Pathophysiology of Hydrops Fetalis

Michael Apkon

Hydrops fetalis occurs when the rate of interstitial fluid production by capillary ultrafiltration
exceeds the rate of interstitial fluid return to the circulation via lymphatic vessels. Developmental
differences in the microcirculation and lymphatic system of the fetus, as compared with mature
subjects, renders the fetus susceptible to interstitial fluid accumulation. These differences include
greater capillary permeability, more compliant interstitial compartment, and greater influence of
venous pressures on lymphatic return. The balance between interstitial fluid production and
removal is most commonly disrupted as a consequence of homeostatic mechanisms serving to
preserve adequate systemic delivery of metabolic substrate when cardiocirculatory function is
impaired. The pathophysiology of two conditions of impaired cardiocirculatory function, atrial
tachycardia and severe anemia, serve as examples of the mechanisms by which these homeostatic
mechanisms perturb the balance of interstitial fluid movement.
Copyright 9 1995 by W.B. Saunders Company

H ydrops fetalis, the excessive accumulation a physiological framework to examine the factors
of interstitial fluid in the fetus, is a condi- governing fluid movement between the vascular
tion that results in considerable rates of morbidity and interstitial spaces and will explore, in partic-
and mortality. The list of abnormalities associated ular, the unique features of the fetal microcir-
with hydrops is protean, encompassing congenital culation that facilitates fluid movement in utero.
malformations, infections, tumors, and acquired Additionally, the article discusses the compen-
disorders. Although many disorders have a clear satory mechanisms that preserve adequate oxy-
association with hydrops fetalis, in many cases, gen and nutrient delivery to the tissues at the
the actual cause of excessive fluid accumulation expense of increased interstitial fluid accumu-
remains obscure. One might consider interstitial lation. Finally, the pathophysiological mecha-
fluid to be a transudate of plasma, which, via nisms in two experimental models of hydrops fe-
lymphatic vessels, is eventually returned to the talis will be considered.
vascular space (Fig 1). Accordingly, hydrops may
be considered a result of an imbalance between
the rate of interstitial fluid formation and the Regulation o f Fluid M o v e m e n t Between
rate at which this fluid is returned to the circu- Vascular and Interstitial Spaces
lation. Thus, hydrops might result from any con-
Transcapillary Filtration
dition that accelerates the rate of transudation
from the vascular space or delays the return of The rate of interstitial fluid formation is deter-
lymph to the circulation. This disrupted balance mined by the balance of forces governing water
of fluid accumulation and clearance from the in- movement across the capillary endothelium, as
terstitial space may represent a primary distur- originally described by Starling I (Fig 1). The
bance in the mechanisms governing fluid flux, as driving forces for water extravasation from the
in the case of lymphatic malformations. More vascular space is the hydrostatic pressure within
frequently, however, the balance of fluid move- the capillary (Pc) and the colloid oncotic pressure
ment is disrupted as a consequence of compen- of the interstitial fluid (a'i). Similarly, the forces
satory mechanisms invoked by the fetus to pre- driving the return of water from the interstitial
serve adequate delivery of oxygen and other vital
suhstrates to the tissues in the face of impaired From the Division of Critical Care, Departmentof Yale University
cardiocirculatory function. One approach in School of Medicine, New Haven, CT.
planning a rational diagnostic and therapeutic Address reprint requeststo Michael Apkon, MD, PhD, Department
of Pediatrics, Yale UniversitySchool of Medicine, PO Box 208064,
strategy in hydrops fetalis is to consider the New Haven, CT 06520-8064.
pathophysiological basis of interstitial fluid ac- Copyright 9 1995 by W.B. Saunders Company
cumulation. Therefore, the author will develop 0146-0005/95/1906-0009505.00/0

Seminars in Perinatology, Vol 19, No 6 (December), 1995: pp 437-446 437

488 Michael Apkon

Artery Interstitium
O0
0
0 00 o
(Pi)o
0 O00r 0

Capillary
(Pc) o 000 ~176
o o o
o

Oo~C~176
o
Vein
L Lymphatic

Figure 1. Interstitial fluid balance. Interstitial fluid is formed by ultrafiltration of plasma at the capillary. Capillary
ultrafiltration is driven by the difference between capillary (Pc) and interstitial (Pi) hydrostatic pressures. Fluid
reabsorption by the capillary is driven by the difference between the capillary 0re) and interstitial (Tri) colloid
oncotic pressures, measures of the concentrations of oncotically active molecules, or, the reflection coefficient,
and Lp, the hydraulic permeability, are intrinsic properties of the capillary wall that characterize the efficacy of
such driving forces to produce water flux (.Iv). Interstitial fluid is returned to the circulation via the lymphatic
system. The outflow pressure of the lymphatic pressure is the venous pressure. The "manometers" indicate the
relative pressures in each compartment.

