386, 1958.
Soskin, S., H. Essex, J. Herrick, and F. Mann. Mecha-
nism of regulation of the blood sugar by the liver. Am. J.
Physiol. 124:558, 1938.
Stewart, J., J. Stephens, M. Leslie, B. Portin, and W.
Schenk. Portal hemodynamics under varying experimen-
tal conditions. Ann. Surg. 147: 868, 1958.
Tankel, H., and F. Hollander. The relation between
pancreatic secretion and local blood flow : A review.
Gastroenterology 32: 633, 1957.
Thompson, J., and J. Vane. Gastric secretion induced by
histamine and its relationship to the rate of blood flow.
J. Physio/., London 121 : 433, 1953.
Trapold, J. Effect of ganglionic blocking agents upon
blood flow and resistance in the superior mesenteric artery
of the dog. Circulation Research 4: 718, 1956.
Vidt, D., A. Bredemever, and L. Sapirstein. Effect of
ether anesthesia on cardiac output, blood pressure, and
distribution of blood flow in albino rat. Circulation Research
7:759. '959-
Walder, D. Arteriovenous anastomoses of the human
stomach. Clin. Sci. 1 1 : 59, 1952.
Walder, D. Some observations on the blood flow of the
human stomach. In : Ciba Found. Symp., Visceral Circulation.
I952-
. Weaver, M. Studies on the visceral vasomotor responses
to intravenous injection of purified pancreatic secretin.
Am. J. Physiol. 85: 410, 1928.
Weiner, D. Kinetics of Distribution of D20 in the Tissues of
the Canine Ileum (Thesis). Minneapolis: Univ. of Minne-
sota, 1 96 1 .
Wolf, S., and H. Wolff. Human Gastric Function. New-
York: Oxford Univ. Press, 1943.
Woods, G., V. Nelson, and E. Nelson. The effect of
small amounts of ergotamine on the circulatory response
to epinephrine. J. Pharmacol. Exptl. Therap. 45: 403, 1932.
CHAPTER 43
The renal circulation
EWALD E. SELKURT
Department of Physiology, Indiana University
School of Medicine, Indianapolis, Indiana
CHAPTER CONTENTS
Functional Architecture of the Renal Circulation
Arterial System
Venous System
Glomerular Circulation
Juxtaglomerular Complex
Blood Supply to the Medullary Zones
Renal Lymphatic System
Nerve Supply to the Kidney : Anatomical Aspects
Extrinsic Nerves
Intrinsic Innervation
Distribution of Osmotic Constituents in the Kidney: The
Countercurrent Hypothesis
Renal Blood Volume: The Intrarenal Hematocrit
Metabolic Aspects
Oxygen Utilization
Heat Production
Pressure Gradients in the Renal Vascular Circuit
Pressure Gradients
Critical Closure; Yield Pressure
Intrarenal Pressure
Measurement of Renal Blood Flow
Methods
Critique of the Clearance Method
Renal Blood Flow Values
Extrinsic Regulation of Renal Blood Flow
Neurogenic Control
Humoral Control; Pharmacologic Agents
Anatomy of the Renal Circulation
History
Mechanism of Autoregulation
Present Status of the Trueta Juxtamedullary Shunt
Morphological Evidence
Functional Evidence; Interpretations Based on Clearance
Rate
Role of the Medullary Circulation in Diuresis and Anti-
diuresis
Response of Renal Blood Flow in Physiological Stress
Exercise
Posture and Orthostatic Hypotension
Renal Hypoxia and Ischemia
Hypercapnia and Acidosis
Hemorrhagic Hypotension and Shock
Concluding Remarks
the introduction of the concept of the counter-
current osmotic multiplier system to the kidney by
Wirz el al. (345-349) as a means of explaining
urinary concentration and dilution has apparently
initiated a phase of re-evaluation of classical renal
functional concepts which promises to be far reaching
in scope. Recent critical reviews, while pointing out
gaps in our knowledge, have nevertheless opened up
exciting vistas and new pathways for research (169,
171, 289, 314)- The countercurrent concept rests
rather firmly on findings in the rat and hamster;
however, significant anatomical differences in the
kidneys of the dog and man require that this hypoth-
esis be intensively tested in these and other species.
Only about one-eighth of the nephrons of the human
kidney appear to have the long medullary loops of
Henle requisite for the mechanism (245). Most lie in
the cortex, and have straight, short, thin segments, or
indeed, none at all (fig. 1). The dog, however, has
long loops and long, thin medullary segments, yet
its kidneys are not remarkably different from those of
the human in concentrating power.
The renal circulation has been found to play a
unique and important role in the composite picture of
the countercurrent mechanism. Knowledge of the
distribution of blood to the cortex and medulla has
assumed pre-eminent importance. Older ideas have
had to be revised. The vasa recta, considered originally
as a medullary shunt by Trueta et al. (311), assume
H57
H58
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
CORT
--PAP D
fig. I. Anatomical distribution of nephron types in the hu-
man kidney. CORT.: cortex; O.ST.: outer stripe or band;
O.Z.: outer zone of medulla; INST.: inner stripe or band;
MED.: medulla; IN.Z. : inner zone of medulla; PAP.D.: pap-
illary duct. [After Peter (245).]
new importance as the vascular counterpart of the
countercurrent system of the nephrons. Longley et al.
(191) look upon the vasa recta as retia mirabilia
conjugata (similar to the retia mirabilia of the swim
bladder of fishes), especially endowed to function as a
countercurrent multiplier system. Recent studies
(166, 309) have shown that the flow of blood through
these vessels appears to be significantly slower than
through the cortical circulation, apparently a func-
tional adaptation to the optimal operation of the
countercurrent mechanism.
The question of the possible role of the phenomenon
of the autoregulation of the renal circulation in the
countercurrent system has been raised. Speculatively,
it would appear undesirable for rapid fluctuations in
blood flow to occur through the zone of hypertonicity,
and the over-all constancy of renal blood flow may
thus be an adaptation to insure stability in this system.
This article will include largely the developments
in renal circulatory physiology since Homer Smith's
review in 1940 of The Physiology of Renal Circulation
(286). This era has seen the ascendancy of the
clearance method for measurement of renal blood
flow, the waxing and waning of the Trueta juxta-
medullary shunt mechanism, the development of a
growing interest in the mechanism of renal circulatory
autonomy, and the unfolding of the countercurrent
hypothesis of kidney function with important implica-
tions for the renal circulation.
FUNCTIONAL ARCHITECTURE OF THE
RENAL CIRCULATION
Limitation of space precludes the consideration of
the anatomy of the renal circulation on the broad
comparative basis that it warrants. Rather, major
emphasis will be placed on the salient features of
circulation in the dog, the species in which a significant
proportion of the functional studies have been made,
with appropriate references to other species, especially
human, when needed for full development of a given
topic.
Arterial System
Major distribution of the renal artery in the dog is
shown in figure 2 [from plastic injection corrosion
studies of von Kiigelgen et al. (322)]. Figure 3 shows
division of the interlobar artery into primary,
secondary, and tertiary arcuate arteries, from which
spring the interlobular arteries. The afferent arterioles
usually supply only one glomerulus, but rarely may
branch to supply 2 to 4 glomeruli with a total of
200,000 per kidney. This is compared to estimates
ranging from 600,000 to 1,700,000 in each human
kidney (213, 216, 318).
specialized arterial circuits. Spanner (290, 291),
Trueta et al. (311), Baker (6), and von Ki'igelen &
Passarge (323) have found peculiarly coiled vessels
(which arise from the interlobar arteries) in the renal
sinus of dog, cat, and human. These spiral vessels,
THE RENAL CIRCULATION
1459
ventr.
fig. 2. Horizontal section through the dog kidney. RI and
RII. renal artery and primary branches; Arc I and Arc II:
primary and secondary arcuate arteries; IL: interlobar artery;
ILl: interlobular artery; Caps: capsular artery, U: aorta. [After
von Kiigelgen et al. (322).]
100 to 150 ii in diameter, form a plexus which supplies
the calycine mucosa and the renal papilla (fig. 4).
Baker contends that they anastomose with the vasa
recta. It is important to emphasize that these vessels
participate with the vasa recta system (vide infra) in
supplying blood to the papillary zone containing the
tips of the loops of Henle, the site of maximal osmotic
concentration. Their long, coiled length delivers blood
into the vasa recta system at low pressure (6). Arterio-
arterial anastomoses occur in this system, an exception
to the usual pattern of end arteries found in the
divisions of the renal artery.
Venous System
Deferring discussion of the glomerular and capillary
circulation, attention is directed to the venous
system in figure 5. Note the sparsity of interlobular
veins relative to the interlobular arteries (a ratio of
20 to 1). Their function appears to be to connect the
superficial and deep venous systems of the cortex (V.
corticalis superficialis and V. corticalis profunda),
into which the capillaries drain. The upper fifth of
the cortex appears to be an "arterial-free'" zone (af
in fig. 5), so that the upper glomeruli are overlayed by
only venous channels (stellate veins, superficial
cortical veins), and prevenous capillaries (cortex
corticis). Puncture of glomeruli for this reason has
been unsuccessful in the dog. A "venous-free"' zone
(vf) also exists, free of cortical veins (superficial and
deep), and occupied only by occasional interlobular
veins.
venous sinuses; veno-venous anastomoses. Venous
sinuses or sinusoids lying in the connective tissue
adjacent to the pelvis of the human kidney were
observed by Spanner (290, 291) and by Barrie
et al. (12). Spanner described them as isolated
accumulations of large venous plexi arranged super-
ficially along the walls of the minor calyces of the
renal pelvis. Trueta el al. also described in the same
zone of the human kidney many vessels of large caliber
which unite interlobar veins (veno-venous anasto-
moses). These vessels lie closely adjacent to the outer
surfaces of the walls of the calyces of the renal pelvis,
and the capillaries of the pelvic mucosa drain into this
complex system. They may offer a clue to the phenom-
enon of pyelovenous backflow sometimes seen after
retrograde pyelography. Veno-venous anastomoses
fig. 3. Scheme of the finer arterial supply
of the dog kidney, gl: Glomerulus with vas
afferens. [After von Kiigelgen et al. (322).]
1460
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
CORTEX
fig. 4. The medullary and papillary blood supply of the
kidney. [After Baker (6).]
are common in the dog kidney between interlobar
veins, between arcuate veins, and between stellate and
arcuate veins (322).
arteriovenous anastomoses. The venous sinusoids
have been described as the site of numerous arterio-
venous anastomoses in the human kidney by Spanner.
Barriet et at. (12) believe that such arteriovenous
connections occur between the aforementioned
spiral arteries and the sinusoids, the latter emptying
presumably into the interlobar veins, but admit that
open communications between the spiral arteries and
the sinusoids are extremely difficult to demonstrate.
Trueta et at. have disputed Spanner's findings, and
Baker (6) found only an insignificant number in
confirmation of Trueta. Nor could von Kiigelen
et at. (322, 323) and Christensen (54) find them in
the dog kidney.
In summary, arterio-arterial and veno-venous
anastomoses occur commonly in the kidneys of man
and dog. Although arteriovenous anastomoses prob-
ably exist, their occurrence is infrequent. Func-
tionally, direct A-V shunting of blood must therefore
be negligible, and blood passes generally through a
capillarv circuit (cortical peritubular plexus or
medullary vasa recta system). In confirmation,
Piiper & Schiirmeyer (249) found that the intact
dog kidney passed only 1.5 per cent of 19 n wax
spheres injected into the renal artery, 0.3 per cent
of 30 n, and 0.08 per cent of 38 ^ size. Denervation,
and injection of KCN, novacaine, and histamine did
not influence the results. Although Simkin et at.
(283) recovered glass spheres up to 440 m in size from
the renal vein of excised human kidneys, they did not
indicate what portion these represented of the total
injected.
venous valves and valvelike structures. These
have been described by von Kugelgen et at. (320-322)
in the dog, swine, and human kidney. They are
V.s fellah.
fig. 5. The blood supply of the cortex, in-
cluding the venous system, a/: Arterial-free
zone of the cortex; vf: vein-free zone of the
cortex. [After von Kugelgen et at. (322).]
V. corti calis
superficial/!
/cortical is
profunda
V. interlobularis
V.interlobans
THE RENAL CIRCULATION
I 46 I
located in the renal vein at its entrance to the vena
cava, and in the main branches of the renal vein (the
latter not as a rule in man). They are found also at
the orifices of the interlobar veins, arcuate veins, and
occasionally just before the opening of the capsular
(stellate) veins into the interlobular veins.
Koester el al. (162) have found, in both human and
dog kidneys, structures in the veins which might act
as effluent constrictions, which they described in
terms of "stenoses" and '"sinusoidal cushions."
Stenoses are common at the ostia of smaller tributaries
entering interlobar veins in the human kidney. They
seem to be composed of a dense collagenous framework
lined with endothelium; usually muscle is present as
a proliferation of the media of the vessel. In the dog,
stenoses appear occasionally along the course of the
interlobars and primary tributaries to the renal vein;
however, they are present primarily at the con-
fluence of arcuates with interlobars and of the
interlobular with arcuates.
At the confluence of arcuates with one another to
form an interlobar vein in the human kidney, the
sinusoidal cushions usually appear, sometimes at
the confluence of the interlobulars with arcuates.
These structures characteristically contain venous
sinuses (which connect with interlobular and medul-
lary veins) in the connective tissue matrix. These
structures are often interlaced with smooth muscle.
Their appearance is said to resemble erectile tissue
(162). In the dog, they are less extensive and lie
primarily close to the arcuate-interlobar junction.
Smooth muscle in the cushions of this species is very
inconspicuous or absent.
It is worthy of emphasis that the aforementioned
structures are not valves in the sense of those found in
systemic veins, although many of those pictured by
von Kiiglegen et al. (320, 321) exhibit a cusplike
organization. In any event, their designation as
"effluent constrictors" at present best describes their
function, although the functional significance is hard
to assess. The relatively high pressure found in the
arcuate veins of dogs (24 mm Hg) by Swan el al.
(302) appeared to give functional evidence of a point
of increased resistance at the arcuate-interlobar
junction. When the catheter was withdrawn into the
interlobar vein, pressure decreased immediately to 7
mm Hg. Brun et al. (41) found wedged catheter
pressures averaging 18 mm Hg in the human kidney;
the pressure in the renal vein averaged 5.6 mm Hg.
It was concluded that the wedged renal vein pressure
equalled arcuate venous pressure and hence very near
to the pressure in the peritubular capillaries and
interlobular veins. According to Koester et al. (162),
the effluent constrictors keep the kidney "functionally
distended" with fluids; they state also that these
structures cause smaller vessels of the vascular system
(venous channels?) to widen in bore, reducing
resistance to blood flow. The logic of this can be
doubted since this at best would only compensate for
the initial resistance imposed. Nor is this supported by
physiological studies in which venous pressure has
been experimentally elevated (119, 122, 123, 233,
273, 281, 32g), under which circumstance over-all
renal resistance in fact increases, possibly by a
"venous-arteriolar" reflex.
Glomerular Circulation
The studies of Boyer (29), Elias et al. (82), Hall
(125, 126), Johnston (154), and Kurlz & McManus
( 1 70) show that the glomerular capillaries are not
simple loops, but form a freely branching, anastomotic
network (fig. 6). More specifically, larger through
channels exist with an associated capillary network
of smaller anastomotic channels. Hall has suggested
that this may afford a structural basis for the skim-
ming of plasma relatively freed of cells into the net-
work of small capillaries, while the greater mass of
blood cells directly and rapidly flows through the
lobule to the efferent arterioles as an axial stream.
fig. 6. Glomerular capillary supply, showing anastomotic
connections. [After Elias (82).]
1462
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 7. Electron microscope reconstruction
of the glomerular filtering membranes. [After
Hall (125).]
INTRACAPILLARY SPACE
(LUMEN)
LAMINA FENESTRATA
(LINING NETWORK).
PODOCYTE
(COVERING CELL)
EXTRACAPILLARY SPACE
(CAPSULAR SPACE)
This may facilitate the filtration process by slowing
flow and reducing turbulence.
Figure 7 shows a sectional diagram of the struc-
tures forming the filtration apparatus as developed
by electron microscopy (125). The pores in the capil-
lary endothelium (.05 fi thick) (lamina fenestrata)
are too large (0.1 /x) to restrain the plasma constitu-
ents. Rather, they expose the ultrafiltration mem-
brane, the lamina densa, to the free flow of plasma
by removing the endothelial cytoplasmic barrier.
Although the lamina densa (glomerular basement
membrane), 0.1 n thickness, exhibits differences in
stratification (244), it is probable that it is a homoge-
nous layer; pores that have been noted are probably
artifactual. It appears to be the limiting membrane
for restraint of plasma proteins and cells.
The podocytes (foot cells) of the visceral layer of
Bowman's capsule rest on the lamina densa with
thousands of foot processes (pedicels). Hall has sug-
gested that they may play an important part in the
regulation of filtration. The space between the pedicels
may be narrow enough ( 1 00 A) to be a limiting dimen-
sion in restriction of plasma proteins ("slit pore").
Hall suggested that the foot processes may be nar-
rowed or widened thereby exposing a greater or lesser
area of basement membrane, although a mechanism
by which such changes could be brought about has
not so far been proposed. However, it is conceivable
that changes in caliber of the capillaries as a function
of internal pressure (vis a tergo) may alter the spacing
of the pedicels. Using as a basis the observations on
the frog glomerulus, Elias et al. (82) described another
possible method of regulation. They observed that
the position of the glomerular blood channels is not
constant and undergoes changes (e.g., transverse
displacement) in relation to the foot processes. Thus,
a group of pedicels may be active while a blood
channel is located under them, and later at rest
(when that blood channel has shifted to a new loca-
tion).
The permeability of the filtering membrane of the
kidney has been repeatedly studied by determining
the renal plasma clearance of molecules of varying
sizes. Wallenius (326), for example, by fractional
hydrolysis of dextran, produced and separated sub-
stances with a wide range of molecular sizes and
shapes and examined the facility with which they
passed into the urine (fig. 8). He calculated that the
pore radius in the dog glomerular membrane may
range from 18 A to 50 A. These findings are in accord
with the anatomical evidence. The findings of Gie-
bisch et al. (100) are in essential agreement. The
ratio of dextran clearance to circulation clearance
fell markedly at a molecular weight of ca. 50,000.
Juxtaglomerular Complex
Two structural entities at the vascular pole of the
glomerulus, the juxtaglomerular apparatus and
macula densa, have been thought to be related in
some way to the control of blood pressure or salt
balance and thus to be concerned with renal hyper-
tension (310). One of these, the juxtaglomerular
apparatus (JGA), is a thickening of the media of the
afferent glomerular arterioles (polkissen) (fig. 9).
The cells of the JGA become swollen, afibrillar in
THE RENAL CIRCULATION
1463
I0(
>
L_
60
\
±
\
0
W
40
5?
0 \
0
O
■\ 8
0
n
MOLECULAR, WEIGHT x l|o-3
1
0 2
0 30 40 50
fig. 8. The relationship of the molecular
weight of dextran to percentage filtered. [After
VVallenius C326).]
fig. 9. The juxtaglomerular
complex of the kidney. JGA:
juxtaglomerular apparatus (Pol-
kissen) ; MD : macula densa of
distal convoluted tubule. [Cour-
tesy of B. S. Garber (unpub-
lished).]
appearance, and contain granules (periodic acid-
Schiff reaction) which vary in amount in various
forms of experimental hypertension and with varia-
tions in sodium intake. Pathological states which
produce renal ischemia, such as the crush syndrome,
cause similar changes and are accompanied by in-
crease in blood pressure (no). A role in the regulation
of autonomy has been invoked for this structure (121,
274, 281, 308, 330). In general, when the kidney is
exposed to hypertensive blood pressures, the granu-
larity decreases; if the blood pressure is decreased, the
granularity tends to increase. Tobian (310) feels that
these cells act as "stretch receptors," changing their
rate of secretion inversely with degree of stretch of the
walls of the arterioles.
The changes in granularity in the JGA cells are
1464
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
related to extractable renin or its precursor. It is now
conceived that renin is an inhibitor of an antihyper-
tensive function of the kidney ( 163). An explanation of
hypertension caused by renal ischemia would be as
follows: if the renal artery is constricted, the JGA
cells, being stretched less, would increase the secretion
of renin. This would then inhibit the cells subserving
the antihypertension function of the kidney, and
systemic blood pressure would rise.
The cells are also influenced by electrolytes. A diet
low in sodium increases granulation in dogs, cats, and
rats (both intact and hypophysectomized). Increase
in salt intake decreases granularity (310). Although a
complex interrelationship with the adrenal cortex
is probable, a simpler explanation offered by Tobian
is that a low salt intake favors a decrease in blood
volume, reduced blood pressure, and decreased
stretch of the JGA cells (increased secretion). A high
salt diet would have the opposite effect. On the basis
of the above scheme, a decreased sodium diet should
ultimately result in increased blood pressure, based
upon the increased granularity of the JGA. This may
indeed be a compensatory mechanism to maintain
blood pressure in the face of lowered blood pressure
resulting from decreased plasma and extracellular
volumes caused by low sodium intake.
The other structure of importance is a portion of
the distal convoluted tubule near the vascular pole,
the macula densa. McManus (194-196) has suggested
that the JGA, macula densa, and associated struc-
tures be together called the juxtaglomerular "com-
plex." This epithelial plaque appears to have a
reversed polarity from the rest of the tubule, in the
sense that the Golgi apparatus is between the nucleus
and the attached pole of the cell, (contiguous to the
vasculature), rather than between the nucleus and the
lumen. The suggestion made by McManus (196) and
supported by Garber el al. (98) was that these cells
abstracted from the contents of the lumen and trans-
mitted this material to the cells of the arterioles. It is
relevant to point out that a site of active sodium re-
absorption is found in the vicinity of the distal con-
voluted tubule. It has been suggested that the JGA
and macula densa form a regulatory system capable
of responding to osmotic pressure changes (and
possibly hydrostatic pressures), in turn modifying
glomerular filtration in a self-regulatory manner.
This interesting hypothesis needs experimental
verification, particularly in view of the contention
by de la Pefia & de Castro (70) that structures re-
sembling the macula densa were found in apposition
to efferent arterioles in the human kidney. It is worthy
of note that afibrillar cells containing granules, similar
to those in the afferent arterioles, have been noted in
efferent arterioles (194).
Blood Supply to the Medullary Zones
Edwards (79) has described two types of efferent
arterioles which exist in the juxtamedullary zone of
the human kidney. One out of four to five glomeruli
has a "corticomedullary" efferent arteriole to capil-
laries of the juxtamedullary parenchyma (fig. 10).
The others (about 180,000 per human kidney)
have long, descending arterioles (arteriolae rectae
spuriae). These go on to the capillaries. One type
forms networks around the tubules, the other goes
on to the vasa recta system (fig. 1 1). The venae rectae
return to the arcuate veins.
Note in table 1 the greater total muscle volumes
in the medullary efferent arterioles as compared with
the cortical. The total volume of muscle in the wall
of the afferent and efferent arterioles was 0.124 ml
and in the medulla 0.169 ml (79). Christensen (54)
found the diameters of the juxtamedullary vasa
efferentia of the dog kidney about the same as those
in the cortical vasa efferentia, contrary to the findings
of Trueta et al. (311) who state that the caliber of the
juxtamedullary efferent arterioles greatly exceeds the
Interlobular ... s.
\
k
Corhcal
Eff. arteriole
Corhco- medullary
Eff. arteriole
^8i
s
5>
Medullary
Eff. arteriole
fig. 10. The blood supply of the juxtamedullary zone. [After
Edwards (79).]
THE RENAL CIRCULATION
[465
Capsule
Cortex
Outet zone
of medulla
Inner zone
of medulla
fig. 11. The vasa recta system of the kidney. (Courtesy of
A. A. Maximov & VV. Bloom, Textbook of Histology. Philadel-
phia: Saunders, 1957.)
cortical efferents. The impression is definitely gained
that the arterioles supplying the medullary zone are
not low-resistance vessels as originally suggested by
Trueta, but appear to be sites that could offer con-
siderable resistance to blood flow, thus resulting in a
significant drop in pressure gradient. One would
anticipate on this basis that hydrostatic pressure in
the vasa recta system would be very low were it not
for the relatively high venous pressure found in the
arcuate veins. Gottschalk & Mylle (112) and Wirz
(347), by direct puncture of cortical peritubular
capillaries in the rat, found an average of ca. 1 6 mm
Hg (range: 14.0 to 20.0 mm Hg), which is well below
the oncotic pressure of the plasma protein when one
considers that glomerular filtration concentrates the
protein. If this applies to the vasa recta, it would be
favorable for optimal operation of the countercurrent
system. The water abstracted from the collecting
ducts moves into the vasa recta because of the gradient
of chemical potential established by the colloid os-
motic pressure of the plasma proteins.
Direct connections from the arcuate arteries into
the medullary zone (arteriolae rectae verae) have
been found in dog and man (18, 54, 217, 221, 311)
but appear to be rare. Likewise, Ludwig's arterioles,
branches from the afferent arterioles in the cortico-
medullary zone passing directly into the medullary
peritubular capillaries, are very infrequent in the
dog, cat, and man. They are very rare (295) in the
rat kidney. Oliver (236) and More & Duff (217) did
not find them in normal human kidneys.
Renal Lymphatic System
A greater lymphatic system (cortical) exists, and
there is a lesser (medullary) system which follows the
table I . Averaged Measurements in Microns of Cortical and Medullary Arterioles per Kidney
and the Total Volume in cm3 of the Muscle Composing Their Walls1
Arteriole
Afferent
Efferent
Total Xo.
I ,000,000
820,000
Length per
Arteriole
277
200
Total Length
Cortex
277,000,000
164,000,000
External Diam. Luminal Diam.
26
16
13
12
Total Muscle
Vol. in cm3
O.II
O.OI4
Medulla
Efferent 180,000 600 108,000,000 33 18 0.065
1st brs.-3 540,000 400 216,000,000 24 14 0.064
2nd brs.-6 1,080,000 300 324,000,000 15 9 0.04
1 Specimen calculation such as was used to obtain the muscle volume given in the last column of the above table.
Afferent arteriole: w X 132 X 277,000,000 = 147,087,000,000 ^i3 = total volume
7r X 6.52 X 277,000,000 = 37,1 18,000,000 m3 = lumen volume
Total volume less lumen volume = 109,989,000,000 m3 = 0.11 cm3 [After Edwards (79).]
1466
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
course of the vasa recta (251). They begin blindly
in two locations: closely adjacent to the capsules of
cortical glomeruli and beneath the mucosa of the
papilla (fig. 12). The cortical lymphatic capillary
networks do not have a functional relationship to
the glomeruli. There is apparently no entry (248).
The lymphatics from the cortex with an inter-
lobular course drain toward the arcuate vessels.
Those from the medulla, draining the vasa recta, join
the cortical branches at the arcuate level, then pass
out with the interlobar vessels toward the renal pelvis.
After converging at the hilum of the kidney, they
course as perivascular channels to the cysterna chyli
and thoracic duct (230, 261). Valves appear to be
lacking in the lymphatics of the renal parenchyma
but are present in the large trunks of the renal sinus.
Figures for the volume of lymph produced by the
fig. 12. The black threadlike lines indicate the greater and
lesser lymphatic systems of the human kidney. The arrows show
the probable direction oflymph flow;a: capsule; b: interlobular
vein; c: interlobular artery; d: glomerulus; e: arcuate artery;/:
arcuate vein, »: interlobar artery; h: interlobar vein; i: papilla.
[After Rawson (251).]
kidney are scarce. Single capsular lymphatics of the
dog yield flows of ca. 1 ml per hour (176, 177, 304).
An estimate from the data of Schmidt & Hayman
(266) yields a total of ca. 7 ml per hour per kidney.
Available anatomical evidence indicates that the
capsular lymphatics join with the cortical and medul-
lary lymphatics. Lymph flow increases when the
kidney is subjected to osmotic diuretics (106, 230,
266). It is enhanced by ureteral obstruction. Pyelolym-
phatic backflow is evidenced by the fact that dye has
been shown to move from the pelvis into the lymphat-
ics with increased intrapelvic pressure. Elevation of
renal venous pressure in the dog by 1 4 to 35 cm results
in approximately 3-fold to 5-fold increase in lymph
flow (123, 177). Marked increase in lymphatic pres-
sure accompanies venous obstruction (157). Elevation
of arterial pressure does not markedly increase lymph
flow from hilar vessels: 0.023 ml per min at 58 mm
Hg to 0.039 ml Per mm at '57 mrn Hg (124).
It is of considerable interest that the lymph is high
in sodium, chloride, and urea content compared to
plasma and thoracic duct lymph (176, 304). LeBrie
& Mayerson (176) have found Na and CI concentra-
tions of 162 and 140 per liter, respectively, compared
to 145.7 anc' 1 10.5 in the plasma, and 145.6 and 1 2 1 .3
in the thoracic duct lymph. Interestingly, the K
content does not differ significantly. These findings
support the countercurrent hypothesis, for it is to be
expected that these concentrations will be elevated
as a result of the contribution of the medullary
lymphatics which drain the papillary zone of hyper-
osmolarity of the kidney. K. is not a significant con-
tributor to this hyperosmolarity (267). This is further
supported by the low glucose content of this fluid
relative to plasma (304), suggesting an important
source beyond the proximal convoluted tubules.
An interesting avenue of investigation of the counter-
current mechanism thus appears to be afforded by a
study of the renal lymphatics.
The renal lvmph protein concentration averages
2.9 g per 100 ml as compared to 5.83 g per 100 ml for
the plasma proteins (177). Evidently the renal lym-
phatics subserve an important function for operation
of the countercurrent system by draining off excessive
protein filtered by the vasa recta, which might other-
wise accumulate in the interstitial spaces of the me-
dulla. Removal of such protein would act to maintain a
more favorable gradient of movement of interstitial
fluid into the vasa recta, attracted by the relatively
higher oncotic pressure.
THE RENAL CIRCULATION
I467
fig. 13. Renal plexuses of the human kidney, anterior aspects. 1 : Hypogastric nerves. 2: Middle
spermatic and ureteric nerve. 3 : R. spermatic artery. 4 : Renal branch from sup. hypogastric plexus.
5: Lumbar splanchnic nerve. 6: Sup. ureteric nerve. 7: Communication between renal plexus and
spermatic nerve. 8: Small renal ganglion. 9: Renal branch from lumbar sympathetic trunk. 10: Post,
renal ganglion. 1 1 : Aorticorenal ganglion. 12 : Communication between suprarenal and renal plexuses.
1 3 Right coeliac ganglion. 14: R. phrenic nerve. 15: Post, vagal trunk and coeliac di v. i6:Ant. vagal
trunks. 17: Esophagus. 18: L. phrenic nerve. 19: Greater (thoracic) splanchnic nerve. 20: Lesser
(thoracic) splanchnic nerve. 21 : Lowest (thoracic) splanchnic nerve. 22: Sup. mesenteric ganglion.
23: Post, renal ganglion. 24: Intermesenteric nerves. 25: Renal branches from lower ends of inter-
mesenteric nerves. 26 : Lumbar sympathetic trunk. 27 : L. ureter. 28 : Inf. mesenteric plexus. 29 : Sup.
hypogastric plexus. [After Mitchell (211).]
NERVE SUPPLY TO THE KIDNEY: ANATOMICAL ASPECTS
Extrinsic Nerves
It is generally agreed that the major nerve supply
to the kidney has its origin largely from the twelfth
thoracic to the second lumbar ganglia of the sym-
pathetic nervous system in man (51), and in the dog
from T4 through L2, but most abundantly from
T10 through T12 (31). Relative to its size, the kidney
receives a more profuse and widespread supply than
almost any other viscus. Mitchell (211) has written
comprehensively on the anatomical aspects of the
nerve supply to the human kidney, with an extensive
historical review, and with good anatomical illustra-
tions to which the reader is referred. Christensen
el a/. (55) have given a detailed description of the
innervation of the cat kidney. The renal nerves of
the human kidney are derived from the following:
celiac plexus. The renal branches arise from the
celiac or aorticorenal ganglia, and contain sympa-
thetic and almost certainly parasympathetic fibers.
Most investigators feel that the posterior vagal trunk
supplies the kidney via the celiac plexus, although
several have indicated that it may pass directly to the
renal plexus (fig. 13).
thoracic splanchnic nerve. The greater (superior
thoracic) splanchnic nerve occasionally, and the
lesser (middle thoracic) splanchnic nerve almost
invariably send direct filaments to the aorticorenal
ganglion or renal plexus, while the least (inferior
thoracic) splanchnic nerve ends in the renal plexus.
1468
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
lumbar splanchnic NER\ES. Direct branches to the
renal plexus may arise both from the first and second
lumbar ganglia, or from the adjacent portions of the
sympathetic trunk. These branches are inconsistent,
especially the ones from the second lumbar ganglion.
When present, they join the posterior root of the
renal plexus often close to the terminations of the
lowest splanchnic nerve, and sometimes they end in
the posterior renal ganglia.
intermesenteric nerves. The renal branches from
the upper parts of the intermesenteric nerves run
almost directly to the renal hilum. The uppermost
ones are associated with those arising from the celiac
plexus and occasionally from the superior mesenteric
ganglion, and those a little lower down are connected
with the origins of the superior spermatic nerves.
Other branches originate from the lower ends of
the intermesenteric nerves or from the superior
hypogastric plexus (presacral nerve). Except near
their terminations they are separate from the other
renal nerves; they communicate with the superior
and middle spermatic nerves, and are apparently
distributed mainly to the renal pelvis and upper
ureter. Mitchell suggests that the ureter, renal pelvis,
and renal collecting tubules may receive their para-
sympathetic supply via these nerves through the
caudal (pelvic splanchnic) rather than cranial (vagal)
outflow.
The various renal nerves unite in a plexiform
manner around the renal artery. Xo filaments of any
size lie anterior to the vein or posterior to the pelvis,
and the plexus splits into subsidiary plexuses which
accompany the branches of the renal artery into the
kidney. A few filaments accompany the superior and
inferior renal capsular veins. All nerves of the plexus
do not cluster intimately around the renal artery, but
several approach only the branches of this vessel in
the actual hilum of the kidney. Both preganglionic
and postganglionic nerve fibers exist in the renal
plexus. The renal nerves and plexus form multiple
intercommunications with many other autonomic
nerves and plexuses.
Ganglia of varying size are invariably located in
the plexus, and the posterior renal ganglion is the
largest and most constant. Many of the ganglia are
of microscopic size; these are more numerous in
infantile kidney specimens, although not absent in
adult kidneys.
Intrinsic Innervation
Mitchell (212) has emphasized the difficulties
involved in the differential staining of intrinsic nerve
fibers; reticular fibers are especially troublesome and
caution must be exercised in properly separating
them from the nerves. Errors of interpretation have
resulted from improper identification. The following
description is largely from Mitchell and based upon
the innervation in man.
The main renal plexus divides into large bundles
which accompany the branches of the renal artery
into the kidney, giving off interlobar, arcuate, and
interlobular nerves corresponding to the divisions of
the artery. They may lie adjacent to the arteries or be
imbedded in the adventitia; some spread out in the
adventitia and media, but others leave the vessels to
run between the tubules. The nerve fibers in the
adventitia do not all end in the artery, but may
re-emerge into the perivascular space. Definite nerve
filaments or endings were not found in the intima of
any of the renal vessels, and no fine nerve plexuses
were detected around the peritubular capillaries. The
fibers are unmyelinated, according to Mitchell.
nerve supply to cortex. Nerve fibers are much
more common in the cortex than in the medulla, and
probably more frequent among the convoluted
tubules. They are derived from the small bundles of
nerve fibers associated with the interlobular arteries.
At irregular intervals strands detach themselves from
the parent bundles and pass between the tubules
where they are connected by occasional anastomoses,
and filaments may be traced to the limits of the
cortex.
Many fibers appear to end as free, fine, beaded
filaments on the basement membranes or between the
tubular cells. Others give off short side branches which
end in globular or fusiform swellings. Endings have
been seen on the basement membranes and between
the cells, but the presence of intracellular endings is
doubtful. This aspect has been controversial: some
workers believe that nerve fibers may end within the
tubular cells (133, 161).
The glomeruli receive offshoots from the inter-
lobular nerves and other filaments which are derived
from adjacent interlobular nerve bundles. These
offshoots run alongside afferent arterioles, and wind
around them in a spiral fashion, supplying these
vessels. Knoche (161) has demonstrated a terminal
reticulum on the specialized cells of the juxtaglomeru-
lar apparatus (polkissen cell). Filaments proceed to
the glomerulus, and Mitchell contends that the
resultant strands fade away on the capsule or may
end in a small series of slight varicosities. Dark
staining strands (Romane's stain) have been seen
within the glomeruli, but identification as nerves was
THE RENAL CIRCULATION 1 469
not absolutely positive. Knoche, on the other hand
(using Bielschowsky-Gros stain), described a promi-
nent terminal reticulum surrounding the glomerular
capillaries. Since there are presumably no contractile
elements of the smooth muscle type in the glomerular
zone, the role of these nerve fibers becomes highly
problematical. He also found nerve filaments around
the efferent arterioles. Harman & Davies (133) saw
nerve endings in the perivascular tissue of the glomeru-
lus, and hypothesized an afferent function for them.
Knoche also described a prominent reticulum
surrounding the complex made up of the macula
densa, the polkissen, Goormaghtigh's "cell aggre-
gate" and paravascular and paraportal cell clumps,
and suggested that together they formed a receptor-
effector zone to adjust glomerular filtration to the
variations in blood pressure.
medullary nerve supply. Comparatively few nerve
fibers pass into the medulla as compared with the
cortex, and most of these are located in the boundary
zones. They reach the medulla largely through
offshoots from the nerve bundles alongside the arcuate
arteries, and by accompanying the arteriolae rectae
spuriae. A point of control of the vasa recta system
could reside here.
renal afferents. Beyond the possibility of pres-
soreceptors in the field of the juxtaglomerular cell
groups, the renal tissue is generally nonsensitive to
afferent stimulation. Swelling gives renal pain via
stretch of the capsule, which has pain afferents. It is
said that some afferent sensory function is localized in
the papillary tissue and in the pelvis.
distribution of osmotic constitutents in the
kidney; the countercurrent hypothesis
The osmotic constituents of the kidney are arranged
so that they are isotonic with blood in the cortex,
then rise to three to four times this concentration at
the tip of the papilla. VVirz et al. (345) have demon-
strated by a cryoscopic method (disappearance of ice
crystals as observed by a polarizing microscope) that
points of equal osmotic pressure form shells concentric
to the tip of the papilla (fig. 14), and are parallel to
the interzonal boundary. The important osmotic
constituents are sodium, chloride, and urea, as
revealed by the analysis of Ullrich & Jarausch (312)
(figs. 15 and 16), and supported by the findings of
Schmidt-Nielsen & OTJell (267).
A uniform distribution of osmotic constituents
max. 700
IsotoniscfiO
fkj. 14. Distribution of osmotic constituents in the kidney
(hamster). A.Z.: outer zone of medulla: I.Z.: inner zone of
medulla. [After Wirz et al. (345).]
between the loops of Henle, vasa recta, and collecting
tubules has been proved by micropuncture (346,
348) in the hamster and rat (fig. 17) [Gottschalk &
Mylle (1 13)-] The "hairpin" loop arrangement of the
loops of Henle and vasa recta has been the basis for
the formulation of the countercurrent multiplier
system concept for urinary concentration and dilution
(132). (For discussion of this principle as applied to
kidney function, see 131, 169, 171, 259, 289, 314.)
Briefly, the principle of the countercurrent system
as it applies to the medullary loop of Henle system is
as follows: sodium, by an active process, and chloride,
as the result of an electrochemical gradient thus
established, are believed to be transported out of the
relatively water-impermeable ascending limb of the
loop of Henle into the interstitium of the medulla
until a gradient of ca. 200 mOsm per kg. H20 has
been established (fig. 17). This single effect is multi-
plied as the fluid of the thin part of the descending
limb comes into osmotic equilibrium with the inter-
stitial fluid by diffusion of water out (and probably
by the diffusion of some NaCl into the descending
limb), thus raising the osmolarity of the fluid rounding
the hairpin loop into the ascending limb. The in-
'47°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
100-
jrmol/ml
Jrin
ostnol
Hz-Konz
^imol/mJ.
2, IS
13
1,66
es
1.73
31
1,66
90
20
13
32
umol/ml
too -
Urin
osmol ZVKonz.
^mol/ml
62
112
356
212
fig. 15. Na and CI concentration in kidney tissue (Mmole/ml of tissue fluid,) in hydropenic (solid
lines) and diuretic idashed lines) dogs. To the right are the osmotic pressures, and Na and CI concen-
trations of bladder urine taken shortly before the tissue analysis. [(After Ullrich & Jarausch (312).]
125.5
106
creased concentration here, also raising that in the
interstitium now favors further movement of fluid out
of the descending limb, further increasing concen-
tration, and so on.
In this fashion, an increasing osmotic gradient is
established in the direction of the tip of the papilla,
and yet at no level is there much osmotic difference
among the luminal fluid, interstitial fluid, and blood.
The collecting ducts in the presence of pituitary
antidiuretic hormone (ADH) are believed to be water
permeable and somewhat Na-impermeable (net
transport small, although there may be diffusion
into and active transport out). This results in diffusion
of water out of the collecting ducts into the hyper-
osmotic medullary interstitium, and ultimately into
the vasa recta to be carried away until fluid in the
collecting ducts becomes correspondingly concen-
trated. The role of the vasa recta will be dealt with in
greater detail below.
The view is currently favored that ADH
•
Q
•
•
0
J
;
0
•
&-
I
M
J
I
•
•
k
•
<§>
-ami-
nohippurate (PAH) (287, 288). Then CD or CPAh is
nearly equivalent to plasma flow. The fact that ex-
traction is not complete is interpreted as indicating
that a small fraction of blood does not perfuse ex-
cretory tissue: this would include capsule and inert
supportive tissue, medullary tissue (loops of Henle,
collecting ducts), calycine mucosa, and pelvis. On
this basis, Smith has referred to this as the "effective"
plasma or blood flow.
A considerable amount of study has been made of
the extraction ratios of Diodrast and PAH. A repre-
sentative group of findings is shown in table 5.
Although ED and £pAH seem comparable in the
dog, £PAH is more efficient in man than ED . This was
particularly the experience of Bergstrom et al. (16)
who made simultaneous comparisons (EPAB = 0.90;
Ely = 0.74). Differences in kinetics of erythrocyte to
plasma shift for PAH ma\ lie involved, and this is
probably less important for man than the dog. A
factor to be considered is that the animal work has
done largely under anesthesia, while the human sub-
jects were unanesthetized.
The Fick principle can be employed with any sub-
stance cleared by the kidney which shows a measura-
ble A-V difference. Obviously, the smaller the A-V
difference, the more prone to error the calculation
will be. Thus, phenol red, urea, mannitol, and inulin
have been employed, but have considerably smaller
A-V differences than Diodrast and PAH.
the nitrous oxide method. This is an adaptation of
the method employed for the measurement of cerebral
blood flow and involves inhalation by the subject of
nitrous oxide, and uptake from the blood by the
kidney. The Fick principle is employed (58, 68).
N20
100V, ,-.
f'(A*
VJdt
RBFS„0 is the blood flow per 100 g kidney tissue
per minute as measured by N20 uptake; VcV-S is
the kidney uptake of N20 per g tissue during the time
from O to /' (time of blood-tissue equilibrium); and
Ac and Vc are arterial and renal venous concentra-
tions, respectively, which finally become equal at
time I'. S is the partition coefficient between blood
and tissue (assumed to be unity in this instance).
1 III-: RENAL CIRCULATION
I481
The method yields an average of 3.2 ml per min
per g kidney weight in anesthetized dogs (58) and in
man (68). Comparison of this method with a direct
method ( bubble flowmeter) in the do? under various
physiological conditions shows that the two yield
flow values which are not significantly different (58).
An obvious advantage is that the nitrous oxide method
can be employed during conditions of anuria. A
similar application using radioactive krypton (Kr85)
has been employed during anuria (40).
Critique of the Clearance .Method
CRITERIA FOR APPLICATION OF CLEARANCE. Some of the
criteria which must be met in order for the clearance
of a substance such as Diodrast or PAH to measure
accurately renal plasma flow are: a) change of volume
of blood in passage is negligible (i.e., urine and lymph
flow not excessive), b) concentration of substance in
blood is constant, or the rate of change of concentra-
tion is uniform for midpoint collection, <) rate of
urine flow should be sufficiently large and constant
so that it may be representative of the urine in the
nephrons, d) the substance should not be formed or
altered in the kidney, e) all blood in the renal vein
should pass through the kidney (and not enter via
shunts).
FACTORS WHICH MIGHT INVALIDATE THE CLEARANCE
method, a) Oliguria or marked fluctuations in urine
flow such as might accompany rapid changes in blood
pressure. If there is stagnation, or rapid fluctuation
of the urine flow in the nephrons, the collected sample
will not reflect the true excretion, and the midpoint
plasma sample will lack validity. b) Rapid changes in
plasma concentration, preventing establishment of
equilibrium among blood, interstitial fluid, tubular
cells, and tubular urine, c) Renal storage of substance
in tubular cells or interstitial fluid, d) Storage of sub-
stance in the erythrocytes in excess of the plasma
concentration, so that its simple outward diffusion
through the plasma adds appreciably to the amount
actually carried by the plasma leading to an errone-
ous plasma flow figure, e) Impairment of the PAH
tubular transfer mechanism, leading to an erroneously
low plasma flow figure.
ADEQUACY OF Cd AND CPAH AS MEASUREMENT OF RENAL
plasma flow. Under stabilized conditions that fulfill
the criteria explained previously, good correspond-
ence of clearance to direct methods is obtained.
Selkurt (269) found that BFPAH averaged 91 per cent
of the simultaneously measured direct blood flow
(venous outflow method). The difference was at-
tributable to incomplete extraction of PAH. Conn
& Markley (57) compared renal blood flow in anes-
thetized dogs as measured indirectly by the Fick
principle (PAH clearance) to blood flow measured
directly by bubble flowmeter. The ratio of indirect
to direct values averaged 1.025. Employment of the
Fick principle corrects for incomplete extraction and
yields total blood flow. Schwalb et al. (268) made a
similar comparison and found a ratio ot 1.06 ±
0.17. But after the kidney was poisoned with Na
fluoride, the agreement did not hold. Then flow
measured by the bubble flowmeter was often much
higher than that measured by PAH clearance. Since
the use of the Fick method should correct for incom-
plete extraction due to impairment of the PAH secre-
tory mechanism, the authors believed that PAH was
stored in the kidney (possibly in the tubular cells)
so that excretion (I'V) was low, relative to the ap-
parent removal.
Reubi et al. (255) compared simultaneous Fick
plasma flows for PAH, mannitol, endogenous creati-
nine, and thiosulfate. For example, the ratio between
Cpah ;'£pAH and CM/EM varied between 1 .54 and
0.645. Disparities were further exaggerated by in-
jection of epinephrine and histamine causing rapid
transients in blood pressure and urine flow. Suggested
causes for the discrepancies were: differences in the
extraction and blood flow in separate kidneys; intra-
renal extraction; conjugation or breakdown of PAH,
mannitol, creatinine, and thiosulfate; removal of part
of the substances from the kidney through lymphatic
vessels, thus bypassing the renal vein; changes in
the permeability of the red cells to the test substances;
or, finally analytical difficulties. Balint & Fekete (8)
found great disparities between direct blood flow and
the Fick method (C'pah, £pah) in hemorrhagic hypo-
tension, hemorrhagic shock, and shock from pyloric
obstruction in dogs. The indirect method was always
lower by varying degrees than the direct method.
Since errors are compounded by the analysis of
Reubi et al., it would be more desirable to compare the
indirect methods against a direct method in tests for
fidelity under experimental conditions. Under cir-
cumstances of rapidly changing blood flow resulting
from nerve stimulation or action of vasoactive drugs,
as has been suggested, the clearance method may not
accurately follow direct flow. Study & Shipley (298)
found excellent agreement between the Fick method
(Diodrast) and direct flow (rotameter in renal vein)
during control conditions. During stimulation of the
[ (,'lj
HANDBOOK OP II I YSIOI.OGY
CIRCULATION II
renal nerves, resulting in a 53 per cent reduction in
direct flow, the calculated RBF was from 1 to 70
per cent of the true values because of reduction in
mine How and incomplete excretion (UV). All calcu-
lated flows exceeded the direct flows on cessation of
stimulation as stored urine was washed out. They
emphasized the need to correct for possible shifts of
Diodrast (or PAH) from erythrocytes to plasma during
the venous sampling. To the extent that this occurs,
the Eu or £PAH will be vitiated, and the Fick applica-
tion inaccurate. Phillips et al. (247) have given meth-
ods for correction of PAH shift. Whole blood extrac-
tion eliminates errors incurred by the shift from cells
to plasma, and Bergstrom et al. (16) have found
the use of radioactive Diodrast (containing I131)
particularly helpful in this respect. The possibility has
been examined that opening of vascular shunts not
perfusing excretory tissue might occur following nerve
stimulation or drug action and invalidate the clear-
ances. Epinephrine (0.1 /xg) in rabbits caused the
£PAH to fall to negative values in seven of nine cases
[average for the seven, — 26.6 c1 (214)]. This was
restored in 10 to 40 min. The negative values have
been explained by a return to the venous outflow
of stored PAH (interstitial fluid of papillary zone?).
Injections of Thorotrast in these pictured the possi-
bility of juxtamedullary shunting of blood. Ephedrine
produced a similar picture in cats [India ink injection
(189)], but Lofgren points out that the picture of
cortical ischemia and medullary filling could result
from congestion of the vasa recta following contrac-
tion of venous effluent constrictors, rather than from
opening of a bypass and increased flow. Mover et al.
(222) employed sciatic stimulation and epinephrine
in dogs and rabbits. With sciatic stimulation, blood
flow decreased ca. 36 per cent, but renal venous blood
never became arterialized, as the original Trueta
shunt operation would demand. In fact, the A-V oxy-
gen difference actually increased. India ink distributed
fairly equally throughout cortex and medulla after
nerve stimulation. The rabbit kidney after epineph-
rine, however, appeared to confirm the appearance
of cortical ischemia and subcortical injection. But the
latter does not necessarily mean increased medullary
flow. Epinephrine and histamine caused a maximum
decrease of E,,XH of only 1 1 .4 per cent in the human
kidney (254).
In an interesting experiment Cargill (48) infused
human serum albumin into patients. EPAH invariably
decreased significantly, even as CPAH increased.
C\n rose proportionally to (.',. ul , so that the filtered
fraction remained constant. These results could readily
be explained by increased shunting of blood through
the medullary vasa recta system.
EXTRACTION RATIO AS A TEST OF VALIDITY OF THE
clearance method. The extraction ratio has been
one of the measurements which yields insight into
the efficacy of the tubular transfer process or the
adequacy of tubular vascular perfusion. It is reduced
during shunting of blood away from the tubular
secretory sites, or as the result of actual impairment
of the transport mechanism. Some of the physiological
and pathological conditions in which renal extraction
has been evaluated follow.
£pAH is not reduced by abdominal compression
which elevates control venous pressure from ca. 6 mm
Hg to 18 mm Hg (32). This lack of effect on £PAH
may be due to the probability that transmural renal
venous pressure would not be changed by this maneu-
ver (329). Werko et al. (334) found no change in
£PAH during the renal ischemia induced by tilting.
£PAH may be normal or only slightly impaired in
essential hypertension. A series examined by Reubi
& Schroeder (254) averaged 0.84, including one of
69.8. CargilPs (49) series of hypertensive patients
including those with nephrosclerosis averaged 0.79
(0.58-0.91). The lowered values are associated with
reduction of CHAH below 300 ml per min. In anemia,
there is only a slight decrease of the ratio (256, 305).
In nine observations on subjects with no renal pathol-
ogy but in congestive heart failure, Merrill (202)
found only two below 0.85 (0.64, 0.63); Edelman
et al. (78) reported an average of 0.90 (0.88-0.91)
in ten congestive heart failure subjects.
In nephritis there may be considerable reduction
in the extraction ratio. Bradley et al. (33) obtained
values for £PAH ranging from 0.58 to 0.76 in six sub-
jects with chronic glomerulonephritis. It may be
supposed that in the course of disorganization of the
renal vascular pattern, channels are established in
which blood flows from the artery to vein without
exposure to functional tubular tissue (abnormal shunts
or destroyed excretory tissue). Marked reduction in
£PAH (as low as 0.034 and 0.106) was noted with
tubular damage resulting from carbon tetrachloride
poisoning (284). £PAH decreased during acidemia
which developed during the apnea of diffusion respira-
tion in dogs (27). The control EPAH of 0.86 at pH
7.4 decreased to 0.53 at 7.05.
Renal ischemia and anoxic damage resulting from
hemorrhagic shock will impair extraction. Twenty
THE RENAL CIRCULATION
1483
minutes of renal ischemia in dogs resulted in a reduc-
tion in £pAH from 0.74 to 0.59 (269). Control flow
(Cpah) which gave 91 per cent of the simultaneous
direct blood flow measurement decreased to 30
per cent of the direct flow as a consequence of ische-
mia. Recovery occurred in 85 min. After 2 hours of
ischemia (246), £pAH (control, 0.90-0.94) was reduced
to 0.1 1 to 0.43.
Phillips et al. (247) found adequate extraction of
PAH until renal plasma flow was reduced below 7
ml per min during hemorrhagic hypotension and
then clearances no longer reflected plasma flow
accurately. Corcoran & Page (62) stated that CD did
not have value as a measure of plasma flow during
severe, prolonged hypotension, nor immediately
after restoration of blood pressure by transfusion.
Diodrast clearance fell progressively on repeated
hemorrhage and transfusion, until in some instances
negative extraction values were obtained (as low as
— 1 .59 compared to control of 0.757). C*n transfusion,
an "over-shooting" of clearances beyond the control
was observed during the early stages as a result ot
washing out of material accumulated in the interstitial
fluid and stagnant urine during hypotension. Selkurt
(270) compared CPAH with a direct blood flow method
in dogs during hemorrhagic hypotension and shock.
/;-Aminohippurate clearance virtually ceased during
hypotension (60-40 mm Hg) as direct flow fell to 1 1
per cent of control. On transfusion, although direct
flow was rapidly restored to near control, blood flow
calculated from CPAH averaged only 39 per cent of
direct flow, as the result of anoxic tubular impairment.
£PAH during hypotension was low and variable with
numerous negative extraction values (range, —0.750
to 0.543 during a 90 min period at 60 mm Hg, and
— 1.50-0.285 during 45 min at 40 mm Hg). After
transfusion, £VAH partially recovered, averaging
0.406 (0.03-0.69) compared to the control of 0.73.
Clearly the hypotensive anoxia had invalidated the
CpAH clearance as a measure of plasma flow probably
because of consequent tubular damage.
The negative extraction during hypotension has
been explained as the result either of back diffusion of
PAH from the lumina of damaged nephrons into
venous blood (270), or of absorption into the renal
venous blood of PAH accumulated during the period
of hypotension and impaired urinary excretion (62).
Again, this may be PAH concentrated in the vasa
recta and interstitial fluid in proximity (counter-
current mechanism), and will thus imply continued,
table 6. Clearance Data in Mammals and
Renal Blood Flow in Dog
Per g K\V
ml min
Ci„
CpAH
Per kg Body Wt
ml, m in
Cln Cpah
Per i.7) m-
BSA ml min
Cln CpAH
A. Clearance Data in Mammals*
Rat
o-75
2-75
6.00
22.0
7°
253
Rabbit
0.66
2.50
3.12
18.2
87
5' 2
Dog
0.62
1. 91
4.29
■3-5
146
460
Man
0.46
2 -33
'■97
10. 0
118
600
B. Renal Blond Flow in Dog\
RPP RBF
RPF
RBF
RPF
RBF
Unanesthetizedf
Anesthetized§
2.08
1.89
3.80
3-4°
■2-5
13.0
22.7
23-4
463
845
K.W = kidney weight, BSA = body surface area.
* [From Smith (287).] f (From Handbook of Circulation,
WADC Tech. Rpt. 59-593, '959) X 220 Observations;
direct venous outflow; urea, phenol red, and PAH extrac-
tion. §58 observations; (pentobarbital and chloralose) :
direct venous outflow, rotameter, and bubble flowmeter.
but reduced, perfusion of the medullary zone, with
cortical ischemia.
Renal Blood Flow Values
Data have been culled from two important sources
in the summary presented in table 6. It will be noted
that the dog appears to have the lowest CPAh per
gram kidney weight. The rat's value for CPAh is least
per 1 .73 m- body surface area, increasing progressively
in the series to the value in man. The dog has the
highest filtration rate (CIn) relative to the effective
plasma flow (C'pah), giving a filtration fraction (FF) of
0.32. In man this is 0.20.
In summary, as Smith has repreatedly stressed, the
clearance methods yield adequate information on
renal hemodynamics only under conditions of relative
stability of flow. They cannot accurately follow rapid
changes of blood flow, and changes in pathological
states (e.g., shock kidney) seriously handicap their
utility.
EXTRINSIC REGULATION OF RENAL BLOOD FLOW
A eurogenic Control
The thoracolumbar sympathetic supply is a rich
source of vasoconstrictor fibers for the kidneys. The
1484
HANDBOOK OF PHYSIOLOGY
ClkCILATION II
vagus apparently contains no vasomotor fibers to the
kidney, and no evidence exists for vasodilator fibers
in this circuit. Hence, the vasomotor status of the
kidney is maintained by variations in vasoconstrictor
tone.
the question of renal autonomy. Considerable
controversy has revolved around the question of
whether or not a continued flow of impulses passes
to the arterioles, or whether such regulation is absent
in the basal state, to be invoked only in emergency
states of heightened sympathetic nervous system
activity. Early investigators, working in anesthetized
animals, appeared to demonstrate a "denervation
hyperemia." In view of the fact that ample evidence
exists that morphine, ether, chloroform, urethane, and
pentobarbital anesthesia depress renal blood flow
to varying degrees, probably due to enhanced activity
of the sympathetic system and adrenal medulla, it is
understandable that removal of the neurogenic source
of renal vasoconstrictor activity would result in a
relative hyperemia, e.g., with unilateral denervation.
Earlier work in this area has been reviewed by Smith
(287) and Carstensen & Holle (51).
When clearance techniques are employed in
trained, unanesthetized dogs, which have recovered
well from the effects of surgical denervation of one
kidney, or denervation and transplantation of one
organ, function is equal in the experimental and con-
trol kidneys. This includes concordance of glomerular
filtration rate (creatinine or inulin clearance), plasma
flow (Diodrast or PAH clearance), and indeed, diure-
tic activity and electrolyte excretion (17, 35, 139,
197. 257> 299)-
Carstensen & Holle (51) performed sympathetec-
tomies at the levels of the first and second lumbar
vertebrae (Li and L2) in patients suffering with
arteriosclerotic obliterans and endarteritis obliterans.
Clearances of phenolsulfonphthalein (PSP), creati-
nine, and PAH were measured before and after the
operation. Although individual results were quite
variable, the average changes were not significant:
endogenous creatinine clearance for glomerular
filtration rate (GFR), 127 ± 42 before; 136 ± 59,
after; PAH clearance, 340 ±67 before; 366 ± 92
after. Unilateral sympathectomy (from T8 to Li,
and greater and lesser splanchnics) in patients with
essential hypertension did not increase blood flow
on the operated side (104), and both kidneys re-
sponded by an equal reduction in flow after Adrenalin
administration.
Smith et al. (285) demonstrated in normal, un-
operated human subjects that spinal anesthesia up
to T5 or higher did not produce renal hyperemia as
measured by the Diodrast clearance, nor did it have
any other consistent effect on the renal circulation.
They concluded that the renal blood flow is normally
determined by autonomous, intrinsic activity of the
renal arterioles and is not dependent upon the tonic
activity in the sympathetic pathways, which show
continued action potentials (263).
It must be emphasized that despite its inherent
autonomy of circulation, the kidney will respond with
vasoconstriction during enhanced activity resulting
from direct electrical stimulation of the renal nerves
in the dog, rabbit, cat, and rat (81, 148, 167, 222,
298, 319); this is reversed by a variety of sympatholytic
drugs (81, 319). Studies of blood distribution in the
rabbit kidney supplemented with India ink injection
techniques, revealed that the resulting ischemia was
largely cortical, and that the blood supply to the
medullary zones was not noticeably altered (25).
Houck (148) examined the effect of electrical stimu-
lation of the renal nerves of anesthetized dogs on
blood flow (PAH clearance), filtration rate (creatinine
clearance), and Tm (tubular maximum) of PAH and
glucose (G). By relating filtration rate to unit of
tubular excretory tissue (filtration/TmPAH), and
reabsorptive tissue (filtration/TmG), and the per-
fusion of active tubules with blood (RBF/TmPAH ,
it was discerned that regions of ischemia were pro-
duced, with random closure of nephrons. This was
verified by the distribution of India ink injected into
the renal artery. The evidence was that the effects
resulted predominantly from constriction of the af-
ferent arterioles. Blood was not shunted from the
cortex to the medulla. Study & Shipley (298) also
believe the effects are largely on the afferent arterioles.
They too found no evidence of shunting.
Strong afferent stimulation (acute tracheal com-
pression, sudden clamping of an upper or lower
extremity, and sciatic nerve stimulation (25, 65,
66) ) likewise caused renal vasoconstriction on a
reflex basis. Pain caused by intense cold stimulation
of the hand, or pressure headaches, resulted in de-
creased clearance of Diodrast and of inulin to a lesser
degree (filtration fraction increased), while blood
pressure increased (35 1 ) .
More subtle reflex mechanisms have been dis-
cerned. Bladder distention in chloralosed cats gave
reflex increases in blood pressure; apparently the
kidney participated in the vasoconstriction as mani-
fested by decreases in volume (229). Bilateral splanch-
nectomy abolished the viscerovascular response.
THE RENAL CIRCULATION
I48;
An ureterorenal reflex was described by Hix (146).
When a catheter distended the ureter in the dog,
ipsilateral plasma flow and GFR decreased. The
afferent stimulus facilitated further decrease in these
functions during emotional stimulus. Anesthetization
of the ureter or surgical denervation abolished the
reflex. The physiological significance of such viscero-
vascular reflexes is not apparent; but it can be sug-
gested that the circumstances evoking the reflex
(bladder distention, ureteral irritation) are such that
cessation of urine production would be beneficial,
at least temporarily.
CENTRAL NERVOUS SYSTEM (CNS) CONTROL OF RENAL
blood flow. Evidence exists that a representation of
control of the renal circulation exists in the cerebral
cortex. Smith (286) in 1940 presented an example of
psychogenic renal vasoconstriction, as evidenced by a
marked decrease in the Diodrast clearance; the inulin
clearance decreased only slightly, so that FF increased.
Meehan recently has confirmed the observation that
emotional states will cause a decrease in renal plasma
flow (PAH clearance) (200). Cort (66) observed
reduction of A-V oxygen and carbon dioxide differ-
ences in the cat kidney during stimulation of the
supraorbital cortex, signifying reduction in flow.
Hoff et al. (147) acutely stimulated two cortical foci
in cats on the right and left anterior sigmoid gyri, or
applied more diffuse chronic stimuli to the rostral
surface of the cranium. Ischemia of the renal cortex
(revealed by India ink injection) resulted, with little
effect on the renal medulla. When denervated, the
kidney was passively engorged as the blood pressure
rose. Chronic stimulation for a number of days led to
tubular degeneration as a result of the continued
ischemia.
Wise & Ganong (350) stimulated the hypothala-
mus, pons, and medulla oblongata of pentobarbital-
ized dogs with chronically implanted electrodes.
Effects on glomerular filtration, and excretion of water
and electrolytes were studied. Influence on GFR was
variable: stimulation of the dorsal medulla just lateral
to the midline led to a rise in blood pressure with an
associated decline in GFR and urine volume, abol-
ished by renal denervation. Stimulation of an area in
the obex, in and near the area postrema, led to a rise in
GFR and urine volume, without significant change in
blood pressure. Other points stimulated in the brain
stem (medulla, pons, midbrain, and posterior hypo-
thalamus) had no effects on GFR and electrolyte
excretion, even though some stimuli caused changes
in blood pressure.
Thus, the CNS control of the renal circulation is
intimately wrapped up in the general problems of
higher regulation of the cardiovascular system. As
these become worked out, better insight into renal
control will eventuate (241).
Humoral Control ; Pharmacologic Agents
adrenergic. /-Epinephrine and arterenol (levartere-
nol, norepinephrine) are both active vasoconstrictors
of the renal vasculature. The comparative potency,
and the site of action is not entirely settled, depending
upon technique employed, e.g., indirect clearance
techniques with intravenous injection, or direct flow
studies with intra-arterial injection. The latter
method, employed by Spencer et al. (292) has an
obvious advantage in that local effects can be ob-
served without demonstrable alteration of systemic
pressure. Flow measurement was made with an elec-
tromagnetic flowmeter in dogs. Table 7 shows the
effect of the same dosage of epinephrine and arterenol
as measured by the volume of blood shunted from the
kidney. Only at a 10 ng dose is the difference signifi-
cant, and at this dose epinephrine appears to be the
more effective.
Werko et al. (335) compared the effects on clear-
ances (CIn and CPAH) done in man. The substances
were given in approximately the same dosage during
two experimental periods, following control. An
attempt to assess the differential site of action was
made by application of the formula of Gomez (105)
for calculation of regional vascular resistance. The
average values appear in table 8.
As Spencer et al. found, the differences between the
action of these two adrenergic drugs are not great, and
here arterenol appears to be the more effective. For
both, the greatest degree of resistance change was in
the afferent arterioles. Maxwell et al. (199) injected
1.0 to 1.5 mg of epinephrine intramuscularly in
human subjects, and noted a decrease of 13 per cent
TABLE 7. Effect of Epinephrine and Arterenol on
Renal Blood Flow in the Dog
Dose,
No. of
Paired
Observa-
tions
Avg. m! of Blood
Shunted by:
Mean Difference
/-Epinephrine Arterenol
*g
Epineph-
rine
Arterenol
Avg.
24-5
u-3
68.3
SE
P
I
3
10
9
9
10
58.5
69.8
217.O
33-8
58 -4
148.7
21 .2
IO.6
18.4
°-3
0.36
0.005
[After Spencer et al. (292).]
i486
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
table 8. Comparative Effects of Epinephrine and Arterenol on Human Renal J 'aseular Dy
namies
Vascular Resistanct
in Dynes- sec -cm"5
Cid
RPF
FF
VK
RAa
RAe
RV
Total resistance
X 10'
X 10'
X W
X 10'
Control
"5
665
0.18
IOO
2.71
I .60
2.89
7-19
Arterenol 20.4 pg/min
in
498
O.22
121
5-42
2.13
3-73
1 1 .27
Control
128
660
O.iq
IOO
2-53
1.92
2.88
7-33
Epinephrine 24.5 pg/min
129
553
O.23
"3
4.04
2.29
3-54
9-85
[After Werko et at. (335)]
in CIn and 40 per cent in RPF; FF increased by 39
per cent. Also employing Gomez' calculation, they
computed that the greatest resistance increase was in
the venular and venous component, and suggested
that this contributed significantly to the kidney vol-
ume increase that had been noted some time ago
from epinephrine, the so-called "paradoxical ex-
pansion" of Richards and Plant. The work of Mehrizi
and Hamilton (201) in the dog kidney has confirmed
this conclusion for arterenol. Note that in the data of
Werko et al. (table 8), RAe (efferent arteriolar resist-
ance) and RY (venous resistance) increased equally
after epinephrine, and the change was most pro-
nounced in RAa (afferent arteriolar resistance).
Studies were made in which measurements of
TmG and TmpAH were combined in the dog and
human (149, 208). In the dog, the ratio GFR TmG
decreased significantly, due to over-all reduction in
filtration rate in each glomerulus, rather than nephron
shutdown, according to Houck (149). But Mills
et al. (208) state that GFR and TmG decrease together
with epinephrine and arterenol, implying nephron
closure; this might be anticipated with higher dosage.
In the human, the dosage employed (0.243 /ig kg min
of epinephrine, 0.321 fig kg min of arterenol) caused
no significant alteration of GFR or TraPAH, although
RBF fell to 63 per cent of control. Changes in EPAH
were never observed by several groups of workers.
sympathomimetic drugs. Several sympathomimetic
substances have been studied for comparative effects
on renal blood flow (5, 1 14, 293). None of a series of
rapidly acting vasodepressors, such as isoproterenol
and ethylarterenol, injected into the renal artery of
dogs (flow measured by electromagnetic flowmeter)
induced vasoconstriction (114, 293). Epinephrine,
tried in this series, caused the most potent constric-
tion. The amylbutyl and isobutylamine derivatives of
arterenol, even in quite large doses, were devoid of
any renal vasomotor action despite the fact that they
exhibited definite vasodepressor actions on the sys-
temic circulation. It was concluded that the renal
circulation does not exhibit sympathetic inhibitory
receptors.
Aviado et al. (5) grouped a number of drugs into
four categories, based upon effects observed in the dog
kidney by intrarenal arterial or systemic intravenous
injection. Direct flow was measured by rotameter.
Type A: Drugs which are capable of constricting renal
vessels when injected into the renal artery or when
given intravenously: levarterenol, epinephrine, phen-
vlephrine, metaraminol, methoxamine, and nephazo-
line. Type B: Drugs which constrict when injected
into the renal artery but, when injected intravenously,
constriction is not always encountered : ephedrine,
phenylpropanolamine, hydroxyamphetamine, and
compound 45-50. Type C: These have no important
actions when injected into the artery. When injected
intravenously, renal blood flow is increased because
of their systemic pressor effect : methamphetamine,
pseudoephedrine, amphetamine, pholedrine, methyl-
aminoheptane, tuaminoheptane, mephentermine, and
phenylpropylmethylamine. Type D: Drugs that have
a local dilator action; when they are injected intra-
venously, renal blood flow is decreased as a result of
arterial depressor action: isoproterenol, nylidrin,
isoprophenamine, methoxyphenamine, and cyclo-
pentamine. Spencer (293) reported a weak constrictor
action by isoproterenol, but this was not encountered
in the above series because smaller doses were used.
A similar analysis of various sympathomimetic
drugs on renal hemodynamics has been made recently
by Milhetal. (210), employing clearance in normoten-
sive and hypotensive drugs. In this series, mephen-
termine had the least effect on GFR and RBF, and
methoxamine the greatest.
ganglionic blocking agents. Ganglion-blocking
drugs interfere with the reflex adjustments of the
circulation. They block the vasoconstrictor pathways
THE RENAL CIRCULATION
I487
which control peripheral resistance and venous return
and hence prevent the rise in blood pressure which
results from such maneuvers as clamping the carotids,
cutting the pressoreceptor nerves, the Valsalva maneu-
ver, and the cold pressor test. Essentially, they elimi-
nate efferent nervous influences which keep blood
pressure up. The fall in blood pressure which results
from their administration may, in part, be due to
decrease in cardiac output. Peripheral vasodilatation
and increased flow may occur, e.g., in the limbs, in
the presence of a fall in blood pressure, or at least
in the absence of a rise. There appears to be little or
no direct effect on the vasculature. Responses of the
splanchnic organs, including the kidney, may be quite
different: decreased blood pressure is accompanied by
decrease in flow.
Hexamethonium chloride. Mover et al. (225) gave 2
to 5 mg per kg (iv) and observed an average drop
of 1 38 to 97 mm Hg in the blood pressure in 20 dogs
anesthetized with pentobarbital or chloralose. Glo-
merular filtration rate showed no change (46-45
ml min), RPF decreased from 182 to 1 72 ml per min,
and FF varied from 0.26 to 0.28. Renal vascular
resistance (RVR) decreased from 0.57 to 0.48, and
TmG showed little alteration (161 -155 mg/'min).
In patients (220), with a greater fall in blood pres-
sure at the dosage used, blood pressure fell to 66
per cent of control. CIn decreased to 78 per cent, and
CTpah was maintained at 98 per cent of control, signify-
ing decrease in RYR. One must consider the possibil-
ity that renal autonomy may account for this, rather
than dilatation due to drug action. In another series,
in normotensive and hypertensive patients (209),
blood pressure decreased to 80 to 85 per cent of con-
trol following dosage of 5 to 75 mg. In half no change
or actual increase in RVR occurred, so that GFR and
RPF were reduced at the time of maximum decrease of
blood pressure. In the other half, RYR decreased so
that GFR and RPF were maintained despite the fall
in blood pressure. There was no effect on EPAS.
Arjonad (trimethaphan camphorsulfonate) has a
greater depressing effect on renal function (GFR and
RPF) than hexamethonium, due to greater reduction
of blood pressure (227); RYR is not significantly
altered in normals. In patients with nephrosclerosis
(243) CpAH , originally reduced, tended to be main-
tained despite a fall in blood pressure to 40 to 50 per
cent of control; C?in was noticeably reduced. The
intensity of renal vasoconstriction in dogs, produced
by clamping the aorta or limb trauma, was relieved
by Arfonad blockade (26).
Other blocking agents. Pendiomid (azamethonium
chloride), administered to patients with no vascular
disease at the rate of 2 to 6 mg per min for several
hours (avg 250 mg over 3 hours), caused blood pres-
sure to fall from 97 to 71 mm Hg. Renal blood flow
decreased in about the proportion of the decline in
blood pressure, with no significant change in RVR
(228). Ecolid (chlorisondamine) in hypertensives
shows maintained renal blood flow despite significant
fall in blood pressure (77). Tetraethylammonium
bromide is a renal vasoconstrictor, according to Aas
& Blegen ( 1 ) as revealed by the more marked fall in
Cpah than systemic pressure. Priscoline (tolazoline),
state Young et al. (353), is a vasodepressor and a renal
constrictor both in humans and dogs. Marked decre-
ment in GFR occurs, along with somewhat lesser
decrease in RPF, and they believe the major site of
action is on the afferent arterioles. After denervation
of the kidney, Priscoline has no effect, and Young
and his group suggest that the drug causes afferent
arteriolar stimulation via the sympathetic innervation.
Ilidar (azapetine) directly injected into the renal
artery has no effect up to a dose of 3 mg; above this,
it is a constrictor (292). Regitine (phentolamine)
is both vasodepressor and vasodilator in dogs (226)
at infusion rates of 3 mg/kg for over 5 min RBF
increased from 307 to 341 ml per min, despite a de-
cline in blood pressure from 134 to 102 mm Hg.
Dibenzyline (phenoxybenzamine) was injected into
one renal artery of the dog, followed by infusion of
arterenol systemically. Marked decrease in GFR
and RPF occurred in the untreated kidney, but not
in the treated kidney (129). Dibenamine (N,n'-
dibenzyl-B-chloroethylamine hydrochloride) caused
definite remission of enhanced vasomotor tone result-
ing from hemorrhage in dog (34), but did not alter
the outcome of hemorrhagic shock.
other vasoactive drugs. Apresoline (hydralazine) is
a vasodepressor which reduces vascular resistance in
the kidney. It is not a ganglionic blocker, but its exact
mechanism of action is unknown, although it has been
suggested that it may antagonize neurohumoral sub-
stances (such as serotonin, pherantosin, and angio-
tensin) which are believed to affect blood pressure.
Table g shows its effects in normal subjects and acute
nephritics (dosage, 0.2-2.5 mg/kg, orally).
While improving normal blood flow, Apresoline
unaccountably increased vascular resistance of the
nephritics, despite a fall in systemic pressure. That
hypertension per se was not basic to this response is
shown by the effects in essential hypertensives studied
by Gjorup & Hilden (101). While mean blood pres-
1488
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
table 9. Effect of Apresoline on Renal Hemodynamu s
in Normal and Nephritic Humans
GFR RPF RBF FF PA RVR
A oi mat
[idol V
'35
560
872
-44
92
0. 108
After
i 1 1
665
1 1 16
Xephritic
• ' 73
74
0.067
Ui-I'oi i-
95
594
928
. 164
126
0.150
After
60
456
7'7
. 1 20
99
0.197
[After Etteldorf et al. (86).]
table 10. Effects of Serotonin on Renal Hemodynamu *
in the Unanesthetized Dog
Dose
GFR
ml/min
Urine Vol.
ml, min
RPF
ml/min
P
Before
After
Before
After
Before
After
5
72
71
3-2
3-3
226
227
10
74
74
3-5
■•7
227
238
>o. 10
15
73
64
3-9
1 .0
225
257
4 A
OC c
4 C-*^
31
1.60
3~-
O °
0
/
1.20
Hi
yf'
.80
-«8
\ .•»' ^~- " 4 B
uj£
\w < ^
.40
- en
HI
K
^48
240
M _ ,«*""
^_ c
f
160
-|l
/
^S
/ ,.»--"^a»:fi— - 4 A
80
l.*^^
*r'S
f 4 B
100
-
/
Ill
/
80
as
" 3
/
>
/ ' C
(/> —
, / 4A
UJ o*
/ X/
60
Lq:X
1- E
/ ..•/"
V)
f dr
40
/ -V
(rt
?//
H
20
"
",>^- 1 1 1 1 1
0 40 80 120 160 200 240
RENAL ARTERY PRESSURE ( mmHq )
fig. 25. Renal circulatory autonomy in the dog as a func-
tion of tissue pressure (needle puncture). Symbols 4A and 4C
represent curves obtained with blood perfusions before and
after 4% dextran (4B). [After Hinshaw et al. (143).]
increase in pressure. It was concluded that the auto-
regulation occurred because of increased accumula-
tion of extravascular fluid resulting from enhanced
filtration at high pressure, which compressed low
pressure vessels. Significantly, blood volume estimated
from mean transit time (T-1824 X mean blood flow)
was shown to decrease slightly at pressures in the
range 100-200 in the autoregulating kidney (188),
but volume increased in the K.CN poisoned kidney.
Hence the weight change is likely due largely to ex-
travascular fluid accumulation.
Analysis of regional resistance changes has been
attempted by Hinshaw et al. (145), on the basis of
certain assumptions. The first was that a stabilized
ureteral pressure after occlusion was a measure of
the Bowman's capsule extravascular pressure. Then,
glomerular capillary pressure should equal this pres-
sure plus the plasma oncotic pressure (20 mm Hg
in this series). Another assumption was that intra-
renal venous pressure (postperitubular capillary seg-
ment) was equal to tissue pressure (IRP), (which has
been shown to be correct for the arcuate veins at
elevated venous pressure) plus the plasma oncotic
pressure. The authors have formulated the regional
resistances as follows [reprinted with permission from
Hinshaw et al. ( 1 45)] :
PRE -GLOMERULAR SEGMENT
= RA-UP- COP
POST -GLOMERULAR SEGMENT
(a) EFFERENT ARTERIOLAR SEGMENT
UP-TP
F
(b) post- peritubular capillary segment
(venous segment)
_ tp -i- cop -rv
f
RA= RENAL ARTERY PRESSURE UP" URETERAL PRESSURE
TP. TISSUE PRESSURE COP 'COLLOID OSMOTIC PRESSURE
RV- ORIFICE RENAL VENOUS PRESSURE
Autoregulation was shown to persist during occlu-
sion of the ureters (144), as indeed it does during
venous pressure elevation (119, 123, 281). The above
estimates of regional resistance are applicable, then,
to the autoregulation manifested during ureteral
occlusion in the isolated perfused kidney. In a range
of 100 to 191 mm Hg renal arterial pressure the fol-
lowing average changes occurred : preglomerular
resistance, —4 per cent; postglomerular, +101 per
cent (in the latter value, most is attributable to the
postperitubular capillary segment). Under these spe-
cial circumstances, afferent arteriolar control seems
unimportant, and it is the influence of increased
IRP on compressible postglomerular vessels that
appears to dominate.
Although this hypothesis is ingenious in its applica-
tion, the special circumstance of the measurements
will make it difficult to apply to the normally func-
tioning kidney. It is well to recall that the fundamental
precept, i.e., that IRP varies with arterial pressure,
has not been uniformly accepted by all investigators.
If the above hypothesis is correct, decapsulation of
the kidney should have a significant influence on the
autoregulatory mechanism. In this, investigators are
not in agreement. Bounous et al. (28) after careful
I I'll
HANDBOOK OF PHYSIOLOGY-
CIRCULATION II
decapsulation procedures, found that autoregula-
tion was indeed abolished (fig. 26). Haddy et al. (124)
illustrate several experiments which offer support of
this: two pressure-flow curves are more linear after
decapsulation than before. But Miles & DeWardener
(206) found no difference between the IRP of the
control and that of the decapsulated kidney. Elevation
of IRP by mannitol diuresis and elevation of venous
pressure caused approximately equal increases in
IRP in the decapsulated kidney and in the paired
control. In an extreme situation, following K.CN
treatment and elevation of perfusion pressure by
pump to 300 mm Hg, IRP increased ca. 100 mm Hg
in both control and decapsulated kidneys.
In summary, the tissue pressure theory is attractive
in some respects, but since it concerns a purely phvsi-
cal mechanism it is hard to square with the lack of
autoregulation in kidneys treated with procaine,
KCN, and papaverine, in the oil-perfused kidney, or
even in dead kidneys. Implicitly, it dispenses with
the need for afferent arteriolar control, but a con-
siderable body of evidence supports the possibility of
such control.
the myogenic theory. The principal evidence for
this theory comes from the behavior of the renal blood
flow during rapid changes in perfusion pressure. An
example taken from the work of Semple &
DeWardener (281) appears in figure 27. Flow was
measured with an electromagnetic flowmeter. Note
I60>
5CV
PV
30SEC -v-
fig. 26. Effect of decapsulation on autoregulation in the dog.
C: control; V: bilateral section of cervical vagosympathetic
chains in the neck; S: renal denervation; I): renal decapsula-
tion. [After Bounous et al. (28).]
fig. 27. Renal circulatory adjustment following sudden in-
crease in arterial perfusion pressure i/J.)i PI': renal venous
pressure. [After Semple & DeWardener (281).]
the immediate "overshoot" of flow as pressure is
raised, followed by return to a flow level somewhat
below the control within 60 sec, and then stabilization
at the control flow but at a pressure some 50 mm Hg
higher than during the control. On occasion, rhythmic
rapid fluctuations in flow were observed after pressure
elevation before stabilization occurred, a "hunting"
phenomenon.
When the elevation was done in progressive steps,
the overshoot was proportional to the pressure eleva-
tion, but returned in each instance to approximate
the control level [see fig. 28 (308)]. It is of interest
that the levels of flow, reached instantaneously after
pressure change, fall on a curve describing the
pressure-flow relationship in the same kidney after
paralysis of smooth muscle activity with papaverine
(x x in the figure).
Likewise, when pressure was dropped in steps, flow
decreased immediately in a passive manner, but in
30 to 60 sec readjusted to the previous level [fig. 29
(120)]. In this series, constancy of flow was main-
tained down to 70 mm Hg, then fell off rapidly.
Thurau & Kramer (307) have analyzed in an in-
teresting fashion the correlation of total blood flow,
superficial (cortical peritubular) capillary blood con-
tent, and weight change in response to rectangular
pressure increments. The results are illustrated in
figure 30. Capillary blood content was measured by
an "infrared reflectometer'" technique. Note the
typically instantaneous overshoot of flow as pressure
is increased, followed by stabilization. (Allowance
must be made for the possibility that an overshoot
artifact by the rotameter may contribute to the initial
rise.) This appears to be a function of the initial tonus
THE RENAL CIRCULATION 1 495
300 --
ZOO mmHg 300
fig. 28. Immediate and stabilized relationship of renal
blood flow to perfusion pressure. [After Thurau & Kramer
(308).]
¥,0
3,0
10
10
0
1 T
'g min
/
/
/
/s
x¥
x3
/
*Z
>
/
x7
/
7
•
/ ,
^
7 .
S
&
. y
V if
3^-
d>
/ S
,//
/-r
/
DmTnHg
! 1 1
1
0
eo
¥0
so
so
700
720
7W
fig. 29. Immediate and stabilized response of renal blood
flow to decrease in perfusion pressure. O: Immediate flow; •:
stabilized flow; X : immediate response to restoration of pressure,
then return to control ( •) cluster, at upper end of curve. The
numbers indicate sequence of response. [After Grupp et
al. (120).]
of the vascular smooth muscle; when low, overshoot
was greater than when tonus was high. The capillary
volume increases transiently during the phase of
overshooting (increase is with downward deflection of
the galvanometer) accompanying the initial passive
expansion of arteries and arterioles as pressure is
suddenly increased. Then, as total flow settles to
BLOOD PRESSURE
BLOOD FLOW
GALV.
DEFLECTION
WEIGHT
120
fig. 30. Immediate and delayed adjustment to rectilinear
increase in blood pressure — renal blood How, superficial corti-
cal blood volume (galvanometer deflection) — and kidney
weight. (Downward deflection of galvanometer indicates in-
crease in volume.) Weight change is an approximation of trend.
[After Thurau & Kramer (307).]
lower levels after the onset of the myogenic contrac-
tion, capillary content decreases somewhat.
Upon decrement of pressure, flow decreases mark-
edly below the control, indicative of the contracted
state of the resistance vessels. Then normal flow is
slowly restored as the myogenic response recedes.
Capillary blood content during this decrement in
flow also decreases significantly, then is restored as
total flow rises.
The weight change may show a triphasic response
in experiments with more prolonged stages: a) an
initial rapid increment as blood surges into the re-
laxed vessels; b) a transient drop as the myogenic re-
sponse occurs; and c) a secondary rise. The last may
be the result of increased transudation of fluid through
the capillaries at the elevated pressure and increased
flow.
The dynamic reactivity implied in these fairly-
rapid adjustments corresponds to the type of reac-
1 4g6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
tivity anticipated from the smooth muscle of the
vasculature. That a vital phenomenon is involved is
supported by the action of a number of agents known
to impair smooth muscle activity: papaverine will
eliminate autoregulation (306, 308), as will K.CN
(188, 207, 233) and theophylline (121). Procaine has
been cited earlier. Certain anesthetics, such as numal
(120, 121) and chloralhydrate (330) impair auto-
regulation as will ethanol (260).
Both cooling and perfusion of the kidney with oil
remove autoregulation (328, 330). Hemorrhage de-
presses autoregulation (73, 271, 331). Anoxia created
by perfusion of the kidney with perfusion fluid [20 9i
plasma-80 % polyvinylpyrollidone (PVP)-Locke's solu-
tion] subjected to helium rather than oxygen appeared
to impair autoregulation somewhat; flow increased
more with pressure increments than with comparable
increments during the control (331 ).
It is well to point out that autoregulation may be
impaired in another manner and may, in part, ex-
plain the apparent loss of response in hemorrhage.
Under this circumstance and with adrenaline and
hypertonic fluid infusions, the smooth muscle of the
vasculature becomes highly tonic, and responses to
increments in pressure become much reduced (307).
Then the pressure-flow curve becomes convex to the
pressure axis and resembles the pressure-flow curves
obtained in the hind limb and other organs, which
usually have a higher resistance than the kidney.
aberrant resltlts. Several investigations may be
cited in which typical autoregulation was not ob-
served, but in which pressure-flow curves were linear
or convex to the pressure axis. This includes the work
of Ohler et al. (235) in the rat. Indications of a high
degree of vascular tone are seen in the low flows in
many preparations and the high flow intercepts on the
pressure axis. It will be recalled that others have re-
ported the more common concave-to-pressure-axis
curve in the rat (333), indicating autoregulation.
Likewise, the work of Langston et al. (172, 173)
manifested a convex-to-pressure-axis relationship of
flow in dogs. Again, flow per gram of tissue was low
(less than 2 ml min g) at normal pressure, suggesting
a highly tonic preparation. The high pressure inter-
cept for flow also suggested this. In the first report
(172) flow appeared to be only about 10 ml per min
(total per kidney) at 60 mm Hg. In the second report
(173), in the control series, the zero flow intercept
lay between 20 and 40 mm Hg; flow at 100 mm Hg
in most preparations was less than 1 .5 ml per min per
g. Furthermore the method of perfusion suggested the
possibility of a source of technical error. The kidney
was perfused via an isolated segment of aorta at the
level of the renal artery. Hardin et al. (130) used a
similar technique, and found the same convex rela-
tionship. However, when they carefully tied small
lumbar arteries leaving this segment of the aorta, the
pressure-flow curve assumed the more commonlv
found contour, concave to the pressure axis (fig. 31).
SIGNIFICANCE OF THE MYOGENIC RESPONSE. Bayliss ( I 3),
a number of years ago, called attention to a myogenic
response to sudden changes in pressure both in de-
nervated organs and segments of artery (carotid), and
attributed it to alterations in tonus of smooth muscle
in the arteries in response to change in tension.
Wachholder (324) studied isolated segments of equine
carotid, and observed contractions following sudden
increases in pressure occurring with a latency of
usually 10 to 20 sec (8 sec was the shortest). The con-
traction phase lasted 20 to 60 sec. Burgi (44) utilized
bovine mesenteric artery segments, but saw distinct
responses in only 23 per cent of his tests; weak re-
sponses occurred in 9 per cent, and in 12 per cent the
response had so great a latency that it was deemed
questionable; 56 per cent showed no response.
Folkow (89, 90) has placed the suggestion of Bayliss
and others on a firmer footing. His experiments,
utilizing the dog hind limb preparation, under condi-
tions which apparently controlled possible neurogenic
350
'■Sr
300
1 7
• LUMBAR ARTERIES occluded
250
X
• LUMBAR ARTERIES patent
E
E
200
UJ
"0:
150
3
CO
to
UJ
■ (Z
0.
1.0 J
-—~7
100
8^/
""7
50
.9
FLOW
m l/mln
0 25 50 75 OO 125 150 175 200
fig. 31. Perfusion pressure as a function of rate of blood flow
through both kidneys of a dog before and after occlusion of the
lumbar arteries. Numbers are renal vascular resistance in mm
Hg/ml/min. (After Hardin et al. (130).]
THE RENAL CIRCULATION
'497
and humoral factors, gave support to the myogenic
theory, and he concluded "Vascular tone is in its
basic origin myogenic, though strongly influenced by
external factors."
That this property of the smooth muscle of the kid-
ney arterioles is typical of smooth muscle elsewhere
is shown by the work of Bozler (30). Isolated segments
of ureter were subjected to sudden increases in in-
ternal pressure. This created electrical potentials
which produced, at first, local responses; if the poten-
tial was strong enough, a conducted response resulted.
He found that the greater the pressure, the steeper
was the local potential change and the shorter the
delay for the onset of conduction. Bulbring (43) has
shown that stretch of smooth muscle cells of the taenia
coli acted as a stimulus for increased myogenic auto-
maticity, and that the element of the smooth muscle
cell sensitive to stretch was closely combined with the
properties of the tension-producing element.
The group of investigators that support the
myogenic theory to explain autoregulation of the
kidney favor the afferent arteriole as the site of regula-
tion. More specifically, the myocytes of the juxta-
glomerular apparatus appear to be a likely point of
control (274, 331). In conclusion, the myogenic
theory seems most attractive as an explanation of
autoregulation of the renal circulation, but it is likely
that acceptance of one theory to the exclusion of some
of the others would be an oversimplification. The
challenging prospect remains to integrate properly
the several possibilities into a unified concept which
might operate in the intact, unanesthetized animal in
normal circulatorv homeostasis
PRESENT STATUS OF THE TRUETA
JUXTAMEDULLARY SHUNT
It was postulated by Trueta et al. (311) that diver-
sion of renal blood from its usual cortical route to the
"less resistant and more capacious medullary circuit"
(198) (probably not true by currently known facts)
was a physiologic mechanism which was involved in a
number of abnormal circulatory states. These in-
cluded reflex anuria, anuria associated with incom-
patible blood transfusion, crush injury, blackwater
fever, etc., Pitressin inhibition of water diuresis, the
renal ischemia of shock or that induced by fright or
adrenaline, in the reduction in tubular excretion
following protracted renal ischemia, and in the genesis
of essential hypertension.
These investigators had reported that during renal
ischemia the arterial pulse may be seen in the renal
vein, and that the renal venous blood may acquire an
arterial color. It was their belief that the juxtamedul-
lary glomeruli and vasa recta circuit may afford
veritable shunts between the renal artery and vein "...
a diversion of blood from the cortex, the most active
part of the kidney, to the medullars- pathway, with a
possible increased speed of flow through these
channels."
Such shunts should therefore cause a reduction in
renal oxygen A-V difference. Furthermore, this would
shunt blood away from the zone ol greater metabolic
activity in the cortex to the juxtamedullary vasa recta
and loop of Henle system, with less efficient perfusion
of the proximal tubular secretory sites. This should
cause a decrease of ED or £PAH with an increase, or
no necessary decrease, in total blood flow as measured
by direct methods or perhaps by the Fick method.
Finally, if shunts open which bypass glomeruli, £In
should decrease without a decrease in blood flow, or
should decrease more markedly than blood flow
(assuming continued adequacy of filtration pressure).
Morphological identification of the shunt should be
possible with injections of India ink, Prussian blue, or
radiopaque material such as Thorotrast. The Trueta
evidence consisted mainly of appearance of India ink
or Thorotrast in higher concentration in the juxta-
medullary region when injected intravascularly during
sciatic nerve stimulation or epinephrine action. But
great care must be exercised in attempts to interpret
rate or volume of flow by appearance of the injection
mass. Thus, contraction of venous effluent constrictors
could give the appearance of congestion of the medul-
lary circuit, in the face of an actual reduction of flow.
Incomplete filling due to faulty injection could give
the appearance of vasoconstriction in the cortex.
Morphological Evidence
Injection studies have been controversial, being
interpreted either in favor of the original hypothesis or
against it. This has been largely a matter of interpreta-
tion of what are often quite similar pictures.
Montague & Wilson (214), correlating Thorotrast
injection studies with clearance data (£PAH) in rabbits,
believed they saw evidence of a juxtamedullary shunt
after epinephrine injection. This was accompanied
by marked decreases in EFxn (mostly negative, and
averaging — 26.6^). Herdman & Jaco (138) par-
tially constricted one renal artery of rabbits, and in-
jected India ink 3 days to 5 weeks later. They found
the ink chiefly in the inner cortex and juxtamedullary
1498
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
zone, suggesting the diversion of blood accompanying
cortical ischemia. Tracheal occlusion in rabbits (92)
caused paling of the kidney and decrease in volume.
Sections of innervated kidneys removed at the height
of anoxia showed anemia of the cortical portions. In
the denervated, there was no such "diversion" of flow,
but it was possible to produce it in these by injections
of epinephrine. Arcadi & Farman (2) state that in
rabbits, blood was diverted almost exclusively to the
cortical circuit by pilocarpine and magnesium sulfate
injection and most prominently by water diuresis.
Dehydration, on the other hand, caused the accumula-
tion of ink in the medulla. The findings of Kuhlgatz
(168) in rats were very similar with respect to the
findings on diuresis and dehydration. Mover et al.
(222, 223) although finding uniform distribution of
India ink in dog and rabbit kidneys during sciatic
stimulation, nevertheless report cortical ischemia and
subcortical accumulation of ink after injection of 0.2
mg of epinephrine into the rabbit, but this finding does
not have to be interpreted as demonstrating the open-
ing of medullary shunts.
Insull et al. ( 1 50) have pointed out that an adequate
filling pressure is needed for validity of the injection
methods. When Prussian blue was injected into fresh
rabbit kidneys at a pressure of 50 cm HoO, good
filling of the entire kidney, including cortex, was ob-
served; at 25 cm HoO, filling of the juxtamedullary
glomeruli and vasa recta only occurred.
In some experiments involving sciatic stimulation,
stimulation of the perirenal plexus, or hemorrhage,
the juxtamedullary glomeruli were uniformly stained,
while the peripheral glomeruli were not. This oc-
curred only when renal blood flow was low. They in-
terpreted this as regional cortical ischemia but not as
increased flow through the medulla.
The conclusions of Block et al. (24) and Kahn rt a/.
(156) are similar. Block et al. tried stimulation of renal
nerves in rabbits, clamping of the artery, injection of
constrictor agents, and sciatic nerve stimulations.
They concluded that a pale cortex and a medulla
filled with blood were not evidence that blood was
flowing largely through the medulla: the medulla
may be congested even though flow has stopped. Kahn
et al. found during sciatic stimulation in the rabbit a
normal distribution of ink in 8 of 1 1 animals. In 3,
however, the peripheral cortex had no ink, and the
juxtamedullary glomeruli and the medullary vessels
were well filled.
In summary, the rabbit kidney during various types
of strong afferent stimulation or during epinephrine
action mav demonstrate a cortical ischemia, with
maintained flow in the medulla. This cannot be
interpreted as diversion of flow to the medulla, and
particularly not as increased flow through this zone.
Nevertheless, the anatomical evidence of a dual circu-
lation is good, and there is some good functional
evidence of this.
Daniel et al. (6g) have made excellent serial angio-
grams in cats and dogs which follow the progress of
Thorotrast through the kidney. The material passes
very rapidly through the cortex, but the diffuse
shadow of the medulla persists long after the veins
have emptied, demonstrating a much slower per-
fusion of the vasa recta system of the medulla. This
conforms with the studies of Kramer et al. (166) who
used a photoelectric technique. They bring evidence
that the role of the medullary vasa recta system may be
unique in connection with the role of the counter-
current system, and the dilution and concentration of
the urine. Other evidence is at hand in support of
this, and will be taken up in a later section.
Functional Evidence; Interpretations
Based on Clearance Data
Scher (264) used a heated-thermocouple technique
in dogs, rabbits, and cats, one thermocouple being
placed in the cortex and the other in the medulla.
Although quantitative interpretations must be made
cautiously, focal flow paralleled total renal blood flow
during action of epinephrine and arterenol, acetyl-
choline, and stimulation of periarterial plexus.
The clearance data are concerned mainly with
changes in A-V oxygen difference, and in £PiH and
Eln. Mover et al. (223), during marked reduction in
flow (—46%) resulting from prolonged sciatic stimu-
lation in dogs, found the A-V oxygen difference in-
creased from 1.7 to 3.3 vol per cent; in rabbits, from
3.1 to 6.8 vol per cent in a group in which venous
flow was measured, and 3.5 to 9.2 vol per cent in a
venipuncture group. With epinephrine, resulting in
increased blood pressure and reduced flow (as much
as 70% reduction) in rabbits, A-V oxygen increased
as much as 313 per cent. Other investigtions de-
tected no significant alteration of oxygen extraction
in dogs or man during epinephrine injection ( 1 49, 224,
254)-
Epinephrine in dogs gave no evidence of a medul-
lary shunt, since £PAH, ECT, extraction of oxygen and
TmG did not significantly decrease in presence of
moderate to marked decrease in GFR, urine flow,
and RBF (149). Mover & Handley (224) observed no
reduction in EPXH in dogs, but TmG decreased due to
THE RENAL CIRCULATION
'499
table 1 1 . Effect of Albumin Infusion on Renal Function in Man
\-75 S dlb. (300 ml) infused in 10-24 m'"l
U. Inn
After
'; Increase
Cln
CpAH
Total
KM
EpAB
Med.
RPF
Cln
CpAH
Total
RPF
I I'M!
Med.
RPF
Ci»
CPAH
Total
RPF
Med.
RPF
Normal
l65
863
1044
82.7
181
189
I I 23
1586
71.2
463
'4
3'
52
256
Normal
112
741
820
9O.4
79
124
IOOI
1280
78-3
279
1 1
35
56
354
Normal
130
638
788
8l .O
■5°
I38
764
1250
6l .O
486
6
20
59
324
Hypertension
IO9
375
429
87-3
54
112
491
662
74-3
'7'
3
3'
54
316
Hypertension
IOO
542
*57
82.5
"5
IO3
693
938
73-7
245
3
27'5
43
213
Hypertension
36
■45
250
58.O
105
39
154
353
43 -6
'99
8
6
42
190
Chronic nephritis
56
265
4'3
64.O
148
32
24O
533
45'
293
-43
-9.0
29
198
Nephrotic synd.
66
444
493
90.0
49
1 10
757
IOIO
75 0
253
67
70
205
5'6
Avg.
86
26.5
67 -5
296
Total RPF (total renal plasma flow) = (Cpah/Epab)' 100.
Cpah- [After Cargill (48).]
Med. RPF (medullary renal plasma flow) = Total RPF —
glomerular closure during epinephrine and arterenol
infusion. Epinephrine and histamine caused a de-
crease in ispAH of only 1 1 .4 per cent at the maximum
in the human (254). Tilting, with resultant increased
sympathetic activity as evidenced by reduction in
RPF (CPAH £pah)j induced no change in £PAH (39)-
An interesting exception, and at present the only
type of positive evidence of increased flow through
the medullary circuit, is supplied by Cargill (48) on
the effects of iv administration of serum albumin on
renal function of human subjects in water diuresis.
The results are summarized in table 1 1 . Note that
■Epah decreased and the calculated medullary plasma
flow increased from no to 298 ml per min. Michie
et al. (204) have supplied excellent confirmation of
these results. In their studies, the constancy of TmG
and TmpAH suggested that no nephrons were shut
off as RPF increased up to 200 per cent. They sug-
gested that this was due to opening of intrarenal
shunts without diversion of cortical blood. Barker et
al. (11) concur with the observations of Michie et al.
They found also that A-V oxygen decreased by 30 to
40 per cent as the total renal blood flow increased, a
fact consonant with the above interpretation.
Role of the Medullary Circulation
in Diuresis and Antidiuresis
The Oxford workers originally suggested juxta-
medullary diversion of blood, with diminished filtra-
tion and greater reabsorption of water in the thin
segment, as a mechanism explaining the antidiuretic
action of Pitressin, spontaneous changes in urine flow
in the erect and supine position, during emotional
excitation, and in other circumstances involving
endogenous ADH secretion. The medullary diversion
of blood was first proposed as an explanation of anti-
diuresis by Frey (96) in 1934.
From this it has been assumed by Maxwell et al.
(198) that changes in EPAH should accompany diure-
sis. This was not the case in a series of human sub-
jects presented by them with a range of urine flow
from 0.68 to 19.7 ml per min. £]>AH remained within
a normal range of 0.88 to 0.96. However, their ob-
servations of entire diuretic cycles were few. Further-
more, inhibition of water diuresis with doses of
Pitressin as high as 2000 to 5000 milliunits per hour
caused no significant alteration in EPAH, Ein, and
A-V oxygen difference.
The application of the intrarenal photoelectric
technique for measuring regional dye transit time
(T-1824) has disclosed interesting new facts about
this mechanism. The technique as applied by Kramer
et al. (166, 309) is shown in figure 32. Normally,
mean transit time averages 27.7 sec for the medulla,
and 2.5 sec for the cortex of the canine kidney. When
the perfusion pressure was elevated (carotid sinus de-
nervation and vagal block by narcosis, or by pump),
medullary transit time was markedly reduced, while
cortical transit time and total renal blood flow (ro-
tameter) remained constant. With this, urine volume
increased noticeably. Two representative experiments
appear in table 12 (308). From such evidence, they
conclude that the juxtamedullary glomeruli and the
vasa recta system do not demonstrate autoregulation
of flow. (It is possible that enhanced perfusion of the
medulla at higher pressure via shunts of other types,
e.g., the spiral arteries, arteriolae rectae verae, or
1500
IIAMlBi )( iK OF I'HYSIOI OCA-
CIRCULATION II
Kobe/
Rinden-
Refleklomeler
Y\^j Photozelle
R
Harnabfluss
fig. 32. Method for measuring cortical and medullary cir-
culatory transit time. [After Kramer el al. (166).]
table 12. Effect of Increased Perfusion Pressure on
Regional Transit Time of the Kidney
Renal Art.
Pressure
mm Hg
Med. Trans.
Time (tpM)
sec
Cortical
Trans. Time
U.,R) sec
tPM
1PR
Blood Flow
ml/g/min
Urine Vol.
ml/min
Experiment
'°5
■95
205
16.5
12. 1
1 1 .2
4.2
4-5
4-5
0.25
2.50
3-5°
Experiment 1 1
140
32-3
3'
IO.4
4.1
0.30
l65
27-5
3-3
8-3
4.0
o-75
210
22.9
3-i
7-4
4-3
2.25
[After Thurau et al. (308).]
Ludwig's arterioles, could produce this effect.) The
total volume flow, they contend, is small relative to
total flow, and is within the error of the method of
measurement. Nevertheless, it suffices to "washout"
the osmotic gradient established in the critical long
vasa recta loops and accompanying loops of Henle in
the papillary zone. With this, the mechanism for
concentration of the urine becomes limited; and
diuresis ensues. Selkurt (276) has shown that this
type of diuresis is accompanied by enhanced sodium
excretion.
In support of this hypothesis are the effects of water
diuresis and ADH action (fig. 33). Note the marked
decrease in medullary-plasma transit time (tpM)
with diuresis, and the return during ADH action.
These effects are thought to be the result of changes in
blood viscosity brought about during water diuresis
(decreased concentration of albumin and cells in the
vasa recta) or increased ADH action. It will be re-
called that with water diuresis, lack of ADH activity
permits the urine to remain hypotonic: the osmotic
gradient is dissipated and, with it, no concentration
of blood constituents occurs: blood viscosity decreases,
and transit time is reduced. The vasopressor activity
of ADH (arginine-vasopressin) may conceivablv be
involved in regulation of blood flow in this circuit.
The critical point of change in t,,M occurs when
Cosm — V — o, as revealed in two representative
experiments in figure 34. When free water clearance
(Ch2o) begins, T„M reaches a rather constant, minimal
value.
The failure of Maxwell et al. (198) to note changes
in £pAH with diuresis and antidiuresis may have oc-
curred because the above changes in flow are small
enough not to be discernible in the normal range of
variation of the EpXH measurement.
An explanation of the results of Goodyer et al. ( 108,
log) may fall into line with the above findings. During
nonshocking hemorrhage during which arterial
pressure was kept constant, sodium excretion declined
in the absence of measurable changes in glomerular
filtration rate or renal plasma flow. (Data on urine
volume were not supplied, but this must certainly
have declined.) Measurement of intrarenal hematocrit
led them to conclude that intrarenal redistribution of
blood flow may have occurred, involving diversion of
plasma to cell-poor capillaries (or to lymphatic
spaces). This could involve the above mentioned vasa
recta mechanism, and obviouslv would be the con-
verse of the above cited experiments involving in-
creased renal perfusion pressure.
In summary, a newly recognized and important
function of the vasa recta system as a counterpart of
?r
I'
ml/min
80r 1
few
- *:
sec
80
Wasserdiurese
nachADH
tpMs*
"/Posm
y ■
/ &
60
' W
9(1
^\
" " 0 2 3 CStd.
fig. 33. Mean medullary transit time of T-1824 (tpM)
during diuresis and after ADH; U/P„sm: osmolar concentration
ratio of urine to plasma; GF: glomerular filtration. [After
Thurau et al. (309).]
THE RENAL CIRCULATION
I 50I
80
sec
60-
W
20
0% s
\°
0
- a.
1-
4
ml/min
rH20 "•—
-*Ch2o '
2
fig. 34. Mean transit time during
diuresis and antidiuresis as a function of
CHoO (water concentration); T§20: os-
motic water deficit, i.e., below equilib-
rium point. [After Thurau et al. (309).]
the loop of Henle in the countercurrent system is indi-
cated, and this becomes the modern role of the former
Trueta juxtamedullary shunt.
RESPONSE OF RENAL BLOOD FLOW
TO PHYSIOLOGICAL STRESS
The dog and man differ significantly in the response
of the renal blood flow to exercise. The canine kidney-
shows a considerable amount of autonomy of circula-
tion during exercise to a degree which results in sig-
nificant reduction in blood flow in man. Blake (23)
exercised dogs on a treadmill at the rate of 2.5 mph
for 40 min and observed no significant changes in
Cpah and C'cr- In one of three dogs tested, when "emo-
tional" stimulus was superimposed (a loud horn),
Cpah decreased from 161 to 137 ml per min, then re-
turned. CCr did not change. Greater effects were noted
on sodium excretion, which decreased during the
emotional response. Carlin et al. (50) ran their dogs
at 5.6 to io mph on a 150 grade for 7 to 20 min; pulse
rate was often over 160 per min, and respiratory
rate, over 300 per min. Yet there was no change in
Cpah or CIn, and sodium excretion did not change sig-
nificantly.
Recumbent human subjects pedaling the equiva-
lent of 0.5 kg weight at 60 cycles per min, which
doubled their resting oxygen consumption, showed a
20 per cent reduction in RPF, while GFR remained
unchanged, as did £PAH (42). Sodium excretion de-
creased by 20 per cent. Chapman et al. (52) worked
their normal male subjects on a treadmill for 16-min
periods. The following decreases in CPAH were noted :
at 3.0 mph at o grade, 6 per cent; 3.0 mph at 5 per
cent grade, 17 per cent; at 3.5 mph at 10 per cent
grade, 25 per cent below resting control. Work was
continued for another 16-min period, with the follow
ing decreases: 15, 27, and 35 per cent, respectively.
Recovery was incomplete after 40 min. In a subse-
quent study (53), the above results were confirmed
and, in addition, the work period (3 mph at j'~,
grade) was prolonged to 3 hours; during the second
and third hours, CPAh decreased no more than it had
during the first hour ( — 18.510 —33.7 % below control).
Recovery occurred in about an hour. Radigan &
Robinson (250) observed that exercise (3 mph on a 5 ' .
grade) produced a 42 per cent decrease in RPF, but
the Cln did not change when the environmental tem-
perature was cool (21 C); when the work was done
in a hot environment (50 C), C'In decreased by 16.5
per cent, with a 36 per cent decrease in CPAH. In
another study, subjects who had run the 440-yard
dash at full speed had reductions of 18 to 54 per cent
below control in CD, and exhibited also decrease in
CIn (10). The apparent blood flow remained reduced
for 10 to 40 min postexercise.
Harpuder et al. (134) compared different grades of
work in different postures on CPAH. Light work (3500
kg-m) in the erect or sitting position had no significant
effect. At ca. 4800 kg-m, CPAH decreased to 0.85 ±
.08 of control in the supine, as compared to 0.69 ± .04
in the erect position. At 9120 kg-m in the erect posture,
the reduction was 0.55 ± .10. At the peak of exer-
cise, blood pressure had risen from 1 14/72 to 164/82
mm Hg, and heart rate from 64 to 1 42 per min. They
point out that with a normal renal blood flow about
1 liter per min (2o'<' of the cardiac output), a saving
of 0.5 liter per min is made available for the circula-
tion of active tissues. White & Rolf (339) similarly-
analyzed the effects of running exercise With brief
maximum exercise, RPF decreased to 20 per cent of
control, and under extreme conditions they predicted
that almost 1 liter per min of blood was made available
or the active tissues.
The cardiac patient shows more marked renal re-
1502
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
sponses to exercise than the normal subject. A degree
of exercise (recumbent) on a bicyle ergometer which
had no influence on normal subjects (70 kg-m min)
had the effects shown in table 13 on cardiac patients
[Werko et al. (336)]. Patients were grouped into cate-
gories based on heart size [Group A, 512 ml m- BSA;
Group B, 796 ml/m2 (no right heart failure) ; Group
C, 807 ml per m2 (with right heart failure)]. Note
particularly reduction in RBF, percentage of cardiac
output, and increase in renal vascular resistance. The
work of others is in support of this (94, 155, 203).
Evidently heart failure lays additional stress on com-
pensatory mechanisms. When exercise is added, more
intense neurogenic and hormoral influences serve in
shunting blood away from the kidney.
The kidney may not suffer as much as might be
anticipated during the curtailment of flow in exer-
cise, at least in heart disease, in view of the findings
of Bishop et al. (21). They have discovered that the
A-V oxygen difference of the kidney of the cardiac pa-
tient may increase, during exercise, in apparent con-
tradiction to the generally accepted "flow-limited"
characteristic of its circulation. The renal A-V oxygen
in 12 cardiac patients (mostly rheumatic heart disease
but none in congestive heart failure), averaged 2.03
vol per cent; this increased to a mean of 3.31 vol per
cent during exercise. In one, an increase to 12.40 vol
per cent was recorded.
Posture and Orthostatic Hypotension
In normal young males, CPAh in the sitting position
is 0.91 ± .04 of that recorded with the subject supine;
in the erect position, it is 0.85 ±0.14 (242). Tilting of
the subject from the horizontal similarly produces re-
duction in glomerular filtration rate, e.g., 127-120
to 98-93 ml per min (6o° tilt), recovering to 126, 1 1 7,
and 1 12 ml per min (39).
Motionless standing, or tilting of the subject lying
quietly on a tilt-table, leads to progressive venous
stagnation, reduced cardiac output, and neurogenic
vasoconstriction until the cerebral circulation becomes
inadequate, at which point syncope occurs.
When tilting is done in increments from horizontal
to 60 °, RPF and C,„ progressively decrease; this is
more marked when reflex compensatory mechanisms
are good, as manifested by well-sustained blood
pressure, than when blood pressure is not sustained
and fainting is imminent (36, 60, 72, 286, 334). Pa-
tients prone to orthostatic hypotension in particular
manifest the latter responses (36). £PAH usually is not
altered (36, 334). Filtration fraction tends to increase,
suggesting predominantlv efferent arteriolar constric-
tion (60, 287).
Two different types of responses are shown in table
14. In table 14 A, compensation was good, and arterial
blood pressure was well maintained. In B, in a patient
subject to orthostatic hypotension, renal blood flow
"opened up" as syncope ensued.
The type of response seen in table 14 B was also
shown during fainting produced by cuffing the lower
appendages, plus venisection (up to 500 ml) (72). In
all cases, when blood pressure fell, CPAh decreased
but the calculated renal vascular resistance decreased.
Thus, the kidney participates in the more widespread
splanchnic vasodilatation which occurs during syncope
(14)-
It is of interest that the medullary flow increases in
the subject (table 14 B) during the failure of vascular
compensation. Although this type of calculation of
medullary flow must be accepted with some reserva-
tion, the data suggest that vascular resistance de-
creases more markedlv in the medullary circulation,
table 1 3. Effect of Exercise on Renal Hemodynamics in Cardiac Patients
Brachial BP
mm Hg
Cardiac
Index
Cln
ml/min
CPAH
ml/min
FF
%
RPF
ml, min
% Cardiac
Output
R dynes-sec
cm5
(
jroup A
N = 26 C
N=i4 E
9°-9
+6.2
3-59
+ .07
108 7
-2-5
(
404.8
-485
iroup B
27.8
+3-'
731
-90
12.2
-3-4
IO5.6
+ 22.9
N=7 C
N=7 E
97-7
+ 10.2
2. 1 1
+0.86
76.3
-4.0
(
216.4
-36.8
~iroup C
34-9
+3-4
398
-71
10.8
-3-4
183.0
+ 24.8
N=7 C
N=7 E
96-3
+5-4
i-94
+0.26
83.0
-8.0
204.6
-23-7
43-3
+2.6
403.6
-48.O
1 1 .2
-1.8
'99-7
+65 -7
[After Werko et al. (336).
THE RENAL CIRCULATION
I5°3
table 14. Effect of Tilting on Renal Vascular Compensation in Man
Position
Cln
ml min
CpAH
ml min
RPF"
ml min
Medullary6
Flow ml/min
Pac
mm Hg
Trued FF
Epah
RKe
A. Adequate cardiovascular compensation'
Horizontal
141
588
620
32
94
0.227
O.96
.091
30 degree
109
543
573
28
116
0.190
0.88
. 121
45 degree
77
235
262
27
104
O.327
0.90
.238
Horizontal
'52
766
79°
24
98
°'93
o-97
•°75
B. Ina
ienuate cardiovascular comi
'icnsation*
Horizontal
69
354
5'4
160
92
0. 140
0.63
.108
15 degree
62
34"
515
'74
77
0. 120
0.68
.090
25 degree
66
358
580
222
57
0. 1 10
°-59
■059
30 degree
54
293
425
132
45
0.130
0.68
.064
■ RPF = Cpah/Epah- ''Medullary flow: RPF — Cpah- cPa in Part B calculated from diastolic pressure + '3 pulse
pressure. dTrue FF == Ci„/RPF. 'Hemat. 40% (assumed); RK = (Pa - PV)/RBF, in mm Hg/ml/min. '[After
VVerko et at. (334).] «[After Brodwall (36).]
supporting the notion of differential blood flow to
cortex and medulla.
VVerko et al. (334) found little increase in the renal
A-V oxygen difference during tilting, so that renal
oxygen consumption tended to decrease. The subject
in table 14 A showed a decrease from 25 ml per min
to 10 ml per min at a 45° tilt, but this was the greatest
change in the series. The data thus support the con-
cept of the flow-limited nature of the renal circula-
tion.
Renal Hypoxia and Ischemia
hypoxia. Analysis of the response of renal blood flow-
to hypoxic states has been complicated by the varied
techniques and experimental conditions employed,
and the varying degrees of hypoxia. Thus, the whole
organism may be expected to show a different re-
sponse than the isolated organ to hypoxia (278).
Ischemia presents not only the problem of reduced
oxygen supply, but also of accumulation of metabolites
in the organ.
Caldwell et al. (47) gave to subjects oxygen intakes
as low as 9.3 per cent for periods of 5 to 17 min. No
consistent effects in CPAh and CTn were observed;
blood pressure only occasionally showed a slight
tendency to increase. Berger et al. (15) experimentally
reduced arterial oxygen tension from ca. 97 mm Hg
to ca. 50 mm Hg in humans. CPAh increased by an
average of 13 per cent ( — 5.2 to +22.8) with no sig-
nificant change in CIn or blood pressure. Acute ex-
posure of dogs to simulated altitudes of 18,000 feet
(79.4 mm Hg partial pressure of oxygen) and 24,000
feet (61 .6 mm Hg oxygen) caused an increase in renal
plasma flow at 18,000 feet, but this generally de-
creased below ground level values at 24,cco fee
(193). The seal, when subjected to 10 per cent oxygen
or asphyxia, reacts with a significant decrease in RPF
and GFR (192), accompanied by marked slowing of
the heart and increases in blood pressure. It would
appear that mild hypoxic states unaccompanied by
systemic reflex vascular alterations manifest a slight
renal hyperemia, but more severe hypoxic states or
asphyxia trigger vasoconstrictor reflexes in which the
kidney participates. Other experimental approaches
confirm this. Bing & Knudsen (19) by direct observa-
tion of the mouse kidney noted blanching (cortical
ischemia) to occur in the range of 6.5 to 10.5 vol per
cent oxygen in the inspired air (average 7 vol %),
with arterial oxygen tension at 42 per cent of control.
It was concluded that a reflex was involved, with
centers sensitive to hypoxia in the spinal cord, for it
persisted after cord section at T4, but was lost with
renal denervation. It is likely that the medullary
centers activated by impulses from the chemoreceptors
in the carotid sinus and aortic arch would also par-
ticipate.
acute renal ischemia. Numerous experiments have
been made in which the duration of ischemia has lasted
from several minutes to 3 and 4 hours. Short-term
occlusion results in rather transient hemodynamic
effects, while longer ischemia involves varying degrees
of tubular damage which must be taken into account
in interpreting clearance data. Short bouts (10-20
min) produce small but variable effects with rapid
recovery. No change in blood flow (246), or small de-
creases (76, 269), or even increases (87) have been
reported. Since brief ischemia has been reported to
cause reactive hyperemia (262, 293) and when clear-
1504
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
anccs arc used a washout •"overshoot" could occur
(62, 298), some reasons for the variability can be per-
ceived.
Thirty minutes of ischemia (unilateral) in dogs (76)
resulted in a fall in CPAH and C,„ to less than 50 per
cent of control, but recovery occurred in 30 min. No
significant change in EFAS occurred, indicating con-
tinued authenticity of the clearance. Following 45
min of ischemia, CPAH and CIn were still much reduced
1 35 min after release, but now the validity of the
clearance could be questioned because of possible
tubular damage. Unquestionably, 2 hours of ischemia
results in marked and persistent reduction of £pAH
(control 0.90-0.94, to o. 1 1-0.43) (246).
The mechanism of the persistent ischemia is of
great interest. One speculation is that prolonged
clamping of the artery results in the excessive produc-
tion and accumulation of pressor substances which act
locally (269). Other possible factors should be con-
sidered. A locally activated, persisting reflex, either
intrarenal (93) or caused by mechanical compression
of the arteries and intramural nerve fibers, could be
involved. It is noteworthy that renal artery blockade
does not have the same effect as venous blockade.
Neely & Turner (231) found that renal blood flow in
the dog kidney (direct method) after 1 hour of uni-
lateral occlusion of the artery was reduced to 44 per
cent of control immediately after release, but was re-
stored to 79 per cent 1 hour later. Venous occlusion
for 30 min resulted in a decrease to 58 per cent of con-
trol and blood flow remained at this value 1 hour
later. Combined occlusion of artery and vein also
resulted in poor recovery of flow. With venous
occlusion, a persistent weight increase occurred; but
with combined artery and vein occlusion the weight
was constant, so that congestion of the kidney did not
appear to be the answer. The possibility of intra-
vascular thrombosis was raised.
With prolonged ischemia, tubular damage, uremia,
and death in renal failure is the outcome. Hamilton
et al. (128) found that anesthetized dogs with the
right kidney previously removed uniformly survived
2 hours of clamping of the left renal artery, and some
survived ischemia of 3 to 4 hours. When the kidney
was cooled to 5 to 17 C, percentage survival was im-
proved even with longer periods of ischemia, because
of greatly reduced cellular metabolism (20).
The pattern of recovery of renal clearance following
2 hours of clamping of the remaining kidnev after
unilateral nephrectomy in the clog is nicely shown in
the work of Friedman et al. [table 15 (97)].
It can be seen that the clearances lose their validity
for measuring plasma flow because of tubular damage,
revealed by the low £PAh and reduced TmPAH. When
plasma flow is estimated by the Fick application, al-
though reduced to less than half of control 3 hours
after ischemia, blood flow is fairly well restored in 24
hours. The low CCr is probably the result of continued
back diffusion of creatinine, so that the FF has little
meaning for some time after ischemia.
Hypercapnia and Acidosis
Dowds et al. (75) studied the effects of progressive
hypercapnia in anesthetized dogs rebreathing from a
spirometer flushed with pure oxygen to prevent
hypoxia. During about 2.5 hours, the carbon dioxide
content of the inspired air increased to an average of
16.8 vol per cent (13.5-19.9). This was accompanied
by a marked increase in respiratory rate. Arterial
blood carbon dioxide increased from 35 vol per cent
to an average peak of 52.8 vol per cent. Heart rate
slowed and blood pressure declined about 10 per cent
below the control. In this range of carbon dioxide in-
crease, C'pah and C'cr did not change remarkably; if
anything they decreased with the fall in blood pres-
sure. Brooker et al. (37) subjected dogs to 30 per cent
carbon dioxide in oxygen for 30-min periods. All dogs
became acidotic, with decreased urinary output. Blood
flow decreased to an average of 45 per cent of control
(25 to 64 % ). Renal resistance increased from 0.68 to
1. 1 7 mm Hg per ml per min, despite a fall of blood
pressure to 93 per cent of control. In similar experi-
ments, Stone et al. (296) studied the effects on intact
and pharmacologically denervated kidneys of anes-
thetized dogs. With carbon dioxide inhalation, blood
table 15. Effects of 2-Hour Renal Ischemia
1 "rural
; t hours
24 hours
5-8 days
2 weeks
3 monlhs
CCr (ml/min)
59
1 -9
8.6
13-5
12.7
63.O
Cpah (ml/min)
182
5-
27.9
44-7
62.5
174.O
FF
Ep Ml
Gp mi Epah (ml/min)
0.320
0.66
276
0.470
0.048
108
0.400
0 . 1 42
196
0 . 209
0 . 263
170
0.203
0.327
191
O.366
O.84
208
TmpAH (mg/min)
.3.8
1 .4
3'7
3-23
6.0
[After Friedman el al . (97).]
THE RENAL CIRCULATION
I5°5
pH decreased from ca. 7.45 to 7.10. Respiration rate
tripled, but blood pressure fell slightly. Renal blood
flow showed an average reduction to 24.3 per cent of
control (range 11-45%) at tne er,d °f tne 30-min
inhalation period. This was accompanied by oliguria
or anuria. Denervation of the kidney apparently pre-
vented the marked decrease in flow observed in the
intact kidney and urine production continued. The
authors concluded that the increased renal vascular
resistance was a reflex component of a more gen-
eralized vasoconstrictor response to high carbon
dioxide. Franklin et al. (93) by the visual "blanching"
technique in rabbits inspiring gas mixtures of in-
creased carbon dioxide content (up to 25 %) saw
blanching (cortical ischemia) when the blood carbon
dioxide content had increased to 140 per cent of con-
trol. The response was abolished by nerve section,
confirming the reflex nature of the phenomenon.
Hvpercapnia is an important factor contributing to
the marked reduction in renal flow which results
during diffusion respiration (297). In this, with re-
spiratory arrest resulting from excess anesthetic action
or curare, oxygen is taken into the lungs by the con-
tinued removal of the gas from the alveoli by
hemoglobin uptake in the pulmonary circulation.
Breathing of pure oxygen for 1 hour prior to onset of
respiration arrest is essential ("denitrogenation").
Blood content of carbon dioxide rises progressively,
since it is not removed by the quiescent lungs. After
30 min of apnea, blood flow had decreased to 28 per
cent of the control. Blood pressure had fallen an aver-
age of 23 mm Hg during this time and renal resistance
increased by 230 per cent. In a denervated series
(nerve block), these changes in renal blood flow and
resistance were restored to the decreasing control
trend. Again, a central origin of the renal ischemia
was predicated. Bohr et al. (27), although demon-
strating a lessened trend for CPAH to decrease in the
denervated kidney, nevertheless observed significant
decreases (PAH to 38 % of control with blood pressure
decrease from 1 16-95 mm Hg). Therefore, circulatory
pressor substances must be released in greater amounts
to contribute to the vasoconstriction.
It seems reasonable to conclude that the reduction
in renal blood flow during hvpercapnia and acidosis
is centrally mediated. It is nevertheless surprising
that in none of these investigations was there recorded
an increase in blood pressure. From the reported
facts it would appear that the preponderant effect of
hvpercapnia and the accompanying acidemia was a
reflex increase in renal vascular resistance in the face
of an actual fall in arterial blood pressure, a con-
comitance of events difficult to reconcile. It may be
that anesthesia alters the normal response. Also, it
must be kept in mind that the direct peripheral vascu-
lar action of carbon dioxide is dilatory (e.g. on vessels
of skeletal muscle), which action may become pre-
ponderant. This does not preclude the possibility that
other tissues, such as the kidney, respond only by con-
striction.
Hemorrhagic Hypotension and Shock
HEMORRHAGE AND HEMORRHAGIC SHOCK. Acute
hemorrhage provokes responses in the renal circula-
tion which are typical of general compensatory mecha-
nisms set into play, viz. reflex vasoconstriction, and
shunting of blood to other tissues in order to com-
pensate for low blood flow. In the case of the kidney,
if blood loss is great enough, this means shutdown of
renal excretory function which, if prolonged, might
have serious consequences to the organism. Moreover,
a prolonged period of anoxic hypotension will impair
the function of the tubular epithelium, adding to the
problem of shock the probability of renal failure and
uremia.
Following acute hemorrhage, the kidney's circu-
lating autonomy aids in reestablishing flow. Heine-
mann et al. (136) bled anesthetized dogs 1.3 to 3.9
per cent of body weight; blood pressure fell by 5 to 59
mm Hg to levels 91 to 51 per cent of mean control
values. Renal blood flow (based on CPAh) decreased
more than the blood pressure, signifying vasoconstric-
tion. In four representative experiments, RBF de-
clined from 16. 1 ( 1 5.6—16.6) to 4.5 (0.3-10.5) ml per
kg per min. In three animals, while hypovolemia and
hypotension were maintained, blood flow was re-
stored autonomously to 16.5 (11. 7-19. 5) ml per kg
per min in 25 to 70 min. Goodyer & Jaeger (107)
found similar responses to moderate hemorrhage in
anesthetized dogs, followed by restoration of flow.
Denervated kidneys showed a lesser decrease after
hemorrhage than the paired intact kidney, but both
showed compensatory restoration, indicating that the
autonomy is intrinsic. Dibenzyline selectively injected
into one renal artery reduced its responsiveness to
hemorrhage compared to the control side (129).
Phillips et al. (247) found that rapid, massive hemor-
rhage in anesthetized dogs was accompanied by almost
complete cessation of RBF (CPah/£pah)- If hemor-
rhage was not too great, arterial pressure rose as the
result of extrarenal vasoconstriction, and renal blood
flow was restored but at a figure less than before
hemorrhage. In the recovery phase, the kidney ap-
1506
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
peared to be favored at the expense of the rest of the
circulation. The cycle could be repeated with addi-
tional hemorrhages. Ultimately, peripheral vasocon-
striction failed to maintain an adequate systemic
pressure, and renal plasma flow and glomerular filtra-
tion fell to low values. It was inferred that at this
stage afferent arteriolar vasoconstriction closed the
renal circulation in an effort to maintain circulation
to vital centers. In view of the possible unreliability
of the Fick method (CPAB/EPAB) during hypotension
and shock according to Balint & Fekete (8), this in-
terpretation may not be warranted. The indirect
method gave them much lower values than did the
direct, giving the erroneous impression of marked in-
crease in renal vascular resistance and marked de-
crease in the renal fraction of the cardiac output.
Corcoran & Page (62) induced hemorrhagic shock
in anesthetized dogs by controlled bleeding to main-
tain pressure at about 60 mm Hg for 70 min, followed
by transfusion. This cycle was repeated two or three
times. Clearances (CD or Cin) decreased to zero or
nearly so during hypotension. Repeated reduction
and restoration of blood pressure led ultimately to a
permanent reduction in renal blood flow. Since this
phenomenon occurred in dogs with denervated kid-
neys, it was suggested that the reduced function was
the result of appearance in the blood of vasoconstrictor
substances. Again, caution must be exercised in inter-
pretation because of the unreliability of indirect
methods.
Selkurt (270) noted the persistence of a small flow
of blood through the kidneys of anesthetized dogs
(direct venous outflow) subjected to hemorrhage of
2 to 5 per cent of body weight to bring blood pressure
to consecutive 60 and 40 mm Hg pressure stages,
held 90 and 45 min, respectively. Clearances could
not be followed because of extreme oliguria and
anuria. Renal vascular resistance (calculated from
direct flow) was not excessively increased early in the
60 mm Hg pressure stage period [experimental/
control = avg. 1.15 (0.73-2.07)], possibly because of
operation of renal autonomy. But at the end of the
period at 40 mm, the ratio averaged 3.04 (1.53-6. 15).
Enhanced vasoconstrictor activity as the result of
additional hemorrhage, plus increased release of cir-
culating pressor materials, such as catecholamines
(325) or serotonin (63), could have accounted for the
enhanced vasoconstriction.
On transfusion, renal vascular resistance returned
almost to control value, but increased again second-
arily as normovolemic shock developed. Terminally,
this was as great as 4.7 times the control Because of
variable degrees of tubular damage, the clearances of
PAH and creatinine could not be relied upon for
accurate measurement of plasma flow and GFR after
transfusion.
tourniquet and traumatic shock. Allowing for the
time factor and sequences of events, the changes in-
duced in renal function by tourniquet shock in dogs
are much the same as observed after hemorrhage.
Corcoran el al. (63) applied leg tourniquets tight
enough to block venous return but not necessarily
arterial inflow. RPF and GFR progressively fell until
at 90 min they were 25 per cent of control. Blood
pressure decreased about 25 per cent, with increased
hematocrit ratio. On release of the tourniquet which
had been in place for 200 min, blood flow might re-
cover for a time, then decline again if shock ensued.
With development of shock ED, which had remained
normal, declined to 0.50. Because flow decreased
somewhat, even in the denervated kidney, the vaso-
constriction must have been partly of humoral origin.
Increased release of serotonin was considered as a
possibility. Catecholamine output could have been
enhanced.
The effects of tourniquet application and limb
crushing in anesthetized dogs was studied by Eggleton
et al. (80), the tourniquets being left in place for 4 to 5
hours. On release, blood pressure fell and urine flow-
ceased. With gum acacia infusion to restore pressure,
the creatinine clearance still remained about one-
third of control. The basis for the reduced creatinine
clearance was not satisfactorily explained, but afferent
impulses to vasomotor centers, and release of humoral
substances which might be vasoconstrictor to the
kidney could be considered as possibilities. Back
diffusion through damaged tubules did not appear
likely under the circumstances of their experiment.
Fleming & Bigelow (88) made direct visual observa-
tions of cortical blood flow of kidneys with crushing
injury to the hind legs. They saw agglutination of the
cells in vessels of 20 to 30 ju size as large clumps In
capillaries, the clumps were seen intermittently ob-
structing the lumen, often causing stasis or even ap-
parent reversal of flow.
traumatic injuries in man. Lauson el al. (175) re-
ported renal function studies in shock of varied etiolo-
gies in man but mostly resulting from hemorrhage
and skeletal trauma. Keeping in mind the limitations
of the clearance methods for measurement of blood
flow and filtration rate under the unfavorable condi-
tions that apply in shock, general conclusions emerge
THE RENAL CIRCULATION
I507
which are consonant with the findings in the experi-
mental animal. Filtration rate rapidly declined at a
mean arterial pressure of 60 mm, and often ceased en-
tirely between 40 and 50 mm Hg. Estimated blood
flow was drastically reduced, and calculated renal re-
sistance was high. The renal fraction of the cardiac
output was much below the normal range, but the
criticism of Balint & Fekete (8) must be kept in mind.
Despite this, and its autonomy under or during
such circumstances, it is tempting to conclude that the
renal circulation in man is subordinate to the welfare
of the body as a whole.
CONCLUDING REMARKS
The kidney is an organ characterized by a high
volume of blood flow resulting in a narrow A-V oxy-
gen difference despite a high rate of oxygen utiliza-
tion. The A-V oxygen difference tends to remain
constant in the face of minor fluctuations in flow, but
at very low rates of flow, an increase in the A-V oxy-
gen difference has been observed by several investi-
gators. The remarkable autonomy of the renal
circulation may be an adaptation to insure steady de-
livery of oxygen to the renal tissue.
The constancy of flow appears to be desirable for
another reason. The countercurrent system for the
concentration and dilution of the urine operates
optimally with constant blood flow. When this has
been experimentally altered (e.g., increased flow
through the medullary circuits) the osmotic stratifica-
tion in the vasa recta and loop of Henle system has
been dissipated, and concentrating power impaired.
Indications are that a countercurrent multiplier
system for concentration of serum albumin exists in
the vasa recta, a mechanism which would aid inter-
stitial fluid uptake and removal into the systemic
circulation. The interesting interrelationship of the
vasa recta system to the loop of Henle system in the
role of water and salt absorption merits much further
study, particularly in the direction of quantitative
measurement of regional flow (cortical versus medul-
lary), and the factors which alter it.
REFERENCES
1. Aas, K., and E. Blegen. The effect of tetraethylammo-
nium bromide on the kidney. Lancet 1 : 999, 1949.
2. Arcadi, J. A., and F. Farman. Experimental studies
and clinical aspects of the renal circulation. J. Urol. 62 :
756. >9490-
3. Asheim, A., C. G. Helander, and F. Persson. Studies
on renal function in dogs. Extraction values for PAH
obtained by percutaneous catheterization and clearance
studies on single kidneys. Acta Physiol. Scand. 44: 103,
io958-
4. Astrom, A. A study of pressure-time curves obtained in
the occluded renal artery in cats at different venous pres-
sures. Acta Physiol. Scand. 49: 10, i960.
5. Aviado, D. M., Jr., A. L. Wnuck, and E. J. DeBeer.
The effects of sympathomimetic drugs on renal vessels.
J. Pharmacol. Exptl. Therap. 124: 238, 1958.
6. Baker, S. B. de C. The blood supply of the renal papilla.
Brit. J. Urol. 31 : 53, 1959.
7. Baker, W. P., and L. A. Woods. A study in the dog of
renal clearance of morphine and the effect of morphine
on PAH clearance. J. Pharmacol. Exptl. Therap. 120: 371,
'957-
8. Balint, P., and A. Fekete. Das Verhalten des Minuten-
volumens und der Nierendurchblutung bei stagnierender
Hypoxic Pfliigers Arch. ges. Physiol. 270: 575, i960.
9. Barac, G. Effet renal de la bradykinine chez le chien.
Compt. Rend. Soc. Biol. 151 : t 77 1 , 1957.
10. Barclay, J. A., W. T. Cooke, R. A. Kenney, and
M. E. Nutt. The effect of exercise on renal blood flow
in man. J. Physiol., London 104: 14P, 1946.
11. Barker, H. G., J. K. Clark, A. P. Crosley, Jr., and
A. J. Cummins. The effect of salt poor human albumin on
renal oxygen consumption in man. Am. J. Med. Sci.
218: 715, 1949.
12. Barrie, H. J., S. J. Klebanoff, and G. W. Cates.
Direct medullary arterioles and arteriovenous anasto-
moses in the arcuate sponges of the kidney. Lancet 258/1 :
23, i95°-
13. Bayliss, VV. M. On the local reactions of the arterial wall
to changes in internal pressure. J. Physiol., London 28: 220,
1902.
14. Bearn, A. G., B. Billing, O. G. Edholm, and S.
Sherlock. Hepatic blood flow arid carbohydrate changes
during fainting. J. Physiol., London 1 15: 442, 1951.
15. Berger, E. Y., M. Gladstone, and S. A. Horwitz.
The effect of anoxic anoxia in the human kidney. J. Clin.
Invest. 28 : 648, 1 949.
16. Bercstrom, J., H. Bucht, and B. Josephson. Determina-
tion of renal blood flow in man by means of the radio-
active Diodrast and renal vein catheterization. Scand.
J.Clin. & Lab. Invest. II: 71, 1959.
17. Berne, R. M. Hemodynamics and sodium excretion of
denervated kidney of anesthetized and unanesthetized
dog. Am. J. Physiol. 171: 148, 1952.
18. Bialestock, D. The extra-glomerular arterial circulation
of the renal tubules. Anal. Record 129: 53, 1957.
19. Bing, J., and P. J. Knudsen. Effects of severe hypoxia, or
fright on renal blood flow on normal and shocked mice.
Acta Pathol. Microbiol. Scand. 35: 39, 1 951 .
20. Birkeland, S., A. Vogt, J. Krog, and C. Semb. Renal
1508
I! Willi! 1! Ik ill I'll1! -,11 ll ! II. \
CIRCULATION II
circulatory occlusion and local cooling. J. Appl. Physiol.
'4: "7. '959-
ai. Bishop, J. M., O. L. Wade, and K. W. Donald. Changes
in jugular and renal arterio-venous oxygen content differ-
ence during exercise in heart disease. Clin. Sci. 17: 611, 41.
1958.
22. Blaokmore, \V. A., V. E. Wilson, and T. R. Sherrod.
The effect of histamine on renal hemodynamics. J.
Pharmacol. Expll . Therap. 109: 206, 1953.
23. Blake, W. D. Effect of exercise and emotional stress on 42.
renal hemodynamics, water and sodium excretion. Am. J.
Physiol. 165: 149, 1 95 1.
24. Block, M. A., K. G. Wakim, and F. C. Mann. Certain
features of the vascular beds of the cortico-mcdullary 43.
and medullary regions of the kidney. A. M. A. Arch.
Pathol. 53: 437, 1952.
25. Block, M. A., K. G. Wakim, and F. C. Mann. Circula- 44.
tion through the kidney during stimulation of the renal
nerves. Am. ./. Physiol. 169: 659, 1952. 45.
26. Boba, A., S. R. Powers, Jr., and A. A. Stein. Studies on
renal vasoconstrictor response. Anesthesiology 20: 268,
■959-
27. Bohr, V. C, R. J. Ralls, and R. E. Westermever.
Changes in renal function during induced apnea of diffu- 46.
sion respiration. .4m. J. Physiol. 194: 143, 1958.
28. Bounous, G., M. Onnis, and H. B. Shumacker. The 47.
abolition of renal autoregulation by renal decapsulation.
Surg. Gynecol. Obstet. 111:682, i960.
29. Boyer, C. C. The vascular pattern of the renal glomerulus 48.
as revealed by plastic reconstruction from several sec-
tions. .4/ia/. Record 1 25: 433, 1 956.
30. Bozlf.r, E. The response of smooth muscle to stretch. 49.
Am. J. Physiol. 149: 299, 1947.
31. Bradford, J. The innervation of the renal blood vessels.
J. Physiol., London 10: 358, 1889. 50.
32. Bradley, S. E., and G. P. Bradley. The effect of in-
creased intra-abdominal pressure in man. ./. Clin. Invest.
26: 1010, 1947. 5'-
33. Bradley, S. E., J. J. Curry, and G. P. Bradley. Renal
extraction of /i-aminohippurate in normal subjects and
in essential hypertension and chronic diffuse glomerulo- 52.
nephritis. Federation Proc. 6: 79, 1947.
34. Brandfonbrenner, M., and H. M. Geller. Effect of
Dibenamine on renal blood flow in hemorrhagic shock.
Am. J. Physiol. 171: 482, 1952. 53.
35. Bricker, N. S., R. A. Straffon, E. P. Mahoney, and
J. P. Merrill. The functional capacity of the kidney
denervated by autotransplantation. J. Clin. Invest. 37 :
185, 1958. 54.
36. BrodwaLL, E. K. A study of renal function in orthostatic
hypotension. Circulation 21 : 38, i960. 55-
37. Brooker, W. J., J. S. Ansell, and E. B. Brown, Jr.
Effect of respiratory acidosis on renal blood flow. Surg.
Forum 10: 869, i960. 56.
38. Brull, L., D. Louis-Bar, and H. Lybeck. The action
of chronic denervation and of the use of ganglioplegic
and sympatholytic agents on the barosthetic device of
the renal artery. Acta Physiol. Scand. 34: 175, 1955. 57.
39. Brun, C, E. O. E. Knudson, and F. Raaschou. The
influence of posture on kidney function. Acta Med. Scand.
122:315, 1945.
40. Brun, C, C. Crone, H. G. Davidsen, J. Fabricius, 58.
A. T. Hansen, N. A. Lassen, and O. Munch. Renal
blood flow in anuric human subjects determined by the
use of radioactive krypton85. Proc. Soc. Expll. Biol. Med.
89:687, '955-
Brun, C, C. Crone, H. G. Davidsen, J. Fabricius,
A. T. Hansen, N. A. Lassen, and O. Munch. Renal
interstitial pressure in normal and in anuric man: Based
on wedged renal vein pressure. Proc. Soc. Expll. Biol. Med.
91 : '99. '956-
Bucht, H., J. Ek, H. Eliasch, A. Holmgren, and B.
Josephson. The effect of exercise in the recumbent posi-
tion on the renal circulation and sodium excretion in the
normal individual. Acta Physiol. Scand. 28: 95, 1953.
Bulbring, E. Correlation between membrane potential,
spike discharge, and tension of smooth muscle. J. Physiol.,
London I 28 : 200, 1 955.
Burgi, S. Zur Physiologie und Pharmakologie der iiber-
lebenden Arterie. Hcli. Phiswl Acta 2: 345, 1944.
Burnett, C H., E. L. Bloomberg, G. Shortz, D. W.
Compton, and H. K. Beecher. A comparison of the
effect of ether and cyclopropane anesthesia on the renal
function in man. J. Pharmacol. Expll. Therap. 96: 380,
'949-
Burton, A. C. On the physical equilibrium of small
blood vessels. Am. J. Physiol. 164: 319, 1951.
Caldwell, F. T., D. Rolf, and H. L. White. Effects of
acute hypoxia on renal circulation in man. J. Appl.
Physiol. 1 : 597, 1949.
Cargill, W. H. Effect of I.V. administration of human
serum albumin on renal function. Proc. Soc. Exptl. Biol.
Med. 68: 189, 1948.
Cargill, W. The measurement of tubular plasma flow
in the normal and diseased kidney. J. Clin. Invest. 28:
533. 1949-
Carlin, M. R., C. B. Mueller, and H. L. White.
Effects of exercise on renal blood flow and sodium excre-
tion in dogs. J. Appl. Physiol. 3: 291, 1950.
Carstensen, G., and F. Holle. Anderungen der intra-
renalin Hamodynamik nach lumbarer Sympathektomie.
Arch. Klin. Clur. Langenbecks 290: 440, 1959.
Chapman, C B., A. Henschel, J. Minckler, A. Fors-
gren, and A. Keys. The effect of exercise on renal plasma
flow in normal male subjects. J. Clin. Invest. 27: 639,
1948.
Chapman, C. B., A. Henschel, and A. Forsgren. Renal
plasma flow during moderate exercise of several hours
duration in normal male subjects. Proc. Soc. Exptl. Biol.
Med. 69: 170, 1948.
Christensen, G. C. Circulation of blood through the
canine kidney. -4m. J. Vet. Research 13: 236, 1952.
Christensen, K., E. Lewis, and A. Kuntz. Innervation
of the renal blood vessels in the cat. J. Comp. Xeurol. 95:
373. 195>-
Coller, F. A., V. L. Rees, K. N. Campbell, V. L. Iob,
and C. A. Moyer. Effect of ether and cyclopropane
anesthesia upon renal function in man. Ann. Surg. 118:
7'7, 1943-
Conn, H. L., Jr., and K. Markley. Simultaneous com-
parison of renal blood How as measured by the Fick
principle and the bubble flow meter. Am. J. Physiol. 160:
547> i95°-
Conn, H. L., Jr., W. Anderson, and S. Avena. Gas
THE RENAL CIRCULA I [( IN
I509
diffusion technique for measurement of renal blood How
with special reference to the intact anuric subject. J. Appl.
Physiol. 5:683, 1953.
59. Corcoran, A. C, H. VV. Smith, and I. H. Page. The
removal of Diodrast from the blood of the dog's explanted
kidney. Am. J. Physiol. 143: 108, 1941.
60. Corcoran, A. C, J. S. Browning, and I. H. Page.
Renal hemodynamics in orthostatic hypotension. J. Am.
Med. Assoc. 119: 792, 1942.
61. Corcoran, A. C, and I. H. Page. Effects of anesthetic
dosage of pentobarbital sodium on renal function and
blood pressure in dogs. Am. J. Physiol. 140: 234, 1943.
62. Corcoran, A. C, and I. H. Page. Effects of hypotension
due to hemorrhage and blood transfusion on renal func-
tion in dogs. J. Expll. Med. 78: 205, 1943.
63. Corcoran, A. C, R. D. Taylor, and I. H. Page. Imme-
diate effects on renal function of the onset of shock due to
partially occluding limb tourniquets. Ann. Surg. 118:
871, '943-
64. Corcoran, A. C, G. M. C. Masson, F. del Greco,
and I. H. Page. 5-Hydroxy-tryptamine (serotonin): Its
lack of specific renal action. Arch, intern, pharmacody-
namic 97: 483, 1954.
65. Cort, J. H. Post -traumatic anuria. Am. J. Physiol. 164:
686, 1 95 1 .
66. Cort, J. H. Effect of nervous stimulation on the arterio-
venous oxygen and carbon dioxide difference across the
kidney. Xature 171: 784, 1 953
67. Craig, F. N., F. E. Visscher, and C. R. Houck. Renal
function in dogs under ether or cyclopropane anesthesia.
Am. J. Physiol. 143: 108, 1 945.
68. Crosley, A. P., Jr., J. F. Brown, J. H. Huston, D. A.
Emanuel, H. Tuchman, C. Castillo, and G. G. Rowe.
The adaptation of the nitrous oxide method to the deter-
mination of renal blood flow and in vivo renal weight in
man. J. Clin. Invest. 35: 1340, 1956.
6g. Daniel, P. M., C. N. Peabody, and M. M. L. Pritchard.
Observation on the circulation through the cortex and
medulla of the kidney. Quart. J. Expll. Physiol. 36: 199,
'951-
70. de la Pena, A., and F. de Castro. Structure and arrange-
ment of the "macula densa" in the human kidney. Urol.
Intern. 10: 171, 1 960.
71. De Langen, C. D. Intrarenal pressure. Acta Med. Scand.
'57: ^79- '957-
72. DeVVardener, H. E., and R. R. McSwiney. Renal
hemodynamics in vaso-vagal fainting due to hemorrhage.
Clin. Sci. 10: 209, 1 95 1.
73. DeWardener, H. E., and B. E. Miles. The effect of
hemorrhage on the circulatory autoregulation of the
dog's kidney perfused in situ. Clin. Sci. I 1 : 267, 1952.
74. Dole, V. P., K. Emerson, Jr., R. A. Philips,
P. Hamilton, and D. D. Van Slyke. The renal extrac-
tion of oxygen in experimental shock. Am. J. Physiol.
'45 337. '94°-
75. Dowds, E. G., E. W. Brickner, and E. E. Selkurt.
Renal response to hypercapnia. Proc. Soc. Expll. Biol.
Med. 84: 15, 1953.
76. Dutz, H., and G. Kreizschmar. Die Veranderungen
in der Funktion beider Nieren nach einseitiger voll-
standigcr Isthamie. Zeit. f.d. ges. expll. Med. 123: 497,
'954
77. Ebner, C. M., and C. Y. Morita. The effect of chlori-
sondamine on renal hemodynamics in hypertensive pa-
tients. Am. J. Med. Set. 233: 424, 1957.
78. Edelman, I. S., B. W. Zweifach, D. J. \V. Escher, J.
Grossman, R. Mokotoff, R. E. Weston, L. Leiter,
and E. Shorr. Studies on VED and VDM in blood in
relation to renal hemodynamics and renal oxygen extrac-
tion in congestive heart failure. J. Clin. Imest. 29: 925,
I95°-
79. Edwards, J. G. Efferent arterioles of glomeruli in the
juxtamedullary zone of human kidney. Anal. Record 125:
5a«. '956.
80. Eggleton, M. G., K. C. Richardson, H. O. Schild,
and F. R. VVinton. Renal damage due to crush injury
and ischemia of the limbs of the anesthetized dog. Quart.
J. Expll. Physiol. 32: 89, 1944.
81. Eicholtz, F., R. Taugner, and W. Braun. Untersuch-
ungen zur Behandlung Renaler Ischamien. Arch, intern.
pharmacodynamic 98: ! 18, 1954.
82. Elias, H., A. Hossman, I. B. Barth, and A. Solmor.
Blood flow in the renal glomerulus. J. Urol. 83: 790, i960.
83. Emanuel, D. A., J. Scott, R. Collins, and F. J. Haddy.
Local effect of serotonin on renal vascular resistance and
renal flow rate. Am. J. Physiol. 196: 1 122, 1959.
84. Emery, E. VV., A. H. Gowenlock, A. G. Riddell, and
D. A. K. Black. Intrarenal variations in haematocrit.
Clin. Sci. 18: 205, 1959.
85. Enger, R., F. Linder, and H. Sarre. Die VVirkung
quantitativ abgestufter Drosselung der Nierendurch-
blutung auf den Blutdruck. Z. ges. Exptl. Med. 104: 1,
I938.
86. Etteldorf, J. N., J. D. Smith, C. P. Tharp, and A. H.
Tuttle. Hydralazine in nephritic and normal children.
Am. J. Diseases Children 89: 451, 1956.
87. Fajers, C. M. On the effect of brief unilateral renal
ischemia. Acta Pathol. Microbiol. Scand. Suppl. 106, 1955.
88. Fleming, J. F. R., and VV. G. Bigelow. Microscopic
observations on the living mammalian kidney: The effect
of crush injuries, shock and adrenalin on the cortical
blood flow. Surgery 30 : 994, 1 95 1 .
89. Folkow, B. Intravascular pressure as a factor regulating
the tone of the small vessels. Acta Physiol. Scand. 1 7 : 289,
'949-
90. Folkow, B. A study of the factors influencing the tone of
denervated blood vessels perfused at various pressures.
Ada Physiol. Scand. 27: 99, 1952.
91. Forster, R. P., and J. P. Maes. Effect of experimental
neurogenic hypertension on renal blood flow and glomer-
ular filtration rates in intact denervated kidneys of unanes-
thetized rabbits with adrenal glands demedullated. Am. J.
Physiol. 150: 534, 1947.
92. Franklin, K. J., L. E. McGee, and E. Ullman. Anoxic
diversion of the renal cortical blood flow. Proc. Soc. Expll.
Biol. Med. 71 : 339, 1949.
93. Franklin, K. J., L. E. McGee, and E. A. Ullman.
Effects of severe asphyxia on the kidney and urine flow.
J. Physiol., London 112: 43, 1951.
g4. Freeman, O. VV., G. VV. Mitchell, J. S. Wilson, F. VV.
Fitzhugh, and A. J. Merrill. Renal hemodynamics,
sodium and water extraction in supine exercising normal
and cardiac patients. J. Clin. Imest. 34: 1109, 1955.
1510
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
95. Fregler, G. Measurement of renal blood flow and heat
production. Arch, intern, physiol. el biochem. 66: 662, 1958.
96. Frey, E. Der Mechanismus der Harneindickung und
Harnverdiinnung. Arch, expll. Pathol. Pharmakol. 177: 134,
1934-
97. Friedman, S. M., R. L. Johnson, and C. L. Friedman.
The pattern of recovery of renal function following renal
artery occlusion in the dog. Circulation Research 2: 231,
'954-
98. Garber, B. S., F. W. McCoy, E. R. Hayes, and B. H.
Marks. Pharmacological studies on the renal juxta-
glomerular apparatus. Arch, intern, pharmacodynamic 121 :
275. !959-
99. Gibson, J. G., A. M. Seligman, W. C. Peacock, J. C.
Aub, J. Fine, and R. D. Evans. The distribution of red
cells and plasma in large and minute vessels of the normal
dog, determined by radioactive isotopes of iron and
iodine. J. Clin. Invest. 25 : 848, 1 946.
100. Giebisch, G., H. D. Lauson, and R. F. Pitts. Renal
excretion and volume of distribution of various dextrans.
Am. J. Physiol. 178: 168, 1954.
101. Gjorup, S., and T. Hilden. The effect of hydralazine
(Apresoline) in kidney function and sodium excretion.
Scand. J. Clin. & Lab. Invest. 8: 273, 1956.
102. Glaser, H, D. Laszlo, and A. Schurmeyer. Uber die
Durchblutungsregulation der Niere. Arch, expll. Pathol.
Pharmakol. 167: 292, 1932.
103. Glauser, K. F., and E. E. Selkurt. Effect of barbiturates
on renal function in the dog. Am. J. Physiol. 168: 469,
I952-
104. Goldring, W., and H. Chasis. Sympathectomy and uni-
lateral nephrectomy in the treatment of hypertensive
disease. Med. Clin. North Am. p. 751, May, 1949.
105. Gomez, D. M. Evaluation of renal resistances, with special
reference to changes in essential hypertension. J. Clin.
Invest. 30: 1 143, 1 95 1.
106. Goodwin, W. E., and J. J. Kaufman. Renal lymphatics:
II: Preliminary experiments. J. Urol. 76: 702, 1956.
107. Goodyer, A. V. N., and C. A. Jaeger. Renal response
to non-shocking hemorrhage: Role of the autonomic
nervous system and of the renal circulation. Am. J.
Physiol. 180: 69, 1955.
108. Goodyer, A. V. N., L. R. Mattie, and A. Chetrick.
Renal response to non-shocking hemorrhage. Sodium
retention at constant perfusion pressure. Proc. Soc. Expll.
Biol. Med. 97: 422, 1958.
109. Goodyer, A. V. N., L. R. Mattie, and A. Chetrick.
Renal response to non-shocking hemorrhage: The role
of intrarenal shunt. Am. J. Physiol. 193: 360, 1958.
no. Goormagtigh, N. The renal arteriolar changes in the
anuric crush syndrome. Am. J. Pathol. 23: 513, 1947.
ill. Gottschalk, C. W. A comparative study of renal inter-
stitial pressure. Am. J. Physiol. 169: 180, 1952.
112. Gottschalk, C. W., and M. Mylle. Micropuncture
study of pressures in proximal tubules and peritubular
capillaries of the rat kidney and their relation to ureteral
and renal venous pressures. Am. J. Physiol. 185: 430, 1956.
113. Gottschalk, C. W., and M Mylle. Micropuncture
study of the mammalian urinary concentrating mech-
anism : Evidence for the countercurrent hypothesis.
Am. J. Physiol. 196: 927, 1959.
114. Green, H. D., and J. H. Kepchar. Control of peripheral
resistance in major systemic vascular beds. Physiol. Revs.
39:6l7> >959-
115. Grupp, G., and K Hierholzer. Der 02 — Verbrauch
von Nierengewebe verschiedener Zonen. Z. Biol. 109:
'97. '957-
116. Grupp, G., and K. Heyn. Der Warmererlust uber die
Oberflache der Niere. Z. Biol. 1 1 o : 476, 1 958.
1 17. Grupp, G., and S. J. Janssen. Untersuchungen uber die
Warmebildung der Niere. Pftiigers Arch. ges. Physiol. 267:
58, 1958.
118. Grupp, G., and H. Heimpel. Zum Problem der "reak-
tiven Hyperamie" der Niere. Pflugers Arch. ges. Physiol.
267: 426, 1958.
119. Grupp, G. Das Verhalten der Selbsteuerung des Nieren-
kreislaufs und der VVarmbildung der Niere auf Erhohung
des Venen Druckes. Z. ges. expll. Med. 131 : 174, 1959.
120. Grupp, G., H. Heimpei , and K. Hierholzer. Uber die
autoregulation der Nierendurchblutung. Pflugers Arch. ges.
Physiol. 269: 149, 1959.
121. Grupp, G. Uber den Einfluss von Narcotica and vaso-
konstriktorisch wirkenden Pharmaka auf die Autoregula-
tion der Nierendurchblutung. Arch, expll. Pathol. Pharma-
col. 235: 261, 1959.
122. Haddy, F. J. Effect of elevation of intraluminal pressure
on renal vascular resistance. Circulation Research 4: 659,
1956.
123. Haddy, F. J., J. Scott, M. Fleischman, and D. Emanuel.
Effect of changes in renal venous pressure upon renal
vascular resistance, urine and lymph flow rates. Am. J.
Physiol. 195: 97, 1958.
124. Haddy, F. J, J. Scott, M. Fleischman, and D. Emanuel.
Effect of change in flow rate upon renal vascular resist-
ance. Am. J. Physiol. 195: ill, 1958.
125. Hall, V. Further studies of the normal structure of the
renal glomerulus. Proc. Sixth Ann. ConJ. Nephrotic Syndrome.
New York: National Nephrosis Foundation, 1954, pp.
i-39-
1 26. Hall, V. The protoplasmic basis of glomerular filtration.
Am. Heart J. 54 : 1, 1957.
127. Hall, P. W\, and E. E. Selkurt. Effects of partial graded
venous obstruction on electrolyte clearance by the dog's
kidney. Am. J. Physiol. 164: 143, 1951.
128. Hamilton, P. B., R. A. Phillips, and A. Hiller. Dura-
tion of renal ischemia required to produce uremia. Am. J.
Physiol. 152:517, 1948.
129. Handley, C. A., and J. H. Mover. Unilateral renal
adrenergic blockade and the renal response to vasode-
pressor agents and to hemorrhage. J. Pharmacol. Exptl.
Therap. 1 1 2 ■ 1 , 1 954.
130. Hardin, R. A., J. B. Scott, and F. J. Haddy. Relation-
ship of pressure to blood flow in dog kidney. Am. J. Physiol.
199: 1 192, i960.
Hargitay, B., W. Kuhn, and H Wirz. Ein Modell-
versuch zum Pioblem der Harnkonzentrierung. Helv.
Physiol, et Pharmacol. Acta 9: C26, 1 951 .
Hargitay, B., and W. Kuhn. Das Multiplikations Prinzip
als Giundlage des Harnkonzentrierung in der Niere.
Z. Elecktrochem. 55: 539, 1951-
133. Harman, P. J., and H. Davies. Intrinsic nerves in the
mammalian kidney. J. Comp. Neurol. 89: 225, 1948.
134. Harpuder, K, M. Lowenthal, and S. Blatt. Periph-
»3'-
132.
THE RENAL CIRCULATION
IjII
eral and visceral vascular effects of exercise in the ere< I
posture. J. Appl. Physiol. II: 185, 1957.
135. Hartman, H., S. L. 0rskov, and H. Rein. Die Gefas-
reaktionen der Niere in Verlaufe allgemeiner Kreislauf
Regulationsvorgange. Pfliigers Arch. ges. Physiol. 238 : 239,
1937-
136. Heinemann, H. O., C. M. Smvthe, and P. A. Marks.
Effect of hemorrhage on estimated hepatic blood flow
and renal blood flow in dogs. Am. J. Physiol. 174: 352,
■953-
137. Hemingway, A., and A. Schweitzer. The excretion of
diodone by the isolated perfused kidney. J. Physiol.,
London 1 02 : 49 1 , 1 944.
138. Herdman, J. P., and N. T. Jaco. The effect of renal
artery constriction on the renal blood How. Brit. J. Exptl.
Pathol. 31 : 806, 1950.
139. Hiatt, E. P. The effect of denervation on the filtration
rate and blood flow in dog kidneys rendered hyperemic
by the administration of pyrogen. Am. J. Physiol. 136:
38, 1942.
140. Hilger, H. H., J. D. Klumper, and K. J. Ullrich.
Wasserruckresorption und Ionentransport durch die
Sammelrohrzellen der Saugetierniere. Pfliigers Arch. ges.
Physiol. 267: 218, 1958.
141. Hinshaw, L. B., S. B. Day, and C. H. Carlson. Tissue
pressure and critical closing pressure in the dog kidney.
Am. J. Physiol. 196: 1132, 1959.
142. Hinshaw, L. B., S. B. Day, and C. H. Carlson. Tissue
pressure as a causal factor in the autoregulation of blood
flow in the isolated perfused kidney. Am. J. Physiol. 197:
3°9. '959-
143. Hinshaw, L. B., H. M. Ballin, S. B. Day, and C. H.
Carlson. Tissue pressure and autoregulation in the dex-
tran perfused kidney. Am. J. Physiol. 197: 853, 1959.
144. Hinshaw, L. B., and C. H. Carlson. Mechanism of auto-
regulation in isolated perfused kidney. Proc. Soc. Exptl.
Biol. Med. 103: 373, i960.
145. Hinshaw, L. B., R. D. Flaig, C. H. Carlson, and N K.
Thuong. Pre- and postglomerular resistance changes in
the isolated perfused kidney. Am. J. Physiol. 199: 923,
i960.
146. Hix, E. L. Uretero-renal reflex facilitating renal vaso-
constrictor response to emotional stress. Am. J. Physiol.
■92: '9'. '958-
147. Hoff, E. C, J. F. Kell, Jr., N. Hastings, D. M. Sholes,
and E. H. Gray. Vasomotor, cellular and functional
changes produced in the kidney by brain stimulation.
J. Neurophysiol. 14: 317, 1 951 .
148. Houck, C. R. Alteration of renal hemodynamics and
function in separate kidneys during stimulation of the
renal artery nerves in dogs. Am. J. Physiol. 167: 523, 1951.
149. Houck, C. R. Alterations in renal hemodynamics and
function during the intravenous injection of epinephrine
in the dog. Am. J. Physiol. 166: 649, 1951.
150. Insull, VV., Jr., I. G. Tillotson, and J. Hayman, Jr.
Distribution of blood in the rabbit's kidney. .4m. J. Physiol.
163:676, 1950.
