Feto-placental circulatory competence

Eur. J. Ohsret.

~wecol.

Reprod.

Bid,

21 (1986)

15-26

Elsevier

Feto-placental P.J.H.M. Ikpurtmentsof ’

circulatory

Reuwer

‘, W.C. Nuyen*, H.J.M. A.A. Haspels ’ and H.W.

Obstetrics,

’ Electronics,

Cwculutron.

’ Medical

Physrcs

Unrrvrsit,~ Hospital Accepted

for publication

competence Beijer 4, R.M. Bruinse ’

and ’ E~p~rrmmtol

Utrecht. 4 July

Heethaar

Lahorutor~Jor

‘. Prrrphrrrrl

The, Nrthwlmd.~ 1985

Summary

Doppler assessment of the pulsatility index (PI) of umbilical artery blood velocities offers promise as a simple. non-invasive method for early diagnosis of placental failure. The present report identifies the hemodynamic factors determining this PI, based on current knowledge of feto-pl.acental physiology. It is postulated that the umbilical artery PI is mainly determined by the ratio of capillary to arterial resistances in the fetal placental circulation. However, the biological PI variations on a minute by minute basis are caused by short-term variations in blood pressure pulsatility. The ratio of resistances declines markedly with advancing pregnancy. indicating that the site of predominant resistance determining the volume of placental blood flow gradually migrates from the peripheral capillary to the arterial compartment. After 28 wk of gestation, umbilical artery PI values decrease below 1, reflecting the normal reserve capacity of the functional capillary bed of the placenta. PI values above 1 in the last trimester reflect a critical enhancement of placental capillary resistance, which must impede blood flow to functional chorionic villi. It is concluded that this Doppler technique may provide a simple method for clinical assessment of placental circulatory compelence. Doppler:

umbilical

arteries:

pulsatility

index;

fetal

hemodynamxs:

placental

circulation

Introduction Non-invasive Doppler recordings of pulsatile blood velocities (umb. aa.) introduce new possibilities for an early diagnosis

Reprint Utrecht,

requests P.J.H.M. Keuwer, Department CathariJnesingel 101. 3511 GV Utrecht,

002%2243/86/$03.50

0 1986 Elsevier

Science

of Obstetrics The Netherlands.

Publishers

and

in umbilical of placental

Gynecology.

B.V. (Biomedical

Division)

Unlverslty

arteries failure

Hospital

16

- - - ---3yst

(5)

PI = s-d m

mean(m) diast

(d) time

Fig. 1. Determination

of the pulsatility

-

index (PI)

[l-4]. A simple and promising modality is the assessment of the pulsatility index (PI), defined as systolic minus diastolic divided by the mean blood velocity (Fig. 1). The PI from umb. aa. is an intra- and inter-observer reproducible variable [5], decreasing during normal pregnancy (Fig. 2). In growth-retarded fetuses, increased PI values can be found several weeks to months before fetal compromise is clinically suspected on the basis of current diagnostics [2]. The diagnostic significance of this new method is now under evaluation in several research centers (Commission of the European Communities, 1984 [6]).

x0-

PI

2.0I.2.6-I

2.4-

(_ 2.2-

\ \ \ \ \

---f

mean +stl n=23

\ \ \ ' \ \

2.0-

~

\ l\ \

I 1.6-

\ \ \ \

I 1.6-

\ \ \

I 1.4-

I.21 I l.O-

0.6-

0.6I 0.41.

-7_1-1-1~l-rpl :6yio

Fig. 2. Variation

24

20

in PI during

weeks 32

36

pregnancy.