space to the intravascular space is the hydrostatic Jv = CFC(Ap - ~A~r)

pressure o f the interstitium (Pi) and the plasma
colloid oncotic pressure (a'D. The colloid oncotic where Ap is the net hydrostatic pressure gradient
pressure in the vascular and interstitial spaces (Pc - Pi) and ATr is the colloid osmotic pressure
depends on the concentration o f osmotically ac- gradient between the capillary and interstitial
tive particles in the compartment, and the efficacy fluid (Trc - 7ri).
with which oncotic pressure differences influence Several features o f the fetal microcirculation
water m o v e m e n t depends on the extent to which may facilitate transcapillary water m o v e m e n t
m o v e m e n t o f osmotically active molecules across c o m p a r e d with adult subjects. 2 First, fetal cap-
the capillary is restricted (as described by ~, the illaries a p p e a r to be m o r e p e r m e a b l e to plasma
colloid osmotic reflection coefficient). The less proteins, resulting in a lower reflection coeffi-
p e r m e a b l e the e n d o t h e l i u m is to a particular cient for oncotically active solute. The effect of
molecule (ie, the higher the tr), the greater the this enhanced solute permeability is that for any
effect o f colloid oncotic pressure differences. A given solute concentration difference across the
corollary of this relation is that a m o r e permeable capillary endothelium, the colloid oncotic pres-
endothelium will allow a greater solute flux to sure difference will drive water less effectively
accompany, via solvent drag, water movement. f r o m the interstitium to the vascular space. Thus,
The net rate o f fluid m o v e m e n t depends on the transcapillary flux and the distribution o f body
capillary surface area and its hydraulic conduc- water will be less sensitive to alterations in plasma
tivity (Lp). The p r o d u c t o f capillary surface area or interstitial colloid osmotic pressure. This, in
and hydraulic conductivity, r e f e r r e d to as the part, explains the finding that plasma protein re-
capillary filtration coefficient (CFC) provides a duction in chronically instrumented fetal sheep
measure of units o f volume translocated p e r unit does not affect body water content. 3 Second, the
time per millimeters o f m e r c u r y driving force. capillary filtration coefficient is fivefold higher
Thus, Jv, the fluid flux across the capillary, may than that of adults. The increased capillary fil-
be described by the following equation: tration coefficient in the fetus results in an in-
 Pathophysiologyof Hydrops 439

creased water flux for a given driving force. flow (normalized for body mass) in fetal sheep is
Third, the interstitial space o f the fetus is more four to five times greater than those o f adult
compliant than that of more mature subjects. sheep. In addition to differences in lymphatic ca-
That is to say, that the interstitial space o f the pacity between fetal and mature subjects, there
fetus is capable o f receiving a greater a m o u n t o f are differences in the regulation o f lymphatic
fluid with a smaller increase in the interstitial hy- flow. For example, lymphatic flow rates d e p e n d
drostatic pressure. Because it is the interstitial on the outflow pressure for the lymphatic system,
hydrostatic pressure that opposes the capillary which is the venous pressure (Fig 2). In mature
hydrostatic pressure and drives water from the subjects, lymphatic flow is relatively constant until
capillary, a given increase in capillary hydrostatic the outflow pressure increases to approximately
pressure may result in greater interstitial fluid 8 mm Hg. At greater pressures, lymphatic flow
accumulation before it is checked by a corre- falls and eventually ceases when the outflow
sponding increase in interstitial hydrostatic pres- pressure nears 25 mm Hg. In contrast, fetal lym-
sure. phatic flow begins to decrease when the outflow
In the short term, facilitated transcapillary pressure is greater than 0 m m H g and ceases
flow has important consequences in blood vol- when the outflow nears 10 m m Hg. Because the
ume regulation. For example, transcapillary flow normal fetal central venous pressure is typically
from the interstitium is likely an import mecha- greater than 0 mm Hg, basal lymph flow is limited
nism restoring blood volume after acute hem- in the normal fetus and is expected to be limited
orrhage. MeUander and others 4'5 have shown that further whenever central venous pressure in-
in adult cats, acute hemorrhage induces altera- creases. The factors governing lymphatic flow
tions in microcirculatory pressures, which results through abnormal lymphatic structures are more
in reabsorption rather than filtration by the cap- poorly understood.
illaries. This "autotransfusion" probably con-
tributes to restoring the blood volume and
preserving cardiac output. In adult sheep, res- Circulatory Failure as a Cause o f Hydrops
toration o f full blood volume after a 30% hem-
Although increases in capillary permeability, de-
orrhage requires 24 to 48 hours. 6 In contrast,
creases in plasma colloid oncotic pressure, and
full blood volume o f fetal sheep is restored over
impaired lymphatic function may each contribute
3 to 4 hours. 7 The more rapid recovery o f blood
to interstitial fluid accumulation in the hydropic
volume in the fetal subjects is consistent with a
fetus, circulatory dysfunction with associated in-
greater capillary filtration coefficient and a
creases in venous pressures may be a more com-
greater interstitial compliance. Similarly, facili-
mon mechanism causing noninunune hydrops. ~~~4
tated transcapillary flow contributes to blood
Elevation in venous pressure is a manifestation
volume regulation during blood volume expan-
of several o f the normal homeostatic mechanisms
sion. For example, saline infusion results in blood
that preserve organ perfusion in the face o f a
volume expansion. However, this expansion is
critical imbalance between the supply and de-
transient as fluid redistributes across the capil-
mand for oxygen and other vital nutrients. An
lary. In adult sheep, 30% o f administered saline
understanding o f the pathophysiology o f hydrops
remains in the vascular space 30 minutes after
u n d e r such conditions o f supply-demand imbal-
infusion. In fetal sheep, in contrast, only 6% o f
ance requires examination o f the determinants
the infused saline remains in the vascular space, s
o f substrate delivery to the tissues as well as an
This suggests that for a given amount o f fluid
understanding o f the homeostatic mechanisms
added to or retained by the fetus, a smaller frac-
contributing to elevations o f venous pressures.
tion will reside in the vascular space and a larger
fraction will reside in the interstitial space. Circulatory Failure as an Imbalance
Lymphatic Return Between Substrate Supply and Metabolic
Demand
Enhanced capillary filtration o f water to the in-
terstitial space predicts that lymphatic flow in the Adequate delivery o f metabolic substrate to the
fetus must be increased also. Indeed, Brace et tissues depends not only on the overall cardiac
al 9A~ have shown that the thoracic duct lymph output, the concentration o f substrate within the
440 Michael Apkon