151. Janssen, S., and G. Grupp. Undersuchungen iiber die
Temperaturverteilung in der Niere des Hundes. Arch,
exptl. Pathol. Pharmakol. 230: 245, 1957.
152. Josephson, B., L. Werko, and H. Bucht. Renal extrac-
tion of Diodrast in man. Scand. J. Clin. & Lab. Invest. 2:
■49. »95°-
153. Josephson, B., H. Bucht, J. Ek, and L Werko. Renal
extraction, its depression, and the tubular storage of PAH
in the healthy and the diseased human kidney. Scand. J.
Clin. & Lab. Invest. 4: 1, 1952.
154. Johnston, W. B. A reconstruction of a glomerulus of the
human kidney. Anat. Anz. 16: 260, 1899.
155. Judson, W. E., W. Hollander, J. D. Hatcher, and
M. H. Halperin. The effects of exercise on cardiovascular
and renal function in cardiac patients with and without
heart failure. J. Clin. Invest. 34: 1546, 1955.
156. Kahn, J. R., L. T. Skeggs, and N. P. Shumway. Studies
of the renal circulation. Circulation 1 : 445, 1 950.
157. Katz, Y. J. Some factors affecting renal lymphatic pres-
sure. Circulation Research 6: 452, 1958.
158. Kessler, R. H., O. P. A. Heidenreich, and R. F. Pitts.
Evaluation of the cell separation hypothesis of autoregula-
tion of renal blood flow and filtration rate : Glucose titra-
tions in normal and anemic dogs. .4m. J. Physiol. 191 : 501 ,
1957-
159. Kinter, W. B., and J. R. Pappenheimer. Renal extrac-
tion of PAH and Diodrast — I131 as a function of arterial
red cell concentration. Am. J. Physiol. 185: 391, 1956.
160. Kinter, W. B., and J. R. Pappenheimer. Role of red
blood corpuscles in regulation of renal blood flow and
glomerular filtration rate. Am. J. Physiol. 1 85 : 399, 1 956.
161. Knoche, H. Uber die feinere Innervation der Niere des
Menschen. Z. Zelljoisch. 36: 448, 1951.
162. Koester, H. L., J. C. Locke, and H. G. Swann. Effluent
constrictions in the renal vascular system. Texas Rpts.
Biol, and Med. 13: 251, 1955.
163. Kolff, W. J., I. H. Page, and A. C. Corcoran. Patho-
genesis of renoprival cardiovascular disease in dogs. .4m. J.
Physiol. 178: 237, 1954.
164. Kramer, K., and F. R. Winton. The influence of urea
and of change in arterial pressure on the 02 consumption
of the isolated kidney of the dog. J. Physiol., London 96:
87, '939-
165. Kramer, K., and K. J. Ullrich. GvSattingung und
Hb-Gehalt des Capillarblutcs der Nierenrinde. Pfliigers
Arch. ges. Physiol. 267: 251, 1958.
166. Kramer, K., K. Thurau, and P. Deetjen. Hamody-
namik des Nierenmarks: Capillare Passagezeit, Blutvolu-
men, Durchblutung, Gewebshamatokrit und 02-Ver-
brauch des Nierenmarks in situ. Pfliigers Arch. ges. Physiol.
270: 251, i960.
167. Kubicek, VV. G., F. J. Kottke, D. J. Laker, and M. B.
Visscher. Renal function during arterial hypertension
produced by chronic splanchnic nerve stimulation in the
dog. .4m. J. Physiol. 174: 397, 1953.
168. Kuhlgatz, G. Intrarenale Blutverteilung der Ratteniere
in Durst und Wasserversuchen. Pfliigers Arch. ges. Physiol.
256: 1, >952-
169. Kuhn, VV. Haarnadelgegenstromprinzip als Grundlage
der Harnkonzentvierung in der Niere. Klin. Wochschr.
37:70, 1959-
170. Kurtz, S. M., and J. F. A. McManus. A reconsideration
of the development, structure, and disease of the human
renal glomerulus. .4m. Heart J. 58: 357, 1959.
171. Lamdin, E. Mechanism of urinary concentration and
dilution. A.M. A. Arch. Internal Med. 103: 644, 1959.
'.:>'-
HANDBOOK OK PHYSIOLOGY
CIRCULATION II
i 7_- Langston, J. B., A. C. Guyton, and W. ). Gillespie, 192
Jr. Acute effect of changes in renal arterial pressure and
sympathetic blockade on kidney function. Am. J. Physiol.
'97:595. '959-
1 73. Langston, J. B., A. C. Guyton, and W. J. Gillespie, Jr. 193
Autoregulation absent in normal kidney but present after
renal damage. Am. J. Physiol. 199: 495, i960. 194
174. Lassen, N. A., J. B. Longley, and L. S. Lilienfield.
Concentration of albumin in the renal papillae. Science 195
128: 720, 1958.
175. Lauson, H. D., S. E. Bradley, and A. Cournand. The
renal circulation in shock. J. Clin. Invest. 23: 381, 1944. 196.
176. LeBrie, S. J., and H. S. Mayerson. Composition of
renal lymph and its significance. Proc. Soc. Exptl. Biol. Med.
100: 378, 1959. 197.
177. LeBrie, S. J., and H. S. Mayerson. Influence of elevated
venous pressure on the flow and composition of the lymph. 198.
Am. ./. Physiol. 198: 1037, i960.
178. Levy, M. N. Influence of variations in blood flow and of
dinitrophenol on renal oxygen consumption. Am. J. 199.
Physiol. 196: 937, 1959.
179. Levy, M. N., and G. Sauceda. Diffusion of oxygen from
arterial to venous segments of renal capillaries. Am. J.
Physiol. 196: 1336, 1959. 200.
180. Levy, S. E., R. A. Light, and A. Blalock. The blood
flow and 02 consumption of the kidney in renal hyper- 201.
tension. Am. J. Physiol. 122: 38, 1938.
181. Levy, S. E., and A. Blalock. The effects of unilateral
nephrectomy on renal blood flow and Ol, consumption 202.
of unanesthetized dogs. Am. J. Physiol. 122: 609, 1938.
182. Lewis, A. E., R. D. Goodman, and E. A. Schuck. Organ
blood volume measurement in normal rats. J. Lab. Clin.
Med. 39: 704, 1952. 203.
183. Lilienfield, L. S., J. C. Rose, and F. A. Porfido. Evi-
dence for a red cell shunting mechanism in the kidney.
Circulation Research 5: 64, 1957. 204.
184. Lilienfield, L. S., N. A. Lassen, and J. C. Rose. Diverse
distribution of red cells and plasma albumin in anatomical
regions of the kidney. J. Clin. Invest. 37: 912, 1958.
185. Lilienfield, L. S., and J. C. Rose. Effect of blood pres-
sure alterations on intrarenal red cell-plasma separation. 205.
J. Clin. Invest. 37:11 06, 1 958.
186. Lilienfield, L. S., J. C. Rose, and N. A. Lassen. Diverse
distribution of red cells and albumin in the dog kidney. 206.
Circulation Research 6:810, 1 958.
187. Livesay, W. R., and J. H. Mover. The renal hemody- 207.
namic effects of a xanthine compound, diethylaminoethyl
theophylline hydrochloride (Parephyllin). J. Pharmacol.
Exptl. Therap. 109: 123, 1953.
188. Lochner, \V , and Ochwadt, B. Uber die Beziehung 208.
zwischen arteriellen Druck, Durchblutung, Durchfluss-
zeit und Blutfullungan der isolierten Hundenniere. Pfliigers
Arch. ges. Physiol. 258: 275, 1954. 209.
189. Lofgren, F. The influence of ephedrine on the renal
circulation. Urol. Intern. 8: 142, 1959.
190. Longley, J. B., N. A. Lassen, and L. S. Lilienfield.
Tracer studies in renal medullary circulation. Federation 210.
Proc. 17: 99, 1958.
191. Longley, J. B., W. G. Bonfield, and D. C. Brindley.
Structure of the Rete Mirabile in the kidney of the rat
as seen with the electron microscope. ./. Biophys. Biochem. 211.
Cytol. 7 : 1 03, 1 960.
Lowrance, P. B., J. F. Nickel, C. McC. Smythe, and
S. E. Bradley. Comparison of the effect of anoxic anoxia
and apnea on renal function in the harbor seal. J. Cell-
ular Cornp. Physiol. 48: 35, 1956.
McDonald, R. K., and V. C. Kelley. Effects of altitude
anoxia on renal function. Am. J. Physiol. 1954: 193, 1948.
McManus, J. F. A. The juxtaglomerular apparatus.
Lancet 2: 394, 1942.
McManus, J. F. A. Apparent reversal of position of the
Golgi element in the renal tubules. Xalure 1952: 417,
1943-
McManus, J. F. A. Element in the cells of the first and
second convoluted tubules of the cat kidney. Quart. J.
Mnroscop. Sci. 85: 97, 1944.
Maluf, N. S. R. Role of the renal innervation in renal
tubular function. Am. J. Physiol. 139: 103, 1943.
Maxwell, M. H., E. S. Breed, and H. W. Smith. Sig-
nificance of renal juxtamedullary circulation in man.
Am. J. Med. 9: 216, 1950.
Maxwell, M. H., D. M. Gomez, A. P. Fishman, and
H. W. Smith. Effects of epinephrine and typhoid vaccine
on the segmental vascular resistance in the human kidney.
J. Pharmacol. Exptl. Therap. 109: 274, 1953.
Meehan, J. P. Central nervous system control of the renal
circulation. .4m. Heart J. 6: 318, i960.
Mehrizi, A., and VV. F. Hamilton. Effect of leverterenol
on renal blood flow and vascular volume in the dog. Am.
J. Physiol. 197: 1 1 15, 1959.
Merrill, A. J. Edema and decreased renal blood flow in
patients with chronic congestive heart failure, evidence of
"forward failure'" as the primary cause of edema. J. Clin.
Invest. 25: 389, 1946.
Merrill, A. J., and W. H. Cargill. Effect of exercise
on the renal plasma flow and filtration rate of normal and
cardiac subject. J. Clin. Invest. 27: 272, 1948.
Michie, A. J., N. Gimbel, C. Riegel, and M. Ragni.
Opening of intrarenal A-V shunts without cortical is-
chemia by sudden administration of salt-poor concen-
trated human serum albumin. J. Appl. Physiol. 3 : 472,
I951-
Miles, B. E., and H. E. DeWardener. Renal vasocon-
striction produced by ether and cyclopropane anesthesia.
J. Physiol., London 118: 140, 1952.
Miles, B. E., and H. E. DeWardener. Intrarenal pres-
sure. J. Physiol., London 123: 131, 1954.
Miles, B. E., M. G. Ventom, and H. E. DeWardener.
Observations on the mechanism of circulatory autoregu-
lation in the perfused dog's kidney. J. Physiol., London
123: '43> '954-
Mills, L. C, J. H. Mover, and J. M. Skelton. The effect
of norepinephrine and epinephrine on renal hemody-
namics. .4m. J. Med. Sci. 226: 653, 1953.
Mills, L. C, and J. H. Moyer. The acute effects of
hexamethonium on renal hemodynamics in normotensive
and hypertensive human subject. Am. ./. Med. Sci. 226:
1. >956-
Mills, L. C, J. H. Mover, and C. A. Handley. Effects
of various sympathicomimetic drugs on renal hemody-
namics in normotensive and hypotensive dogs. Am. J.
Physiol. 198: 1279, i960.
Mitchell, G. A. G. The nerve supply to the kidney.
Acta Anat. 10: 1, 1950.
THE RENAL CIRCULATION
1513
•212. Mitchell, G. A. G. The intrinsic renal nerves, Acta Anal.
13: ". I95L
213. Moberg, E. Anzahl und Grosse der Glomeruli renales
beim Menschen. Z. nul.roskop.-ana/. For sell. 18: 271, 1929.
214. Montague, F. E., and F. L. Wilson, Jr. Effect of epi-
nephrine on Na-hippurate extraction by the rabbit kid-
ney. Am. J. Physiol. 159: 581, 1949.
215. Montgomery, A. V., J. C. Davis, Jr. J. M. Prine, and
H. G. Swann. The intrarenal pressure. ./. Exptl. Med.
92:637. '95°-
216. Moore, R. A. The total number of glomeruli in the nor-
mal human kidney. Anal. Record 48: 153, 1931.
217. More, R. H., and G. L. Duff. The renal arterial vascu-
lature in man. Am. J. Pathol. 27: 95, 1951.
218. Morel, F. F., M. Guinnehault, and G. Amiel. Mise en
evidence d'un proces d'echange d'eau par contre-courant
dans les regions profondes du rein de hamster. Helv.
Physiol. Acta 18: 183, i960.
219. Morgan, D. P. Hematocrit value of blood expressed from
the isolated perfused kidney. Am. J. Physiol. 197:571, 1959.
220. Morris, G. C, J. H. Mover, H. B. Snyder, and B. W.
Haynes. Vascular dynamics in controlled hypertension.
Ann. Surg. 138: 706, 1953.
221. Morrison, D. M. A study of the renal circulation, with
special reference to its finer distribution. Am. J. Anal.
37:53, 1926.
222. Moyer, J. H., H. Conn, K. Markley, and C. F. Schmidt.
Hemodynamics of the renal circulation. Am. J. Physiol.
'95:582. '949-
223. Mover, J. H., H. Conn, K. Markley, and C. F. Schmidt.
Attempt to demonstrate vascular bypasses in the kidney
(the Trueta phenomenon). Am. J. Physiol. 161 : 250, 1950.
224. Mover, J. H., and C. A. Handley. Norepinephrine
and epinephrine effect on renal hemodynamics. Circula-
tion 5:91, 1952.
225. Mover, J. H., R. A. Huggins, C. A. Handley, and L. C.
Mills. Effect of the hexamethonium chloride on cardio-
vascular and renal hemodynamics and in electrolyte
excretion. J. Pharmacol. Exptl. Therap. 106: 157, 1952.
226. Moyer, J. H., C. A. Handley, and R. A. Huggins.
Cardiovascular and renal hemodynamic responses to
2-(N' p-tolyl-N'-m-hydroxy-phenylaminomethyl) Imid-
azoline hydrochloride (Regitine). J. Pharmacol. Exptl.
Therap. 108: 240, 1953.
227. Moyer, J. H., W. R. Livesay, and R. A. Seibert. The
effect of blood pressure reduction with Arfonad on renal
hemodynamics and the excretion of water and electro-
lytes. Am. Heart J. 48: 817, 1954.
228. Mover, J. H., R. McConn, and G. C. Morris. Effect
of controlled hypotension with Pendiomid (as used in
surgery) on renal hemodynamics and water and electro-
lyte excretion. Anesthesiology 16: 355, 1955.
229. Mukherje, S. R. Effect of bladder distention on arterial
blood pressure and renal circulation: role of the splanch-
nic and buffer nerves. J. Physiol., London 138: 307, 1957.
230. Murphy, J. J., M. K. Myint, W. H. Rattner, R. Klaus,
and J. Shallow. The lymphatic svstem of the kidney.
Proc. North Cent. Soc. Am. Urol. Assoc, p. 64. 1958.
231. Neely, W. A., and M. D. Turner. The effect of arterial,
venous, and arteriovenous occlusion on the renal blood
flow. Surg. Gynecol. Obslct. 1 08 : 669, 1 959.
232. Ochwadt, B., and J. Schmier. Uber Temperature und
Kreislaufsmessungen in verschiedenen Abschnitten der
Hundenniere. Pflugers Arch. ges. Physiol. 258: ig, 1954.
233. Ochwadt, B. Zur Selbststeuerung des Nieren-Kreislaufes.
Pflugers. Arch. ges. Physiol. 262: 207, 1956.
234. Ochwadt, B. Durchflusszeiten von Plasma und Erythro-
cytes intrarenal Hamatokrit und Widerstandregulation
der isolierten Niere. Pflugers Arch. ges. Physiol. 265: 7,
'957-
235. Ohler, W., O. Harth, and W. Kreienberg. Die
Abhangigkeit der Nierendurchblutung vom arteriellen
Blutdruck bei der Ratte. Pfliigers Arch. ges. Physiol. 269 :
274, '959-
236. Oliver, J. Architecture 0/ the Kidney in Chronic Brighl's
Disease. New York : Hoeber, 1 939.
237. Olsen, N. S., and H. A. Schroeder. Oxygen tension
and pH of the renal cortex in acute ischemia and chronic
hypertension. Am. J. Physiol. 163: 181, 1950.
238. Opitz, E., and D. H. Smyth. Nierendurchblutung bei
Reizung des Carotissinus. Pflugers Arch. ges. Physiol. 238:
&33> !937-
239. Page, I. H., and J. W. McCubbin. Renal vascular anil
systemic arterial pressure responses to nervous and chemi-
cal stimulation of the kidney. Am. J. Physiol. 173: 411,
'953-
240. Pappenheimer, J. R., and W. B. Kinter. Hematocrit
ratio of blood within mammalian kidney and its sig-
nificance for renal hemodynamics. Am. J. Physiol. 1 85 : 377,
1956.
241. Pappenheimer, J. R. Central control of renal circulation.
Physiol. Revs. 40: Suppl. 4, 35, i960.
242. Papper, E. M., and S. H. Ngai. Kidney function during
anesthesia. Ann. Rev. Med. 7: 213, 1956.
243. Parrish, A. E., J. Klek, and J. F. Fazekas. Renal and
cerebral hemodynamics with hypotension. Am. J. Med.
Sci- 233: 35, 1957.
244. Pease, D. C. Electron microscopy of the vascular bed of
the kidney cortex. Anal. Record 121 : 701, 1955.
245. Peter, K. Untersuchungen uber Bau und Entwicklung der
Niere. Jena: Fischer, 1927.
246. Phillips, R. A., and P. B. Hamilton. Effect of 20, 60
and 120 minutes of renal ischemia on glomerular and
tubular function. Am. J. Physiol. 152: 523, 1948.
247. Phillips, R. A., V. P. Dole, P. B. Hamilton, K. Emerson,
Jr., R. Archibald, and D. D. Van Slyke. Effects of
acute hemorrhage and traumatic shock on renal function
of dogs. Am. J. Physiol. 1 45 : 3 1 4, 1 946.
248. Pierce, E. C. Renal lymphatics. Anat. Record go: 315, 1944.
249. Piiper, J., and E. Schurmeyer. Uber den Nachweis
von arterio-Venosen Anastomosen in der Hundenniere.
Pflugers Arch. ges. Physiol. 261 : 543, 1955.
24ga.PoLOSA, C, and W. F. Hamilton. Blood volume and
intravascular hematocrit in different vascular beds. Am.
J. Physiol. 204: 903, 1963.
250. Radigan, L. R., and S. Robinson. Effects of environ-
mental heat stress and exercise on renal blood flow and
filtration rate. Am. J. Physiol. 159: 585, ig4g.
251. Rawson, A. J. Distribution of the lymphatics of the
human kidney as shown in a case of carcinomatous
permeation. A.M. A. Arch. Pathol. 47: 283, ig4g.
252. Rein, H. Vasomotorische Regulationen. Ergeb. Physiol.
32: 28, 1 931.
• j'4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
-■-,). Rennie, D. W., R. B. Reeves, and J. R. Pappenheimer.
Oxygen pressure in urine and its relation to intrarenal
blood flow. Am. J. Physiol. 195: 120, 1958.
2 -,4. Reubi, F. C, and H. A. Schroeder. Can vascular
shunting be induced in the kidney by vasoactive drugs?
J. Clin. Invest. 28: I 14, 1949.
255. Reubi, F. C, H. A. Schroeder, P. H. Futcher, and
C. Reubi. A discrepancy between renal extraction and
urinary excretion of various substances (para-amino-
hippurate, mannitol, creatinine, and thiosulphate) in man.
J. Appl. Physiol. 3: 63, 1950.
256. Reubi, F. Objections a la theorie de la separation intra-
renale des hematics et du plasma (Pappenheimer). Helv.
Med. Acta. 25: 516, 1958.
257. Rhoads, C. P., D. D. Van Slvke, A. Hiller, and A. S.
Alving. The effects of novocainization and total section
of nerves of the renal pedicle on renal blood flow and
function. Am. J. Physiol. 110:392, 1934.
258. Ritter, E. R. Pressure/flow relations in the kidney.
Alleged effects of pulse pressure. Am. J. Physiol. 168: 480,
'952-
259. Robinson, J. R. Reflections on Renal Function. Springfield,
111. : Thomas, 1954.
260. Rosen felt, S., and A. L. Sellers. Pressure-flow studies
in the isolated artificial heart-lung perfused mammalian
kidney. Am. J. Physiol. 199: 499, i960.
261. Rusznvak, I., M. Foldi, and G. Szabo. Lymphatics and
Lymph Circulation. New York : Pergamon Press, 1 960, pp.
1 14-120.
262. Sarre, H., and E. Ansorge. Uber die reaktive Hyperamie
der Niere. Pflugers Arch. ges. Physiol. 242 : 79, 1 939.
263. Schaefer, H. Discussion of "Central Control of Renal
Circulation." Physiol. Revs. 40: Suppl. 4, 45, i960.
264. Scher, A. M. Focal blood flow measurements in cortex
and medulla of the kidney. Am. J. Physiol. 167 : 539, 1951.
265. Scher, A. M. Mechanism of autoregulation of renal
blood flow. Nature, 184: Suppl. 17, 1322, 1959.
266. Schmidt, C. F., and M. M. Hayman. Lymph formation
in the dog kidney. Am. J. Physiol. 91 : 157, 1929.
267. Schmidt-Nielsen, B., and R. O'Dell. Effect of diet on
distribution of urea and electrolytes in the kidneys of
sheep. Am. J. Physiol. 197: 856, 1959.
268. Schwalb, J., J. Hernandez-Richter, E. Gross, and
K. Kotsianos. Vergleichende experimentelle Nieren-
durchblutung mit den Bubble Flow Meter und mit der
Clearance der />-aminohippursaure. Z. ges. Exptl. Med.
■3°:5°5. '958-
2fig. Selkurt, E. E. Comparison of renal clearances with
direct renal blood flow under control conditions and
following renal ischemia. Am. J. Physiol. 145: 376, 1946.
270. Selkurt, E. E. Renal blood flow and renal clearance
during hemorrhagic shock. Am. J. Physiol. 145: 699, 1946.
271. Selkurt, E. E. The relationship of renal blood flow to
effective arterial pressure in the intact kidney of the dog.
Am. J. Physiol. 147: 537, 1946.
272. Selkurt, E. E. Measurement of renal blood flow. Methods
in Medical Research. Chicago: Yr. Bk. Pub., 1: 191, 1948;
Ibid. 5: 150, 1952.
273. Selkurt, E. E., P. W. Hall, and M. P. Spencer. Re-
sponse of renal blood flow and clearance to graded partial
obstruction of the renal vein. Am. J. Physiol. 157:40, 1949.
274. Selkurt, E. E., P. W. Hall, and M. P. Spencer.
275-
276.
277.
278.
279.
280.
281.
283
284
285
286
287
288
289
290.
292
293
294'
Influence of graded arterial pressure decrement on renal
clearance of creatinine, />-aminohippurate and sodium.
Am. J. Physiol. 159: 369, 1949.
Selkurt, E. E. Physiologic mechanisms of the kidney in
relation to anesthesia. J. Am. Assoc. Nurse Anesthetists 17:
242, 1949-
Selkurt, E. E. Effect of pulse pressure and mean arterial
pressure modification in renal hemodynamics and the
handling of electrolytes and water. Circulation 4: 541, 1951.
Selkurt, E. E., M. Brandfonbrenner, and H. M.
Geller. Effects of ureteral pressure increases on renal
hemodynamics and the handling of electrolytes and water.
Am. J. Physiol. 170:61, 1952.
Selkurt, E. E. Influence of hypoxia on renal circulation
and on excretion of electrolytes and water. Am. J. Physiol.
172:700, 1953.
Selkurt, E. E. Sodium excretion by the mammalian
kidney. Physiol. Revs. 34: 287, 1954.
Selkurt, E. E. Der Nierenkreislauf. Klin. Wochschr. 33:
Jahr. 15/16, No. 15, 359, 1955.
Sf.mple, S. J. G, and H. E. DeWardener. Effect of
increased renal venous pressure on circulatory "auto-
regulation" of isolated dog kidneys. Circulation Research 7 :
643. '959-
\ Shipley, R. E., and R. S. Study. Changes in renal blood
flow, extraction of inulin, glomerular filtration rate, tissue
pressure, and urine flow with acute alterations of renal
arterial blood pressure. Am. J. Physiol. 167: 676, 1951.
Simkin, B., H. C. Bergman, H. Silver, and M. Prinz-
metal. Renal arteriovenous anastomoses in rabbits, dogs
and human subjects. Arch. Internal Med. 81 : 115, 1948.
Sirota, J. H. Carbontetrachloride poisoning in man. I.
The mechanisms of renal failure and recovery. J. Clin.
Invest. 28: 1412, 1949.
Smith, H. W., E. A. Rovenstine, W. Goldring, H.
Chasis, and H. A. Ranges. The effect of spinal anesthesia
on the circulation in normal, unoperated man with
reference to the autonomy of the arterioles, and especially
that of the renal circulation. J. Clin. Invest. 18: 319, 1939.
Smith, H. W. The physiology of renal circulation. Harvey
Lectures Ser. 35 : 1 66, 1 939-40.
Smith, H. W. The Kidney: Structure and Function in Health
and Disease. New York: Oxford Univ. Press, 1951.
Smith, II. W. Principles of Renal Physiology. New York:
Oxford Univ. Press, 1956.
Smith, H. W. The fate of sodium and water in the renal
tubules. Bull. N. Y. Acad. Med. 35: 293, 1959.
Spanner, R. Der Abkurzungskreislauf der mcnschlichen
Niere: Beitrag zur Kenntnis der Leistungsweiteilung ihre
Gefassystems. Klin. Wochschr. 16: 1421, 1937.
Spanner, R. Uber Geffasskurzschlusse in der Niere.
Yerhandl. anal. Ces. Jena. 45: 81 (Erganzungsheft, Anat.
Ans., 85), 1937.
Spencer, M. P., A. B. Denison, and H. D. Green. The
direct renal vascular effects of epinephrine and nor-
epinephrine before and after adrenergic blockade. Circu-
lation Research 2: 537, 1954.
Spencer, M. P. The renal vascular response to vaso-
depressor sympathomimetics. J. Pharrncol . Exptl. Therap.
116: 237, 1956.
Spinazzoi.a, A. J., and T. R. Sherrod. The effect of
THE RENAL CIRCULATION
1515
serotonin (5-hydroxytryptaminc) on renal hemodynamics.
J. Pharmacol. Exptl. Therap. 119: 114, 1957.
295. Still, J. W., and E. R. Whitcomb. An investigation of
renal shunts in rats. Am. J. Physiol. 178: 399, 1954.
296. Stone, J. E., J. Wells, \V. B. Draper, and R. \V.
Whitehead. Changes in renal blood flow in dogs during
the inhalation of 30 per cent carbon dioxide. Am. J. 313.
Physiol. 194: 115, 1958.
297. Stone, J. E., R. L. Irwin, C. D. Wood, W. B. Draper.
and R. W. Whitehead. Renal blood flow in dogs during 314.
diffusion respiration. J. Appl. Physiol. 1 4 : 405, 1 959.
298. Study, R. S., and R. E. Shipley. Comparison of direct
with indirect renal blood flow, extraction of inulin and 315.
Diodrast before and during acute renal nerve stimulation.
Am. J. Physiol. 163:442, 1950. 316.
299. Surtshin, A. C, C. B. Mueller, and H. L. White.
Effect of acute changes in glomerular filtration rate in 317-
water and electrolyte excretion : mechanism of dener-
vation diuresis. Am. J. Physiol. 169: 159, 1952.
300. Swann, H. G., A. V. Montgomery, and J. S. Lovvry.
Effect of renal venous occlusion on intrarenal pressure.
Proc. Soc. Exptl. Biol. Med. 76: 773, 1951. 318.
301. Swann, H. G, V. Moore, and A. V. Montgomery.
Influence of arterial pressure on intrarenal pressure. Am.
J. Physiol. 168:637, 1952. 319.
302. Swann, H. G., B. W. Hink, H. Koester, V. Moore,
and J. M. Prine. The intrarenal venous pressure. Science
"5:64. '952-
303. Swann, H. G., L. Valdivia, A. A. Ormsby, and W. T. 320.
Witt. Nature of fluids which functionally distend the
kidney. J. Exptl. Med. 104: 25, 1956.
304. Swann, H. G., A. A. Ormsby, J. B. Delashaw, and 321.
W. W. Tharp. Relation of lymph to distending fluids of
the kidney. Proc. Soc. Exptl. Biol. Med. 97: 517, 1958.
305. Thompson, D. D., F. Kavalier, R. Lozano, and R. F. 322.
Pitts. Evaluation of the cell separation hypothesis of
autorcgulation of renal blood flow and filtration rate: 323.
blood flow, filtration rate, and PAH extraction as function
of arterial pressure in normal and anemic dogs. Am. J.
Physiol. 191 :493, 1957. 324-
306. Thurau, K., and K. Kramer. Der Einfluss des Gefass-
tonus und des Haematokrit des Perfusions — Fliissigkeit
auf die Autoregulation des Nieren-kreislaufs. Pfliigers 325'
Arch. ges. Physiol. 268: 43, 1958.
307. Thurau, K., and K. Kramer. Die Reaktionsweise der
glatten Muskulatur der Nierengefasse auf Dehnungsreize
und ihre Bedeutung fur die Autoregulation des Nieren- 326-
kreislaufes. Pfliigers Arch. ges. Physiol. 268: 188, 1959.
308. Thurau, K, and K. Kramer. Weitere Untersuchungen
zur myogenen Natur der Autoregulation der Nieren- 327-
kreislaufes. PJliigers Arch. ges. Physiol. 269: 77, 1959.
309. Thurau, K., P. Deetjen, and K. Kramer. Hamodynamik
des Nierenmarks : Wechselbeziehung zwischen vascularem
und tubularem Gegenstromsystem bei arteriellen Druck- 328.
steigerung, Wasserdiurese, und osmotischer Diurese.
Pfliigers Atch. ges. Physiol. 270: 270, i960. 329-
310. Tobian, L. Interrelationship of electrolytes, juxtaglomer-
ular cells and hypertension. Physiol. Rers. 40: 280, i960.
311. Trueta, J., A. E. Barclay, P. M. Daniel, K.J. Frank- 330.
lin, and M. M. L. Prichard. Studies of the Renal Circu-
lation. Oxford (England) : Blackwell Scientific Publ., 1947.
312. Ullrich, K. J., and K. H. Jarausch. Untersuchungen 331.
zum Problem der Harnkonzentrierung und Verdiinnung.
Uber die Verteilung der Electrolyten (Na, K, Ca, Mg,
anorg. Phosphat), HarnstofT, Aminosaiiren und exogenen
Kreatinin in Rinde und Mark der Hundeniere bei
verschiedenen Diuresezustandcn. Pfliigers Arch. ges. Physiol.
2°2:537. 1956-
Ullrich, K. J., and G. Pehling. Aktiver Natrium Trans-
port und Saucrstoffverbrauch in der aiisseren Markzone
der Niere. Pfliigers Arch. ges. Physiol. 267: 207, 1958.
Ullrich, K. J. Das Nierenmark: Stuktur, Stoffwechsel,
und Funktion. Ergeb. Physiol, u Exptl. Phaimakol. 50: 433,
'959-
Ullrich, K. J. Uber die Funktion des Nierenmarkes.
Deut. Med. Wochschr. 84: 1 197, 1959.
Unna, K. Artericller Druck und Nierendurchblutung.
Pfliigers Arch. ges. Physiol. 235: 515, 1935.
Van Slyke, D. D., C. P. Rhoads, A. Hiller, and A. S.
Alvinc. Relationships between urea excretion, renal
blood flow, renal oxygen consumption, and diuresis. The
mechanism of urea excretion. Am. J. Physiol. 109: 336,
1934-
Vimtrup, B. Number, shape, structure and surface area
of glomeruli in man and animals. Am. ./. Anat. 41 : 123,
1928.
von Bubnoff, M., D. Hoffman, E. Schmid, and R.
Taugner. Zur sympatholytischcn, adrenolytischen und
noradrenolytischen Wirkung Phenothiazine. Arch, exptl.
Pathol. Pharmakol. 224 : 443, 1 955.
von Kiigelgen, A., and H. Greinemann. Die Klappen in
den menschlichcn Nierenvenen, besonders an der Miin-
dung der Nierenbeckenvenen. Z. Zellforsch., 47 : 648, 1958.
von Kugelgen, A., and S. Zuleger. Nachweis von
Venenklappen in der Niere von Hund, Schwein und
Mensch. Z. Zellforsch. 47: 327, 1958.
von Kugelgen, A., B. Kuhlo, W. Kuhlo, and J. Otto.
Die Gefassarchitektur der Niere. Stuttgart : Thieme, 1 959.
von Kugelgen, A., and E. Passarge. Das Nierenbeckan-
gefassystem als extraglomenularer Blutweg. Z. Anat.
Entwicklungsgeschiclite 122:86, i960.
Wachholder, K. Haben die rhythmischen Spontankont-
raktionen der Gefasse einen nachweisbaren Einfluss auf
den Blutstrom? Pfliigers Arch. ges. Physiol. 190: 222, 1921.
Walker, W. F., M. Sheretettin Zileli, F. W. Reutter,
W. C. Shoemaker, D. Friend, and F. D. Moore.
Adrenal medullary secretion in hemorrhagic shock. Am.
J. Physiol. 197: 773, 1959.
Wallenius, G. Renal clearance of dextran as a measure of
glomerular permeability. Acta Soc. Med. Upsalien. 59:
Suppl. 4, 1 -9 1, 1954.
Warren, J., E. Brannon, and A. Merrill. A method of
obtaining renal venous blood in unanesthetized persons
with observations on the extraction of 02 and sodium
/>-aminohippurate. Science 100: 108, 1944.
Waugh, W. H. Flow as a function of arterial pressure in
the oil-perfused kidney. Circulation Research 6: 107, 1958.
Waugh, W. H., and W. F. Hamilton. Increased una1
venous pressure and extrarenal pressure on renal vasculai
resistance. Circulation Research 6: 116, 1958.
Wauch, W. H. Myogenic nature of autoregulation of
renal blood flow in the absence of blood corpuscles.
Circulation Research 6: 363, 1958.
Waugh, W. H., and R. G. Shanks. Cause of genuine
1516
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
autoregulation of the renal circulation. Circulation Research
8: 871, i960.
332. Weaver, A. N, C. T. McCarver, and H. G. Swann.
Distribution of blood in the functional kidney. J. Exptl.
Med. 104: 41 , 1956.
333. Weiss, C, H. Passow, and A. Rothstein. Autoregulation
of flow in isolated rat kidney in the absence of red cells.
Am. J. Physiol. 196: 1 1 15, 1959.
334. Werko, L., H. Bucht, and B. Josephson. The renal
extraction of PAH and oxygen in man during functional
changes of the circulation. Scand. J. Clin. & Lab. Invest.
[1: 32I> !949-
335. Werko, L., H. Bucht, B. Josephson, and J. Ek. The
effect of nor-adrenaline and adrenaline on renal hemo-
dynamics and renal function in man. Scand. J. Clin. &
Lab. Invest. 3: 255, 1951.
336. Werko, L., E. Varnauskas, H. Eliasch, J. Ek, H.
Bucht, B. Thomasson, and J. Bergstrom. Studies on the
renal circulation and renal function in mitral valvular
disease. I. Effect of exercise. Circulation 9: 687, 1954.
337. Werko, L., E. Varnauskas, J. Ek, H. Bucht, B.
Thomasson, J. Bergstrom, and H. Eliasch. Studies on
the renal circulation and renal function in mitral valvular
disease. II. Effect of Apresoline. Circulation 9: 700, 1954.
338. White, H. L. Observations on the behavior of Diodrast
in the dog. Am. J. Physiol. 130: 454, 1940.
339. White, H. L., and D. Rolf. Some effects of exercise and
of some other influences on the renal circulation in man.
Am. J. Physiol. 152: 505, 1948.
340. Winton, F. R. Intrarenal pressure. J. Physiol. , London 78:
9p. 1933-
341. Winton, F. R. The influence of changes in the arterial
pressure on the intrarenal pressure in the isolated mam-
malian kidney. J. Physiol., London 87: 18P, 1936.
342. Winton, F. R. Intrarenal pressure and renal blood flow.
With discussion by H. G. Swann. Trans, yd Conf. Josiah
Macy, Jr. Found, p. 51, 1951.
343. Winton, F. R. Pressures and flows in the kidney. Modern views
on the secretion of the urine. ( The Cushny Memorial Lectures)
Boston: Little, Brown, 1956, p. 61.
344. Winton, F. R. Present concepts of the renal circulation.
A.M. A. Arch. Internal Med. 103: 495, 1959.
345. Wirz, H., B. Harcitav, and W. Kuhn. Lokalization des
Konzentrierungs-prozessen in der Niere durch directe
Kryoskopie. Helv. Physiol, et Pharmacol. Acta 9: 196, 1951.
346. Wirz, H. Der osmotische Druck des Blutes in dern Nieren-
papillae. Helv. Physiol, et Pharmacol. Acta 1 1 : 20, 1 953.
347. Wirz, H. Druckmessung in Kapillaren und Tubuli der
Niere der Ratte. Helv. Physiol, et Pharmacol. Acta 13: 42,
■955-
348. Wirz, H. Der osmotische Druck in den corticalen Tubuli
der Ratten Niere. Helv. Physiol, et Pharmacol. Acta 14: 353,
■956-
349. Wirz, H. Die Niere als Regulator des osmotischen Druckes
Mod. Probl. Paediat. 6: 86, i960.
350. Wise, B. L., and W. F. Ganong. Effect of brain-stem
stimulation on renal function. Am. J. Physiol. 198: 1291,
i960.
351. Wolf, G. A. Effect of pain on renal function. Research
Publ. Assoc. Nervous Research Mental Diseases 23: 358, 1943.
352. Yamada, S. I., and A. Astrom. Critical closing pressure
and vasomotor tone in the hind leg and kidney of the cat.
Am. J. Physiol. 196: 213, 1 959.
353. Young, W. G., Jr., J. S. H. Harris, and W. C. Sealy.
Production of neurogenic afferent renal vasoconstriction
in humans and dogs by 2-benzyl-4,5-imidazoline HC1
(Priscoline). J. Appl. Physiol. 3: 77, 1950.
CHAPTER 44
Blood supply to the heart
DONALD E. GREGG
LLOYD C. FISHER
Department of Cardiorespiratory Diseases, Walter Reed Army Institute
of Research, Walter Reed Army Medical Center, Washington, D. C.
CHAPTER CONTENTS
Functional Anatomy
The Myocardium
Coronary Arteries
Myocardial Arterioles and Capillaries
Myocardial Veins
Collateral Circulation
Congenital Anomalies
The Cardiac Nerves
Lymphatic Drainage of the Heart
Preparations and Methodologies of Special Interest in the
Study of the Heart and Its Coronary Circulation
Preparations
Coronary Flow Methods (Animals)
Coronary Flow (Man and Animals)
Distribution of Myocardial Blood Flow-
Arterial Circuit
The Venous Circuit
Possible Use of Left Coronary Artery Flow Together with
the Chemical Composition of Coronary Sinus Blood as an
Index of Left Ventricular Metabolism
Physical Determinants of Coronary Flow
Determinants of Normal Myocardial Metabolism
Basal Data
Response of the Coronary Circulation to Various Stimuli
Resting State
Reactive Hyperemia
Heart Rate
Heart Doing No External Work
Ventricular Volume or Fiber Length
Blood Pressure
Chemical Composition of the Blood
Transfusion
Anemia
Nervous Influences
Hormones
Exercise and Excitement
Valvular Disease
Hypertensive Cardiovascular Disease
Heart Failure
Hemorrhagic Shock
Hypothermia
Hyperthermia
Summary
Drugs Versus the Coronary Circulation
Coronary Artery Disease
Natural Responses of the Normal but Overstressed Portion of
the Myocardium
Coronary Artery Collateral Circulation
FUNCTIONAL ANATOMY
the historical knowledge of the heart's integral
blood supply parallels knowledge of the broader scope
of the cardiovascular system in toto. Thus commenc-
ing with Galen's designation of the term "coronary
arteries," it nevertheless remained for Harvey (1645)
to show accurately that channels existed in the walls
of the heart for its own nourishment. Interarterial
anastomoses were demonstrated by Lower in 1671
using fluid injection techniques, and in 1 704 the
ventricular branches of the coronary arteries were
visualized by a corrosion technique introduced by
Ruysch. Connections between the arteries and the
cardiac cavities were shown in 1 706 by Vieussens
using saffron injections into the coronary arteries, and
between the cardiac veins and the cardiac chambers
in 1 708 by Thebesius using air injected through the
coronary sinus. That these cavitary communications
were, in fact, different channels was not well docu-
mented until the twentieth century when phylo-
genetic studies by Grant (142), and mammalian
studies by Wearn (382) established the existence of
intramyocardial trabeculae and sinusoids which
separated the veins (Thebesian) from the arterial
circuit (arterioluminal), and contributed their own
communications (arterio-sinusoidal) to the cavities.
The more recent introduction of radiographic tech-
niques (84, 352) for visualization of coronary arteries
in intact humans and animals, or in pathologic speci-
mens (18, 37, 338), and of cast-digestion techniques
(18, 172, 173, 190-192, 258) for permanent reproduc-
1517
i5i8
HANDBOOK OF PHVSIOI.I « :\
CIRCULATION II
lions of normal and pathologic channels have
further advanced and clarified the interarterial and
transarterial communications and their branches.1
The Myocardium
ventricles. Gross dissection studies (108, 153, 232)
reveal a rather consistent and orderly arrangement in
mammals with quantitative differences overshadowed
by qualitative similarities. Every muscle fascicle
originates from the fibrous rings at the base, the super-
ficial fibers descending toward and penetrating the
apex to form the vortex spirals, and then looping
upward as the deeper fibers which ascend along the
endocardial surface to reinsert in the annulus fibrosis.
Thus, the two ventricles are encompassed by figure-
of-eight bands of muscle with origins and insertions
at the base and a fulcrum at the apex. The muscular
interventricular septum receives part of these fascicles
while an intermediate layer encircles only the left
ventricle, also adding to the septum.
There are probably no true cleavage planes between
isolated fascicles but, rather, the ventricle represents
a single muscle mass dividing and branching into
intercommunicating fascicles. In any one plane, how-
ever, the fibers are more or less constant, the epicardial
fibers running perpendicular to the endocardial
fibers at any given point.
atria. Nearly all fibers arise and insert into the A-V
rings, but some fibers merge and disappear on the
muscular coats of the great veins.
Interatrial fibers form the septal areas while the
auricles and pectinate regions are largely intra-atrial
fascicles. There are two simple layers — an inner
horizontal and an outer vertical — bound by much
intertwining interdigitation.
1 :i >\ducting system. Commencing with the sino-atrial
node at the superior vena cava and right atrium,
specialized myomeric conducting tissue traverses the
right side of the interatrial septum to the locus of the
atrioventricular node (153, 399). The latter is situated
on the atrial side of the base of the tricuspid valves'
medial leaflet, above the coronary sinus and between
1 No attempt has been made to give a complete bibliography
which would involve consideration of many thousands of publi-
cations. Except for an occasional lead article, the older work is
considered by referring to some 40 to 50 reviews, monographs,
and symposia. Direct but incomplete reference is made to the
more recent work not covered in such summaries. By this
means, most of the important work in the field can be found by
the interested reader although direct reference may no! In-
made to it.
the limbus fossa ovalis and the medial leaflet. From
the A-V node, the bundle of His penetrates the fibrous
A-V ring and runs in the posterior membranous
interventricular septum, branching into the right and
left bundles at this site or in the upper muscular
septum. The right bundle branch is solitary in its
course through the septum to the base of the modera-
tor band, while the left bundle subdivides into many
branches. The terminations of each bundle form many
fine fasciculi intimately applied to the endocardium
before merging with the contractile myocardium.
Coronary Arteries
The course and distribution of the major coronary
arteries in all mammalian subgroups is remarkably
similar and intergroup differences are less pronounced
than intragroup variations. The basic anatomic pat-
terns are thus comparable from the smallest to the
largest mammals, i.e., from rodents to whales (63,
78, 142, 298).
There are two coronary arteries, right and left,
arising respectively from the right anterior and left
anterior aortic sinuses of Valsalva. The ostia are
situated above the reflections of the semilunar valves,
the right coronary in man being 35 ° to the right, and
the left coronary 650 to the left of the anteroposterior
axis of the body (258).
left coronary artery. This vessel courses in epi-
cardial areolar tissue anteriorly and to the left in the
auriculoventricular groove, between the pulmonary
artery and the left auricular appendage, and bifur-
cates into the anterior descending and circumflex
branches (fig. 1). These two branches are quite con-
stant in all species, the bifurcation occurring 1 to 1.5
cm (84, 1 89-191, 258) from the ostium in man, and
2-4 mm in dogs and smaller mammals (36, 64, 153,
290). In dogs and rabbits, but not in man, monkeys,
or higher primates (63, 64, 78, 172, 173, 189, 258), a
septal artery arises just prior to, at, or not uncom-
monly, just beyond the bifurcation on the descendens
or circumflex, in that order. Small branches from the
left coronary artery are frequently present passing to
the pulmonary conns and left atrium, and in the
rabbit, branches from both left and right coronary
arteries supply the major portion of the vasa vasorum
of the pulmonary artery (351). A third primary
division has also been described arising between the
above and supplying the anterior left ventricle (348).
The anterior descendens follows the anterior inter-
ventricular sulcus toward the apex and is of variable
length, terminating prior to, at, or beyond the apex.
BLOOD SUPPLY TO THE HEART
I5!9
fig. I. Vinylite cast of a human heart. Anterolateral aspect
of left ventricle following digestion of muscle. M.P.A. = main
pulmonary artery; R.A. = right atrium; L.A.D. = left anterior
descending coronary artery; L.C. = left circumflex coronary
artery; G.C.V. — great cardiac vein, P.I.I'. = posterior inter-
ventricular vein. [From James (191).]
In humans it terminates 40 per cent of the time at the
apex, and in 60 per cent, ascends 2 cm or more in the
posterior longitudinal sulcus, while in rabbits it
rarely reaches the apex (64, 84, 153, 191, 258). It is
covered by bridges of ventricular myocardium for
most of its course (292). There are from two to seven
ventricular branches, the large left ventricular
branches coursing over the anterior surface toward the
apex, the small right ventricular branches crossing the
interventricular groove to supply a narrow band of
muscle and to anastomose with right coronary
branches. Anastomoses exist with anterior ventricular
branches of the left circumflex coronary artery and
at the apex, with the latter's marginal branch and
the posterior descending artery whether of circumflex
or right coronary origin (18, 189, 191, 258, 337).
Septal branches penetrate deeply from the underside
of the vessel all along its course in the anterior sulcus.
In humans, primates, and pigs, these branches are not
supported as in dogs and rabbits by an individual
septal artery arising from the main left coronary or
origin of the descendens. A fairly constant branch to
the pulmonary conus region exists in most species.
The left circumflex follows the auriculoventricular
groove to the left, coursing under the left auricular
appendage and terminating at a variable distance
from the posterior longitudinal sulcus. It is largely an
epicardial vessel, surrounded by areolar and adipose
tissue, and rarely covered by muscular loops (292). In
dogs it almost always reaches or crosses the crux of the
posterior sulcus, terminating as the posterior descend-
ing artery, whereas in pigs (64, 289) it rarely does so.
In man, higher primates, and rabbits, the vessel
usually ends at the obtuse margin (63, 78, 191, 258).
An average of three anterior ventricular branches and
three atrial branches occurs in man and dogs (36,
64), the former coursing to the apex to anastomose
with the anterior descendens branches. Posteriorly,
communications exist with the right coronary either
from the posterior descendens or the marginal
branches. In the dog a branch of the left circumflex at
the posterior crux passes deep to supply the A-V node
and His bundle (172, 173).
right coronary artery. The main right coronary
artery arises from a single ostium in its aortic cusp,
but not infrequently, especially in dogs and primates,
smaller ostia of accessory branches are also present
(63, 258, 290). The right coronary passes anteriorly
behind the pulmonary artery and follows the respec-
tive auriculoventricular groove to the right (acute)
margin of the heart. In dogs and rabbits it usually
terminates here as the marginal branch, whereas in
pigs and man it invariably (93 %) reaches the posterior
crux to become the posterior descending artery (64,
191, 289). In its course it gives off an average of three
atrial branches, one of which, the dorsal (posterior)
right atrial artery, is the major supply to the S-A
node in man and dog (172, 190), and three to five
right ventricular branches. Posteriorly, in man and
pigs, a branch to the A-V node is given off at the
crux, corresponding to the branch from the circum-
flex in dogs (172, 189, 190, 246, 258, 399). A constant
branch to the pulmonary conus frequently arises from
an accessory ostium.
Although it is evident that the course and distribu-
tion of the coronary arteries is basically similar in the
various species mentioned, the ramifications are such
as to permit a breakdown into patterns of dominance.
Thus, in all species, the entire anterior and lateral
left ventricle is supplied by the left coronary branches,
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and the free right ventricular wall by the right
coronary artery. The most variable area is posterior,
and it is by virtue of the communication of the poste-
rior descending artery with either the left or right
coronary artery, or both, that the designation "domi-
nant" pattern has arisen (18, 37, 84, 153, 258, 348).
Hence, dogs are universally left coronary dominant,
the left circumflex branches supplying the posterior
left and right ventricles, the posterior septum and
A-V node. This pattern is the least common in man
and pigs, approximating 20 per cent of cases in the
former. Pigs are generally right coronary dominant,
while man, both living and autopsied, and the higher
primates manifest this pattern half of the time and a
balanced circuit in approximately a third (37, 63,
84, 153, 348). In the perfused human heart, however,
this pattern of dominance is not found (370).
The secondary divisions of the major coronary
branches in man are consistently different over the
two ventricles, the branches of the anterior descendens
and left circumflex arising at acute angles and cours-
ing to the apex, while those of the right coronary
arise at right angles and course toward the anterior
interventricular sulcus (191). The terminal branches
are likewise different; those over the left ventricle are
perpendicular to the epicardial course, while those
over the right ventricle are parallel. Once the arteries
penetrate the myocardium, they lose their tortuosity
and linearly follow the muscular grain in a plane
between the superficial and deep muscle layers (153,
191, 258).
The functional supply to the conducting tissue
bears further comment since it has been shown in
man and dogs that mortality and morbidity are in-
creased when ligation of vessels includes septal and
nodal arteries (5, 61, 246). Of the three atrial branches
from each coronary artery, the cristal branch of the
dorsal right atrial artery is the major supply to the
S-A node in man (60-70%) and dogs; rich anastomo-
ses exist with the ventral left atrial artery in 75 per
cent of dogs whereas, in man, this latter vessel is the
major supply to the S-A node in 40 per cent (1 53, 1 72,
173, 189, 190, 399). In rats and dogs (172, 174), and
possibly in man, anastomoses with extracardiac ves-
sels are readily shown at the junction of the superior
vena cava and right atrium. Rats have a dual blood
supply to the heart, the atrial and S-A nodal vessels
stemming from cardiaco-mediastinal branches of the
internal mammary and subclavian arteries, while the
ventricles, A-V node, and parts of the atria are sup-
plied by the coronary arteries.
Recent vinylite cast techniques have shown a
consistent artery to the region of the A-V node arising,
BRANCH TO
AV NODE
R. CORONARY A
LEFT ANT DESC.
CORONARY A.
POST
DESC.
CORONARY
A.
fig. 2. Drawing of the blood supply of the normal human
interventricular septum. Note the preponderance of supply by
the left anterior descending coronary artery and the U-turn of
the posterior right coronary artery which gives off the branch
to the atrioventricular node. [From James (189).]
in man, from that coronary artery which crosses the
posterior crux (189-191, 246). Thus, in 80 percent,
this was the right coronary, the left in 10 per cent, and
from both in another 10 per cent. In 100 per cent of a
large series in dogs (246), a similar vessel arose at the
crux from the left circumflex coronary. This vessel,
variously named the posterior septal artery and ramus
septi fibrosi, courses along the base of the interatrial
septum and penetrates the annulus fibrosus to supply
the His bundle and upper interventricular septum
(246, 399) (fig. 2). In its course it freely anastomoses
with atrial vessels, predominantly the dorsal left
atrial artery and, from below, the anterior septal
arteries.
The interventricular septum receives its blood sup-
ply from the anterior septal artery and penetrating
branches of the anterior and posterior descending
arteries (36, 172, 173, 189, 190, 246, 258). The former
is well developed in dogs but in man and higher
primates it is somewhat vestigial, although easily
identified as the first and largest branch of the anterior
descendens.
The anterior branches in man are 40 to 80 mm in
length, supply the anterior two-thirds to three-fourths
of the septum, and penetrate near the right ventricular
side remaining under the right ventricular endo-
cardium before terminating deeper in the septum.
BLOOD SUPPLY TO THE HEART
I 52 I
The posterior branches are shorter, up to 15 mm; they
supply the posterior one-third of the septum and
anastomose with the anterior branches. In dogs, how-
ever, the anastomoses are deficient and the large
anterior septal artery's superior and inferior divisions
supply the central portion of the upper two-thirds of
the septum including the moderator band and lower
His bundle (36, 64, 246, 290, 399). The more distal
bundle branches are supplied by the penetrating
vessels.
coronary blood volume. Available information re-
garding coronary blood volume (artery through
coronary vein content) is incomplete and quite ap-
proximate. In humans, at postmortem, average values
in both sexes range from about 2 to 6 ml per 100 g
heart weight (310). In the arrested dog and cat heart
and isolated beating dog heart, values approximate 6
to 8 ml per 100 g heart muscle (129).
Myocardial Arterioles and Capillaries
As the superficial arteries penetrate the myo-
cardium, they bifurcate or trifurcate disproportion-
ately so that the parent vessel and diameter grow
gradually smaller while the daughter vessels narrow-
rapidly, terminating in the capillary network (294,
295). The deeper arterioles lose the internal elastic
membrane and subendothelium present in the more
superficial layers and contain a single layered intima
and a media one to two muscle layers thick. As the
arteriole narrows, its muscularis becomes discontinu-
ous and the muscle cells decrease in frequency with
increasing distance from the arteriole. This latter
vessel, the metarteriole, is continuous at its distal end
with the simple endothelial tube characterizing the
capillary. A group of one or more, usually three, mus-
cle cells at the proximal end of a capillary constitutes
a sphincter and denotes the precapillary.
Recent studies have suggested that the myocardial
capillaries are not all functional at all times as was
previously believed (294, 295). It has been shown that
the metarterioles and precapillary sphincters can
close off the capillary lumen. Thus, during sphincteric
contractions, the nucleus of the endothelial cell under-
lying the sphincter becomes rounded and is forced
into the lumen of the vessel thereby occluding it. Dur-
ing relaxation and in those regions where there are no
sphincters, the nucleus is flattened along the wall and
the lumen is open. The demonstration of nerve fibers
accompanying the vessels in the areolar connective
tissue and terminally "splaying'' to surround the
myocardial cells and sphincters, and the absence of
any such supply to "true" capillaries, lends support
to a changing dynamic state of capillary patency and
function. Moreover, the demonstration of anastomotic
connections between arterioles, metarterioles, pre-
capillaries, and venules in both man and dogs suggests
arteriovenous shunting as an integral component of
the myocardial capillary circulation.
In the newborn human and rabbit there is approxi-
mately one myocardial capillary per four myocardial
fibers, corresponding to 4,000 capillaries per mm2 of
tissue (382). In the human adult the ratio of capillaries
to fibers approaches 1:1, while the capillary concen-
tration approximates 3,000 to 4,000 per mm2 of tissue,
both values being fairly constant over a wide age span
(6, 125, 382). The capillary diffusing area per cm3 of
tissue averages 1,145 cm2 m children, and 1,184 cm2
in adults (6). An analysis of tissue from various ven-
tricular areas reveals similar capillary densities and
surface areas for the human left ventricle, right ven-
tricle, and papillary muscle, whereas the interventric-
ular septum shows a decrease in both these parameters.
While the maximum diffusing distance is calculated
to be 8 /u in all of the above areas, that to the con-
ducting system proper is appreciably greater.
In contrast to the septal myocardium, there is a
scanty capillary supply to the A-V node and His
bundle in sheep and cattle (125). Capillaries and
conducting fibers are not intimately connected and
are often separated by wide spaces of connective tissue.
In the His bundle, capillaries are located outside the
dense band of fibers with the central nuclei far from
the source of blood. Other investigations in dogs and
humans have shown a well-developed system of sinu-
soids anastomosing with capillaries, veins, and arteries
which traverse the annulus fibrosus and supply the
A-V node and common bundle (365, 366).
Exchange of metabolites in myocardial capillaries
has received anatomic amplification and clarification
by electron microscopic techniques (iii, 270, 286).
The endothelial cells form a continuous capillary and
arteriolar lining without any evidence of intercellular
or intracellular pores. Many vesicles or caveolae are
concentrated under the cell membranes facing both
the capillary lumen and pericapillary spaces, and are
believed to represent continuous invagination and
pinching off of the plasma membrane which then
crosses the cell and liberates nutrients, metabolites,
and other materials (1 1 1, 270, 286). Injected colloidal
gold particles have been photographed concentrating
along the luminal side, engulfed and transported
across the cell in vesicles, and finally, phagocytized by
macrophages in the pericapillary spaces. This trans-
port mechanism has been variously termed "pino-
1522
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
cytosis" (286) and "cytopempsis" (270), the latter
being preferred since it does not imply actual utiliza-
tion of the transported substances by the endothelial
cell.
Myocardial I 'etm
There are twice the number of venous as arterial
channels in the heart, their density in the left ventricle
greatly exceeding that of the right (191), and they
have been subdivided into superficial and deep cir-
cuits (153, 365, 366). The superficial left ventricular
veins parallel the arterial branches and course toward
the base of the heart to empty into the great cardiac
vein anteriorly, and its continuation in the left auric-
uloventricular groove, the coronary sinus, poste-
riorly. The latter empties into the right atrium in the
posterior-inferior interatrial septum located between
the medial end of the inferior vena cava and A-V
ring, and receives subsidiary trunks up to its orifice
(183). The anterior cardiac veins drain the right
ventricle and are smaller, frequently solitary, trunks
which empty individually into the right atrium just
above the A-V valves (153).
The deeper venous circuit has communications
with both atrial and ventricular cavities via Thebesian
and sinusoidal channels (153). Myocardial sinusoids
or trabeculae are especially rich in the ventricular
walls and maintain communications with arterioles,
capillaries, venules, and the heart cavities (64, 153,
189, 191, 365, 366) These sinusoids are lined by a
single layer of endothelium and range from 40 to 75 ji
in dogs, and 60 to 90 jx in newborn humans in the
septal myocardium. In dogs and pigs there is a mas-
sive formation of sinuses in the left ventricular wall
communicating with the cavity (64).
Collateral Circulation
As noted earlier (vide supra), intercoronary anasto-
moses were first demonstrated by Lower using a
watery injection of dye, and in 1803 von Haller re-
ported on the extracardiac communications of the
coronary arteries, utilizing the same techniques (153,
382). The latter were principally channels from the
base of the pulmonary artery and veins, root of the
aorta and venae cavae, and other basal (usually
atrial) vessels, to vessels in the intrapericardial re-
flections. These vessels are largely from the internal
mammary artery via the pericardiacophrenic branch,
but communications also exist with the bronchial
arteries.
In lower vertebrates the blood supply to the heart
is nearly all extracardiac in origin, whereas the rat
maintains a dual supply of both intracardiac and
extracardiac origin. In normal mammalian hearts the
extracardiac communications are of the order of small
arterioles and capillaries and are anatomically and
physiologically insignificant.
Intracardiac coronary anastomoses in human
hearts and those of various laboratory animals have
been the subject of numerous pathologic and experi-
mental investigations during the past decade. Col-
lateral arterial communications in normal hearts have
been anatomically divided into those stemming from
the same major coronary artery, i.e., intracoronary,
and those between the right and left coronary-
branches, i.e., intercoronary (18, 258). All mammalian
species show some intercoronary anastomoses, espe-
cially over the anterior left ventricle, while inter-
coronary anastomoses vary appreciably between
species, the dog's being fairly vvell developed, the
pig's poorly, and man's quite variable; the greater
proportion occurring in the muscular interventricular
septum (18, 189).
Functional collateral channels, as opposed to
anatomic communications, have been defined for
mammalian hearts as those above 40 tx in diameter,
i.e., those which do not traverse a capillary bed (37,
38, 338). High viscosity fluids do not penetrate vessels
below 40 /x and, utilizing this technique or those with
graduate spheres above 35 tx, 6 to g per cent of normal
human hearts have adequate collateral channels (38,
2gi, 411). Conversely, latex casts have shown lux-
uriant anastomoses ranging from 20 to 350 /x in all
normal hearts and in all myocardial areas below the
subepicardial layer of muscle (18). Thus, while the
experimental and pathophysiological approaches to
this problem will be more fully discussed in a later
section, the disparity between the functional state,
i.e., its physiologic competency, and the nonfunctional
state, i.e., its anatomic patency, becomes more obvious
and the reason for the designation of the coronary
arteries as "end arteries" more apparent (153, 404).
While the above discussion has dealt mainly with
arterial collaterals, venous collateral channels freely
communicate over the surface of the heart (153), in-
cluding those between the anterior cardiac veins of
the right ventricle and the left ventricular coronary
sinus system. Extracardiac communications of the
cardiac veins are not uncommon especially in lower
mammals, and are usually related to the persistence
of the left caval or cardinal veins. In the pig, large
communications may exist between the hemizygous
vein and the great cardiac vein, the latter also having
substantial epicardial connections with the anterior
BLOOD SUPPLY TO THE HEART
'523
cardiac veins (64). Of greater concern, however, is the
existence of communications between the ventricular
cavities and the trabecular sinusoids via the The-
besian, arterioluminal, and arteriosinusoidal vessels
(153). Although dyes and particulate matter have
been recovered from ventricular myocardium follow-
ing intracavity injections (153), this has only occurred
experimentally with a) high ventricular end-diastolic
perfusion pressures, b) congenital aortic and pulmonic
valvular atresia with intact septa, and c) in the ar-
rested heart or one in which the heart stopped before
its removal. As a result of simple pressure differentials,
the dye or particle moves into the myocardium, while
for the same dynamic reasons only the reverse could,
and indeed does, occur in the actively beating heart
(i53. 404)-
Congenital Anomalies
Variations in the course and number of nutrient
vessels to the myocardium are not uncommon and,
as with other organ systems, are usually of no physio-
logic concern. However, the acceptance of the clinical
syndrome of the aberrant left coronary artery as part
of a group of congenital coronary arteriovenous
fistulae, and its recent physiologic documentation, has
prompted this brief digression into those embryologic
and phylogenetic ramifications relating to the coro-
nary arteries.
The lowest orders of vertebrate hearts have no well-
defined myocardial blood supply. Thus, the single-
chambered ventricle of the lamprey nourishes its
myocardium via extensive intramyocardial sinusoids
in direct communication with the ventricular cavity
(142). The arterial supply to vertebrate orders below
reptiles arises from cranial and caudal vessels coursing
through the cardiac ligaments. Reptilia maintain a
single cranial supply of vessels which are related to the
fishes' epibranchial and hypobranchial vessels, the
latter disappearing and moving caudally with the loss
of the gills (8).
Mammalian coronary arteries arise from primordial
buds in the truncus arteriosus during the 5th week of
gestation. At this time, the endocardial cushions and
longitudinal ridges are also forming, respectively
dividing the heart and truncus into two channels. The
heart has been actively beating and forcing blood
through the systemic circulation since the 3d week,
and the heart itself is nourished by the sinusoido-
luminal channels (8, 101, 140). In the fetal rabbit,
endothelial-lined trabecular spaces spiral toward the
surface forming capillaries and epicardial vessels. The
latter join with venous cords growing caudally in the
epicardium from the sinus venosus to form the first of
the myocardial vessels. Arterial buds form a few days
later and spread as a solid column of cells to the
bulbus cordis, with subsequent extensions and branches
to the lateral areas. As these epicardial arteries
enlarge, the sinusoids decrease in size by a condensa-
tion and compression of the myocardial cortex, finally
becoming capillaries (140). In lampreys and lower
fishes, and in certain human congenital anomalies,
this condensation does not occur, the spongy trabecu-
lar network remaining undisturbed (141, 382). In
higher fishes and mammalia there is an outer, con-
densed, capillary-containing layer supplied by epi-
cardial vessels, and an inner trabecular layer with
retained cavitary communications. Thus, the varia-
tions in the number and site of the coronary ostial
anlagen will determine the final origin of the coronary
arteries, while variations in the epicardial course and
degree of myocardial condensation may determine the
eventual communications. These anomalies have
recently been presented as follows (101).
a) Coronary arteries arising from the aorta and
supplying the heart in normal, albeit variable, fashion
without abnormal communications. This includes
those with single ostia and single coronary arteries,
common sinus, accessory ostia, and ostia elsewhere in
the aorta. In a recent large series, such anomalies oc-
curred in 52 cases of 18,950 autopsies for an incidence
of 2.75 per 1,000 (4). Reviews of single coronary
arteries in man have stressed the absence of clinical
symptoms except those related to associated cardio-
vascular anomalies (4, 308, 350). However, the
anomalous distribution seems to predispose to early
sclerotic changes and myocardial infarction, the
average age of death in adults being 45 years. In one
series (308), all cases of myocardial infarction, fibrosis,
or ischemia were related to the absence of a left
coronary artery, i.e., the presence of a single right
coronary artery.
b) Coronary arteries supplying blood to grossly
abnormal hearts in which congenital pulmonary or
aortic atresia exists in conjunction with intact ven-
tricular septa and intact A-V valves. Ventricular
blood is forced from the cavities via myocardial
sinusoids which anastomose in the epicardium with
the coronary arteries. This type, fortunately, is rare.
c) Coronary arteries distributing blood abnormally.
These may be via left-to-right arteriovenous shunts
into the right heart chambers, cardiac veins, or pul-
monary artery, or via arterioluminal shunts into the
left heart chambers (fig. 3).
1524 HANDBOOK OF PHYSIOLOGY -^ CIRCULATION II
Atretic HV.
fig. 3. Anomalous coronary artery commu-
nications: A: retrograde flow from right ven-
tricular cavity to epicardial coronary arteries
via myocardial sinusoids in presence of pul-
monic (or aortic) atresia with intact ventricular
septum and competent auriculoventricular
valves. B: composite illustration of aortic
communication with the cardiac chambers via
the coronary arteries. C: communication of the
aorta with the pulmonary artery via aber-
rantly coursing coronary arteries. D: anoma-
lous origin of the left coronary artery from the
pulmonary artery. [From Edwards (101).]
ronary a
Congenital coronary arteriovenous fistulae have
been demonstrated in humans at thoracotomy, or
preoperatively utilizing angiocardiography and coro-
nary arteriography (104, 356). While gasometric
analyses may suggest a left-to-right shunt similar to
septal defects or a patent ductus arteriosus, ausculta-
tory findings have more often suggested the latter.
Clinical symptoms and signs, present in half the
cases, reflect a high output congestive failure, the
shunts averaging 40 per cent of the cardiac output
(356). The embryologic defect is probably a per-
sistence of myocardial sinusoids although the large,
sometimes aneurysmal, dilatation and veinlike thin-
ning of the arterial wall is a "common feature to all
arteries proximal to an arteriovenous shunt," and
may, therefore, be a secondary rather than a primary
alteration (101). A recent review now totals 71 cases
(104).
The anomalous left coronary artery has recently
become a subject of increasing clinical and physio-
logic interest, not only because it is the most common
of the congenital coronary artery aberrations and
readily diagnosed with modern clinical techniques,
but also because of the controversy concerning the
direction of blood flow in the aberrant vessel. There
have been over 60 cases reported in various reviews on
this anomaly, approximately one-fourth occurring in
adults in whom an apparent attenuation of the patho-
physiologic process is manifested. As the truncus is
dividing into aorta and pulmonary artery (5th week of
gestation), the primordial coronary ostial buds have
already been established and the growth of solid
arterial cords has commenced (8, 6g) (fig. 4). The
predominant finding of normal and equal-sized
aortae and pulmonary arteries strongly implicates a
malposition anteriorly of the left coronary artery as
the primary developmental defect, but the occurrence
of hypoplastic aortae, in rare cases, does not negate
the possibility of an abnormal division of the truncus
arteriosus.
BLOOD SUPPLY TO THE HEART
'525
AORTIC VALVE
TRUNCUS ARTERIOSUS
NORMAL
LEFT CORONARY
ARTERY
ANOMALOUS
PULMONARY VALVE
fig. 4. Diagrammatic representation of the normal and the anomalous origin of the left coronary
artery following torsion and division of the truncus into aorta and pulmonary artery. [From George
& Knowlan (127).]
The anatomic abnormality was first described for
an aberrant right coronary artery in 1886 (101). At
that time, the suggestion of reversal of flow in the
aberrant vessel was postulated because of the tortuous,
dilated nature of the arteries involved and a simple
reflection on the pressure differential between the two
circuits. The anatomic aberration of the left coronary
was described in ign and the clinical syndrome of
infarcts in 1933 (101). Electrocardiograph findings
suggest a recent anterior or anterolateral myocardial
infarction (58, 69, 10 1, 127, 210), while angiocardiog-
raphy or cine-angiocardiography reveals a normal
right ventricle and pulmonary artery and a dilated,
thinned left ventricle without evidence of filling of the
left coronary artery from the pulmonary artery; retro-
grade aortography reveals a dilated right coronary and
late filling of the left coronary (from right coronary
collaterals). The aberrant artery in both adults and
infants is a thin-walled veinlike vessel with an atro-
phied media. Grossly visible right-to-left coronary
anastomoses were present in 27 per cent of the adult
specimens.
Using pathologic specimens and surgical observa-
tions, but without definitive physiologic data for
support, Edwards earlier proposed a hypothesis
sustaining the concept of retrograde flow and refuting
that of antegrade flow from the pulmonary artery
(101). Physiologic proof of the retrograde nature of
flow in the aberrant left coronary has been presented
at thoracotomy in a preoperatively diagnosed 2}/%-
month-old child (325). Prior to ligation of the vessel
at its origin from the pulmonary artery, the pressure
in the left coronary artery was 30/15 mm Hg, rising
to 75 mm Hg systolic distally after occlusion, while a
simultaneous pulmonary artery mean pressure was 25
mm Hg. Arterial saturations in the corresponding
vessels were 100 and 76 percent, respectively. A post-
ligation rise of 30 mm Hg systolic pressure, and a
decreased paradoxical bulge of the left ventricular
infarct area, as blood now traversed rather than
shunted away from the myocardial bed, lends final
support to the retrograde flow thesis. In contrast to
the invariably fatal outcome within the first year of
life, this patient is alive and asymptomatic.
The Cardiac Nerves
The nerve supply to the heart is mediated through
the cardiac plexuses located above the base and be-
tween the aortic arch and tracheal bifurcation (397).
Vagal, sympathetic, and dorsal root fibers intermingle
and tend to lose their identity as they decussate into
right and left halves before entering the pericardium.
Functionally, however, they are best divided into
sensory and autonomic functions.
The sensory afferent fibers originate in thoracic
1526
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
dorsal ganglia. They are largely unmyelinated in their
myocardial course (153, 397), and supply the pain-
sensitive areas in the pericardium, connective tissue,
adventitia, and walls of the heart, terminating as fine
beaded nerve fibers and loops similar to those in the
skin and skeletal muscle. Sensory axons traveling in
sympathetic plexuses and through the lower two
cervical and upper four thoracic sympathetic ganglia
complete the afferent limb of the pain reflex. In both
man and dog, ablation of the stellate and upper four
thoracic ganglia, or upper four dorsal thoracic spinal
roots, completely blocks the pain pathway (397 ).
These neurons send fibers via the posterior spinal
roots which synapse in the posterior spinal horn with
secondary fibers running in the spinothalamic tracts
and terminating in the posterior-ventral nucleus of
the thalamus (fig. 5). While connections to the cortical
somatic sensory areas exist, these only modify the
reaction to, rather than the perception of, cardiac
pain.
The autonomic innervation includes both an af-
ferent and efferent vagal and sympathetic supply.
Vagal parasympathetics are mediated by the cardiac
plexus and stem from both right and left vagi and the
recurrent laryngeal nerves. A large portion of both
afferent and efferent fibers is distributed to the great
vessels superior to the heart, while the greatest part of
the remainder supply the interatrial septum and the
sino-atrial and A-V nodal areas (12, 76, 153, 368,
397). The large number of fibers in the latter areas
contrasts with the paucity of fibers supplied to the
atrial muscles via atrial arteries and the even smaller
number found in the ventricles. Using veratrum alka-
loids, only the left coronary artery system, i.e., the left
ventricle, has been found to contain afferent vagal
ganglia which contribute to the Bezold-Jarisch reflex
while, conversely, no efferent vagal supply is present
in either ventricle (12, 76, 368). The sympathetic
efferent discharge is largely to ventricular muscle and
coronary arteries and contains both cardiomotor
and vasomotor fibers, while atrial efferents are pre-
dominantly to the S-A node and are cardio-accelera-
SUP
CERVICAL GANGLION!^
MIDDLE
CERVICAL GANGLIONJ
UPPER SENSORY
NEURON IN
SPINOTHALAMIC
TRACT
LEG
PARAVERTEBRAL CHAIN OF
SYMPATHETIC GANGLIA
DIRECT THORACIC
CARDIAC NS.
OCCASIONAL CONNECTING
RAMUS TO
SPINOTHALAMIC TRACT
fig. 5. Illustration of the cardiac nerves and their central communications. Parasympathetic ef-
ferent and afferent fibers from the vagus and recurrent nerves join the cardiac plexuses at the base of the
heart. [From White (397).]
BLOOD SUPPLY TO THE HEART
1527
tor. Since staining techniques have been notably poor
in differentiating vagal and sympathetic terminals,
most of the available functional neural anatomy stems
from physiologic and pharmacologic observations
(12, 76).
Lymphatic Drainage of the Heart
The myocardial lymphatics arise at the periphery of
the capillaries and drain into deep and superficial lym-
phatic plexuses. They lie, respectively, immediately
subjacent to the endocardium and epicardium, the
former draining toward the surface to join in the
formation of lymphatic trunks. The vessels course in
the anterior and posterior longitudinal sulci and con-
dense to form left and right common trunks. The left
trunk passes between the pulmonary artery and left
atrium, and the right behind the pulmonary artery,
both terminating in the "cardiac Kmph node." This
node is well delineated in the dog and is regularly
found between the innominate artery and superior
vena cava (86, 287 I.
Recent studies have indicated a pathologic simi-
larity between experimentally induced myocardial
fibrosis secondary to chronic obstruction of the com-
mon lymph trunks and idiopathic endocardial
fibroelastosis or endomyocardial fibrosis (261).
PREPARATIONS AND METHODOLOGIES OF SPECIAL
INTEREST IN THE STUDY OF THE HEART
AND ITS CORONARY CIRCULATION
Many of the various preparations, procedures, and
instruments have been considered in previous reviews
(7, 10, 136, 146, 149, 152, 153, 299, 384, 400).
Preparations
The coronary circulation has been studied with the
heart in various degrees of deviation from the normal
state. These preparations include the heart-lung, the
isolated heart, and the open or closed-chest animal or
human with anesthesia. The use of the nonworking
isolated perfused heart by Langendorff in 1895 (221)
and by Porter (293), in which arterial inflow and
venous outflow could be measured, laid the ground-
work for our understanding of the coronary circula-
tion. An early bottleneck to the study of the coronary
circulation in the isolated heart was the lack of an
efficient means of oxygenating the blood. The isolation
of the heart connected to its lungs (215), and subse-
quent use of this preparation by many others (10)
contributed extensively to our knowledge of the heart
and coronary circulation. There are many variations
of this procedure but, in general, the heart and lungs
are removed in such a way that the cerebral circula-
tion and the vagal and sympathetic nerves remain
connected to the heart while the venous return, car-
diac output, ventricular volume, heart rate, aortic
and pulmonary resistances, atrial, ventricular and
arterial pressures, and the chemical composition of the
blood can be separately altered and controlled and
even cardiac biopsies made. Early in its use, Morawitz
& Zahn (272) developed a cannula for insertion into
the coronary sinus via the right atrium. The flow
through it was presumed to quantitate total venous
return from the vessels of the heart. Although this
idea was later shown not to be true, the investigation
was important for it enabled the experimenter to
study the coronary sinus fraction of coronary venous
outflow not only in the isolated heart but also in the
heart beating in situ. An artificial lung was substituted
by Evans et al. in 1934 (107), and since then the de-
velopment of such devices and preparations has been
rapid, permitting total coronary venous flow measure-
ment and fractionation of coronary sinus drainage and
noncoronary sinus drainage in the working and non-
working isolated heart. Some of the better arrange-
ments are as follows: a) Coronary venous drainage is
pumped through an oxygenator into the coronary
arteries, b) The heart is isolated in a manner similar
to the classical heart-lung preparation except that
instead of returning the blood to the right atrium and
through the lungs to the left atrium, the left ventricle
usually discharges its blood through a resistance into a
reservoir from which it returns to the left atrium. This
is a closed system except for the escape of blood
through the coronary vessels into the right heart which
receives no other blood (335). This coronary venous
blood may be ejected through the pulmonary artery
and collected, or it can be separated into coronary
sinus and non-coronary sinus fractions by coronary
sinus cannulation. In either case the blood goes into
the venous system of a donor dog whose arterial system
is connected to the reservoir, c) The isolated beating
heart doing no external work but with its nerves and
cerebral circulation intact may be studied within the
chest of dog (or man) by directing systemic venous
return through a pump oxygenator into the aorta, thus
bypassing the heart. Total coronary venous drainage
can be measured in the pulmonary artery or it can be
fractionated by also collecting separately coronary
sinus flow. The effect of systolic and diastolic ventricu-
1 528
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
lar distention on cardiac energetics and coronary flow
can also be gauged (in the dog) by inserting into the
left ventricle a balloon inflated to different degrees of
fullness (328). The blood may be oxygenated by
passing it through a mechanical oxygenator system, an
autogenous lung, or through a donor, human or
animal. It is believed that isolated hearts generally are
in varying degrees of failure (performance character-
istics less than those of a normal heart within the chest)
and that this can be prevented by a continued inter-
change of its blood with that of a supporting dog or
human. The latter arrangement has been used by
Sarnoff (335) and by Garcia-Ramos, Rosenblueth,
and their associates (126, 313). A possible explanation
for this phenomenon is the loss of myocardial cate-
cholamines in the isolated heart preparation and its
replenishment by the cross-perfusion technique (201).
In any case, in these varying types of isolated hearts
doing work, the coronary vascular bed is largely di-
lated, for the coronary flow is greatly increased and
the coronary A-V oxygen difference greatly decreased
over the values for hearts working within the chest.
Finally, it is of considerable interest that the heart
beat and carbohydrate metabolism of the isolated dog
heart can be maintained for prolonged periods by per-
fusion of the coronary circulation, not with blood, but
with gaseous oxygen (55).
Wiggers (796) was one of the first to study the hemo-
dynamics of the coronary circulation and energetics
of the heart beating and working in situ in the anes-
thetized open-chest dog, in which the coronary vessels
were naturally perfused from the aorta. By right heart
bypass, the extracoronary sinus venous drainage can
be quantitated. In this, systemic venous return by-
passes the right atrium and ventricle into a reservoir
from which it is pumped into the peripheral portion
of the pulmonary artery. The coronary venous drain-
age can be collected by a tube in the right atrium or in
the central portion of the divided pulmonary artery
(309)-
While in some of these preparations the coronary
circulation is naturally perfused from the aorta at its
prevailing pressure, for many investigations it is
desirable to have coronary artery perfusion at constant
controlled flow rates or at constant perfusion pressures,
or both, the pressures being different from the pre-
vailing aortic pressure. Various expedients have been
devised to achieve these ends. The simplest arrange-
ment is to connect the peripheral end of a coronary
artery or a branch to a blood reservoir at an appro-
priate elevation so that it drains into the artery by
gravity (249). In another arrangement, air expansion
chambers are used to permit constant pressure per-
fusion (153). Similarly, one end of a pump may be
connected to a local arterial source and the other end
of the pump to the coronary artery. Either the per-
fusion pressure or flow rate can be varied separately.
For more complicated systems, pump or pump
oxygenator systems such as already indicated for
right heart and total heart bypass are used, in which
the coronary perfusion pressure is constant. These ap-
proaches have the important advantage that they
enable the investigator to study separately the periph-
eral and myocardial factors that regulate flow. As a
further separation of those peripheral parameters
which determine flow and metabolism in the myo-
cardium, the coronary arteries of the isolated heart,
heart-lung preparation, and of the open-chest dog,
may be perfused at constant pressure or flow rate,
first while the heart is beating, and then while it is in
prolonged diastole as the result of stoppage from pro-
longed cervical vagal stimulation or intracoronary
injection of acetylcholine, potassium chloride or
citrate (17, 249, 311, 323).
Finally, with the advent of methodologies not re-
quiring use of anticoagulants or the insertion into a
vessel of a flow metering device, reasonably satis-
factory measurements have been made in the resting
and active dog (159, 212) and the resting human
(92. 3r9)-
Coronary Flow Methods (Animals)
phasic flow. These were designed with the hope of
analyzing the factors affecting coronary flow which
are too rapid in action to be studied effectively by
mean flow measurements. They record the instan-
taneous flow at the point of their insertion into a blood
vessel. The vascular bed of the heart is made up not
only of vessels within the myocardial wall, but also of
vessels lying on the surface of the heart. Since the
change in mean vessel bore during a cardiac cycle in
the superficial vessels (in which the flow measurement
is made) is presumably different from that of the
deeper vessels, such a device measures a combination
of "intramural" and ''extramural" flow. It does not,
therefore, necessarily indicate correctly the intramural
flow at all times. Comparison, however, of the arterial
blood pressure with the flow in late diastole in such
recordings is the only means known to the author by
which change in the active vasomotor state of the
coronary bed can be estimated when the coronary
arteries are naturally perfused from the aorta.
Some of the major earlier phasic flow methods
BLOOD SUPPLY TO THE HEART
!529
which were used on isolated hearts or anesthetized
animals have been: a) estimation of the phasic dif-
ference between the central and peripheral coronary
pressure curves during a cardiac cycle (151, 153);
b ) the recording of movement of the free end of a
bristle mounted in the wall of a tube of fixed diameter
through which coronary flow occurs (302); c) meas-
urement of cooling by air of a heated platinum wire
mounted in the neck of a bottle partially filled with
blood, the lower part of which is connected to the
coronary circulation (10); d) measurement of cyclic
movement of various foreign substances (toluene,
mercury droplet) inserted into a coronary artery (10);
e) recording of the small pressure drop in a pressurized
air-blood chamber as blood flows from the base of the
reservoir into a coronary artery (93, 153);/) measure-
ment of the lateral pressure difference above and
below an area of constriction (orifice) in a metal tube
inserted into a coronary artery (153); g) recording of
the upstream and downstream pressure difference in
a metal tube inserted into the coronary sinus (196).
Finally, the electromagnetic flowmeter has been
successfully applied to the coronary circulation in the
dog (217, 395). A square-wave type of electromag-
netic flowmeter has been found quite useful in coro-
nary flow studies in open-chest sacrifice dogs (81 ), but
because of their necessary size, they have not yet been
chronically implanted on the coronary arteries. The
sine-wave type can be miniaturized, and flow trans-
ducers of aspirin-tablet size or smaller have been suc-
cessfully applied for periods of weeks to the right
coronary artery, the main left coronary artery and its
major branches, of the conscious and active dog (212).
For further discussion of flowmeters see Chapter 38
in this Handbook.
mean flow. The most accurate measurement of mean
coronary venous outflow is by its collection in a
graduate, and of coronary arterial inflow by reading
the graduations on a calibrated reservoir. More
sophisticated devices have been developed and applied
to dog and man. Some of the more important in the
dog are: a) timing visually or photoelectrically the
passage of an air bubble through a glass tube of known
length and volume which is placed between the cut
ends of a coronary artery through which flow is being
measured [bubble flowmeter (90)]; b) recording the
position of a '"float" in a vertical tapered tube through
which coronary blood is flowing [rotameter (153,
345)]; c) recording the temperature difference of two
thermojunctions mounted in a plastic sleeve of con-
stant cross section through which coronary blood is
flowing (thermostromuhr — though its ultimate re-
liability in many circumstances has been questioned)
(153); d) recording the heat clearance or the tempera-
ture difference of a reference cold thermocouple and
an electrically heated thermocouple inserted into the
myocardium ( 14;^ ).
Coronary Flow (Man and Animals)
Variations of the Fick principle and coronary cine-
angiography have been used to attack the coronary
blood flow problem in man. The first major advance
came with the use of nitrous oxide inhalation for
determining blood flow draining into the coronary
sinus. As compared to direct measurement of coronary
blood flow, the method [see previously cited reviews:
(35, 92, 154, 319)] shows a reasonable accuracy, and
in humans has furnished almost all our information
regarding coronary blood flow. Another variation of
the Fick principle has been used to estimate myo-
cardial blood flow in the animal and in man. Studies
in the rat and dog have indicated that when intra-
venous slug injections of the radioisotopes K42 or Rb8B
are made, the following occurs: the isotopes have a
large volume of distribution within the myocardium,
and, for at least 1 min after a single intravenous
injection of the isotope, their coronary venous drain-
age is negligible compared with their initial deposi-
tion; the extraction ratios of the heart and whole body
for the isotope are identical. By determining cardiac
output by means of this isotope injection, and at the
same time determining the fraction of the injected
isotope taken up by the myocardium (animal sacrifice
and direct counting) within this minute, it is possible
to estimate total myocardial blood flow in the dog
and rat (185, 234, 330). By comparing the isotope
concentrations in different myocardial areas, the
regional flow distribution can also be estimated. These
results could have a reasonable accuracy. The obsta-
cles, however, to the use of such a method in man
without coronary sinus catheterization are formidable.
While the isotope is being infused intravenously at a
rate designed to keep a constant arterial concentration,
it might be possible to estimate, by radiation detection
over the precoidium and by direct counts on the
blood, the increments in myocardial Rb86 content and
its concentration in the coronary sinus blood. How-
ever, the isotope extraction at different coronary
blood flow rates is not constant (reported extractions
vary from 40 to 70%) and may vary with duration of
the perfusion. As yet, these difficulties have not been
resolved (245, 274).
'53°
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
The indicator dilution technique for coronary blood
flow in man is based on the fact that, when cardiac
output is being estimated by means of a device placed
over the precordium to pick up the specific activity of
a tracer substance such as I131 following rapid intra-
venous injection, the curve produced during the first
circulation of the radioactivity, in addition to having
two well-defined peaks representing passage of radio-
activity through the right and left sides of the heart,
respectively, may also have a third small peak closely
following that attributed to left ventricular activity.
This appears at a time which could represent myo-
cardial blood flow (343, 381 ). Unfortunately, because
of an insufficient time lag, it is difficult to differentiate
the peak of precordial radioactivity related to myo-
cardial flow from other rapid changes in precordial
activity, such as that resulting from the preceding
passage of blood through the left side of the heart, or
that due to subsequent recirculation from the most
rapid noncoronary circuits (252). Until the true
coronary precordial peak in radioactivity can be more
sharply defined, it is difficult to place reliance on data
obtained with this method as representing coronary
flow. A variation of the application of the isotope
dilution technique to the problem arises from the
observation in both humans and dogs that concentra-
tion curves of radioactivity recorded over the heart,
following rapid intravenous injection of small boluses
of I131, have slower disappearance rates than those
obtained by sampling directly from a peripheral
artery, such as the femoral (260), the difference being
due presumably to the coronary flow.
Finally, special techniques have been developed
for selective catheterization of individual coronary
arterial branches in intact closed-chest dogs and
unanesthetized human subjects by means of special
catheters which permit intracoronary injection of
radiopaque materials and roentgenological visualiza-
tion of the coronary vessels (352, 393). Coronary cine-
angiography is, at the present time, in a somewhat
embryonic stage of development as a research or
diagnostic method. Perfusion with the media in ap-
propriate volume and concentration apparently does
not result in development of anginal pain, EGG evi-
dence of myocardial ischemia, or photographic evi-
dence of coronary vasoconstriction. However, funda-
mental hazards in the application of this technique
lie in the possibility of inadvertent mechanical occlu-
sion of a coronary artery with the catheter tip and the
known moderate vasodilator and cardiotoxic proper-
ties of the contrast media. From an investigative point
of view, coronary arteriography, even when recorded
continuously by motion picture photography of the
fluorescent screen, does not provide a measure of
coronary flow and vascular resistance. Such data can,
however, provide evidence of change in size and
number of visible arterial vessels after administration
of various physiological and pharmacological agents.
In the presence of a constant blood pressure, such
changes would indicate local changes in the vasomo-
tive states although it cannot be currently determined
whether these alterations are active or passive. In
addition, it is used to study the existence of and change
in intercoronary arterial collateral channels in life,
since the origin and distribution of collateral channels
as small as 100 ju can be well demonstrated. From a
diagnostic point of view, selective opacification of
individual coronary arteries provides information on
the length and exact location of partial and complete
occlusive lesions in major vessels as small as 1 mm in
diameter.
DISTRIBUTION OF MYOCARDIAL BLOOD FLOW'
Arterial Circuit
As blood is ejected from the left ventricle, it simul-
taneously enters both coronary ostia and flows via the
epicardial coronary arteries to their respective myo-
cardial beds. By direct measurement in open-chest
dogs, the left coronary and right coronary arterial
inflow approach 85 and 15 percent, respectively ( 1 53)-
The same relationship also exists in the dog heart-lung
preparation and perfused, fibrillating heart, lending
physiologic support to the anatomically designated
left coronary artery dominance in dogs. While the
direct measurement of coronary arterial distribution
in man is unknown, coronary arteriography in un-
anesthetized patients has demonstrated variations in
volume of the various coronary arterial beds which
correspond quite well with postmoitem anatomic
studies (84).
Utilizing the bubble flowmeter (90) in both open-
and closed-chest acute experiments in dogs, an average
left coronary inflow of 65 ml per 100 g left ventricular
tissue per minute was found with no significant dif-
ference between the two groups. The left circumflex
was found to supply an average of 40 per cent, and the
anterior descendens about 26 per cent by weight of the
left ventricle. Similar values have been obtained with
rotameters and in the intact unanesthetized dog with
electromagnetic flowmeters. Figures for the contribu-
tion of the anterior septal artery flow in the dog can
BLOOD SUPPLY TO THE HEART
I531
be estimated by observing the decrease in total left
coronary inflow after occlusion of the septal artery or
by direct cannulation (265 and Gregg, unpublished
observations). In either method, the volume of flow
is from 11 to 21 per cent of total left coronary flow.
Radioactive cations (Na22, K42, Mg2S, Rb86, Fe56)
and anions (P32, I131) and D20 have been applied to
the coronary circulation as a means of determining
distribution of blood and plasma flow, and metabo-
lism of the involved myocardial bed. Tissue uptake
and turnover rates of the radioactive substances have
revealed a heterogeneous myocardial distribution (21,
122, 1 97, 234, 244, 245). All left ventricular areas in-
cluding base, apex, septum, and free walls have a 50 to
100 per cent higher uptake and turnover rate than
the right ventricle and atria. In descending order of
activity are the right ventricle, left atrium, right
atrium, His bundle and, lastly, the sino-atrial and
atrioventricular nodes. In most instances, the myo-
cardial uptake is nearly instantaneous since a plateau
is reached after a single systemic circulation and,
thereafter, remains relatively constant with only-
minor differences between the 20-sec and 10-min
determinations. D20 similarly reaches equilibrium
between plasma and tissue water after a single circula-
tion and can also be calculated within 10 to 20 sec
following injection (197).
Radio-rubidium (Rb'*6) has been found to be the
most versatile for myocardial flow determinations
because of its long half-life (T1 2 = 19.5 days), rapid
myocardial uptake (in exchange for intracellular
potassium), and relatively fixed myocardial extraction
despite varying arterial concentrations (230, 245). In
addition to the tissue concentrations, Rb86 and Na22
and D20 have been used for coronary blood flow
determinations, and in those instances where checks
against a standard reference method (i.e., N20 and
flowmeters) were done, good correlations were ob-
tained (274). Flow values vary from 0.4 to 1.6 ml
per g per min, with an average of 0.7 to 1 .0 ml per g
per min for dog and man, while in rats values four
times this have been found, supposedly related to the
four-fold greater energy output of the rodent myo-
cardium, i.e., 1 .00 joules per g per min versus 0.27
joules per g per min (185).
The Ye,
Circuit
In addition to the regional differences in rate of up-
take, there also exists a concentration gradient be-
tween the endocardial and epicardial surfaces, the
former having the higher uptake and turnover of
radioactive cations (244). The disparity is most
marked in the right ventricle since the concentration
of Thebesian vessels is highest in this chamber, and
also, a favorable pressure gradient exists for blood to
flow from the myocardium to the cavities during
systole. It has therefore been argued that this is sup-
portive evidence for utilization of the deep vascular
communications of the heart. The role played by the
deep vascular structures, however, is probably quite
small for several reasons. Balance studies in which an
attempt was made to measure coronary inflow and
outflow simultaneously with rotameters in the super-
ficial coronary vessels of the open-chest dog have
shown that a) coronary sinus flow ceases when both
the right and left coronary arteries are occluded with
the heart beating in situ; b) the left coronary artery
accounts for all but 5 to 10 per cent of coronary sinus
outflow; c) 80 to 85 per cent of left coronary inflow is
reflected in the coronary sinus outflow while some of
the remainder is accounted for by the anterior cardiac
veins; d) 90 per cent or more of the right coronary
inflow drains via the anterior cardiac veins; and
e ) there is no evidence of significant Thebesian drain-
age of the right coronary system (153). These studies
in the open-chest dog are technically quite difficult
and although recovery is usually of the order of 80 to
85 per cent (300), comparison of total coronary inflow
with outflow in the superficial veins is subject to
considerable error. At the same time, in other experi-
ments following acute coronary sinus ligation it was
observed that although the lateral wall of the left
ventricle was markedly congested, portions of the
interventricular septum showed less evidence of con-
gestion. This observation of 20 years ago was not fol-
lowed up until recently when it was found that the
portion of left coronary inflow (about 15%) not re-
covered in the coronary sinus could be largely ac-
counted for by the fact that a portion of the left
anterior atrial artery flow drains into the left atrium,
and that most of the septal artery and some branches
of the left descendens artery which perfuse the septum
drain into the right ventricular cavity (265, 266). The
finding concerning drainage of the left atrial coronary
flow is in line with observations with an illuminated
cardioscope in humans and dogs at the time of cardiac
surgery, that very small streams of dark blood can be
seen entering the left atrium but not the left ventricle
(53)-
The deep drainage channels could have an im-
portant functional role if they served as arterial
channels from the left ventricular cavity to the myo-
cardium during coronary artery constriction or
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
occlusion, or as venous channels for the whole myo-
cardium in the presence of extensive superficial vein
constriction or occlusion. Regarding the first situation,
although essentially complete occlusion of the coro-
nary arteries in human beings has been found at
autopsy (227), the presence or extent of development
of extracardiac arterial collaterals is not known. In
addition, with temporary functional separation of one
or both coronary arteries from the aorta, no blood
flow from the ventricles into the superficial coronary
venous system can be demonstrated and the hearts do
not survive (153). When dye is injected into the right
ventricle in acute experiments, extensive capillary
injection on the surface of both ventricles occurs if
right ventricular pressure is artificially made to exceed
left ventricular pressure (153). Although this could
have occurred through the Thebesian channels of the
interventricular septum (265), the anterior cardiac
veins were not excluded as a portal of entry for the
dye. Regarding the second situation, with acute
closure of all grossly visible anterior cardiac veins or
of the coronary sinus, or both, a considerable reduction
in right and left coronary inflow occurs (153). Al-
though the heart, following acute closure of both the
coronary sinus and anterior cardiac veins, becomes
exceedingly hemorrhagic and progressively weaker,
such hearts may survive up to 2 hours. Dogs in which
both superficial venous systems have been chronically
occluded in a two-stage operation have survived for
periods of months. However, that significant drainage
occurs through such a route could not be verified,
since, at postmortem examination, these hearts
exhibited numerous superficial cardiac veins of con-
siderable size which were not previously apparent, and
several large extracardiac venous anastomoses, the
aggregate cross section of which was estimated to be
adequate for venous drainage of the entire heart ( 1 53).
Until the intracardiac and extracardiac arterial and
venous collaterals which appear with coronary arterial
or venous ligation have been excluded as flow chan-
nels, any conclusion regarding the utilization of deep
coronary venous drainage channels in diseased hearts
is difficult to reach.
Possible Use of Left Coronary Artery Flow Together
with the Chemical Composition of Coronary Sinus
Blood as an Index of Left Ventricular Metabolism
It is not possible to quantitate accurately the
metabolism of the right ventricle in dog (or man)
because its superficial anterior cardiac veins have
many exits into the right atrium and their contained
blood is grossly contaminated by blood from the left
coronary artery. However, a large drainage of the
left myocardium occurs into the coronary sinus and
the latter is accessible. Hence, the question of whether
the chemical composition of coronary sinus blood
together with left coronary inflow can be used as an
index of quantitative changes in metabolism of the
left ventricle is a very practical and important con-
sideration because of the widespread use by the basic
experimenter and the clinical investigator of these
measurements for this purpose. To justify such usage,
experimental evidence must show, first, that most of
left coronary inflow drains into the coronary sinus
and that the latter is not significantly contaminated
by drainage from the right coronary artery and,
second, that its chemical composition approximates
that portion of the blood coming from the left coronary
artery which does not flow through the coronary
sinus.
In the open-chest dog in which no great effort is
made to avoid obstruction at the ostium, the per-
centage recovery in the coronary sinus of left coronary
inflow varies from 64 to 83 per cent in any one dog,
and shows little variation from dog to dog (153)- By
use of a special cannula which collects all the blood
draining into the coronary sinus without obstruction
to any of its veins, the percentage of left coronary
artery inflow recovered in the coronary sinus is quite
high (80-90 %) and reasonably constant during the
induction of a variety of physiological variables and
drug injections (300). In the open-chest dog, the
right coronary artery contributes not more than 2 to 3
per cent, or 1 to 2 ml per min, to the coronary sinus
flow, and this only occasionally. This has been deter-
mined by observing minimal changes in coronary
sinus flow when the right coronary artery is clamped
in the presence of an elevated right ventricular pres-
sure from pulmonary artery stenosis, when right
coronary artery clamping is superimposed on a pre-
existing occlusion of the left coronary artery (153),
and by observing only minimal changes in the optical
density of coronary sinus blood following massive
injection of Evans blue dye into the right coronary
artery (300).
The investigation of whether the coronary sinus
fraction of blood is representative in chemical compo-
sition of total left coronary venous return started with
the experiments of Evans & Starling in 1913 (106) and
has continued to the present time. Actually, investiga-
tions during this period did not directly attack the prob-
lem (158). In these experiments, the effect of increased
right ventricular pressure was determined on flow
BLOOD SUPPLY TO THE HEART
'533
and oxygen content of coronary sinus blood and of the
remaining coronary venous blood including that
from the right coronary artery. Obviously, these ob-
servations are germane only to the problem of whether
an increase in right ventricular metabolism asso-
ciated with increased right ventricular pressure is
reflected in the coronary sinus blood (195, 264). This
might not be expected because of the very small
drainage of right coronary flow into the coronary
sinus. These experiments are certainly not germane
to the problem of whether the two coronary venous
drainage fractions from the left coronary artery have
the same chemical composition, for such measure-
ments were not made.
This question for the left myocardium has been
answered by simultaneously and continuously meas-
uring, under different circumstances, left coronary
artery flow, and the flow and oxygen content of the
two coronary venous fractions derived from the left
coronary artery. In these experiments, the systemic
venous return bypassed the right heart, and the right
coronary artery was generally clamped. The oxygen
uptake calculated on the basis of left coronary artery
flow times the difference between the arterial and
coronary sinus oxygen content agrees quite well with
the oxygen uptake based on the sum of the respective
volume flows and the oxygen content of the two left
coronary venous drainage fractions. This is effected
by a combination of a generally lower oxygen content
in the coronary sinus and a considerably greater
coronary sinus flow (300). Hence, in the open-chest
dog a combination of left coronary artery flow and
coronary sinus arteriovenous oxygen difference gives
a reasonably precise value for uptake of oxygen by
the left ventricle.
From data such as these it has been reasonably-
assumed that measurement of coronary sinus flow
could be substituted for left coronary inflow and,
together with the coronary sinus, A-V oxygen differ-
ence could also serve as an index of metabolic events
in the left myocardium of man and beast. The authors,
however, in no way recommend this procedure.
Although widely used in man, it has never been
demonstrated that the flow, composition, and sources
of coronary sinus blood fulfill the requirements as laid
down and found to exist in the dog. [Actually in
early experiments with the isolated dog heart sig-
nificant right coronary artery drainage into the
coronary sinus was demonstrated (92).] In addition,
accurate measurement of coronary sinus flow is
extremely hazardous whether done indirectly by
means of the nitrous oxide method or directly by
cannulation. In the first case there is the ever present
danger of contamination with right atrial blood. In
the second instance, without knowledge of the in-
vestigator, coronary sinus flow may be reduced 1 >\
shrinkage and partial closure of the sinus. This
diminishes only slightly the left coronary inflow,
which now drains preferentially by the anterior
cardiac veins.
PHYSICAL DETERMINANTS OF CORONARY FLOW
Coronary flow is related to the pressure difference
(effective pressure) between the central coronary
artery (identical to aortic pressure) and the right
atrium divided by the sum of the viscous resistances
to flow in the epicardial portion of the artery and in
the peripheral coronary bed. Viscous resistance to
flow, aside from change in hematocrit, is mainly
governed by the mean caliber of the coronary vascular
bed. Since the arterial resistance is negligible, the
mean coronary diameter and, hence, flow are con-
trolled by the effective intravessel pressure and by two
peripheral mechanisms, i.e., active changes in the
state of the small mass of intramural smooth muscle
built into the coronary vessels, and the mechanical or
passive effect on flow exerted during ventricular
systole by the large muscle mass around the coronary
vessels.
Insight into the complexity of the integrating action
of central and peripheral flow determinants has been
obtained from the recording of the peripheral coro-
nary pressure and the phasic or moment-to-moment
changes in coronary inflow and outflow in the epi-
cardial arteries and veins (151, 153, 158, 212, 301).
These curves were obtained from the open-chest dog
and from the resting unanesthetized dog some days
postoperatively, after implanting an electromagnetic
flowmeter on the left coronary artery (fig. 6). At the
onset of isometric contraction of the left ventricle in
the unanesthetized dog, there is an abrupt decrease in
left coronary inflow and, although at times backflow
may appear, a considerable forward flow generally
persists throughout systole. With the rise in aortic
pressure, forward flow increases initially and rapidly,
only to decrease to a new intermediate level in late
systole. With the onset of isometric relaxation,
coronary flow increases significantly, peaking at
early diastole and then declining progressively. These
demarcations of flow are much less obvious in the
right coronary inflow pattern, which roughly re-
sembles the prevailing aortic pressure curve. The flow
[534
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
AORTIC PRESS
LEFT COR FLOW
CARDIAC OUTPUT
105 mm Hq MEAN
225 ml/min
90 ml/IOOg/min
2 8 ml/HEART
BEAT
3000 ml/mm
37 ml/HEART
BEAT
HEART RATE 80
fig. 6. Reproduction of a retrace of an original record taken
in the conscious dog, 14 days postoperative, showing phasic
aortic blood pressures recorded by a strain gauge connected to
a chronically implanted aortic catheter, and phasic left coronary
artery flow and stroke cardiac output by electromagnetic flow-
meters chronically implanted, respectively, on the main left
coronary artery and ascending aorta. (Unpublished observa-
tions.)
patterns just indicated for the conscious dog are
similar to those in the open-chest dog except that in
the left coronary artery of the latter, systolic flow is
minimal and backflow is usually present during iso-
metric contraction (153).
The flow patterns of the left coronary artery are a
complex of events happening in the total distribution
of flow in the left myocardium and a small portion of
the right ventricle. Regional variations of flow pattern
might be expected based on anatomical and func-
tional differences in the areas supplied. Flow patterns
of the main left coronary and its circumflex and
descendens branches are essentially similar. Phasic
flow, however, in the left anterior atrial artery shows
a forward flow in both systole and diastole with the
flow pattern resembling an aortic pressure pulse (349).
About 40 per cent of this arterial flow (5 '"< of left
circumflex flow) drains into the left atrium (266).
Patterns of flow through the canine septal artery are
not available. It would, however, be predicted that
the pattern would differ from that in the circumflex
and descendens by having a much smaller systolic
flow since this artery has essentially no epicardial
component. Most of the flow in this very small artery
drains into the right ventricle (265).
The finding of a significant and variable coronary
flow during systole in the left coronary artery of the
unanesthetized dog deserves further comment. In the
past, the view based on work in the open-chest dog
has been that flow in the left coronary artery is very
small during systole, that it does not vary significantly
with different dynamic conditions, and that it can
be accounted for largely on the basis of radial en-
largement of the epicardial vessels and their filling
during ventricular contraction (153). This meant
that events in systole could be and were largely ig-
nored and that the only important considerations for
regulation of left coronary flow were happenings
during diastole. Recent work using chronically im-
planted electromagnetic flowmeters indicates that
although the coronary flow in systole in the unanesthe-
tized dog at rest can, at times, be rather small, in
many dogs it may approximate 30 per cent of that
during diastole. In the presence of mild exercise, it
does not appreciably increase, but following release of
coronary artery occlusion, and during excitement and
chronic stimulation of the cardiac sympathetic nerves,
the volume of systolic flow increases 300 to 400 per
cent, as does the diastolic flow, the ratio between the
two remaining about the same (139, 212, 301 ). Finally,
in irreversible hemorrhagic shock, late in the period
of spontaneous cardiovascular decay after blood
reinfusion, the systolic flow may approach that during
diastole for an equivalent time interval, and eventually
the flow pattern may resemble somewhat the prevail-
ing aortic pressure pulse with most of the coronary
flow occurring in systole rather than in diastole ( 1 59)-
The proper explanation of these findings awaits
future experimentation (fig. 7).
The preceding account indicates that the coronary
bed has a fluctuating resistance to flow. Flow curve
inspection shows the obvious importance of left
ventricular contraction in controlling coronary flow,
because during systole left coronary flow is reduced
while coronary sinus flow is increased. The increase
in coronary sinus flow suggests that ventricular con-
traction acts to aid coronary flow by massaging blood
through its wall; the reduction in coronary inflow sug-
gests that it acts to throttle coronary flow. The
answer depends upon the relative changes of inflow
and outflow volume during systole. Unfortunately,
this is impossible to determine because of the incom-
plete and variable drainage of the left coronary artery
through the coronary sinus. However, actual meas-
urements in the left coronary artery of the open-chest
dog show that the peripheral coronary maximal
systolic and minimal diastolic pressure values approxi-
mate 80 20 mm Hg, and inflow is cut off at these
pressure levels when the left coronary artery is per-
fused through its distal end under constant pressure.
BLOOD SUPPLY TO THE HEART 1 535
BLOOD
PRESS
mm Hg
®
19
21
©
CIRC.
COR
FLOW
ml/min
0
/W-l_ P*
FLOW
STROKE
VOLUME
ml
Hr
AzA.
A^i_A
fig. 7. Reproduction of retraces from an original record taken in a resting unanesthetized dog
some days postoperative showing the effect of irreversible hemorrhagic shock on phasic blood pressure
and phasic stroke left circumflex coronary flow, using a strain gauge and electromagnetic flowmeter
as in fig. 6. A — early; B — midway; C — late in the period of spontaneous hemodynamic decay fol-
lowing reinfusion. (L'npublished observations.)
In the right coronary artery, the contour and time
relations of the peripheral coronary pressure curve
are similar but the values for systole and diastole and
for the cut-off of flow are considerably lower ( 1 53).
Separation and quantitation of the determinants
of coronary flow lying within the myocardial wall,
i.e., the vascular and extravascular muscle, are of
extreme importance. Various methods have been
proposed and used, but they have been only partially
successful. The problem of determining the relation-
ship of blood flow to active vasomotor changes,
irrespective of whether the effect on the intrinsic
muscles of the coronary vessels is mediated through
the blood stream or is secondary to metabolic changes
in the surrounding myocardium, is especially difficult.
It is not known how much coronary flow might change
with a given change in coronary perfusing pressure
without an associated active change in the vasomotor
state of the bed. Determination of active variations in
vasomotor tone in the coronary bed is further compli-
cated by uncontrollable mechanical factors. Varia-
tions may occur in the respective durations of systole
and diastole during which the rates of flow per unit
of time may be quite different and thus obscure any
active vasomotor changes.
By analysis of phasic inflow curves, however,
change in the vasomotor state can be separately and
roughly estimated. A critical point on a coronary
inflow curve is selected in late diastole, at which
time the rate of change of the volume-elastic and
myocardial compression forces is presumed to be
minimal (153)- At this point, extravascular forces are
at a minimum, the rate of flow reflecting the vaso-
motor state of the coronary bed, and the ratio of the
aortic pressure to the simultaneously existing rate of
flow is then determined. A shift in the diastolic ratio
is taken to represent active constriction or dilatation
of the coronary bed (41, 146). It has also been sug-
gested that change in the extravascular compressing
force during systole can be estimated by comparing
the diastolic ratio with the ratio of blood pressure to
coronary flow at a point in late systole when extra-
vascular support is maximal and flow reflects the
combined effect of myocardial compression and the
existing vasomotor state (146). At this time, the rate of
change of the volume-elastic and myocardial com-
pression forces is presumed to be minimal. Use of
such a systolic point has as yet no experimental
verification.
The problem of determining the magnitude of
extravascular support has been approached in differ-
ent ways. It has been suggested that intramural
pressure can be used as a measure of extravascular
compression, and attempts have been made to quanti-
>536
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
tate the pressure developed within the wall of the left
ventricle during systole and to use it as a measure of
extravascular support. To do this, pressure pulses
have been recorded from a myocardially imbedded
vessel (or myocardial fluid pocket connected to a
recording manometer). However, experimental work
indicates that although these pressures may indicate
directional changes in extravascular compression,
they are, in part, artifactually produced and, hence,
do not approximate the correct values for intramural
pressure (153).
A method recently developed has given some in-
formation on this point (323). Continuous measure-
ments are made in the open-chest dog while the left
coronary artery is perfused with blood under a con-
stant pressure. First it is done in the beating heart,
and then during ventricular asystole induced by
vagal stimulation, or by disconnecting an external
pacemaker which drives the ventricles (complete
atrioventricular heart block having been surgically
produced previously). By either means, the mechani-
cal effects of ventricular contraction are largely
removed. Induction of ventricular asystole by vagal
stimulation always increases immediately (within
1 sec) left and right coronary inflow. Thus, ventricu-
lar contraction acts to impede coronary flow through
the ventricular wall. The extent of the rise of flow is
taken to represent the magnitude of the mechanical
or passive factors limiting coronary flow. The mag-
nitude of this mechanical throttling effect on cor-
onary flow during systole normally varies from
31 to 300 per cent and averages about 50 per
cent. The new flow level represents that state of
coronary dilatation related to the condition of the
intrinsic smooth muscle of the coronary vessels at
the prevailing coronary pressure. The relative con-
tribution of extravascular and intravascular re-
sistance to an increase of coronary flow has been
tested under the different conditions of increasing
heart rate, decreased arterial blood oxygen satura-
tion, aortic constriction, transfusion, and drug injec-
tions. In all instances, the major portion of a flow
increase is through active dilatation and not through
reduction in extravascular resistance. The largest
reduction (40 % ) in extravascular resistance is from a
decrease in arterial oxygen saturation (155, 236).
DETERMINANTS OF NORMAL MYOCARDIAL METABOLISM
The ability of the heart to do work depends basi-
cally on its biochemical activity leading to muscular
contraction. Cardiac muscle has been found to have
basic chemical patterns similar to those of other
muscle. The catabolism of fat, carbohydrate, and pro-
tein produces free energy, about half of which is
dissipated as heat and half is captured as phosphate-
bond energy which is used for muscle cell work and
for various anabolic activities such as synthesis of
glycogen, lipids, proteins, and enzymes. These cata-
bolic and anabolic reactions proceed simultaneously
under the influence of a complex system of enzymes,
coenzymes (from the vitamin B complex), and
hormones.
Coronary sinus catheterization studies in man and
dog have indicated that the heart is able to choose its
fuel from a variety of foodstuffs. These include
mainly glucose, lactate, pyruvate, fatty acids (non-
esterified) and, to a lesser extent, acetate, ketone
bodies, and amino acids. To determine their quanti-
tative contribution to the energy production of the
heart, i.e., its oxygen consumption, measurements
have been made of their cardiac extraction (coronary
artery — coronary sinus difference), their total uptake
[coronary flow X (coronary artery — coronary sinus
difference of substance)], and the myocardial respira-
tory quotient (coronary sinus — arterial carbon dioxide
difference; coronary artery — coronary sinus oxygen
difference). Excellent correlation has been demon-
strated between the myocardial respiratory quotient
and the myocardial uptake of substance. The extent
to which each substrate contributes to the energy
requirement of the heart in vivo is influenced by its
concentration (above threshold) in arterial blood. In
addition, the state of nutrition of the organism mark-
edly influences the kind of substrate used for energy
production of the heart. Under postprandial condi-
tions, or after glucose infusion, myocardial metab-
olism is mainly glucose, lactate, and pyruvate, since
its respiratory quotient approximates 0.9 with a high
extraction of carbohydrate and a negligible uptake of
amino acids. Even the substitution of 5 to 10 per cent
oxygen for the normal 2 1 per cent in the inspired air
does little to change carbohydrate uptake by the
normal heart. During overnight fasting, the heart
derives much of its energy from fat, as indicated by a
myocardial respiratory quotient of 0.80 with a low
extraction and uptake of carbohydrate. With pro-
longed fasting, the extraction coefficient for carbo-
hydrate practically disappears, those for fatty acids
and ketones are maximal and the respiratory quotient
is 0.70. As regards the uptake of oxygen, the coronary
A-V oxygen differences in man vary linearly with the
arterial oxygen content through a range from mild
BLOOD SUPPLY TO 1 UK HEART
[537
anemia to marked polycythemia so that the myo-
cardial extraction coefficient (A-V), A is constant.
In addition to patterns of myocardial metabolism in
the normal heart, other metabolic changes have been
reported in some pathological and diseased states.
Patients with heart failure and decreased cardiac
work due to valvular disease show an increased carbo-
hydrate uptake by the heart with a normal extraction
of lactate and pyruvate and increased glucose exti ac-
tion. The heart in the patient with diabetes appears
to derive most of its energy from fat even in mild
cases with a postabsorptive respiratory quotient ol
about 0.7 and an increased uptake of fatty acids and
a decreased carbohydrate uptake.
Thus, the heart demonstrates broad flexibility in
the utilization of substrate for energy production
without a change in work performance or work
capacity. This makes it largely independent of fluctua-
tions in its chemical environment. There is no evi-
dence that substrate lack occurs in any clinical
situation to the extent that it embarrasses the cardiac
work capacity. Similarly, the metabolic disturbances
such as diabetes mellitus which alter the fuel mixture
available to the heart do not also alter cardiac func-
tion. It is, however, well to defer detailed considera-
tion of other data because an interpretation must be
based on the assumption that oxidation of foodstuffs
to carbon dioxide and water is the sole factor in the
determination of the myocardial respiratory quotient
and of the myocardial extraction and uptake of these
compounds including oxygen. Without doubt, stor-
age of and or conversion into other compounds is
occurring concurrently, and these activities are
expecially prominent in the presence of a changing
cardiac level of activity or changing levels of blood
substrate (16, 32, 33, 74, 1 16, 133, 169, 278, 279).
BASAL DATA
In the resting state, the coronary data for dog and
man agree. With the left ventricular cardiac work
index approximating 3.0 to 4.6 kg-m, left coronary-
flow approximates 72 to 85 ml per 100 g of left
ventricle per min (118, 153, 307). In the anesthetized
open-chest dog, values as high as 600 ml per 100 g
left ventricle per min have been recorded when the
left heart has been stressed by a combination of
catecholamine injection and aortic constriction (344).
Left coronary flow values in the unanesthetized dog
during maximal natural stresses are not yet available
but during moderate treadmill exercise and following
excitement, the coronary flow has approximated
that in the open-chest dog (212). As indicated under
physical determinants of coronary flow, the frac-
tionation of the volume flow between systole and
diastole is somewhat variable, but in the left coronary-
artery of the unanesthetized dog the systolic volume
flow very often approximates 25 to 30 per cent of the
diastolic flow under semibasal conditions, as well as
during excitement, exercise, and reactive hyperemia
(i59a)-
In the anesthetized dog, the circulation time from
the central coronary artery to the coronary sinus
approximates 4.5 sec (260). In normal patients, the
coronary transit time (with I131 injection) varies from
6.5 to 1 1 sec. Exercise and nitroglycerin, which in-
crease coronary flow (nitrous oxide method), decrease
the transit time (increased coronary flow velocity)
while the Valsalva maneuver, which increases the
circulation time, decreases coronary flow (135).
Flow values for the right coronary artery in a good-
sized open-chest dog approximate 10 to 15 ml per
min. In the resting, unanesthetized dog, the values
are similar (unpublished observations). The volume of
systolic flow generally exceeds the diastolic volume
flow for an equivalent time period and very often ex-
ceeds total diastolic flow (153, and unpublished obser-
vations). Values per gram of myocardium and the
flow responses to natural stresses of everyday life are
not known.
Although each ventricle can remove essentially all
oxygen from the coronary blood in its passage through
the myocardium, normally, for the left ventricle
(also the right), about two-thirds is extracted with an
arteriovenous difference of 1 1 to 14 ml, and a coro-
nary sinus value of 5 to 6 ml. This extraction changes
little, i.e., less than 10 to 20 per cent with increased
stress (except following catecholamine injection,
anoxia, and anemia, in which it decreases), indicating
that the oxygen supply is well balanced with metabolic
demands (208).
Oxygen uptake per 100 g left ventricle (coronary
flow X coronary A-V 02 difference) is 8 to 10 ml
per min in the open-chest dog, the anesthetized closed-
chest dog with normal blood pressure and cardiac
output, and in the resting unanesthetized dog and
human. Maximum values calculated in the open-
chest dog approximate 60 ml per 100 g per min. In
the unanesthetized active dog under the influence of
mild exercise and excitement, values are not avail-
able.
With present poor methodology, separation of oxy-
gen usage between systole and diastole can only be
'538
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
made by measuring oxygen uptake, first in the beating
heart during repetitive systoles and diastoles, and then
in the relaxed heart or during prolonged diastole,
thus obtaining the oxygen usage during systole by
difference. Estimation of the metabolism of the myo-
cardium in the absence of a heart beat, that is during
prolonged diastole, has been made in the vagus-ar-
rested heart (see section on Physical Determinants of
Coronary Flow). The oxygen saturation of the arterial
blood and coronary sinus blood is also measured
continuously. This permits left coronary arteriovenous
oxygen difference as well as coronary inflow to be
measured continuously, first in the beating heart and
then in the stopped heart until a new equilibrium is
established, usually within 20 to 25 sec (155, 156,
249). As coronary inflow rises immediately with
asystole, the oxygen saturation of blood in the coro-
nary sinus also rises, thus greatly reducing the coro-
nary arteriovenous oxygen difference. Calculations
in many experiments show that as the result of the
combination of an increased coronary flow and a
decreased coronary arteriovenous oxygen difference,
the oxygen usage per 100 g of left ventricle per min
decreases from the average control level of 8. 1 ml in
the working beating heart to 2.3 ml in the resting
heart, or to 30 per cent of the control. This oxygen
consumption in diastole is about one-third that in
systole for an equivalent time period (249).
Attention is also directed to the values for oxygen
usage obtained in the same type of preparation but in
which the external work of the heart is reduced to
zero by other means. In the potassium-stopped heart,
the oxygen usage of 2 ml during diastole is about the
same as in vagal asystole. In the beating heart emp-
tied by suction and hemorrhage, and in the heart
with induced ventricular fibrillation, the oxygen
usages of 3.4 and 3.8 ml are much greater (249) (see
the paragraphs under Heart Rate for more detailed
consideration).
The metabolism of the heart is predominantly
aerobic. With abrupt vagal stoppage, however, during
constant pressure perfusion of the coronary arteries,
an excess of oxygen (oxygen debt) over that in the
asystolic state is taken up by the heart from the onset
of asystole to the time of appearance of the final
resting metabolism. This volume of oxygen, which
is quite small (estimated as 8% compared to the
maximum oxygen debt for an equivalent weight of
skeletal muscle of man), might be greater in a heart
working to capacity. Whether under prolonged hy-
poxia the anaerobic component of myocardial metab-
olism can be extended has not been determined
(]55)-
As in any muscle, the mechanical efficiency of the
left ventricle is estimated by dividing its external
work by the difference between its oxygen consump-
tion during activity and during its resting state. Pub-
lished data (31) which indicate efficiency approxi-
mating 10 to 20 per cent in the normal heart include
only the first two measurements. Since the resting
metabolism is considerable and variable, and such
values are generally not available, interpretation of
the relation of cardiac work to oxygen uptake is diffi-
cult.
RESPONSE OF THE CORONARY CIRCULATION
TO VARIOUS STIMULI
The information has been obtained from the un-
anesthetized dog and man and from the anesthetized
open- or closed-chest dog.
Resting Stalt-
As already pointed out, the levels of coronary flow
and oxygen usage of the myocardium are quite
fluctuant, varying greatly with the different types of
preparation and the prevailing stimuli. For values of
coronary flow and myocardial metabolism representa-
tive of the basal or resting state, selection of data in
the dog and man must be restricted to those in which
the systemic arterial blood pressure, cardiac index,
cardiac work index, heart rate, and body oxygen
uptake roughly approximated those figures for the
resting state. In the abnormal or diseased state in
human beings, data have been restricted to those in
which systemic blood pressure, cardiac index, cardiac
work index, and heart rate approximate values re-
garded as acceptable for the basal state when there
was no known reason for it to be elevated. These
criteria exclude a considerable volume of published
work, especially in man. While most of the data
comparing left coronary flow to the oxygen usage
per 100 g left ventricle per min, for the resting state
in man and dog, have been obtained by the N20
method, the excellent correlation of A-V oxygen dif-
ferences yielded by this method with those from the
more precise methods used on the dog supports the
accuracy of the former when properly used.
BLOOD SUPPLY TO THE HEART
[539
Reactive Hyperemia
Reactive hyperemia is considered to be the excess
blood flow (over the control flow that normally would
have occurred) following release of a coronary artery
occlusion. The coronary bed is extremely reactive to
the stimulus of anoxia. After release of temporary
occlusion of a coronary artery, even for as short a
time as 2 to 3 sec, left coronary arterial flow increases
almost immediately in the isolated heart, heart-
lung preparation, the anesthetized open-chest dog,
and the unanesthetized dog (66, 67, 147, 301). The
augmented flow exists throughout systole and diastole.
The flow response occurs without necessarily any
change in blood pressure or heart rate and before any
impairment of myocardial contraction occurs in the
area rendered ischemic. Beyond 30 to 60 sec of occlu-
sion, the area bulges during systole (153)- Depending
upon the duration of the occlusion, the peak flow
response (100-300% of control flow) does not usually
occur immediately upon release of the occluded
coronary artery, but reaches a maximum some time
during the first half minute of reactive hyperemia and
may last up to 4 min (fig. 8). The volume of reactive
hyperemia blood flow, its duration and peak flow
values increase with lengthening periods of left
circumflex arterial occlusion up to 120 sec. The
theoretical blood flow "debt" (control blood flow X
duration of occlusion) seems to be always greatly over-
paid (average 2 1 9 c'< ) in the presence of periods of
occlusion lasting from 5 sec to 180 sec. The reactive
hyperemia responses in skeletal muscle vascular beds
are similar (222) except that the blood flow debt is
variably over- or underpaid (410). In other vascular
beds, such as the superior mesenteric artery, this re-
sponse is much smaller than in the myocardium; in
the renal (150), it is essentially nonexistent.
The presence of reactive hyperemia has not been
satisfactorily explained. Its purpose must be to supply
,200
£ 100-
E
£89
HEART RATE = 162
90
82
88
COR SINUS 56 | y^~
_______ 13.0 cc 0
>-fa— -i**"^
5.6
tA
24r
16
6 2
0 DEBT 31 cc
11.8 cc
6 2
SECONDS
30 60 90 120 150 180
fig. 8. Diagrammatic redrawing of arterial blood pressure (upper tracing), left coronary blood
How (next lower tracing), and coronary sinus oxygen saturation (third tracing down), before, during,
and after release of 30 sec of left coronary artery occlusion in the open-chest dog. Lowest curve repre-
sents Oj consumption of left myocardium (flow times A-V O2) in ml/min, calculated from above
experimental data. Arrows A and B represent, respectively, beginning and end of measurements of
reactive hyperemic blood flow (RHBF) and its average A-V 02 difference used in calculation. [From
Coffman & Gregg (67).]
i54"
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the metabolic needs accumulated during the anoxic
period. The large increase in systolic as well as dia-
stolic flow within the first few seconds, before there is
time for a change in myocardial contractility, in-
dicates that massive active vasodilation has taken
place which overcomes systolic flow resistance. With
a longer period of occlusion, there must be considered
the possibility of flow through arteriovenous shunts
and through vessels probably near the epicardial
surface, whose surrounding myocardium is now "tired"
and does not so strongly oppose flow.
The oxygen consumption during myocardial re-
active hyperemia is measured by determining the
left coronary artery blood flow (rotameter) and the
difference in oxygen saturation of the arterial and
coronary sinus blood (measured continuously with
an optical densitometer). The theoretical oxygen
"debt" (control oxygen consumption X duration of
left coronary artery occlusion) is overpaid for 15- and
30-sec but slightly underpaid for 10-sec occlusions
(67). The rate of oxygen consumption during the
increased blood flow period is greater than in the
control state, showing that the myocardium has been
stimulated to take up more oxygen. The basic hy-
pothesis governing the calculation of the oxygen
"debt" in these studies is somewhat erroneous, for the
oxygen in the blood in the coronary vascular bed
during arterial occlusion, the metabolic rate during
the circulatory stasis, and changes in cardiac work are
not considered. As further evidence that the myo-
cardium develops an oxygen deficit, i.e., that
anaerobic metabolism occurs, it has been found that
lactic acid increases in the coronary sinus blood,
often in comparison to pyruvic acid levels following
the period of anoxia (67). These observations have
been confirmed in the unanesthetized dog with the
aid of chronically implanted (3-14 days postoperative)
electromagnetic flowmeters and coronary sinus
sampling tubes (280).
The contracting myocardium can withstand much
shorter periods of arterial occlusion and oxygen
deficit than resting skeletal muscle, and repays its
oxygen "debt" with an increased blood flow but with a
decreased A-V oxygen difference.
Heart Rate
Early reports indicated that myocardial oxygen
usage increases in the heart-lung and isolated heart
when heart rate spontaneously changes or when it is
increased by warming the sinus node or by driving the
heart electrically, but the evidence was conflicting
concerning the effect of rhythm of the heart on coro-
nary blood flow. In the above preparations, increase
in heart rate either increases, decreases, or does not
affect coronary flow do). More recent investigations
with somewhat better techniques and methodologies
confirmed this finding of correlation of oxygen usage
with heart rate in these preparations and showed a
higher energy cost of external work at the faster heart
rate (369). The oxygen observations were extended
to the empty heart beating in the open-chest dog
(249). In the latter preparation, electrically induced
ventricular tachycardia after section of the bundle of
His (23, 236, 355) or electrically induced auricular
tachycardia at rates somewhat higher than those
naturally occurring generally increases aortic blood
pressure, cardiac output, and cardiac work, while the
stroke volume and stroke work decrease (225).
Simultaneously, minute coronary flow and oxygen
usage increase, coronary resistance decreases, oxygen
extraction is unchanged, but the coronary flow and
oxygen consumption per beat decrease (23). Com-
parable results were obtained in normal human
subjects with atropine-induced cardio-acceleration
(137) and in the anesthetized closed-chest dog with
electrically induced auricular tachycardia, except
that systemic blood pressure, cardiac output, and
cardiac work did not rise (256). Since acceleration of
of the heart means proportionally greater time per
beat and per minute in systole than in diastole, and
since in systole coronary flow is less than in diastole,
it would be anticipated that increased heart rate per
se should reduce coronary flow. Since it does not, it
must be that increased flow is due to arteriolar dilata-
tion resulting from the increased metabolic activity.
Actual measurements indicate that as heart rate
increases, extravascular resistance rises but intra-
vascular resistance falls to a greater extent, indicating
a fall in net coronary resistance (236). The same trend
of flow and oxygen usage per beat and per minute
also occurs at the faster heart rate when minute
cardiac work is held constant or when comparisons
are made at the same level of stroke work. This means
that cardiac acceleration can augment the energy
metabolism of the myocardium without manifesta-
tion of the extra energy as work (23, 225). Data on
alteration in the coronary circulation following a
naturally occurring change in heart rate are limited
to the observation of increased coronary flow with
increased heart rate (92). Hence, the value of these
observations in relation to natural changes in heart
rate arising from local changes in the circulation of
BLOOD SUPPLY TO THE HEART
■541
the sinus node naturally occurring remains to be
determined.
In the open-chest dog, various arrhythmias, either-
occurring naturally or induced by electrical means,
by mechanical stroking of the heart, or by aconite
application, significantly reduce systemic blood pres-
sure and coronary flow when the irregularity is
marked or the rate rapid (above 190 per min) (72,
387). These arrhythmias include incomplete heart
block, premature auricular and ventricular systoles,
auricular fibrillation and flutter, auricular and ven-
tricular tachycardia.
Heart Doing Xo Externa! Work
Knowledge of the metabolic state of the heart
doing no external work is important because: a)
ventricular fibrillation and asystole are frequent
experimental and clinical occurrences; b) in the
empty beating heart or the asystolic arrested heart,
the magnitude of oxygen utilization could seriously
affect the potential for normal external efficiency of
the myocardium; c) with the advent of open-heart
surgery one must be certain, in the hearts made
dynamically quiescent by means of cardiac bypass,
ventricular fibrillation, or ventricular arrest, that the
metabolic requirements are met by the available
oxygen and myocardial damage does not occur.
The relative length of time the A-V node and myo-
cardium can withstand complete ischemia and still
function normally on return of their blood supply
has been studied in dogs whose hearts were maintained
on an extracorporeal circulation. Myocardial anoxia
(by clamping the coronary inflow) leads to somewhat
earlier damage to the myocardial muscle than to the
conducting system, for after 80 to 90 min of anoxia,
blood pressure cannot be maintained on removal of
the heart from the extracorporeal circulation, while
conduction is normal after as much as 100 min of
anoxia. The former is due to development of an
unusual firmness of the left ventricular muscle which
is not reversible upon reperfusion of the heart (65).
At the same time, ventricular distensibility, as re-
vealed by ventricular pressure-volume curves, is
sharply reduced (161). If ventricular fibrillation is
induced without maintenance of coronary flow, myo-
cardial substances such as adenosine triphosphate
and glycogen (which are maintained with coronary
perfusion) fall progressively within 15 to 30 min and
are partially resynthesized upon reinstitution of
perfusion (288). The oxygen usage has been deter-
mined for the left ventricle, the external work of
which has been reduced to zero by four different
procedures — vagal stimulation, intracoronary po-
tassium injection, ventricular fibrillation, and hemor-
rhage combined with suction to give an empty but
beating heart. Results have been rather variable for
the same procedure in the hands of different investi-
gators and generally one investigator has used only
two of the procedures. For example, the values for
oxygen usage for 100 g myocardium during fibrilla-
tion vary from 3.7 to 14.6 ml, in the empty beating
heart from 1.5 to 3.5 ml, during vagal asystole from
0.8 ml to 3.7 ml, intracoronary potassium injection
from 1.4 to 2.5 ml (17, 22, 26, 29, 179, 192, 249,
268, 288, 369). However, comparing the four pro-
cedures in the same series of experiments using the
open-chest dog, the oxygen uptake per 100 g left
ventricle in the working heart is 8 to 10 ml per min,
the resting metabolism (absence of heart beat, cardiac
output, and arterial blood pressure) during cardiac
arrest by vagal stimulation or potassium injection,
approximates 2.5 ml per 100 gleft ventricle per min, or
about 25 to 30 per cent of that at the prior working
level (249). The metabolism of the nonworking (but
slowly beating) heart obtained by rapid exsanguina-
tion is about 3.4 ml, and of the fibrillating heart 3.8
ml. Where measured, oxygen values were the resultant
of a simultaneous coronary flow increase and coro-
nary A-V oxygen decrease. Although the relative
values may hold, too much stress should not be placed
on the absolute values. While the various determi-
nants of each are probably still largely unknown,
knowledge is sufficient to indicate that each should
be widely variable. For example, values for the empty
beating heart are grossly affected by the prevailing
heart rate; values for the fibrillating heart depend
upon the type and frequency of fibrillation and upon
the ventricular diastolic size (268); values for the
vagus-arrested heart or following removal of an
artificial pacemaker are not affected by ventricular
systolic or diastolic size but vary greatly with the
previous level of metabolism in the working heart.
With a large elevation of myocardial metabolism by
intracoronary artery injection of epinephrine or
norepinephrine, the values are especially high and
may equal 50 per cent of those in the control state
(249)-
I entmular I olume or Fiber Length
Correlation of the left ventricular diastolic or
systolic fiber length (volume of a ventricle), or ven-
tricular tension, with the coronary flow and oxygen
'542
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
usage has been the subject of many serious and ex-
cellent investigations, but largely with the use of the
isolated heart. The hypothesis that the oxygen used
during ventricular systole is largely determined by
ventricular diastolic volume but not by ventricular
tension has been in vogue for many years. In 1927,
Starling & Visscher (354) showed that myocardial
oxygen consumption correlated with the changing
diastolic ventricular volume to the point of diminished
stroke volume from excessive ventricular distention,
but that the oxygen consumption of the heart had no
relation to its systolic volume. In isolated strips of
mammalian myocardium, the oxygen consumption at
rest increases significantly as the muscle length is
extended, but the tension developed as a result of
lengthening has a negligible effect on the oxygen
usage (396). Recently, others have shown an excellent
correlation of oxygen usage and diastolic volume in
the excised beating or fibrillating heart with perfused
coronary arteries (269). These observations also
apply to the heart beating within the chest. For
example, following partial constriction (by means
of a snare) of the pulmonary artery or the aorta
central to the two coronary ostia, the oxygen usage of
both ventricles, and the blood flow in right and left
coronary arteries, are increased, even the flow in
systole being considerably augmented in the right
coronary artery (153). However, in the excised heart
or the heart in situ which has been stopped by cervical
vagal stimulation or potassium injection, this relation
does not hold. Here large changes in the volume of
blood within the ventricular cavities can occur with-
out alteration of the oxygen consumption (249, 268).
This observation ties in with the fact that in the
open-chest dog the working heart's oxygen usage and
its coronary flow are not determined by the filling
pressure (atrial pressure) or the end-diastolic pressure
or volume, for at any given filling pressure the oxygen
and coronary flow can vary widely (46, 47). In ex-
periments in which the isolated beating heart with
perfused coronary arteries has been made to contract
isovolumetrically or isobarically, the myocardial
oxygen consumption is best correlated with peak
systolic pressure or systolic volume (269).
that in the right atrium into which the coronary
blood drains approximates o to 8 mm Hg. It would be
expected that elevation of systemic venous or right
atrial pressure would decrease both right and left
coronary inflow. However, the influence of these
pressures on coronary flow is difficult to study for the
changes induced in them lead to other cardiodynamic
alterations. An approach to the problem has been
made by studying the effect of constriction or ligation
of the coronary venous drainage system on coronary
inflow. With the heart beating in situ, mild elevation
of pressure in the coronary veins draining the left
coronary artery by coronary sinus constriction may
decrease only slightly coronary inflow and increase
coronary A-V oxygen difference. Acute coronary
sinus closure causes congestion of the left ventricle
(but not of the right ventricle, atria, or a portion of
the interventricular septum), a greatly elevated venous
pressure in the coronary sinus and great cardiac vein
often approximating or exceeding aortic systolic
pressure (153), but the flow reduction in the left
coronary artery or its major branches is only moder-
ate, averaging 8 per cent in 10 dogs. However, the
venous outflow measured simultaneously in several
major anterior cardiac veins increases greatly. Simi-
lar responses occur when the major venous drainage
channels of the right heart, the anterior cardiac
veins, are occluded in acute experiments; right coro-
nary inflow decreases from o to 63 per cent, averaging
21 per cent in eight different experiments (1 53)-
In acute experiments, pulmonary artery constric-
tion in the presence of previous ligation of the anterior
cardiac veins still causes a significant augmentation
of right coronary inflow. Finally, occlusion of both
the coronary sinus and all grossly visible anterior
cardiac veins reduces inflow further, but the hearts
generally survive and coronary inflow increases with
increased load. Even with chronic ligation of the
anterior cardiac veins and the coronary sinus, the
peripheral coronary venous pressure returns toward
normal within 30 days (1 53).
From these observations, it does not seem likely
that a considerable elevation of right atrial pressure
will influence significantly coronary inflow in the
normal heart.
Blood Pressure
coronary' venous pressure. The venous pressure in
the great cardiac vein of the anesthetized dog, with
or without open chest, approximates (10-15)7(0-5)
mm Hg (153); the values for the coronary sinus and
anterior cardiac veins are considerably lower, while
coronary arterial pressure. The mechanisms con-
cerned in alterations of coronary flow following acute
elevation or depression of central coronary pressure
have been only partially elucidated. Before con-
sidering the effect of coronary perfusion pressure on
coronary flow, attention is called to the experimental
BLOOD SUPPLY TO THE HEART
1543
fact that in the presence of a declining coronary
perfusion pressure, coronary flow ceases even when
coronary perfusion pressure is still sizeable, an ob-
servation also made in the renal and mesenteric beds
(150) (see values under Physical Determinants of
Coronary Flow). Coronary inflow (right or left
coronary artery) immediately increases throughout
the cardiac cycle with a rising perfusion pressure and
decreases with a falling perfusion pressure in all
preparations studied.
In both right and left hearts, however, there is no
set relationship between coronary flow and change in
central coronary perfusion pressure, the effect on
flow of a given pressure change varying from zero to
a maximum. The degree of change and its duration
will depend upon the extent of passive and active
changes in resistance within the myocardium asso-
ciated with the alteration of perfusion pressure. In
the heart doing no external work (empty, beating,
or fibrillating), the change in coronary flow is sizeable
with moderate change in coronary perfusion pressure,
but various relationships are observed. The resistance
may be semilogarithmic (80, 283, 341), i.e., it de-
creases with increasing flow, or at the highest flow
rates resistance may be constant or may increase.
Associated changes in resistance in the coronary bed
can be demonstrated when the vessels are perfused at
various pressures with the cardiac work kept con-
stant or not varying greatly. In the open-chest dog,
there is a rapid and marked change in coronary
flow within a few seconds following change of the
perfusion pressure. The induced change in coronary
flow may remain for some time (1 to 2 min) or it
may return toward, to, above, or below the control
flow level, thus showing large resistance changes in
the coronary vascular bed (89, 158, 249, 315). A
similar autoregulation of blood flow in the presence of
a mechanically induced change in perfusion pressure
has been observed in other regions such as the kidney
(363) and skeletal muscle (353).
It is not surprising that in these various situations,
most of which are highly abnormal, a variable rela-
tion of pressure to flow exists. It is believed that these
changes represent automatic shifts in the size of the
coronary vascular bed and in vascular resistance
(passive and active blood vessel changes) which
serve to meet the metabolic needs of the myocardium.
Whether they are related to the oxygen supply and
demand, to the relative amount of metabolites washed
away, or to some other control, is not known.
One of the largest changes in coronary flow from
altered coronary perfusion pressure occurs during
aortic constriction with the heart beating and working
in situ. In general, in such instances in which a change
in coronary flow resulting from alteration of coronary
perfusion pressure is associated with a change in
ventricular stress (ventricular size or systolic pres-
sure, or both), the coronary A-V oxygen is the same
or increased slightly while the oxygen consumption
changes considerably in the same direction as the
flow. Since the heart rate (generally fast) does not
alter greatly, both stroke coronary flow and stroke
coronary oxygen usage show large increases.
The oxygen uptake of a heart in the open-chest dog
performing external work can apparently be altered
by changing abruptly or gradually the level of a
constant coronary perfusion pressure by 5 to 35 mm
Hg for periods up to several minutes. The apparent
oxygen uptake of the left ventricle increases signifi-
cantly when coronary perfusion pressure increases.
When coronary perfusion pressure decreases, oxygen
uptake decreases. This change in oxygen uptake by
the myocardium associated with the opposite change
in coronary A-V oxygen occurs in the presence of a
constant arterial blood pressure, heart rate, stroke
volume, and cardiac work. Similarly, in isolated
hearts not performing external work, the change in
coronary flow from alteration of coronary perfusion
pressure is counterbalanced by a shift in the coronary
A-V oxygen in the opposite direction, but the oxygen
uptake is changed significantly especially at the
higher levels of coronary perfusion pressure. As yet,
experimental testing has not been able to ascribe
this apparent change in oxygen uptake to an arti-
factual happening (158, and unpublished observa-
tions on the isolated heart). These findings, which
have been confirmed (7), would seem to make suspect
various observations, especially in the isolated heart,
in which perfusion pressure has been varied.
Many observers have reported that a given increase
in work of the heart, produced by raising aortic
pressure (aortic clamp) while holding cardiac output
constant, results in a much higher coronary flow and
oxygen usage per minute and per beat than when a
similar increase in cardiac work is effected by ele-
vating cardiac output through increased venous
return at a constant aortic blood pressure (208).
This discrepancy between the relative oxygen costs of
pressure and flow work is observed in the isolated
heart as well as in the dog with a complete circulation.
In experiments with the isolated, supported heart
with a constant heart rate, an increase in left ven-
tricular work, by augmenting cardiac output while
at the same time lowering aortic pressure by aortic
1544
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
clamp release, results in a large increase in cardiac
work associated with a decrease in oxygen usage
(334). These observations taken together stress the
importance of the aortic pressure and of the develop-
ment of tension by the heart in the control of oxygen
utilization.
Chemical Composition of the Blood
The chemical composition of the blood and tissue
fluids within the myocardium has been found to be of
great importance in determining the volume of
coronary flow.
asphyxia. Asphyxia, in which the carbon dioxide
content of the blood increases and the oxygen con-
tent decreases simultaneously from cessation of
breathing, is accompanied by a large increase in
coronary inflow in the anesthetized dog. Within 30
to 60 sec after cessation of respiration, the flow in both
systole and diastole increases, averaging about 200
per cent, and this occurs before any significant change
in aortic pressure or heart rate ( 1 47 ).
hypoxia. When the oxygen content of fully saturated
arterial blood of normal hematocrit is decreased by
exposing it to successive mixtures of oxygen and
nitrogen, containing progressively less oxygen (100%
O2 down to 5 % 02), the resultant arterial hypoxia
induces profound increases (200-300 %) in coronary
arterial inflow in both systole and diastole in the
fibrillating heart, isolated heart, heart-lung prepara-
tion, and anesthetized dog, but the oxygen consump-
tion is not changed (26, 113, 147, 167, 182). In the
anesthetized dog, as the arterial oxygen saturation
decreases, the coronary A-V oxygen difference and
coronary sinus oxygen content are decreased as the
oxygen extraction increases. For example, starting
with a control coronary sinus oxygen content of 4.2
vol per cent and saturation of 20 per cent, the coro-
nary A-V oxygen difference may be decreased to 3
vol per cent and the coronary sinus oxygen to 1 vol
per cent, while the oxygen extraction may be in-
creased to 95 per cent. Similar findings are reported
in man (181). Eventually, heart rate, blood pressure,
and cardiac output may be elevated, presumably
from the marked increment of cardiac contractility
arising from stimulation of the sympathetic nervous
system (408) since anoxia has only a depressant
effect on the completely isolated heart (242). The
increase in coronary flow precedes any change in
these parameters and maximal coronary dilatation
occurs when the arterial saturation falls to about 20
per cent of normal.
Since the flow effects of systemic anoxia produced
by artificial respiration with air and nitrogen and of
local myocardial anoxia by underperfusion or by
cyanide injection (147) into the coronary artery
are similar, it is concluded that they all depend upon
the anoxia produced, and probably upon the presence
of this anoxia in the myocardium. Since the blood
pressure does not change and the ratio of pressure to
flow increases throughout the cardiac cycle, it is
also concluded that anoxia causes a relaxation of the
walls of the coronary vessels. To what extent this is
active, that is, a direct effect on smooth muscle of the
coronary vessels, and to what extent, if any, extra-
vascular support has been lowered, cannot be as-
certained by these experiments. By using the technique
already described of prolonged vagal stoppage of the
heart for separation and fractionation of flow deter-
minants, coronary perfusion with saturated blood
has been compared to perfusion with somewhat
unsaturated blood. The resulting flow increase in the
beating heart, in the latter instance, is shown to be
about equally divided between a decreased extra-
vascular compression and an active vessel relaxation
(158)-
The ultimate cause of the decrease in coronary
vascular resistance in the presence of hypoxemia is
not known, but it could arise from a direct action of
low arterial oxygen content of the blood on the smooth
muscle of the coronary vessels (182) or from hypoxia
of the myocardium. To differentiate these possi-
bilities, open-chest experiments have been made
on fibrillating dog hearts, involving coronary per-
fusions with blood at varying levels of saturation and
at high perfusion pressures which increased the coro-
nary flow until coronary sinus blood became relatively
rich in oxygen (26). In the presence of a quite high
coronary perfusion pressure, considerable lowering of
arterial oxygen content (to 10 vol per cent) does not
increase coronary flow. An increase in flow occurs
only at coronary sinus oxygen levels less than 5.5
vol per cent. Since the coronary sinus oxygen content
probably closely reflects tissue oxygen content, this
finding suggests that arterial oxygen content is not
critical in the regulation of coronary flow but that
coronary vasodilation in hypoxemia is related to
myocardial hypoxia (myocardial oxygen content).
Finally, experiments are reported in which the coro-
nary arteries of the isolated heart are perfused with a
fully saturated hemoglobin solution whose oxygen
content is varied by dilution. In this situation, as the
BLOOD SUPPLY TO THE HEART
1 545
oxygen content was varied from 18 vol per cent down
to 2 vol per cent by dilution with Ringer-Lockc\
solution, the coronary flow increased although the
intravascular oxygen tension at the level of the
arterioles was kept constant (164).
metabolites. The mechanism whereby hypoxia
operates to increase coronary flow remains obscure.
Presumably, metabolites accumulate but their nature
and possible effectiveness are unknown. Experiments
dealing with this problem in which an extracorporeal
circulation of blood is used must be cautiously evalu-
ated. Dog blood contains potent vasoconstrictor and
vasodilator substances. The red cells, especially,
contain a potent vasodilator substance (adenosine
triphosphate). This and other substances are readily-
made active by hemolysis resulting from minute
mechanical trauma and agitation (60). In the heart-
lung preparation, coronary flow generally progres-
sively increases as the experiment continues, and
substances accumulating in the coronary venous
blood were originally thought to cause vasodilatation
when reinfused into the coronary arteries (17).
Hilton & Eicholtz (182), however, could not confirm
this in the isolated heart, for replacement of the blood
that had circulated for some time by fresh defibrinated
blood did not significantly alter flow.
It is not clear whether or not vasodilator substances
exist in the coronary sinus blood of the heart beating
within the chest in sufficient concentration to alter
coronary blood flow. In recent experiments, blood
draining normal, hypoxic, or overperfused (perfusion
pressure considerably greater than aortic pressure)
hearts has been oxygenated in a dog lung or on a
screen and perfused at a controlled pressure through a
rotameter into a test coronary artery of the same or
second dog, or unoxygenated coronary sinus blood
has been perfused into an isolated beating frog heart.
These experiments have failed to demonstrate sub-
stances having vasoactive, inotropic, or chronotropic
properties in the coronary sinus blood (193). On the
other hand, injection of coronary sinus blood obtained
during cardiac sympathetic nerve stimulation causes
a moderate coronary dilatation at the same blood
pressure and heart rate (277).
Intracoronary injection of intermediate metabolites
will increase coronary blood flow. Histamine, metabo-
lites such as adenosine, adenylic acid, and breakdown
products of nucleic acids increase coronary flow in
the perfused heart, the human heart-lung preparation
and heart in situ (10, 277). However, it has never been
demonstrated that the concentration of these sub-
stances increases within the myocardium during
anoxia or increased effort of the heart. Studies on
relative coronary vasodilator potency show that
adenine is relatively inactive, while adenosine tri-
phosphate and adenosine diphosphate have approxi-
mately four times the potency of adenosine mono-
phosphate, adenosine, and uridine triphosphate
(406, 407). In addition to vasodilator properties,
some of the purine and pyrimidine derivatives have
been demonstrated to have positive inotropic action
in the normal and failing heart (52). It is not sur-
prising, therefore, that in the open-chest dog with
constant pressure perfusion of the coronary arteries,
some of these substances (ATP and UTP) with
inotropic and vasodilator properties also increase the
myocardial oxygen consumption (407). However,
since the elevation in coronary blood flow is greater
than that necessary to meet the increased oxygen
demand, i.e., the coronary A-V oxygen is decreased
considerably, the action of such compounds is prob-
ably largely on the coronary vessels, and only second-
arily on metabolic rate.
Although it has not been possible to demonstrate
in the coronary sinus blood substances having vasoac-
tive, chronotropic, or inotropic properties, this does
not necessarily rule out an active role of metabolites
in regulating coronary flow. As pointed out by Berne
(28), although adenosine and adenine nucleotides are
not recoverable in the coronary sinus, derivatives of
adenosine such as inosine and hypoxanthine appear
in the coronary sinus blood during periods of myo-
cardial hypoxia, and vasoactive concentrations of
adenosine added to coronary arterial blood are re-
coverable in the coronary sinus only as inosine and
hypoxanthine. Thus, the possibility should be enter-
tained that in hypoxia, myocardial nucleotides give
rise to adenosine which diffuses out of cardiac cells,
induces vasodilatation, but is deaminated and split
before separation from the blood can be effected.
acidosis and alkalosis. In the isolated heart or heart-
lung preparation, acidosis induced by administering
C02 or lactic acid dilates the coronary blood vessels,
for coronary flow may increase ( 1 82 ) despite a marked
reduction in the rate, output, arterial blood pressure,
and contractile force of the myocardium (273). In an
intact preparation in the presence of severe respiratory
acidosis (from C02 administration), or a fixed acidosis
(from infusion of HC1 solution), coronary flow in-
creases (114), remains constant (134), or decreases
(go, 147), while systemic dynamics (blood pressure,
heart rate, and cardiac output) are not largely
1546
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
changed. In contrast to its direct depressant effect
on the isolated perfused heart, a C02 concentration
in the respired air of the open-chest dog of at least 8
per cent has a stimulating influence on the myocar-
dium for the contractile force may not be lowered and
myocardial function curves are not depressed (267).
Presumably, this arises from its stimulating influence
on the sympatho-adrenal system since acidotic heart
failure in the heart-lung preparation can be reversed
by administration of sympathomimetic amines. The
divergent results with hypercapnia may, therefore,
merely express the variable weighting of the antago-
nistic effects on the intact animal circulation, of a
direct depressant effect on the myocardium and
smooth muscle of the coronary vessels, and of an
indirect stimulating effect on the same structures
through the sympatho-adrenal system. In the intact
dog, hypocapnia (arterial C02 less than 35 vol per
cent) is without effect on systemic and coronary
hemodynamics and the ventricular function curve is
unaltered (1 14, 267).
In the isolated heart doing no work, alkalosis in-
duced by varying the bicarbonate concentration of
the perfusing solution, or following CO» administra-
tion, depresses inotropic and chronotropic cardiac
activities but increases coronary flow. These changes
in cardiac function are effected only when the pH of
the perfusate is altered (248). Sodium bicarbonate
infusion in the intact anesthetized dog stimulates
both systemic and coronary hemodynamics. In the
presence of an essentially constant heart rate and
arterial blood pressure, and a marked increase in
cardiac output, the coronary flow and oxygen usage
per minute and per heart beat are elevated consider-
ably (134). The mechanisms concerned are not
known. Nor has it been established whether these
effects of acidosis and alkalosis on the coronary cir-
culation in the intact dog arise from changes in pCO>,
HC03, or from the resulting change in pH, since in no
instance has the pH been separately controlled.
potassium, calcium. There has not been sufficient
experimentation to determine the role, if any, of the
electrolytes in regulating the coronary circulation.
Increased potassium concentration (150%) in the
blood perfusing the fibrillating heart of the Langen-
dorff preparation increases coronary flow, while
larger intracoronary concentrations in the open-
chest dog also produce similar increases (88). In the
heart-lung preparation, isolated heart, or heart within
the chest, very large potassium concentrations suffi-
cient to arrest the heart reduce coronary inflow (249,
259). In the latter instance, the oxygen usage is also
greatly reduced as a result of a decreased flow and
oxygen extraction. In the open-chest dog, upon the
addition of pharmacological amounts of Ca gluconate,
coronary flow and oxygen consumption increase and
oxygen extraction decreases without much change in
blood pressure and heart rate (115).
Transfusion
Augmentation of ventricular load by increasing
venous return, and, hence, circulating blood volume
through infusion has a clinical counterpart in the load
placed upon the human heart by transfusion or by an
aortacaval fistula. During transfusion, systolic and
diastolic heart size, ventricular stroke volume and
stroke work, atrial and ventricular end-diastolic
pressures and arterial blood pressure all increase as
the heart rate slows considerably. When the coronary
arteries are perfused at a constant pressure or with a
normal pulsatile aortic pressure, an increase in the
cardiac output or cardiac work through augmentation
of venous return to the heart beating in situ augments
moderately the coronary flow and oxygen usage per
minute and per heart beat while the aortic pressure
rises (46, 90, 152, 153). The increasing coronary flow
is partially explainable on a mechanical basis since
the slowing of the heart should increase coronary
flow per beat and per minute by increasing the dia-
stolic time period during which coronary flow is
greater. The coronary flow and oxygen are used
economically, for the ratio of stroke cardiac work to
stroke oxygen consumption increases. However, it
has been repeatedly shown in the denervated heart
and heart-lung preparation that coronary arterial
inflow and coronary sinus outflow are reduced some-
what or unchanged by alterations in cardiac output
as long as the resistance against which the ventricles
contract is unaltered (10). These findings have been
extended to the open-chest dog in which coronary
flow and oxygen usage may not change or may rise
only slightly in the presence of a constant arterial
blood pressure and marked increase in cardiac out-
put (46). In the isolated supported heart, the coronary
flow and oxygen usage per minute and per heart
beat can be made to decrease in response to a lowering
of arterial blood pressure, achieved by release of an
aortic constriction during the augmentation of cardiac
output which follows an increase in venous return
(334)-
Since the average individual is transfused only on
rare occasions, it is not known whether such, informa-
BLOOD SUIM'LY TO THE HEART
r547
tion can be used to explain happenings in the coro-
nary circulation of the normal heart exposed to the
stresses of everyday life. At least, in the latter instance
the systemic dynamics are quite different from those
listed above as occurring with transfusion. For ex-
ample, in exercise and excitement, while the heart
rate increases greatly and the duration of systole
decreases, cardiac size may decrease and stroke
volume and atrial and ventricular diastolic pressures
may undergo only limited changes.
Anemia
The coronary system participates actively in the
circulatory adjustments to anemia. For hemoglobin
values of 10 g or more, the systemic circulation is
essentially unaltered and the compensation of the
coronary system to the decreased oxygen-carrying
capacity is similar to that with hypoxia, i.e., an
increased coronary flow without change in oxygen
uptake. When the hemoglobin values reach 6 to 8 g,
the response of the systemic circulation is manifested
by tachycardia, increased cardiac output and cardiac
work, and a fall in peripheral resistance. The coronary
flow may now triple; coronary venous blood may
contain less than 2 vol per cent oxygen, the coronary
arteriovenous oxygen difference may be 4 ml or less,
and oxygen uptake may be considerably increased.
The increase in coronary flow is related in part to the
decreased blood viscosity, and in larger part to the
active dilatation associated with myocardial hypoxia,
which in turn arises from the low hematocrit and
from the increased metabolism. Ultimately, myo-
cardial failure will occur in severe anemia when the
coronary vessels have approached maximal dilatation
and cannot further compensate for the decreased
oxygen-carrying capacity of the blood either by
increased flow or by increased oxygen extraction. In
the presence of coronary stenosis associated with
anemia, the effect of coronary arteriolar dilatation in
increasing flow is minimized by the high fixed resis-
tance of the stenotic artery, and myocardial depression
and failure occur at lesser degrees of anemia (31, 57).
Very little information is available regarding the
coronary circulation in the presence of polycythemia
vera. The expanded red cell mass has been associated
with a considerable reduction in coronary blood flow
and an increased oxygen extraction without change
in oxygen usage. Allocation of these changes to an
enhanced oxygen-carrying capacity or to a viscosity
effect has not been made (305).
Xarvits Influences
The control of the coronary circulation by para-
sympathetic and sympathetic nerves has been the
subject of intensive investigation and dispute. Many
experiments, however, have been interpreted with
difficulty since cardiac output and cardiac work were
not determined, and heart rate and arterial blood
pressure which affect coronary flow and oxygen
usage varied widely (go, 187, 340, 405). There is
some evidence to indicate that the over-all nature and
extent of neural cardiogenic control is some degree of
coronary vasoconstriction since a) an outstanding
characteristic of the isolated heart or heart-lung
preparation is a very high coronary blood flow and
low myocardial oxygen extraction, and b) in chronic
dogs the procedure of pericoronary denervation
results in a relative increase in coronary flow and
decrease in oxygen extraction (43). For the most part,
the nervous system influences on the coronary cir-
culation have been studied by observing the coronary
flow, oxygen usage, and contractility responses fol-
lowing electrical stimulation or severance of the
nerves. Although such procedures are not paralleled
by normal occurrences in the animal, the observed
responses are presumed to indicate the functions
which the nerves are capable of exercising in the
intact animal. Further difficulty in interpretation
arises from the fact that the specific effects upon the
heart muscle and on the coronary vascular system
are largely experimentally inseparable and only the
net effect can be observed. Differences in methods
and preparations are additional variables which may
account for the discordant results of different in-
vestigators.
vagus. Recent studies of the effect of the vagus nerves
on the heart have gradually clarified our view of their
effect on the coronary circulation. Early evidence
indicated that the vagus nerves contain both dilator
and constrictor fibers (401). That the vagus exerts a
vasoconstrictor effect is based on observations that
abolition of the parasympathetic pathways in the
heart-lung preparation (by mechanical and chemical
means) results in augmentation of heart rate or
coronary flow, while stimulation of the peripheral
ends of the cut vagi decreases coronary flow (10).
The evidence that it exerts a vasodilator effect arises
from the observation that in the fibrillating heart with
coronary arteries perfused with blood under a con-
stant pressure, vagal section usually decreases coro-
nary inflow but vagal stimulation usually increases
i548
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
coronary inflow (206). It would seem unlikely that
the vagus would act as a vasoconstrictor since its
chemical transmitter, acetylcholine, is a coronary
vasodilator.
Such studies, of course, do not define the action of
the vagi in the intact animal. In the open-chest dog
with the heart rate maintained at a constant value
by an electric pacemaker, section of these nerves or
stimulation of their cut ends does not usually evoke
significant change in coronary inflow, coronary
sinus flow (orifice meter, bubble flowmeter, rotam-
eter), or in coronary A-V oxygen difference, while
blood pressure and cardiac output are essentially
unchanged (83, go, 236, 339). In all preparations,
ventricular contractility is usually not changed (367).
These observations are in accord with the apparent
lack of vagal fiber distribution to the ventricular
myocardium (54, 262). At times, vagal stimulation
(378) has been observed to cause reduction in cardiac
output, blood pressure, and coronary inflow, an
effect which was abolished by atropine. Although
this effect is ascribed by the authors to a negative
inotropic effect of vagal stimulation on the ventricular
muscle, it could also have resulted from diminution of
vigor of atrial contraction thus reducing ventricular
filling, a finding of different investigators (336).
sympathetic. Stimulation of the stellate ganglion or
its cardiac branches in the anesthetized open-chest
dog or in the unanesthetized resting dog increases
mean flow in both right and left coronary arteries
(83, 139, 153). This augmentation lasts for minutes
and persists long after any augmentation of heart
rate or blood pressure (if such occurs) has returned
to normal. Figure 9 is a response of the left circumflex
flow obtained with an electromagnetic flowmeter in a
resting unanesthetized dog a few days after probe
implantation. In this dog, in which the left stellate
ganglion had been previously disconnected from the
thoracic sympathetic chain and spinal nerves to
eliminate peripheral effects of excitation, stimulation
of the common ansa subclavia initially and transiently
decreases coronary flow throughout the cardiac cycle
without change in blood pressure or heart rate. Within
a few seconds, at an unchanged blood pressure but
increased heart rate, mean coronary flow increases as
the result of an increase in diastolic flow and in spite
of a depression of systolic flow with appearance of
backflow. Almost immediately, thereafter, coronary
flow increases greatly throughout the entire cardiac
cycle, with the disappearance of backflow. This
pattern of response to stellate stimulation may occur
with or without spontaneous elevation of blood pres-
STELLATE G. STIM.
RECOVERY
BLOOD
PRESS
MEAN COR
FLOW/MIN
SYST. COR
FLOW/MIN
MEAN COR. FLOW
SYST COR FLOW
fig. 9. Reproduction of sections from a continuous record in a conscious dog a few days postopera-
tive, showing effect of left stellate ganglion stimulation on phasic arterial blood pressure and stroke
left circumflex coronary inflow using a strain gauge and electromagnetic flowmeter as in fig. 6.
Connections of stellate to sympathetic chain severed at time of operation. [Granata el al. (139).]
BLOOD SUPPLY TO THE HEART
'549
sure and heart rate, or when the aortic blood pressure
(in open-chest dogs) is artificially controlled and
compensated to the control level. This indicates that
the factors of heart rate, blood pressure, and work
are not indispensable to the natural mechanism
through which the flow increase is mediated. The
major portion of the flow increase is related to acti\<-
dilatation since blood pressure does not necessarily
rise. The net flow increase rests in part, however, on a
mechanical basis, for the duration of systole (in which
flow is less than in diastole) is reduced and thus, at
the same heart rate, the period of time occupied by
diastole (in which rate of flow is greater), is increased
considerably. About 30 per cent of the flow increase
is estimated to be due to this shortening of systolic
and lengthening of diastolic time per beat or per
minute (93). Concurrently, left ventricular oxygen
usage, cardiac output, and cardiac work, either per
minute or per heart beat, increase while the systolic
and diastolic dimensions of the heart decrease. Since
the blood pressure and heart rate do not necessarily
change during the flow increase, while duration of
systole shortens, the stroke coronary flow does not
correlate with pressures developed by the ventricle
or with directional trends in ventricular tension
calculated thereon.