40

17

Blood flow pulsatility in the umb. aa. is generally assumed to reflect placental resistance to flow. Anatomical support for this qualitative assumption was provided by Giles and coworkers [7], who demonstrated a significant correlation between increased pulsatility of umb. a. flow and microvascular obliterations in the placenta. However, these microscopical observations do not elucidate functional relationships and do not allow a quantitative interpretation on all the dynamic factors determining the PI in umbilical arteries. The present report defines such functional and numerical relationships between The hypothesis is based on current the PI and feto-placental hemodynamics. knowledge of fetal placental morphology, physiology and hemodynamics. Hemodynamic studies on peripheral arteries in adults have demonstrated a strong correlation between the PI of an artery and its downstream impedance to flow [Kg]. However, a simple extrapolation to the umb. aa. would disregard several unique characteristics of the fetal placental circulation. Fetal blood pressure, circulating blood volume and flow resistance change during fetal growth. The umbilical circulation receives about half the combined ventricular outputs, but, unlike other vascular beds, it possesses no nervous innervation nor responsive arterioles for auto-regulation of blood flow [lo]. Apparently, short-term variations in umbilical flow result only from changes in fetal cardiac output and blood pressure, while the gestational age-related decrease in placental resistance produces the longer-term increase in flow needed for fetal growth [ll]. Normal fetal growth and well-being is secured by the continuous extension of the placental vascular bed providing a vascular reserve capacity with wide margins of safety. Even a considerable obstruction of as much as 30% of the placental vascular bed is not necessarily associated with fetal malnutrition or hypoxia [12]. Regarding these unique anatomical and functional aspects, a more detailed description of the umbilical circulation is mandatory in order to comprehend its hemodynamics. The umbilical circulation For a full exposition of the vascular anatomy the reader is referred to the work of Bee [13], Crawford [14] and Ramsey and Donner [15]. It should be noted that the nomenclature is not employed uniformly among these authorities. We summarize only those morphological aspects with direct relevance to the hemodynamic interpretation. The 2 umbilical arteries are usually of equal size and each supplies predominantly one half of the placenta (Fig. 3). The 2 umbilical arteries anastomose at the chorionic plate. In this way arterial pressure in both parts of the placenta is equalized. In the chorionic plate, the arteries lose their muscular coats and ramify radially. At each division, a proportion of them turns downwards into the substance of the placenta. Each such artery, still of relatively large size, is contained in a primary trunk which, after a variable distance downward or laterally. ends in a cotyledon. These cotyledonary arteries are end-vessels with 3 orders of arterial divisions. There are no anastomoses between arteries in the same or contiguous cotyledons.

18

/

\

19

Regarding placental growth, it is essential to realize that this basic configuration of the arterial tree is completely developed even in the first trimester. Of course the diameters of the vessels increase during growth, but the number of cotyledons and the 3 orders of arterial branches remain constant from the 12-16 wk [14]. From this point, placental growth is characterized by a continuous proliferation of the terminal capillary bed (Fig. 3). This peripheral growing-end comprises an extensive capillary network with small sprouting capillary units covered by a thin layer of syncytium. These sprouts are the chorionic villi where oxygen and metabolites are exchanged between the fetal and maternal blood streams. New syncytial buds, vascularized with capillary sprouts from older villi, are continuously formed until term, although the rate of proliferation slows down in the last trimester. In this way, the functional transfer area of chorionic villi, as well as the preceding capillary plexus, grows continuously until term. The thickness of the syncytial layer decreases with advancing placental maturity, while the villous capillaries become sinusoid-shaped with ultimately larger diameters than the supplying vessels. It will be apparent that the early developed arterial compartment and the continuously growing capillary bed must be regarded as two distinct hemodynamic entities. Fetal placental hemodynamics Pressure-flow relationships have been measured directly in the larger fetal vessels in animal experiments. The peripheral hemodynamics must be extrapolated with emphasis on the morphological characteristics. The complexity of the capillary bed (Fig. 3) points to a more sophisticated physiology than mere flow from artery to capillary to vein. The capillary plexus provides numerous ‘extra villous shunts’, which are not true arteriovenous shunts but alternative capillary channels set up in the course of a single supply line [15]. In this way, flow to functioning villi is maintained in the case of an impediment to the main route at any point. Most of the mature larger villi do not empty their oxygenated blood into the capillary plexus, but drain by means of separate venules. The facultative ‘extra villous shunts’ as well as the large size of the capillary plexus relative to the villi should serve as a buffer to protect the villi from marked or extreme fluctuation in blood pressure [16]. The blood pressure in the villi must be low, at an approximately equal level to the maternal blood pressure in the surrounding intervillous space, in order to permit free transfer of water and metabolites in either direction. This intervillous pressure is about lo-15 mmHg [17], and the same pressure range was demonstrated in the umbilical veins [lo]. During fetal growth, systemic arterial pressure as well as the circulating volume of flow increase steadily, while the umbilical venous pressure remains relatively low (lo-15 mmHg) with only a comparatively small rise of few mmHg [lo]. At any gestational age, the pulsatile character of arterial pressure (systolic and diastolic levels) continues at approximately systemic level through the umbilical arteries [10,16]. The blood pressure in the villi, which equals the venous pressure, is hardly or no longer pulsatile. Therefore a considerable compliance must be present in the arterial branches and/or the capacious capillary network.

