ELSEVIER Early Human Development 39 (1994) 15-26

Ductus venosus velocity waveforms in appropriate and small for gestational age fetuses

Giuseppe Rizzo* a, Alessandra Capponi a, Domenico Arduini b, Carlo Romanini a

aDepartment of Obstetrics and Gynecology. Universitri di Roma “Tor Vergata”, Policlinico Nuovo S.

Eugenio, P.le Umanesimo, 10, 00144 Rome, Italy bDepartment of Obstetrics and Gynecology. Universith di Ancona, Ancona, Italy

Received 20 December 1993; revision received 27 May 1994; accepted 30 May 1994

Abstract

The objective of this study was to evaluate differences in ductus venosus velocity waveforms between appropriate and small for gestational age fetuses by using a new index based on the ratio between systolic and atria1 peak velocities. Ductus venosus velocity waveforms were cross-sectionally recorded in 164 appropriate for gestational age fetuses at 16-42 weeks of ges- tation and in 97 small for gestational age fetuses free from structural and chromosomal abnor- malities between 24-36 weeks of gestation. Small for gestational age fetuses were divided according to the Doppler findings in arterial peripheral vessels: group A (n = 33), normal ratio between umbilical artery and middle cerebral artery Pulsatility Indices; group B (n = 41), um- bilical artery/middle cerebral artery ratio > 95th centile but presence of end diastolic flow in umbilical artery; group C (n = 23), umbilical artery/middle cerebral artery ratio >95th centile and absence of end diastolic flow in umbilical artery. Eighteen small for gestational age fetuses (10 from group B and 8 from group C) were also serially studied until delivery due to fetal distress. Ductus venosus velocity waveforms were recorded at the level of its origin from um- bilical vein and the ratio between systolic and atria1 peak velocities (systolic/atria1 ratio) calculated. In appropriate for gestational age fetuses, systolic/atria1 ratio values significantly decrease with gestation. No significant differences were found in systolic/atria1 ratio between appropriate for gestational age fetuses and group A small for gestational age fetuses while, after correction for gestational age, significantly higher values were found in group B (P _( 0.01) and group C (P I 0.001) fetuses. Among these fetuses, those with systolic/atria1 ratio above the 95th confidence interval showed a poorer perinatal outcome. No relationships were found between systolic/atria1 ratio and Pulsatility Index values from fetal arterial peripheral vessels, while a positive relationship was found with the percentage reverse flow in inferior vena cava. In fetuses serially followed, the systolic/atria1 ratio progressively increas-

* Corresponding author.

0378-3782/94/!§07.00 0 1994 Elsevier Science Ireland Ltd. All rights reserved SSDI 0378-3782(94)01562-4

16 G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26

ed approaching the onset of abnormal fetal heart rate patterns. The systolic/atria1 ratio allows the evaluation of ductus venosus hemodynamics in small for gestational age fetuses and this index may be useful in the monitoring of such fetuses.

Keywords: Color Doppler; Ductus venosus; Fetal circulation; Fetal growth retardation

1. Introduction

Ductus venosus is an important regulator of venous blood flow distribution in fetal circulation [l]. Ductus venosus, by providing a by-pass of the hepatic circula- tion, allows the well-oxygenated umbilical venous blood to reach the inferior vena cava and the left ventricle directly through the foramen ovale and then to vital organs such as the heart and brain [2]. Experimental animal models have shown, during hypoxia, a dramatic increase in blood flow shunted through the ductus venosus in the attempt to maintain optimal oxygenation of the vital organs [3].

The advent of high resolution ultrasound machines together with color and pulsed Doppler equipment has allowed the recording of blood flow velocity waveforms from the ductus venosus in normal fetuses and in fetuses with congenital heart diseases [4-71. To the best of our knowledge, no data are currently available on the human fetus during hypoxia.

Moreover, by studying velocity waveforms in normal and small for gestational age (SGA) fetuses, we noted some changes in the waveform morphology in particular at the level of the ratio between peak velocities during systole and atria1 contraction. We speculated that this index might be modified in SGA fetuses. Objectives of this study were: (1) to construct reference limits for gestation in normal pregnancies; (2) to evaluate differences, if any, in SGA fetuses grouped according to the character- istics of Doppler velocity waveforms in umbilical artery and middle cerebral arteries; (3) to evaluate changes in this index in serial recordings.