The evidence is well founded that stimulation of the
sympathetic nerves to the heart aids greatly in main-
taining and augmenting the rate and contractility of
the heart, as shown by improvement in the atrial and
ventricular function curves from any given atrial,
ventricular end-diastolic pressure (263, 336), or fiber
length, and in the gradient of the aortic pressure and
stroke volume curves (11, 153, 211). In no experiment
has coronary inflow been found to increase without
experimental evidence of increased vigor of contrac-
tion, increased cardiac work, and metabolism. The
mechanism by which this is accomplished has not
been completely identified. The possibility that
adrenal secretion is responsible for the cardiac stimu-
lating effect has been largely discounted experi-
mentally (153)- The facts that a) administration of
acetylcholine to an atropinized heart results in libera-
tion of an adrenaline-like substance (186); /;) that
this substance is also normally present in heart ex-
tracts (376); and c) that stimulation of cardiogenic
sympathetic fibers sets free an adrenaline-like sub-
stance (284, 346), all support the view that a dominant
role is played by myocardial catecholamines.
The experiments of Eckstein et al. (93) offer evi-
dence that this process is very wasteful, for the increase
in the work of the heart is not essential to the asso-
ciated increase in coronary inflow. Stimulation of the
accelerator nerves in the open-chest dog produces an
increase in vigor of contraction, cardiac output,
cardiac work, coronary blood flow, and oxygen con-
sumption. However, simultaneous nerve stimulation
and inflation of a left auricular balloon to reduce the
external work of the heart below the control value is
likewise followed by increased vigor of contraction,
increased coronary flow and increased oxygen con-
sumption. Thus, the adrenaline-like substance re-
leased by nerve stimulation would appear to increase
oxygen consumption directly.
The preceding observations do not preclude the
possibility that the major influence of the sympathetic
cardiac nerves may be to exert a direct vasomotor
influence on the coronary vessels, the initial temporary
coronary vasoconstriction which invariably occurs
being overpowered by metabolic dilator influences
associated with the type of stimulation. It is, however,
most difficult to establish and identify conclusively,
by experimental means, the separate effects of nervous
influences upon the myocardium and coronary vessels
because the physiological functions of these structures
are so intimately related that their individual re-
sponses can be secondarily modified, each by the
other. In the innervated fibrillating heart, stellate
stimulation at times decreases coronary flow, while
nerve section increases coronary flow (206). Recent
evidence indicates the existence and functional im-
portance of coronary vasomotor fibers the action of
which in previous investigations was presumably
obscured because of a nonselective type of nerve
stimulation (199, 358). As other workers have shown,
excitation of high threshold postganglionic (stellate
or inferior cervical) cardiosympathetic fibers with
high voltage and high frequency stimulation causes
profound alterations in myocardial metabolism which
could not be prevented by ergotamine or atropine.
However, appropriate stimulation of preganglionic
fibers, using low voltage and low frequency, leads
either to coronary vasoconstriction (decreased coro-
nary flow and blood pressure and increased coronary
A-V oxygen without alteration in oxygen consump-
tion), or to coronary vasodilation (increased coronary
flow, decreased coronary A-V oxygen without blood
pressure elevation and without alteration in cardiac
metabolism). The magnitude and importance of these
direct vasomotor effects remain to be determined.
In the author's laboratory, preliminary attempts have
failed to demonstrate these direct vasomotor influences
of the cardiac sympathetic fibers in the unanesthe-
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
tized dog with a chronically implanted electromag-
netic flowmeter on the coronary artery.
To the author's knowledge, the existence of a
tonic action on the coronary circulation of the intact
animal, attributable to the parasympathetic and
sympathetic nerve fibers, has never been demon-
strated, the section of these nerves in the open-chest
dog leading only to nonspecific changes. Study of this
problem should be made through the application to
the coronary circulation of recent techniques for
coronary neurectomy and extrinsic cardiac denerva-
tion on chronic dogs (43, 71).
reflex control. Our knowledge of the role of the
central nervous system in regulation of the coronary
circulation in health and disease is sufficient to war-
rant discussion but is certainly insufficient to draw
firm conclusions. This arises from a lack of proper
experiments in which studies of both the reflexes and
the coronary hemodynamic responses have been
simultaneously made. A proper demonstration de-
pends on the observations that the vagus and sym-
pathetic nerves to the heart are tonically active as
far as coronary blood flow regulation is concerned, or
can be made so by the induction of adequate stimuli
arising either within the heart or peripherally.
Different observations support the view that there
are receptors in the distribution of the canine left
coronary artery. Injection of veratridine into a left
coronary artery (but not the right coronary), going
only to the left ventricular muscle and in amounts
insufficient to affect the systemic circulation upon
direct injection into the left ventricle, causes an
abrupt fall in blood pressure and heart rate (76, 77).
( lirculatory depression which may follow selective
augmentation of central coronary pressure near the
left coronary orifice and the initial part of the cle-
scendens artery, or which may occur during coronary
sinus occlusion, is relieved by vagal section (132, 187,
285).
The evidence is equivocal that flow in one coronary
artery can be influenced reflexively and adversely by
impulses arising in another occluded coronary artery.
Various supportive observations suggest that noxious
intercoronary reflexes can be made to occur: a)
Ligation of a coronary artery is stated to cause reflex
spasm or vasoconstriction in the other coronary artery
resulting in fatal ventricular fibrillation (231). This
is presumed to be abolished by bilateral vagosym-
pathetic blockade (228). The infarction after coronary
ligation is increased with vagal stimulation and is
prevented by local anesthesia of the vessel wall at the
site of the ligature (220, 223). b) In the anesthetized
closed-chest dog with visualization of the coronary
artery bed by cinefluorography after injection of a
radiopaque dye, selective embolization (lycopodium
spores) of a coronary artery branch results in a marked
decrease in size of the nonembolized coronary artery
bed and in coronary sinus flow (165). On the other
hand, a) West et al. (392), using techniques similar to
those of Guzman, failed to find evidence of reflex
coronary vasoconstriction following coronary em-
bolization. /)) In the open-chest dog, following liga-
tion of the right or left coronary artery, the coronary
flow (rotameter) rises and resistance falls in the un-
occluded coronary artery, such flow augmentation
presumably resulting from an anatomical and func-
tional overlap of the right and left coronary arteries
(153, 282, 377). c) In the unanesthetized resting dog,
some days after implantation of an electromagnetic
flowmeter on the main left coronary artery or a major
arterial branch, temporary (10-30 sec) occlusion of a
left coronary artery branch results either in no change
or an increase in blood flow in the artery in which
flow is being measured ( L. C. Fisher, unpublished obser-
vations). It must be remembered that these experiments
with negative results follow considerable dissection of
the coronary arteries, and actually represent con-
ditions which deviate extensively from the normal
nerve state. Thus they do not rule out the possibility
of reflex coronary vasoconstriction occurring in small
localized regions of the myocardium after coronary
occlusion.
Changes in coronary blood flow which might result
from extracardiac stimuli would be of great clinical
interest, and their demonstration might aid in eluci-
dating the mechanism of the relationship between
angina pectoris and its various incitants, such as
eating, abdominal distention, cold, and exercise. The
claim is made that many diverse afferent stimuli affect
the coronary circulation. Prolonged experimental
neurosis in monkeys produced by conflicting condi-
tioned reflexes or selected brain stimulation can pro-
duce ECG changes identical to those of human
ventricular ischemia (250). Stimulation of many
afferent nerves, distention of the stomach, gall bladder
and esophagus, and cutaneous pain, all are presumed
to decrease coronary flow in the anesthetized dog,
while elevation of cerebral blood pressure and carotid
sinus pressure decreases coronary flow in the inner-
vated heart-lung preparation (blood pressure and
heart rate kept constant) (153). In these experiments,
the recording devices and data were generally in-
sufficient to establish that no changes occurred in
BLOOD SUPPLY TO THE HEART
'Sj1
heart rate, cardiac output, blood pressure, length of
systole and diastole, each of which could alter cardiac
metabolism, work and coronary flow separately or in
combination. Actually, in experiments with adequate
flow methods, a) increase in intragastric or intrabiliary
tension gives a variable flow response (bubble flow-
meter and rotameter) but always in the same direction
as the blood pressure change (90, 285, 405); b) dermal
contact with ice water in the anesthetized dog fails to
produce reflex constriction of the coronary arteries
(24, 347). The sight, smell, and ingestion of food and
tilting the head down all increase coronary flow
(electromagnetic flowmeter) concurrently with an
augmented heart rate and blood pressure (301). It is
therefore unreasonable to maintain, as has been done,
that such agents have caused active vasoconstriction
or vasodilatation in the coronary bed and that such
changes are necessarily largely controlled through
nervous reflexes. This is especially so since in each
experiment the effects of the stimuli were not generally
tested after as well as before the cutting of the car-
diac nerves.
Thus, while reflexes to the coronary circulation from
the heart or extracardiac visceral structures certainly
do exist and may be important in normal physiology
and pathological physiology, we must wait for the
future to show their exact function.
Hormones
norepinephrine and epinephrine. Since the generally
accepted theory of autonomic nerve transmission is
based on the liberation of acetylcholine and epineph-
rine-like substances, the coronary flow effects with
these agents are of particular interest in connection
with coronary innervation.
The action of epinephrine on the coronary blood
flow has been investigated extensively. In most dog
preparations, including the fibrillating heart (27),
heart-lung preparation (179), the open-chest dog (82,
83, 1 12, 146), and the unanesthetized dog a few days
postoperative to flowmeter implantation (301),
intracoronary artery injection of epinephrine and
norepinephrine increases coronary blood flow. In
the latter two preparations, their effect on the coro-
6 fiq EPINEPHRINE I V
CONTROL ] 12 !
16 SEC
RECOVERY
i
/H.
/^
90
123
197
2 2
162
35
120
120
196
1.6
163
33
J\A
/"N/\
114
114
119
1.0
90
29
fig. 10. Reproduction of sections from original record taken in a conscious resting dog some days
postoperative, showing the effect of rapid intravenous injection of 6 /xg epinephrine on phasic arterial
blood pressure and stroke left circumflex How, using a strain gauge and electromagnetic flowmeter
as in fig. 6. [Rayford el a/. (301).]
■552
HANDBOOK OF PHYSIOIcx.Y
CIRCULATION II
nary flow pattern is similar to the sustained effect
obtained during stimulation of cardiac accelerator
nerves, i.e., an increased blood flow throughout the
cycle (fig. 10). In all animal preparations, as well as
in man, myocardial contractility is increased to a
marked degree as indicated by the intensity of fibril-
latory movements in the fibrillating heart, by de-
pression of the isometric and systolic portions of the
phasic inflow curve in the dog, and by an increase in
myocardial contractile force as measured by a strain
gauge arch in man and animal (131). Both intra-
venous and intracoronary artery injections increase
cardiac oxygen consumption, in the first instance by
increasing coronary flow and decreasing coronary
A-V oxygen difference (112), in the second instance
by increasing coronary flow and increasing coronary
A-V oxygen difference (27). This occurs even in the
vagus-stopped heart (249). With very small doses,
coronary inflow may increase without any change in
blood pressure or heart rate and with increased coro-
nary A-V oxygen difference. With larger doses, as
the systemic effects of the substance (increased aortic
blood pressure, cardiac output, and changing heart
rate) become evident, the coronary and metabolic
effects are exaggerated (90).
From the preceding it can be seen that there is
general agreement that these substances produce
coronary vasodilatation. The flow increase is the net
result of an augmented extravascular support tending
to decrease coronary flow, a metabolic dilator effect
tending to increase coronary flow and any direct
effect the compounds may have on the coronary
vessels. There is, however, disagreement and con-
fusion regarding the respective magnitude of each
separate effect. There is little doubt that with the
larger doses, most of the flow increase is due to the
large increase in myocardial metabolism. However,
one point of view has it that these substances are
primarily coronary vasoconstrictors, their vasodilator
action arising secondarily from a hypoxic state of the
myocardium as a result of their stimulating effect on
the myocardial metabolism. The evidence for this is
that epinephrine causes an initial and transient de-
crease in coronary flow in the fibrillating heart and
in the beating heart (as does cardiac sympathetic
nerve stimulation). Elevation of extravascular support
as a cause of the early impediment to flow here ap-
parently does not occur since intramyocardial pres-
sure does not rise in the fibrillating heart (27), and
extravascular resistance falls somewhat in the vagus-
stopped heart (236). Unfortunately, since the duration
of the period of constriction is so fleeting and the flow
effect so mild, the view is very difficult to document.
acetylcholine. Acetylcholine, intra-arterially, in-
creases coronary blood flow in the dog in the fibril-
lating heart and in the heart-lung preparation (10).
In the open-chest dog, intravenous injection of acetyl-
choline decreases aortic pressure and coronary flow,
and increases heart rate as a result of a decreased
systemic peripheral resistance (339). Intracoronary
artery injection of effective doses of this hormone, and
also intravenous injection (provided the blood pres-
sure is mechanically compensated by an aortic clamp
and the heart electrically driven following surgical
induction of an A-V heart block), increases consider-
ably left coronary inflow and coronary sinus flow,
and decreases the left ventricular function curve (90,
339, 405). If the intracoronary dose is properly chosen,
this response occurs without a significant effect on the
systolic blood pressure, heart rate, systolic 'diastolic
time interval, cardiac output, cardiac work, but the
myocardial oxygen consumption per minute and per
heart beat increases. The increased coronary flow is
completely abolished by atropine (405). Since the
mechanical and metabolic factors which could in-
fluence coronary flow are thus excluded, the increase
in coronary flow represents a true coronary vasodila-
tation (see fig. 13). The relation between myocardial
oxygen consumption and left ventricular work is not
changed. Consequently, the induced depression of
myocardial contractility or work per unit of filling
pressure is not associated with any change in myo-
cardial efficiency (work per unit of oxygen consump-
tion).
thyroid. The myocardium participates in the in-
crease in oxygen consumption characteristic of all
body tissues in thyrotoxicosis (398). This hyper-
metabolism is accompanied by an increase in coronary
blood flow, a decrease in coronary vascular resistance,
and an increase in oxygen consumption per minute
and per beat. Since there is an increase in oxygen
usage per beat, cardiac oxygen utilization is pre-
sumably related not only to the increase in heart
rate but to the general hypermetabolism of the myo-
cardium as well (230, 317).
Hypothyroidism in man has been shown to be
associated with a reduction in heart rate, cardiac
output, arterial blood pressure, and body oxygen
usage. In vitro studies of experimentally induced
hypothyroidism have demonstrated a reduction in
oxygen consumption of the myocardium (130). Con-
trolled experimental inactivation of the thyroid by-
use of I131 in the dog leads to standardized changes in
the systemic circulation (342). In addition, coronary
sinus flow (N20 method) and left ventricular oxygen
BLOOD SUPPLY TO THE HEART
'553
consumption are reduced. Atropine raises each of these
parameters fas well as the heart rate) to normal.
These experiments are difficult to interpret. Howevei ,
since the stroke coronary flow and stroke coronary
oxygen usage are unaffected by hypothyroidism, the
reduced flow and oxygen consumption are probably
related in part to the altered myocardial metabolism.
pitressin. There is agreement that Pitressin increases
resistance to flow in the coronary circulation (146,
153, 218, 384). In the revived human heart perfused
by the Langendorff method, and in the perfused dog
heart in ventricular fibrillation, Pitressin decreases
coronary flow. In the open- (146) or closed-chest dog,
coronary inflow decreases, the reduction occurring
throughout the cardiac cycle in the presence of an
increased central coronary pressure and, sometimes, a
mild reduction in heart rate. Selective angiography
demonstrates visible vasoconstriction of the superficial
coronary arterial tree (see fig. 13) (393). Although it
seems reasonably sure that this hormone decreases
coronary flow by a direct constrictive action on the
coronary bed, simultaneous studies have not been
made of the associated work and metabolism, and the
possibility of a reduced metabolic influence has not
been ruled out. If Pitressin has a direct action on the
coronary vessels, presumably it acts at the arteriolar
level. This is so since in the isolated perfused rabbit
heart (1 76) Pitressin does not change the intracellular
and extracellular Na and K values. If resistance
increased at the venules or distal end of the capillaries,
one might expect an increase in the extravascular
space.
Exercise and Excitement
Most of the information thus far considered is
based upon observations obtained from the resting
human and the anesthetized, open-chest dog. It is not
known to what extent it applies to normal situa-
tions, since the information has been obtained either
under conditions far removed from normal, as the
result of insults from anesthesia, surgery, and trauma
in the last situation, and hence, it does not contain
information from normal humans and animals as to
the regulation of the coronary circulation exposed to
the stresses of everyday life such as exercise, excite-
ment, and positional changes. For example, in exercise
and excitement, the heart rate is greatly increased.
It is disturbing that in only two of all the conditions
of stress in which heart rate increases, in the open-
chest dog, do the stroke coronary flow and stroke
coronary oxygen increase. These are in thyrotoxicosis
and with cardiac sympathetic nerve stimulation. In
the others, stroke coronary flow and stroke coronary
oxygen decrease. This would mean that coronary flow
and oxygen usage are completely limited by the
heart rate. For example, if the heart rate is tripled,
coronary flow can only be increased three times. It is
difficult to conceive that the heart works in this way,
but rather that additional mechanisms can also
increase the coronary flow per heart beat.
Accordingly, considerable effort has been ex-
pended to make appropriate measurements in the
normal state. It is not to be expected that new pa-
rameters of control will necessarily exist in these
stresses imposed by everyday life, but it is possible that
their weighting will be quite different. Early observa-
tions indicated that in man (240) and in the dog
(105) left coronary flow and myocardial oxygen
consumption increased, while the coronary arterio-
venous oxygen showed little change. More recently, an
appropriate flowmeter has been applied to the
coronary system of an essentially normal animal.
Initially, it was believed that no flowmeter would
operate properly if applied directly to the ventricle
on the surface of the heart because of its violent
motion. Therefore, in large dogs a systemic artery,
either the carotid or internal mammary, was anasto-
mosed by a nonsuture technique to the left circumflex
coronary artery branch so that a flowmeter could
later be mounted on it in a quiescent region off the
surface of the heart. Angiograms and postmortem ex-
amination of the hearts indicated the patency of the
anastomoses and the normalcy of the other coronary
vessels and the myocardium. Of y^ dogs, 6 died of
technical errors on the table or shortly thereafter,
3 died of thrombosis at the site of the anastomosis, 2 to
1 3 days postoperative. The remaining 24 dogs were
sacrificed 1 2 to 24 months later. Prior to sacrifice an
electromagnetic flowmeter, modeled after that of
Kolin (217), was placed on the anastomosed internal
mammary artery and the coronary blood flow meas-
ured daily for periods up to 2 months. Zero blood
flow was obtained when desired by temporarily
occluding the flow by means of a special rubber pneu-
matic cuff placed around the internal mammary
artery just distal to the flow transducer at the time of
its implantation (171).
These preliminary experiments in 1958 were
encouraging. Electromagnetic flowmeters of the
sine-wave type, but greatly modified and improved
from the standpoint of miniaturization, sensitivity,
stability, and ruggedness, were constructed (212a).
The flow probes used on the left coronary artery were
necessarily somewhat smaller than an aspirin tablet,
'554
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
since the maximum space available for implantation
on the main left coronary approximates 2 to 2.5 mm.
A large electromagnetic flowmeter implanted on the
ascending aorta (or pulmonary artery) gave simul-
taneous cardiac output per heartbeat. For phasic
arterial pressure a plastic tube filled with heparin
was implanted in the aorta just beyond the aortic
flow transducer.
Strips of record in figure 1 1 illustrate the flow
through the left circumflex artery anastomosed to the
internal mammary artery in a large greyhound at
standing rest, and running on a treadmill at 1 2 mph
on a 5 per cent grade for 3 min (mild to moderate
exercise for such a dog). As the heart rate almost
triples, the mean coronary flow also triples. Despite
this, stroke coronary flow does not increase but
decreases mildly. The fact that the stroke coronary
flow did not increase with exercise cannot be ascribed
to increased resistance through the much longer
anastomosed circuit because tests in acute experi-
ments showed that coronary flow was the same in
the long and short circuits up to levels of about 650
ml per min (344). Similar flow changes during
exercise occur in the main left coronary artery and
in the descendens and circumflex branches without
anastomosis (212). This suggests that the coronary
flow is limited by the heart rate. It is expected (as
yet without proof) that with quite heavy exercise,
the coronary flow per heart beat will rise.
The coronary flow response to varying degrees of
excitement is quite different from that to exercise.
In the experiment illustrated (fig. 12), a dog at rest
underwent spontaneous excitement. As the heart
rate increases from 98 to 250 beats per min, but
without change in mean blood pressure, the left
circumflex coronary flow increases from 94 to 344
ml per min, and despite the shortened diastole
during which most of the coronary flow occurs, the
coronary flow is more than tripled. In contrast to the
response in exercise, however, stroke coronary flow-
increases from 1.0 to 1.3, this occurring during systole
and diastole. Later, as a moderate increase in blood
pressure occurs, the stroke coronary flow is approxi-
mately tripled. Similar flow responses to excitement
have been observed in the main left coronary artery
with or without a large blood pressure change. These
experiments indicate that the heart is able to obtain
an increased coronary flow during excitement not
only because of the increased number of heartbeats,
but also because of an increased flow per heartbeat.
\'alvular Disease
It is difficult, if not impossible, to duplicate human
valvular disorders experimentally because of lack of
methodologies to assess accurately the degree of
insufficiency and stenosis in both man and beast, to
measure the coronary blood flow, and to make the
REST
(STANDING)
180
120
60
0
•"-I
EXERCISE
RECOVERY
5 SEC
fo
l\
10 SEC
20 SEC
5 SEC
:
M
HYPEREMIA
m
(5 SEC OCCLUSION)
HEART
RATE
MEAN
FLOW
STROKE
COR FLOW
80
45.5
0.57
172
54 7
0.32
214
70.5
0.33
220
110.7
0 48
146
108
0.74
89
45
0.51
fig. 1 1. Reproduction of sections of records taken from an exercising dog showing phasic coronary
flow obtained by an electromagnetic flowmeter mounted on an internal mammary artery anasto-
mosed to the left circumflex coronary artery. Anastomosis performed 1 7 months earlier and flow
probe implanted 7 weeks before. Large greyhound at standing rest, running 3 min on treadmill at
10 mph, recovery, 5-sec occlusion of anastomosis to observe reactive hyperemia. [Khouri ft al.
(212).]
79
108
1.37
BLOOD SUPPLY TO THE HEART
1 555
SPONTANEOUS EXCITEMENT
CONTROL I 25 SEC
HEART
RATE
MEAN
ART B P
MEAN COR
FLOW/MIN
STROKE COR
FLOW (cc)
DIASTOLIC
FLOW/MIN
SYSTOLIC
FLOW/MIN
98
115
94
1.0
76
250
122
314
13
258
56
117
137
340
29
276
64
77
126
200
2.6
177
23
72
113
74
10
66
85
fig. 1 2. Reproduction of sections from a continuous record in a conscious dog some days post-
operative, showing effect of excitement on mean arterial blood pressure and stroke left circumflex
flow, using strain gauge and electromagnetic flowmeter as in fig. 6. [Rayford et al. (301).]
experiments of a long-term chronic nature. It is not
known, therefore, what application to the clinical
situation can be made of present experiments.
AORTIC STENOSIS, PULMONARY HYPERTENSION, PUL-
MONARY EMPHYSEMA AND COR PULMONALE. In the
past, the coronary effects of stenosis of the aortic
valves, cor pulmonale, pulmonary emphysema, and
pulmonary hypertension associated with mountain
sickness have not been studied in humans, largely for
lack of an adequate method. Our information on
these events thus comes largely from the dog.
In experiments with the isolated heart, elevation of
right ventricular pressure by constriction of the
pulmonary artery or elevation of left ventricular
pressure by aortic constriction (coronary perfusion
pressure kept constant) has been demonstrated to
cause a reduction in blood flow to the myocardium
of the right and left heart, respectively (153, 205).
The flow decrease is attributed to the dominant
effect of the direct mechanical inhibitive action of the
increased vigor of the heart or the establishment of
an unfavorable pressure gradient for right coronary
drainage or both. In studies of the isolated supported
dog heart, when coronary perfusion pressure (aortic)
is kept constant, elevation of the resistance to right
ventricular output (an increase in cardiac work)
does not affect total coronary outflow (313). This
means that either the heart is performing the work
much more economically or there is a large safety
factor in the oxygen to be extracted. The latter is, of
course, true in the isolated heart in which the ex-
traction may be only 20 per cent, but in the normal
heart no such wide margin of safety is available,
extraction being of the order of 75 per cent. Therefore,
it remains to be seen whether this dissociation of
cardiac work and coronary flow in the isolated heart
applies to a normal situation.
Acute elevation of right ventricular pressure by
pulmonary artery constriction in the open-chest dog
with constant heart rate is followed by a maintained
increase up to 4 hours in systolic as well as diastolic
blood flow in the right coronary artery in the presence
of the same or some lowering of the aortic or central
coronary artery pressure. In addition, venous outflow
in the anterior cardiac veins increases greatly (1 53).
During the sustained response, both right ventricular
work and metabolism increase, the former being a
result of the increased pulmonary arterial pressure
and a small decrease in cardiac output, the latter
elevation resulting from a combination of an increase
in right coronary flow and a greater oxygen extrac-
1556
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
tion from the right coronary blood. A high degree of
coronary dilatation has obviously occurred since
right coronary artery flow has increased throughout
the cardiac cycle (especially in systole), in the
presence of the same or a lower central coronary
arterial pressure (153). The mechanisms responsible
cannot be identified with certainty. They could be
the opening of closed or partially closed capillaries
and arterioles, the increased passage of blood through
arteriovenous shunts, or increased metabolism. It is
probably not explainable on the basis of myocardial
hypoxia, since, if the right coronary flow is increased
300 to 400 per cent by a constant but very high
infusion pressure, the flow increases still further when
right ventricular pressure is elevated. Regardless,
however, of the mechanism of coronary flow increase,
elevation of right ventricular pressure can also be
shown to have a flow-reducing effect antagonistic to
the flow-promoting mechanisms. In the presence of
an adequately maintained central coronary pressure,
the sustained flow increase and decrease are preceded
and followed by transient periods of flow reduction
and elevation. The initial temporary decrease in flow
is attributed to the dominant influence of augmented
extravascular mechanical compression on the
coronary vessels. The subsequent appearance of a
sustained increased flow observed shortly thereafter
indicates that the effect of coronary dilatation has
exceeded the flow reducing effect of increased extra-
vascular compression. The immediate and transient
flow increase following abrupt lowering of intra-
ventricular pressure is a rough index of the extent to
which flow had previously been retarded by aug-
mentation of extravascular compression.
Concurrent with the elevation of right coronary
flow, left coronary flow and its drainage into the
coronary sinus are significantly elevated (153, 196).
The increase might be of significant magnitude to be
determined in human subjects in the presence of an
elevated right ventricular pressure, although this has
not been found (312). It might be conjectured that
the source of a portion of this increased left coronary
blood is increased flow through the ventricular
septum, much of which normally drains into the right
ventricular cavity but which, because of high right
ventricular pressure, might be diverted into the
coronary sinus (265).
Such responses of flow and metabolism in the
right myocardium to elevation of its cavity pressure
are not peculiar to it. Elevation of left ventricular
pressure by an aortic constriction central to the
coronary ostia, i.e., between the aortic valves and
coronary ostia to stimulate aortic stenosis, gives
trends for coronary flow and myocardial metabolism
of the left ventricle identical to those found in the
right ventricle (1 53).
These maintained changes in the coronary circula-
tion could well be the early response in the human
being to gradual moderate stenosis of the correspond-
ing valves.
aortic insufficiency. In patients with aortic in-
sufficiency and lacking disease of the coronary ostia
or arteries, the presence of chest pain resembling that
due to myocardial ischemia is generally assigned to a
reduction in coronary blood flow, this arising pre-
sumably from a reduction of the mean aortic or
central coronary diastolic pressure. In the open-chest
dog reversible aortic regurgitant flow has been
accurately produced without valve injury, metered
and controllably varied, while at the same time
metering cardiac output. Aortic regurgitant flows in
excess of the dog's resting cardiac output resulted in a
marked decrease of effective cardiac output, a rise of
peripheral resistance and left ventricular end-diastolic
pressure, and a marked depression of the left ventricu-
lar pressure curve without significant change in mean
left atrial pressure. No coronary flows were measured
(390). In early experiments in the open-chest dog in
rather poor condition (144, 153) reversible aortic
insufficiency (umbrella-type aortic valve expanders)
sufficient to lower aortic diastolic pressure decreases
mean left coronary flow as a resultant of an increased
systolic flow and a markedly reduced diastolic flow.
On the other hand, Foltz et al. (119), from measure-
ments on anesthetized dogs two or three days after
the aortic cusps had been torn, found a considerable
increase in coronary flow and myocardial oxygen
usage. This latter finding has been confirmed by
Wegria in acute experiments, and West in chronic
dogs (386, 394). In patients with reduced diastolic
pressure, wide pulse pressure and varying degrees of
left ventricular enlargement, those without angina or
failure have normal coronary hemodynamics; those
with angina have a reduced coronary flow and cardiac
oxygen usage; and those in failure have an increased
coronary flow and oxygen usage (constant coronary
A-Y oxygen difference) (303). Such observations are
difficult to interpret because of their small number,
the lack of adequate control data, and the possibility
of complicating disease of coronary ostia or arteries,
or both.
mitral stenosis. The general hemodynamic effects
from mitral stenosis include increased wedge pressure,
pulmonary arterial pressure, right ventricular work,
BLOOD SUPPLY TO THE HEART
1557
and decreased systemic blood pressure, cardiac index,
and cardiac work index. In small groups of human
subjects, a normal or decreased coronary blood flow
has been reported (216). In a large group of females,
the above systemic changes have been found to be
associated with a decrease in left ventricular coronary
blood flow, increased coronary oxygen extraction,
and decreased left ventricular oxygen utilization per
unit of myocardium, as compared with normal
females but not as compared with normal males
(321). This depression of the left coronary circulation
in the presence of a lowered activity of the left
ventricle would be expected. Acceptance of these
data, although in line with those previously reported
in the intact dog (118), should possibly be deferred
until previous work indicating that the left coronary
circulation in the female is maintained at a con-
siderably higher level than in the male (320) is
confirmed.
mitral insufficiency. Mitral regurgitation has been
experimentally produced in the open-chest dog by
permitting blood to flow externally from the left
ventricular apex through a flowmeter into the left
atrium during systole. Such controlled regurgitant
flows, up to three times the resting cardiac output,
are tolerated with only slight or mild alterations of
effective cardiac output, aortic, left atrial and left
ventricular pressures, total peripheral resistance, and
the effective left ventricular function curves (45). In
anesthetized open-chest dogs, acute mitral insuffi-
ciency of variable severity, produced by means of an
umbrella-type valve spreader so as to allow a partial
or incomplete return of the aortic flow to its control
level, results in a moderate increase in coronary blood
flow and myocardial oxygen usage and a reduced
efficiency. Presumably, the left ventricle expends a
significant amount of energy in regurgitating blood
into the left atrium during mitral insufficiency (388).
No comparable studies of this nature are available in
humans.
aortic coarctation. With simulation of clinical
coarctation by acute mechanical constriction of the
thoracic aorta just beyond the left subclavian artery,
venous return to the heart by way of the inferior vena
cava is decreased but compensatory flow through
various branches of the aortic arch may increase,
with a resultant maintained cardiac output and
elevated left ventricular work load. With greater
aortic constriction, the net cardiac output decreases,
causing the cardiac work to decrease. In either case,
the coronary dilatation and increased flow arise in
large part from active changes in the bore of the
coronary bed related to the metabolic demands, and,
in part, passively from the increased blood pressure
and moderately decreased heart rate (90, 207). The
cardiac oxygen consumption is increased much more
by this augmentation of pressure work than with an
equal increase of volume work following transfusion
(335). No chronic studies of aortic coarctation have
been made because, owing to development of col-
lateral circuits, the aorta may be first partially and
then completely constricted at the arch without
permanent development of hypertension proximal to
the occlusion. In human coarctation not much
change is reported in coronary flow and oxygen
uptake, but this might be expected because systemic
pressure is only mildly elevated (31). However, if true,
the deviation might be explained by the fact that in
these hearts, which are hypertrophied, there are
fewer capillaries per unit of muscle to carry the oxygen
and flow.
Hypertensive Cardiovascular Disease
An exception to the general picture of coronary
compensation to increased systemic stress appears to
be the response of the chronically hypertensive heart.
In essential hypertension, with a normal cardiac
output and elevated systemic blood pressure, the
coronary flow and oxygen consumption per 100 g
myocardium are unaltered while coronary resistance
increases. This increased resistance is shared with the
renal and cerebral circulations. Since these hearts are
generally hypertrophied, total coronary flow and
oxygen usage are probably increased. This deviation
is explainable if it is assumed that such hearts with
known coronary artery disease have an increased
amount of perfused fibrotic tissue (31, 316).
Heart Failure
Although the underlying mechanisms for various
types of heart failure may be different, the basic
hemodynamic manifestations of cardiac failure are
similar from causes such as congestive heart failure,
anemia, anoxia, hemorrhagic shock, myocardial
infarction, hyperthyroidism, and beriberi. Experi-
mentally, such hearts exhibit depressed Starling or
ventricular function curves (increased ratio of end-
diastolic ventricular volume or ventricular filling
pressure to stroke work), and show the characteristic
optimum beyond which further stretching reduces
the force of contraction and leads to myocardial
failure. In acute heart failure in the open-chest dog,
'558
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
with progressive deterioration of the right myocardium
from pulmonary artery stenosis, the changes in
coronary flow and oxygen usage per minute and per
beat may be in the same direction (increase) as those
described for the nonfailing right myocardium, but of
lesser magnitude (see section on Valvular Disease).
If the heart failure is severe enough, extravascular
compression can become dominant over any active
coronary dilatation from metabolic processes, and
coronary flow and oxygen usage may be normal or
decrease, with the oxygen extraction at times reaching
go per cent (153)- The coronary circulation in the
heart, failing with severe aortic stenosis, undergoes
similar changes (Gregg, unpublished data). When
acute heart failure and chronic congestive failure
simulating the human condition are induced by
surgical complete heart block, changes in left coronary
flow and ventricular oxygen consumption also rather
closely parallel alterations in the reduced left ventricu-
lar work (355). In each instance, the mechanical
efficiency of the myocardium drops, the total energy
of liberation (oxygen consumption) being fairly well
maintained, but the work falls off. The isolated
mammalian heart or heart-lung is also characterized
by deterioration of mechanical efficiency and on the
same basis (74, 243). In chronic left heart failure
due to rheumatic, arteriosclerotic, and hypertensive
heart disease, the coronary circulation apparently
responds by a slight increase in oxygen usage through
maintenance of the left coronary flow and an in-
creased coronary A-V oxygen difference. This
corresponds with the changes indicated for the right
heart in an early stage of failure. As is true for the
heart-lung or isolated heart, such hearts have con-
siderable difficulty in transforming released energy
into realizable work. Studies of the coronary circula-
tion in high-output failure from excessive transfusion
or a chronic aorta-caval fistula are not available. In
the anesthetized open-chest dog, however, an acute
arteriovenous fistula sufficient to increase cardiac
output and cardiac work causes considerable aug-
mentation of stroke coronary flow and stroke coronary
oxygen even in the presence of a sizeable decrease in
arterial blood pressure (389).
When acute heart failure induced by pulmonary
artery constriction has advanced to the stage where
systemic blood pressure is low, left ventricular size is
small, right ventricular size is large, and release of
the constriction does not restore ventricular working
capacity, then the use of arterial transfusion with a
pump temporarily promotes functional recovery of
the heart (increase in arterial and coronary perfusion
pressure, coronary flow, cardiac output, cardiac work,
and myocardial vigor, decrease in cardiac size and
coronary A-V oxygen difference). Veno-arterial
pumping accentuates and makes more permanent
these beneficial changes. Since an increase in coronary
flow invariably precedes recovery of the heart, it
suggests that it is a primary stimulus through an
effect on myocardial metabolism for increased
ventricular performance and decrease in heart
size (14).
Postmortem specimens from human patients may
show myocardial edema (increased water, Na and CI
content per unit of myocardium) in the presence of
congestive heart failure, acute infarction, and ischemic
areas. Very often, however, previous drug administra-
tion, together with agonal and postmortem changes,
makes interpretation difficult. The isolated dog heart
does not develop edema when directly perfused from
a donor dog, but increased myocardial water content
is found after acute cardiac injury (over-distended
ventricle), excessive perfusion pressure, increased
coronary venous pressure, and by perfusion with
blood from a disposable bag oxygenator system. It
occurs spontaneously in the failing heart-lung prepa-
ration and in chronic heart failure produced experi-
mentally by thoracocaval constriction, pulmonary
stenosis, and tricuspid insufficiency, separately or
together. The mechanism or mechanisms involved are
unknown, but in these chronic preparations, elevation
of right atrial pressure seems to be a major pathogenic
factor in its formation. Whether its presence con-
tributes to abnormal cardiac function or whether its
prevention or reversal is a therapeutic objective in
the management of heart disease is a moot question
(329. 409 )•
Hemorrhagic Slunk
Standardized oligemic shock in dogs is characterized
during the hypotensive phase by a decrease in
cardiac output, systemic blood pressure, cardiac
work, stroke volume and stroke work, and by an
increase in heart rate and an adequate central venous
pressure. Coronary flow and coronary resistance
are greatly decreased but the coronary flow fraction
of cardiac output is increased (102). Coronary flow is
generally greater and the resistance generally less than
can be accounted for by a simple decline in arterial
blood pressure (281). At the same time, the oxygen
uptake decreases and the coronary arteriovenous
oxygen difference is generally unchanged (166). The
coronary response to sustained hypotension through
BLOOD SUPPLY TO THE HEART
1559
spinal anesthesia or injection of procaine and Etamon
is similar (168). With partial or complete restoration
to normal systemic blood pressure by reinfusion
(intra-arterial and intravenous routes are equally
effective) (56, 361), coronary flow is greater and flow
resistance is less than at an equivalent arterial blood
pressure in the preshock state.
The fact that early in the hypotensive phase neither
ventricular end-diastolic pressure nor atrial pressure
rises indicates that the functional capacity of the
heart is adequate for the work performed. However,
that myocardial depression or failure is partially
responsible for the hemorrhagic shock syndrome is
suggested by different observations, a) After prolonged
hypotension, there may be evident cardiac dilatation
and elevated left and right atrial pressures with the
heart eventually proceeding to ventricular fibrillation
or standstill (331). With spontaneous cardiovascular
decay after reinfusion, the atrial pressure may be at a
normal or elevated level despite large cardiac output
reduction (402). b) During prolonged oligemic
hypotension, as the animal starts to take up blood
from the reservoir to maintain its falling blood
pressure, the atrial pressure may rise to very high
levels (163). Gross and microscopic evidence of
myocardial injury appears in both reversible and
irreversible shock. Such myocardial depression could
be caused by an insufficient coronary flow during
either the hypotensive or the post-hemorrhagic
periods. The high coronary flow during the restora-
tion period would seem to preclude an inadequate
coronary flow as an adequate explanation. During
the hypotensive period, the actual coronary flow is
greatly curtailed. The problem is whether the
associated sizeable reduction in coronary resistance
is sufficient to permit enough blood to reach the
myocardium to prevent it from failing. In some
instances, at least, this loss of myocardial contrac-
tility is consequent upon an insufficient coronary
flow, since the relation of atrial pressure to cardiac
size can be reversed by increasing left coronary flow
mildly with a pump, without change in either the
hypotension or blood volume (331).
The work just discussed has been largely restricted
to studies in experimental animals exposed to anes-
thesia, surgery, and varying amounts of traumatic
insult. More proper studies might be conducted in
intact conscious dogs; this is possible with methodology
now available. This type of study has been made with
the use of modified and improved electromagnetic
flowmeters which were chronically implanted on the
left coronary artery as well as the aorta and various
systemic arteries (159). The experiments confirm
previous findings that, of all the arterial beds, only
the coronary shows a decreased vascular resistance
during hemorrhagic irreversible shock, and add new
information regarding the compensatory behavior of
the left coronary vascular bed. The coronary pressure-
flow ratio moderately increases during hemorrhage,
progressively decreases during the hypotensive period
as the coronary flow increases spontaneously, and is
temporarily restored during the reinfusion. During
the irreversible period, in which the coronary flow is
fairly well maintained, the pressure-flow ratio again
drops. The resistance, however, to coronary flow is
somewhat less during the period of spontaneous decay
than during the initial hypotensive period. These
pressure-flow changes may have their explanation in
certain characteristic changes in the coronary flow
pattern. The phasic flow pattern, initially some dis-
tance above the zero flow line throughout the cardiac
cycle, moves closer to the zero flow line during
hemorrhage, and backflow may appear during
systole. The magnitude of the flow pattern, however,
increases, indicating increased vigor of contraction.
As the hypotensive period progresses, flow is re-
established in systole and increased somewhat in
diastole. Following reinfusion, and late in the period
of spontaneous hemodynamic decay, the flow pattern
may resemble somewhat the prevailing aortic pres-
sure pulse with the systolic flow equal to or exceeding
the diastolic flow. The mechanisms whereby coronary
systolic flow is thus preferentially enhanced are not
known.
Hypothermia
The circulatory and metabolic adjustments of the
heart during hypothermia have been partially
explored (87). When the body temperature is dropped
from 37 C to 20-28 C, by immersion hypothermia or
by cooling the systemic arterial blood flow, the
associated changes that occur which tend to reduce
the coronary flow are a diminution in blood and
muscle temperatures, cardiac output, heart rate,
cardiac work, and oxygen usage by the heart, an
increased blood viscosity and a greatly lengthened
period of ventricular systole. The coronary A-V
oxygen difference remains about normal or decreases
(103, 128, 175, 180, 322). Opposing these factors are
the relaxation of the major coronary vessels, which is
known to occur with hypothermia, and dilatation of
the coronary bed caused by the hypotension per se
(25, 177). As a resultant of these determinants,
[560
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
coronary flow is decreased at low temperature.
However, the per cent reduction in cardiac output is
greater than that in coronary flow, which results in an
increase in the coronary fraction of cardiac output at
temperatures of 25 to 26 C (103, 322). There is little
change, a decrease, or an increase in peripheral
resistance in the coronary bed, whereas in the systemic
bed an increase in peripheral resistance invariably
occurs (1 77, 322). A constant or increasing mechanical
efficiency is usually observed in the open- or closed-
chest dog (128, 175) although it has been reported to
fall (103, 198). Similarly, in the heart-lung prepara-
tion or isolated heart, the mechanical efficiency is
fairly constant when cardiac work per beat (same
stroke volume and arterial blood pressure) and
heart rate are constant (15). Myocardial function
appears to be adequate and myocardial hypoxia not
to exist (198). However, many hearts are apparently
not too far from failure because if total venous inflow
occlusion (which decreases coronary flow close to
zero) is now added to permit open cardiotomy,
myocardial failure supervenes, as evidenced by-
elevation in mean right atrial pressure and post-
mortem findings. This trend can be reversed by
perfusion of the coronary system with small volumes
of oxygenated blood (239).
Hyperthermia
The systemic dynamic changes resulting from
elevation of body temperature by fever or external
application of heat (hot baths, diathermy) are quite
similar in man and dog, are well documented, and
include considerable elevation of heart rate, blood
pressure, cardiac output, right and left ventricular
work, a decreased peripheral resistance and a con-
stant stroke volume and stroke work. The little
information available on the associated coronary
changes indicates that a large elevation of body
temperature (up to 105 F) by means of diathermy,
in the closed-chest dog, increases considerably
coronary blood flow, myocardial oxygen usage,
coronary A-V oxygen difference, increases mildly the
stroke coronary oxygen usage, decreases external
efficiency, and leaves unchanged the stroke coronary
flow and coronary resistance (257). In open-chest
dogs with an initial hypotensive systemic blood
pressure, diathermy has no effect on coronary flow
(253). In the heart-lung preparation, when the
myocardium is warmed, coronary flow and oxygen
usage are increased; but they are not when the
coronary blood is warmed (10).
Summary
Over the years, the basic mechanisms affecting
coronary flow and oxygen usage have been related
experimentally to various parameters, and the state-
ment is often made that there is one controlling or
unifying influence for coronary flow per heart beat
and also one for oxygen usage of the left myocardium
per cardiac cycle, a) Directional changes in stroke
coronary flow correlate with stroke coronary oxygen
usage. This is so, however, because normally most
oxygen is removed from the coronary blood and the
level of coronary sinus oxygen is usually fairly con-
stant under stress, i.e., it does not change more than
10 to 15 per cent. In those instances in which the
coronary arteriovenous oxygen difference increases
or decreases by this amount, it does not greatly
affect the relation of coronary flow to oxygen usage
since the change is very small relative to the magni-
tude of the coronary flow change, but this does not
document a functional correlation between these
parameters, b) The stroke coronary flow and stroke
coronary oxygen correlate fairly well with the stroke
work under a variety of conditions of changing
systemic stress, but it is possible to so regulate experi-
ments that the response of the coronary circulation is
dissociated from stroke work, c) The exceptions to
the usual correlation of stroke work and coronary
oxygen usage constitute a group of conditions in
which the outflow channels of the two ventricles have
been restricted in some manner. In these, one can
show excellent correlation of the stroke coronary flow
and oxygen usage with the mean systolic arterial
blood pressure alone, or with the product of the
systolic blood pressure and the duration of systole,
the so-called "tension-time index." However, experi-
ments in the unanesthetized dog during exercise and
excitement do not always support this view. In
addition, there is quite a list of determinants that
have been thought to be fundamental. Attempts
have been made to relate coronary flow and oxygen
usage to the mean arterial blood pressure, ventricular
filling pressure or mean atrial pressure, ventricular
diastolic volume or fiber length, tension within the
ventricular wall, oxygen tension of the arterial blood,
oxygen tension within the myocardium, action of
local metabolites or vasodilating substances. Possibly,
the best correlation of all should be with the reduction
of cytochrome oxidase and the needs of the hydrogen
transport system. Probably no final answer is avail-
able. Final decision as to whether any of these
determinants of coronarv flow or oxygen usage are
BLOOD SUPPLY TO THE HEART
1561
primary or empirical must await the necessary
measurements under normal conditions of stress
without anesthesia or surgical insult.
DRUGS VERSUS THE CORONARY CIRCULATION
The pertinent literature has been reviewed (7, 13,
49- 68. 145. '53. 209, 238, 275, 384, 413). Con-
sideration will be given here only to the effects of a
few selected drugs on the coronary circulation in the
normal state and in the presence of coronary artery
disease.
Drugs may be effective in altering the normal and
collateral myocardial blood supply by a direct effect
on the vasomotor state of the vessels, by an increase
or decrease in central blood pressure, by myocardial
stimulation or depression, by a change in the cardiac
workload through extracardiac phenomena, or by
electrolyte, pH or gaseous alterations of the blood
perfusing the coronary bed. It is interesting to know
whether a drug affects the extravascular and intra-
vascular resistances of the coronary bed, but it is
more important to know its effect upon the supply of
oxygen to the myocardium, the oxygen used by the
myocardium, and the efficiency of the heart in the
use of its oxygen for the work performed. In addition,
a pharmacological agent may be able to improve the
oxygen utilization for external work of the heart
without an increase in coronary blood flow or oxygen
extraction.
In order to properly evaluate an agent, the follow-
ing information is necessary: a) coronary blood flow,
b) arteriovenous oxygen difference across the coronary
bed, c) blood pressure, d) cardiac output, e) myo-
cardial contractility, and /) heart rate. From these
data, the myocardial oxygen availability and usage,
cardiac work and efficiency can be calculated.
Few drugs have been completely studied. The
pharmacologic agents will be considered from the
standpoint of: a) the effects of therapeutic or poten-
tially therapeutic drugs on the normal myocardium
undergoing normal or excessive stress, and b) the
effects of nontherapeutic drugs on the normally
stressed myocardium. In the normal and hyper-
tensive heart, the ganglionic blockers such as hexa-
methonium decrease both cardiac work and myo-
cardial oxygen availability but not the oxygen usage
(73, 162). Nitroglycerin has no apparent direct effect
on the amount of free energy released with each
contraction of the myocardium either before or after
partial coronary artery occlusion, but rather reduces
hemodynamic workload by a decrease in left atrial
filling pressure (75). Experiments in dogs with
sodium nitrite or nitroglycerin injected into the left
coronary artery show that the coronary arteries and
their small branches dilate (fig. 13) (393), and that
the flow increases greatly in both systole and diastole
and in the presence of a decreased central coronary
pressure, cardiac output, cardiac work, a constant
heart rate and only a slight decrease in the systolic:
diastolic ratio. Hence, the conclusion is inescapable
that these drugs exert a vasodilating action on the
coronary vessels (41). This could arise from a direct
effect of the drug on the coronary vessels since cardiac
metabolism is not increased (decreased coronary
A-V oxygen difference and increased coronary flow)
(94, 333). Experiments in normal man with nitro-
glycerin, however, show an increased coronary blood
flow with an increased myocardial oxygen uptake
(constant coronary A-V oxygen difference and
increased coronary flow), decreased cardiac work and
decreased cardiac efficiency (42). Furthermore, in
patients with coronary artery disease, this drug does
not increase coronary flow while, with a steady
oxygen extraction, it decreases cardiac work (de-
creased blood pressure and cardiac output) (138).
These data raise the old question of the applicability
of knowledge obtained in normal animal or human
studies to the diseased states. If these studies should
be confirmed in patients during anginal attacks, other
theories for the action of nitroglycerin must be con-
sidered. One theory holds that nitroglycerin blocks
the anoxia-inducing effect of the catecholamines on
the heart (296), but an antiadrenergic action could
not be demonstrated for this drug (94). One might
postulate that a decrease in cardiac work secondary
to the decrease in blood pressure, in the presence of a
stable oxygen consumption, may be helpful to the
myocardium despite a calculated decrease in myo-
cardial efficiency. Present calculations include only-
evaluation of the external efficiency of the heart. If
such hearts are using all the oxygen they could
extract at a given workload, then a decrease in this
work, at the same level of oxygen consumption, might
be beneficial.
With the xanthines mean coronary flow is increased,
this being the net result of a marked increase during
diastole and a decrease during systole which occurs
in the presence of a normal or mildly decreased blood
pressure and without significant change in cycle
length or systolic: diastolic ratio. Visually, the heart
shows increased vigor and its metabolism and work
are increased (41). Nikethamide acts similarly by
1562
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 13. Effects of coronary
arterial injection of Pitressin,
nitroglycerin, and acetylcholine
in the anesthetized dog. Catheter
in the first part of the anterior
descendens branch (left lateral
view). Radiopaque material in-
jected in all cases 5 cc. A : control
angiogram (BP 150/115; HR 90)
before Pitressin. B: (BP 145/115;
HR 85) 1 min after Pitressin in-
jection (0.008 units/kg). C: con-
trol angiogram (BP 175/125;
HR 75). D: (BP 165/115 HR
78) after nitroglycerin injection
(5 MgAg). E: control (BP 177/
115; HR 84). F: (BP 170/117;
HR 84) immediately after acetyl-
choline injection (0.04 MgAg).
[Modified after West & Guzman
(393)0
increasing the oxygen available to the myocardium
but at the expense of an increased oxygen usage and
cardiac work (91). It would probably be preferable
to increase the oxygen availability without stimulation
of myocardial oxygen metabolism, as has been shown
for nitroglycerin and papaverine in animal studies
(75, 117, 214). Hydralazine, which may precipitate
anginal attacks in hypertensive patients, possesses
what are usually considered the desired properties
for a coronary vasodilator: it increases coronary
blood flow, increases oxygen availability, and does
not alter the oxygen uptake of the myocardium (316).
BLOOD SUPPLY TO THE HEART
1563
It would appear that our concepts of the "ideal"
agent for coronary artery disease must be revised.
Information is confusing regarding the coronary
effects of khellin, a drug used in the eastern Medi-
terranean regions since ancient times in renal colic
and ureteral spasm. Interest in its possible cardiac
effect arose as the result of the discovery that, orally
or intravenously, it acts for many hours as an ex-
tremely potent coronary vasodilator in the heart-lung
preparation and in the heart in situ. In the doses
used, it has no effect on the general blood pressure
and does not increase the oxygen requirements of
the heart, i.e., it acts only to relax the intrinsic smooth
muscle of the coronary arterioles (153). However,
others (109) did not find it effective on the coronary
or systemic circulation of the anesthetized dog.
There is little doubt that digitalis augments
ventricular contractility and the peripheral circulation
(48, 314). However, in normal human subjects,
strophanthus apparently has a deleterious effect
since it decreases cardiac work and efficiency without
altering coronary flow, oxygen supply, or usage of
the myocardium (31). Conversely, in the patient with
congestive heart failure, it has a salutary effect by
acting to increase cardiac work and efficiency without
using more oxygen or altering the coronary circula-
tion. This is another case in which the action of a
drug is entirely different in the normal subject from
what it is in the diseased subject.
Ever since Favarger (110), in 1887, claimed that
excessive tobacco smoking produced coronary vaso-
constriction which, repeated over many years,
gradually resulted in organic heart disease, tobacco
has been considered an important cause of coronary
disease. The alterations of the ECG (T-wave depres-
sion and sagging of the S-T segment) and of the BCG
in the normal heart, or the heart with coronary
arterial disease, that follow inhalation of tobacco
smoke or administration of nicotine, have been
generally thought to result either from coronary-
arterial constriction or from an increase in the work
of the heart beyond the capacity of the coronary
arteries to supply the necessary metabolic require-
ments of the myocardium. Observations on the
normal heart of the anesthetized dog do not support
this view. Intracoronary nicotine injection greatly
increases myocardial contractility, and in dogs pre-
treated with Dopa, nicotine increases considerably
the myocardial catecholamine concentration (202).
The circulatory responses to administration of
cigarette smoke or nicotine generally include elevation
of heart rate, blood pressure, cardiac output, cardiac
work, left coronary flow, myocardial oxygen usage,
and a decrease in coronary vascular resistance. The
coronary oxygen extraction may be decreased, and
often the oxygen usage may be transiently unchanged
or decreased (213). These responses can all be blocked
by injection of tetraethylammonium chloride. These
effects parallel those observed with epinephrine
injection and are presumably related to its release.
The response to nicotine of the coronary flow, in dogs
with coronary insufficiency from coronary arterial
ligation or gradual coronary artery narrowing, is
considerably less than in dogs with normal coronary
arteries (20); in the isolated atherosclerotic rabbit
heart, it decreases coronary flow (364). In normal
man, earlier findings indicate that cigarette smoking
increases coronary blood flow and heart rate, and
decreases coronary vascular resistance in the presence
of reduced systemic dynamics, whereas in patients
with coronary artery disease, smoking causes no
appreciable change in coronary flow and myocardial
oxygen consumption in the presence of increased
heart rate, blood pressure, cardiac output, and cardiac
work (304). This suggests that the electrocardio-
graphic changes observed during smoking are the
result of a relatively deficient oxygen supply to the
myocardium in the presence of increased oxygen
needs (increased cardiac work). Later reports from
the same laboratory indicate, in subjects both normal
and with coronary artery disease, that smoking
increases heart rate, blood pressure, and left ventricu-
lar work but does not alter coronary flow or cardiac
oxygen usage (306). Decision as to the action of
smoking and nicotine in human subjects must be
deferred until the various neurohumoral responses
evoked are better understood and better methodology
is available.
Certain agents which increase coronary blood flow
will not be discussed because of the scant information
available. Isoproterenol, histamine, antihistamines,
heparin, Dicumarol, ethanol, 5-hydroxytryptamine,
Amplivix, and RA-8 increase coronary flow in animal
preparations (62, 200, 224, 238, 299, 384). Reports
are conflicting regarding Metrazol, quinidine,
quinine, and morphine (299, 318, 384). Only two
drugs, Pitressin and angiotensin, decrease coronary
blood flow without a decrease in central blood
pressure (150).
For clinical use in angina pectoris, drugs are
evaluated in patients by methods involving their
ability to alter the electrocardiographic response to
an exercise test or by drug-placebo (double blind)
studies (70, 157). The selection of patients for these
1564
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
trials is very difficult due to the variability of the
disease and its response to many extraneous factors.
Nitroglycerin remains the only universally preferred
treatment. Other agents, including the long-acting
nitrates, have received some favorable but also many
unfavorable reports. The mode of action of the mono-
amine oxidases is unknown although the clinically
beneficial but toxic iproniazid has been demonstrated
to increase coronary flow and depress cardiac con-
traction in isolated hearts (160). Anticholesterol
agents and thyroid are used in the hope of decreasing
the atherosclerotic process, but long-term studies are
needed to assess their value. However, study of
hyperthyroid patients shows that their myocardial
oxygen utilization is increased more than would be
expected from the increased heart rate (3 17)-
Table 1 is a compilation of data concerning the
action of the vasopressor agents most commonly used
in cardiogenic shock. Since most of these studies were
performed in normal animals or man, the question
again arises whether the information can be applied
to the diseased state. Against such application is the
demonstration that drugs may act differently in a
normal, as compared to a failing heart, i.e., mephen-
termine may exert a myocardial oxygen-conserving
effect in failing animal hearts but an oxygen-wasting
action in normal hearts (391). Since the etiology of
cardiogenic shock is undetermined (decreased pe-
ripheral resistance, or myocardial failure or both),
controversy exists regarding its treatment. The
question concerns whether an agent should be used
which not only increases coronary flow and blood
pressure (methoxamine), but also stimulates the
myocardium and, therefore, increases myocardial
oxygen consumption (epinephrine, levarterenol,
metaraminol, and mephentermine) (13, 237). It
appears that clinical results favor the latter concept,
for levarterenol has met with the most success. Also,
in animal studies the vasopressor agents which stimu-
late myocardial contractility and lower atrial pressure
are more beneficial to the "failing" heart (332).
However, methoxamine has been shown to be of
use for increasing the blood pressure and coronary
flow in hemorrhagic shock in animals, and also in
cardiogenic shock in patients (13). A useful agent for
cardiogenic shock should increase coronary flow,
stimulate myocardial contractility, and raise the
blood pressure and cardiac output, but should not
increase myocardial oxygen consumption in relation
to its workload. It was pointed out above that myo-
table 1. Pressor Agents Used in Cardiogenic Shock*
Drug
Route
Blood
Pressure
Systolic
Diastolic
Heart
Rate
Myo-
cardial
I'mM r ci]
Con-
traction
Coro-
nary
Blood
Flow
Myo-
cardial
o2
Uptake
Coro-
nary
A-V C:
Differ-
ence
Cardiac
Output
Cardiac
Work
Cardiac
Effi-
ciency
Total
Periph-
eral Re-
sistance
Right
Atrial
Pres-
sure
Epinephrine
1. V.
+ /-
+ (±)
(+)
(+)
(+)
(-1
+
+
(-)
—
(Adrenaline)
I. c.
(+)
c+)
(+)
(+)
Levarterenol
I. V.
+ / +
-<±)
(+)
(+)
(+>
(-)
-(±)
+ (+.)
(-)
+
(-)
(Levophed)
I. c.
(+)
(+)
(+)
(+)
Ephedrine
1. V.
I. c.
+/o
+
(±)
(+)
+ (±)
+
-
Metaraminol
I. V.
+ / +
_
(+)
(+>
o(±)
+
+
(-)
(Aramine)
I. c.
(+)
(+)
Mephentermine
I. V.
+ / +
—
(+)
(+)
(+)
(-)
0(±)
+
(-)
+ (0)
(")
(Wyamine)
I. c.
(+)
<+)
Phenylephrine
I. V.
+ /+
—
(+)
( + )
±
+
+
+ t
(Neosynephrine )
I. c.
(+)
(+)
Methoxamine
I. V.
+ / +
—
(0)
(+)
(-)
(-)
(+)
(+)
(Vasoxyl)
I. c.
(O)
(0)
* Results in man. Figures in parentheses indicate if dogs react differently or if only dog results are available.
Key: I. V. = intravenous; I. C. = intracoronary ; + = increase; o = no change; — = decrease; ± = variable effects or
conflicting data; t = venous pressure.
BLOOD SUPPLY TO THE HEART
1^6
3»3
cardial contractility and oxygen consumption may be
dissociated. Such drugs may be available but more
experimental proof is necessary.
CORONARY ARTERY DISEASE
The basic pathological lesion in coronary artery
disease is the atheroma which eventually leads to
narrowing or occlusion of the coronary artery lumen
by progressive intimal thickening, intimal ulceration,
hemorrhage or superimposed thrombosis. Thrombosis
on an arteriosclerotic basis (43%), arteriosclerosis
with and without infarction (41%), and intramural
hemorrhage (8'c), also presumably on an arterio-
sclerotic basis, account for about go per cent of
coronary artery lesions (379). Coronary artery
narrowing or occlusions are limited to the three
main coronary arteries (50 '"< ) and their primary
branches (50%), and are almost entirely epicardial
(100). The lesions are localized, segmental, and
multiple (avg. 2.5/heart), and 70 per cent occur
within 3 to 4 cm of the coronary ostia (39, 184).
As a result of this occlusive process, the myocardial
circulation is reduced to a variable degree, depending
upon the nature and extent of the lesion and the
extent to which intercoronary artery collateral
development takes place. Serious consequences occur
when the extent of the former is large or the latter
mechanism fails to compensate for the ischemic
changes produced by the atherosclerotic process.
The following, singly or together, may then take
place: angina pectoris, myocardial infarction,
mechanical failure, or sudden death.
Since Heberden's classic description in 1 768 of the
syndrome of angina pectoris, much effort by medical
investigators has been directed toward this problem.
While there are some dissenting voices (297), general
consideration indicates that the production of pain
arises from stimulation of sensory cardiac nerve
endings which, in turn, arises from imbalance in the
heart between supply and demand of oxygen. Sensory
nerve endings of the heart (and aorta) are present in
the myocardium, endocardium, and epicardium,
and in the adventitia of the coronary arteries. Their
associated neurones converge in the periarterial
plexus of the coronary arteries, continue through the
superficial and deep cardiac plexuses and course in
the middle and inferior cardiac nerves to join the
corresponding cervical ganglia of the sympathetic
chain. These centrally bound fibers then descend to
the upper thoracic ganglia and reach their cells in the
spinal ganglia by passing through the white rami
communicantes into the first thoracic and upper 4 or
5 intercostal nerves. They cross to the opposite spino-
thalamic tract and course through the brain stem to
the thalamus (397, 401).
From the preceding, it is obvious that pain could
be relieved in different ways: by raising the cerebral
level or threshold for pain perception, by attenuation
of factors in the environment that lead to stimulation
of the cardiac pain end organs, by the induction of
proper coronary vasodilatation. However, the physio-
logical evaluation of angina pectoris, and of the
effects of medical and surgical therapy on it, is
limited to study of the relief of the angina of effort in
cases where attempts are made to delete the sub-
jective element of pain, and to the measurement in
equivocal cases of the coronary flow response to
vasodilator drugs such as nitroglycerin to determine
the ability of the coronary bed to dilate on demand.
The latter is predicated upon the experimental
finding in advanced coronary artery disease of
fixation of the coronary flow when challenged by
nitroglycerin (44). Whether the ability of a drug to
diminish anginal episodes or to improve the electro-
cardiographic response in exercise is an objective
measure of positive benefit to a stressed myocardium
is still debatable. This is so because it is not known
to what extent the influence of the physiology of
sensation on angina has been removed, and the
assumption must necessarily be made that the electro-
cardiographic response correctly indicates myocardial
hypoxia or ischemia.
As yet, experimental studies directly attacking the
problem of coronary atherosclerosis have not been
productive in elucidating the mechanism of or
prevention of the lesion. However, the functional
consequences and compensatory physiological re-
sponses to controlled experimental coronary con-
striction and occlusion, or to the loss of functional
myocardial areas in acute and chronic animals, have
been extensively investigated.
No standardizable preparation with coronary
artery constriction or occlusion similar to that of the
human has been worked out for the experimental
animal. Naturally occurring or experimentally
induced coronary lesions (dog and rabbit) are similar
in many respects to the human lesions, but the
endothelium remains intact and ulceration and
thrombus formation do not occur. For acute or
chronic experiments, an artery may be tied off
abruptly and completely or partially, by inserting a
probe between the artery and suture (98), or acute
1566
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
(and at times chronic) preparations can be made by
the intracoronary artery injection of lvcopodium
spores (392) or plastic microspheres (2, 34). More
gradual constriction of the lumen, however, may
reduce the incidence of ventricular fibrillation,
minimize infarction, and augment collateral develop-
ment. The introduction of intracoronary clots, the
induction of coronary thrombosis by electrical means
(194, 327), the application to the artery of adjustable
Goldblatt clamps, irritant rings or bands of cellophane
or bakelite, osmotic clamps, or swelling casein rings,
can all ultimately lead to complete coronary artery
occlusion (123, 362, 375). Unfortunately, by none of
these methods can the time of complete occlusion be
known in vivo, nor could the per cent reduction in
flow be predicted even if the extent of local reduction
in vessel lumen were known. As in other vessels, the
effectiveness of a given localized constriction in
reducing flow may be large or small and will vary in
inverse relation to the peripheral resistance of the
vascular bed and lumen of the constricted segment,
and in direct relation to the flow velocity, blood
viscosity, and axial length of the constricted area
(153)-
The hearts of persons afflicted with the clinical
signs and symptoms of coronary artery disease, or of
animals in which coronary insufficiency has been
experimentally induced, generally present a dual
problem. The area of the heart with a normally
functioning coronary arterial system carries much,
if not most, of the burden of metabolism and work of
the poorly nourished myocardium, in addition to its
own. If the handicapped area of myocardium is
large, then the normal portion of the myocardium is
heavily loaded and stressed in its efforts to carry the
total performance of the heart. In the remaining area
of the myocardium, i.e., that handicapped by
sclerosed vessels, or vessels not carrying a normal
supply of oxygen to the myocardium, the supply of
blood and oxygen is too small.
Natural Responses oj the Normal but Overstressed
Portion of the Myocardium
If the ligation or constriction of a coronary artery
is severe enough, useful function is lost within 1 min
in the myocardium fed by it, since the muscle mass
which was shortening during systole now bulges and
lengthens (359). The fact that the area lengthens
rather than shortens during systole does not mean
that the area is not viable, but rather that although
attempting to shorten, the force it exerts is so weak as
to be overbalanced by the intraventricular pressure
which distends it. Since, as will be discussed later,
the collateral flow does not increase for some hours,
any early natural cardiac compensation must occur,
not by improvement of the circulation in the affected
area, but through enhanced action of the normal
myocardium which is not involved. Loss of contract-
ing blocks of muscle following coronary artery
occlusion not only reduces the total myocardial force
available for raising intraventricular tension, but
some of this pressure is spent in stretching the ischemia
area and thus is lost for expelling blood into the aorta.
The immediate consequences of this, producing a
hypodynamic ventricle, are a reduction in left
ventricular systolic pressure, aortic pulse pressure,
systolic and diastolic pressures, duration of systole,
and, especially, stroke volume and stroke work. In
this situation, left coronary inflow decreases con-
siderably (120, 385). However, within a few minutes,
the normal portion of the heart may put into opera-
tion compensatory mechanisms by means of which
dynamic conditions are largely restored to normal,
provided the normal myocardium is in a good
responsive condition. In this situation of increased
cardiac work per unit of functioning myocardium,
coronary flow, arteriovenous oxygen difference, and
metabolism of the left ventricle increase. The increase
in oxygen uptake is equal to, and at times can be
much more than, that lost by the deletion of non-
contractile muscle.
However, not all hearts react as well because the
viable portion of the myocardium may not initially
respond to stretch, or the same lack of response may
occur later after an initial salutary response. This has
been especially studied in dog hearts in which inter-
ference with the coronary circulation has been by
coronary ligature, or by intracoronary injection of
plastic microspheres (2, 33). This leads to acute or
progressive heart failure associated with profound
hypotension, decreased cardiac output and stroke
volume, and the clinical signs and symptoms of a
shock-like state similar to that which occurs following
the loss of blood or plasma. The clinical inference
that this is due to supervention of local coronary
spasm or peripheral circulatory failure has not
received experimental support. Most evidence indi-
cates that no primary insufficiency of tfie resistance or
capacity vessels exists, nor even any noxious reflex to
which the cause of shock could be attributed, nor does
such shock arise, apparently, from reflex coronary
constriction in the nonoccluded coronary artery (57,
153, 233, 254). The experiments of Kuhn et at. (219)
BLOOD SUPPLY TO THE HEART
1567
could, however, be interpreted differently. There are
many reasons to favor the view that in this situation,
circulatory failure not due to severe irregularity of
the heart beat is due successively to: a) defection of
useful contractions in the ischemic area, b) a loss of
contractile energy through expansion of the affected
area and d) failure of the still viable fractions to
compensate adequately.
Since protracted hypotension can, at times, lead to
myocardial damage and failure, and since experi-
mentally the coronary collateral flow varies passively
with the systemic blood pressure, attempts have been
made to improve such hearts experimentally and
clinically by drugs and a venoarterial perfusion.
The state of the heavily stressed normal myo-
cardium could be improved with drugs either by
increasing its oxygen supply or by using the available
oxygen more economically. The major mechanism
for increasing the oxygen supply is by increased
coronary flow since, normally, the oxygen is largely
extracted from blood passing through the myo-
cardium. The drugs would have to promote coronary
flow in the heavily loaded normal myocardium in
which oxygen usage, coronary flow, and coronary
A-V oxygen difference are already at a high level.
Whether any drug has the desired type of dilatation
(active myocardial vessel relaxation, decreased extra-
vascular compression, minimal increase in metab-
olism and cardiac work, minimal effect on other
vascular beds), and whether it also increases ventricu-
lar efficiency remains to be determined. In the normal
dog, drugs such as papaverine, nitroglycerin, epineph-
rine, aminophylline, Coramine, and khellin augment
the myocardial coronary flow and oxygen supply, but
generally at a considerable expense to the heart
through decreased coronary sinus oxygen (with
nitroglycerin coronary sinus oxygen is increased),
and increased cardiac work and metabolism. In
normal man, sublingual nitroglycerin leads to an
increased myocardial oxygen usage (increased
coronary flow and constant coronary A-V oxygen
difference), with little or no change in cardiac output
and cardiac work, and with a decreased efficiency
(42). In patients at rest, with coronary artery disease,
coronary flow is normal. Following nitroglycerin,
coronary flow and oxygen usage are unchanged but
systemic blood pressure, cardiac work, and cardiac
output are reduced; hence, coronary resistance is not
changed while efficiency is decreased (138). It would
thus seem that the dilator capacity of the coronary-
tree with coronary artery disease is exhausted. The
mechanism whereby nitroglycerin relieves pain is not
that of general coronary dilatation and is unknown.
The incidence of cardiogenic shock complicating
acute myocardial infarction has been reported as 1 2
per cent, and mortality associated with this com-
plication may be in excess of 80 per cent (3, 121).
Vasopressor drugs have been widely employed in
this situation (see table 1 for details). The improve-
ment that occurs in the human heart with drugs such
as neosynephrine and norepinephrine, in the presence
of coronary insufficiency and infarction, arises
because of a good dynamic response in the normal but
overstretched myocardium. This presumably aug-
ments the coronary collateral flow by increasing the
coronary perfusion pressure and by making the heart
smaller (see section on coronary collaterals).
Although the use of vasopressor agents may
reduce mortality in myocardial infarction with
shock, at least half fail to respond. In such patients,
extracorporeal circulatory support is being tried
whereby blood is pumped from a convenient vein to
an artery (14, 357). The major objective is to produce
a sustained increase in aortic pressure and, hence, an
increase in coronary, cerebral, and other important
regional circulations, and yet, without an increase in
left ventricular work that might cause further cardiac
deterioration. Conclusive evidence of the benefit of
this procedure has not yet been obtained. In dogs
subjected to coronary embolization, use of a closed-
chest extracorporeal circulation with blood transfer
from the veins to the abdominal aorta has been
effective in restoring central aortic pressure only if
the aorta is occluded beyond the pump (219).
Coronary Artery Collateral Circulation
PREPARATIONS AND METHODOLOGIES FOR COLLATERAL
flow in animal and man. Most studies have been
prophylactic in nature, i.e., a potential stimulus has
been applied to the normal coronary circulation
without interruption of coronary flow to determine
whether, following subsequent coronary artery ob-
struction, the coronary collateral flow will be in-
creased. In only a few instances has the effect on
collateral flow of different variables been studied some
time after creation of coronary insufficiency. It is
unfortunate that a standardized preparation of
coronary insufficiency has not been generally em-
ployed since this is the situation existing in man with
coronary artery disease.
The experimental tools for study of the collateral
circulation leave much to be desired. In the experi-
i568
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
mental animal, these are concerned with measure-
ments of the effects of various prophylactic and,
occasionally, postcoronary occlusion procedures on
the electrocardiogram, mortality, size of infarcts,
exercise tolerance, the coronary artery pressure be-
yond a region of coronary artery occlusion (the
so-called peripheral coronary pressure), and finally,
on the injectable and functional collaterals in the
presence of coronary insufficiency or occlusion. All
are difficult to evaluate because of the considerable
variability in the size of the naturally occurring
collateral circulation. The latter difficulty can be
significantly reduced but not eliminated by using
only animals showing large T-wave inversion and
S-T segment depression during temporary coronary
artery ligation. Experimental indications are that the
size of the injectable collateral bed and the level of
the peripheral coronary pressure correlate well with
direct collateral flow measurements (95). The latter
measurement has been widely used and has given
considerable information (9, 15:3). The collateral
flow (retrograde or backflow) is determined by
collecting the volume of blood flowing externally
from a tube inserted into the peripheral end of a
centrally occluded coronary artery. This is flow-
before it has passed through a capillary bed, i.e., it is
fully oxygenated, and is presumably somewhat too
large because, in the measurement, it drains against
atmospheric pressure whereas, functionally, the
collateral blood must flow against the peripheral
coronary resistance beyond the occlusion. Collateral
flow can also be measured under selected circum-
stances as it enters the myocardium, or after it has
passed through a capillary bed and appears in the
coronary sinus. This can be done when extracardiac
tissue with a vascular stalk has been previously
applied to the heart to stimulate collateral develop-
ment. The collateral inflow can be measured acutely
in the open-chest dog by interposing a rotameter in
the vascular stalk, or chronically by applying an
electromagnetic flowmeter to the extracardiac arterial
pedicle. The collateral contribution to the coronary
sinus is estimated by measuring the decrease in sinus
flow after clamping the potential extracardiac source
of collateral flow.
Recent investigations indicate caution in the use of
the directly measured collateral flow. 0) Rb86
clearance studies estimate collateral flow as two to
three times the directly measured backflow, thus
suggesting that in addition to functioning inter-
arterial channels, other vessels communicate with the
ischemic zone at the arteriolar and capillary levels
(235, 247). This method, however, cannot be used for
estimating changes in collateral flow because of the
unknown and variable extraction ratio of this sub-
stance in the ischemic area, b) The small portion
(possibly 15%) of left coronary artery inflow not
recoverable in the coronary sinus or anterior cardiac
veins has been largely accounted for by drainage of
the septal artery and some branches of the left
descendens into the right ventricular cavity (265).
Thus, in the presence of coronary artery occlusion,
some blood might perfuse portions of the septum
retrogradely during systole when the pressure gradient
might be favorable.
Tests of coronary collateral function in life in the
normal and diseased heart of man have been largely
restricted to monitoring changes in the electro-
cardiogram to exercise tolerance and angina and,
after death, to injection of the coronary collateral
circulation at autopsy with opaque viscous material
(338). In those individuals with an occluded coronary
artery ramus and undergoing a coronary operation,
it would appear feasible to use as an index of collateral
flow the coronary pressure beyond the occlusion,
which can be measured by simple needle insertion.
This technique used so successfully in animals has not
been attempted in man. Finally, coronary angio-
graphic studies by Sones (352), and others, have
demonstrated collaterals in both normal and ab-
normal hearts. Whether this technique has a future
in the study of the development and regression of
collaterals and atheromatous lesions remains to be
seen (184).
From the preceding it can be seen that, because of
our poor methodology, and especially because the
direct or indirect measurement of collateral flow has
not as yet been made in man, objective evidence of
positive benefit to the heart cannot come primarily
from observations after experimental or surgical
maneuvers or coronary surgery in man, but must
come from the effect of various procedures on coronary
collateral function in other animals.
NATURAL RESPONSES OF THE CORONARY COLLATERAL
circulation. The natural responses of the coronary
circulation of animals during experimental coronary
artery constriction and occlusion, which, presumabh ,
also happen in the heart of man, have been studied
extensively.
Considerable reduction in the lumen of a coronary
artery can occur with minimal or no permanent
change in coronary flow. This is so because the
coronary resistance to flow measured beyond a point
BLOOD SUPPLY TO THE HEART
1569
of occlusion (by ligation) of a left coronary artery
branch is considerable, being about 30 20 mm Hg,
and the central coronary resistance is quite low (153).
The effect of a central constriction on coronary flow
is a function of how much the resistance imposed by-
it is in relation to the resistance in the coronary bed.
When the flow to the bed is reduced by central
constriction, the peripheral vessels dilate as a result
of the associated ischemia and the flow may tend to
increase, the combined result of which will be a new
equilibrium. Hence, since the peripheral resistance in
the coronary bed is constantly changing and will be
decreased by the anoxia induced by the central
constriction, and since the effect on flow of any central
constriction of lumen is a function of how much that
resistance is, in relation to peripheral resistance, no
predictions can be made as to the effect on blood
flow when the coronary artery is constricted by known
amounts. Since the peripheral resistance is relatively
high, generally sizeable reductions in lumen are
necessary before inflow decreases. Thus, the reduction
of lumen of a coronary vessel may be of little func-
tional importance to the vascular bed supplied by
that vessel when the rate of flow is normal or some-
what low, but the same constriction can seriously
limit flow to the same bed just at the time when the
requirements of the latter are greatest and flow would
otherwise be much greater (153). Obviously, how-
ever, this compensatory dilatation of the coronary bed
in the presence of constriction of its central coronary
artery has a limit, and flow through it will ultimately
fall significantly. In part because of this, the heart
has a remarkable ability to retain viability of its
muscle beyond a constriction, and significant changes
in the electrocardiogram do not occur until coronary
inflow is reduced approximately 70 per cent (383).
Studies have been made of how quickly such an
ischemic area with its potential collateral supply of
oxygen becomes nonviable. Admittedly, tests for
viability are crude. However, if the criteria used are
an absence of local action currents, failure of local
conduction, and lack of movement in the presence of
generalized ventricular fibrillation, then viability
does not usually continue beyond an hour, although
occasionally the presence of local action currents and
excitability may persist from 2l 2 to 7 hours (403).
The return of normal myocardial function has been
studied also after reinstitution of coronary flow in dog
hearts maintained anoxic for prolonged periods on an
extracorporeal circulation. Hearts maintained anoxic
for up to 100 min can maintain their blood pressure
on removal from the extracorporeal circulation (65).
Within 1 min after occlusion of a left coronary
artery branch, the intracoronary pressure beyond this
point drops to about 30 20 mm Hg and useful
function is lost, for the muscle now lengthens during
systole of the left ventricle (359). When, however,
the peripheral end of this ligated coronary artery is
permitted to bleed externally, collateral arterial
blood appears immediately, averaging about 3.0 ml
per min for about 50 g of potentially infarcted myo-
cardium, and this blood can be shown to come from
the other nonoccluded coronary arteries (153). The
collateral communications are largely in the epicardial
areas (40). Probably not more than 2.4 ml of this
blood (containing 0.5 ml oxygen) would perfuse the
myocardial bed if the collateral flow were not per-
mitted to bleed externally. This is because of the
peripheral resistance existing beyond the point of
occlusion and averaging 20 or so mm Hg. That most
of this calculated collateral flow actually traverses
the capillary bed is evidenced by the fact that the
electrocardiogram improves when the collateral flow
is not permitted to bleed externally (96, g8).
Most of these hearts with occlusion of a major left
coronary artery branch die within a number of
hours. For example, experimental ligation of the left
circumflex coronary artery may give mortalities of
70 per cent or more (170). Other hearts are more
fortunate, for if they survive the first few hours, then,
for some completely unknown reason, within 12
hours collateral flow starts to rise, doubling within 2
days, and within 3 to 4 weeks it may approximate 40
to 100 per cent of normal inflow into that coronary
artery. Almost all the collateral flow comes from the
unoccluded coronary arteries. The myocardial fibers
which were lengthening early after occlusion now
shorten in systole. Concurrently, the peripheral
coronary pressure increases to values somewhat less
than the normal central coronary pressure and the
myocardium shortens during systole (153).
MEANS OF EXPERIMENTALLY CHANGING COLLATERAL
FLOW EARLY AFTER CORONARY OCCLUSION. The level
of collateral flow with its oxygen content is estimated
to be about 40 per cent of that calculated as necessary
to maintain indefinitely the viability of this myo-
cardium, since the oxygen uptake of 50 g of a heart
with perfused coronary arteries at rest and doing no
external work approximates 1.2 ml, as compared to
the immediately available collateral oxygen supply of
0.5 ml (249). Hence, it is important to try to increase
immediately this collateral flow or backflow. Except
for one report on the positive effect of nitroglycerin
'57o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
(229), this level of backflow has not been made to
increase for 8 to 1 o hours by drugs or by any known
physiological means, such as increased heart rate,
increased flow in the other coronary arteries, in-
duction of hypoxia or hypoxemia in the other coronary
arteries (204). Why the collateral flow remains fixed,
why the anastomoses function as a set of inert tubes,
and why they do not exhibit vasomotion or participate
in the vasodilatory response of the normal coronary
bed are not known. This situation contrasts with the
rapid development of collaterals in other vascular
beds such as the femoral and carotid arteries (153).
This retrograde flow can, however, be greatly
reduced by excessive stretch of the myocardium and
reactive hyperemia in the other nonoccluded coro-
nary artery branches. The improvement that occurs
with drugs such as neosynephrine and norepinephrine
in the human heart in the presence of coronary
insufficiency and infarction could result from an
increase in the oxygen supply to regionally ischemic
muscle (337), and from augmentation of the col-
lateral flow through increase of the coronary per-
fusion pressure and a smaller heart size. Spasm of the
coronary arteries with diminished blood flow is also
frequently invoked to explain the onset of episodes of
angina pectoris and of reduction in collateral flow.
However, no firm conclusion can yet be drawn as to
whether flow in one coronary artery can be in-
fluenced reflexly and adversely by impulses arising
from an intra- or extracardiac source (see the section
on Reflexes).
Lysis of coronary thrombi induced experimentally
can be observed to follow fibrinolytic therapy.
Whether this will change the evolution of early
myocardial infarction and result in salvage of ischemic
tissue without collateral development has not yet
been determined (276).
collateral stresses to which the hearts have been
previously exposed.
Evidence, largely from the classical work of Zoll
et al. (411), indicates that the incidence of injectable
coronary artery collaterals is quite small in normal
human hearts, but is greatly increased in the presence
of coronary artery constriction or occlusion. There is
also evidence in different species that nature adopts
prophylactic measures to protect some hearts against
subsequent coronary artery occlusion. The coronary
vessels appear to be capable of setting up or enlarging
anastomoses between themselves without the stimulus
of coronary occlusion or insufficiency. Presumably,
this is due to some form of antecedent stress. In these
hearts, stresses, some known but mostly unknown,
prophylactically enhance the potential collateral
circulation without the stimulus of coronary occlusion
or constriction. These are exemplified in man by the
increase in the incidence of the injectable coronary
arterial collateral bed in the presence of hypertrophy,
valvular disease, cor pulmonale, anemia, and prob-
ably high altitude (38, 412; also Rotta, personal
communication). This is exemplified in the pig by an
increase in the injectable collaterals in the presence of
anemia (39, 412), and in the dog by an increase in
both the injectable and functional collaterals in the
presence of high altitude (Rotta, unpublished ob-
servations), and transfused anemia (97). No good
experimental evidence exists, however, to indicate
that physical exercise per se augments prophylacti-
cally the collateral flow as measured in a normal
coronary artery immediately after its occlusion.
The injectable coronary collateral bed, however, is
stated to increase in exercised rats (360). Individuals
who escape serious consequences from coronary
occlusion may well be those whose collaterals have
been previously expanded by such means.
RESPONSE OF THE CORONARY COLLATERAL CIRCULATION
TO NATURALLY OCCURRING PROPHYLACTIC STIMULI.
As indicated earlier, the intercoronary arterial com-
munications are generally small in normal man,
fewer than 10 per cent having anastomoses with
diameters of 40 p. or more (39). However, others using
corrosion and injection techniques found anastomoses
of greater size and with greater frequency (18, 226,
371). The coronary arterial tree of the pig is strikingly
similar to that of man (96, 289), while in the dog the
anastomoses are larger and more frequent. These
differences in collateral function might be explained
on a technical basis or as fundamental species varia-
tions; however, they could be related to the types of
MEDICAL, PHYSIOLOGICAL, AND SURGICAL ATTEMPTS TO
IMPROVE THE CORONARY ARTERY COLLATERAL CIRCU-
LATION prophylactically. Either before or after
establishment of coronary insufficiency, it should be
possible to improve the state of the heart of dog or
man by augmentation of the coronary artery col-
lateral circulation which naturally functions, by
retrograde perfusion of the ischemic coronary bed
with arterial blood, or by elevation of the ventricular
fibrillation threshold. In man, in addition, positive
and subjective benefit could arise through psychogenic
effects which are not necessarily related to the heart.
Chronic experiments have produced no good
evidence to indicate that any drug promotes collateral
BLOOD SUPPLY TO THE HEART
'57'
flow or reduces the size of infarcts produced by subse-
quent coronary artery ligation (380, 404). The alleged
favorable effect on survival of the use of drugs such as
papaverine or quinidine is better explained by their
known action in raising the fibrillation threshold and
in reducing myocardial excitability (384).
The capable experimental coronary surgeon has
been able to improve considerably on the state of
such hearts. Much of the advancement in the surgical
and physiological fields has arisen from the pioneer
investigations and stimulus of Beck (19, 50, 255).
The procedures used include section of the cardiac
sympathetic nerves (178), induction of myocardial
hypoxia by various manipulations of the coronary
venous system or by a coronary fistula (19, 79, 95,
158), production of mechanical and chemical
pericarditis between the epicardium and pericardium
to use the extracardiac anastomoses (19, 178),
application of extracardiac tissue to the heart (271,
324, 372, 373), internal mammary artery ligation
(99), sham operations (1, 19, 85), coronary endarterec-
tomy (241, 326), and coronary artery bypass (171 ).
Many of these procedures in the experimental
animal are of positive benefit to the heart and give
immediate or sustained protection against subsequent
ligation of a major coronary artery ramus. Ligation
of a major ramus of the left coronary artery causes
about a 70 to 90 per cent mortality within the first
1 to 2 hours, and chronically there is considerable
infarction (95). When partial or complete occlusion of
the coronary sinus precedes coronary artery ligation,
or when a portion of the coronary bed is perfused in
retrograde fashion by connecting the coronary sinus
to an artery, the immediate mortality is reduced con-
siderably. With the exception of section of cardiac
sympathetic fibers and internal mammary artery
ligation, most other procedures — chronic coronary
venous maneuvers, application of various chemical
and mechanical irritants, separately or in combina-
tion, and application of extracardiac tissue to the
heart, generally lead to a significant reduction in
mortality and infarction (there are, however,
exceptions) (124). There is an increase in the in-
jectable and functional collaterals with the chronic
coronary venous maneuvers and with the application
of mechanical and chemical irritants to the heart.
The level of collateral flow, 5 to 12 ml in most in-
stances, considerably exceeds the control retrograde
flow of 3 ml with acute artery ligation alone. Accord-
ingly, it is deduced that these surgical maneuvers
give sustained, and in the case of the coronary venous
maneuvers, immediate protection against ligation of a
major coronary artery branch. The retrograde flow in
the chronic experiments equals or exceeds that
estimated to be necessary to maintain viability.
Cardiac benefit from these procedures could arise
from retrograde flow of blood from the superficial
veins through the capillary bed, from development of
intra- and extracardiac collaterals, or from elevation
of the ventricular fibrillation threshold, thus giving
nature time to develop additional collaterals to
sustain the heart. There are no critical experiments
to prove that with the acute coronary venous maneu-
vers, protection against fibrillation and death is
supplied by blood flowing in a retrograde direction
from coronary vein to capillary to ventricular cavity.
Acute perfusion of the coronary sinus with arterial
blood at or near aortic blood pressure, or acute
ligation of the coronary sinus, results in venous con-
gestion of the left heart with an increased coronary
venous pressure, at times equal to the aortic pressure,
a diffuse myocardial hemorrhage (with the exception
of the septum which remains pink in color), and a
sizeable reduction in left coronary inflow and cardiac
output. When the peripheral portion of the occluded
coronary artery is permitted to bleed externally, the
measured backflow is of highly reduced blood and the
volume is increased greatly (to 15 ml or more) over
that which occurs with acute coronary artery ligation
alone (153). It is very important to know that this
blood can be shown to have traversed the capillary
bed of the occluded coronary artery in a reverse
direction. However, proof is lacking that, when the
ligated coronary artery is not permitted to bleed
externally, flow from the superficial coronary veins is
diverted through the capillary bed of the left myo-
cardium and then into the left ventricular cavity.
Actually, the development of extreme myocardial
embarrassment, together with the fact that most of
left coronary artery inflow and the blood entering the
coronary sinus from the shunt can now be recovered
in the anterior cardiac veins of the right ventricle
(153), offers not quite certain evidence that the deep
ventricular drainage channels are not used. However,
the high values for venous pressure in the coronary
sinus and the augmentation of peripheral vascular
pressure and retrograde flow which appear in the
left coronary artery immediately after left coronary
venous ligation decrease after a time interval (up to
30 days) to values only slightly above normal (153).
The observation that these procedures can elevate
the ventricular fibrillation threshold suggests but
does not prove that this is a major mechanism of
protection. In hearts with chronic application of these
1572
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
various latter maneuvers, protection in large part,
and in many instances, is probably afforded by the
augmented collateral circulation. For example, with
an aorta-coronary sinus shunt, the backflow of i o to
12 ml of arterial blood exceeds that calculated to be
necessary for viability and persists for at least a year
and even after loss of function of the shunt (95). But
since most hearts following coronary artery ligation
die within 24 hours, since the usual retrograde flow
observed with these procedures is not large, and since
sham operation involving manipulation of the heart
at times increases collateral flow or gives sustained
protection against coronary occlusion, or both, the
possibility must be entertained that there may be no
specific effect of some of the maneuvers; they may
act by raising the ventricular fibrillation threshold
thus giving time for collaterals to develop.
In some procedures that apply extracardiac tissue
to the heart, such as a pedical skin flap (271), or an
internal mammary artery ligation (99) or its myo-
cardial implantation (324), the collateral flow does
not increase. These studies, however, are incomplete.
Further work should be done to determine, in
addition to the usual arterial collateral flow measure-
ments, whether blood actually flows from the extra-
cardiac tissue through the capillary bed of the
myocardium into the coronary sinus or other coronary
venous outflow channels. Despite some positive find-
ings, no firm conclusion can be drawn (372, 373).
ATTEMPTS TO IMPROVE THE COLLATERAL CIRCULATION
AFTER CORONARY ARTERY OBSTRUCTION IN ANIMALS
and man. As already indicated, immediate or early
augmentation of the coronary collateral circulation,
beyond that occurring naturally following marked
coronary constriction or occlusion, has not been
demonstrated. Neither experimental estimation of a
favorable delayed or chronic collateral response
(decrease in infarct size and increased injectable
collateral bed or collateral flow) to drugs has been
demonstrated (379, 412). However, the following
evidence of positive benefit has been reported : a)
Treadmill exercise, when added to a pre-existing
coronary insufficiency, appears to increase the
collateral flow to a level greater than with coronary
constriction alone (98). b) In the presence of aneroid-
induced chronic left coronary insufficiency and
epicardectomy, the addition of a mammary artery
implant or application of an Ivalon sponge is stated to
greatly extend the survival time of the dog and to
increase the functional communications of the
ischemic bed with the left ventricular cavity and
extracardiac arteries. This benefit does not follow the
use of cardiopneumopexy or the applications of
various other irritants to the myocardium (374). c)
Experimental attempts have been made to improve
the blood supply to the normal heart and the heart
with infarction (intracoronary injection of plastic
microspheres) by altering the time of arrival of the
arterial pressure pulse so that the systolic pulse
arrives during diastole, the period of greater flow (59,
188, 203). The procedure is reported to greatly
reduce the mortality rate from the myocardial in-
farction and to increase the injectable collateral bed.
d) When a pulmonary artery to left atrial shunt is
added to an already existing chronic occlusion of the
left circumflex branch, coronary angiograms and
vinyl acetate casts show a more rapid collateral
filling and a greater vascularity, respectively, than
following coronary artery occlusion alone (30).
Most of the procedures designed to promote
collateral development, including the sham opera-
tion, have been applied to the heart of man suffering
from coronary artery disease. All appear to increase
to some extent the work and exercise tolerance and to
decrease cardiac pain (19, 178, 374). The summary of
over 600 patients on whom the Beck operation was
performed may serve as an example (51). These
observations are not necessarily explained on the
same basis of the improvement in the collateral
circulation of the dog which follows such procedures.
This is because in the dog most surgery precedes
coronary artery ligation and is designed to promote
collaterals in the presence of a normal coronary
circulation, whereas, in the human, surgery follows
coronary artery occlusion and is designed to promote
collateral circulation after the coronary insufficiency
has been naturally established. In man, hypoxia, the
greatest known vessel dilator, and a natural stimulus
to collateral development, has already been working
for many months. Since human coronary surgery
which follows coronary occlusion has as yet little
counterpart in animal experiments, attempts should
not be made to interpret these human coronary
experiments on a physiological basis.
The explanation of the results in man is not clear.
Patients treated surgically by epicardial phenoliza-
tion, poudrage, cardiopneumopexy, and bilateral
internal mammary artery ligation, although showing
marked relief of angina, do not show electrocardio-
graphic improvement or an increase in coronary flow,
or a decrease in coronary vascular resistance following
nitroglycerin (44). Undoubtedly, some subjects are
protected and live longer because of the known experi-
BLOOD SUPPLY TO THE HEART
1573
mental fact that handling the heart raises the ventricu-
lar fibrillation threshold. Some may be improved by
procedures such as de-epicardialization which could
obliterate the afferent pathways for pain. However,
results of the sham operation of Adams (1) and
Dimond (85), involving only a skin incision, strongly
suggest that much of the positive benefit is on a
psychogenic basis.
Coronary endarterectomy which has been applied
to man is on a sound physiological basis and its
purpose is entirely different from the preceding. The
surgeon directly reestablishes coronary flow through
the original coronary artery by removing its athero-
sclerotic plug. It does not require collateral develop-
ment and should be effective provided there exists a
gross coronary insufficiency of blood beyond the
obstruction, provided the vessel remains patent and
thrombi do not form, and provided there are no
sizeable atherosclerotic lesions beyond the region of
the occluded coronary artery. It is quite doubtful
that these criteria can be met (182). Preliminary
experiments with the use of endarterectomy for
coronary occlusion were apparently initially favorable
to the patients, relieving their angina, and improving
their electrocardiograms and work tolerance (241,
326). However, most of these patients have died, and
no evidence is available that at autopsy the endarter-
ectomized artery has remained patent. Many more
operations will have to be performed to establish the
possible merit of this procedure in humans.
Finally, bypass of a length of an occluded coronary-
artery by anastomosis of its peripheral patent end to a
systemic artery has not yet been attempted in man.
In dogs, a nonsuture anastomosis by intima-to-intima
contact between the left coronary artery and the left
internal mammary artery has been highly successful
(171). In almost all the dogs (24 of 33), the anasto-
moses have been demonstrated to be patent and
without myocardial infarction as evidenced by gross
observation, angiography, and measurement of
coronary blood flow through the anastomosis up to
the time of dog sacrifice (12-24 months after opera-
tion). Other technical achievements in this area
include chronic anastomoses of two branches of the
left subclavian artery to the peripheral and central
ends, respectively, of the left circumflex coronary, the
central end of the main left coronary being tied
(unpublished observations), and end-to-end anasto-
mosis of the central end of the main left coronary
artery to the peripheral end of the left subclavian
artery (251). Since anastomosis of a coronary artery
branch to a systemic artery is almost always successful
in the dog in which the anastomosed vessels are only
2 to 3 mm diameter, there should be no difficulty at
all in the human heart in which the coronary artery
branches have a much greater diameter. This proce-
dure might, therefore, have an application in the
creation of a permanent new blood supply in the
presence of coronary artery disease in man. One
should not, however, overlook a probably late com-
plication to successful coronary endarterectomy or
coronary bypass in man. In the presence of such a
large new blood supply, the existing collateral flow
will disappear. If another coronary occlusion subse-
quently occurs, the patient will be in difficulty,
having lost his collaterals.
REFERENCES
1. Adams, R. Internal-mammary-artery ligation for coronary
insufficiency. An evaluation. New Engl. J. Med. 258: 1 13,
1958-
2. Agress, C. M., H. F. Glassner, M. J. Binder, and J.
Fields. Hemodynamic measurements in experimental
coronary shock. J. Appl. Physiol. 10: 469, 1957.
3. Agress, C. M. Management of coronary shock. Am. J.
Cardiol. 1 : 231, 1958.
4. Alexander, R. W., and G. C. Griffith. Anomalies of
the coronary arteries and their clinical significance. Circu-
lation 14: 800, 1956.
5. Allen, J. B., and J. R. Laadt. The effect of the level of
the ligature on mortality following ligation of the circum-
flex coronary artery in the dog. Am. Heart J. 39: 273,
'95°-
6. Altman, P. L. Handbook of Circulation. Natl. Acad. Sci-
Natl. Research Council. Philadelphia: Saunders, 1959.
7. American Heart Association. Symposium on the Coronary
Circulation, Chicago, 1962. Submitted for publication.
8. Arey, L. B. Developmental Anatomy (5th ed.). Philadelphia:
Saunders, 1950.
9. Anrep, G. V., A. Blalock, and M. Hammouda. The
distribution of blood in the coronary blood vessels. J.
Physiol., London 67: 87, 1929.
10. Anrep, G. V. Studies in cardiovascular regulation. Lane
Medical Lectures. Med. Sci. 3: 199, 1936.
11. Anzola, J., and R. F. Rushmer. Cardiac responses to
sympathetic stimulation. Circulation Research 4: 302, 1956.
12. Aviado, D., R. G. Pontius, and C. F. Schmidt. The
reflex respiratory and circulatory actions of veratridine
on pulmonary, cardiac and carotid receptors. J. Phar-
macol. Exptl. Therap. 97: 420, 1949.
13. Aviado, D. M. Cardiovascular effects of some commonly
used pressor amines. Anesthesiology 20: 71, 1959.
'574
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
14. Bacaner, M., J. E. Connolly, and D. Bruns. The cor-
onary blood flow as a critical determinant of cardiac
performance and cardiac size. Am. J '. Med. 311 392, 1961.
15. Badeer, H., and A. Khachadurian. Role of bradycardia
and cold per se in increasing mechanical efficiency of
hypothermic heart. Am. J. Physiol. 192: 331, 1958.
16. Ballard, F. B., W. H. Danforth, S. Naegle, and R. J.
Bing. Myocardial metabolism of fatty acids. J. Clin.
Invest. 39: 717, i960.
17 Barcroft, ) , and W. E. Dixon. The gaseous metabolism
of the mammalian heart. Part I. J. Physiol., London 35:
182, 1906-7.
18. Baroldi, G., O. Mantero, and G. Scomazzoni. The
collaterals of the coronary arteries in normal and patho-
logic hearts. Circulation Research 4: 223, 1956.
ig. Beck, C. S. Symposium on Coronary Artery Disease:
Blood supply to ischaemic myocardium distal to the oc-
clusion of a coronary artery. Diseases of Chest 31 : 243,
'957-
20. Bellet, S., J. W. West, U. C. Manzoli, O. F. Muller,
and P. Rossi. Effect of nicotine on the coronary blood
flow in the presence of coronary insufficiency: an experi-
mental study in dogs. Ann. A'.!'. Acad. Sci. 90: 317, i960.
21. Bercu, B. A., W. H. Danforth, E. E. Pund, Jr., and
G. A. Diettert. Radioactive sodium for the measurement
of myocardial blood flow. J. Clin. Invest. 37: 877, 1958.
22. Berglund, E., R. G. Monroe, and G. L. Schreiner.
Myocardial oxygen consumption and coronary blood
flow during potassium induced cardiac arrest and during
ventricular fibrillation. Acta Physiol. Scand. 41 : 261, 1957.
23. Berglund E., H. G. Borst, F. Duff, and G. L.
Schreiner. Effect of heart rate on cardiac work, myo-
cardial oxygen consumption and coronary blood flow in
the dog. Acta Physiol. Scand. 42: 185, 1958.
24. Berne, R. M. Effect of dermal contact with cold on the
coronary circulation. Proc. Sol. E.xpll. Biol. Med. 84: 150,
1953-
25. Berne, R. M. The effect of immersion hypothermia on
coronary blood flow. Circulation Research 2: 236, 1954.
26. Berne, R. M., J. R. Blackmon, and T. H. Gardner.
Hypoxemia and coronary blood flow. J. Clin. Invest. 36 :
1 101, 1957.
27. Berne, R. M. The effect of epinephrine and norepineph-
rine on the coronary circulation. Circulation Research 6:
644. >958-
28. Berne, R. M. Release of adenine nucleotide derivatives
from the hypoxic heart: possible role in regulation of
coronary blood flow. Am. J. Physiol. 204: 317, 1963.
29. Beurens, A., R. J. Bing, and C. Sparks. Metabolic
studies on the arrested and fibrillating perfused heart.
Am. .1. Cardiol. 1 : 103, 1958.
30. Bilgutav, A. M , L. H. Sanchez, D. L. Siegal, and
C. W. Lillehei. Effect of pulmonary artery-left atrium
shunts on ischemic hearts — experimental study and clinical
application. Stir!;. Forum 12: 229, 1961.
31. Bing, R. J. The coronary circulation in health and disease
as studied by coronary sinus catheterization. Bull. A . )'.
Acad. Med. 27: 407, 1951.
32. Bing, R. J. Myocardial metabolism. Circulation 12: 635,
'955-
33. Bing, R. J. The metabolism of the heart. Harvey Lectures
New York: Acad. Press, 1954-1955, p. 27.
34-
35-
3"-
37-
38.
39-
40.
4'-
42-
43-
44-
45-
46.
47
49-
50.
51'
Bing, R. J., A. Castellanos, E. Gradel, A. Siegel, and
G Lupton. Enzymatic, metabolic, circulatory and path-
ological studies in myocardial infarction. Trans. Assoc. Am.
Physicians 69 : 1 70, 1 956.
Bing, R. J., H. K. Hellems, and T. J. Regan. Measure-
ment of coronary blood flow in man. Circulation 22: I,
i960.
Blair, E. Anatomy of the ventricular coronary arteries
in the dog. Circulation Research 9: 333, 1961.
Blumgart, H. L., P. M. Zoll, A. S. Freedberg, and
D. R. Gilligan. The experimental production of inter-
coronary arterial anastomoses and their functional sig-
nificance. Circulation 1:10, 1950.
Blumgart, H. L. Anatomy and functional importance
of intercoronary arterial anastomoses. Circulation 20:812,
'959-
Blumgart, H. L., and P. M. Zoll. Pathologic physiology
of angina pectoris and acute myocardial infarction. Cir-
culation 22: 301, i960.
Bobb, J. R. R., D. C. Kunze, W. McCall, Jr., and
H. D. Green. Location of communications between
cognate bed of descending ramus of left coronary and
adjacent collateral vascular beds. Proc. Soc. Exptl. Biol.
Med. 69 : 115, 1 948.
Bover, N. H., and H. D. Green. Effects of nitrites and
xanthines on coronary inflow and blood pressure in
anesthetized dogs. .4m. Heart J. 2 1 : 199, 1 941.
Brachfeld, N., J. Bozer, and R. Gorlin. Action of
nitroglycerin on the coronary circulation in normal and
mild cardiac subjects. Circulation 19: 697, 1959.
Brachfeld, N., R. G. Monroe, and R. Gorlin. Effect
of pericoronary denervation on coronary hemodynamics.
Am. J. Physiol. 199: 174, i960.
Brachfeld, N., and R. Gorlin. Physiologic evaluation of
angina pectoris. Diseases of Chest 38: 658, i960.
Braunwald, E., G. H. Welch, Jr., and S. J. Sarnoff.
Hemodynamic effects of quantitatively varied experi-
mental mitral regurgitation. Circulation Research 5: 539,
■957-
Braunwald, E., S. J. Sarnoff, R. B. Case, W. N.
Stainsby, and G. H. Welch, Jr. Hemodynamic deter-
minants of coronary flow: effect of changes in aortic
pressure and cardiac output on the relationship between
myocardial oxygen consumption and coronary flow.
Am. J. Physiol. 192: 157, 1958.
Braunwald, E., R. L. Frve, and J. Ross, Jr. Studies on
Starling's Law of the Heart. Circulation Research 8 : 1 254,
i960.
Braunwald, E., R. D. Bloodwell, L. I. Goldberg,
and A. G. Morrow. Studies on Digitalis. IV. Observa-
tions in man on the effects of digitalis preparations on the
contractility of the non-failing heart and on total vascular
resistance. J. Clin. Invest. 40: 52, 1961.
Bretschneider, H. J. Neue Pharmaka zur Behandlung
der Koronarinsuffizienz. Deut. Mid- Wochschr. 86: 1649,
1 961.
Brofman, B L. Symposium on Coronary Artery Disease:
Surgical treatment of coronary artery disease; medical
management and evaluation of results. Diseases of Chest
3' : ^53, '957-
Brofman, B. L. Long term influence of the Beck opera-
BLOOD SUPPLY TO THE HEART
1 )/ i
tion for coronary heart disease. Am. J. Cardiol. 6: 259,
i960.
52. Buckley, N. M., K. K. Tsuboi, and N. J. Zeig. Inotropic 7 1 .
effects of purines and pyrimidines on the isolated heart.
Circulation Research g: 242, 1961.
53. Butterworth, R. F. The venous drainage of the left 72.
atrium. J. Anal. 88: 131, 1954.
54. Garlsten, A., B. Folkow, and C. A. Hamberger.
Cardiovascular effects of direct vagal stimulation in man. 73.
Ada Physiol. Scand. 41 : 68, 1957.
55. Carter, D., and D. C. Sabiston, Jr. Myocardial metabo-
lism during perfusion of the coronary circulation with
gaseous oxygen. Surgery 49: 625, 1961.
56. Case, R. B., S. J. Sarnoff, P. E. Waithe, and L. C. 74.
Sarnoff. A comparison of the effect of intra-arterial and
intravenous blood infusion on coronary blood flow in
hemorrhagic shock. J. Am. Med. Assoc. 152: 208, 1953. 75.
57. Case, R. B., E. Berglund, and S. J. Sarnoff. Ventricu-
lar Function. VII. Changes in coronary resistance and
ventricular function resulting from acutely induced 76.
anemia and the effect thereon of coronary stenosis. Am. J.
Med. 18: 397, 1955.
58. Case, R. B., A. G. Morrow, W. Stainsby, and J. O.
Nestor. Anomalous origin of left coronary artery: The 77.
physiologic defect and suggested surgical treatment.
Circulation 17: 1062, 1958. 78.
59. Casten, G. G., VV. P. Murphy, and J. C. Alley. Aug-
mentation of diastolic arterial pressure by mechanical
means : Effect on coronary sinus flow. Circulation 1 6 : 866, 7g
1957-
60. Chambliss, J. R., J. Deming, K. Wells, W. W. Cline,
and R. W. Eckstein. Effects of hemolyzed blood on
coronary blood flow. Am. J. Physiol. 163: 545, 1950.
61. Chardach, W. M., F. J. Bolgan, K. C. Olson, A. A. 80
Gage, and W. E. Farnsworth. The mortality following
ligation of the anterior descending branch of the left
coronary artery in dogs: An experimental study. Ann. 81
Surg. 141 :443, 1955.
62. Charlier, R. Un nouveau dilatateur coronarien de
synthese. Acta Cardiol. Suppl. 7., 140, 1959. 82
63. Chase, R. E., and C. F. DeGaris. Arteriae coronariae
(cordis) in the higher primates. Am. J. Phys. Anthropol.
24:427> "939-
64. Christensen, G. C, and F. D. Campeti. Anatomic and 83
functional studies of the coronary circulation in the dog
and pig. Am. J. Vet. Research 20: 18, 1959.
65. Coffman, J. D., F. B. Lewis, and D. E. Gregg. Effect
of prolonged periods of anoxia on atrioventricular con- 84
duction and cardiac muscle. Circulation Research 8: 649,
i960. 85
66. Coffman, J. D., and D. E. Gregg. Reactive hyperemia
characteristics of the myocardium. Am. J. Physiol. 199.
1 1 43, i960.
67. Coffman, J. D., and D. E. Gregg. Oxygen metabolism 86
and oxygen debt repayment following myocardial is-
chemia. .4m. J. Physiol. 201 : 881, 1961.
68. Coffman, J. D., and D. E. Gregg. Pharmacology in Blood 87
Vessels and Lymphatics, edited by D. I. Abramson. New
York: Acad. Press, 1962.
69. Cohen, H., and S. Siew. Aberrant left coronary artery. 88
Circulation 20: 918, 1959.
70. Cole, S. L., H. Kaye, and G. C. Griffith. Assay of
anti-anginal agents. I. A curve analysis with multiple
control periods. Circulation 15: 405, 1957.
Cooper, T., J. W. Gilbert, R. D. Bloodwell, and
J. R. Crout. Chronic extrinsic cardiac denervation by
regional neural ablation. Circulation Research 9: 275, 1 96 1.
Corday, E., H. Gold, L. B. deVera, J. H. Williams,
and J. Fields. Effect of the cardiac arrhythmias on the
coronary circulation. Ann. Internal Med. 50: 535, 1 959-
Crumpton, C W., G. G. Rowe, G. O'Brien, and Q. R.
Murphy, Jr. The effect of hexamethonium bromide
upon coronary flow, cardiac work and cardiac efficiency
in normotensive and renal hypertensive dogs. Circulation
Research 2 : 79, 1 954.
Danforth, W. H., F. B. Ballard, K. Kako, J. D.
Choudhury, and R. J. Bing. Metabolism of the heart
in failure. Circulation 21: 112, i960.
Darby, T. D., and E. E. Aldinger. Further studies of
the effects on myocardial energy utilization elicited by
nitroglycerin. Circulation Research 8 : 1 00, 1 960.
Dawes, G. S. Studies on veratrum alkaloids, receptor
areas in coronary arteries and elsewhere as revealed by
use of veratridine. J. Pharmacol. Exptl. I'herap. 89 : 325,
'947-
Dawes, G. S., and J. H. Comroe, Jr. Chemoreflexes from
the heart and lungs. Physiol. Revs. 34: 167, 1954.
Day, S. B., and J. A. Johnson. The distribution of the
coronary arteries of the rabbit. Anal. Record 132: 633,
1958.
Day, S. B., and C. W. Lillehei. Experimental basis for a
new operation for coronary artery disease; a left atrial-
pulmonary artery shunt to encourage the development of
interarterial intercoronary anastomoses. Surgery 45: 487,
■959
Day, S. B., and J. A. Johnson: Pressure-flow relation-
ships in the isolated perfused rabbit heart. Am. J. Physiol.
.96: 1289, 1959.
Denison, A. B., Jr., M. P. Spencer, and H. D. Green.
A square wave electromagnetic flow meter for application
to intact blood vessels. Circulation Research 3 : 39, 1 955.
Denison, A. B., Jr., S. Bardhanabaedya, and H. D.
Green. Adrenergic drugs and blockade on coronary
arterioles and myocardial contraction. Circulation Re-
search 4:653, 1956.
Denison, A. B., Jr., and H. D. Green. Effects of auto-
nomic nerves and their mediators on the coronary cir-
culation and myocardial contraction. Circulation Research
°:°33. 1958.
DiGuglielmo, L., and M. Guttadamo. Anatomic varia-
tions in the coronary arteries. Acta Paediat. 41 : 393, 1954.
Dimond, E. G., C. F. Kittle, and J. E. Crockett.
Comparison of internal mammary artery ligation and
sham operation for angina pectoris. Am. J. Cardiol. 5:
483, i960.
Drinker, C. K., and J. M. Yoffey. Lymphatics, Lymph
and Lymphoid Tissue. Cambridge, Mass. : Harvard Univ.
Press, 1 94 1 .
Dripps, R. D. (editor). Proceedings of Symposium — The
Physiology oj Induced Hypothermia. Washington, D. C. :
Natl. Acad. Sci.-Natl. Research Council, 1956.
Driscol, T. E., and R. M. Berne. Role of potassium in
regulation of coronary blood flow. Proc. Soc. Exptl. Biol.
M,d. 96:505, 1957.
>76
HANDBOOK OF PHYSIOIOCY
CIRCULATION II
90
91.
89. Eckel, R., R. W. Eckstein, M. Stroud, and VV. H.
Pritchard. Effects of over and underperfusion upon
coronary arterial blood flow. Federation Proc. 8: 38, 1949
ECKENHOFF, J. E., J. H. 1 1 AFKENSCHIEL, AND G. M.
Landmesser. The coronary circulation in the dog. Am.
J. Physiol. 148:582, 1947.
ECKENHOFF, J. E., AND J. H. H A FKENSCHIEL. The effect
of nikethamide on coronary blood flow and cardiac
oxygen metabolism. J. Pharmacol. Exptl. Therap. gi : 362,
'947-
92. Eckenhoff, J. E., J II Hafkenshiel, M. H. Uarmel,
W. T. Goodale, M. Lubin, R. J. Binc, and S. S. Kety.
Measurement of coronary blood How by the nitrous oxide
method. Am. J. Physiol. 152: 356, 1948.
93. Eckstein, R. W., M. Stroud III, R. Eckel, C. V.
Dowling, and W. H. Pritchard. Effects of control of
cardiac work upon coronary flow and oxygen consump-
tion after sympathetic nerve stimulation. Am. J. Physiol.
163:539. '95°-
94. Eckstein, R. W., VV. B. Newberry, J. A, McEachern,
and G. Smith. Studies of the anti-adrenergic effects of
nitroglycerin on the dog heart. Circulation 4: 534, 1951.
95. Eckstein, R. VV., and D. S. Leighninger. Chronic effects
of aorta-coronary sinus anastomosis of Beck in dogs.
Circulation Research 2 : 60, 1 954.
96. Eckstein, R. VV. Coronary interarterial anastomoses in
young pigs and mongrel dogs. Circulation Research 2 : 460,
'954-
97. Eckstein, R. VV. Development of interarterial coronary
anastomoses by chronic anemia. Disappearance following
correction of anemia. Circulation Research 3: 306, 1955.
98. Eckstein, R. VV. Effect of exercise and coronary artery
narrowing on coronary collateral circulation. Circulation
Research 5: 230, 1 957.
99. Eckstein, R. VV, and R. E. Hurley. Effect of bilateral
internal mammary artery ligation on coronary circula-
tion in dogs. Circulation Research 7:571, 1 959.
100. Edwards, J. C, C. Burnsides, R. L. Swarm, and A. I.
Lansing. Arteriosclerosis in the intramural and extra-
mural portions of coronary arteries in the human heart.
Circulation 13:235, 1956.
1 01. Edwards, J. E. Anomalous coronary arteries with special
reference to arteriovenous-like communications. Circula-
tion 17: 1 00 1 , 1 958.
102. Edwards, VV. S., A. Siegel, and R. J. Bing. Studies on
myocardial metabolism. III. Coronary blood flow, myo-
cardial oxygen consumption and carbohydrate metabo-
lism in experimental hemorrhagic shock. ./. Clin. Invest.
33: 1646, 1954.
103. Edwards, VV. S., S. Tuluy, W. E. Reber, A. Siegel,
and R. J. Bing. Coronary blood flow and myocardial
metabolism in hypothermia. Ann. Surg. 139: 275, 1 954.
104. Engle, M. A., E. I. Goldsmith, G. R. Holswade, H. P.
Goldberg, and F. Glenn. Congenital coronary arterio-
venous fistula. New Engl. ./. Med. 264: 856, 1961.
105. Essex, H. E., J. F. Herrick, E. J. Baldes, and F. C.
Mann. The effects of exercise on the coronary blood flow,
heart rate and blood pressure of trained dogs with de-
nervated and partially denervated hearts. Am. J. Physiol.
138:687, 1943.
106. Evans, C. L., and E. H. Starling. The part played by
107.
108.
109.
"3-
114.
11b.
117.
118.
119.
123.
124.
the lungs in the oxidative processes of the body. J. Physiol.,
London 46: 413, 1913.
Evans, C. L., F. Grande, and F. V. Hsu. Two single
heart oxygenator systems for the heart. Quart. ./. Exptl.
Physiol. 24: 283, 1934.
Eystek, C. J. The muscular architecture of the ventricles
and atria of hog and dog hearts. Dissertation Abst. 14:
216, [954.
Farrand, R. L., and S. M. Horvath. Effects of khellin
on coronary blood flow and related metabolic functions.
Am. ./. Physiol. 196: 391, 1959.
Favarger, H. Die chronische Tabakvergiftung und ihren
Einfluss auf das Herz und den Mogen. Wien. klin.
Wochschr. Nr. 11 -14, 1887.
Fawcett, D. VV. The fine structure of capillaries, arterioles
and small arteries. Symposium on F'actors Influencing Ex-
change of Substances Across Capillary II 'all. Proc. Conf. Micro-
circul. Physiol. Pathol. Urbana: Univ. Illinois Press, 1959.
Feinberg, H., and L. N. Katz. Effect of catecholamines,
1-epinephrine and 1-norepinephrine on coronary flow
and oxygen metabolism of the myocardium. Am. J.
Physiol. 193: 151, 1958.
Feinberg, H., A. Gerola, and L. N. Katz. Effect of
hypoxia on cardiac oxygen consumption and coronary
flow. Am. J. Physiol. 195: 593, 1958.
Feinberg, H., A. Gerola, and L. N. Katz. Effect of
changes in blood CO» level on coronary flow and myo-
cardial oxygen consumption. Am. J. Physiol. 199: 349,
i960.
Feinberg, H., E. Boyd, and L. N. Katz. Calcium effect
on performance of the heart. Am. J. Physiol. 202: 643,
1 962 .
Fishman, A. P. (guest editor). The myocardium — its
biochemistry and biophysics. Circulation 24: 324, 1961.
Foltz, E. L., S. K. Wong, and J. E. Eckenhoff. Effects
of certain "cardiac stimulant" drugs on coronary circula-
tion and cardiac oxygen metabolism. Federation Proc. 7 :
219, 1 948.
Foltz, E. L., R. G. Page, VV. F. Sheldon, S. K. Wong,
VV. J. Tuddenham, and A. J. Weiss. Factors in variation
and regulation of coronary blood flow in intact anesthe-
tized dogs. Am. J. Physiol. 162: 521, 1950.
Foltz, E. L., M. Wendel, and J. W. West. Effects of
aortic insufficiency on coronary blood flow and cardiac
oxygen consumption. Federation Proc. 12: 44, 1953.
Freis, E. D., H. VV. Schapner, R. L. Johnson, and G. E.
Schreiner. Hemodynamic alterations in acute myocardial
infarction. I. Cardiac output, mean arterial pressure,
total peripheral resistance, "central" and total blood
volumes, venous pressure and average circulation time.
,/. Clin. Incest. 31 : 131, 1952.
Friedberg, C. K. Cardiogenic shock in acute myocardial
infarction. Circulation 23: 325, 1 961.
Frolkis, V. V., and V. I. Milko. The uptake of radio-
active phosphorus (P32) by various structures of the heart.
Bull. Expil. Biol. Med., U.S.S.R., English Trans!. 48: 842,
'959-
Gage, A. A., K. C. Olson, and W. M. Chardach. Ex-
perimental coronary thrombosis in the dog. Description
of a method. .Inn. Surg. 143: 535, 1956.
Gage, A. A., K. C. Olson, and W. M. Chardach.
BLOOD SUPPLY TO THE HEART
'577
'*5
uf.
127
129.
130.
'3'-
132.
■33-
134
i35-
.36.
■37-
138.
'39-
140.
141.
142.
■43-
144
Cardiopericardiopexy. An experimental evaluation. Ann.
Surg. 147:289, 1958.
Gallo, P. A study on the topographical and quantitative
nlations between capillaries and fibers of the conduction
system of the heart and on their functional significance.
Cardiologia 29: 241, 1956.
Garcia-Ramos, J., J. Alanis, and A. Rosenblueth.
Estudios sobre la circulacion coronaria. I. Factores extra-
vasculares. Arch. inst. cardial. Mix. 20: 474, 1950.
George, J. M., and D. M. Knowlan. Anomalous origin
of the left coronary artery from the pulmonary artery in
an adult. New Engl. J. Med. 261 : 993, 1959.
Gerola, A., H. Feinberg, and L. N. Katz. Myocardial
oxygen consumption and coronary blood flow in hypo-
thermia. Am. J. Physiol. 196: 719, 1959.
Gibson, J. G., A. M. Seligman, W. C. Peacock, J. C.
Aub, J. Fine, and R. D. Evans. The distribution of red
cells and plasma in large and minute vessels of the normal
dog, determined by radio-active isotopes of iron and io-
dine. J. Clin. Invest. 25: 848, 1946.
Goh, K. O., and R. D. Dallam. Oxygen consumption of
the auricles, right and left ventricles of the normal, hypo-
thyroid and hyperthyroid rat heart. Am. J. Physiol. 188:
5 '4. '957-
Goldberg, L. I., R. D. Bloodwell, E. Braunwald,
and A. G. Morrow. The direct effects of norepinephrine,
epinephrine and methoxamine on myocardial contractile
force in man. Circulation 22: 1125, 1960.
Gonzalez, H., and D. Erlij. Un reflejo circulatorio de
origen coronario. Arch. inst. cardiol M'ex. 28 : 404, 1 958.
Goodale, W. T., R. E. Olson, and D. B. Hackel. The
effects of fasting and diabetes mellitus on myocardial
metabolism in man. Am. J. Med. 27: 212, 1 959.
Goodver, A. V. N., W. F. Eckhardt, R. H. Ostberg,
and M. J. Goodkind. Effects of metabolic acidosis and
alkalosis on coronary blood flow and myocardial metabo-
lism in the intact dog. Am. J. Physiol. 200: 628, 1961.
Gorlin, R., and J. P. Storaasli. Transcoronary circu-
lation time: A new method of evaluating the coronary
vascular system. Circulation 14: 943, 1956.
Gorlin, R. Coronary blood flow. Methods in Med.
Research 7 : [21 , 1 958.
Gorlin, R. Studies on the regulation of the coronary
circulation in man. I. Atropine-induced changes in
cardiac rate. Am. J. Med. 25: 37, 1958.
Gorlin, R., N. Brachfeld, C. MacLeod, and P. Bopp.
Effect of nitroglycerin on the coronary circulation in
patients with coronary artery disease or increased left
ventricular work. Circulation ig: 705, 1 959.
Granata, L., A. Huvos, and D. E. Gregg. Hemodynamic
changes in coronary and mesenteric arterial beds following
sympathetic nerve stimulation. Physiologist 4 (No. 3) :
42, 1961.
Grant, R. T. Development of the cardiac coronary
vessels in the rabbit. Heart 13: 261, 1926.
Grant, R. T. An unusual anomaly of the coronary vessels
in the malformed heart of a child. Heart 13: 273, 1926.
Grant, R. T., and M. Regnier. The comparative anat-
omy of the cardiac coronary vessels. Heart 1 3 : 285, 1 926.
Grayson, J., and D. Mendel. Myocardial blood flow in
the rabbit. Am. J. Physiol. 200: 968, 1961.
Green, H. D., and D. E. Gregg. Changes in the coronary
circulation following increased aortic pressure, augmented
cardiac output, ischemia and valve lesions. Am. J. Physiol.
1 30 : 1 26, 1 940.
145. Green, H. D. Effect of Pitressin, the nitrites, epinephrine
and the xanthines on coronary flow in mammalian hearts.
In : Blood, Heart and Circulation. Washington, D.C. : Am.
Assoc. Advance. Sci. Publ. 13, 1940, p. 105.
146. Green, H. D., R. Wegria, and H. H. Boyer. Effect of
epinephrine and Pitressin on the coronary artery inflow
in anesthetized dogs. J. Pharmacol. Exptl. Therap. 76 :
37°, I942-
147. Green, H. D., and R. Wegria. Effects of asphyxia, anoxia
and myocardial ischemia on the coronary blood flow.
Am. J. Physiol. 135: 271, 1942.
148. Green, H. D. Circulation — Blood Flow Measurement.
Methods in Medical Research. Chicago: Yr. Bk. Pub., 1948,
vol. 1, pp. 66-253.
149. Green, H. D. Circulatory system: methods. Med. Physics
2 : 208, 1950.
150. Green, H. D., and J. H. Kepschar. Control of systemic
resistance in major systemic vascular beds. Physiol. Revs.
39:6[7. '959-
151. Gregg, D. E. Phasic blood flow and its determinants in
the right coronary artery. Am. J. Physiol, iiq: 580, 1937.
152. Gregg, D. E. Phasic changes in flow through different
coronary branches. In: Blood, Heart and Circulation. Wash-
ington, D.C. : Am. Assoc. Advance. Sci. Publ. 13, 1940,
p. 81.
153. Gregg, D. E. The Coronary Circulation in Health and
Disease. Philadelphia : Lea & Febiger, 1 950.
154. Gregg, D. E., F. H. Loncino, P. A. Green, and L. J.
Czerwonka. A comparison of coronary flow determined
by the nitrous oxide method and by a direct method using
the rotameter. Circulation 3: 8g, 1951.
155. Gregg, D. E. Some problems of the coronary circulation.
Verhandl. deut. Ges. Kreislaujforsch. 21: 22, 1955.
156. Gregg, D. E., and D. C. Sabiston, Jr. Current research
and problems of the coronary circulation. Circulation 13:
916, 1956.
157. Gregg, D. E., R. C. Batterman, L. N. Katz, W. Raab,
and H. I. Russek. Experimental methods for the evalua-
tion of drugs in various disease states. Part II. Angina
pectoris. Ann. N. Y. Acad. Sci. 64 : 494, 1 956.
158. Gregg, D. E. Regulation of the collateral and coronary
circulation of the heart. Circulation. Proceedings Harvey
Tercentenary Congress. Oxford: Blackwell Sci. Publ., 1958,
p. 168.
159. Gregg, D. E. Hemodynamic factors in shock. Proc. Intern.
Symp. Shock. Sweden : Saltjobaden, 1961 .
1 59a. Gregg, D. E., E. M. Khouri, C. R. Rayford, L.
Granata, and A. Huvos. The systolic component of
coronary arterial inflow in the active unanesthetized dog.
Proc. Intern. Union Physiol. Sci. Leiden: 1962, vol. 11.
160. Griffith, G. C. Amine oxidase inhibitors. Their current
place in the therapy of cardiovascular disease. Circulation
22:1 156, i960.
161. Griggs, D. M., Jr., P. R. Holt, and R. B. Case. Serial
pressure-volume studies in the excised canine heart. Am.
J. Physiol. ig8: 336, ig6o.
162. Grob, D., W. R. Scarborough, A. A. Kattus, and H. G.
Lang ford. Further observations on the effects of auto-
I 378 HANDBOOK OF PHYSIOLOGY
CIRCULATION II
nomic blocking agents in patients with hypertension.
Ciri ulation 8: 352, 195;).
163. Guyton, A. C, and J. VV. Crowell. Dynamics of the 183
heart in shock. Federation Pine. 20: 51, 1961.
164. Guz, A., G. S. Kurland, and A. S. Freedberg. Relation
of coronary How to oxygen supply. Am. .1 . Physiol. 199: 184
179, i960.
165. Guzman, S. V., E. Swenson, and M.Jones. Intercoronary 185
reflex : demonstration by coronary angiography. Circula-
tion Research 10: 739, 1962.
166. Hackel, D. B., and W. T. Goodale. Effects of hemor- 186
rhagic shock on the heart and circulation of intact dogs.
Circulation 1 1 628, 1955.
167. Hackel, D. B., and G. H. Clowes, Jr. Coronary blood
flow and myocardial metabolism during hypoxia in
adrenalectomized-sympathectomized dogs. Am. J. Physiol, 187.
186: 111, 1956.
168. Hackel, D. B., S. M. Sancetta, and J. Klfinerman. 188.
Effect of hypotension due to spinal anesthesia on coronary
blood flow and myocardial metabolism in man. Circula-
tion 13: 92, 1956. 189.
169. Hackel, D. B. Effect of insulin on cardiac metabolism of
intact normal dogs. Am. J. Physiol. 199: 1135, i960. 190.
170. Hahn, R. S., and C. S. Beck. Revascularization of the
heart. A study of mortality and infarcts following multiple 191.
coronary artery ligation. Circulation 5: 801, 1952.
171. Hall, R J , E. M. Khouri, and D. E. Gregg. Coronary- 192.
internal mammary artery anastomosis in dogs. Surgery
50:560, 1961.
172. Halpern, M. H. Arterial supply to the nodal tissue in 193.
the dog heart. Circulation 9: 547, 1954-
173. Halpern, M. H. Blood supply to the atrioventricular
system of the dog. Anal. Record 121 : 753, 1955.
1 74. I [alpern, M. H. The dual blood supply of the rat heart. 194.
Am ./. Anal. 1 in : 1 , 1957.
175. Hansen, A. T., B. F. Haxholdt, E. Husfeldt, N. A.
Lassen, O. Munck, H. Rahbek Sorensen, and K.. 195.
Winkler. Measurement of coronary blood flow and
cardiac efficiency in hypothermia by use of radioactive
krypton 85. Scand. J. Clin. & Lah. Invest. 8: 182, 1956.
176. Hanson, K. M, and J. A. Johnson. The effect of Pitressin 196.
on the isolated perfused rabbit heart. Am. J. Physiol. 190:
a'> '957-
177. Hardin, R. A., J. B. Scott, and F. J. Haddv. Effect of 197.
cardiac cooling on coronary vascular resistance in normo-
thermic dogs. .4m. J. Physiol. 199: 163, i960.
178. Harken, D E., H. Black, J. F. Dickson, and H. E. Wil- 198.
son HI. Dcepicardialization : A simple, effective surgical
treatment for angina pectoris. Circulation 12: 955, 1055.
: ;«i Hashimoto, K, T. Shigei, S. Imai, Y. Saito, N. Yago, '99
I. Vei, and R. E. Clark. Oxygen consumption and
coronary vascular tone in the isolated fibrillating dog
2 ( M 1
heart. Am. ./. Physiol. 198: 965, i960.
180. Hegnauer, A. H., and H. E. D'Amato. Oxygen con-
sumption and cardiac output in the hypothermic dog.
Am. J. Physiol. 178. 138, 1954. J(M
181. Hellems, H. K, J. \V. Ord, F. N. Talmers, and R. C
Christensen. Effects of hypoxia on coronary blood flow
and myocardial metabolism in normal human subjects. >ll2
Circulation 16:893, 1957.
182. Hilton, R., and F. Eichoi.tz. The influence of chemical
factors on coronary circulation. J Physiol., London 59:
4' S, '9^5-
Hellerstein, H. K., and J. L. Orbison. Anatomic
variations of the orifice of the human coronary sinus.
Circulation 3:514, 1 95 1 .
Henry Ford Hospital. Symposium on the Etiology of Myo-
cardial Infarction. Boston: Little, Brown. In press.
Hershgold, E. J., S. II. Steiner, and L. A. Sapirstein.
Distribution of myocardial blood flow in the rat. Circula-
tion Research 7: 551, 1959.
Hoffman, F., E. J. Hoffman, S. Middleton, and J.
Talesnik. The stimulating effects of acetylcholine on the
mammalian heart and the liberation of an epinephrine-
like substance by the isolated heart. Am. J. Physiol. 144:
l89, '945-
Ikeda, M. The nervous control of the coronary circula-
tion. Japan. Circulation J. 21: 1, 1957.
Jacobev, J. A., W. J. Taylor, G. T. Smith, R. Gorlin,
and D. E. Harken. A new therapeutic approach to
acute coronary occlusion. Surg. Forum 12: 225, 1961.
James, T. N., and G. E. Burch. Blood supply of the
human interventricular septum. Circulation 17: 391, 1958.
James, T. N., and G. E. Burch. The atrial coronary
arteries in man. Circulation 17: 90, 1958.
Jamfs, T. N. The arteries of the free ventricular walls in
man. Anat. Record 136: 371, i960.
Jardetzky, O., E. A. Greene, and V. Lorber. Oxygen
consumption of the completely isolated dog heart in
fibrillation. Circulation Research 4: 144, 1956.
Jflliffe, R. W., C. R. Wolf, R. M. Berne, and R. W.
Eckstein. Absence of vasoactive and cardiotropic sub-
stances in coronary sinus blood of dogs. Circulation Re-
search 5: 382, 1957.
Jennings, R. B., and W. B. Wartman. Production of an
area of homogeneous myocardial infarction in the dog.
I .1/ .! Arch. Pathol. 63:580, 1957.
Jochim, K. Vascular and extravascular factors influencing
coronary blood flow. In: Blood, Heart and Circulation.
Washington, DC: Am. Assoc. Advan. Sci. Publ. 13,
1940, p. 94.
Johnson, J. R., and C. J. Wiggers. The alleged validity
of coronary sinus outflow as a criterion of coronary reac-
tions. Am. J. Physiol. 118: 38, 1 937.
Johnson, J. A., V. Gott, and F. Welland. Perfusion
rates of brain, intestine and heart under conditions of
total body perfusion. Am. J. Physiol. 200: 551, 1961.
Jude, J. R., L. M. Haroutunian, and R. Folse. Hypo-
thermic myocardial oxygenation. Am. J. Physiol. 190:
57. '957-
Juhasz-Nagy, A., and M. Szentivanyi. Separation of
cardioaccelcrator and coronary vasomotor fibers in the
dog. Am. .1. Physiol. 200: 125, 1961.
Kadvtz, R. Die pharmakologischcn Eigenschaften der
Neuen Coronarer-weiternden Substanz 2,6-Bis(dia-
ethanolamino) - 4,8 - dipiperidinopyrimido(5 - 4 - d)
Pyrimidin. Arzneimittel-Forsch. 9: 39, 1 959.
K.AKO. K , J D. Choudhury, and K J. Bing. Possible
mechanism of decline in mechanical efficiency of the
isolated heart. J. Pharmacol Exptl. Therap. 130: 46, i960.
Kako, K., A. Ciirysohou, and R. J. Bing. Factors af-
fecting myocardial storage and release of catecholamines.
Circulation Research 9: 295, 1 96 1.
BLOOD SUPPLY TO THE HEAR!
'579
206.
207.
203. Kantrowitz, A., and A. Kantrowitz. Experimental 221.
augmentation of coronary flow by retardation of the
arterial pressure pulse. Surgery 34: 678, 1953. 222.
204. Kattus, A. A., and D. E. Gregg. Some determinants of
coronary collateral blood flow in the open-chest dog.
Circulation Research 7: 628, 1959.
205. Katz, L. N., K. Jochim, and A. Bohning. The effect of 223.
the extravascular support of the ventricles on the flow in
the coronary vessels. Am. J. Physiol 122: 236, 1938. 224.
Katz, L. N., and K. Jochim. Observations on the inner-
vation of the coronary vessels of the dog. Am. J. Physiol.
126: 395, 1939. 225.
Katz, A. M., L. N. Katz, and F. L. Williams. Regula-
tion of coronary flow. Am. J. Physiol. 180: 392, 1955.
208. Katz, L. N., and H. Feinberg. The relation of cardiac
effort to myocardial oxygen consumption and coronary 226.
flow. Circulation Research 6 : 656, 1 958.
209. Katz, L. N. Cigarette smoking and cardiovascular dis- 227.
ease. Circulation 22: 160, i960.
210. Keith, J. D. The anomalous origin of the left coronary
artery from the pulmonary artery. Brit. Heart J. 2 1 : 1 49, 228.
'959-
si I. Kelso, A. F., and W. C. Randall. Ventricular changes
associated with sympathetic augmentation of cardio-
vascular pressure pulses. Am. J. Physiol. 196: 731, 1959- 229.
212. Khouri, E. M., D. E. Gregg, R. J. Hall, and C. R.
Ravford. Regulation of coronary flow during treadmill
exercise in the dog. Physiologist 3 (No. 3): 93, i960.
212a. Khouri, E. M., and D. E. Gregg. Miniature electro- 230.
magnetic flow meter applicable to coronary arteries.
./. Appl. Physiol. 18: 224, 1963.
213. Kien, G. A., and T. R. Sherrod. Action of nicotine and
of smoking on coronary circulation and myocardial
oxygen utilization. Ann. A'.)'. Acad. Set. 90: 161, i960. 231.
Kisin, I. E. The influence of certain pharmacological
agents, used in the treatment of stenocardia, on the
coronary circulation. In : New Data on the Pharmacology of
the Coronary Circulation. Moscow: U.S.S.R. Acad. Med. 232.
Sci., i960, vol. 11.
Knowlton, F. P., and E. H. Starling. The influence of
variations in temperature and blood pressure on the
performance of the isolated mammalian heart. ./. Physiol., 233.
London 44: 206, 1912.
Koboyashi, I. G. I., A. Nakanishi, S. Murav, M. Shiba
K. Kato, Y. Tachenchi, H. Yasuda, and Y. Mikano. 234.
Studies on coronary circulation in man by method of
coronary sinus catheterization. Japan. Circulation J. 20.
299. '956- 235,
Kolin, A. Circulatory system : methods, blood flow deter-
mination by electromagnetic method. In: Medical Physics
(O. Glasser, ed.). Chicago: Year Book Pub., 3: 141, i960.
218. Kountz, W. B., and I. R. Smith. The flow of blood in the
-30
coronary arteries in pathological hearts. J. Clin. Invest.
H- '47. '938-
219. Kuhn, L. A., F. L. Gruber, A. Frankel, and S. Kupfer.
Hemodynamic effects of extracorporeal circulation. Cir-
culation Research 8 : 1 99, 1 960. ■* ' '
220. Kuzmina-Prigradova, A. V. Collateral circulation
after ligation of the anterior descending coronary artery,
and effect of vagal stimulation. Experiments on dogs. 238
Bull. Exptl. Biol. Med., U.S.S.R. English Transl. 42: 67,
1956. 2 39
214.
2'5-
216.
217.
Langendorff, O. Untersuchungen am iiberlebenden
Saugethierherzen. Pfliigers Arch. ges. Physiol. 61 : 291, 1895.
Lanier, J. 1 , II. D. Green, J. Hardawav, H. D.John-
son, and W. B. Donald. Fundamental difference in
reactivity of the blood vessels in skin compared with those
of muscle. Circulation Research 1 : 40, 1953.
Lapin, V. A. Pathogenesis of myocardial infarction. Bull.
Exptl. Biol. Mid.. U.S.S.R English Transl. 40: 19, 1955.
Lasker, N., T. R. Sherrod, and K. F. Killam. Alcohol
on the coronary circulation of the dog. J. Pharmacol.
Exptl. Therap. 113: 414, 1955.
Laurent, D., C. Bolenf.-Wii i i wis, F. L. Williams,
and L. N. Katz. Effect of heart rate on coronary flow
and cardiac oxygen consumption. Am. J. Physiol. 185.
355- '956-
Laurie, W., and J. D. Woods. Anastomoses of the coro-
nary circulation. Lancet 2: 812, 1958.
Leary, T., and J. T. Wearn. Two cases of complete
occlusion of both coronary orifices. Am. Heart J. 5: 412,
1930.
Lebedinskii, A. V., V. I. Medvedev, and I. A. Peimer.
Importance of Coronary Spasm in the Pathogenesis of Coronary
Insufficiency. Sukhumi : Nauk. Nauk Akad. Med. S.S.S.R.
'954. P- 32-
Leighninger, D. S., R. Rueger, and C. S. Beck. Effect
of glyceryl trinitrate 1 nitroglycerin) on arterial blood
supply to ischemic myocardium, ,4m. J. Cardiol. 3: 638,
'959-
Leight, D., V. DeFazio, F. N. Talmers, T. J. Regan,
and H. K. Hellems. Coronary blood flow, myocardial
oxygen consumption and myocardial metabolism in
normal and hyperthyroid human subjects. Circulation
'4:9°> '95°-
LeRoy, G. V., G. K. Fenn, and N. C. Gilbert. The
influence of xanthine drugs and atropine on the mortality
rate after experimental occlusion of a coronary artery.
Am. Heart J. 23: 637, 1942.
Lev, M., and C S. Simkins Architecture of the human
ventricular myocardium: technique for study using a
modification of the Mall-MacCallum method. Lab.
Invest. 5 : 396, 1 956.
Levy, M. N., and A. L. Frankel. Vasomotor responses
to acute coronary occlusion. Am. J Physiol. 172: 427,
■953-
Levy, M. N., and J. M. DeOliveira. Regional distribu-
tion of myocardial blood flow in the dog as determined
by RbK,;. Circulation Research 9: 96, ig6i.
Levy, M. N., E. S. Imperial, and H. Zieski. Collateral
blood flow to the myocardium as determined by the
clearance of rubidium86 chloride. Circulation Research 9:
1035. '961-
Lewis, F. B., J. D. Goffman, and D. E. Gregg. Effect
of heart rate and intracoronary isoproterenol, levarterenol
and epinephrine on coronary flow and resistance. Circula-
tion Research g: 89, 1961.
Livesav, W. R., J. H. [over, and D. W. Chapman.
The cardiovascular and renal hemodynamic effects of
Aramine. Am. Heart J. 47: 745, 1954.
Lochner, W., and E. Witzleb. Probleme dec Coronar-
durchblutung. Berlin: Springer -Verlag, 1958.
Lombardo, T. A., L. R. Radigan, and C. Morrow.
i58o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Myocardial failure in experimental hypothermia. Cir-
culation Research 5: 22, 1 957.
240. Lombardo, T. A., L. Rose, M. Taeschler, S. Tuluy,
and R. J. Bing. The effect of exercise on coronary blood
flow, myocardial oxygen consumption and cardiac effi-
ciency in man. Circulation 7:71, 1 953.
241. Longmire, W. P., J. A. Cannon, and A. A. Kattus.
The surgical treatment of angina pectoris. Arch. Internal
Med. 1 04 : 886, 1 959.
242. Lorber, V., and G. T. Evans. Mechanical response of
the isolated mammalian heart to anoxia. Proc. Soc. Expll.
Biol. Med. 54: 1, 1943.
243. Lorber, V. Energy metabolism of the completely isolated
mammalian heart in failure. Circulation Research 1 : 298,
'953-
244. Love, W. D., and G. E. Burch. Differences in the rate
of Rb86 uptake by several regions of the myocardium of
control dogs and dogs receiving 1-norepinephrine or
Pitressin. J. Clin. Invest. 36: 479, 1 957.
245. Love, W. D., and G. E. Burch. Influence of the rate of
coronary plasma flow on the extraction of Rb86 from
coronary blood. Circulation Research 7: 24, 1959.
246. Lumb, G., R. L. Shocklett, and W. A. Dawkins. The
cardiac conduction tissue and its blood supply in the
dog. .4m. J. Pathol. 35:467, 1959.
247. MacLean, L. D., P. H. Hedenstrom, and S. K. Young.
Distribution of blood How in the canine heart. Proc. Soc.
Exptl. Biol. Med. 107: 786, 1 96 1.
248. McElroy, Wm. T., Jr., A. J. Gerdes, and E. B. Brown,
Jr. Effects of CO2, bicarbonate and pH on the perfor-
mance of isolated perfused guinea pig hearts. Am. J.
Physiol. 195: 412, 1958.
249. McKeever, W. P., D. E. Gregg, and P. C. Canney.
Oxygen uptake of the non-working left ventricle. Circula-
tion Research 6: 612, 1958
250. MAGAK1AN, G. 0.,D. I. MlMNOSHVILI, AND G. I. KoKOIA.
Eksperimental'noe izuchenie patogeneza gipertonii i
koronarnoi nedostatochnosti. Klin. Med., U.S.S.R. 34:
3°. '956-
251. Mamiya, R. T., T. Cooper, V. L. William, J. G. Mudd,
and C. R. Hanlon. Distal relocation of the origin of the
left coronary artery by subclavian left coronary anasto-
mosis. Surg. Gynecol. Obstet. 113: 599, 1961.
252. Marchioro, T., A. Feldman, J. C. Owens, and H.
Swan. Measurement of myocardial blood flow. Indicator-
dilution technique. Circulation Research g: 541, 1961.
253. Mart, J. A., and J. R. Miller. The effect of diathermy
on coronary flow: an experimental study in dogs. Am.
Heart J. 29: 390, 1945.
254. Matthes, K. Myocardial shock. Ciba Foundation Sym-
posium Shod. Sweden: Saltjobaden, 196 1.
255. Mautz, F. R. Anatomical and physiological considera-
tions in the development of a collateral circulation to the
myocardium. Diseases of Chest 31 : 265, 1957.
256. Maxwell, G. M., C. A. Castillo, D. H. White, Jr.,
C. W. Crumpton, and G. G. Rowe. Induced tachycardia:
Its effect upon coronary hemodynamics, myocardial
metabolism and cardiac efficiency of the intact dog. J.
Clin. Incest. 37: 1413, 1958.
257. Maxwell, G. M., C. A. Castillo, (.'.. W. Crumpton,
and G. G. Rowe Hyperthermia: Systemic and coronary
circulation changes in the intact dog. Am. Heart J. 58:
854. 1959-
258. May, A. M. Surgical anatomy of the coronary arteries.
Diseases of Chest 38: 645, i960.
259. Melville, K. I., and I. Mazurkiewicz. Actions of
potassium and calcium on coronary flow and heart con-
tractions with special reference to the responses to epi-
nephrine and norepinephrine. J. Pharmacol. Exptl. Therap.
118: 249, 1956.
260. Mena, I., A. A. Kattus, M. A. Greenfield, and L. R.
Bennett. Effect of coronary blood flow on radioisotope
dilution curves measured by precordial scintillation
detection. Circulation Research 9:911, 1961.
261. Miller, A. (., R. Pick, and L. N. Katz. Ventricular
endomyocardial pathology produced by chronic cardiac
lymphatic obstruction in the dog. Circulation Research 8.
941, i960.
262. Mitchell, G. A. G. The innervation of the heart. Brit.
Heait J. 15: 159, 1953.
263. Mitchell, J. H., R. J. Linden, and S. J. Sarnoff.
Influence of cardiac sympathetic and vagal nerve stimu-
lation on the relation between left ventricular diastolic
pressure and myocardial segment length. Circulation
Research 8:11 00, 1 960.
264. Moe, G. K., and M. Visscher. The distribution of
coronary blood flow. In: Blood, Heart and Circulation.
Washington, DC: Am. Assoc. Advance. Sci. Publ. 13,
1940, p. 100.
265. Moir, T. W., R. W. Eckstein, and T. E. Driscol. Phasic
and mean blood flow in the canine septal artery and an
estimate of systolic resistance in deep myocardial vessels.
Circulation Research 12: 203, 1963.
266. Moir, T. W., R. W. Eckstein, and T. E. Driscol.
Thebesian drainage of the septal artery. Circulation Re-
search 12 : 212, 1963.
267. Monroe, R. G., G. French, and J. L. Whittenberger.
Effects of hypocapnia and hypercapnia on myocardial
contractility. Am. J. Physiol. 199: 1121, i960.
268. Monroe, R. G., and G. French. Ventricular pressure-
volume relationships and oxygen consumption in fibrilla-
tion and arrest. Circulation Research 8 : 260, 1 960.
269. Monroe, R. G., and G. N. French. Left ventricular
pressure -volume relationships and myocardial oxygen
consumption in the isolated heart. Circulation Research 9:
362, 1 96 1.
270. Moore, D. H., and H. Ruska. The fine structure of
capillaries and small arteries. J. Biophys. Biochem. Cytol. 3 :
457- '9^7-
271 Moran, R., C. G. Neumann, J. Wedel, J. Lord, P. W.
Stone, and J. W. Hinton. Revascularization of the heart
by tubed pedicle graft of skin and subcutaneous tissue.
Plastic Reconstruc. Surg. 10: 295, 1952.
272. Morawttz, P., and A. Zahn. Untersuchungen iiber den
Coronarkreislauf. Devi. Arch. khn. Med. 116: 364, 1914.
273. Nahas, G. G., and M. Cavert. Cardiac depressant effect
of COo and its reversal. Am. J. Physiol. 190: 483, 1 957.
274. Nolting, D., R. Mack, E. Luthy, M. Kirsch, and C.
Hogancamp. Measurement of coronary blood flow and
myocardial rubidium uptake with Rb86. J. Clin. Invest.
37:921, 1958.
275. Nuki, B. The pharmacology of the coronary circulation.
Japan. Circulation J. 21 : 279, 1957.
BLOOD SUPPLY TO THE HEART
276. NvDICK, I., P. Ruegsegger, R. AbARQUEZ, E. E.
Cliffton, and J. S. LaDue. The effect of fibrinolytic
agents on myocardial infarction. Progr. Cardiovascular
Diseases 3:13, i960.
277. Okinaka, S., M. Ikeda, K. Hashiba, K. Murata, J.
Kanedo, T. Ozawa, H. Nitani, Z. Ishimi, J. Fujii, Y.
Takeda, K. Kuramoto, M. Tsuji, and F. Terasawa.
Studies on the control of coronary circulation, Part I.
The effect of the stimulation of the nerves on the coronary
circulation. Part II. The humoral effect on the coronary
circulation. Am. Heart J. 56: 319, 1958.
278. Olson, R. E., and D. A. Piatnek. Conservation of energy
in cardiac muscle, in metabolic factors in cardiac con-
tractility. Ann. N.Y. Acad. Sci. 72: 466, 1959.
279. Olson, R. E. Myocardial metabolism in congestive heart
failure. J. Chronic Diseases g: 442, 1959.
280. Olsson, R. A., and D. E. Gregg. Reactive hyperemia
characteristics of the myocardium. Federation Proc. 21 :
1 06, 1962.
281. Opdyke, D. F., and R. C. Foreman. A study of coronary
How under conditions of hemorrhagic hypotension and
shock. Am. J. Physiol. 148: 726, 1947.
282. Opdyke, D. F., and E. E. Selkurt. A study of alleged
intercoronary reflexes following coronary occlusion.
Am. Heart J. 36: 73, 1948.
283. Osher, W. J. Pressure-flow relationship of the coronary
system. Am. J. Physiol. 172: 403, 1953.
284. Outschoorn, A. S., and M. Vogt. Nature of cardiac
sympathin in the dog. Brit. J. Pharmacol. 7: 319, 1952.
285. Ozawa, T. Studies on the reflex mechanism in relation
to coronary circulation. I. The effect of distention of the
gall bladder on coronary circulation. II. Pressoreflex
arising from the left coronary artery. Japan. Circulation J.
23: 126, 137, 1959.
286. Palade, G. E. Blood capillaries of the heart and other
organs. Circulation 24: 368, 1961.
287. Patek, P. R. The morphology of the lymphatics of the
mammalian heart. Am. J. Anat. 64: 203, 1939.
288. Paul. M. H., E. O. Theilen, D. E. Gregg, J. B. March,
and G. G. Casten. Cardiac metabolism in experimental
ventricular fibrillation. Circulation Research 2: 573, 1954.
289. Paul, M. H., L. R. Norman, P. M. Zoll, and H. L.
Blumgart. Stimulation of interarterial coronary anasto-
moses by experimental acute coronary occlusion.
Circulation 16:608, 1957.
290. Pianetto, M. B. The coronary arteries of the dog. Am.
Heart J. 1 8 : 403, 1 939.
291. Pitt, B. Interarterial coronary anastomoses. Occurrence
in normal hearts and in certain pathologic conditions.
Circulation 20: 81 6, 1959.
292. Polaceky, P. Svalove mutsky a poutka na vencityck
tepnoch u clovela. Ceskoslov. morfol. 7: 119, 1959.
293. Porter, W. T. The vasomotor nerves of the heart. Boston
Med. Surg. J. 134: 39, 1896.
294. Provenza, D. V., and S. Scherlis. Demonstration of
muscle sphincters as a capillary component in the human
heart. Circulation 20: 35, 1959.
295. Provenza, D. V., and S. Scherlis. Coronary circulation
in dog's heart: demonstration of muscle sphincters in
capillaries. Circulation Research 7. 318, 1959.
296. Raab, W., and E. Lepeschkin. Anti-adrenergic effects of
nitroglycerin on the heart. Circulation 1: 733, 1950.
297. Raab, W. Neurohormonal factors in the origin and treat-
ment of angina pectoris. Experimental methods for the
evaluation of drugs in various diseased states. Ann. V. )'.
Acad. Sci. 64: 528, 1956.
298. Race, G. J., W. L. J. Edwards, E. R. Halden, H. E.
Wilson, and F. J. Luibel. A large whale heart. Circula-
tion 19: 928, 1959.
299. Ratnoff, O. D., and M. Plotz. The coronary circula-
tion. Medicine 25: 285, 1946.
300. Rayford, C. R., E. M. Khouri, F. B. Lewis, and D. E.
Gregg. Evaluation of use of left coronary artery inflow
and oxygen content of coronary sinus blood as a measure
of left ventricular metabolism. J. Appl. Physiol. 14: 817,
1959-
301. Rayford, C. R., A. Huvos, E. M. Khouri, and D. E.
Gregg. Some determinants of coronary flow in intact
dogs. Physiologist 4 (No. 3): 92, 1961.
302. Rebatel, F. Recherches experimentales sur la circulation dans
les arteres coronaires. Paris, 1872.
303. Regan, T. J., F. N. Talmers, R. C. Christensen, T.
Wada, and H. K. Hellems. Coronary blood flow and
myocardial metabolism in aortic insufficiency. Circulation
14:987, 1956.
304. Regan, T. J., H. K. Hellems, and R. J. Bing. Effect of
cigarette smoking on coronary circulation and cardiac
work in patients with arteriosclerotic coronary disease.
Ann. N.Y. Acad. Sci. 90: 186, i960.
305. Regan, T. J., M. J. Frank, P. H. Lehan, and H. K.
Hellems. Influence of red cell mass on myocardial blood
flow and oxygen uptake. Clin. Research 8: 367, i960.
306. Regan, T. J., M. J. Frank, J. F. McGinty, E. Zobl,
H. K. Hellems, and R. J. Bing. Myocardial response to
cigarette smoking in normal subjects and patients with
coronary disease. Circulation 23: 365, 1961.
307. Regan, T. J., K. Binak, S. Gordon, V. DeVazio, and
H. K. Hellems. Myocardial blood flow and oxygen con-
sumption during postprandial lipemia and heparin-
induced lipolysis. Circulation 23: 55, 1961.
308. Roberts, J. T., and S. D. Loube. Congenital single coro-
nary artery in man. Am. Heart J. 34: 188, 1947.
309. Rodbard, S., G. R. Graham, and F. Williams. Con-
tinuous and simultaneous measurement of total coronary
flow, venous return and cardiac output in the dog. J.
Appl. Physiol. 6: 311, 1953.
310. Rodriguez, F. L., and S. L. Robbins. Capacity of human
coronary arteries: a postmortem study. Circulation 19:
57°. '959-
311. Rohde, E. Stoffwechseluntersuchungen am uberlebenden
Warmbliiterherzen. 1. Zur Physiologie des Herzstoff-
wechsels. Z. physiol. Chan. 68: 181, 1910.
312. Rose, L. B., and Hoffman, D. L. The coronary blood
flow in pulmonary emphysema and cor pulmonale. Cir-
culation Research 4: 130, 1956.
313. Rosenblueth, A., J. Alanis, R. Rubio, and G. Pilar.
Relations between coronary flow and work of the heart.
Am. J Physiol. 200: 243, 1961.
314. Ross, J., Jr., E. Braunwald, and J. A. Waldhausen.
Studies on digitalis. II. Extracardiac effects on venous
return and on the capacity of the peripheral vascular bed.
J. Clin. Invest. 39: 937, i960.
315. Ross, J., Jr., P. W. Mosher, and R. F. Shaw. Auto-
regulation of coronary blood flow. Circulation 24: 1025,
1 96 1.
HANDBOOK OF PHYSIOKK.Y
CIRCULATION II
316. Rowe, G. G., J. H. Huston, G. M. Maxwell, A. B.
VVeinstein, H. Tuchman, and C. W. Crumpton. The
effects of 1-hydrazinophthalazine upon coronary hemo-
dynamics and myocardial oxygen metabolism in essential
hypertension. J. C.lin. Invest. 34: 696, 1955.
317. Rowe, G. G., J. H. Huston, A. B. Weinstein, H. Tuch- 334.
man, J. F. Brown, and C. W. Crumpton. The hemo-
dynamics of thyrotoxicosis in man with special reference
to coronary blood flow and myocardial oxygen metabo-
lism. ./. Clin. Invest. 35: 272, 1956.
318. Rowe, G. G., D. A. Emanuel, G. M. Maxwell, J. F. 335.
Brown, G. Castillo, B. Schuster, Q. R. Murphy, and
C. W. Crumpton. Hemodynamic effects of quinidine :
including studies of cardiac work and coronary blood
flow. J. Clin. Invest. 36: 844, 1957. 336.
319. Rowe, G. G. The nitrous oxide method for determining
coronary blood flow in man. Am. Heart J. 58: 268, 1959.
320. Rowe, G. G., C. A. Castillo, G. M. Maxwell, and
( ' \V Crumpton. The comparison of systemic and coro- 337.
nary hemodynamics in the normal human male and
female. Circulation Research 7: 728, 1959.
321. Rowe, G. G., G. M. Maxwell, C. A. Castillo, J. H.
Huston, and C. W. Crumpton. Hemodynamics of mitral 338.
stenosis with special reference to coronary blood flow
and myocardial oxygen consumption. Circulation 22 :
559. !96o. 339.
322. Sabiston, D. G, E. O. Theilen, and D. E. Gregg. The
relationship of coronary blood flow and cardiac output
and other parameters in hypothermia. Surgery 38 : 498,
■955- 34°.
323. Sabiston, D. C, Jr., and D. E. Gregg. Effect of cardiac
contraction on coronary blood flow. Circulation 15: 14,
'957- 34'-
324. Sabiston, D. C, Jr., J. P. Fauteux, and A. Blalock.
An experimental study of the fate of arterial implants in
the left ventricular myocardium. With a comparison of 342
similar implants in other organs. Ann. Surg. 145:927, 1957.
325. Sabiston, D. C, Jr., C. A. Neill, and H. B. Taussig.
The direction of blood flow in anomalous left coronary
artery arising from the pulmonary artery. Circulation 22 : 343.
591, i960.
326. Sabiston, D. C. Coronary endarterectomy. Am. Surgeon
26: 219, i960. 344.
327. Salazar, A. E. Induction of coronary thrombosis in the
intact closed chest dog. Circulation Research 9: I 351, 196 1.
328. Salisbury, P. F., C. E. Cross, and P. A. Rieben. Reflex 345
effects of left ventricular distention. Circulation Research
8:530, i96°- 346-
329. Salisbury, P. F., C. E. Cross, K. Katsuhara, and P. A.
Rieben. Factors which initiate or influence edema in the
isolated dog's heart. Circulation Research 9: 601, 1961.
330. Sapirstein, L. A. Regional blood flow by fractional dis- 347.
tribution of indicators. Am. J. Physiol. 193: 161, 1958.
331. Sarnoff, S. J., R. B. Case, P. E. Waithe, and J. P.
Isaacs. Insufficient coronary flow and myocardial failure 348.
as a complicating factor in late hemorrhagic shock. .4m.
./. Physiol. 1 76 : 439, 1 954. 349.
332. Sarnoff, S. J., R. B. Case, E. Berglund, and L. C.
Sarnoff. Ventricular function. V. The circulatory effects
of aramine; mechanism of action of "vasopressor" drugs 350.
in cardiogenic shock. Circulation 10: 84, 1954.
331 Sarnoff, S. J., R. B. Case, and R. Macruz. Observa- 351
tions on the vasodilator properties of urine. I. Comparison
of the effect of human urine and nitroglycerin on coronary
resistance and myocardial oxygen consumption on the
isolated supported heart preparation. Circulation Research
6:5"- I958
Sarnoff, S. J., E. Braunwald, G. H. Welch, Jr., R. B
Case, W. N. Stainsby, and R. Macruz. Hemodynamic
determinants of oxygen consumption of the heart with
special reference to the tension-time index. .4m. J. Physiol.
192: 148, 1958.
Sarnoff, S. J., R. B. Case, G. H. Welch, E. Braunwald,
and W. N. Stainsby. Performance characteristics and
oxygen debt in a non-failing metabolically supported
isolated heart preparation. .4m. J. Physiol. 192: 141, 1958.
Sarnoff, S. J., S. K. Brockman, J. P. Gilmore, R. J.
Linden, and J. II. Mitchell. Influence of cardiac sym-
pathetic and vagal nerve stimulation on atrial and ven-
tricular dynamics. Circulation Research 8: I 108, i960.
Sayen, J. J., A. H. Katcher, W. F. Sheldon, and C. M.
Gilbert, Jr. Effect of levarterenol on polarographic
myocardial oxygen, the epicardial electrocardiogram and
contraction. Circulation Research 8: log, i960.
Schlesinger, M. J. An injection plus dissection study of
coronary artery occlusions and anastomoses .4m. Heart J.
>5: 528, 1938.
Schreiner, G. L., E. Berglund, H. G. Borst, and G
Monroe. Effects of vagus stimulation and of acetylcholine
on myocardial contractility, oxygen consumption and
coronary flow in dogs. Circulation Research 5: 562, 1957.
Scott, J. G, and T. A. Balourdas. An analysis of coro-
nary flow and related factors following vagotomy, atropine
and sympathectomy. Circulation Research 7: 162, 1959.
Scott, J. B., R. A. Hardin, and F. J. Haddy. Pressure-
flow relationships in the coronary vascular bed of the
dog. .4m. ./. Physiol. 199: 765, i960.
Scott, J. G, T. A. Balourdas, and M. N. Cross. The
effect of experimental hypothyroidism on coronary blood
How and hemodynamic factors. Am. J. Cardiol. 7 : 690,
1 96 1.
Sevelius, G., and P. C. Johnson. Myocardial blood flow
determined by surface counting and ratio formula. J.
Lab. Clin. Met/. 54:669, 1959.
Shaw, R., C. R Rayford, and D. E. Gregg. Patterns of
phasic blood flow in the left coronary artery. Physiologist
2 (No. 3): 105, 1959.
Shipley, R. E., and C. Wilson. An improved recordine
rotameter. Proc. Soc. Exptl. Biol. Med. 78: 724, 1951.
Siegel, J. H., J. P. Gilmore, and S. J. Sarnoff. Cate-
cholamines in coronary venous blood before and during
stimulation of the stellate ganglion. Federation Proc. 19:
108, i960.