20

The pressure drop between the fetal aorta and the umbilical vein points to vascular resistance in the interjacent system [lo], in which 2 compartments must be distinguished on morphological grounds: the resistance in the early developed arterial compartment and the resistance in the continuously growing capillary bed.

Mathematical

model

Starting from the pressure drop between the fetal aorta and the chorionic villi we will demonstrate that the pulsatility of blood flow (PI,) in the umbilical arteries is mainly determined by the ratio between the capillary and arterial resistances. As shown in the appendix (Fig. 7), the umbilical circulation can be simplified to only a few concentrated components (Fig. 4): 1. The arterial resistance (R,) as a concentrated representation of the distributed resistances in the non-proliferating arterial compartment; 2. The resistance (R2) in the proliferating capillary periphery; 3. The combined vascular compliance (C). Fig. 4 illustrates the pulsatile aortic pressure (P,,,,,) propelling the flow in the umb. aa. (Q) through the arterial resistance (R,), the compliance (C) and the capillary resistance (R2). In the chorionic villi the pressure (P,;,,;) is hardly or no longer pulsatile. At a given total resistance, Q is determined by the pressure gradient between the fetal aorta and the villi: i.e. the pressure gradient across R, (Fig. 4). Thus the pulsatility of the flow Q in the umb. aa. equals the pulsatility of this pressure gradient. Elementary mathematics (see appendix) now yields a simple relationship between PI,, the model resistances and the pulsatility of fetal aortic pressure (PI,) (Fig. 5). Note that PI, depends directly on PI,, while at a given PI, there is a linear relationship between PI, and the ratio of capillary to arterial resistances. The PI, is fairly invariant, as demonstrated below, thus the ratio of resistances will be the predominant factor determining PI, values obtained in Doppler examinations.

pulsatilo gradient

Fig. 4. Pressure

curves in the simplified

pressure across

RI

model of the umbilical-placental

circulation.

See text for details.

21

Fig. 5. Relationship

Aortic pressure

between

PI,

pulsatility

and model resistances

and pulsatility

of fetal aortic pressure

PI,

in relation to fetal growth

Mean aortic pressure rises during fetal growth, but pulse pressure rises proportionally, leaving the PI, apparently unchanged. We estimated PI, values of about 0.5 from pulsatile pressure data of mid-pregnancy as well as term fetal lambs [18]. Casuistic blood pressure registrations in human fetuses of 200-360 g [19,20] also yielded PI, values between 0.4 and 0.6. In human adults the PI,, also ranges between 0.4 and 0.6, as well as in newborns, even prematures and small-for-date babies with birthweights below 750 g [21-231. Apparently, the margins of PI, are invariant during growth. However, on a minute-by-minute basis, small variations between invariant margins of 0.4 and 0.6 will be present, inherent to short-term pressure variations in the active fetus. Much higher PI, values would correspond with too extreme pulse pressure with too low mean pressure levels and a diastolic aorta pressure approaching zero. Such a situation is not compatible with life, neither in adults nor in the fetus, simply because of inadequate perfusion of vital organs. We conclude that the relevant trajectory for PI, in the diagram of Fig. 5 runs between reasonable margins from 0.4 to 0.6. As a consequence of a fairly constant PI,, Doppler PI, measurements give direct information as to the ratio of placental capillary to placental arterial resistances. Testing