0.31 , , , I [ l I I I
i
[

!iiiiiiiiiiiii!iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii
.0~176 ... ..... ..............

,r
iiiiiii!i!iiiiii
iiiiiiii!!i
ili
.c:_ 0.2
E
Fetal
E

t~

0
0.1 Figure 2. Left thoracic
Q. duct lymph-flow in ovine
Adult

/
E fetus and adult. Measure-
._1 ments were made in animals
with chronically implanted
...........~176 , . . . - . . . . . . . . - . . . . . . . . . . . , lymphatic and vascular
catheters. The normal lym-
phatic outflow pressure is
the central venous pressure
0 k- which is typically 3 to 4 mm
-20 -10 0 10 20 30 40 Hg in the ovine fetus. Data
are mean -+2 X SE. (Re-
Outflow pressure (mmHg) printed with permission. 2)

blood and the appropriateness of the distribution cardiac o u t p u t is the p r o d u c t o f heart rate and
o f blood flow a m o n g various organs, but also on stroke volume. Furthermore, stroke volume is
the metabolic demands o f the tissue. Table 1 lists determined by the degree o f cardiac filling (ie,
examples o f clinical conditions that have been the end diastolic volume) and the effectiveness
associated with n o n i m m u n e hydrops and that o f contraction (ie, the ejection fraction). Thus
may decrease cardiac o u t p u t or increase the de- deficiencies in heart rate, cardiac filling, or con-
mand for blood flow. Conditions are g r o u p e d tractile function could impair cardiac output.
according to the primary mechanism by which I m p a i r m e n t o f cardiac filling may result from
supply-demand imbalance occurs. Recall that an inadequate intravascular volume, an impedi-

Table 1. Causes of Supply-Demand Imbalance for Metabolic Substrate Delivery

Disturbance Mechanism Examples
Decreased cardiac output Inadequate cardiac filling Decreased ventricular compliance,
pericardial effusion, tachyarrythmias,
hemorrhage, hepatitis or peritonitis
causing extravasation of fluid from
vascular space
Inadequate or impeded ejection Myocarditis, myocardial infarction, asphyxia,
polycythemia, premature closure of the
ductus arteriosus, valve dysfunction
Inadequate heart rate Congenital heart block
Increased demand for flow Decreased oxygen content Anemia, bypoxemia
Maldistribution of flow Arteriovenous malfornaation
Increased metabolic rate Thyrotoxicosis
Modified with permission)8
 Pathophysiology of Hydrops 441