Simonson, E. Clinical progress. Russian research on the
role of visceral reflexes in coronary insufficiency. Circula-
tion 22 : 1 179, i960.
Singer, R. The coronary arteries of the Bantu heart. S.
African Med. J. 33: 310, 1959.
Smith, J. R., and I. C. Layton. The flow of blood supply-
ing the cardiac atria. Proc. Soc. Exptl. Biol. Med. 62 : 59,
1946.
Smith, J. C. Review of single coronary artery with report
of two cases. Circulation 1:11 68, 1 950.
Sobin, S. S., W. G. Frasher, Jr., and H. M. Tremer.
BLOOD SUPPLY TO THE HEART
1583
Vasa vasorum of the pulmonary artery of the rabbit.
Circulation Research 11 : 257, 1962.
352. Sones, F. M., Jr. Cinecardwangiography . Clinical Cardio-
pulmonary Physiology. New York: Grune & Stratton, i960, 370.
pp. 130-144.
353. Stainsby, W. N., and E. M. Renkin. Autoregulation of
blood flow in resting skeletal muscle. Am. J. Physiol. 201 :
117. '96'- 37'-
354. Starling, E. H., and M. B. Visscher. The regulation of
the energy output of the heart. J. Physiol., London 62:
243. '927-
355. Starzl, T. E., and R. A. Gaertner. Chronic heart
block in dogs. A method for producing experimental 372.
heart failure. Circulation 12: 259, 1955.
356. Steinberg, I., J. S. Baldwin, and C. T. Dotter. Coro-
nary arteriovenous fistula. Circulation 17: 372, 1958.
357. Stuckey, J. H., M. M. Newman, C. Dennis, E. H. Berg,
S. E. Goodman, C. C. Fries, K. E. Karlson, M. Blumen- 373.
field, S. W. Weitzner, S. W. Binder, and A. Winston.
The use of the heart-lung machine in selected cases of
acute myocardial infarction. Surg. Forum 8: 342, 1958. 374.
358. Szentivanyi, M., and A. J. Nagy. A new aspect of the
nervous control of the coronary blood vessels. Quart. J.
Exptl. Physiol. 44:67, 1959.
359. Tennant, R., and C. J. Wiggers. Effect of coronary 375.
occlusion on myocardial contraction. Am. J. Physiol. 112:
35". '935-
360. Tepperman, J., and D. Pearlman. Effects of exercise and 376.
anemia on coronary arteries of small animals as revealed
by the corrosion-cast technique. Circulation Research 9 :
576, '961- 377
361. Theilen, E. O., M. H. Paul, and D. E. Gregg. A com-
parison of effects of intra-arterial and intravenous trans-
fusions in hemorrhagic hypotension on coronary blood 378
flow, systemic blood pressure and ventricular end-diastolic
pressure. J. Appl. Physiol. 7 : 248, 1 954.
362. Thornton, J. J., and F. R. Mautz. Experimental
methods for producing chronic, progressive, coronary 379
arterial occlusion. Am. Heart J. 19: 404, 1940.
363. Thurau, K., and K. Kramer. Die Reaktionweise der
glatten Muskulatur der Nierengefasse und Dehnungeriese
und ihre Bedeutung fur die Autoregulation des Nieren- 380.
kreislaufes. Pflugers Arch. Ges. Physiol. 268: 188, 1959.
364. Travell, ]., S. H. Rinzler, and D. Karp. Cardiac
effects of nicotine in the rabbit with experimental coronary 38 1 .
atherosclerosis. Ann. N.Y. Acad. Sci. 90: 290, i960.
365. Truex, R. G, and M. J. Schwartz. Venous system of
the myocardium with special reference to the conduction 382,
system. Circulation 4: 881, 1 95 1.
366. Truex, R. C, and A. W. Angelo. Comparative study
of the arterial and venous systems of the ventricular myo- 383
cardium with special reference to the coronary sinus.
Anat. Record 113: 467, 1952.
367. Ullrich, K. J., G. Riecker, and K. Kramer. Das
Druckvolumdiagrams des Warmbluterherzens. Iso- 384
metrische gleichgewichtskurven. Pflugers Arch. Ges. Physiol.
259:48l, 1954- 385
368. Uvnas, B. Central cardiovascular control. In: Handbook
of Physiology, edited by J. Field and H. W. Magoun.
Washington, D.C. : Am. Physiol. Soc, 1960, Sect. 1, Vol.
11, p. 1 131 . 386.
369. Van Citters, R. L., W. E. Ruth, and K. R. Reismann.
Effect of heart rate on oxygen consumption of isolated
dog heart performing no external work. Am. J. Physiol.
■9I:443. 1957-
Vasko, J. S., and D. C. Sabiston. A study of predomi-
nance of human coronary arteries determined by arterio-
graphic and perfusion technics. Am. J. Cardiol. 8: 379,
1 96 1.
Vastesaeger, M. M., P. P. Van Der Straeten, J.
Friant, G. Caudaele, A. Ghys, and R. M. Bernard.
Les anastomoses intercoronariennes telles qu'elles ap-
paroissent a la coronarographie postmortem. Acta Cardiol.
■2: 365. '957-
Vidone, R. A., J. L. Kline, M. Pitel, and A. A. Liebovv.
The application of an induced bronchial collateral circu-
lation to the coronary arteries by cardiopneumonopexy.
II. Hemodynamics and the measurement of collateral
flow to the myocardium. Am. J. Pathol. 32: 897, 1956.
Vineberg, A., and G. C. McMillan. The fate of the
internal mammary artery implant in the ischemic human
heart. Diseases of Chest 33 : 64, 1 958.
Vineberg, A., and T. D. Deliyannis. Myocardial nu-
trition after the Ivalon sponge operation. The return of a
400 million year old septum. Can. Med. Assoc. J. 80: 948,
'959-
Vineberg, A., B. Mahauti, and J. Litvak. Experimental
gradual coronary artery constriction by aneroid con-
strictors. Surgery 47 : 765, 1 960.
Von Euler, U. S. Presence of a sympathomimetic sub-
stance in extracts of the mammalian heart. J. Physiol.,
London I 05 : 38, I 946.
Wang, H. H., C. W. Frank, D. M. Kanter, and R.
Wegria. An experimental study on intercoronary re-
flexes. Circulation Research 5: 91, 1957.
Wang, H. H., M. R. Blumenthal, and S. C. Wang.
Effect of efferent vagal stimulation on coronary sinus
outflow and cardiac work in the anesthetized dog. Circu-
lation Research 8: 271, i960.
Wartman, W. B. Factors concerned in narrowing or
occlusion of coronary vessels. In: Blood, Heart and Circula-
tion, Washington, D.C: Am. Assoc. Advance. Sci. Publ.
'3- '94°. P- 122.
Wartman, W. B., L. A. Campbell, and R. L. Craig.
The effect of ACTH on experimental myocardial infarcts.
Circulation Research 3: 496, 1955.
Waser, P., and Hunzinger, W. Radiocirculographische
untersuchung des Coronarkreislaufes mit Na24. Cardio-
logia 22:65, '953-
Wearn, J. T. Morphological and functional alterations
of the coronary circulation. Harvey Lectures. 1 939-40, pp.
243-270.
Wegria, R., M. Segers, R. P. Keating, and H. P.
Ward. Relationship between the reduction in coronary
flow and the appearance of electrocardiographic changes.
Am. Heart J. 38: 90, 1949.
Wegria, R. Pharmacology of the coronary circulation.
Pharmacol. Revs. 3: 197, 1 95 1 .
Wegrla, R., C. W. Frank, G. A. Misrahy, H. Wang, R.
Miller, and R. B. Case. Immediate hemodynamic
effects of acute coronary artery occlusion. Am. J. Physiol.
i77: '23. >954-
Wegria, R., G. Muelheims, J. R. Golub, R. Jreissaty,
and J. Nakano. Effect of aortic insufficiency on arterial
1584
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
blood pressure, coronary blood flow and cardiac oxygen 399.
consumption. J. Clin. Invest. 37: 471, 1958.
387. Wegria, R . C. W. Frank, H. Wang, and J. Lammer-
ant. The effect of atrial and ventricular tachycardia on 400.
cardiac output, coronary blood flow and mean arterial
blood pressure. Circulation Research 6: 624, 1958.
388. Wegria, R., G. Muelheims, R. Jreissatv, and J. Na- 401.
kano. Effect of acute mitral insufficiency of various de-
grees on mean arterial blood pressure, coronary blood
flow, cardiac output and oxygen consumption. Circulation 402.
Research 6: 301, 1958.
389. Wegria, R., Nakano, J., J. C. McGiff, D. F. Roch- 403.
ester, M. R. Blumenthal, and T. Muraviev. Effect of
arteriovenous fistula on mean arterial blood pressure, 404.
coronary blood flow, cardiac output, oxygen consump-
tion, work and efficiency. Am. J. Physiol. 193: 147, 1958. 405.
390. Welch, G. H., Jr., E. Braunwald, and S. J. Sarnoff.
Hemodynamic effects of quantitatively varied experi-
mental aortic regurgitation. Circulation Research 5 : 546, 406.
'957-
391. Welch, G. H, Jr., E. Braunwald, R. B. Case, and S. J.
Sarnoff. The effect of mephentermine sulfate on myo-
cardial oxygen consumption, myocardial efficiency and 407.
peripheral vascular resistance. -4m. J. Med. 24: 871, 1958.
392. West, J. W., T. Kobayashi, and F. S. Anderson. Effects
of selective coronary embolization on coronary blood flow 408.
and coronary sinus venous blood oxygen saturation in
dogs : With special reference to coronary reflexes. Circula-
tion Research 10: 722, 1962. 409.
393. West, J. W., and S. V. Guzman. Coronary dilatation
and constriction visualized by selective angiography.
Circulation Research 7: 527, 1959.
394. West, J. W., II. Wendel, and E. L. Foltz. Effects of 410.
aortic insufficiency on circulatory dynamics of the dog.
Circulation Research 7: 685, 1959.
395. Wetterer, E. Eine neue Methode zur Registrierung der 411.
Blutstromungsgeschwindigkeit am uneroffneten Gefass.
Z. Biol. 98: 26, 1937.
396. Whalen, W. J. Some factors influencing oxygen con-
sumption of isolated heart muscle. Am. J. Physiol. 198: 412.
"53. '96°-
397. White, J. C. Cardiac pain. Anatomic pathways and
physiologic mechanisms. Circulation 16: 644, 1957. 41 3.
398. Whitehorn, W. V., and W. C. Ullrich. Influence of
thyroid hormone on respiration of cardiac tissue. Am. J.
Physiol. 171: 407, 1952.
Widran, J., and M. Lev. The dissection of the atrioven-
tricular node bundle and bundle branches in the human
heart. Circulation 4: 863, 1951.
Wiggers, C. J. The physiology of the coronary circulation.
In: Diseases of the Coronary Arteries and Cardiac Pain, edited
by R. L. Levy. New York: Macmillan, 1936, pp. 57-109.
Wiggers, C. J. The physiology of pain. In: Diseases of the
Coronary Arteries and Cardiac Pain, edited by R. L. Levy.
New York: Macmillan, 1936, pp. 163-183.
Wiggers, C. J. The Physiology of Shock. New York: Com-
monwealth Fund, 1950, pp. 253-286.
Wiggers, C. J. The problem of functional coronary col-
laterals. Exptl. Med. Surg. 8: 402, 1950.
Wiggers, C. J. The functional importance of coronary
collaterals. Circulation 5: 609, 1952.
Winburv, M. M., and D. M. Green. Studies on the
nervous and humoral control of coronary circulation.
Am. J. Physiol. 170: 555, 1952.
Winburv, M. M., D. H. Papierski, M. L. Hemmer, and
W. E. Hambourger. Coronary dilator action of the
adenine-ATP series. J. Pharmacol. Exptl. Therap. 109: 255,
■953-
Wolf, M. W., and R. M. Berne. Coronary vasodilator
properties of purine and pyrimidine derivatives. Circula-
tion Research 4: 343, 1956.
Woods, E. F., and J. A. Richardson. Effects of acute
anoxia on cardiac contractility. Am. J. Physiol. 196: 203,
'959-
Yankopoulos, N. A., J. O. Davis, E. Cotlove, and
M. Trapasso. Mechanism of myocardial edema in dogs
with chronic congestive heart failure. Am. J. Physiol. 199:
603, i960.
Yonce, L. R., and W. F. Hamilton. Oxygen consump-
tion in skeletal muscle during reactive hyperemia. Am. J.
Physiol. 197: 190, 1959.
Zoll, P. M., S. Wessler, and M. J. Schlesinger. Inter-
arterial coronary anastomoses in the human heart, with
particular reference to anemia and relative cardiac
anoxia. Circulation 4: 797, 1951.
Zoll, P. M., and L. R. Norman. The effects of vasomotor
drugs and of anemia upon interarterial coronary anas-
tomoses. Circulation 6: 832, 1952.
Zakusov, V. V. (editor). New Data on the Pharmacology of
the Coronary Circulation. Moscow: Acad. Med. Sci., U.S.-
S.R., Inst. Pharmacology and Chemotherapy, i960,
vol. 11.
yd
CHAPTER 45
Maternal blood flow in the uterus and placenta1
S. R. M. REYNOLDS Department of Anatomy, University of Illinois College of Medicine, Chicago, Illinois
CHAPTER CONTENTS
Comparative Anatomy of Uteri
Angiogenesis in the Uterus
Vascular Connections of the Uterus
Functional Implications of Venous Drainage
Comparative Anatomy of the Placenta
Types of Placentas
Placental Structure and Placental Exchange
Vascularity and Accommodation of the Products of Conception
Menstruation
Hormones and the Uterine Vasculature
Uterine Contraction and Blood Flow
Body Posture and Uterine Contractility
Estrogen and Uterine Blood Vessels
Uterine Innervation
Pregnancy and the Uterine Circulation
among the several orders of mammals, no organ
in the body is more varied in form and size than the
uterus. One may not properly speak of "the uterus"
as an organ in which identical physiological activities
take place in the fulfillment of the purpose for which
a uterus exists. True, the uterus permits implantation
of fertilized blastocysts, accommodates the products
of conception for a normal span of development, and
then delivers to the outside world an organism or
organisms that can survive. Specialized adaptations
exist among mammals in the form and function of
the various uterine and placental types. Such varia-
tions in uterine and placental structures, coupled
with specialized variations in cyclic activity, serve to
render them quite different from one another while
achieving a common goal, namely, the production
of living offspring.
As a student of physiology, man tends to be an-
1 Prepared and typed under USPHS Grant RG 4728.
thropocentric; he employs many kinds of animals
exhibiting many types of mechanisms but, while ex-
amining comparative basic processes, he hopes to
understand himself. It is necessary, therefore, that
insofar as present knowledge permits, the several
types of form and function be considered.
COMPARATIVE ANATOMY OF UTERI
It is axiomatic in developmental biology that
ontogeny repeats phylogeny. This is to say, as one
passes from species A to species Z there are grades of
morphological complexity that can be seen. Similarly,
in the development of species Z, all or most of the
essential elements of species A, B, C, D, . . . Z are
observable in transition from a simple type to a
complex type of structure. Although this is an over-
simplification of the situation, it is generally true and
it is as easily demonstrable for the uterus as with any
organ in the body [Reynolds (198)].
Like so many viscera of the body, the uterus may be
characterized as starting as paired symmetrical tubes,
part of the mullerian duct system. In monotremes
(e.g., Echidnae), marsupials (e.g., Marsupialae), and
some rodents (e.g., Leporidae) at least, two uteri
remain separate throughout life, arising cephalad at
the caudal end of the fallopian tube and terminating
caudally with independent cervical openings in the
vagina. In other species, the caudal ends of the ducts
fuse mesially to form a single cervical opening in the
vagina. Examples of this are seen in certain rodents
(e.g., Mus rattus and norvegicus), carnivores (e.g.,
Canis), ungulates (e.g., Ovis, Bovis, and Equidae)
and many others. Continuing the extent of mesial
fusion to the ultimate degree, the primates normally
have a single uterus, which receives two fallopian
1585
i586
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
tubes, and a cervix. These several types of uteri are
recognized as the uterus duplex, the uterus bicornis,
and uterus simplex; the latter representing, para-
doxically, the most complicated organization of all,
being, as it is, the fusion of paired simple ducts into a
single complex organ. Just as increasing degrees of
complexitv of organization may be seen throughout
the phyla so, in the development of the uterus simplex,
all the transitional stages of development from the
duplex to the simplex form are recognized. Persistence
of incompletely formed uteri as malformations
sometimes complicates the parturitional process. No
one has produced experimentally arrested fusion or
partial development of the uterus simplex, probably
for the simple reason that no studies of experimental
teratogenesis have been made in primates.
Since the several classes of uteri have in common an
orderly and progressively more complex organiza-
tion, one might anticipate that there would be an
FIG. I. Comparative types of uteri from the uterus duplex to
the uterus simplex found in various mammals: .-1, monotreme
[Echidna aculeala); B, marsupial (Didelphis virginiana); C, rodent
(rabbit); D, carnivore (dog); E, ungulate (mare); /•', primate
(Macacus rhesus). [After Rudolph and Ivy, taken from Reynolds
(198)-]
orderly and progressively more complex organization
of the vasculature of the uterus among different
animals. So there is.
ANGIOGENESIS IN THE LTERUS
All organogeny takes place around a primary
\ ascular organizational pattern [Evans (71)]. Blood
vessels begin as a diffuse capillary network, some
channels of which become more and more prominent,
larger and structurally more complex as arterial and
venous pathways come into being [Thoma (233)].
Why this is so is not clear, although Thoma has
postulated that the process is governed in part by the
hemodynamic load imposed upon certain parts of the
delicate capillary system. As these pathways become
more defined, they give rise to still further differ-
entiation of more peripheral branches. The sizes and
angles of these branches are related to certain physical
relationships that were first laid down by Hess (105)
on thermodynamic grounds, and first given substance
experimentally by Reynolds (197) in the developed
vascular tree. However, more than simple hemody-
namics is involved, since Price (177) has shown that
organogenesis can take place in tissue culture only if
a semisolid medium is used, but not if a liquid medium
is employed. Thus the dependence or role of vascular
development as a contributory mechanism to or-
ganogenesis is seen to be unessential for primary
organization, but to be essential for subsequent devel-
opment.
VASCULAR CONNECTIONS OF THE UTERUS
The common vascular denominator for all uteri is
the pattern of vascular supply of blood to, and drain-
age of blood from, the uterus. This was certainly seen
by Aristotle, by Vesalius, and by Hunter. It was not
stressed as a vascular complex, apparently, until the
early part of this century by Byron Robinson (211).
This author compared in different species the vascular
circle that starts in the aorta by way of the anterior
division of the internal iliac arteries on each side, or
may arise in common with the vaginal, umbilical, or
middle rectal arteries. The uterine arteries descend
in the fat at the base of the broad ligament and, going
between the layers of that ligament, pass to the uterus,
following a tortuous course. They run along the
mesial sides of the uterus, giving off branches to the
bodv of the uterus along the way. At the cephalic end
UTERINE BLOOD FLOW
1587
fig. 2. Arrangements of arte-
rial and venous pathways to
uterus simplex (left) and uterus
duplex (right). The "circle" of
the arterial pathway is shown as a
continuous pathway from the
aorta, ovarian artery, uterine
artery, hypogastric artery (uterus
simplex), femoral artery, and
aorta. A similar circle exists in
the venous connections. [After
Byron Robinson (211).]
of the uterus, the arteries anastomose with a branch of
the ovarian artery, one on each side. The uterine
arteries supply, therefore, part of the vagina, the
uterus, and fallopian tubes on each side.
Since the ovarian arteries arise from the aorta just
below the renal arteries, it will be seen that there is,
indeed, a large communicating arterial circle supply-
ing the uterus on each side of the midline. In the case
of partial or complete uterine fusion during develop-
ment there is further connection of the finer arterial
branches from both sides in the body of the uterus
[Faulkner (77, 78)]. The arcuate arteries of the uterus
lie in the zona vascularis in the myometrium. Myomas
in the smooth muscle of the uterus are singularly
deficient in blood supply (see fig. 3) [Faulkner (77),
Holmgren (1 13)].
We observe in this arrangement that the uterus
simplex is supplied from two primary arterial sources,
on each side, and that where the form of the uterus
permits, there is free union between these. It is possible
to see that the uterus, which increases along with its
blood vessels many times over in size during gestation,
is assured of a reasonably large and constant head of
arterial pressure at all times.
The morphology of the venous drainage of the
uterus is equally important for the physiological
changes that take place in the uterus and its circula-
tion. The uterine veins, without valves, arise from
within the tissues of the uterus and enter the broad
ligament at numerous points in increasingly large
venous channels as smaller ones have united along
the way. However, there are four main venous paths
of exit from the uterus [Bieniarz (31)]. In the common
laboratory animals, the veins of the broad ligaments
unite and form rather uncomplicated plexuses in the
broad ligament [Reynolds ( 1 98)] receiving veins from
the uterus along the way. The parametrial veins join
with ovarian veins to drain blood toward their point of
entrance into the inferior vena cava on the right side
and the renal vein on the left side.
As with many parts of the venous system, the
drainage connections are complex, rather than dia-
grammatically simple as is commonly believed. In
the primate, these relations are more complicated
than in the usual laboratory animals. For example, in
the broad ligament of primates, there is an extensive
pampiniform plexus, having multiple connections
with the pudendal veins, the several rectal plexuses,
the internal iliac veins, and inferior vena cava. Not
only may uterine blood move toward the heart
through the inferior vena cava, but through the
internal and superficial epigastric veins to the mam-
mary and internal costal veins as well. Blood may
also, if pressure within or upon the venous system
requires it, flow to the ascending lumbar veins, by
way of segmental connections, to the azygous and
hemiazygous veins.
The ovarian vein enlarges greatly during pregnancy
[Borell & Fernstrom (38), Hodgkinson (no)]. There
is much current interest in this subject, first stressed
i588
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 3. Arterial pattern of uterus. Cleared preparation. One
small myoma on right side. [From Faulkner (77).]
by Davidsohn (64) many years ago. Further assurance
of venous return exists by way of multiple venous
connections to the vertebral vein system, as Batson
(27, 28) has emphasized (see also Jeffcoate (117).
The ovarian "vein" in part is in reality the tubo-
ovarian pampiniform plexus which pours blood into
the inferior vena cava from the right side, or the
renal vein on the left. Other interconnections exist,
also. However, sudden occlusion of the inferior vena
cava can cause separation of the placenta [Mengert
et al. (150)]. Donnelly (66) relates noninduced pla-
cental separation to gross abnormalities of the
placenta.
All in all, within the uterus itself and in the venous
systems of the abdominal cavity and body wall, there
are abundant intercommunications, so that oppor-
tunity for obstruction to the venous drainage of uterine
blood is minimized. Barcroft & Rothschild (21)
emphasized this with regard to the rabbit; Bieniarz
(31) has stressed it in relation to the human. Oughtred
& Reynolds (165) demonstrated the operation of the
collateral abdominal and somatic venous systems in
the dog when the inferior and superior vena cava
were blocked at various levels.
FUNCTIONAL IMPLICATIONS OF VENOUS DRAINAGE
Problems that arise from malfunctioning of the
venous system are recognized. In general, they are
twofold. In animals having an erect posture, at least
in the human, pelvic congestion which is correctable
by operative procedures is known [Taylor (230),
Curtis et al. (58)]. These contribute to endocrine
disorders and a variety of clinical entities. The second
group of functional disorders is related to distribution
of the vascular loads upon the circulatory system in
late pregnancy [Bieniarz (30)]. When the placenta in
the human is implanted high in the uterus, drainage
by way of the ovarian pathways predominates. When
this happens, albuminuria, hypertension, and even
toxemia frequently occur. A continuous discharge of
several hundred milliliters of blood per minute into
the vena cava or renal vein may complicate renal and
adrenal blood flow, Bieniarz (31) postulates. Ligation
of the vena cava above the renal veins affects renal
function and possibly adrenal gland activities as well
[Karaev (126)]. When, on the other hand, the pla-
centa is implanted low in the uterus simplex, placenta
praevia and hemorrhage more commonly occur
[Bieniarz (30)]. The former condition is more frequent
in primigravidas, the latter, in multiparas.
Although ovarian vein physiology has been im-
plicated by deduction in the incidence of toxemia
[Bieniarz (30)], other mechanisms are suspect, also.
Placental ischemia as a factor is considered by Page
(168). Saito (214) produced toxemic signs in animals
with human placental extracts and this is said to be
an allergic reaction by Lin ( 136) who sensitized rats to
placental tissue by injection of placental tissue 5
months before. However, killing of the fetuses in
hypertensive rats leads to lowering of blood pressure
[Page (166)]. Examining the problem experimentally,
Ogden et al. (162) placed Goldblatt clamps on uterine
arteries in pregnant rabbits and observed that pro-
gressive hypertension developed promptly. This was
relieved by removal of the clamps. Grollman (87)
observed that induced hypertension in rats is reduced
by normal pregnancy, but not by pseudopregnancy.
(See below discussion of the placenta as an A-V
fig. 4 (facing page). The uterine drainage system during pregnancy (sagittal view). Different
visceral and parietal venous drainage routes are shown in different colors. [Reprinted by permission
from Bieniarz (31).]
Cranial caval vein
Thoracoabdominal network
Intercostal veins
Internal mammary vein
Caudal caval vein
Portal vein
Epigastric superficial vein "^Ka%
Epigastric internal vein
\ Dural and cerebral sinuses
')
J)
y
SI
J^V- Vertebral vein system
Segmental communications
,IAJ7 " rt8 I Thoracic longitudinal
lf.Ql left \ veins
Tubo-ovarian pampiniform
plexus
S}\ Caudal caval vein
Ascending lumbar vein
Mesenteric inferior
^t\ VI -A Sacral vein
\WJjjf~ |~ " ~ Superior rectal
Medial rectal j plexuses
Inferior rectal I
Pudendal vein
Caudal caval system.
Uterovenal visceral circulation.
Portal circulation.
The anterior parietal abdomino-thoracal communications.
Retroperitoneal and retropleural communications to the cranial caval system.
Vertebral vein system to the cerebral sinuses.
fig. 4. See legend on facing page.
UTERINE BLOOD FLOW
I 591
Vertebral plexus +0 brain
anas+omosi
por ro-ovaric
«
MATERNAL SIDE
EPITHELIO-
CHORIAL
SYNDESMO
CHORIAL
ENDOTHELIO-
CHORI AL
HEMO-
CHOR IAL
UTERINE BLOOD FLOW
1593
0 4 0.5 0.6 0.7
PERIOD OF GESTATION IN TENTHS OF TOTAL
-100
fig. 8. Transfer rate of sodium during gestation in six species. [From Flexner & Gellhorn (81).]
ventral, the other on the dorsal aspect of the interior
of the uterus. However, only one of these receives the
umbilical vessels and is called, therefore, the primary
placenta. The secondary placenta receives its vessels
from continuations of a number of umbilical vessel
branches on the chorioallantoic surface of the primary
placenta that pass between the amnion and chorion
laeve to the secondary placenta, rather like the con-
nection between the cotyledons of the ungulate
placenta.
The "unitary" structure of the fully developed
discoidal placenta, of the human, rhesus monkey,
and other species is the cotyledon [Wilkin (245)].
This is a vascular unit which is fetal (see further
discussion below). By implication, the maternal
placental structures are fitted to the cotyledon. In a
sense they are. The decidual plate with its septa blocks
off smaller areas of the fetal portion of the placenta as
ridges or folds of tissue about a number of the cotyle-
dons. However, the ridges of the septa do not make
connection with the fetal tissues in a way that makes
discrete, unitary compartments, or chambers. The
maternal vascular compartments interconnect deep
in the placenta beneath the chorioallantoic plate of
the placenta, so that there is a continuum of the
maternal vascular area. The entire area is spoken
of as a maternal lake or intervillous space. To the
extent that there is continuity, this is true, but there
are innumerable attachments of fetal vessels covered
with chorionic tissue to the basal part of the placenta
in the lake on the decidual plate and on the septa
[Wilkin (245)]. The crypts or pockets between septal
folds become more numerous as pregnancy advances.
In early pregnancy, isolated lakes of maternal blood
in the trophoblast merge and fuse as one may imagine
pockets of gas in aging cheese might fuse to form
larger pockets. This involves the entire syncytial
trophoblast in the area of implantation, embedded in
the decidua basalis. As the placenta enlarges and
undergoes morphogenesis and the uterus enlarges
pari passu, the characteristics of the placenta change
until it is complete in form, after the fourth month of
pregnancy. Only a small proportion of the interior of
the uterine surface then is involved. Zonary placentas
completely surrounding the fetus are found in some
species. These and still other patterns of gross struc-
ture and vascular arrangements have been described
by Mossman (158) and by Amoroso (5).
Because there is such a variety in the types of
placentas and since these undergo important struc-
tural changes during pregnancy, a physiologist who
measures uterine blood flow during pregnancy may
not properly speak of blood flow except as the latter
relates to an evolving set of morphological relation-
ships with respect to a given form of placenta. It is
to be borne in mind also that the largest part of the
gravid uterus is not associated with the placenta; it,
too, must be supplied with blood. There are, in a
sense, two uterine blood flows, one to the placenta in
its various forms, the other to the uterine tissues which
are undergoing enormous growth, stretching and
change of shape. To measure total uterine blood flow
1594
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 9. Circumferential arteries of nonpreg-
nant rabbit uterus. These anastomose with
each other laterally and supply the uterine
wall. [From Reynolds (199).]
tells what the circulatory load of the gravid uterus
is on the maternal cardiovascular system but it fails
to tell how this is related to supplying fetal needs,
on one hand, and uterine tissue needs, on the other.
among many types of placentas. In this way, it
becomes possible to account for the occurrence of
fetal erythrocytes in the maternal blood and vice
versa (see below).
PLACENTAL STRUCTURE AND PLACENTAL EXCHANGE
Flexner and associates [Flexner & Gellhorn (81),
see Reynolds (198)] have found wide differences in
the rates of transfer of given electrolytes across various
types of placentas. Thus, in the ninth decile of gesta-
tion the transfer of Xa is 0.026 mg per hour per gram
of placenta in the so-called epitheliochorial placenta
and 6 to 8 mg per hour per gram of placenta in the
hemochorial placenta. Intermediate rates of transfer
are found in the syndesmochorial and endothelio-
chorial placentas. Although it is believed by some
that Grosser's classification of placental types has
outlived its usefulness [cf Amoroso (6)], it is possible
that the classification expresses a general tendency
toward morphological organization and functional
capability that is not entirely negated by dwelling on
details of isolated microscopic fields either in the
several types of placentas or in any one placenta. One
may accept the fact that diere are differences in
morphology, but these are not sharply defined either
within one placenta at various stages of pregnancy or
VASCULARITY AND ACCOMMODATION OF THE
PRODUCTS OF CONCEPTION
The adaptation of the vasculature of the nonpreg-
nant uterus to the changes of gravidity is met in
various ways in the several types of uteri: through
growth and enlargement of the blood vessels [Orsini
(164), Wermbter (241), Schwarz & Hawker (218)]
and by physical rearrangement of blood vessels as
the uterus enlarges and changes the spatial orienta-
tion of the uterine blood vessels, especially as the
uterus is finally stretched in the latter part of preg-
nancy. In all nonpregnant uteri numerous arteries
are tortuous, coiled, or undulating [Ramsey (180)].
These tortuosities in the blood vessels permit their
extension to accommodate in part the increase in
size of the uterus. In sheep, one or two of the uterine
arteries approach each cotyledon. The vessels divide
into five or six trunks and pursue a tortuous course
in the submucosa before dividing again and entering
the cotyledon [Barcroft & Barron (19)]. The early
pattern of these structures seems not to change during
fig. 10 (facing page). Distribution of fetal arteries and veins to cotyledons of placenta of Pere
David's deer. Note that arteries and veins are of about equal size and number, indicating about
equally rapid flows of blood in them. 1. Three injected cotyledons. Nearly natural size. 2. Section
through middle of placentome. Masson stain. Natural size. 3. Drawing of three injected villi, re-
moved from placentome. Central vessels of the villus and the intraepithelial capillary network are
shown. 4. Section through fetal zone of placentome showing stem villus and to vessels. X50. Left
and right, thin strands of maternal connective tissue from which epithelium is removed. [From Harri-
son & Hamilton (95). Courtesy Cambridge University Press.]
fig. 10. See legend on facing page.
UTERINE BLOOD FLOW
'597
pregnancy except that the uterus is distended as
pregnancy advances and increases steadily.
Harrison & Hamilton (95) have demonstrated
especially well the relation of maternal and fetal
blood vessels to each other in a deer. (See below for
discussion of the fetal blood vessels in the hemochorial
placenta.)
In the monkey, as the endometrium becomes
thinner because the uterus is distended by the con-
ceptus, the arteries become extended and the number
of arterial connections with the intervillous spaces
increases by development of smaller branches. Sub-
sequently, the ends of adjacent vessels coalesce to
form terminal dilations from which a single opening
with a large accumulation of lining cells passes
through the basal decidual plate into the maternal
lake of blood [Ramsey (183)]. In the human, similar
arrangements exist [Lundgren (137)]. Arterio-arterial
shunts in the uterus exist among these vessels [Heckel
& Tobin (100), Reynolds (199)]. As pregnancy
advances, the number of arterial openings into the
villous lake decreases substantially. Uterine blood
vessels, along with all other tissues of the uterus, grow
by hypertrophy and hyperplasia of their component
parts during pregnancy [Reynolds (198)]. The cause
of this is partly hormonal, partly the result of disten-
tion of the tissues. Hormones and distention interact
to effect the growth response of the uterus during
pregnancy (198).
For many reasons, therefore, the physiologist who
would study blood flow would be well advised to
appreciate the complexity of structural and functional
factors involved in the uterus in different functional
states.
The adaptation of the uterine vasculature to the
uterus during pregnancy has been especially well
studied in the rabbit [Reynolds (199)], which is
typical of the uterus duplex and, with some modifi-
cations, to the uterus bicornis. A certain parallel
exists also with changes that occur in pregnancy in the
uterus simplex [Ramsey (187)].
In the rabbit, the vascular system of the uterus
consists of large vascular channels which intercom-
municate freely, both longitudinally along the meso-
metrial border of the uterus and circularly along the
length of the uterus. This is well shown by Orsini
(164) in the hamster. Of these large channels, the
mesometrial arcuate vein is concerned primarily with
draining the region of the uterus to which placental
sites are attached. The lateral arcuate veins on each
side of the uterus drain the larger part of the uterine
wall. Each of these vascular beds is supported by the
same incoming arteries and drained by the veins of
the broad ligament. The implication of such an ar-
rangement is clear. If, due to distention, the blood
flow is reduced in one area (i.e., the peripheral vas-
cular resistance increases), the flow of blood to the
ARTERIES
MARGIN OF INTERVILLOUS SPACE
■
A
i
/
TERMINAL SAC
ARCUATE ARTERY
PERITONEAL SURFACE
[B] 6th WEEK
a
3-d WEEK
MARGIN OF INTERVILLOUS SPACE
BASE OF ENOOMETRIU
PERITONEAL SURFACE
5^
A
MARGIN QF INTERVILLOUS SPACE ,
SPIRAL ARTERY
(PARTIALLY
UNCOILED)
[D]l5'h WEEK
SP pal ajjter. [uncoiled;
MARGIN OF INTERVILLOUS SPACE
6th WEEK
ARCUATE ARTERY
ff]l9"> WEEK
fig. 1 1 . Pattern of arterial supply to basal plate of the monkey placenta at different stages of
pregnancy. [Permission of Ramsey (181).]
1598 HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
VEINS
ARCUATE VEIN
fig. 12. Pattern of veins draining basal plate of placenta in the monkey at various stages of
pregnancy. [Permission of Ramsey (181).]
other vascular bed increases so that it receives a larger
portion of the incoming arterial blood to the uterus.
MENSTRUATION
The morphological and physiological mechanisms
of menstruation have been the subject of considerable
interest, particularly since precise knowledge of
uterine rhythmic cycles was first established some
forty years ago. Reviews on the subject in a modern
context began to appear when Hartman (96) dis-
cussed the subject of intermenstrual bleeding.
Bartelmez (23) gave us the first comprehensive re-
view of the subject, however, and this was extended
and revised in the light of later information by
Reynolds (196), by Smith & Smith (222) and by
Kaiser (124). Since that time, remarkably little
attention has been directed to the problem.
In the uteri of certain but not all primates there are
numerous coiled arterioles in the endometrium [Daron
(62), Dalgaard (59), Kaiser (121), Bartelmez (26)]
which are demonstrable in conventional tissue sec-
tions [Kaiser (121)]. These arterioles undergo in-
crease in coiling throughout the menstrual cycle,
reaching maximum development prior to menstrua-
tion [Daron (62), Kaiser (121)]. Van Wagenen (237)
notes that it is vasoconstriction of these vessels [not
coiling, as commonly supposed (61)] which causes
ischemia. Then, due to hormonally induced changes
within the tissues [Smith & Smith, (222)], there is
loss of tissue fluid and thinning or regression of the
endometrium, and this results in congestion and
stasis of blood in the coiled vessels [Markee (144)].
Subsequently, the superficial layers of the en-
dometrium degenerate, slough off in an irregular but
spreading manner as menstruation takes place.
Sloughing begins when the endometrium is about
one-half of the peak thickness prior to the end of the
cycle. Bleeding is by capillary seepage, by reflux
venous hemorrhage [Markee (144), Daron (63),
Bartelmez (22)], and occasionally by brief arterial
spurting of blood [Markee (144)]- Menstruation
begins in localized areas and extends to others to
involve the entire area [Strassman (228), Phelps
( 1 72)]. A role of arteriovenous shunts in the menstrual
process is both alleged [Schlegel (216), Dalgaard
(59)] and denied [Bartelmez (24)], but the evidence
at hand seems to favor the former view [see Hertig
& Rock (104)]. The physiological effect of coiling the
arterioles of the uterus can hardly be different from
that of coiling of vessels in the ovary and the testicle.
Here, it lowers the blood pressure to the tissues beyond
the coil [Reynolds (200), Waites & Moule (240)].
Since menstruation has been shown to be asso-
ciated with local hemodynamic changes within the
tissues concerned, some investigators have attempted
to induce profound hemodynamic disturbances with a
view to causing uterine bleeding. This has been done
[Van Wagenen & Zuckerman (238), Markee et a!.
(145)] but not invariably so [Emmel et al. (69)]. The
effect of cord transection depends upon a proper
effective estrogen level.
The critical factors in determining what manipula-
UTERINE BLOOD FLOW
'599
tions or hormone treatments will produce menstrua-
tion are related to the nature of the vasculature of the
endometrium. This is affected by the previous
menstrual history of the organism since each cycle
brings about some residual changes through growth of
the vascular tree which persist to affect the next
cyclic bleeding [Phelps (173)].
In addition to the idea that changes in the uterine
vasculature affect and modify the menstrual process,
Smith (220) and Smith & Smith (221) hold to the
view that a toxic substance is produced within the
endometrium, secondary to the premenstrual
ischemia, and that this toxin leads to the breakdown
of the tissues and the ensuing menstrual discharge.
Prostigmin, a vasodilating substance, has been
shown to cause uterine bleeding in nonpregnant
women [Soskin et al. (223)], but not in pregnant
women. Kaiser (123) failed to observe a similar
response in the rhesus monkey. This drug also failed
to affect estrogen-induced hyperemia in endometrial
ocular transplants in the rabbit [Kaiser (125)].
In the nonpregnant endometrium of both rabbits
and monkeys, there are rhythmic constrictions and
dilations of the minute vessels which are independent
of the nervous system [Markee (142, 144)]. Under the
influence of estrogen there is persistent hyperemia of
the endometrium [Pompen (176), Markee (143)].
The significance of the rhythmic vascular changes,
both as to cause and as to function, are unknown.
They are, apparently, unique to endometrial vessels,
although estrogens do have profound effects on
somatic minute vessels [see Reynolds (198)], especially
integumentary, in rabbits and humans [Reynolds &
Foster (206, 208)], and in the nasal mucosa [Mac-
kenzie (139)], as well. The retinal circulation is also
modified in women, manifesting itself by scotomata
that change in position with change in posture [Evans
(72)]; this is marked in the last half of the menstrual
cycle.
Of importance to the vascular architecture in the
endometrium is the seldom emphasized fact that the
tissue in which these structures lie is loose and spongy.
The vascular elements are developed out of all pro-
portion to the immediate vascular needs of the tissue
[Reynolds (196)]. It is clear that the vascular arrange-
ment is adapted to the future needs of supplying and
invading the implanting trophoblast [Hasner (98),
Bartelmez (25)]. This instance is not unique in
developmental biology where nature has repeatedly-
contrived to anticipate future needs by prior organi-
zation of mechanisms. When a trophoblast fails to
develop, the complex vascular structure ot the endo-
metrium breaks down since it cannot be maintained
in the face of the requirements of cyclic endocrine
activity in which ovulation is the focal point of the
pattern. This endometrial cycle occurs, even though
ovulation may not occur. Moreover, most uteri,
even including those of some primates [Kaiser (120),
Hamlett (93), Goodman & Wislocki (85)], do not
manifest endometrial sloughing even though they
exhibit microscopic bleeding; instead, the endometrial
vessels undergo an ebb and flow of cyclic growth and
regression unaccompanied by profound menstrual
process. In either event, the local vasculature changes
cyclically. In some species, such as rats, hamsters,
and guinea pigs, the cyclic occurrence of localized
areas of hyperemia within the uterus is evident (see
below) .
HORMONES AND THE UTERINE VASCULATURE
The endometrial hyperemia, indeed the entire
uterine hyperemia, that occurs periodically has a
metabolic basis under endocrine control. Estrogen
augments the amount of acetylcholine found in the
uterus [Reynolds (193, 203)] and in the nasal mucosa
as well [Reynolds & Foster (207)]. However, it was
later found that it is the change in cholinesterase
which accounts for this [Everett & Sawyer (73),
Herschberg (103)]. It is also reported that the hypere-
mia is associated with alterations in the amount of
histamine or histamine-like substances in the uterus
[Kaiser (122)]. It appears that one can only say that
there is a change in vasoreactive tissue constituents
under the influence of estrogen and it is probable that
more than one substance is involved.
UTERINE CONTRACTION AND BLOOD FLOW
The circulation in the uterus, like that in all mus-
cular viscera, functions in the face of contraction and
relaxation of the muscular components of the organ.
It must serve with great efficiency as the uterus under-
goes great change in size and shape during pregnancy.
The consequences of contractions, growth, and dis-
tention upon blood flow in the uterus require con-
sideration. Certainly, clamping of the blood supply
to the uterus elicits uterine contractions, as Rorhrig
showed many years ago [see Reynolds (198)]. In
this respect, myometrium is no different from intestine
or other smooth muscle. More delicate, however, is
the observation that low arterial blood pressure is
associated with an increase in frequency and ampli-
tude of uterine contractions [Kunisima (132), Robson
i Goo
HANDBOOK OF PHYSIOLOGY
CIRCULATION
fig. 1 3. Schematic representation of arterial
supply to portions of uterus simplex (monkey,
human). [From Reynolds (.196).]
Spiral Artery
Venous Lake
Gland
Capillary bed
Basal Artery
Radial Artery
Arcuate Artery.
Arcuate Vein
Peripheral Artery
Functtonal
ENDOMETRIUM
Basal
Myo-Endometnal
border
MYOMETRIUM
& Schild (212)], while an induced higher blood
pressure has an opposite effect. Ahlquist & Wood-
bury (2) found in cats that when intrauterine pressure
reaches 60 to 70 mm Hg, uterine blood flow virtually
ceases. This is reminiscent of the report by Moir
(156) that when intrauterine pressure exceeds arterial
blood pressure, a woman feels ischemic uterine pain.
It may be that myometrial smooth muscle acts in
concert with that of the uterine blood vessels them-
selves. In myometrial studies, adrenaline and nor-
adrenaline cause decreased uterine blood flow in
rabbits and guinea pigs, associated with uterine con-
tractions [Dornhorst & Young (67)], but an action on
uterine blood vessels was not eliminated. The conse-
quence of strong uterine contractions on the systemic
circulation are shown by the fact that undulatory
changes in arterial blood pressure occur as the post-
partum uterus contracts [Franklin (82)].
For many years, speculation existed concerning
the effect of uterine contractions on the flow of blood
in the placenta of the human. Two possibilities
existed: a) that the contraction squeezes blood out of
the placenta as water may be squeezed out of a
sponge [Kermauner (128), Grosser (89, 90)], and
b) that as the intervillous space pressure builds up,
veins are at first occluded, then pressure increases in
the intervillous space as it increases in the amniotic
cavity [Keiffer (127), Wagner (239), Pryztowski
(179)]. Meanwhile, blood remains in the placenta to
meet the needs of maternal-fetal exchange during
uterine contraction. There is now no doubt that the
second view is correct. This was suggested indirectly
by the work of Woodbury et al. (250) and shown
clearly by Woodbury et al. (251 ), Alvarez & Caldeyro
Barcia (3), Caldeyro Barcia (50), and by Pryztowsky
(179) in women and by Ramsey et al. (188)
n monkeys.
BODY POSTURE AND UTERINE CONTRACTILITY
Perhaps the most telling observation about the
effect of uterine circulation on uterine contractions is
the observation made in women that a change in
posture modifies the quality of uterine contractility.
When a woman in late pregnancy or in labor lies on
her back, uterine contractions of a given frequency
and intensity (i.e., change of intrauterine pressure)
are seen [Williams (246), Caldeyro Barcia et al. (51)].
When she assumes a semireclining posture, or turns
on her side, the contractions become slower and more
intense. With a view to studying the role of compres-
sion of the inferior vena cava in the recumbent
position, pressures were recorded simultaneously in a
woman in the lower and upper parts of the vena
cava [Caldeyro Barcia et al. (51)]. The weight of the
gravid uterus on the retroperitoneal surface caused a
disassociation of the venous pulse pressures in the two
parts of the vein; with the woman on her side, the
venous pulse waves became synchronous, and the
quality of uterine contractions changed.
ESTROGEN AND UTERINE BLOOD VESSELS
Another indication of the relation between uterine
contractility and uterine blood flow lies in the ob-
servation that, following estrogen withdrawal, the
UTERINE BLOOD FLOW
l60I
EMBARAZO NORMAL DE TERMINO
PARTO INDUCIDO PITOCIN l/V p^ U. por min
EMBARAZO NORMAL DE TERMINO
PARTO INDUCIDO PITOCIN l/V ,-?rn U. por min
DECUBITO DORSAL
mmHg f£\
N.510
^VAjf^VENA
DECUBITO
mmHg
30
200
LATERAL DERECHO
ILIACA
PRESION
AMNIOTICA
VENA CAVA
W SUPERIOR
PRESION
ARTERIAL
minulos
N.510
VENA
ILIACA
PRESION
AMNIOTICA
VENA CAVA
SUPERIOR
PRESION
ARTERIAL
! minutos
fig. 14. Effect of body position on pressure in the iliac vein, superior vena cava during late
pregnancy and arterial blood pressure. .4.- on back. Note bimodal pressure peaks. B: on side. Note
single simultaneous pressure peaks in upper and lower vena cava, synchronous with acme of uterine
contractions and lower systolic blood pressure. [Permission of Caldeyro Barcia el al. (51).]
uterus becomes less hyperemic and gradually loses its
contractility [Reynolds (191)]. Within an hour after
estrogen is injected there is an intense hyperemia
[Markee (142), Pompen ( 1 76)]. Twelve or more hours
later the myometrium becomes active [Reynolds
(190)]. Beginning activity depends upon synthesis of
actomyosin in the uterus [Csapo (57)]; this is related
to a rise of aerobic metabolism of the uterus [MacLeod
& Reynolds (140)]. There seems to be an assumption
that this is solely in the smooth muscle of the myo-
metrium. It is possible, however, that smooth muscle
in the uterine blood vessels is equally estrogen-
dependent; this has not been investigated. Certainly,
with prolonged estrogen withdrawal the blood vessels
of certain parts of the uterine vasculature show a
reversible hyaline degeneration [Okkels & Engle
(163), Kahn & Laipply (.119)]. All parts of the vas-
culature are not equally affected. The very first
effect of estrogen on the uterine vasculature is to
cause capillary dilation [Pompen (176), Fagin &
Reynolds (74)]. Its role in affecting the larger vessels
seems to have attracted very little attention although
stilbestrol raises the arterial pressure in female but not
in male rats [Hill (106)].
When estrogen given to rabbits is combined with
progesterone in relatively massive doses, profound
hyperemia of the uterus cccurs [Gillman (84)].
Extensive sloughing of the endometrium results.
Estrogen alone increases capillary permeability
[Hechter et al. (99)] which is associated during the
first 6 hours with an increase in the relative wet
weight of the uterus of ovariectomized rats. Later the
relative dry weight increases progressively to a maxi-
mum about 24 hours after the injection [Astwood
(13)]-
The mechanism of the estrogen-induced hyperemia
is indicated by the fact that estrogen increases the
acetylcholine-content of rabbit uteri within 1 hour
[Reynolds (193)], and its concentration in the uterus
changes during pregnancy [Reynolds & Foster (205)].
One group of workers failed to confirm the response
in rabbits [Emmens et al. (70)] for unknown reasons.
Even so, estrogen seems to affect the acetylcholine of
the uterus by altering its cholinesterase content
[Herschberg (103), Sawyer & Everett (215), Everett
& Sawyer (73)]. Pompen (176), it will be recalled,
found that the uterus in situ does not become hypere-
mic under estrogen if atropine is administered. Kaiser
(125), however, failed to observe this if the endo-
metrium is transplanted, and without an innervation.
I t >( 12
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Sturgis (229) found that anything which alters
uterine blood flow in the monkey may affect the rate
of fluid formation in the uterine lumen (glandular
secretion?). This lead has never been followed up,
or studied in relation to endometrial physiology or
cytology.
In the absence of large amounts of estrogen, but
not in ovariectomized rabbits or guinea pigs, the
capillaries of the endometrium exhibit a rather rapid
rhythmic blanching and blushing [Markee (141,
142)]. This phenomenon is unrelated to myometrial
activity. Since capillaries lack contractile elements,
it must be, therefore, a manifestation of arteriolar
activity.
UTERINE INNERVATION
There is a rich sympathetic innervation to the uterus
[Reynolds (198), Krantz (130)]. The parasympathetic
innervation is limited, so far as is known, to the region
of the cervix. Despite the nerve supply, a uterus which
is denervated by transplantation to the ventral peri-
toneal wall possesses all the normal nongravid re-
activity of the uterus in situ [Reynolds & Kaminester
(209)]. This is not to say that the innervation is with-
out effect upon the vasculature [Reynolds & Kami-
nester (210)]. Rather, the hormones are independent
of the innervation in their action on the uterus. Cer-
tainly, fright causes vasoconstriction, but it is possible
that this is a hormonal effect of blood-borne epineph-
rine [Markee (142)]. Few data exist which suggest,
much less show, what the normal role of the vasomotor
innervation to the uterus is. The existence of vaso-
motor nerves, however, as entities separate from the
nerves which supply the myometrial smooth muscle
has been amply shown [Medowar (149)]. Vasodilator
fibers may exist since cholinergic sympathetic fibers
to the uterus seem to have been demonstrated, as
well as adrenergic ones [Burn (46)]. Moreover,
atropine has been seen to reduce the hyperemic effect
of estrogen on the uterus in situ [Pompen (176)] but
not in denervated endometrial transplants in the eye
[Kaiser (125)].
PREGNANCY AND THE UTERINE CIRCULATION
Pregnancy imposes an array of special requirements
upon the uterine circulation. These are, as mentioned
above, responses to growth of the conceptus and
spatial adjustments that are of great magnitude.
At the outset of pregnancy, the uterine hyperemia
of estrus gives way to a quality in the circulation in
the uterus prior to implantation and for a time after
which renders the uterus bluish in appearance as if
the circulation were turgid. This is seen with the
uterus in situ [Barcroft & Rothschild (21)] and in
ocular grafts [Neumann (160)]. Aside from the
generalized uterine hyperemia referred to above,
localized hyperemia, more marked in some parts of
the endometrium than in others, has long been known
and suspected to be related to implantation. This was
reported in the human by His (108), Ilitschmann &
Adler (109), Strahl & Beneke (227), Delporte (65),
Teacher (231), Falkiner & Fleming (76), and Wilkin
(242). More refined examination of this in controlled
experiments on animals awaited the work of Bacsich
& Wyburn (15-17) in the guinea pig and more re-
cently in the rat [Williams (246), Holmes & Davis
(112)] and hamster [Orsini (164)]. Perhaps the most
remarkable instance of localized implantation is found
in the South African shrew, Elephantulus myurus
jamesoni, which has a uterus duplex. This species
sheds dozens of ova at each ovulation; all become
fertilized but only two become implanted, one in
each uterus in a region of remarkable vascular de-
velopment [van der Hoist & Gillman (115)].
The meaning of the above relationships is being
studied by Boving (36, 37). Implantation occurs in
the vicinity of a single capillary loop lying beneath
the endometrial epithelium. Attachment (in the
rabbit) takes place when the abembryonic pole of the
blastocyst develops a sticky substance that is lacking
over the embryonic pole. This substance is related to a
gradient of alkalinity occurring within the blastocyst
and is associated with the differential in production of
metabolites between the embryonic pole and the
abembryonic pole of the blastocyst. The concept is
that the blood flowing through the capillary loop
removes C02 about as fast as it is produced, leaving
behind on the surface of the blastocyst a calcium-
proteinate residue that is sticky. Carbonic anhydrase
is present in high concentration in the endometrium
[Lutwack-Mann & Laser (138)]. The epithelium of
the endometrium breaks down when attachment
occurs [Boving (37)]- In intraocular transplants in
rats, trophoblast causes a breakdown of capillaries
[Grobstein (86)] as it does in the endometrium
[Mossman (158)].
From this time until the period of uterine conver-
sion (see above), when the pregnant uterus changes
from spheroidal to an elongating form, the uterine
vasculature undergoes enlargement and its blood
UTERINE BLOOD FLOW
l6o3
fig. 15. Local vasodilating
action of estrogen in endome-
trium of guinea pig. A, anti-
mesometrial, B, lateral; C, mes-
ometrial. [From Bacsich &
Wyburn (15).]
volume increases [Orsini (164), Reynolds (199)]
coinciding with die period of most rapid uterine
enlargement. From this time until near term, there is
a period of diminished blood in the vascular bed
until term, when a period of partial hemostasis
supervenes [Barcroft & Rothschild (21), Reynolds
(192, 196)]. These changes are supported by studies
of bits of transplanted endometrium to the anterior
chamber of the eye [Neumann (160), Krichesky
('30].
When the uterus is in situ, the blood vessels over the
conceptus give evidence of hypertrophy and the
tortuous course of the uterine arteries progressively
changes as they straighten out [Reynolds (199)].
This is associated with local distention of the tissues
by the conceptus. Distention is a factor in uterine
hypertrophy [Reynolds (198)]. The veins, showing no
initial tortuosities, can only adapt by growth, stretch-
ing, and proliferation. In the uterus duplex, the
vessels in the interconceptus sites show no such
changes. Only as the spheroidal conceptuses enlarge
and encroach upon the interconceptus sites do the
blood vessels there become involved in extension and
stretching. These processes continue until a phase of
maximum spheroidal size is attained. At this time,
vessels that have been crowded from about each
conceptus toward the interconceptus sites lie close
together; those that lie around a conceptus are
stretched and, in any one area, sparse. Within a very-
short period of time (in the order of hours), the rapidly
enlarging conceptus breaks out of its spheroidal
shape as it pushes along the uterine lumen into less
distended regions of the uterus. When this happens,
the vessels of the interconceptus region slip with the
tissues in which they lie over the conceptus, much as a
stocking is slipped up a leg. After this, the enlargement
of the conceptus is solely by elongation, without
further increase in diameter. This elongation con-
tinues until shortly before term, at which time a second
limit of distention is reached and stress is placed upon
the circulation for a second time. In any event, at a
time when fetal growth and demands upon the cir-
culation are great, the uterine blood vessels merely
become rearranged so as to minimize the hemo-
dynamic work of the maternal circulatory system in
supplying the uterus and its contents.
How is blood flow in the uterus modified as these
changes take place? By measuring local circulation
times [Reynolds (194, 199)], it has been found that as
the spheroidal conceptus enlarges there is a progres-
sive decline in the circulation rate until the time of
conversion. Just prior to conversion, there is a pro-
found hemostasis in the tissues about a conceptus.
Upon release of tissue tension by the act of conversion,
a sudden increase in local circulation rate takes place,
approaching that observed at the start of pregnancy.
As gestation nears its end, there is a second decline in
blood flow concomitant with the longitudinal stretch-
ing of the uterus prior to parturition.
The flow characteristics described above relate to
the flow in the maternal vessels of the uterine wall,
not to the other part of the uterine circulation, which
goes to the placenta. Here, there must be adequate
flow at all times, otherwise the fetus will be
endangered. No objective study has been made of the
manner by which this is accomplished, but it has been
speculated that the governing factor is the changing
shape of the pregnant uterus combined with tension
in the uterine tissues [Reynolds (195)]. Blood flow is
reduced to the tissues of the uterus which are most
concerned with change of shape in order to accommo-
date products of conception, and at the same time
blood is directed toward the placenta, since both
parts of the system are supplied by the same arteries at
the border of the mesometrium. This is to say that as
the peripheral vascular resistance increases in one
1604
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
_RPI _,
fig. 16. Injected arteries of rabbit uterus on 12th day (/),
16th (2), 20th (3, side view; ,/, antimesometrial view), 22nd
(5 and 6) 24th (7 and ) of gestation. Note diminishing in-
jectibility of blood vessels as uterus reaches maximum spheroidal
distention on day 22. [From Reynolds (199).
part of a common vascular bed, flow to the other area
is favored, since the peripheral vascular resistance
there, while unchanged, is relatively less than in the
first area. Study of the vascular rearrangements in the
pregnant monkey uterus show that a comparable
pattern of vascular arrangements occur in the uterus
simplex [Gillespie et al. (83), Ramsey (187)].
Physiologists have long been concerned with uterine
blood flow. Barcroft and his associates measured
total uterus blood volumes and blood flows in the
uterus of the rabbit throughout pregnancy [Barcroft
et al. (20), Barcroft & Rothschild (21)]. It was found
that both increase but in dissimilar patterns. From a
content of about 2 ml at the outset of pregnancy in the
rabbit, the uterine blood volume increases to more
than 30 ml by day 27, whereupon it declines 50 per
cent in the next 3 days. The blood flow increases to
two peaks of 30 ml per min on the 20th and 27th
days (gestation 31 days), with the decline in total
blood flow to less than 20 ml on day 24. It will be
seen that the local deprivation of blood in uterine
tissues described in the regional studies above were
reflected also in the total blood flow' [Reynolds (192)].
Page (167), using an indirect method of reasoning
based on facts, shows that in the ninth month of
pregnancy in women there is a decline of nearly one-
half in uterine blood flow.
The turnover of blood in the pregnant rabbit uterus
based on the flow divided by the volume, shows a
progressive increase from about 60 per cent turnover
on the 1 2th day of pregnancy to 1 75 per cent turnover
on day 20 with a sustained 75 to 85 per cent turnover
after uterine conversion. Comparing the turnover
characteristics with the factors of uterine growth and
distention, one sees how affected by or related to these
factors the uterine circulation is [Reynolds (192)].
Recent studies have been directed toward the meas-
urement of total uterine blood flow in sheep and
humans. These have been of three types. In the sheep
and humans, arterial and venous blood sampling and
application of the Fick principle have been used. In
30
c c s/min
cc s
•
<7
A
20
10
'\
days
fig. 17. A: blood flow (x) and blood volume (•) in rabbit
uterus during pregnancy. B: percentage turnover of blood (•)
related to uterine growth (stippled area) and intrauterine
pressure (x). [From Reynolds (192), based in part on Barcroft
et al. (20).]
UTERINE BLOOD FLOW
[605
Case N 821
mmHg
minutes
- 90
10
15
20
25
30
3T
TD-
TT
To
fig. 18. Effect of uterine contraction in women on maternal circulation. Uterine pressure is
shown on two sensitivity scales (0-100 mm Hg; 0-20 mm Hg to show effect of tonus in latter record
on vena cava pressures, superior vena cava, and retrohepatic). Note decrease in inferior caval pres-
sure as uterus contracts when tonus is high but not when tonus is low. Note bradycardia during
contraction. (Permission of Bieniarz el al., XXI Int. Cong. Physiol. Sciences, Buenos Aires, Aug.
'959-)
the ewe, continuous monitoring; of uterine blood flow
in a uterine artery has been done with an electro-
magnetic flow meter. The third method involved in-
jection of radioactive sodium into the intervillous
spaces and measuring the disappearance rate.
Metcalfe el al. (154) found by use of a modified
Fick principle (142) that the blood flow to the non-
gravid uterus in sheep and goats is 25 ml per min.
Slightly more than half way through pregnancy on
the 80th day, the flow increases to 200 ml per min.
At term, flow is more than 1 liter per min, a sub-
stantial increase in blood flow for an organ. The
surface area of the fetal portion of the human placenta,
it may be noted parenthetically, is estimated to be
about 15 m2 [Christoffersen (53)]; the increase in
placental villous surface area throughout pregnane}-
is described by Wilkin & Bursztein (245).
Metcalfe et al. (153) have related the increase in
blood flow as observed by them and others to the
fetal demand, in the rabbit, human, and ungulate
(sheep, goat). The relations are summarized as
follows :
Rabbit
Human
Ungulate
Uterine blood flow per
•25
156
283
kilogram fetal weight
(ml/kg/min)
i6o6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
(Retrohepatic
Region )
INFERIOR °
VENA CAVA 80
PRESSURE 40
(Retrouterine -
Region)
80
Case N. 889
AMNIOTIC
FLUID
PRESSURE
FEMORAL 120-(-|
ARTERY
PRESSURE
80
to\
fig. 19. Blocking of inferior (retrouterine) vena cava with strong uterine contraction. Note
fall of arterial blood pressure as venous pressure rises. [Permission of Caldeyro Barcia el al. (51).
See fig. 14.]
The above differences in maternal hemodynamic
work are not related to the oxygen utilization of the
fetus, as shown by the same authors:
Uterine oxygen consump-
tion per kilogram fetal
weight (ml/kg/min)
Rabbit Human Ungulate
8.3 7.4 8.9
The syndesmochorial placenta in terms of uterine
work to supply a given weight of placenta appears
to be far less efficient than the hemochorial type of
placenta.
On admittedly less sure grounds, observations on
human uteri have been made. Techniques of
sampling, numbers of subjects, assumptions regarding
uterine weights and other uncertainties all contribute
to the interpretation of the data, and a number of
workers have entered into this uncertain field. These
reports are discussed by Metcalfe et al. (152). In
their series, 13 single fetus pregnancies were studied.
They found the blood flow to be of the order of 500
ml per min and the O2 consumption of the uterus and
its contents, 25 ml per min. In one twin pregnancy,
both values were about doubled. It appears that the
fetal load, or drain upon the uterus, may be a kev
factor in determining how much blood flows to the
uterus.2 This is shown bv the following data:
Single
Twin
Uterine blood flow
460
I 150
Uterine 02
25
48
Uterine CO»
20
47
Uterine RQ
O.80
0
If this is for nonidentical twin pregnancies [not stated
by Metcalfe et al. (153)], the result is understandable:
there are two placentas to be supplied. If there were
but one placenta, the results with respect to blood
flow are less clear. Romney et al. (213) have added
data in the human also.
With respect to the increase in uterine blood flow,
Ramsey et al. (188) have pointed out, it is about half
as great as the increase in maternal renal blood flow
during pregnancv.
The problem of the utero-placental circulation as a
2 Since this review was written, an important paper (H.
Wulf. Der Gasaustausch in der reifen Plazenta des Menschrn.
Z. Geburlshilfe u. Gynakol. 152:
-134, 1962) with extensive
literature review on gas exchange in the placenta has been
published. While dealing mainly with gas exchange, it discusses
the causes (including anatomical and physiological) of utero-
umbilical oxygen and carbon dioxide tension gradients.
UTERINE BLOOD FLOW
1607
physiological burden on the circulation led Burwell
(47-49) to regard it as an arteriovenous shunt. The
blood -volume increase of the human uterus that
occurs is reported by Caton et al. (52). The effect on
the maternal heart rate, cardiac output, and blood
volume are comparable to the effect of a major arterio-
venous shunt in the cardiovascular system. In the
latter part of pregnancy increases are seen in resting
heart rate, cardiac output, circulation rate and blood
volume [Hamilton (91), Palmer & Walker (169)].
In the last few weeks of pregnancy there begins to be a
z 4
" i!60 h
1180
-: x / \
• — •
^ UTERINE CONTRACTION
MINUTES __
^__ 1 // ET^a
10
20
60
90
fig. 20. Average uterine blood flow at term in uterus of ewe.
The time scale does not refer to the duration of uterine con-
tractions or to the duration of the decrease in flow. Although
recordings were taken continuously, for the reason of space
economy only one single contraction, with two blood flow
readings, is given for each rate of oxytocin infusion. In early
labor, uterine contractions occurred every 8-10 min and lasted
for about 20-25 sec. Each contraction was accompanied by a
decrease in flow. When the rate of oxytocin was doubled, con-
tractions occurred every 4-5 min and lasted for 35-40 sec, and
the flow was more reduced. With a further increase in the
infusion rate, the contractions lasted for 45-60 sec and the
reduction in flow was more marked. Note the rebound in flow
during uterine relaxation. [Assali et al. (10).]
decline in each of these as maternal oxygen consump-
tion increases. The cause of these declines is not
known, but one is reminded of the profound uterine
ischemia seen in rabbits (vide supra) toward the end
of pregnancy. The concept of the gravid uterus actin?
as an A-V shunt was first set forth in 1938 [Burwell
(49)] and supported by the later studies cited above.
The most recent summary is by Burwell (47). The
circulatory load as a pathophysiological mechanism
upon the circulation after birth when the A-V shunt
is removed has drawn the attention of Schwarz (217)
and the ischemia of the parturient uterus noted by
Thoma (234).
Assali et al. (7, 8) were the first investigators to
catheterize the uterine vein in women for the purpose
of withdrawing blood samples that could be employed,
when combined with simultaneous arterial samples,
to use the Fick principle (nitrous oxide) in estimating
uterine blood flow. Studying women in normal preg-
nancy, Assali et al. (9) found that the flow was 15 ml
per 100 g of tissue per min. This value, approximating
150 ml per min, is surely low (see above). However,
with the same method, they observed a decrease to
9 ml per 1 00 g of tissue per min in the first 24-hour
postpartum period. Before commencing these studies,
Assali (7) reviewed and criticized methods used
previously.
A new departure in measurement of uterine blood
flow was reported by Assali et al. (10) who monitored
blood flow in a uterine artery of the pregnant ewe
with an electromagnetic flowmeter during spontane-
ous and induced labor. Uterine contractions, spon-
taneous or induced, were accompanied by a significant
decrease in blood flow which was more or less pro-
portional to the strength of the contraction. During
Pr«- 1
pactum^
delivery
6 7 8
DAY POST-PARTUM
10
-/>-
14
fig. 21. Average uterine blood How
obtained during the puerperium in the
ewe. Delivery of the placenta occurred
between day 1 and day 2. Note the
precipitous fall in flow which occurred
after the expulsion of the placenta. The
progressive decrease during the post-
partum period coincided with uterine
involution. [Assali et al. (10).]
i6o8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
AMNIOTIC
fig. 22. Identical pressures in amniotic
fluid and intervillous space of placenta in the
monkey with the uterus contracted or relaxed.
[Permission of Ramsey et at. (,i 88.]
5 10
C-772 138 DAYS
AMNIOTIC
Average tonus 16
Average amplitude 18 8
I. VS.
Average tonus 4 I
Average omplitude:20 8
fig. 23. Failure of India ink to reach a sinus
structure at margin of monkey placenta. [Per-
mission of Ramsey (185).]
fig. 24. Injected lymphatics in uterus of a
nearly mature rhesus monkey. Note paucity in
superficial endometrium. (From Wislocki &
Dempsey. Anal. Record 75: 341, 1939.)
M vometrium'
uterine relaxation, reactive hyperemia set in. After
labor, there was a precipitous fall in uterine blood
flow which declined gradually with subsequent uterine
involution.
The action of oxytocic, vasopressor, and vaso-
depressor drugs on postpartum uterine blood flow-
was studied by the same method [Assali et al. (11)].
Both natural and synthetic oxytocics in large doses
cause an initial rise, followed by a marked decline in
uterine artery blood flow. Epinephrine produces no
significant change in uterine blood flow, although
norepinephrine increases the diastolic pressure and
mean blood flow. Apresoline (a sympatholytic agent)
increases blood flow substantially.
In a study of uterine blood flow and uterine metabo-
lism in women, Assali et al. (12) report that blood
UTERINE BLOOD FLOW
1609
fig. 25. Pattern of distribution of radiopaque dye injected
into aorta by way of femora] artery in the monkey pregnant
for 1 1 1 days. Serial radiographs taken at 3 - ' 2 _ 4, 5, and 6 sec
after start of injection. Insert at lower left is of marked out
portion of picture above, enlarged 4 times. The arrows indicate
spurts of dye in the intervillous space. SA, spiral arteries of
endometrium; HA, hypogastric artery; (7.4, uterine artery;
RA, renal artery. [Permission of Ramsey f 1 88 ).]
flow values determined by the electromagnetic flow-
meter and the nitrous oxide method were in good
agreement. Uterine blood flow increases from 50 ml
per min in the 10th week of gestation to 190 ml per
min at the 30th week. The flow of blood per unit of
uterine tissue was determined to be relatively constant
throughout pregnancy. The rate of increase in the
rate of blood flow and oxygen consumption of the
"uterus" exceeds that of the fetus; it is surmised that
the placenta absorbs the difference.
A number of studies of placental blood flow in
women have been reported. Browne & Veal (44) were
the first to use the injection of \a-4 for this purpose.
They injected it into the intervillous space of normo-
tensive and hypertensive women and estimated in the
former a flow of 600 ml per min; in the latter, about
200 ml per min. Similar differences were found in
the uptake of Na24 in the myometrium by Johnson &
Clayton (1 18). Variations in growth of the conceptus
and associated changes in shape of the uterus affect
markedly uterine and placental blood flow [Browne
(43)]. After fetal death, the placenta cannot be
localized by the Na'-4 method; placental blood flow
must decrease substantially. In later pregnancy, the
placental flow of the maternal blood exceeds by
three times the flow necessary to maintain the fetus.
The work of Browne (43) likewise suggests that as
maternal blood pressure diminishes in normal pa-
tients, placental blood flow increases by some en-
hancing mechanism, perhaps the A-V shunt of
Burwell. This is presumed to be a protective mecha-
nism, analogous to a renal shunt type of mechanism.
L^ The intervillous space pressure is equal to that of
amniotic fluid pressure or very close to it [Alvarez &
Caldeyro Barcia (3), Pryztowski (178), Hellman
et al. (102)]. Interestingly, fetal capillary blood
pressure in the placenta is considerably higher
[Reynolds (202)]. It is probably the association of
several factors that permits the escape of fetal blood
constituents into the intervillous space, and into
maternal blood. One is the high fetal capillary blood
pressure just noted. Another is the progressive thinning
of the trophoblast layer as the placenta ages, with
loss of the cytotrophoblast layer and with the capil-
laries coming to lie next to the thin syncytium. Still
another is the ever diminishing size of the villi as they
increase in number. Combined, these factors permit
some escape of fetal erythrocytes into the maternal
circulation [Naesland (159), Mengert et al. (151),
Bromberg et al. (42)]. Maternal erythrocytes do not
normally enter the fetal circulation [Mittelstrass &
Horst (155)]- How the exchange of water and other
substances occurs between the maternal circulation
where the pressure is low and the fetal circulation,
where the pressure is high, has been considered
theoretically by Wilkin (243, 244). Under certain
conditions, simple diffusion occurs; under others, the
process depends upon active transport mechanisms
[Huggett & Hammond (116)].
The connection between maternal uterine blood
and amniotic fluid is still a moot question, despite
intensive study. That there is a rapid and voluminous
l6l0 HANDBOOK OF PHYSIOLOGY ^CIRCULATION II
fig. 26. Pattern of distribution of India ink
in monkey placenta perfused by way of the
aorta. [Permission of Ramsey (183).]
■>
interchange is not questioned [Flexner & Gellhorn
(80), Hellman et al. (101), see Reynolds (198) for
review]. It was shown long ago [Paton et al. (170)]
that the volume of amniotic fluid in any species is
nearly constant at any given stage of normal preg-
nancy [Hammond (94), Reynolds (198), Lell (134),
Wislocki (249), McCafferty (146)]. Although some
amniotic fluid surely passes from the fetus to the
amniotic sac [Reynolds (201)], water passes by an
extraplacental route as well [Paul et al. (171)]. More-
over, maternal emboli of amniotic fluid detritus are
known to occur [Bachman (14)]. Sfameni (219) has
reviewed the lymphatic circulation in the vascular
relations between the mother and fetus. The theo-
retical aspects of the subject are reviewed by Plentl
(174, 175) and McCance & Dickinson ( 1 47).
Knowledge of the manner by which maternal
blood reaches the placenta has received much study
in recent years. Maternal blood reaches the placenta,
of course, by endometrial branches of the uterine
arteries. Blood is drained from the placenta by endo-
metrial branches of the uterine veins. These are
largely anatomical studies based upon injection of
India ink or other media into the aorta or the femoral
vein followed by sections and study of the injected
regions [Ramsey (181)]; by injection-corrosion prep-
arations; and by serial radiography [cf Ramsey et al.
(188)]. Direct injections of uterine vessels of excised
PRIMARY PLACENTAL'''
• Totol orteries 7
SECONDARY PLACENT,
C-750- I23days
o Total veins 19
• Total arteries 10
o Total veins 18
fig. 27. Total arterial and venous openings in the placenta of the monkey in Lite pregnancy.
[Permission of Ramsey (184).]
UTERINE BLOOD FLOW
l6l I
pregnant uteri have also been made. Evidence shows
that the blood enters the intervillous space of the
placenta in spurts and diffuses into relatively localized
areas where, circulating about the placental villi,
it leaves the spaces by nearby veins. However, simul-
taneous blood samples from different parts of the
intervillous space yield the same blood oxygen levels
[McGaughey et a!. (148)]. There is no appreciable
Spanner type of circulation toward the chorial plate
and then to the margin of the placenta where it is
carried off through a marginal sinus. The latter
structure does not, in fact, exist. There are veins that
drain various parts of the margin of the placenta but
veins also drain the basal plate and the septa as well.
The number of vessels supplying the placenta has
received recent attention. The number of arteries
emptying into the placenta in late pregnancy per unit
area is less than at an earlier time [Boyd (41)]. The
number of arterial openings into the human inter-
villous space is about 300 for 25,000 mm- at term and
about 1 20 for 6000 mm2 in the fourth month. More-
over, the lumens of the arterioles are much reduced in
size by an accumulation of intimal tissue near the
orifice [Ramsey (184)]. In elephantulus, each pla-
centa is supplied by three small arterioles [van der
Horst (114)]. The number of veins draining the
placenta is about double that of the arteries [Ramsey
(182, 186)]. Radiographic (serial) studies of the
entry of blood into the placenta are reported in
monkeys [Ramsey et al. (188)] and women [Borell et
al. (38, 40), Fernstrom (79), Hormann (111), Hart-
nett (97)].
The question of the pathway of maternal bood
flow vis-a-vis the fetal blood flow in the hemochorial
placenta has received consideration. Barcroft & Bar-
ron (18, 19), Wimsatt (248), and Mossmann (158)
incline to the view that incoming maternal arterial
blood (oxygenated) encounters incoming fetal blood
(reduced) and, running parallel to the point at which
the streams part, the maternal blood gives up oxygen
to the fetal blood along the way. Noer (161) has
shown this to be true in an artificial model when
acid ions, dyes, and dextrose are used. To apply the
principle of countercurrent flow to the hemochorial
placenta, as Spanner (224, 225) has done, is in error,
as a number of observations show [Stieve (226),
Ramsey (183, 185), Fernstrom (79), Borell et al. (39),
Hilleman (107), Kladetzky-Haubrich (129)]. It does
not apply in the labyrinthine placenta of the rat
[Hamilton & Boyd (92), B0e (32, 33)] or hamster
[Adams & Hilleman (1)], or in the placenta of the
sow [Amoroso (4)].
Extensive studies of the arrangements of fetal
cotyledons and of the blood vessels within them have
been made. The gross vascular arrangement in the
hemochorial placenta shows the cotyledon to be a
tuft, arising from a single stem artery. It sends an-
choring branches to the basal plate. Free villi are
given off from the anchoring villi and from recurrent
free villi that pass toward the chorionic plate from
the anchoring sites [Wilkin (244), Crawford (54)].
Gross relations are described by Falkiner (75), Earn
& Nicholson (68), Crawford (56), ten Berge (232),
Thoyer-Rogart & Harris (235), Vernete & Esteba-
Caballera (236), Lemtis (135), Danesino (60, 61),
and La Have (133). One author claims the fetal
vessels are densest near the decidual plate [ten Berge
(232)], but this is denied by Beker & van Steenis (29)
Vasculature of human chorionic villi
(after BfJe)
Superficial capillary network
Paravascular network
fig. 28. Schematic representation of major and minute
vessels in villus of human placenta. [Permission of B0e (33).]
l6l2
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and Crawford (56). Within die fetal villus, the es-
sential vascular arrangement is one of a large plexus
rather than a series of simple capillary networks
[Crawford (56), B0e (33-35)].
In concluding a review of the maternal blood flow
in the uterus it is appropriate to note that where the
pregnant uterus is concerned, the fetal circulation
cannot properly be separated from the uterine cir-
culation. They are a complex and in a sense a single
unit, one affecting the other in the developing changes
that take place. The conceptus affects uterine growth
and all that that entails, while at the same time the
uterus affects the development of its contents. The
complexity of the relationships between uterine
growth, vascularity, and fetal development are re-
viewed elsewhere in this context (203).
REFERENCES
1. Adams, F. W., and M. H. Hilleman. Morphogenesis of
vitelline and allantoic placentae of the golden hamster
(Cricetus auratus). Anal. Record 108: 363-384, 1950.
2. Ahlquist, R. P., and R. A. Woodbury. Influence of
drugs and uterine activity upon blood flow. Federation
Proc. 6: 305, 1947.
3. Alvarez, H., and R. Caldevro Barcia. Fisiopatologia
de la contraccion uterina y sus aplicaciones en la clinica
obstetricia. Segundo Congreso Latino-Americano de Obstetricia
y Ginecologia. Brazil, 1954.
4. Amoroso, E. C. The vascular relations in the placenta of
the sow. Quoted by Huggett and Hammond, 116. J.
Physiol., London 18: 1, 1947.
5. Amoroso, E. C. The Physiology of Reproduction (3rd ed.),
edited by A. S. Parkes. New York: Longmans, vol. 2, 1952.
6. Amoroso, E. C. The biology of the placenta. Trans.
5th Conf. on Gestation, edited by C. Villee. New York:
Josiah Macy, Jr., Found., 1959, pp. 15-72.
7. Assali, N. S. Measurement of uterine blood flow and
uterine metabolism. I. Critical review of methods. Am.
J. Obslet. Gynecol. 66: 3-10, 1953.
8. Assali, N. S., R. A. Douglass, Jr., W. W. Baird, D. B.
Nicholson, and R. Suyemoto. Measurement of uterine
blood flow and uterine metabolism. II. The techniques of
catheterization and cannulation of the uterine veins and
sampling of arterial and venous blood in pregnant women.
Am. J. Obstet. Gynecol. 66: 11 -17, 1953.
9. Assali, N. S., R. A. Douglass, Jr., W. W. Baird, D B.
Nicholson, and R. Suyemoto. Measurement of uterine
blood flow and uterine metabolism. Am. J. Obslet. Gynecol.
66: 248-253, 1953.
10. Assali, N. S., A. Dasgupta, A. Kolin, and L. Holms.
Measurement of uterine blood flow and metabolism. V.
Changes during spontaneous and induced labor in un-
anesthetized pregnant sheep and dogs. Am. J. Physiol.
614-620, 1958.
1 1. Assali, N. S., K. Dasgupta, and A. Kolin. Measurement
of uterine blood flow and metabolism. VI. Effects of
oxytocin, vaso-pressor and vaso-depressor drugs on the
blood How to the post partum uterus of unanesthetized
sheep. Am. J. Obstet. Gynecol. 78: 313-321, 1 959.
12. Assali, N. S., L. Raurama, and T. Peltonen. Measure-
ment of uterine blood flow and uterine metabolism. VIII.
Uterine and fetal blood flow and oxygen consumption in
early human pregnancy. Am. J. Obstet. Gynecol. 79 : 86-
98, i960.
13. Astwood, E. B. Six-hour assay for quantitative deter-
mination of estrogen. Endocrinology 23: 28-31, 1938.
14. Bachman, C. Maternal pulmonary embolism by amniotic
fluid (editorial). .4m. J. Obstet. Gynecol. 43: 164-165, 1942.
15. Bacsich, P., and G. M. Wyburn. Observations on the
oestrous cycle of the guinea pig. Proc. Roy. Soc, Edinburgh
60 : 33-39. '94°.
16. Bacsich, P., and G. M. Wyburn. Cyclic variations in the
vascular architecture of the uterus of the guinea pig.
Trans. Roy. Soc, Edinburgh 60: 79-86, 1940.
17. Bacsich, P., and G. M. Wyburn. Hormonal analysis of
the cyclic variations in the vascular architecture of the
uterus of the guinea pig. Trans. Roy. Soc, Edinburgh 60:
465 (part II), 1940-1941.
18. Barcroft, J., and D. H. Barron. Circulation in the
placenta of the sheep. J. Physiol., London 100: 208, 1942.
19. Barcroft, J., and D. H. Barron. Observations upon
the form and relations of the maternal and fetal vessels in
the placenta of the sheep. Anal. Record 94: 569-595, 1946.
20. Barcroft, J., W. Herkel, and S. Hill. Rate of blood
flow and gaseous metabolism of uterus during pregnancy.
J. Physiol., London 77: 184-206, 1933.
21. Barcroft, J., and P. Rothschild. The volume of blood
in the uterus during pregnancy. J. Physio/., London 76:
447-459, 1932.
22. Bartelmez, G. W. Histological studies on the menstruat-
ing mucous membrane of the human uterus. Contrib.
Embryol. Carnegie Inst. 24: 131 -186, 1933.
23. Bartelmez, G. W. Menstruation. Physiol. Revs. 17: 28-
72, 1937-
24. Bartelmez, G. W. Premenstrual and menstrual ischemia
and the myth of endometrial arterio-venous anastomoses.
Am. J. Anal. 98: 69-95, 1956.
25. Bartelmez, G. W. The phases of the menstrual cycle and
their interpretation in terms of the pregnancy cycle. Am.
J. Obslet. Gynecol. 74: 931 -955, 1957.
26. Bartelmez, G. W. The form and function of the uterine
blood vessels in the rhesus monkey. Contrib. Embryol.
Carnegie Inst. 36: 153-181, 1957.
27. Batson, O. V. The function of the vertebral veins and
their role in the spread of metastases. Ann. Surg. 112:
138" '49. '94°-
28. Batson, O V. The vertebral vein system. .4m. J. Roent-
genol. 78: 195-212, 1957.
29. Beker, J. C, and C. van Steenis. Arterial circulation in
normal and pathological conditions. Ned. Tijdsch Verl.
Gynaecol. 32: 154-158, 1927.
30. Bieniarz, J. The patho-mechanism of late pregnancy
toxemia and obstetrical hemorrhages. I. The contradic-
tion in the clinical picture of placenta praevia depending
UTERINE BLOOD FLOW
l6.3
on the placental site. Am. I Obstet. Gynecol. 75: 444-453,
'958-
31. Bieniarz, J. Venous drainage from the uterus. Trans.
5/A Conj. on Gestation, edited by C. Villee. New York :
Josiah Macy, Jr. Found., 1959, pp. 109-130.
32. B0e, F. Studies on placental circulation in rats. I. Vascular
pattern illustrated by experiments with India ink. Acta.
Endocrinol. 5: 356-367, 1950.
33. B0E, F. Studies on placental circulation in rats. II. Vas-
cular pattern illustrated by corrosion preparations. Acta
Endocrinol. 5: 369-375, 1951.
34. B0E, F. Studies on vascularization of the human placenta.
Acta Obstet. Gynecol. Scand. (Suppl. 5) 32: 1-92, 1953.
35. B0e, F. Vascular morphology of the human placenta. In :
The mammalian fetus: Physiological aspects of develop-
ment. Cold Spring Harbor Syrnp. Quant. Biol. 19: 29-35, 1954.
36. Boving, B. G. Internal observation of rabbit uterus.
Science 116:211-214, ' 952
37. Boving, B. G. Implantation. Ann. N. Y. Acad. Sci. 75:
700-725, 1959.
38. Borell, U., and I. Fernstrom. The ovarian artery; an
arteriographic study. Acta Radiol. 42: 253-265, 1954.
39. Borell, U., I. Fernstrom, and A. Westman. Eine
arteriographische Studie des Placentarkreislaufs. Ger-
burtsh. Frauenheilk. 18: 1-9, 1958.
40. Borell, U., I. Fernstrom, and A. Westman. Hormonal
influence on the uterine arteries, an arteriographic study
in the human. Acta Obstet. Gynecol. Scand. 32: 271-284,
■953-
41. Boyd, J. D. Physiology of the utero-placental circula-
tion. Trans. 2nd Conf. on Gestation, edited by C. Villee.
New York: Josiah Macy, Jr., Found., 1955, pp. 170-171.
42. Bromberg, Y. M., M. Salzberger, and A. Abrahamov.
Transplacental transmission of fetal erythrocytes with
demonstration of fetal hemoglobin in maternal circulation.
Obstet. and Gynecol., U.S.S.R. 7: 672-674, 1956.
43. Browne, J. C. McG. Utero-placental physiology. Cold
Spring Harbor Syrnp. Quant. Biol. 19: 60-70, 1954.
44. Browne, J. C. McC, and N. Veal. Method of locating
placenta in intact uterus by means of radioactive sodium.
J. Obstet. Gynaecol. Brit. Empire 57 : 566-568, 1950.
45. Browne, J. C. McC, and N. Veal. The maternal pla-
cental blood flow in normo-tensive and hyper-tensive
women. J. Obstet. Gynaecol. Brit. Empirebo: 141-147, 1953.
46. Burn, J. H. On vasodilator fibers in the sympathetic, and
on the effect of circulating adrenaline in augmenting the
vascular response to sympathetic stimulations. ./. Physio/.,
London 75: 144-160, 1932.
47. Burwell, C. S. Circulatory adjustments to pregnancy.
Bull. Johns Hopkins Hosp. 95: 1 15-149, 1954.
48. Burwell, C. S. Utero-placental circulation in mammals.
Trans. 2nd Conf. on Gestation, edited by C. Villee. New York :
Josiah Macy, Jr., Found., 1955, p. 195.
49. Burwell, C. S. Placenta as a modified arteriovenous
fistula, considered in relation to the circulatory adjust-
ments to pregnancy. Am. ,/. Med. Sci. 195: 1-7, 1938.
50. Caldeyro Barcia, R. Trans. 1st Conj. on Physiol. Prema-
turity, edited by J. Lanman. New York: Josiah Macy, Jr.,
Found., 1953.
51. Caldeyro Barcia, R., L. Norica Gcerra, L. Cibile,
H. Alvarez, J. Poseiro, S. Pose, Y. Sica-Blanco, C.
Mendez-Bauer, C. Fieletz, and V. Gonzalez-Panizza.
Effect of position changes on the intensity and frequency
of uterine contractions during labor. Am. J. Obstet. Gynecol.
80: 284-290, 1 96 1.
52. Caton, W. L., C. C. Roby, D. E. Reid, R. Caswell,
C. J. Maletako, R. G. Flaherty, and J. G. H. Gibson.
The circulating blood volume and body hematocrit in
normal pregnancy and the puerperium, by direct meas-
urement using radioactive red cells, II. Am. J. Obstet.
Gynecol. 61 : 1 207-1217, 1951.
53. Christoffersen, A. K. La superficie des villosites chori-
ales du placenta a la fin de la grossese; etude d'histologie
quantitative. Compt. rend. soc. biol. 117: 641-644, 1934.
54. Crawford, J. M. Fetal placental circulation. III. Anat-
omy of cotyledons. J. Obstet. Gynaecol. Brit. Empire 63 : 542-
547. '956-
55. Crawford, J. M. Fetal placental circulation. II. Gross
anatomy. J. Obstet. Gynaecol. Brit. Empire. 63 : 87-90, 1956.
56. Crawford, J. M. Fetal placental circulation. IV. The
anatomy of the villus and its capillary structure. J. Obstet.
Gynaecol. Brit. Empire 63: 548-552, 1956.
57. Csapo, A. Function and regulation of the myometrium.
Ann. N. Y. Acad. Sci. 75: 790-808, 1959.
58. Curtis, A. H., B. J. Anson, F. L. Ashley, and T. Jones.
The blood vessels of the female pelvis in relation to gyneco-
logical surgery. Surg. Gynecol. Obstet. 75: 421-423, 1942.
59. Dalgaard, J. B. The blood vessels of the human endo-
metrium. Acta Obstet. Gynecol. Scand. 26: 342-378, 1946.
60. Danesino, V. Blocking and arterio-venous anastomosis
arrangements in the human placenta. Arch, ostet. e gmecol.
55: 251-272, 1950.
61. Danesino, V., and K. Wiedermann. A microscopic study
of the arrangement and characteristics of the fetal vessels
in the human placenta. Arch, ostet. e ginecol. 55: 471-495,
195°-
62. Daron, G. H. The arterial pattern of the tunica mucosa
of the uterus of Macacus rhesus. Am. J. Anat. 58: 349-
419. '936-
63. Daron, G. H. The veins of the endometrium (Macacus
rhesus) as a source of the menstrual blood. Anat. Record
67 (Suppl. 3): 13, 1937.
64. Davidsohn, S. Ueber die Arteria uterina insbesondere
fiber ihre Beziehungen zum unterei Uterinsegment.
Morphol. Arbeiten 2: 663-671, 1893.
65. Delporte, F. Contributions a /'etude de la nidation de I'oeuf
humain et de la physio/ogie du trophoblasle (Thesis). Brussels,
1912.
66. Donnelly, G. C Gross abnormalities of placenta asso-
ciated with bleeding in pregnancy. Am. J. Obstet. Gynecol.
61 : 910-913, 1951.
67. Dornhorst, A. C, and I. M. Younc. Action of adrenaline
and nor-adrcnaline on the placental and fetal circulations
in the rabbit and guinea pig. J. Physiol., London 118:
282-288, 1952.
68. Earn, A. A., and D. Nicholson. The placental circula-
tion, maternal and fetal. Am. J. Obstet. Gynecol. 63: 1-5,
I952-
69. Emmel, V. M., R. V. VVorthington, and E. Allen.
Attempts to induce menstruation by operative ischemia in
monkeys. Endocrinology 29: 330-335, 1941.
70. Emmons, C. W., F. C. MacIntosh, and D. Richter.
Oestrogens and acetycholine. J. Physiol., London 101 :
460-664, 1943.
71. Evans, H. M. On the development of the aortae, cardinal
and umbilical veins, and the other blood vessels of verte-
1 6i 4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
brate embryos from capillaries. Anat. Record 3: 498-518, 94
1909.
72. Evans, J. N. A scotoma associated with menstruation.
Am. J. Ophthalmol. 24: 507-518, 1 94 1 . 95.
73. Everett, J. W., and C. H. Sawyer. Effects of castration
and treatment with sex steroids on the synthesis of serum
cholinesterase. Endocrinology 39: 323-343, 1946.
74. Fagin, J., and S. R. M. Reynolds. The endometrial 96.
vascular bed in relation to rhythmic motility with a con-
sideration of the function of intermittent contractions of
oestrus. Am. J. Physiol. 117: 86-91, 1936. 97.
75. Falkiner, N. M. Placental circulation. Proc. Roy. Soc. Med.
37: 417-425. I943"'944-
76. Falkiner, N. McE, and J. B. Fleming. Uterine vascular 98.
changes in menstruation and pregnancy. Irish J. Med.
Set. No. 286: 739-749, 1949. 99-
77. Faulkner, R. L. The blood vessels of the myomatous
uterus. Am. J. Obstet. Gynecol. 47: 185-197, 1944.
78. Faulkner, R. L. An injection study of uterine blood 100.
vessels. -4m. J. Obstet. Gynecol. 49: 1-9, 1945.
79. FERNSTRom, E Arteriography of the uterine artery. Acta 101.
Radiol. Suppl. 122: 1, 1955.
80. Flexner, L. B., and A. Gellhorn. The transfer of water
and sodium to the amniotic fluid of the guinea pig. Am. J.
Physiol. 136:757-961, 1942.
81. Flexner, E. B., and A. Gellhorn. The comparative 102.
physiology of placental transfer. Am. J. Obstet. Gynecol. 43:
965-974. '942-
82. Franklin, K. J. L'ndulatory changes of uterine origin in 103.
the arterial blood pressure. J. Physiol., London 84: 342-
343. :935-
83. Gillespie, E. C, E. M. Ramsey, and S. R. M. Reynolds. 104.
The pattern of uterine growth during pregnancy. Am. J.
Obstet. Gynecol. 58: 758-764, 1949.
84. Gillman, J. Profound vascular changes induced in the
uterus of the castrated rabbit by combinations of estradiol 105.
benzoate and progesterone. Endocrinology 29: 336-342,
1941.
85. Goodman, L., and G. B. Wislocki. Cyclical uterine bleed-
ing in a New World monkey (Ateles Geoffroyi). Anat. 106.
Record 61: 379-387, 1935.
86. Grobstein, C. Production of intraocular hemorrhage by
mouse trophoblast. J. Exptl. Zool. 114: 159-174, 1950. 107.
87. Grollman, A. J. Effect of pregnancy on course of experi-
mental hypertension. Am. J. Physiol. 151 : 373-379, 1947.
88. Grosser, O. Friihentwicklung, Eihautbildung und Pla-
centation des Menschen und der Saugetiere, Deutsche 108.
Frauenheilkunde. Deutsche Frauenh. (Band V). New York:
Bergman, 1927.
89. Grosser, O. Uber die Bedeutung des intervillosen 109.
Raumes. Arch. Gyndkol. 137: 681-689, 1929.
90. Grosser, O. Human and comparative placenfation in-
cluding early stages of human development. Lancet 1 :
999. '933- II0-
91. Hamilton, H. F. H. Cardiac output in normal pregnancy
as determined by Cournand right heart catheterization 111.
technique. J. Obstet. Gynaecol. Bnt. Empire 56: 548-553,
1949. 112.
92. Hamilton, W. J., and J. D. Boyd. Observations on the
human placenta. Proc. Roy. Soc. Med. 44: 489-496, 1951.
93. Hamlett, G. W. D. Reproduction in American monkeys. 1 13.
E Estrous cycle, ovulation and menstruation in cebus.
Anat. Record 73: 171-187, 1939.
Hammond, J. The changes in the reproductive organs of
the rabbit during pregnancy. Trans. Dynamics of Develop-
ment 10: 93-103, 1935.
Harrison, R. J , and W. J. Hamilton. The reproductive
tract and the placenta and membranes of Pere David's
deer {Elaphurus davidianus, Milne Edwards). J. Anat. 86:
203-224, 1952.
Hartman, C. G The homology of menstruation. New
observations of intermenstrual bleeding in the monkey.
J. Am. Med. Assoc. 92: 1992-1995, 1929.
Hartnett, L. J. Visualization of maternal circulation at
the site of the placenta J. Missouri State Med. Assoc. 44 :
754-756, 1947.
Hasner, E. Endometritis Vasculare Cyklus (Thesis). Copen-
hagen : Det Berlingske Bogtrykeri, 1946.
Hechter, O., L. Krohn, and J. Harris. The effect of
estrogen on the permeability of the uterine capillaries.
Endocrinology 29: 386-392, 1941.
Heckel, G. P., and C. E. Tobin. Arteriovenous shunts in
the myometrium. Am. J. Obstet Gynecol. 71 : 199-205, 1956.
Hellman, L. M., L. B. Flexner, W. S. Wilde, G. J.
Vosburgh, and ). H. Proctor. Permeability of the
human placenta to water and the supply of water to the
human fetus as determined with deuterium oxide. .4m. J.
Obstet. Gynecol. 56: 861-868, 1948.
Hellman, L. M., V. Tricomi, and O. Gupta. Pressures
in the human amniotic fluid and intervillous space. Am. J.
Obstet. Gynecol. 74: 1018-1021, 1957.
Herschberg, A. O. Contribution a V etude de la Regulation
Physiologique du Systeme Acetylcholine-Cholinesterase (Thesis).
Paris: Imprimerie Union, 1946.
Hertig, A. T., and J. Rock. Two human ova of the
pre-villous stage, having an ovulation age of about eleven
and twelve days respectively. Conlrib. Embryol. Carnegie
Inst. 29: 127-156, 1941.
Hess, W. R. Die Verteilung von Querschnitt, Wieder-
stand, Druckgefalle und Stromgeschwindigkeit im Blut-
kreislauf. In: Handb. d. Norm. u. Pathol., Physiol., edited by
A. Bethe. VIE Berlin: Springer, 1928, pp. 904-933.
Hill, H. C, ]r. Effect of diethylstilbestrol upon the
systolic blood pressure of normal rats. Proc. Soc. Exptl.
Biol. Med. 63: 458-459, 1946.
Hilleman, H. H. The organization, histology and cir-
culatory pattern of the near term placenta of the Guinea
baboon Papia cynocephalus. Oregon State Studies Monogr. Zool.
9'. '955-
1 1 is, W. Die Umschliessung der menschlichen Frucht wahrend
der fruhesten Zeiten '936-
i6i6
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
157. Mossman, H. W. Rabbit placenta and the problem of
placental transmission. Am. J. Ana!. 37: 433-497, 1926.
158. Mossman, H. \V. Comparative morphogenesis of the
fetal membranes and accessory uterine structures. Contrib.
Embryo/. Carnegie Inst. 26: 129-246, 1937.
159. Naesland, J. Research on the permeability of the placenta
with the aid of blood group determination, radio-active
corpuscles and elliptocytes. Nord. Med. 29: 589-592, 1946.
160. Neumann, R. Uterus-Kammer-Transplantationen Verp-
flanzung von Endo- und Myometrium in die Vordere
Augenkammer. Arch. Gynakol. 157: 548-581, 1934.
161. Noer, R. A study of the effect of flow direction on pla-
cental transmission using artificial placentas. Anal. Record
96:383-389. IQ46-
162. Ogden, E., G. J. Hildebrand, and E. W. Page. Rise of
blood pressure during ischemia of the gravid uterus.
Proc. Soc. Exptl. Biol. Med. 43 : 49-5 1 , 1 940.
163. Okkels, H., and E. T. Engle. Studies on the finer struc-
ture of the uterine blood vessels of the Macacas monkey.
Acta Pathol. Microbiol. Scand. 15: 150-168, 1938.
164. Orsini, M. W. The vascular knot of the hamster uterus;
the placental arterial supply and its changes during
gestation and postpartum involution. ./. Morphol. 100:
565-600, 1957.
165. Oughtred, O. W., and S. R. M. Revnolds. Collateral
pathways utilized upon ligation of the inferior vena cava
at different levels in the dog. Surg. Gynecol. Obstet. 1 1 :
63-70, i960.
166. Page, E. W. Relation of fetus and placenta to decline of
hypertension in pregnant rats. Am. J. Obstet. Gynecol. 53 :
275-278, 1947.
167. Page, E. W. Discussion of uteroplacental circulation in
mammals. Trans. 2nd Conf. on Gestation, edited by C. Villee.
New York : Josiah Macy, Jr., Found., 1955, p. 210.
168. Page, E. W. In: Trans. 5th Conf. on Gestation, edited by C.
Villee. New York: Josiah Macy, Jr., Found., 1959, pp.
122-124.
169. Palmer, A. J., and A. H. C. Walker. The maternal
circulation in normal pregnancy. J. Obstet. Gynaecol. Brit.
Empire 56: 537-547, 1949.
170. Paton, D. N., B. P. Watson, and J. Ken. On the source
of the amniotic and allantoic fluids in mammals. Trans.
Roy. Soc, Edinburgh 46: 71, 1907.
171. Paul, W. M., T. Enns, S. R. M. Reynolds, and F. P.
Chinard. Sites of water exchange between the maternal
system and the amniotic fluid of rabbits. J. Clin. Invest.
35 : 634-640. 1 946-
1 72. Phelps, D. H. The experimental production of menstrual
anomalies. Endocrinology 39: 105-119, 1946.
1 73. Phelps, D. H. Endometrial vascular reactions and the
mechanism of nidation. Am. J. Anat. 79: 167-197, 1946.
174. Plentl, A. A. The origin of amniotic fluid. Trans.
4th Conf. on Gestation, edited by C. Villee. New York :
Josiah Macy, Jr., Found., 1957, p. 71.
175. Plentl, A. A. The dynamics of the amniotic fluid. Ann.
IV. Y. Acad. Sci. 75: 744-761, 1959.
176. Pompen, A. W. M. De Invloed van Menformon op der Baar-
moeder (Thesis). Amsterdam: 1933.
1 77. Price, D. Influence of hormones on sex differentiation in
cxplanted fetal reproductive tracts. Trans, yd Conf. on
Gestation, edited by C. Villee. New York : Josiah Macy,
Jr., Found., 1956, pp. 175-186.
178. Pryztowsky, H. Fetal blood studies VII. The oxygen
pressure gradient between the maternal and fetal bloods
of the human in normal and abnormal pregnancy. Bull.
Johns Hopkins Hosp. 101 : 48, 1957.
179. Pryztowsky, H. Fetal blood studies. VIII. Some obser-
vations on the transient fetal bradycardia accompanying
uterine contractions in the human. Bull. Johns Hopkins
Hasp. 102 : 1 -7, 1958.
180. Ramsey, E. M. The vascular pattern of the endometrium
of the pregnant rhesus monkey. Anat. Record 97 : 363, 1947.
181. Ramsey, E. M. The vascular pattern of the endometrium
of the rhesus monkey (Macaca mulatta). Contrib. Embryol.
Carnegie Inst. 33: 1 13-148, 1949.
182. Ramsey, E. M. Venous drainage of the placenta of the
rhesus monkey (Macaca mulatta). Contrib. Embryol.
Carnegie Inst. 35: 151-174, 1954.
183. Ramsey, E. M. Circulation in the maternal placenta of
primates. Am. J. Obstel. Gynecol. 67: 1-14, 1954.
184. Ramsey, E. M. Physiology of the utero-placental circu-
lation. Trans. 2nd Conf. on Gestation, edited by C. Villee.
New York: Jo=iah Macy, Jr., Found., 1955, pp. 174-175.
185. Ramsey, E. M. Circulation in the maternal placenta of
the rhesus monkey and man, with observations on the
marginal lakes. Am. J. Anat. 98: 159-189, 1956.
186. Ramsey, E. M. Circulation in the placenta. Trans. 5th
Conf. on Gestation, edited by C. Villee. New York: Josiah
Macy, Jr., Found., 1958, pp. 102-103.
187. Ramsey, E. M. Vascular anatomy of the utero-placental
and fetal circulation. Proc. Josiah Macy, Jr. Found. CIOMS
Conf. on Oxygen Supply to the Human Fetus. Springfield,
111. : Thomas, 1957.
188. Ramsey, E. M., G. W. Corner, Jr., N. W. Donner, and
H. M. Stran. Radioangiographic studies of circulation in
the maternal placenta of the rhesus monkey: preliminary
report. Proc. Natl. Acad. Sci. US. 46: 1 003-1 008, i960.
189. Ramsey, E. M. Vascular adaptations of the uterus to
pregnancy. Ann. X. Y. Acad. Sci. 75: 726-745, 1 959-
190. Reynolds, S. R. M. Studies on the uterus. V. The in-
fluence of the ovary on the motility of the uterus of the
unanesthetized rabbit. Am. J. Physiol, 97: 706-774, 1932.
191. Reynolds, S. R. M. The nature of uterine contractility.
Physiol. Revs. 17: 304-334, 1937.
192. Reynolds, S. R. M. Haemodynamic factors in the uterus
during the latter part of gestation. Nature 140: 546, 1937.
193. Reynolds, S. R. M. Acetylcholine content of uteri before
and after administration of oestrin to ovariectomized
rabbits. J. Physiol., London 95: 258-268, 1939.
194. Reynolds, S. R. M. Relation of maternal blood flow
within the uterus to change in shape and size of the con-
ceptus during pregnancy; physiological basis of uterine
accomodation. Am. J. Physiol. 148: 77-85, 1947.
195. Reynolds, S. R. M. Differential uterine tensions and the
flow of blood through the uterus during pregnancy.
Federation Proc. 6- 188, 1947.
196. Reynolds, S. R. M. The physiologic basis of menstrua-
tion; a summary of current concepts. J. Am. Med. Assoc.
135: 552-557. '947-
197. Reynolds, S. R. M. Morphological determinants of the
flow-characteristics between an artery and its branch,
with special reference to the ovarian spiral artery of the
rabbit. Acta Anal. 5: 1-16, 1948.
198. Reynolds, S. R. M. Physiology of the Uterus (2nd ed.).
New York: Hoeber, 1949.
199. Reynolds, S. R. M. Adaptation of maternal uterine blood
UTERINE BLOOD FLOW
161 7
vessels and uterine accommodation of the products of 222
conception. Contrib. Embryo/. Carnegie Inst. 33: 1-18, 1949.
200. Reynolds, S. R. M. The vasculature of the ovary and
ovarian function. Recent Prog. Hormone Research 5: 65-100, 223
1950.
201. Reynolds, S. R. M. A source of amniotic fluid in the
lamb, the nasopharyngeal and buccal cavities. Nature
>72: 3°7. '953- 224
202. Reynolds, S. R. M. Hemodynamic characteristics of the
fetal circulation. Am. J. Obstet. Gynecol. 68: 69-80, 1954.
203. Reynolds, S. R. M. Gestation mechanisms. Ann. A7. J'. 225.
Acad. Sci. 75: 691-699, 1959.
204. Reynolds, S. R. M., and F. I. Foster. Acetylcholine- 226,
equivalent content of the uterus and placenta in rabbits.
.4m. J. Physiol. 127 : 343-346, 1939.
205. Reynolds, S. R. M., and F. I. Foster. Species differences 227.
in the cholinergic action of estrogens. Am. J. Physiol.
131 : 200-202, 1939. 228.
206. Reynolds, S. R. M., and F. I. Foster. Peripheral vas-
cular action of estrogen in the human male. J. Clin. 229.
Invest. 18:649-655, 1939.
207. Reynolds, S. R. M., and F. I. Foster. Acetylcholine-
equivalent content of the nasal mucosa in rabbits and
cats. Am. J. Physiol. 131 : 422-425, 1940. 230.
208. Reynolds, S. R. M., and F. I. Foster. Peripheral vas-
cular action of estrogen, observed in the ear of the rabbit.
J. Pharmacol. Exptl. Therap. 68: 173-184, 1940. 231.
209. Reynolds, S. R. M., and S. Kaminester. Motility of the
transplanted denervated uterus. Am. J. Obstet. Gynecol.
3°: 395-402. 1935. 232.
210. Reynolds, S. R. M., and S. Kaminester. The peripheral
motor sympathetic innervation to and within the uterus. 233.
Am. J. Physiol. 112: 640-648, 1935.
211. Robinson, B. Arteria Uterina Ovarica: The Utero-ovarian
Artery or the Genital Vascular Circle. Chicago: Colegrove, 234.
1903.
212. Robson, J. M., and H. O. Schild. Effect of drugs on blood 235.
flow and activity of the uterus. J. Physiol., London 92:
9-19, 1938.
213. Romney, S. L., J. Metcalfe, D. E. Reid, and C. S. 236.
Burwell. Blood How of the gravid uterus. Ann. N. Y.
Acad. Sci. 75: 762-791, 1959.
214. Saito, S. Pure human placental extracts causing symp- 237.
toms of toxemia in late pregnancy. J. Japan. Obstet.
Gynecol. Soc. (Eng. ed.). 3: 131, 1956.
215. Sawyer, C. H., and J. W. Everett. Effects of various 238.
hormonal conditions in the intact rat on the synthesis of
serum cholinesterase. Endocrinology 39: 307-322, 1946.
216. Schlegel, J. U. Arteriovenous anastomoses in endo-
metrium in man. Acta Anal. 1 : 284-325, 1945-1946. 239.
217. Schwarz, O. H. Blood pressure changes following de-
livery. Am. J. Obstet. Gynecol. 6: 656-672, 1923. 240.
218. Schwarz, O. H., and W. O. Hawker. Hyperplasia and
hypertrophy of uterine vessels during various stages of
pregnancy. Am. J. Obstet. Gynecol. 60:967-976, 1950. 241.
219. Sfameni, P. The lymph circulation in the vascular re-
lations between mother and fetus. Monit. Zool. Hal. 56
(Suppl.): 338, 1948.
220. Smith, O. W. Menstrual toxin. I. Experimental studies.
Am. J. Obstet. Gynecol. 54: 201-21 1, 1947. 242.
221. Smith, O. W., and G. S. Smith. Evidence that menstrual
'toxin'' and canine "necrosin" are identical. Proc. Soc.
Exptl. Biol. Med. 59: 1 19-121, 1945.
Smith, O. W., and G. S. Smith Studies concerning the
cause and purpose of menstruation. J. Clin. Endocrinol.
6 : 483-492. 1946.
Soskin, S., H. Wachtel, and O. Hechter. Treatment
of delayed menstruation with prostigmin, therapeutic
test for early pregnancy. J. Am. Med. Assoc. 114: 2090-
2091, 1940.
Spanner, R. Mutterlicher und kindlicher Kreislauf der
menschlichen Placenta und seine Strombaknen. Z. Anal.
Entwicklungsgeschichte 105: 163-242, 1935.
Spanner, R. Circulation of the human placenta. Am. J.
Obstet. Gynecol. 71 : 350-362, 1956.
Stieve, H. Uber den Abfluss des Blutes aus dem inter-
villosen Raum der menschlichen Placenta. Vorlaulige
Mitteilung. Z. Gyndkol. 64: 1570-1582, 1940.
Strahl, R., and R. Beneke. Ein jungen menschlicher Em-
bryo Wiesbaden. Wiesbaden: Bergmanon, 1910, p. 292.
Strassman, P. Placenta praevia. Arch. Gyndkol. 67: 112-
275, 1902.
Sturgis, S. H. Method for obtaining uterine fluid from
the monkey: effect of pilocarpine, atropine, physiological
salt solution and adrenalin. Endocrinology 31 : 664-672,
1942.
Taylor, H. C, Jr. Pelvic pain based on a vascular and
autonomic nervous system disorder. Am. J. Obstet. Gynecol.
67: 1177-76, 1954.
Teacher, J. H. On the implantation of the human ovum
and the early development of the trophoblast. J. Obstet.
Gynaecol. Brit. Empire 31: 166-217, ' 924-
Ten Berge, B. S. Capillaraktion in der Placenta. Arch.
Gyndkol. 186: 253-256, 1955.
Thoma, R. Der mittlere Durchflussmenge der Arterien
des Menschen als Funktion des Gefassradius. Pfliigers
Arch. ges. Physio/., 194: 385-406, 1922.
Thoma, H. Ischaemia of the parturient uterus. Am. J.
Obstet. Gynecol. 15:853-857, 1928.
Thoyer-Rocart, J., and A. Martin. A study of fetal
circulation in the placenta by iniection of synthetic resins.
Gynecol, el Obstet. 55: 255-256, 1956.
Vernete, G., and J. Estaba-Caballera. A study of the
morphology of the premature placenta. Acta Ginecol.,
Madrid 5 : 48 1 , 1 954.
Van Wagenen, G. Uterine bleeding of monkeys in re-
lation to neural and vascular processes: spinal transection
and menstruation. Am. J. Physiol. 105: 473-486, 1933.
Van Wagenen, G, and S. Zuckerman. Uterine bleeding
of monkeys in relation to neural and vascular processes:
II. Spinal-cord transection and the oestrin-level. Am. J.
Physiol. 106: 416-422, 1933.
Wagner, G. A. Der intervillose Raum. Arch Gyndkol. 137:
699-708, 1929.
Wattes, G. M. H., and G. R. Moule. Blood pressure in
the internal spermatic artery of the ram. J. Rcprod. and
Fertility 1 : 223-229, i960.
Wermbter, F. Uber den Umbau der Uterusgefasse in
verschiedenen Monaten der Schwangerschaft erst- und
mehrgebarender Frauen unter Beriicksichtigung des
Verhaltcns der Zwischemubstanz der Arterienwande.
Virchow's Arch, pathol. Anal. 257: 249-283, 1925.
Wilkin, P. Some aspects of the vascularization of the
human endometrium during the luteal phase of the men-
strual cycle. Bull. soc. roy. beige, gynecol. obstet. 25: 402-412,
'955-
i6i8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
243. Wilkin, P. Study of physical factors determining pla-
cental permeability. Bull, federation soc. gynecol. el obstH.
langue franf. 9: 33, 1957.
244. Wilkin, P. Morphogenese. In: Le Placenta Humain, edited
by P. Snoeck. Paris: Masson, 1958, pp. 60-67.
245. Wilkin, P., and M. Bursztein. Quantitatise study of the
development of placental exchange surface during preg-
nancy. Bull, federation soc. gynecol. et obstet. langue franc. 9:
37- >957-
246. Williams, E. A. Abnormal uterine action in labor. J.
Obstet. Gynaecol. Brit. Empire 59: 635-641, 1952.
247. Williams, M. F. The vascular architecture of the rat
uterus as influenced by estrogen and progesterone. Am.
J. Anat. 83: 247-308, 1948.
248. Wimsatt, W. A. The placentation of a vespertilionid bat,
Myotis lucifugus lucifugus. Am. J. Anat. 77: 1-52, 1945.
249. Wislocki, G. B. On the volume of the fetal fluids in sow
and cat. Anat. Record 63: 183-192, 1935.
250. Woodbury, R. A., W. F. Hamilton, and R. Torpin.
The relationships between abdominal, uterine and ar-
terial pressures during labor. Am. J. Physiol. 121: 640-
649. '938-
251. Woodbury, R. A., W. F. Hamilton, B. E. Abreu, and
R. Torpin. Effects of posterior pituitary extracts, oxytocin
(pitocin) and ergonovine hydracrylate (Ergotrate) on
uterine, arterial, venous and maternal effective pressures
in pregnant humans. J. Pharmacol. Exptl. Therap. 80:
256-263, 1944.
CHAPTER 46
The fetal and neonatal circulation
MAUREEN YOUNG Department of Medicine, St. Thomas's Hospital, London, England
CHAPTER CONTENTS
The Fetal Placenta
Implantation
Placentation and Placental Function
Early Development of the Cardiovascular System
Peripheral Circulation
The Heart
Congenital Malformation
Course of the Circulation of the Fetus
Regional Blood Flow
Hepatic Blood Supply and the Ductus Venosus
Fetal Heart
Heart Rate. Regulating Mechanisms
Cardiac Output
Arterial Blood Pressure
Systemic Pressure
Pulmonary Artery Pressure
Development of the Cardiovascular Reflexes and the
Responses to Asphyxia and Hormones
Fetal Placental Blood Flow
Effective Perfusion Pressure. Resistance of the Placental
Circulation
Umbilical Blood Flow
Oxygen Environment and Requirements of the Fetus
Influence of Hypoxia and Asphyxia on the Fetus
Hemoglobin
Blood Flow-
Oxygen Consumption
Heart Rate During Reduction in Maternal Placental Blood
Flow
Changes in Fetal Circulation at Birth and in the Neonatal
Period
Umbilical Cord; Ductus Venosus
Fetal Channels in the Thorax
Pulmonary Vascular Resistance, Arterial Pressure, and
Blood Flow
The Heart
Systemic Pressure; Cardiovascular Reflexes, and Peripheral
Resistance
Viability
Congenital Heart Disease
thirty years have passed since the inspiration and
eloquence of Sir Joseph Barcroft gave the functional
development of the cardiovascular system its place
in circulatory physiology. In the intervening vears
histochemical techniques and the electron microscope
have shown how complex is the placental structure
between the maternal and fetal circulations: advances
in knowledge of transport mechanisms and the use of
isotopically labeled compounds begin to clarify the
active processes occurring within this structure: the
pathways of the circulation "in utero,"' in both the
placenta and the fetus, and the changes of the latter
at birth, have been confirmed : studies of the develop-
ment of the regulatory mechanisms of the fetal circula-
tion have been extended into the neonatal period
(70, 83); finally, obstetricians and pediatricians have
accumulated circulatory information on the human
infant which demonstrates the value and limitations
of applying observations from one species to another.
Detailed reviews have been written on each of these
subjects both from the historical viewpoint and that of
comparative physiology: it falls to this chapter to do
justice to the main facts with particular reference to
the higher mammals and to the human infant.
In acute experiments on the fetus with an intact
placental circulation, the possibilities of departure
from the physiological state are even more numerous
than in the grown animal. Understanding of the precise
influence of the disturbances due both to the anesthetic
and to removal from the uterine environment awaits
the development of intrauterine techniques such as the
chronic implantation of electrodes and catheters.
Many workers have tried to minimize these disturb-
ances by working on the fetus delivered into a saline
bath at 37 C; however, particular attention must be
paid to the position of the fetus in relation to the
placenta; further, interference with the maternal
1619
I 620
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
placental blood flow and therefore the internal en-
vironment of the fetus, consequent upon the retraction
of the cut uterine muscle around the uterine vessels,
still occurs. In the sheep, on which many of the
studies have been made, the uterine muscle is not so
reactive to mechanical stimuli as the uterus of the
rabbit, the guinea pig, or the monkey and man.
Spasmolytics have not been used to prevent this re-
sponse of the uterine muscle. The local application
of procaine or papaverine is used to prevent spasm of
the umbilical vessels during cannulation or sampling
of blood, but these maneuvers are best carried out on
the abdominal trunks of the vessels, which are less
contractile, or on the placental tributaries, in order
not to interfere with the main blood flow.
Barcroft's (25) theme that it is dangerous to argue
from species to species about the relative stage of
physiological development in utero and at birth is
most applicable to a consideration of the cardiovascu-
lar system. It will become apparent that, in the spe-
cies which have been most extensively studied, func-
tional development does not depend upon gestational
age but corresponds most nearly to the requirements
of the newly born when it is, however, never so ad-
vanced as in the adult.
THE FETAL PLACENTA
Implantation
What forces compel the fertilized ovum to satisfy
its high nutritive requirements in the superficial
layers of the endometrium? Or does the endometrial
epithelium have the power to attract inert particles
the size of the 7-day human blastocyst (106)? What
are the mutual relationships between the trophoblast
and the endometrium once contact between the two
has been established? Both the anatomical and ex-
perimental aspects of implantation are beautifully
described, for many species, by Hamilton et al.
(102) and by Boyd & Hamilton (45). Under normal
conditions the blastocyst in utero implants at a defi-
nite size, at a prescribed time, and in special sites
(43); the presence of progesterone is essential hut the
mechanism of its action is unknown (148). Fawcett
et al. observe that, "the individual potentialities of the
ovum and uterine mucosa should not be thought of as
mutually exclusive but mutually supporting and
neither is "chiefly" responsible for implantation"
(94). These potentialities may, however, be observed
quite independently : the mouse ovum, once it has
reached a certain size, is capable of implanting ran-
domly in extrauterine sites such as the anterior cham-
ber of the eye and the abdominal cavity, regardless
of the sex of the host. The trophoblast causes extra-
vasation of blood in these sites before cellular inva-
sion has taken place and the substance responsible
must be actively penetrating, for secondary implan-
tation sites start to proliferate in the macaque before
any erosion of the uterine surfaces (191) and, in the
human, congestion also appears on the opposite side
of the uterus to the site of implantation (105). The
active substance may be a product of metabolism
of the dividing blastocyst, even carbon dioxide itself;
or it may be chorionic gonadotropin, known to
appear first at the time of implantation. Evidence
for a penetrating action of chorionic gonadotrophin
is suggested by perfusion experiments on full-time
human placentas; citrate metabolism is enhanced by
estradiol only when chorionic gonadotrophin is also
added to the perfusing fluid (180). The initial re-
sponses of the endometrium may also be observed in
Fetal Capillary
I v 1
Epithelio- Maternal Syndesmo- Maternal Endothelio- Haemo-
chorial Tissues chorial Capillary chorial chorial
fig. I. Histological types of placenta arranged to emphasize the progressive breaking down of the
barrier between the maternal and fetal circulations. [Redrawn by Amoroso (8).]
Maternal Haemoendo-
Blood thelial
THE FETAL AND NEONATAL CIRCULATION
[621
» J&I3
the absence of a fertilized ovum. In the rat, but not
in the guinea pig, the endometrium is able to implant
inert objects, such as glass beads the size of the blasto-
cyst normally implanted (36). Electrical and me-
chanical stimulation of the pregnant rat uterus can
produce the formation of a maternal placenta, iden-
tical in structure with the decidua of pregnancy;
these deciduomata may bleed into the uterine cavity
and the early normal extravasation of blood in the
endometrium is therefore probably not dependent
upon the fetal trophoblast (125, 170).
The implanted blastocyst probably receives its
nourishment for a short time from the glycogen,
lipid, and other materials stored in the stromal
cells of the uterine mucosa which have become en-
larged by the decidual response; in some ruminants
the mucosa secretes uterine milk for this purpose. The
formation of the true placenta, containing the fetal
and maternal circulations is due to a balance in activ-
ity of the fetal trophoblast and maternal decidua; any
disturbance in this balance may result in the rejection
of the blastocyst or nonspecific, even malignant,
growths. Progesterone and estrogen are required for
placental and fetal development, the ovary and
placenta itself contributing to the production of these
hormones to varying extents in the different species
(10): an adequate placenta may develop in the per-
itoneal cavity and viable infants be delivered at lap-
arotomy (81); it has been shown in the mouse and the
cat that the fetus is not necessary for the development
of the placenta (11, 141).
Placentation and Placental Function
Amoroso (8) describes fully the structure of the
tissue which separates the fetal from the maternal
blood streams in mammalian placentas, following the
classification of Grosser (100) modified by Mossman
(138). Figure 1 demonstrates how this scheme is
based on the number of tissue lavers between the two
fig. 2.A: section through a villus from a human placenta
of 9 weeks gestation. (X 7,040.) The surface projections and a
bulbous promontory are illustrated. In the apical part of the
syncytium some large vacuoles, filled with granular material,
are seen. These are interpreted as absorption vacuoles formed
by pinocytosis. They are different from the smaller vesicles with
homogenous centers which are thought to be ergastoplasmic.
B: section through a thin portion of a human villus from a
placenta delivered at term. (X 10,110.) There are well-de-
veloped microvilli on the surface of the syncytium. Beneath the
syncytium is a zone of lighter cytoplasm. Such a zone has been
shown to be continuous with residual Langhans cells. The base-
ment membranes are present, separated by a space in which
collagenous fibrils can be seen. Beneath the second of these is
the endothelium of a fetal capillary. [From Amoroso (9).]
[622
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
circulations. The "thinner" placentas start with
six cell layers between the maternal and fetal blood,
as in the epitheliochorial placenta, and there is a
progressive breakdown of tissue during development
which chiefly involves the three maternal layers.
Figure 2A and B shows that, even after the number
of cell layers has been determined in the human
placenta, there is a further reduction in depth of the
remaining tissue. The thickness of this placental
"barrier" will be one of the factors influencing the
rate of transfer of substances, such as the respiratory
gases, water, and electrolytes, dependent upon
simple diffusion for their transfer; but most of the
exchanges between the mother and fetus will take
place by active transport, and the composition of the
cytoplasm rather than the depth of the barrier is
likely to be the more important factor (63, 143);
the evidence suggests that placental tissue has a high
oxygen consumption, 10 ml per kg per min, probably
higher than the fetus itself (17, 111). Histochemical
techniques have identified the cytoplasmic content of
the cells with a variety of proteins, enzymes, lipids,
and carbohydrates (8, 190); the cytoplasmic structure
is transient, and changes as the functional capacity of
the fetal metabolic processes develop during gestation
(181). The varying structure of the barrier in different
«.'. '■■/-•.■.■-^J'.^*.'
. '■^■.,\irV-:- Maternal artery :;*.
v ' • ' *
•2>"~
1M ...,„/¥/.. .
Maternal vein feS&y, ■'■..■
r:...\vC\
Muscular coat
fig. 3. Arrangement of blood vessels in the placentome of
the sheep placenta. [Redrawn by Amoroso (8) after Barron.]
species has been a constant target for speculation and
probably represents adaptations concerned with
differences in intermediary metabolism and the re-
quired rate of growth of the fetus.
In the hemochorial placenta, in which the maternal
endothelium is absent, vesicles of maternal plasma
may be transported across the trophoblastic cells into
the fetal blood stream; these vesicles are formed by a
fusion of the microvilli of the syncytium, or pinocyto-
sis (128), and probably enable the transfer of whole
protein molecules, possibly those responsible for the
passive immunity of the fetus (22, 64). Other special
mechanisms occur in the pregnant uterus for trans-
ferring materials to the fetus; the fetal membranes in
the rabbit are able to transfer immune proteins, se-
creted by the uterine glands (46) and the endometrial
cups of the pregnant mare secrete gonadotrophin (9).
Finally, the functional capacity of any placenta
will also depend upon the maternal and placental
blood flows. In most animals the number of chorionic
villi and the placental weight increase rapidly after
implantation and reach a maximum while the
fetus is differentiating and before the major increase
in weight gain (25). The opportunity for exchange
between the two circulations will be limited by the
efficiency of these chorionic villi, the disposition of the
maternal and fetal blood vessels in relation to each
other, and the blood flows on either side of the pla-
cental barrier. Bumm (48) suggested that a counter-
current flow mechanism might exist to facilitate ex-
change across the barrier in the human placenta, and
Mossman (137) demonstrated that suitable anatomical
arrangements of the blood vessels were, in fact, to be
found in the ground squirrel and in the rabbit. Figure
3 shows the probable direction of the two blood
streams in the sheep; fetal blood, passing through the
chorionic vessels has the opportunity of exchanging
with the maternal arterial blood before leaving in the
umbilical vein for the fetus. Similar arrangements
exist in all species with a labyrinthine placenta (8).
In the hemochorial placenta of the primate the prin-
ciple of countercurrent flow is insured functionally:
the maternal arteries enter the intervillous space
through funnel-shaped openings, and the blood is
projected up to the base of the chorionic villi to ex-
change with fetal blood leaving for the umbilical vein
(153). There has been much controversy over this
circulation through the years but recently elegant
radiological demonstrations by Borell et al. (38) in the
human, and by Ramsey (153) in the macaque
THE FETAL AND NEONATAL CIRCULATION
l62 ;
monkey, leave little doubt that there is a blood flow
mechanism in these placentas which approaches the
efficiency of the countercurrent methods (31).
EARLY DEVELOPMENT OF THE
CARDIOVASCULAR SYSTEM
Streeter's label for the fetus as a whole, "Open
for business during alterations'" is most readily ex-
tended to the cardiovascular system: this is the first
organ system to reach a functional state in the embryo,
it supplies all the embryonic tissues and undergoes
rapid and extensive alterations during the develop-
ment of the organs.
Peripheral Circ illation
Again, Hamilton et al. provide a detailed account
of the morphology of the development of the mam-
malian cardiovascular system (102). But, what deter-
mines the appearance of isolated endothelial cords,
first in the yolk sac area, and then in the embryo,
with the eventual formation of diffuse plexuses?
Why do lumina develop in these cords and why do
larger channels form? What is responsible for the
elaboration of the neighboring mesenchyme into
the tunica media and adventitia? Why does the heart
form and become differentiated to direct the blood
through these channels? None of these questions can
be fully answered but, following the study of the his-
togenesis of the arteries in the chick embryo, Hughes
discusses the many factors which can influence the
development of blood vessels (113). The primitive
endothelial network is formed before the circulation
begins and is determined by genetic factors: in con-
trast, the development of the main vessels within the
capillary network is dependent upon a circulation
and the dynamic relationship between the structure
of vessels and the rate, direction and pressure of
blood within them is probably acquired early in em-
bryonic life. There are no hemodynamic measure-
ments with which to substantiate this statement, but
the classical relation between function and structure
is to be found in Benninghof and Spanner's descrip-
tion of the acardiac fetus with a normal twin (35);
all the arteries of the acardiac fetus, including the
aorta and common carotid, possessed the structure
of peripheral muscular vessels because they were
physiologically peripheral arteries of the normal
Premature- 30w«(« 9*
Full Term 3e-do««ks 3 months
x COLLAGEN • ELASTIN oELASTIN+COLLAGEN
fig. 4. The pattern of fibrous protein distribution in the
major vessels of the human infant at 30 weeks gestation, full
term, and 3 months post partum (Cleary, unpublished).
twin, whose heart circulated the blood in both fetuses.
The mammalian ductus arteriosus, on the other
hand, is a particular example of a muscular artery
joining two adjacent elastic arteries. The adult pattern
of fibrous protein distribution is present by 30 weeks
in the major vessels of the human fetus (fig. 4) ; elastin
exceeds collagen in the thoracic aorta but the propor-
tion of each is reversed in the abdominal aorta. The
percentage of elastin increases to a maximum 3
months after delivery.
The capillary networks are coarse in young em-
bryos and become more delicate and numerous dur-
ing development, but at different times in the various
tissues (142, 152) and it would be instructive to cor-
relate the degree of vascularization with the oxygen
requirements of the organs. The richness of distribu-
tion of the capillary bed will be of special importance
in the lungs, brain and cardiovascular system of the
prematurely born and postmature young.
The Heart
Ebert et al. describe the initial phases in heart
formation, from experimental evidence in the chick
and rat embryos (87, 96). In the prestreak embryo
the capacity for heart formation is widely distributed
and pulsating cardiac muscle may develop in tissue
culture of peripheral and posterior regions of the
blastoderm: later, this capacity is more restricted and
is finally limited to two definite regions which sub-
sequently fuse in the head process stage embryo.
In vitro these cells will develop into a rounded mass
of cardiac muscle and the onset of contractility is
swift and associated with the appearance of glycogen,
but not with definite mvofibrils or cross striations.
[624
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
But the morphogenesis of heart chambers can only
take place "in embryo" demonstrating, again, the
probable importance of spacial factors in develop-
ment; striated myofibrils are present by the time the
contractile activity is sufficient to circulate the blood.
Ebert (86) has shown by immunochemical techniques
that cardiac myosin, similar to that found in the
adult, is widely distributed in the very early embryo
and that the restriction of the heart-forming area
during development is accompanied by a limitation
of the synthesis of this specific protein and the com-
mencement of the capacity for synthesis in these
areas. These facts do not explain the first appearance
of myosin or actin, the origin of the nonpropagated
contractions which begin in the ventricles, or the
probable dependence of the developing cardiac mus-
cle metabolism chiefly upon anaerobic glycolysis
(>77)-
Congenital Malformation
A knowledge of the metabolic processes responsible
for embryonic differentiation would provide the
foundation for a better understanding of the causes
of congenital malformation and help to enable their
prevention. In human pediatrics, congenital ab-
normalities appear in 1 per cent of all live births
(162) and now contribute to 20 per cent of the neo-
natal deaths in countries where the infant death rates
are low (185); malformation of the cardiovascular
system is second to malformation of the nervous sys-
tem in causing this mortality. The causes of congenital
abnormality may be genetic but are chiefly due to
environmental factors (147) and there is a wealth of
descriptive information on the influence of a wide
variety of experimental procedures and chemical
substances which are teratogenic (119); each is usu-
ally effective at a certain stage of development and
may influence the organogenesis of one or a number
of the systems. Abnormalities of the cardiovascular
system in human infants are mainly associated with
the rubella virus: in the experimental animal meta-
bolic inhibitors and nutritional deficiencies of the
mother, especially of the vitamins, may produce
abnormalities; the high concentration of riboflavin
in fetal blood and the transfer mechanism which
exists in the placenta for this vitamin is probably
related to its high requirements in fetal metabolism
(131). Acute anoxaemia, due to maternal exposure
to carbon monoxide gas is known to be teratogenic
in the human infant (115) but chronic hypoxia,
though it may be effective in animals (119), is diffi-
cult to establish as teratogenic in the human. The
incidence of congenital malformation of the cardio-
vascular system is no greater in infants born to women
living at high altitudes than it is at sea level (147),
demonstrating that the adaptive processes enabling
life at lower oxygen tension also ensures an adequate
oxygen supply in developing tissues. Another aspect
of chronic hypoxia, reduction in maternal placental
blood flow, probably has most important conse-
quences for the infant in such conditions as toxemia
of pregnancy: in animals a reduced maternal placen-
tal blood supply causes "runting" and the over-all
size of the fetus is small, but there appear to be no
definite congenital abnormalities (132); this may
possibly be explained by the reduction in supply of
nutritive material without any alteration in balance
of the essential constituents. These epidemiological
and etiological facts cannot, yet, explain why only
5 to 30 per cent of the infants, born of mothers in-
fected with rubella during the first trimester, de-
velop malformations of the cardiovascular system
(95) or why the disturbance of organogenesis presents
itself in diverse forms. For example, why does normal
but misplaced growth of the large blood vessels
occur? What stimulates growth of the septum secun-
dum causing premature closure of the foramen ovale
and why is there too little reabsorption of the septum
spurium leaving Chiari's net (144)?
COURSE OF THE CIRCULATION IN THE FETUS
The probable course of the fetal circulation in the
mammal, once the major channels have developed,
is illustrated in figure 5. The most arterial blood
circulates from the placenta, in the umbilical vein,
to the liver which it perfuses; this blood leaves the
hepatic vein to join venous blood from the caudal
part of the body in the inferior vena cava. In some
species, notably the human, the monkey, and the
sheep, a proportion of the blood in the umbilical
vein short-circuits the liver and passes straight into
the inferior vena cava through the ductus venosus.
As it enters the heart the inferior caval stream is
divided by the crista dividens of the foramen ovale
(fig. 6) ; most of the blood flows straight into the left
auricle, where it mixes with a small volume of pul-
monary venous blood and passes into the left ven-
tricle, whence it is pumped mainly to the head and
upper extremities. A smaller stream of inferior caval
blood is directed to the right auricle, mixes with ve-
nous blood from the coronarv sinus and from the
UMBILICAL
VEIN
THE FETAL AND NEONATAL CIRCULATION 1 625
s.v.c.
fig. 5. Fetal circulation and probable course of the blood
through the fetal heart. [After G. S. Dawes (Bell el al. Textbook
of Physiology and Biochemistry, 5th ed., 1961).]
upper part of the body, carried by the superior vena
cava, and passes into the right ventricle; most of this
blood short-circuits the lungs, through the ductus
arteriosus, and passes to the descending aorta to sup-
ply the lower extremities or become oxygenated in the
placenta.
The presence of the ductus arteriosus and the
foramen ovale and their functional significance,
allowing the two ventricles to work in parallel, did
not escape William Harvey: "Thus, in the embryo,
while the lungs are idle and devoid of activity or
movement, as though they did not exist, Nature uses
the two ventricles of the heart as one for the trans-
mission of the blood." Harvey used the fetal circula-
tion to support his general thesis of the circulation of
the blood. Barclay et al. (27) review the history of the
anatomical evidence for the present concept of the
fetal circulation : Sabatier, nearly two hundred years
ago, observed that the foramen ovale did not lie be-
tween the two atria, but at the junction of the two
venae cavae with the left auricle, and directed the
inferior caval blood into the left auricle; it was he who
CRISTA
OIVIOENS
PULMONARY
VEINS
I.V.C.
VALVE OF
FORAMEN
OVALE
fig. 6. Diagram of the great veins to show that in the fetus
the inferior vena caval blood divides into two streams, one of
which enters the right atrium while the other passes through
the foramen ovale into the left atrium. [From Dawes (66).]
first suggested the figure-of-eight-like course for the
circulation shown in figure 5. Shortly afterwards,
Wolf also found that the two atria were not in com-
munication with each other, that the inferior vena
cava lay between them with openings in each, and
that the relationship of these communications was
such that the major portion of the inferior caval
stream would pass into the left auricle. It was not
until 1939 that the pathways of the inferior and su-
perior caval streams, in the chest and heart, were
actually observed in the sheep by Barclay et al., using
rapid serial radiography following the injection of
radiopaque substances (26). Similar observations have
been made, most elegantly, in the full-time human
infant by Lind and Wegelius who were able to make
the injections and perform the angiocardiography be-
fore the first breath (129). The latter have also con-
firmed the functional relationship between the venae
cavae and the atria in early nonviable infants at thera-
peutic abortion.
Regional Blood Flow
Was Sabatier correct in suggesting that the brain is
supplied by the most arterial blood? How much mix-
ing is there of the superior and inferior caval blood
in the right auricle? How much pulmonary venous
blood is added to the inferior caval blood in the left
atrium? Huggett, who was the first to carry out ex-
periments on the living fetus with an intact placental
circulation, found that the oxygen content of the
carotid artery exceeded that of the umbilical artery
in goats (112); Barcroft observed a 10 to 20 per cent
1626
I! Will',' II >k Ml I • I I -i Ml il I >,,\
CIRCULATION II
100
FORAMEN
OVALE
BODY
L HEART
70
130
130
R HEART
100
■200-
4-70-
v m i/-
DUCTUS
ARTERIOSUS
fig. 7. To show that both sides of the fetal heart work in
parallel; the approximate volume of blood flow through the
principal vessels, in the lamb, is indicated in ml/kg/min.
(From G. S. Dawes. Changes in the circulation at birth. Brit.
Med. Bull. 17: 149, 1 961.)
difference in saturation in favor of the carotid artery
in the sheep (25). Everett and Johnson injected
labeled phosphorus into the superior or the inferior
vena cava and, from its partition in the left and right
atria, were also in favor of the Sabatier hypothesis
(91). Evidence which suggests that the upper half
of the body may require a better oxygen supply is
provided by Spratt who added metabolic inhibitors
to the developing chick embryo in vitro; he concluded
that the developing nervous system depended pri-
marily upon oxidative metabolism, in contrast with
the heart which depended chiefly upon anaerobic
glycolysis (177). It was also shown, by tissue slice
technique in the sheep that the requirements of the
brain per gram of tissue increased during the last
third of gestation but that the proportional oxygen
uptake of the brain per kg of body weight remained
constant, at about five times the adult value (53).
Eranko and Karvonen, however, could find no
difference in the number of hemopoietic foci between
the lower and upper limb bone marrow of fetuses
which might be expected if the oxygen tensions of
the two bloods were different (90). Dawes and his
colleagues observed that the oxygen content of the
carotid artery only exceeded that of the umbilical
artery by 6 per cent, when both were sampled simul-
taneously, in the lamb (76) and the monkey fetus
(71); greater differences were observed following
hemorrhage, constriction of the umbilical cord, or
hypoxia, especially in young fetuses (68). These ob-
servers also approached the problem more quan-
titatively by estimating, simultaneously, the oxygen
content of the blood in the two venae cavae and,
after the two streams have mixed, in the pulmonary
trunk in the sheep at term. Similar analyses were
applied to the three other positions in the fetal cir-
culation where blood of differing oxygen content
meet, namely the upper part of the inferior vena
cava, the left atrium and the junction of the ductus
arteriosus with the descending aorta. From these
measurements they were also able, by making certain
assumptions, to calculate the blood flow in all the
principal vessels as a fraction of the cardiac output
(fig. 7) : they concluded that the similarity of the oxy-
gen content of the blood supplying the upper and
lower extremities could easily be accounted for.
Measurements of the regional blood flows and oxygen
utilizations are needed to prove the hypothesis that
the course of the circulation in the fetus is designed to
ensure the supply of the most arterial blood to the
brain and coronary circulation.
The blood flow through the various fetal organs
and through the placenta will vary both quantita-
tively and relatively to one another during growth,
and this theme has been well developed by Barcroft
(25) and Barron (28). What are the relative propor-
tions of the cardiac output which perfuse the fetus
and the placenta? What is the magnitude of the pul-
monary blood flow during development? Barcroft &
Kennedy (24) found the relative distribution of the
blood between the fetus and the fetal placental cir-
culation in the sheep to change during growth in
such a manner that when the embryo was young, the
greater part of its blood volume was in the placenta;
halfway through gestation, when the placenta had
reached its full size, the position was reversed and the
amount of blood in the placenta remained constant
while that in the fetus increased. The anatomical
limit having once been set, the rate of turnover of the
blood in the placenta will become increasingly im-
portant and the fetal heart does not "keepe holiday"
(William Harvey) but has an increasing responsibil-
ity to meet the demands of growth; the increase in
cardiac output and vasomotor tone will ensure the
gradual rise in arterial pressure upon which the
umbilical blood flow will depend. Barcroft (25) esti-
mated that at least 50 per cent of the combined car-
diac output perfused the placenta in the goat and
in the sheep near term, and Dawes el al. (76) cal-
culated a figure of 57 per cent. How this proportion
changes during gestation is not known. Cineradio-
graphic observations in the sheep (26) and human
infant (129) suggest that the blood flow through the
fetal lungs is a small proportion of the combined
cardiac output during intrauterine life. Since the
development of blood vessels is dependent upon
THE FETAL AND NEONATAL CIRCULATION
i * >-> 7
genetic and environmental factors as well as the
pressure within them, the fact that the great vessels
do enlarge and the pulmonary vascular bed does
increase must mean that there is an increasing vol-
ume of flow during intrauterine life; however, it is
not known whether the proportion of this blood flow
to the total cardiac output changes or remains con-
stant during development. Dawes et al. (76) have
estimated that, in the near term sheep fetus, about 10
per cent of the combined ventricular output perfuses
the lungs.
Hepatic Blood Supply and the Ductus Venosus
The liver is probably supplied by the most oxy-
genated blood in the body; the umbilical vein carries
well-oxygenated portal blood from the placenta
and a hepatic branch leaves, before the ductus veno-
sus, to supply the left lobe of the liver, nearly two-
thirds of the whole organ. The volume of this flow
is large, representing over 50 per cent of the cardiac
output (since the umbilical blood flow is about 57%
of the cardiac output in the lamb and probably only
a small proportion passes through the ductus veno-
sus), and the oxygen tension is unlikely to be greatly
reduced by mixture with hepatic arterial blood.
Emery (89) found more hemopoietic foci in the right
side of the liver than in the left and degenerative
changes are observed more frequently on the right
side of the liver at autopsy in stillborn infants and
following neonatal deaths (101). It is possible that the
reduction in oxygen supply to the liver, following
birth, is a factor in the development of physiological
icterus (178).
The presence of a ductus venosus is not universal
but it is patent in the lamb and monkey and in the
human infant at term (27, 71, 129); at the junction
with the umbilical vein, the vessel possesses a muscular
sphincter, which is innervated by postganglionic
branches of the vagus nerve. Cineangiography sug-
gested to Barclay et al. (26) that only a small propor-
tion of the umbilical venous blood flow passed through
the ductus venosus in the lamb but no direct meas-
urements have yet been made. It has been suggested
that the sphincter closes in response to a rise in um-
bilical venous return to the heart (155); conversely
it may regulate hepatic blood flow itself or the placen-
tal blood flow since the main resistance to the um-
bilical blood flow resides in the liver. A large flow
through the ductus venosus would ensure a good
supply of arterialized blood to the head but the fact
that a ductus venosus is not always present suggests
that no special mechanism exists for supplying the
brain with the most arterialized blood. Experimental
occlusion of the ductus venosus in the mature lamb,
caused no significant change in arterial blood pres-
sure, heart rate, or carotid arterial 0> saturation
(12). Rostral to the ductus venosus the umbilical
vein continues as the portal sinus and joins the portal
vein where arterialized and venous blood meet in
unknown quantities.
FETAL HEART
The development of activity in the mammalian
heart has been observed in hanging drop cultures of
whole embryonic rat vesicles (96) : the earliest con-
tractions occurred in the left ventricle and were fol-
lowed by a slower rhythm in the right ventricular
tube; when the two ventricles joined the left became
the pacemaker. The auricles beat a little later and the
sinus venosus last, finally bringing the ventricular
rhythm under their control at an early stage in de-
velopment.
Recording of the electrical activity of the heart
in utero has not been frequently attempted in experi-
mental animals but would provide both fundamental
knowledge of the development of the propagated
impulse and enable the fetal heart rate to be counted
with minimal disturbance during growth. The im-
pulse is large enough to record in rabbit and guinea
pig fetuses, 15 g in weight (32, 133). Recording of the
ECG of the human fetal heart in utero, using leads
placed in the mother's vagina or rectum, or on her
abdomen has been employed for many years to
monitor the heart rate, particularly during difficult
labors (127); the method is not, however, widely
used and it is possible that the electrophonocardio-
graph will be simpler and less subject to interference
from the maternal heart (175). Electrocardiograms
obtained from human infants at Cesarean section
show all the deflections characteristic of the adult as
early as the second month; in the full-term infant
there is a small right ventricular preponderance
corresponding to the slightly greater relative weight
of this ventricle at birth. The left ventricle starts to
exceed the right 3 months after birth and by 6 months
of age the deflections are usually identical with those
of the adult; this is due both to growth of the left
ventricle and to involution of the right ventricle
(121). The T wave is frequently of low amplitude at
birth and becomes negative shortly afterward (178);
the sign may be reversed again by the administration
1 628
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
of adrenaline and the observers attributed this to a
rise of pulmonary arterial pressure.
Heart Rate. Regulating Mechanisms
In the smaller animals systematic correlation of
the fetal heart rate with age has not been frequently
made and when the uterus is opened the data may be
questionable on account of cooling and hypoxia
(189). In the guinea pig and the rabbit the heart
rate has been counted from ECG records taken with
the uterine wall intact and the maternal abdomen
opened under saline at 37 C (133); the heart rates in
both increased from 160 to 320 per min during 20 to
67 days in guinea pig and 25 to 31 days in the rabbit
fetuses, but the range was wide, possibly on account
of intrauterine hypoxia. The maternal heart rates
also varied widely and within the same range as the
full-term fetus. In the guinea pig fetus a slight slow-
ing occurred when the uterine wall was incised and
allowed to contract around the large vessels supply-
ing the placenta. In the rabbit fetus postmaturity
did not influence the heart rate (134). The heart
rate of the monkey fetus at Cesarean section is 140
to 1 70 beats per min (71 ).
Barcroft and colleagues (25) counted the lamb
heart rate in utero with a stethoscope and found
that it increased during the first two-thirds of preg-
nancy to 150 beats per min and thereafter fell slowly
to 128 beats per min at term; in a larger series in
which the heart rates were obtained from blood
pressure tracings, the rate rose throughout gestation
to about 200 beats per min at term (39). The ewe
heart rate is normally 100 to 120 per min. In two
studies in humans the fetal heart rate was also ob-
served to be faster in midfetal life, 156 per min, than
just before birth, when the average was 142 per min.
These differences are small, however, and probably
not significant; in one study the counts were made
with a stethoscope (1 76) and in the second from ECG
recordings (192).
The pattern of changes in fetal heart rate in utero
will be determined by the rate of development of the
pacemaker rhythm and the onset of subsequent vagal
restraint. The anatomical pathways of the parasym-
pathetic system are laid down early, and vagal fibers
may be observed in the A-V bundle in a 6-weeks-
old human fetus (184), before the inhibitory response
of the isolated cardiac muscle to acetylcholine is ob-
served (194). Vagal tone is not apparent in utero or
subsequently in the guinea pig, rabbit, or cat: the
full-term fetal heart rates are the same as in the adult,
and in the cat the heart rate is uninfluenced, in both
the newborn and the adult, by section of the vagus
nerves (114); the latter, however, is not particularly
good evidence. Vagal tone was considered by Bar-
croft and colleagues (25) to be present in the sheep
fetus toward term, for they found that bilateral sec-
tion of these nerves increased the heart rate; this,
however, was not confirmed by Born et al. (39).
Stimulation of the peripheral cut end of the vagus
nerve will cause bradycardia in the sheep fetus half-
way through gestation, though the heart will re-
spond to intravenous acetylcholine earlier (74).
The isolated fetal heart is very sensitive to acetyl-
choline but it is not possible to correlate this with the
age of the fetus (19). The influence of atropine on
fetal heart rates and a comparison with its action in
the adult of the same species is practically unknown;
late in intrauterine life atropine in the fetal circula-
tion causes an acceleration of the fetal guinea pig
heart (97). In the pregnant woman atropine in the
maternal circulation ( 1 1 o) abolishes asphyxial fetal
bradveardia, but there is no evidence for its influence
on the normal heart rate nor independent evidence
for its placental transfer.
The sympathetic pathways are known to be laid
down early in development in the kitten and the hu-
man fetus (44, 103); the lamb heart is able to acceler-
ate in response to intravenous adrenaline two-thirds
of the way through the gestation period and at term
the sensitivity of the fetal heart is little different from
that of the adult in both the sheep (74) and the rabbit
(70) ; earlier observers frequently found a decreased
sensitivity and this was possibly due to anoxia (194).
Again, the isolated heart is sensitive to adrenaline
and noradrenaline and there is no correlation with
the age of the fetus (19).
Cardiac Output
Fetal cardiac output was measured in the goat
both cardiometrically and using the Fick principle
by Barcroft and his colleagues (25); they estimated
that it increased from 1 1 3 ml per kg body weight per
min at 89 days of age to 193 ml per kg body weight
per min at 150 days, full term. Dawes et al. (75) cal-
culated that the cardiac output of both ventricles in
the lamb at term was 235 ml per kg per min, know-
ing the umbilical blood flow and estimating that it
formed about 57 per cent of the combined cardiac
output. Assali et al. (17) made similar calculations
in human fetuses of 9 to 28 weeks gestation and found
the cardiac output to be 200 ml per kg per min;
THE FETAL AND NEONATAL CIRCI LATION
[629
their assumptions a) that the umbilical blood flow is a
constant fraction of the cardiac output during this
period of rapid growth, and h) that the umbilical
blood flow forms the same proportion of the cardiac
output in both the human and in the sheep fetus have,
to date, no foundation.
ARTERIAL BLOOD PRESSURE
Systemic Pressure
The rate of increase in systemic arterial pressure
during gestation varies among the species and the
final values at term correspond most nearly to the
requirements of the newly born: for instance, in the
helpless newborn of the rat and rabbit the mean pres-
sure in the carotid artery is only 30 mm Hg after 21
and 31 days of gestation, respectively (49, 70); the
newborn kitten and puppy are also born with ar-
terial pressures of 30 mm Hg after 67 days (114)
while the active guinea pig is born with an arterial
pressure of 50 mm Hg after a similar time in utero.
Arterial pressures of 60 to 70 mm Hg are observed in
the newborn lamb and kid following 147 days ges-
tation (25), and in the newborn human babe after
an intrauterine life of twice this duration (196).
The rhesus monkey has a mean arterial pressure of
about 55 mm Hg at birth after 160 days gestation
(71). An increase in arterial pressure during intra-
uterine life must assist in increasing the umbilical
blood flow and the opportunity for exchange between
the mother and fetus; however, this is only one means
of meeting the increasing demands of growth, and
the potentialities of the placental and fetal tissues
vary among the species (98).
The course of the rise in arterial blood pressure
during intrauterine life is shown in figure 8 for the
lamb. It is impossible to assess the relative parts
played by alterations in cardiac output and the de-
velopment of vasomotor tone in contributing to
these changes. After about 90 days gestation in the
lamb, when the arterial pressure rises more rapidly,
the heart rate continues to increase but no cardiac
output measurements are available; Barcroft's (25)
results in the goat suggest that there may be an
increase in cardiac output in relation to body weight
from go days onward and, as the umbilical blood
flow in the lamb decreases in relation to body weight
during the same period, the mean body blood flow is
probably increased. However, the cardiac output at
term is greater per kg of body weight than in the
adult and the low arterial pressure may be accounted
for by a low peripheral resistance: as will be seen,
this low resistance is probably due to low tonic ac-
tivity of both nervous and chemical regulating mecha-
nisms.
Pulmonary Artery Pressure
In utero, before the lungs are inflated with air
there is no good reason why the pulmonarv vascular
resistance should be widely different from the vascu-
lar resistance elsewhere in the growing fetus. Ardran
et al. (13) find in the lamb that the pressure in the
left pulmonary artery is about 5 mm Hg higher than
that in the carotid artery, which suggests that the
vascular resistance in the lungs before birth is pos-
sibly slightly higher than the combined resistance of
the fetal tissues and the placenta; this has recently
been confirmed by Assali et al. (16). In keeping with
these observations are the findings that the thickness
of the walls of the two ventricles is approximately the
same during development, with a slight preponder-
ance of the right over the left, in the lamb and in the
human infant at birth (66).
Development of the Cardiovascular Reflexes and the
Responses to Asphyxia and Hormones
The anatomical pathways for the cardiovascular
reflexes are laid down early in development in both
the human fetus (44) and in the cat (103), but, as
predicted by Barcroft, though the machinery is
ready it may not be functional and the stage of
gestation at which the cardiovascular reflexes are
0
70_
X
0
MB *
•
•
O • •
0 %
E
• 0
0
E
0
50 _
a.
CD
_J
O
•
0
•
0 0
i .•
<
•
0
0
tr
UJ
a.
•
°.
•
0*T>0
0
0
0
<
•
*•< •
8 .•
30_
\
•
•
0
•
• •
a
•
•
GESTATION
AGE
IN DAYS
1
1
1 1
1
1
1
1
1
60
8O
IOO
I20
140
fig. 8. Systemic blood pressure of fetal lambs, under
dialurethane (o) or pentobarbitone (•) anesthesia. [From
Dawes (66).]
1630
HANDBOOK (J1- I'HYSIOI.OOY
CIRCULATION II
operative varies among the species. The earlier
work is described by Barcroft (25) and the later by
Dawes and his colleagues (39, 70, 71, 74).
The responses to asphyxia and to the intravenous
administration of hormones have been most gener-
ally used to determine the activity of the cardio-
vascular system in the fetus: the low resistance of the
placenta, the low arterial oxygen tension, and the
fetal course of the circulation must also influence the
final operation of the reflexes. Quantitative data are
difficult to obtain when both the peripheral and cen-
tral mechanisms have not yet reached a steady rela-
tionship with each other. However, in the lamb, the
steeper rise in arterial pressure which occurs from
90 days onward approximately coincides with the
development of increasing responsiveness to asphyxia,
as judged by the rise in arterial pressure and heart
rate (39) : further, the removal of sympathetic tone,
following the injection of a ganglion-blocking agent
such as hexamethonium causes a greater fall of
blood pressure toward term. The tone of the vasomo-
tor mechanisms is probably not fully developed at
birth for the mean arterial pressure is about 40 mm
Hg lower than in the adult sheep. In the rabbit,
cat, and dog, with low arterial pressures at term, the
vasomotor mechanisms are probably still less de-
veloped at birth (70, 1 14).
The pattern of the response of the developing
cardiovascular system to asphyxia alters with gesta-
tional age. The bradycardia which follows either the
occlusion of the umbilical cord or the administration
of low oxygen tensions to the mother is probably
brought about in a variety of ways. In the early fetus
of all species cardiac slowing is delayed, it is due to
the direct effect of the hypoxia on the pacemaker
and is the cause of the ensuing hypotension; this
depression of the pacemaker is the final cause of
death at any age when hypoxia is prolonged. Later in
development, a transient bradycardia of swift onset
is observed, which is due to stimulation of the medul-
lary vagal center; later still, this slowing is succeeded
by a tachycardia, due to stimulation of the medullary
sympathetic center. This response is enhanced by
cutting the vagus nerves. The third type of brady-
cardia is reflex in origin and occurs in response to
the rise in blood pressure when vasomotor, baro-
receptor, and chemoreceptor activity is developed;
the bradycardia seen in the fully developed lamb or
human fetus is probably reflex in origin provided the
asphyxia is of short duration. It is noteworthy that
prolonged asphyxia or hypoxia reduces the heart
rate to between 60 to 80 beats per min in most species;
this rate is sustained for varying periods before ar-
rhythmia occurs. Both the tachycardia and the
reflex bradycardia during asphyxia may be enhanced
by the activity of the adrenal medulla (54). Bradv-
cardia is also the primary response to hypoxic stimu-
lation of the chemoreceptors in the adult animal (60,
61); however, if the brain is also hypoxic, tachycardia
usually results from hypoxia of these areas and this
tachycardia is enhanced it their oxygen supply is
increased.
Quantitative data relating the oxygen saturation
of the fetal blood at which the changes in heart rate
take place in utero have been provided in the near-
term lamb by Born et til. (39), and by Reynolds &
Paul (161). These observers are not entirely in
agreement with each other; both administered
nitrogen containing low concentrations of oxygen to
the mother under barbiturate anesthesia. Born el al.
delivered their lambs by Cesarean section and ob-
served an increase in heart rate and blood pressure
during the administration of 7.5 to 5.0 per cent oxy-
gen to the mother which caused the fetal carotid
arterial oxygen saturation to fall to 50 to 35 per cent;
bradycardia did not occur until the arterial oxygen
saturation was below 20 per cent for some minutes
(fig. 9). Reynolds and Paul's lambs were kept in
utero and blood pressures and blood samples were
obtained from branches ot the umbilical vessels
exposed through a small uterine incision. Their
results were not so clear cut for fetal tachycardia or
bradycardia might be observed following the ad-
ministration of 13 per cent and 10 per cent oxygen
to the mother and 6 per cent oxygen usually caused
fetal bradycardia; it is to be noted that the adminis-
tration of 13 per cent oxygen reduced the arterial
oxygen saturation to 30 per cent, a figure which was
obtained by Born et til. with much lower oxygen mix-
tures. In the guinea pig, also anesthetized with
Nembutal, the administration of 10 per cent oxygen
to the mother caused a slight fall in fetal heart rate,
during the last third of gestation, while 6 per cent
oxygen always caused marked bradycardia (97, 133).
The absence of cardiac acceleration in response to
asphyxia early in gestation is not due to the inability
of the young heart to increase its rate, for it will
respond to adrenaline early in development: in the
lamb the heart is more sensitive to adrenaline than
are the peripheral vessels for tachycardia occurs at a
time when the increase in pressure is relatively small;
later the rise in blood pressure is greater and the
increase in heart rate diminished, due to the de-
velopment of baroreceptor reflex activity (74)- Acetyl-
THE FETAL AND NEONATAL CIRCULATION
I 63 I
fig. 9. Response of the fetal heart rate and arterial pressure
in the lamb during hypoxia. The rise in blood pressure is accom-
panied by a) tachycardia during the administration of 10' ", < >
to the mother, b) bradycardia during ventilation with 6% Oj.
[From Born et at. (39)-]
choline also causes bradycardia and hypotension
early in gestation. Dawes and his colleagues consider
that the range of effectiveness, per kg of body weight,
of those autonomic drugs does not differ from 60 to
160 days in the lamb, and is about the same as in
the adult for both the lamb and the fetal rabbit (70,
74). In the lamb at term, with ventilation established,
adrenaline and noradrenaline cause a greater rise in
blood pressure following occlusion of the umbilical
cord, when the low resistance circuit of the placenta
is absent. Equal doses of these drugs are also more
effective when injected into the femoral vein and pass
straight to the left side of the heart and to the coro-
nary circulation, than after injection into the jugular
vein when the drug has first to traverse the lungs.
Since suprarenal venous blood enters the inferior
vena cava, there will be the possibility of a rise in
the fetal blood pressure and an increase in the placen-
tal blood flow during stress; the effectiveness of sym-
pathomimetic amines liberated during the stress of
asphyxia in the fetus may be limited by the reduced
responsiveness of the cardiovascular system during
asphyxia (195). The adrenal glands and accessory
organs contain a pressor substance early in develop-
ment in the sheep (54) and in the human infant
(186). It is also known that sympathomimetic amines
are released into the adrenal veins during asphyxia
by about 90 days gestation in the sheep; this libera-
tion is due to the direct action of asphyxia on the
adrenal medulla. The splanchnic nerves do not
take part in the release until shortly before term.
It is perhaps significant that noradrenaline predomi-
nates for its pressor activity is the greater and the
stimulating action on metabolism apparently more
effective than that of adrenaline in the young animal
(136).
The catecholamine concentration of human fetal
heart, kidney, and lung during the first trimester
was found to be roughly similar to that in adult
organs though the brain contained much smaller
concentrations than in the adult (99); again, norepi-
nephrine predominated suggesting its early appear-
ance at sympathetic nerve endings, but no dopamine
was found in any tissue studied. 5-Hydroxytrypta-
mine is found in the blood platelets of the fetal guinea
pig two-thirds of the way through gestation and is
still lower than the adult at term (174); the brain
levels, however, approximate to those of the adult
at term (120). The high estrogen and progesterone
content of fetal blood (4) may influence both cardio-
vascular development and the responses of the vessels.
The action of many other pharmacological substances
on the fetus has been reviewed recently (20).
The importance of cardiovascular regulating
mechanisms to the fetus in utero is questionable;
asphyxia and hemorrhage are probably the only
stresses which the fetus encounters. The responses to
hemorrhage have not been frequently studied and
Mott suggests that they might possibly be a better
indication of the homeostatic capacity of the fetal
circulation than the response to asphyxia since the
fetus is more resistant to hypoxia than the adult
(140).
FETAL PLACENTAL BLOOD FLOW
Effective Perfusion Pressure. Resistant e 0]
the Placental Circulation
The effective perfusion pressure across the fetal
placental circulation increases as the arterial pressure
rises with gestational age. Figure 10 shows some com-
parative values for umbilical arterial and venous
pressure measurements in the lamb (25). Reynolds &
Paul (160) have observed umbilical venous pressures
as high as 35 mm Hg in the lamb at term; the reason
for this rise in umbilical venous pressure may be re-
lated to an increased resistance to flow in the fetal
liver, through which most of the umbilical blood
passes during development. The sphincter of the
1632
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
1 "T"
1
SO
—
• Umbilical art
0 Umbilical veir
>ry
1
1
00
X
E
B
< 1
u
3
to
•
<
1
»
•
•
u
Pu
1
'
20
1
4
O
J
j
i
0
c
1
)
i
0
<
1
i
0
Fetal age -days
1 . 1
50 70 90 110 130 150
fig. 10. Pressures in umbilical veins and arteries at successive
fetal ages in the lamb. [From Barcroft (25).]
ductus venosus may regulate this pressure and thus
both placental and hepatic blood flows in this species.
From figure 10 it can be seen that the pressure drop
across the placenta is about 40 mm Hg in the lamb at
term and, in comparison, the pressure drop in the
systemic circulation is of the order of 60 mm Hg; since
about 60 per cent of the cardiac output goes to the
placenta the resistance in the fetal placental circula-
tion is about half that of the fetal systemic circulation.
The resistance in the fetal liver is still lower than in the
placenta for the greater part of the umbilical blood
flow traverses this organ with a pressure fall of only
20 mm Hg. No arteriolar regulating mechanism has
been described in the chorionic villi but B0e (37) has
demonstrated, in the human placenta, the existence
of a shunt mechanism within the villous circulation
which may possibly open up during asphyxia and
increase the fetal placental reserve. The walls of the
capillaries in the chorionic villi have neither smooth
muscle nor a nerve supply, but in teased specimens
their endothelium has been observed to undergo
spontaneous rhythmic movements and to be con-
stricted by histamine and acetylcholine (179).