the hemodynamic

interpretation

Now, the hypothesis must be tested in view of current concepts of fetal placental morphology and physiology. For this reason, the PI, values obtained by Doppler examinations throughout normal pregnancy [2] are plotted within the relevant trajectory of PI, (0.4-0.6) in our theoretical diagram (Fig. 6). Note that the ratio of resistances R,/R, declines with advancing normal pregnancy. This decline is not regular but occurs at a decelerating rate, in particular in the last trimester.

Fig. 6. Plq values PI, = 0.4-0.6.

obtained

by Doppler

examinations

throughout

normal

pregnancy

plotted

within

The vascular resistance in the placenta declines with advancing pregnancy [lo]. According to our model the resistance diminution is not homogeneous; R, decreases far more than R,. This mathematical result is in good agreement with the morphology of the growing placenta; R, decreases due to gradually increasing diameters of a fixed number of arterial branches, while R, decreases by an enormous enlargement of the terminal capillary bed. The rate of peripheral proliferation slows down in the last trimester, explaining the decelerating rate of the declining ratio. Corroborative evidence for the validity of our model is given by the range of biological variation observed in repeated PI, measurements in the same fetus at a given gestational age. This biological variation was explained by short-term but limited alterations of pressure (PI,), due to fetal activities. In early pregnancy, PI, variations (between 0.4 and 0.6) must induce more pronounced variations in PI, than the same PI,, alterations will do during later pregnancy, since the PI, trajectory in our diagram is narrowing towards term (Fig. 6). Indeed, the biological PI, variations observed in Doppler examinations are far more pronounced in early than in more advanced gestation [5]. In the last trimester, the ratio R/R, declines below 1, supporting the current concept of the reserve capacity of the placental circulation (see next section). In conclusion, the Doppler results interpreted according to this hemodynamic hypothesis are in good agreement with current knowledge on placental morphology and physiology.

Physiological and clinical relevance The postulate that the PI, of umbilical arteries is determined by the ratio of capillary to arterial resistances in the placental circulation gives rise to some fascinating conclusions with regard to fetal physiology and the possibility of monitoring placental circulatory competence using Doppler ultrasound.