ment to venous return, or to a decrease in ven- for increased flow. Maldistribution o f blood flow,
tricular compliance. A relative deficiency in in- as might occur in the presence o f an arteriove-
travascular volume may exist as a consequence nous malformation, also creates a need for in-
o f either fluid loss from the vascular space or creased total blood flow because a portion o f the
from an increase in vascular capacity. Fluid may total cardiac o u t p u t is distributed to provide ex-
be lost from the vascular space to the interstitial cessive flow to some tissues at the expense o f
space if capillary filtration increases as a result flow to o t h e r tissues. Finally, increases in meta-
o f either an increase in capillary permeability or bolic rate, as may occur in metabolic disturbances
a decrease in plasma colloid oncotic pressure as such as thyrotoxicosis, must be met by increases
discussed previously. Alternatively, blood may be in substrate delivery.
lost from the vascular space as a result o f fetal- In the fetus as well as the mature subject, a
maternal or twin-twin transfusion. Furthermore, n u m b e r o f homeostatic responses are activated
fetal hemorrhage (ie, intracranial hemorrhage) locally and systemically by inadequate cardiac
causes blood loss from the vascular space. output. 17.18 In addition to their beneficial effects,
Even when intravascular volume is normal, these compensatory mechanisms may also be
cardiac filling will be diminished if venous return maladaptive in so far as they contribute to fluid
is impeded. Pericardial effusions, cardiac tumors, retention by the fetus and extravasation o f fluid
and subendocardial fibroelastosis each decrease from the vascular space, thereby contributing to
the effective ventricular compliance, resulting in the development o f hydrops.
a smaller end diastolic ventricular volume for a
given venous pressure. Tachyarrhythmias also Compensatory Mechanisms
impart an impediment to cardiac filling by lim- Local and systemic cardiovascular regulation
iting the duration o f diastole. preserves homeostasis via three mechanisms.
Given adequate cardiac filling, ejection o f First, local mechanisms enhance the efficiency o f
blood may be impaired by depressed cardiac oxygen extraction and use by the tissues. Second,
muscle function as may occur from in utero systemic and local mechanisms act together to
myocarditis, myocardial infarction or hypoxemia. redistribute blood flow to more metabolically ac-
Moreover, ejection may be impaired when ven- tive organs such as the brain and heart. Third,
tricular outflow is impeded. Thus, valvular mal- systemic mechanisms act to enhance cardiac out-
formations such as calcified aortic valve or ab- put by augmenting blood volume and myocardial
normalities of the puimonic valve compromise performance.
the ability o f the ventricle to eject. Premature Local mechanisms enhance the efficiency o f
closure o f the ductus arteriosus might also be oxygen extraction in part by recruiting previously
expected to compromise right ventricular ejec- closed capillaries. 19 In adult subjects, capillary
tion by forcing the total right ventricular o u t p u t recruitment allows enhanced oxygen extraction
to be ejected through the high vascular resistance by increasing surface area for oxygen diffusion,
of the fetal pulmonary circulation. In fetal lambs, by decreasing the distance between the capillary
acute ductal occlusion caused a decrease in right and tissue, and by increasing the transit time o f
ventricular output (RV), increases in RV systolic blood cells within the tissues. It is not known to
and diastolic dimensions, and tricuspid valve re- what extent the capillary bed o f the fetus is fully
gurgitation) 5 Interestingly, premature closure o f recruited at rest or if capillary recruitment is an
the foramen ovale may also create a relative de- important homeostatic mechanism during de-
ficiency in RV function by necessitating that the velopment.
RV discharge the total systemic venous return as At times o f insufficient cardiac o u t p u t for
right to left transatrial flow is blocked) 6 metabolic demands, flow is redistributed by a
Decreases in the concentration o f metabolic combination o f local and systemic mechanisms.
substrate, particularly oxygen, in the blood cre- Many organs are able to locally regulate their
ates a requirement for more blood flow to pro- blood flow by varying the resistance o f the con-
vide a constant amount o f substrate delivery. ducting vessels. This regulation occurs as a result
Thus, decreases in hemoglobin concentration or of metabolic feedback from the tissues to the
in hemoglobin saturation result in less oxygen conducting vessels and by direct effects o f per-
delivery per unit o f blood and hence a demand fusion pressure and blood flow on the blood ves-
442 Michael Apkon

sels themselves. As organs differ in their ability cortex, which in turn p r o m o t e s sodium reab-
to regulate flow in this manner, local mechanisms sorption. The effect o f these responses is that
will serve to redistribute blood flow to those or- fluid is retained by the kidneys, in p a r t aug-
gans with m o r e potent autoregulation. Systemic menting the circulating blood volume.
mechanisms involving neural and humeral factors Blood volume may also be a u g m e n t e d if trans-
also serve to redistribute b l o o d flow. Specifically, capillary hydrostatic forces are altered to favor
stimulation o f the sympathetic nervous system fluid r e a b s o r p t i o n f r o m the interstitium. Mean
via activation o f low pressure atrial stretch re- capillary pressure could decrease as a direct result
ceptors and arterial b a r o r e c e p t o r s results in pe- o f decreases in arterial or venous pressures. Al-
ripheral vasoconstriction. Similarly, plasma con- ternatively, m e a n capillary pressure could de-
centrations of vasopressin, angiotensin II, and crease even though arterial pressure is increased
other n e u r o e n d o c r i n e factors are increased and if the arterial pressure is dissipated by arteriolar
cause arteriolar constriction. As vascular beds constriction, as occurs with sympathetic nervous
differ in their sensitivity to sympathetic stimula- system stimulation. It is likely that e n h a n c e m e n t
tion, blood flow is redistributed to those tissues in transcapillary reabsorption by these mecha-
with a lower sensitivity such as the heart and nisms contributes to the restoration o f fetal blood
brain. In chronically i n s t r u m e n t e d fetal sheep, it volume after a c u t e h e m o r r h a g e . 26"27
has b e e n possible to show that these mechanisms Increases in blood volume result in increases
function during gestation. The b a r o r e c e p t o r re- in myocardial filling as blood is distributed a m o n g
flex, for example, a m e a s u r e o f autonomic ner- different c o m p a r t m e n t s (ie, arteries, capillaries,
vous system function, is active t h r o u g h o u t the veins, and cardiac chambers) according to the
second half of gestation, 2~ although the sensitivity relative compliance and capacitance o f each
o f the reflex to changes in mean arterial blood c o m p a r t m e n t . Even if blood volume is not in-
pressure seems to increase with age. It is also creased, myocardial filling can be increased by
possible to show increases in fetal plasma con- redistributing blood a m o n g these various com-
centrations of n o r e p i n e p h r i n e , renin, and vaso- partments. The venous system has a large capac-
pressin in response to circulatory stress such as itance that is regulated by the sympathetic ner-
hemorrhage. 7'21 That these mechanisms are im- vous system. Increases in sympathetic tone cause
p o r t a n t to cardiovascular homeostasis is sup- venoconstriction, which increases venous pres-
p o r t e d by the finding that blockade o f angioten- sures and redistributes blood f r o m the veins to
sin II (whose formation is stimulated by increased the heart. Such increases in fetal venous pressure
plasma renin activity) or vasopressin activity have b e e n shown during circulatory stress such
magnifies the decrease in fetal m e a n arterial as hypoxia. 2s
pressure that occurs with h e m o r r h a g e . 22 The net Increases in myocardial filling are translated to
effect o f these regulatory mechanisms in the fetus increases in cardiac output because of the Frank-
is that during times o f circulatory stress such as Starling mechanism, 29 which relates changes in
hypoxia or h e m o r r h a g e , b l o o d flow and oxygen end-diastolic fiber length to changes in stroke
delivery to the brain, heart, and adrenal glands volume. T h e t e r m preload reserve has b e e n used
are preserved at the expense of that to the kidney, to describe the augmentation in cardiac output
gut, and m u s c l e . 23'24 that may be obtained with increases in myocardial
The same systemic mechanisms that serve to filling. In the fetal lamb, as in the adult, increases
redistribute flow also act to enhance cardiac out- in mean atrial pressure or end-diastolic pressure
put by augmenting blood volume and myocardial result in increases in right and left ventricular
performance. For example, renal arteriolar va- output 3~ (Fig 3). However, the increase in ven-
soconstriction decreases renal perfusion. How- tricular o u t p u t for a given increase in filling
ever, glomerular filtration may be maintained by pressure is substantially diminished at filling
selective constriction o f the efferent arteriole. 25 pressures greater than approximately 3 to 8 m m
The kidney responds to this decrease in perfusion Hg. W h e t h e r or not the Frank-Starling mecha-
by increasing the proximal reabsorption o f the nism provides preload reserve in the fetus de-
glomerular filtrate. Additionally, vasopressin acts pends on the normal mean atrial pressure relative
to increase water r e a b s o r p t i o n and angiotensin to this inflection point in the output-pressure re-
II stimulates aldosterone release f r o m the adrenal lationship. It appears, at least in fetal lambs, that
 Pathophysiology of Hydrops 443