When the placenta has reached its maximum
weight in the lamb, at 80 days gestation, the pressure
drop across the placenta is 25 mm Hg; during the last
third of intrauterine life this pressure drop only in-
creases by a further 15 mm Hg while the umbilical
140
50
MM HG.
AORTIC -UV PRESSURE
fig. 1 1 . Pressure How curves for the fetal placental circula-
tion at 90, 115, and 140 days gestation in the sheep. [From
Dawes (69).]
blood flow is increasing tenfold. From pressure-flow
measurements in the umbilical circulation (fig. 11)
Dawes concludes that the increase in flow is chiefly
brought about by a decrease in placental vascular
resistance; at the end of term no further decrease in
resistance occurs and the increasing flow is dependent
on the rising pressure gradient (69).
Umbilical Blood Flow
Umbilical blood flow has been measured in the
sheep fetus the most frequently and by a variety of
methods. Cooper el at. (56) used the venous occlusion
plethysmograph (see fig. 12) and found that the blood
flow per kg of fetal weight ranged from 250 ml per
min at 60 days gestation to about 1 30 ml per min at
term, 147 days. The actual flows and their decrease
in relation to body weight are in good agreement
with the later observations of Acheson el al. (1) using
the same technique but a different breed of sheep
(fig. 13). Reynolds et al. (158) made a few measure-
ments of the blood flow in the umbilical artery
in the lamb by cineangiography and concluded that
there was no reduction in relation to body weight at
the end of term; this suggests that the fall in blood
flow using the plethysmograph may be an artifact
due to the greater sensitivity of the umbilical vessels
at this time. In the guinea pig the venous occlusion
plethysmograph gave values of 45 to 108 ml per kg
per min, with no tendency to change as the fetal
weight increased (172): the arterial pressure is lower
THE FETAL AND NEONATAL CIRCULATION
1633
Soft rubber
seal
To volume
recorder
Perspex curtain Perspex window
hanging into /for direct
saline to / observation of
make fluid seal / umbilical vessels
Pneumatic
bag for
compression
of cord
Umbilical
cord
Electric lamp
Tinned brass tank
fig. 12. Section through a fetal plethysmograph at the
point of entry of the umbilical cord. The umbilical cord lies on
a gently curved perspex strip. [From A. D. M. Greenfield. A
foetal plethysmograph. J. Phyiiol., London 108: 158 (Fig. 2),
1949]
J00
O
-s
0
<
0
ui
a
5?
-1
O
u.
^•Oy
0
UJ
200
. 0
0
0
•
X^-^j5
<■
UJ
0:
m~
0
-— _ r^ '---. •*
(J
O z
"^4
0 c*
7 — -^^ ° 0 --^
-l<
vrc---:^5~~~
too
-31
3
•
GESTATION AGE (DAYS)
0 ,
fig. 13. Umbilical blood flow in the lamb, per kg body
weight during gestation. [Data of K. E. Cooper el al. (O )
G. H. Acheson el al. (• ).] The thin continuous curved line
indicates the weight increase per cent per day. [From Acheson
el al. (1).]
than in the sheep and the pressure gradient between
artery and vein is likely to be smaller and, assuming
that the vascular resistances are similar, this will
account for the lower placental blood flow. When
these umbilical blood flow rates are compared with
growth curves it is observed that 5.5 liters of blood
are required to lay down 1 .0 g of fetal tissue in the
sheep, as compared with only 1.3 liters in the guinea
pig, (98). Using an electromagnetic flowmeter Assali
et al. (17) found the umbilical arterial flow in nine
human fetuses of 12 to 28 weeks gestation to range
between 94 and 127 ml per kg per min. It is remark-
able that vessels so contractile as those in the cord
have yielded, on the whole, reproducible results.
Each worker has been most aware of the experimen-
tal errors involved in his measurements. Dawes &
Mott (72) also point out that the venous occlusion
plethysmograph has a disadvantage in the present
application, for when the umbilical vein is temporarily
occluded the return to the heart must be reduced;
they found that a velodyne flowmeter, providing a
direct measure of flow, inserted into the vein in the
abdomen gave results which were higher than those
obtained by the plethysmograph.
The umbilical blood flow may be increased at the
end of term in the lamb by reducing the fetal ar-
terial oxygen saturation (39). This is probably mainly
due to the rise in arterial pressure caused by the re-
sponse of the fetal vasomotor center to the altered
chemical composition of the blood. Reynolds & Paul
(160) found that no rise in umbilical venous pressure
accompanied the rise in umbilical arterial pressure
and suggested that the tone of sphincter of the ductus
venosus was decreased in response to the increased
umbilical venous flow. The injection of adrenaline
into the femoral or jugular vein of the fetus causes an
increase of umbilical blood flow which is proportional
to the rise in arterial blood pressure (74) (fig. 14).
Isolated umbilical vessels are very sensitive to the
vasoconstrictor action of adrenaline and these results
suggest that the hormone is destroyed before it reaches
the umbilical vessels; no figures are available to
show how the hormone influences the resistance in
the placenta. The umbilical blood flow is reduced
100
80
60
40
20
§5
o
- o
o
.0
h E
"o
V
O
•o
O o°
• cfo
o
Increase of systemic blood pressure (%)
l l I i i i l_
20 40 60
fig. 14. Increase in umbilical blood flow following the in-
jection of adrenaline (•) or noradrenaline (O) in the mature
fetal lamb. [From Dawes el al. (74).]
1634
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
when the arterial pressure falls following the injec-
tion of hypotensive drugs such as acetylcholine and
hexamethonium into the fetal circulation (74) and
following severe hvpoxaemia of 10 to 1 r, min duration
(69)-
Oxygen Requirements and Environment of the Fetus
Primarily dependent upon the maternal placental
circulation, the fetal heart provides an umbilical
blood flow which, under normal conditions, main-
tains a steady oxygen consumption of 4 to 6 ml per
kg body weight per min in both the lamb during the
last half of gestation ( 1 ) and in the human fetus of 9
to 28 weeks gestation (17): this represents an oxygen
consumption in relation to weight comparable with
the adult and the constancy is remarkable in view of
the changing oxygen utilization of the various organs,
and their varying weights in relation to each other,
during development. Huckabee et al. ( 1 1 1 ) point
out that without a knowledge of the anaerobic metab-
olism of the fetus it is impossible to obtain an accurate
estimate of the energy requirements of growth from
the quantity of oxygen consumed alone; however,
there is no good evidence for anaerobic metabolism
in the normal fetus for blood lactate levels are com-
parable with the adult (71). Huckabee et al. also
point out that the metabolic rate of the fetus, if it
were known, is not synonymous with the metabolic
rate required for the life and growth of the fetus and
the metabolic needs of the placenta must be included.
These observers found in the goat, as did Assali et al.
(17) in the human, that the oxygen consumption
of the pregnant uterus was about 10 ml per kg per
min; the calculations were made lrom uterine blood
flow and A-V O2 differences. But, while Assali et al.
consider the placenta to have a greater oxygen
consumption than the fetal tissues, Huckabee et al.,
from uterine oxygen utilization measurements after
fetal death, suggest that this may be the reverse;
the latter estimate of fetal oxygen consumption, as
approximately 10 ml per kg per min, would agree
with determinations of the minimal oxygen consump-
tion of the newborn lamb. However, Dawes and
Mott have shown that such a high oxygen consump-
tion is characteristic of the newborn only and is
attained at different ages in the different species;
further, they have demonstrated that this increase in
oxygen consumption is not dependent upon the raised
arterial oxygen saturation following the establishment
of respiration for it does not occur in immature lambs
delivered by Cesarean section and artificially ven-
tilated (72).
What is the oxygen environment of the fetal tissues
in utero? Recently, Misrahy et al. (135) have measured
the oxygen availability (a02) in fetal brain and kid-
ney, in a number of species under Nembutal anes-
thesia. Nondiffusion limited polarographic elec-
trodes, too n in diameter, with a circumferential
recording surface, 2 mm in width, were inserted into
the tissues, with little disturbance of the uterine wall;
the aO; ranged between 18 per cent and 30 per cent
of the diffusion current in air, corresponding to 30
to 45 mm Hg O2 and was similar to the maternal
tissue oxygen tensions measured in the same manner.
Misrahy el al. consider these readings to represent
the rate of oxygen transport between the capillaries
and the active cells. The tension of oxygen in the
arterial blood of the fetus is, however, probably con-
siderably lower than that in the maternal blood. It is
difficult to assess the values for arterial oxygen satura-
tion in utero for when the uterus is opened and um-
bilical vein samples collected the placental circulation
is impaired to an unknown extent. Westin (187) has
shown, by hysterophotography, that the oxygen
saturation of the fetal blood is probably high in 14
to 18 week human fetuses for the skin is pink and the
umbilical vein arterial in color in utero, and Dawes
and his colleagues ( 1 , 39) have observed arterial
oxygen saturations as high as 74 per cent in the lamb
near term. In the human at term, blood collected from
the choriodecidual space by placental puncture has
been reported to have a mean pO> of 38 mm Hg;
the p()j in umbilical vein blood is probably 10 mm
lower (31). These results in the human should, how-
ever, be regarded with reserve for there is no means
of knowing whether the sample of blood obtained
comes from the choriodecidual space or a uterine
vein.
How is equality of oxygen availability to the fetal
and adult tissues attained in spite of the low arterial
oxygen saturation in the former? The mechanisms
appear to be, for the most part, similar to the adult
response to low arterial oxygen saturations. First,
the fetal blood has a greater affinity for oxygen than
the maternal blood; this is a property of the fetal
hemoglobin and its environment in the red cells
which enables fetal blood to leave the placenta with a
greater oxygen saturation than the maternal blood
at low oxygen tensions. The factors involved in the
transfer of 0> and CO» between the maternal and
fetal circulations are clearly outlined by Barron &
Meschia (30) and Bartels et al. (31). Second, there is a
THE FETAL AND NEONATAL CIRCULATION
l635
steady rise in the blood hemoglobin in fetuses of all
species during gestation. Most are born with levels
which are higher than that of the mother (29) and
erythropoietic concentration is known to be high in
the cord blood of many species (6, 41). Third, the
average blood flow through the fetal tissues is high.
This has not been compared with the adult values
for each individual tissue but estimates of the fetal
cardiac output in the lamb are high, as already de-
scribed and amount to an average tissue flow of
about 120 ml per kg per min, which is at least twice
the flow in the adult sheep. These cardiac output
measurements have been calculated indirectly from
umbilical blood flow measurements and the dis-
tribution of blood within the fetus with an open chest
and are, therefore, probably an underestimate.
Approximate calculations for the human fetus also
suggest that the average body blood flow is high.
INFLUENCE OF HYPOXIA AND ASPHYXIA
ON THE FETUS
The effects of a prolonged reduction in oxygen
supply have been observed in fetuses born to mothers
at high altitude and the possibility of hypoxia as a
cause of congenital malformation has already been
discussed : experimentally the influence of acute hy-
poxia, produced by maternal breathing of low-
oxygen gas mixtures has been studied the most
frequently. The results of true asphyxia may be ob-
served during marked impairment of the maternal
placental circulation or the mechanical obstruction
of the umbilical vessels.
Hemoglobin
The possibility of a rise in blood hemoglobin
concentration, in response to a reduced oxygen
supply, first attracted the attention of Joseph Bar-
croft (25) who correlated the fetal hemoglobin level
with the percentage saturation of the umbilical vein
blood with oxygen, in the lamb at term. This idea
has proved most controversial clinically (118, 182),
particularly because subsequent investigators did not
heed Barcroft's warning concerning the difficulties of
collecting a good specimen of umbilical vein blood,
and his awareness of the variety of conditions which
might bring about rapid changes in the oxygen satu-
ration of cord blood. Neither the oxygen saturation
of the blood in the umbilical vessels at birth, nor
the total level of hemoglobin in the blood and the
relative proportion of fetal hemoglobin, contributing
to this, have proved to assist in the interpretation of
either the extent or duration of any impairment of
the intrauterine environment. However, it is now
certain that fetal hemopoietic tissues can respond
when oxygen availability is reduced for the young
born to llamas at 15,000 ft have higher blood hemo-
globin concentrations than those born at sea level
(150). It is interesting to speculate on this response
at altitude in the fetus: as described, erythropoietic
concentration is high in cord blood at sea level and
may represent the fetal response to low arterial oxy-
gen tensions despite the adequate availability of
oxygen to the majority of fetal tissues; adult hemo-
poietic tissue, however, will also respond at 25,700 ft
(7,830 m) when the arterial oxygen tension is reduced
to 33 mm Hg (151), a value which is normal for the
fetus. Born et al. have shown that fetal hemoglobin
concentrations increase during acute hypoxia in the
lamb, which suggests that red cells may have been re-
leased from the spleen or there may only have been a
loss of plasma to the extracellular space (39).
Blood Flow
It is doubtful if the possibility of increasing the
tissue and placental blood flow in response to a re-
duction in oxygen supply is significant in the fetus:
in the lamb, younger than 60 days of gestational age,
both umbilical and tissue blood flows will probably
fall since there are no reflex mechanisms to elevate
the blood pressure and the depleted oxygen supply
will cause bradycardia. Later, when the cardio-
vascular reflexes begin to be developed, a rise in
arterial pressure and carotid and umbilical blood
flow is observed in response to asphyxia or low oxy-
gen tensions; these responses occur when the fetal
arterial oxygen saturation is reduced to 50 to 35
per cent, following the administration of 7.5 to 5.0 per
cent oxygen in nitrogen to the mother (39). The
increase in carotid and umbilical blood flows prob-
ably occurs at the expense of the blood supply to the
major portion of the body, for it is unlikely that the
cardiac output increases: there is no experimental
evidence to support this statement, but it is known that
hypoxia does not increase the cardiac output in the
newborn lamb (57); this lack of response is possibly
related to the very high cardiac output at this time,
for the fetal cardiac output at term is at least twice
that of the adult per kg body weight. The decrease
in oxygen consumption of the hind quarters in the
lamb during hypoxia may be evidence for peripheral
vasoconstriction.
[636
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
12 -
100
fig. 15. Oxygen consumption per kilogram body weight at
different arterial oxygen saturations. (•) Fetal lambs; (©)
lambs less than 1 day old; (o) lambs 1-10 days old after the
rise in minimal oxygen consumption (95ro confidence limits
are shown); horizontal lines indicate the range of observations.
[From Cross el al. (57).]
Oxygen Consumption
The fetus has a third mechanism of defense at
low oxygen tensions, that of lowering its oxygen con-
sumption. Cross et al. have shown that this occurs
when the umbilical arterial oxygen falls to 50 per cent
saturation in the lamb, and the effect increases as the
arterial oxygen tension is reduced still further (57)
(fig. 15). The fall in oxygen utilization may be due
primarily to the decrease in blood flow to the majority
of tissues, as discussed above, for it is accompanied
by an accumulation of lactic acid and depletion of
tissue glycogen stores (171). It would be interesting
to know if the tissue temperature falls as it does in the
adult when the blood flow and oxygen supply are re-
duced to muscle (169). A reduction in oxygen con-
sumption with low arterial oxygen tensions is not
readily demonstrated in the adult animal, for the
cardiac output increases and the heart is liable to
sudden failure before low oxygen tensions are reached
(1). The inability of the fetal cardiac output to in-
crease, and the capacity of the heart to continue to
beat during asphyxia, must be important for survival
during birth.
The oxygen consumption of newborn animals, at
their neutral temperature, increases after birth at
intervals which vary with the species (67, 72). In
the lamb, the minimal oxygen utilization is trebled
within 24 hours of delivery to correspond with the
metabolic requirements of its surface area, and usu-
ally no shivering occurs (72): this recently acquired
increase in oxygen consumption is not well main-
tained when the arterial oxygen saturation is lowered.
Hill has also observed that the increase in oxygen
consumption without shivering, in response to a
low environmental temperature, is particularly sus-
ceptible to hypoxia (107).
Heart Rate During Reduction in Maternal
Placental Blood Flow
The influence of asphyxia on the fetal heart rate
in utero and its relationship to the degree of reduc-
tion in maternal placental blood flow, or the short-
term placental reserve, has important practical
applications. The physiology of the response of the
fetal cardiovascular system to asphyxia has already
been discussed. It is generally agreed that tachy-
cardia is the first indication of intrauterine asphyxia
at term in the human infant (108), and in the lamb.
Born et al. observed that tachycardia did not occur in
the lamb until the umbilical arterial oxygen satura-
tion was reduced to 50 to 35 per cent, during the
administration of 7.0 to 5.5 per cent oxygen to the
maternal sheep; bradycardia, most usually associated
with intrauterine asphyxia, was not observed until
the oxygen saturation reached 20 per cent (39).
The influence of the accumulation of carbon dioxide,
occurring in asphyxia, is not known. The time
course of both the cardiac acceleration and slowing
observed experimentally and clinically is very vari-
able, depending upon the rate of onset and degree
of asphyxia induced, the existing oxygen environ-
ment and the previous asphyxial history. For in-
stance, a sustained acceleration is readily observed
as the tension of oxygen administered to the maternal
animal is gradually lowered, but it is only transient
when nitrogen is inspired by the mother or when
the cord is tied; frequently acceleration does not
precede the bradycardia in the latter circumstances.
Hon (108) has described two time courses for fetal
bradycardia in the human infant during labor: the
one, which he describes as physiological, occurs fol-
lowing a uterine contraction and most usually in
vertex presentations; the heart slows briefly and
recovers swiftly within 15 sec. The second, which
Hon calls pathological, has a longer time course and
is considered to be possible evidence of previous
asphyxia or a permanent reduction in uterine blood
flow. An example of the influence of limiting the
THE FETAL AND NEONATAL CIRCULATION
I&37
Intra- uterine
Uterus open
22
24
26
1 2 3 4 S 6 20
Time (mln)
fig. 16. In the guinea pig, of 54 days gestation, a) the fetal
heart slows when the uterus is opened; h) marked fetal brady-
cardia occurs following the maternal injection of noradrenaline
only after the uterine wall is opened. [From Martin & Young
(■33)-]
blood supply to the fetus upon the susceptibility to
further asphyxia as shown in figure 16; after the
uterine wall was opened, to expose the fetus, the
uterine muscle contracted away from the incision
around the uterine blood vessels, and cardiac slowing
could be produced with a smaller dose of vasocon-
strictor substance in the maternal circulation than
when the fetus was in utero.
Reduction in the uterine blood flow giving rise to
these various patterns of fetal heart rate changes may
be brought about in many ways. It follows the injec-
tion of either hypotensive (82, 194) or vasoconstrictor
(34) drugs into the maternal circulation; the effect
will be reversible or not according to the dose given
and the duration of action of the pharmacological
substance. Uterine blood flow may be markedly
reduced temporarily, by adrenaline or noradrenaline
in the maternal circulation, and this may be one
reason for the poor placental transfer of noradrenaline
which has been observed (168). In the guinea pig the
uterine blood vessels become sensitized to the action
of adrenaline as gestation proceeds and following the
administration of both estrogen and progesterone
(133). It is possible that there is a reduction in ma-
ternal placental blood flow before conversion of the
uterus from the spherical to oval shape (154, 157).
The influence of uterine contraction on the fetal
heart rate has been most extensively studied and de-
pends upon the duration of the contraction, its fre-
quency and upon the intrauterine pressure developed
(7, 15, 109). It is probable that the first effect of any
uterine contraction will be to upset the functional
countercurrent mechanism which enables the mater-
nal arterial blood in the intervillous space to reach
the base of the chorionic villi and enable final ar-
terialization of the umbilical venous blood. Hendricks
et at. (104) have made simultaneous recordings of the
intra-amniotic and intervillous pressures in the hu-
man and consider the sequence of events on the
maternal side of the placenta to be complicated be-
fore the oxygen supply to the fetus is impaired during
a contraction. They observed the pressures in the
intervillous pool and the amniotic cavity to be about
equal, both when the uterus was relaxed and during
systole; the increase in the intervillous pressure lagged
behind the intra-amniotic pressure rise during con-
traction. It is suggested that the intervillous volume is
slightly reduced during the early phase of contrac-
tion; but once the intra-amniotic pressure exceeds
that in the uterine vein, venous drainage will cease
and the intervillous volume become expanded as
the arterial inflow continues. The oxygen supply,
though slowed, may continue for a considerable time
during contraction, and the spongy structure of the
placenta and the large venous sinuses allows local
pressure differences to be distributed and prevent
retroplacental hemorrhage. The increased pressure
in the intervillous space will be transmitted to the
fetal vessels and, added to a reduced oxygen supply,
there will be a reduction in umbilical blood flow
as the resistance increases. Reynolds & Paul (159)
observed in the lamb, in utero, that rhythmic con-
tractions of low intensity which caused a rise of intra-
amniotic pressure of about 5 mm Hg caused an
equal rise in fetal blood pressure but no change in
heart rate: manual pressure on the uterus or the
application of weights, from 1 kg upward, caused
a rise of arterial pressure in the fetus which exceeded
the rise in amniotic pressure; this was asphyxial in
origin and accompanied by bradycardia. Strong
contractions induced by Pitocin, which raised the
intra-amniotic pressure more than 10 mm Hg, gave
similar vascular responses in the fetus (15, 159).
The quantitative relationship between reduction
in maternal placental blood flow and the appearance
of fetal bradycardia has been supplied in the sheep
by Adams et al. (2); no change in fetal heart rate was
observed until the uterine blood flow, measured with
an electromagnetic flowmeter, was reduced to about
one-third of the control level following the injection
of adrenaline into the maternal circulation. This
relationship was readily predicted from the heart
rate changes occurring during the administration of
i638
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
low oxygen tension mixtures to the mother in both
the sheep and the guinea pig (39, 133)- It appears
that there is no species difference for the sensitivity
of the fetal heart to hypoxia in utero. Hon has also
described intermittent fetal bradycardia during de-
livery which he considered to be unrelated to altera-
tion in placental blood flow and due to either com-
pression of the cord (109) or medullary asphyxia.
The bradycardia during cord compression had a
long time course, but swift physiological brady-
cardia was frequently observed in vertex presenta-
tions and could be related to the degree of cervical
dilatation and was possibly caused by the rise in
intracranial pressure. This is an old clinical observa-
tion and Harvey Gushing also observed bradycardia
in adult animals during experimental asphyxia of
the medulla (59).
The great ability of the fetus to survive asphyxia
is still not understood (139, 171) and it is not known
whether the ultimate damage to the tissues is mainly
due to the absence of oxygen and, therefore, the
supply of energy, or to the fall in pH as the lactic
acid accumulates. Whittam (188) has shown that
anoxic fetal kidney slices maintain their potassium
content better than adult tissue and, if this is true
for both the heart and the brain, it possibly explains
the maintenance of their excitability and activity
for long periods during asphyxia. Mott stresses the
importance of the maintenance of a circulation during
anoxia so that glucose, from the liver glycogen, may
be supplied to all the tissues, and lactic acid removed
(139): liver glycogen is partially mobilized during
anoxia and the brain and heart both suffer a large
reduction in glycogen content; in the young fetus
total lactate production can be accounted for by the
loss of carbohydrate from the heart. The survival
time of the fetal heart is directly related to its carbo-
hydrate stores which are larger than those of the adult
(fig. 17); these reserves may be depleted by repeated
episodes of hypoxia which may have a cumulative
effect.
CHANGES IN THE FETAL CIRCULATION AT BIRTH
AND IN THE NEONATAL PERIOD
Umbilical Curd: Ductus Venosus
The detailed structure of the umbilical cord varies
widely among the species, but all the arteries and
veins have thick muscular walls and lack a nerve
supply (27); the horse and rabbit have separate
sphincters in the region of the umbilical ring (194).
The isolated umbilical and placental vessels are very
reactive: constriction occurs in response to cooling,
stretching, or handling, the presence of the sympa-
thetic autonomic drugs in the perfusion fluid and
high oxygen tensions; relaxation occurs in the pres-
ence of low oxygen tensions and high CO-; tensions.
Rogers (163) observed the phenomenon of "pressure
spasm," a complete but temporary occlusion follow-
ing an increased and sustained perfusion pressure;
this response to pressure is also observed in dener-
vated systemic vessels (33). Following a natural birth
there will, therefore, be many factors combining to
ensure an effective closure of the umbilical vessels.
Intrauterine asphyxia might be expected to impair
the effectiveness of these stimuli and, recently, it
has been observed that the cord continues to pulsate
for long periods in infants following a difficult de-
livery (80).
40
30
20
IO
„ monkey
puineapig
<
01
\-
0
0 1
K
0
u
IX
•0
O
01
" 7
en
111
B
0
* — '
-0
U
1
¥-
' o
<
a
0
>
-<
I
0
In
<*
<
<
0
0 s
1
sheep
STAGE OF GESTATION
20
fig. 17. Cardiac glycogen in different
species before and after birth. [From Sheliev
(171).]
THE FETAL AND NEONATAL CIRCULATION
■639
IO
I
e
a.
z>
i/>
to
UJ
at
a.
z
<
UJ
2
CORD TIED
VENTILATION
BEGUN
M I NUTE. S
O 5 IO 15
fig. 18. Ventilation of the lungs of a mature fetal lamb
caused the mean left atrial pressure to rise above the pressure
in the inferior vena cava (IVC) ; occlusion of the umbilical
cord caused the IVC pressure to fall. Both, therefore, contribute
to the rapid reversal of the pressure gradient across the foramen
ovale, resulting in its closure after birth. (Modified from G. S.
Dawes. Changes in the circulation at birth. Brit. Med. Bull.
17: 152, 1961.)
The mechanism for the functional closure of the
ductus venosus is unknown but it is important that
this should take place early in neonatal life. The
formation of an Eck fistula, with the portal blood
short-circuiting the liver, passing through the portal
sinus and straight into the inferior vena cava, might
explain the hypoglycemia and icterus which some-
times occurs in the neonatal period, especially in pre-
mature infants. Cardiac catheterization through the
umbilical vein depends upon anatomical patency of
the ductus venosus; there is evidence that it is either
closed or absent in about 30 per cent of newborn
infants (167). However, when the ductus venosus is
patent it may be visualized by radiopaque sub-
stances up to 12 days after birth (146).
Fetal Channels in the Thorax
The first breath initiates the changes in course of
the blood streams in the heart: expansion of the lungs
decreases the resistance in the small vessels and the
resulting increase in pulmonary blood flow raises
the left atrial pressure above that in the inferior vena
cava, closing the foramen ovale functionally (fig. 18);
this closure is assisted by the fall in the inferior vena
caval pressure due to the temporarily reduced venous
return to the heart, following occlusion of the um-
bilical vessels. The whole volume of inferior caval
blood now joins the superior caval blood in the right
atrium to maintain the high pulmonary blood flow.
As a result of the reduced pulmonary vascular re-
sistance the pulmonary arterial pressure falls below
the systemic level and blood flow through the ductus
arteriosus is diminished.
The radiological studies in the sheep and the hu-
man infant (26, 129), at first suggested that when
respiration is off to a flying start the functional clos-
ure of both the foramen ovale and the ductus ar-
teriosus is immediate. However, anatomical closure is
not complete for some weeks and there is evidence
that blood may flow through both these channels,
probably intermittently, for about a fortnight after
birth; this is demonstrated in the angiocardiographic
studies and in dye dilution curves which, in normal
babies, are characteristic of pathological states with a
patent ductus (149). More direct evidence for a
patent ductus with a left-to-right shunt has been
obtained in mongol (117) and in normal infants (3);
during cardiac catheterization it was found that blood
obtained from the pulmonary artery contained more
oxygen than that collected from the right auricle
and, in addition, the pulmonary arterial pressures
were higher than expected. Dawes and his colleagues
have measured this flow in the lamb and find it con-
siderable (76). Blood flowing through the wide open
ductus arteriosus creates no murmurs, but as the
vessel constricts there is turbulence of the swiftly
flowing stream and murmurs attributed to this can
be heard in both the sheep (76) and the human in-
fant (50). The direction of this shunt may be from
left to right or right to left according to the relative
pressures in the pulmonary and aortic trunks. Follow-
ing expansion of the lungs, the pulmonary arterial
pressure falls relative to the systemic pressure and
there is the possibility of a left-to-right shunt; if this
occurs, the work of the left heart will be increased
but, during recirculation of the blood through the
lungs, there is a further opportunity for oxygen up-
take which is advantageous when the ventilation is
poor. During asphyxia or crying the pulmonary ar-
terial pressure rises and may exceed the systemic
pressure causing the possibility of a right-to-left
flow again, when the lower half of the body will
probably receive blood of a lower oxygen content
than the upper half (88).
The wall of the ductus arteriosus has a sphincter-
like structure and the musculature a poor nerve
supply. In the lamb fetus the lumen is nearly as
large as the pulmonary artery and descending aorta
and the blood flow through it approximately one-
third of the combined output of the two ventricles (76) ;
1640
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
LEFT ATRIUM
LEFT PULMONARY
FLOW
PULMONARY
VASCULAR
RESISTANCE
MINUTES FROM DELIVERY 90
fig. ig. Changes in the circulation on ventilating the
fetal lung: a) artificial positive pressure ventilation of the lungs
caused a large fall of pulmonary vascular resistance, an increase
in pulmonary flow and a fall in pulmonary artery pressure.
b) Temporary occlusion of the ductus arteriosus caused a rise in
femoral pressure and a fall in pulmonary pressure and flow,
showing that blood had been flowing from the aorta into the
pulmonary trunk. The figures 35 and 79 indicate the carotid
arterial Os% saturation. (From G. S. Dawes. Changes in the
circulation at birth. Brit. Med. Bull. 17: 151, 1 96 1 . )
with the reduction in pulmonary arterial pressure
the flow is diminished and the wall constricts. Closure
of the ductus is not dependent upon its nervous con-
nections and will occur following inflation of the
lungs provided the oxygen tension is high and, like
the umbilical vessels, it will dilate when the blood
oxygen tension is low (27, 40); constriction can,
however, occur during asphyxia and this may be due
to the release of sympathetic amines. The responses
of the ductus arteriosus and the cord vessels are com-
mon to all unstriped muscle, and the exemption of
the neighboring aorta and pulmonary artery is due
to the preponderance of elastic fibers in the tunica
media of the latter vessels.
In the fetal lamb in utero the right atrial pressure
is 1 to 1 1 2 cm H2O higher than the left atrial pressure
(fig. 18). This pressure difference occurs because
the pulmonary venous return to the left auricle is
small and only about one-ninth of that returning to
the right side of the heart; 75 per cent of the inferior
caval blood is directed bv the valve of the foramen
ovale and redistributes the venous return to maintain
the left ventricular output and systemic and placental
blood flows. Following inflation of the lungs the de-
crease in pulmonary vascular resistance enables the
pulmonary blood flow to treble and as the interatrial
pressure difference is reversed the foramen oval closes
(77); Dawes and his colleagues consider that clamp-
ing the cord before the first breath, thus reducing
temporarily the inferior caval flow, might be sufficient
to lower the right atrial pressure and assist closure of
the foramen ovale. However, maintenance of its
closure will depend upon the increased pulmonary
venous return. In small animals the preponderance of
left-over-right atrial pressure is difficult to demon-
strate within the first 24 hours of birth, but develops
during the subsequent days and weeks; the main-
tenance of this pressure difference which is observed
throughout life is probably the combined influence of
filling and elasticity of the two ventricles.
Pulmonary Vascular Resistance, Arterial
Pressure, and Blood Flow
The way in which the first breath initiates the re-
duction in pulmonary vascular resistance is not yet
fully explained. Using a density flowmeter, Dawes
et al. measured the blood flow in the left pulmonary
artery of lambs delivered by Cesarean section; follow-
ing positive pressure ventilation with air, oxygen, or
nitrogen they observed a three- to fourfold increase in
pulmonary blood flow, a decrease in arterial pressure
and calculated a tenfold decrease in pulmonary vascu-
lar resistance (13, 78). Distention of the lungs with
warm saline was not found to increase the pulmonary
blood flow and no change in circulatory pattern
probably takes place during respiratory effects in
utero when amniotic fluid is known to enter the
lungs (65). Dawes' conclusion that the decrease
in pulmonary vascular resistance was primarily
due to the mechanical factors associated with ventila-
tion was questioned, recently, by Cook et al. (55)
following observations in newborn lamb preparations
in which the two lungs were ventilated separately
and the pulmonary vessels perfused at a constant
pressure; alveolar hypoxia and hypercapnia caused
vasoconstriction in the pulmonary circulation which
was more marked than that observed in the adult lung
(84) and ventilation with nitrogen alone gave variable
results — possibly on account of differences in local
( !( )_. tension. Recent observations have shown that
both an increase in arterial and alveolar pOL> and a
reduction in pCOj contribute toward the increase in
THE FETAL AND NEONATAL CIRCULATION
I 641
50
E 40
<
_,UI
0= 30
—
20
aSC icH
AGE IN DAYS
5 15 25 35
fig. 20. Right ventricular systolic pressure in 15 puppies
during the first 5 weeks of neonatal life. [From Rudolph el al.
(166).]
pulmonary blood flow at the onset of pulmonary
ventilation in the lamb; similar changes in blood
chemistry, and vasodilator drugs, also increase blood
flow in the unexpanded fetal lung (73) which would
suggest that a decrease in vascular resistance is not
necessarily due to uncoiling of vessels as suggested by
Reynolds (156). Once started, the increase in blood
flow itself together with the raised left atrial blood
pressure may help to maintain a low pulmonary
vascular resistance as it does in the adult lung (42).
Pulmonary vascular resistance has been calculated in
the human infant at birth and during the first 3 weeks
of life from measurements of the pulmonary artery
pressure and cardiac output, determined by the Fick
principle (164). Within a few hours of delivery the
pulmonary vascular resistance is about 550 dynes per
sec per cm-s in comparison with an assumed fetal
value of 8,000 dynes sec cm-5. This neonatal value
is still considerably higher than that found in the
older infant and the adult, but is already much less
than the systemic vascular resistance; it declines to
the adult level by 6 months of age by which time
the walls of the pulmonary arterioles are reduced in
thickness (62). The lung blood volume does not
change immediately following this large drop in re-
sistance (66) but a considerable increase has been
demonstrated within the first 24 hours of life in the
guinea pig (92).
The pulmonary arterial pressure is reduced by
about a half to approximately 35 mm Hg during
the immediate postnatal period in both the lamb
and the human infant, and in the puppy (165, 166).
The final reduction in pressure occurs gradually over
the following weeks (fig. 20). The thickness of the
walls of the two ventricles is nearly equal in the
fetus with a slight preponderance of the right
chamber: while the pulmonary vascular resistance
and arterial pressure are falling and the systemic
vascular resistance and pressure are rising in the
newborn period, the right ventricular wall decreases
in thickness and the left ventricular wall increases in
thickness; in the human infant these changes are
nearly completed within the first month of life (121).
The Heart
The immediate changes in heart rate in the human
infant following a normal birth are variable and
transient (173). In the lamb delivered by Cesarean
section the heart rate slows when the cord is clamped
(66): this may be reflex in origin for the arterial
pressure is raised, but may also be due to the direct
effect of asphyxia on the pacemaker; the bradycardia
is followed by tachycardia once respiration and
oxygenation of the blood are established. During the
first 2 days after birth the heart rate of the human
infant is usually lower than in utero, about 120
beats per min, and rises during the first week of
life (14, 21, 198). The temporary bradycardia is
possibly the combined effect of the low body tempera-
ture during this period (52) and the residual effect
of perinatal asphyxia. In the newborn kitten and
puppy the heart rate varies widely, ranging from
180 to 260 per min during the first 15 weeks of life
(114). The heart rate of the newborn monkey is
205 ± 20 (sd) beats per min (1 16).
The heart volume has been measured radiologically
in the human infant (124) and found to have an
average value of 48 ml in 55 infants on the day of
birth; during the first hour of life there was an in-
crease in volume with a return to the immediate
postnatal level within 3 hours. During the subsequent
4 days of life the volume diminished by 25 per cent
and thereafter increased; the decrease in heart size
was more pronounced in premature babies. These
early changes in heart volume and the enlargement
of the heart which occurs following birth asphyxia
(51) need to be made simultaneously with other
circulatory measurements, for a better understanding
of the events taking place.
Cardiac output measurements in newborn human
infants by the dye dilution technique (149) and using
the Fick principle (3) provide a very wide range of
values 180 to 850 ml per min, which is probably ex-
plained by the patency of the fetal channels; this is a
1642
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Cord clamped after 3 mm _^*
J<
Cord clamped immediately
\y^
Cord clamped during delivery
AGE (hr )
I 2 3 4 24
fig. 21. Three representative records of the changes in the
systolic pressure of normal babies during the first 24 hours of
life. [From Ashworth & Neligan (14).]
factor which must influence all but the cardiometer
results in animals, which also have their inherent
disadvantages. The average cardiac output of the
human infant, 540 ml per min, corresponds to a
value of 180 ml per kg per min and is about double
the value in the adult per kg body weight; the cardiac
index is 2.5 liters per min per m2. Assuming a new-
born heart rate of 140 per min, the stroke volume
will be approximately 4 ml. It may be noted that if
the estimates of cardiac output in utero are correct,
200 ml per kg per min (17), the value does not change
in the neonatal period and no increase in oxygen
consumption is observed (58). On the other hand,
there is evidence for an increase in cardiac output
following birth in the lamb: Cross et al. (57) made
calculations using the Fick principle and obtained
values of 325 ± 30 ml per kg per min, which com-
pared with the near term intrauterine estimate of
235 ml per kg per min for both ventricles; a single
ventricle has therefore increased its output three-
fold. This increase may be the response to the raised
oxygen consumption which occurs in the lamb at
birth or it may be the expression of the better measure-
ments which are possible after birth.
Systemic Pressure, Cardiovascular Reflexes
and Peripheral Resistance
When the changes in systemic arterial pressure at
delivery are measured, a discrepancy exists between
the lamb and the human baby; namely, a small
transient rise of pressure is observed following the
initiation of ventilation or occlusion of the cord in
the lamb (66) while, remarkably, in the human
infant no change of pressure is seen (183, 196). There
are many possible explanations for this difference:
first, the different types of maternal placental circu-
lation; second, the influence of contraction of the
uterine muscle on this circulation; and third, the
alteration of distribution of blood between the
placenta and fetus before the arterial measurements
are made. The arterial pressure measurements in
lambs have all been made on fetuses delivered by
Cesarean section and, as the sheep uterus is not very
reactive to surgery, the maternal and consequently
the fetal placental circulations are probably not
greatly impaired. When the lamb is delivered vagin-
ally the maternal placental blood flow is not reduced
during labor and does not decrease until separation
of the placenta some hours after delivery of the fetus
(15). The temporary rise in pressure observed is
therefore probably due to the removal of the low
resistance circuit of the placenta, and a small rise in
arterial pressure following cord occlusion has also
been observed in the rhesus monkey delivered by
Cesarean section (71). In the adult animal the reduc-
tion of a circulating bed even of the same resistance
as the total vascular bed, raises the arterial blood
pressure (23). In the human, contraction of the uterus
during labor probably reduces both the maternal and
fetal placental blood flow and therefore much of the
low resistance circuit of the placenta is gradually
removed before the arterial pressure measurements
are made as the cord is tied. Other hemodynamic
factors, such as the relative distribution of blood
between the fetus and the placenta and the relative
proportions of the cardiac output which traverse
the placenta, might also influence any change in
systemic pressure at birth. In the lamb, at term, only
1 5 per cent of the total blood volume is to be found
in the placenta and 60 per cent of the cardiac output
traverses this vascular bed. The human placenta
contains 30 per cent of the total circulating blood
volume at term (173), but the portion of the cardiac
output perfusing it is not known.
Ashworth & Neligan (14) have used the con-
ventional inflatable cuff and manometer and a
sensitive pulse indicator to measure the arterial
pressure in the newborn infant's arm, and report
marked changes in systolic pressure within the first
24 hours of life. The initial pressures, within 2 min
of delivery, ranged from 1 16 to 52 mm Hg and there
was subsequently a fall of up to 54 mm Hg (fig. 21);
delay in clamping the cord postponed this fall, but
THE FETAL AND NEONATAL CIRCULATION
'643
SO
RABBIT
/
CAT /
SHEEP /MONKEY
0. /
' /
6O
- I /
/
E /
/ /
* /
/ /
°" /
/
CD /
/
40
" J
/ y
_-' „-"
- 1
K 1
- — "
Ul
•
1- 1
20
or /
- < i
DAYS
1
FROM
1
CONCEPTION
1 1
40 SO I20 160
fig. 22. Arterial blood pressures before ( ) and after), )
birth, showing the continuous course of the rise with increasing
age, in the rabbit, cat, sheep, and monkey. (Modified from
G. S. Dawes. Changes in the circulation at birth. Brit. Med.
Bull. 17: 150, 1 96 1.)
did not influence its magnitude. It is tempting to
suggest that the wide range of initial pressures is due
to varying degrees of asphyxia during birth, but no
proof exists for this explanation. The pressures rise
gradually during the second day of life and during
the subsequent weeks.
Once the temporary interruptions of parturition
are over, the mechanisms which have been responsible
for the gradual rise in arterial pressure throughout
gestation will, probably, be extended into the neo-
natal period: these mechanisms are, however, likely
to be modified by the different internal environment
of the young free animal, as compared with the
fetus, and by many other factors which will vary
with the species; orthostatic factors and the mode of
life will be among these. The rate of rise in arterial
pressure is rapid in small animals and the mean pres-
sure is about doubled during the first 6 weeks of life,
approaching the adult level; in the sheep and monkey
the rise is slower (fig. 22). In the human infant, who
has been a repeated subject for blood pressure meas-
urements, the rise is slow during the first 9 months
of life (fig. 23) and continues well into adolescence
and throughout adult life (194). [But see also (193).
Ed.]
The newborn is a more satisfactory experimental
subject than the fetus for, under experimental condi-
tions, an established respiration provides a more
constant internal environment than the placental
circulation. The differences between the cardio-
vascular responses of newborn and adult animals are
of a quantitative rather than a qualitative nature:
£lOO
S so
a.
8 so
o
-I
3
y
_j
o
H
to
z
<
40
20
ft* *
54 BABIES
<
□
9
DAYS
6
WEEKS
3 6
MONTHS
fig. 23. Mean arterial blood pressures at birth and during
the first few months of life in normal infants. [From Holland &
Young, Brit. Med. J. 2: 1331, 1956.]
in the newborn monkey there is evidence for func-
tional baroreceptor and chemoreceptor activity,
yet bradycardia and hypotension still follow acute
hypoxia (71). In the young growing rabbit (70, 83),
kitten, and puppy ( 1 1 4) a gradual increase in vasocon-
strictor tone in the systemic circulation can be
demonstrated by the responses to asphyxia and to the
injection of hexamethonium. Downing's (83) observa-
tions show that the threshold for baroreceptor stimu-
lation in young rabbits is about 40 mm Hg.
Hutchinson el al. (11 4) have also demonstrated in the
newborn kitten and puppy that the carotid sinus-
cardiac center mechanism will respond to a rise in
pressure but not to a fall. These findings may be ex-
plained by Landgren's (126) observations that 40
mm Hg is just within the recording range of the
baroreceptors; any stimulus which raises the pressure
will elicit a response, especially if the pulse pressure
is also increased (85), but a further fall will be in-
effective. As the resting arterial pressure rises and
approaches the maximum sensitivity range of the
baroreceptors, 85 to 100 mm Hg, the reflexes
become more active; for instance, following the injec-
tion of adrenaline the percentage decrease in heart
rate increases in relation to the percentage rise in
blood pressure in the growing rabbit and kitten. The
direct action of adrenaline on the heart could only
be demonstrated in the youngest animals following
doses so small that the blood pressure did not rise
sufficiently to elicit a reflex bradycardia. In contrast,
the young kitten heart was found to be more sensitive
to acetylcholine than the peripheral vessels; with
small doses a marked bradycardia accompanied the
fall in blood pressure in the kitten, while reflex
1644
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
3D
i
E
E
2.0
©
0-
c r
a- E
J0.5-
#6
I Or- ©
90
\
o
l>
no 130
Gestation age in days
fig. 24. Increase in distensibility of fetal lamb lungs with
age. A: tidal air/peak intratracheal pressure = distensibility;
B: distensibility per kg body weight plotted against age. [From
Dawes (66).]
mechanisms are not very active in the newborn infant
(197). The physiological activity of the vasomotor
sympathetic mechanism to the skin blood vessels is,
however, well developed at birth and quite compa-
rable with that of the adult — a fact which was dem-
onstrated clearly by Day (79) who showed, by
conductivity measurements, that the circulatory re-
sponses to changes in environmental temperature were
as effective as in the adult in maintaining body tem-
perature. These observations have recently been am-
plified by Briick (47).
Renal blood flow appears to be low in the sheep
fetus (5) and the newborn infant (173) when com-
pared with the adult on a body weight basis; PAH
clearance was used in these measurements but nothing
is known of the secretory capacity of the tubules for
this substance. Unilateral renal artery stenosis, with
fatal arterial hypertension of 180 mm Hg, has been
observed in a newborn infant ( 1 30) suggesting that
the renin-hypertensinogen mechanism is active early
in life in man and may account for the hypertension
above the lesion with coarctation of the aorta.
Viability
tachycardia followed the hypotension occurring with
comparable doses in the adult.
In the newborn monkey, the carotid sinus reflexes
are functional and occlusion of the carotid arteries
has been shown to cause a rise in the arterial pressure,
which is abolished by cutting the carotid sinus nerves
(71) : but acute anoxia causes a fall in arterial pressure
suggesting that the vasomotor center itself is not very
active: there was, however, a rise in arterial pressure
in the fetus in response to asphyxia. In the newborn
baby the mean arterial pressure is about 40 mm Hg
below the mean pressure of the adult and the avail-
able evidence shows that, once the immediate read-
justment of birth are complete, there is a low periph-
eral resistance (196); the cardiac output per kg of
body weight is about double that of the adult, the
blood flow to the extremities is likewise double and
the cerebral blood flow is high (123). Low tonic
activity of both the chemical and reflex regulating
mechanisms are probably concerned, and the develop-
ment of these will probably contribute relatively
more to the gradual rise in arterial pressure during
growth than the cardiac output, which declines in
relation to body weight. Recently, records of arterial
blood pressure changes during replacement trans-
fusions, when the blood volume was reduced rapidly
by 10 per cent, demonstrated that the baroceptor
Viability, in its narrowest sense, may be considered
as the capacity of the newborn to establish correct
200
150
100
50
1 38 days gestation age
• •
• •
CD
0 <- 2 •
o
£
sCP
50r
94 days gestation age
-§*
0L. * »
Pulmonary arterial perfusion pressure (mm Hg)
I 1 ■ ■ 1 1
0 10 20 30 40 50
fig. 25. Perfusion of isolated lungs of two fetal lambs,
mature, above; nonviable, below. Pressure flow diagrams were
constructed before ventilation (o) and about 15 min later (•).
Following ventilation, there is a large decrease of pulmonary
vascular resistance in the mature lamb and almost no change
in the premature. [From Dawes (66).]
THE FETAL AND NEONATAL CIRCULATION
'645
pulmonary ventilation and perfusion to provide full
oxygenation of the blood in order to maintain the
necessary oxygen supply to the body tissues; it is,
therefore, closely linked with lung development
which occurs relatively late in intrauterine life.
Most of the quantitative data relating this develop-
ment to length of gestation are, again, supplied by
Dawes and his colleagues in the lamb fetus. At 90
days of age fluid starts to collect in the alveolar
spaces of the lungs (93). Shortly afterward the dis-
tensibility of the lungs begins to increase so that, at
a given inflation pressure, older lambs obtain more
tidal air per kg body weight (fig. 24); ventilation
also starts to cause a decrease in pulmonary vascular
resistance and there is a larger blood flow for the same
perfusion pressure (fig. 25). By 110 days gestational
age, about 28 to 30 weeks on the human scale, arti-
ficial ventilation can raise the arterial oxygen satura-
tion to 95 per cent and independent existence is
possible; in the nonviable premature this cannot occur
and death is due to asphyxia.
The development of many other physiological
mechanisms must also influence the successful opera-
tion of ventilation and perfusion of the lung tissues.
Among these will be the level of the arterial blood
pressure, closure of the foramen ovale and ductus
arteriosus, and the presence of sufficient surface
active substance to prevent collapse of the expanded
alveoli (145); Avery & Mead (18) have found the
surface activity of lung extracts from premature
infants to be only one-third of that from normal
full-time lungs.
Congenital Heart Disease
The transition from the fetal to the adult course of
the circulation may not take place because of in-
herent congenital abnormality or be protracted on
account of a difficult labor and the ensuing asphyxia.
Rowe (164) gives a concise account of the physical
signs and the physiology of both, together with the
possibilities of their treatment (122). The physio-
logical disturbances accompanying congenital mal-
formations may be divided into three groups: /) a
left-to-right shunt through a patent ductus arteriosus,
2) the retention of a fetal type of flow through both
ductus arteriosus and foramen ovale, and 3) simple
intracardiac arteriovenous shunts. Only the first
group have normal arterial oxygen saturations.
The physical signs of congenital heart disease are
frequently difficult to distinguish from the transient
abnormalities due to respiratory disturbances at
birth; Rowe also divides these infants into three main
groups. In the first are those who do not breathe
readily at birth and who have a murmur due to
patency of the ductus arteriosus: Burnard (51) has
observed a midsystolic murmur in 70 per cent of
such infants, and considers it due to swift turbulent
flow through a ductus only partially constricted on
account of asphyxia; the direction of this flow will
depend upon the relative pressures in the pulmonary
and aortic trunks. Lind & Wegelius (129) have
angiocardiographic evidence for delayed closure of
the ductus arteriosus following asphyxia neonatorum.
In the second group, apnea may develop suddenly
following normal respiration of a few hours to 4 weeks
duration; a loud continuous murmur, due to a left-
to-right shunt through the opened ductus arteriosus
is heard. The third group of premature infants, with
classical respiratory distress syndrome have, on ac-
count of the high lung resistance, a pulmonary ejec-
tion click to the second heart sound; during the
recovery phase a midsystolic sound is also heard as
the ductus arteriosus narrows.
REFERENCES
1. Acheson, G. H., G. S. Dawes, and J. G. Mott. Oxygen
consumption and the arterial oxygen saturation in new-
born lambs. J. Physiol., London 135: 623, 1957.
2. Adams, F. H., N. Assali, M. Cushman, and A. Westen-
sten. Interrelationships of maternal and foetal circulations.
I. Flow-pressure responses to vasoactive drugs in sheep.
Pediatrics 27 : 627, 1 961.
3. Adams, F. H., and J. Lind. Physiologic studies on the
cardiovascular status of normal infants (with special
reference to the ductus arteriosus). Pediatrics 19: 431, 1957.
4. Aitken, E. H., R. V. Eton, B. Eton, J. R. K. Preedv,
and R. V. Short. Oestrogen and progesterone levels in
foetal and maternal plasma at parturition. Lancet 2 :
1096, 1958.
5. Alexander, P. P. and D. A. Nixon. The foetal kidney.
Brit. Med. Bull. 17: 112, 1961.
6. Althoff, H., and H. Werner. Vorkommen und Bedeu-
tung der Erythropoetine der Erythroblastosis foetalis. Acta
Haematol. 18: 126, 1957.
7. Alvarez, H., and R. Caldevro. Heart rate of the human
foetus in utero. Proc. yd Intern. Congr. Med. Electronics.
London 1960. In press.
8. Amoroso, E. C. Placentation. Marshall's Physiology of
Reproduction (3rd ed.). 1952, vol. II, 127.
1646
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
g. Amoroso, E. C. The comparative anatomy and histology 30.
of the placental barrier. Gestation, edited by L. B. Flexner,
Trans. 1st Conf. New York: Josiah Macy, Jr., Found.,
'954. P- "9- 3'-
10. Amoroso, E. C. Endocrinology of pregnancy. Brit. Med.
Bull, n : 117, 1955.
11. Amoroso, E. C. The biology of the placenta. Gestation, 32.
edited by C. A. Villee. Trans. 5th Conf. New York
Josiah Macy, Jr., Found., 1958, p. 15. 33.
12. Amoroso, E. C, G. S. Dawes, J. C. Mott, and B. R.
Rennick. Occlusion of the ductus venosus in the mature
foetal lamb. ./. Physiol , London uq 64, 1955. 34.
13. Ardran, G. M., G. S. Dawes, M. M. Pritchard, S. R.
Reynolds, and D. G.J. YVyatt. The effect of ventilation
of the foetal lungs upon the pulmonary circulation. 35.
J. Physiol., London 113: 12, 1952.
14. Ashworth, A. M., and G. A. Neligan. Changes in the
systolic blood pressure of normal babies during the first
twenty-four hours of life. Lancet 1: 804, 1959. 36.
15. Assali, N. S., K. Dasgupta, K. K.olin, and L. Holm.
Measurement of uterine blood How and uterine mctabo- 37.
lism. Am. J. Physiol. 195: 614, 1958.
16. Assali, N. S , S. A. Marabel, and N. Sehgal. Pulmonary 38.
and ductus arteriosus circulation in the fetal lamb before
and after birth. Am. J. Physiol. 202: 536, 1962.
17. Assali, N. S., L. Rauramo, and T. Peltonen. Uterine 39.
and fetal blood flow and oxygen consumption in early
human pregnancy. Am. J. Obstet. Gynec. 79: 86, i960.
18. Avery, M. E., and J. Mead. Surface properties in rela- 40.
tion to atelectasis and hyaline membrane disease. A.M. A.
J. Diseases Children 97: 517, 1959.
ig. Baker, J. B. E. Some observations upon isolated perfused
human foetal hearts. J. Physiol., London 120: 122, 1953. 41.
20. Baker, J. B. E. The effects of drugs on the foetus. Pharmcol.
Revs. 12: 37, i960.
31. Balard, P. Modifications, evolutives du pouls et de la 42.
tension arterielle chez le nouveau-ne, dans les premiers
jours de la vie, etudiees par l'oscillometrie. Compt. rend.
Soc. Biol. 73: 483, 1912.
22. Bangham, D. R , K. R. Hobbs, and R. J. Terry. Selec- 43.
tive placental transfer of serum proteins in the rhesus.
Lancet 2:351, 1 958. 44.
Barcroft, H. Cardiac output and blood distribution.
J. Physiol., London 71 : 280, 1931.
Barcroft, J., and J. A. Kennedy. The distribution of 4.5.
blood between the foetus and the placenta in sheep.
./. Phyiiol., London 95: 173, 1939.
BARCROFT, J. Researches on Prenatal Life. Oxford: Blackwell, 46.
1946.
Barclay, A. E., J. Barcroft, D. H. Barron, and K. J. 47.
Franklin. A radiographic demonstration of the circula-
tion through the heart in the adult and in the foetus 48
and the identification of the ductus arteriosus. Brit. J.
Radiol. N. S. 18: 505, m ;i|
Barclay, A. E., K. J. Franklin, and M. M. L. 49,
Pritchard. The Foetal Circulation and Cardiovascular System
and the Changes that They t 'ndergo at Birth. Oxford : Blackwell,
1944. 50.
28. Barron, D H. The changes in the foetal circulation at
birth Physiol. Reus. 24: 277, 1944. 51
29. Barron, D. H. In: Blood and Othei Body Fluids. Washing-
ton, D.C. : Fed. Am. Sues, for Exper. Biol., 1961, p. 114. 52
23-
24.
25-
26.
27.
Barron, D. H., and G. Mesciiia. A comparative study
of the exchange of respiratory gases across the placenta.
Cold Spring Harbor Syinp. Quant. Biol. 19:93, 1954.
Bartels, H., VV. Moll, and J. Metcalfe. Physiology of
gas exi h.inge in the human placenta. Am. J. Obstet.
Gynec. 84: 1 7 1 4, 1962.
Bauer, D. J. The effect of asphyxia upon the heart rate
of rabbits at different ages. J. Physiol., London 93: go, 1938.
Bayliss, VV. M. On local reactions of the arterial wall to
changes of internal pressure. J. Physiol., London 28: 220,
1902.
Beard, R. VV. Response of the human foetal heart and
maternal circulation to adrenaline and noradrenaline.
Brit. Med. J. i : 443, 1 962.
Benninghof, A., and R. Spanner. Das Gefassystem eins
( )c,n dins Untersuchungen liber der Einfluss des Blut-
stroms auf die Gefassentwicklung. Morphol. Jahrb. 61 : 380,
!929-
Blandau, R. J. Experimental implantation in the rat
and guinea pig. Anal. Record 97: 322, 1947.
B0e, F. Vascular morphology of the human placenta.
Cold Spring Harbor Symp. Qjiant. Biol. 19: 2g, ig54-
Borell, U., I. Fernstrom, and A. VVestman. Eine
arteriographische Studie des Plazentarkreislaufs.
Geburtsch. Frauenheilk. 18: 1, 1 958.
Born, G. V., G. S. Dawes, and J. C. Mott. Oxygen
lack and autonomic nervous control of the foetal circula-
tion in the lamb. J. Physiol., London 134: I4g, igs6.
Born, G. V. R., G. S. Dawes, J. C. Mott, and B. R.
Rennick. Constriction of the ductus arteriosus caused by
oxygen and by asphyxia in newborn lambs. J. Physiol.,
London 132:304, 1956.
Bornsdorff, E. On the presence of erythropoietins in the
plasma from sheep fetuses during the latter half of gesta-
tion. Acta Physiol. Scand. 18: 51, 1949.
Borst, H. G., M. McGregor, M. Whittenberger, and
E. Berglund. Influence of pulmonary arterial and left
atrial pressures on pulmonary vascular resistance. Circula-
tion Research 4: 393, 1956.
Boving, B. G. Blastocyst-uterine relationships. Cold Spring
Harbor Syinp. Quant. Biol. 19:9, 1 954.
Boyd, J. D. Development of the human carotid body-
In: Contribution to Embryology. Washington: Carnegie
Inst. 26: 1937.
Boyd, J. D., and W. J. Hamilton. Marshall's Physiology
of Reproduction (3rd ed.l. 1952, vol. II, I, London: Long-
mans, Green.
Brambell, F. W. R , W A. Hemmings, and M. Hender-
son. Antibodies and Embryos. London: Athlone Press, 1951.
Bruck, K. Temperature regulation in the newborn
infant. Biol. Neonatorum 3: 65, 196 1.
Bumm, E. Ueber die Entwicklung des mutterlichen Blut-
kreislaufes in der menschlichen Placenta. Arch. Gyndkol.
43: 181, 1893.
Burlingame, P., J. A. Long, and E. Ogden. The blood
pressure of the fetal rat and its response to renin and
angiotonin. Am. J. Physiol. 137: 473, 1942.
Burnard, E. D. A murmur from the ductus arteriosus
in the newborn baby. Brit. Med. J. 1 : 1495, 1959.
Burnard, E. D. Changes in heart size in the dyspnoeic
newborn baby. But. Med. J. 1 : 1495, 1 959-
Burnard, E. D., and K. W. Cross. Rectal temperatures
THE FETAL AND NEONATAL CIRCULATION
1647
59
60.
61
in the newborn after birth asphyxia. Brit. Med. ./. 2:
"97. >958-
53. Carlyle, A. An integration of the total oxygen consump-
tion of the sheep foetus from that of the tissues. J. Physiol., 75.
London 107: 355, 1948.
54. Comline, R. S., and M. Silver. The release of adrenaline
and noradrenaline from the adrenal glands of the foetal 76.
sheep. J. Physiol., London 156: 424, 1961.
55. Cook, C. D., P. A. Drinker, H. N. Jacobson, H. Levin-
son, and L. B. Strang. Factors determining the increase
in pulmonary blood flow on ventilation of the foetal lamb 77.
lung. J. Physiol.. London 166: 9P, 1963.
56. Cooper, K. E., A. D. M. Greenfield, and A. St. G.
Huggett. Umbilical blood flow in the foetal sheep. 78.
J. Physiol., London 108: 160, 1949.
57. Cross, K. W., G. S. Dawes, and J. C. Mott. Anoxia,
oxygen consumption and cardiac output in newborn 79.
lambs and adult sheep. ./. Physiol., London 146: 316, 1959.
58. Cross, K. W., J. P. Tizard, and D. A. Trvtrall. The
gaseous metabolism of the newborn infant. Acta Pediat. 80.
46:265, 1957.
Cushinc, H. Quoted by D. H. Barron, in Oxygen Supply
to the Human Foetus. C.I.O.M.S. Symposium. Oxford:
Blackwell, 1959. 81.
Daly, M. de B., and M. J. Scott. The effects of stimu-
lation of the carotid body chemoreceptors on the heart 82.
rate in the dog. J. Physiol., London 144: 148, 1958.
Daly, M. de B., and J. M. Scott. The effects of hypoxia
on the heart rate of the dog with special reference to the 83.
contribution of the carotid body chemoreceptors. J.
Phvsiol., London 145: 440, 1958. 84.
62. Dammann, J. F., and C. Ferencz. The significance of
the pulmonary vascular bed in congenital heart disease. 85.
Am. Heart J. 52: 7, 1956.
63. Dancis, J. The placenta. J. Pediat. 55: 85, 1959.
64. Dancis, J., and M. Shafran. The origin of plasma
proteins in the guinea pig foetus. J. Clin. Invest. 37: 1093,
1958. 86.
65. Davis, M. E., and E. L. Potter. Intra-uterine respiration
of the human foetus. J. Am. Med. Assoc. 131: 1 1 94, 1946.
66. Dawes, G. S. Changes in the circulation at birth and the 87.
effects of asphyxia. Recent Advances in Pediatrics, edited by
D. Gairdner, London: Churchill, 1958, p. 1.
67. Dawes, G. S. Oxygen consumption and hypoxia in the 88.
newborn animal. Ciba Found. Syrnp. on Somatic Stability in
the Newly Born. 1961, p. 170.
68. Dawes, G. S. Changes in O? supply within the foetal lamb. 89.
J. Physiol., London 159: 44 P, 1961.
69. Dawes, G. S. The umbilical circulation. Am. J. Obstel. go.
Gynec. 84: 1634, 1962.
70. Dawes, G. S., J. J. Handler, and J. C. Mott. Some
cardiovascular responses in foetal, newborn and adult 91.
rabbits. J. Physiol., London 139: 123, 1957.
71. Dawes, G. S., H. M. Jacobson, J. C. Mott, and H. J.
Shelley. Some observations on foetal and newborn 92.
rhesus monkeys. J. Physiol., London 152: 271, i960.
72. Dawes, G. S., and J. C. Mott. The increase in oxygen
consumption of the lamb after birth. J. Physiol., London 93.
'46: 295, 1959.
73. Dawes, G. S., and J. C. Mott. Vascular tone of the foetal
lung. J. Physiol. London 164: 465, 1962. 94.
74. Dawes, G. S., J. C. Mott, and B. R. Rennick. Some
effects of adrenaline, noradrenaline and acetyl choline
on the foetal circulation in the lamb. J. Physiol., London
134: '39. 1956-
Dawt.s, G. S., J. C. Mott, and J. G. Widdicombe. The
foetal circulation in the lamb. J. Physiol., London 126:
503. '954-
Dawes, G. S., J. C. Mott, and J. G. Widdicombe.
Patency of the ductus arteriosus in newborn lambs and
its physiological consequences. J. Physiol., London 128:
344, 1955-
Dawes, G. S., J. C. Mott, and J. G. Widdicombe.
Closure of the foramen ovale in newborn lambs. J.
Physiol., London 128: 384, 1955.
Dawes, G. S., J. C. Mott, J. G. Widdicombe, and D.
G. Wyatt. Changes in the lungs of the newborn lamb.
J. Physiol., London 121 : 141, 1953.
Day, R. Respiratory metabolism in infancy and in child-
hood. Regulation of body temperature of premature
infants. Am. J. Diseases Children 65: 376, 1943.
Desmond, M. M., J. L. Kay, and A. L. Megarity. The
phases of transitional distress occurring in neonates
associated with prolonged pulsating umbilical cord.
J. Pediat. 55: 131, 1959.
Dixon, R. C, and D. B. Stewart. Advanced extra-
uterine pregnancy. Brit. Med. J. 2: 1 103, 1960.
Dornhorst, A. C, and I. M. Young. The action of
adrenaline on the placental circulation in the rabbit and
guinea pig. J. Physiol., London 118: 282, 1952.
Downing, S. E. Baroreceptor reflexes in newborn rabbits.
./. Physiol., London 150: 201, i960.
Duke, H. N., and G. de J. Lee. Regulation of blood flow
through the lungs. Brit. Med. Bull. 19: 71, 1963.
Ead, H. W., J. H. Green, and E. Neill. A comparison
of the effects of pulsatile and non-pulsatile blood flow
through the carotid sinus on the reflexogenic activity
of the sinus baroceptors in the cat. J. Physiol., London
118509. '952-
Ebert, J. D. An analysis of the synthesis and distribution
of the contractile protein myosin, in the development of
the heart. Proc. Natl. Acad. Sci. 39: 333, 1953.
Ebert, J. D , R. A. Tolman, A. M. Mun, and J. R.
Aleright. Molecular basis of the first heart beats. Ann.
New York Acad. Sci. 60: 965, 1955.
Eldridge, F. L., H. N. Hultgren, and M. E. Wigmore.
The physiologic closure of the ductus arteriosus in new-
born infants. J. Clin. Invest. 34: 987, 1955.
Emery, J. L. The distribution of haemopoietic foci in
the infantile human liver. J. Anal. 90: 293, 1956.
Eranko, O., and M.J. Karvonen. Conditions of erythro-
poiesis in the fore and hind legs of foetal sheep. Ann.
Paediat. Fenniae. 1: 179, 1954-1955.
Everett, N. B., and R. J. Johnson. Use of radioactive
phosphorus in studies of foetal circulation. Am. J. Physiol.
162: 147, 1950.
Everett, N. B., and B. S. Simmons. The magnitude of
the increase in the pulmonary blood volume of the
postnatal guinea pig. Anal. Record 119: 329, 1954.
Faure-Fremiet, E., and J. Drogoiu. Le developpement
du poumon foetal chez le mouton. Arch. Anal. Microscop.
I9:4H. I923-
Fawcett, D. W., G. B. Wislocki, and C. M. Waldo.
The development of mouse ova in the anterior chamber
i648
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
of the eye and in the abdominal cavity. Am. J. Anal.
8':4I3. '947-
95. Fraser, F. C. Causes of congenital malformations in
human beings. J. Chronic Diseases 10: 97, 1 959.
96. Goss, C. M. First contractions of the heart without
cytological differentiation. Anal. Record 76: 19, 1940.
97. Greenfield, A. D. M., and J. T. Shepherd. Cardio-
vascular responses to asphyxia in the foetal guinea pig.
J. Physiol., London 120: 538, 1953.
98. Greenfield, A. D. M., J. T. Shepherd, and R. F.
Whelan. The relationship between the blood flow in the
umbilical cord and the rate of foetal growth in the sheep
and the guinea pig. J. Physiol., London 115: 158, 1951.
99. Greenberg, R. E., and J. Lind. Catechol amines in
tissues of the human foetus. Pediatrics 27: 904, 1961.
100. Grosser, O. FruhentwiclJing, Eihauthildung und Placentation
des Menschen und der Sdugciiere. Munchen : J. F. Bergmann,
I927-
101. Gruenwald, P. Pathology of perinatal distress. Arch.
Pathol. 60: 150, 1955.
102. Hamilton, W. J., J. D. Boyd, and H. VV. Mossman. In:
Human Embryology (2nd ed.). Cambridge: W. Heffer,
'952-
103. Hammond, W. S. The development of the aortic arch
bodies in the cat. Am. J. Anal. 6g: 265, 1941.
104. Hendricks, C. H., E. J. Quilligan, C. VV. Tyler, and
G. J. Tucker. Pressure relationships between the inter-
villous space and the amniotic fluid in human pregnancy.
Am. J. Obstet. Gynecol. 77: 1028, 1959.
105. Hertig, A. T., and J. Rock. Two human ova of the
previllous stage having an ovulation age of about eleven
and twelve days. In : Contribution to Embryology. Washing-
ton: Carnegie Inst. 29: 127, 1941.
106. Hertig, A. T., and J. Rock. Two ova of the previllous
stage having a developmental age of about 7 and 9 days
respectively. In: Contribution to Embryology. Washington:
Carnegie Inst. 31 : 65, 1945.
107. Hill, J. R. The oxygen consumption of newborn and
adult mammals. Its dependence on the oxygen tension
in the inspired air and on the environmental temperature.
J. Physiol., London 49: 346, 1 959.
108. Hon, E. H. The electronic evaluation of the foetal heart
rate. Am. ./. Obstet. Gynecol. 75: 1215, 1958.
109. Hon, E. H. Observations on pathologic foetal brady-
cardia. Am. J. Obstet. Gynecol. 77: 1084, 1959.
1 10. Hon, E. H., A. H. Bradfield, and O. W. Hess. The vagal
factor in foetal bradycardia. Am. J. Obstel. Gynecol. 82 :
291, 1961.
111. Huckabee, VV. E., J. Metcalfe, H. Prystowsky, and
D. H. Barron. Blood flow and oxygen consumption of
the pregnant uterus. Am. J. Physiol. 200: 274, 1961.
112. Huggett, A. St. G. Foetal blood-gas tensions and gas
transfusion through the placenta of the goat. J. Physiol.,
London 62 : 373, 1927.
1 13. Hughes, A. F. VV. The histogenesis of the arteries of the
chick embryo. ./. Anal. 77: 266, 1943.
114. Hutchinson, E. A., C. J. Perctval, and I. M. Young.
Cardiovascular responses in the growing kitten and
puppy. Quart. J. Exptl. Physiol. 47: 201, 1962.
115. Ingalls, T. H. Environmental factors in causation of
congenital anomalies. Ciba Found. Symp. Congenital Mal-
formations i960, p. 51.
116.
117.
119.
123.
124.
■25-
126.
127.
128.
130.
■31
132.
'33-
■34-
'35-
136.
■37-
Jacobson, H. N., and W. F. Windle. Responses of
foetal and newborn monkeys to asphyxia. J. Physiol.,
London 153:447, i960.
James, L. S., and R. D. Rovve. The pattern of response
of pulmonary pressures in newborn and older infants to
short periods of hypoxia. J. Pediat. 51 : 5, 1957.
James, L. S., I. M. VVeisbrot, C. E. Prince, D. A. Hola-
day, and V. Apgar. The acidbase status of human infants
in relation to birth asphyxia and onset of respiration.
J. Pediat. 52: 379, 1958.
Kalter, H., and J. VVarkany. Experimental production
of congenital malformation in mammals by metabolic
procedure. Physiol. Revs. 39:69, 1959.
Karki, N., R. Kuntzman, and B. B. Brodie. Norepineph-
rine and serotonin brain levels at various stages of onto-
genetic development. Federation Proc. 19: 282, i960.
Keen, E. N. The postnatal development of the human
cardiac ventricles. J. Anal. 89: 484, 1955.
Keith, J. D., R. D. Rovve, and P. Vlad. Heart Disease in
Infancy and Childhood. New York: Macmillan, 1958.
Kennedy, C, and L. Sokoloff. An adaptation of the
nitrous oxide technique to the study of the cerebral
circulation in children; normal values for cerebral blood
flow and cerebral metabolic rate in childhood. J. Clin.
Invest. 36: 1 130, 1957.
KjELLBERC, R. S., V. RUDHE, AND R. ZoTTERSTROM.
Heart volume variations in the neonatal period. Acta.
Radiol. 42: 173, 1954.
Krehbiel, R. H. Cytological studies of the decidual
reaction in the rat during early pregnancy and in the
production of deciduomata. Physiol. Zoo. 10: 212, 1935.
Landgren, S. On the excitation mechanism of the carotid
baroreceptors. Acta Physiol. Scand. 26 : 1 , 1952.
Larks, S. D. Fetal electrocardiography. Springfield, 111.:
Thomas, 1961.
Lewis, VV. H. Pinocytosis. Bull. Johns Hopkins Hasp. 49:
17. 1931'
Lind, J., and C. Wegelius. Human foetal circulation:
changes in the cardiovascular system at birth and disturb-
ances in the postnatal closure of the foramen ovale and
ductus arteriosus. Cold Spring Harbor Symp. Qjtant. Biol.
'9: '°9. '954-
l.[rsDQUisT, A., and G. VVallgren. L'nilateral artery
stenosis and fatal arterial hypertension in a newborn
infant. Acta Paediat. 51 : 575, 1962.
Lust, J. E., D. D. Hagerman, and C. A. Villee. Trans-
port of riboflavin by human placenta. J. Clin. Invest. 313:
38, 1954-
McLaren, A., and D. Michie. Congenital runts. Ciba
Found. Symp. Congenital Malformations i960, p. 178.
Martin, J. D., and I. M. Young. The influence of gesta-
tional age and hormones on experimental foetal brady-
cardia. J. Physiol., London 152: 1, i960.
Martin, J. D., and I. M. Young. Experimental foetal
bradycardia in the post mature rabbit. Australian J. Obstet.
Gynaecol. In press.
Misrahy, G. A., A. V. Beran, J. F. Spradley, and V. P.
Garwood. Foetal brain oxygen. Am. J. Physiol. 199:
959. i960.
Moore, R. E. Thermoregulation in newborn animals.
Ciba Found. Symp. Adrenergic Mechanisms, i960, p. 469.
Mossman, H. VV. The rabbit placenta and the problem
of placental transmission. Am. J. Anat. 37: 433, 1926.
THE FETAL AND NEONATAL CIRCULATION
1649
138. Mossman, H. W. Comparative morphogenesis of the
foetal membrane and accessory uterine structures In :
Contribution to Embryology. Washington: Carnegie Inst. 26:
■29. '937-
139. Mott, J. C. The ability of the young mammals to with-
stand total oxygen lack. Brit. Med. Bull. 17: 144, 1961.
140. Mott, J. C. The stability of the cardiovascular system.
Ciba Found. Symp. on Somatic Stability in the Newly Born.
1 96 1, p. 192.
141. Newton, W. H. Pseudo-parturition in the mouse and the
relation of the placenta to postpartum oestrus. J. Physiol.,
London 84: 196, 1935.
142. Niemineva, K., and L. Tervila. On the capillary bed
of the human foetal cerebellar hemispheres. Acta Anal.
'9: 2°4. 1953-
143. Page, E. S. Transfer of material across the human pla-
centa. Am. J. Obstet. Gynecol. 74: 705, 1957.
144. Patten, B. M. Varying developmental mechanisms in
teratology. Pediatrics 19: 734, 1957.
145. Pattle, R. E. Properties, function and origin of the
alveolar lining layer. Nature 175: 1125, 1955.
146. Peltonen, T., and L. Hirvonen. The ductus venosus.
Acta Paediat. 52: 202, 1963.
147. Penrose, L. S. Genetic causes of malformation and the
search for their origins. Ciba Found. Symp. Congenital
Malformations, i960, p. 22.
148. Phelps, D Endometrial vascular reactions and the
mechanisms of nidation. Am. J. Anat. 179: 167, 1946.
149. Prec, K. J., and D. E. Cassels. Dye dilution curves and
cardiac output in newborn infants. Circulation 1 1 : 789,
1955-
150. Prvstowskv, H., A. Hellegers, G. Meschia, J. Met-
calfe, \V. Huckabee, and D. H. Barron. Blood volume
of foetuses carried by ewes at high altitude. Quart. J.
Expll. Physiol. 45: 292, i960.
151. Pugh, L. G. C. E. Physiological and medical aspects of
the Himalayan scientific and mountaineering expedition,
1960-61. Brit. Med. J. 2: 621, 1962.
152. Raiha, C. E. Tissue metabolism in the human foetus.
Cold Spring Harbor Symp. Quant. Biol. 19: 143, 1954.
153. Ramsey, E. M. Circulation in the intervillous space of the
primate placenta. Am. J. Obstet. Gynecol. 84: 1649, 1962.
154. Reynolds, S. R. M. Adaption of uterine blood vessels
and accommodation of the products of conception. In :
Contribution to Embryology. Washington: Carnegie Inst. 33:
1, 1949.
155. Reynolds, S. R. M. Circulatory adaptions to birth and
their clinical implications. Am. J. Obstet. Gynecol. 70: 148,
ass-
ise. Reynolds, S. R. M. The fetal and neonatal pulmonary
vasculature in the guinea pig in relation to haemody-
namic changes at birth. Am. J. Anat. 98: 97, 1956.
157. Reynolds, S. R. M., and M. M. Cliff. A dose-stress
response of adrenaline affecting foetuses at a critical time
in pregnant rabbit. Anat. Record 134: 379, 1959.
158. Reynolds, S. R. M., F. W. Light, Jr., G. M. Ardran,
and M. M. L. Pritchard. Qualitative nature of pulsatile
flow in umbilical blood vessels with observations on flow
in the aorta. Bull. Johns Hopkins Hasp. 91 : 83, 1952.
159. Reynolds, S. R. M., and W. M. Paul. Circulatory re-
sponses of the foetal lamb "in utero" to increase of intra-
uterine pressure. Bull. Johns Hopkins Hosp. 97: 383, 1955.
160. Reynolds, S. R. M., and W. M. Paul. Pressures in
umbilical arteries and veins of the foetal lamb "in utero."
Am. J. Physiol. 193: 257, 1958.
161. Reynolds, S. R. M., and W. M. Paul. Relation ol
bradycardia and blood pressure of the foetal lamb "in
utero" to mild and severe hypoxia. Am. ./. Physiol. 193:
249. 1958-
iiu Richards, M. R., K. K. Merritt, M. H. Samuels, and
A. Langmann. Congenital malformations of the cardio-
vascular system in a series of 6,053 infants. Pediatrics 15:
'2. 1955-
163. Rogers, A. F. Irritability of the arteries of the human
umbilical cord. (Thesis) Bristol, England, 1948.
164. Rowe, R. D. Clinical Observations of Transitional Circula-
tions. Adaption to Extrauterine Life. Columbus, Ohio: Ross
Laboratories, 1959, p. 33.
165. Rowe, R. D., and L. S. James. The normal pulmonary
arterial pressure during the first year of life. ./. Pediat. 51 :
1. '957-
166. Rudolph, A. M., R. A. M. Auld. R. J. Golinko, and
M. H. Paul. Pulmonary vascular adjustments in the
neonatal period. Pediatrics 28: 28, 1961.
167. Rudolph, A. M., J. E. Drorbaugh, P. A. M. Auld,
A. J. Rudolph, A. S. Nadas, C. A. Smith, and J. P.
Hubbell. Circulation in the respiratory distress syndrome.
Pediatrics 27: 551, 1 96 1 .
168. Sandler. M., C. R. J. Ruthven, S. F. Contracter, C.
Wood, R. T. Booth, and J. H. M. Pinkerton. Trans-
mission of noradrenaline across the human placenta.
Nature 197: 598, 1963.
169. Scholander, P. Experimental Studies on Asphyxia in Animals.
Oxygen Supply to the Human Foetus. C.I.O M.S. Symposium.
Oxford: Blackwe.ll, 1959, p. 267.
170. Selye, H., and T. McKeown. Studies on the physiology
of the maternal placenta in the rat. Proc. Roy. Soc. London,
Ser. B. 119: 1, 1935-36.
171. Shelley, H. J. Glycogen reserves and their changes at
birth and in anoxia. Brit. Med. Bull. 17: 127, 1961.
172. Shepherd, J. T., and R. F. Whelan. The blood flow in
the umbilical cord of the foetal guinea pig. J. Physiol.,
London 115:1 50, 1 95 1 .
173. Smith, C. A. The Physiology of the Newborn Infant (3rd
ed.). Oxford: Bla< kwell Sci. Publ., 1959, p. 122.
174. Smith, S. E., R. S. Stacey, and I. M. Young. 5-HT con-
centrations in the gut and platelets of the developing
guinea pig. Chem. Pharm. In press.
175. Smythe, C. N., and J. L. Farrow. Present place in
obstetrics for foetal phonocardiography and electro-
cardiography. Brit. Med. J. 2: 1003, 1958.
176. Sontag, L. W., and T. W. Richards. Soc. Res. Child.,
Devel. Monog. 3, 1938, p. 4.
177. Spratt, N. T. Nutritional requirements of the early chick
embryo. Biol. Bull. 99: 120, 1950.
178. Stern, L., and J. Lind. Cardiovascular disease: perinatal
circulation. Ann. Rev. Med. 11: 113, i960.
179. Ten Berge, B. S. Capillair-activitie in placenta-vlokken:
de invloed van histamine en acetylcholine; de invloed
ophet beloop van de zwangerschap. Ned. Tijdschr. Geneesh.
99 : 3556, 1955-
180. Troen, P., and E. E. Gordon. Perfusion studies of the
human placenta. I. Effect of estradiol and human cho-
1650
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
rionic gonadotropin on citric acid metabolism. J. Clin.
Invest. 37: 1516, 1959.
181. Villee, C. A. The intermediary metabolism of human
foetal tissues. Cold Spring Harbor Symp. Quant. Biol, ig:
186, 1954.
182. Walker, J., and A. C. Turnbull. Oxygen Supply to the
Human Foetus. C.I. O. M.S. Symposium. Oxford: Blackwell,
1959. P- '55-
183. Wallgren, G., P. Karlberg, and J. Lind. Studies of
the circulatory adaption immediately after birth. Acta
Paediat. 49: 843, i960.
184. Walls, E. W. Development of specialized conducting
tissue of human heart. J. Anal. 81 : 93, 1947.
185. Warkany, J. Congenital malformations and pediatrics.
Pediatrics 19:725, 1957.
186. West, G. B., D. M. Shepherd, R. B. Hunter and A. R.
MacGregor. The functions of the organs of Zuckerkandl.
Clin. Sci. 12: 317, 1953.
187. Westin, B. Technique and estimation of oxygenation of
the human foetus in utero by means of hystero-photography.
Acta Paediat. 46: 117, 1957.
188. Whittam, R. Sodium and potassium movements in
kidney cortex slices from newborn animals. J. Physiol.,
London 153:358, i960.
189. Windle, W. F. Physiology of the Foetus. Philadelphia:
Saunders, 1940.
190. Wislocki, G. B., and E. W. Dempsey. The chemical
histology of the human placenta and decidua with
reference to the mucopolysaccharides, glycogen, lipids
and acid phosphatase. Am. J. Anat. 83: 1, 1948.
191. Wislocki, G. B., and G. L. Streeter. Placentation of
the macaque (Macaca mulatta) from the time of implan-
tation until the formation of the definitive placenta. In:
Contribution to Embryology. Washington : Carnegie Inst. 27 :
', I938.
192. Wong, M., and D. E. Cassels. The foetal electrocardio-
gram A.M. A. J. Diseases Children 99: 4, ig6o.
193. Woodbury, R. A., M. Robinow, and W. F. Hamilton.
Blood pressure studies on infants. Am J. Physiol. 122:
472. '938-
194. Young, I. M. The uterine, placental and foetal circula-
tions. In : The Control of the Circulation of the Blood, edited
by R.J. S. MacDowall. London: Dawson, 1956, vol. 2,
p. 184.
195. Young, I. M. Some observations on the mechanism of
adrenaline hypernoea. J. Physiol., London 137: 374, 1957.
196. Young, I. M. Blood pressure in the newborn baby. Brit.
Med. Bull. 17:1 54, 1 96 1 .
197. Young, I. M., and D. C Cottom. (To be published.)
198. Young, I. M., and W. W. Holland. Some physiological
responses of the neonatal arterial blood pressure and
pulse rate. Brit. Med. J. 2: 276, 1958.
CHAPTER 47
The flow of blood through bones and joints
WALTER S. ROOT Department of Physiology, College of Physicians and Surgeons, New York City
CHAPTER CONTENTS
Bones
Long Bones
Vertebrae
Flat Bones
Nerve Supply of Bone
Blood Flow in Bone
Oxygen in the Blood of Bones
Intramedullary Pressure
Temperature of Bone Marrow
Joints
The Blood Supply
The Nerve Supply
Blood Flow Through Joints
Nervous Control of Joint Blood Vessels
bone must not be thought of as an inert substance,
but rather as one of the highly specialized tissues of
the body, consisting of active cells which respond
promptly to physiological demands upon the skeletal
and hematopoietic systems. The cells are sensitive to
nutritional and functional processes, and differ in
their reactions from those of other tissues only because
of the rigidity and stability of the intercellular de-
posits of mineral salts.
Contrary to general belief, bone is a relatively
vascular tissue (68). This concept is borne out by the
rapidity with which substances injected into bone
marrow appear in the general circulation (105, 106).
Large infusions can be administered in a short time
(4), and this route of giving fluid has been useful in
dealing with infants (43) and in treating patients in
conditions of hemorrhage and shock (101).
BONES
The anatomical features of the vascular system in
bones are classified as long bones, flat bones, or
vertebrae.
Long Bones
Most studies of the long bones have been made
upon the femur or the tibia-fibula. The long bones
receive blood from three sources: a) the nutrient
artery or arteries entering the bone in the shaft, b)
blood vessels entering the ends of the bone, and c)
blood vessels penetrating the periosteum (fig. 1).
Radiological observations in which the arteries are
injected with radiopaque substances show that the
principal nutrient artery of the femur traverses the
cortex inclined toward the knee (16). No branches
are given to the cortex in the nutrient canal. On
entering the medulla, the artery divides into ascend-
ing and descending limbs which, with few subdivi-
sions, pass to either end of the bone. In general, larger
arteries are visualized as sharply defined and tortuous
channels in the proximo-distal axis, and are com-
paratively few in number.
Medullary arteries can be traced to the meta-
physeal region where they break up into numerous
fine vessels which join across the line of union at the
epiphyseo-metaphyseal synostosis with others derived
from the epiphyseal arteries (fig. 2). Arterial twigs
from the main medullary arteries can be seen to pass
more or less transversely toward the endosteal aspect
of the compactum where they recurve and course for
a short distance in the peripheral medullary zone.
Here they anastomose with one another, and give rise
to fine vessels which pierce the endosteal face of the
compactum and arborize irregularly in the inner
cortical zone (16).
There has been some difference of opinion con-
cerning the vascular supply to the metaphyseal re-
gion. Thus, Weinmann & Sicher (111) state that the
nutrient artery supplies the central part of the meta-
physis, its more peripheral parts being fed by meta-
physeal arteries derived from the periosteum. On the
1651
rfea
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
PRINCIPAL NUTRIENT
ARTERY AND VEIN
PERIOSTEAL CAPILLARIES -
IN CONTINUITY WITH '
CORTICAL CAPILLARIES
other hand, Trueta & Harrison (107)
believe that in the adult human femur
the nutrient artery does not reach the
metaphyseal region which is wholly sup-
plied by metaphyseal arteries. Studies
on rabbit embrvos and young rabbits metaphyseal arteries
AND TERMINALS OF THE
indicate that in the earliest stages of de- medullary arterial
SYSTEM ^«^
velopment the metaphysis receives blood
only from the nutrient artery. Later,
metaphyseal arteries derived from the
periosteum take over the supply of the
peripheral region, the extent of the area
supplied bv them increasing progres-
sively (85).
The arteries supplying the proximal
part of the femur and the acetabulum
are the lateral femoral circumflex, the
medial femoral circumflex, the obtu-
rator, the superior gluteal, the inferior
gluteal, the first perforating artery, and
the nutrient artery of the femur (62).
Branches of these enter the head and
neck of the femur through small foram-
ina, or enter by way of the fovea cen-
tralis, or are carried to the head of the
femur in the ligamentum teres.
In children the foveal vessels assume
a small role in supplying blood to the
femoral head (108, 114). When these
arteries do penetrate to the ossification
center they are probably supplemen-
tary (62). In the adult the foveal ar-
teries are larger and usually nourish the
femur. Apparently, arteries enter the
femoral head through the ligamentum section. [From Brookes (13).]
teres in the majority of cases (108, 1 14),
but this supply is supplementary to the vital supply
from the capital arteries (62).
All investigators have emphasized the importance
of the capital branches of the medial femoral circum-
flex artery for the nutrition of the femoral head (109).
The terminal branches enter the femoral head at the
articular rim, just posterior to the superior and in-
ferior poles of the femoral neck. Most investigators
have found that the superior posterior branches are
larger and more numerous than the inferior posterior
arteries. Usually, no arteries enter the femoral head
or neck anteriorly (109). The nutrient artery cannot
be traced past the marrow cavity (62).
Vessels enter the distal end of the femur through
three groups of foramina: supracondylar, condylar,
and intercondylar (99). In each some 10 to 35 for-
END-ARTERIAL
TERMINALS
VENOUS SINUSOIDS V
.--'METAPHYSEAL VEINS
MEDULLARY
SINUSOIDS
INTERFASCICULAR
CAPILLARIES IN MUSCLE
CENTRAL VENOUS
CHANNEL
LARGE EMISSARY VEIN
TRANSVERSE EPIPHYSEAL
VENOUS CHANNEL
fig. i . Diagram of vascular organization of rat tubular bone in longitudinal
amina are present. Condylar arteries perforate the
cortex and ramify within the spongiosa. Terminal
branches of the middle geniculate arteries pass
through the intercondylar foramina and are distribu-
ted to the central parts of the epiphysis. Rami arising
from the superior, lateral, and middle geniculate
arteries pass through the anterior and posterior
supracondylar nutrient foramina, and are distributed
to the distal end of the diaphysis. The generous vascu-
lar supply explains the lack of ischemic necrosis after
fractures of the lower end of the femur.
The blood supply to the cortex or compactum of
long bones runs in longitudinal canals known as
Haversian canals. In man the canals vary from 25 to
125, averaging 50 yu in diameter, but larger ones are
also seen (70). Although these canals run longitudi-
FLOW OF BLOOD THROUGH BONES AND JOINTS
1653
Periosteal arteriole
and venae comitantes'
Periosteal
capillaries
Medullary
artery
Cortical
-x" capillaries
Endosteal
capillaries
Medullary
sinusoids
Central
enous sinus
fig. 2. The blood vascular organization of diaphyseal tu-
bular bone represented diagramatically in transverse section.
[From Brookes (14).]
fig. 3. A longitudinal section of cortical bone showing the
anastomosing and branching Haversian canals. The com-
municating canals between the Haversian canals are demon-
strated. [From Jaffe (70).]
nally, they do not run vertically for more than short
distances soon deviating from a straight line. The
canals form a continuously anastomosing and ramify-
ing network (fig. 3). Beneath the articular cartilage
at the upper and lower ends of a bone, the canals run
transversely to the long diameter of the bone. Near
the surface of the bone, Haversian canals communi-
cate with the canals of the ground lamellae which
open to the external surface of the bone, and the
innermost canals lead into the medullary cavity. Re-
cently the term "macrocanalicular system" has been
used to refer to the system in mineralized tissue which
is made up of Haversian and anastomosing Volkmann
spaces (68). It is generally stated that one or two
capillaries are present in an Haversian canal (79). A
single endothelial tube surrounded by a slight
adventitia has been described in Volkmann's canals.
Lexer (86) had emphasized the role of the periosteal
arteries in bone nutrition, but more recent studies
indicate that the periosteal circulation may be scanty
(20), periosteal arteries being found rarely or only
with difficulty (i). Also, the notion of a periosteal
arterial penetration of compact bone has recently
been rejected as a result of microradiographic analy-
sis in the rabbit ( 16), and in the rat and human fetus
(13, 14). This opinion is further strengthened by a
study of nonischemic adult tubular bone (15). Ac-
cording to certain investigators (13-16), normal
diaphyseal blood flow is centrifugal, that is, passing
from the medullary arterial system outward through
the cortex into the periosteal and interfascicular capil-
laries of muscle. Drainage of compact bone is effected
either by way of periosteal capillaries or through
medullary sinusoids and the central venous channel.
Apparently, the vascular systems of bone and peri-
osteum are united, but only at the capillary level.
This would explain the survival of outlying bone cells
seen by Marneffe (88) in rat diaphysis nourished by
the periosteum alone. It also provides a basis for the
development of a collateral circulation as found in
Johnson's experiments in dogs (72).
The obliquity of Haversian canals has been noted
by Cohn & Harris (29) and others. Brookes (13),
studying rat femora and tibiae as a whole, was able
to show how cortical vascular obliquity is in opposing
senses at either end of a long bone, the two regions
meeting by abrupt directional changes. In the adult
human tibia this change takes place at the inferior
metaphysis and may well be a factor in the delayed
healing of fractures at this site, where a rich venous
outflow would predispose to recurrent hematoma
formation.
It seems likely that the normal arterial supply to
cortical capillaries is mediated by medullary end
arteries. This conception is supported by the findings
of Eletto (42) who noted the lack of anastomoses be-
tween branches of the principal nutrient artery in the
medulla. It would help to explain the occurrence of
irregular bone cell necrosis in the cortex produced by
the injection of particulate suspensions (75, 76), or by
interruption of the principal nutrient artery (10).
Epiphyses also seem to contain discrete circumscribed
ib-,4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
vascular zones with little functional overlap, the ob-
struction of which on either arterial or venous sides
is a factor in osteochondritis juvenilis (19).
The role of the periosteum in bone regeneration
and the incorporation of blood vessels from the
periosteal vascular network during bone growth in
width can still be accepted in that the vitality of the
osteogenic layer of young periosteum is maintained
by the osteogenic capillary layer fed by periosteal
arteries (54).
As visualized radiographically, the venous system
differs somewhat from the arterial system. Thus, a
solitary longitudinal channel of wide caliber in an
approximately central medullary position can be
traced from one end of a bone to the other. Since this
central canal lacks a muscular tunica media (88), it
may properly be called a central venous sinus in the
medulla (16). At the trochanteric fossa the central
venous channel is joined by tributaries from the lesser,
third, and greater trochanter as well as by a vessel
p.issing down the neck from the head of the femur.
The central vein is joined by the principal nutrient
vein a short distance below the nutrient canal and
passes characteristically as a single vessel down to the
inferior metaphysis where it anastomoses with an
ascending branch of the middle geniculate vein.
Sometimes it divides into two stems at the mid-shaft
level. The central venous channel has numerous
transverse branches radiating toward the endosteum
and these drain the sinusoids of the medulla. An
endosteal line marking the junction of the medullary
sinusoids with the cortical capillaries can be seen.
Branemark (11), who has been able to visualize
bone circulation in the living rabbit, states that the
bone marrow arteriole divides dichotomously into
capillaries. These run to sinusoids which are sometimes
hexagonal, sometimes spindle shaped. Sinusoids may
unite to form sinusoidal systems. The sinusoids are
drained by venules into collecting venules which
empty into the central veins. The sinusoids vary
rhythmically in their degree of dilation. Blood cells
may bypass a sinusoid by flowing through a shunting
capillary directly into a venule. In some instances,
cells appear to hug the vessel wall of one half of a
sinusoid apparently without disturbing flow in the
other half. Capillaries stemming from marrow ar-
terioles enter the Haversian canals to supply endos-
teal parts of diaphyseal bone. The capillaries then
swine; back into the marrow to empty into sinusoids
or directly into collecting venules. Blood flow in bone
capillaries is fairly steady, and the velocity of flow is
higher than in marrow capillaries.
In the pigeon, "transitional capillaries" (37, 38)
connect the arteries to the venous sinusoids. The
capillary link is extremely circumscribed, and it is not
until the venous sinusoidal anastomoses are reached
that the blood spreads out in lacing and interlacing
vessel tufts, thence to be directed from the tuft-like
branchings into larger and larger vessels eventually to
enter the central longitudinal vein almost at right
angles.
There seems little doubt that the extensively dis-
tributed, spacious, thin-walled venous sinusoids nor-
mally form the principal functioning vascular bed for
the actively circulating blood in marrow, i.e., they
correspond to the capillaries of other organs.
In pigeons in which the marrow is made hypo-
plastic by starvation, one can see, between the fat
spaces, well-outlined and clearly defined channels
which constitute a most extensive system of capil-
laries (37). Many of these appear to be nonpatent and
functionally dormant as far as the active blood circu-
lation is concerned. These capillaries come off the
venous sinusoids by way of conical openings, and seem
to be continuous with them. They are not capillaries
in the sense of an arteriovenous transition, but instead
extend from venous channel to venous channel; they
are intersinusoidal. The same intersinusoidal semi-
collapsed channels have been reported in the marrow
of the ribs of the white rat 138), and are believed to be
present in the dog (40).
Three theories as to the nature of the circulation in
adult marrow have been advanced : a) by Rindfleisch
(98) who believed that the blood spaces are lined by
parenchvma alone and have no endothelial cells; b)
Langer (81), on the other hand, thought of the mar-
Spinal
branches
Ddrsc^i
medial a.
Aorta-
Calcified
spongiosum
Cartilaginous
Anterolateral a.
fig. 4. Diagram of a transverse section through the mid-
body of the vertebra of a six-month-old fetus. [From Ferguson
(46).]
FLOW OF BLOOD THROUGH BONES AND JOINTS
165
00
row as an entirely closed vascular system; and c)
Bunting (18) pictured vessels lined with epithelium,
but with openings at various points communicating
directly with the medullary parenchyma. A survey of
the literature indicates no general agreement among
those who have studied the subject.
I 'a librae
In the lumbar and thoracic regions the aorta gives
off paired segmental arteries. These penetrate the
anterior groups of spinal muscles and continue pos-
teriorly in the horizontal plane to pass on either side
of the vertebral body and lie in direct contact with
the anterior and lateral wall of this structure (fig. 4)
to which small branches are contributed (56). Each
artery gives off a large branch in the trough formed
by the vertebral body and the transverse process.
This branch traverses the intervertebral foramen and
divides into three terminal arterioles (113). One of
these passes to the posterior surfaces of the two adja-
cent vertebral bodies. A second runs to the spinal
cord and its meninges. The third supplies the pos-
terior vertebral processes and surrounding soft
structures.
The first branch mentioned above divides within
the spinal canal, one terminus running upward and
medially across the posterior surface of the vertebral
body under the posterior spinal ligament, to enter a
foramen about the center of the body. The other
terminus runs downward and medially to a similar
entrance in the center of the body of the next distal
vertebra. Thus, there are four diagonal arteries, two
from each side converging to enter the center of the
posterior surface of each vertebra, either through a
common foramen or through separate foramina. The
arterioles may coalesce or remain separate before
radiating to all parts of the centrum (113). The main
dorsal vertebral artery and the right and left antero-
lateral arteries appear to end in the middle of the
developing osseous spongiosa. No arterial branches
can be demonstrated beyond the center of the verte-
bral body. Irregular vascular canals can be seen in
the spongiosa and a diffuse network of thin-walled
channels is present in the surrounding cartilaginous
zone. Very small vessels perforate the cartilaginous
plate and tiny capillary channels permeate the can-
nulus fibrosa. The branches to the cord anastomose
freely with the anterior and posterior spinal arteries
which lie on the respective surfaces of the cord ex-
tending from within the skull to the end of the cord.
The pedicles, transverse processes, articular facets,
and lamina have a good arterial blood supply through
the anastomosing branches of the posterior rami from
the paired segmental arteries (46). Similar anatomical
arrangements are true for the cervical vertebrae (56).
The intervertebral disc tissues appear to offer an
important focus for degeneration as they are always
farthest from the arterial supply (46).
Each lumbar vertebral body is drained by four
main venous trunks. Two leave the body, one on
either side from an anterolateral position at a level
just above the midline; two emerge as paired vessels
from the bony foramen in the center of the posterior
vertebral wall (110). The two posterior veins empty
at once into the anterior longitudinal meningo-
rhachidian veins of the posterior external plexus.
Direct connection with the corresponding lumbar
veins is made through the spinal rami of the latter.
The two anterolateral veins from the vertebral body
empty directly into the lumbar veins. The lumbar
veins passing horizontally are in direct communica-
tion with three great longitudinal or vertical venous
systems: a) posteriorly with the posterior external
venous plexus which extends vertically within the
spinal canal, external to the spinal cord membranes;
b) with the azygos or hemiazygos systems, and c) with
the inferior vena cava.
Within the vertebral body the two posterior and
the two anterolateral veins meet to form a large
reservoir. Although these veins have the usual venous
structure beyond the vertebral periosteum, the wall
structure is replaced by a limiting membrane of
flattened endothelial cells within the body. Radiating
peripherally from the central venous basin are many
irregular columnar spaces which occupy approxi-
mately forty per cent of the entire vertebral body.
The more peripheral parts of the venous spaces con-
tain within their lumens a very large proportion of
hematopoietic tissue, together with reticuloendo-
thelial elements. Hematopoietic tissue is occasionally
found within the lumen of the central venous space
(no).
Flat Bones
The mandible appears to be the flat bone concern-
ing which the most information is available. The
periosteum and outer circumferential portions of the
osseous mandible are supplied by such adjacent
arteries as the facial, submental, inferior alveolar,
mylohyoid, mental, masseric, lateral pterygoid, medial
pterygoid, temporal and sublingual branch of the
lingual (36). One or more nutrient foramina passing
,(>-,<>
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
through the marrow space are evident above the
genial tubercle, and multiple small foramina are
usually seen in the small triangular area on the inner
face of the ramus below the mandibular notch and
above or on the endocondylar and endocoronoid
ridges.
Within the mandibular canal the inferior dental
artery gives rise to blood vessels which pass upward
toward the alveolar border (28). Some vessels pass
toward the lower border, but these are few in number.
The lower border of the mandible is supplied mainly
by periosteal vessels. The inferior dental artery within
the mandibular canal is surrounded by numerous
vessels, presumably venae comitantes.
Some eleven areas of cortical bone are recognized
(36), the regions being based upon the direction of
the canals in the Haversian mesh. There is little evi-
dence of the presence of lacunae or canaliculi in the
adult.
Nerve Supply of Bone
Perhaps the most complete and careful study of the
innervation of bone has been carried out by Kuntz &
Richins (78). According to their study the absence of
any nerve fibers not in close proximity to blood
vessels, in preparations in which excision of the dorsal
root ganglia had resulted in degeneration of the af-
ferent fibers, leads to the conclusions: a) that the
parenchymatous tissue of the bone marrow is devoid
of direct afferent innervation, and />) that, in prep-
arations of normally innervated bone marrow, un-
myelinated fibers which exhibit no obvious relation-
ships to blood vessels represent unmyelinated afferent
fibers or the unmyelinated terminal portions of
myelinated ones.
afferent fibers. The conception of some sensory
innervation of the bone marrow is supported by the
common clinical observation that puncture of bone
gives rise to pain, and the finding that many afferent
fibers in bone fall within the caliber range of the
pain-conducting fibers is in full agreement with this
view. Schleicher (100) noted that when blood plasma
was infused into the sternum a sharp pain was felt
about the infusion area when the pressure of the in-
coming fluid was greater than the intramedullary
pressure. In many persons with multiple myeloma
and metastatic bone lesions, distinct bone pain is asso-
ciated with sudden straining or coughing (94). It is
possible that the sudden elevation of intramedullary
pressure shown to result from such effort results in
distortion of the arteries and arterioles bearing
sensory nerves and that this produces pain.
When the sympathetic nerve fibers are caused to
degenerate by removing the appropriate sympathetic
ganglia, afferent fibers are found running to blood
vessels. They are present in relationships which indi-
cate that they are also connected with receptors im-
bedded in the parenchyma of the marrow (78). Foa
(47, 48) suggests that the afferent fibers may play a
part in the reflex regulation of functional activity of
bone marrow, and Chiray et al. (24) have shown that
the intramedullary injection of certain substances will
produce reflex changes in blood pressure.
efferent fibers. The relation of the sympathetic
innervation to bone circulation has been studied
either by sectioning or stimulating sympathetic fibers.
It seems clear that cutting sympathetic fibers causes
vasodilation and hyperemia (39, 40, 48, 60), and that
stimulation produces vasoconstriction (39, 48, 60,
1 12).
Hurrell (69) has traced nerve fibers into and along
Haversian canals of adult bone into two-thirds of the
thickness of the shaft of a cat's femur. Some end
blindly in the bone matrix; others, in close connection
with osteocytes. He suggests tentatively that the nerve
fibers found may be the two ends of a reflex arc
governing bone growth and maintenance. In this con-
nection Coppo (31) has reported a decrease in the
percentage of ash, and a modification of its composi-
tion, 8 to 1 5 davs after denervation of bone. Neverthe-
less, all experiments on animals have shown that uni-
lateral sympathectomy by itself has no observable
effect on bone growth (3, 21). The results obtained
from experiments on the effect of sympathectomy on
the healing of fractures are equivocal. Some investi-
gators have found healing to be accelerated (30),
whereas others have seen no effect (89). According to
Corbin & Hinsey (32), bones and joints are not sup-
plied with nerves having specific trophic functions.
Blood Flow in Bone
The circulation in bone is sufficient to supply the
normal variations of physiological processes, but often
fails to respond to the extreme insults of trauma or
infection. The delayed union of comminuted bumper
fracture of the tibia, and the extensive involvement of
the shaft of the long bones in osteomyelitis are classi-
cal examples. In situations where end arteries are
present, as at the metaphyseal side of the epiphyseal
plate, infarction is common. Aseptic infarction is
FLOW OF BLOOD THROUGH BONES AND JOINTS
1657
found in nutritional diseases such as scurvy and
rickets, and septic infarction with abscess formation
in tuberculosis and other forms of bacteremia (23).
Arterial blood from the terminal arborizations in
the cortex, derived from the medullary arterial sys-
tem, empties into a vascular lattice contained in the
canals of Havers and Volkmann (16). Here the circu-
lation is probably very sluggish and, besides move-
ment up and down the diaphysis, blood may be
shifted into either medulla or periosteum depending
upon functional variations in opposing muscles and
hematopoietic activity in the marrow. According to
Branemark (11) blood flow in bone capillaries is
fairly steady and the velocity is greater than in mar-
row capillaries. Externally, the vascular lattice of the
cortex connects with the osteogenic capillary layer;
internally, with the medullary sinusoids. The former
route to the systemic veins is direct and probably
drains most of the blood circulating in the cortex. The
latter route is indirect, through the sinusoids, into the
central venous sinus, and thence via the nutrient vein
at the bone extremities into periarticular veins (16).
Lamas et al. (80) have pointed out that a relatively-
slow blood flow in bone should be expected, since
none of the three functions of long bones, mechanical
support of the body, storage of calcium salts, and
hematopoiesis, needs a rapid circulation. In this con-
nection, it may be noted that the arrangement of
blood vessels within bone favors a slow circulation.
Thus, the nutrient artery describes many curves
before entering bone and after dividing into ascend-
ing branches which run through the marrow, it ends
in wide blood spaces close to the epiphyses. These
blood spaces are in close contact with thin-walled
veins of wide caliber. This arrangement of blood
vessels reduces the pressure and speed of circulation
in the arteries, and enables the vein to carry sub-
stances quickly away from the blood spaces: hence
the efficacy of therapeutic injections into marrow.
Few quantitative studies of blood flow through
bone have been made, [ones (73), who studied the
uptake rate of a radioactive colloid which is highly
selected by marrow cells, found a minimal circula-
tion through the red marrow in the rabbit amounting
to 7 per cent of the circulating blood volume per min.
A corresponding value for man would be about
300 ml.
The blood flow through human bone has been
studied by measuring the rate of clearance of I131 from
bone marrow (94). Xo apparent correlation was
found between marrow clearance rates and hemo-
globin level, leucocyte count, body temperature, or
blood pressure. The intravenous injection of hexa-
methonium bromide, a ganglionic blocking agent,
reduced the clearance rate of marrow, and this ap-
pears to be directly related to the fall in blood pres-
sure. Conversely, the injection of Paredrine resulted in
a distinct increase in the clearance rate from the
marrow, presumably as a result of the sympathomi-
metic action of the drug. A decreased flow through
the perfused tibia of the dog can be produced by
stimulating sympathetic nerve fibers or by adding
Adrenalin to the perfusion fluid (39).
Plethysmographic measurements of blood flow
through the normal humerus of man have been made
by Edholm et al. (41). They report a blood flow
through the nutrient artery of 0.5 to 1.0 ml per 100 ml
of bone per min. They point out that this value may
represent only half the total flow, for the periosteal
vascular supply is not included. They calculate on the
basis of these measurements that the total skeletal
blood flow should be 74.5 ml per min, although they
concede that this is bound to be an underestimation,
for bones with an active marrow are more vascular
than the humerus. The above measurements are much
lower than the values of 3.5 to 41 ml per 100 g bone
per min found in perfusion experiments using the
tibia of the dog (40).
hyperemia. Blair (7) has suggested that alternating
ischemia and hyperemia maintain normal bone calcifi-
cation and aid healing after a fracture. In this con-
nection it should be noted that hyperemia has long
been thought to be the physiological basis for localized
deossification. Thus, Leriche & Policard (83) state:
"If by any process whatever, the activity of the circu-
lation is increased in the vicinity of bone, the latter
becomes rarefied." Also, Greig (52) has written:
"Every trauma of bone is followed by a reactionary
local hyperemia, and every disease resulting in bone
rarefaction or decalcification is accompanied by a
more or less copious and prolonged increase of the
arterial and capillary circulations." De Lorimer et al.
(87) interpret their radiological studies showing areas
of bone rarefaction as being the result of localized
reflex hyperemia. They believe this may be produced
by trauma even of minor degree as in simple contu-
sions, sprains, or overenergetic physical therapy, or
by infection or neoplasms.
Although the above statements seem to rest more
on logic than on observable facts, it seems clear that
vascular resorption of bone is related to definite circu-
latory changes. Thus, Miller & cle Takats (91), who
carried out plethysmographic studies of blood flow on
i658
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
1 2 patients suffering from painful osteoporosis follow-
ing injury, found an increased blood flow amounting
to 5 to 60 per cent in the affected limb. Also, in severe
inflammatory processes an actual coalescence of
several Haversian canals takes place, the bony parti-
tion between them disappearing with the formation
of large spaces in bone containing numerous blood
\ essels and considerable vascular granulation. The
process is not confined to the Haversian system, but
also involves the spongy tabeculae. The marrow also
becomes extremely vascularized, as a result of the
proliferation of the existing vessels (71).
reduction of blood supply. According to Haslhofer
(58) the richness of anastomoses in long bones pre-
vents bone infarction even when the blood supply to
the nutrient artery is interrupted. On the other hand,
Axhausen & Bergmann (2) present clinical instances
of aseptic bone necrosis which they ascribe to interrup-
tion of local blood supply. Also, Phemister (96, 97)
has published radiographic and pathologic descrip-
tions of lesions which he considers to be the result of
marrow infarction.
Since the production of bone and marrow infarcts
in animals was generally considered impossible by
conventional means, earlier investigators resorted to
extensive stripping procedures or to the production of
multiple small emboli designed to occlude large num-
bers of capillaries. Thus, Brunschwig (17) attempted
to produce infarction of the marrow of the femur in
dogs by stripping the entire periosteum and simulta-
neously cutting the nutrient artery. Despite this ex-
tensive trauma, no evidence of infarction was seen in
adult dogs. Among the injection experiments are
those of Wollenberg (115) who injected talc into the
femoral artery of dogs and observed areas of necrosis
in metaphyses and epiphyses. Bergmann (6), on the
other hand, could find no changes in epiphyses, al-
though he saw widespread necrosis of cortical bone
after the injection of particles of silver suspended in
gum arabic, and Kistler (75-77) could find no in-
farcts following ligation of the nutrient artery of the
femur in rabbits. However, the injection of suspen-
sions of charcoal in acacia and of masses of aggluti-
nated bacteria, under unmeasured but admittedly
high pressure, produced areas of necrosis in the center
of the metaphyses.
Huggins & Wiege (65) seem to have been the first
to report changes following occlusion of the nutrient
artery only. In both mature and immature rabbits
ligation of the nutrient vessels to the femur was fol-
lowed in all instances by infarction of the marrow.
Although in a few cases there was some periosteal and
endosteal reaction above the operative site, no evi-
dence of bone infarction was found. In a recent study
Brookes (12) has shown that occlusion of the princi-
pal nutrient canal of the femur in day-old rabbits
produces an initial shortening, followed by equaliza-
tion, and then a final absolute shortening of 3 per cent
in the occluded femur.
Variations in nutrition of the growth cartilage will
cause shortening or lengthening of long bones. Thus,
interruption of the medullary arteries and diversion
of blood to the growth cartilage presumably will ac-
count for the increased growth rate observed in long
bones affected by a variety of conditions, such as
fractures, chronic infections, and tumors (45). Occlu-
sion of the nutrient canal may result in diversion of
blood into the epiphyseal and metaphyseal arteries,
with possible ischemia of the diaphysis. It is probable
that while collateral circulation is developing in the
bone extremities, a diminution in the femoral blood
supply to the growth cartilages occurs, thus account-
ing for the growth lag of occluded femora noted by
Brookes (12) in the first 30 days. With the establish-
ment of a collateral circulation by means of anasto-
moses between the metaphyseal arteries and the
principal nutrient artery, blood flow near the growth
cartilage is increased bringing about equalization in
the length of occluded and normal femora in the
intermediate growth phase. In the final phase, 120 to
150 days, a relative decrease in the nutrition of the
growth cartilages must occur to account for the 3.7
per cent retardation in femoral growth seen at ma-
turity. The reason for this is not clear. However, it
seems probable that towards the end of growth the
collateral circulation is not quite able to furnish the
same quantity of blood to die medullary artery as
when the nutrient artery is also available as a supply
channel (12). These results may be compared with
the evidence of bone lengthening after fractures in
children (22) in whom presumably the same local
vascular mechanism is active that determined the
growth curve of occluded femora in rabbits.
It is generally thought that the disturbances on the
venous side produced by obliterating the vein accom-
panying the nutrient artery is slight, because of the
profuse venous drainage at the bone extremities (57)
and at the surface of the diaphysis (16).
Oxygen in the Blood of Bones
Ham (53) has shown by measurements in the dog's
radius that bone cells, if they are to survive, can gen-
FLOW OF BLOOD THROUGH BONES AND JOINTS
'659
erally be no farther than 100 n from a nutrient vessel.
Actually few quantitative measurements of the oxy-
gen content of the blood circulating through bone
have been made, and these have been concerned with
the blood taken from marrow. Thus, Grant & Root
(51) punctured the humerus of unanesthetized dogs
and determined the oxygen saturation and tension in
the first 0.15 ml of blood removed. The oxygen satu-
ration of bone marrow blood is similar to that of blood
drawn from the jugular vein. The oxygen contents
and capacities, and the hematocrit values decrease to
about the same extent in jugular and bone marrow
blood after a 30 per cent hemorrhage, and recovery
takes place at the same rate. Similar studies have been
carried out on normal man and patients suffering
from anemia and polycythemia vera (5, 102), and pa-
tients with primary and secondary polycythemia
(59). The oxygen saturation of marrow blood was
found to be similar to that of normal man; that of
polycythemic individuals appears to be greater than
normal.
The factors controlling the oxygen tension and
oxygen supply to "erythrogenic nests" are not well
understood. It is likely that the oxygen consumption
of these growing cells is one consideration, and that
the quantity and partial pressure of oxygen in the
blood perfusing the area constitutes another factor.
The bulk of marrow blood appears from histological
evidence to be contained in the venous sinusoids. It
seems reasonable to think that most of the mature
erythrocytes found in the sample obtained by needle
puncture are derived from the sinusoids. Presumably
the erythrogenic nests are in diffusion equilibrium
with the blood in the sinusoids even though the
growing erythroid cells may be sealed off from the
sinusoids as proposed by Doan (37, 38).
Intramedullary Pressure
The pressure within the medullary canal has been
measured in various bones in different animals by a
number of investigators. In animals, such as the dog
and cat, it varies between 20 and 1 15, averaging some
50 mm Hg (8, 60, 74, 82). According to Petrakis (95),
the pressures in the marrow of patients without
marrow disease are uniformly low. Thus, in the
sternum the pressures were approximately atmos-
pheric, ranging from 2/0 to 17/15 mm Hg. Tocantins
& O'Neill (106) report intramedullary pressures of
50 to 120 mm H20 (3.7-8.9 mm Hg) in the human
sternum. Petrakis (95) notes that human marrow
pressures are lower than those seen in lower animals,
and suggests that the difference may be attributed to
the effects of anesthesia. The pressures measured in
the region of the diaphysis are said to be definitely
higher than those found near the epiphyses (fig. 5)
(103).
Marrow pressure records show a definite, but small,
pulse pressure (60, 74, 82, 103). This observation may
be of practical importance, for Miles (90) reports
that the absence of such fluctuations in pressure in the
femoral head of patients indicates necrosis of this
structure.
In addition to changes in pulse pressure, records of
marrow pressure show rhythmic fluctuations corre-
sponding to respiration (fig. 5) (60, 103). Also, slower
rhythmic variations in pressure, presumably Traube-
Hering waves, are sometimes seen (8).
Rasgone, Vater, and Marbarger (see 74) concluded
that the marrow behaves as a semiclosed cavity and
that changes in intramedullary pressure are dependent
upon the volume of blood within the marrow cavitv.
When the venous return is obstructed, the mean
pressure of the marrow increases and the pulse pres-
100
0
x
5
50
RESP
DIAPHYSIS EPIPHYSIS
FEMUR
fig. 5. Bone marrow pres-
sure in diaphysis and epiphysis,
and respiration rate recorded
simultaneously in the dog. [From
Stein (103).]
Il
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
sure decreases (8, 103), whereas occlusion of the
arterial supply to the bone decreases both mean and
pulse pressures (8, 60, 103). Fracture of both sides of
the femur causes the intramedullary pressure to fall
to zero (8). Kaiser and his co-workers (74) demon-
strated a direct correlation between altitude and a
fall in marrow pressure in dogs, and experimentally
confirmed the fact that the marrow cavity acts as a
semiclosed cavity.
This conception is supported by the study of
Petrakis (95) who showed that the Valsalva maneuver
in man produces a rise in systemic venous pressure,
thus reducing the venous outflow from the marrow
cavity and causing an increase in marrow pressure
and a decrease in pulse pressure. Human marrow
pressure varies with respiration and body position.
Thus, sternal pressures were found to be near or
below atmospheric pressures in recumbent, nonleu-
kemic patients, and to vary with the respiration. The
more distal iliac sites did not respond to the respira-
tory effects of forced breathing, but required the more
strenuous effects of the Valsalva effort or of coughing
to evoke changes. The higher pressures obtained in
the iliac crest are presumably a result of die erect
position of man. The effects of changes in body posi-
tion and of alterations in respiration on marrow
pressures indicate that, under conditions of normal
activity, intramedullary pressure varies considerably
and is passively aflected by changes in venous pres-
sure resulting from these activities.
In this connection, it is interesting to note that in-
creasing the blood volume by the intravenous infu-
sion of large quantities of saline causes a gradual in-
crease of marrow pressure (8). On the other hand, a
decrease in blood volume produced by hemorrhage
causes a slow fall in medullary pressure (60).
Although marrow pressure does not ordinarily re-
flect changes in mean systemic arterial pressure (74),
decapitated cats and cats with acute spinal injury
have low femoral arterial pressures and do show low
marrow pressures (60). Chronic "spinal" cats, in
which the systemic pressure has returned to the levels
seen before the spinal cord was cut, had bone marrow
pressures similar to those found in unoperated animals.
Stimulation of the cut peripheral end of either vagus
nerve produced the usual slowing of the heart and fall
in svstemic pressure. This was associated with a re-
duction in bone marrow pressure. Changes in marrow
pressure induced by stimulation of the central ends
of the cut vagi or of the central end of the cut femoral
or sciatic nerve, or by making an incision in the ab-
dominal wall, were small and unpredictable. The
increase in systemic arterial pressure produced bv
occluding both carotid arteries was associated with a
rise in marrow pressure.
Stimulation of the cut peripheral end of the ab-
dominal sympathetic chain isolated from its connec-
tions with the spinal cord produced a fall in marrow
pressure within the femur. A similar reduction in
marrow pressure of the mandible occurs when the
peripheral end of the cut cervical sympathetic cord is
stimulated. The fall in marrow pressure caused by
excitation of sympathetic nerve fibers has been used
to trace the pathway by which such fibers reach
specific bones (1 12).
To determine whether the sympathetic innervation
of the marrow vessels is constantly influenced by tonic
impulses, the abdominal sympathetic chain of one
side was removed with strict aseptic precautions.
Several days after recovery from the operation, simul-
taneous measurement of the marrow pressures
showed no difference. However, when the experiment
was repeated using cats made decerebrate by ligating
both common carotid arteries and the basilar artery,
the pressure in the denervated femur was found to be
25 mm Hg higher than its innervated control (60).
Stimulation of the peripheral end of the cut splanch-
nic nerve produced the usual prolonged rise in femoral
arterial pressure, whereas the marrow pressure of the
femur was greatly reduced (fig. 6) and recovered
only as the systemic pressure returned to the control
value (60). The same phenomena can be reproduced
by the intravenous injection of Adrenalin (fig. 7)
(8, 60, 82). The Adrenalin effect can be reversed by
the prior injection of Hydergine (60). Other drugs
which produce an increase in systemic pressure and a
reduction in marrow pressure are norepinephrine
(60), Pituitrin (8, 82), Neo-Synephrine (8), Syneph-
rine (8), and tyramine (8).
A rise in systemic blood pressure with a simulta-
neous increase in marrow pressure is produced by the
fig. 6. Effect of stimulating the peripheral end of the cut
splanchnic nerve on the femoral arterial pressure (upper record)
and the marrow pressure (lower record) in the femur of the
cat. [From Herzig & Root (60).]
FLOW OF BLOOD THROUGH BONES AND JOINTS
[66l
mmHg
300 , vvrts,)..^
200
T.rt
i I
IBSSS&
■■■
■
ESSSEE1
ESEEEBSEE
■■■
■
iiiiii
iiiiili""
Be
a
liilli
ill !Ii>"i
■liiljl
iiiilH"!
!E"
I
Hl(i3r^ .™
■■■■■■■■■■
ill
1
-——a.. t+ikiit.^.,. ....
Mil
IIEEBiRBBI
■■■"
■
loOII
llm S
11
■
Ml
i
IkUHl
1151=5—-'
pi*
mm
p
•
eiiiil
«■«■■«!
H
■
iiilll
■■
■
U
mill
lllllilli
II
1
fig. 7. Effect of injecting in-
travenously 1 ml of 1 150,000
Adrenalin on the femoral ar-
terial pressure (upper record)
and the marrow pressure (lower
record) in the femur of the cat.
[From Herzig & Root (60).]
administration of ephedrine (82), Benzedrine (8, 60),
ergotamine tartrate (8) or nicotine (8). A reduction
in both systemic and marrow pressure follows the in-
jection of histamine (8, 82), acetylcholine (8), sodium
nitrate (8), and amyl nitrite (60).
Patients with leukemia and multiple myeloma dif-
fer from nonleukemic patients in having elevated
mean marrow pressures and increased pulse pressures
(95). In patients with acute leukemia in whom the
highest pressures are found, dicrotic notches are pres-
ent in the pulse waves suggesting a lowered pe-
ripheral resistance in the marrow circulation. The
degree of anemia could not be correlated with the
mean pressures, nor with the pulse pressures in the
marrow. The pressure data confirm the increased
vascularity in the marrow in some forms of leukemia
as demonstrated by the clearance of I131 from the
marrow (94).
Temperature of Bone Marrow
This subject is of some importance, for it is generally
believed that hematopoiesis requires the maintenance
of a high bone marrow temperature (33). According
to Huggins et al. (64), the more centrally placed
bones of the extremities, the cranial bones and the
sternum in the rat, rabbit, and pigeon have tempera-
tures similar to that of the peritoneal cavity, whereas
the temperature of the peripheral bone marrow of the
extremities may be lower by 4 to 8 C or more. In
adult man the red marrow is exclusively limited to
the bones of the body trunk and head as well as the
proximal portions of the limbs (92). Chemical ac-
tivity of the marrow does not affect the thermal con-
dition appreciably. On the other hand, the heat of
muscular activity of the limbs increases marrow
temperature.
Huggins & Blocksom (63) showed that an increase
in bone marrow temperature of the outlying bones
led to the replacement of yellow by red marrow. They
found a close correlation between the development of
cellular marrow and a temperature level similar to
that of the deep peritoneal cavity. However, they were
uncertain whether this is a primary effect upon tissue
metabolism, or a secondary vasomotor effect.
Petrakis (93), who studied the temperature in the
sternum, iliac crest, tibia, spinous process, and verte-
bral body of ten patients, found temperatures ranging
from that of the rectum in the vertebral body mar-
row, to 4 C below this in other bones. He interprets
this to mean a lack of precise temperature regulation
in hematopoietically active bone marrow. No corre-
lation was found between temperature and cell type.
JOINTS
Gardner's excellent review (50) should be con-
sulted for general information concerning the physi-
ology of movable joints.
The Blood Supply
According to Davies (34) little has been added to
the first description of the circulus vasculosus by
William Hunter (66) who in 1743 wrote: "All around
the neck of the bone there is a great number of
Arteries and Veins which ramify into smaller Branches
and communicate with one another by frequent
Anastomoses like those of the mesenterv. This might
be called the Circulus vasculosus, the vascular border of
the Joint."
At the articular margins, the capillaries form deli-
cate anastomosing loops comparable in pattern with
those seen in the mesentery. The blood supply of the
synovial membrane and capsule communicate freely
with the periosteal and epiphyseal supply; hence, the
1 66a
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
shaft of the bone forms one nutritional unit, and the
joint cavity and adjoining epiphysis form another.
For this reason, Harris (55) uses the term circulus
vasculosis arliculi et epiphyseos to emphasize the nutri-
tional interdependence of the joint and the epiphysis.
The venous drainage has received little attention.
According to Testut (104), Sappey described the
veins as characterized by frequent anastomoses,
tortuosities, and varicosities. Testut remarks on their
voluminous nature. Occasional valves are seen in the
large veins, even in the more superficial parts of the
synovial membranes.
The delicate nature of the synovial membrane and
its blood supply is indicated by the fact that small
extravasations of blood into the joint cavity are often
found in animals, and some extravasation follows
such a simple procedure as puncture of the joint (35).
The Nerve Supply
Chief among the features of the synovial membrane
is its sensitivity to pain. Localization is often not
highly accurate. To what degree the synovial mem-
brane responds to other sensations, such as tension or
pressure, is uncertain. Medullated and nonmedullated
nerves entering the joint with the blood vessels form a
plexus in the synovial membrane. The nonmedullated
fibers in large part innervate the blood vessels and are
probably of sympathetic origin. The effects of sympa-
thectomy on the vascular supply of joints remains
obscure, and in view of the paradoxical effects re-
corded by Engel (44) they need investigation. The
synovial membrane shows an abundance of free nerve
endings presumably subserving pain; end organs
possibly concerned with proprioceptive impulses are
variously described as Ruflini, Golgi, Mazzoni,
looped or knotted types. Pacinian corpuscles are not a
characteristic feature of the synovial membrane.
Gardner (49) failed to find them, and Davies (34)
confirms this.
Blond Flow Through Joints
Attempts have been made to determine the blood
flow through joints by measuring the intra-articular
temperature (61), by the application of the plethysmo-
graph to a knee segment (9), by using a bubble flow-
meter (25), and by means of an electromagnetic
flowmeter (26, 27).
In adult anesthetized dogs weighing 9.5 to 31 k",
the blood flow through the knee joint amounts to 1.5
to 7.0 ml per min (26). The temperature of the joint
must be markedly raised to obtain a measurable in-
crease in flow, changes of 10 C having little effect.
Even a high external temperature of 60 to 65 C in-
creases the blood flow only 1 5 to 57 per cent. Rapid
cooling of the joint with ice packs causes the flow to
decrease and remain fairly constant at half that of the
control. However, sometimes the flow falls, rises, and
then falls again; this is a type of behavior also de-
scribed for skin vessels exposed to low temperatures
(84). Removal of the ice packs is followed by a slow
return of blood flow to the control level, partly owing
to the delay in returning to the normal temperature.
Nevertheless, a delay in the return of blood flow also
occurs when the joint is quickly restored to the control
temperature.
According to Horvath & Hollander (61), joint
blood vessels dilate in response to cold and constrict
when exposed to heat. This finding is not supported
by the work of Hunter & Whillams (67) who used the
same technique. The latter found that joint tempera-
ture fell when their subject was exposed to cold, and
this they attributed to a reflex superficial vasodilata-
tion resulting in a short period of excessive heat loss.
On the other hand, Cobbold & Lewis (26) believe
the decrease in intra-articular temperature on expo-
sure to cold is the result of the constriction of joint
vessels. In their plethysmographic study of blood flow
through the knee segment, Bonney*/ a/. (9) found that
cooling the area resulted in a decrease in blood flow,
and heating produced the reverse effect. Since they
found a similar decrease in flow when the circulation
to the skin of the segment was suppressed by Adrenalin
iontophoresis, and further observed that after this
procedure cooling no longer decreased the blood flow,
Bonney et al. suggest that a different reaction to cool-
ing may occur in articular than does in superficial
vessels. The direct measurement of flow by Cobbold
& Lewis (26) does not support this view. These in-
vestigators believe that the results reported by Bonney
et al. are complicated by the presence of other tissue
such as muscle.
Nervous Control of Joint Blood J'essels
The innervation of joints has been fully reviewed
by Gardner (50). The blood vessels of the knee joint
receive vasoconstrictor fibers by way of the articular
nerves. Section of these increases blood flow some
50 per cent above the resting level (27). Stimulation
of the peripheral cut end of this nerve produces vaso-
constriction and a decreased blood flow. When the
carotid arteries were occluded below the carotid sinus,
the usual increase in systemic pressure was seen, but
FLOW OF BLOOD THROUGH BONES AND JOINTS
1663
no change in outflow from the joint was observed.
Repetition of carotid occlusion after sympathectomy
always produced an increase in flow. Lowering of
systemic pressure by hemorrhage results in a decrease
in joint blood flow which may cease if the blood loss
is great enough.
The vasoconstrictor action of Adrenalin, measured
as a decrease in blood flow (27), can also be produced
by means of Noradrenalin. Such effects are seen in
both innervated and denervated vessels. Acetylcholine
has a marked vasodilator action on innervated and
denervated vessels.
REFERENCES
1. Anseroff, N. J. Die Arterien der langen Knochen des
Menschen. Z. Anal. Entwick. lungsmech. 103: 793, 1934.
2. Axhausen G., and E. Bergmann. Die Ernahrungsunter-
brechungen am Knochen. In: Handbuch der speziellen
pathologischen Anatomie und Histologic, edited by F. Henke
and O. Lubarsch. Berlin: Springer, 1937, 9: Pt. 3, 118.
3. Bacq, Z. M. The action of abdominal sympathectomy
on the growth of the albino rat and the weight of the
genital organs. Am. J. Physiol. 95: 601, 1930.
4. Bailey, H. Impending death under anesthesia. Lancet 1 :
5. '947-
5. Berk, L., J. H. Burchenal, T. Wood, and W. B. Castle.
Oxygen saturation of sternal marrow blood with special
reference to pathogenesis of polycythemia vera. Proc. Soc.
Exptl. Biol. Med. 69: 316, 1948.
6. Bergmann, E. Theoretisches, Klinisches und Experi-
mentelles zur Frage der aseptischen Knochennekrosen.
Deul. Z. Chit. 206: 12, 1927.
7. Blair, H. C. The alteration of blood supply as a cause
for normal calcification of bone. Surg. Gynecol. Obstet. 67 :
413. '938-
8. Bloomenthal, E. D., W. II. Olson, and H. Necheles.
Studies on bone marrow cavity of the dog. Fat embolism
and marrow pressure. Surg. Gynecol. Obstet. 94: 215, 1952.
9. Bonnev, G. L. W., R. A. Hughes, and O. James. Blood
flow through the normal human knee segment. Clin. Sci.
i! : 167, 1952.
10. Bragdon, J. H., L. Foster, and M. C. Sosman. Experi-
mental infarction of bone and marrow. Am. J. Pathol. 25:
709. '949-
11. Branemark, P. I. Vital microscopy of bone marrow in
rabbit. Scand. J. Clin. & Lab. Invest. 11:5, 1959.
12. Brookes, M. Femoral growth after occlusion of the prin-
cipal nutrient canal in day old rabbits. J. Bone and Joint
Surg. 39B: 563, 1957.
13. Brookes, M. The vascular architecture of tubular bone
in the rat. Anat. Record 132: 25, 1958.
14. Brookes, M. The vascularization of long bones in the
human foetus. J. Anat. 92: 261, 1958.
15. Brookes, M. The vascular reaction of tubular bone to
ischaemia in peripheral occlusive vascular disease. J.
Bone and Joint Surg. 42B: 1 10, i960.
16. Brookes, M., and R. G. Harrison. The vascularization
of the rabbit femur and tibiofibula. J. Anat. 91 : 61 , 1957.
17. Brunschwig, A. Experimental infarction of bone mar-
row. Proc. Soc. Exptl. Biol. Med. 27: 1049, 1930.
18. Bunting, C. H. The regulation of the red blood cell sup-
ply. In : Contribution to Medical and Biological Research. New
York: Hoeber, 191 9, vol. 11, p. 824.
J3
24
ig. Burrows, H. J. Coxa plana with special reference to its
pathology and kinship. Brit. J. Surg. 29: 23, 1941.
20. Caeiro, J. C, and Y. H. Mainetti. La circulacion dia-
fisiaria en los huesos largos. Su importancia en la etiologia
de las endo-arterosis. Prenna Med. Argentina 18: 156, 1932.
21. Cannon, W. B., H. F. Newton, E. M. Bright, V.
Menkin, and R. M. Moore. Some aspects of the physi-
ology of animals surviving complete exclusion of sym-
pathetic nerve impulses. Am. J. Physiol. 89: 84, 1929.
22. Chandler, F. A. Local overgrowth. J. Am. Med. Assoc.
109: 1411, 1937.
Chandler, F. A. Observations on circulatory changes
in bone. Am. J. Roentgenol. 44: 90, 1940.
Chirav, M., L. Justin-Besancon, R. Benda, C. Debrav,
and M. Lacour. Influence des injections intramedullaires
osseuses sur la pression arterielle du chien. Ann. mid.,
Paris 46: 267, 1940.
25. Cobbold, A. F., and O. J. Lewis. Blood flow to the knee
joint of the dog. Effect of heating, cooling and Adrenalin.
J. Physiol., London 132: 379, 1956.
26. Cobbold, A. F., and O. J. Lewis. The nervous control
of joint blood vessels. J. Physiol., London 133: 467, 1956.
27. Cobbold, A. F., and O.J. Lewis. The action of Adrenalin,
noradrenalin and acetylcholine on blood flow through
joints. J. Physiol., London 133:472, 1956.
28. Cohen, L. Methods of investigating the vascular archi-
tecture of the mandible. J. Dental Research 38 : 920, 1 959.
29. Cohn, J., and W. H. Harris. The three dimensional
anatomy of the Haversian system. J. Bone and Joint Surg.
40.^:419, 1958.
30. Colp, R., and S. Magee. Experiences with periarterial
sympathectomy in fractures of the lower extremity. J.
Am. Med. Assoc. 97: 1069, 1931.
31. Coppo, M. Investigations on the mineral composition of
bones. Results, Conclusions. Arch, intern, pharmacodyn. 50:
3a8. 1935-
32. Corbin, K. B., and J. C. Hinsey. Influence of the nervous
system on bone and joints. Anat. Record 75: 307, 1939.
33. Cowdry, E. V. Textbook of Histology, 4th ed. Philadelphia:
Lea & Febiger, 1950, chapt. vi.
34. Davies, D. V. Synovial membrane and synovial fluid of
joints. Lancet 251 (2): 815, 1946.
35. Davies, D. V. Observations on the volume, viscosity and
nitrogen content of synovial fluid, with a note on the
histological appearance of the synovial membrane. J.
Anat. 78: 68, 1944.
36. Dempster, W. T., and D. H. Enlovv. Patterns of vascular
channels in the cortex of the human mandible. Anat.
Record 135: 189, 1959.
1664
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
37. Doan, C. A. The capillaries of the bone marrow of the
adult pigeon. Bull. Johns Hopkins Hosp. 33: 222, 1922.
38. Doan, C. A. The circulation of the bone marrow. In:
Contributions to Embryology, no. 67, 14: 27, 1922.
39. Drinker, C. K., and K. R. Drinker. A method for main-
taining an artificial circulation through the tibia of a dog
with a demonstration of the vasomotor control of the
marrow vessels. Am. J. Physiol. 40: 514, 191 6.
40. Drinker, C. K., K. R. Drinker, and C. C. Lund. The
circulation in the mammalian bone marrow. Am. J.
Physiol. 62 : 1, 1922.
41. Edholm, O. G., S. Howarth, and J. McMichael. Heart
failure and blood flow in osteitis deformans. Clin. Sci.
5: 249, 1945.
42. Elletto, L. Ricerche Topografiche e radiografiche sulla
circolazione arteriosa delle grandiossa lunghe degliarti,
nell' uomo. I. Arto superior. Arch. ital. anal. e. embriol. 31:
569. 1933-
43. Elston, J. T., R. V. Jaynes, D. H. Kaump, and W. A.
Erwin. Intraosseous infusions in infants. Am. J. Clin.
Pathol. 17, 143, 1947.
44. Engel, D. The influence of the sympathetic nervous sys-
tem on capillary permeability. J. Physiol., London 99:
161, 1 94 1.
45. Ferguson, A. B. Surgical stimulation of bone growth by
a new procedure. J. Am. Med. Assoc. 100: 26, 1933.
46. Ferguson, A. B. Some observations on the circulation
in foetal and infant spines. J. Bone and Joint Surg. 32A:
640, 1950.
47. Foa, P. P. Study on the innervation of the bone marrow.
I. Anatomy. Univ. Mich. Med. Bull. 9: 9, 1943.
48. Foa, P. P. Study on the innervation of the bone marrow.
II. Physiology. Univ. Mich. Med. Bull. 9:19, 1943.
49. Gardner, E. Nerve terminals associated with the knee
joint of the mouse. Anal. Record 83: 401, 1942.
50. Gardner, E. Physiology of the movable joints. Physiol.
Revs. 30: 127, 1950.
51. Grant W. C, and \V. S. Root. The relation of Oo in
bone marrow blood to post hemorrhagic erythropoiesis.
Am. J. Physiol. 150: 618, 1947.
52. Grieg, D. M. Clinical Observations on the Surgical Pathology
0/ Bone. Edinburgh: Oliser & Boyd, 1931.
53. Ham, A. W. Some histophysiological problems peculiar
to calcified tissue. J. Bone and Joint Surg. 34A: 701, 1952.
54. Ham, A. W. Histology (2nd ed.). Philadelphia: Lippincott.
'953-
55. Harris, H. A. Bone Growth in Health and Disease. London:
Oxford Univ. Press, 1933.
56. Harris, R. S., and D. M. Jones. The arterial supply to
the adult cervical vertebral bodies. J. Bone and Joint Surg.
386:922, 1956.
57. Harrison, R. G., and H. H. Grossman. The fate of
radiopaque media injected into the cancellous bone of
the extremities. J. Bone and Joint Surg. 37B: 150, 1955.
58. Haslhofer, L. Kreislaufstorungen des Knochens. In:
Handbuch der speziellen pathologischen Anatomic und Histologic,
edited by F. Henke and O. Lubarsch. Berlin: Springer,
'937, 9: Pt 3. P- 87-
59. Hecht, H. H., and A. J. Samuels. Observations on the
oxygen content of sternal bone marrow with reference to
polycythemic states. Federation Proc. 11 : 68, 1952.
60. Herzig, E., and W. S. Root. Relation of sympathetic
nervous system to blood pressure of bone marrow. Am. J.
Physiol. 196: 1053, 1959.
61. Horvath, S. M., and J. L. Hollander. Intra-articular
temperature as a measure of joint reaction. J. Clin. Invest.
28: 469, 1949.
62. Howe, W \V Jr., T. Lacev, and R. P. Schwartz. A
study of the gross anatomy of the arteries supplying the
proximal portion of the femur and acetabulum. J. Bone
and Joint Surg. 32A : 856, 1 950.
63. Huggins, C, and B. H. Blocksom, Jr. Changes in out-
lying bone marrow accompanying local increase of tem-
perature within physiological limits. J. Exptl. Med. 64:
^53. '936-
64. Huggins, C, B. H. Blocksom, and W. J. Noonan. Tem-
perature conditions in the bone marrow of rabbit, pigeon
and albino rat. Am. J. Physiol. 115: 395, 1936.
65. Huggins, C, and E. W'iege. The effect on bone marrow
of disruption of the nutrient artery and vein. Ann. Surg.
110:940, 1939.
66. Hunter, W. Of the structure and diseases of articulating
cartilages. Phil. Trans. 42: 514, 1743.
67. Hunter, J. and M. G. Whillams. A study of the effect of
cold on joint temperature and mobility. Can. J. Med. Sci.
29: *55. '951-
68. Hurley, L. A., and C. W. Miller. Demonstration of the
marrow vascular space (macrocanicular system) of bone;
technique for production of three dimensional plastic
anatomical models. A. M. A. Arch. Pathol. 68: 615, 1959.
69. Hurrell, D. J. The nerve supply of bone. J. Anal. 72:
54, '937-
70. Jaffe, H. L. The vessel canals in normal and pathological
bone. Am. J. Pathol. 5: 323, 1929.
71. Jaffe, H. L. The resorption of bone. Arch. Surg. 20: 355,
1930.
72. Johnson, R. W. Jr. A physiological study of the blood
supply of the diaphysis. J. Bone and Joint Surg. 9: 153,
I927-
73. Jones, H. B. Respiratory system: nitrogen elimination.
In: Medical Physics II: Chicago: Yr. Bk. Pub., 1950, p. 860.
74. Kalser, M. H., H. K. Ivy, L. Prevsner, J. P. Mar-
barger, and A. C. Ivy. Changes in bone marrow pres-
sure during exposure to simulated altitude. J. Aviation
Med. 22 : 286, 1 95 1 .
75. Kistler, G. H. Formation of bone by periosteum after
experimental infarction by embolism of femur in rabbits.
Proc. Soc. Exptl. Biol. Med. 31 : 218, 1934.
76. Kistler, G. H. Sequences of experimental bacterial in-
farction of femur in rabbits. Surg. Gynecol. Obstet. 60: 913,
'935-
77. Kistler, G. H. Effect of circulatory disturbances on the
structure and healing of bone. Injuries of the head of the
femur in young rabbits. Arch. Surg. 33: 225, 1936.
78. Kuntz, A., and C. Richins. Innervation of the bone mar-
row. J. Comp. Neurol. 83: 213, 1945.
79. Lacroix, P. Organization of Bones. London: Churchill,
■95i
80. Lamas, A., D. Amado, and C. Da Costa. La circulation
du sang dans l'os. Presse med. 54: 862, 1946.
81. Langer, K. Ueber die Blutgefasse der Knochen des
Schaedeldaches und der harten Hirnhaut. Denkschr. Kgl.
Akad. H'iss., Math, naturwiss. Kl., rVien 37: 217, 1877.
82. Larsen, R. M. Intramedullary pressure with particular
FLOW OF BLOOD THROUGH BONES AND JOINTS
1665
reference to massive diaphyseal bone necrosis. Ann. Surg.
108: 127, 1938.
83. Leriche, R., and A. Policard. Les prob/emes de la physi-
ologic Normale ft Pathologique de Vos. Paris: Masson, 1926.
English translation by S. Moore and J. A. Key. St.
Louis: Mosby, 1928.
84. Lewis, T. Observations upon the reactions of the vessels
of the human skin to cold. Heart 15: 177, 1929.
85. Lewis, O. J. The blood supply of developing long bones
with special reference to metaphyses. J. Bone and Joint
Surg. 38B: 928, 1956.
86. Lexer, E. Weitere Untersuchungen iiber Knochenar-
terien und Bedeutung fur krankhafte Vorgange. Arch.
klin. Chir. 73: 481, 1904.
87. Lorimer, A. A. de, M. L. Minear, and H. B. Boyd.
Reflex hyperemia deossification regional to joints of the
extremities. Radiology 46: 227, 1946.
88. Marneffe, R. de. Recherches morphologiques et ex-
perimentales sur la vascularization osseuse. Acta chir. belg.
50: 469, 568, 681, 1 95 1.
89. McMaster, P. E., and N. W. Roome. The effect of
sympathectomy and of venous stasis on bone repair. J.
Bone and Joint Surg. 16: 365, 1934.
90. Miles, J. S. The use of intramedullary pressures in the
early determination of aseptic necrosis in the femoral
head. J. Bone and Joint Surg. 37 A : 622, 1955.
gi. Miller, D. S., and G. de Takats. Post traumatic dys-
trophy of the extremities : Sudeck's atrophy. Surg. Gynecol.
Obstet. 75: 558, 1942.
92. Neumann, E. Das Gesetz der Verbreitung des gelben und
roten Markes in den Extremitatenknochen. Centr. med.
Wiss. 20: 321, 1882.
93. Petrakis, N. L. Temperature of human bone marrow.
J. Appl. Physiol. 4: 549, 1952.
94. Petrakis, N. L., S. P. Masowedis, and P. Miller. The
local blood flow in human bone marrow in leukemia and
neoplastic diseases as determined by the clearance rate
of radio-iodide (I131). J. Clin. Invest. 32: 952, 1953.
95. Petrakis, N. L. Bone marrow pressures in leukemic and
non-leukemic patients. J. Clin. Invest. 33: 27, 1954.
96. Phemister, D. B. Changes in bones and joints resulting
from interruption of circulation. I. General considerations
and changes resulting from injuries. Arch. Surg. 41 : 436,
1940.
97. Phemister, D. B. Changes in bones and joints resulting
from interruption of circulation. II. Non-traumatic lesions
in adults with bone infarction ; arthritis deformans. Arch.
Surg. 41: 1455, 1940.
98. Rindfleisch, G. E. Ueber Knochenmark und Blut-
bildung. Arch, nukroskop. Anal. u. Entwicklungsmach. 17:1,
21, 1880.
99. Rogers, W. M., and H. Gladstone. Vascular foramina
and arterial supply of the distal end of the femur. J. Bone
and Joint Surg. 32A: 867, 1950.
100. Schleicher, E. M. On the "conical openings" in the
wall of venous sinusoids and their relation to the so-called
erythrogenic capillaries in the bone marrow of man.
Anat. Record 95: 379, 1946.
101. Schnall, M. D., and R. J. Heffernan. Intrasternal in-
fusions in obstetrical hemorrhage. Am. J. Surg. 68: 44,
'945-
102. Schwartz, B. M., and D. Stats. Oxygen saturation of
sternal marrow blood in polycythemia vera. J. Clin.
Invest. 28: 736, 1949.
103. Stein, A. H., H. C. Morgan, and F. C. Reynolds. Varia-
tions in normal bone marrow pressures. J. Bone and Joint
Surg. 39A: 1 1 29, 1957.
104. Testut, L. Vaisseaux et nerfs des tissues conjonctifs fibreux,
sereux, et osseux. Anatomie et physiologic. (These d'agregation.)
Paris : Masson, 1 880.
105. Tocantins, L. M. Rapid absorption of substances in-
jected into bone marrow. Proc. Soc. Exptl. Biol. Med. 45:
292. '94°-
106. Tocantins, L. M., and J. F. O'Neill. Infusion of blood
and other fluids into the circulation via the bone marrow.
Proc. Soc. Exptl. Biol. Med. 45: 782, 1940.
107. Trueta, T., and M. A. M. Harrison. The normal vas-
cular anatomy of the femoral head in adult man. J. Bone
and Joint Surg. 35B: 442, 1953.
108. Tucker, F. R. Arterial supply to the femoral head and
its clinical importance. J. Bone and Joint Surg. 31B: 82,
■949-
109. Vereby, K. Die Blutversorgung des Femurkopfes. Anat.
Anz. 93: 225, 1942.
1 10. Wagner, G., and E. P. Pendergrass. Intrinsic circula-
tion of the vertebral body. Am. J. Roentgenol. 27: 818, 1932.
111. Weinmann, J. P., and H. Sicher. Bone and Bones. Funda-
mentals of Bone Biology. London: Kimpton, 1947.
112. Weiss, R., and W. S. Root. Innervation of the vessels
of the marrow cavity of certain bones. Am. J. Physiol. 197:
1255. '959-
113. Willis, T. A. Nutrient arteries of the vertebral bodies.
J. Bone and Joint Surg. 31 A: 538, 1949.
114. Wolcott, W. E. The evolution of the circulation in the
developing femoral head and neck. Surg. Gynecol. Obstet.
77:61, 1943.
1 1 5. Wollenberg, G. A. Die aseptische Knochennecrose und
ihre Bedeutung fur die Knochen- und Gelenkchirurgie.
Acta Chir. Scand. 60: 369, 1926.
CHAPTER 48
Dynamics of the pulmonary circulation
ALFRED P. FISHMAN
Department of Medicine, Columbia University,
College of Physicians and Surgeons, New York City
CHAPTER CONTENTS
The Growth of Ideas
Bridge Between Two Ventricles
Role in External Respiration
Analysis of Pulmonary Hemodynamics
Alveolar-Capillary Gas Exchange
Comparative Physiology
Experimental Animals and Test Preparations
Functional Anatomy
Blood Vessels
Overlap of distensibility and resistance characteristics
Large pulmonary vessels
Small muscular pulmonary vessels
Capillaries
Extravascular Smooth Muscle
Systemic Blood Supply of the Lung
Venous Admixture
Pulmonary Vasomotor Nerves
Pulmonary Blood Flow
Normal Values
Uneven Pulmonary Blood Flow
Pulmonary Vascular Pressures
Recording
Hydrostatic Reference Level
Pulmonary Arterial Pressure
Pulmonary Venous and Left Atrial Pressures
Pulmonary Arteriovenous Pressure Gradient
Pulmonary Wedge Pressures
Influence of Intrathoracic Pressure on Pulmonary Vascular
Pressure
Transmural Versus Luminal Pressures
Pulmonary Blood Volume
Measurement of Pulmonary Blood Volume
Stewart-Hamilton : indicator dilution
Newman: exponential downslope
Bradley: equilibration curves
Changes in Pulmonary Blood Volume
Lung volumes
Mechanics of breathing
Radioactive tracers
Miscellaneous
Normal Values for Pulmonary Blood Volume
Variations in Pulmonary Blood Volume
Partition of Pulmonary Blood Volume
Hemodynamic Interrelations
Distensibility and Resistance
Distensibility arid Capacity : Pressure -Volume Relationships
Resistance: Pressure-Flow Relationships
Meaning of Pulmonary Vascular Resistance
Practical Recognition of Pulmonary Vasomotricity
Blood Flow Through Each Lung Separately
Critical Closure
Potential and Kinetic Energy
Pulmonary Capillary Circulation
Pulmonary Capillary Pressure
Rate of Pulmonary Capillary Blood Flow-
Nature of Pulmonary Capillary Blood Flow
Size of Pulmonary Capillary Bed
Pulmonary Capillary Blood Volume
Resistance and Distensibility
Time Spent by Blood in Pulmonary Capillaries
Pulmonary Capillary I lematocrit
Transcapillary Exchange
Miscellaneous Hemodynamic Phenomena
Pulmonary Arterial Pulse-Wave Velocity
Pulmonary Circulation Time
Influence of Respiration on Pulmonary Circulation
Spontaneous Breathing
Inllation of the Lungs
Positive Pressure Breathing
Negative Pressure Breathing (Pleural)
Negative Pressure Breathing (Intrapulmonary)
Cough
Prolonged Expiration
Forced Expiration (Valsalva)
Occlusion of a Pulmonary Artery
Effects of Exercise on Pulmonary Circulation
Pulmonary Blood Flow
Blood Flow and Oxygen Uptake
Arteriovenous Oxygen Difference
Pulmonary Vascular Pressures
Pulmonary Blood Volume
Pulmonary Vascular Resistance
Miscellaneous Mechanical Influences
Heart Rate
1667
1 668
HANDEOOK OF PHYSIOLOGY
CIRCULATION II
"Bronchomotor Tone"
Mechanical Compression (Atelectasis)
Hypertonic Solutions
Pulmonary Vasomotor Activity
Respiratory Gases
Acute Hypoxia
Chronic Hypoxia
Acute Hyperoxia
Acute Hypercapnia
Acute Acidosis
Alveolar Hypoventilation
Pulmonary Vasomotor Reflexes
Pulmonary Vasomotor Waves
Effect of Drugs
Predominantly Passive Effects
Pulmonary Vasoconstrictors
Pulmonary Vasodilators
Cardiopulmonary Disorders
Pulmonary Arterial Hypertension
Restricted vascular bed
Increase in pulmonary blood flow
Increase in pulmonary venous pressure
Pulmonary arterial vasoconstriction
Cor Pulmonale
Pulmonary Edema
Pulmonary Hypotension
Pulmonary Arteriovenous Fistula
Pulmonic Stenosis
Pulmonic Valvular Insufficiency
(242). By this arrangement, the lung is equipped to
operate efficiently over a wide range of metabolic ac-
tivities: the enormous expanse of alveolar-capillary
surface is capable of increasing during activity (346)
and the geometric distribution of airways and blood
vessels favors the continued balance of alveolar venti-
lation and pulmonary capillary perfusion even during
strenuous exertion (7). Finally, governing the coordi-
nated performance of this respiratory apparatus is a
complicated system of ventilatory and circulatory
controls; these succeed, despite the phasic and
asynchronous nature of the ventilation and circula-
tion, in stabilizing the gaseous composition of the
alveolar gas and in ensuring adequate perfusion of
the gas-exchanging surfaces.
In addition to participating in external respiration,
the pulmonary circulation also performs several me-
chanical functions as a consequence of its architecture
and location. Thus, as the bridge between the two
sides of the heart, it is in a position to serve as a
reservoir of blood for the left ventricle and to control
left ventricular output by varying the pulmonary
venous return (183). .Similarly, as a consequence of
their position at the outlet of the right ventricle, the
smaller pulmonary vessels constitute a filter for sys-
temic venous particles of all kinds, including the
normal formed elements of the blood (1).
the pulmonary circulation is part of an elaborate
tonometric system for external respiration; it exists
for the perfusion of the lungs rather than for their nu-
trition. And, as a consequence of its anatomical dispo-
sition with respect to the pulmonary airways and air
spaces, the lung operates as a respiratory organ
rather than as an air sac.
The lung appeared in the vertebrate phylum. In
principle, it is designed to bring a thin stream of
venous blood into gaseous equilibrium with a large
volume of alveolar gas; however, in construction and
in efficiency, it varies from class to class. Thus, in the
amphibia — vertebrates which bridge the gap between
the water and the land — the lung resembles a large
bulla (fig. 1 ) ; this inefficient construction apparently
suffices for the low metabolic requirements of the
amphibia for oxygen. At the other extreme is the
complex lung of the large and vigorous terrestrial
vertebrates; in such a lung, septation and alveolation
have created a porous structure, composed mainly of
myriads of microscopic air spaces; suspended in the
walls of these tiny air spaces are the pulmonary capil-
laries to which the pulmonary arterial tree delivers
the entire right ventricular output for arterialization
fig. 1. Alveolar structure of the frog lung. Each lung con-
sists of a large central cavity surrounded by numerous small
chambers of varying size. The alveolar walls are outlined by
the vessels which they contain. (Prepared in collaboration with
H. O. Heinemann.)
DYNAMICS OF PULMONARY CIRCULATION
.669
fa
r
h
fig. 2. Structure of the heart according to Leonardo da Vinci
(98). The diagram shows the (nonexistent) pores in the ven-
tricular septum, an essential component of the Galenical con-
cept of the motion of the blood.
In recent years, the biological role of the pulmo-
nary circulation and lungs has been emphasized
(294). For example, the pulmonary vascular endo-
thelium seems to contribute enzymes, such as lipo-
protein lipases, to the perfusing blood (1). Mast cells,
which are abundant in the lungs of many species, are
believed to add a wide variety of substances, including
heparin, histamine, hyaluronic acid, and serotonin
(230). Finally, the walls of the pulmonary blood ves-
sels and pulmonary tissue may also neutralize certain
endogenous substances (e.g., serotonin) which could
exert noxious effects if they gained access to the left
heart and systemic circulation (158).
THE GROWTH OF IDEAS
The large pulmonary vessels were known to
Herophilus of Alexandria in the fourth century, B.C.
(228). But not until the time of Harvey (1578-1667)
did dispassionate evidence begin to establish the
structure and function of the pulmonary circulation
(90, 91,1 79). For convenience, the origins and growth
of the modern ideas will be sketched under four sepa-
rate headings: a) the appreciation and proof that the
pulmonary vascular tree constitutes a closed circuit
between the right and left hearts, b) the awareness
that the lungs are concerned with external respira-
tion, <) the systematic analysis of pulmonary hemo-
dynamics, and d) the coordinated description of
alveolar-capillary gas exchange. While such a presen-
tation of the growth of ideas has the advantage of
brevity, its lack of historical detail ignores foretellers
such as Ibn Xafis (291), Servetus (91), and Mayow
(144), whose clairvoyance could only be appreciated
retrospectively; it also exaggerates the contributions
of the "'finishers," whose discoveries crowned the
concepts and efforts of others.
Bridge Between Two Ventricles
In Harvey's time, Galenical misconception (fig. 2)
and philosophic speculation still predominated. Al-
though some of Harvey's predecessors and contempo-
raries had realized that there were no ventricular
pores by which right ventricular blood could bypass
the lungs and that the pulmonary artery was too
large to serve only as a nutrient vessel, their preoccu-
pation with the idea of the vascular system as the
generator of essential spirits blinded them to the
motion of the blood in the vessels. For Harvey, the
idea of the pulmonary circulation as a bridge between
the two ventricles was an essential component of his
theory of the unidirectional circulation of the blood;
he verified his theory by direct experiment and pro-
posed "porosities" as the final links between the
arteries and veins (195). In 1661, a few years after
Harvey's death, Malpighi provided the final proof of
pulmonary vascular continuity by visualizing the
passage of blood from the pulmonary arteries to the
veins by way of the pulmonary capillaries (280).
Role in External Respiration
Harvey was concerned solely with the mechanical
aspects of the circulation of the blood. His concept
did not deal with the prevalent notions about the
role of the lungs either as a source of material for the
generation of the vital spirit or as a refrigerating
device to control the innate heat of the heart (166).
Two years after Harvey's death, Lower (fig. 3), using
Hooke's new respiration pump, showed that blood
became arterialized as it passed from the right to the
left side of the heart (172). The idea that ingredients
of air, rather than air itself, were the basis of external
respiration had to await new developments in chemis-
try and in physics. A century was to pass before: a)
Black ( 1 788—1867), Lavoisier (1743-1794), and
Rutherford (1 753-1814) identified the three respira-
tory gases; b) Lavoisier proved that oxygen rather
than air was essential for life; and c) Lavoisier and
Laplace likened respiration to combustion (90, 91,
144). Indeed, not until the mid-nineteenth century
was it appreciated that combustion occurred in the
tissues rather than in the lungs and that hemoglobin
167O HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
fig. 3. Richard Lower (163 1-1691). In his book, Tractatus
De Corde, he described his experiments with Hooke's respiration
pump. These experiments proved that venous blcod becomes
arterialized in traversing the lungs and that bleed absorbs a
vital chemical substance from the air (172).
was involved in the transport of oxygen from the lungs
to the tissues (90a, 144).
Analysis of Pulmonary Hemodynamics
Although certain aspects of the regulation of the
pulmonary circulation — such as the influence of the
respiration — were under experimental scrutiny by the
middle of the eighteenth century (180), the system-
atic study of pulmonary hemodynamics could not
begin without practical methods for measuring pul-
monary vascular blood pressures and flow (189, 433).
These became available about a century later: pul-
monary arterial pressures were first measured in the
laboratory of Carl Ludwig (fig. 4) in the i85o's, by
using the recording mercury manometer in open-chest
dogs (27); shortly thereafter, more elaborate measure-
ments were made in the intact horse (90a, 91). In
1870, A. Fick (fig. 4) pointed out how measurements
of respiratory gas exchange could be used to calculate
the volume rate of pulmonary blood flow (130). But,
without ready access to mixed venous blood, the
direct Fick principle offered little promise of becoming
a popular method for measuring the pulmonary blood
flow in either the intact animal or man.
fig. 4. Pioneers in hemodynamic measurements. Left: Carl Ludwig (1816-18951 introduced
graphic recording of blood pressure in 1847. Right: Adolph Fick (1829-1901), student of Ludwig,
in 1870 described the use of respiratory gas exchange for the measurement of cardiac output in intact
animal or man. [After Rothschuh (361).]
DYNAMICS OF PULMONARY CIRCULATION
1671
During the next seventy years, a wide variety of in-
genious experimental preparations and new tech-
niques were used to gain information about the remote
pulmonary circulation: a) artificial experimental
conditions were devised to control some respiratory
and circulatory parameters so that others could be
measured (40, 1 57) ; b) high-fidelity manometric sys-
tems were invented and used to register the details of
the pulmonary vascular pressure pulses (183); c)
cannulae were placed during open thoracotomy so
that pulmonary arteriovenous pressure gradients
could be measured in the closed-chest animal imme-
diately after operation (225) ; d) angiostomy cannulae
were devised so that pulmonary vascular blood pres-
sures could be recorded at will in intact, unanesthe-
tized animals (183, 187); and e) indirect methods
were developed for the estimation of the pulmonary
blood flow in intact animals or man (see Chapter 1 7).
This three-quarters of a century of steady progress
underwent sudden acceleration in the logo's. In 1929,
Forssmann demonstrated on himself that a catheter
could be safely threaded by way of a peripheral vein
into the right heart (142); shortly thereafter, Klein
measured the pulmonary blood flow by the direct
Fick principle in man (90a). By World War II, the
stage was set for Cournand, Richards, and their co-
workers to begin their systematic studies of the pul-
monary circulation in man under natural conditions
(92). And, since the ig4o's, right heart catheterization
has been used for the sampling of mixed venous blood,
for the injection of contrast material and test sub-
stances into the pulmonary circulation, and for the
recording of blood pressures from the right side of
the heart; the technique has also provided access to
the venous effluent from special organs and regions
of the body and has led to the techniques of left heart
catheterization.
Alveolar-Capillary Gas Exchange
As indicated previously, the pulmonary circulation
is predominantly built for alveolar-capillary gas ex-
change. Up to the turn of the present century, the
precise nature of alveolar-capillary gas exchange was
unclear; particularly uncertain was the mechanism
by which oxygen traversed the alveolar-capillary
interfaces: some held that oxygen was secreted by the
alveoli (177); others maintained that diffusion alone
was involved (10). The issue was finally settled in
favor of diffusion by August and Marie Krogh (243).
These studies by the Kroghs also paved the way for
measuring the rate of pulmonary capillary blood flow
using soluble, inert gases as tracer substances (343).
To complete the picture of the coordinated circula-
tory-respiratory mechanism for external gas exchange
(fig. 5), more had to be learned of the physiological
behavior and of the physicochemical properties of the
blood. To this end, Barcroft (fig. 6) provided pre-
cise experimental information concerning the dis-
sociation of oxyhemoglobin (9); L. J. Henderson
(fig. 6) and his collaborators analyzed blood as a
physicochemical system and defined its role in the
exchange of the respiratory gases between the
atmosphere and the tissues (201, 297).