2. Materials and methods

After informed consent from the mothers, ductus venosus velocity waveforms were recorded in 164 appropriate for gestational age (AGA) fetuses from un- complicated pregnancies at 16-42 weeks of gestation (Table 1). Ninety-seven SGA fetuses were also studied at 24-36 weeks. All pregnancies were singleton and accurately dated by early second trimester ultrasonographic examination.

All SGA fetuses fulfilled the following criteria: (1) abdominal circumference below the 5th centile (fetuses with a gestational age ~28 weeks) or an ultrasonographic estimated fetal weight below the 10th centile (fetuses with a gestational age 228 weeks); (2) absence of congenital or chromosomal abnormalities; (3) postnatal con- firmation of a birthweight below the 10th centile for Italian population standards PI.

According to the Doppler findings in arterial peripheral vessels, SGA fetuses were divided into three groups: group A (n = 33) normal ratio between umbilical artery

Tabl

e I

Cha

ract

eris

tics

of t

he

fetu

ses

stud

ied

Para

met

er

AG

A

(n=

164)

SG

A

grou

p A

(I

I = 3

3)

SGA

gr

oup

B

(n=4

1)

SGA

gr

oup

C

(n =

23)

Ges

tatio

nal

at

Dop

pler

re

cord

ing

age

28.3

zt

1.3

5 (w

eeks

)

Ges

tatio

nal

at d

eliv

ery

(wee

ks)

age

40.2

f

1.89

C

esar

ean

sect

ion

for

feta

l di

stre

ss

0 (0

%)

No.

(%

)

Birt

hwei

gth

(g)

3567

zt

z 435

U

mbi

lical

ar

tery

pH

-

5’

Apg

ar

scor

e c7

N

o.

(%)

0 (0

%)

Neo

nata

l de

ath

No.

(%

) 0

(0%

)

Dat

a ar

e pr

esen

ted

as m

ean

f SD

. or

N

o.

and

%.

31.4

5 zt

3.2

8 30

.04

f 3.

16

21.7

f

2.19

38.9

zt

2.1

I 33

.9

f 3.

21

30.1

f

4.32

0

(0%

) 30

(73

.1%

) 19

(82

.6)

2434

f

276

1555

f

324

914

f 24

5 7.

293

f 0.

037

7.22

1 f

0.03

9 7.

199

f 0.

041

0 (0

%)

8 (1

9.5%

) I I

(47

.8%

) 0

(0%

) 3

(7.3

%)

7 (3

0.4%

)

18 G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26

and middle cerebral artery Pulsatility Indices (PI); group B (n = 41) umbilical artery/middle cerebral artery PI ratio > 95th centile for our reference limits for gesta- tion [9] and presence of end diastolic flow in umbilical artery; group C (n = 23) um- bilical artery/middle cerebral artery PI ratio >95th centile and absence of end diastolic flow in umbilical artery (Table 1). All the fetuses were cross-sectionally studied.

Some SGA fetuses of groups B and C were also studied longitudinally until deliv- ery. In 23 SGA fetuses (35.9%) in groups B and C, delivery occurred within 6 days from the first Doppler recording for maternal or fetal indications. Thus, no further Doppler examinations were performed. In the remaining 41 cases, expectant man- agement was possible and serial Doppler recordings were performed at weekly inter- vals. In order to obtain a homogeneous group, we retained the data from the 18 fetuses (10 from group B and eight from group C) who developed antepartum late heart rate decelerations and had been studied on at least three occasions, 1 week apart.

In the 41 pregnancies managed expectantly, heart rate monitoring was performed daily for a minimum duration of 45 min from the time of the first Doppler recording onwards. At the onset of late decelerations during Braxton Hicks contractions, an elective cesarean section was performed, based on the close association between these decelerations and fetal hypoxemia and (mild) acidosis [9]. The median interval between the last Doppler recordings and delivery was 2 days (range O-4). Doppler data were not considered for the clinical management except as an indicator for in- tensive monitoring.