1.5

. . - - - - - RV
I

_._...-- L V
Figure 3. Ventricular stroke
volume as a function of mean
atrial pressure. These com-
posite left (LV) and fight
ventricular (RV) function
curves measured in ovine
fetuses were obtained by
measuring stroke volumes
ej

o.,//
during rapid blood rein- /
fusion after phlebotomy.
Lines represent the "best i t t i I J * i i I i * * , I
fits" to the rising and pla-
teau segments of the curve. 0 5 10 15
(Reprinted with permission?~) Mean atrial pressure (torr)

normal mean atrial pressures are sufficiently high tion surface area. Increases in venous resis-
that preload reserve is limited. tance that are p r o p o r t i o n a t e l y g r e a t e r than in-
creases in arteriolar resistance will also increase
Maladaptive Effects of Compensatory mean capillary pressure. T h e s e increases in
Mechanisms mean capillary pressure are maladaptive in that
Although the c o m p e n s a t o r y mechanisms dis- they result in extravasation o f fluid f r o m the
cussed previously preserve balance between vascular space and interstitial fluid deposition.
substrate supply and d e m a n d , at least in the Such decreases in blood volume have b e e n
most metabolically active tissues, homeostasis shown experimentally in fetal lambs d u r i n g cir-
occurs at the expense o f increased venous culatory stress such as o c c u r s in fetal hyp-
oxia.2S,s7
pressures, impaired organ function, and ulti-
mately interstitial fluid accumulation. Eleva- In addition to alterations in hydrostatic pres-
tion in venous pressures has b e e n shown in sure, colloid oncotic pressure has also been sug-
h u m a n I~'s~ and lamb s436 fetuses with hydrops gested to be altered as a consequence o f circu-
secondary to impaired circulatory function. latory compromise. In adults, for example,
Increases in venous pressure may c o n t r i b u t e to venous congestion and decreased hepatic per-
interstitial fluid accumulation by two mecha- fusion have been suggested to impair synthesis
nisms, as discussed previously: (1) increasing o f albumin, ss the predominant oncotically active
the mean capillary hydrostatic pressure, caus- plasma protein. Moreover, hepatic synthetic
ing increased capillary filtration and, (2) in- function may be further reduced by increased
creasing the outflow pressure for lymphatic extramedullary hematopoeisis in response to
return, causing a decrease in lymphatic flow. chronic fetal hypoxia.
Even if venous pressures are not increased, Increases in capillary permeability could
mean capillary pressures can be increased as a also contribute to increased transcapillary fluid
result o f local and systemic c o m p e n s a t o r y m o v e m e n t at times o f circulatory compromise.
mechanisms for insufficient circulation. For Capillary permeability may increase as a result
example, precapillary vasodilation could ele- o f hypoxia-induced endothelial injury. Alter-
vate mean capillary pressure by b e t t e r trans- natively, capillary permeability may be m o d u -
mitting arteriolar pressure to the capillaries and lated by circulating n e u r o h o r m o n a l factors re-
could recruit previously closed capillaries, leased via homeostatic mechanisms o r by the
contributing to an increase in capillary filtra- host defense against infection. For example,
444 Michael Apkon