The regulation of alveolar-capillary gas exchange
came under serious experimental scrutiny in the
1940's. In 1946, Euler & Liljestrand (125) proposed
that the local concentration of the respiratory gases
within the lung — a function of local ventilation-
perfusion relationships — might regulate, in turn, the
distribution of the pulmonary blood flow; the many
experiments subsequently performed by others to test
this hypothesis (132) will be considered later in this
chapter. Interest in alveolar-capillary gas exchange
was also stimulated in the 1 94o's from another direc-
tion, i.e., from the practical exigencies of aviation
medicine in World War II; from this practical interest
developed theoretical models, quantitative formula-
tions, and graphic representations which have not
only helped to resolve old problems in alveolar-
capillary gas exchange but also to point up new ones
(267, 327, 345).
COMPARATIVE PHYSIOLOGY
There are exceedingly few hemodynamic measure-
ments in the nonmammalian vertebrates. In the
fishes and sharks, the mean blood pressure in the
ventral aorta (to the gills) is of the order of 30 mm Hg;
as blood traverses the gills, blood pressure drops to
reach a slightly lower level in the dorsal (systemic)
aorta (57).
This arrangement of the circulation in the fishes is
unfavorable for the systemic circulation. The hemo-
dynamic situation of the systemic circulation begins
to improve in the Amphibia and Reptilia in which the
pulmonary artery is separate from the aorta and over-
rides a functionally single ventricle. Among the
Reptilia, pulmonary arterial blood pressures have
been measured in the turtle (352) and in the snake
(222). In both of these species, the patterns of blood
1672
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
BLOOD FLOW X HbOg CAPACITY
DIFFUSING CAPACITY
fig. 5. Nomogram illustrating the respiratory and circulatory interplay involved in delivering
an adequate supply of oxygen to the arterial blood. The dashed-line rectangle represents the situa-
tion at rest; the solid-line rectangle represents the situation during exercise. At rest, this subject has
an oxygen requirement (VuJ of 250 nil/min. Starting in the left upper quadrant, and moving from
quadrant to quadrant in a counterclockwise direction, Vo, is shown to be met by a cardiac output of
5 liters/min, an oxygen capacity of 20 ml/ 100 ml, a Hb flow/capillary area of 5 and a diffusing
capacity of 20 ml/min/mm Hg. Also pictured are the corresponding arteriovenous differences in
oxygen content and saturation of blood (A-V AO?), as well as the mean alveolar-capillary diffusing
gradient for oxygen (mean A alveolar-capillary Po2). During exercise, as the oxygen requirement
increases (Vo« = 1250 ml/min), these variables undergo appropriate change. A similar nomogram
could be constructed for oxygen uptake in the tissues. [Based on Barcroft (9) and Lilienthal el al.
(266).]
■- NORMAL
— EXERCISE
pressure are qualitatively similar: the systolic pres-
sures in the pulmonary and systemic circulations are
identical; the pulmonary arterial diastolic pressure is
lower than the systemic arterial diastolic pressure, due
to the presence of a spiral valve between the systemic
artery and the ventricle (352). In the turtle, the pul-
monary arterial pressure is of the order of 35/12
(352). When systemic vascular resistance increases,
blood is diverted through the ventricular septal
defect to the pulmonary circulation (443).
The ventricular septum is complete and the two
circulations are entirely separate in the homeothermic
mammals and birds. Among the mammals and birds,
pulmonary arterial pressures have been measured in
a wide variety of species including man, dog, cat,
guinea pig (114), chicken (352), and calf (199). In
the chicken and calf, the pulmonary arterial pressure
is generally of the same order of magnitude as in the
clog, cat, and man, i.e., of the order of 20 to 30 mm
Hg systolic and 10 to 12 mm Hg diastolic; in the
DYNAMICS OF PULMONARY CIRCULATION
'673
J^iA*tZ&u
fig. 6. Joseph Barcroft (1872-1947) (left) and Lawrence J.
Henderson (1878-194.;) (right) photographed in September
1936. (Courtesy of D. B. Dill.)
guinea pig, it is somewhat lower ( 1 14), whereas in the
pig, horse, cow, and steer it is often considerably
higher (112, 199J.
The pulmonary blood flow has also been measured
in various conventional and unconventional labora-
tory animals including the goat (10), the horse (446),
and the cow (112, 199). Although, in general, the
larger species have the larger pulmonary blood flows,
there is no consistent interspecies relationship be-
tween pulmonary blood flow on the one hand and
either body surface area or weight on the other (22).
Whether the disparities represent real biological dif-
ferences, or the inadequacies of weight and body sur-
face area as standards of reference, or artifacts arising
from trying experimental situations, remains to be
decided.
EXPERIMENTAL ANIMALS AND TEST PREPARATIONS
The normal pulmonary circulation has only been
studied in a few dogs and humans. Instead, most of the
observations have been made on the pulmonary cir-
culations of anesthetized animals, of artificial prep-
arations, and of patients with heart and lung disease.
Each of these three categories is a major deviation
from normal : the use of anesthetized animals succeeds
admirably in excluding the elements of anxiety and
cooperation; but its substitutes, instead, blunted
vasomotor responses and changing levels of metabo-
lism, respiration, and circulation (257). Artificial
preparations, such as isolated vascular rings or iso-
lated lungs, certainly allow remarkable control of
mechanical parameters and may uncover influences
which are obscured in the intact organism; but, by
severing nervous connections, by failing to pass pul-
monary blood through other vital organs, and by
depending on abnormal perfusates, impaired nutrient
circulations, deteriorating heart and lungs, and ab-
normal gas exchange, they may introduce not only
discernible — but also hidden — artifacts (95, 141, 183).
Finally, while the study of patients with heart and
lung disease may be revealing to physicians who are
attempting to gain insights into the mechanism of
heart strain and failure, the results from these "ex-
periments of nature" can rarely be used to predict
the behavior of the normal pulmonary circulation,
since both heart and lung disease tend to exaggerate
the influence of mechanical factors and to alter the
structure of the pulmonary blood vessels.
"Species difference" is a standard apology for
atypical responses of the pulmonary circulation to
diverse stimuli (95). Occasionally, the basis for this
excuse is a distinctive morphological characteristic
(72, 171, 199). For example, the small muscular pul-
monary arteries of the rabbit contain much thicker
media than do the corresponding vessels of rat, cat,
and man (fig. 7). It is easy to imagine that contraction
of such hypertrophied muscle could evoke the
"gnarly" distortion of the rabbit's vascular tree which
follows the infusion of large quantities of norepineph-
rine (61); on the other hand, it is somewhat more
difficult to imagine such an intense vascular response
in species with muscle-poor precapillary vessels.
Species difference may reside in physiological as
well as in anatomical peculiarities (218). For exam-
ple, the rabbit is notoriously vagotonic, whereas the
cat, dog, and man are generally regarded as sympa-
thotonic; moreover, the pulmonary circulation of the
rabbit also appears to be more susceptible to the
effects of pharmacological agents, such as histamine,
than is the pulmonary circulation of the dog (442).
Such anatomical and physiological peculiarities occur
throughout the animal kingdom, complicating the
transfer of information from one species to another
(95. 239> 352)-
FUNCTIONAL ANATOMY
The dependent position of the pulmonary circula-
tion— within the lung and thorax on the one hand,
1674
HANDBOOK (II- PHYSIOLOGY
CIRCULATION II
60
^
^
M
v
v
53
V^v PULMONARY
49
xx ARTERY
•V
44
\
~~^ \
«39
Q
Id
5 35
\ \-6UINEA PIG
fe30
\ \
\ \
£26
UJ
z 22
RABBIT-\ ^.
v COW, PIG v \
vx_Z 1 \
*•<«. MAN ^«0 \
o
V ^>/ V>N V
I l7
x NcatN \ \\
13
8
4
"** — — — — —
M
1 1 1 1 1
PULMONARY
VEIN
220
176
132
88
44 22
DIAMETER
220 176
OF VESSELS
fig. 7. Relationship between vascular calibers and medial thickness in different species. Both the
small pulmonary arteries and the small pulmonary veins are well developed in the guinea pig, cat,
calf, and pig. On the other hand, in man and in the rabbit, only the arterial muscle, and in the rat
only the venous muscle, is well developed. [Redrawn after Takino (392).]
and between the two ventricles on the other — subjects
it to a variety of mechanical influences. Consequently,
the following appraisal of the functional anatomy of
the pulmonary circulation will take into account not
only those features of the vascular tree which deter-
mine pulmonary vascular distensibility and resistance
to perfusion but also the extravascular structures
which may, under appropriate conditions, modify or
obscure the natural properties of the pulmonary
vessels (288).
Unless expressly indicated, the anatomical descrip-
tions which follow derive largely from the examination
of the lungs of normal man at sea level. It is likely
that, in most respects, the generalizations about struc-
ture, and particularly about the relationships between
structure and function, apply almost as well to the
cat and to the dog. However, much more has to be
learned before the generalizations from normal man
at sea level can be applied directly either to other
test animals, such as the rabbit and the cow, or to
normal native residents at high altitudes, or to sea
level residents with abnormal pulmonary vessels or
parenchyma (99, 392).
Blood Vessels
OVERLAP OF DISTENSIBILITY AND RESISTANCE CHARAC-
TERISTICS. In the systemic circulation, the term
"arteriole" is synonomous with "resistance" vessel.
Characteristically, the systemic arteriole has a heavy
coat of circular smooth muscle and a high ratio of
wall thickness to lumen diameter. With respect to
size, ''systemic arteriole" generally refers to vessels of
300 to 400 fi or less, depending on the organ in which
they are found (50). On the other hand, in the low-
pressure pulmonary circulation, the anatomical
counterpart of the systemic arteriole does not exist.
This lack of sphincteric precapillary vessels has several
implications: a) that other small vessels may con-
tribute appreciably to the pressure drop between the
pulmonary artery and veins; b) that the small pul-
monary vessels may also serve as storage vessels,
changing caliber passively with the pulmonary blood
volume; and c) that under appropriate conditions,
each of the small vascular segments may constitute
the dominant pulmonary vascular resistance to
blood flow.
large pulmonary vessels. The pulmonary artery
rapidly subdivides into terminal branches which
have thinner walls and wider bores than the cor-
responding branches of the systemic arterial tree. The
media of the main pulmonary artery is about half
as thick as that of the aorta; the elastic fibers are
short and far less orderly than in the aorta (198).
The smooth muscle appears to insert on the elastic
DYNAMICS OF PULMONARY CIRCULATION
■675
fibers, suggesting an arrangement capable of con-
trolling either the pressure of the wall on its contents
or the volume of blood contained within the large
vessel (198). In general, the structure of the large
vessels seems better suited for varying their distensi-
bility than their geometry; nonetheless, the possibility
exists that constriction of large vessels may also effect
pulmonary vascular resistance to blood flow (126).
In the normal human lung the pulmonary arteries
are end-arteries, continuing without branching to the
level of the first alveoli in the walls of the respiratory
bronchioles (292, 421). Unfortunately, too little is
known of the pattern of branching to serve as a re-
liable basis for predicting the distribution of resistance
along the length of the pulmonary vascular tree
(140, 169, 205). The arterial portion of the pulmonary
circulation lies adjacent to the bronchial tree; indeed,
in the region of the respiratory bronchiole, arterial
bifurcations straddle the airway (117). Consequently,
the arterial branches are more susceptible to passive
distortion by the conducting airways than are the
venous branches which are situated at the periphery
of the lobule.
The pulmonary veins are end-veins (421). Their
musculo-elastic components are more irregularly
dispersed than those of the corresponding pulmonary
arteries; their media contain more collagenous fibers.
At the entry of the veins into the left atrium, exten-
sions of cardiac muscle become incorporated into
the venous walls. The suggestion has been made that
under certain experimental conditions, these muscular
extensions may act as "throttles" (56, 121, 392, 394).
small muscular pulmonary vessels. From the point
of view of vasomotor activity, three concepts are
generally held: /) vascular smooth muscle is pre-
requisite for active change in caliber; 2) during a
change in vasomotor tone, the small, muscle-con-
taining vessels are the site of changed resistance; and
j) the thicker the media, the more apt is the vessel to
constrict, the less apt is it to undergo passive dilation,
and the more likely is it to offer appreciable resistance
to perfusion (59, 141).
The anatomical characteristics of the small muscu-
lar pulmonary vessels are illustrated in figure 8. The
upper half of this figure depicts the structure of ex-
ceedingly small (30 n) pulmonary vessels: in neither
the pulmonary "arteriole" or venule is smooth
muscle discernible; by way of contrast, the coat of
smooth muscle in the systemic arterioles is readily
apparent. The lower half of this figure contrasts a
small pulmonary artery and a small pulmonary
vein — each about 50 ^ in diameter — with a systemic
arteriole of approximately the same size; pulmonary
arterioles of this size are to be found at the level of
the alveolar ducts and alveoli, buried in pulmonary
tissue (118). It may be seen that the pulmonary
arteriole contains only a thin rim of smooth muscle;
in the corresponding pulmonary venule of 55 /j, no
smooth muscle can be recognized; on the other hand,
the systemic arteriole contains a thick media. It is
difficult to imagine the pulmonary vessels shown in
figure 8 as the sites of intense vasoconstriction.
Somewhat better suited for vasomotor activity
are the larger precapillary vessels. These "small
muscular arteries" range from 100 to 1000 n in
diameter (403), contain well-formed media, and lie
adjacent to the respiratory bronchioles. They are
usually separated from the pulmonary tissue by
perivascular lymph spaces and their muscular coats
thin as they proceed peripherally to the vicinity of
the alveolar ducts. From these muscular vessels, the
pulmonary arterioles generally arise at right angles
so that the configuration of muscle at their origins
often appears sphincteric (118, 196).
The corresponding venules of 100 to 1000 y. lie
at the periphery of the lobule. And, in contrast to
the small muscular arteries, smooth muscle is either
poorly organized or absent and the elastic fibers are
irregular and indistinct. Consequently, even pul-
monary veins up to 1000 /* in diameter seem to be
poorly equipped for vasomotor activity.
capillaries. At the alveolar border, the precapillary
vessel subdivides to form a racemosing network of
capillary segments sandwiched between adjacent
alveolar walls (fig. 9, insert) (292). Whether these
capillaries lie free between the alveoli or indent
them — a structural distinction relevant to estimates
of pericapillary pressure — is uncertain.
The capillary circulation has certain distinctive
features: a) each of these capillary segments is ap-
proximately 10 to 14 p in length and 7 to 9 n in
diameter (422); b) except in congested lungs the red
cells pass through in single file (fig. 9); c) the capillary
networks in different parts of the lung differ with
respect to the length, caliber, and number of con-
stituent vessels (162, 292); d) "pores," presumed on
physiological grounds to exist in the pulmonary
capillary wall, have not been seen by electron mi-
croscopists; e) chemical analyses have failed to settle
if the capillary wall is predominantly aqueous or
lipoid in nature (375); /) there appear to be neither
contractile cells around the capillaries nor smooth
1676
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
eem
fig. 8. Structure of small pre- and postcapillary vessels. Upper half. Comparison of a 30 fi pre-
capillary vessel (left) and postcapillary vessel (center) with systemic arterial branches (right) of the same
size. The pulmonary pre- and postcapillary vessels are structurally similar; they are strikingly differ-
ent from the systemic arterial branches. Lower half. Comparison of a 57 11 pulmonary artery (left) and
a corresponding pulmonary vein (center with systemic arteriole (right) of the same size. Muscle cannot
be identified in the vein, ec, endothelial cells; ef, elastic libers; ™, smooth muscle; iem, internal clastic
membrane; eem, external elastic membrane; ac, alveolar capillaries. (Elastic tissue stain. 585 X)
[Courtesy of E. R. Weibel (132).]
muscle in the capillary walls. Without such con-
tractile elements, it is unlikely that the capillaries
can contract actively in the conventional manner of
muscle-containing vessels (400); on the contrary,
capillary lumens are more apt to be passively nar-
rowed by swelling of endothelial cells, by perivascular
transudates (114), by raised alveolar pressures (351),
and by the pushes and pulls of adjacent structures
(397)-
According to Weibel, the "typical" alveolus in
man is more like the cell of a honeycomb than a
sphere (422). It measures approximately 200 to
250 fi in diameter. Each alveolus is lined by a con-
tinuous epithelium (40 to 65 m,u thick) which changes
its submicroscopic appearance upon appropriate
stimulation (276, 375). In the human lung approx-
imately 300 million alveoli are juxtaposed to ap-
proximately the same number of capillary segments.
After the age of 8 years, an increase in the size of the
lung seems to involve an increase in the dimensions
of existing alveolar-capillary units rather than in
their number (422).
The thickness of the alveolar-capillary interface is
of the order of 285 to 640 rn.fi (375): not all of the
DYNAMICS OF PULMONARY CIRCULATION
1677
fig. 9. Electron microphotograph of human lung. The red
cells (RBC) are shown passing single hie through a pulmonary
capillary (CAP) between adjacent alveoli (ALV). 19,370 X.
[Courtesy of Dr. Councilman Morgan.] Insert: Network of cap-
illaries in the walls of the sacculi alveolares. 330 X- [From
Miller (292).]
70 m2; at three-quarters of the total lung capacity it
increases further (to the order of 90 m2) (422).
Extr avascular Smooth Muscle
Pulmonary smooth muscle is contained not only
in the vessels but also in the tracheobronchial tree
and in the pulmonary tissue. In man, the neatly
organized tracheobronchial smooth muscle continues
down to the mouths of the alveoli (fig. 10) where it
is in a position to influence passively the pressure in
the alveoli and, thereby, the caliber of the capillaries
in the alveolar walls (8, 196). Although parenchymal
smooth muscle is apparently plentiful in the amphi-
bian and reptilian lung (220, 236), and in patients
with chronic pulmonary disease (265), the quantity
and arrangement of this parenchymal smooth muscle
in the normal human lung is unknown. Nonetheless,
because of its close association with the elastic network
of the lung, parenchymal smooth muscle may con-
ceivably affect vascular calibers directly by con-
tiguity and, indirectly, by changing the pulmonary
lung volume and distensibility. Moreover, since the
musculo-elastic system of the lung is nourished by the
bronchial arteries, the possibility exists that agents
which reach the lungs by way of the systemic circu-
lation may change pulmonary vascular dimensions
through their effects on extravascular, rather than
intravascular, smooth muscle.
alveolar surface is ordinarily used for gas exchange;
nor is all of the capillary circumference in contact
with alveolar wall (196). The portion of the available
capillary surface which is actually used appears to
vary with the total lung volume, the degree of capil-
lary filling and the size of the alveoli. At a volume
corresponding to three-quarters of the total lung
capacity, the capillary network occupies 60 per cent
of the alveolar surface and the capillary blood volume
is of the order of 200 to 250 cm3 (422).
Over the years, anatomical measurements of the
capillary surface area have provided exceedingly
variable results: values have ranged from 50 m- to
140 m2 (132). Some of this discrepancy is undoubtedly
attributable to methodological differences (143), to
the uncertainties of reconstructing the lung on the
basis of small sections, and, particularly, to the failure
to specify the lung volume at which the measure-
ments were made. The recent measurements by
Weibel indicate that at the resting position of the
lung, the capillary surface is of the order of 50 to
Systemic Blood Supply of the Lung
In the normal human and canine lung, the bron-
chial arteries arise from intrathoracic systemic
arteries and deliver oxygenated blood to the walls
of the tracheobronchial tree, the supporting frame-
work of the lungs and the walls of the pulmonary
arteries and veins (133). Accordingly, they are
nutrient arteries. In contrast to pulmonary arteries
of equal caliber, the bronchial arterial walls are
thick and their innervation is plentiful. In the normal
lung, bronchial venous blood drains largely into the
azygous veins but some also enters the pulmonary
veins (263, 445).
The quantity of blood carried to the lungs by the
bronchial arteries is difficult to measure precisely;
the complexity of the problem may be inferred from
the wide variety of experimental approaches which
have been attempted in both dog and man (93, 133
214, 368). Nonetheless, despite inevitable differences,
the results of these diverse trials suggest that the
bronchial arterial flow ordinarily constitutes only an
i6?8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 10. Schematic representation of
the disposition of smooth muscle in the
terminal portions of the respiratory
tree. The cut muscle appears as small
nodular structures of various sizes and
shapes; it ends at the mouths of the
alveoli. [After Baltisberger (8).]
«r •*
t ft. J i . rvs To* *" \ x.
exceedingly small fraction of the cardiac output and
that the effect of the bronchial circulation on the
behavior of the normal pulmonary circulation is
negligible. The evidence for the latter conclusion is
of three general types: /) the difficulty encountered
by anatomists in finding bronchial-pulmonary ar-
terial communications in the normal human or
canine lung except by elaborate injection techniques
(421); 2) the measurements in the dog during arti-
ficial perfusion of the lungs which indicate that the
normal bronchial arterial flow is of the order of 1
to 2 per cent of the cardiac output (55, 423); and
3) the measurements in intact man which indicate
that the normal bronchial arterial blood flow is too
small to be measured by conventional techniques
(139, 152). It should be noted that under some ex-
ceedingly artificial experimental circumstances, the
bronchial circulation in the dog has been found to
exert an appreciable hemodynamic effect on the
pulmonary circulation (95). However, because of the
unusual experimental conditions, these results seem
to indicate the ultimate potential of the bronchial
circulation rather than its actual performance in life.
The systemic blood supply of the lung undergoes a
remarkable proliferation in various disorders of the
heart and lungs (79, 98, 133, 263) : old vessels enlarge
and become tortuous; new vessels appear and join
with the old to form bizarre Medusaean patterns.
Moreover, in contrast to the normal lung, in which
precapillary communications between the two circu-
lations are difficult to demonstrate (292, 420), the
enlarged precapillary anastomoses between the
DYNAMICS OF PULMONARY CIRCULATION 1 679
NORMAL
NUTRIENT
©
BRONCHIECTASIS
PA
C2)
HEMODYNAMIC?
TETRAD
©PA
COi
RESPIRATORY ?
®
PROXIMAL
BV
0 mm Hg 5 mm Hg
NORMAL
10 mm Hg 5 mmHg
RIGHT HEART FAILURE
fig. 1 1 . Pulmonary collateral circula-
tion. Upper half: The arterial portion. A:
usual nutrient function; B: expansion to
constitute a hemodynamic burden, as in
diffuse suppurative disease; C: participa-
tion in external respiration when systemic
arterial hypoxemia coexists with inade-
quate pulmonary arterial blood flow.
Lower half: The venous portion. D: usual
emptying of proximal bronchial veins; E:
alternate emptying of proximal bronchial
veins when right atrial pressures exceed
left atrial pressures. [After Fishman (133).]
pulmonary and systemic circulations are grossly
visible (264).
The proliferation generally does not affect the
entire collateral circulation in a uniform manner
(fig. 11). Thus, in the portion of the lung which lies
adjacent to an area of pulmonary inflammation, as
well as in the lung with a severely compromised
pulmonary arterial blood supply, it is the arterial
portion of the collateral circulation which expands;
on the other hand, the venous portion of the collateral
circulation undergoes the more striking expansion in
certain types of pulmonary emphysema and in mitral
stenosis (263). If the expanded collateral circulation
becomes sufficiently large — as in diffuse suppurative
disease of the lung (263) — it may carry large volumes
of blood, and transmit systemic blood pressures, to
the point of constituting a hemodynamic burden on
the pulmonary circulation.
It should be noted that as the collateral circulation
proliferates, the difficulties in measuring the rate of
collateral blood flow also grow. Particularly trouble-
some, from the technical point of view, are the
multiple origins of the collateral arterial branches on
the one hand and the alternate venous outlets on the
other (fig. 1 1). Indeed, on account of this anatomical
arrangement, it is difficult to measure volumetrically
the total collateral blood flow even in the open-chest
dog in which the heart and thoracic vessels are
i68<
HANDBOOK 11F PHYSIOLOGY
CIRCULATION II
exposed for cannulation. Another problem is the
prediction, on a priori grounds, of the hemodynamic
behavior, i.e., the blood flow, the driving pressure,
and the resistance to perfusion of anastomotic channels
which vary so in length, caliber and, possibly, in
tone. However, one anatomical aspect of the ex-
panded collateral arterial circulation does lend itself
to physiological exploitation: the precapillary anas-
tomoses make it possible to measure that part of
the collateral arterial inflow which reaches the gas-
exchanging surfaces of the lung and is available for
respiratory gas exchange, i.e., the "effective" col-
lateral blood flow (31, 139).
I 'enous Admixture
In the normal pulmonary circulation a small
quantity of venous blood traverses anatomical chan-
nels to bypass the gas-exchanging surfaces of the
lungs, thereby reducing the oxygen tension of periph-
eral arterial blood. This shunt has diverse anatomical
origins: bronchial veins, anterior cardiac veins,
Thebesian veins, portal veins, mediastinal veins and
pulmonary arteries; in normal man and dog the
volume of shunted blood is generally considered to
be of the order of 2 per cent of the cardiac output
(13. 23> 349)-
In the rabbit and guinea pig, arteriovenous chan-
nels have been seen on the surface of the transil-
luminated lung (219). In the dog, the evidence for
such shunts is less direct and there is considerable
dispute concerning their size (349). Two types of
observations favor a large size: /) glass spheres, up
to 500 fi in diameter, reach the left heart following
injection into the pulmonary artery (322) ; 2) radi-
opaque material, forcefully injected through the
side vent of a wedged pulmonary arterial catheter,
traces a cine-angiographic course suggestive of short-
circuits (331). Opposed is the experimental evidence
that these channels are closer to 25 n than to 500 n
in diameter (38, 167, 349). A reasonable interpreta-
tion of the disparate results in the dog is that the
experimental conditions determine the degree of
patency of these channels and that ordinarily these
channels are virtually closed (95).
The situation is somewhat more tenuous for the
human lung: on the one hand, large glass spheres
(up to 500 n in diameter) also traverse the isolated
human lung (322); on the other, is the inability of
painstaking histological examination to disclose the
channels (421) and the failure of physiological meas-
urements to obtain the high values for venous ad-
mixture which would be consistent with the presence
of large, patent channels (23). If arteriovenous chan-
nels do exist in the normal human lung, they seem to
allow very little blood flow under ordinary conditions.
The small anatomical shunt in the normal animal
or man stands in marked contrast to the large size
which it may achieve in certain clinical states, such
as congenital right-to-left intracardiac shunts and
pulmonary hemangiomatosis (155, 226). Appreciable
shunting has also been demonstrated in those patients
with cirrhosis of the liver who develop portal-pul-
monary venous communications (63).
Pulmonary \'asomotor Nerves
There is no doubt about either the existence of
pulmonary vasomotor nerves or their ability to
change pulmonary vascular calibers when appro-
priately stimulated; only their physiological meaning
can be questioned (96, 382).
The pulmonary vasomotor nerves have been most
intensively studied in the dog: both vasodilator and
vasoconstrictor fibers have been identified in the
upper sympathetic chain and in the vagus nerves
(95). Because of the complicated intermingling of
these fibers — not only with each other but also with
bronchial and cardiac fibers — electrical stimulation
often fails to separate vagal from sympathetic effects
on the one hand, and vasomotor from bronchomotor
and cardiac effects on the other (95).
For comprehensive reviews of pulmonary and
pulmonary vascular innervation the reader is referred
elsewhere (96, 267). A few aspects are particularly
relevant to considerations of pulmonary hemody-
namics: a) the large pulmonary arteries and veins
are more richly innervated than their smaller counter-
parts (81, 147, 392); b) the muscular arteries and
arterioles are more richly innervated than the cor-
responding veins and venules (392); c) nerve endings
reach the medial and subendothelial layers of the
large arteries and veins (392); d) sensory nerves and
receptors have been identified in the airways and in
the large pulmonary arteries and veins (6, 80) ; e )
the bronchial arteries are more richly innervated
than any other pulmonary vessels (292); and /) the
nerve supply to the bronchi exceeds that of the pul-
monary vessels (392).
As a pharmacological device for estimating the
concentration of adrenergic nerve endings in the
different parts of the pulmonary vascular tree, Euler
& Lishajko (126) compared the concentrations of
norepinephrine in the central and in the peripheral
DYNAMICS OF PULMONARY CIRCULATION
l68l
portions of the human pulmonary vascular tree. In
keeping with the anatomical evidence for a pre-
dominant distribution of nerves to the larger pul-
monary vessels, they found that the large pulmonary
vessels (of the dog and cow) contain larger quantities
of norepinephrine than do the small pulmonary
vessels. The greater concentration of nerves in the
region of the large pulmonary vessels is consistent
with the notion that the pulmonary vascular bed is
better innervated for tensing its large vessels than for
shrinking the caliber of its small ones (403). However,
this attractive idea, which is based on anatomical
observations, is inconclusive on several accounts:
a) the display of an abundant innervation provides
no measure of either the number or the nature of the
impulses which the nerves transmit; b) consecutive
muscular segments of a single pulmonary vascular
unit may be differently affected by a stimulus (95) ;
and c) because of its mixed embryological origin in
endoderm and mesoderm, pulmonary vascular in-
nervation may possess subtle, and as yet undisclosed,
features.
PULMONARY BLOOD FLOW
In subjects with a normal heart and circulation,
the pulmonary blood flow, the pulmonary capillary
blood flow and the output of each ventricle (the
cardiac output) represent virtually identical quan-
tities. In previous chapters of this book, the cardiac
output is considered with respect to its measurement
(Chapter 17) and control (Chapters 15 and 16); the
present section will confine itself to resting measure-
ments of pulmonary blood flow, leaving for subse-
quent sections the pulmonary capillary blood flow
and the behavior of the cardiac output during
exercise.
Normal Values
For the sake of comparison, cardiac output in man
is generally expressed per square meter of body sur-
face area (cardiac index): in one representative
study, the average cardiac index of a group of basal,
postprandial, supine human adults was 3.12 liters/
min/m2 (sd ±0.40); the corresponding oxygen uptake
of this group was 138 ml min/m2 sd ±14 (87).
Unfortunately, even in adults, body surface area is
not an ideal standard of reference; it becomes even
less reliable when subjects of different age, sex, and
body build are compared, since the •'normal"
values have been derived from a select portion of the
adult population. In the unanesthetized dog, the
cardiac output per minute is of the order of 150
ml per kg (12). It should be emphasized that there
are exceedingly few such measurements on the un-
anesthetized dog and the values which do exist are
far from consistent (303).
Excitement (discomfort or anxiety) may artificially
increase the "basal" cardiac output. This fact has
been illustrated by measurements on the unanesthe-
tized dog prior to, and following, treadmill exercise:
the resting cardiac output, while awaiting the start
of treadmill exercise, was higher than the resting
cardiac output after the exercise was finished (12).
Excitement may continue to operate during the test
periods. Fortunately, there are objective criteria
which can be used to detect the existence of disturb-
ing emotional influences; these include tachycardia,
a high respiratory exchange ratio of the expired gas,
a high oxygen uptake, and a high pH of systemic
arterial blood (136). Transient episodes of emotional
stress are apt to introduce appreciable errors into
steady-state measurements of flow, particularly by
the Fick principle (412, 439); on the other hand,
sustained excitement will artificially increase the
cardiac output. In the latter instance, the normality
of the cardiac output can be appraised by comparison
with the simultaneously measured oxygen uptake
(fig. 12). Ordinarily, the arteriovenous difference
for oxygen is of little help in such an appraisal since
its variations at rest approximate the limits of analytic
error (e.g., average of 38.4 sd ±6.3 ml per liter
(334))-
Uneven Pulmonary Blood Flow
The pattern of distribution of the right ventricular
output throughout the lung has been examined in
several different ways: a) direct inspection of the
pulmonary blood vessels; b) fractional, or continuous,
analysis of the alveolar component of expired air;
c) bronchospirometry or regional sampling of alveolar
air; d) external scintillation counting following the
breathing of radioactive gases; and e) the use of
conceptual models to explain actual respiratory
gas exchange.
Direct inspection of the lung for the determination
of the pattern of the pulmonary blood flow has been
practiced for at least 90 years (54)- Three types of
observations have been made: /) the examination of
the surface of the exposed lung in the living animal
(21, 105, 419), 2) the postmortem examination of the
1 68a
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
excised lung following the injection of tracer materials
(323, 405), and 3) roentgenography, including
angiocardiography (202). These approaches are all
qualitative but have clearly established two features
of the pulmonary circulation : the distribution of the
pulmonary blood flow is ordinarily quite uniform
but that it may be drastically modified by appro-
priate stimulation (fig. 13); structural abnormalities
are prepotent over physiological influences in deter-
mining the course taken by the blood (427). The
direct observations have also been used to account
for a variety of otherwise inexplicable clinical phe-
• •
♦ ♦
• • • .. mr i
•• • ♦
•Supine
+ Upright
OXYGEN UPTAKE (cc/min)
400 800 1200 1600 2000
pig. 12. Relationship between oxygen uptake and cardiac output at rest (supine), during supine
exercise, and during upright exercise. The diagonal line (at the far left) is based on the method of
least squares and represents spontaneous variations in the cardiac output in 56 normal subjects at
rest. This line lies to the left of the exercise data and has a steeper slope. For any given oxygen uptake,
the cardiac output is lower during upright exercise than during supine exercise. [After Reeves el al.
(336)-]
2400
2800
fig. 13. Variations in the distribution of the pulmonary blood flow. For each experiment, filtered
India ink was injected into a marginal ear vein of an unanesthetized rabbit loosely restrained in its
normal body position. .4.- uniform distribution of the India ink; B: "patchy' distribution following
introduction of a cardiac catheter into the right ventricle via the right external jugular vein (local
procaine anesthesia). [After Tuller el al. (405).]
DYNAMICS OF PULMONARY CIRCULATION
l683
nomena, including the bizarre "'butterfly" shadows
of pulmonary edema (202) and the maintenance of
the virtually normal oxygenation of peripheral arterial
blood in patients with atelectasis and pneumonia (85).
Much more relevant to the performance of the
lung in gas exchange is the distribution of the pul-
monary capillary blood with respect to alveolar
volume, alveolar ventilation, and pulmonary diffusing
surfaces. In the normal lung, these parameters are
ordinarily quite precisely balanced (131, 145, 427);
in disease, the upsets may be quite striking (58).
Two approaches are in popular use for relating pul-
monary capillary perfusion to alveolar ventilation:
the determination of the pattern of change in the
alveolar composition of a respiratory gas (284) and
of the respiratory exchange ratio (428) during a
single expiration; the determination of the rate of
increase in the peripheral arterial oxygen saturation
(314) and tension (131) during oxygen breathing.
It has been pointed out elsewhere that each of these
approaches has its own uncertainties (428).
The comparison of blood flow through different
parts of the lungs has generally involved either
bronchospirometry or regional sampling of alveolar
gas. Bronchospirometry has been particularly fruitful
in comparing the perfusion of the two lungs; thus,
simultaneous measurements of the uptake of each
lung separately have disclosed that ordinarily each
lung receives a share of the cardiac output which is
proportional both to its gas volume (29) and to its
ventilation (203). Accordingly, in man, the right
lung receives 55 per cent of the cardiac output (135);
this fraction is decreased when the subject turns on
his side so that the left lung is down (29).
Bronchospirometric comparisons of oxygen uptake
have also disclosed that gravity rearranges the
distribution of the blood flow within each lung: as
the human subject stands, the oxygen uptake of the
lower lobes increases at the expense of the upper
lobes, indicating a preferential distribution of blood
flow to the lower lobes; the change in the pattern of
the blood flow occurs even though the distribution of
ventilation is little altered by the change in posture
(286). It should be noted that the use of broncho-
spirometry to detect changes in regional blood flow
presupposes that all parts of the lungs are breathing
the same inspired mixture; when different parts of
the lungs are given different inspired gas mixtures to
breathe, the procedures and calculations grow much
more complicated since all of the variables in the
Fick equation — instead of only the oxygen uptake — ■
have to be determined (135, 221, 410).
Hemodynamic measurements (247) and analyses
of alveolar gas have been consistent with the broncho-
spirometric measurements. For example, the alveolar
gas analyses have shown that: a) the oxygen tension
of the upper lobes exceeds that of the lower lobes
(285, 330); b) the carbon dioxide tension of the upper
lobes is less than that of the lower lobes (284, 330,
388), and c) the respiratory exchange ratio of the
upper lobes exceeds that of the lower lobes (428).
All these observations are consistent with the clinical
belief that high oxygen tension in the apices of the
lungs, resulting from inadequate perfusion with
respect to ventilation, is responsible for the apical
localization of pulmonary tuberculosis (344). They
also indicate that if intrapulmonary baroreceptor
mechanisms for rearranging pulmonary blood flow do
exist at the pulmonary bases, they are easily over-
whelmed by the mechanical effects of gravity.
The combination of xenon133 and external counting
was originally used to estimate the distribution of
inspired air (32, 234). Subsequently, oxygen15 (107)
and then oxygen'Mabeled carbon dioxide (427) were
introduced to relate the distribution of the perfusion
to the distribution of the inspired air. In addition to
confirming that in the seated normal subject the
lower lobes are much better perfused than the upper
(8:1) (428), these studies also expressed, in quantita-
tive terms, the spectrum of ventilation-perfusion
ratios which exist in the lungs of upright normal
man, and showed how the ratios gradually convert
from high to low values as the base of the lung is
approached. Moreover, although these inhomoge-
neities have inevitable consequences for the gas
tensions in the regional alveoli and capillaries, they
were shown to have little significance for the efficiency
of the lung in oxygen uptake or carbon dioxide output.
Finally, the intrapulmonary distribution of air and
blood was demonstrated to become much more
uniform when the normal subject assumed the supine
position or when mechanical influences, such as
anatomical restriction of the lower pulmonary
vascular bed by congestion and fibrosis, counteracted
the tendency of gravity to direct blood to the lower
lobes in the upright position (107).
Because the normal lung is too inhomogeneous
and too complicated to be treated in simple mathe-
matical terms, conceptual models of alveolar-capillary
gas exchange have been adopted as practical tools
for assessing the adequacy of pulmonary capillary
perfusion. One particularly useful model has been
the homogeneous "ideal" lung, a figurative lung to
which actual inhomogeneities can be referred (266,
[684
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
PULMONARY
ARTERY
SYSTEMIC
ARTERY
60-
X
- 1
Va
Q
7
i
1
5
Va 1
vV1
— i 1 — ■ I
E 40 -1
- ~\
o
MIXED VENOUS
BLOOO
Va 5
<
1
20 -
INSPIRED
-* — m- — i—
0
1 f—
— 1 —
Pa
— (—
°2
mrr
— y
Hg
—
CAS ■>.
— 1 1 1
40
60
80
VENOUS ADMIXTURE
fig. 14. Model of the lungs. Any inhomogencity of ventilation and perfusion is represented as
"virtual venous admixture." Pulmonary arteriovenous shunts appear as •'anatomical venous ad-
mixture." According to this model, the alveolar-arterial difference in oxygen tension may be sub-
divided into three components: diffusion, virtual venous admixture, and anatomical venous ad-
mixture. As indicated in the text, this is an oversimplification. [After Briehl & Fishman (51).]
327, 345): the standard tactic is to express deviations
from homogeneity in terms of their effect on the
alveolar-arterial differences in oxygen tension ("A-a
gradient"). By partitioning the A-a gradient into
three components (fig. 14), it is possible not only to
identify the venous admixture component, but also
to separate it into anatomical and "virtual" portions;
the "virtual" venous admixture is then an expression
of the inhomogeneity of pulmonary capillary perfusion
with respect to other gas-exchanging parameters
(14. 23, 51).
The picture which has emerged from this type of
approach is illustrated in figure 15: alveoli which
are excessively perfused for their ventilation (VA/Q
< 0.8) contribute to the virtual venous admixture;
those which are perfused but nonventilated (VA/Q.
= o) appear as anatomical venous admixture; those
which are excessively ventilated for their perfusion
(VA/Q. > 0.8) contribute to the "physiological" dead
space, the "alveolar" dead space, and to the alveolar-
arterial gradient for carbon dioxide (347* 377)-
While this model is the basis of much of contempo-
rary thinking about the distribution of blood flow
with respect to gas exchange, it is known to be inade-
quate on several practical and theoretical accounts:
a) the fractionation of A-a gradient is technically
difficult and apt to be imprecise, especially in patients
with diffuse pulmonary disease; b) the model does
not recognize other inhomogeneities, e.g., between
perfusion and diffusing capacity or between stroke
output and pulmonary capillary blood volume, which
100
120
140
160
fig. 15. Hypothetical distribution of alveolar ventilation-
perfusion ratios (VA/Q) within the normal human lung. Values
for VA/Q range from zero at the mixed venous blood point
(perfusion but no ventilation) to infinity at the inspired air
point (ventilation without perfusion). According to this model,
the VA/Q ratio of each alveolus fixes its respiratory exchange
ratio (R) as well as its gas tensions (Pan„, Pa02, and PacoJ.
[Based on Riley & Cournand (345) and Rahn (327).]
consequently appear as imbalances between ventila-
tion and perfusion (319, 413); and c) basic assump-
tions, such as the type of statistical distribution of
ventilation-perfusion ratios among the alveoli may
be erroneous (128).
In practice, the mixing formula shown in figure
16 is generally applied to data obtained during
ambient air breathing to determine the total venous
mixture, i.e., the sum of the anatomical and the
virtual; by repeating the measurements during
high-oxygen breathing, the virtual component is
minimized so that the venous admixture consists
almost entirely of the anatomical component (23).
DYNAMICS OF PULMONARY CIRCULATION
l685
PULMONARY
ARTERY
SYSTEMIC
ARTERY
(Cq)
Qs ca _ cc
Q Cy — Cc'
fig. 16. Schematic representation of the lung to illustrate
the components of the venous admixture and the calculation of
the venous admixture as a fraction of the cardiac output.
Q, = venous admixture (anatomical plus "virtual"); Q =
cardiac output; Ca, Cc, C; = oxygen content of arterial, end-
capillary, and mixed venous blood, respectively. Furthermore,
by administering enriched-oxygen mixtures, the total venous
admixture (Qs/Q) can be subdivided into its anatomical and
"virtual" portions.
Attempts have also been made to measure anatomical
venous admixture in other ways, e.g., the simultaneous
intravenous injection of T-1824 and KrS5; unfor-
tunately, such methods are most reliable when the
anatomical venous admixture is large, i.e., greater
than 1 5 per cent of the cardiac output (155).
In recent years, relationships between pulmonary
capillary perfusion and other gas-exchanging param-
eters have been clarified in many different ways:
a) the determination of alveolar-arterial gradients
for nitrogen (67, 232); b) the quantification of the
role played by parameters other than ventilation and
perfusion in determining virtual venous admixture
(319); c) the comparison of anatomical dead spaces
with physiological and alveolar dead spaces (347,
377); d) the analysis of the pulmonary elimination
of intravenously injected radioactive tracers (173);
and e) by the creation of new and more elaborate
models (52, 131). With the growth of understanding
of these interplays has come the fuller appreciation
of the extent to which they may affect conventional
tests of pulmonary performance and calculations of
pulmonary resistance.
PULMONARY VASCULAR PRESSURES
Recording
The characteristics of adequate manometric systems
as well as the limitations of the cardiac catheter in
reproducing the intravascular and intracardiac
pressure pulses are considered elsewhere in this
volume. However, it should be emphasized that
with modern, hi-fidelity recorders and manometers,
it is generally the catheter attached to sensing ele-
ment, rather than the manometric system, which
limits the capacity of the apparatus to duplicate
faithfully the pressure pulse. It is also noteworthy
that even though blood pressure recorded from the
end of a catheter in the pulmonary artery is suffi-
ciently exact for most physiologic purposes, it fails to
measure the lateral pressure in the vessel by a small,
but variable, amount.
Hydrostatic Reference Level
For the measurement of absolute pressures within
the thorax, correction is made for the hydrostatic
pressure difference between the intrathoracic site
from which pressure is being recorded and the exter-
nally-placed sensing element of the manometer
(169). For this purpose, the plane of the sensing
element is set in relationship to both the heart and
to some thoracic landmark. In practice, different
hydrostatic zero levels have been adopted : most
popular are levels 5 cm below the angle of Louis
(253) and 10 to 12 cm above the tabletop (103).
While the different reference levels do complicate
the comparison of data from different laboratories,
each is a perfectly reliable standard for comparing
consecutive measurements in a single animal.
Several uncertainties creep into the use of fixed
external references to obtain absolute values of
intrathoracic blood pressures, particularly in patients
who are dyspneic from cardiac or pulmonary disease.
Thus, in subjects with large hearts or unusual con-
figurations of the chest, it may be difficult to estimate
precisely the difference between the external reference
plane and the intracardiac site of reference (253);
moreover, even in normal subjects, the heart changes
position during each cardiac cycle. Consequently, it
seems reasonable to view pulmonary vascular and
intracardiac pressures which are measured in this
way as accurate only to within a few mm Hg.
Unfortunately, many intuitively attractive solutions
to the problem of "zeroing" are not feasible: the
tip of the cardiac catheter, as localized by X ray,
cannot, per se, serve as the zero reference plane; nor
is it a simple matter to "zero" an intracardiac
manometer which is built into the tip of a cardiac
catheter. Nonetheless, despite these difficulties in-
herent in the choice of reference levels for absolute
1 686 HANDBOOK OF PHYSIOLOGY v-~ CIRCULATION II
fr/V A,
i \ ! V \ I
hi ijiiMii y \
fig. 17. Simultaneous aortic (AO) and pulmonary arterial (PA) pressure pulses recorded from a
human subject with a normal circulation during open thoracotomy. The arrows indicate the be-
ginning of ejection and the end of protodiastole in the aorta and pulmonary artery. (Paper speed,
25 mm/sec on left; 75 mm/sec on right. Interval between time lines, 0.04 sec.) [After Braunwald
etal. (47).]
values, any of these reference levels will suffice for
consecutive measurements in a single experiment.
Pulmonary Arterial Pressure
With minor differences, the contour of the pul-
monary arterial pulse mirrors that at the root of the
aorta: as may be seen in figure 17 (47), the pul-
monary arterial pressure pulse is small in amplitude
as compared to the aortic pulse and characteristically
displays a rapid rise to a rounded peak during systole,
a brisk small incisura and a gradual decrease in
pressure during diastole (182, 225). The "classical"
pulmonary arterial curves are more apt to be recorded
in pulmonary hypertensive states than in pulmonary
normotensive states; at the lower levels of pulmonary
arterial pressure, distorting artifacts are exceedingly
common. Not shown are the corresponding records
of the velocity of the blood flow: in contrast to the
pressure-velocity relationships in the aorta, the
pulmonary arterial pressure-velocity curves are
quite similar: the velocity of blood flow in the pul-
monary artery lags slightly behind the pulmonary
arterial pressure (156).
Ordinarily, the pulmonary arterial mean pressure
in man (87), dog (303), cat (132), and the rabbit
(239) averages one-fifth to one-sixth that in the
systemic circulation. In man, the level of the pul-
monary arterial pressure seems to increase slightly
with age (101). There is no fixed relationship between
the pressures in the two circuits. In man, before the
onset of systole, the pulmonary arterial pressure is
of the order of 7 to 12 mm Hg; during systole it
rises abruptly to 20 to 30 mm Hg; the corresponding
mean pressure is of the order of 12 to 15 mm Hg
(87, 103). In the dog, pulmonary arterial pressures
tend to be somewhat higher, so that a mean pressure
of 20 mm Hg is not unusual (187).
Pulmonary Venous and Left Atrial Pressures
Blood pressures have been recorded directly from
the left atrium and pulmonary veins in dog (187)
and man (45, 92, 299). The pulmonary venous
pressure pulse is a record of left atrial events, indi-
cating that the pulmonary arterial pressure pulse has
been damped out by the small pulmonary vessels.
As in the systemic veins and right atrium, the a, c, and
v waves are clearly defined (187); but, in contrast
to the right heart, the summit of the v wave is usually
the highest part of the pressure pulse, and pressure
variations during the cardiac cycle are greater in the
left atrium and pulmonary veins. Thus, in both the
unanesthetized and anesthetized dog, pulmonary
venous pressures during a single cardiac cycle range
between 3 and 12 mm Hg (187). In intact, unanes-
thetized man the mean left atrial pressure is of the
order of 4 to 5 mm Hg (47). Although physiologic
observations (121) arc accumulating to support
the anatomic impression (56, 392) that the pulmonary
venous-left atrial junctions can act as sphincters,
final proof, in the form of suitably recorded pulmo-
nary venous-left atrial pressure gradients or differences
between the contours of the pulmonary venous and
left atrial pressure pulses have as yet not been pub-
lished.
Until recently, measurements of left atrial and
pulmonary venous pressures in intact animals were
DYNAMICS OF PULMONARY CIRCULATION
1687
confined to dogs fitted with angiostomy cannulae
(150, 225); in recent years, these pressures have
been measured in both intact animal and man by
every conceivable route: right heart catheterization
in patients with congenital atrial defects, direct
cardiac puncture, transbronchial puncture, trans-
thoracic puncture, and intracardiac transseptal
puncture (45, 299).
Pulmonary Arteriovenous Pressure Gradient
In man, cat, and dog, the pressure drop across the
pulmonary vascular bed is of the order of one-tenth
of the pressure drop across the systemic circulation.
The pulmonary arterial-left atrial pressure gradient
is maximal early in systole (fig. 18); it decreases late
in systole and may even approach zero if diastole is
sufficiently prolonged (182). Unfortunately, since
both the pulmonary arterial and pulmonary venous
pressure pulses have different origins (right and left
sides of the heart, respectively), it is not possible to
predict the shape of the pressure pulses of the inter-
vening vascular bed from the pulmonary arterial-
pulmonary venous pressure gradient.
Pulmonary Wedge Pressures
The pulmonary arterial wedge pressure is recorded
by advancing a cardiac catheter until its tip occludes
a terminal branch of the pulmonary artery; flow then
fig. 18. The pulmonary vascular pressure gradient. Upper
curve: record of the pulmonary arterial pressure pulse of an un-
anesthetized, unoperated dog; blood pressure 35/12 mm Hg,
mean 20 mm Hg. Lower curve : record of the pulmonary venous
pressure pulse; blood pressure 2 to 12 mm Hg. Middle curve:
differential manometer record of pulmonary arterial minus
pulmonary venous pressure, i.e., the gradient of pressure driving
blood through the pulmonary vascular system. [After Hamilton
(182).]
stops in the vascular segment beyond the tip of the
catheter: the pressure transmitted by the intervening
static column of blood presumably approximates
closely the pressure in the first communicating
pulmonary veins in which flow persists. Pulmonary
venous wedge pressure is recorded by impacting a
catheter (passed retrograde) in a pulmonary vein.
Originally (200), it was believed that the wedged
pulmonary arterial pressure could be used as a meas-
ure of pressure in the pulmonary capillary bed. It is
now clear that the wedged arterial catheter registers
more remote events, i.e., events in the large pulmonary
veins and, unless the "throttles" actually operate, in
the left atrium (83). In both dog and man — with
normal pulmonary circulation or with pulmonary
venous congestion — the mean pulmonary arterial
wedge pressure and the mean left atrial pressure are
nearly identical (83). In the normal animal and man,
the level of the arterial wedge pressure is of the order
of 5 to 9 mm Hg (103); in patients with pulmonary
venous congestion from mitral stenosis, it parallels the
left atrial and pulmonary venous pressure.
The validity and meaning of the arterial wedge
pressure have been the subjects of considerable debate
(26). Various criteria have been adopted for deciding
if a wedge pressure is a reliable measure of the level of
the left atrial pressure; these include higher pulmonary
arterial mean and diastolic pressures than the re-
corded wedge pressure, the withdrawal of fully oxy-
genated blood from the impacted catheter, the snap of
the catheter as it is withdrawn from the wedge posi-
tion and a characteristic configuration of the wedge
tracing (83, 103). No single one of these criteria en-
sures a reliable measure of left atrial pressure,
particularly when pressure is changing rapidly (26).
Indeed, even when all criteria are met, the left atrial
pressure may be poorly transmitted due to a faulty
wedge position of the catheter (fig. 19) (22).
The use of the arterial wedge pressure as a measure
of the level of left atrial pressure is on sounder footing
than its use to record the contour of the left atrial
pressure pulse. Only in states of pulmonary venous
congestion is the wedge catheter apt to reproduce
cyclic events in the left atrium (83, 113). Interpreta-
tion of changes in contour is particularly troublesome
when artifacts are present; these artifacts tend to be
most pronounced during exercise and deep breathing.
Blood pressure falls in the pulmonary artery distal
to an occlusive balloon and assumes the nondescript
character of a wedge pressure (fig. 20) (49, 42). The
level of this distal pulmonary arterial pressure
corresponds to that in the left atrium and fluctuates
1 688 HANDBOOK OF PHYSIOLOGY ■*■ CIRCULATION II
fig. 19. Various positions of the "wedged" catheter redrawn from pulmonary wedge arterio-
grams. A: the catheter is wedged in an artery which is slightly smaller than the catheter tip; the
lumen of the artery is in direct line with the lumen of the catheter. B: the catheter is wedged at a
bifurcation of an artery of the same size as the catheter tip. C: the tip of the catheter impinges against
the wall of a sharply angulated artery. D: The catheter is wedged at a point where the artery divides
into three or more branches. E: the catheter is incompletely wedged. The injected dye regurgitates
around the catheter outlining the artery proximal to the catheter tip. Positions A, B, and D are
favorable for recording wedge pressure; positions C and E are not. [After Bell el al. (22).]
with changes in the left atrial pressure. The tracing
shows no left atrial or pulmonary venous events but
does display respiratory swings.
Pulmonary venous wedge pressures have also been
recorded in the dog (435), in normal human subjects
(84, 248), and in patients with atrial septal defects. In
the normal dog the pulmonary venous wedge pressure
approximates mean pressure in the pulmonary artery
(435) '■> m patients, with pulmonary hypertension, the
pulmonary arterial mean pressure is much higher
than the pulmonary venous wedge pressure, pre-
sumably due to the interposed high vascular resistance
(84).
In brief, neither the pulmonary arterial wedge
pressure nor the pulmonary venous wedge pressure
provides a measure of the pulmonary capillary
pressure. However, with care and under appropriate
circumstances, the pulmonary arterial wedge pres-
sure does provide an approximate measure of the
pulmonary venous, and usually, of the mean left
atrial pressure; it can then be used to estimate the
driving pressure across the entire pulmonary vascular
bed and to calculate the resistance to perfusion offered
by the small pulmonary vessels.
Influence of Intrathoracic Pressure on Pulmonary
Vascular Pressure
Pressure in an intrathoracic vessel is not a simple
concept. In order for such a pressure to have meaning,
it must be related to a reference level, i.e., atmospheric
or pleural pressure. If the manometer which records
the pressure is balanced against atmospheric pressure,
all pressure changes within the thorax arising from the
ventilation — for example, a cough (fig. 2 1 ) — will be
immediately propagated across the walls of the pul-
monary vessels and heart to the incompressible blood
which they contain; the intrathoracic pressure changes
will, therefore, be registered as an integral part of
the pressure pulse. However, pressures recorded in
this way ("luminal" pressures) provide no measure
of the pressure which distends the vessels ("trans-
mural" pressures) : during a cough, while the pressure
recorded by a manometer balanced against atmos-
pheric pressure rises precipitously, a manometer
balanced against pleural pressure shows that the
transmural pressure has remained virtually unchanged
(190).
Values for the pleural pressures have been obtained
DYNAMICS OF PULMONARY CIRCULATION
1689
in various ways, including direct measurements from
gas pockets and balloons within the pleural or
mediastinal spaces (82) and indirect estimates from
the esophagus (287). It is generally conceded that
mm
Hg
20 r
10
0
-i
20rl
10
0
BEFORE
,#www
^q^AA/VVWVV
WfflffifflA
m»gMUi***0mm*mm tu
AFTER
IiVWi'Mt-m-i
fig. 20. Effect of occluding the right pulmonary artery on
blood pressures distal and proximal to the occlusive balloon.
Before occlusion, blood pressures are identical in the main and
right pulmonary arteries. After occlusion, the distal pressure
falls to the level of pulmonary wedge pressures (left atrial
pressure), pressure in the main pulmonary artery proximal to
the balloon increases by approximately 5 mm Hg. ( Unpub-
lished observations of M. Brandfonbrenner, A. Himmelstein,
G. M. Turino, and A. P. Fishman.)
even the direct methods may fail to provide precise
measurements of the pressures which are operating
at the surface of the particular pulmonary vessels
under consideration: the pressure within the pleura
may not be entirely uniform (82, 127); the extramural
pressures along the length of the vascular tree may
differ from segment to segment and from the pleural
pressure, depending on the location of the segment,
i.e., intrapericardial, intrapulmonary, or juxta-
alveolar. The use of indirect measures, which provides
reliable measures of pleural pressures in some experi-
mental and clinical conditions, fails in others (287).
Transmural \ersus Luminal Pressures
During each respiratory cycle, the changing pleural
pressures (fig. 22) affect all intrathoracic vessels except
those apposed to alveoli. Consequently, for the
alveolar capillaries, the pressure which determines
their caliber, i.e., transmural pressure, is customarily
calculated as the difference between (estimated)
intracapillary and alveolar pressure; the transmural
pressure of all other vessels is calculated as the differ-
ence between the luminal and the pleural pressure
(fig. 23) (61, 233). The practical difficulties in esti-
mating perivascular pressure from pleural pressure
have been indicated above; pericapillary pressures
also have an element of uncertainty because of the
prospect that tissue forces, such as alveolar surface
tension, may decrease pericapillary pressure to sub-
atmospheric levels.
Depending on the purpose of the observation,
pulmonary vascular pressures are referred either to
atmospheric or to pleural pressure. Considerable
fic. 21. Differential pressure record of a "cough."' The lowest tracing is from a mouthpiece into
which a forcible expiration was made. The middle record is that of luminal systemic arterial pressure.
The upper record is a differential record of the middle minus the lower record. The mouthpiece
record is assumed to show pressure changes nearly identical with intrathoracic pressure changes;
the differential record indicates the stresses which the intrathoracic arteries undergo. [After Hamilton
etal. (190).]
169O HANDBOOK OF PHYSIOLOGY "- CIRCULATION II
T.M.
M
IPP
I ! lilt i: i ll!l I ! 1,1 lltlil'l I "1 I' I
bJ"***"1* -kJ
I Hi! I !:
REST- 21% 02
fig. 22. Effects of breathing 5' <
CO2 and of exercise on the pleural
pressures (IPP) and esophageal pres-
sures (EP) of a human subject. All
pressures are in mm Hg. [After Fishman
et al. (132).]
IPP
""''! — T-
confusion has arisen from the indiscriminate use of
transmural pressures for luminal pressures in the
calculation of pulmonary vascular resistance. It
should be emphasized that as long as left atrial pres-
sure exceeds alveolar pressure, the measurement of
the driving pressure across the lung requires only the
simultaneous measurements of luminal pulmonary
arterial and venous pressures — no matter what the
intrathoracic pressure may be.
PULMONARY BLOOD VOLUME
The pulmonary vasculature constitutes a distensible
reservoir, interposed between the right and left heart.
The volume of blood which it contains is of interest
on three separate accounts: /) the mechanical be-
havior of the lungs; 2) the efficiency of gas exchange;
and j) the sustained return of pulmonary venous
blood to the left heart. In large part, the volume of
blood contained in the lungs at any instant is de-
termined passively by the balance between pulmonary
inflow, i.e., between the output of the two ventricles;
it is also influenced considerably by the ventilation.
Whether an element of self-control is also provided by
pulmonary vasomotor activity, particularly on the
part of the veins (305) or of hypothetical venous
sinuses (381), is uncertain.
fig. 23. Difference between luminal pressures (referred to
atmosphere) and transmural pressures (referred to perivascu-
lar pressure) along the length of the pulmonary vascular tree.
The shaded area represents the luminal pressure. In the capil-
laries (PC), which are exposed to alveolar pressure, the luminal
and transmural pressures are virtually identical. On the other
hand, in the pulmonary artery (PA) and vein (PV), the trans-
mural pressure exceeds the luminal pressure by the pleural
(perivascular) pressure.
Measurement of Pulmonary Blood Volume
For convenience, the methods for measuring pul-
monary blood volume may be sorted according to
whether they are designed to measure the pulmonary
blood volume or a change in pulmonary blood
volume.
In the isolated lung or in thoracotomized animals,
the pulmonary blood volume is available for direct
mensuration (293, 384); but, because of the surgical
manipulations and the drastic experimental condi-
tions, the measured volume may differ considerably
DYNAMICS OF PULMONARY CIRCULATION
l6gi
from the volume which prevails under more natural
conditions. In intact animal or man, indicator-dilu-
tion techniques have been commonly used to approxi-
mate the size of the pulmonary blood volume.
stewart-hamilton: indicator dilution. This is an
indicator-dilution method (fig. 24) which entails the
introduction of a test substance into the venous side
of the circulation and the registration, from a systemic
artery, of its changing concentration with time (18,
hi, 184). This application was first proposed by-
Stewart (184), who held that the product of the flow
and the appearance time of the injected substance is
a measure of the capacity of the bed through which
the flow takes place; this idea was shared by Blumgart
and Weiss (184). Hamilton and collaborators (186)
showed that the mean circulation time rather than
the shortest circulation time should be used to calcu-
late the volume of blood in the vascular bed between
the point of injection and the point of sampling. In a
simple model, in which the entire stream passes the
points of injection and of sampling, the idea that the
product of the flow and the mean circulation time
measures the intervening volume is not only ac-
ceptable intuitively, but has also been checked in
models (186) and proved mathematically (444).
Since the mean circulation time is approximately the
time ordinate corresponding to the center of gravity
of the time-concentration curve (fig. 25A), the substi-
tution of the median for the mean circulation time
may introduce considerable error into the calcula-
tion (184).
The injection into the venous circulation coupled
with sampling from a peripheral artery defines only a
"central blood volume"; its limits are wide and vague:
it includes not merely the blood volume between the
needles, but also the volume of blood contained in the
other branches of the venous and arterial trees having
equivalent circulation times. It is a virtual volume
which corresponds to an anatomical volume only
under ideal conditions: if mixing of blood and tracer
is complete and uniform, if the system contains neither
stagnant nor sequestered blood and if there are no
preferential channels which are operating to short-
circuit the system. The use of mathematics to con-
struct a continuous infusion curve from the single in-
jection curve involves identical premises and does not
make the measurement of the pulmonary blood vol-
ume any more definitive. The continuous infusion of
a tracer substance into the central circulation does
provide an alternate approach for measuring the
pulmonary blood flow and central blood volume
(fig. 25B) (444); however, as in the case of the single
injection, the results promise to be less precise for
volume than for flow (444).
When the test substance is injected into a peripheral
vein (instead of into the pulmonary artery), the cen-
tral blood volume includes the whole cardiac blood
volume. Many different radiological techniques have
been applied to the measurement of the cardiac
blood volume in dog and man (151). Despite theoreti-
cal reservations of various kinds — such as the diffi-
culty in separating the contribution of cardiac cavities
and walls to the radiographic picture of the heart —
the radiographic cardiac volume in dogs was found to
correspond, within 10 per cent, to the directly meas-
Jt
.6.4 mg/Lit
<
tr
t-
o
z
o
o
20
I 0 -
INJECT E D
TIME
Av CONC
6 0
14.11
I 5 sec
m g
FLOW =
CB V
I 1.70 mg/Lit
14.11
11.70 x
MCT
I 5
1 1 sec
= 4.82
Lit/min
4.82
60
x I I = 880 ml
SECONDS
fig. 24. Concentration-time curve inscribed by densitometer through which peripheral arterial
blood was drawn at a constant rate following injection of T-1824 into the pulmonary artery of a
normal human subject. At t = o, the indicator was injected. The calibration marks at the top of the
record indicate that a deflection of 1 cm is equal to a concentration of 6.4 mg of dye per liter of
blood. From such a record, the pulmonary blood flow, the mean circulation time (MCT), and the
central blood volume can be calculated as shown.
[692
HANDBOOK OF PHYSIOLOGY "> CIRCULATION II
40r
10 15
Seconds
fig. 25. Schematic representations of concentration-time curves following injection of indicator
into central circulation. A: single injection curve. Recirculation of indicator occurs at arrow. The
dashed lines illustrate likely extremes of extrapolation of the downlimb to zero during the first cir-
culation of indicator. The shaded arrow represents the relative difference between the two esti-
mates of blood flow based on the two extrapolations. The vertical lines fi and £2 represent the two
estimates of mean transit time based on the two extrapolations. If recirculation occurs earlier, so that
the shape of the downlimb is uncertain, considerable errors may be introduced by the extrapolation.
B: constant injection curve. The times at which the indicator just appears and recirculates are identi-
cal with those in panel A. The dashed lines represent likely extrapolations to a plateau concentra-
tion. The shaded area between Pmax-i and Pmax-s represents the difference between estimates of
area above the extrapolated buildup concentration curves. The problem of recognizing the point of
recirculation is the same as for the single injection curve of panel A. [After Zierler (444).]
10 15
Seconds
ured volumes (184); moreover, following epinephrine
overdosage, the radiographic cardiac volume was
found to constitute an unusually large fraction of the
central blood volume (151, 184). By substituting
cineradiography of the opacified intracardiac volumes
for conventional radiography, the precision of the
radiographic approach has been greatly enhanced
(73, 170, 367); this modification promises a reliable
measure of the volumes of the individual chambers in
normal man and dog. It remains to be seen if precise
measurements of this type can also be made in pa-
tients with pulmonary congestion and cardiomegaly.
Recently, the central blood volume has been ex-
perimentally narrowed to the pulmonary blood vol-
ume by the use of two catheters — one in the
pulmonary artery and the other in the left atrium.
Once placed, the catheters have been put to different
uses: a) for injecting a tracer substance into the pul-
monary artery and for sampling from the left atrium
(246, 293), and b) for injecting tracer substances into
both the pulmonary artery and left atrium, and
sampling from the brachial artery, thereby determin-
ing the mean pulmonary arterial-left atrial transit
time (106, 278). Although the second of these ap-
proaches was designed to circumvent the theoretical
possibility of incomplete mixing in the left atrium, the
values for the pulmonary blood volume by both ap-
proaches have been not only similar, but also sur-
prisingly low.
newman: exponential DOvvNSLOPE. The time-con-
centration curve of injected substance typically has a
descending exponential limb. According to Newman
(304), the slope of this line measures the volume of a
model through which water is perfused if there is
instantaneous and complete mixing of injected dye
and perfusate. If several chambers are perfused in
series, the slope indicates the volume of the largest.
Assuming that the lung volume is the largest of those
concerned in the circulation, Newman used the slope
to obtain a measure of the pulmonary blood volume.
In 1932, Hamilton et al. (186) had evolved an equa-
DYNAMICS OF PULMONARY CIRCULATION
1693
tion similar to that of Newman, but rejected the idea
that the volume term in that equation could stand
for a significant physiological volume because, in the
physiological circuit, there is neither instantaneous
nor complete mixing of dye with all of the blood in
either heart or lungs. It now seems that the 1932 view
is correct (1 1 1, 283, 417).
bradley: equilibration curves. The method origi-
nally devised by Bradley et al. for the estimation of
splanchnic blood volume (41) has been applied by
others to the estimation of the pulmonary blood vol-
ume (326). The method entails the determination of
the amount of tracer substance contained in the sys-
tem at equilibrium (cardiac output X arteriovenous
difference X equilibration time) divided by the
equilibration concentration of tracer. From the point
of view of application to the lungs, the most vul-
nerable part of the equation is the arteriovenous
difference. Experiments with models have shown that,
in contrast to the splanchnic circulation, the pul-
monary circulation is not suited for this type of
equilibration method (46). Consequently, it is difficult
to place much confidence in the measurements in man
which find that all three methods — the Stewart-
Hamilton, Bradley, and Newman — provide com-
parable values for the pulmonary blood volume
(326), particularly when there are other theoretical
and practical reasons to expect discrepancies ( 1 1 1 ) .
Changes in Pulmonary Blood Volume
Many different approaches have been used to detect
a change in pulmonary blood volume. They include:
a) lung volumes, b) mechanics of breathing, c) radio-
active tracers, d) teeter board.
lung volumes. In normal subjects the vital capacity
is less in the supine than in the upright position. Al-
though part of this decrease may reflect a change in
the position and tone of the diaphragm (296, 388),
an increase in the pulmonary blood volume also seems
to be involved since measures which interfere with
systemic venous return to the lungs minimize, or
prevent, the decrease in vital capacity (188).
Clinically, a low vital capacity is found in pulmonary
congestion (406). However, in such patients, par-
ticularly if pulmonary venous hypertension has been
prolonged and severe, the lung volumes may be more
encroached upon by pulmonary edema and fibrosis
than by an expanded pulmonary blood volume (238,
378).
mechanics of breathing. Pathologists have long been
aware that the chronically congested lung is a stiff
lung (415). In 1934, Christie and Meakins showed by
measurements of pleural pressure in vivo that the
chronically congested lung requires a greater dis-
tending force than the normal lung (287). Since then,
more elaborate ways of measuring and expressing
pulmonary distensibility, such as "compliance"
(change in lung volume per unit change in pleural
pressure) have come into general use for the study of
both acute and chronic pulmonary congestion; for
the sake of safety and expediency, and at some sacrifice
of accuracy, esophageal pressures have been substi-
tuted for pleural pressures (fig. 26) (287).
The effects of acute pulmonary engorgement on
pulmonary distensibility have been examined in ani-
mals (35, 146) and in man (33, 406). Such studies
have shown that pulmonary venous hypertension has
a considerably greater effect in reducing pulmonary
compliance than does either pulmonary arterial
hypertension or an increase in pulmonary blood flow
(35); moreover, a decrease in vital capacity parallels
a decrease in pulmonary compliance (406). But these
studies have also clarified some of the uncertainties
which attend the use of a change in compliance as a
TIDAL
VOLUME
ml
ESOPHAGEAL PRESSURE
cm H20
200
60
25
20
03
COMPLIANCE
2U
cm H20
WORK AGAINST
NON -ELASTIC
RESISTANCE
(%)
WORK OF
BREATHING
Kq m
lit VE
fig. 26. Comparison of the pulmonary pressure-volume dia-
gram of a normal subject with that of a patient with severe
pulmonary congestion due to mitral stenosis. In the congested
lung, the compliance (AV/AP) is approximately a third of
normal and the resistance to air flow is normal. If pulmonary
congestion is accompanied by edema of the airways ("cardiac
asthma"), both the increased resistance to air flow and the
stiffer lungs contribute to the inordinate work of breathing.
[After Turino & Fishman (406).]
1 694
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
qualitative measure of the state of the pulmonary
blood volume: an increase in pulmonary interstitial
fluid during acute pulmonary venous hypertension
may be indistinguishable from associated increase in
pulmonary blood volume (378); the discrepancies
between esophageal pressure and pleural pressure
are exaggerated in the supine position since medi-
astinal contents compress the esophagus to yield
artificially high values for pleural pressures; changes
in the lung volume may, per se, affect apparent pul-
monary distensibility (287).
Despite the limitations of methodology and the un-
certain distinction between an expanded pulmonary
blood volume on the one hand and its consequences
on the other, pulmonary mechanics in pulmonary con-
gestion continues to attract attention on several physio-
logical accounts. For example, mechanical work and
energy cost of moving congested lungs has proved to
be abnormally high; moreover, in some obscure way,
stiff lungs seem to set the characteristic breathing pat-
tern (rapid frequency, small tidal volume) of pulmo-
nary congestion (177, 406).
radioactive tracers. Change in the radioactivity
of a portion of the lung field after the intravenous
administration of a radioactive tracer has been used
as a measure of the change in pulmonary blood vol-
ume under various experimental conditions (271,
425). The validity of this approach rests heavily on
the assumption that the external detector continues
to survey an unchanged geometry throughout the
control and test periods. It is difficult to prove that
this assumption is fulfilled in experiments which in-
volve either respiratory maneuvers or changes in body
position (425).
miscellaneous. Some experiments require only the
recognition of a change in thoracic (instead of pul-
monary) blood volume. For such experiments, the
critically balanced teeter board has served as a useful
device to detect a shift in the center of gravity of the
body as blood is displaced from one end of the body
to the other (fig. 27) (86, 154, 395). Also, the "cardio-
pneumogram'' has provided an approach to the
changes in thoracic blood volume during each cardiac
cycle (185).
Normal Values for Pulmonary Blood J'olume
It is meaningless to use the central blood volume —
with its vague boundaries and its potential for internal
rearrangement — as a measure of the pulmonary blood
volume as long as the test substance is injected into a
peripheral vein (184, 307). The first step to narrow
the boundaries of the central blood volume was the
pulmonary arterial injection of the test substance
(coupled with peripheral arterial sampling) ; under
these conditions, the central blood volume approxi-
mates 20 to 25 per cent of the total circulating blood
volume (224, 249, 250). The second step was to couple
the pulmonary arterial injection either with sampling
from the left atrium or with the injection of a second
Fig. 27. The teeter board for detect-
ing shifts in regional blood volumes.
The records show that during acute hy-
poxia (.4) the position of the center of
gravity of the body remains unchanged;
on the other hand, during the infusion
of noradrenaline (B), the center of
gravity shifts cephalad. CAL = cali-
bration by placing a 200-gram weight at
the angle of Louis. [After Fritts el al.
(■54)-]
DASH-POT
DAMPER
COUNTER-
WEIGHT
SPRING
rm
KYMOGRAPH
/TmrnyvTWir^^
START
HYPOXIA
t t t
5 MINUTES 20 MINUTES STOP
HYPOXIA
\EIk3
vfy iM^0mmmmmimiiiimmmmmmmn
START
NOR-ADRENALINE
t t
5 MINUTES STOP
NOR-ADRENALINE
DYNAMICS OF PULMONARY CIRCULATION
l69c
tracer into the left atrium. By these techniques, the
pulmonary blood volume is of the order of 10 per
cent of the total circulating blood volume (106, 246,
293)-
It is surprising how closely the latter indicator-dilu-
tion value of 10 per cent in intact man corresponds to
the more direct measurements in animals, i.e., dog,
rabbit, and rat (293). The indicator-dilution value of
10 per cent also coincides with estimates based on
pulmonary vascular dimensions in the dog (169).
Variations in Pulmonary Blood Volume
The pulmonary blood volume increases under a
heterogenous group of conditions (fig. 28) : a) an in-
crease in pulmonary blood flow (224, 250); b) infla-
tion of an antigravity suit (33); c) negative (pleural)
pressure breathing (397) ; d) the assumption of the
supine position (381); e) systemic vasoconstriction
from a variety of causes (64, 154, 186, 372); /) im-
mersion in water (188); g) clamping of the pul-
monary veins (114); and h) left ventricular failure
and mitral stenosis (238).
Conversely, a decrease in pulmonary blood volume
occurs during venesection and reduced cardiac output
(184), positive pressure breathing and the Valsalva
maneuver (44), systemic vasodilatation from warming
(369, 381) and the assumption of the upright posture
(247,381).
Partition of Pulmonary Blood I 'olume
One particularly hazy aspect of the pulmonary cir-
culation is the pattern in which the pulmonary
arteries, capillaries, and veins share the pulmonary
blood volume under natural conditions, and the wax-
in which this pattern is modified either by physiologi-
cal stimuli or by disease. A few beginnings have been
made: anatomical measurements in the dog suggest
that the capacity of the pre- and postcapillary pul-
monary vascular segments is approximately the same
(169); observations on the isolated lung, while failing
to define precise anatomical boundaries, have suc-
ceeded in disclosing how the pulmonary blood volume
may be reapportioned in response to mechanical in-
fluences (124, 315, 317, 324) and to special stimuli
(116, 305). However, there is no obvious way to ap-
ply these experimental observations to the arrange-
ment of the pulmonary blood volume in life.
HEMODYNAMIC INTERRELATIONS
Distensibility and Resistance
In previous sections, pulmonary blood flow, vol-
ume and pressures were considered separately. The
analysis of their interplay is a much more complicated
matter. Generally speaking, the aim of such an
analysis is to relate the static and dynamic properties
of the pulmonary vascular tree to its architecture and
to the structure of its walls. Until recently, investiga-
tors were preoccupied with the model of the pul-
monary circulation which pictured it as the
hemodynamic analog of an electrical d-c circuit and
which viewed the pulmonary blood flow as though it
were continuous and steady (169); for testing this con-
ceptual model, the isolated lung seemed ideal on the
mm
Hg
2
UP
t
LEG RAISING
mmHg MI H
DOWN 80_
'■^ih^^h^Mm^^^
SUPINE EXERCISE
80-
UP DOWN
1, * bdfi^^ttlKittAfttaM a
^^fmMW0^
0-.
fig. 28. Effects of leg raising and supine exercise on pulmonary arterial blood pressure. Leg raising.
In the normal subject (upper left), leg raising is without appreciable effect on the pulmonary arterial
pressure ; in the patient with tight mitral stenosis (upper right), leg raising elicits a considerable in-
crease in pressure. Supine exercise. In normal subject (lower left), exercise increases pulmonary arterial
pressure by a few mm Hg; in the patient with tight mitral stenosis (lower right), the increase in pressure
is much more striking. [After Turino & Fishman (406).]
1696
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
fig. 29. Continuous records of pulmonary arterial pressure
(P) and flow (Q) from a closed-chest, unanesthetized dog.
Blood pressure was recorded through a polyvinyl tube (encased
in nylon) inserted through the wall of the main pulmonary
artery about 1 cm distal to pulmonic valve. Blood flow was re-
corded by an electromagnetic flow meter (modified Kolin
type) placed proximal to the bifurcation of the main pulmo-
nary artery. Tubing and wires placed surgically and led to out-
side between scapulae. For calibration of flow meter, snares
around right and left pulmonary arteries were tightened to
arrest pulmonary arterial flow. (Courtesy of L. Fisher and
D. E. Gregg.)
assumption that it simulated the geometry and dis-
tensibility of the pulmonary circulation in vivo.
Within the last few years, investigators have begun to
take a more realistic view of the pulmonary circula-
tion, recognizing that hemodynamic events within it
vary from instant to instant and that phasic differ-
ences between pressure and flow (fig. 29) are
important; in order to treat these phasic events, they
have resorted to a model based on electrical a-c
theory (71, 177, 438). However, for the moment, this
approach is handicapped by the technical difficulties
of recording pulsatile pulmonary blood flow, espe-
cially in living systems (153, 237).
Distensibility and Capacity:
Pressure- 1 olume Relationships
Because of the manner in which the pulmonary
circulation is incorporated into the lung, the term
"pulmonary vascular distensibility" is a composite
one : it connotes not only the elastic properties of the
vascular walls but also the tone of their smooth
muscle, the perivascular air pressures, the effects of
hidden forces such as alveolar surface tension (78,
312), the presence of excessive interstitial fluid (239),
and the mechanical distortions of adjacent pulmonary
tissues (146). As in the systemic circulation, the dis-
tensibility characteristics are customarily expressed
as the change in vascular volume per unit change in
transmural pressure. However, in contrast to the
systemic circulation, the small precapillary vessels
are thin-walled and easily distensible, thereby con-
tributing to the pressure-volume characteristics of the
pulmonary arterial tree (350). This participation of
the pulmonary '"resistance" vessels in the "capaci-
tance" function of the pulmonary circulation is of
hemodynamic significance: for example, without
pulmonary arteriolar sphincters, a larger fraction of
the right ventricular stroke volume is apt to escape
from the pulmonary arterial tree during and just
after each systole than from the systemic arterial tree
(240); also, during bradycardia the pulmonary ar-
terial pressure may fall to the level of pulmonary
venous pressures (187).
The distensibility characteristics of the pulmonary
circulation, and of its individual segments, have been
determined under a wide variety of experimental
conditions, using greatly different types of prepara-
tions. These studies have led to certain generaliza-
tions: a) the pressure-volume characteristics of the
entire vascular tree (fig. 30) and of the large pul-
monary vessels resemble those of a large systemic
vein (148, 211,
290); b) as in other distensible
structures, the blood pressure at any volume is higher
when the system is being filled than when it is being
emptied ("hysteresis," "delayed compliance," "stress-
50--
fig. 30. Pressure-volume relationship of the pulmonary
vascular bed in the dog. To construct this curve, blood was
withdrawn at 10-sec intervals after initially elevating pressure
in the system to approximately 60 mm Hg. [After Sarnoff &
Berglund (371 ).]
DYNAMICS OF PULMONARY CIRCULATION 1 697
relaxation") (318, 337, 371); c) the pulmonary venous-
left atrial segment is less distensible than the systemic
venous-right atrial segment (87, 272, 309); d) the
successive segments of the pulmonary vascular tree
differ considerably in distensibility [the veins and
arteries are more distensible than the capillaries ( 1 24,
318)]; and e) although the aorta and pulmonary
artery are of approximately the same caliber in life,
the range of maximum distensibility for the pul-
monary artery (10 to 40 mm Hg) is much lower than
for the aorta (182). Unfortunately, measurements of
pulmonary vascular distensibility in intact animal or
man have not yet become practical or reliable (71,
293)-
These generalizations about pulmonary vascular
distensibility help to explain some physiological
features of the pulmonary circulation. For example,
the small pulse pressure in the pulmonary artery
seems to be a consequence of both the marked distensi-
bility of the pulmonary arterial tree, which prevents
a considerable rise in pressure as the right ventric-
ular stroke volume is ejected, and the low pul-
monary vascular resistance, which allows more blood
to escape from the pulmonary arterial tree during
each systole and causes the pressure to fall earlier
during systole (102, 182). The greater distensibility
of the pulmonary than the systemic arterial tree also
helps to account for the slower velocity of the pulse
wave in the pulmonary artery (250 cm/sec) than in
the aorta (300 cm/sec).
The unusual distensibility of the small pulmonary
vessels, i.e., of the pulmonary "resistance" vessels
affects their hemodynamic behavior. For example,
as the pulmonary blood volume is expanded (251,
437), small pulmonary vessels share in this increase,
leading to an increase in their transmural pressures,
passive dilatation of their lumens and a decrease in
their resistance to blood flow; since the arterial,
capillary and venous portions of the small pulmonary
vessels have different capacities and pressure-volume
characteristics, the increase in pulmonary blood vol-
ume will not be equally apportioned among these
vascular segments. Moreover, the distensibility char-
acteristics and capacities are such that at low pul-
monary vascular volumes and pressures, each in-
crement in blood volume will raise the blood pressure
less, and passively dilate the vessels more, than at
high levels. This hemodynamic behavior is particu-
larly relevant to those experiments in which an under-
standing of the passive characteristics of the
pulmonary vascular tree and of its segments is pre-
requisite for interpreting a change in calculated pul-
monary vascular resistance in terms of pulmonarv
vasomotor activity (69, 101).
Resistance: Pressure-Flow Relations/u/n
It has been noted above that, for the sake of ex-
pediency, flow through the pulmonary circulation is
conventionally treated as though it were steady. Ac-
cordingly, and by analogy with Ohm's law, the ratio
of the drop in mean pressure across the pulmonarv
circulation (AP) to the mean blood flow (Q) is used
as a measure of pulmonary vascular resistance. This
idea of resistance is unambiguous when applied to
rigid tubes perfused by a homogeneous fluid flowing
in a laminar stream: under these special conditions,
the plot of AP against Q, is linear and passes through
the origin, i.e., it is predictable and interpretable in
physical terms. Complexities are introduced when
these concepts are extended to the pulmonary (as
well as to the systemic) circulation: the vascular bed
is a nonlinear, visco-elastic, frequency-dependent
system perfused by a complicated non-Xewtonian
fluid; moreover, the flow is pulsatile so that inertial
factors, reflected waves, pulse wave velocity, and in-
terconversions of energy become relevant considera-
tions (156, 277). In such a system, resistance varies
with pressure and flow; plots of AP against Q, are not
linear and do not pass through the origin (61, 169,
175). And, as the result of the many different active
and passive influences which may affect the relation-
ship between AP and Q, the term "resistance" is
bereft of its original physical meaning: instead of
representing a fixed attribute of blood vessels, it has
assumed physiological meaning as a product of a set
of circumstances.
table i. Representative Values for a Normal Human
Subject in the Basal State
Pulmonary blood flow 6.0 liters/min
3.1 liters/min/m2 BSA
Pulmonary blood pressures
s/d.m
Right atrium
3/2. 2
mm Hg
Right ventricle
20/0
mm Hg
Pulmonary artery
20/9,15
mm Hg
Pulmonary wedge
6
mm Hg
Left atrium
7/3.5
mm Hg
Pulmonary vascular
O.I*
R units
resistance
* Calculated from the data in this table: R = (15 — 5)/
(6000/60) = 0.1 R units. t R units express calculated
resistance as mm Hg/(ml/sec); to convert to C.G.S. units
(dynes sec cm-5), the value in R units is multiplied by 1328.
i6g8
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Meaning of Pulmonary Vascular Resistance
Generally speaking, the pulmonary circulation —
which receives the same blood flow as the systemic
circulation at one-fifth the blood pressure — is a low-
resistance circuit. In the normal pulmonary circula-
tion, the pulmonary vascular resistance is ordinarily
of the order of o. i to 0.3 R units (table 1). But in
evaluating data for resistance, at least three separate
problems are involved: /) the precise measurement
of the parameters invoked in the equation for re-
sistance, i.e., pressure drop across the pulmonary
vascular bed divided by the rate of pulmonary blood
flow; 2) the decision as to whether a change in calcu-
lated resistance means a change in pulmonary vascu-
lar caliber; and 3) the interpretation of a change in
caliber in terms of the mechanism which effected it,
i.e., vasomotor or passive (61, 132).
With respect to the values substituted in the equa-
tion for resistance, it is self-evident that calculations
of pulmonary vascular resistance, which are to be
meaningful in vasomotor terms or even in terms of
vascular caliber, presuppose accurate measurements
of blood flow and pressures. Under certain stressful
conditions, such as exercise, acute hypoxia, and acute
hypercapnia, heightened respiratory excursions com-
plicate the precise measurement of pressures, and pul-
monary blood flow is easily miscalculated. Moreover,
Permutt and co-workers have recently likened the
pulmonary vessels to a series of Starling valves and
warned against blind faith in the left atrial (or pul-
monary venous) pressure as a measure of pulmonary
outflow pressure. In particular, they have stressed
that any situation in which alveolar pressure exceeds
pulmonary venous pressure, by creating a discon-
tinuity in pressure between the capillaries and the
pulmonary veins, invalidates the use of the pulmonary
venous pressure for the calculation of total pulmonary
vascular resistance (315, 354). Accordingly, just as the
studies of West et al. (427) suggest a spectrum of
ventilation-perfusion relationships in the lung of up-
right man, the model of Permutt et al. (unpublished
observations) suggests a distribution of the de-
terminants of resistance to perfusion, depending on
the relationships of pulmonary arterial, left atrial, and
alveolar pressures in the different parts of the upright
lung. The precise relationships between the normal
imbalances between ventilation and perfusion on the
one hand, and the interplay of alveolar and pul-
monary vascular pressure on the other, remain to be
elucidated.
With respect to the second problem, i.e., the equa-
tion of a change in calculated pulmonary vascular
resistance to a change in pulmonary vascular calibers,
there are at least two different types of enigmas. One is
the possibility that a change in "anomalous viscosity,"
which is customarily disregarded, may masquerade
as a change in caliber (197, 260, 430); since this
source of confusion is most apt to become appreciable
when pulmonary blood flow drops to exceedingly low
levels, the practice of ignoring it seems reasonable as
long as levels of pulmonary blood flow are of the same
order of magnitude as that ordinarily encountered
in vivo. The other is the equivocal anatomical mean-
ing of a change in caliber, since a change in geometry
may arise not only from a change in the diameters of
patent vessels but also a change in the number of
parallel paths which are being perfused (267).
Finally, before pulmonary vasomotricity can be
invoked, it is axiomatic that all conceivable passive
mechanisms for affecting vascular calibers (table 2)
be taken into full account. One such passive mecha-
nism, particularly likely during artificial ventilation,
is the mechanical distortion of the vessels by adjacent
lung tissue at abnormal lung volumes (397)- Another,
more universal, source of confusion is an undetected
change in transmural pressure operating subtly to
table 2. Factors Conceivably Involved in a Change
in Pulmonary Vascular Resistance
MECHANICAL (PASSIVE)
Passive cardiocirculatory effects
1 . Back pressure from left atrium or pulmonary veins
2. Change in pulmonary blood flow
3. Change in pulmonary blood volume
4. Bronchial collateral circulation
a) Nutrition of nerves, ganglia, and smooth muscle
b) Patency of collateral circulation
Passive respiratory effects
1 . Change in alveolar pressures
a) Tone of bronchial smooth muscles
bl Secretions of bronchial glands
c) Alveolar surface tensions
2. Change in intrathoracic pressures
3. Tone of interstitial smooth muscle
4. Pericapillary edema
VASOMOTOR (ACTIVE)
Originating from without the lungs
1. Autonomic nervous system (including systemic chemo-
receptors)
2. Catecholamines
Originating within the lungs
1. "Critical" closure of small muscular vessels
2. Intravascular chemoreceptors
3. Chemical stimuli (directly on vascular muscle)
4. Deranged vascular metabolism
5. Local reflexes
DYNAMICS OF PULMONARY CIRCULATION
1699
200-
150-
P vs R
Q vs R
R mmHg/ml/min
-I 1 1-
FiG. 31. Passive changes in pulmonary' vascular resistance
(R) at different pulmonary arterial pressures (P) and at dif-
ferent pulmonary blood flows (Q). Pulmonary venous pressure
remains constant throughout. As How and pressure decrease,
resistance increases. [Based on Edwards (119).]
which would be expected to obtain were it not for the
stimulus (fig. 32); and b) the continuous registration
of the pressure gradient across the pulmonary vascu-
lar tree and of the systemic blood pressure, before
and after the injection of a pharmacological agent
into the pulmonary artery (fig. 33).
The use of pressure-flow points to recognize vaso-
motor activity requires that the passive pressure gra-
dient-flow relations be known or predictable. It is
difficult to compare the published relationships in the
pulmonary with those in the systemic circuit because
conventionally the data do not cover the same range.
Pressure-flow plots for systemic beds include zero
pressure and zero flow, while the conventional presen-
tation of pulmonary data start with "normal" pressure
and flow and plot the fractional excess of one against
the fractional excess of the other. Qualitatively, the
modify vascular caliber and resistance. For example,
an increase in transmural pressure — arising from an
increase in either pulmonary- arterial or left atrial
pressure — passively widens the vessels and decreases
their resistance (fig. 31); conversely, a drop in trans-
mural pressure increases vascular resistance (36, 69,
366). Consequently, a change in resistance is not a
reliable sign of pulmonary vasomotricity when trans-
mural pressures change. Indeed, at different levels of
transmural pressure, calculated resistance may re-
main unaltered even though pulmonary vasomotor
tone has altered considerably (61). Considerations
such as these have had two major effects on experi-
mental design and interpretation: a) many have
urged that the use of ohmic resistance be abandoned
in favor of more straightforward presentation of
pressures and the corresponding flows, and b) others
have insisted on stringent experimental criteria, such
as constant flow (fig. 31 ), left atrial, alveolar and intra-
pleural pressures before interpreting a change in
pulmonary arterial pressure.
Practical Recognition of Pulmonary Vasomotricity
Dissatisfaction with the use of calculated resistance
(a ratio) to detect an active change in vascular caliber
has encouraged the use of graphic representations
which relate blood flow to the pressure gradient that
effects it (70). For example, the recognition of pul-
monary vasomotor activity has been attempted bv:
a) the comparison of experimentally determined pul-
monary vascular pressure-flow points, obtained after
applying a stimulus, with the pressure-flow curve
20 --
10 --
2 I % 0 .
0 HI (-
3.5
4.5
5.5
0.3
0. I --
E /
12
•/. 0. ©
EXERCISE
21 V. 0.
3.5
4.5
Lit/min
55
fig. 32. Detection of a decrease in pulmonary vascular cali-
ber from pulmonary arterial flow-pressure curves and from
pulmonary vascular flow-resistance curves. For these curves,
mild exercise was used to increase pulmonary blood flow pas-
sively and acute hypoxia was used as the test stimulus. A. Dur-
ing exercise, pulmonary arterial pressure increased as blood
flow increased; during acute hypoxia, an equivalent increase
in pressure occurred without an appreciable increment in
blood flow. B. During exercise, calculated resistance decreased;
conversely, during acute hypoxia, calculated resistance in-
creased even though blood flow (and presumably all
other respiratory and circulatory parameters) remain un-
changed. [Based on Fishman et al. (132).]
1700
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
PA
MM HG
20
LP
10 SEC
M$tm
sxmmmmm
fig. 33. Blood pressures in the pulmonary artery, pulmonary vein, and aorta following the in-
jection of 5 mg of acetylcholine into the pulmonary artery of a human subject during open thora-
cotomy. The time of injection is indicated by the upright arrow. (Unpublished observations of A. G.
Jameson and A. P. Fishman.)
lower portion of the pulmonary arterial pressure-flow
curve in the rabbit (416), the dog (119, 252, 434),
and man (89) resembles that of systemic beds: an
increase in pressure is associated with a parabolic
increase in flow; the inscribed curve is convex to the
pressure axis. The upper portion of the pulmonary
plot shows an opposite inflection which does not ap-
pear in systemic beds. Quantitatively, the pulmonary
and systemic arterial curves differ not only in the level
of the arterial pressure but also in the large increments
in blood flow evoked by slight increments in pul-
monary arterial pressure (at constant left atrial pres-
sure).
Curves depicting the relationship between the
driving pressures and flow are difficult to establish
for either intact dog or man since it is impractical to
increase pulmonary blood flow without simultaneously
modifying the behavior of the respiration, the heart,
and the systemic circulation. However, Lategola has
succeeded in drawing a passive pressure-flow curve
for the pulmonary vascular tree of the intact dog,
using values obtained in the course of graded occlusion
of the pulmonary arterial tree by balloon-tipped
catheters (252). This curve appears as the solid line
in figure 34. The shape of this curve is generally in-
terpreted as showing that: a) as flow increases, re-
sistance decreases; and b) beyond a transition phase
(AQ of approximately 250 per cent), resistance be-
comes constant. Moreover, the length of the gently
sloping portion of the curve is regarded as a measure
of the maximum calibers, both of the patent vessels
and of those available to open in parallel; the start
of the steeply ascending portion is thought to occur
when the system begins to behave as though it were
comprised of rigid tubes (89, 252). It should be noted
that while the general shape of the pressure-flow rela-
tionship seems beyond cavil, the precise levels of flow
at which the tubes appear to become rigid are not as
300-
200-
<
<
100-
0 -
3o
A 6 %
1 1
100
200
300
fig. 34. Relationship between pulmonary blood flow' and
pulmonary arterial pressure in dog and man. Note that the
origin represents normal or control levels (not zero levels) of
both pressure and flow. The line is redrawn after Lategola
(•252) and is based on data obtained during graded occlusion
of the pulmonary artery tree in the dog. The shaded area rep-
resents corresponding measurements in normal man during
balloon occlusion of one pulmonary artery both at rest and
during mild exercise (42). The individual points represent
observations on human subjects during supine exercise. Open
circles: mild exercise (382); solid triangles: graded exercise (149);
open triangles: mild exercise after pneumonectomy (89).
convincingly established (281) and the final slope
must be considered in the assessment of constancy of
resistance.
Superimposed on the pressure-flow curve of the
dog is a shaded envelope which includes the points
obtained during similar occlusion of a pulmonary
arterv in man (42); in order to exceed the increments
in blood flow obtainable at rest, the human subjects
performed mild leg exercise during the occlusion of
one pulmonary artery. It may be seen that the en-
velope of human points closely follows the horizontal
DYNAMICS OF PULMONARY CIRCULATION
I 70I
portion of the dog's pressure-flow curve; un-
fortunately, in this study, sufficiently high flows to
define the steep portion of the curve were not
achieved. However, the original measurements by
Cournand and co-workers on human subjects after
pneumonectomy (open triangles) suggest that the rest
of the human pressure-flow curve may also resemble
that of the dog (89). More observations in both man
and dog at higher levels of flow are obviously needed;
unfortunately, patients with congenital left-to-right
shunts, who may, from a priori considerations of their
large pulmonary blood flows, appear to be logical
candidates for such measurements, are usually found
to be unsuitable for pressure-flow curves because of
complicating pulmonarv vascular disease and ana-
tomical defects which preclude precise measurements
of pulmonary blood flow.
There are three interesting side lights to the curve
illustrated in figure 34. The first is the difference
between this parabolic curve of the normal subject
and the linear relationship between pressure and
flow which has been described for patients with ab-
normal vascular beds (132); this difference suggests
that those animal or isolated-lung experiments which
find a linear relationship between pulmonary arterial
pressure and flow may be dealing with abnormal, or
overfilled, lungs (132). The second is the relationship
between the sharp inflection of the curve and the
maximum diffusing capacity; it has yet to be estab-
lished whether maximal dilatation of the pulmonary
capillary bed, i.e., the achievement of the maximum
diffusing capacity, coincides with the abrupt increase
in the pulmonary arterial pressure (267, 407). The
third deals with the use of graded exercise to construct
the pressure-flow curve in intact animal or man. It
may be seen that during mild to moderate exercise in
man, the pressure-flow points overlap those obtained
during graded occlusion in the dog; during heavier
exercise, the coincidence of human and animal points
is not as exact. These discrepancies raise the possi-
bility that strenuous exertion may sufficiently alter
passive determinants, i.e., transmural pressures and
left atrial pressure, to invalidate the use of such exer-
cise for the construction of a reference curve which is
supposed only to depict the uncomplicated conse-
quence of increasing flow on pressure (132). On the
other hand, the use of mild to moderate exercise for
this purpose seems valid on several scores: a) the
mean left atrial pressure (104) and mean pleural
pressures are little affected by light exercise (132),
b) the pressure-flow curses obtained during light
exercise and the passive curves obtained from iso-
lated lungs are quite similar (119, 416), and c) the
exercise points correspond to those obtained during
graded occlusion of the pulmonary artery (42, 53,
101).
The second way of identifying pulmonary vasocon-
striction is particularly applicable to the use of
pharmacological agents; it has the advantage of circum-
venting many of the restrictions outlined for steady-
state measurements. It involves (fig. 33) the single
injection of a pharmacological agent into the pul-
monary circulation of the intact animal or man and
the continuous registration of the pressure drop across
the lungs, the heart rate, and the systemic blood pres-
sure during the single pulmonary circulation, i.e.,
before recirculation. In this way, the effect of the
agent appears as a change in pulmonary arterial
pressure before flow can change and before the agent
can affect the systemic circulation (187). An alternate
way of accomplishing the same end for steady-state
experiments is the continuous infusion of an agent,
e.g., acetylcholine (192, 441) which is destroyed
within the lungs during the course of a single circula-
tion.
Blood F/ozv Through Each Lung Separately
After application of a unilateral stimulus, such as
hypoxia (135), or the unilateral infusion of acetylcho-
line (89), the partition of flow between the two lungs
is a measure of the relative resistances of the two sides
since the pressure gradient across the lungs is identical
on the two sides. Although the idea of using one lung
in this way, as a control for the other, is intuitively-
attractive, the experiments are generally technically
difficult, particularly if bronchospirometry is involved.
Critical Closure
Small muscular blood vessels of the systemic circu-
lation are believed to be inherently unstable so that
they are inclined to spring shut — concentrically and
completely — when their intraluminal pressure drops
below a critical value. This "critical closing pressure'1
has been proposed as a measure of the tone of vascular
smooth muscle, i.e., of vasomotor activity: the level
of the "critical closing pressure" increases as wall
tension increases and as the size of the vessel decreases.
Critical closing pressure is manifested experimentally
by the arrest of flow despite an appreciable perfusion
pressure. By similar reasoning, the muscular small
vessels spring open when "critical opening pressures"
are exceeded (60, 165).
1 702
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
The concept of critical opening and closure has
been invoked to account for certain puzzling responses
of the pulmonary circulation (429). These include the
exceedingly gradual increase in pulmonary arterial
pressure during graded exercise (267), the relative
stability of the pulmonary arterial blood pressure
during hemorrhage (160), and the pressure gradient
between the pulmonary artery and left atrium as the
left atrium falls below 7 mm Hg (36, 366).
However, it is more difficult to prove the operation
of critical opening and closure of small vessels in the
pulmonary, than in the systemic, circulation. The
difficulties are of several different kinds: a) mechanical
influences, e.g., local changes in transmural pressure,
may open and close vessels independent of vasomotor
activity; b) the effects of anomalous viscosity are apt
to be more pronounced and to simulate changes in
vascular calibers in a low-pressure circulation; c) the
pulmonary arterioles are thin-walled, wide-lumened
and, in general, poorly constructed to spring shut;
d) there are generally alternate, and equally con-
vincing mechanisms to account for pulmonary vascu-
lar behavior (272); and e) experiments specifically
designed to look for signs of critical closure have not
always been able to find them (273, 434).
At present, the experimental evidence for critical
opening and closure of small pulmonary vessels — a
vasomotor phenomenon — is inconclusive. If the phe-
nomenon does occur, it seems to do so when the
pressure gradient between the pulmonary artery and
left atrium is exceedingly low, i.e., of the order of 1
to 2 mm Hg (42); moreover, it does not seem to affect
equally all small vessels of comparable dimensions
(273, 366). In general, transmural pressures are more
apt to be involved in the closure and opening of small
pulmonary vessels than is vasomotor activity. It
would be of interest to examine such closed pul-
monary vessels to see if their lumens are slits (mechani-
cal collapse) or circles (concentric obliteration by
vasoconstriction ) .
Potential and Kinetic Energy
Mechanical energy is imparted by the right ven-
tricle to the blood perfusing the pulmonary circulation
in two forms, kinetic and potential energy. At rest,
the kinetic energy factor in the pulmonary circulation
is of the order of 10 per cent or less of the total; on
the other hand, both in normal subjects during exer-
cise and in patients with pulmonic stenosis or left-
to-right shunts, the kinetic energy factor may increase
to over 50 per cent of the total (320, 338).
The usual calculation of resistance deals only with
the drop in potential energy (pressure) across the
system. It does not take into account the fact that
as blood courses down the pulmonary vascular tree,
part of the kinetic energy is reconverted to pressure
energy as the area of the bed increases; a small frac-
tion is dissipated as heat arising from the friction of
blood flow (20). In experimental pulmonic valvular
insufficiency, the unusually rapid blood flow and
turbulence in the pulmonary artery may produce a
drop in pressure across the pulmonic valve (123).
These considerations suggest that at rest, when the
kinetic energy factor is small and of the same order of
magnitude in the pulmonary arteries and veins, the
drop in potential energy (pressure) between the
pulmonary arteries and veins provides a rough meas-
ure of the decrease in mechanical energy across the
pulmonary vascular bed; on the other hand, in
normal subjects during exercise, and in patients with
cardiac abnormalities characterized by large stroke
volumes and rapid rates of pulmonary blood flow,
the pressure gradient across the pulmonary circula-
tion does not provide an adequate measure of the
mechanical energy delivered to the system.
PULMONARY CAPILLARY CIRCULATION
Pulmonary Capillary Pressure (Pc)
Since a direct method for measuring Pc pressures
is not available, the level of the Pc pressure is generally
estimated from the pulmonary arterial diastolic
pressure on the one hand, and the mean left atrial
pressure, on the other. In the normal subject these
limits set the mean Pc pressure at approximately 10
mm Hg.
Rate of Pulmonary Capillary Blood Flow (Q,c)
In the normal animal or man the rate of pulmonary
capillary blood flow is virtually identical with the
right ventricular output; in left-to-right shunts or
extensive collateral arterial circulations, Qc exceeds
the right ventricular output. An earlier chapter has
analyzed the methods used to measure the cardiac
output. Of special interest to the present section is
the use of inert soluble gases not only to measure the
rate of pulmonary capillary blood flow in man but
also to explore the nature of the pulmonary capillary
flow. Throughout this section it will be assumed that
physiological measurements of pulmonary capillary
DYNAMICS OF PULMONARY CIRCULATION
'7°3
flow need not be measuring only the flow through
anatomic pulmonary capillaries. The physiologic
measurements may also be including the flow through
other small pulmonarv vessels that participate in the
uptake of the inert gas. However, this distinction be-
tween the anatomic and the physiologic pulmonary
capillary is more meaningful with respect to relating
the gas-exchanging characteristics of the small pul-
monary vessels to their hemodynamic behavior than
with respect to the measurement of the cardiac
output.
The principle underlying the use of inert gases to
measure pulmonary blood flow was enunciated by
Bornstein in 1910 (343). Unfortunately, he chose an
insoluble gas, i.e., nitrogen, as the test gas. In 191 2,
Krogh & Lindhard (240) substituted the soluble
inert gas, nitrous oxide, for nitrogen and devised an
experimental protocol, involving respiratory maneu-
vers, to obtain the values needed for the equation,
Qc = vw>n,,o-FaN!0
in which Qc = pulmonary capillary blood flow
per minute
VN;0 = volume of N>0 absorbed per
minute (BTPS)
XN2O = Ostwald's coefficient of solubility
of nitrous oxide in blood at 37 C
FaN20 = mean fraction of N2O in alveolar
gas during the test (BTPS).
Since the coefficient of solubility (X) is constant,
the variables involved in the calculation of the flow
are two: /) the volume of N20 absorbed per minute
(VN!o); and 2) the mean alveolar fraction of N2O
during the test (Fav,,,)- Subsequently, it was shown
that there are several practical limitations to the
Krogh and Lindhard method; these include: a) the
need to complete the test before recirculation of the
test gas; in normal man the pulmonary recirculation
time is of the order of 1 1 ± 3 sec (74, 343); b) the
difficulty in obtaining simultaneous measurements of
the different variables involved in the equation; c)
the dilemma of distinguishing the uptake of the gas
by the pulmonary tissues from the uptake by the
pulmonary capillary blood; and d) the unsubstanti-
ated use of "correction factors" (273, 343). Despite
these reservations, the inert-gas methods do provide
accurate measurements of pulmonary capillary blood
flow (Qc) in resting subjects if proper precautions
are taken. However, during exercise and in chronic
pulmonary disease they become less reliable. The
current consensus appears to be that despite the
attractive simplicity of these tests, their most reliable
use, even in resting patients with normal lungs, is for
consecutive measurements of Qc.
Interest in the use of soluble, inert gases to measure
pulmonary capillary blood flow lagged once the direct
Fick and Stewart-Hamilton methods were stand-
ardized into clinically useful techniques. However, it
revived when Lee and DuBois substituted the body
plethysmograph for the spirometer to measure the
rate of uptake of nitrous oxide: this ingenious mod-
ification of the Krogh and Lindhard method promised
not only to provide the usual measure of the rate of
capillary blood flow per minute but also of the rate
of flow at any instant (254). Unfortunately, there are
practical difficulties inherent in the use of the bodv
plethysmograph for the measurement of instantaneous
flow. These limitations have led to the development
of modified plethysmography: techniques in man
(37, 418) and a modified cardiopneumographic
method in the dog (185, 298).
A ature of Pulmonarv Capillary Blood Flow
Whether pulmonary capillary flow is steady or
pulsatile is critical for the understanding of both
pulmonary hemodynamics and gas exchange (275,
319). For example, if the linear velocity of the blood
flow through the alveolar capillaries were to vary
during the cardiac cycle without compensatory
changes in other parameters, e.g., diffusing capacity
and capillary blood volume, the equilibration between
alveolar gas and capillary blood might well be dis-
turbed (143).
A standard of reference for assessing the nature of
the pulmonary capillary blood flow is the prevalent
idea that blood flow through the systemic capillaries
is ordinarily continuous and devoid of major oscilla-
tions from the mean. This idea is consistent with two
features of the systemic circulation: a) the interplay
between arterial distensibility and arteriolar re-
sistance, so that the systemic "windkessel" maintains
flow during diastole; and b) the varying path lengths
between the root of the aorta and the capillaries. Of
these two influences, the windkessel effect is held to
be the more important.
It is much more difficult to predict the nature of
the pulmonary capillary flow. On the one hand,
marked surges of capillary flow following systole
(pulsatile flow) might be expected to occur on at
least two accounts: /) the relatively small capacity of
the pulmonary arterial tree as compared to the
systemic arterial tree; and 2) the relatively low re-
1 704
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
sistance of the small precapillary vessels (102). On
the other hand is the fact that, somewhere en route to
the pulmonary veins, the pulmonary arterial pressure
pulse is damped out so that the pulmonary venous
pressure pulse ordinarily reflects only left atrial
events. The problem then devolves into deciding the
degree to which the flow pattern in the pulmonary
capillaries resembles the pattern of instant-to-instant
changes in the pulmonary vascular pressure gradient
and of flow in the pulmonary artery.
If it is assumed that the pattern of pulmonary
capillary blood flow is uniform throughout the lung,
direct inspection of the surface capillaries of the lung
should afford some insight into the nature of the
pulmonary capillary flow. In 1733, while examining
the exposed frog lung, Stephen Hales observed that
not only was blood "sensibly accelerated at each
svstole in the finest capillaries, but also in their cor-
responding capillary veins, tho' not in their larger
trunks" (178). This was the first declaration that
pulmonary capillary flow — at least in the frog — was
pulsatile. However, 200 years later, observations on
transilluminated lung of the cat indicated that the
pattern of capillary flow in the mammalian lung was
distinctly different from that of the amphibian lung;
thus, instead of pulsatility, YVearn and co-workers
stressed intermittency, a phenomenon attributable to
the opening and closing of pulmonary precapillary
vessels (419). The rarity of pulsatile flow in the sur-
face capillaries of the mammalian lung has since been
confirmed by others (21).
Opposed to these direct observations on the mam-
malian lung are the results obtained by plethysmo-
graphic techniques in man (37, 254, 418). Not only
do they picture pulmonary capillary flow as regularly
and vigorously pulsatile, but they also depict a flow
pattern in the pulmonary capillaries which corre-
sponds more closely to the instantaneous changes in
the blood pressure gradient across the lungs than to
the flow pulse in the pulmonary artery (fig. 29)
(270). Moreover, unless one postulates species
differences among mammals, these results in man
also challenge the notion that examination of the
surface capillaries of the lung is a useful index of the
pattern of flow in the bulk of the pulmonary capil-
laries.
Even though the majority of the plethysmographic
studies agree that pulmonary capillary blood flow is
pulsatile in man, they are not entirely consistent
with respect to the form of the capillary flow pulse.
For example, not only do the published records
differ with respect to the amplitude of peak flow,
but they also display different contours for the flow
pulse: some, but not all, the records indicate that
flow is interrupted for much of each cardiac cycle;
often, pulmonary capillary blood flow seems to
reverse; finally, the reported flow patterns generally
vary from beat-to-beat. Undoubtedly, at least part of
this variability is attributable to practical difficulties
inherent in the plethysmographic techniques (342).
In addition, as may be gleaned from figure 35, in-
accuracies are inescapable in the matching and
analvsis of the air and nitrous oxide records.
ECG
AIR
©
-kv_X
feo%N20 -vi
(20% 02 \/*S^
4.5 —
1.5 —
D1
18 —
12
6 —
fig. 35. The pattern of the pulmonary capillary blood flow
according to the pncumocardiographic (A) and plethysmo-
graphic (B and C) methods. From above downward, the elec-
trocardiogram (ECG), the air record (AIR), the nitrous oxide
record (8o*>c NjO, 20% O.), the difference between the air
record and the nitrous oxide record (VN„0), and the rate of
pulmonary capillary blood flow (Qc) in liters per minute. A:
actual pneumocardiograms obtained from an anesthetized,
curarized dog during arrested respiration. Mechanical events
are eliminated in the process of point-by-point subtraction of
the nitrous oxide pneumocardiogram from the air pneumo-
cardiogram, yielding a record of instantaneous changes in air-
way pressures due only to the volumes of nitrous oxide removed
by the perfusing blood (VN-,0). Differentiation of the volume-
uptake record provides a record of the rate of uptake of nitrous
oxide and, therefore, of the pulmonary capillary blood flow
(Qc)- Except for the unexplained dip in the Qc record, pulmo-
nary capillary flow appears to be continuous and largely con-
fined to the systolic portion of the cardiac cycle. B and C:
hypothetical plethysmographic records to compare strongly
pulsatile but continuous capillary flow (B) with weakly pulsa-
tile but continuous capillary flow (C). Not illustrated is the
possibility of interrupted capillary flow, i.e., that capillary flow-
may actually stop (drop to zero) for part of each cardiac cycle.
[After Morkin el al. (298).]
DYNAMICS OF PULMONARY CIRCULATION
'7°5
Size o] Pulmonary Capillary Bed
Although the extent of the pulmonary capillary
bed in a living subject cannot be expressed in absolute
units, a change in area can be detected from con-
secutive measurements of the pulmonary diffusing
capacity. Such measurements, using either oxygen or
carbon monoxide as the test gas have shown that:
a) not all of the available area is in use at rest; and
b) the capillary area involved in gas exchange in-
creases progressively under a variety of circum-
stances, e.g., exercise. The major mechanism involved
in increasing the area is an increase in transmural
pressure (357). The precise way in which this distend-
ing pressure is increased varies from circumstance to
circumstance. Thus, in some conditions, an increase
in capillary blood volume is involved; in others, the
perivascular pressures may decrease; at high lung
volumes, the capillaries may even be passively-
stretched. As expected from theoretical considera-
tions, the diffusing capacity is little affected by
changes in pulmonary blood flow; only when pul-
monary blood flow is severely curtailed does the
diffusing capacity decrease (358, 407). Pneumo-
nectomy generally (143), but not invariably (89) de-
creases the diffusing capacity.
The maximum diffusing capacity is of interest as
a measure of the maximum available pulmonary
capillary area. It has been suggested that this max-
imal value is reached at a level of blood flow which
corresponds to the steep inflection of the flow-pressure
curve (fig. 34). However, the experimental support
for this hypothesis is inconclusive (348).
Pulmonary Capillary Blood Volume (Q.,.)
On the basis of measured differences between the
rates of reaction of carbon monoxide with hemoglobin
solutions at different oxygen tensions, and the calcu-
lation of the average time spent by blood in traversing
the pulmonary capillaries, the volume of blood in the
pulmonary capillaries of the normal resting subject
was originally calculated to be of the order of 60 to
75 ml (364, 365); during severe exercise, Qc increased
somewhat (to approximately 90 to 100 ml) (223,
364). More recent measurements and calculations of
the same type have raised the value of the resting
Qc to approximately 100 ml, both for the normal
subject (15, 244) and for the patient with mitral
stenosis (15). At present, there is no way to decide
how much of the reported variability is artificial
(143, 365) and how much is a consequence of either
\
1.5 en
O to
_1 0:
- 200
TTTTTTTfTTf
PRE CAPILLARY
1 T 1 1 1 1 1 1 1
POSTCAPILLAR Y
fig. 36. Hypothetical relationship between blood pressure
(solid line,) and cross-sectional area (shaded) in the pulmonary
vascular bed of the dog. The blood pressure is represented as
per cent of initial value. Vascular diameters at key points are
indicated by asterisks. According to this schema, the major drop
in blood pressure occurs in the region of the pulmonary capil-
laries (shaded spike). [Based on Schleier (373).]
biological differences between subjects or the effects
of the breathing maneuvers which are part of the
tests (244). A variety of agents and procedures seem
to be capable of passively altering Qc (262).
In the dog, anatomical measurements suggest that
the pulmonary capillaries may contain 10 per cent
of the total volume of blood between the right ven-
tricle and left atrium, i.e., of the order of 1.2 per cent
of the total circulating blood volume (169). For the
human lung, Weibel found that Qc varies with the
lung volume and with the degree of capillary filling.
In his preparation, which involved negative (pleural)
pressure inflation of the lung and fixation in formalin
vapor, Qc was 150 to 200 ml, i.e., approximately
twice the volume obtained by physiological measure-
ments. Part of the difference between the anatomical
and physiological measurements may be the degree of
inflation of the lung (422).
Resistance and Distensibility
How much of the pulmonary vascular resistance to
perfusion lies in the pulmonary capillary bed is a
matter of opinion (66). The prevalent notion is that,
under ordinary conditions, the resistance function
resides in the small, muscular precapillary vessels.
On the other hand, calculations of resistance based
on anatomical measurements and assumptions have
raised the possibility that a large part (up to half) of
1706
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
the pressure drop across the pulmonary vascular bed
may occur in the capillaries (fig. 36) (373). Con-
sistent with the latter view are some physiological
observations on the isolated lung (318). While such
calculations and observations cannot define the major
site of pulmonary vascular resistance under natural
conditions, they do emphasize that the evidence
favoring the precapillary segments is not on firm
footing.
Under certain circumstances, the capillaries do
become the major site of pulmonary vascular re-
sistance. The most common circumstance is the
artificial increase in alveolar pressure in the course of
positive pressure breathing (191, 306). It is also
possible to imagine such a role for the capillaries in
those natural circumstances in which the left atrial
pressure happens to fall below alveolar pressures.
Such a condition presumably exists in the apical
alveoli of the upright human subject and is exag-
gerated when the standing subject takes a deep
breath.
In isolated lungs perfused at physiological levels
of blood pressure, the pulmonary capillaries appear to
be less distensible than the larger pulmonary vessels
(124). However, physiologically meaningful measure
ments of the distensibility characteristics of the
pulmonary capillaries are difficult to obtain for a
variety of technical reasons, including the inacces-
sible location of the capillaries, the difficulty of
making static measurements under in vivo, dynamic
conditions, the difficulty in reproducing the natural
capillary pressures in experimental preparations, and
the uncertainty concerning the distensibility charac-
teristics of the pericapillary tissues because of the
propensity of the isolated lung to develop pulmonary
edema.
Time Spent by Blood in Pulmonary Capillaries
Stephen Hales seems to have been the first (1733)
to pay serious attention to the rate at which blood
flows through the pulmonary capillaries; coupling
direct observation with simple arithmetical calcula-
tion, he estimated this rate to be approximately 1.4
mm per sec in the frog lung (1 78). Curiously enough,
the elaborate techniques of the twentieth century
(cine-microphotography with lamp black as a tracer
in the exposed lung) have provided similar values for
the cat (1 to 2 mm/sec) (414).
The latter observations on the cat are the only
direct observations on the time spent by particles in
the pulmonary capillaries (o. 1 sec). All other esti-
mates represent calculations and assumptions based
on measurements of alveolar capillary gas exchange.
The most commonly cited values are those of Rough-
ton, based on the kinetics of the combination of
carbon monoxide and hemoglobin in man. Originally,
this approach indicated that, on the average, a unit
of blood spends approximately three-quarters of a
second in gas exchange in the pulmonary capillary
at rest, and somewhat less during exercise (364);
subsequent refinements in methodology have sug-
gested that the contact time at rest may be a little
longer, i.e., of the order of 1 sec (223, 317, 365).
Other calculations, based on the analysis of the
alveolar-arterial oxygen gradient have yielded lower
values: 0.18 sec in the dog (295) and 0.23 to 0.5 sec
in resting man (295, 396); however, theoretical con-
siderations suggest that this approach tends to under-
estimate the time of contact (318). Finally, anatomical
considerations have led to the low contact time of
0.1 sec (301). At first encounter, this is a discouraging
span of values. But, in view of the wide variations in
methodology, assumptions, and types of calculations,
this range of about 0.1 to 1.0 sec in resting man is
surprisingly small and, when duly weighed, the
generally accepted value of three-quarters of a sec
to 1 sec at rest seems quite reasonable. Obviously
lacking are simultaneous measurements of contact
time by the physiological methods and by direct
observation of the exposed lung in the same animal
(143)-
The time spent by a unit of blood in the pulmonary
capillary depends on various hemodynamic influences.
Paramount among these is the relationship between
the stroke output of the right heart and the pulmonary
capillary blood volume; in resting man these two
values are of the same order of magnitude. Another
determining influence is the nature of the pulmonary
capillary blood flow: pulsatile pulmonary capillary
flow causes some red cells to spend less time in the
pulmonary capillaries than others. Finally, values
based on alveolar-capillary gas exchange and the
rate of combination of test gases with hemoglobin
may misjudge actual contact times; for example,
actual time would be expected to differ from calcu-
lated time if, as is customarily done, the calculations
assume that the hematocrit of pulmonary capillary
blood is identical with that of blood sampled from
large systemic vessels (143, 317).
These theoretical considerations may have practical
meaning. Ordinarily, the time spent by each unit of
blood in the pulmonary capillary is more than ample
for complete oxygenation, both at rest and during
DYNAMICS OF PULMONARY CIRCULATION
I707
exercise. However, it is conceivable that, under
certain pathological conditions, such as those which
involve a huge pulmonary blood flow through a
curtailed pulmonary vascular bed, the contact time
may be too brief. Indeed, an inadequate contact
time has been invoked to account for peripheral
arterial hypoxemia in resting patients with multiple
pulmonary emboli and in exercising patients with
"alveolar-capillary block" (359)- However, such
explanations are not entirely convincing since, in
most of these pathologic states, other equally con-
vincing explanations for peripheral arterial hy-
poxemia, e.g., opening of pulmonary arteriovenous
shunts, also exist.
Pulmonary Capillary Hematocrit
In the 1930's, Fahraeus and Lindquist pointed out
that blood flowing in capillary tubing has a greater
ratio of plasma to red cells than does blood in wider
streams (18). Since then, it has been repeatedly shown
that the hematocrit of blood in many organs is less
than in the large vessels which enter and leave them.
The lung appears to be no exception: measurements
of the hematocrit of blood obtained from whole organ
homogenates (161) as well as comparisons of transit
time of tagged red cells and plasma (332) indicate
that the small vessel hematocrit is regularly lower
than that of the large vessels, ranging from 1 7 per
cent less at rest to 13 per cent less during exercise.
This difference may affect not only the hemodynamic
behavior of the pulmonary capillary circulation but
also measurements of alveolar-capillary gas exchange,
such as the pulmonary diffusing capacity for carbon
monoxide, and derivative values, such as the time
spent by blood in the pulmonary capillary and the
pulmonary capillary blood volume (143, 318).
Transcapillary Exchange
Until recently, considerations of the transcapillary
movements of water and electrolytes emphasized
their bulk transfer and dealt largely with the balances
between hydrostatic and oncotic pressures (Starling's
law of capillary exchange). Recognizing that the
pulmonary capillaries were unique in being "bathed
in air rather than in water," such considerations of
bulk transfer were sufficient to account for the normal
"dry" lungs as well as the "wet" lungs of clinical
pulmonary edema. Within the last few years, trans-
capillary exchange by diffusion has also been taken
into serious account (76). Still incomplete is the
definition of the role of the pulmonary lymphatics
with respect to the water which escapes into the
pulmonary interstitium and alveoli.
Certain aspects of the transcapillary exchange of
water seem well established. For example, the osmotic
pressure of the plasma colloids (expressed figuratively
as an "oncotic pressure") of approximately 25 mm Hg
normally suffices to prevent bulk loss of fluid from
the pulmonary capillaries, even in the hydrostatically
dependent portions of the lung. Also, while trans-
capillary molecular exchange rates by diffusion may
be of the order of the cardiac output, no net fluid
transfer occurs. Ethanol and injected carbon dioxide
behave like isotopic water. As expected, the indi-
cator-dilution curves for T-1824 (which does not
leave the capillaries) and for water (which undergoes
rapid to-and-fro transcapillary exchange) are quite
different (76). Indeed, labeled water resembles the
inert gases in behaving as though the barrier did not
exist.
Quite unexpected is the similarity between the
T-1824 indicator-dilution curve and the correspond-
ing curves for urea and for the highly diffusible
phosphate, potassium, sodium, and chloride ions
(16, 77, 360). Virtual identity of these curves has
been interpreted to mean that: a) in contrast to the
enormous pulmonary volume of distribution of
sodium and chloride at equilibrium (134), the volume
of dilution available to urea and to the diffusible
ions during a single circulation is confined either to
the strict pulmonary vascular volume or to the pul-
monary vascular volume plus an additional, circum-
scribed perivascular volume into which these sub-
stances diffuse and then promptly return, and b)
because of the ready permeability of the barrier
(probably the basement membrane) to water as well
as its relative impermeability to ions and urea, the
barrier is aqueous rather than lipid in nature.
MISCELLANEOUS HEMODYNAMIC PHENOMENA
Pulmonary Arterial Pulse-Wave Velocity
The pulse-wave velocity is related to relative,
rather than to absolute, distensibility (see Chapter
24). Estimates of the speed at which the pulse wave
travels along the length of the pulmonary arterial
tree vary considerably. The discrepancies are at-
tributable to three causes: inadequate methodology;
the experimental difficulty of controlling the hemo-
dynamic influences which modify the speed of the
[708
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
pulse wave, e.g., initial volume or diastolic pressure;
and species differences in distensibility. Thus, one
set of values for the pulmonary arterial pulse-wave
velocity in the rabbit averages 200 cm per sec (124);
this method is based on measurements of pulmonary
vascular distensibility and capacity. On the other
hand, much lower values (83 cm sec) have been
calculated by an alternate approach which involves
the registration of phase shifts of a single harmonic
component of the pulse wave as it traverses a known
distance (140a). Similarly, in the dog, mean velocities
have averaged 250 cm per sec in one studv (311)
and 400 cm per sec in another (225). Finally, in man,
kymographic studies have indicated a velocity of
200 cm per sec in the main pulmonary artery and
275 cm per sec in the peripheral branches (71). It
should be noted that each of these methods has its
own peculiar problems: thus, some are troubled by
the need for the precise measurement of in vivo
distances between points on the pulmonary artery
(71, 225); others have to overcome the difhcultv of
attempting static measurements under dynamic con-
ditions of flow (124).
Because of the practical limitations and assump-
tions involved in the experimental approaches, none
of these values seem to offer more than an order of
magnitude. However, with a single exception (225),
they are consistent with the notion that the pulse-
wave velocitv in the distensible pulmonarv arterial
tree is somewhat less than in the aorta.
Pulmonary Circulation Tune
Measurements of circulation times are in common
clinical use for the recognition of heart failure. This
use depends on the arrival of a test substance at a
chosen site in sufficient concentration to be detected;
the value obtained, i.e., the ''appearance time," is
related, in a complex way, to the mean circulation
time used for the calculation of central blood volume
(184, 313). It is clear that the precise value for cir-
culation time will be influenced not only by technical
peculiarities, e.g., by rate of injection, nature of the
test substance, and sensitivity of the detector, but
also by physiological events. For example, if the
pulmonary circulation lies between the sites of in-
jection and sampling, an increase in its blood volume
will dilute the test substance excessively and delay
its recognition at the test site. It is, therefore, not
surprising that values for circulation times from
different laboratories are frequently inconsistent.
One study in the dog (Stewart principle) found the
total circulation time to be approximately 1 1 sec,
and the pulmonary circulation time to average
approximately half of the total (184). Other studies
indicate that the pulmonary circulation time (pul-
monary artery to vein) is somewhat less, i.e., about
3 to 4 sec (306, 329). The circulation time for red
INCREASED PRESSURE DECREASED PRESSURE
IN L IN T
fig. 37. Two-chamber model to illustrate the effects of vary-
ing pressures around a collapsible rubber tube on its dimen-
sions. The tube (a, c, b) runs through one rectangular chamber
(T-T) and is exposed for a limited extent (c) to the pressure
of the other chamber (L). In .-1 and B, the two ends of the per-
fusing system (a, b) are outside of the chambers; in C and D,
the entire system is contained within chamber T-T. For the
sake of clarity, only A has been lettered; the arrow in each
figure indicates the chamber subjected to a change in pressure
and the direction of change. A : the pressure in chamber L is
greater than atmospheric, the pressure in chamber T-T is
atmospheric. The aspect of the tube exposed to L is collapsed.
B: the pressure in chamber T-T is less than atmospheric; the
pressure in L is atmospheric. The whole length of tube within
chamber T-T is increased in diameter; the portion exposed to
L is less dilated than the remainder of the tube. C and D : two
different but equivalent conditions. As long as the reservoirs
are contained within chamber T-T, the same effect is obtained
by balancing positive pressure in L against atmospheric pres-
sure in T-T (C) or by balancing atmospheric pressure in L
against negative pressure in T-T (D). In either case, the result
is identical with that illustrated in .4. By analogy, this model
suggests that: /) when alveolar pressure \L) is raised (.1), the
transmural pressure of adjacent vessels (c) is decreased; 2) the
situation of two reservoirs (pulmonary arterial and venous
pressures) outside of the pleural cavity (B) corresponds to
negative (pleural) pressure inflation as well as to normal
respiration; 3) when pleural pressure is decreased {B), the
transmural pressures of larger vessels are increased more than
those of the capillaries, and ./> the situation of two reservoirs in
the pleural cavity (C and D) is a physical identity with positive
pressure inflation. [Based on Quincke & Pfeiffer (324).]
DYNAMICS OF PULMONARY CIRCULATION
I709
cells is less than that of the plasma [cell transit time
plasma transit time = 0.91 ± .05 (310)]. In man,
the total (systemic plus pulmonary) circulation time
is of the order of 15 to 18 sec (184). Approximately,
one quarter to one third of the total time is spent in
traversing the pulmonary circulation; the circulation
time between the right ventricle and the pulmonary
capillaries is estimated to be 2 to 3 sec (87). Despite
the many measurements of the pulmonary circula-
tion time, both in normal subjects and in patients
with cardiopulmonary disorders, there is still an
inadequate fund of information concerning the pre-
cise pulmonary vascular pathways traversed by the
test substance between the sites of injection and
sampling (75).
INFLUENCE OF RESPIRATION ON
PULMONARY CIRCULATION
In the pulmonary circulation, blood pressures,
volume, and flow change during each breath. The
precise nature of these changes has been debated for
two centuries (50, 324). Nonetheless, many aspects
remain unsettled largely because of the technical
difficulties involved in simultaneously recording
transient respiratory and circulatory events in the
intact animal or man.
The attempts to circumvent the technical diffi-
culties have created problems of their own: a) the
recourse to simplifying physical models (fig. 37)
and artificial preparations has led to dubious gener-
alizations about natural breathing (61, 354); b)
the experimental control of some respiratory in-
fluences at the expense of others, has tended to
exaggerate the physiological importance of some
parameters while denying others — such as the degree
and type of inflation — their full due (215, 380); and
c) complicated experimental designs have created
artificial situations in which the usual calculation of
pulmonary vascular resistance either does not apply
or is very difficult to translate into terms of pul-
monary vascular dimensions (315, 354).
Spontaneous Breathing
During inspiration, as pleural pressure becomes
more negative, luminal pressure (referred to atmos-
phere) decreases. On the other hand, transmural
pulmonary arterial pressures — systolic, diastolic, and
mean — increase. During expiration, these changes are
reversed.
There is no unanimity concerning the mechanisms
responsible for the increase in pulmonary arterial
transmural pressure during inspiration. Most certain
is an increase in pulmonary blood flow, arising from
the decrease in intrathoracic pressure and from the
increase in systemic venous return which it promotes
(17, 49); much more equivocal is a reduction in the
outflow from the pulmonary vascular bed so that the
pulmonary blood volume is increased (225, 253).
Such a combination of increased inflow and reduced
outflow would imply pulmonary vascular distension
and, hence, a decrease in pulmonary vascular
resistance.
However, there are experimental results which do
not fit this picture: a) under some circumstances, in-
spiration has been found to increase — rather than to
decrease — pulmonary vascular resistance (49, 1 1 5,
309); b) measurements of transmural atrial pressures
suggest that the pulmonary veins empty uninterrupt-
edly during inspiration (187); and e) experiments on
models and dogs indicate, that, under circumstances
which promote an unusual emptying of the extra-
thoracic veins, the veins may collapse during inspira-
tion as they enter the thorax, thereby preventing an
increase in venous return (308).
At least part of the divergent opinions about the
effects of inspiration on the pulmonary circulation
seem to arise from failure to take full cognizance of the
experimental setting: during an ordinary quiet
breath, pulmonary blood flow and volume do appear
to increase; if resistance does change, the change is
small (354). Moreover, collapse of extrathoracic veins
is not apt to occur under ordinary physiological cir-
cumstances even though it may conceivably occur in
the resting subject who is breathing with enormous
tidal volumes (354).
As long as fluctuations in intrathoracic pressure
are small and venous return to the right heart remains
ample throughout the respiratory cycle, the pulmo-
nary arterial pressure pulses are fairly uniform. How-
ever, in clinical conditions associated with low
systemic venous return, in chronic pulmonary disease
(fig. 38), during exercise and during voluntary deep
breathing, marked swings do occur in the pulmonary
arterial pressure pulses. These reflect not only the
swings in intrathoracic pressure but also changes in
blood flow, volume, and resistance (187, 253).
During natural expiration, the filling of the right
heart is decreased as intrathoracic pressures approach,
or even exceed, caval pressures; the pattern described
for inspiration is reversed. In patients with pulmonary
disease, in whom expiration has become an active
I 7IO
HANDBOOK OF PHYSIOLOGY
CIRCULATION' II
fig. 38. Blood pressures re-
corded from the pulmonary
artery (PA) and right atrium
(RA) during quiet breathing in
a patient with chronic bronchitis
and emphysema.
30 [—
mm Hg
20
5
0
PA
I 0 —
MMjuW
M
M
RA
process, venous return to the thorax may become ob-
structed as positive intrathoracic pressures are im-
posed on central venous pressures. However, until the
systemic venous valves (in the external jugular, sub-
clavian, axillary, and femoral veins) become incompe-
tent from central venous congestion, the rise in central
venous pressure is not transmitted to the peripheral
systemic veins. Therefore, during a forced expiration,
peripheral venous pressure rises only gradually, rep-
resenting the gradual filling of a distensible system
which is obstructed at its thoracic venous outlet.
of the pulmonary vascular bed (397);/) the balance
of alveolar, pleural, left atrial, and pulmonary ar-
terial pressures which is required to make calculated
changes in resistance meaningful (354); and g) the
probable insignificance of alveolar surface tension in
determining pulmonary vascular resistance during
either positive (intrapulmonary) or negative (pleural)
pressure inflation of the lungs (398). This list also
serves to emphasize the fallibility of extrapolating
from artificial inflation of the lung to spontaneous
breathing.
Inflation of the Lungs
In 1 87 1, Quincke and Pfeiffer reported that positive
(intrapulmonary) inflation of the lungs decreases the
pulmonary blood volume and increases the resistance
to flow (324). Since then, physiologists have debated —
on the basis of a wide variety of experiments, models,
animal preparations, and intact animals — whether
resistance to perfusion increases as the lungs are in-
flated and if positive pressure inflation exerts the same
effects as negative (pleural) pressure inflation (39,
431). It now seems that the discordant results were to
be expected because of the nature of the experiments
and of the models (62). The principal bases for dis-
agreement seem to have been: a) the uncertain mean-
ing of the model under study (354); b) the failure to
distinguish between transmural pressure and luminal
pressure in determining vascular calibers (61); c) the
mechanical increase in resistance at exceedingly low
lung volumes, possibly due to kinking or collapse of
small vessels (61); d) the mechanical increase in re-
sistance at high degrees of pulmonary distension as
resistance vessels are stretched (354); e) the influence
of the pressure-volume behavior of the lung, and of
its enclosed pulmonary vasculature, on the resistance
Positive Pressure Breathing
In the isolated lung, or in the open-chest animal,
inflation of the lung from the collapsed position is
associated with an initial decrease in resistance as the
lung is moderately inflated, followed by an increase
in resistance as the lung is distended further (62).
Such U-shaped curves have been taken to represent:
a.) a decrease in resistance to blood flow as vessels in
the collapsed lungs are unkinked and opened (62),
followed by h) an increase in resistance due to both a
decrease in transmural distending pressure as alveolar
pressures increase (191, 215) and mechanical distor-
tion of the resistance vessels at the high lung volumes
(2I5> 397)-
In the closed-chest animal, positive pressure breath-
ing affects the pulmonary circulation by increasing
alveolar pressure and impeding systemic venous
return to the lungs: the systemic venous-right atrial
pressure gradient is decreased, thereby decreasing the
filling of the right ventricle and right ventricular out-
put (87, 253); the volume of the heart and pulmonary
vessels decreases. As a result of the combination of a
decreased right ventricular output and a sustained
left ventricular output, the pulmonary blood volume
DYNAMICS OF PULMONARY CIRCULATION
I7II
V*TI ON
fio. 39. Schematic representation of the influence of positive pressure breathing on the right
ventricular (RV) and systemic arterial (BA) pressures of a normal human subject. The mask pressure
(M) appears above the pressure pulses. All pressures are in mm Hg. During inflation, as mask (and
pleural) pressures increase, the pulse pressure in the right ventricle progressively decreases; con-
comitantly, the systemic arterial pulse pressure increases. During expiration, the reverse occurs. The
prompt return of the mask pressure to ambient pressure accounts for the ability of expiration to com-
pensate for the inspiratory deficit in blood flow. [After Richards el al. (340).
decreases (395) and the pulmonary vascular resist-
ance increases (49). Although the imposed pressure
raises pulmonary vascular luminal pressures the trans-
mural pressures are virtually unaffected (253), and
the pressure gradient along the length of the pulmo-
nary vascular tree remains essentially unchanged
(187, 256). The circulatory changes arising from the
imposed pressure reverse promptlv once the lungs
are vented to atmosphere (fig. 39).
In systemic hypotensive states, positive pressure
breathing may precipitate circulatory collapse if
compensatory mechanisms are insufficient to sustain
the venous return to the right heart (49). Not only
are the output of the right heart and the pulmonary
blood volume reduced, but the normal balance be-
tween alveolar perfusion and alveolar ventilation is
also upset so that portions of the lung become ex-
cessively ventilated with respect to perfusion (160).
A variety of mechanical devices are in common use
for intermittent positive pressure breathing (426).
Their effects on the circulation are functions of the
degree and duration of the cyclic swings which they
induce in intrathoracic pressure. The cardiac output
generally falls (11) in proportion to the mean increase
in intrathoracic pressure; in practice, the cardiac out-
put may be kept at control levels by using cycling
devices which operate to: a) inflate the lungs gradu-
ally to peak pressure; b) decompress the lungs sud-
denly by venting them to atmosphere; and c) allow a
longer period of exposure to atmospheric pressure
than to positive pressure.
Negative Pressure Breathing (Pleural)
As pressure around the collapsed isolated lung is
artificially decreased, its resistance to perfusion de-
creases (61, 354). The changes in resistance which
accompany further inflation of this type are unsettled.
Thus, some have found only a continuing decrease in
resistance as the lung is expanded by progressive
decrements in "pleural pressure" (62); others have
found U-shaped curves in which the initial drop in
resistance as the lung begins to expand (pleural pres-
sure — 5 to — 10 cm H^O) is succeeded by an increase
in resistance as the pleural pressure decreases further
(pleural pressure —10 to —25 cm H20) (354, 397).
The nadir in resistance occurs at half-maximal lung
volume.
The mechanisms proposed to account for these
divergent results are enlightening. The initial de-
crease in resistance — to which all agree — has been
attributed to either an increase in transmural pressure
or to the unkinking of "gnarly" vessels (61 ). Different
explanations have been used to account for the di-
vergent results at high levels of inflation: those who
find a continued drop in resistance ascribe it to the
continued increase in transmural pressure as "pleural"
pressure drops (62); those who find that resistance
finally increases believe that at lung volumes exceed-
ing 50 per cent of maximal, mechanical distortion of
the pulmonary vessels — a function of lung volume
rather than of transmural pressure — is involved (354,
397). While these studies leave unsettled the question
of the behavior of the pulmonary vascular resistance
as the lung is progressively inflated, they do serve as
a reminder that the transmural pressures, which ac-
count satisfactorily for passive changes in caliber at
moderate degrees of inflation, may be supplanted by
other mechanical influences, e.g., stretching or col-
lapse, in determining vascular calibers at extreme
inflation or deflation.
Negative Pressure Breathing (Intrapulmonary)
"Snorkel" breathing is characterized by a lower
pressure within the lungs than around the body (49)-
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
4 OH
I00|
60
fig. 40. Effects of a prolonged forced expiration on systemic arterial pressures {upper tracing) and
right ventricular {lower /racing) pressures in a patient with chronic pulmonary emphysema and
fibrosis. Ten seconds after the start of expiration (solid arrow), the amplitude of the RV pressure
pulse begins to increase, after several beats, there is a progressive rise in the femoral arterial systolic,
diastolic, and pulse pressures. This suggests that as the flow of blood into the thorax is impeded, the
volume of the peripheral venous reservoir slowly increases until the venous pressure becomes suffi-
ciently great again to increase the right heart filling and output, in spite of the continued elevation
of the intrathoracic pressure. During the succeeding inspiration (hollow arrow), a further augmenta-
tion of the right ventricular pulse pressure occurs (last four beats). [After Lauson, el at. (253).]
i!
IIIIIIIIIIIIMMIIHIIIIIMMIim|N|||»l|IU}H
lliu^..; :^»i«
lUwMwi1
BA
O -20
20
h,
OES
fig. 41. Effect of the Valsalva maneuver on brachial arterial pressure (BA) and "transmural"
pulmonary arterial pressure (dPA). The changes in intrathoracic pressure, measured as the esopha-
geal pressure (OES), indicate the onset, duration, and end of the expiratory effort. All pressures are
in mm Hg. The overshoot in the systemic arterial response following the Valsalva maneuver is
ascribed to reflex vasoconstriction. On the other hand, the pattern of change in pulmonary arterial
pressure is attributed to mechanical events, i.e., to alterations in venous return and right ventricular
output. [After Lee et at. (255).]
At the start of negative (intrapulmonary) pressure
breathing the systemic venous return to the lungs
and the pulmonary blood volume increase (49). In
contrast to positive pressure breathing, the negative
intrapulmonary pressures distend the intrapulmonary
vessels (256). At the small lung volumes, associated
with continuous negative pressure breathing, atelec-
tasis may develop.
Cough
During a cough, pressure referred to atmosphere
rises simultaneously and equally in the thorax
(fig. 21), abdomen, and cerebrospinal canal (190).
The increase in pressure (which may transiently reach
levels of 1 50 mm Hg) does not strain the intrathoracic
(255) or abdominal or cerebrospinal vessels, and does
not, per se, affect the pressure gradient which drives
blood along the pulmonary vascular tree. However,
it is propagated to the peripheral arterial tree where
it causes a marked increase in the transmural distend-
ing pressure (190).
Prolonged Expiration
During a prolonged expiration (fig. 40), the ampli-
tude of the pulmonary arterial (and right ventricular)
pressure pulse first decreases and then gradually in-
creases; the systemic arterial blood pressures undergo
a similar pattern of change. This sequence has been
DYNAMICS OF PULMONARY CIRCULATION
ni3
interpreted as reflecting the gradual increase in
peripheral venous blood volume and pressure during
the sustained expiration until adequate filling of the
right, and then the left, heart is restored (187, 253).
Forced Expiration (Valsalva)
The effects of the Valsalva maneuver (forced ex-
piration against a closed glottis or a column of water
30 to 40 cm high) (fig. 41 ), has been more intensively
studied in the systemic circulation than in the pulmo-
nary circulation (255). Shortly after the start of the
maneuver, the distending pulmonary arterial pressure
falls abruptly as the filling pressure of the heart is
reduced by the increased intrathoracic pressure (187,
253); it remains low during the period of strain. Upon
release of the expiratory effort, pulmonary arterial
mean and pulse pressures "overshoot" the prestrain
level, but to a lesser extent than in the systemic
arteries. During the maneuver, considerable quanti-
ties of blood may be displaced from the thorax to the
periphery (253). The systemic arterial overshoot
seems to involve a combination of an increased
cardiac output and vasoconstriction; although some
believe that these same mechanisms are involved in
the pulmonary arterial overshoot, the evidence for
pulmonary vasoconstriction is much more tenuous
than for systemic vasoconstriction (255).
OCCLUSION OF A PULMONARY ARTERY
In principle, occlusion of larger and larger portions
of the pulmonary arterial tree provides a simple tool
BEFORE
LEFT
RIGHT
DURING
I
£ :
'
J._;jt"'-
E=Si5P
- 1" '■
fig. 42. Effect of complete occlusion of one pulmonary artery
on ipsilateral oxygen uptake. Before inflation of the occlusive
balloon in the right pulmonary artery (left panel), both lungs
share almost equally in the oxygen uptake. The middle panel
shows the occlusive balloon, inflated with Diodrast, positioned
at the end of a cardiac catheter in the right pulmonary artery.
After complete occlusion, the oxygen uptake by the right lung
ceases. [After Fishman et al. (139J.J
for testing the passive effects of an increase in pulmo-
nary blood flow on pulmonary arterial pressures
(fig. 34) and pulmonary vascular resistance. In the
open-chest dog, graded occlusion of the pulmonary
vascular tree is easily performed (194). The situation
is much more complicated in the closed-chest dog or
man in whom balloon-tipped, venous catheters are
guided, under fluoroscopic control, into a pulmonary-
artery; in this experimental situation additional tech-
niques, such as bronchospirometry, are required to
establish the degree of occlusion which has been
accomplished (fig. 42).
In wondrous contrast to the catastrophic effects
occluding the pulmonary vascular tree by emboli, in-
flation of a balloon in one pulmonary artery is entirely-
innocuous: the metabolic rate, the total cardiac out-
put, the systemic arterial and left atrial blood pres-
sures, and the heart rate remain unchanged; the total
ventilation rarely increases by more than 10 per cent
(53, 101). However, even this slight change in total
minute ventilation helps to adapt the alveolar
ventilation to the altered pulmonary capillary per-
fusion (390).
The first studies of the pulmonary circulation fol-
lowing occlusion of one pulmonary artery- were made
in 1876 on the open-chest dog (132). These indicated
that the pulmonary arterial pressure, measured
proximal to the site of occlusion, increased by 50 per
cent following interruption of the blood flow to one
lung. Subsequently, a similar procedure in other
open-chest animals found lesser increases, ranging from
zero in the rabbit ( 1 25) to 20 per cent in the cat ( 1 25).
More puzzling than these divergent results in the dif-
ferent species is the fact that in the intact dog, oc-
clusion of one pulmonary artery by a balloon-tipped
catheter has also produced variable effects: on the
one hand are results indicating that pulmonary ar-
terial pressure remains essentially unchanged (68,
101); other results indicate an increase in pressure of
the order of 33 per cent (252, 259). However, since
none of the experiments in the dog verified the degree
of pulmonary arterial occlusion produced by the in-
flated balloon, it seems reasonable to assume that in-
flation of the balloon was not equally successful in
producing complete occlusion in the different dogs,
and that the larger increments in pulmonary arterial
pressure — i.e., of the order of 33 per cent (252, 259) —
represent the more complete occlusions.
In man the results have been more consistent: after
complete occlusion of one pulmonary artery, the
pulmonary arterial pressure (primarily systolic)
proximal to the occlusion increases, the pulmonary-
i7i4
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
blood volume on the patent side also increases, the
pulmonary circulation time decreases and the pul-
monary vascular resistance decreases (42, 101). Be-
cause of the configuration of the trachea in the dog,
the completeness of the unilateral interruption of
blood flow is more readily checked by bronchospi-
rometry in man (fig. 42); when the right ventricular
output has been shown to be completely diverted to
one lung in man, the pulmonary arterial pressure has
been found to increase by 30 to 40 per cent (5 to
7 mm Hg), and the calculated resistance to fall to 40
or 50 per cent of the initial value (42, 53, 101). The
increase in pulmonary arterial pressure seems en-
tirely attributable to the passive consequences of an
augmented pulmonary blood flow and volume.
EFFECTS OF EXERCISE ON PULMONARY CIRCULATION
Exercise is a practical expedient for increasing pul-
monary blood flow. However, by comparison with
unilateral occlusion of one pulmonary artery, it suffers
the disadvantage of simultaneously evoking changes
in the ventilation, in the performance of the heart,
and in a variety of circulatory parameters.
Experimentally, dog and man have been exercised
in various different ways: electrical stimulation,
ergometer, push-pedal, and treadmill. The workload
imposed by the exercise, as well as the efficiency with
which the exercise is performed, varies with the type
of exercise, the position in which it is performed, and
the familiarity with the exercise (7, 12). For practical
reasons, the work load is generally inferred from the
increase in oxygen uptake rather than measured
directly (216).
Pulmonary Blood Flow
The first measurements of the cardiac output
during exercise were made by foreign gas methods
(240). Since then, with few exceptions (3), the foreign
gas methods have been superceded by indicator-
dilution methods and by applications of the Fick
principle.
For accuracy, the use of the Fick method during
exercise requires that the oxygen uptake at the mouth
provide a precise measure of the oxygen uptake by-
pulmonary capillary blood, and that the pulmonary
arteriovenous oxygen difference be constant. These
criteria are most apt to be satisfied when respiration
and circulation become stable, i.e., when minute
ventilation, respiratory exchange ratio, oxygen up-
take, heart rate, arteriovenous oxygen difference, and
cardiac output no longer vary with time. Unfortu-
nately, all these parameters do not stabilize simulta-
neously (108, 109, 136). Thus, the oxygen uptake at
the mouth, and the arteriovenous oxygen difference
mav level off after 1 min of heavy exercise (up to 1200
ml min nr), whereas the respiratory exchange ratio
and the ventilation require much longer to reach a
plateau. As long as the respiratory gas exchange is
unstable, it is difficult to be sure that the measure-
ment of the oxygen uptake at the mouth provides a
reliable value for the numerator of the Fick equation,
i.e., of the oxygen taken up by blood perfusing the
pulmonary capillaries. On the other hand, when
both the respiration and circulation become stable —
usually within 3 min in normal subjects performing
light, supine exercise (O2 uptake up to 400 ml min/
m'2) — the prospect of an accurate measurement of
pulmonary blood flow is increased (212). Obviously,
during heavy exercise (O2 uptake greater than 1000
ml min m2), it may become difficult to achieve a
steady state; indeed, as exhaustion is approached, the
Fick method may become completely unreliable. It
is apparent that these considerations do not support
the practice of applying the Fick principle to the
measurement of the cardiac output during brief
periods of heavy exercise (108, 109). It would be re-
assuring if this use of the Fick principle were validated
by another independent method, such as the Stewart-
Hamilton, which, in principle, requires a briefer
steady state.
Blood Flow and Oxygen Uptake
In the unanesthetized dog (12) and in normal man
(104, 149, 335), an increment in oxygen uptake (AY0„)
of 100 ml is usually associated with an increment in
cardiac output (AQ.) of 600 to 800 ml. However, both
lower (101, 109, 258) and higher (101 ) ratios of AQ. /
AVo, have also been observed. One likely explanation
for at least part of the variability is that different
degrees of approach to the '"basal" state were achieved
prior to exercise in the different studies: it has been
shown that in those studies in which a serious attempt
is made to achieve a basal pre-exercise state, the
ratio AQ/AVoj may even exceed 1 liter of cardiac
output per 100 ml of oxygen uptake (132); conversely,
when a nonbasal state exists prior to the exercise, the
ratio AQ./AV0„ may fall below 600 ml (109). This
point is emphasized by the solid line of figure 12,
which indicates that for a given level of oxygen uptake
the cardiac output during excitement is higher than
DYNAMICS OF PULMONARY CIRCULATION
!7I5
during exercise; this effect may continue, to an un-
predictable degree, during mild exercise.
In heart failure the ratio AQ/AV02 is often abnor-
mally low, i.e., less than 600 ml increase in flow per
100 ml increase in oxygen uptake. A sti iking dissocia-
tion between AQ, and AV0, follows the exhibition of
dinitrophenol, so that the cardiac output continues
at basal levels even though oxygen uptake increases
tremendously (216).
Arteriovenous Oxygen Difference
The pulmonary arteriovenous oxygen difference
increases during exercise. However, in contrast to the
roughly linear relation between cardiac output and
oxygen uptake during graded exercise, the relation
between the arteriovenous oxygen difference and
oxygen consumption is clearly hyperbolic (109, 382).
Whether the oxygen requirements of the tissue are met
predominantly by an increase in cardiac output or by
a greater extraction of oxygen from each unit of blood
perfusing the tissues seems to depend, at least in part,
on the type of exercise, the body position in which the
exercise is performed, and the ambient temperature
(7, 336). Parenthetically, it is of interest that during
graded exercise (up to 2000 ml/min/m2) trained and
untrained subjects increase cardiac output and widen
arteriovenous differences for oxygen in an identical
fashion (149).
Pulmonary Vascular Pressures
Because of the difficulty in measuring intrathoracic
pressures, pulmonary vascular pressures are conven-
tionally referred to atmosphere. In only one study
were they also referred to esophageal pressures (109);
this study suggested that conventional luminal pres-
sures tend to underestimate slightly the transmural
pressures. Before considering the change in pulmo-
nary artery pressure during exercise, it is relevant to
recall that: a) pulmonary vascular pressures are
difficult to measure accurately during exercise since
respiratory swings are marked and the records are
apt to be distorted by artifacts, and b) especially
during severe exercise, shifts in mid-position of the
lung and changes in compliance confuse the recogni-
tion of the mechanisms involved in a change in pres-
sure (109).
Until a few years ago, because of the practical
difficulties in measuring small changes in pressure
during exercise, it was uncertain if light (supine)
exercise elicited an increase in pulmonary arterial
pressure (104, 208, 346). Indeed, an appreciable
increase was believed to occur only at levels of exer-
cise which tripled the cardiac output (89). However,
recent refinements in manometric methods, coupled
with the substitution of continuous pressure recording
for the tedious process of measuring and integrating
individual pressure pulses, have established that the
pulmonary arterial pressure increases (by 3-5 mm Hg)
even during light supine exercise (132, 370, 383).
The behavior of the pulmonary arterial pressure
during light exercise is quite stereotyped (fig. 28) : at
the start of the exercise, the (luminal) pulmonary
arterial pressure increases abruptly by 3 to 5 mm Hg.
As exercise is continued, a plateau is reached, gener-
ally 1 to 2 mm Hg less than peak values (132, 382).
The increase in systolic pressure exceeds the increase
in diastolic pressure. As a rule, the higher the pre-
exercise level of the pulmonary arterial pressures, the
higher the values reached during exercise. Immedi-
ately after the exercise, the pulmonary arterial pres-
sure often falls below control, resting values (iog, 132,
382).
The pulmonary arterial flow-pressure points ob-
tained by different investigators during graded exer-
cise are superimposed on the pressure-flow line of
figure 34. At the lower grades of exercise, the points
fall along the flow-pressure curve obtained in the
course of progressive curtailment of the pulmonary
vascular bed by balloon-occlusion. At the higher
grades of exercise the pulmonarv arterial pressure at
any given level of blood flow tends to exceed the
corresponding pressure during balloon-occlusion.
Direct measurements of the left atrial pressure
during exercise in intact man or dog have not been
reported. On the other hand, in dogs exercised by
electrical stimulation of the extremities, the left atrial
pressure remains unchanged (125); unfortunately, the
level of exercise in these experiments is unknown. In
man the slight increments of the pulmonary arterial
pressure during mild exercise suggest that if the left
atrial pressure does increase, the increase cannot
exceed a few mm Hg. The pulmonary ''wedge" pres-
sure is unaffected by mild exercise but may increase
slightly during severer exercise (104).
Pulmonary Blood Volume
There is considerable indirect evidence to indicate
that the pulmonary blood volume increases during
supine exercise: the central blood volume increases
(48, 101), the pulmonary compliance decreases (279),
and, except for the muscles (250), the regional blood
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
volume decrease (48). However, such evidence ap-
plies only to the steady part of the exercise; more
difficult to ascertain is the pattern of change in the
pulmonary blood volume from the start to the finish
of the exercise. In this regard, the most convincing clue
is the characteristic sequence of changes in the pulmo-
nary arterial pressure (fig. 28); this pattern is con-
sistent with an abrupt increase in the pulmonary blood
volume at the start of the supine exercise, a gradual
stabilization at below-peak values as exercise is con-
tinued, and a prompt fall to below resting values
when exercise is arrested (iog, 132). How the in-
creased blood volume is apportioned among the
different vascular segments of the lung is unknown;
however, the pulmonary capillaries apparently share
in the increase (364).
Pulmonary Vascular Resistance
The calculated pulmonary vascular resistance either
remains unaltered (iog) or, more often, decreases
(101, 346) during light to moderate (supine) exercise.
Although it is generally believed that the decrease in
calculated resistance at these levels of exercise repre-
sents both the widening of patent pulmonary vessels
and the opening of closed vessels (208, 241, 346), the
particular mechanisms which are responsible for this
change in vascular geometry remain speculative.
Three reasonable alternatives come to mind : active
pulmonary vasodilatation, passive dilatation by the
decrease in pleural pressure, or passive dilatation by
the increase in luminal pulmonary arterial blood
pressure. Of the three alternatives, the passive increase
in pulmonary arterial luminal (and transmural) pres-
sure seems sufficient — without invoking vasomotric-
ity — to account for the widening and opening of the
pulmonary vessels at these low grades of exercise (187).
During heavy exercise, as the pulmonary blood
flow is more than tripled, the pulmonary vascular
resistance (calculated on the basis of an assumed left
atrial pressure) is described as becoming constant
(252). As pointed out previously, the pulmonary
vascular tree is then pictured as behaving as though it
were comprised of "rigid tubes"; the "rigid tubes,"
in turn, are envisaged as wide-open, low-resistance
vessels with elastic fibers stretched to tighten their
collagen "jackets" (187). Generally speaking, the
calculations of pulmonary vascular resistance during
heavy exercise on the basis of pulmonary arterial
pressure and flow (fig. 34) are consistent with this
view. However, this interpretation of calculated
resistance is handicapped by the lack of assurance
concerning the simultaneous behavior of the left
atrial pressure, the pulmonary blood volume and the
pleural pressures. Indeed, without information about
these critical parameters, the ratio of pulmonary
arterial pressure to pulmonary blood flow during
heavy exercise may represent either an increase or a
decrease in pulmonary vascular resistance. Finally, as
indicated previously, not only may the kinetic energv
in the pulmonary artery exceed the potential energy
at high rates of pulmonary blood flow, but intercon-
versions of potential and kinetic energy are bound to
occur along the length of the pulmonary vascular tree
(20). Consequently, during heavy exercise, even the
ratio of the pulmonary vascular pressure gradient
(potential energy gradient) to the pulmonarv blood
flow need not provide a reliable index of pulmonary
vascular dimensions.
MISCELLANEOUS MECHANICAL INFLITENCES
Heart Rate
In normal dog and man, speeding up of the heart
rate by atropine, is ordinarily without appreciable
effect on pulmonary vascular blood pressures or blood
flow (424); in some instances the cardiac output may
increase by 40 to 50 per cent (168). In patients with
"tight" mitral stenosis, even the slight increment in
cardiac output induced by atropine may suffice to
precipitate pulmonary edema by elevating pulmonary
venous and pulmonary capillary pressures.
Slowing of the heart has been produced by vagal
stimulation in dogs: as the heart rate drops to one-
half or one-third of the initial value, the cardiac out-
put falls and the pulmonary venous pressure rises
(65). A similar combination of bradycardia, low
cardiac output, and high pulmonary venous pressure
also occurs when intracranial pressure is considerably
increased; in this case, the occurrence of pulmonary
edema is often potentiated by left ventricular failure
from intense systemic vasoconstriction. Measures
which prevent the bradycardia or left ventricular
overwork also protect against the pulmonary edema
of increased intracranial pressure (65).
"Bronclwmotor Tone"
This colloquialism refers to a state of partial con-
traction of bronchial smooth muscle (95, 132, 351).
An increase in "bronchomotor tone" may conceivably
affect pulmonary vascular dimensions in at least three
DYNAMICS OF PULMONARY CIRCULATION
'7'7
different ways: /) by mechanical distortion of the
pulmonary arterial tree and of the large pulmonary
veins which lie adjacent to the tracheobronchial tree;
2) by raising intra-alveolar pressure to compress the
pulmonarv capillaries and to increase, thereby, their
resistance to perfusion; and j) by increasing the
"elastance" of the lung (ig, 305, 351 ), i.e., the elastic
forces which are developed during each respiratory
cycle. Many experimental (95, 305) and clinical
observations attest to the capacity of the bronchial
smooth muscle to undergo drastic changes in tone in
response to appropriate stimulation; this severe type
of bronchospasm poses no problem in recognition.
More troublesome is the prospect that subtle changes
in "bronchomotor tone" mav escape detection (351).
As a general approach, bronchomotor tone may
reasonably be considered to remain unchanged during
the course of an experiment if: a) clinical evidences of
bronchial obstruction or dyspnea do not appear; b)
the ventilatory pattern remains unchanged; and c)
the mechanical properties of the lungs remain un-
altered (132, 153). When the nature of the experiment
precludes such clinical and experimental stability,
decision as to the influence of altered bronchomotor
tone on pulmonary hemodynamics falls to the experi-
menter.
Following complete collapse, only 10 to 15 per cent
of the cardiac output perfuses the collapsed lung (316).
Several different ways have been used to trace the
sequential changes in perfusion following bronchial
obstruction : the change in peripheral arterial oxygen-
ation (316), the change in "venous admixture'" (23)
and angiography (85). Although not entirely con-
sistent (5), the results seem to indicate that within an
hour after the bronchial obstruction, the blood flow
to the nonventilating lung is apt to decrease by 30 to 40
per cent (316). In time, the blood flow to the non-
ventilating lung decreases further; up to a month may
be required for nearly all mixed venous blood to be
excluded from the atelectatic lung and for systemic
arterial oxygenation to return toward normal values
(85). By way of contrast, the spontaneous restoration
of systemic arterial oxygenation in patients with
pneumonia or pneumothorax is more often a matter
of days than of weeks.
The observation has been made that the pulmonarv
collateral circulation may proliferate in atelectatic
areas. However, the strong possibility exists that
complications of atelectasis, such as pulmonary infec-
tion, rather than the mechanical collapse, per se, are
responsible for the expanded collateral circulation
(316).
Mechanical Compression (Atelectasis)
It is well known that mechanical factors influence
the caliber of the pulmonary blood vessels and their
resistance to blood flow. For example, at a given
hydrostatic pressure head, moderately inflated lungs
contain more blood (wider vascular calibers) than do
either collapsed or markedly distended lungs (146);
similarly, pneumothorax not only decreases the air
content of the lungs but also shrinks the vascular
calibers (380).
Atelectasis is generally believed to affect the pulmo-
nary circulation by mechanical compression. It has
been produced experimentally by bronchial obstruc-
tion (85, 94), pneumothorax (380) and sustained
hypoventilation (359). The changes following bron-
chial obstruction have been most intensively studied:
following complete occlusion of a bronchus, gas is
absorbed at a rate set by the composition of the gas,
the surface area and the rate of perfusion of the af-
fected area (94). As the gas content of the lung de-
creases, the mechanical compression of the pulmo-
nary blood vessels — particularly of the capillaries in
the collapsed alveoli — diverts the blood flow from the
atelectatic to the unaffected parts of the lung (85, 380).
Hypertonic Solutions
A particularly puzzling phenomenon has been the
occurrence of pulmonary arterial hypertension fol-
lowing the injection of hyperosmotic solutions, e.g.,
20 per cent sodium chloride, into a peripheral vein
(28). Different mechanisms have been proposed to
account for this pressor response, including selective
constriction of the superior pulmonary veins at their
entry into the left atrium (120). Recently, microscopic
examination has shown that intravascular red-cell
agglutination occurs after the injection of highly con-
centrated salt and sugar solutions, raising the possi-
bility that luminal obstruction, rather than vasocon-
striction, may underlie the pulmonary pressor response
to hypertonic solutions (333, 376).