A color and pulsed Doppler machine (Ansaldo Esaote Hitachi 560 Genova, Italy) was used to record Doppler velocity waveforms from the umbilical artery and vein, descending aorta, middle cerebral artery, inferior vena cava and ductus venosus. The techniques of recording have been reported in detail elsewhere [6,10,11].

In particular, ductus venosus velocity waveforms were recorded from a transverse section of the upper fetal abdomen, at the level of its origin from the umbilical vein. The ductus venosus velocity waveforms exhibit a biphasic pattern, with a first peak concomitant with ventricular systole, a second peak during ventricular diastole, and a nadir occurring during atria1 contraction (Fig. 1). Care was taken to keep the angle of insonation between the Doppler beam and the presumed direction of flow close to zero as assessed by the color flow mapping and recordings with an angle of insonation greater than 20” were rejected.

For each vascular district, 10 consecutive heart cycles were selected during periods of fetal rest without breathing movements and the values measured were averaged. The following variables were measured with the aid of a computer-interfaced digitizer pad (Cardio 800, Kontron, Oxford, UK): (1) Pulsatility Index (PI) [ 121 from the umbilical artery, descending aorta and middle cerebral artery; (2) percentage reverse flow in inferior vena cava [lo]; (3) notation on the presence or absence of repetitive pulsations in the umbilical vein at the end of diastole during fetal apnea [13]; (4) ratio between peak velocities during systole and atria1 contraction (systolic/atria1 ratio) at the level of the ductus venosus.

All the recordings were performed by one of the authors (G.R.). As previously

.G. Rho et al. /Early Huti. Dev. 39 (1994) 15-26 19

established, the intra-observer coefficient of variation in our laboratory was 4.9% for the ductus venosus systolic/atria1 ratio [6].

Reference limits for gestation of ductus venosus systolic/atria1 ratio (mean and 95th individual confidence intervals) were constructed by regression analysis in AGA fetuses. The optimal polynomial model was selected on the basis of the max- imal coefficient of correlation and the presence of coefficients differing from zero. Kolmogorov D-tests were used to examine the adequacy of normal (gaussian) approximations for the empirical distributions.

Comparison between the systolic/atria1 ratios of SGA fetuses and newly developed nomograms for AGA fetuses was performed by the one-sample l-test. Since this index may change with advancing gestation, data were also expressed as the number of standard deviations from the respective normal means for gestation (delta values). Relationships between the systolic/atria1 ratio and the other Doppler indices measured were tested by simple and multiple regression analysis after controlling gestational age by analysis of covariance. The longitudinal changes in the systolic/ atria1 ratio were evaluated by the analysis of variance for repeated measurements. Further statistical analysis included the unpaired r-test, Wilcoxon rank sum test and Fisher’ exact test. P I 0.05 was considered significant.

3. Results

In AGA fetuses, the systolic/atria1 ratio significantly decreases with gestation (constant, 3.35; slope, -0.049; S.D., 0.490; r, 0.474; P < 0.001) (Figs. 1,2). When the systolic/atria1 ratio was compared with the other Doppler indices tested, a signiti- cant relationship (r, 0.372; P C 0.01) was noted with the % reverse flow in the inferior vena cava even when gestational age was taken into account by analysis of covariance. No significant relationships were found with the PI values from the

Fig. 1. Blood flow velocity waveforms from an AGA fetus at 16 weeks of gestation (systolic/atria1 ratio, 2.56) (left panel) and at 40 weeks of gestation (systolic/atria1 ratio, 1.64) (right panel).

20 G. Rho et al. /Early Hum. Dev. 39 (1994) 15-26

5

4.5

4

3.5 0 .I 2 3

1 3 2.5 G! u ‘S O 2 s 2

1.5

1

0.5

0

14 16 18 20 22 24 26 28 30 32 34 36 38 40 42

gestational age (weeks)

Fig. 2. Reference range (mean and linear regressions of individual 95% confidence intervals) for gestation of systolic/atria1 ratio from the ductus venosus obtained from the cross-sectional study of I64 AGA fetuses.

umbilical artery, descending aorta and middle cerebral artery when the data were corrected for gestational age.