atrial natriuretic factor, released in response Given the decrease in cardiac o u t p u t during
to atrial distension, may directly increase cap- pacing, it is not surprising that the arterial Po~
illary permeability. 39 decreases somewhat from approximately 18 to
15 mm Hg. Decreases in arterial Po2 were not
observed in similar experiments by Gest et a l Y
Experimental Models of Hydrops Fetalis It seems unlikely that this modest degree o f hyp-
Atrial Tachycardia oxia resulted in increases in capillary permeability
because plasma protein, albumin, and Na § con-
Recognition o f i n t r a u t e r i n e supraventricular centrations do not change significantly with pac-
tachycardia as a cause o f nonimmune hydrops ing. Gest et a135 measured the turnover rate of
suggested that hydrops might be c r e a t e d in ex- radiolabeled albumin in the plasma and found
perimental subjects by atrial pacing. Stevens et no evidence for increased capillary permeability.
al 4~ adapted the chronically instrumented fetal Stevens et al 4~ f o u n d that the histopathology
lamb preparation to electrically pace the left o f those fetuses who had received atrial pacing
atrium for prolonged periods o f time. Atrial pac- suggested that edema formation is a result o f ve-
ing at 300 bpm resulted in evidence o f fluid ac- nous congestion: many animals had cardiomegaly
cumulation in all animals studied: 1 o f 13 study or hepatomegaly. Elevations in inferior vena cava
animals developed massive edema (hydrops)~ Ev- pressure were showed directly in the study by
idence o f fluid accumulation was not observed Gest et al. 35 Interestingly, the increase in venous
in a set o f control animals that did not receive pressures does not necessarily reflect an increase
atrial pacing. in total blood volume. Plasma volume may ac-
It is likely that these pathological changes re- tually decrease somewhat with a constant red
sult from the decrease in cardiac o u t p u t that ac- blood cell volume. This suggests that the com-
companies rapid atrial pacing. In this regard, it pensatory mechanisms preserving tissue perfu-
is important to distinguish between the conse- sion during tachycardia redistribute the blood
quences o f abnormal rhythms with rapid atrial between the various vascular spaces and redis-
rates, and the consequences o f tachycardia per tribute fluid from the vascular to the interstitial
se. Anderson et al have distinguished between space. Furthermore, increases in venous pressure
the effects of sinus tachycardia and rapid atrial must reflect increases in v e n o m o t o r tone and de-
pacing in chronically catheterized fetal lambs. 41'42 creased venous compliance. This does not mean
In fetal lambs, increases in spontaneous heart that fluid accumulation by the fetus does not take
rate were accompanied by increases in right ven- place. Rather, it suggests that if fluid retention
tricular end-diastolic dimension and o u t p u t at a occurs it is poorly effective at increasing blood
constant stroke volume, but decreases in left volume and restoring circulatory function.
ventricular end-diastolic dimension and stroke The elevation in venous pressure contributes
volume; left ventricular o u t p u t also increased. to edema not only by forcing fluid from the vas-
However, when the heart rate was increased by cular space, but also by impeding the return of
atrial pacing, a decrease in stroke volume and fluid via the lymphatics. Gest et a143 found that
end-diastolic dimension was observed for both if the thoracic duct lymph flow in fetal lambs was
ventricles. Interestingly, ventricular o u t p u t only collected against a pressure equal to the baseline
decreased significantly in the ventricle ipsilateral central venous pressure, then electrically induced
to the paced atrium. The alterations in myocar- tachycardia caused a 50% increase in lymph flow.
dial filling appear to result from an inadequate If, however, the lymph flow outlet pressure was
diastolic intervalbecause the rate-related changes increased to a pressure equal to the venous pres-
were reversed during test beats introduced at sures measured during atrial pacing, then lymph
longer interbeat intervals. One reason t h a t the flow decreased to baseline rates. Thus it appears
cardiac o u t p u t is better preserved during spon- that in tachyarrhythmia-induced hydrops, edema
taneous tachycardia is because the mechanisms formation results from (1) increases in microcir-
that normally accelerate the heart beat, such as culatory pressure resulting f r o m increased ven-
the sympathetic nervous system, also augment o m o t o r tone a n d / o r decreased ventricular com-
cardiac filling (by increasing venomotor tone) and pliance, and (2) impairment o f lymphatic flow
the force o f contraction. resulting from increased venous pressure. Inter-
 Pathophysiologyof Hydrops 445