23

Obviously, blood flow to an organ is mainly determined by the site of the highest vascular resistance. In all circulatory systems of healthy adults this flow controlling resistance is located in responsive arterioles at organ level, just preceding the functional capillary bed. The resistance of the supplying arteries is always lower. except for cases of obstructive arterial diseases. In the growing placenta, however, the predominant vascular resistance appears to migrate gradually from functional organ level to the arterial supply line. Before 28 wk, vascular resistance is highest at organ level (R2/R, > l), as in all other systems, although a capillary network instead of responsive arterioles precedes the functional tissue (chorionic villi). The volume of flow to functional tissue is mainly determined at organ level. In the last trimester, however, the ratio R,/R, decreases below 1, pointing to an unique characteristic: the resistance at functional organ level (R,) has decreased even below the arterial resistance (R,), which itself already is low. Now. the arterial supply line, having the predominant resistance, mainly determines the volume of flow to the placental functional tissue. This is not achieved by an increasing arterial resistance, but it results from an abundant decrease of peripheral vascular resistance. This indicates an abundance of peripheral capillary bed, reflecting the well-known reserve capacity [ll] of the placental vascular bed. As long as capillary obstructions or insufficient villous maturation do not raise the total capillary resistance (R,) above the total arterial resistance (R,), the volume of flow to functioning villi will hardly be diminished, since it will continue at the maximum rate allowed for by the predominant arterial resistance. This critical turning point (R,/R, = 1) reflecting the presence or absence of capillary reserve capacity, corresponds to PI, values in the umb. aa. of approx. 1 (0.8-1.2) (Fig. 6). Higher mean PI, values in the last trimester correspond to a resistance ratio above 1, reflecting loss of capillary vasculature below its reserve capacity. This must diminish the total flow to the diminished amount of remaining functioning villi. This just indicates beginning placental vascular insufficiency, and the fetus must be regarded at risk. Such an incipient disorder of placental flow might be compensated by an increased extraction of oxygen and nutrients, emphasizing that placental circulatory compromise is not necessarily synonymous with fetal compromise. The duration and seriousness of placental circulatory failure, which is reflected by the degree of PI, enhancement, will eventually determine whether fetal well-being is compromised. General conclusions It is concluded that the PI of Doppler recordings from umb. aa. gives direct information as to the circulatory competence of the fetal placenta. Short-term biological PI variations are caused by small variations in fetal aortic pulse pressure due to fetal activities. To diminish the impact of these PI, variations on PI, values, averaging PI, values of several steady-state Doppler recordings was recommended [5]. The gestational age-related decline of mean PI, values reflects a decreasing vascular resistance in the terminal capillary bed, but the relationship is complex. Mean PI, values below 1 in the last trimester must reflect the normal reserve capacity of the placental capillary bed, indicating placental circulatory competence. Increased mean PI, values (exceeding 1 after 28 wk) reflect a critical loss of

24

placental peripheral vasculature below its reserve capacity. Thus must diminish total blood flow to functioning chorionic villi, indicating the beginning of circulatory incompetence with a fetus at risk. The degree and duration of placental vascular incompetence will determine whether symptoms of fetal compromise, such as fetus malnutrition or hypoxia, will eventually become evident. Preliminary Doppler results confirm these theoretic postulates; pregnancies with PI, values below 1 after 28 wk have normal fetal outcomes, while increased PI, values may be associated with clinically evident placental failure [2].

Appendix

Model considerations Modelling the umbilical circulation (Fig. 3) results in an analogous circuit using compliances and resistances as shown in Fig. 7. Within the biological range of Pa,,,,, and within the range of PI, observed ic our patients, the computer simulations, using this detailed model, yielded identical (variation within 5%) results (that is identical PI, as a function of changing capillary resistances) to those from simulations using the simplified model (Fig. 7, bottom) based upon a lumped compliance and only 2 distinct resistance compartments: R, in the non-proliferating arterial tree and R, in the continuously proliferating capillary periphery. Moreover, the more detailed simulations confirmed that the effects of inertia were negligible.

Mathematics The pulsatility of the flow Q equals the pulsatility of the pressure gradient (PI,,,.,, - P,,,,,) shown at the bottom of Fig. 4. This pulsatile pressure gradient is shown in more detail in Fig. 8.

arterial

tree

capillary

bed

venous

of the umbilical Fig. 7. A detailed and a simplified model resistance in the proliferating capillary periphery.

circulation.

R,, arteriai resistance;

R,,

25

Fig. 8. Pulsatile

pressure

gradient.

See text for details

Now we define the PI of fetal aortic pressure PI,=

(PI,,):

p. - P,l

~

P.l”l

The compliance is large enough to reduce the pulsatile character of aortic pressure to a non-pulsatile level in the chorionic villi (P,,,,,) (see text). This means that the with a mean value P,,,,. P,,,,,, depends pressure P,,,,, across R, is almost non-pulsatile, on the resistances R, and R, as well as on the mean aortic pressure (Pan,): R* P”,,,= ___ .P,,” R, +Rl Here the superposition principle, Under laminar flow conditions Q in the umbilical arteries:

Simple calculation

Q,=,

now yields for systolic (Q,), diastolic

p -p p, - P”“, Pd- pm,Q,=+.EL . Q