PULMONARY VASOMOTOR ACTIVITY
It has been shown in a previous section that the
pulmonary circulation is equipped with vascular
smooth muscle and nerves, and that the pulmonarv
circulation has the ability to vasoconstrict and to
vasodilate. Much more difficult to decide is whether
1 718
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
this capacity for vasoconstriction is actually used by
the intact animal under natural conditions: the
strength of such a decision depends on the degree to
which all conceivable passive influences on the pulmo-
nary circulation have to be appraised and found
wanting. And the list of potential passive influences
is apt to be longer for experiments performed under
natural conditions than under contrived experimental
conditions (387).
Recognizing that final proof of the operation of
pulmonary vasomotor activity under natural condi-
tions is difficult to obtain, there is, nonetheless,
reasonable evidence to indicate that it does occur;
this evidence is of two types: /) the response of the
pulmonary circulation to acute hypoxia and to acute
acidosis and 2) the occurrence of pulmonary vaso-
motor reflex activity. More uncertain is the occurrence
of pulmonary vasomotor waves.
Respiratory Gases
The first experiments devoted to the effects of the
respiratory gases on the pulmonary circulation were
concerned with asphyxia. Although these experiments
did show that asphyxia elicited pulmonary hyperten-
sion, they made no attempt to distinguish which of
the respiratory gases was responsible for the rise in
pressure. Thirty years later, acute hypoxia, per se,
was shown to be capable of eliciting pulmonary
hypertension in the anesthetized dog (132). There-
after, interest in the respiratory gases was episodic
until 1946, when Euler and Liljestrand demonstrated
that acute hypoxia and acute hypercapnia evoked an
increase in pulmonary arterial pressure in the anesthe-
tized cat and that this pressor response occurred in
the face of an unchanged or a decreased left atrial
pressure (125, 268). Although these experiments were
inconclusive in some respects — e.g., the lack of meas-
urement of pulmonary blood flow or peripheral
arterial oxygen saturation or pH — they constituted
a landmark in the study of the pulmonary circulation
because of the clairvoyant hypothesis which they sug-
gested and the subsequent work which they stimu-
lated. The hypothesis consisted of three parts: /) that
a change in the composition of the inspired gas is
capable of eliciting an increase in resistance and that
this increase in resistance stems, in turn, from pulmo-
nary vasoconstriction; 2) that this vasoconstriction is
mediated by local vasomotor responses rather than
by reflexes involving the extrapulmonary portions of
the autonomic nervous system; and 3) that the vaso-
motor effects of the respiratory gases serve to adjust
alveolar perfusion to alveolar ventilation. Subsequent
experiences have done more to supply and clarify
details than to alter the general structure of the hy-
pothesis; in particular, they have distinguished be-
tween the effects of acute hypoxia, acute hypercapnia,
and respiratory acidosis on the pulmonary circulation.
Acute Hypoxia
A reduction in the fraction of oxygen in the inspired
air — generally below 1 2 per cent — has elicited an in-
crease in pulmonary arterial pressure in every species
in which it has been tested (132). In the intact, un-
anesthetized animal and man this pressor response
generally occurs when the oxygen saturation of
peripheral arterial blood drops below 80 per cent
(136). The associated increase in mean pressure is of
the order of 4 to 8 mm Hg (fig. 43) (300). Only a
small part of this increase in pressure is attributable to
an increase in pulmonary blood flow: during the
breathing of a 10 per cent oxygen mixture the in-
crease in flow rarely exceeds 30 per cent (132). Since
the usual passive determinants of pulmonary vascular
pressure — pulmonary blood volume (154), ventila-
fic. 43. Effect of acute hypoxia on
pulmonary arterial pressure. During
acute hypoxia, the systolic, diastolic,
and mean blood pressures increase. The
heart rate also increases and respiratory
fluctuations in the luminal pressures
(referred to atmosphere) become more
marked.
H.P.
H
A A A A r\ A
\J KJ Sj V \J
21% op
30
5-
0
W^£
12% o2
DYNAMICS OF PULMONARY CIRCULATION
1719
Mixed
Venous P02
Pulmonary
Venous P02
mm Hk
mm Hg
40
100
3°
105
3°
45
35
35*
table 3. Representative Oxygen Tensions oj Precapillary
and Postcapillary Pulmonary I 'essels During
I 'at urns Experimental Circumstant es
Rest; ambient air
Moderate exercise; ambient air
Bilateral hypoxia; 12% 02
Unilateral hypoxia; 5% O:
* Pulmonary venous P,,„ on the opposite side, i.e., the
hyperoxic side, exceeds 100 mm Hg.
tion (132), and left atrial pressure (125, 303) —
undergo too little change to affect the level of pulmo-
nary arterial pressure, the increase in the blood pres-
sure gradient across the lungs is generally acknowl-
edged to involve an active increase in pulmonary
vascular resistance, i.e., vasoconstriction.
In essence, the evidence for vasoconstriction during
acute hypoxia falls into three categories (132): /) the
disproportionate increase in the pressure gradient
across the lung with respect to the increment in
pulmonary blood flow (fig. 32) (132); .' ) the redistri-
bution of the pulmonary blood flow in favor of the
high-oxygen lung during unilateral hypoxia (209,
328, 408); and 2) the vasodilator effects of infused
acetylcholine during bilateral (153) and unilateral
(go) hypoxia. Despite this cumulative evidence, not
all are convinced that acute hypoxia elicits pulmo-
nary vasoconstriction (353). However, although the
evidence against pulmonary vasoconstriction is not
very substantial, it does serve to recall: a) that the
magnitude of the changes in pulmonary vascular
blood pressure is small; b) the possibility that subtle
extraneous influences, such as constriction of the
extravascular smooth muscle may mimic vasocon-
striction; and c) that the effects of acute hypoxia on
the pulmonary circulation are easily overwhelmed
by known mechanical influences, such as gravity
('32)-
The particular vascular segment, or segments, in-
volved in the pulmonary vasoconstriction has been
sought in many ways. At the moment, the experi-
ments performed under exceedingly artificial condi-
tions favor postcapillary vasoconstriction. On the
other hand, experiments in intact animals with di-
nitrophenol (which selectively lowers precapillary
oxygen tension) have demonstrated precapillary vaso-
constriction (Bergofsky et al., unpublished observa-
tions). No evidence has yet been adduced to indicate
that pulmonary venous-left atrial junctions constrict
during hypoxia. The opinion of the author is that
both the precapillary small vessels and the post-
capillary small vessels can constrict if exposed to a
sufficient degree of hypoxia (132). An idea of the
oxygen tensions which exist in the pre- and post-
capillary segments under various conditions is given
in table 3.
The notion that the small pulmonary muscular
vessels, regardless of location, constrict when exposed
to a sufficiently intense hypoxic stimulus implies that
during ambient air breathing the hypoxic mixed
venous blood may set the tone (albeit slight) of the
pulmonary ''arterioles" and, thereby, the level of the
pulmonary arterial pressure; this tonic effect would
presumably be heightened during exercise (as mixed
venous blood becomes more unsaturated) unless the
arterioles were passively widened by mechanical
influences. The experiments involving hypoxia by
airway are also complicated. In these, the prospect
exists that the hypoxic mixture may affect the pre-
capillary as well as the postcapillary segments; none-
theless, the postcapillary segments would be more
drastically affected because mixed venous blood is
ordinarily low in oxygen tension. Finally, vasocon-
striction of either segment could account for some
rearrangement of the pulmonary blood flow in pa-
tients with maldistribution of air and blood even
though mechanical influences would be expected to
be prepotent.
Several other aspects of the pressor response to
acute hypoxia warrant special emphasis: <;) the in-
crease in vascular resistance in the isolated perfused
lung — which is devoid of neurohumoral influences, of
a collateral circulation, and of extrapulmonary re-
flexes— indicates that hypoxia acts locally, i.e., either
by a direct chemo-effect on the vessel, or by way of an
intrapulmonary reflex, rather than by way of extra-
pulmonary controls (116, 305); b) the persistence of
the pressor response after ergotamine and atropine
favors a direct rather than a reflex action (116, 125);
c) severe hypoxia or anoxia, as commonly used in the
isolated lung or in artificial preparations, may not
represent the same biochemical stimulus to vascular
smooth muscle as tolerable levels of hypoxia in animal
or man (132); d) the pulmonary vasoconstriction
evoked by hypoxia has to be reconciled with the fact
that hypoxia dilates most intact vascular beds, con-
stricts isolated vessels, and dilates the placental
vessels (137); e) the biochemical mechanism by which
acute hypoxia causes smooth muscle to constrict has
not been elucidated (116);/) the catecholamines are
not involved in the pressor response to moderate
hypoxia (163); andjj) the vasoconstriction evoked by
I 720
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
©
©
ONE YEAR RESIDENT
NATIVE RESIDENT (".-} = NATIVE "SEROCHE'
fig. 44. Schematic representation of
respiratory and circulatory measure-
ments of man at altitude (14,900 feet).
The horizontal line in each panel rep-
resents a typical sea-level value. [Based
on Rotta et at. (362).]
cog
mm Hq
SATURATION
/o
m I /kg
8L00D
HEMOGLOBIN
gm/IOOml
®K
■®
(S)
20-
00-
11 I
80-
E 3 ■
1 ~
J j»
•
' ) 1 1
CARDIAC
HEART
PULMONARY
SYSTEMIC
INDEX
RATE
ARTERY
ARTERY
MEAN
MEAN
PRESSURE
PRESSURE
acute hypoxia is easily overcome by mechanical in-
fluences, such as gravity (2).
Chronic Hypoxia
Chronic hypoxia and hypoxemia are regular fea-
tures of life at high altitudes. At the Fifth Annual
Conference on Research in Emphysema, held at
Aspen, Colorado, June 15-18, 1962, Pefialoza, Sime,
Banchero, and Gamboa enlarged upon the earlier
observations of Hurtado and co-workers at Moro-
cocha (Peru) (altitude of 14,900 feet, atmospheric
oxygen tension of 80 mm Hg) (fig. 44) (362). They
confirmed, on the basis of right heart catheterization
in 38 native residents of Morococha, that mild pulmo-
nary hypertension (of the order of 41 15, 28 mm Hg)
coexisted with normal cardiac output (average of
3.71 1/min/m2) and with normal pulmonary wedge
pressure and heart rate. During strenuous supine
exercise (four- to fivefold increase in oxygen uptake),
the doubling of blood flow (from 3.65 to 7.49 1 min/
m2) was associated with a doubling of the pulmonary
arterial pressure (from 41/15, 29 to 77/40, 60). In
resting subjects, the breathing of 35 per cent oxygen
(or the infusion of acetylcholine) reduced the pulmo-
narv arterial pressures somewhat (by 20 to 25 per
cent) but not quite to normal sea level values. Restudy
of 1 1 altitude dwellers after two years at sea level dis-
closed that the blood gases, the respiration, and the
circulation had returned to virtually normal sea level
DYNAMICS OF PULMONARY CIRCULATION
[72 1
values at rest but that the increment in pulmonary
arterial pressure during exercise was still excessive for
the increment in blood flow.
Penaloza et al. also found that young children (1-5
years of age), born and raised at altitude, had more
marked pulmonary hypertension (of the order of
58/32, 45 mm Hg) than older children (of the order
of 41/18, 28 mm Hg) and adults; in this respect,
the youngsters at altitude differed strikingly from
their counterparts at sea level who achieved normal
pulmonary arterial pressures during the third to sixth
month of life.
Pulmonary hypertension and right ventricular en-
largement are characteristic not only of acclimatized
man but also of acclimatized cattle (199). Malaccli-
matization to altitude results in "mountain sickness,"
both in man and in animals. Although mountain
sickness is not a distinct clinical entity, at least two
different types have been identified, i.e., "brisket
disease" in cattle and "seroche" in man (table 4);
both seem to originate in alveolar hypoventilation.
Seroche is manifested by polycythemia, easy fatiga-
bility, and respiratory distress during exertion; its
physiological hallmarks are severe hypoxemia, hyper-
capnia, polycythemia, and pulmonary hypertension
(9, 362). Removal of the native suffering from seroche
to sea level results in a prompt clinical improvement
and, within 2 months at sea level, in return of blood
gases, circulation, and respiration at rest to virtually
normal values except for a slight residual pulmonary
hypertension (Penaloza et al.). On the other hand,
the clinical picture of brisket disease is dominated by
the consequences of severe pulmonary hypertension
and cor pulmonale, i.e., by severe right ventricular
failure, functional tricuspid insufficiency, and de-
pendent edema of the brisket. Although the clinical
pictures of seroche and brisket disease overlap some-
table 4. Chronic Mountain Sickness in Cattle and in Man*
Cattle Man
SYNONYM BRISKET DISEASE SEROCHE
Altitude 8,000 to 1,200 ft. > 12,000 ft.
Prepotent Pulmonary vaso- Severe hypoxemia
mechanism constriction and hypercapnia
Major conse- Severe pulmonary Polycythemia; mod-
quences hypertension; erate pulmonary
cor pulmonale hypertension
Clinical Congestive heart CNS disturbances;
disability failure lassitude, fatigue,
dyspnea
* Based on observations of Hecht et al. (199).
what, hypoxemia and polycythemia are far less
striking in the animals than in man (table 4).
At least part of the severe pulmonary hypertension
of brisket disease is attributable to the sphincteric
construction of the small precapillary pulmonary
vessels in cattle; presumably this anatomical arrange-
ment not only affords unusual intrinsic resistance to
blood flow but also effects an intense pulmonary
precapillary vasoconstriction in response to mild
hypoxia (199). It is not yet settled if postcapillary
events (left heart failure, constriction of the pulmonary
veins, or "throttles") are also involved in severe
brisket disease.
Thick pulmonary precapillary vessels are also found
in native residents at high altitudes. Thus, the small
precapillary vessels are thicker than their counter-
parts at sea level and smooth muscle extends further
down the pulmonary vascular tree at altitude than at
sea level. This medial hypertrophy of the pulmonary
precapillary vessels suggests that the pulmonary
hypertension of man at altitude originates in morpho-
logic changes as well as in vasomotor activity (Arias-
Stella and Saldana, unpublished observations).
Acute Hvperoxia
Enrichment of the inspired air with oxygen, or the
substitution of 100 per cent oxygen for air, is without
appreciable effect on the normal pulmonary circula-
tion (132). This lack of effect is consistent with the
notion that the resistance vessels of the normal pulmo-
nary circulation — despite the normal unsaturation of
mixed venous blood — ordinarily have little "tone."
On the other hand, oxygen-rich mixtures have been
shown to partially relieve pulmonary hypertension of
chronically hypoxic and hypoxemic animals and man
(132, 164, 389). The effectiveness of oxygen-rich mix-
tures as pulmonary vasodilators in hypoxemic states
has led to the use of oxygen-rich mixtures to relax
hypertonic pulmonary vascular smooth muscle in
nonhypoxemic states. However, there is no apparent
reason to suspect that hvperoxia will dilate pulmo-
nary vessels which are not hypoxemic.
Acute Hypercapnia
At first encounter, the published accounts of the
effects of breathing 5 to 10 per cent C02 in air on the
pulmonary circulation are utterly confusing. Un-
doubtedly, part of the confusion arises from the failure
to take into account the peculiarities of the different
preparations and experimental conditions. The situa-
tion is improved by sorting the results according to
i 72 a
HANDBOOK OF PHYSIOLOGY -" CIRCULATION II
♦ Q • 2 87 L/min
pH- 7 38
L Ac " 10 mgm V.
Q ■ 2 94 L/min
pH -7 21
L Ac ■ 56 mgm °/.
fig. 45. A continuous record of the pulmonary arterial (PA) and left ventricular (LV) blood
pressures in the dog prior to, and during, an infusion of 0.3 M lactic acid. The arrow above the
pressure tracing indicates the start of the infusion. The values for cardiac output (Ql, blood pH and
blood lactate concentration (L Ac) on the left were obtained during the control period; those on the
right are after 3 min of the infusion. Time lines occur at i-sec intervals; the duration of the entire
record is 3 min 15 sec. [Unpublished records of Bergofsky et at. (24).]
four categories: isolated lungs, unilateral hypercapnia,
controlled ventilation, and spontaneous ventilation
(132). But, even though results tend to be consistent
within each category, the differences between cate-
gories may be quite striking. Thus, in spontaneously
breathing animals and man, acute hypercapnia is
generally without effect on pulmonary hemodynamics
(132); conversely, in anesthetized animals which are
being passively ventilated, acute hypercapnia usually
increases pulmonary vascular resistance (24). Recent
observations have suggested a basis for this disparity:
for example, during anesthesia and controlled CO2
breathing — when the ventilatory response to inspired
CO2 is limited by the apparatus — respiratory acidosis
is common; on the other hand, during spontaneous
breathing — when the increase in ventilation is quite
marked — respiratory acidosis is ordinarily mild. In
the next section it will be shown that severe acidosis
increases pulmonary vascular resistance. Accordingly,
the effects of breathing CO2 on the pulmonary circula-
tion appear to depend on the degree of acidosis which
it produces.
Acute Acidosis
For a long while, observations on the isolated lung
(1 16, 305) and on the lungs perfused in situ (423) led
to the opinion that acidosis played no role in the
regulation of the pulmonary circulation. Recently, this
view was challenged by experiments on similar
preparations which not only indicated that acute
acidosis is capable of eliciting an increase in pulmo-
nary vascular resistance, but also suggested that it
might be involved in the pulmonary vascular response
to acute hypoxia (269). That acidosis can also elicit
an increase in the pulmonary vascular resistance in
the intact anesthetized dog is shown in figure 45; in
these animals, the pressor response seems to arise
from pulmonary vasoconstriction and to depend upon
the degree of acidosis rather than upon specific
anions (24, 30). It should be noted that this constrictor
effect of acidosis on pulmonary vascular smooth mus-
cle stands in marked contrast to the inhibitory effects
of acidosis on systemic vascular smooth muscle (402).
Another use of alkali and amine buffer has been to
test the idea that acidosis may underlie the pulmonary
arterial pressor response to acute hypoxia (269). This
idea could not be substantiated in normal man (24).
Instead, the conclusion was reached that acute
hypoxia and acute acidosis constitute independent
stimuli for pulmonary vasoconstriction; however, it is
conceivable that in subjects with regional hypoventi-
lation, the two separate stimuli may act synergistically
to divert pulmonary blood flow to the well-ventilated
portions of the lung (24).
Alveolar Hypoventilation
Alveolar hypoventilation may be uniform, as in
patients with kyphoscoliosis or extreme obesity (25),
or spotty, as in chronic bronchitis and emphysema.
During ambient-air breathing, the designation "alveo-
lar hypoventilation" implies a combination of alveolar
hypoxia and hypercapnia; the state achieves clinical
significance when sufficiently severe to produce sys-
temic arterial hypoxemia and respiratory acidosis (138).
Experimentally, it has been deliberately induced by
artificial underventilation during general anesthesia
(359). Largely on the basis of such experiments, it has
DYNAMICS OF PULMONARY CIRCULATION
1/2 3
been claimed that alveolar hypoventilation elicits
pulmonary arterial hypertension not only by wa\ of
acute hypoxia but also by an hypothetical "alveolo-
vascular reflex." Since mechanical underventilation
of the lungs involves an element of mechanical col-
lapse of alveoli as well as a change in the alveolar gas
composition, the operation of this special alveolar
reflex is difficult to prove. Nonetheless, this hypotheti-
cal reflex is consistent with the consensus that pulmo-
nary vasomotor effects of acute hypoxia (by airway)
are independent of systemic arterial hypoxemia (204).
Pulmonary Vasomotor Reflexes
The plentiful supply of autonomic nerves to the
lungs and of nerve fibers to the pulmonary blood
vessels has stimulated the search for direct evidence
of pulmonary reflex activity. This search has dis-
closed that numerous afferent vagal fibers and baro-
receptor endings exist in the large pulmonary arteries,
that the impulse activity of the pulmonary barorecep-
tor fibers varies with the pulsatile blood pressure in
the pulmonary artery and that the receptors are active
at the usual levels of pulmonary arterial pressure (80).
On the other hand, since the efferent limbs have not
yet been traced either to their conjunction with affer-
ent limbs or to their endings in effector cells, the
proof of their existence consists entirely of indirect
physiologic observations, e.g., systemic vasodilation as
pulmonary arterial pressure increases (6).
The pulmonary circulation is believed to participate
in a wide variety of mechano- and chemoreflexes
(207, 393). Some of these hypothetical reflexes are
conceived to be purely local, e.g., pulmonary veno-
arteriolar (6, 101, 231, 409) or alveolar-vascular
(359); these local reflexes are inaccessible for direct
appraisal. Much more tangible are the remote reflexes.
Three types of remote reflexes have been exten-
sively studied. The first is a reflex from the pulmo-
nary vessels to the systemic circulation. With rare
exception (261), this type of reflex has been ''depres-
sor" in nature, evoking bradycardia and systemic
arterial hypotension in response to a wide variety of
stimuli; the stimuli have included an increase in static
pressure at either end of the pulmonary vascular tree
or along its whole length (6, 110), chemoreflexes of
different kinds (99), and pulmonary vascular hypo-
thermia (159)-
The second type of remote reflex is a combined or
chain reflex from the pulmonary arteries to the small
pulmonary vessels on the one hand (4, 32 1 ) and to the
respiratory apparatus on the other (432). With few
exceptions (101), such a reflex has customarily been
invoked to account for the dramatic clinical syndrome
which follows multiple pulmonary emboli, i.e., the
pulmonary hypertension, the rapid shallow breathing,
the bronchoconstriction, and the decrease in periph-
eral arterial oxygenation (227). However, many of the
links in this reflex chain reaction remain speculative.
More precisely defined, but much less meaningful
with respect to function, are the chemoreflex path-
ways which connect the pulmonary arterial tree with
the respiratory apparatus (99).
The third type of remote reflex runs from the
reflexogenic areas of the carotid arterial bifurcations
and aortic arch to the pulmonary circulation (206).
To create the proper experimental setting for the
demonstration of these feeble reflexes, Daly and Daly
were obliged to resort to the "vasosensory controlled
perfused living animal" preparation in which the
pulmonary and systemic circulations could be sepa-
rately controlled. In this special preparation, intense
pressor stimulation of the systemic baroreceptors
evoked pulmonary vasodilatation; perfusion of the
carotid chemoreceptors with hypoxic or venous blood
(during interrupted bronchial arterial flow) evoked
pulmonary vasoconstriction (96). The authors are
careful to point out that the elaborate controls re-
quired to demonstrate the existence of these reflex
pathways obscure the meaning of these reflexes for the
live, intact organism (95).
Pulmonary Vasomotor Waves
Rhythmic oscillations in systemic arterial blood
pressure (Traube-Hering-Mayer waves) were first
described toward the close of the nineteenth century
(404). Although the consensus since then has favored
the view that these systemic waves reflect the rhythmic
activity of the medullary vasomotor center, not always
has irradiation from the respiratory to the vasomotor
center been excluded. Most often, the Traube-
Hering-Mayer waves have been encountered in ab-
normal or deteriorating experimental preparations;
even in the same preparation the pattern of the waves
tends to vary with respect to frequency and to ampli-
tude (229).
Infrequently, the swings in systemic arterial blood
pressure were found to be associated with swings in
pulmonary arterial blood pressure (125). And, on
rare occasion, the pulmonary arterial swings occurred
either without (379), or with barely perceptible (125),
systemic arterial waves. In these few instances, other
passive effects were not entirely excluded.
•7*4
II \M>HOOK OF I'HYsioI ( IGY
CIRCULATION II
mmHq H. F
80
10 SEC
PA
40
li I L
WW
mmm
10 SEC
t i
I 20
BA
80
40
fig. 46. Spontaneous rhythmic fluctuations in pulmonary arterial blood pressure in a young
woman with primary pulmonary hypertension. The pulmonary arterial systolic pressure is identical
with the systolic pressure in the right ventricle. The pulmonary arterial pressure waxes and wanes.
Each cycle is 1 10 sec long; the pulmonary arterial systolic pressure ranges from 57 to 74 mm Hg; the
diastolic pressure ranges from 26 to 36 mm. The pressure changes are not accompanied by parallel
changes in heart rate. The brachial arterial pressure is somewhat low. The slow cyclic variations in
pulmonary arterial pressure have no counterparts in the systemic blood pressure. (Unpublished
observations by A. P. Fishman and A. G.Jameson.)
Recently, a pulmonary arterial rhythm, unaccom-
panied by fluctuations in systemic arterial blood pres-
sure, was observed in an unanesthetized woman with
primary pulmonary hypertension (fig. 46). In this
subject, systemic hypotension coexisted with pulmo-
nary hypertension, a combination which presumably
favors the occurrence of isolated pulmonary arterial
pressure waves in the experimental animal (379).
However, while it is intuitively attractive to accept
the pulmonary arterial waves in this subject as a
manifestation of a central vasomotor rhythm — super-
imposed by the central nervous system on local pulmo-
nary vascular controls — the probability remains that
the fluctuations in pulmonary arterial pressure may
merely reflect the passive consequences of rhythmic
changes in systemic hemodynamics (125).
EFFECTS OF DRUGS
Interests in the pharmacology of the pulmonary
circulation differ: at one extreme is the use of drugs to
display the capacity of the pulmonary vessels to
undergo a change in "tone"; this has led to the stuck
of isolated perfused lungs and vascular rings. At the
other extreme is the effect of a particular drug on the
pulmonary circulation under natural conditions; this
has involved the study of the unanesthetized intact
animal or man in whom the ventilation, circulation,
and the coordinating neurohumoral systems are all
intact. Between these extremes are many shades of
interest which are not always defined or self-evident
from the experimental protocols.
It is generally difficult, in intact animal and man,
to separate the direct, local vasomotor effects of a
drug on the pulmonary circulation from its indirect,
passive effects originating from afar, i.e., in the
systemic circulation, the left heart or the ventilation.
Theoretically, this distinction should be easily made
if the pharmacological agent, such as acetylcholine,
is rapidly destroyed by contact with blood (122):
minute quantities of acetylcholine are infused into a
peripheral vein or into the pulmonary artery at a rate
carefully adjusted to avoid the classical circulatory
picture of systemic vasodilatation and cardiac inhibi-
tion; the action of the drug is then presumed to be
confined to the pulmonary circulation. During this
venous or pulmonary arterial titration with acetyl-
choline, steady-state measurements are made, not
only of pulmonary blood flow and pressures, but also
of other relevant respiratory and circulatory param-
eters (153).
Of more universal applicability is the procedure of
injecting a drug into the pulmonary artery while re-
cording blood pressures simultaneously from the
pulmonary artery, the pulmonary vein (or left
atrium), and a systemic artery as the drug traverses
the pulmonary circulation (fig. 33) (150, 187). At
first, the use of this approach required open thora-
cotomy for the cannulation of the pulmonary vessels
DYNAMICS OF PULMONARY CIRCULATION
1725
(225). However, the gradual progression from angios-
tomy cannulae (187) to combined right and left
heart catheterization (92) has made it feasible to
record simultaneously the blood pressures at both
ends of the pulmonary circulation as well as the blood
pressure in a systemic artery in the intact, unanesthe-
tized animal and man.
Predominantly Passive Effects
Certain familiar drugs seem to affect the pulmo-
nary circulation of the intact animal or man pre-
dominantly by way of the systemic circulation.
Cardinal examples are the effects of digitalis and
quinidine in subjects with left heart failure: digitalis
reduces the pulmonary hypertension of left heart
failure by improving myocardial performance; quini-
30--"
20--
10 -
15
10
5
.16--
.08--
- PA
»
-i -
P "wedge"
T_.__.
~2
r"
L"I :
_T_
J
Rp
■ V
^
-N
AMBIENT
AIR
HYPOXIA
(So02:74%)
NE
(20-67,ug7min)
fig. 47. Comparison of the effects of acute hypoxia and of
infusing norepinephrine on pulmonary vascular pressures and
resistance. PA = mean pulmonary arterial pressure;
P "WEDGE" = mean pulmonary arterial wedge pressure;
Rp = mean pulmonary vascular resistance. The solid circles
represent the average values for the group of 13 normal sub-
jects; the vertical bars represent the range. During hypoxia, the
increase in pulmonary arterial pressure was not associated with
an increase in "wedge" pressure; during norepinephrine infu-
sion, an increase in pulmonary arterial pressure was invariably
associated with an increase in "wedge" pressure. Accordingly,
calculated resistance increased during hypoxia and decreased
during norepinephrine infusion. [After Goldring et at. (163).]
dine, on the other hand, improves the emptying of
the left heart by decreasing systemic vascular resist-
ance rather than by a direct action on the myocard-
ium (87, 129).
The effects of epinephrine on the pulmonary circu-
lation have long been disputed, primarily because of
the occurrence of simultaneous changes in both the
systemic and pulmonary circulations (97, 187, 436).
However, granting that a direct pulmonary vasocon-
stricting effect can be demonstrated in special prep-
arations (355), in the intact animal the increase in
pulmonarv arterial pressure evoked by epinephrine
is almost exclusively a consequence of passive back
pressure from the left heart and systemic circulation
(187). Indeed, in the dog, excessively large doses of
epinephrine reproduce the sequence elicited by the
intracisternal implantation of fibrin (64, 372), in-
cluding pulmonary hypertension and pulmonary
edema from left heart failure. And, consistent with
the prepotent effects of epinephrine on the systemic
circulation in the dog, is the observation that in the
turtle (single ventricle), intravenous epinephrine in-
creases systemic vascular resistance without affecting
pulmonary vascular resistance (443).
Levarterenol (/-norepinephrine) also elicits an in-
crease in pulmonary arterial pressure. As in the case of
epinephrine, the capacity for pulmonary vasoconstric-
tion can be demonstrated by special techniques (355)-
However, in intact man, the increase in pulmonary
arterial pressure elicited by levarterenol is predomi-
nantly, if not exclusively, passive, i.e., secondary to an
increase in left atrial pressure (fig. 47) (163).
Histamine elicits a complex series of ventilatory and
circulatorv effects. In the isolated lung, it elicits
vasoconstriction (442); the intensity of this response
varies with the species, dose, and preparation. In
intact man, tolerable doses, which are sufficient to
elicit systemic hypotension, are without discernible
effect on pulmonary hemodynamics; whether tolera-
ble doses are inadequate to provoke pulmonary vaso-
constriction, or whether vasoconstriction does occur
and is neutralized by some concomitant passive
effects, remains unsettled (4). Finally, the induction of
severe systemic hypotension in the dog by large
quantities of histamine is associated with passive
pulmonary hypotension (100).
Pulmonary I asoconstrictors
The systemic circulation is far more sensitive to the
usual vasoconstrictor agents than is the pulmonary
circulation. However, a host of apparently unrelated
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
substances have been categorized as pulmonary vaso-
constrictors. These include serotonin (5-hydroxy-
tryptamine) (34, 302, 356) adenosine triphosphate
(4), small quantities of hypertonic saline (28), bacte-
rial endotoxins (245) and alloxan (4).
Serotonin, which has captured physiological and
clinical imaginations on many different accounts, is
also generally held to be a uniquely effective pulmo-
nary arterial and venous vasoconstrictor (43, 227).
The evidence rests largely on animal experiments
(4, 366) and on the behavior of isolated lungs (158)
since "safe" doses in intact man are without discerni-
ble pulmonary vascular effect (193). At first en-
counter, the disprepancies between the effects of
serotonin in animals and man might be attributed to a
"species" difference; however, even this refuge is un-
certain because of the diffuse and bizarre effects of
serotonin: a) its tendency to produce "temporary em-
boli,'' so that an increase in pulmonary arterial pres-
sure and an increase in pulmonary vascular resistance
may arise from transient occlusion of small pulmo-
nary vessels as well as from vasoconstriction (235), b)
its bronchoconstricting effects (366), c) its respiratory
and Bezold-like circulatory effects in the intact animal
(193), and d) the discrepancies between the doses of
serotonin used in the different experiments. Finally,
the ineffectiveness of tolerable closes of serotonin as a
pulmonary vasoconstrictor in man is consistent with
the lack of pulmonary hypertension in patients with
serotonin-secreting tumors (399).
Aside from their apparent proclivity for the venous
side of the pulmonary circulation, these pharmaco-
logical pulmonary "vasoconstrictors" have little in
common. For example, in contrast to serotonin and
alloxan, endotoxin requires contact with blood to be-
come effective (4). Moreover, hypertonic saline (333,
376), as well as serotonin, has been shown to conglu-
tinate red cells (235). It is clear that much remains
unknown about these pulmonary vasoconstrictors.
Pulmonary Vasodilatm 1
Interest in pulmonary vasodilators has been stimu-
lated both by the clinical need for therapeutic agents
to relieve pulmonary hypertension and the physio-
logic concern with pulmonary vasomotor tone. Aviado
(4) has sorted the substances which have been tested
into five groups: /) musculotropics (aminophylline);
2) parasympathomimetics (acetylcholine); 3) sym-
pathomimetics (isoproterenol); -/.) adrenergic blockers
(tolazoline); and 5) ganglionic blockers (hexa-
methonium). A good number of these have been
used to treat systemic hypertension, indicating the
complex hemodynamic changes which may be ex-
pected to complicate the interpretation of their
effects on the pulmonary circulation. It is also note-
worthy that none have yet found a place in the treat-
ment of pulmonary hypertension, and that, except
for acetylcholine, none have provided any fresh
insights into the regulation of the pulmonary circula-
tion.
Acetylcholine has achieved clinical pre-eminence
as a pulmonary vasodilator. This reputation arises
largely from recent experiences with pulmonary
hypertensive patients since previous studies on intact
animals, artificial preparations and normal man have
been contradictory (132). As mentioned previously,
the experiments which have adduced evidence for a
pulmonary vasodilating effect of acetylcholine have
exploited the rapid destruction of acetylcholine by
the cholinesterases of the blood to restrict the action
of acetylcholine to the pulmonary circulation (122,
401). The experiments have involved either a single
injection of acetylcholine into the venous circulation
or pulmonary artery (187, 192, 441), or a continuous
infusion of acetylcholine into the pulmonary arterv
at a rate (0.5 mg/min) insufficient, at least by con-
ventional tests, to affect either the lungs, the respira-
tion, the left heart, or the systemic circulation (153).
The evidence that acetylcholine elicits pulmonary
vasodilatation includes: a) a decrease in the pressure
gradient across the lungs in pulmonary hypertensive
subjects in whom the pulmonary vessels are pre-
sumably hypertonic (192, 282, 441); b) a partial or
complete reversal of the anticipated increase in
calculated pulmonary vascular resistance during
acute hypoxia (153); c) the prevention of the increase
in unilateral resistance during unilateral hypoxia by
the infusion of acetylcholine on the hypoxic side
(go); and d) a decrease in the peripheral arterial
oxygenation of patients with supposed ventilation-
perfusion abnormalities (383). The last effect is
generally believed to reflect the diversion of mixed
venous blood to hypoventilated portions of the lungs
as local hypoxic vasoconstriction is relieved by the
acetylcholine; however, alternate explanations such
as the opening of arteriovenous shunts or atelectasis
have also been proposed. It has been indicated
elsewhere that while these experiments on man are
consistent with a pulmonary vasodilating effect of
acetylcholine, they are not entirely convincing (132).
Other vasodilator substances and autonomic block-
ing agents (including spinal anesthesia) have also
been used in the attempt to elicit pulmonary vaso-
DYNAMICS OF PULMONARY CIRCULATION
1727
dilation, especially in pulmonary hypertensive sub-
jects (4, 369, 386). Granting that these agents are
often capable of relieving pulmonarv arterial hvper-
tension, it has yet to be shown that their hypotensive
effect represents pulmonary vasodilation.
CARDIOPULMONARY DISORDERS
Pulmonary Arterial Hypertension
According to the range of normal values described
earlier in this chapter, pulmonary arterial hyper-
tension exists when pulmonary arterial pressures
exceed approximately 30/18 mm Hg. Even such
mildly hypertensive levels have been found to strain
the heart if continued for a lifetime (363). Moreover,
subjects with "high-normal" pulmonary arterial
pressures at rest often become pulmonary hyper-
tensive when blood flow is increased acutely, as by
occlusion of a pulmonary artery (68) or by exercise
(132); the latter observations suggest that pulmonary
arterial hypertension occurs frequently in the course
of daily activities once pulmonary arterial pressures
reach "high-normal" levels at rest.
The causes of pulmonary arterial hypertension
may be conveniently sorted into four categories.
Three of these are mechanical (passive) : reduction
in the extent and distensibility of the pulmonary
vascular bed, increase in pulmonary blood flow, and
increase in pulmonary venous pressure; the fourth
is vasoconstriction (active). Before considering these
mechanisms separately, it should be noted that
pulmonary hypertension is more often the conse-
quence of a complex interplay of mechanisms than
of any single influence operating independently.
Moreover, in patients with cardiopulmonary dis-
orders, it is generally easier to single out the pre-
potent mechanism than to try to quantify the relative
contributions of all the mechanisms that could con-
ceivably be involved (88).
restricted vascular bed. In normal animal and
man, almost two-thirds of the lungs have to be re-
moved before pulmonary arterial pressures reach hy-
pertensive levels (89, 252). By way of contrast, there
are many pulmonary diseases which surreptitiously
reduce the number and caliber of small pulmonary
vessels and modify the distensibility of the remaining
vessels, so that even a normal pulmonary blood flow
is associated with marked pulmonary hypertension.
Examples of such diseases are pulmonary emboli,
pulmonary arteritis, interstitial fibrosis and granu-
loma, bullous emphysema, and "primary pulmonary
hypertension" (341). The architecture of the thorax
may also limit the capacity and expansibility of the
pulmonary vascular bed : thus, in subjects with severe
kyphoscoliosis, the combination of a dwarfed pulmo-
nary vascular bed and an adult cardiac output
predisposes to pulmonary hypertension (25).
increase in pulmonary blood flow. It has been
indicated previously, that in the normal pulmonary
circulation, the cardiac output has to be tripled
before pulmonary hypertensive levels are reached
(8g). In patients with congenital cardiac defects and
left-to-right shunts, pulmonary hypertension may
occur even at lower blood flows because of anatomical,
and possibly functional, changes in the vessels. An
especially interesting situation obtains in patients in
whom both the pulmonarv and systemic circulations
communicate, as in the reptilian heart, with the left
ventricle. In this case, the partition of the left ven-
tricular output between the two circulations is a
function of their relative resistances: in time, if
pulmonary vascular resistance to perfusion increases,
the pulmonary blood flow will diminish even though
the level of pulmonary hypertension remains un-
changed (440).
INCREASE IN PULMONARY VENOUS PRESSURE. The tWO
previous causes of pulmonary arterial hypertension
are unrelated to the level of the pulmonary venous
pressure. But, in the 1 30 years since James Hope, it
has become common knowledge that pulmonary
venous hypertension leads to pulmonary arterial
hypertension (213). Etiologically, pulmonary arterial
hypertension of this type generally originates either
in mitral valvular disease or left heart failure. In
dogs with acute or subacute (up to ten months)
mitral stenosis, the increment in pulmonary arterial
pressure appears to be a direct consequence of back
pressure: as pulmonary venous pressure and pulmo-
nary blood volume increase, the pulmonary capillary
and arterial pressures also rise, but not to the same
degree as the pulmonary venous pressure; since the
decrease in the pressure gradient is associated with an
unchanged cardiac output, the calculated pulmonary
vascular resistance decreases (176). Clinical counter-
parts of this experimental situation are rare but do
occur; they are characterized by complete restoration
of the pulmonary arterial blood pressure to normal
as the pulmonary venous hypertension is relieved.
The more common clinical situation is one in
1728
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
which the pulmonary arterial pressure is inordinately
high for the level of the pulmonary venous pressure,
the pulmonary blood volume is not abnormally
large, and relief of the pulmonary venous hypertension
does not completely restore the pulmonary arterial
pressure to normal levels. In such patients, the
persistence of pulmonary arterial hypertension after
relief of the back pressure is attributable to secondary-
effects, i.e., to anatomical changes in the lungs and
vessels, possibly abetted by constriction of the small
pulmonary arteries (238).
One prevalent notion about chronic pulmonary
venous hypertension is that it elicits "protective"
vasoconstriction of pulmonary precapillary vessels.
Although it is self-evident that pulmonary arterial
hypertension must occur if sufficient forward flow is
to continue in the face of pulmonary venous hyper-
tension, the teleological advantage of pulmonary
precapillary vasoconstriction is not entirely clear:
heightened "arteriolar" tone would increase the
pressure work of the right heart and only reduce
capillary blood pressure if it succeeded in reducing
the pulmonary capillary blood flow. Teleologically,
the prevention of undue filtration pressures in the
pulmonary capillaries would be more economically
accomplished by quieting the heart rather than by-
increasing the right ventricular work. Indeed, the
inability to resolve the question of protective vaso-
constriction again emphasizes that pulmonary vaso-
motor activity is exceedingly difficult to recognize
in the abnormal pulmonary circulation, particularly
when structural changes have extended beyond the
vessels into the surrounding lung.
PULMONARY ARTERIAL VASOCONSTRICTION. In normal
dog, cat, and man, pulmonary arterial vasocon-
striction rarely evokes more than a mild pulmonary
hypertension. On the other hand, in cattle, con-
traction of the sphincteric pulmonary arterioles
often effects dramatic increases in pulmonary arterial
pressure (199). This correlation between vascular
structure and the intensity of the pulmonary vascular
response raises the prospect that pulmonary vascular
disease may, by thickening vascular media, enable
the small muscular vessels to contract with unusual
vigor. However, this ingenious notion has yet to be
critically tested (181).
Cor Pulmonale
Pulmonary hypertension attracts clinical attention
when it causes the right heart to enlarge (dilate or
BRONCHITIS
AND
EMPHYSEMA
/
\
RESTRICTED
PU LMONARY
VASCULAR BED
ALVEOLAR
HYPOVENTILATION
\
/
PU LMONARY
H Y PERTENSION
COR
PULMONALE
HEART
FAILURE
fig. 48. The evolution of cor pulmonale and right heart
failure in chronic pulmonary emphysema. Alveolar hypo-
ventilation contributes to pulmonary hypertension by way of
hypoxia and respiratory acidosis: hypoxia elicits pulmonary
vasoconstiiction, polycythemia, hypervolemia, increased blood
viscosity, and increased cardiac output; acidosis elicits pulmo-
nary vasoconstriction.
hypertrophy) or to fail. The term cor pulmonale is
generally reserved for right ventricular enlargement
which originates either in diffuse pulmonary disease
or in ineffective performance of the chest bellows. As
a rule, pulmonary hypertension underlies cor pul-
monale; in some types of lung disease, particularly
those associated with hypoxemia, the abnormal
pressure work of the right heart may be supple-
mented by an abnormally high cardiac output, i.e.,
flow work (89, 339, 341).
It has become clear that the genesis of cor pul-
monale is to be sought in the mechanisms which
ordinarily determine the normal pulmonarv arterial
pressure; only the combinations and the prepotent
influences differ. For example, in diffuse interstitial
diseases of the lung (e.g., "alveolar-capillary block")
anatomic changes in pulmonary vessels and paren-
chvma operate without benefit of increased flow or
hypoxia. On the other hand, in the concentric alveolar
hypoventilation of extreme obesity, respiratory
paralysis or kyphoscoliosis, hypoxia and respiratory
acidosis elicit pulmonary hypertension in subjects
with normal lungs. Finally, in the most common
DYNAMICS OF PULMONARY CIRCULATION
1729
fig. 49. Dye-dilution curve inscribed
by densitometer from peripheral artery
following injection of Evan's blue dye
(T-1824) into the pulmonary artery.
The upper curve is normal. The short
appearance time and abnormal initial
deflection of the lower curve are charac-
teristic of pulmonary arteriovenous
shunts.
cause of cor pulmonale, i.e., chronic bronchitis and
emphysema, a combination of anatomic changes,
hypoxia and acidosis are involved : destruction of
alveolar capillaries sets the stage by restricting the
pulmonary vascular bed, generally without evoking
pulmonary hypertension; the picture is completed by
disturbances of alveolar ventilation and perfusion —
usually incidental to an acute bronchitis — which
superimpose the vasoconstriction of hypoxia and
respiratory acidosis on the structural changes (fig. 48)
(34')-
It is clinically and physiologically important to
recognize the occurrence of right heart failure in
patients with cor pulmonale. Prior to heart failure,
the enlarged right ventricle functions normally: it is
filled by an atrial inflow pressure of a few mm Hg, it
empties approximately half of its volume during each
ejection and it increases its output during exercise in
accord with metabolic requirements. The first signs
of right ventricular failure generally appear during
exercise: as ventricular emptying is compromised,
the mean filling pressure increases to abnormal levels
(7-10 mm Hg) and the increase in cardiac output is
no longer commensurate with the increase in oxygen
uptake (341 ).
Pulmonary Edema
In 1878, Welch produced pulmonary edema in
rabbits by either ligating the aorta or compressing
the left ventricle. He attributed the pulmonary edema
to the pulmonary congestion and pulmonary venous
BEFORE VALVULOTOMY
AFTER VALVULOTOMY
NK
to M M U U M A
MM
Hg
-18O
40
J 0
vL^UUUUUUUuU^
fig. 50. Blood pressures recorded during open thoracotomy in a patient with pulmonic stenosis.
Before valvulotomy. Marked right ventricular hypertension coexists with systemic hypotension. The
pulmonary arterial pressure pulse is vibratory. After valvulotomy. The right ventricular pulmonary
hypertension has been considerably relieved. The pulmonary arterial pressure has increased and the
pressure pulse is characteristic of pulmonic insufficiency. [After Himmelstein, el al. (210).]
'- ;o
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
hypertension which followed the induced (transient)
imbalance between the outputs of the two ventricles
(411).
Since then, many other experimental procedures
have been used to produce pulmonary edema:
vagotomy, vagal stimulation, intravenous infusion of
fluid, left heart failure, increase in intracranial
pressure, exhibition of epinephrine, and exhibition of
ammonium chloride (411)- These share a common
denominator: an excessively high pulmonary venous
and capillary pressure. Accordingly, they are con-
sistent with Welch's hypothesis; and, the origin of
the pulmonary edema which these procedures effect
is to be regarded in the light of Starling's law of
transcapillary exchange (1 74).
It would be misleading to imply that an inordinate
filtration pressure in the pulmonary capillaries
underlies all types of experimental and clinical
pulmonary edema. For example, the pulmonary
edema caused by a-naphthylthiourea does seem to
depend on an increase in capillary permeability
(114); an increase in capillary permeability has also
been postulated to account for the bilateral pulmonary
edema which follows the injection of a starch sus-
pension into a lobar pulmonary artery (227). How-
ever, continued emphasis on hemodynamic balances
promises to be rewarding for several reasons: a) the
usual forms of pulmonary edema do seem explicable
in terms of the usual determinants of transcapillary
exchange of water, solutes, and colloids, i.e., in terms
of Starling's law (174, 385, 411); b) many earlier
types of so-called "'neurogenic" pulmonary edema
disappeared when subjected to analysis in terms of
conventional hemodynamic parameters (64, 372);
< ) uncertainties as to the precise mechanisms involved
in special types of pulmonary edema are bound to
prevail until elusive parameters, such as capillary
pressure, volume, and permeability on the one hand,
and the role of the lymphatics on the other, can be
precisely measured and defined in quantitative terms
(325); and d) mysterious influences should only be
given credence when the local hemodynamic and
physicochemical mechanisms operating across capil-
lary walls have been taken into full account, and found
wanting (217).
Pulmonary Hypotension
During bleeding to the point of systemic arterial
hvpotension as well as during traumatic and his-
tamine shock, the circulating blood volume, the
cardiac output, and the central venous pressures
decrease (100, 289, 391). However, despite the pro-
gressive decline in systemic arterial and left atrial
blood pressures, the pulmonary arterial pressure
tends to stabilize at approximately two-thirds of its
initial value. This stability presumably involves the
gradual closure of portions of the pulmonary vascular
tree as intraluminal pressures in these areas fall. As a
result of the preferential closure of certain portions of
the pulmonary vascular tree during systemic hypo-
tension, the affected portions of the lungs become
excessively ventilated for their perfusion, leading to
an appreciable arterial-alveolar difference in carbon
dioxide tension (of the order of 8 mm Hg) and to the
creation of an "alveolar dead space." Restoration of
the circulating blood volume raises the pulmonary
arterial pressure to supracontrol values even though
slight systemic arterial hypotension persists (160).
Pulmonarx Arteriovenous Fistula
The surgical production of a pulmonary arterio-
venous anastomosis is associated with a decrease in
systemic arterial oxygenation and in pulmonary
arterial (mean) pressure. The subsequent course of
the experimental animal, as well as the natural
history of the human subject with a pulmonary
arteriovenous fistula (155), is determined by the size
of the shunt and the degree of systemic arterial
hypoxemia which it effects. If systemic hypoxemia is
sufficiently marked, a considerable polycythemia will
ensue leading, in turn, to an increase in the viscosity
of the blood, an increase in the resistance to blood
flow through the usual resistance vessels and the
diversion of more and more of the right ventricular
output through the low-resistance shunt (fig. 49).
Pulmonic Stenosis
A hindrance to the exit of blood from the right
ventricle occurs commonly as a congenital cardiac
malformation; either the valve or the infundibulum
or the main pulmonary artery may be the seat of the
stenosis. Experimentally, stenosis of the pulmonary
artery has been produced in different ways (12). In
all, severe narrowing of the lesion is necessary before
the right ventricle becomes strained.
In the absence of an abnormally large blood flow
across the pulmonary valve, the physiologic hallmark
of pulmonic stenosis is right ventricular hypertension
coupled with a systolic blood pressure gradient
between the right ventricle and pulmonary artery
(fig. 50). In acute animal experiments, a constriction
DYNAMICS OF PULMONARY CIRCULATION
1 731
of the pulmonary arterial lumen of at least 40 per
cent is needed to raise systolic pressure appreciablv
in the right ventricle; greater degrees of constriction
are needed to produce right ventricular failure, i.e.,
dilatation of the right heart, abnormally high end-
diastolic pressures in the right ventricle, and tri-
cuspid regurgitation (12). Parenthetically, it may be
noted that pulmonic stenosis is an excellent physio-
logical tool for stimulating the proliferation of the
pulmonary collateral arterial circulation (263, 264).
Pulmonary I 'alvular Insufficiency
Pulmonary valvular insufficiency has been
produced experimentally in dogs (123) and during
remedial cardiac surgery in man (fig. 50). After
avulsion of the valve, not only does the pulmonary
arterial diastolic pressure fall to right-ventricular
diastolic levels, but a systolic right ventricular-
pulmonary arterial pressure gradient may also appear.
This gradient is a consequence of unusually rapid and
turbulent flow during systole rather than of pulmonic
stenosis (123).
Pulmonic insufficiency is generally regarded as a
benign lesion : in dogs, performance on the treadmill
as well as end-diastolic pressures in the right ven-
tricle remains normal after months of exercise and
despite right ventricular systolic pressures approxi-
mating 100 mm Hg (12). However, pulmonic in-
sufficiency may bring the heart closer to the brink of
its reserve so that an additional lesion, e.g., tricuspid
insufficiency may precipitate overt heart failure (12).
Whether the cardiac reserve is sufficient to tolerate
pulmonary valvular insufficiency for a lifetime, or
only for a few years, remains to be established.
REFERENCES
1. Ambrus, C. M., J. L. Ambrus, G. C. Johnson, E. W.
Packman, W. S. Chernick, N. Back, and J. W. E.
Harrison. Role of the lungs in regulation of the white
blood cell level. Am. J. Physiol. 178: 33, 1954.
2. Arborelius, M., Jr., G. Lundin, L. Svanberc, and J. G.
Defares. Influence of unilateral hypoxia on blood flow
through the lungs in man in lateral position. J. Appl.
Physiol. 15:595, i960.
3. Asmussen, E., and M. Nielsen. The cardiac output in
rest and work determined simultaneously by the acetylene
and dye injection methods. Acta Physiol. Scand. 27: 217,
'952-
4. Aviado, D. M. The pharmacology of the pulmonary
circulation. Pharmacol. Revs. 12: 159, 1690.
5. Aviado, D. M. Effects of acute atelectasis on lobar blood
flow. Am. J. Physiol. 198: 349, i960.
6. Aviado, D. M., Jr., and C. F. Schmidt. Reflexes from
stretch receptors in blood vessels, heart and lungs. Physiol.
Reus. 35:247, 1955.
7. Bainbridge, F. A. The Physiology of Muscular Exercise.
(3d ed.), rewritten by A. V. Bock and D. B. Dill. London:
Longmans, Green, 1931.
8. Baltisberger, W. Ueber die glatte Muskulatur der
menschlichen Lunge. Z. Anal. Entwicklungschichle 61 :
249, 1 92 1.
9. Barcroft, J. The Respiratory Function of the Blood. Part I.
Lessons from High Altitudes. Cambridge: Cambridge Univ.
Press, 1 913.
10. Barcroft, J. Features in the Architecture of Physiological
Function. London: Cambridge, 1934.
11. Barer, G. R., and E. Nusser. Pulmonary blood flow in
the cat. The effect of positive pressure respiration. J.
Physiol., London 138: 103, 1957.
12. Barger, A. G, V. Richards, J. Metcalfe, and B.
Gunther. Regulation of the circulation during exercise.
Am. J. Physiol. 184:613, 1956.
13. Bartels, H., R. Beer, E. Fleischer, H. J. Hoffheinz,
J. Krall, G. Rodewald, J. Wenner, and I. Witt.
Bestimmung von Kurzschlussdurchblutung und Diffu-
sionkapazitat der Lunge bei Gesunden und Lungen-
kranken. Pfliigers Arch. ges. Physiol. 261 : 99, 1955.
14. Bartels, H., and G. Rodewald. Die alveolar-arterielle
Sauerstoffdruckdifferenz und das Problem des Gasaustau-
sches in der menschlichen Lunge. Pfliigers Arch. ges.
Physiol. 258: 163, 1953.
15. Bates, D. V., C. J. V'aris, R. E. Donevan, and R. V.
Christie. Variations in the pulmonary capillary blood
volume and membrane diffusion component in health
and disease. J. Clin. Incest. 39: 1401, 1960.
16. Bauman, A., M. A. Rothschild, R. S. Yalow, and S. A.
Berson. Pulmonary circulation and transcapillary ex-
change of electrolytes. J. Appl. Physiol. 11: 353, 1957.
17. Baxter, I. G, and J. VV. Pearce. Simultaneous measure-
ment of pulmonary arterial flow and pressure using con-
denser manometers. J. Physiol., London I 15: 41 0, 1 95 1.
18. Bavliss, L. E. Translocation of solutes in animals and
man. In: Deformation and Flow in Biological Systems, edited
by A. Frey-Wyssling. New York: Interscience, 1952, p.
355-
19. Bayliss, L. E., and G. W. Robertson. The visco-elastic
properties of the lungs. Quart. J. Exptl. Physiol. 29: 27,
■939-
20. Bazett, H. C, and P. Bard. The pulmonary circulation
and the respiratory variations in the systemic circulation.
In: Medical Physiology (10th ed.), edited by P. Bard. St.
Louis : Mosby, I 956, p. 205.
21. Bekauri, N. V., A. I. I l'in a, and A. V. Tonkikh. Con-
cerning the physiology of the pulmonary circulation.
Direct visualization of the pulmonary circulation in
warm blooded animals. Fiziol. Zhur., V.S.S.R. 40: 295,
!954-
22. Bell, A. L. L., Jr., W. F. Haynes, Jr., S. Shimomura,
1732
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and D. P. Dallas. Influence of catheter tip position on
pulmonary wedge pressures. Circulation Research IO. 215,
1962.
23. Berggren, S. M. The oxygen deficit of arterial blood
caused by nonventilating parts of the lung. Ada Physiol.
Scand. 4 Suppl. 1 I : 1942.
24. Bergofsky, E. H., D. E. Lehr, M. A. Tuller,
M. Rigatto, and A. P. Fishman. The effects of acute
alkalosis and acidosis on the pulmonary circulation. Ann.
N.Y. Acad. Sci. 99:626, 1961.
25. Bergofsky, E. H., G. M. Turino, and A. P. Fishman.
Cardiorespiratory failure in kyphoscoliosis. Medicine 38:
263, 1959-
26. Bernstein, W. H., E. M. Fierer, M. H. Laszlo, P.
Samet, and R. S. Litwak. The interpretation of pul-
monary artery wedge (pulmonary capillary) pressures.
Brit. Heart J. 22: 37, i960.
27. Beutner, A. Ueber die Strom und Druckkrafte des Blutes
in der Arteria und Vena Pulmonalis. Z. rat. Med. n. F.
2:97, 1852.
28. Binet, L., and M. Burstein. Sur les effets vasomoteurs
locaux du serum sale hypertonique injecte par voie intra-
arterielle. J. physiol. pathol. gen. 44: 217, 1952.
29. Bjorkmann, S. Bronchospirometrie Acta Med. Scand.
Suppl. 56: 1, 1934.
30. Bjurstedt, H., G. Liljestrand, and G. Matell. Ex-
periments on pulmonary circulation and gas exchange.
In : Problems oj Pulmonary Circulation. Ciba Foundation
Study Group No. 8, edited by A. V. S. de Reuck and M.
O'Connor. Boston: Little, Brown, 1961, p. 63.
31. Bloomer, VV. E., W. Harrison, G. E. Lindskog, and
A. A. Liebow. Respiratory function and blood flow in
the bronchial artery after ligation of the pulmonary
artery. Am. J. Physiol. 157: 317, 1949.
32. Bolt, VV., and H. Rink. Studien zur regionalen Analyse
der Lungenventilation und Lungenzirkulation. Thorax-
chirurgie 29: 5, 1958.
33. Bondurant, S, J. Mead, and C. D. Cook. A re-evalua-
tion of effects of acute central congestion on pulmonary
compliance in normal subjects. J. Appl. Physiol. 15: 875,
i960.
34. Borst, H. G., E. Berglund, and M. McGregor. The
effects of pharmacologic agents on the pulmonary cir-
culation in the dog. Studies on epinephrine, norepi-
nephrine, 5-hydroxytryptamine, acetylcholine, histamine
and aminophylline. J. Clin. Invest. 36: 669, 1 957.
35. Borst, H. C, E. Berglund, J. L. Whittenberger, J.
Mead, M. McGregor, and C. Collier. The effect of
pulmonary vascular pressures on the mechanical proper-
ties of the lungs of anesthetized dogs. J. Clin. Invest. 36:
1708, 1957.
63. Borst, H. G., M. McGregor, J. L. Whittenberger,
and E. Berglund. The influence of pulmonary arterial
and left atrial pressures on pulmonary vascular resistance.
Circulation Research 4: 393, 1956.
37. Bosman, R., A. J. Honour, G. de J. Lee, R. M.
Marshall, and F. D. Stott. Instantaneous pulmonary
blood flow measurement in man. J. Physiol., London 159:
15P, 1961.
38. Bostroem, B., and J. Piiper. Uber arterio-venose Anas-
tomosen und Kurzschlussdurschblutung in der Lunge.
PHugers Arch. ges. Physiol. 261 : 165, 1955.
39. Bowditch, H. P., and G M. Garland. The effect of the
respiratory measurements on the pulmonary circulation.
J. Physiol., London 2:91, 1879.
40. Bradford, J. R., and H. P. Dean. The pulmonary cir-
culation. J. Physiol., London 16: 34, 1894.
41. Bradley, S. E., P. A. Marks, P. C Reynell, and J.
Meltzer. The circulating splanchnic blood volume in
dog and man. Trans. Assoc. Am. Physicians 66 : 294, 1953.
42. Brandfonbrenner, M., G. M. Turino, A. Himmelstein,
and A. P. Fishman. Effects of occlusion of one pulmonary
artery on pulmonary circulation in man. Federation Proc.
■7: 19. I958-
43. Braun, K , and S. Stern. Pulmonary and systemic blood
pressure response to serotonin : role of chemoreceptors.
Am. J. Physiol. 201 : 369, 1 96 1.
44. Braunwald, E., J. T. Binion, W. L. Morgan, Jr., and
S. J. Sarnoff. Alterations in central blood volume and
cardiac output induced by positive pressure breathing
and counteracted by metaraminol (Aramine). Circulation
Research 5: 670, 1957.
45. Braunwald, E., E. C. Brockenbrough, C. J. Frahm,
and J. Ross, Jr. Left atrial and left ventricular pressures
in subjects without cardiovascular disease. Observations
on eighteen patients studied by transseptal left heart
catheterization. Circulation 24: 267, 1961.
46. Braunwald, E., A. Cournand, and A. P. Fishman.
Evaluation in a model of Stewart-Hamilton and Bradley
methods for measurement of volume of vascular segments.
Federation Proc. 14: 17, 1955.
47. Braunwald, E., A. P. Fishman, and A. Cournand.
Time relationship of dynamic events in the cardiac cham-
bers, pulmonary artery and aorta in man. Circulation
Research 4: 100, 1956.
48. Braunwald, E., and E. R. Kelly. The effects of exercise
on central blood volume in man. J. Clin. Invest. 39: 413,
i960.
49. Brecher, G. A. Venous Return. New York: Grune
& Stratton, 1956.
50. Brenner, O. Pathology of the vessels of the pulmonary
circulation. Arch. Internal Med. 56: 211, 1935.
51. Briehl, R. W., and A. P. Fishman. Principles of the
Bohr integration procedure and their application to
measurement of diffusing capacity of the lung for oxygen.
J. Appl. Physiol. 15: 337, i960.
52. Briscoe, W. A. A method for dealing with data concern-
ing uneven ventilation of the lung and its effects on blood
transfer. J. Appl. Physiol. 14: 29 1, 1 959.
53. Brofman, B. L., B. L. Charms, P. M. Kohn, J. Elder,
R. Newman, and M. Rizika. Unilateral pulmonary
artery occlusion in man. Control studies. J. Thoracic
Surg. 34:206, 1957.
54. Brown-Sequard, C. E. On the production of hemorrhage,
anemia, edema, and emphysema in the lungs by injuries
to the base of the brain. Lancet 1:6, 1871.
55. Bruner, H. D., and C. F. Schmidt. Blood flow in the
bronchial artery of the anesthetized dog. Am. J. Physiol.
148:648, 1947.
56. Burch, G. E., and R. B. Romney. Functional anatomy
and "throttle valve" action of the pulmonary veins.
Am. Heart J. 47: 58, 1954.
57. Burger, J. W., and S. E. Bradley. The general form of
the circulation in the dogfish, Squalus Acanthias. J.
Cellular Comp. Phvsiol. 37: 389, 1951.
58. Burrows, B., A. H. Niden, C. Mittman, R. C. Talley,
DYNAMICS OF PULMONARY CIRCULATION
'733
59
63
and W. R. Barclay. Non-uniform pulmonary diffusion 78.
as demonstrated by the carbon monoxide equilibration
technique: experimental results in man. J. Clin. Invest.
39 : 943. l°f>°- 79-
Burton, A. C. Relation of structure to function of tissues
of wall of blood vessels. Physiol. Revs. 34: 6ig, 1954.
60. Burton, A. C. On the physical equilibrium of small
blood vessels. Am. J. Physiol. 164: 319, 1951 . 80.
61. Burton, A. C. The relation between pressure and flow
in the pulmonary bed. In: Pulmonary Circulation, edited by
W. Adams and I. Veith. New York: Grune & Stratton,
1959. P- a6- 8l-
62. Burton, A. C, and D. J. Patel. Effect on pulmonary
vascular resistance of inflation of the rabbit lung. J.
Appl. Physiol. 12: 239, 1958.
Calabresi, P., and W. H. Abelman. Porto-caval and
porto-pulmonary anastomoses in Laennec's cirrhosis and 82.
heart failure. J. Clin. Invest. 36: 1257, 1957.
64. Cameron, G. R., and S. N. De. Experimental pulmonary
oedema of nervous origin. J. Pathol. Bacteriol. 61 : 375, 83.
>949-
65. Campbeli , G. S., F. J, Haddy, W. L. Adams, and M. B.
Visscher. Circulatory changes and pulmonary lesions in
dogs following increased intracranial pressure, and the 84.
effect of atropine upon such changes. Am. J. Physiol. 158:
96. '949- 85-
66. Campbell, H. The resistance to the blood flow. J. Physiol.,
London 23: 30 1, 1 898.
Canfield, R. E., and H. Rahn. Arterial-alveolar N? 86.
gas pressure differences due to ventilation-perfusion
variations. J. Appl. Physiol. 10 : 165, 1957. 87.
Carlens, E., H. E. Hansen, and B. Nordenstrom.
Temporary unilateral occlusion of the pulmonary artery.
J. Thoracic. Surg. 22:527, 1951. 88.
Carlill, S. D., and H. N. Duke. Pulmonary vascular
changes in response to variations in left auricular pres- 89.
sure. J. Physiol., London 1 33 : 275, 1956.
70. Carlill, S. D. , H. N. Duke, and M. Jones. Some ob-
servations on pulmonary hemodynamics in the cat. J.
Physiol., London 136: 112, 1957. go.
71. Caro, C. G., and D. A. McDonald. The relation of
pulsatile pressure and flow in the pulmonary vascular
bed. J. Physiol., London 157:426, 1961. 90a
72. Castigli, G. I vasi sanguigni del polmone di Bos taurus.
Arch. Ital. anat. embriol. 59: 283, 1954.
73. Chapman, C. B., O. Baker, J. Reynolds, and F. Bonte.
Use of biplane cinefluorography for measurement of 91.
ventricular volume. Circulation 18: 1 1 05, 1958.
74. Chapman, C. B., H. L. Taylor, C. Borden, R. V.
Ebert, and A. Revs. Simultaneous determinations of the
resting arterio-venous oxygen difference by the acetylene 92.
and direct Fick methods. J. Clin. Invest. 2g: 651, 1950.
75. Chidsey, C. A. hi, H. W. Fritts, Jr., G. P. Zooche, A.
Himmelstein, and A. Cournand. Effect of acetylcholine 93.
on the distribution of pulmonary blood flow in patients
with chronic pulmonary emphysema. Malattie cardio-
vascolari 1 : 15, i960. 94.
76. Chinard, F. P., and T. Enns. Transcapillary pulmonary
exchange of water in the dog. Am. J. Physiol. 178: 197, g5.
'954-
77. Chinard, F. P., T. Enns, and M. F. Nolan. Diffusion g6.
and solubility factors in pulmonary inert gas exchanges.
J. Appl. Physiol. 16. 831, 1 96 1.
67
68
69
Clements, J. A., R. F. Hustead, R. P. Johnson, and I.
Gribetz. Pulmonary surface tension and alveolar stability.
J. Appl. Physiol. 16:444, 1 96 1.
Cockett, F. B., and C. C. N. Vass. A comparison of the
role of the bronchial arteries in bronchiectasis and in
experimental ligation of the pulmonary artery. Thorax
6: 268, 1 95 1.
Coleridge, J. C. G., and C. Kidd. Relationship between
pulmonary arterial pressure and impulse activity in
pulmonary arterial baroreceptor fibres. J. Physiol., London
'58: '97. >96'-
Coleridge, J. C. G., C. Kidd, and J. A. Sharp. The
distribution, connexions and histology of baroreceptors
in the pulmonary artery, with some observations on the
sensory innervation of the ductus arteriosus. J. Physiol.,
London 1 56 : 59 1 , 1 96 1 .
Coleridge, J. C. G., and R. J. Linden. The measure-
ment of effective atrial pressure. J. Physiol., London 126:
3°4. 1954-
Connolly, D. G, J. W. Kirklin, and E. H. Wood. The
relationship between pulmonary artery wedge pressure
and left atrial pressure in man. Circulation Research 2
434. >954-
Connolly, D. C, and E. H. Wood. The pulmonary vein
wedge pressure in man. Circulation Research 3: 7, 1955.
Coryllos, P. N., and G. L. Birnbaum. The circulation
in the compressed, atelectatic and pneumonic lung.
A.M. A. Arch. Surg. 19: 1346, 1929.
Cotton, F. S. Studies in center of gravity changes. Aus-
tralian J. Exptl. Biol. Med. Sci. 8: 53, 193 1.
Cournand, A. Recent observations on the dynamics of
the pulmonary circulation. Bull. N.Y. Acad. Med. 23:
27, '947-
Cournand, A. Cardio-pulmonary function in chronic
pulmonary disease. Harvey Lectures 46: 68, 1950.
Cournand, A. Control of the pulmonary circulation in
normal man. In: Circulation. (Proceedings of the Harvey
Tercentenary Congress), edited by J. McMichael. Oxford:
Blackwell Sci. Pub. 1958, p. 218.
Cournand, A. Pulmonary circulation. Its control in man,
with some remarks on methodology. Am. Heart J. 54:
172, 1957-
Cournand, A. Historical development of the concepts of
pulmonary circulation. In: Pulmonary Circulation, edited
by W. Adams and I. Veith. New York: Grune & Strat-
ton. 1959. P- '•
Cournand, A. Air and blood. A historical account of
their conjunction in the lungs. In : The Circulation of the
Blood, Men and Ideas, edited by A. P. Fishman and D. W.
Richards. New York : Oxford Univ. Press. In press.
Cournand, A., and H. A. Ranges. Catheterization of the
right auricle in man. Proc. Soc. Exptl. Biol. Med. 46: 462,
1941.
Cudkowicz, L., W. H. Abelmann, G. E. Levinson, G.
Katznelson, and R. M. Jreissaty. Bronchial arterial
blood flow. Clin. Sci. 19: 1, i960.
Dale, W. A., and H. Rahn. Rate of gas absorption during
atelectasis. Am. J. Physiol. 170: 606, 1952.
Daly, I. de B. Intrinsic mechanisms of the lung. Quart.
J. Exptl. Physiol. 43: 2, ig58.
Daly, I. de B., and M. de B. Daly. The nervous control
of the pulmonary circulation. In: Problems of Pulmonary
Circulation, Ciba Foundation Study Group No. 8, edited
:734
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
bv A. V. S. de Reuck and M. O'Connor. Boston: Little,
Brown, 1961.
97. Daly, M. de B., and C. P. Luck. The effects of adrenaline
and noradrenaline on pulmonary hemodynamics with
special reference to the role of reflexes from carotid sinus
baroreceptors. J. Physiol., London 145: 108, 1959.
98. Da Vinci, Leonardo. Quaderm D'Anatomia. I. Tredici
Fogli Delia Royal Library di Windsor. Christiania;
Dybwad. Folio 3, Recto, MCMXI.
99. Dawes, G. Reflexes originating in the pulmonary cir-
culation. In: Pulmonary Circulation, edited by W. Adams
and I. Veith. New York: Grune & Stratton 1959, p. 57.
100. Delaunois, A. L., R. Kordecki, H. Polet, and J.
Rvzewski. Cardiac output, arterial blood pressure and
pulmonary arterial pressure in histamine shock. Arch,
intern, pharmacodynamic 120: 114, 1959.
101. Denolin, H. Contribution a l'etude de la circulation
pulmonaire en clinique. Acta Cardiol. Suppl. X, 1961.
102. Deuchar, D. C, and R. Knebel. The pulmonary and
systemic circulations in congenital heart disease. Brit.
Heart J. 14: 225, 1952.
103. Dexter, L., J. VV. Dow, F. W. Haynes, J. L. Whitten-
berger, B. G. Ferres, W. T. Goodale, and 11. K.
Hellems. Studies of the pulmonary circulation in man at
rest. Normal variations and the interrelation between
increased pulmonary blood flow, elevated pulmonary
arterial pressure, and high pulmonary (capillary) pres-
sure. J. Clin. Invest. 29: 602, 1950.
104. Dexter, L., J. L. Whittenberger, F. W. Haynes, W. T.
Goodale, R. Gorlin, and C. G. Sawyer. Effect of exer-
cise on circulatory dynamics of normal individuals. J.
Appl. Physiol. 3:439. ' 95 • •
105. Dirken, M. N. J., and H. Heemstra. The adaptation of
the lung circulation to ventilation. Quart. J. Exptl. Physiol.
34:2'3. I948-
106. Dock, D. S., W. L. Kraus, L. B. McGuire, J. W.
Hyland, F. W. Haynes, and L. Dexter. The pulmonary
blood volume in man. J. Clin. Invest. 40: 317, 1 96 1 .
107. Dollery, C. T., and J. B. West. Regional uptake of
radioactive oxygen, carbon monoxide and carbon dioxide
in the lungs of patients with mitral stenosis. Circulation
Researc h 8 : 765, i960.
108. Donald, K. W., J. M. Bishop, and O. L. Wade. A study
of minute to minute changes of arteriovenous oxygen
content difference, oxygen uptake and cardiac output
and rate of achievement of a steady state during exercise
in rheumatic heart disease. J. Clin. Invest. 33: 1 146, 1954.
109. Donald, K. W., J. M. Bishop, G. Gumming, and O. L.
Wade. The effect of exercise on the cardiac output and
circulatory dynamics of normal subjects. Clin. Sci. 14:
37. '955-
110. Donnet, V., P. Zwirn, and J. L. Ardisson. Pressure
sensitivity of the pulmonary arteries in the dog. Signifi-
cance of Schwiegk's reflex. Compt. rend. soc. biol. 145: 736,
'951-
111. Dow, P. Estimations of cardiac output and central blood
volume by dye dilution. Physiol. Revs. 36: 77, 1956.
112. Doyle, J. T, J. L. Patterson, Jr., J. V. Warren, and
D. K. Detweiler. Observations on the circulation of
domestic cattle. Circulation Research 8: 4, i960.
113. Doyle, J. T, J. S. Wilson, E. H. Estes, and J. V.
Warren. The effect of intravenous infusion of physiologic
saline solution on the pulmonary arterial and pulmonary
capillary pressure in man. ./. Clm. Invest. 30: 345, 1951.
1 14. Drinker, C. K. Pulmonary Edema and Inflammation. Cam-
bridge: Harvard Univ. Press, 1945.
115. DuBois, A. B., and R. Marshall. Measurements of
pulmonary capillary blood flow and gas exchange
throughout the respiratory cycle in man. J. Clin. Invest.
36: 1566, 1957.
116. Duke, H. N. Observations on the effects of hypoxia on
the pulmonary vascular bed. J. Physiol., London 135:
45. '957-
117. Dunnill, M. S. An assessment of the anatomical factor
in cor pulmonale in emphysema. J. Clin. Pathol. 14: 246,
1961.
118. Edwards, J. E. Functional pathology of the pulmonary
vascular tree in congenital heart disease. Circulation 15:
164, 1957.
119. Edwards, W. S. The effects of lung inflation and epi-
nephrine on pulmonary vascular resistance. Am. J.
Physiol. 167:756, 1 95 1.
120. Eliakim, M., S. Stern, and H. Nathan. Site of action of
hypertonic saline in the pulmonary circulation. Circula-
tion Research g: 327, iq6i.
121. Eliakim, M., and D. M. Aviado. Effects of nerve stimu-
lation and drugs on the extrapulmonary portion of the
pulmonary vein. J. Pharmacol. Exptl. Therap. 133: 304,
1961.
122. Ellis, L. B., and S. Weiss. A study of the cardiovascular
responses in man to the intravenous and intraarterial
injection of acetylcholine. J. Pharmacol. Exptl. Therap. 44 :
235. "932-
Ellison, R. G, W. J. Brown, Jr., E. E. Hague, Jr.,
and W. F. Hamilton. Physiologic observations in experi-
mental pulmonary insufficiency. J. Thoracic Surg. 30:
633. '955-
Engleberg, J., and A B. DuBois. Mechanics of pul-
monary circulation in isolated rabbit lungs. Am. J. Physiol.
196:401, 1959.
Euler, U. S. v., and G. Liljestrand. Observations on
the pulmonary arterial blood pressure in the cat. Acta
Physiol. Scand. 12:301, 1946.
126. Euler, U. S. v., and F. Lishajko. Catechol amines in the
vascular wall. Acta Physiol. Scand. 42: 333, 1958.
127. Farhi, L. E., A. B. Otis, and D. F. Proctor. Measure-
ment of intrapleural pressure at different points in the
chest of the dog. J. Appl. Physiol. 10: 15, 1957.
128. Farhi, L. E., and H. Rahn. A theoretical analysis of the
alveolar-arterial O; difference with special reference to
the distribution effect. J. Appl. Physiol. 7: 699, 1955.
129. Ferrer, M. I., R. M. Harvey, L. Werko, D. T.
Dresdale, A. Cournand, and D. VV. Richards, Jr.
Some effects of quinidine sulfate on the heart and circula-
tion in man. Am. Heart J. 36: 816, 1948.
130. Fick, A. Ueber die Messung des Blutquantums in den
Herzventrikeln. Sitzung, July 1870. Verh. phys.-med. Ges.
Wiirzb. N.F. 2: XVI, 1872.
131. Finley, T. N. The determination of uneven pulmonary
blood flow from the arterial oxygen tension during nitro-
gen washout. J. Clm. Invest. 40: 1727, 1961.
132. Fishman, A. P. Respiratory gases in the regulation of
the pulmonary circulation. Physiol. Revs. 41: 214, 1961.
133. Fishman, A. P. The clinical significance of the pulmonary
collateral circulation. Circulation 24: 677, 1961.
123.
124.
!25-
DYNAMICS OF PULMONARY CIRCULATION
'735
134. Fishman, A. P., E. L. Becker, H. VV. Fritts, Jr., and
H. O. Heinemann. Apparent volumes of distribution of
water, electrolytes and hemoglobin within the lung.
Am. J. Physiol. 188:95, ]957-
135. Fishman, A. P., A. Himmelstein, H. W. Fritts, Jr., and
A. Cournand. Blood How through each lung in man dur-
ing unilateral hypoxia. J. Clin. Invest. 34: 637, 1955.
136. Fishman, A. P., J. McGlement, A. Himmelstein, and A.
Cournand. Effects of acute anoxia on the circulation
and respiration in patients with chronic pulmonary dis-
ease studied during the steady state. J. Clin. Invest. 31 :
77°, 1952.
137. Fishman, A. P., M. H. Maxwell, C. H. Crowder, and
P. Morales. Kidney function in cor pulmonale, with
particular reference to changes in renal hemodynamics
and sodium excretion during variation in level of oxy-
genation. Circulation 3: 703, 1951.
138. Fishman, A. P., G. M. Turino, and E. H. Bergofsky.
The syndrome of alveolar hypoventilation. Am. J. Med.
3: 333. '957-
139. Fishman, A. P., G. M. Turino, M. Brandfonbrenner,
and A. Himmelstein. The effective pulmonary collateral
blood flow in man. J. Clin. Invest. 37: 107 1, 1958.
140. Fleisch, A. Die Beziehung zwischen Stamm-und Astquer-
schnitt im Arteriensystem. Z. Anal. Entwicklungeschichte.
64:543. 1922.
140a.FLEisc.HNER, F. O., F. J. Romano, and A. A. Luisada.
Studies of fluorocardiography in normal subjects. Proc.
Soc. Exptl. Biol. Med. 67: 535, 1948.
141. Folkow, B. Nervous control of the blood vessels. Physiol.
Revs. 35:629, 1955.
142. Forssmann, W. Die Sondierung das rechten Herzens.
Klin. Wochschr. 8: 2085, 1929.
143. Forster, R. E. Exchange of gases between alveolar air
and pulmonary capillary blood : pulmonary diffusing
capacity. Physiol. Revs. 37: 391, 1957.
144. Foster, M. Lectures on the History 0/ Physiology. Cambridge:
Cambridge Univ. Press, 1 901.
145. Fowler, W. S. Intrapulmonary distribution of inspired
gas. Physiol. Revs. 32: 1, 1952.
146. Frank, N. R. Influence of acute pulmonary vascular
congestion on recoiling force of excised cats' lung. J.
Appl. Physiol. 14: 905, 1959.
147. Franklin, K. J. A Monograph on Veins. Springfield, 111.:
Thomas, 1937.
148. Frasher, W. G., and S. S. Sobin. Distensible behavior of
pulmonary artery. Am. J. Physiol. 199:472, 1960.
149. Freedman, M. E., G. L. Snider, P. Brostoff, S.
Kimelblot, and L. N. Katz. Effects of training on re-
sponse of cardiac output to muscular exercise in athletes.
J. Appl. Physiol. 8: 37, 1955.
150. Friedberg, L., L. N. Katz, and F. S. Steinitz. The
effect of drugs on the pulmonary and systemic arterial
pressures in the trained, unanesthetized dog. J. Pharmacol.
Exptl. Therap. 77:80, 1943.
151. Friedman, C. E. Heart volume, myocardial volume, and
total capacity of the heart cavities in certain chronic
heart diseases. Acta Med. Scand IOO; Suppl. 257, 1951.
152. Fritts, H. VV., Jr., P. Harris, C. A. Chidsev hi, R. H.
Clauss, and A. Cournand. Validation of a method for
measuring the output of the right ventricle in man by
inscription of dye-dilution curves from the pulmonary
artery. J. Appl. Physiol. 11:362, 1957.
153. Fritts, H. W., Jr., P. Harris, R. H. Clauss, J. E.
Odell, and A. Cournand. Effect of acetylcholine on the
human pulmonary circulation under normal and hypoxic
conditions. J. Clin. Invest. 37: 99, 1958.
154. Fritts, H. W., Jr., J. E. Odell, P. Harris, E. W. Braun-
wald, and A. P. Fishman. Effects of acute hypoxia on the
volume of blood in the thorax. Circulation 22: 216, 1960.
155. Fritts, H. VV., Jr., A. Hardewtg, D. F. Rochester, J.
Durand, and A. Courand. Estimation of pulmonary
arteriovenous shunt-flow using intravenous injection of
T-1824 dye and KR". ./. Clin. Invest. 39: 1841, i960.
156. Fry, D. L. Methods of flow estimation by pressure sensing
techniques. IRE Trans, on Med. Electronics. ME-6: 264,
■959-
157. Fuhner, H., and E. H. Starling. Experiments on the
pulmonary circulation. J. Physiol., London 47: 286, 1913.
158. Gaddum, J. H., C. O. Hebb, A. Silver, and A. A. Swan.
5-hydroxytryptamine pharmacological action and destruc-
tion in perfused lungs. Quart. J. Exptl. Physiol. 38- 255,
■953-
159. Galletti, P. M., P. F. Salisbury, and A. Rieben. In-
fluence of blood temperature on the pulmonary circula-
tion. Circulation Research 6: 275, 1 958.
160. Gerst, P. H., C. Ratteneorg, and D. A. Holaday.
Effects of hemorrhage on pulmonary circulation and
gas exchange. J. Clin. Invest. 38: 524, 1 959.
161. Gieson, J. G. 11, A. M. Seligman, VV. C. Peacock, J. C.
Aug, J. Fine, and R. D. Evans. The distribution of red
cells and plasma in large and minute vessels of the normal
dog, determined by radioactive isotopes of iron and
iodine. J. Clin. Invest. 25:848, 1946.
162. Giese, W. Uber die Endstrombahn der Lunge. In:
Lungen und Kleiner Kreislauf. Bad Olynhausner Gesprdc/ie I.,
S- 45-53- Berlin: Springer, 1957.
163. Goldrinc, R. M., G. M. Turino, G. Cohen, A. G.
Jameson, B. G. Bass, and A. P. Fishman. The catechola-
mines in the pulmonary arterial pressor response to acute
hypoxia. J. Clin. Invest. 41 : 121 1, 1962.
164. Goldring, R. M., G. M. Turino, D. H. Andersen, and
A. P. Fishman. Cor pulmonale in cystic fibrosis of the
pancreas. Circulation 24: 942, 1961.
165. Gomez, D. M. Hemodynamique et Angiocinitique. Paris:
Hermann, 1941.
166. Good field, G. J. The Growth oj Scientific Physiology. Lon-
don: Hutchinson, i960.
167. Gordon, D. B., J. Flasher, and D. R. Drury. Size of
the largest arteriovenous vessels in various organs. Am. J.
Physiol. 173:275, 1953.
168. GORTEN, R., J. C. GUNNELLS, A. M. WeISSLER, AND E. A.
Stead, Jr. Effects of atropine and isoproterenol on cardiac
output, central venous pressure, and mean transit time of
indicators placed at three different sites in the venous
system. Circulation Research g: 979, 1 96 1.
169. Green, H. D. Circulatory system: physical principles.
In: Medical Physics, edited by O. Glasser. Chicago: Yr.
Bk. Publ., ig55, vol. 1 and 2.
170. Gribbe, P., L. Hirvonen, J. Lind, and C. Wegelius.
Cineangiocardiographic recordings of the cyclic changes
in volume of the left ventricle. Cardiologia 34: 348, 1959-
171. Guteysse-Pellissier, M. A. Sur les vaisseaux pulmo-
naires a fibres striees des petits mammiferes. Compt. rend.
Acad. Sci. 205: 1176, 1937.
172. Gunther, R. T. Early Science in Oxford, De Corde by
736
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Richard Lower, London, 1669, with introduction and
translation by K J. Franklin. Oxford: Oxford Univ.
Press, 1932, vol. IX.
173. Gurtner, H. P., W. A. Briscoe, and A. Cournand.
Studies of the ventilation-perfusion relationships in the
lungs of subjects with chronic pulmonary emphysema,
following a single intravenous injection of radioactive
Krvpton (KRsi). I. Presentation and validation of a
theoretical model. J. Clin. Invest. 39: 1080, i960.
174. Guyton, A. G, and A. W. Lindsev. Effect of elevated
left atrial pressure and decreased plasma protein concen-
tration on the development of pulmonary edema. Ciuula-
tion Research 71 : 649, 1959.
175. Haddy, F. J., and G. S. Campbell. Pulmonary vascular
resistance in anesthetized dogs. Am. J. Physiol. 172: 747,
'953-
176. Haddy, F. J., A. L. Ferrin, D. W. Hannon, J. F. Alden,
W. L. Adams, and I. D. Baronoesky. Cardiac function
in experimental mitral stenosis. Circulation Research I : 219,
■953-
177. Haldane, J. S., and J. G. Priestley. Respiration. New
Haven: Yale Univ. Press, 1935.
178. Hales, S. Statical Essays: Haemastaticks, 1733 (3rd ed.).
London: Wilson and Nicol, 1769, vol. 2 pp. 66-67.
179. Hall, A. R. The Scientific Resolution, 1500-1800. Boston:
Beacon Press, 1954.
180. Haller, A. v. Elementa Physiologiae Corporis Humani,
Lausanne: Marci-Michael, Bousquet, 1757, vol. 2, book
6, section 4, S VIII, p. 330.
181. Halmagyi, D. F. J. Cardiorespiratory effects of experi-
mental lung embolism. J. Clin. Invest. 40: 1785, 1961.
182. Hamilton, W. F. Section on Circulatory System: Lungs.
In: Medical Physics, edited by O. Glasser. Chicago: Yr.
Bk. Pub., 1950, vol. 2, p. 207.
183. Hamilton, \V. F. Pressure relations in the pulmonary
circuit in blood, heart and circulation. Publ. Am. Assoc.
Advance. Sei. 13: 324, 1 940.
184. Hamilton, W. F. The physiology of the cardiac output.
Circulation 8: 527, 1953.
185. Hamilton, W. F., and E. A. Lombard. Intrathoracic
volume changes in relation to the cardiopneumogram.
Circulation Research I : 76, 1953.
186. Hamilton, W. F., J. W. Moore, J. M. Kinsman, and
R. G Spurling. Studies on the circulation. IV. Further
analysis of the injection method and of changes in hemo-
dynamics under physiological and pathological condi-
tions. ,4m. J. Physiol. 99: 534, 1932.
187. Hamilton, W. F., R. A. Woodbury, and E. Vogt. Dif-
ferential pressures in the lesser circulation of the unanes-
thetized dog. Am. I . Physiol. 125: 130, 1939.
188. Hamilton, W. F., and J. P. Mayo. Changes in the vital
capacity when the body is immersed in water. Am. J.
Physiol. 141 : 51, 1944.
189. Hamilton, W. F. , and D. W. Richards. The output of
the heart. In: The Circulation of the Blood, Men and Ideas,
edited by A. P. Fishman and D. W. Richards. New York :
Oxford Univ. Press. In press.
190. Hamilton, W. F., R. A. Woodbury, and H. T. Harper,
Jr. Arterial, cerebrospinal and venous pressures in man
during cough and strain. Am. J. Physiol. 141 : 42, 1944-
191. Harasawa, M., and S. Rodbard. Ventilatory air pres-
sure and pulmonary vascular resistance. Am. Heart J. 60 :
73, i960.
192. Harris, P. Influence of acetylcholine on the pulmonary
arterial pressure. Brit. Heart J. ig: 272, 1957.
193. Harris, P., H. W. Fritts, Jr., and A. Cournand. Some
circulatory effects of 5-hydroxytryptamine in man. Circula-
tion 21: 1 1 34, 1 960.
194. Harrison, R. W., W. E. Adams, W. Beuhler, and E. T.
Long. Effects of acute and chronic reduction of lung
volumes on cardiopulmonary reserve. Arch. Surg. 75: 546,
!958-
195. Harvey, W. Movement of the Heart and Blood in Animals.
Translated by K. J. Franklin. Springfield, 111.: Thomas,
'957-
196. Hayek, H. v. Die Menschliche Lunge. Berlin: Springer,
'953-
197- Haynes, R. H., and A. C. Burton. Role of the non-
Newtonian behavior of blood in hemodynamics. Am. J.
Physiol. 197: 943, 1959.
198. Heath, D., J. W. DuShane, E. H. Wood, and J. E.
Edwards. The structure of the pulmonary trunk at dif-
ferent ages and in cases of pulmonary hypertension and
pulmonary stenosis. J. Pathol. Bacterial. 77:443, 1959.
■ 99- Hecht, H. H., H. Kuida, R. L. Lange, J. L. Thorne,
and A. M. Brown. Brisket disease. II. Clinical features
and hemodynamic observations in altitude-dependent
right heart failure of cattle. Am. J. Med. 32: 171, 1962.
200. Hellems, H. K, F. W. Haynes, and L. Dexter. Pulmo-
nary "capillary" pressure in man. J. Appl. Physiol. 2 : 24,
•949-
201. Henderson, L. J. Blood: A Study in General Physiology.
New Haven: Yale Univ. Press, 1928.
202. Herrnheiser, G., and K. F. W. Hinson. An anatomical
explanation of the formation of butterfly shadows. Thorax
9: '98, 1954-
203. Hertz, C. W. Die Durchblutungsgrosse hypoventilierter
Lungenbezirke. Verh, dent. Ges. Kreislauforsch. 21. 447,
1955-
204. Hertz, C. W. Untersuchungen iiber den Einfluss der
alveolaren Gasdrucke auf die intrapulmonale Durch-
blutungsverteilung beim Menschen. Klin. W'ochschr. 34:
472. '956-
205. Hess, W. R. Das Prinzip des kleinsten Kraftverbrauches
in Dienste hemodynamischer Forschung. Arch. Anal.
Physiol. Physiol. Abt. 1914, p. I.
206. Heuvel-Heymans, G. M. v. d., and A. L. Rovati. Carotid
sinus baroreceptors and pulmonary hemodynamics. Arch,
intern. Pharmacodynamic 121 : 169, 1959.
207. Heymans, G, and E. Neil. Refiexogenic Areas of the Cardio-
vascular System. London: Churchill, 1958.
208. Hickam, J. B., and W. H. Cargill. Effect of exercise on
cardiac output and pulmonary arterial pressure with
cardiovascular disease and pulmonary emphysema. J.
Clin. Invest. 27: 10, 1948.
20g. HlMMELSTEIN, A., P. HARRIS, H. W. FrITTS, Jr., AND A.
Cournand. Effect of severe unilateral hypoxia on the
partition of pulmonary blood flow in man. J. Thoracic
Surg. 36: 369, 1958.
210. Himmelstein, A., A. G. Jameson, A. P. Fishman, and
G. H. Humphreys ii. Closed transventricular valvulo-
tomy for pulmonic stenosis. Surgery 42: 121, 1957.
211. Hochrein, M., and C. J. Keller. Beitrage zur Blutzirku-
lation in kleinen Kreislauf. Arch, exptl. Pathol. Pharmak.
■64: 529. 552, I932-
212. FIolmgren, A., and B. Pernow. The reproducibility cf
DYNAMICS OF PULMONARY CIRCULATION
1737
213.
Q14.
cardiac output determination by the direct Fick method
during muscular work. Scand. J. Clin. Lab. Invest. 12:
224, i960.
Hope, J. A Treatise on the Diseases of the Heart (3rd London
ed.). Philadelphia: Lea and Blanchard, 1846.
Horisberger, B., and S. Rodbard. Direct measurement
of bronchial arterial flow. Circulation Research 8: 1 149,
i960.
215. Howell, J. B. L., S. Permutt, D. F. Proctor, and R. L.
Riley. Effect of inflation of the lung on different parts of
pulmonary vascular bed. J. Appl. Physiol. 16: 71, 1961.
216. Huckabee, W. E. Relationships of pyruvate and lactate
during anaerobic metabolism. III. Effect of breathing
low-oxygen gases. J. Clin. Invest. 37: 264, 1958.
217. Hultgren, H., VV. Spickard, and C. Lopez. Further
studies of high altitude pulmonary oedema. Brit. Heart J.
24:95. '962-
218. Irving, L. Respiration in diving mammals. Physiol. Revs.
19: 112, 1939.
219. Irwin, J. W., W. S. Burrage, C. E. Aimar, and R. W.
Chestnut, Jr. Microscopical observations of the pulmo-
nary arterioles, capillaries, and venules of living guinea
pigs and rabbits. Anal. Record Iig: 391, 1954.
220. Ivanov, K. P. Central control of active pulmonary tonus.
Sechenov Physiol. J. U.S.S.R. 43: 790, 1957.
221. Jacobeus, H. C, and T. Brlce. A bronchospirometric
study on the ability of the human lungs to substitute for
one another. Acta Med. Scand. 105: 211, 1940.
222. Johansen, K. Circulation in the three-chambered snake
heart. Circulation Research 7: 828, 1959.
223. Johnson, R. L., Jr., W. S. Spicer, J. M. Bishop, and
R. E. Forster. Pulmonary capillary blood volume, flow
and diffusing capacity during exercise. J. Appl. Physiol.
>5:893> '960-
224. Johnson, S. R. The effect of some anesthetic agents on
the circulation in man : with special reference to the sig-
nificance of pulmonary blood volume for circulatory
regulation. Acta Chir. Scand. Suppl. 158, 1951.
225. Johnson, V., W. F. Hamilton, L. N. Katz, and W.
Weinstein. Studies on the dynamics of the pulmonary
circulation. Am. J. Physiol. 120:624, 1937.
226. Jose, A. D., and W. R. Milnor. The demonstration of
pulmonary arteriovenous shunts in normal human sub-
jects, and their increase in certain disease states. J. Clin.
Invest. 38: 1915, 1959.
227. Kabins, S. A., J. Fridman, J. Neustadt, G. Espinosa,
and L. N. Katz. Mechanisms leading to lung edema in
pulmonary embolization. Am. J. Physiol. 198: 543, 1960.
228. Keele, K. D. Three early masters of experimental medi-
cine— Erasistratus, Galen and Leonardo da Vinci. Proc.
Roy. Soc. Med. 54: 577, 1961.
229. Kelly, W. D., and M. B. Visscher. Observations on
blood flow during spontaneously occurring Traube-
Hering waves. Proc. Soc. Exptl. Biol. Med. 98: 597, 1958.
230. Kelsall, M. A., and E. D. Crabb. Lymphocytes and Mast
Cells. Baltimore: Williams & Wilkins, 1959.
231. Kleinermann, L., T. Ghitescu, I. Busu, N. Enescu, and
A. Lupu. Der Einfluss der Ausdehnung des linken Vor-
hofes auf den pulmonalen Arteriendruck. Cardiologia 31 .
475. '957-
232. Klocke, F. J., and H. Rahn. The arterial-alveolar inert
gas ("N2") difference in normal and emphysematous
subjects, as indicated by the analysis of urine. J. Clin.
Invest. 40: 286, 1961.
233. Knebel, R., and E. Wick. Uber die Bestimmung des
transmuralen Druckes des Herzens und der intrathoraka-
len Gefasse. Z. Rrtislaufforsch. 46: 271, 1957.
234. Knipping, H. W., W. Bolt, H. Venrath, H. Valentin,
H. Ludes, and P. Endler. Eine neue Methode zur
Pruflung der Herz- und Lungenfunktion. Die regionale
Funktionsanalyse in der Lungenund Herzklinik mit
Ililfe des radioaktiven Edelgases Xenon 133 (Isotopen-
Thorakographiej. Deut. Med. Xoc/ischr. 80: I 146, 1955.
235. Kniseley, W. H., J. M. Wallace, and W. A. Addison.
"Temporary" pulmonary embolization caused by in-
travenous injections of 5-hydroxytryptamine. Federation
Proc. 17: 88, 1958.
236. Kobayasi, S., and S. Furuya. Effects of histamine and
curare upon the pulmonary muscular tone in isolated
lungs of the Japanese toad. Acta Med. Biol., Niigata 8:
251, i960.
237. Kolin, A., and R. T. Kado. Miniaturization of the elec-
tromagnetic blood flow meter and its use for the recording
of circulatory responses of conscious animals to sensory
stimuli. Proc. Natl. Acad. Sci. 45: 1 312, 1959.
238. Kopelman, H., and G. de J. Lee. The intrathoracic
blood volume in mitral stenosis and left ventricular failure.
Clin. Sci. 10: 383, 1951.
239. Korner, P. I. Circulatory adaptations in hypoxia.
Physiol. Revs. 39: 687, 1959.
240. Krogh, A., and J. Lindhard. Measurements of the blood
flow through the lungs of man. Skand. Arch. Physiol. 27 :
100, 1912.
241. Krogh, A. Anatomy and Physiology of Capillaries. New
Haven: Yale Univ. Press, 1929.
242. Krogh, A. The Comparative Physiology of Respiratory Mech-
anism. Philadelphia: Univ. Pennsylvania Press, 1 941 .
243. Krogh, M. Diffusion of gases through the lungs of man.
J. Physiol., London 49: 271, 1914.
244. Kruh0ffer, P. Lung diffusion coefficient for CO in
normal human subjects by means of C140. Acta Physiol.
Scand. 32: 1 06, 1 954.
245. Kuida, H., L. B. Hinshaw, R. P. Gilbert, and M. B.
Visscher. Effect of gram-negative endotoxin on pulmo-
nary circulation. Am. J. Physiol. 192: 335, 1958.
246. Kunieda, T. Determination of pulmonary blood volume
in patients with mitral valve disease by T-1824 dye
method. Kokyd to Junkan 3: 510, 1955.
247. Lagerlof, H., H. Eliasch, L. Werko, and E. Berglund.
Orthostatic changes of the pulmonary and peripheral
circulation in man. Scand. J. Clin. Lab. Invest. 3: 85, 1951.
248. Lagerlof, H., and L. Werko. Studies on the circulation
of blood in man. VI. The pulmonary capillary venous
pressure pulse in man. Scand. J. Clin. Lab. Lnvest. I : 147,
'949-
24g. Lagerlof, H., L. Werko, H. Bucht, and A. Holmgren.
Separate determination of the blood volume of the right
and left heart and the lungs in man with the aid of the
dye injection method. Scand. J. Clin. Lab. Invest. I: 114,
'949-
250. Lammerant, J. Le Volume Sanguin des Poumons. Brussels:
Arscia, 1957.
251. Lanari, A., and A. Angrest. Pressure-volume relation-
ship in the pulmonary vascular bed. Acta Physiol. Latinam.
4: 116, 1954.
1738
II WDI'.i ii >K ' '1 l'in Ml H i ii. -i
CIRCULATION II
252. Lategola, M. T. Pressure-flow relationships in the dog 271.
lung during acute, subtotal pulmonary vascular occlusion.
Am. J. Physiol. 192:613, 1958.
253. Lauson, H. D., R. A. Bloomfield, and A. Cournand. 272.
The influence of the respiration on the circulation in man.
Am. J. Med. 1 1615, 1946.
254. Lee, G. de J., and A. B. DuBois. Pulmonary capillary-
blood flow in man. J. Clin. Invest. 34: 1380, 1955. 273.
255. Lee, G. de J., M. B. Matthews, and E. P. Sharpey-
Schafer. The effect of the Valsalva manoeuvre on the
systemic and pulmonary arterial pressure in man. Bril. 274.
Iharl J. 16: 31 I, 1954.
256. Lenfant, G, and B. Howell. Cardiovascular adjust-
ments in dogs during continuous pressure breathing. J.
Appl. Physiol. 15:425, i960. 275.
257. Leusen, I., and G. Demeester. Variations de la resistance
vasculaire pulmonaire au cours d'une anesthesie pro-
longee. Arch, intern, physiol. 61 : 553, 1953.
258. Leusen, I., G. Demeester, and J. J. Bouckaert. In- 276.
fluence du travail musculaire sur la circulation et la res-
piration chez le chien. Acta Cardiol. 13: 153, 1958. 277
259. Leusen, I., G. Demeester, and K. Vuvlsteek. Effets de
l'occlusion d'une branche de l'artere pulmonaire chez 278.
le chien. Acta Cardiol. 12: I, 1957.
260. Levy, M. N., S. H. Brind, F. R. Brandlin, and F. A.
Phillips, Jr. The relationship between pressure and flow 279.
in the systemic circulation of the dog. Circulation Research
2:372, 1954-
261. Lewin, R. J., C. E. Cross, P. Rieben, and P. F. Salis- 280.
bury. Stretch reflexes from the main pulmonary artery
to the systemic circulation. Circulation Research g: 585,
1961. 281.
262. Lewis, B. M., W. T. McElroy, E. J. Hayford-Welsing,
and L. C Samberg. The effects of body position, gan-
glionic blockade and norepinephrine on the pulmonary
capillary bed. J. Clin. Invest. 39: 1345, i960.
263. Liebow, A. A., M. R. Hales, and W. E. Bloomer. Rela- 282.
tion of bronchial to pulmonary vascular tree. In: Pulmo-
nary Circulation, edited by W. Adams and I. Veith. New
York: Grune & Stratton 1959, p. 79.
264. Liebow, A. A., M. R. Hales, W. Harrison, W. Bloomer, 283.
and G. E. Lindskog. The genesis and functional implica-
tions of collateral circulation of the lungs. Yale J. Biol.
and Med. 22: 637, 1950. 284.
265. Liebow, A. A., W. E. Loring, and W. E. Felton. The
musculature of the lungs in chronic pulmonary disease. 285.
Am. J. Pathol. 29: 885, 1953.
266. Lilienthal, J. L., Jr., R. L. Riley, D. D. Proemmel,
and R. E. Franke. An experimental analysis in man of 286.
the oxygen pressure gradient from alveolar air to arterial
blood during rest and exercise at sea level and at altitude. 287,
Am. J. Physiol. 147: 199, 1946.
267. Lilienthal, J. L., Jr., and R. L. Riley. Diseases of the 288.
respiratory system. Circulation through the lung and
diffusion of gases. Ann. Rev. Med. 5: 237, 1954. 289.
268. Liljestrand, G. Regulation of pulmonary arterial blood
pressure. Arch. Intern. Med. 81 : 162, 1948.
269. Liljestrand, G. Chemical control of the distribution of 290.
the pulmonary blood flow. Acta Physiol. Scand. 44: 216,
>958-
270. Linderholm, H., P. Kimbel, D. H. Lewis, and A. B.
DuBois. Pulmonary capillary blood flow during cardiac 291.
catheterization. J. Appl. Physiol. 17: 135, 1962.
Lindsey, A. W., and A. C. Guyton. Continuous recording
of pulmonary blood volume: pulmonary pressure and
volume changes. Am. J. Physiol. 197: 959, 1959.
Little, R. C. Volume pressure relationships of the pulmo-
nary -left heart vascular segment. Evidence for a "valve-
like" closure of the pulmonary veins. Circulation Research
8:594. i960.
Lloyd, T. C, Jr., and G. W. Wright. Pulmonary vas-
cular resistance and vascular transmural gradient. J.
Appl. Physiol. 15: 241, i960.
Lochner, W. Weitere Lntcrsuchungen iiber den Eigen-
stoffwechsel der Lunge, insbesondere eine Freisetzung
veresterter Fettsauren. Pjiugers Arch. ges. Physiol. 272 : 180,
i960.
Lochner, \V\, H. Bartels, R. Beer, M. Mochizuki, and
G. Rodewald. Untersuchung des Gasaustausches am
isolierten durchbluteten Lungenlappen des Hundes. Pflii-
gers Arch. ges. Physiol. 264: 294, 1957.
Low, I. N. Electron microscopy of the rat lung. Anal.
Record 113: 437, 1952.
McDonald, D. A. Blood Floiv in Arteries. London: Arnold,
i960.
McGaff, C. J., A. D. Jose, and W. R. Milnor. Pulmo-
nary, left heart and arterial volume in valvular heart
disease. Clin. Research 7: 230, 1959.
McIlroy, M. B., R. Marshall, and R. V. Christie.
The work of breathing in normal subjects. Clin. Sci. 13:
127. "954-
Malpighi, M. De pulmonibus observationes anatomiae.
Bologna, 166 1. Translated by J. Young. Proc. Roy. Soc.
Me,/. (Part I) 23: 7, 1929-30.
Marshall, R. J., Y. Wang, H. J. Semler, and J. T.
Shepherd. Flow, pressure and volume relationships in
the pulmonary circulation during exercise in normal dogs
and dogs with divided left pulmonary artery. Circulation
Research 9: 53, 1 961.
Marshall, R. J., H. F. Helmholz, Jr., and J. T. Shep-
herd. Effect of acetylcholine on pulmonary vascular
resistance in a patient with idiopathic pulmonary hyper-
tension. Circulation 20: 391, 1959.
Marshall, R. J., and J. T. Shepherd. Interpretation
of changes in ''central" blood volume and slope volume
during exercise in man. J. Clin. Invest. 40: 375, 1961.
Martin, C. J., and A. C. Young. Ventilation-perfusion
variations within the lung. J. Appl. Physiol. 1 1 : 371, 1957.
Martin, C. J., F. Cline, Jr., and H. Marshall. Lobar
alveolar gas concentrations: effect of body position. J.
Clin. Invest. 32: 617, 1953.
Mattson, S. B., and E. Carlens. Lobar ventilation and
oxygen uptake in man. J. Thoracic Surg. 30: 676, 1955-
Mead, J. Mechanical properties of lungs. Physiol. Revs.
41 : 281, 1961.
Merkel, H. Structure and function of the pulmonary
circulation. Z. Kreislaufforsch. 38: 705, 1949.
Merriman, J. E. The pulmonary circulation in hemor-
rhagic shock. In: Shock and Circulatory Homeostasis, edited
by H. D. Green. New York: Macy, 1954.
Meyer, W. W., and P. Schollmeyer. Die Volumendehn-
barkeit und die Druck-Umfang-Beziehungen des Lungen-
schlagader-Windkessels in Abhangigkeit vom alter und
pulmonalen Hochdruck. Klin. Wochschr. 35: 1070, 1957.
Meyerhof, M. Ibn an-Nafis and his theory of the lesser
circulation. Isis 23: 100, 1935.
DYNAMICS OF PULMONARY CIRCULATION
'739
292. Miller, W. S. The Lung. Springfield, 111. : Thomas, 1947. 312.
293. Milnor, W. R., A. D. Jose, and C. J. McGaff. Pulmo-
nary vascular volume, resistance and compliance in man. 313.
Circulation 22: 1 30, i960.
294. Mitchell, A. M., and A. Cournand. The fate of circu-
lating lactic acid in the human lung. J. Clin. Invest. 34: 314.
47'. '955-
295. Mochizuki, M., and J. Fukuoka. The diffusion of oxygen
inside the red cell. Japan. J. Physiol. 8: 206, 1958.
296. Moreno, F., and H. A. Lyons. Effect of body posture on 315.
lung volumes. J. Appl. Physiol. 16: 27, 1961.
297. Morgan, W. O. P., and C. D. Murray. Oxygen ex-
change, blood and circulation; coordinated treatment of
factors involved in oxygen supply on basis of diffusion 316.
theory. J. Biol. Chem. 65: 419, 1925.
298. Morkin, E., O. R. Levine, F. O. Bowman, and A. P.
Fishman. The nature of pulmonary capillary blood flow
and gas exchange. J. Clin. Invest. 41 : 1386, 1962. 317.
299. Morrow, A. G., E. Braunwald, and J. Ross, Jr. Left
heart catheterization : an appraisal of techniques and
their applications in cardiovascular diagnosis. Arch. In- 318.
ternal Med. 105: 645, i960.
300. Motley, H. L., A. Cournand, L. Werko, A. Himmel- 319.
stein, and D. Dresdale. The influence of short periods
of induced acute anoxia upon pulmonary artery pressures
in man. Am. J. Physiol. 150: 315, 1947. 320.
301. Muller, A. Bemerkungen zum Gasaustausch in den
Lungen. Helv. Physiol, el Pharmacol. Acta 3: 203, 1945.
302. Nahas, G. G., and I. MacDonald. Effects of norepi-
nephrine and 5-hydroxytryptamine on the pulmonary 321.
circulation of the spinal dog. Am. J. Physiol. 196: 1045,
■959-
303. Nahas, G. G, M.. B. Visscher, G. VV. Mather, F. J. 322.
Haddy, and H. R. Warner. Influence of hypoxia on the
pulmonary circulation of nonnarcotized dogs. J. Appl.
Physiol. 6: 467, 1954. 323.
304. Newman, E. V., M. Merrill, A. Genecin, C. Monge,
W. R. Milnor, and W. P. McKeever. The dye dilution
method for describing the central circulation. Circulation 324.
4:735. I951-
305. Nisell, O. I. Some aspects of the pulmonary circulation 325.
and ventilation. Intern. Arch. Allergy 3: 142, 1952.
306. Nordenstrom, B. Contrast examination of the cardio-
vascular system during increased intrabronchial pressure.
Acta Radiol. Suppl. 200, i960. 326.
307. Nylin, G., and S. Hedlund. Blood flow and pool in
heart, lungs and brain. In: Circulation. Proceedings of The
Harvey Tercentenary Congress, edited by J. McMichael.
Springfield, 111.: Thomas, 1958. 327.
308. Opdyke, D. F., and G. A. Brecher. Effect of normal and
abnormal changes of intrathoracic pressure on effective
right and left atrial pressures. Am. J. Physiol. 160: 556, 328.
'95°-
309. Opdyke, D. F., H. F. Van Noate, and G. A. Brecher.
Further evidence that inspiration increases right atrial 329.
flow. Am. J. Physiol. 162: 259, 1950.
310. Parrish, D., D. E. Strandess, Jr., and J. W. Bell. Dif-
ferences between plasma and red cell flow characteristics
of pulmonary vascular bed. Am. J. Physiol. 200: 619, 330.
1 961.
31 1. Patel, D. J., D. P. Schilder, and A.J. Mallos. Mechan-
ical properties and dimensions of the major pulmonary
arteries. J. Appl. Physiol. 15: 92, i960. 331.
Pattle, R. E. The formation of a lining him by foetal
lungs. J. Pathol. Bacterial. 82: 333, 1961.
Pearce, M. L., A. E. Lewis, and M. R. Kaplan. The
factors influencing the circulation time. Circulation 5 :
583. '952-
Perkins, J. F., Jr., W. E. Adams, and A. Flores. Arterial
oxygen saturation vs. alveolar oxygen tension as a measure
of venous admixture and diffusion difficulty in the lung.
J. Appl. Physiol. 8: 455, 1 956.
Permutt, S., J. B. L. Howell, D. F. Proctor, and R. L.
Riley. Effect of lung inflation on static pressure-volume
characteristics of pulmonary vessels. J. Appl. Physiol. 16:
64, 1 96 1.
Peters, R. M., W. E. Loring, and W. H. Sprunt. An
experimental study of the effect of chronic atelectasis on
pulmonary and bronchial blood flow. Circulation Research
7: 3'. '959-
Piiper, J. Grosse des Arterien-, des Capillar- und des
Venenvolumens in der isolierten Hundelunge. Pjlugers
Arch. ges. Physiol. 269: 182, 1959.
Piiper, J. Die funktionellen Abschnitte des Lungenge-
faszsystems. Beitr. Silikose-Forsch. 1960.
Piiper, J. Variations of ventilation and diffusing capacity
to perfusion determining the alveolar-arterial O* differ-
ence: theory. J. Appl. Physiol. 16: 507, 1961.
Prec, O., L. N. Katz, L. W. Sennett, R. Rosenman,
A. P. Fishman, and W. Hwang. Determination of the
kinetic energy of the heart in man. Am. J. Physiol. 159:
483, '949-
Price, K. C, D. Hata, and J. R. Smith. Pulmonary
vasomotion resulting from miliary embolism of the lungs.
Am. J. Physiol. 182: 183, 1955.
Prinzmetal, M., E. M. Ornitz, Jr., B. Simkin, and
H. C. Bergman. Arteriovenous anastomoses in liver,
spleen and lungs. Am. J. Physiol. 152: 48, 1948.
Pritchard, M. M. L., P. M. Daniel, and G. M. Ardran.
Peripheral ischemia of the lung. Brit. J. Radiol. 27: 93,
!954-
Quincke, H., and E. Pfeiffer. Ueber den Blutstrom in
den Lungen. Arch. Anal. Physiol. S, 90, 187 1.
Rabin, E. R., and E. C. Meyer. Cardiopulmonary effects
of pulmonary venous hypertension with special reference
to pulmonary lymphatic flow. Circulation Research 8: 324>
i960
Rabinowitz, M., and E. Rapaport. Determination of
circulating pulmonary blood volume in dogs by an ar-
teriovenous dye equilibration method. Circulation Research
2:525. >954-
Rahn, H. A concept of mean alveolar air and the ventila-
tion-blood flow relationships during pulmonary gas ex-
change. Am. J. Physiol. 158:21, 1949.
Rahn, H., and H. T. Bahnson. Effect of unilateral hy-
poxia on gas exchange and calculated pulmonary blood
flow in each lung. J. Appl. Physiol. I : 105, 1953.
Rahn, H., R. C. Stroud, and H. Meier. Radiographic
anatomy of heart and pulmonary vessels of the dog with
observations of the pulmonary circulation time. J. Appl.
Physiol. 5: 308, 1952.
Rahn, H., P. Sadoul, L. E. Farhi, and J. Shapiro.
Distribution of ventilation and perfusion in the lobes of
the dog's lung in the supine and erect position. J. Appl.
Physiol. 8: 417, 1956.
Rahn, H., R. C. Stroud, and C. E. Tobin. Visualization
1740
HANDBOOK OF PHYSIOLOGY ^- CIRCULATION II
of arteriovenous shunts by cinefluorography in the lungs 350.
of normal dogs. Proc. Soc. Exptl. Biol. Med. 80: 239, 1952.
332. Rapaport, E., H. Kuida, F. W. Haynes, and L. Dexter.
PuUnonary red cell and plasma volumes and pulmonary 351.
hematocrit in the normal dog. Am. J. Physiol. 185: 127,
1956.
333. Read, R. C, J. A. Johnson, J. A. Vick, and M. W. 352.
Meyer. Vascular effects of hypertonic solutions. Circula-
tion Research 8: 538, i960. 353.
334. Reeves, J. T., R. F. Grover, G. F. Filley, and S. G.
Blount, Jr. Cardiac output in normal resting man. J.
Appl. Physiol. 16: 276, 1 961. 354.
335. Reeves, J. T., R. F. Grover, G. F. Filley, and S. G.
Blount, Jr. Circulatory changes in man during mild
supine exercise. J. Appl. Physiol. 16: 279, 1 96 1.
336. Reeves, J. T., R. F. Grover, S. G. Blount, Jr., and 355.
G. F. Filley. Cardiac output response to standing and
treadmill walking. J. Appl. Physiol. 16: 283, 1961.
337. Remington, J. W. Extensibility behavior and hysteresis
phenomena in smooth muscle tissue. In: Tissue Elasticity,
edited by J. W. Remington. Washington D.C.: Am. 356.
Physiol. Soc, 1 957.
338. Remington, J. VV., and W. F. Hamilton. The evaluation
of the work of the heart. Am. J. Physiol. 1 50: 292, 1947. 357.
339. Richards, D. W. The contributions of right heart cathe-
terization to physiology and medicine, with some ob-
servations on the physiopathology of pulmonary heart 358.
disease. .4m. Heart J. 54: 161, 1957.
340. Richards, D. W., A. Cournand, and H. L. Motley.
Effects on circulatory and respiratory functions of various
forms of respirator. Trans. Assoc. Am. Physicians 59: 102, 359.
1946.
341. Richards, D. W., and A. P. Fishman. Cor pulmonale in
chronic pulmonary emphysema. In: Pulmonary Emphysema, 360.
edited by A. L. Barach and H. A. Bickerman. Baltimore.
Williams & Wilkins, 1956, chapt. 15.
342. Rigatto, M., and A. P. Fishman. The pulsatile nature of
the pulmonary capillary blood flow. J. Clin. Incest. 39: 361.
1626, i960.
343. Rigatto, M., G. M. Turino, and A. P. Fishman. Deter- 362.
mination of the pulmonary capillary blood flow in man.
Circulation Research 9: 945, 1 96 1.
344. Riley, R. L. Apical localization of pulmonary tuberculo- 363.
sis. Bull. Johns Hopkins Hasp. 106 232, i960.
345. Riley, R. L., and A. Cournand. "Ideal" alveolar air 364.
and the analysis of ventilation-perfusion relationships in
the lung. J. Appl. Physiol. I: 199, 1949.
346. Riley, R. L., A. Himmelstein, H. L. Motley, H. M.
Weiner, and A. Cournand. Studies of the pulmonary 365.
circulation at rest and during exercise in normal indivi-
duals and in patients with chronic pulmonary disease.
Am. J. Physiol. 152: 372, 1948.
347. Riley, R. L., S. Permutt, S. Said, M. Godfrey, T. O.
Cheng, J. B. L. Howell, and R. H. Shepard. Effect of
posture on pulmonary dead space in man. J. Appl. Physiol. ;U>
'4:339. '959-
348. Riley, R. L., R. H. Shepard, J. E. Cohn, D. G. Carroll,
and B. W. Armstrong. Maximal diffusing capacity of 367.
the lungs. J. Appl. Physiol. 6: 573, 1954.
349. Ring, G. C, A. S. Blum, T. Kurbatov, W. G Moss,
and W. Smith. Size of microspheres passing through 368.
pulmonary circuit in the dog. Am. J. Physiol. 200: 1191,
1961.
Roach, M. R., and A. C. Burton. The reason for the
shape of the distensibility curves of arteries. Can. J.
Biochem. and Physiol. 35: 681, 1957.
Rodbard, S. Bronchomotor tone. A neglected factor in
the regulation of the pulmonary circulation. Am. J. Med.
>5: 356. '953-
Rodbard, S., F. Brown, and L. N. Katz. The pulmonary
arterial pressure. Am. Heart J. 38: 863, 1949.
Rodbard, S., and M. Harasavva. The passivity of the
pulmonary vasculature in hypoxia. Am. Heart J. 57:
232, 1959-
Roos, A., L. J. Thomas, Jr., E. L. Nagel, and D. C.
Prommas. Pulmonary vascular resistance as determined
by lung inflation and vascular pressures. J. Appl. Physiol.
16: 77, 1961.
Rose, J. C, E. D. Freis, C. A. Hufnagel, and E. A.
Massulo. Effects of epinephrine and norepinephrine in
dogs studied with a mechanical left ventricle. Demon-
stration of active vasoconstriction in the lesser circulation.
Am. J. Physiol. 182: 197, 1955.
Rose, J. G, and E. J. Lazaro. Pulmonary vascular
responses to serotonin and effects of certain serotonin
antagonists. Circulation Research 6: 283, 1958.
Rosenberg, E., and R. E. Forster. Changes in diffusing
capacity of isolated cat lungs with blood pressure and flow.
J. Appl. Physiol. 15: 883, i960.
Ross, J. C, R. Frayser, and J. B. Hickam. A study of
the means by which exercise increases the pulmonary
diffusing capacity for carbon monoxide. J. Clin. Invest.
3°: 9l6. '959-
Rossier, P. H., A. A. Buhlmann, and K. Wiesinger.
Respiration, edited and translated by P. C. Luchsinger
and K. M. Moser. St. Louis: Mosby, i960.
Rothschild, M. A., A. L. Davis, M. Oratz, and S. S.
Schreiber. Pulmonary transcapillary exchange of NaM
and P32-labelled phosphate in pulmonary emphysema.
J. Clin. Invest. 38: 2224, !959-
Rothschuh, K. E. Geschichtr der Physiologic. Berlin:
Springer, 1953.
Rotta, A., A. Canepa, A. Hurtado, T. Velasquez,
and R. Chavez. Pulmonary circulation at sea level and
at high altitudes. J. Appl. Physiol. 9: 328, 1 956.
Rotta, A., and A. Lopez. Electrocardiographic patterns
in man at high altitudes. Circulation 19: 719, 1959.
Roughton, F. J. W. The average time spent by the blood
in the human lung capillary and its relation to the rates
of CO uptake and elimination in man. Am. J. Physiol.
143:621, 1945.
Roughton, F. J. W., and R. E. Forster. Relative im-
portance of diffusion and chemical reaction rates in
determinining rate of exchange of gases in the human
lung, with special reference to true diffusing capacity of
pulmonary membrane and volume of blood in the lung
capillaries. J. Appl. Physiol. II : 290, 1957.
Rudolph, A. M., and P. A. M. Auld. Physical factors
affecting normal and serotonin-constricted pulmonary
vessels. Am. J. Physiol. 198: 864, i960.
Rushmer, R. F., and N. Thal. Factors influencing stroke
volume : a cinefluorographic study of angiocardiography.
Am. J. Physiol. 168: 509, 1952.
Salisbury, P. F., P. Weil, and D. State. Factors in-
fluencing collateral blood flow to the dog's lung. Cir-
culation Research 5: 303, 1957.
DYNAMICS OF PULMONARY CIRCULATION
I74I
369. Sancetta, S. M., R. B. Lynn, F. A. Simeone, and R. \Y.
Scott. Hemodynamic changes in humans following in-
duction of low and high spinal anesthesia. Circulation 6:
559. '952-
370. Sancetta, S. M., and L. Rakita. Response of pulmonary-
artery- pressure and total pulmonary resistance of un-
trained convalescent man to prolonged mild steady-
state exercise. J. Clin. Invest. 36: 1 1 38, 1957.
371. Sarnoff, S. J., and E. Berglund. Pressure volume char-
acteristics and stress relaxation in the pulmonary vas-
cular bed of the dog. Am. J. Physiol. 171 : 238, 1952.
372. Sarnoff, S. J., E. Berglund, and L. C. Sarnoff. Neuro-
hemodynamics of pulmonary edema. III. Estimated
changes in pulmonary blood volume accompanying
systemic vasoconstriction and vasodilatation. J. Appl.
Physiol. 5:367, 1953.
373. Schleier, J. Der Energievebrauch in der Blutbahn.
Pfliigers Arch. ges. Physiol. 173: 172, 1919.
374. Schlicher, L., V. Peiper, H. Krug, and H. Bohme.
Die VVirkung des transmuralen Drukes auf den arteriellen
und venosen Raum der Strombahn der isolierten Kanin-
chenlunge. Z. ges. exptl. Med. 131 : 443, 1959.
375. Schulz, H. Die Submikroskopische Anatomic und Pathologic
der Lunge. Berlin: Springer-Verlag, 1959.
376. Semler, H. J., J. T. Shepherd, and H. J. C. Swan.
Pressor effect of hypertonic saline on pulmonary circula-
tion. Circulation Research 7: 1011, 1959.
377. Sevxrinchaus, J. W., and M. Stupfel. Alveolar dead
space as an index of distribution of blood flow in pul-
monary capillaries. J. Appl. Physiol. 10: 335, 1957.
378. Sharp, J. T. The effect of body position change on lung
compliance in normal subjects and in patients with con-
gestive heart failure. J. Clin. Invest. 38: 659, 1959.
379. Sharpey-Schafer, E. The influence of the depressor
nerve on the pulmonary circulation. Qiiart. J. Exptl.
Physiol. 12: 372, 1 920.
380. Simmons, D. H., L. M. Linde, J. H. Miller, and R. J.
O'Reilly. Relation between lung volume and pulmonary
vascular resistance. Circulation Research g: 465, 1961 .
381. Sjostrand, T. Volume and distribution of blood and then-
significance in regulating the circulation. Physiol. Revs.
33 :202> '953-
382. Slonim, N. B., A. Ravin, O. J. Balchum, and S. H.
Dressler. The effect of mild exercise in the supine posi-
tion on the pulmonary arterial pressure of five normal
human subjects. J. Clin. Invest. 33: 1022, 1954.
383. Soderholm, B., and L. Werko. Acetylcholine and the
pulmonary circulation in mitral valvular disease. Brit.
Heart J. 21: I, 1959.
384. Spehl, E. De la Repartition du Sang Circulant dans UEcono-
mie. (These d'Agregation.) Brussels: Lebeque, 1883.
385. Starling, E. H. Physiological factors involved in the
causation of dropsy. Lancet I : 1267, 1896.
386. Stern, S., and K. Braun. Blood gas changes following
priscoline administration in mitral stenosis and chronic
lung diseases. Am. J. Cardiol. 7: 188, 1961.
387. Strubell-Harkort, A. Vasomotorische Einflusse und
Druckverhaltnisse im kleinen Kreislauf. Verhandl. deut.
Ges. Kreislaufforsch. 8: 123, 1935.
388. Svanberg, L. Influence of posture on the lung volumes,
ventilation and circulation in normals. Scand. J. Clin.
Lab. Invest, g, suppl. 25, 1 957.
389. Swan, H. J. C, H. B. Burchell, and E. H. Wood. Effect
of oxygen on pulmonary vascular resistance in patients
with pulmonary hypertension associated with atrial
septal defect. Circulation 20 : 66, 1959.
3go. Swenson, E. W., T. N. Finley, and S. V. Guzman.
Unilateral hypoventilation in man during temporary
occlusion of one pulmonary artery. J. Clin. Invest. 40 :
828, 1 96 1.
391. Takacs, L., Z. Nagy, and K. Kali ay. Pulmonary cir-
culation in shock. Acta Physiol. Acad. Sci. Hung. 1 1 : 233,
■957-
392. Takino, M. Vergleichende Studien uber die histologische
Strucktur der Arteriae und Venae pulmonales, die
Blutgefassnerven der Lunge und die Nerven der Bronchien
bei verschiedenen Tierarten, besonders iiber die Be-
ziehung der Blugefassnerven zu den glatten Muskeln der
Blutgefasse. III. Mitteilung. Acta Schol. Med. Univ. Kioto
1 5 : 32 1 , 1932-1933-
393. Takino, M. Der Lungenarterienstammrefiex. Japan. J.
Med. Sci. HI Biophysics. 9:67, 1 943.
394. Takino, M., and Y. Ezaki. Uber die Besonderheiten der
Arteriae und Venae pulmonales bei verschiedenen Tieren,
besonders beim Menschen. V. Mitteilung. Acta Schol.
Med. Univ. Kioto 17: 1, ig35-
395. Tenney, S. M. Fluid volume redistribution and thoracic
volume changes during recumbency. J. Appl. Physiol.
■4: 129. '959-
396. Thews, G. Die Sauerstoffdiffusion in der Lunge. Ein
Verfahren zur Berechnung der GvDiffusionzeiten der
Kontaktzeit und des GvDiffusion-Faktors. Pfliigers Arch,
ges. Physiol. 265: 154, 1957.
397. Thomas, L. J., Jr., Z. J. Griffo, and A. Roos. Effect of
negative-pressure inflation of the lung on pulmonary
vascular resistance. J. Appl. Physiol. 16-451, 1961 .
3g8. Thomas, L. J., Jr., A. Roos, and Z. J. Griffo. Relation
between alveolar surface tension and pulmonary vas-
cular resistance. J. Appl. Physiol. 16: 457, 196 1.
399. Thorson, A., and O. Nordenfelt. Development of
valvular lesions in metastatic carcinoid disease. Brit.
Heart J. 21:243, 1959.
400. Tiemann, W., and A. Daiber. Beobachtungen an den Lun-
gencapillaren. II. Teil. Z. ges. exptl. Med. 86: 464, 1933.
401. Tiffeneau, R., and M. Beauvallet. Role de la destruc-
tion intrapulmonaire de ['acetylcholine. Effects locaux
et generaux des aerosols acetylcholiniques. Compt. rend.
Soc. Biol. 138:747, 1944.
402. Tobian, L., S. Martin, and W. Eilers. Effect of pH on
norepinephrine-induced contraction of isolated arterial
smooth muscle. Am. J. Physiol. 196: 998, 1959.
403. Tondury, G., and E. Weible. Anatomie der Lungen-
gefasse. In: Erg. d. Ges. Tuberkulose — und Lungenforschung,
edited by S. Engel, L. Heilmeyer, J. Heim, and E. Ueh-
linger. Stuttgart: Thieme, 1958, vol. 14, p. 59.
404. Traube, L. Ueber periodische Thatigkeits-Aeusserungen
des vasomotorischen und Hemmungs-Nervencentrums.
In : Gesammelte Beitrage zur Pathologic und Physiologic. Berlin :
August Hirschwald, 187 1, vol. 1, chapt. 21, p. 387.
405. Tuller, M. A., D. E. Lehr, and A. P. Fishman. Induced
alterations in the distribution of pulmonary blood flow.
Federation Proc. 20: 106, 1961.
406. Turino, G. M., and A. P. Fishman. Congested lung. J.
Chronic Diseases 9: 510, 1959.
407. Turino, G. M., M. Brandfonbrenner, and A. P. Fish-
man. The effect of changes in ventilation and pulmonary
1742
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
blood flow on the diffusing capacity of the lung. J. Clin,
invest. 38: 1 186, 1959.
408. Ulmer, W. von, and A. Wenke. Bronchospirometrische
Untersuchungen zur Frage der gasspannungsabhangigen
Durchblutungsregulation der Alveolarkapillaren. Arch.
Kreislaujforsch. 26:256, 1957.
409. Van Bogaert, A., L. Fannes, J. Buytaert, H. de
Munck, H. V. Genabeck, F. van der Henst, and J.
Vandael. Hypertension aiterielle pulmonaire aprcs
ligature d'une ou de plusieurs veines pulmonaires. Arch.
Mai. coeur et vaisseaux 46: 289, 1953.
410. Venrath, H., R. Rotthof, H. Valentin, and W. Bolt.
Bronchospirographische Untersuchungen bei Durch-
blutungestorungen im kleinen Kreislauf. Beilr. Klin.
Tuberk. 107:291, 1952.
411. Visscher, M. B., F. J. Haddy, and G. Stephens. The
physiology and pharmacology of lung edema. Pharmacol.
Revs. 8: 389, 1956.
412. Visscher, M. B., and J. A. Johnson. The Fick principle:
Analysis of potential error in its conventional application.
J. Appl. Physiol. 5:635. 1953.
413. Visser, B. F., and A. 11. J. Mass. Pulmonary diffusion
of oxygen. Phys. Med. Biol. 3: 264, 1959.
414. Vogel, H. The blood velocity in lung capillaries. Helv.
Physiol, et Pharmacol. Acta 5: 105, 1947.
415. von Basch, S. Ueber eine Funktion des Capillardruckes
in den Lungenalveolen. Wien. Med. Blatter 15: 465, 1887.
416. Wagner, R. Ueber de Beziehungen zvvischen Pulmo-
nalisdruck und Minutensolumen. Z. Biol. 88: 25, 1928.
Wang, Y., J. T. Shepherd, and R. J. Marshall. Evalua-
tion of the slope-volume method as an index of pulmonary
blood volume. J. Clin. Invest. 39: 466, i960.
Wasserman, K., and J. H. Comroe, Jr. A method for
estimating instantaneous pulmonary capillary blood
flow in man. J. Clin. Invest. 41 : 410, 1962.
Wearn, J. T., A. C. Ernstene, A. W. Bromer, J. S.
Barr, W. J. German, and L. J. Zschiesche. The normal
behavior of the pulmonary blood vessels with observations
on intermittence of the flow of blood in the arterioles and
capillaries. Am. J. Physiol. 106: 236, 1934.
Weibel, E. Early stages in the development of collateral
circulation to the lung in the rat. Circulation Research 8:
353. '96°-
Weibel, E. Die Blutgefassanastomosen in der Mensch-
lichen Lunge. Z Zellforsch. 50: 653, 1 959.
Weibel, E. R., and D. M. Gomez. The architecture of the
human lung. Science 137: 3530, 1962.
Weil, P. E., P. F. Salisbury, and D. State. Physiological
factors influencing pulmonary artery pressure during
separate perfusion of systemic and pulmonary circulation
in dog. Am. J. Physiol. 191 : 453, 1957.
424. Weissler, A. M., J. J. Leonard, and J. V. Warren.
Effects of posture and atropine on the cardiac output.
J. Clin. Invest. 36: 1656, 1957.
425. Weissler, A. M, B. H. McCraw, and J. V. Warren.
Pulmonary blood volume determined by a radioactive
tracer technique. J. Appl. Physiol. 14: 531, 1959.
426. Werko, L. The influence of positive pressure breathing
on the circulation in man. Acta Med. Scant! . Suppl. 193,
■947-
427. West, J. B., C. T. Dollery, and P. Hugh-Jones. The
use of radioactive carbon dioxide to measure regiona 1
417.
418.
4>9-
420.
421
422.
423-
428.
429-
43°
431-
432-
433'
434-
435-
436-
437-
438.
439-
440.
441
442.
443-
blood flow in the lungs of patients with pulmonary dis-
ease. J. Clin. Invest. 40 : I, 1 96 1.
West, J. B., K. T. Fowler, P. Hugh-Jones, and T. V.
O'Donnell. The measurement of the inequality of ventila-
tion and of perfusion in the lung by the analysis of single
expirates. Clin. Set. 16:549, 1957.
Wezler, K., and W. Sinn. Das Stromungsgesetz des
Blutkreislaufes. Ar-jieiniitlrl-Porsch. J Beihejt. 3: 1953.
Whittaker, S. R. F., and F. R. Winton. Apparent
viscosity of blood flowing in isolated hindlimb of dog, and
its variation with corpuscular concentration. J. Physiol.,
London 78:339, 1933.
Whittenberger, J. L., M. McGregor, E. Berglund,
and H. G. Borst. Influence of state of inflation of the
lung on pulmonary vascular resistance. J. Appl. Physiol.
15: 878, i960.
Whitteridge, D. Multiple embolism of the lung and
rapid shallow breathing. Physiol. Revs. 30:475, 1950.
Wk.gers, G. J. The regulation of the pulmonary cir-
culation. Physiol. Revs. I : 239, 1 92 1.
Williams, M. H., Jr. Relationships between pulmonary
artery pressure and blood flow in the dog lung. Am. J.
Physiol. 179:243, 1954.
Wilson, R. H., W. Hoseth, and M. Dempsey. The inter-
relations of the pulmonary arterial and venous wedge
pressure. Circulation Research 3: 3, 1955.
Witham, A. C, and J. W. Fleming. The effect of epi-
nephrine on the pulmonary circulation in man. J. Clin.
Inrat. 30: 707, 1 95 1.
Witham, A. G, J. W. Fleming, and W. L. Bloom. The
effect of the intravenous administration of dextran on
cardiac output and other circulatory dynamics. J. Clin.
Invest. 30: 897, 1951.
Wmmersley, J. R. The mathematical analysis of the
arterial circulation in a state of oscillatory motion. Wright
Air Develop. Center Tech. Rept. WADC-Tr 56-61 4, 1958.
Wood, E. H., D. Bowers, J. T. Shepherd, and I. J.
Fox. O2 content of "mixed" venous blood in man during
various phases of the respiratory and cardiac cycles in
relation to possible errors in measurement of cardiac
output by conventional application of Fick method. J.
Appl. Physiol. 7: 621, 1 955.
Wood, P. The Eisenmenger Syndrome or pulmonary
hypertension with reversed central shunt. Brit. Med. J.
2:70'. 755. ]958-
Wood, P., E. M. Besterman, M. K. Towers, and M. B.
McIlroy. Effect of acetylcholine on pulmonary vascular
resistance and left atrial pressure in mitral stenosis. Brit.
Heart J. 19:279, 1957.
Woodbury, R. A., and W. F. Hamilton. The effect of
histamine on the pulmonary blood pressure of various
animals with and without anesthesia. J. Pharmacol. Exptl.
Therap. 71 : 293, 1 941.
Woodbury, R. A., and G. G. Robertson. The one
ventricle pump and the pulmonary arterial pressure of
the turtle: the influence of artificial acceleration of the
heart, changes in temperature, hemorrhage and epi-
nephrine. Am. J. Physiol. 137:628, 1942.
DYNAMICS OF PULMONARY CIRCULATION '743
444. Zierler, K. L. A simplified explanation of the theory of nalen Venennetze. Sitzungsb. k. Akad. Wissen. Malhnaturw.
indicator -dilution for measurement of fluid flow and 0.84:110, 1882.
volume and other distributive phenomena. Bull. Johns 446. Zuntz, N., and O. Hagemann. Untersuchungen fiber den
Hopkins Hosp. 103: igg, 1958. Stoflfwechsel des Pferdes bei Ruhe und Arbeit. Land-
445. Zukerkandl, E. Uber die Anastamosen der Venae pul- wirtsch. Jahr. Z. U'iss. Landwirtschaft sy (Erganzungsband
monales mit den Bronchialvenen und mit dem mediasti- III): 1, 1898.
INDEX
Index
[£ LIBRAR
^V MASS.
Abdominal pressure
venous pressure, 1123
Absorption
general formulation, 961
injury and, 994
extracellular pathway, 1065- 1066
intracellular pathway, 1 065-1066
pathways, discussion of, 1 066
Acetabulum
arteries of, 1652
Acetate ion
arteriolar size and, 948
entrance rate into muscle, 1 1 38
Acetylcholine
arterial diameter and, 807
as coronary vasodilator, 1548
as neuromuscular transmitter, 1366
atrophine antagonism, 1552
blood flow and, 953
cardiac effects, 1552
coronary circulation and, 1562
cutaneous blood How and, 1 346
distribution in abdominal viscera, 141 7
gastric blood flow and, 1446
hepatic blood flow and, 1419
in uterus, estrogen and, 1599
intestinal blood flow and, 1448
joint blood flow and, 1663
mesenteric blood flow and, 1451
pancreatic blood flow and, 1450
pressure-flow and, 952
pulmonary vascular pressures and, 1700
release of norepinephrine by, 1420
secretion in arteriovenous anasto-
moses, 1256
segmental resistance and, 952
skeletal muscle blood flow and, 1 364
skin circulation and, 1 334, 1 345
smooth muscle potential and, 1 155
splanchnic circulation and, 1419
splenic blood flow and, 1450-1451
vascular capacity and, 953
vascular effects of, 972
vascular volume and, 955
venodilator responses, 1094
Acetylsalicylic acid
bradykinin release and, 1242
erythromelalgia and, 1242
Acidosis
acute, pulmonary resistance and, 1722
renal blood flow and, 1504
Adipose tissue
fat metabolism in, 1 175
Acrocyanosis
basic defect in, 1 231 -1232
characterization, 1230
detailed description, 1231
differential diagnosis, 123 1
pathology, 1231
picture of, 1230
sympathectomy in, 1232
symptoms, 1230
vascular mechanisms in, 1231
Actinomyosin
as contractile system, 871
Adenosine triphosphate
intestinal blood flow and, 1448-1449
skin circulation and, 1346
vasodilatation and, 1338
Adenylic acid
pressure-volume relations and, 1085
venodilator responses, 1094
ADH : see Antidiuretic hormone
Adipose tissue
nature of, 1 173
Adrenaline: see Epinephrine
Adrenergic agents
'747
action on vascular beds, 949
venoconstrictor responses, 1 094
Adrenergic blocking agents
hepatic circulation and, 1421
Adrenocortical hormones
atherosclerosis and, 1201
filtration coefficients and, 999
lipid metabolism, iaoi
Adrenocorticotropic hormone
fat mobilization and, 1174-1175
Adventitial cells: see Rougct cells
Aging
aorta and, 875
arterial distensibility and, 808, 876
arterial ratio of radius to wall thickness
and, 876
arterial retraction and, 813
arterial stretch and, 876
blood pressure regulation and, 876
blood vessel wall stiffness and, 809
collagen fibers and, 868
decrease of hysteresis with, 877
elastic tissue and, 86g
peripheral vascular disease and, 12 17
serum cholesterol level and, 11 98
smooth muscle degeneration and, 875
vascular smooth muscle and, 872, 875
venous valves and, 882
wall thickness and, 813
Albumin
appearance in lymph, 1043
binding of fatty acids, 1 1 78
capillaries permeability to, 1013
concentration in human plasma, 974
disappearance from plasma, 1043
egg, molecular sieving of, 1017
extra- and intravascular masses, 1042
infusion, renal function and, 1499
in lymph, 1042
i748
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
in lymph and plasma, 1063
urea and, 1064
molecular weight, 974
osmotic pressure-concentration curves,
972
osmotic pressure due to, 974
permeability of muscle capillaries to,
1013
Alcohol
hepatic blood How and, 1400
Aldosterone
shock and, 1 1 16
Alexander, R. S.
The peripheral venous system, 1075-
1098
Alkali metal ions
ionic radii, 1 137
relative entrance rates into muscle,
.138
relation ion size to electric field mo-
bility, 1 138
Alloxan diabetes
atherosclerosis and, 1201
lipid metabolism in, 1201
Altitude
circulatory and respiratory measure-
ments and, 1 720
Alveoli
capillary gas exchange, 167 1
hypoventilation
alveolovascular reflexes, 172 3
in chronic bronchitis, 1722
in various conditions, 1722
pressure, transmural vessel pressure
and, 1 708
relationship to capillaries, 1676
structure of, 1676
ventilation, determination of, 1682
ventilation-perfusion ratios, 1684
Ammonium
ionic radii of, 1 137
mobility in electric field, 11 38
relative diameter, 1 1 38
Ammonium ion
entrance rate into muscle, 1 138
Amniotic fluid
maternal uterine blood and, 1609
pressure, placental pressure and, 1608
Amyl nitrite
splenic blood flow and, 1450
Amyloid disease
proteins of edema fluids in, 982
Analogue computer techniques
differentiation, 841
Anaphylactic shock
blood and lymph in, 1059
lymphatic system in, 1058
Anesthesia
blood vessel dilation and, 900
hepatic blood flow and, 1405
mesenteric blood flow in, 1443
plasma K response to epinephrine and,
1 152
renal blood flow and, 1488
Angina pectoris
causes of, 1550
origin of pain, 1 565
response to drugs, 1564
treatment of, 1572
Angiogenesis
general aspects, 1 258
in transparent chamber, 1259
Angiotensin (angiotoninl
plasma Na concentration and, 1153
skeletal muscles and, 1355
vascular volume and, 955
Aniline
clearance by rat stomach, 1443
Anions
vascular smooth muscle tension and,
1 '59
Anoxia
capillary filtration coefficients and, 9Q8
capillary permeability and, 997
collateral development and, 1 260
coronary blood vessels and, 1544
effects on fetus, 1635-1638
fetal blood flow and, 1635
fetal hemoglobin and, 1635
fetal resistance to, 1638
filtration coefficients and, 989
intermittent claudication and, 1217
muscle blood flow and, 1378
pulmonary edema and, 1053
reactive hyperemia and, 1375
Anterior pituitary hormones
atherosclerosis and, 1 202
cholesterol metabolism and, 1 202
lipid metabolism and, 1 202
Anticipation
venous pressure and, 1090
Antidiuresis
mean regional transit time, water con-
centration and, 1 50 1
renal medullary circulation in, 1499
Antidiuretic hormone
permeability of collecting ducts and,
1470
urine concentration and, 1472
Antipyrine
capillary clearance blood flow and,
1023
diffusion of, 1002
Aorta
abdominal
arterial flow pulse in, 849
backflow component, 852
circumference, 846
flow pulses, 852
mean peak flow, 846
phase relationships, 851
resonant flow wave, 851
acetylcholine and, 1700
aging and, 875
analogy to nylon stocking, 806
arch
blood distribution of, 849
pressoreceptors of, 875
ascending
blood flow and, 845, 859
circumference, 846
diastolic flow curve, 845
flow pattern, 859
flow pulse through, 845
hemodynamics in, 845, 846, 859
mass-acceleration effects, 846
mean peak flow, 846
mean peak velocity, 846
pressure-flow relationship, 846
velocity profile, 845
as reservoir, 799, 824-835
baroreceptor in, 874
blood distribution in arch, 851
branches of, flow in, 852
central pulse contour, 824-826
diastole, backflow and, 847
distensibility
measurement, 801-814
pressure volume curves and, 8 1 2
electrical analogue, 820
electrolyte levels in, 1 141
extension-release of, 877
foam cell in, 1 170
function as conduit, 814-824
functions of, 799
histological considerations of, 804-806
hysteresis loop, 802-804
length, pulsatile changes in, 813
muscle contraction in, 807
Na and K of, 1 1 4 1
phase relationships in, 772
physiology, 799-838
pressure differences with ventricle, 781-
782
pressure-flow relations in, 820
pressure pulse compared to pulmonary,
1686
pressure pulses, reconstruction of com-
parative values, 825
pressure-volume diagram in, 875
pressure-volume diagram, muscular
arteries and, 879
pulse at various positions, 829
pulse contour in, Valsalva maneuver
and, 831
reflected waves in, 827
resonance
exit tubes and, 827
femoral system and, 828
in visceral arteries and, 829
mode of, 828
origin of, 829
requirements for, 827
standing waves in, 826-829
resonating system and, 827-828
■749
response of related rings, 807
response to stretch, 873
simulated, flow velocity curve, 822
tension-length curve, 801-802
thoracic
arterial flow pulse in, 849
blood flow in, 860
circumference, 846
distensibility modulus, 813
mean peak flow, 846
mean peak velocity, 846
pressure-flow values, 823-824
pressure-volume diagram, 873
pressure-volume relationships, 812-
813
stretch curves of, 802, 811
stretch-release curves, 812
vaso vasorum in, 884
wall thickness, pulsatile changes in, 813
Aortic flow
axial
pressure difference and, 847
pulse in, 845
computer solution to, 847
curves, 82 1
at rest, 773
during sympathetic stimulation, 773
in exercise, 772
pressure and, 772
stroke volume and, 82 1
ductus arteriosus and, 861
during exercise, 846
measurements in, 800
pattern, 820
pressure changes and, 82 1
piessure relationship in values, 848
pulse in, 845
pulse, in stenosis, 859
relationships in, 849
see also Aorta; Cardiac output
Aortic regurgitation
aortic flows and, 1 556
backflow wave in, 860
brachial pulse contour in, 831-832
discussion of, 860
left ventricular ejection pulse and, 860
flow-pressure gradient in, 860
myocardial ischemia, 1556
reversible aortic insufficiency, 1 556
Aortic stenosis
aortic arch pressure and, 860
elevated ventricular pressure and, 1555
flow curve and, 860
flow pulse in, 859
stroke volume and compensatory
mechanisms, 860
Apresoline : see Hydralazine
Aramine : see Metaraminol
Arbacia egg
filtration coefficients in, 992
Arfonad: see Trimethaphan camphor -
sulfonate
Arrhythmias
coronary blood flow and, [541
effectiveness of heart and, 11 04
Arterial disease
chronic, vasoconstrictor mechanism,
1237
obliterative, vascular dilatability and,
1223
occlusive, pain in, 12 18
vasoconstrictor mechanisms in acute,
1236
Arterial spasm
basis for improvement, 1237
elastic recoil and, 1237
possible role of serotonin, 1237
Arterial system
abnormal, communications, 1256
analogy approach, 844
experimental testing of, 850
frequency response, characteristics,
848-849, 851
gas embolism, skin circulation and,
1337
harmonics of, 818-821
harmonics, phase lag and, 817
low-pass filtered hydraulic supply, 847
nonpassive network, 843
physical properties, 841
pressure -pulse augmentation, 848-849
resonant frequency, 848-849
resonant-network model, 847, 850
resonant wave amplitude, 848-849
segmentation of, 843
size range, 841
transient, response method, 844
transmission line model, 849, 851, 852
values of elements, 843
windkessel model, electrical analogue
of, 848
Arteries
anatomy of central and peripheral, 880
arcuate patterns, 905
arrangement of longitudinal muscles in,
878
arteriolar branch enlargement at ori-
gin, 896
branches, angles of, 896
changes after extirpation, 801
collateral, wall structure in, 1 268
contraction
propulsion of blood and, 871
relaxation of, and, 87 1
distensibility
age and, 808, 875
in living vessels, 804
muscle contraction and, 806
elastic
anatomy, 877
characteristics of, 873
discussion, 863-877
distensibility, 873
hysteresis in, 879
interaction of elements, 876
mechanical behavior, 878
model of, 876
pressure-volume diagrams, 879
role of smooth muscle, 877
schema of behavior, 876
tension muscles in, 874
visco-elastic and plastic behavior, 877
evaluation of status, 1223
excised, stretched rings, 802
from red pulp of spleen ampulla, 912
harmonics, factors affecting, 820
length and diameter changes of, 813
length changes, H 1 j
ligation, after sympathectomy, limb
temperature and, 1264
major
function of, 799
histology, 804
loss of blood pressure in, 801
physiology, 799-838
mediation of reflex vasospasm, 1236
muscular
anatomy, 877
discussion, 872, 877-880
hysteresis in, 879
mechanical behavior of, 878, 879
pressure-volume diagrams, 879
schema of behavior, 879
smooth muscle in, 880
stress-relaxation in, 872
structure, 877
muscle ring, response to loading, 1084
nutrient
ligation, marrow infarction and, 1658
physiology, 1677
peripheral
pulse contours, 828
vascularization of wall, 885
vasa vasorum of, 885
precapillary branch, camera-lucida
drawing of, 898
pressure-volume diagram, 873, 879
ratio of radius to thickness, age and, 876
resonance in, 826
ring and longitudinal muscle arrange-
ment, 878
ring muscles of, 873
spasm of, myogenic factor, 1236
spontaneous rhythmical contractions,
898
standing waves in, 826
stretch
age and, 876
hysteresis loop, 802
tension muscles of, 873
valves, structure, 781
vascular tone, source of, 1409
vasa vasorum in, 885
wall thickness of, 812
Arterioles
arcuate systems of, 896
identification of, 909
receptor sites for blood flow control, 949
terminal, contractions in, 899
i?5c
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
terminal, definition, 891
terminal, vasomotion of, 897
vasomotion in, 8g6, 900
Arterioportal anastomoses
in liver, 922
Arteriovenous anastomoses
abnormal description, 1 256
blood How and, 1261
body temperature and, 921
classification, 1251
collateral circulation and, 1257, 1261
compared to A-V bridges, 92 1
criteria, 909
factors affecting, 972
graphic reconstruction, 1253
in cochlea, 916
in conjunctiva, 908-910
in ilial submucosa, 907
in lungs, 880
in panniculus carnosus, 902
in rabbit ears, 92 1
in skeletal muscle, 902, 1358
in stomach, 905
in submucosal plexus, 906
in tongue, 1254
in various tissues, 921
microcirculation and, 899
nor m, il
development, 1254
distribution, 1252
factors affecting, 1 255
fate, 1254
function, 1252
role, 1256
size, 1252
structure, 1252
opening of, blood flow and, 1257
precapillary, 1251
spontaneous rhythm, 92 1
structure, 880, 92 1
temperature regulation and, 922
traumatic or surgically induced, 1257
wall structure, 880
Arteriovenous bridges
blood flow in, 918
compared to arteriovenous anasto-
moses, 92 1
description, 918
Arteriovenous communications
cutaneous arterial spider, 1256
fistula, musical qualities of, 859
fistula, traumatic, 1257
in rabbit ear, 1255
in spleen, 912
normally occurring, 1252-1256
surgically induced, 1257
Arteriovenous pathways
antimesenteric area, anastomoses in,
905
arcuate patterns in terminal vascular
bed, 897
definitions, 891
in bat wing, 895-898
in bulbar conjunctiva, 908-910
in cochlea, 915-918
in hamster cheek pouch, 900
in lung, 913-915
in mesentery, 900
in myocardium, 903-904
in rabbit ear, 898-goo
in skeletal muscle, 900-903
in skin, 904
in spleen, g 10-9 13
in stomach and intestine, go4-go8
structure of in terminal vascular beds,
895-923
techniques for microscopic observation
of, 892-895
through terminal vascular beds, 923
Artificial membranes
filtration coefficients through, 992
Artificial perfusion
pressure gradients, resistance change
and, 1089
Artificial respiration
lymph flow and, 1052
Ascites
liver, lymphatics and, 1051
Atherogenesis
chickens and, 1201
cholesterol-induced, 1201
cholesterol-induced corticoids and,
1202
Atherogenic diets
thyroid activity and, 1 200
Atheromatous lesion
accumulation of lipophages, 1 1 70
Atherosclerosis
/-epinephrine and, 1202
Atheroma
derivation of constituents, 1188
lipid constituents, 1 188
Atherosclerosis
adipose tissue and, 11 73- 1175
adrenocortical activity and, 1201
alloxan diabetes and, 1 201
anterior pituitary hormones and, 1 202
basic vascular lesions, 1 1 g8
blood clotting and, ii8g-iigo
blood lipids and, 11 87-1 189
chickens
production of, 1201
sex differences, 1204
coronary heart disease and, 1 1 98
complications, 11 98
definition, 1 167, 1 197
diabetes and, 1 187, 1200
diet and, 1 171-1 172
dietary fats and, 1 167
elastic tissue and, 1168
emotional factors and, 1 198
emotional stress and, 1208
etiology of, 1 1 67
experimental
hyaluronidase and, 1202
in rabbit, 1 i6g
factors affecting, Ii8g
fat absorption, digestion and, 1 1 72-
"73
feeding patterns and, 1207
foam cells and, 1 168
heredity and, 1205-1206
hormones and, 1 197-1205
hypertensive patients and, 1169
hypothyroidism and, 1 187
ingestion of food and, 1 171
initial lesion, 1 168
insulin and, 1200
interplay of factors, 1208
lipid metabolism and, 1 167-1190
lipoprotein diffusion and, 1046
metabolic disturbance in, 1197
pathogenesis, 1 168-1 1 7 1
pathology, 1 168
physical activity and, 1206
serum cholesterol level and, 1 197
serum lipids and, 11 75-1 180
sex differences, 1 203-1 204
sex hormones and, 1203
special manifestations, 1 188
species differences, 1 205
strain differences, 1205
stress and, 1206-1208
thrombus formation and, 1 189
Atria
contraction, valve closure and, 781
filling
changes of venous return and, 786
during ventricular systole, 786
factors controlling, 787
pericardium and, 7g3
ventricular contraction and, 788, 794
X wave and, 786
function of, 769
left and right compared, 789
muscle fascicles, 1518
myocardial fibers in, 1518
pressures
characteristics, 776
complex graphical analysis of, 1 1 1 8
drop after contraction, 776
lymph flow and, 1052
pulmonary edema and, 1052
recording of, 776
V point, 776
X point, 776
Y point, 776
Z point, 776
role of, 768
system, steady -state oscillation, 844
systole, functions, 787
systolic reserve volume, 786
volume, 783-786
at different stages of cardiac cycle,
785
difficulties of determining, 785
INDEX
■751
exercise and, 786
terminology, 785
under various conditions, 782
various activities and, 785-786
Atrioventricular valves
asynchrony in closure, 790
closing, 775
mechanism of closure, 780
movement of, 781
passive movements of, 779
structure, 780
Atrium, left
pressure in
acute left heart failure and, 1 120
acute right heart failure and, 1121
lung edema and, 991
relation to right at birth, 1640
venous pressure and, 1 1 18-1 1 19
Atrium, right
pressure in
acute left heart failure and, 1 1 20
acute right heart failure and, 1 121
blood transfusion and, 1 1 1 4
blood volume and, 1 1 14
cardiac decompensation and, 1 1 1 7
cardiac output and, 1 105
congestive heart failure and, 11 16
disease and, 1710
drugs and, 1 564
exercise and, 1 1 13
heart recompensation and, 1 1 17
hemorrhage and, 1 1 1 4
muscular exercise, 1 1 1 3
myocardial damage, 1 1 15
open chest and, 11 15
peripheral venous pressure and, 1 127
rapid transfusion and, 1 1 1 3
relation to left at birth, 1640
shock and, 1 1 14
simplified graphical analysis of, 11 03
sympathetic stimulation and, 1112
vasomotor tone, 1 1 1 3
venous collapse and, 1 107
venous pressure and, 1 1 26
venous return and, Ilia
Atropine
estrogen-induced hyperemia and, 1601
intestinal blood flow and, 1^-17
resistance in vascular bed and, 948
skin blood flow and, 1340
Autonomic nervous system
anatomical arrangement, 1723
blockade, hepatic circulation and, 1420
distribution of, 1526
fat mobilization and, 1 1 75
myogenic activity and, 880
Autoregulation
myogenic theory, 946
perfusion pump schema for, 943
Axon reflex
capillary pressure and, 996
femoral artery and, 1 380
Bader, H.
The anatomy and physiology of the
vascular wall, 865-889
Barcroft, H.