No significant differences were found in the ductus venosus systolic/atria1 ratio be- tween AGA and group A SGA fetuses (mean difference, 0.055 S.D.; not significant) (Fig. 3). On the other hand, significantly higher values were found in group B (delta value, 1.282 S.D.; t = 3.121; P < 0.01) and group C (delta values, 7.634; t = 5.832; P < 0.001) SGA fetuses (Fig. 3). Furthermore, in five SGA fetuses in group C, flow velocity during atria1 contraction was absent and reversed in three fetuses (Fig. 4). For this condition, a ratio equal to infinity, a value of 10 was arbitrarily assigned. This value was selected on the basis of our experience which has shown that in the presence of forward flow during atria1 contraction, the systolic/atria1 ratio is always lower than nine.

G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26 21

SGA group A SGA group B

0.5 .

o- . - * . . n ’ 22 24 26 28 30 32 34 36

gestational age (weeks)

o.;l . . . _ . . , 22 24 26 28 30 32 34 36

gestational age (weeks)

SGA group C

9.

8.

7. 2 6

I .e P 5. y 2 4. $

3.

2.

1 .

??

? ?

??0 0

o- 22 24 26 28 30 32 34 36

gestational age (weeks)

Fig. 3. Ductus venom systolic/atria1 ratios of SGA fetuses in groups A, B and C plotted on reference range.

22 G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26

Fig. 4. Blood flow velocity waveforms from an SGA fetus of group B at 28 weeks of gestation (systolic/atria1 ratio, 4.14) (left panel) and an SGA fetus of group C at 28 weeks of gestation with reverse flow during atrial contraction (right panel).

Like AGA fetuses, SGA fetuses also showed a significant relationship between the systolic/atria1 ratio and the percentage reverse flow in the inferior vena cava when either all SGA fetuses were considered together (r = 0.289; P < 0.05) or when the analysis was restricted to groups B and C (r = 0.464; P < 0.005). However, when the analysis was limited to group A (r = 0.201; not significant), no significant rela- tionships were recorded between these two venous indices. No significant relation- ships were present for the PI values in the umbilical artery, descending aorta and middle cerebral artery. Umbilical vein pulsations were present in 12 SGA fetuses in group C (eight of these fetuses also had absent or reversed velocities during atria1 contraction at the level of the ductus venosus) and those fetuses exhibited higher systolic to atria1 values than the remaining fetuses in group C (9.141 f 1.523 vs. 5.023 f 2.893; P < 0.001).

When SGA fetuses in groups B and C were divided according to the presence of a systolic/atria1 ratio above or below the upper 95th confidence interval, a shorter interval between the Doppler recordings to delivery (median, 6; range O-22 vs. medi- an, 18; range 4-37; P < O.Ol), a lower umbilical artery pH at birth (7.198 f 0.023 vs. 7.231 f 0.043; P < O.OOl), and a higher neonatal mortality (8/31 vs. 2/33; P < 0.001) were seen in the former situation. Limiting the analysis to fetuses in group B, these differences became significant for the time interval between the Doppler re- cording and delivery (median, 12 range l-22 vs. median, 20; range 6-37; P < 0.05) and umbilical artery pH (7.212 i 0.033 vs. 7.247 f 0.041; P < 0.02).

Finally, the SGA fetuses followed longitudinally showed a progressive increase in the systolic/atria1 ratio until the onset of late fetal heart rate decelerations (F= 26.123; P < 0.001) (Fig. 5).

G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26 23

10

9

8

I I I I I I 1

22 24 26 28 30 32 34 36

gestational age (weeks)

Fig. 5. Individual changes in ductus venoms systolic/atria1 ratio in the 18 SGA fetuses studied longitudinally plotted on reference ranges. Values of each individual patients are connected by lines.