estingly, electrically induced tachycardia is par- the rates o f transcapillary fluid movement, it is
ticularly well suited for examining the time course not surprising that hydrops is a c o m m o n conse-
for the development and reversal o f hydrops, be- quence o f many different disease processes. Even
cause the circulatory disturbance can be initiated if one considers hydrops to be most commonly
and terminated instantaneously. Hydrops ap- a consequence o f circulatory insufficiency, the
pears to develop within several hours to a few specific entities leading to circulatory compro-
days after initiating pacing, and many o f the cir- mise are numerous. Understanding the physiol-
culatory disturbances are rapidly reversible after ogy o f interstitial fluid accumulation and the
returning to sinus rhythm. 34's6 pathophysiology o f specific diseases aids in the
advancement o f diagnostic and therapeutic ap-
Anemia proaches to prenatal care. The challenge o f fu-
ture investigations is to develop therapeutic
Anemia as a result o f alloimmunization, hemor-
strategies aimed not only at reversing the under-
rhage, or red blood cell dyscrasia is a c o m m o n
lying disease, but also to ameliorate the conse-
cause o f hydrops. Recently, Blair et al have suc-
quences o f interstitial fluid accumulation.
cessfully produced anemia-associated hydrops in
fetal lambs by partial exchange transfusion. 44 This
experimental model is beginning to provide im- References
portant insight into the pathogenesis o f fluid ac- 1. Starling EH: On the absorption of fluids from the con-
cumulation during anemia. As in electrically in- nective tissue spaces. J Physiol 19:312-326, 1896
duced tachycardia, hydrops in anemia seems to 2. Brace RA: Fluid distribution in the fetus and neonate,
be associated with an increase in venous pressure. in Polin RA, Fox WW (eds): Fetal and Neonatal Physi-
ology. Philadelphia, PA, Saunders, 1992, pp 1288-1298
Plasma colloid oncotic pressure increased some- 3. Moise AA, Gest AI,, Weickmann PH, et al: Reduction in
what in all the anemic animals, suggesting that plasma protein does not affect body water content in
extravasation o f fluid from the vascular space oc- fetal sheep. Pediatr Res 29:623-626, 1991
curs in anemia as well as in electrically induced 4. Mellander S and Lewis DH: Effect of hemorrhagic shock
on the reactivity of resistance and capacitance vessels
tachycardia. It is unlikely that these alterations
and on capillary filtration transfer in cat skeletal muscle.
in oncotic forces contribute to water accumula- Circ Res 13:105-118, 1963
tion in anemia, however, because the observed 5. Maspers M, BjornbergJ, Grande P-O, et al: Sympathetic
increase in plasma oncotic pressure would be ex- alpha-adrenergic control of large-bore arterial vessels,
pected to impede water loss to the interstitium. arterioles and veins, and of capillary pressure and fluid
exchange in whole-organ cat skeletal muscle. Acta Physiol
Interestingly, the likelihood o f developing hy-
Scand 138:509-521, 1990
drops was related to the rapidity with which ane- 6. GrimesJM, Buss LA, Brace RA: Blood volume restitution
mia was produced. after hemorrhage in adult sheep. AmJ Physio1253:R541-
544, 1987
7. Brace RA, Cheung CY: Fetal cardiovascular and endo-
Sumnlal~ crine responses to prolonged fetal hemorrhage. Am J
Physiol 251 :R417-424, 1986
Interstitial fluid accumulates in the fetus when- 8. Brace RA: Fetal blood volume responses to intravenous
ever the rate o f fluid extravasation from the vas- saline solution and dextran. Am J Obstet Gynecol 147:
777-781, 1983
cular space is greater that the rate at which fluid 9. Brace RA, Valenzuela GJ: Effects of outflow pressure
is returned to the venous system by the lym- and vascular volume loading on thoracic duct lymph flow
phatics. The fetus is particularly susceptible to in adult sheep. Am J Physiol 258:R240-244, 1990
such interstitial fluid accumulation by virtue o f 10. Brace RA: Effects of outflow pressure on fetal lymph
developmental differences that facilitate fluid flow. Am J Obstet Gynecol 160:4944497, 1989
11. Kleinman CS, Donnerstein RL, De Vore GR, et al: Fetal
extravasation and cause lymphatic return to be echocariography for evaluation of in utero congestive
more potently influenced by lymphatic outflow heart failure: A techniques for study of nonimmune hy-
pressure. The delicate balance between fluid en- drops. N EnglJ Med 306:568-575, 1982
try and egress from the interstitium can be easily 12. Weiner CP: Umbilical pressure measurement in the eval-
upset as a result o f normal homeostatic mecha- uation of nonimmune hydrops fetalis. Am J Obstet Gy-
necol 168:817-823, 1993
nisms that preserve organ nourishment in the 13. Allan LD, Crawford DC, Sheridan R, et al: Aetiology of
face of circulatory insufficiency. Given the num- non-immune hydrops: The value of echocardiography.
ber of independent parameters that determine B r J Obstet Gynaecol 93:223-225, 1986
446 Michael Apkon