Circulation in skeletal muscle, 1353-
■385
Baroreceptors
extrasplanchnic, 141 6
in young animals, 1643
reflex, in skeletal muscle vessels, 1363
stimulation
skeletal muscle blood flow, 1363
skin circulation and, 1344
Basal data
flow values for, 1537
left ventricle, 1537
Bat wing
arteriovenous pathways, 920
blood flow in capillary bed, 895
microcirculation in, 894, 895
Bernoulli effect
description of, 779
Bernoulli's theorem
applied to flowmeters, 1296
Bier's spots, 1 225
Bile
lymph transport and, 1045
release of alkaline phosphatase from
intestinal mucosa and, 1047
Bile acids
enterohepatic cycle of, 1 180
formation from cholesterol, 1 1 79
formation of, 1 180
sequestrants, characterization, 1186
Birds
protein osmotic pressure in, 975
Bladder distention
skin circulation and, 1 344
Blood
chemical composition of, 1544
cells, injury with perfusion pumps, 1370
clots, relation to thrombi, 1 190
clotting, thrombosis, 1189-1190
coagulation
anaphylaxis reaction, 1059
dietary fat and, 1 189
serum lipids and, 1 168
propulsion by arteries, 871
transfusion
cardiac output and, 1 1 14
right atrial pressure and, 11 14
venous return and, 1 1 14
velocity, measurement of, 1302
viscosity
apparent flow and, 938
apparent in vivo and in vitro, 940
pressure-flow relationships and, 940
Blood brain barrier
capillary permeability and, 1013
Blood flow
acetylcholine and, 953
arterial
arcades, 906
oxygen saturation and, 1 1 25
perfusion pressure and, 936
pressure, extravascular, effect on,
94', 1378
systemic, 845-854
arteriovenous
anastomoses and, 899, 1261
difference pressure and, 937, 938
autoregulation of, 880, 943
A-V bridges in, 918
axial, 842, 1295
equation for, 842
blood-tissue exchange and, 1019, 1023
capillary
clearance and, 1023
contractility and, 923
gas exchange time, 1 706
measurement of, 1702
nets, 896, 897, 898, 899, 900
permeability, 1019-1024
pressure and, 969
rate, hemodynamic influences, 1706
value of, 1702, 1706
changes in velocity, 839
collateral development and, 1 260
critical closing pressure, 1408
debt, repayment of, 1539
determination of vascular behavior
from, 957
edema and, 969
extramural, 1528
flow -pressure curves, 1700
heart-lung preparation, 1527
inflection of wall and, 1 298
laminar, parabolic velocity profile, 1295
lateral pressure and, 941, 1378
lowered impedance in small artery
and, 853
mean transit time, perfusion pressure
and, 954
metabolic control, 1024
nonlaminar flow rules, 858
oxygen consumption, blood flow, tem-
perature and, 1022
perfusion pressure and, 943, 954
phasic, 1528
methods, 1528- 1529
pressure relationships, 1408
pulsatile
arterial inflow, 956
in peripheral arteries, 1296
in pulmonary vessels, 855-856
in systemic arteries, 845-854
in systemic veins, 854-855
methods of measurement, 839-841
rigid and distensible tubes, 814-816
receptors for control in muscle, 949
relation to pressure, equations 814,
88o, 951
175^
HANDBOOK OF PHYSIOLOGY
IJIRCILATION II
spontaneous rhythmical contractions,
898
temperature, O2 consumption and,
1022
thermal conductivity and, 1287
through distensible tube, 824
through rigid system, 824
vasoconstrictor fiber stimulation and,
971
vasomotion and, 926
vascular behavior and, 935
venous
collapsibility and, 1080
pressure and, 1 1 26
system and, 1079
venovasomotor reflex, 1 282
see also Circulatory autoregulation
Blood flow measurements
Aschofl and Wever ring element for,
1285
bubble flowmeter, 1 280-1 281, 1530
capillary, 1 706
cerebral blood flow, 1290
collateral circulation, 1270
computer techniques and, 840-841
dilution methods, 1291
discussion of role, Reynold's numbers
and, 131 1
drop chamber, 1278
during venous occlusion plethysmog-
raphy, 1282
dye studies, precautions, 1 399
electrolytic polarization, 1321
electromagnetic-induction principle,
^V^-lV% '53°
instruments for, 1278-1287
isotopes and, 1530
mathematical relation
to resistance, 939
to pressure, 939
methods, 839-841, 1278-1287
based on Ludwig's principle, 1278-
1279
clearance techniques, 1397
extraction techniques, 1397
nitrous oxide method, 1 290
nuclear magnetic resonance, 1321
outflow measurement, 1279-1280
photoelectric plethysmography, 1283
plethysmography, 1281 -1283
Prandtl's theory, 1311
pulse plethysmography, 1 283
regional determinations, 1022, 1290-
1293
RCA 5734 transducer, discussion of,
1308
Reins thermostromuhr, 1 283-1285
schema of direct -recording flowmeters
1279
strain-gauge occlusion technique, 1282
tachograph, 1283
thermal methods, 1 283- 1 287
traveling markers, 1320
ultrasonic flowmeters, 1318
venous outflow collection, 1278
see also Flowmeters
Blood flow, regional
aorta, 800
ascending aorta and, 845, 859
zero drift, 840
ascending vs. thoracic aorta, 847
bone, 1 656- 1 658
carotid artery flow, 853
conjunctival capillaries, 908
descending thoracic aorta in, 860
gastric, 1443
factors affecting, 1446
muscosa, blood flow in, 1443
mucosal secretion and, 1452
tetranitrate and, 1447
hand
emotion and, 1339
heating of legs and, 1 343
vessel innervation and, 1339
Ilidar, intestine and, 1447
intestine, 1443
intrauterine pressure and, 1600
mesenteric circulation, hematocrit and
blood volume, 1445
pancreas, 1441 -1443
factors affecting, 1450
pancreozymin, 1450
secretion and, 1452
placental, functional implications of
venous drainage, 1588
splanchnic
effect of dopamine on, 141 7
factors affecting, 1429
method, 1392
splenic, 1443
epinephrine and, 955
factors affecting, 1450
sympathetic stimulation and, 955
stria vascularis, 9 1 7
systemic arterial differential pressure
flowmeter and, 345
through portal vein, 1439
umbilical
epinephrine and, 1633
factors influencing, 1633
in lamb, 1633
in man, 1633
in sheep, 1632-1634
norepinephrine and, 1633
uterine
electrolyte exchange, 1594
-pinephrine, effects on, 1637
norepinephrine, 1637
see also Cerebral blood flow; Cutaneous
blood flow; Hepatic blood flow;
Pulmonary blood flow; Renal
blood flow; Skeletal blood flow
Blood pressures
age and, 1643
aortic
cardiac output and, 1534
coronary flow and, 1534
flow during exercise and, 846
arterial
blood flow and, 1378
bone blood flow and, 1657
cerebral venous outflow and, 946
intrarenal pressure and, 1478
lateral pressure, flow and, 941
mesenteric artery, occlusion of, 1449
myocardial damage and, 1 1 1 6
portal venous flow and, 1440
serotonin and, 1089
skeletal muscle blood flow and, 1357
uterine contractions and, 1599
values for, 1714
arteriovenous difference
blood flow and, 937, 938
intestinal blood flow and, 1449
atheroma formation and, 1208
basal tone and, 1 356
bone medullary canal, pressure, and,
1 660- 1 66 1
brachial artery, Valsalva maneuver
and, 171 2
brachial pulse and, 831
capillary, 962-972
changes
after birth, 1642
during birth, 1642
coronary flow
coronary sinus oxygen saturation
and, 1539
epinephrine and, 1551
stellate ganglion stimulation and,
1548
determination of vascular behavior
from, 957
diastolic, vessel wall tension and, 1 139
differentials, equation for, 1 297
digital, response to noradrenaline, 1229
drugs and, 1564
exercise, 846
hyperemia and, 1378
extracellular Na and K and, 1 154
fall, collateral development and, 1260
flow relations
Acetylcholine and, 952
and simultaneous measurement of,
840
source and, 844
hemorrhagic shock and, 1535
histamine and, 970
in bone marrow, 1659
adrenaline and, 1661
nerve stimulation and, 1660
in digital arteries, temperature and,
968
lactic acid and, 1722
lateral
principles of, 1300
I Mil \
!753
simultaneous measurement of, 840
uses of, 1300
leg blood flow and, 1 356
loss in major arteries, 801
mathematical relation
to flow, 937, 939
to resistance, 939
mean circulatory
compared to mean systemic pressure,
1109
measurements of, 1 1 09
mean systemic
acute left heart failure and, 1 1 20
acute right heart failure and, 1 1 2 1
blood volume and, 1 109, 1 1 10
congestive heart failure and, iiiti
definition, 1107, 11 09
factors affecting, 1 1 1 1
value, 1 1 09
vasomotor tone and, 1 1 10
venous return curve and, 1 109-1 1 10
plasma Na and, 1 152
positive pressure breathing and, 1 7 1 1
potassium changes and, 1 153
pressure-pulse contour and, 832
regulation
age and, 876
Na transfer and, 1 153
relation to flow, equations, 814
sodium and potassium
sodium changes and, 1 153
systemic, altitude and, 1720
vasomotion and, 926
vascular behavior and, 935
vessel wall structure and, 1 269
volume relationships, blood injection
and withdrawal, 1085
wall stress and, 875
see also Capillary pressure; Venous
pressure
Blood tissue exchange
blood flow and, 1019, 1023
capillary permeability and, 1019
of ions, 1 02 1
small molecules, 1021
Blood vessels
alpha receptor in, 949
automaticity of
in man, 1 356
physiological condition and, 1356
basal tone of, 1 355
beta receptor in, 949
cochlear perilymph, endolymph and,
9'7
delta dilator receptors in, 949
denervated
adrenaline and, 1341
noradrenaline and, 1341
different types, 872-883
dilated, distensibility pattern, 1084
distensibility and resistance charac-
teristics, 1674
excised, tension-length curves, 804
function, 865
gamma dilator receptors in, 949
innervation of, 1338
self-differentiation, 1 258
small, relation to lymphatic capillaries,
'°37
Blood vessel walls
active transport in, 883
aldosterone and, 1 202
anatomical elements, 866
anatomy of, 865
arrangement of muscle fibers in, 806
basal tone, 880
derivation, 880
myogenic activity and, 880
wall tension and, 880
changes in Raynaud's syndrome, 1227
collagen fibers in, 868
collagen tissue, 866, 868
hysteresis, 868
irreversible elongation, 868
maximal extension and, 868
maximal tensile strength, 868
safety factor, 868
stress and, 865
structure, 868
diffusion in, 883
diffusion limit for nutrition, 883
elastic tissue, 866, 869
aging and, 869
function of, 869
hysteresis, 868
irreversible elongation, 868
maximal extension, 868
maximal tensile strength, 868
structure, 869
elements of, 866-872
endothelium of, 868
cement and, 883
estrogens and, 1204
extensibility
elastic modulus and, 808
measure of, 811
muscle contraction and, 807
of structural elements, 805
fetal, elastin-collagen distribution, 1623
ground substance of, 868
hormones and, 1 1 99
interaction of elements in, 876
location of vasa vasorum, 885
lumen diameter and wall thickness, 866
lymphatics of, 886
model of, 883
muscle of, arterial spasm and, 1236
network of structures, 805
nutrition, 883-886
hypertension and, 886
lymphatics and, 886
of central and peripheral arteries, 880
of venous system, 881
oxygen uptake, 1198
physiology of, 865
properties of, 866
qualities, due to in series arrangement
of elements, 867
relaxation of, 1544
ring muscles in, 869-870
sodium transport in, 884
species differences, 1 198
stiffness, aging and, 809
stress, formula, 865
structure
pressure and, 1269
relation to stretch hysteresis, 803
tension
hydrostatic pressure and, 881
in veins, 882
loading and, 1084
muscles in, 869-870
pressoreceptors and, 874
vascularization, 885, 886
in peripheral arteries, 885
vasa vasorum penetration, 885
venous, arterial supply, 885
see also Vascular smooth muscle; Endo-
thelium
Blood volume
atrial pressure and, 1 1 22
capillary, total blood volume and, 1705
cardiac output and, 1 1 14
cardiovascular dynamics, 1 122
determination of vascular behavior
from, 957
estimation of change, 954
expanded, capillary pores and, 1053
extrinsic influences, 954
factors affecting, 954
maximal effect, 1361
mean systemic pressure and, 1 109, 1 1 10
mesenteric, 1445
muscle circulation vasoconstrictors and,
1361
perfusion pressure and, 954
pulsatile changes in, 956
regional, shifts in, 1694
regulation, lymphatic return and, 1042
resistance to flow and, 956
right atrial pressure and, 11 14
vascular
beds in, 953-954
behavior and, 935
venous return and, 11 10, 11 14, 11 22
curves and, 1 1 10, 1 122
ventricular outputs and, 1 1 2 1
Body build
coronary atherosclerosis and, 1 206
Body position
uterine contractility and, 1600
venous pressure during pregnancy and,
1601
Body size
hepatic blood flow and, 1405
Body temperature
1754
HANDBOOK OF PHYSIOLOGY
CIRCl'LATION II
arteriovenous anastomoses and, 92 1
cutaneous circulation and, 1342
regulation, skin circulation and, 1341
Body temperature
role of skin anastomoses, 1 326-1327
skin blood flow and, 1 339
Bone
bone marrow temperature, 1661
content of human limbs, 1327
cortical, Haversian canals, 1653
erythrogenic nests, oxygen supply to,
1659
intramedullary pressure, 1 659-1 661
marrow, blood pressure in, 1659, 1660
medullary canal pressure, 1659
blood pressure and, 1 660-1 661
blood volume and, 1659- 1660
factors affecting, 1660
Valsalva maneuver and, 1660
nerve supply, 1 656
oxygen content of blood, 1658- 1659
tubular, vascular organization, 1652,
1653
vascularity of, 1651
vertebra, diagram, 1654
Bone blood flow
arterial pressure and, 1657
capillaries in, 1654
in man, 1657
quantitative measures, 1657
Bone circulation
efferent fibers, 1656
flat bones, 1655- 1656
hyperemia, in handling, 1657
infarction in, 1656- 1657
long bones, 1 651 -1655
reduction of, 1658
resorption and, 1657
vertebrae, 1655
Bone marrow
afferent fibers, 1656
oxygen saturation of blood in, 1659
temperature of, 1661
species difference, 1661
Brachial artery, dimensions and flow, 849
pulse
contour in, Valsalva maneuver and,
831
pressure change in, 826
Bradykinin
capillary permeability and, 1061
release, acetylsalicylic acid and, 1242
skin circulation and, 1346
vasodilatation and, 1340
Brecher, G. A.
Functional anatomy of cardiac pump-
ing. 759-798
Bretschneider Pitot meter
principle and formula for, 1301
see also Flowmeters
Brisket disease see Mountain sickness
Bristle flowmeters : see Flowmeters
Brodie Trendelenburg test
description, 1223
Broemser s differential sphygmograph
discussion of, 1 299
principle of, 1302
Bi omosulfophthalein
conjugation of, 1398
extrahepatic removal of, 1 398
metabolism of, 1398
removal from blood, 1397
transfer from blood to bile, 1397
value in clearance studies, 1 400
Bromide
entrance rate into muscle, 1 1 38
Bronchi
diameter in relation to bronchial ar-
teries, 1 266
Bronchial arteries
diameter in relation to diameter of
bronchi, 1 266
flow
collateral vessels, 1679
pulmonary arterial relationships,
1678
nutrient, 1677
Bronchiectasis
pulmonary collateral circulation, 1679
Bronchitis
pulmonary artery pressure and, 1710
right atrium pressure and, 1 7 1 o
Bronchomotor tone
bronchospasm, 1 7 1 7
effects of, 1 7 1 6
elastance of, 1 7 1 7
Bubble flowmeter: see Flowmeters
Bulbar conjunctiva
microcirculation in, 893, 908
Bundle of His
discussion of, 15 18
Burch, G. E.
Peripheral vascular diseases — diseases
other than atherosclerosis, 12 15-
IJ49
Burns
capillary permeability and, 1061
lymph
composition and, 1061
flow and, 1061
protein in, 985, 1061
lymphatic system and, 1061
porosity of capillary wall in, 1000
protein
in interstitial fluid, 985
of lymph and plasma, 1061
vasoconstrictor substance produced,
1 06 1
Caffeine
renal lymph flow and, 1058
Calcium
arteriolar size and, 948
entrance rate into muscle, 11 38
mobility in electric field, 11 38
relative diameter of ion, 1 1 38
smooth muscle tension and, 1 1 57
vascular
effects of, 972
resistance and, 1158
smooth muscle tension and, 1157-
1158
vasoconstriction and, 1 158
Calcium chloride
mean activity coefficients, in given
solution, 1 137
Calorimetry
of skin, as measure of blood flow, 1327
Capacitance
definition, 820
relation to effect of change in resistance,
1 1 08
terminal vascular beds, 935-957
Capillaries
arterial, description, 899
bed
collaterals to, diagram of, 936
density, oxygen diffusion and, 1020
blood and lymph, morphology, 1062
blood flow in, goo
nets, 896, 897
caliber changes, 899
contraction, 923, 924
blood flow and, 923
count in muscle, 1019
definition of, 892
density, conducting tissue of heart and,
152 1
description of, 9 1 o
diagram of average limits, 985
filtration coefficients of, 988-1000
function of, 880
adsorbed protein, 994
growth of, 899
in skeletal muscle, 902
injured, filtration coefficients and, 995
interendothelial junction, 1012
intramural, in venous walls, 885
limb, protein of capillary filtrate, 982
membranes, filtration coefficients
through, 992
mesenteric
effects of various substances on, 998
filtration coefficients of, 998
fluid exchange, 996
microaneurysm in diabetes, 1 200
muscles in, 1358
nets, blood pressure in, 896, 897
origin in various tissues, 920
para-circulations, 986
pores of, 1002, 1 01 1
pore size
calculation, 1005
1 755
diffusion and, 1003
effective radius, 1002
hydrodynamic flow and, 1003
in various areas, 1006, 1043
"pore stretching", 1053
pore structure, 101 1
porosity of injured wall, 999
protein passage through, 986
stasis
permeability to protein and, 997
pressure and, 997
steps leading to, 995
stoppage of flow in, 880
structure, 880
in lungs, 1676
in muscle, 1009
regional differences, 1013-1018
transitional, in pigeon bone, 1654
"true", 1 52 1
venous, leaks in, 1014
walls
exchange of substances through, 961-
1024
filtration and absorption, 961-962
Capillary filtrate
circulations of, 986
production of, 986
protein content, 982, 985
volume formed, 987
see also Lymph; Interstitial fluid
Capillary permeability
anoxia and, 997, 1053
arterial disappearance curves and, ioog
blood brain barrier and, 1013
blood flow and, 971, 1 01 9-1 024
blood-tissue
exchange and, 1019
permeability and, 1023
burns and, 1061
comparison with blood-tissue permea-
bility, 1023
definition, gg2
dilatation and, 1059
factors, 1 06 1
histamine and, 1059
hypoxemia, systemic and, 998
in liver, 1050
in various organs, 1012
leaks and, 1014
lipid-insoluble molecules and, 1007,
1013
lipid-soluble molecules, 1 01 8-1 01 9
local hypoxemia and, 998
lymph formation and, 1040
molecular sieving and, 1005
natural mediators, 1061
osmotic pressure, flux rate and, 1010
pinocytosis and, 1017
platelets and, 995
protein and, in stasis, 997
respiratory gases and, 1018
substance produced in shock and, 1060-
106 1
sympathetic stimulation and, 1 362
tissue metabolism and, gg7
traumatic shock and, 1 05g
urea and, 1064
uterine, estrogen and, 1601
Capillary pressure
blood flow and, g6g
dependence on venous and arterial
pressures, g6g
direct measurement, g64
direct methods, requirements, g63
epinephrine and, g68
factors affecting, g66, g72
nitration coefficients and, ggi
fluid exchange and, 988
functional changes of, g68
histamine and, 996
hydrostatic and isogravimetric, 990
hyperemia and, 968
hypertension and, 968
injury and, gg6
in kidney, g65
in lungs, g66
in nets, 8g6, 897
in retina, 966
in various species, 965
methods of measurement, 962
osmotic pressure of plasma proteins
and, 963, 965
pressure gradient and, 965
protein osmotic pressure and, 9gi
postglomerular, g66
Raynaud"s disease and, g68
spatial relation of capillary bed to heart
and, g67
stasis and, gg7
suction forces, g66
sympathetic stimulation, g7 1
temperature and, g68
variability of, g66
vasodilatation and, g68
venous pressure and, g68
Carbohydrate
absorption and metabolism of, 1 1 7 1
Carbonate
entrance rate into muscle, 1 1 38
Carbon dioxide
capillary filtration coefficients, gg8
cerebral blood flow and, 946
filtration coefficients and, 997
heart and, 1545
mesenteric filtration coefficients and,
igg8
pleural and esophageal pressures and
i6go
reactive hyperemia and, 1375
skin circulation and, 1346
Cardiac blood supply
basal data, 1537-1538
capillaries
density in adults, 1521
density in newborn, 152 1
functional anatomy, 1517-1527
methodology of study, 1527-1530
preparation for study, 1527-1530
see also Coronary blood flow
Cardiac cycle
apex cardiogram, 777-779
atrial pressures, 776
attempts at empirical time correlations,
777
circulatory system cyclical events and,
779
composite graph, 778
description of, 770
diastole and systole, 770
electrocardiogram, 777
intraventricular- streaming, 772
isometric contraction, 770
isovolumetric
relaxation, 773
ventricular contraction, 770
position changes of great vessels, 766
pressure and flow effects during, 769-
775
pressure-volume phase relationships,
774
protosystolic phase, 779
scheme, 770
time intervals, 770, 791
between valvular motions, 7go
vibrocardiogram (apex cardiogram),
777-779
Cardiac decompensation
cardiac output and, 1 1 1 7
right atrial pressure and, 1 1 1 7
venous return and, 1 1 1 7
Cardiac lymph
cardiac work and, 1055
composition, 1055
flow rates, 1 055
obstruction, ECG changes, 1055
Cardiac muscle contractility
coronary blood flow and, I54g
drugs and, 1564
Cardiac nerves
autonomic innervation, 1526
central communications, 1526
plexi
autonomic fibers, 1526
division of, 1525
see also Autonomic nervous system
Cardiac output
acute left heart failure and, 1120
acute right heart failure and, 1121
altitude and, 1720
aortic
flow during exercise, 846
pressure, coronary flow and, 1534
at birth, 1641
blood
1756
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
transfusion and, 1 1 1 4
volume and, 1 1 14
cardiac
decompensation, 1 1 1 7
tamponade and, 1 1 05
change in resistive load and, 1103
closed-chest massage, 761-762
congestive heart failure and, 1 1 16
drugs and, 1564
exercise, 846, 1 1 1 3
feeding and, 1452
fetal, 1628
anoxia and, 1635
heart
failure and, 17 15
recompensation and, 1 1 1 7
hemorrhage and, 1 1 14
hypoefTective heart and, 1 104
limiting factor, 761
local oxygen utilization and, 1 1 26
magnitude, 987
massage, closed-chest, 761-762
muscular exercise, 1 1 1 3
myocardial
damage, 1 1 1 5
infarction, 1 1 15
open chest and, 1105, 11 15
oxygen uptake and, 1682
pressure on outside and, 1 104
rapid transfusion and, 1 1 1 3
right atrial pressure and, I 105
shock and, 1 1 14
simplified graphical analysis of, 1 103
sympathetic stimulation and, 1 1 12
values for, 1720
vasomotor tone, 1 1 1 3
vascular resistance and, 1 1 18
venous return and, 1 1 1 2
Cardiac output curves
cardiac
load and, 1 104
tamponade and, 1 1 05
closed-chest animals and, 11 03
factors affecting, 11 04, 11 05
heart effectiveness as a pump and, 1 104
open -chest
effects, 1 1 04
venous return, 1 1 1 1
curves and, 1 1 03, 1 1 05
Cardiac pumping
development of concept, 759
electrocardiogram and, 777
extracardiac structures and, 761
functional anatomy, 759-798
mechanical analogues, 760
Cardiogenic shock
discussion of, 1564
drug action on, 1 564
pressor agents and, 15(14
vasodepressor drugs, 1 567
Cardiovascular dynamics
analysis
of decompensation, 1 1 1 7
of recompensation, 1 1 1 7
blood volume and, 1 1 10, 1 122
left ventricular weakening and, 1 120
myocardial infarction and, 1 1 15
open chest and, 1 1 15
rapid transfusion and, 1 1 14
right ventricular weakness and, 1121
shock and, 1 1 14
simultaneous analysis of right and left,
1119
sympathetic stimulation and, 1 1 1 2
see also Vascular hydraulics
Cardiovascular reflexes
development of, 1629
Cardiovascular system
development of, 1623-1624
factor affecting, 1624
fetal, species differences, 1620
Carnivores
uterus in, 1585
Carotid artery
arterial flow pulse in, 849
dimensions and flow, 849
flow patterns, 853
muscle tension, electrolytes and, 11 46
pulse-pressure change in, 826
stress-relaxation curves of, 872
Carotid sinus
blood pressure responses, 1094
pressoreceptors of, 874, 875
response, in young animals, 1643
stimulation by noradrenaline, 874
Cartilage
growth, blood supply and, 1 658
Castelli principle
definition of, 1 306
Cat
blood pressure, age and, 1643
capillary blood pressure in, 965
heartbeat at birth, 1641
protein osmotic pressure in, 965
pulmonary vessels in, 1674
Catch mechanism
definition, 871
Catecholamines
coronary heart disease and, 1181
distribution in abdominal viscera, 1417
of blood, in Raynaud's syndrome, 1229
vasomotor adjustments and, 141 7
see also Norepinephrine; Epinephrine
Catheter
wedged, position of, 1688
Cattle
hepatic blood flow in, 1405
mountain sickness in, 1721
Causalgia
characterization, 1233
classification, 1233
initiating factors, 1234
major and minor, 1234
neural elements in, 1 235
pain in, 1234
pathogenesis, 1234
reflex elements in, 1235
sympathectomy in, 1235
thalamic dysfunction in, 1235
theories, 1235
treatment, 1234
vasomotor changes, 1234
X-ray studies, 1234
Celiac plexus
innervation of kidney, 1467
Cell volume
dependence on Na, 1 136
Central nervous system
cardiac nerves, communication, 1526
circulatory, autoregulation
neural control, 946
tissue CO2 tension and, 946
intracapillary oxygen pressure, 1021
reactive hyperemia and, 1329
Cerebral blood flow: see Section 1, vol. 3,
ch. 70-71
adrenergic response, 951
4-aminoantipyrine, I2gi
arterial pressure and, 946
autonomic reflex in, 946
autoregulation, 942
carbon dioxide and, 946
cerebrospinal fluid and, 941
cholinergic response, 951
dilution studies, 1 292
discussion of, 1291
epinephrine and, 950
inadequacies, 1 292
nitrous oxide method, 1290-1291
oxygen and, 946
radioactive krypton and, 129 1
requirements for, 1 292
Cerebrospinal fluid pressure
cerebral blood flow and, 941
Cesium
entrance rate into muscle, 1 1 38
ionic radii, 1 137
mobility in electric field, 1 1 38
relative diameter, 1 1 38
vascular muscle tension and, 1 146
Chemical injury
capillary filtration coefficients, 998
mesenteric capillaries and, 996
Chicken
lipid metabolism and atherosclerosis
in, 1 201
lipoprotein pattern in, 1203
sex hormones, atherosclerosis and, 1 204
Cholesterol
absorption, inhibition of, 1186
biosynthesis, inhibition of, 1 186
conversion to bile acids, 1 1 79
degradation, promotion of, 1 1 86
disposal of, 1 1 79
increased tissue removal of, 1 187
INDEX
i/57
mechanism of lowering, i 187
metabolism
anterior pituitary hormones, 1 202
emotional stress and, 1 208
epinephrine and, 1202
physical activity and, 1206
sex hormones and, 1 203
species differences, 1205
strain differences, 1205
stress and, 1206
thyroid function and, 1 199
saturated fats and, 1 183-1 184
synthesis of, 1 1 79-1 180
Cholinesterase
in uterus, estrogen and, 1599
binding, protein osmotic pressure and,
977
distribution in kidney, 1469
entrance rate into muscle, 1 138
in organs and tissues, 1 141
Chloride space
aorta, 1 1 4 1
bladder, 1141
stomach, 1 141
uterus, 1 1 4 1
vascular tissue, 1141
Chlorisondamine
renal blood flow and, 1487
Chlorothiazide
in treatment of hypertension, 1 142
Chlorpromazine
adrenaline response of blood flow and,
'373
Cholestyramine
mechanism of action, 1 186
Chronic pancreatitis
description, 1200
Chylomicrons
characterization, 1 1 76
direct diffusion of, 1 1 76
lymph in, 1044
of pinocytosis, 1044
Circulation
analyses of
algebraic, 1 102
graphical 1 1 02
history 1 1 02
changes on ventilation of fetal lung,
1640
deep drainage channels, 1 531-1532
factor relationships within circuit, 1 101
hemodynamics, 865
hepatic and systemic relationships, 1426
hydrostatic considerations, 1425
in fetus and, 1624-1627, 1638-1645
neonatal, 1638- 1 645
of submucosal plexus, 906
open chest and, 1 1 15
pairing of arterial and venous, 920
preferential channel, go8
pressure gradient in, 965
schema, 1 100
splenic, species differences in, 913
sympathetic vasoconstrictor nerves and,
1362
time
methodology, 1708
values for, 1537, 1708
Circulatory arrest
nervous changes and, 1225
oxygen consumption in muscle during,
'376
oxygen saturation of venous blood, 1376
recovery after, 1225
results of, 1225
vasoconstrictor and vasodilator sub-
stances and, 1 225
Circulatory autoregulation
collaterals and, 947
discussion of, 1 543
feedback loop, 945
general concept, 945
tissue oxygen tension, 945
hydrogen ion concentration and, 946
mechanisms responsible, 944
nervous system and, 946
physical factors, 947
pressure-flow relationships, 1543
see also Blood flow
Circulus vasculosus
description, 1661
Circumflex artery
branches, 1519
distribution of, 1 5 1 9
Cirrhosis
hepatic lymph and, 1051
Cistera chyli
definition, 1036
Clip needle
definition, 840
Coarctation of aorta
blood flow and, 856, 862
cardiac
O2 consumption, 1557
output and, 1557
changes in flow contours and, 858
experimental, 856-858
flow-pressure relationships, 862
left ventricular workload, 1557
murmur and, 856
Cochlea
blood
velocity in vessels of, g 1 8
vessels of, 917
microcirculation in, 915
spiral
ligament, small vessels in, 915
prominence, species differences in
circulation, gi6
Cold
acclimatization to, 1334
as stress, atherosclerosis and, 1208
prolonged exposure to, 1335
vasodilator response to, in skin, 1 333
venoconstrictor responses, 1094
Cold vasodilatation
acclimatization to, 1334
innervation and, 1334
skin circulation and, 1331
total denervation and, 1341
Collagen diseases
classification, 1237
vasoconstrictor mechanisms, 1237
Collagen fibers
aging and, 868
as safety factor, 868, 880
compared to elastic tissue, 868
extensibility of, 805
in arteries, 873
in scleroderma, 1238
Collagen tissue
contribution to pressure-volume dia-
gram, 873
description, 868
elastic
behavior of, 868
visco-elastic behavior of, 868
Collateral circulation
after arterial occlusion, 1 260
arterial versus venous, 1267
arteriovenous fistulas and, 1261
blood flow measurement, 1270
chemical factors and, 1 263
definition, 1257
denervation and, 1 263
development of, 1 260
effects of, 1 268
experimental alteration, 1569
functional changes, 1522
hormones and, 1265
increase, 1572
factors affecting, 1572
ligation of pulmonary vessels and,
1267
lung in, 1 26 1
lysis of vascular tone and, 1 263
measurement, 1 268
neural factors, 1262
outstanding problems, 1271
rate of development, 1 266
regression, 1 267
structure of vessels, 1 268
terminology, 1259
types of vessels, 1 259
venous channels, 1522
Compliance, arterial
definition, 820
discussion of, 842
Conductance
definition, 841
perfusion pressure and, 954
Congenital anomalies
cardiovascular defects
backflow wave and, in aortic valve
closure, 860
1758
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
classification, 1645
effectiveness of heart and, 1 1 04
heart, 1624
signs of, 1645
types of, 1523
Congestive heart failure
cardiac output and, 1 1 lb
fluid retention, 1 1 16
mean systemic pressure, 1 1 16
renal output and, 1 1 16
restriction of fluid and salt in, 1 1 1 7
right atrial pressure and, 1 1 16
sodium retention and, 1 057
venomotor activity in, 1095
venous pressure, 1057, 11 16
venous return and, 11 16
Conjunctiva
arteriovenous anastomosis in, 908
vasomotion in, 909
Conrad, Margaret C.
Resistance (conductance) and capac-
itance phenomena in terminal
vascular beds, 935-960
Constant-pressure flowmeters
discussion of, 1 303
Contour rule
definition, 858
Convective acceleration
definition, 1296
Cooling
gastric blood flow, 1447
Cor pulmonale
acute bronchitis and, 1729
alveolar-capillary block and, 1728
chronic bronchitis and, 1 728-1729
elevated ventricular pressure and, 1555
emphysema and, 1729
kyphoscoliosis and, 1728
origin of, 1728
pulmonary hypertension and, 1728
right heart failure and, 1 729
Coronary arterial occlusion
bypass, 1573
coronary blood flow and, 1539
experimental alteration of collateral
flow, 1569
improved collateral circulation, 1572
mortality, anatomical location and,
!569
possible treatments, 1572
sinus oxygen saturation and, 1 539
Coronary arteries
adjustment to anemia, 1547
anatomical division of, 1522
anomalous communications, 1524
anterior descendents, 1518
arrangement in myocardium, 1520
basic anatomic patterns, 1518
branches of, 1518, 1519, 1520
catheterization, 1530
circumflex branch, 1519
collateral circulation, 1522
congenital anomalies, 1523
construction, ECG changes and, 1569
C02 and lactic acid, 1545
course of, 15 18
distribution, 15 18
dominance pattern, 1520
estimation of, 1535
functional
anatomy, 1517
supply to conducting tissue, 1520
hemoglobin levels and, 1547
in hypoxemia, 1544
insufficiency, collateral circulation
studies and, 1567
left, 15 18
phasic flow, 1534
receptors in, 1550
ligation
local anesthesia and, 1550
vagal stimulation and, 1550
major branches, 1520
normal and anomalous origins, 1525
occlusion of, 1532
experimental, 1565
origin and distribution, 15 1 9
pressure, augmentation and circulatory
depression, 1550
pressure, perfusion, 1535, ■ 541-1543
relaxation of walls, 1544
right, 1519
secondary divisions of, 1520
sensitivity to anoxia, 1539
species
differentiation, 15 19
variation, 1523
terminal branches, 1520
vasomotion, 1552
vasomotor tone, 1535
Coronary atherosclerosis
body build and, 1 206
effectiveness of heart and, 1 1 04
heart disease and, 1 198
personality and, 1206
Coronary blood flow
acetylcholine, 1552
acidosis and, 1545
action of hormones, 1 55 1
active vasomotor changes and, 1535
adenosine, 1545
adenylic acid, 1545
adrenergic response, 951
alkalosis and, 1545
angina, cardiac failure and, 1556
aortic
arch pressure and, 855
insufficiency, 1556
pressure, cardiac output and, 1534
arterial saturation-critical point, 1544
arteriovenous oxygen difference, 1 536
as index of metabolic rate, 1539
myocardial extraction coefficient
and, 1536
asphyxia and. 1544
augmentation, 1548
autonomic control, 1547
autoregulation, 1543
backflow, 1533
basal data, 1537
bicarbonate concentration and, 1546
blood pressure and, 1539, 1548
epinephrine and, 1551
blood viscosity and, 1547
calcium and, 1546
cardiac
cycle and, 1533, 1548
muscle contractility and, 1549
output and, 1546
cholinergic response, 951
circulation times, 1537
collateral circulation, 1567, 1568, 1570
drugs, 1567
establishment of, 1570
extracardiac vessels, 1520, 1522
functional channels, 1522
intracardial vessels, 1522
measurement of, 1271, 1568
natural responses of, 1568
primary location, 1569
prophylactic stimuli and, 1570
coronary artery disease and, 1565- 1573
determinants of, 1533
drugs and, 1561, 1564
dye-dilution studies, 1292, 1530
elevated ventricular pressure and, 1555
epinephrine and, 950
excitement and, 1554-1555
exercise and, 1553-1554
extra- and intravascular resistance and,
extracardiac stimuli and, 1550
extravascular support, 1535
factors affecting, 1554
Fick principle and, 1529
flow values in right ventricle, 1537
heart rate and, 1540, 1554
hemorrhagic shock and, 1535
histamine, 1545
hypercapnia, 1546
hyperthermia and, 1560
hypocapnia, 1546
hypoxia and, 1544
in animals, 1529
in aortic coarctation, 1557
increased work of heart, 1543
inflow curves, 853
influence of systemic pressure, 1542
in heart failure, 1558
in hemorrhagic shock, 1558
in hypothermia, 1559
in male and female, 1 557
in man, 1529
in mitral insufficiency, 1557
in mitral stenosis, 1557
in myocardium, 1533
in resting state, 1538
INDEX
1759
insufficiency, correction for, 1570
intermediate metabolites and, 1545
left
drainage of, 1534
excitement and, 1534
in irreversible hemorrhagic shock,
■534
mild exercise and, 1534
occlusion and, 1534
patterns, 1534
stimulation of cardiac sympathetic
nerves and, 1534
variance with dynamic conditions,
!534
ventricular pressure and, 855
measurement, 855, 1535, 1553
dilution methods, 1291-1293
isotopes, 1529
mean flow, 1529
nitrous oxide method, 1291
rotameter measurement of, 1291
metabolites, 1545
myocardial distribution of, 1 530- 1 533
nicotine, 1563
nucleic acids, 1545
oxygen and, 945
uptake, 1533, 1537
phasic flow, 1528
physical determinants of, 841, 1533-
"536
Pitressin and, 1553
potassium and, 1546
pulmonary artery constriction and,
"542
purine derivatives, 1545
pyrimidine derivatives, 1545
reflex control of, 1 550
resistance changes and, 1 544
respiratory acidosis, 1545- 1546
response to anoxia, 1 539
retrograde flow, 1570
ventricular fibrillation threshold and,
'57'
sodium bicarbonate infusion, 1546
sodium cyanide and, 945
species variation, 15 19-1523
stellate ganglion stimulation and, 1 ",4.'!
stimulation of, 1548- 1549
stimuli and, 1 538-1561
studies on, 1527
systemic pressure and, 1542
thyroid influence upon, 1552
tonic activity of autonomic nervous
system, 1550
vagus and, 1547
vascular
hydraulics of, 853
pressures and, 853
vasoconstrictors, 1 545
vasodilators and, 1545, 1561
venous occlusion, 1542
ventricular
metabolism, 1532
muscle activity and, 1534
tension, 1 541 -1542
work of heart and, 1 549
see also Cardiac blood supply
Coronary blood volume, 1521
Coronary heart disease
atherosclerosis, 1 1 g8
catecholamines and, 1181
factors associated with, 1 188
personality profile and, 1 181
Coronary sinus
blood
as index of metabolic changes, 1532
arteriovenous oxygen difference, 1533
chemical composition, 1532
chronotropic effects of, 1545
inotropic effects of, 1545
vasodilator substances in, 1545
flow, contribution of right coronary
artery, 1532
pressure in, 1542
Coronary vessels
extravascular mechanical compression
and, 1556
percentage of cardiac output, 762
ventricular
distensibility, 763
function and, 762
Corticoids
cholesterol-induced atherogenesis and,
1202
Cortisone
collateral circulation and, 1265
Cough
differential pressure record of, 1689
Cournand catheter
blood flow and, 1287
Cow
pulmonary vessels in, 1674
Cranial mesenteric artery
blood flow through, 1442
flow, feeding and, 1452
Creep
definition, 1084
description of, 803
Critical closing pressure
blood flow and, 1408
definition, 880
muscle vessels and, 1357
role of, 1409
use of term, 941
Curare
vascular smooth muscle and, 1380
Cutaneous blood flow
acetylcholine and, 1346
adrenergic response, 951
atropine and, 1340
cholinergic response, 951
epinephrine and, 950, 1345
factors affecting, 1 220-1 222
heat loss and, 1 333
histamine and, 1346
humoral agents and, 1344
immersion time and, 1331
measurement, 1284, 132 7- 1328
mental arithmetic and, 1339
nervous control of, 1 337
norepinephrine and, 1 345
posture and, 1 340
relation to pressure, 951
sweating and, 1340
sympathectomy and, 1338
temperature and, 1330
thermal conductivity and, 1287
total, 1220, 1328
venous oxygen saturation and, 1 336
Cutaneous blood vessels
action of humoral agents on, 1344- 1346
arrangement of, 1 326-1 327
innervation of, 1338-1341
nervous control of, 1 337— ' 344
physical disturbances and, 1 329-1 337
reactions to injury, 1 336-1 337
reflex control of, 1 341 -1344
transmural pressure and, 1 330
vasomotor nerves, 1 337—1 338
Cutaneous circulation
acetylcholine and, 1334, 1345
adenosine triphosphate and, 1 346
adrenaline and, 1345
anastomoses, 1326
function, 1327
per square centimeter of surface area,
'3*7
arterial gas embolism and, 1337
arteriovenous anastomoses, 1 326
local cooling and, 1333
autoregulation in, 944
baroreceptor stimulation and, 1344
bladder distention and, 1344
body temperature regulation and, 1 341,
1342
bradykinin and, 1346
capillary blood pressure in, 965
carbon dioxide and, 1346
cold vasodilatation, 1331
color of the skin and, 1328
critical closing pressure and, 1330
deep inspiration and, 1343
determination of the adequacy of, 1219
emotion and, 1343
evaluation by sympathetic innervation,
1223
fainting and, 1 343
frostbite and, 1335
general sensory stimuli and, 1343
gravitational effects, 1221
heat loss, 1333, 1334
histamine and, 1334, '345
humoral agents and, 1 344
hypoglycemia and, 1 344
hypothalamic stimulation, 1367
immersion foot and, 1 335
1760
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
local temperature and, 1 330-1 336
noradrenaline and, 1345
osmotic pressure of plasma proteins
in, 965
oxytocin, 1 346
Pitressin, 1346
posture and, 1344
reactions to injury and, 1 336
reflex control of, 1341
serotonin and, 1345
sympathetic denervation and, 1 340
temperatures, 1330
of the skin and, 1 328
transmural pressure and, 1330
trench foot and, 1 335
triple response, 1336
total denervation and, 1341
ultraviolet light and, 1336
white reaction, 1 336
Cutis marmorata
characterization, 1233
Cyanide
glomerular capillary permeability and,
999
Cyclopropane
renal blood flow and, 1489
Cytopempsis
definition, 884, 1066, 1522
Damping
central pulse contour and, 829
electrical, 841
mechanical, 841
Darcy equation
for viscous flow of fluids, 991
DCA
hypertension and, 1 142
K and, 1142-1 143
Na and, 1 1 42
Na and K of aorta in, 1 150
Na and K measurements in, 1 150
intracellular Na and, 1 150
Denervation
collateral circulation and, 1263
cutaneous heat loss and, 1 334
total
cold vasodilatation and, 1341
skin circulation and, 1341
Denison, A. B., Jr.
Pulsatile blood flow in the vascular
system, 839-864
Density flowmeter: see Flowmeter, Dawes
Dermatomyositis: see Collagen diseases
Desoxycorticosterone acetate : see DCA
Dextran
appearance in lymph, 1043
effective osmotic pressures in capillaries,
994
in lymph and plasma, 1 063
molecular sieving of, 1014
molecular weights
permeability coefficients and, 1 05 1
renal nitration and, 1463
Diabetes mellitus
atherosclerosis and, 1187, 1200
capillary microaneurysm and, 1200
lipid metabolism and, 1 200
peripheral vascular disease and, 1 2 1 7
Diaphragm
pressure in lymphatic plexus, 1065
Dibenamine
intestinal blood flow and, 1447
Diet
fat in, blood coagulation and, 1 189
foodstuffs comprising, 1171
serum
cholesterol and, 1182
lipids and, 1 186
Differential
manometer, natural frequency, 1 299
pressure, measurement, 840
transformer, description, 1307
Diffusion
coefficient
molecular radius and, 1001
osmotic and frictional forces and,
1001
Fick's formulation, 1000
free, 1000
general principles, 1000
glucose, 1001
hydrodynamic flow and, 1002
nutrition of vessel walls by, 883
pore
dimensions and, 1002
size and, 1002, 1003
raffinose, 1009
restricted, 1001
factors in, 1001
simultaneous flow and, 1005
through porous membranes, 1001
transcapillary
lipid-insoluble molecules and, 1007-
1008
quantitation, 1009
water, pore size and, 1 004
Digitalis
congestive heart failure and, 1 1 1 7
Diodrast
renal extraction of, 1480
Diuresis
countercurrent mechanism in, 1473
mean regional transit time, water con-
centration and, 1 50 1
renal medullary circulation in, 1499
Diuretics
renal lymph flow and, 1058
Dog
adrenergic control of renal blood flow,
1485
heartbeat at birth, 1641
hepatic blood flow in, 1405
renal
arterial system, 1459
circulation compared to man, 1501
clearance in, 1483
Donnan effect
protein osmotic pressure and, 976
Donnan equilibrium
in biological systems, 1 1 38
Drop recorders
disadvantages of, 1278
primary use of, 1278
"Dry" lungs
definition, 1052
Ductus arteriosus
aortic flow and, 861
closure, 1639
compensation by left ventricle, 861
direction of flow, 1639
flow through, 861
murmur envelope and, 861
Ductus venosus
closure of, 1639
Dye clearance
through skin, acetylcholine and, 953
Dynamic stretches
definition, 867
Ear
arteriovenous anastomosis in, 1 252
Ecolid : see Chlorisondamine
Edema
blood flow and, 969
formation, hypoxemia and, 998
histamine and, 970
interstitial fluid pressure and, 980
kidney function and, 1057
left atrial pressure and, 991
pulmonary
lymphatics and, 1051
lymph flow and, 1052
traumatic shock and, 1060
venous pressure and, 969
EHBF : see Splanchnic blood flow
Elastic arteries
distensibility of, 873
models, 875
pressure-volume diagram, 873
response to stretch, 871
stress-relaxation in, 872
tension-length diagram, 873
tension muscles in, model of, 874
Elastic incompleteness
definition, 867
Elastic material
behavior with stretch, 866
Elastic modulus
applied stress and, 808
as measure of extensibility, 808
initial arterial diameter and, 808
static versus dynamic, 81 1
wall stiffness and, 809
Elastic tissue
aging and, 869
INDEX
I 761
contribution to pressure-volume
diagram, 873
description, 869
elastic behavior of, 868
extensibility of, 805
function in blood vessels, 869
in thoracic aorta, 873
plastic behavior of, 868
recruitment of fibers in, 876
stress-strain relation, 805
structural arrangement, with smooth
muscle, 873
visco-elastic behavior of, 868
Elastic tubes
internal pressure to radius, 939
Elasticity
definition, 866
description of, 866
properties of blood vessels, 866
Electrocardiogram
after lymphic obstruction, 1 055
cardiac pumping and, 777
fetal, 1627
for division of cardiac cycle, 775
Electrochemical gradients
ions and, 1 138
Electrolytes
mean activity coefficients, 1 137
of vascular tissue, 1141
placental exchange of, 1594
relative entrance rates, 1 1 38
renal
artery pressure and, 1 1 48
vein flow and, 1 148
Electromagnetic -induction principle: see
Flowmeters, electromagnetic
Emotional stress
atherosclerosis and, 1 208
cholesterol metabolism and, 1208
gastric blood flow, 1447
intestinal blood flow and, 1450
lipid metabolism and, 1208
skin circulation and, 1343
vasodilators to human muscle and, 1367
Endothelium
active transport through, 883
contraction of, 924
description, 868
diffusion through, 883
permeability, 11 98
see also Blood vessel walls
Energy production
discussion of, 1536
Enzymes
concentration in lymph, 1047
transportation via lymph, 1047
Ephedrine
cardiac hemodynamics and, 1564
cardiac lymph flow and, 1055
splenic blood flow and, 1450
Epinephrine
arterial diameter and, 807
atherosclerosis and, 1202
biphasic action
in man, 1371
in muscle circulation 1373
biphasic transient dilatation and, 1371
blocking of, by chlorpromazine, 1373
blood pressure
coronary flow and, 1551
in bone marrow and, 1 66 1
bone medullary pressure and, 1660
capillary pressure and, 968
cardiac
hemodynamics and, 1564
lymph flow and, 1055
oxygen consumption, 1551
cardiovascular response to, 141 8
cholesterol metabolism and, 1202
coronary
blood flow and, 1551
vasomotion and, 1551-1552
cutaneous blood flow and, 1345
direct and indirect effects, 1370
effect in various vascular beds, 950
fat mobilization and, 1175
filtration coefficients, 999
gastric blood flow and, 1446
intestinal blood flow and, 1447
isolated aortic rings and, 807
joint blood flow and, 1663
K efflux and, 1 156
lipid metabolism and, 1202
mean systemic pressure and, 1 1 1 1
mesenteric blood flow and, 1451
muscle
blood flow and, 948, 1359
venous outflow and, 1371
neuromuscular transmission and 1363,
peripheral resistance and, 1 1 1 1
plasma K, anesthesia and, 11 52
postpartum uterine blood flow and,
1608
rate of injection, venous return and,
1 1 1 1
renal blood flow and, 1485
residual sustained vasodilatation and,
1371
resistance in vascular bed and, 948
segmental resistance and, 951
sensitivity to lymphatic system, 1040
skeletal muscles and, 1355
blood flow, 1370, 1372
circulation and, 1369
circulation, exercise and, 1374
skin
capillaries and, 904
circulation and, 1345
splenic blood flow and, 955, 1450
stretch curves for aorta and, 802
total systemic effect, 1418
vascular
changes due to, g6g, 970-971
volume and, 955
venous return curves and, 1 1 10
Epiphyses
blood supply of, 1653
Ergotamine
intestinal blood flow and, 1447
Ergotoxine
gastric blood flow and, 1446
Erythermalgia : see Erythromelalgia
Erythrocytes
filtration coefficients, 992
Erythrol
gastric blood flow and, 1447
Erythromelalgia
acetylsalicylic acid and, 1242
cutaneous pain fibers and, 1242
description, 1241
diagnosis of, 1 242
mechanism of, 1241
symptoms, 1241
temperature changes and, 1241
vasodilation and, 1241
Esophageal pressure
carbon dioxide and, 1690
exercise and, 1690
Essential hypertension
definition, 875
smooth muscle tension and, 875
Estrogens
as vasodilating agents, 1603
blood vessel walls and, 1204
uterine
acetylcholine and, 1599
blood vessels and, 1600-1601
histamine and, 1599
uterus and, 1 600-1601
vascularity of endometrium and, 1599
Ether
renal blood flow and, 1489
Excitement
coronary blood flow and, 1555
Exercise
adrenaline and, 1374
aortic flow and, 846
pressure-flow curves and, 772-773
arterial transmission line and, 852
atherosclerosis and, 1206
capillary blood volume, 1705
cardiac output, 11 13
cholesterol metabolism and, 1 206
circulation, skeletal muscle and, 1374
coronary
atherosclerosis and, 1207
blood flow and, 1554
ejection pulse and, 845
heart and, 846
hepatic circulation and, 1426
hyperemia, 1355, 1368, 1376, 1380
pH and, 1379
intestinal blood flow and, 1449
lipid metabolism and, 1206
1762
HANDBOOK OK I'HYMULOGY
CIRCULATION II
lymph
How and, 1050
protein from contracted muscle and,
985
mechanism of effect
on cardiac output, 1 1 13
on right atrial pressure, 1 1 1 3
on venous return, 1 1 1 3
muscle
blood flow and, 1355
metabolism and, 1377
venous oxygen saturation and, 1377
oxygen uptake, cardiac output and,
1682
pleural and esophageal pressures and,
1690
postexercise blood flow and, 1355
protein, in interstitial fluid, 985
pulmonary
circulation, exercise and, 1672
vascular resistance and, 1699
vessel oxygen tension and, 1 7 1 9
wedge pressures and, 966
renal blood flow and, 1501
resistance changes in muscle and, 1379
right atrial pressure and, 1 1 13
running, blood flow in legs and, 1 355
skeletal muscle blood flow after sym-
pathectomy and, 1369
splenic blood flow and, 1451
sympathetic
fibers to muscle and, 1368
impulses to muscle and, 1369
venoconstrictor responses, 1094
venous return and. 1078, 11 13, 1125
cardiac output and, 1113
ventricular volumes during, 783
Extensibility modulus
relation to elastic modulus, 808
Extracardiac stimuli
types of, 1550
Extracellular space
blood pressure changes and, 1 153
ratio to total water, 1 141
volume, chloride measurement of, 1 140
Extracellular fluid volume
inulin measurements of, 1 140
Extraendothelial cells
development of, 899
Extraction
definition, 1397, 1398
Eye
arrangement of lymphatics in, 1038
superficial vascular pattern
nasal quadrant, 909
temporal quadrant, 909
Factor P : see Pain factor
Fainting
neurogenic, venodilatation in, T094
skeletal muscle blood flow and, 1367
skin circulation and, 1343
venous return and, 1 123
Familial hypercholesterolemic xanthoma-
tosis
description of, 1 205
Fat
content of human limbs, 1327
digestion and absorption, 1 172
metabolism
in adipose cells, 1 175
insulin and, 1 1 74
liver and, 1 1 78
mobilization
autonomic nervous system, 1 1 75
hormones and, 1174-1175
synthesis in intestinal mucosa, 1 173
Fats
chemical composition, 1171
Fatty acids
binding to albumin, 1 178
classification, 1183
essential, man and, 1 183
food sources, 1 1 83
free
characterization, 1 1 78
of plasma, relation to diet, 1 1 78
long-chain, absorption, 1045
of diet, cholesterol and, 1 182
Feeding
cardiac output and, 1452
cranial mesenteric artery flow and,
'43-
splenic blood flow and, 1450
Felch, \V. C.
Lipid metabolism in relation to physi-
ology and pathology of athero-
sclerosis, 1 1 67- 1 195
Femoral artery
arterial flow pulse in, 849
dimensions and flow, 846
resonance in, 828
Femur
arteries of, 1652
foveal arteries and, 1652
Fetuin
role in fetal osmotic pressure, 975
Fetus
blood
flow, anoxia, hypoxia and, 1635
oxygen saturation, 1634
pressure in, 1629-1631
volume in relation to placenta, 1626
bradycardia in, 1630
brain, oxygen availability in, 1634
cardiac output, 1628
cardiovascular reflexes in, 1 629
circulation
changes on respiration, 1640
course of, 1 624-1 627
factors controlling, 1623
peripheral, 1623
diagram of great veins, 1625
ductus venosus in, 1627
elastin — collagen in blood vessels, 1623
erythrocytes, entrance into maternal
circulation, 1609
factor enhancing oxygen availability,
1634
heart, 1627-1629
adrenaline and, 1630
diagram, 1626
hemoglobin in, 1635
hepatic blood supply, 1627
hypoxia, asphyxia and, 1635-1638
liver, oxygen supply to, 1627
lung development in, 1645
noradrenaline in, 1631
oxygen consumption of, 1634
organs and, 1623
oxygen requirements of, 1634
plethysmograph of, 1633
pressures in umbilical vessels, 1632
pulmonary artery pressure, 1629
regional blood flow, 1625
relationship between oxygen saturation
and consumption, 1 636
resistance to anoxia, 1638
systemic pressure, 1629
vascularization, 1623
vertebral diagram, 1654
weight
fetal blood flow and, 1605
uterine oxygen consumption and,
1606
Fever
hepatic blood flow and, 1429
Fibrillation
intensity of, 1552
phasic inflow curve, 1552
Fibrinogen
concentration in human plasma, 974
molecular weight, 974
Fibroblasts
filtration coefficients, 992
Fick principle
cerebral blood flow, 1 290
general formula, 1291-1292
nitrous oxide method, 1290
Fick's diffusion formulation, 1000
Filler voltage
definition of, 13 17
Filtration
general formulation, 961
injury and, 994
pressure, pulmonary edema and, 1053
Filtration coefficients
adrenocortical hormones and, 999
adsorbed plasma protein and, 994
capillary pressure and, 991
discussion of term, 991
epinephrine and, 999
factors affecting, 988, 998
injured capillaries and, 995
injury and, 994
INDEX
1763
membranes, various, 992
mesenteric capillaries, factors affecting,
998
methods of measurement, 990
of capillaries, g88-iooo
of tissues, 988-1000
pore size and, 1002
range for capillaries, 99-2
species differences, 992
temperature and, 993
tissue asphyxia and, 997
venous pressure and, 989
Filtration constant : see Filtration coeffi-
cients
Filtration rate
glucose, 1 006
hydrostatic pressure in capillaries and,
990
isogravimetric capillary pressure and,
990
raffinose, 1006
sucrose, 1006
urea, 1006
venous pressure and, 978, 989
Fingers
innervation of skin blood vessels, 1 338
Fisher, L. C.
Blood supply to the heart, 151 7-1584
Fishman, A. P.
Dynamics of the pulmonary circula-
tion, 1667-1743
Flat bones
blood supply to, 1655
Flowmeters
applicability, 1296
based on pressure differences, 1 296
Bernoulli's theorem, application in,
1296
bristle, 1305-131 1
Holzlohner and Bergmann, 1306
modified for large arteries, 1309
schematic diagram of bristle and
pendulum, 1305
standard transducer-tube, 1309
Broemser and Reissinger's cannula,
1298
Broemser's differential sphygmograph,
1302
bubble, 1 280, 1 530
calibration of, 1 294
Dawes, 1280
differential pressure, 845, 1300
equations for, 1297
friction device, 1297
pulsatile flow and, 1297
direct recording, 1279
electromagnetic
a-c modification of, 1 313- 1 314,
1530
calibration of, 1314
coronary blood flow and, 1533
d-c procedure, 1 313
diagram, 1 3 1 2
Faraday's induction law in, 131 2
-induction principle, 1311-1318
Kolin, 1313, 1 316
laminar flow in, 131 2
magnet size, 1315
magnetic field strength, 1315
square wave, 840, 1313
transformer emf, 1 3 1 5
zero changes, 13 16
electroturbinometer, 1 304
frequency characteristics, 1 296
Gaddum venous outflow, 1 278
Hensel's needle, 1286
Lauber's venturi cannula, 1 297
Ludwig's stromuhr, 1279
Pavlov's modification, 1279
silicone filled, 1 279
mechano-electric transducer tube, 1308
natural frequency, I2g6
orifice
constriction of, 1 299
principle, 1299
pendulum, 1305-131 1
Pitot meter, 1300
Cybulski's modification of, 1302
variations of, 1 300
Pitot '"torpedo", 1301
Prandtl's tube, 1301
pressure differentials, 1297
principles, 840
properties of, 840
registration of pressure differences, 1 294
requirements, 839
response to velocity profile changes,
I295
Reynolds number and turbulence, 1295
rotameter, 1 303-1304, 1530
of Shipley and Wilson, 1304
Schroeder's differential — pressure, 1 299
sensitivity, 1295
square-wave, 1 3 1 7
circuits, 1318
magnet sleeve units, 1 3 1 8
theory, constriction and operation,
1 294- 1 32 1
thermal conductivity measurement,
1285
transducer-tube
of Laszt and Miiller, 1 3 1 o
of Miiller, 1310
traveling markers, 1320
turbulence, velocity profile in, 1295
types of, 840
ultrasonic, 131 8-1 320
venturimeter of original type, 1 298
Wretlind, for ascending aorta, 1 300
see also Blood flow measurement
Fluid exchange
capillary blood pressure and, 988
mesenteric capillaries and, 996
Fluid flow
electrical analogues of, 842
in elastic tube, 842
in rigid tube, 842
Fluid movements
symbols of factors, 962
Fluid retention
congestive heart failure and, 1 1 16
venous return curve and, 1 1 16
Food
hepatic blood flow and, 1429
ingestion, metabolic consequences of,
1171
splanchnic blood volume and, 1406
Food fat
body fat and, 1 172
Foot
anastomoses per square centimeter of
surface area, 1327
innervation of skin blood vessels in,
'34°
percentage composition by volume,
■327
skin circulation, local temperature and,
■33'
Foramen ovale
closure of, 1639
Forced expiration
definition of, 1713
overshoot, 1 713
Valsalva maneuver. 1 7 1 3
Forearm
blood flow, body heating and, 1339
innervation of skin blood vessels, 1339
percentage composition by volume,
i327
volume changes, 1093
Fourier
analysis, of pulse wave, 817
series, physiologic meaning, 845
Frank's formula
assumptions necessary, 833
Frequency characteristics
pulsatile flow and, 1296
Friedman, Constance L.
Effects of ions on vascular smooth
muscle, 1 1 35-1 166
Friedman, S. M.
Effects of ions on vascular smooth
muscle, 1 1 35-1 166
Frog
capillary blood pressure in, 965
lung structure, 1668
osmotic pressure of plasma proteins, 965
Frostbite
peripheral vascular disease and, 1 2 1 7
skin circulation and, 1335
Fry's method
formula for, 1 302
measurement of, 1302
Functional residual capacity
definition, 782
764
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
in the intact organism, 784
measurement of, 785
Fused quartz rod
for studies of microcirculation, 893
Galactose
for hepatic blood flow, 1 400
Galletti, P. M.
Functional anatomy of cardiac pump-
ing, 759-798
Globulins
ft, osmotic pressure-concentration
curves, 972
ft concentration in human plasma, 974
y>
concentration in human plasma, 974
molecular weight, 974
osmotic pressure-concentration
curves, 972
osmotic pressure due to, 974
Globulin permeability factors
in lymph, 1061
Glomeruli
capillary
permeability, cyanide and, 999
pressure in, 965
filtration, osmotic pressure of plasma
proteins and, 976
juxtaglomerular complex, 14*14
granularity, 1463
renin and, 1464
membranes
electron microscopy, 1462
molecular sieving in, 1017
podocytes and, 1018
pore size, 1462
Glucose
capillary permeability to, 1013
diffusion of, 1001-1002, 1010
diffusion rates, 1012
filtration rate, 1006
permeability of muscle capillaries to,
1013
Gravity
hepatic circulation and, 1424
Green, H. D.
Resistance (conductance) and capac-
itance phenomena in terminal
vascular beds, 935-960
Greenfield, A. D. M.
The circulation through the skin, 1 525
1351
Gregg, D. E.
Blood supply to the heart, 151 7-1 584
Grim, E.
The flow of blood in the mesenteric
vessels, 1439- 1456
Guinea pig
capillary blood pressure in, 965
cardiac glycogen before birth, 1638
osmotic pressure of plasma proteins, 965
protein osmotic pressure in, 965
pulmonary vessels in, 1674
Guyton, A. C.
Venous return, 1 099-1 133
Haddy technique
application of, 1089
Hair growth
cutaneous circulation and, 12 19
Hamster cheek pouch
microcirculation in, 892, 900
vascular
network, 901
pattern, goi
Hand
anastomoses per square centimeter of
surface area, 1327
blood vessels, pressure-volume dia-
grams of, 878
innervation of skin blood vessels, 1338
percentage composition by volume,
1327
plethysmograph for, 1282
skin circulation, local temperature
and, 1 33 1
Hashim, S. A.
Lipid metabolism in relation to physi-
ology and pathology of athero-
sclerosis, 1 1 67-1 195
Heart
anatomical components, 763
as a pump, 761
effectiveness, 1 1 04
left ventricle, 844
pressure-suction, 789
reciprocating, 760
rotary, 760
backflow, mechanisms preventing, 780
catheterization, 1671
cineangiography, 1530
conduction system, 764, 1584
configuration changes, 764
da Vinci, 1669
development of, 1623
diastole and systole
measurement of, 1538
values for, 1538
diastolic
capacity, definition, 783
reserve volume, definition, 782
differences between right and left
cardiac cavities, 789-791
effectiveness, load pumped against and,
1 104
efficiency, drugs and, 1564
electrical-mechanical event relation-
ships, 777
end arteries in, 1522
end -diastolic level, 784
end-systolic volume, 782, 784
epicardium, anatomy, 792
fetal, 1627-1629
noradrenaline, effect of, 1631
sensitivity to adrenaline, 1 630
filling pressure, 1104
function, simultaneous analysis of right
and left, 1 1 19
functional anatomy of, 759, 798, 1517
glycogen, before birth, 1638
hypereffectiveness of factors, 1 104
hypertrophy, effectiveness of heart and,
1 104
hypothetical ejection curve for, 821
interventricular septum, blood supply,
1520
left, compared to right heart, 789
lymphatics, 1527
lymphatic plexuses in, 1054
macroscopic structures, 762-769
mechanical
analogues, 760
efficiency, 761
properties, 761
Myomeric conducting tissue, 151 8
muscle : see Myocardium
orifice relationships, 764
oxygen tension and, 946
oxygen uptake, 1 541 -1543
cardiac arrest and, 1541
fibrillation and, 1541
oxygen utilization, 1538
aortic pressure and, 1543- 1544
myocardial tension and, 1 543-1 544
vagal-arrest and, 1538
papillary muscles, structure of, 780
pericardium : see Pericardium
phase relationships, 775
recompensation of, 1 1 1 7
cardiac output and, 1117
right atrial pressure and, 1 1 1 7
venous return and, 1 1 1 7
response to increased outflow resist-
ance, 844
right, compared to left heart, 789
role of, 760
schematic representation of, 777
size, stroke volume, systolic peak and,
832
skeleton of, 763
structure, according to Leonardo da
Vinci, 1669
sympathetic nerves, stimulation and,
'549
tamponade, cardiac output curves and,
1105
teratogenic factors, 1624
unilateral disturbance, analysis of, 1 1 18
ventricles, fiber arrangement, 1 5 18
vinylite cast of, 1519
volume
blood, per minute, 761
changes at birth, 1641
Heartbeat
augmented rates, 775
INDEX
65
central venous pressure and, 1 1 24
development of, 1627
venous return and, 1 1 24
Heart disease
compensatory mechanisms, 1566
O; uptake, 1566
valvular
effectiveness of heart and, 1 1 04
elevated ventricular pressure, 1554
valvulitis, lymph How and, 1054
vasodepressor drugs in, 1 567
Heart failure
cardiac work, 1558
congestive, analysis of, 1 1 16
coronary stenosis, 1547
discussion of, 1557
experimental, 1 566
extras ascular compression, 1 558
hemodynamic manifestations of, 1557
in severe anemia, 1547
left
cardiac output and, 11 20
atrial pressures and, 1 120
mean pulmonary pressure and, 1 1 20
mean systemic pressure and, 1 1 20
venous return and, 1 1 20
myocardial mechanical efficiency, 1558
proteins of edema fluids in, 982
right
cardiac output and, 1 1 2 1
atrial pressures and, 1 121
mean pulmonary pressure and, 1 12 1
mean systemic pressure and, 1 1 2 1
pulmonary collateral circulation,
l679
venous return and, 1 121
Starling curves in, 1557
Heart murmurs
axial velocity, 858
critical
internal diameter and, 858
velocity, 858
envelope, definition, 858
examples, 859
general rules, 858
moaning, systolic, 859
musical, origin, 859
nonlaminar flow, 856, 858
normal, 856
recording of, 856
sea gull, definition, 859
shape, definition, 858
turbulence, 858
Heart rate
acceleration, energy metabolism and,
'54°
altitude and, 1720
aortic flow during exercise and, 846
augmented, phase relationships, 775
birth and, 1641
coronary blood flow and, 1540
diastolic oscillations and, 827
drugs and, 1564
exercise and, 846
fetal
maternal placental blood flow and,
1636
regulating mechanisms, 1628
slowing with uterine opening, 1637
in exercise and excitement, 1553
pressure augmentation, 827
rapid, limit to cardiac output, 761
stroke coronary flow and, 1554
uniform, pulse pressure wave and, 817
vascular resistance and, 1540
velocity of pulse wave foot and, 818
ventricular filling and, 775
Heart work
aging and, 875
chemical patterns of, 1 536
coenzymes, 1536
drugs and, 1564
enzymes, 1536
glycogen, 1536
hormones, 1536
index, values in left ventricle, 1537
lipids, 1536
protein, 1536
Hematocrit
pulmonary capillary
during exercise, 1707
during rest, 1 707
splanchnic versus arterial, 1 407
Hemodromograph
definition of, 1306
Hemoglobin
in fetus, 1635
levels, effects of, 1547
X, molecular sieving of, 10 17
Hemorrhage
cardiac output and, 1 1 14
right atrial pressure and, 1 1 14
splenic blood flow and, 1 451
venous return and, 1 1 1 4
ventricular outputs and, 1 121
Hensel needle : see Flowmeters
Heparin
anaphylactic shock and, 1059
anaphylaxis release and, 1059
Hepatectomy
BSP removal and, 1398
Hepatic blood flow
acetylcholine and, 1 41 9
adrenergic response, 951
alcohol as measure, 1 400
cholinergic response, 951
cirrhosis and, 1428
clearance and extraction techniques,
1397
cross section and, 1407
dilution techniques. 1402
direct method for, 1 392
distributional pattern, 1409
epinephrine and, 950, 141 8
factors affecting, 1405
Fick principle and, 1396
galactose as measure, 1400
in dog, 1405
in man, 1405
measured by R-E cell activity, 1401
merits of various dyes, 1400
path length, 1409
perfusion studies of, 1 393
single injection technique and, 1401
thermostromuhr and, 1393
transillumination and, 1393
under control conditions, 1404
viscosity, 141 1
volume and distensibility and, 1412
Hepatic circulation
anatomy, 1 388-1 392
autonomic blockade, 1420
bilaterality of portal flow, 1411
blood
flow and volume, 1407-1414
pressure and, 1 395
volume in, 1 394
capillary nets and, 1 389
catheterization of, 1396
collaterals in, 1 268
critical closing pressures in, 1408
denervation and, 1421
diagram, 1406
differentiation of venous and arterial
now, 1399
distributional pattern, 1409
dye studies, precautions, 1399
dysfunction, 1426-1429
hemorrhage, effect on, 1427
hemodynamic
adjustments, 1414-1426
parameters, 1392
hyperemia
factors affecting, 1429
foods ingested and, 1429
importance, 1387
inflow, arteriolar resistance and, 1407
integration, 1 426-1 429
intraosseous venography, 1394
local biochemical determinants, 1421
lymph and, 1391
methodology, 1 392-1 404
neural determinants, 14 14-141 7
neurohumoral determinants, 141 7
normal parameters, 1404- 1407
physical determinants, 1424
portal venipunctures and, 1394
radiopaque injection masses. 1394
reciprocity of venous and arterial in-
flows. 1388
reflex regulation of, 1416
response to epinephrine. 14 18
sinusoids in, 1390
splanchnic interrelationships, 1427
systemic interrelationships, 1426
turbulence in, 1 4 1 1
1766
HANDBOOK OF PHYSIOLOGY
CIKOI'LATION II
veins of, 1 39 1
sphincter in, 1391
venous
catheterization, 1395
outHow measurement, 1 392
viscosity, 1411
volume
arterial inflow tract, 1412
flow distensibility and, 141-'
Hepatic lymph
formation of, 1050
pathways, 1050
protein
concentration, 1050
distribution. 1050
rate of flow, 1051
volume, 1050
in relation to plasma, 1051
Hereditv
atherosclerosis, 1 205- 1 206
lipid metabolism and, 1205
Hexamcthonium chloride
renal blood flow and, 1487
High altitude
cardiac output, 1720
children and, 1 72 1
exercise, 1 720
Histamine
as a lymphagogue, 1059
capillary
permeability and, 1059, 1061
pressure and, 996
cutaneous blood flow and, 1346
edema and, 970
effects of, 1 725
gastric blood flow and, 144b
hepatic circulation and, 1423
in lymph, 1047
intestinal blood flow and, 1448
pancreatic blood flow and, 1450
reactive hyperemia and, 1329, 1375
skin circulation and, 1334, 1345
smooth muscle potential and, 1155
species difference in action, 1059
splenic blood flow and, 1450
total systemic effect, 1423
vascular
changes due to, 969
pressures and interstitial, 970
venoconstrictor responses, 1094
wheal test, 1223
Histodynamics
principles, 1258
Historical development
concepts, pulmonary circulation, [670
contributors, 1670
Barcroft, 1673
Fick, 1670
Henderson, 1673
Lower, 1670
Ludwig, 1670
Hormones
atherosclerosis and, 1 199
collateral circulation and, [265
estrogen and, 1 600-1 bo 2
lipid metabolism and, 1 199
pregnancy and, 1602-1612
Hoyer-Grosser canal
description, 1252
1 [yaluronic acid
»m-1s, diffusion and, 1021
I lyaluronidase
experimental atherosclerosis and, 1202
Hydralazine
postpartum uterine blood flow and,
1608
renal blood flow and, 1487
I [ydraulic impedance
definition of, 843
discussion of, 843-844
Hydraulic reactance
definition, 844
equation for, 844
in pressure-flow diagram, 844
Hydrochloric acid
mean activity coefficient, in given solu-
tion, 1 137
Hydrochlorothiazide
hypertension and, 1 1 42
Hydrodynamic conductivity : see
Filtration coefficients
Hydrodynamic transcapillary flow
of water, pore size and, 1004
Hydrogen ion
entrance rate into muscle, 1 1 38
mobility in electric field, 1138
relative diameter, 1 1 38
Hydrogen ion concentration
capillary filtration coefficients, 998
change, vascular resistance and, n 59
exercise hyperemia and, 1379
filtration coefficients and, 997
membrane permeability and, 1 159
mesenteric, filtration coefficients of, 998
vascular smooth muscle tension and,
1 1581 '59
1 1\ dropenia
urea distribution in kidney and, 1 47 1
Hydrostatic pressure
peripheral venous pressures and, 1127
venous pressure and, 1 1 26
Hypercapnia
central, venoconstrictor responses, 1094
hepatic circulation and. 1422
intestinal blood flow and, 1449
renal blood flow and, 1504
Hyperemia
capillary pressure and, 968
coronary blood flow and, 1554
exercise, 1355, 1368
anoxia and, 1376
blood pressure and, 1378
bradykinin and, 1380
ischemia and, 1376
lactic acid and, 1379
metabolites and, 1378
skeletal muscle circulation, 1376-
1380
prolonged insufficiency of circulation
1, 1330
reactive
anaerobic metabolism, 1540
arteriovenous shunts and, 1540
blood flow and, 1376
blood pressure and, 1539
centra] nervous system and, 1329
definition, 1539
explanation for, 1539
heart rate and, 1539
histamine and, 1329
in skeletal muscle, 1374-1376
in skin, 1329
lactic acid levels, 1540
length of occlusion time and. 1539
magnitude of, 944
mechanism of, 944
oxygen consumption during, 1540
oxygen debt and, 1540
size and duration, 1329
temperature and, 1329
volume of blood flow, 1539
vascular events during, 1375
Hypertension
acute, Na and K. in, 1 151
capillary pressure and, 968
chronic
brachial pulse in, 832
measurement of Na and K in, 1 150
damping of incisura and, 832
electrolytes in vascular tissue and, 1 150
lipoprotein diffusion and, 1046
Na as primary etiological factor, 1 143
relation of Na and K, 1142, 1143
venous pressure and, 1224
r« also Blood pressures
Hyperthermia
diathermy, 1560
see also Fever
Hyperthyroidism
cardiac effects, 1552
cholesterol metabolism and, 1199
Hypertonic solutions
action of, 1717
Hypoglycemia
skin circulation and, 1344
Hypoproteinemia
protein concentration in extravascular
fluids, 985
protein in interstitial fluid and, 985
Hypotension
acute, Na and K in, 1 151
chronic, measurement of Na and K in,
1 150
hemorrhagic, renal blood flow and,
■5°5
Hypothalamic stimulation
1767
effect on skin and muscle circulation,
■367
skeletal muscle blood flow and, 1359
Hypothermia
discussion of, 1559, 1560
effectiveness of heart and, 1 104
Hypothyroidism
atherosclerosis and, 1 187
cholesterol metabolism and, 1 199
Hypoxemia
capillary permeability and, 998
local edema formation and, 998
I [ypoxia
cardiac contractility, 1544
discussion of, 1544
effects on fetus, 1635- 1638
endothelial proliferation and, 1259
fetal blood flow and, 1635
filtration coefficients and, 997
hepatic circulation and, 1422
intestinal blood flow and, 1449
myocardial nucleotides and, 1545
pulmonary
arterial pressure and, 1053, 1718
vascular pressure and, 1 725
vascular resistance and, 1725
vessel oxygen tension, 17 19
regional blood volume and, 1694
renal blood flow and, 1503
vascular effects of, 972
vasoconstriction, 17 19
venoconstrictor responses and, iog4
Hysteresis
decrease with age, 877
in elastic and muscular vessels, 879
in elastic vessels, dependence upon
stretch velocity, 877
in excised vessels, 803
Hysteresis loop
definition, 867
Ileum
vascular pattern, 907
Iliac artery
arterial flow pulse in, 849
circumference, 846
mean peak
flow, 846
velocity, 846
Immersion foot
description, 1335
skin circulation and, 1335
Index finger
anastomoses per square centimeter of
surface area, 1327
Indicator dye
concentration-time curve, 1 69 1 , 1692
dilution curve, 1729
Indocyanine green
removal from blood, 1 397
transfer from blood to bile, 1397
value in clearance studies, 1400
Inert ance
definition, 820
formula of, 841
Inferior metaphysis
blood supply of, 1653
Inflammation
capillary stasis in, 995
Infusions
effects on protein and lymph flow,
1043-1044
Inhibitory permeability factor, IPF, 1061
Injury
capillary
permeability and, 995
pressure and, 996
filtration coefficients and, 989
mechanical, skin circulation and, 1336
Inlet length
definition, 1 296
Insulin
atherosclerosis and, 1 200
fat metabolism and, 1 1 74
lipid metabolism and, 1200
lnterarterial anastomoses
functional anatomy, 151 7
Intercellular cement
increased intracapillary pressures and,
■°53
Intermesenteric nerve
innervation of kidney, 1468
Intermittent claudication
basic requirements, 1217
definition, 1217
Interstitial fluid
bound and free water and, 978
circulation of, 986
distribution, 978-979
movement of dye in, 979
protein
circulation of, 1041
osmotic pressure, 984
proteins of, 982-1007
Interstitial fluid pressure
breaking point, 981
chronic studies of, 981
cutaneous, 980
description, 977
edema and, 980
factors affecting, 979
in kidney, 982
in various body parts, 980
muscle, 980
null point method, 979
subcutaneous, 980
"tissue pressure", 977-982
venous congestion and, 980
Interstitial pressure
histamine and, 970
Interstitial resistance
skin in, 979
Interventricular system
blood supply to, 1520
Intervertebral discs
blood supply to, 1655
Intestinal mesentery
blood flow in, 1443-1444
Intestine
blood flow, 1443-1444
autoregulation of, 944
factors affecting, 1447
through, 1441, 1442, 1443
blood volume in, 1445
circulation of villi, 907
electrolyte levels in, 1 141
fat synthesis in, 1 173
in vasomotor receptors, 1448
microcirculation in, 904
muscularis, blood flow in, 1 443-1 444
segments, blood flow through, 1442
submucosa, blood flow in, 1 443-1 444
villi, circulation in, 907
Intra-abdominal pressure
hepatic circulation and, 1424
Intraluminal pressure
intestinal blood flow and, 1449
Intramural flow
discussion of, 1528
Intrauterine pressure
blood How and, 1600
Inulin
disappearance from plasma, 1008
permeability
of capillary to, 1013
of muscle capillaries to, 1013
Inulin space
extracellular space and, 1 141
in bladder, 1141
in stomach, 1 141
in uterus, 1 141
in vascular tissue, 1 141
Iodine
protein-bound transport in lymph,
1049
Ions
activity, 1 1 3 7
alkali metal, ionic radii, 1 137
blood-tissue exchange of, 1021
definition, 1 136
hydrated size, 1 137
methods of studying effects, 1 139-1 142
mobility, 1 137
properties, 1 1 36
relative diameters, 1 138
mobility in electric field, size and, 1 138
symbols for concentration and activity,
"37
IPF: see Inhibitory Permeability Factor
Iron
transport in lymph, 1048
Ischemia
renal
blood flow and, 1503
hemodynamics and, 1504
1 768
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Ischemic neuritis
description, 12 18
Isogravimetric capillary pressure
definition, 990
Isolation
atherosclerosis and, 1208
Isometric
contraction
coronary blood flow and. [533
relaxation
coronary blood Bow and. 1533
I sopropyl norepinephrine
intestinal blood How and, 1448
Isoproterenol
vascular volume and, 955
Joint blood vessels
blood pressure and, 1662
nervous control of, 1662
Joints
blood
How, 1662
supply, 1 661 -1662
nerve supply, 1 662
nervous control of blood vessels, 1662
Jugular vein
volume distensibility, 1 083
Juxtaglomerular apparatus (JGA): see
Kidney
Katz, L N.
The role of endocrines, stress, and
heredity on atherosclerosis, 1 197—
1 213
Khellin
coronary effects and, 1563
Kidney
arrangement of lymphatics in, 1038
A-V oxygen difference, 1474
blood volume, 1470- 1474
countercurrent mechanism, 1472
distribution of osmotic constituents,
1469
function
albumin infusion and, 1499
edema, lymph flow and, 1057
renal lymph and, 1058
glomerular filtration of dextran, 1463
heat production, 1475
hydrostatic pressures in nephron, 1477
interstitial pressure, 1478
intrarenal pressure (IRP), factors
affecting, 1478
intratubular and peritubular capillar)
pressures, 1476
juxtaglomerular complex, [462
electron micrograph, 14*13
lymph
circulation, [466
production in, 1466
metabolism, 1474- 1475
nephron types, 1458
nerve supply, 1467- 1469
output
congestive heart failure and, 11 16
in myocardial damage, 1 1 1 6
oxygen utilization, 1 474
blood flow and, 1474
in various zones, 147",
pathological changes in ureter ligation,
.058
pressure gradients in various areas, 1475
safety valve mechanism of lymph, 1058
urea distribution during hydropenia,
1471
volume changes with resistance change,
955
zonary temperature gradients, 147b
Kirchoff's current law-
statement of, 849
Knee
blood flow through, 1662
Kramer. K
Methods of measuring blood flow,
'277-1324
Lacteals
definition, 1036
Lactic acid
blood pressures and, 1722
exercise hyperemia and, 1379
heart and, 1545
Lamb
fetal, systemic blood pressure in, 1 b _■ ct
umbilical blood flow in, 1633
Laminar flow-
discussion of, 1295
Poiseuillc's law, 1 295
velocity distribution form, 1 295
Landis, E. M.
Exchange of substances through the
capillary walls, 961-1034
Leg
filtration coefficients for, 992
Leukemia
medullary pressure and, 1661
Leukocytes
filtration coefficients, 992
Leukotaxine
capillary permeability and, 1061
Levophed : see Norepinephrine
Liebow, A. A.
Situations which lead to changes in
vascular patterns, 1251-1276
Limb fluids
protein in, 985
Lipase
transport in lymph, 11147
Lipid-insoluble molecules
permeability of muscle capillaries to,
1013
restricted diffusion of, 1010
transcapillary movement, 1007
transcapillary movement of, 1008- 1009
Lipid metabolism
adrenocortical activity and, 1201
alloxan diabetes and, 1201
anterior pituitary hormones, ! 202
atherosclerosis and, 1167-1190
caloric intake and, 1207
definition, 1 167
diabetes mellitus and, 1200
emotional stress and, 1208
epinephrine and, 1202
heredity and, 1 205
hormones and, 1 1 99
in chickens, 1 201
insulin and, 1200
physical activity and, 1206
sex hormones and, 1203
stress and, 1206
Lipid-soluble molecules
capillary permeability, 1018
Lipids
absorption of, 1045, 1 1 73
circulation of, in women, 1 203
description, 1 167
lymph transport of, 1044
mobilization, hormones and, 1 1 74
rate of movement from blood, 1044
Lipophage
derivation of, 1 1 70
Lipoproteins
a, 974
oa-acid glycoprotein, 974
(3, 974
/3i-lipid poor euglobulin, 974
ft-metal combining, 974
characterization, 1 1 76
classification, 11 76
interconversion, 11 78
molecular surface, 1177
schematic conception of, 1 177
Lithium
entrance rate into muscle, 1 138
ion diameters, 1138
ionic radii. 1 137
mobility in electric field, 11 38
relative diameter, 1 138
Livedo reticularis
basic defect in. 1233
blood vessel changes in, 1233
characterization, 1 232
classification, 1232
picture of, 1232
primary, description, 1233
secondarv, description 1233
Liver
arrangement of lymphatics in, 1038
autoregulation in, 944
blood volume in, 1445
BSP transfer, 1397
cellular arrangement in, 1390
clearance tests, equation for, 1400
fat metabolism and, 1178
in arterioportal anastomoses, 922
INDEX
I769
in capillary permeability, 1050
lymphatic pathways, 1050
reactive hyperemia in, 1422
sinusoids, structure of, 1391
volume of blood in, 1395
see also Hepatic circulation and Hepatic-
blood flow
Lm
definition, 1397
Local acceleration
definition, 1296
Lochner, W.
Methods of measuring blood flow,
1277-1324
Long bones
central venous sinus. 1654
circulation in, 1651
marrow, 1654
cortical arteries, 1652
medullary arteries of, 1651
periosteal circulation of, 1 653
sinusoids, 1654
sources of blood, 1651
vascular supply of metaphyseal region,
1651-1652
venous
sinusoids in marrow, 1654
system, 1654
Low-pass inverse feedback
definition of. 1317
Low-frequency manometers
definition of, 1 303
Low-pressure receptors
sympathetic vasoconstrictor tone in
muscles and, 1 364
Ludwig's principle
variations of, 1279
Lungs
anatomical structure, 1668
arteriolar diameters, 914
A-V pathways in, 914
capillary
permeability in, 1018
pressure in, 966
pressure, osmotic pressure and, 991
structure in, 1676
surface area, 10 18
collateral circulation in, 1261. 1266
Tetralogy of Fallot. 1 262
rate of development, 1266
fetal
blood flow in, 1641
distensibility, 1644
frog, alveolar structure, 1668
human, electron micrograph of, 1677
ideal
application of, 1683
limitations of, 1683
microcirculation in, 913
model of, 1684, 1685
"ideal", 1683
phylogenetic development, 1668
pressure -tiow diagrams before and after
ventilation, 1644
rate of development, 1266
smooth muscle in, 1678
transcapillary exchange, 1 707
"tropism" of collaterals, 1265
Lupus erythematosus: see Collagen disease
Lymph
absorption
by diaphragmatic lymphatics, 1065
extracellular pathway, 1065
intracellular pathway, 1065
of substances by, 1 064
albumin in
infusion and, 1042
relation to plasma albumin, 1063
atherosclerosis and, 1046
B12 in, 1048
blood volume regulation and, 1042-
1044
chylomicrons in, 1044
circulation, magnitude, 987
coagulation, 1048
fibrinogen and, 1048
prothrombin level and, 1048
vitamin K and. 1048
composition, 1041
from right duct, 1052
dextran in, relation to plasma dextran,
1063
disappearance
of phospholipids, 1045
of triglycerides, 1045
drainage, urine concentration, 1058
enzymes and, 1047
exchange of substances with plasma,
1 040- 1 042
extravascular pool, protein and, 1041
factors affecting protein content, 1041
fatty acid absorption and, 1045-1046
histaminase activity of, 1047
histaminolytic activity of, 1047
iron
content of, 1 049
transport in, 1048
lipase transport in, 1047
lipid absorption and, 1045
lipoproteins of, 1044
methods of study, 1035- 1036
molecular sieving of dextran in. 1014
pericapillary filtrate and, 1037
phospholipid
ratio to lipid diet and, 1046
transport, 1046
plasma
exchange and, 1040
protein-equilibration, 1042
pore concept and, 1040
pressure gradient, 1037
protein-bound iodine in, 1049
protein
circulation and, 1041
compared to intravascular protein,
987
concentration of, 985
in various conditions, 985
movement, 1040
regional differences, 982
relation to plasma, 1046
return to general circulation, 1063
shock and, 1 058-1 061
Starling hypothesis and, 1040
steroids and, 1045
tissue fluid and, 1037-1038
transport
effect on bile, 1045
function, 1044-1050
of nucleotides, 1049
of various substances in, 1044- 1050
urea, and albumin content, 1064
volume flow, before and during infu-
sion, 1042
Lymph flow
anaphylactic sensitization in, 1059
artificial respiration and, 1052
atrial pressure and, 1052
burns and, 1061
edema
kidney function and, 1057
production and, 1060
EKG and, 1055
infusions and, 1043
postural proteinuria and, 1057
pulmonary edema and, 1052, 1054
renal
backflow, in renal disease, 1058
capsular, 1056
compared with urine flow, 1056
diuretics and, 1058
factors affecting, 1056
venous pressure and, 1056
traumatic shock and, 1060
valvulitis and. 1054
vasodilatation and, 968
Lymphatic system
anaphylactic shock in, 1058
anatomical
arrangement, 1527
dissection, 1036
as a homeostatic mechanism, 1035
basement membrane in, 1062, 1066
cannulation, 1036
definition, 1036
development and structure, 1036
direct venous connections, 1039
distribution in tissues, 1038
endothelium, specificity, 1037
histological study, 1062
isotopes in studies of, 1036
liver, 1391
ascites and, 1051
lymphatic return and, 1042
main trunks, anatomic arrangement,
1038
i77"
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
methods of study, 1035
of myocardium, 1 054
of uterus, 1608
origin and drainage, 1527
permeability compared to capillaries,
1062
plexuses
density of, 1038
dI heart. T054
pressure in, diaphragm and, 1065
pressure, 1037
increased venous pressure and, 1038
respiration and, 1 038
pulmonary
anastomotic connections, 1052
anatomical arrangement, 1051
edema and, 1051
radiopaque dyes in studies of, 1036
regional, signilicance, 1049
renal, 1466
anatomy. 1055
obstruction, 1058
pyelonephritis and, 1058
"safety valve" mechanism, 1058
response to trauma, 1040
retention of substances, 1062
right lymph duct
composition of lymph, 1052
origin of contents, 1051
sensitivity to epinephrine, 1040
shock in, 1058
shunts in, 1063
terminology, 1036
thoracic duct, 1050
location, 1039
rate of flow, 1 050
structure, 1039
valve system, 1039
transport
function, 1044
lipid, 1044
valves in, 1037
variability of structures, 1039
venous
origin, 1037
pressure and, 1037
Lymphatic permeabilitv
anatomical structures important in,
1064
colloidal particles, factors involved,
1 06 1 - 1 066
concepts, 1066
intraluminal pressure and, 1064
"leaky pump" concept, 1066
potential physical openings, 1 065
size limitation, 1064
stigmata, stomata and, 1065
test substances for, 1064
theories of, 1065
"toxic" effect of urea, 1064
Lymphatic vessels
capillaries
in myocardium, 1054
relationship to small blood vessels,
'°37
contractility of, 1039- 1040
development and structure of, 1036-
1037
distribution of, 1038- 1039
in heart valves, 1054
permeability pattern in, 1062
protein leakage, 1 062-1 063
shock and, 1050- 1058, 1 058-1 061
Lymphedema
lymph and interstitial protein in, 986
Macrocanalicular system
definition, 1653
Maculadensa: see Glomeruli, juxta-
glomerular complex
Magnesium
arteriolar size and, 948
entrance rate into muscle, 1 138
ion diameters, 1 1 38
mobility in electric field, 1 138
relative diameter, 1 1 38
smooth muscle tension and, 1 158
vascular
resistance, 1 158
smooth muscle tension and, 1 1 57 —
1 158
vasodilatation and, 1158
Magnet-coil current
heat production and, 13 17
Malnutrition
proteins of edema fluids in, 982
Mammals
protein osmotic pressure in, 975
Man
aortic resonance in, 826
arteriovenous anastomosis in ear, 1 252
automaticity in blood vessels, 1356
axillary artery, tension muscles, 869
bile acid formation in, 1 180
biphasic action of adrenaline, 1371
birth
cardiac index at, 1642
cardiac output at, 1641
changes in systolic pressure at, 1642
heartbeat at, 1641
pulmonary vascular resistance at,
1 64 1
blood
lipids and atherosclerosis, 1187
supply to interventricular septum,
1520
vessels in cochlea, 915
blood flow
in foot, 1340
in forearm, 1339
in hand, 1338
bone blood flow in, 1657
brachial pulse in, 831
capillaries of labyrinth, 915
capillary
blood pressure in, 965
networks of the cochlea, 916
cardiac glycogen before birth, 1638
circulation
skeletal muscle, 1363
through skin, 1325
electron micrograph of lung, 1677
essential fatty acids and, 1 1 83
extension-release curves of aorta, 877
filtration coefficients for tissues, 992
flow pulses in, 859-862
hepatic blood flow in, 1405
heredity and atherosclerosis, 1205
interstitial fluid pressure in, 980
lipoproteins in, 1 177
medullary pressure in, 1660
mountain sickness in, 1721
nephron types in kidneys, 1458
osmotic pressure of plasma proteins
in. 965
patterns of circulating lipids, 1203
placental histology, 1621
plasma components, 974
pressure-volume diagram in aorta, 875
protein
content of plasma, 974
osmotic pressure in, 965, 973
pulmonary
artery in, 1674
blood flow, 1697
blood pressures, 1697
vascular resistance, 1697
vein in. 1674
renal blood How
in infants, 1644
trauma and, 1506
renal circulation
adrenergic control, 1 486
compared to dog, 1501
renal clearance in, 1483
skeletal muscle innervation, 1 364, 1 366,
>367
stimulation of arterial baroreceptors
in, 1363
studies on bulbar conjunctiva, 894
sympathetic vasoconstrictor nerves in,
■359
tissue composition of veins, 881
umbilical arterial flow in, 1633
\ in\ lite cast of heart, 1519
Mandible
arterial supply to, 1655
Mannitol
renal
artery pressure and, 1 1 48
lymph flow and, 1058
vein flow and, 1 148
Marsupials
uterus in, 1585
Mayerson, H. S.
The physiologic importance of lymph,
io35-'°73
INDEX
1771
Mechanical compression atelectasis
angiography, 1717
experimental, 1 7 1 7
venous admixture, 1 7 1 7
Mechanical pumps
classification, 760
Membrane
manometers, definition of, 1 303
permeability, ions and, 1 137
potentials, ions and, 1 1 38
Menstruation
mechanisms of, 1 598
role of arteriovenous shunts in, 1598
scotomata and, 1599
Mephentermine
cardiac hemodynamics and, 1564
Mercury-
renal lymph flow and, 1058
Mesenteric artery
pressure-diameter diagrams of, 878
Mesenteric blood flow
adrenergic response, 95 1
arteriovenous anastomoses and, 1444
blood volume, 1445- 1446
cholinergic response, 951
epinephrine and, 950
factors influencing, 1 446-1 452
function and, 1452- 1453
magnitude, 1 439-1 441
organ function and, 1452
partition, 1 441 -1445
to major organs, 1441
to vessels of different sizes, 1 444
Mesenteric circulation
capillary blood pressure in, 965
factors affecting, 1451
function, preferential channels and, 919
organs included, 1439
osmotic pressure of plasma proteins in,
965
pressure gradient in, 964
pressure-volume relations, 1085
Mesentery
camera-lucida outline of capillaries, 919
microcirculation in, 900
veins, plastic cast, 1077
Metabolites
exercise, hyperemia and, 1378
Metabolism
disease and, 1537
of tissues, venous return and, 1 1 25
Metaraminol
cardiac hemodynamics and, 1564
Metarteriole
definition, 892
description, 919
vasomotion in, 925
Methacholine
splenic blood flow and, 1451
Methodology
acute pulmonary engorgement, 1693
cardiopneumogram, 1694
compliance, 1693
mechanics of breathing, 1 693
mercury injection, 1036
pulmonary
arterial hypertension, 1693
distensibility, 1693
venous hypertension, 1693
radioactive tracers, 1694
teeter board, 1694
Methoxamine
cardiac hemodynamics and, 1 564
Microcirculation
alternate routes in, 895
arteriovenous anastomoses, 899
bat wing, 895
bulbar conjunctiva, 908
chambers, 893
cochlea, 915
definition, 891
description, 1 263
hamster cheek pouch, goo
intestine, 904
lung, 9 1 3
mesentery, 900
myocardium, 903
preferential channel, 908
rabbit ear, 898
removable-top chamber for, 893
skeletal muscle, 900
skin, 904
spiral ligament, 916
spleen, 910
stomach, 904
stria vascularis, 916
thoroughfare channel, 918
Mitral insufficiency
discussion of, 1557
experimental mitral regurgitation, 1557
ventricular function curves, 1557
Mitral stenosis
cardiac index, 1557
cardiac work index, 1557
discussion of, 1 556
lymph flow and, 1055
pulmonary
arterial pressure, exercise and, 1695
hypertension, 1727
pressure-volume diagram, 1693
Mitral valve
closure interval, 7go
opening interval, 79°
structure, 780
Moat chamber
for studies of microcirculation, 893
Model
arterial, windkessel, 847
pressure applied around tube, 1708
tension muscles and elastic tissue, 874
Molecular sieving
diffusion, restricted and, 1005
rate of filtration and, 1005
regional differences, 1013
theory of, 1005
Molecules, small
blood-tissue exchange, 102 1
Mollusk smooth muscle
catch mechanism in, 871
Monkey
blood pressure, age and, 1643
cardiac glycogen before birth, 1638
carotid sinus response in, 1644
heart beat at birth, 1641
Monoamine oxidase
in Raynaud's syndrome, 1229
of digital arteries, 1229
Monotremes
uterus in, 1585
Mountain sickness
syndrome of, 1721
Mouse
hepatic blood flow in, 1405
interstitial fluid pressure, 980
Muscle pump : see Pumps
Muscular activity : see Exercise
Myocardial damage
arterial pressure and, 1 1 16
cardiac output, 1 1 1 5
infarction
after coronary ligation, 1550
cardiac output and, 1 1 15
effectiveness of heart and, 1 104
right atrial pressure and, 1 1 15
venous return and, 1 1 1 5
ischemia, viability and, 1569
renal output and, 1 1 16
right atrial pressure, 1 1 1 5
vasoconstriction and, 1 1 16
venous return, 1 1 15
Myocardium
aerobic metabolism, 1538
anoxia
discussion of, 1541
effects of, 1 54 1
ventricular distensibility, 1541
ventricular fibrillation, 1541
arterioles
anatomical structure, 152 1
discussion of, 1 52 1
atrial, 768-769
architecture of, 768
classification, 768
blood flow
distribution of, 1530
methods for determination 1 530
venous circuit, 1531
capillaries, functional activity, 1521
capillary
density, 1521
distribution, 903
exchange of metabolites, 1521
cardiac failure and, 1537
changes of, 1536
chemical patterns of, 1536
1772
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
chordae tendineae, function, 764
coenzymes, 1536
composition of, 762
contractile force, CO2 and, 1546
deep
bulbospiral muscle, arrangement, 767
drainage channels, 1 53 1
sinospiral muscle, arrangement, 767
determinants of normal metabolism,
1 536- 1 537
diabetes, 1537
energy production, 1536
enzymes, 1536
glycogen, 1536
hormones, 1536
ischemia and normal function, 1541
lipids, 1536
lymphatic supply, 1054
mechanical efficiency of, 1 538
metabolism and, 1536
disease, 1537
epinephrine and, 1541
norepinephrine and, 1541
metarteriole, actions of, 1 52 1
microcirculation in, 903
muscle
arrangement in ventricles, 768-769
fascicles, 1518
myocardial respiratory quotient, 1536
oxygen debt, 1538
oxygen uptake
coronary dilatation, 1567
during cardiac arrest, 1541
during fibrillation, 1541
in working heart, 1541
precapillary sphincter, actions ot, 1521
protein, 1536
respiratory quotient, 1536
schematic arrangement of muscle fibers,
768
sinusoids in, functional anatomy, 1517
spiral muscle arrangement, 767
superficial bulbospiral muscle, arrange-
ment, 767
superficial muscle layers, 765
ventricular contraction and, 765
trabeculae, functional anatomy, 1517
veins
density of, 1522
distribution of, 1522
ventricular, 765-768, 1518
architecture of, 765
arrangement, 768-769
classification of muscles, 766
evolution of, 766
fascicles, 15 18
in left ventricle, 1538
vessels, nerve supply to, 1 52 1
vortex spirals, 1518
Myocarditis
effectiveness of heart and, 1 104
Myoglobin
capillary permeability to, 10 13
molecular sieving of, 1017
permeability of muscle capillaries to,
1013
Natriuretic drugs
hypertension and, 1 142
Negative pressure breathing
intrapulmonary, 1 7 1 2
pleural, 171 1
Neomycin
cholesterol levels and, 1 187
Nephrosis
proteins of edema fluids in, 982
Neuromuscular transmission
sympathetic
vasoconstrictors to skeletal muscle,
■363
vasodilators in skeletal muscle, 1 366
Newborn
respiration, course of blood streams
and, 1639
Nicotine
catecholamines and, 1563
Nitrate
entrance rate into muscle, 1 138
Nitrites
venodilator responses, 1094
Nitroglycerin
action of, 1 567
coronary circulation and, 1562
Nitrous oxide
saturation curves, 1291
Nodulus arantii
description, 781
Nonlaminar flow
murmurs and, 858-859
Norepinephrine
arterial diameter and, 807
as neuromuscular transmitter, 1 363
blood pressure rise and, 1 155
bone medullary pressure and, 1660
cardiac oxygen consumption, 1552
carotid sinus stimulation by, 874
coronary
blood flow and, 1552
vasomotion, 1 552
cutaneous blood flow and, 1345
distribution in pulmonary tree, 1680
effects of, 1725
fat mobilization and, 1 1 75
intestinal blood flow and, 1447
mesenteric blood flow and, 145 1
plasma
K and, 1 1 52
Na and, 1 152, 1 153
potentiation by K, 1 146
pulmonary vascular pressure and resist-
ance and, 1725
regional blood volume and, 1694
resistance in vascular bed and, 948
skeletal muscles and, 1355
circulation and, 1369
skin circulation and, 1345
splanchnic circulation and, 141 7
splenic blood flow and, 1 450
total systemic effect, 141 9
vascular
effects of, 969, 970-971
volume and, 955
venous
constriction due to, 1086
distensibility and, iog3
Normotension
damping of incisura and, 832
Nucleotides
transport in lymph, 1049
Nutrient exchange
between maternal and fetal circulation,
1609
Oliguria
renal blood flow and, 1481
Open chest
cardiac output and, 1 1 15
right atrial pressure and, 1 1 1 5
venous return and, 1 1 1 5
Orifice flowmeter
principle and discussion of, 1298
use in veins, 1 299
Orthostatic hypotension
renal blood flow and, 1502
Osler-Weber-Rendu disease
description, 1256
Osmometer
description of, 973
Osmotic equilibrium
description, 1 136
Osmotic flow
osmotic pressure and, 1006
Osmotic pressure
calculated experimentally and on Don-
nan theory, 977
capillaries, dextran and, 994
concentration curves for plasma pro-
teins, 972
dependence on Na, 1 1 36
due to urethan and urea, 1018
factors determining, 1 136
osmotic flow and, 1006
physicochemical aspects of protein, 976
plasma proteins, glomerular filtration
and, 976
protein
anion binding and, 977
concentration and, 975
Donnan effect, 976
factors affecting, 973
fetal, 975
filtration rate and, 989
in various species, 965
measurement, 972
of human plasma, 973
INDEX
'773
of interstitial fluid, 984
predicted and actual, 977
reduced, adaptation due to, 974
species differences, 975
relation to temperature and concentra-
tion, 1 136
renal lymph and, 1057
Osmotic reflection coefficient
definition, 1006
factors influencing, 1007
Ovarian arteries
origin of, 1587
Oxygen
arterial saturation, blood How, 11 25
availability, fetal, 1635
blood-tissue transport of, 1019
cerebral blood flow and, 946
consumption
measurement of, 1540
muscle, blood flow and, 1377
oxygen saturation and, 1636
relation of fetal and placental, 1634
coronary blood flow and, 945
debt, repayment of, 1540
exchange, maternal and fetal blood,
161 1
filtration coefficients and, 997
local utilization
cardiac output and, 1 1 26
venous return and, 1 1 26
mesenteric, filtration coefficients of, 998
of skin blood, as measure of blood flow,
'327
pressure, intracapillary in brain, 1020
pulmonary capillary permeability and,
1018
steady-state radial diffusion of, in tis-
sues, 1020
tension, tissue, arterial occlusion and,
126 1
uptake
cardiac output and, 1682, 1 7 1 4
values for, 1720
Oxytocic drugs
postpartum uterine blood flow and,
1608
Oxytocin
skin circulation and, 1346
Pacinian corpuscles
pressure mediation and, 141 6
structure and function, 141 5
Paddle flowmeter
discussion of, 1 307-1 308
Pain
origin of, 1565
visceral, neural mechanism, 1415
Pain factor
description, 1 2 1 8
Palm
anastomoses per square centimeter of
surface area, 1327
Pancreatic secretion
blood flow and, 1452
Panniculus carnosus
arteriovenous anastomoses in, 902
Pappenhcimer, J . R.
Exchange of substances through the
capillary walls, 961-1034
Para-amino hippuric acid
renal extraction of, 1 480
Paramyosin
as plastic element, 871
Parasympathetic nervous system
inhibition, effectiveness of heart and,
1 104
innervation
of splanchnic viscera, 14 15
of uterus, 1 602
intestinal blood flow and, 1448
lack of control of venous tone, 1077
stimulation, effectiveness of heart and,
1 104
vasodilator nerves cholinergic to mus-
cle, 1368
Pathology
progressive valvular fibrosis, 1054
Pendulum flowmeter
principle and discussion of, 1306
Perfusion pressure
arterial flow and, 936
blood
flow and, 954
volume and, 954
conductance and, 954
mean transit time and, 954
Perfusion pumps
blood cell injury and, 1370
Periarteritis nodosa: see Collagen diseases
Pericardium
anatomy of, 7g2
atrial filling and, 793
filling by heart, 793
function of, 791-794
histology, 792
plastic behavior of, 7g2
pressure, cardiac cycle and, 793
pressure-volume curve, 792
reserve power of heart and, 794
Perimysium
capillary bed of, 902
Peripheral circulation
carbon dioxide and, 11 25
laboratory procedures for examining,
1224
Peripheral vascular diseases
circulatory arrest and, 1 224-1 225
classification of, 1 225-1 226
clinical approach, 12 16-1224
definition, 12 15
malfunction in general, 12 16
mechanisms in, 1226-1242
patient's history, 1 2 1 7
symptoms of arterial disease, 12 1 7
trophic changes in, 12 18
vasoconstrictor, 1226-1 241
vasodilative, 1241-1242
Peripheral venous system
anatomical considerations, 1075-1079
assessment of venomotor activity, 1 087-
1094
hemodynamic relations, 1081
physiological characteristics of, 1079-
1087
venomotor responses, 1094-1095
Peritoneal cavity
absorption of particles from, 1065
Personality profile
coronary atherosclerosis and, 1206
heart disease and, 1 181
Perthes test
description, 1224
Phase-sensitive demodulation
definition, 13 15
Phenoxybenzamine
muscle blood flow and, 948
splenic blood flow and, 1451
Phenylephrine
bone medullary pressure and, 1660
cardiac hemodynamics and, 1564
pressure-volume relationships and. 1085
Phlebothrombosis: see Thrombophlebitis
Physical activity : see Exercise
Pig
pulmonary vessels in, 1674
Phillips, J.
Peripheral vascular diseases — diseases
other than atherosclerosis, 1 2 1 5-
1249
Phosphate
entrance rate into muscle, 1 1 38
renal lymph flow and, 1058
Phospholipids
dietary sources, 1 1 72
transport, total lipid transport and,
1046
Pick, R.
The role of endocrines, stress, and he-
redity on atherosclerosis, 1 197—
1213
Pigeon
transitional capillaries in bone, 1654
Pilocarpine
gastric blood flow and, 1447
Pinocytosis
definition of, 1522
of chylomicrons, 1044
Pituitary hormones
blood volume in relation to fetal, 1626
bone medullary pressure and, 1660
circulation, 1592
arterial supply, 1597, 161 1
sheep, 1622
venous drainage, 1598
comparative anatomy of, 1 591 -1592
cotyledons, vascular arrangements in,
1597. l6l!
'774
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
countercurrent flow in, 1622
fat mobilization and, 1 174
fetal, 1 620- 1 623
capillary blood pressure, 1609
pressure flow curves, 1632
function, 1622
hemochorial, deficiency, 1606
India ink injection of, 1608, 1610
intervillous space pressure, 1609
maternal blood and, 1610
oxygen consumption of, 1622, 1634
Pitressin
coronary circulation and, 1562
effects of, 1553
effects upon Na and K, 1553
gastric blood flow and, 1447
intestinal blood flow and, 1448
mesenteric blood flow and, 1451
plasma Na concentration and, 1 153
skin circulation and, 1346
pituitrin, splenic blood flow and, 1450
pressure, amniotic fluid pressure and,
placental 1608
radiopaque injections of, 1 609
structure, 1592, 1594
arterial and venous openings in, 1610
changes in pregnancy, 1593
exchange of substances and, 1594
unitary, 1593
syndesmochorial, deficiency, 1606
types, 1 592- 1 594
histological, 1620
uterine work necessary, 1606
Placental blood flow
fetal, 1 63 1 -1 635
resistance in, 163 1
maternal, effects of reduction, 1636
uterine contractions and, 1600
Placental exchange
factors affecting, 1622
mechanisms of, 1622
sodium, in various species, 1592, 1593
structure and, 1594
Plasma
albumin, in relation to lymph albumin,
1063
dextran, in relation to lymph dextran,
1063
exchange, lymph and, 1040
human, components of, 974
potassium, epinephrine, anesthesia and,
1 152
proteins
capillary pressure and, 963
glomerular filtration and, 976
in man, 974
osmotic pressure of, 962-977
relation to lymph, 1046
urea, albumin content and, 1064
Plasma skimming, 898
Plasmin
capillary permeability and, 1061
Plastic material
behavior with stretch, 866
Plasticity
definition, 867
description of, 803, 866, 867
properties of blood vessels, 866
Platelets
capillary permeability and, 995
Plethysmograph pulses
factors affecting, 957
Plethysmography
modifications of, 1281
principle of, 1281
Pleural pressure
carbon dioxide and, 1690
exercise and, 1690
measurement of, 1 688-1 689
Podocytes
molecular sieving and, 1017-1018
Poikilotherms
protein osmotic pressure in, 975
Poiseuille's law
blood flow to organs, 865
modified for collapsible tubes, 1080
Poiseuille's equation
origin of, 963
Pore concept
lymph formation and, 1040
see also Capillaries; Pores
Portal vein
anatomy and distribution, 1389
flow
arterial pressure and, 1440
in various species, 1440
through blood flow, 1439
Positive pressure breathing
cardiac output and, 171 1
circulatory collapse and, 171 1
intermittent positive pressure breathing,
171 1
intrathoracic pressure, 171 1
pulmonary blood flow and, 17 10
pulmonary vascular resistance and,
1 7 10-17 1 1
systemic hypotensive states, 1 7 1 1
Posture
cutaneous blood flow and, 1340
renal blood flow and, 1502
skin circulation and, 1344
see also Body position
Potassium
arteriolar size and, 948
blood pressure and, 1 155
entrance rate into muscle, 1 138
exercise hyperemia and, 1380
extracellular, blood pressure and, 11 54
hypertension and, 1142-1143
in various tissues, 1 141
ionic radii, 1 137
mobility in electric field, 1 138
peripheral vascular resistance and, 1 149
potentiation of norepinephrine, 1 1 46
radioactive
blood to tissue permeability, 1023
capillary permeability to, 1023
flow distribution pattern and, 1402
relative diameter, 1 1 38
spontaneous muscle activity and, 1 156
vascular smooth muscle tension and,
1142-1157
vasoconstrictor effect, 1 149
vasodilation and, 1380
Potassium hydroxide
mean activity coefficients, 1 137
Potassium transport
efHux
epinephrine and, 11 56
membrane potentials, 1156
smooth muscle tension and, 1 156
Precapillary sphincter
definition, 892
function, 900
Pregnancy
arterial supply to placenta and, 1597
circulating lipids in, 1203
ovarian vein and, 1587
placental structure and, 1593
uterine
arterial injection during, 1604
blood flow and, 1 593-1 594, 1603,
1609
blood flow and volume during, 1604
circulation and, 1602
vasculature and, 1594, 1603
venous drainage and, 1588, 1598
venous pressure, body position and,
1601
Pressoreceptors
anatomy of, 874
blood pressure and, 875
relation to surrounding tissue, 874
structure and location, 874
Pressure-volume curves
aorta compared to muscular arteries,
879
arteries
elastic, 879
muscular, 879
contractile state of smooth muscle and,
873
drugs and, 1085
in mesenteric artery, 878
in vivo, in hand, 878
pulmonary
normal, 1693
mitral stenosis and, 1693
veins, 882
Primates
uterus in, 1585
Priscoline : sec Tolazoline
Procaine
intestinal blood flow and, 1448-1449
INDEX
1775
Progressive systemic sclerosis: see Sclero-
derma
Pro-PF/IPF system
definition, 1061
Prostigmin
uterine bleeding and, 1599
Protein
absorption
and metabolism of, 1 1 7 1
from interstitial fluid, 983
circulation of, 986
circulation in lymph, 982, 987
concentration, osmotic pressure and,
975
estimates of molecular weight, 973
exchangeable mass of interstitial fluid,
984
hepatic blood flow and, 1429
in extracapillary fluids, 982
in interstitial fluid, 984
in lymph, 985
leakage, infusions and, 1043 -1044
osmotic pressure, capillary blood pres-
sure and, 991
passage through capillary wall, 986
rate of
lymph circulation, 987
movement from blood, 1 044-1 045
Proteinuria
postural, lymph flow and, 1057
Psychic stimulation
venoconstrictor responses, 1094
Pulmonary arterial pressure
acute
left heart failure and, 1 120
right heart failure and, 1 121
age and, 1686
altitude and, 1720
blood flow and, 1696
exercise and, 1695
flow-pressure relationships
in exercise and, 1701
transmural pressures, 1701
hydrostatic reference level, 1685
hypertension, 1727
hypoxia and, 1053, 1 71 8, 1725
in disease, 17 10
in man, 1697
in neonatal period, 1641
lactic acid and, 1722
luminal pressure, 1688
primary pulmonary hypertension and,
1724
mean values, 1686
norepinephrine, 1725
recording, 1685
transmural pressures, 1688
values for low oxygen saturation, 1 7 1 8
Valsalva maneuver and, 1 7 12
vascular resistance and, 1699
Pulmonary artery
acetylcholine and, 1 700
creep in, 805
occlusion, pressure changes, 1689
pressure pulse compared to aorta, 1686
stretch curve of, 805
wall, inertance and resistance in, 843
Pulmonary bed
capillary pulsatile flow in, 856
Pulmonary blood flow
acetylcholine and, 1 726
acute acidosis and, 1722
acute hypercapnia, 17 18, 1721
alveolar-capillary
gas exchange, 167 1
interface, 1676
alveolar gas analysis, 1683
analytical determinations, 1670
anesthetics and, 1673
appearance time, 1 708
arterial pressure and, 1696
arteriovenous
oxygen difference, 1714, 1 7 1 5
pressure gradient, 1687
shunts, indicator dilution curves in,
1729
as a filter, 1668
as a reservoir, 1668
blood vessels, 1674
bradycardia and, 17 16
capillary
blood volume, 1705
Fick method, 1703
flow rate, 1702, 1703, 1706
gas exchange in, 1706
pneumocardiographic recording,
1704
plethysmography recording, 1704
pressure, 1702
Stewart-Hamilton methods, 1703
changes at birth, 1641
chronic hypoxia and, 1720
circulation time, 1708
comparative physiology of, 167 1
critical closing pressure, 1701
critical opening pressures, 1701
distribution pattern in lung, 1 681 -1682
drugs and, 1724, 1725
during expiration, 1709
during inspiration, 1709
epinephrine and, 1725
excitement and, 1681
exercise and, i6gg, 1 714
experimental studies, 1673
Fick principle, 17 14
flow-pressure curves, 1 700
forced expiration and, 1 7 1 3
functional anatomy, 1673
gravity effects, 1683
heart rate and, 1716
high altitude and, 1720
high Oj content, 1721
hypertension and, 1 727
in individual lungs, 1701
in man, 1697
in spontaneous breathing, 1709
inspired air distribution, 1683
intrapulmonary
baroreceptors, 1683
inflation, 1710
in vessels, excluding organs, 1293
intrathoracic pressure, 1709
hypoxia and, 1699, 1718, 1720
kinetic energy, 1702
left ventricle and, 1 293
luminal pressure in, 1709, 1710
lung inflation and, 17 10
measurement of, 1714
negative (pleural) pressure inflation
and, 1 7 1 o
non-Newtonian fluid, 1697
normal values, 1681
oxygen uptake, 17 14
plethysmography, 1704
positive pressure inflation, 17 10
potential energy, 1702
pressure-volume characteristics, 1696
prolonged expiration and, 17 12
pulmonary
artery occlusion, 1 7 1 3
efficiency, 1683
pulse-wave velocity, 1 707
reflex activity and, 1723
respiration, 1 709
respiratory
acidosis, 1722
gases, 1 718
gas exchange, 17 14
right ventricular ejection pulse and, 855
species difference, 1673
Stewart-Hamilton method, 1703, 1714
thoracic aorta, 1293
"tight" mitral stenosis, 1716
transmural pressure, 1699, 1701, 1709,
1710
types of reflexes, 1723
values for, 1681-1685, 1714
valves of, 1 70 1
valvular insufficiency, 1 73 1
vascular pressures, 17 15
vascular resistance, 1697, 17 16
venous admixture, 1 680
wedge pressures, 1687
Pulmonary blood vessels
arteries
as end arteries, 1675
distribution, 1675
pulse control, 1686
pulse -wave velocity, 1697
arteriovenous fistula, 1730
capacity, 1696
capillaries
distensibility of, 1705
resistance in, 1705
smooth muscle, 1675
structure of, 1675
776
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
circulation in, 1675
capillary bed
maximum diffusing capacity, 1705
measurement of, 1705
size of, 1 705
transmural pressure, 1 705
delayed compliance, 1696
distensibility, 1696
hysteresis, 1696
pulmonary "resistance" vessels, 1696
stress relaxation, 1696
veins, 1675
structure of, 1675
venules, structure of, 1675
Pulmonary blood volume
changes in, 1693
estimation
Bradley equilibration curves, 1693
Newman exponential downslope
equation, 1692
in exercise, 171 5
lung volumes, 1693
measurement of, 1691, 1692
methodology, 1693
normal values, 1694
partition of, 1695
pulmonary arterial pressure and, 1715-
1716
Stewart-Hamilton method, 1691
variations in, 1695
Pulmonary circulation
acetylcholine and, 1700, 1726
arterial occlusion, 171 3-1 714
ipsilateral oxygen uptake and, 1 7 1 3
blood volume, 1690- 1695
capillary-
circulation, 1 702-1 707
hematocrit, effect on gas exchange,
1707
hematocrit, values for, 1 707
perfusion, alveolar ventilation and,
1682
pressure, 1052
pressure, wedge pressure and, 1688
cardiopulmonary disorders, 172 7- 1731
collaterals, 1265-1266, 1268, 1679
development of, 1267
comparative physiology, 1671-1673
drugs and, I 724-1 727
dynamics of, 1667-1731
exercise and, 1714-1716
functional anatomy, 1673-1 681
growth of ideas about, 1669—167 1
hemodynamic
interrelations, 1695- 1702
phenomena, 1 707-1 709
interplay with respiration, 1672
large vessels
anatomical structure, 1674
subdivision of, 1674
measurement of collaterals in, 1271
mechanical influences on, 1716-1717
oxygen tensions representative in, 17 19
pressure-volume relationship, 1696
respiration and, 1 709-1 713
serotonin and, 1725- 1726
test preparations for, 1673
vascular
bed, blood pressure and, 1705
distensibility, 1696
pressure gradient, 1687
vasoconstriction
in acute hypoxia, 1 7 1 9
site of, 1 7 1 9
vasoconstrictors and, 1701, 1726
vasodilators and, 1726
vasomotion, 1699, 171 7-1 724
passive mechanism, 1698
pulmonary vasomotricity, 1698
vasomotor
activity, 1 7 1 7
control, 1680
nerves, anatomical arrangement,
1680
reflexes, autonomic nerve supply,
1723
waves, pulmonary arterial pressure
and, 1723
waves, pulmonary arterial rhythm,
1724
waves, Traube-Hering-Mayer waves,
■723
venous pressure
left atrial events and, 1686
measurement of, 1686
values for, 1686
venous return, left ventricular output
and, 1 1 18-1 1 ig
Pulmonary edema
anoxia and, 1053
atrial pressures and, 1052
capillary "pore stretching" and, 1053
causes of, 1 730
experimental, 1729
filtration pressure and, 1053
lymph flow and, 1054
"neurogenic", 1730
pathogenesis of, 1054
Starling's law, 1730
Welch's hypothesis, 1 730
Pulmonary emphysema
cor pulmonale and, 1728
prolonged forced expiration, blood
pressure and, 1 7 1 2
pulmonary artery pressure and, 17 10
right atrial pressure and, 1 710
right heart failure, 1 728
Pulmonary hypertension
arterial pressure and, 1724
bullous emphysema, 1727
causes of, 1727
elevated ventricular pressure and, 1 555
granuloma, 1727
interstitial fibrosis, 1727
left heart failure, 1 727
lymph flow and, 1052
mitral stenosis, 1727
mitral valvular disease, 1727
primary pulmonary hypertension, 1727
pulmonary
arterial vasoconstriction, 1728
arteritis, 1727
blood flow and, 1727
capillary filtration pressure, 1728
emboli, 1727
venous hypertension, 1727
venous pressure and, 1727
restricted vascular bed, 1727
Pulmonary hypotension
alveolar dead space and, 1730
causes of, 1 730
Pulmonary vascular resistance
anomalous viscosity, 1698
blood flow and, 1699
calculation of, 1 698
exercise and, 1699
factors involved, 1698
hypoxia and, 1725
in human infant at birth, 1641
initial breath and, 1640
in man, 1697
norepinephrine, 1725
pulmonary arterial pressures and, 1699
see also Vascular resistance
Pulmonic valve
closure interval, 790
compared to aortic valve, 855-856
differential pressure across, 855-856
experimental stenosis, 1 730
incompetence, blood pressures and,
1729
opening interval, 790
Pulse energy absorption
instances of, 852
Pulse plethysmography
discussion of, 1 283
Pulse technique
discussion of, 1 3 1 9
Pulse wave
aortic and pulmonary compared, 1686
arterial, distribution of, 849
brachial
aortic regurgitation and, 831
in chronic hypertension, 832
varying levels, 831
central
contour, 821
stroke volume and, 833
contour
central and peripheral, 825
characteristics, 830
diastolic, 832
incisural vibration, 832
modification, 824
pressure and, 832
slope of pressure rise, 830
INDEX
'777
systolic peak, 832
systolic profile, 825
derivation of Bramwell and Hill equa-
tion, 816
factors affecting, 819
Fourier analysis of, 817
normal and pathological, 859
origin of, 815
phase lag with fluid displacement, 817
propagation
hypothermia and, 827
resonance, 826
standing waves, 826
secondary causes, 839
systolic length and, 817
velocity
diastolic pressure relation and, 817
heart rate and, 818
of various parts, 819
stiffness modulus and, 819
wall distensibility and, 815
Pumps
characteristics, to duplicate ventricular
contraction, 800
kinetic, 760
positive displacement, 760
reciprocating, definition, 760
rotary, 760
volume, left ventricle as, 844
Purine derivatives
inotropic effects of, 1 545
Purkinje cells
function, 763
Pyelonephritis
renal lymphatic pressure and, 1058
Pyrimidine derivatives
inotropic effects of, 1545
Q wave
interval, 790
Rabbit
blood pressure, age and, 1643
cardiac glycogen before birth, 1638
ear
arteriovenous communications, 1 255
microcirculation in, 898
plexus of regenerated vessels in, 921
hepatic blood flow in, 1405
pulmonary vessels in, 1674
renal clearance in, 1483
uterine vasculature, pregnancy and,
'597
Race
peripheral vascular disease and, 12 17
Radial flow
equation for, 842
Raffinose
diffusion of, 1002, 1009, 1010
filtration rate, 1006
permeability of muscle capillaries to,
1013
Rapela, C. E.
Resistance (conductance) and capaci-
tance phenomena in terminal vas-
cular beds, 935-960
Rat
capillary blood pressure in, 965
hepatic blood flow in, 1405
mesoappendix
circulation of, 920
for studies of microcirculation, 894
osmotic pressure of plasma proteins in,
965
protein osmotic pressure in, 965
pulmonary vessels in, 1674
renal clearance in, 1483
Raynaud's syndrome
basic defect, 1227, 1229
blood catecholamines in, 1229
capillarioscope studies in, 1228
capillary
changes in, 1228
pressure and, 968
changes in hands, 1227
clinical characteristics, 1226
diagnosis of, 1226
differential diagnosis, 1231
digital blood pressure and, 1229
evidence for local defect, 1228
localization of lesions, 1226
mechanisms in, 1227
monoamine oxidase in digital arteries,
1229
secondary characteristics, 1 230
sensitivity to cold in, 1228
symptoms, 1226, 1227
sympathectomy in, 1228
Reactance flow patterns
definition, 853
Rein's thermostromuhr
principle and discussion of, 1283
Remington, J. W.
The physiology of the aorta and major
arteries, 799-838
Renal artery
arterial flow pulse in, 849
circumference, 846
mean peak
flow, 846
velocity, 846
Renal blood flow
abdominal aortic pressure and, 852
acidosis and, 1504
adrenergic
control, 1485
response, 951
anesthetic agents and, 1488
apresoline and, 1488
autonomy of, 1 484
calculated from clearance data, 1 480
cholinergic response, 951
CNS control of, 1 485
computation of, 854
critique of clearance method, 1481
dominant hydraulic elements, 852
epinephrine and, 950
exercise and, 1501
in cardiac patients, [502
extrinsic regulation 1483-1489
factors affecting, 1482
ganglionic blocking agents, i486
heat turnover and, 1476
hemorrhagic hypotension and, 1505
hypercapnia and, 1504
hypoxia, ischemia and, 1503
in vessels, excluding organs, 1293
in young animals, 1644
measurement of, 1479- 1483
neurogenic control, 1 483
nitrous oxide method, 1 480
orthostatic hypotension and, 1502
posture, 1502
renal clearance principle and, 1479
response in physiological stress, 1501-
'5°7
shock and, 1505
sympathomimetic drugs, 1 486
tourniquet and, 1506
traumatic
injuries in man and, 1506
shock and, 1506
vascular compensation in man, tilting
and, 1503
Renal circulation
anatomy, 1 489-1 497
arterial system, 1458
dog, 1459
arteries, volume of muscle in wall, 1465
arteriovenous anastomoses, 1460
autoregulation, 947
cell separation hypothesis, 1491
decapsulation, 1494
factors impairing, 1 496
in vascular beds, 944
intrarenal reflexes, 1490
mechanism of, 1490
metabolic theory, 1490
myogenic theory, 1494, 1496
perfusion pressure and, 1494, 1495,
1496
tissue pressure theory, 1492, 1493
viscosity theory, 1491
capillary pressures in, 965
cell separation hypothesis, 1492
cortical, 1460
critical closing pressure, 1 476
distribution of cells and plasma, 1473
distribution of osmotic constituents,
1 469-1 470, 1472
functional architecture, 1458- 1466
glomerular, 1461
intrarenal hematocrit, 1470
juxtamedullary zone, 1464
medullary
and cortical transit time, 1500
1778
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
and papillary blood supply, 1460
in antidiuresis, 1499
in diuresis, 1499
perfusion pressure and regional transit
time, 1500
pressure gradients, 1475-1479
regional resistance changes, 1 49*3
sinusoidal cushions, 1461
stenoses, definition, 1461
to medullary zones, 1464
Trueta juxtamedullary shunt, 1497-
1501
vasa recta system, 1 465
venous
sinuses, 1459
system, 1459
valves, 1460
veno-venous anastomoses, 1459
Renal clearance
factor affecting extraction ratio, 1 482
in mammals, 1483
Renal disease
pyelolymphatic backflow in, 1058
Renal lymph
capsular, components, 1 056
colloid osmotic pressure and, 1057
composition, 1055, 1466
compared to plasma, 1056
venous pressure and, 1057
flow rate, 1055
glucose concentration, 1 055-1 056
protein concentration, 1056
production, 1466
renal injury and, 1058
sodium concentration, 1056
source of, 1055
urea concentration, 1 055-1 056
Renin
juxtaglomerular complex and, 1464
Reserpine
splanchnic circulation and, 141 7
Reservoir, arterial
reconstruction of extensibility response
in, 816
Resistance
theoretical discussion of, 841
see also Vascular resistance
Resistance flow patterns
definition, 853
Resonance
explanation, 826
model for, 826
Resonant wave
definition, 847
function of, 852
Respiration
central venous pressure and, 1 124
deep inspiration, venoconstrictor
responses, 1094
fetal, circulation changes, 1640
forced expiration, 17 13
in newborn, course of blood and, 1639
inspiration, hepatic circulation and,
1426
pulmonary circulation and, 1672
right ventricular stroke volume, 785
vena cava! flow and, 857
ventricular stroke volume and, 784
see also Negative pressure breathing;
Positive pressure breathing
Respiratory acidosis
effects of, 1545-1546
Respiratory gases
capillary permeability and, 1018
Rest
aortic flow curve at, 773
coronary blood flow and, 1554
oxygen uptake, cardiac output and,
1682
pulmonary vessel oxygen tension and,
■7>9
Reticuloendothelial cells
blood flow, phagocytic action and, 1401
Retina
capillary pressure in, 966
Reynolds number
critical value and inlet length, 1296
Reynolds, S. R. M.
Maternal blood flow in the uterus and
placenta, 1 585-1618
Rheoplethysmography
of digital flow, 1220-1222
Rheumatic fever: see Collagen diseases
Rheumatoid arthritis: see Collagen dis-
eases
Rodents
uterus in, 1585
Root, VV. S.
The flow of blood through bones and
joints, 1651-1665
Rose bengal
removal from blood, 1397
transfer from blood to bile, 1 397
value in clearance studies, 1 400
Rouget cell
capillary contractility and, 923
function, 923
smooth muscle cells and, 924
Rubella virus
cardiovascular abnormalities and, 1624
Rubidium
entrance rate into muscle, 1 138
ion diameters, 1 1 38
ionic radii of, 1 137
mobility in electric field, 1 1 38
relative diameter, 1 1 38
Rutin
vascular effects of, 972
Sartorius muscle
weight, Na and, 1 136
Scleroderma
blood vessels and, 1239
characterization, 1238
organs involved, 1239
picture of, 1 238
Raynaud's phenomena and, 1239
skin pathology in, 1238
vascular signs, 1 239
see also Collagen diseases
Scotomata
menstrual cycle and, 1599
Secretin
pancreatic blood flow and, 1450
Semilunar valve
closure
backflow wave and, in congenital ste-
nosis, 859-860
interval, 790
opening interval, 790
passive movements of, 779
Sensory afferent fibers
distribution, 1526
origin, 1 525-1 526
Septal defects
aortic blood flow, 857
pulmonary blood flow and, 857
ventricular ejection pulses and, 857
Seroche: see Mountain sickness
Serotonin
arterial and venous
effects, 1094
pressures and, 1088
capillary permeability and, 1061
effects of, 1726
gastric blood flow and, 1447
hepatic circulation and, 1423
intestinal blood flow and, 1448
mesenteric blood flow and, 1451
possible role in arterial spasm, 1237
renal hemodynamics and, 1488
skeletal muscles and, 1355
skin circulation and, 1345
vascular changes due to, 969
vessels affected by, 951
venoconstrictor responses, 1094
Serum
permeability of mammalian muscle
capillaries to, 1013
Serum albumin
disappearance from plasma, 1008
Serum cholesterol
aging and, 1 198
atherosclerosis and, 1 1 97
diet and, 1 182
dietary cholesterol and, 1 184
lowering
by dietary means, 1 1 85
mechanism, 1 185
neomycin and, 1 187
thyroxin and, 1186-1187
Serum lipids
atherosclerosis and, 11 75-1 180, 1 187
chain length of dietary fat and, 1 183
classification, 1 1 75
description, 1 1 75
INDEX
'779
dietary
fatty acids and, 1 182
manipulation and, 1 186
diseases affecting, 1 180
factors influencing, 1 180-1 187
melting point of dietary fat and, 1 1 83
"normal" level, 1181
origins of, 1 179
sex and, 1 181
stress and, 1 181
unsaturation of dietary fat and, 1 183
Serum lipoproteins
atherosclerosis and, 1046
Sex
peripheral vascular disease and, 1 2 1 7
serum lipids and, 1 1 8 1
Sex hormones
atherosclerosis and, 1 203
cholesterol metabolism and, 1 203
lipid metabolism and, 1203
Sheep
blood pressure, age, and, 1643
blood vessels of placenta, 1622
cardiac glycogen before birth, 1638
hepatic blood flow in, 1405
Shock
aldosterone and, 1 1 16
cardiac output and, 1 1 14
hemorrhagic
coronary flow, blood pressure and,
'535
discussion of, 1558
myocardial depression and, 1559
oligemic shock, 1558
02 uptake, 1558
irreversible
cardiac output and, 1 1 14
right atrial pressure and, 1 1 14
venous return and, 1114
lymphatic system in, 1058
renal blood flow and, 1505
right atrial pressure and, 1 1 14
tourniquet
renal blood flow and, 1506
"toxic factor" in, 1060
venous return and, 1 1 14
Skeletal muscle
autoregulation, tissue, 945-946
contraction
blood flow and, 1354
intermittent claudication and, 12 17
interstitial fluid pressure, 980
man
fluid pressure in, 980
sympathetic vasoconstrictor tone in,
sympathetic vasodilator fibers, 1366-
1367
metabolism, exercise and, 1377
percentage composition by volume,
1327
percentage of parts of human limbs,
1327
relative entrance rates of electrolytes,
1138
spontaneous activity, K and, 1 156
tetanic contraction, blood flow and,
'354
Skeletal muscle blood flow
adrenaline and, 1359
adrenergic response, 951
after arterial occlusion, 944
after sympathectomy,
effect of exercise, 1 369
anoxia and, 1376
arterial pressure and, 1357
cholinergic response, 951
cholinergic vasodilator nerves and, 1368
contraction and, 1 354
epinephrine and, 948, 950, 1370, 1372
exercise and, 1354, 1355
hypothalamic stimulation and, 1359
in fainting, 1367
in dog, 1 36 1
local temperature and, 1358
oxygen consumption and, 1377
perfusion pressure and, 943
phenoxybenzamine and, 948
posterior root fibers and, 1368
preparation for study of baroreceptors,
■3°3
radioiodine clearance in, 1359
receptors for control, 949
sympathetic stimulation and, 948
sympathetic vasoconstrictor nerves and,
1362
Skeletal muscle circulation
arteriovenous anastomoses in, 902
autoregulation in, 944
basal tone, 1 355-1 358
of vessels in, 1355
capillaries
fine structure in, 101 1
structure of, 1009-101 3
capillary
counts, 1019, 1358
distribution in, 902
exercise, adrenaline and, 1374
function, 1 358-1 359
hyperemia
exercise, 1 376-1 380
reactive, 1374-1376
hypothalamic stimulation, 1367
microcirculation in, 900
nervous control, 1359-1369
oxygen consumption and work, 1378
passage of red cells in capillary from,
9°3
pattern of vessel arrangement, 901
preparation for study, 1360
pressure-flow relations, automaticity
and, 1357
reactive hyperemia in, 944, 1374
resistance changes, exercise and, 1379
resistance of vascular bed, vasoactive
agents and, 948
scheme for, 1359
structure, 1358- 1359
sympathetic
impulses to, exercise and, 1369
vasoconstrictor nerves and, 1 359-
■365. '366
vasodilator nerves and, 1366
sympathomimetic substances and,
'369-'374
temperature -regulating center and,
1368
vasoconstrictors and, 1355, 1359
venous oxygen saturation after exercise,
'377
vessel stretch and, 1357
weight, Na and, 1 136
Skeletal muscle vessels
A-V shunts and, 1358
blood volume and, 1360
capillary filtration in, 1360
critical closing pressure and, 1357
resistance of, 1 360
stimulation of arterial baroreceptors,
1363
structure and function, 1358
sympathetic
vasoconstrictors and, 1360
vasodilator nerves and, 1 366
sympathomimetic substances and, 1369
to fibers and to connective tissue, 1 359
Skin
blood vessels of, 1 326
color, 1328
blood flow and, 1328
temperature and, 1220
dermis, capillary bed of, 904
fluid pressure in, 980
innervation of, 1 338
interstitial fluid pressure, 980
mechanism of vasodilatation in ani-
mals, 1340
microcirculation in, 903
pathology in scleroderma, 1238
percentage composition by volume,
1327
percentage of parts of human limbs,
1327
reactive hyperemia and, 1329
resistance in interstitial, 979
temperature of, 1 328
blood flow and, 1328
color and, 1220
local temperature and, 1332
vasoconstrictor sympathetic nerves in,
■337
venous oxygen saturation in, 1336
see also Cutaneous
Sleep
venodilator responses, 1094
1780
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
Smooth muscle
characteristics, 870
conduction in, 870
degeneration, aging and, 872
hyporesponsiveness to drugs, Na gra-
dient and, 1 145
mechanical properties, 872
pressure-volume diagrams for, 873
response to stretch, 870
spontaneous activity, 880
tension
K and, 1 145
K efflux and, 1 156
membrane potential and, 1 155
Na and, 1 144
Na gradient and, 1 145
spike activity and, 1 157
tonus and, 1 156
see also Vascular smooth muscle
Sodium
blood pressure and, 1 155
deficiency, signs, 1 1 43
distribution in kidney, 1469
entrance rate into muscle, 1 1 38
extracellular, blood pressure and, 11 54
in bladder, 1 141
in stomach, 1 141
intake, juxtaglomerular granules and,
1464
in uterus, 1 141
in vascular tissue, 1 141
ion diameters, 1 1 38
ionic radii, 1 137
mobility in electric field, 1 138
muscle weight and, 1 136
radioactive, clearance as measure of
blood flow, 1328
relative diameter, 1 138
relation to hypertension, 1142
serum level, vascular muscle tension,
"47
smooth muscle tension and, 11 44, 1145
space, in organs and tissues, 1 141
vascular smooth muscle tension and,
[142-1157
Sodium chloride
diffusion of, 1010
diffusion rates, 1012
permeability of muscle capillaries to,
1013
renal
artery pressure and, 11 48
lymph flow and, 1058
vein flow and, 1 1 48
Sodium cyanide
coronary blood flow and, 945
Sodium hydroxide
mean activity coefficients, 1 1 37
Sodium nitrite
gastric blood flow, 1447
Sodium transport
blood vessel walls in, 884
influx
in blood vessel wall, 884
in frog skin, 884
placental, in various species, 1592-1593
Spannmuskeln : see Vascular smooth
muscle
Species differences
atherosclerosis sex hormones and, 1204
cholesterol feeding and, 1 205
vascular caliber and medial thickness,
1674
Spencer, M. P.
Pulsatile blood flow in the vascular sys-
tem, 839-864
Splanchnic blood volume
determination, 1406
factors affecting, 1406
measurements of, 1402, 1403
splenectomy and, 1403
Splanchnic circulation
acetylcholine and, 141 9
blood transit time, 1402
diagram, 1406
hepatic interrelationships, 1428
inflow, arteriolar resistance and, 1407
man and dog, 1405
neurohumoral determinants, 1417
parasympathetic fibers and, 1415
pressures in, 1 405
reflex regulation of, 141 6
resistances in, 1 405
Splanchnic stimulation
intestinal blood flow and, 1447
mesenteric blood flow and, 1451
pancreatic blood flow and, 1450
Splanchnic viscera
innervation of, 141 5
Spleen
arrangement of lymphatics in, 1 038
blood flow, 1 44 1, 1443
factors affecting, 1450
blood volume in, 1445
microcirculation in, 910
removal, splanchnic blood volume and,
■4°3
species differences in circulation, 913
theories of circulation, 910
vascular connections in, 912
vascular volume changes in, 955
Stable loop
definition, 867
Standing waves
explanation, 826
Static stretch
definition, 867
Stellate ganglion
stimulation, 1548
arterial blood pressure, 1548
coronary inflow and, 1548
Steroids
absorption of, 1 045
Stewart-Hamilton method
cardiomegaly, 1692
central blood volume, 1691
mean circulation time, i6gi
pulmonary congestion, 1692
Stomach
blood flow, 1441, 1442, 1443
factors affecting, 144^
blood volume in, 1445
clearance of aniline, 1443
electrolyte levels in, 1 141
microcirculation in, 904
Strain differences
cholesterol feeding and, 1205
Stress
atherosclerosis and, 1 206-1208
cholesterol metabolism and, 1206
cutaneous blood flow and, 1339
epidemiology of, 1 1 8 1
heart rate and, 1553
lipid metabolism and, 1206
serum lipids and, 1 181
stroke coronary oxygen increase and,
1553
Stress-relaxation
behavior, in various arteries, 872
definition, 1084
description of, 803
reversibility, 803
Stretch curve
contribution of structural elements, 806
muscle contraction and, 807
Stroke volume
aorta flow during exercise, 846
calculated, compared to Fick proce-
dure, 834-835
calculation, 833-835
central pressure pulse and, 833
exercise, 846
indirect calculations, 834
relation to diastolic capacity, 784
right ventricular, spontaneous respira-
tion and, 785
under various conditions, 782
Subcutaneous tissue
fluid pressure in, 980
interstitial fluid pressure, 980
percentage of parts of human limbs,
1327
percentage composition by volume,
'327
Subclavian artery
pressure pulse contour, 831
Sucquet-Hoyer canal
description, 1252
Sucrose
blood to tissue permeability, 1023
capillary permeability, 1023
diffusion of, 1002, 1010
disappearance from plasma, 1008
filtration rate, 1006
permeability of, 1023
muscle capillaries to, 1013
[78i
Suction
definition, 787
Sulfate
entrance rate into muscle, 1 138
Sulfuric acid
mean activity coefficients, 11 37
Sweat
evaporation, vasodilatation and, 1337
Sweating
cutaneous circulation and, 12 19
skin blood flow and, 1340
Sympathetic nervous system
collateral circulation and, 1262
coronary blood flow and, 1548
direct action on heart, 1549
fibers of variable intensity, 1549
fibrillating heart, 1549
inhibition, effectiveness of heart and,
1 104
innervation
of kidney, 1467
of splanchnic viscera, 14.15
of uterus, 1602
vasoconstrictor fibers
arteriovenous anastomoses and, 1255
circulation and, 1362
impulse frequency in, 1365, 1366
impulses to skeletal muscle vessels,
1365
in skin, 1337
skeletal muscle circulation, 1359-
'365
tissue fluid volume, 1362
vasodilator fibers
arteriovenous anastomoses and, 1255
in skin, 1337
skeletal muscle circulation, 1366
venous tone and, 1077
Sympathetic stimulation
aortic flow curve and, 773
bone medullary pressure and, 1660
cardiac output, 1 1 12
effectiveness of heart and, 1 104
gastric blood flow and, 1446
heart and, 1549
muscle blood flow and, 948
resistance in vascular bed and, 948
right atrial pressure and, 1 1 12
segmental resistance and, 952
small vessel pressure and, 971
splenic blood flow and, 955, 1450
venous
pressure and, 970
return and, 1 1 1 2
ventricular and lung volumes and, 783
Sympathectomy
cutaneous blood flow and, 1338
extremity temperature after arterial
ligation and, 1264
skin circulation and, 1340
thrombophlebitis and, 1240
vessel spasm and, 1237
Sympathomimetic substances
skeletal muscle circulation and, 1369-
'374
synephrine, bone medullary pressure
and, 1660
Synovial membrane
sensitivity to pain, 1662
Systolic reserve volume
definition, 782
Temperature
arteriovenous anastomoses and, 1 255
blood pressure and, 968
capillary blood pressure and, 966, 968
changes
erythromelalgia and, 1241
in acrocyanosis, 1230-1231
in livedo reticularis, 1233
Raynaud's syndrome and, 1226
circulation arrest and, 1224
cutaneous
blood flow and, 1330
circulation and, 12 19
filtration coefficients and, 988-989, 993
hand, sympathectomy and, 1331
local
muscle blood flow and, 1358
skin circulation and, 1330
of extremity
arterial ligation and, 1264
sympathectomy and, 1264
of skin, as measure of blood flow, 1327
reactive hyperemia and, 1329
regulating center, muscle circulation
and, 1368
skin capillaries and, 904
vasoconstriction and, 1094
vasomotion and, 927
Tendon
percentage composition by volume,
'327
percentage of parts of human limbs,
'3-7
Tension-length relationship
theoretical distensibility
calculated moduli, 81 1
Tension muscles
blood pressure and, 875
connections, 869-870
description, 869-870
essential hypertension and, 875
Tension-time index
definition, 1560
Terminal impedance
energy transference of, 852
Terminal vascular beds
capacitance, 935-957
in submucosa and muscle, 906
resistance, 935-957
vessel and, 936-950
vessel structure in, 895
Tetraethylammonium chloride
effect on catecholamines, 1563
Tetralogy of Fallot
pulmonary collateral circulation and,
1262, 1679
Thebesian vessels
concentration of, 1531
functional anatomy, 1517
Thermal conductivity
measurement, 1 285
of skin, blood flow and, 1327, 1329
Thermal methods
principle of, 1 283
Thermistors
blood flow and, 1287
Thermocouples
blood flow and, 1287
Thermoregulation
arteriovenous anastomoses and, 1256
Thermostromuhrs
disadvantages of, 1 284
temperature profile, 1 284
Thiourea
renal
artery pressure and, 1148
vein flow and, 1 148
Thoracic splanchnic nerve
innervation of kidney, 1467
Thrombophlebitis
acute, 1239
symptoms, 1 240
sympathectomy and, 1240
vasospasm in, 1240-1241
migratory, definition, 1240
peripheral vascular disease and, 1217
phlebothrombosis and, 1239-1240
plethysmograph pulses in, 957
vasoconstrictor mechanisms and, 1239
Thrombosis
atherosclerosis and, 1189
blood clotting and, 1 189-1 190
mechanism of, 1189
Thrombotic thrombocytopenic purpura:
see Collagen diseases
Thrombus
blood clots and, 1 190
structure, 1 igo
Thyroid
activity, atherogenic diets and, 1200
fat mobilization and, 1175
Thyroxin
cholesterol levels and, 1 186-1 187
Thyrotoxicosis
heart rate and, 1554
stroke coronary
flow and, 1553
oxygen usage and, 1553
Tilting
renal blood flow and, 1503
Tissue pressure: see Interstitial fluid pres-
sure
1782
HANDBOOK OF PHYSIOLOGY ^ CIRCULATION II
Tm
definition, 1397
Tobacco
cardiac effects, 1 563
coronary disease and, 1563
Tolazoline
renal blood flow and, 1487
Tongue
arteriovenous anastomoses in, 1 254
Toxemia of pregnancy
ovarian vein physiology and, 1 588
placental ischemia and, 1588
Transformer component
definition, 1315
Transfusion
effects of, 1546
rapid
cardiac output and, 1 1 1 3
right atrial pressure and, 1 1 13
venous return and, 1 1 13
Transillumination
with fused quartz rod, 893
Transmural pressure
calculation, 1689
Traumatic shock
capillary permeability and, 1059
lymph flow and, 1060
participation of the lymphatic system
and, 1059
renal blood flow and, 1506
vasoconstrictor substance and, 1060
see also Shock
Traveling markers
dye, 1320
gas bubbles, 1320
radiopaque material, 1320
tagged corpuscles, 1320
Trench foot
skin circulation and, 1335
stages of, 1 335
Tricuspid valve
closure interval, 790
flow pulse at, 862
opening interval, 790
pulse wave at, 862
regurgitation, differential pressure
method, 862
structure, 780
Trimethaphan camphorsulfonate
renal blood flow and, 1487
Trousseau syndrome : see Thrombophlebi-
tis, migratory
Trueta juxtamedullary shunt
functional evidence, 1498
morphological evidence, 1497
Tumors
origin of blood vessels in, 1 265
Tyramine
medullary pressure and, 1660
Ultraviolet light
skin circulation and, 1336
Umbilical vessels
artery
pulse-wave velocity and rate in, 819
stress-relaxation curves of, 872
closure of, 1638
pressures in, 1632
Ungulates
uterus in, 1 585
Uranium-nitrate injury
renal lymph and, 1058
Urea
blood to tissue permeability, 1023
capillary
clearance blood flow and, 1023
permeability, 1023
concentration in lymph, 1064
diffusion of, 1002, ioio, 1012
distribution in kidney, 1469
filtration rate, 1006
osmotic transients due to, 1018
permeability of, 1023
muscle capillaries to, 1013
renal lymph flow and, 1058
Ureter ligation
pathological changes in kidney, 1058
Urethan
osmotic transients due to, 1018
Urine
concentration, ADH and, 1472
formation
countercurrent system, 1479
intrarenal pressure and, 1479
Uterine blood flow
accommodation of embryo and, 1594-
'598
changes during pregnancy, 1604
contractions and, 1599
drugs and, 1637
during labor, 1607
during puerperium, 1608
fetal load and, 1606
fetal weight and, 1605
postpartum, 1607
pregnancy and, 1603, 1607, 1609
total, methods, 1604, 1605
uterine shape and, 1603
Uterine circulation
accommodation of embryo and, 1594-
1598
arrangement, in various types, 1587
arterial
injection during pregnancy, 1604
pattern of, 1588
supply, in simplex, 1600
arteries
course of, 1586
menstruation and, 1598
comparative, 1586
estrogen and, 1600- 1602, 1603
hormones and, 1599
implantation and, 1602
in nonpregnant rabbit, 1594
pregnancy and, 1594, 1602, 1603, 1604
vascular connections of, 1586- 1588
vascularity, 1598-1599
venous drainage, 1 587
Uterine contractions
arterial blood pressure and, 1 599
blood flow and, 1599
compression, inferior vena cava and,
1600
maternal circulation, 1605
placental blood flow and, 1600
venous pressure and, 1606
Uterine milk
function of, 1621
Uteroplacental circulation
as arteriovenous shunt, 1606
Uterus
accommodation of embryo and, 1594-
1598
angiogenesis in, 1586
arterio-arterial shunts in, 1597
basic function of, 1585
bicornus, definition, 1586
comparative anatomy of, 1585- 1586
comparative types, 1586
contractility
blood flow and, 1 599-1600
body posture, 1600
duplex, definition, 1586
electrolyte levels in, 1 141
endometrium, vascularity of, 1599
estrogen and, 1 600-1 601
hyperemia due to hormones, 1601
innervation of, 1602
lymphatics of, 1608
menstruation and, 1598- 1599
oxygen consumption, fetal weight and,
if Mill
pregnant, radiopaque injections of,
1609
role in implantation, 162 1
simplex, definition, 1586
work of supplying various types of pla-
centa, 1606
Vagal stimulation
gastric blood flow and, 1446
mesenteric blood flow and, 145 1
pancreatic blood flow and, 1450
Vagus stimulation
vasomotor activity, 1 547
Valsalva maneuver
aortic pulse contour in, 831
brachial arterial pressure, 1712
brachial artery pulse contour in, 831
pulmonary arterial pressure and, 1 7 12
Valves
arterial, 781-782
atrioventricular, 780-781
forward work of, 847
INDEX
'783
function of, 779
in lymphatic vessels, 1054
removal, blood pressures before and
after, 1729
time intervals between motions, 790
venoatrial junction, 780
Valvula Eustachii
description, 780
Valvula Thebesii
description, 780
van Itallie, T. B.
Lipid metabolism in relation to physi-
ology and pathology of athero-
sclerosis, 1 167— 1 1 95
van't Hoff's law
deviations, 975
physiological significance of, 975
protein osmotic pressure concentration
curve and, 973
Vasa recta
function of, 1057, 1458
Vasa vasorum
classification, 885
flow resistance, 886
function of, 883
interna, description of, 885
location, 885
pressure and flow in, 886
types, 885
vessel wall penetration, 885
Vasa venarum
anatomy of, 1076
drainage of, 885, 1076
in vascular disease, 1076
Vascular beds
analysis of behavior, 935
autoregulation, 944
definition, 942
blood volume in, 953-954
capacity, methods of studying, 953
compensation for pressure change, ter-
minology, 942
extravascular pressure and pressure-
flow relationships, 941
flow in, epinephrine and, 950
interpretation of behavior, 957
pressure-flow relations, 936
autoregulation and, 942
methods, 936
normal plots, 937
mathematical relationships, 937
venous pressure and, 941
resistance vessels
autoregulation of, 941
chemical effects on, 948
extrinsic control, 948
segmental resistances in, 950-953
extrinsic agents and, 951
perfusion pressure and, 95 1
venous pressure and, 951
Vascular behavior
interpretation of, 957
malfunction, 12 16
physical examination and, 12 18
simple clinical tests of, 1 2 1 8
Vascular capacity: see Vascular beds,
blood volume in
Vascular hydraulics
arterial system, blood flow in branches,
852
axial flow, 842
compliance, 842
dilatability tests of, 1222
distensibility, time dependency in elas-
tic behavior and, 1084
distensible tubes, pulsatile flow in, 815
elements of, 841-845
flow source, pressure source and, 844
future considerations, 794
hydraulic impedance, 843-844
inertance, 841-842
principles of, io7g
problems of pulsatile flow in flexible
tubes, 800
radial flow, 842-843
resistance, 841
total flow in elastic pipe, equation, 843
Vascular resistance
active pulmonary vasodilatation, 17 16
arteriolar, hepatic inflow and, 1407
calcium, magnesium and, 1 158
cardiac output and, 1 1 18
changes in organ volume and, 955
circulatory effects of, 1 1 18
collagen "jackets", 1716
definition, 820
during exercise, 17 16
hypoxia and, 1 544
mathematical relation
to flow, 939
to pressure, 939
muscle circulation vasoconstrictors and,
1 36 1
passive pulmonary dilatation, 17 16
peripheral
major site of, 895
measurement of, 1 139
potassium and, 1 149
venous return and, 1 107, 1 108
postcapillary increase, results of, 956
precapillary increase, results of, 956
pH change and, 1 159
potential energy gradient, 17 16
resistance vessels, 936
terminal vascular beds and, 936-950
terminal vascular beds, 935-957
total peripheral, drugs and, 1564
vasomotor activity, 1 7 1 6
venous
blood flow and, 1 1 26
pressure and, 11 26
return and, 1 r 18
see also Pulmonary vascular resistance
Vascular smooth muscle
adrenaline and, 1373
aging and, 872-875
arrangement, 806, 870
plastic and contractile mechanisms,
87.
autonomic nervous system, 87 1
catch mechanisms, 871
collagen fibers as safety factor, 880
conduction in, 870
contractility, 927
contraction of, 870, 87 1
arterial distensibility and, 806
creep in, 805
curare and, 1380
elastic, viscoelastic, and plastic be-
haviors, 868
elasticity of veins and, 882
excitation in, 870
extensibility of, 805
extracellular ions, water and, 1 140
extracellular space of, 1 157
functional syncytium of, 870
intracellular ions, water and, 1140
in veins, 881
ions and, methods of studying, 1 1 39
longitudinal, in arteries, 1268
measurement problems in, 11 42
methods of studying ion effects, 1139-
1 142
modifying properties of collagen and
elastic tissue, 872
oxygen lack and, 1 1 25
pressure-flow relations and, 870
properties, 870
response to stretch, 870
ring muscles, description of, 870
shift from plastic to viscoelastic be-
havior, 871
spontaneous pendular rhythm and, 870
stress-relaxation of, 872
structure of, 869
tension
anions and, 1 1 59
Ca and Mg and, 1 157
cell volume and, 1 160
changes of milieu and, 1 139
diastolic blood pressure and, 1 1 39
factors, 1 160
K and, 1 1 45, 1 1 46
K infusion and, 1 149
measurement of, 1 1 40
Na and, 1 144, 1 154
Na and K in, 1 151
Na gradient and, 1 145
Na infusion, 1 147
permeability to Na and, 1 160
pH and, 1 158
ratio of K and Na, 1 1 47
role of ions in, 1 142-1 159
sodium and, 1 147
water and, 1 154
1784
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
tension muscle, 869
description of, 870
types of, 869
see also Blood vessel walls
Vasculitis
description of, 1238
see also Collagen diseases
Vasoconstriction
disease syndromes, 1 226
evidence for, 1 7 1 g
oxygenation of blood and, 1125
plasma Na and K and, 1154
veins
factors affecting, 1094
temperature changes and, 1095
Vasodepressor drugs
in heart disease, 1567
postpartum uterine blood flow and,
1608
Vasodilatation
adenosine triphosphate and, 1338
axon reflex pathway, 1338
bradykinin and, 1340
capillary pressure and, 968
erythromelalgia and, 1241
factors affecting, 1094
ganglionic blockers and, 1561
hypertonic solutions and, 1 148
lymph flow and, 968
nitroglycerin and, 1561
oxygenation of blood and, 1 1 25
potassium and, 1380
reflex, 12 19, 1223
sodium and, 1 1 47
sodium nitrite and, 1561
Vasomotion
active and passive, 925
active, environment and, 926
anesthesia and, 925
arterioles in, 900
blood flow and, 926
blood pressure and, 926
definition, 8g2, 925, 928
fluid exchange and, 926
function, 928
in arterioles, 904
in conjunctiva, 909
internal pressure of vessels and, 927
ionic, active, 925
irregular activity, 925
metarterioles in, 925
of arteriovascular vessels, 896
of terminal arterioles, 897
receptors, in intestine, 1448
rhythmical activity, 925
venous pressure alterations due to, 927
Vasomotor tone
arteriovenous difference, pressure, blood
flow and, 937, 938
cardiac output, 113
right atrial pressure, 1 1 13
mean systemic pressure and, 1 1 1 o
myogenic theory, 1409
venous return, 1 1 13
see also Venomotor activity
Vasopressin
skeletal muscles and, 1355
Vasospasm
plethysmograph pulses in, 957
Vasospastic claudication
definition, 1 2 1 7
Veins
action of skeletal muscle upon, 882
arcuate patterns of, 897, 905
arterial pulsation in, 883
as reservoirs, 1078
capacity of, 1078
central, reservoir function, 787
circular muscle in, 881
collapse
blood flow and, 1080
elevated abdominal pressure and,
1 1 23
hemodynamic significance, 1080
hydrostatic shifts and, 1082
right atrial pressure and, 1 1 07
venous return and, 1123
collapsibility, function, 1082
composition of, 881
constricted
distensibility pattern, 1084, 1086
constriction due to levarterenol, 1086
distensibility of, 1082
standardization need, 1085
noradrenaline and, 1093
distensibility pattern, 1084,1086
repeated stretch and, 1085
elastic-muscular system of, 88 1
elongation
spiral twist and, 1083-1084
volume change and, 1083
large, contribution to blood reservoir,
't>79
location of vasa vasorum in, 885
longitudinal muscle in, 881
mechanical properties, 882
morphological differences in wall
structure, 952
muscular arrangement in, 881
nature of constriction, 1086
occlusion, afterdrop, 1 283
phasic pressure changes in, 1082
physiological characteristics, 1079
plastic properties of, 1 085
postural effects upon, 881
pressure alterations due to vasomotion,
927
pressure changes due to anticipation,
1090
pressure -volu me
curves from, 1083
diagrams, 882
proportion of circulating blood in, 900
radial distention, volume change and,
1083
repeated stretch and, 1085
restoration of collapsed to cylindrical
shape, 1 082-1 083
spontaneous pressure waves, 928
stretch curves obtained, 1084
stress relaxation, 1085
structure of, 881
structural support, 883
tone
parasympathetic nerves and, 1077
sympathetic control of, 1077
valves, 1077
aging and, 882
function, 1078
insufficiency and, 882
pumping action, 882
structure of, 882
valvular competence, 1077
vasomotion in, 926
venospasms, 1086
venovenous reflexes, 1095
wall structure in, 881
wall tension in, 882
Velocity inductance
definition of, 1303
Velocity resistance
definition of, 1303
Vena cava
distensibility, collagen fiber recruit-
ment and, 882
inferior, compression, uterine contrac-
tions, and, 1600
superior, pulsatile flow in, 856
Venivasomotor reflex
definition, 141 6
Venomotor activity
adrenergic response, 1087
assessment of, 1087
congestive heart failure and, 1095
direct observation of, 1087
distensibility by volume increment,
1 092-1093
inferences from venous pressure, 1 087
in various species, 927
in vitro studies, 1087
measurement of pressure gradients,
1088
plethysmography and, 1091
psychic stimulation and, 1089
pulse methods, 1090
summary of, 1094
venous distensibility patterns and, 1091
venomotor index, 1 092
Venous blood flow
body acceleration and, 854
peripheral venous pressures and, 1 127
phasic variations, 854
postural changes and, 854
pulsatile flow, 854
systemic, 854-855
INDEX
>785
vena caval
atrial contraction and, 854
heart's action and, 854
normal respiration and, 855
oscillatory pressure gradient, 854
pressure gradient in, 854
respiration and, 857
Venous circuit
methods for determination, 1531
Venous congestion
central, venous pressure and, 1031
interstitial fluid pressure and, 980
peripheral venous pressure and, 1081
protein concentration in extravascular
fluids, 985
protein, in interstitial fluid, 985
Venous pressure
abdominal pressure and, 1 1 23
anticipation and, 1090
body position and, in pregnancy, 1601
capillary pressure and, 968
central
heart beat and, 1 1 24
peripheral venous pressure and, 1 08 1
respiration and, 1 1 24
venous congestion and, 1081
changes, venous return and, 1 1 24
clinical evaluation from, 1224
cold and, logo
coronary, values for, 1542
composition, renal lymph and, 1057
congestive heart failure and, 1057
edema and, 969, 1 240
factors affecting, 972, 11 26
filtration and, 977
coefficient and, 989
rate and, 978, 989
fluid absorption and, 982
human forearm volume and, 1093
hypertension, 1224
intrarenal pressure and, 1478
left atrial pressure and, 1 1 1 8- 1 1 1 9
lymph flow, 969
lymphatic pressure and, 1 037
measurements in an occluded segment,
1089
peripheral
hydrostatic pressure and, 1 127
right atrial pressure and, 1 127
venous congestion and, 1081
venous flow and, 1 127
pulmonary, elevation, lymph flow, and,
1052
pulse methods, 1090
renal lymph flow and, 1056
serotonin and, 1088
sympathetic stimulation and, 970
transverse muscle in veins and, 882
uterine contractions and, 1606
venous return and, 1 1 26
venous system and, 1079
Venous return
acute left heart failure and, 1 1 20
acute right heart failure and, 1 1 2 1
arterial resistance and, 1 1 08
bilateral femoral A-V fistulae and, 1 108
blood
transfusion, 1 1 14
volume and, 1 1 10, 1 1 14
cardiac decompensation, 1 1 1 7
cardiac output and, 1 100, 1 1 12
curves and, 1103-1105, 1111-1118
central pulsation and, 11 24
circuit analysis applied to, 1100-1102
classical analysis, 1 102-1 103
complex graphic analysis of, 1 1 1 8-1 122
congestive heart failure and, 1 1 1 6
definition, 1099
exercise and, 1 1 13
external pressure on heart and, n 04
factors affecting, 1 1 22-1 1 26
factors regulating, 1100
fainting and, 1 1 23
heart, recompensation and, 1 1 1 7
hemorrhage and, 1 1 1 4
local oxygen utilization and, 1 1 26
mean systemic pressure and, 1 109
muscular exercise, 1 1 13, 1 125
myocardial damage, 1 1 15
opening the chest, 1 1 15
peripheral resistance and, 1 107, 1 108
pressure gradient for, 1 107
principles of circuit analysis, 1 100
pulmonary analysis of, 1 1 18
rapid transfusion and, 1 1 1 3
rate of epinephrine injection and, 1 1 1 1
rectification phenomenon and, 1 1 25
right atrial pressure and, 11 06, 11 10,
1 1 12
shock, irreversible and, 1 1 14
simplified graphical analysis, 1 103-
1 1 18
statement of problem, 11 00
studied with external perfusion circuit,
1 106
sympathetic stimulation, 1 1 12
systemic pressure and, 1 1 09
tissue metabolism and, 1 1 25
vascular resistance and, 1 1 1 8
vasomotor tone, 1 1 1 3
venous
collapse and, 1 123
pressures and, 11 26-1 127
pump and, 1 122
resistance and, 1 108
return curves and, 1 105-1 1 1 1
ventricular output and, 1100, 11 20
see also Pulmonary venous return
Venous return curves
blood volume and, 1 1 10, 1 122
cardiac output curves and, 1103, 11 05
epinephrine and, 1 1 10
factors affecting, 1 1 1 1
fluid retention and, 1 1 16
mean systemic pressure and, 1 109-1 1 10
method of recording, 1 1 05
normal, 1 1 06
normal, methods of establishing, 1107
Venous system
blood flow and, 1079
disease, symptoms of, 1218
evaluation of status of, 1 223
flow in systemic veins, 854
peripheral
innervation, 1076
pressure, central venous pressure
and, 1 08 1
structure, 1075
streamlining in, 772
Ventricle, left
ability to maintain entire circulation,
791
as pressure pump, 790
as volume pump, 790
changes in wall after birth, 1641
deceleration of blood rate, 845
linear acceleration of blood, 845
output, pulmonary venous return and,
1 1 1 8-1 1 19
pressure, lactic acid and, 1722
pressure-volume curves, 762
rapid-filling interval, 790
Ventricle, right
ejection pulse, form, 855
pressure, 778, 779, 791
positive pressure breathing and, 1 71 1
rapid-filling interval, 790
systole, function of, 791
Ventricles
asynchrony in systole, 790
blood flow velocity in, 772
changes after birth, 791
compared with piston pump, 784
comparison of left and right, 790
configuration changes, 765
diastolic suction, functional residual
capacity and, 7go
distensibility, coronary vessels and, 763
ejection pulses
comparison of, 857
during deceleration, 846
exercise and, 845
repair of congenital septal defect
and, 857
equilibration of left and right, 1120
expenditure of energy, 847
filling, 786-788
by atrial contraction, 774
forces responsible for, 788
nature of vis a fronte, 789
rapid and slow, 774
sequence, 854-855
suction and, 788
vis a tergo and vis a fronte, 788
function, coronary vessels and, 762
1786
HANDBOOK OF PHYSIOLOGY
CIRCULATION II
functional residual capacity, 788
gross dissection studies of, 1518
metabolism, methods, 1532
output
blood volume and, 1 121
comparison of two sides, 1 1 ig
hemorrhage and, 1121
venous return and, 1 1 20
pressure differences with aorta, 781-782
pressure in, during contraction, 771
rapid and reduced ejection, 772
septal flow, extravascular mechanical
compression and, 1556
stroke volume
respiration and, 784
variations, 784
systolic suction, 787
volume, 782-785
correlation of terms, 784
lung volume and, 783
pertinent terminology, 782
relationship to body weight, 789
under various conditions, 782
various activities and, 782-785
Ventricular contraction
atrial filling and, 788, 794
premature, 781
valve closure and, 780-781
Venturi meter
principles of, 1 297, 1 298
Venules
leaks in, 1014
Vertebrae
blood supply to, 1655
Vertebral arteries
distribution, 1655
Vibrocardiogram
cardiac cycle and, 777
Vis a fronte
analysis of venous return, 1 102
definition, 854
schematic representation, 856
Vis a tergo
analysis of venous return, 1 102
Visco-elasticity
application of term, 803
behavior with stretch, 866
definition, 867
description of, 866
elastic tissue property, 805, 866, 877
illustration, 867
smooth muscle property, 866, 872, 878,
879
tendon, lack of, 805
time-dependent, in umbilical artery,
819
vessel compliance and, 820
Viscous flow patterns
definition, 853
Vitamin B12
transport in lymph, 1048
Water
bound and free in interstitial fluid, 978
diffusion rates, 1012
hydrodynamic flow and diffusion of,
1004
in bladder, 1 141
in stomach, 1 141
in uterus, 1 141
in vascular tissue, 1 141
permeability of muscle capillaries to,
1013
ratio to extracellular space, 1141
relation to sodium hypertension, 11 43
splanchnic blood volume and, 1406
Wedge pressure
use of, 1687
values for, 1687
Wetterer, E.
Methods of measuring blood flow,
1277-1324
Wiedeman, Mary P.
Patterns of the arteriovenous pathways,
89'~933
Windkessel
analogue of arterial system, 825, 847,
848, 873
definition, 873
Windkessel model
arterial system
discussion of, 847
pressure-flow relationship, 847
Work
blood flow, oxygen consumption and,
'3/8
Wyamine : see Mephentermine
X wave
atrial filling and, 786
Xanthines
effect on heart, 1561
Y wave
opening of atrioventricular valves and,
787
Yield pressure
definition, 1477
Young's modulus
definition 809
Zinc sulfate
mean activity coefficients, 1 1 37
tllRlKHl
111