4. Discussion

In AGA fetuses, the systolic/atria1 ratio decreases with advancing gestation. Although the indirect evaluation of ductus venosus does not allow us to establish the hemodynamic causes of these changes, we can speculate that the decrease in the systolic/atria1 ratio is due to a relative increase in blood flow during end diastole. Previous studies have shown that, with advancing gestation, the compliance of the ventricles increases, thus resulting in an improved ventricular filling during diastole [13]. Furthermore, ventricular afterload decreases due to the fall of placental resistances [8]. As a consequence of these two phenomena, the atria1 gradient pres- sure decreases and the return of blood to the heart increases during diastole. The

24 G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26

significant relationship found with the % reverse flow in inferior vena cava (an index believed to be inversely related to atria1 gradient pressure [ 141, irrespective of gesta- tional age) corroborates this hypothesis.

No significant differences in systolic/atria1 ratio were found in group A SGA fetuses with respect to AGA fetuses. The normal Doppler findings from peripheral vessels, together with the normal perinatal outcome, suggest a low growth potentiali- ty as the more likely etiology in such fetuses. Our results prove that the systolic/atria1 ratio is not affected by the smaller size of these fetuses. Besides, several SGA fetuses in groups B and C showed increased systolic/atria1 ratios suggesting an impaired blood flow through the ductus venosus during diastole. Of interest is that, when fetuses are divided according to the presence of systolic/atria1 ratio above or within the reference limits, SGA fetuses with abnormal ratios showed a poorer perinatal outcome suggesting a more severe fetal compromise. This concept is further validated by our longitudinal observations showing a progressive increase in this index approaching the onset of antepartum late decelerations. The significance of the increased systolic/atria1 ratio is again difficult to establish due to the non-invasive nature of our study that does not allow simultaneous recordings of pressure and volume, as in animal models. However, in chronically deteriorating growth retarded fetuses, the high placental resistance, by increasing cardiac afterload, together with the reduced cardiac function, impair both the ejection and the filling of the ventricles resulting in low peak velocities from the outflow tract and high reverse flow in inferi- or vena cava during atria1 contraction [ 10,15,16]. This increased reverse flow may impair the forward flow through the ductus venosus during the atria1 contraction thus increasing the systolic/atria1 ratio. With further deterioration, this process may extend beyond the ductus venosus to the umbilical vein and pulsations may occur [ 16,171. This hypothesis is validated by the relationship found between systolic/atria1 ratio and % reverse flow in the inferior vena cava and by the higher values of the former index in the presence of umbilical vein pulsations. Furthermore, in fetuses in which cardiac output is impaired by structural heart disease, similar changes in venous circulation have been reported [4,7].

The relatively low number of fetuses investigated does not allow us to draw con- clusions on the clinical efliciency of the three venous districts considered (i.e. inferior vena cava, ductus venosus and umbilical vein). The relationship found between the % reverse flow in inferior vena cava and the systolic /atrial ratio suggests a similar significance for these two indices. However, the technique of recording velocity waveforms from inferior vena cava [lo] is more complex than that hereby described for the ductus venosus. Indeed, the former requires a parasaggital section of the fetal trunk while the latter uses the same plane employed for the measurement of abdom- inal circumference. Furthermore, the calculation of the systolic/atria1 ratio does not require planimetering of the velocity waveforms as is necessary for the % reverse flow in the inferior vena cava. Finally, reproducibility of the measurements of the index from the ductus venosus is higher than for inferior vena cava [lo], suggesting that the ductus venosus may be more successfully used in clinical applications. It may be argued that recording of umbilical vein pulsations is simpler and may be obtained with low cost continuous Doppler recordings. However, umbilical vein

G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26 25

pulsations occur in SGA fetuses at a very late stage of compromise already associ- ated or immediately followed by abnormal fetal heart rate patterns requiring prompt delivery [ 16,171. As a consequence, a limited clinical role may be predicted for umbil- ical vein pulsations with respect to the information available from ductus venosus and inferior vena cava.

In conclusion, we described an index from the ductus venosus unrelated to the PI values of fetal peripheral vessels which may allow the evaluation of venous hemody- namics in SGA fetuses.

Acknowledgement

This study was supported by a grant from the Italian National Council for Research (CNR Grant No. 920004 PF41).