14. Holzgreve W, Curry CJR, Golbus MS, et al: Investigation diating fetal blood volume decrease during acute hypoxia.
of nonimmune hydrops fetalis. AmJ Obstet Gyneco1150: A m J Physiol 253:H927-932, 1987
805-812, 1984 29. Frank O: On the dynamics of cardiac muscle. Am Heart
15. Tulzer G, Gudmundsson S, Rotondo KM, et al: Acute J 58:467-478, 1959
fetal ductal occlusion in lambs. Am J Obstet Gynecol 30. Kirkpatrick SE, Pitlick PT, NaliboffJ, et al: Frank-Starling
165:775-778, 1991 relationship as an important determinant of fetal cardiac
I6. Coulson CC, KullerJA: Nonimmune hydrops fetalis sec- output. A m J Physiol 231:495-500, 1976
ondary to premature closure of the foramen ovale. Am 31. Gilbert RD: Control of fetal cardiac output during
J Perinatol 11:439-440, 1994 changes in blood volume. AmJ Physio1238:H80-86, 1980
17. Morin FC, Weiss KI: Response of the fetal circulation 32. Relier MD, Morton MJ, Reid DL, et al: Fetal lamb ven-
to stress, in Polin RA, Fox WW (eds): Fetal and Neonatal tricles respond differently to filling and arterial pressures
and to in utero ventilation. Pediatr Res 22:621-626, 1987
Physiology, Philadelphia, PA, Saunders, 1992, pp 620-
33. Johnson P, Sharland G, Allan LD, et al: Umbilical venous
628
pressure in nonimmune hydrops fetalis: Correlation with
18. Lister G, Apkon M, Fahey JT: Shock, in Emmanoulides
cardiac size. Am J Obstet Gynecol 167:1309-1313, 1992
TA, Riemenschneider TA, Allen HD, Gutgesell HP (eds):
34. Nimrod C, Davies D, Harder J, et al: Ultrasound evalu-
Heart Disease in Infants, Children, and Adolescents, ed ation of tachycardia-induced hydrops in the fetal lamb.
5. Baltimore, MD, Williams and Wilkins, 1994, pp 1725- A m J Obstet Gynecol 157:655-659, 1987
1746 35. Gest AL, Hansen TN, Moise AA, et al: Atrial tachycardia
19. Duling BR, Klitzman B: Local control of microvascular causes hydro ps in fetal lambs. Am J Physiol 258:H 1159-
function: Role in tissue oxygen supply. Annu Rev Physiol 1163, 1990
42:373-382, 1980 36. Gest AL, Martin CG, Moise AA, et al: Reversal of venous
20. Shinebourne EA, Vapaavuori EK, Williams RL, et al: De- blood flow with atrial tachycardia and hydrops in fetal
velopment of baroreflex activity in unanesthetized fetal sheep. Pediatr Res 28:223-226, 1990
and neonatal lambs. Circ Res 31:710-718, 1972 37. Brace RA: Fetal blood volume responses to acute fetal
21. Brace RA, Cheung CY: Fetal blood volume restoration hypoxia. A m J Obstet Gynecol 155:889-893, 1986
following rapid fetal hemorrhage. Am J Physiol 259: 38. Dunn GD, Hayes P, Breen KJ, et al: The liver in congestive
H567-573, 1990 hear failure: A review. AmJ Med Sci 265:174-189, 1973
22. Iwamoto HS, Rudolph AM: Role of renin-angiotensin 39. ValentinJ-P, RibsteinJ, Mimran A: Effect ofnicardipine
system in response to hemorrhage in fetal sheep. Am J and atriopeptin on transcapillary shift of fluid and pro-
Physiol 240:H848-854, 1981 teins. A m J Physiol 257:R174-179, 1989
23. Cohn HE, Sacks EJ, Heymann MA, et al: Cardiovascular 40. Stevens DC, Hilliard JK, Schreiner RL, et al: Supraven-
responses to hypoxemia and acidemia in fetal lambs. Am tricular tachycardia with edema, ascites, and hydrops in
J Obstet Gynecol 120:817-824, 1974 fetal sheep. Am J Obstet Gynecol 142:316-322, 1982
41. Anderson PAW, Killman AP, Mainwaring RD, et al: In
24. ItskovitzJ, Goetzman BW, Rudolph AM: Effects of hem-
utero right ventricular output in the fetal lamb: The effect
orrhage on umbilical venous return and oxygen delivery
of heart rate. J Physiol 387:297-316, 1987
in fetal lambs. A m J Physiol 242:H543-548, 1982
42. Anderson PAW, Glick KL, Killam AP, et al: The effect
25. RobillardJE, Weitzman RE, Burmeister L, et al: Devel-
of heart rate on in utero left ventricular output in the
opmental aspects of the renal response to hypoxia in the fetal sheep. J Physiol 372:557-573, 1986
lamb fetus. Circ Res 48:128-138, 1981 43. Gest AL, Bair DK, Vander Straten MC: Thoracic duct
26. Brace RA: Fetal blood volume responses to acute fetal lymph flow in fetal sheep with increased venous pressure
hemorrhage. Circ Res 52:730-734, 1983 from electrically induced tachycardia. Biol Neonate 64:
27. Brace RA: Fetal blood volume responses to fetal hae- 325-330, 1993
morrhage: Autonomic nervous contribution. J Dev 44. Blair DK, Vander Straten MC, Gest AL: Hydrops in fetal
Physiol 9:97-103, 1987 sheep from rapid induction of anemia. Pediatr Res 35:
28. Brace RA, Cheung CY: Role of catecholamines in me- 560-564, 1994