References

111

121

[31

i41

PI

161

171

PI

[91

1101

IllI

[W

[I31

1141

1151

Edelstone, D.I. and Rudolph, A.M. (1979): Preferential streaming of ductus venosus blood to the brain and heart in fetal lambs. Am. J. Physiol., 237, H724-H729. Reuss, M.L. and Rudolph, A.M. (1980): Distribution and recirculation of umbilical and systemic venous blood flow in fetal lambs during hypoxia. J. Dev. Physiol., 2, 71-84. Bristow, J., Rudolph, A.M., Itskovitz, J. and Barnes, R.J. (1983): Hepatic oxygen and glucose mechanism in the fetal lamb: response to hypoxia. J. Clin. Invest., 71, 1047-1061. Kiserud, T., Eik-Nes, L.R., Blaas, H.G. and Hellevik, L.R. (1991): Ultrasonographic velocimetry of the ductus venosus. Lance& 338, 1139-l 145. Huisman, T.W., Stewart, P.A. and Wladimiroff, J.W. (1992): Ductus venosus blood flow velocity waveforms in the human fetus - a Doppler study. Ultrasound Med. Biol., 18, 33-37. Soregaroli, M., Rizzo, G., Danti, L., Arduini, D. and Romanini, C. (1993): Effects of maternal hyperoxygenation on ductus venosus flow velocity waveforms in normal third trimester fetuses. Ul- trasound Obstet. Gynecol., 3, 115-l 19. De Vore, G.R. and Horenstein, J. (1993): Ductus venosus index; a method for evaluating right ven- tricular preload in the second trimester fetus. Ultrasound Obstet. Gynecol., 3, 338-342. Gagliardi, L., Preve, CU., Crduro di Montezemolo, C., Nattone, G.P. and Piazza, A. (1975): Accrescimento intrauterino ed eta’ gestazionale in un campione di 9774 casi. Ann. Obstet. Gynecol. Med. Perinat., 96, 147-158. Bekedam, D.J., Visser, G.H.A., Mulder, E.J.P. and Poelmann-Weesjes, G. (1987): Heart rate varia- tion and movement incidence in the growth retarded fetuses. The significance of late heart rate de- celerations. Am. J. Obstet. Gynecol., 157, 126-133. Arduini, D. and Rizzo, G. (1990): Normal values of Pulsatility Index from fetal vessels: a cross sectional study on 1556 healthy fetuses. J Perinatal Med., 18, 165-172. Rizzo, G., Arduini, D. and Romanini, C. (1992): Inferior vena cava flow velocity waveforms in appropriate and small for gestational age fetuses. Am. J. Obstet. Gynecol., 166, l27l- 1280. Gosling, R.G. and King, D.H. (1975): Ultrasound angiology. In: Arteries and Veins, pp. 61-98. Editors: A.W. Marcus and L. Adamson, Churchill Livingstone, Edinburgh. Rizzo, G., Arduini, D. and Romanini, C. (1992): Umbilical vein pulsations: a physiological finding in early gestation. Am. J. Obstet. Gynecol., 167, 675-677. Reed, K.L., Sahn, D.J., Scagnelli, S., Anderson, CF. and Shenker, L. (1986): Doppler echocar- diographic studies of diastolic function in the human fetal heart: changes during gestation. J. Am. Coil. Cardiol., 8, 391-395. Reed, K.L., Appleton, C.P., Anderson, C.F., Shenker, L. and Sahn, D.J. (1990): Doppler studies of vena cava flows in human fetuses - insights into normal and abnormal cardiac physiology. Circulation, 8 1, 498-505.

26 G. Rizzo et al. /Early Hum. Dev. 39 (1994) 15-26

[16] R&o, G. and Arduini, D. (1991): Fetal cardiac function in intrauterine growth retardation. Am. J. Obstet. Gynecol., 165, 876-882.

[17] Arduini, D., Rizzo, G. and Romanini, C. (1993): The development of abnormal heart rate patterns after absent end diastolic velocity in umbilical artery: analysis of risk factors. Am. J. Obstet. Gynecol., 168, 43-50.

[18] Indick, J.H., Chen, V. and Reed, K.L. (1991): Association of umbilical venous with inferior vena cava blood flow velocities. Obstet. Gynecol., 77, 551-557.