Optimal management of severe early fetal growth restriction presents one of the greatest challenges in obstetrics. At present, a combination of several different monitoring modalities is used, including arterial and venous Doppler ultrasonography plus assessment of short-term fetal heart rate variation, without consensus on which of these should trigger the decision to deliver. Pathological Doppler waveforms in the ductus venosus (DV) have been found to be associated with adverse outcome in prospective studies1–3.
We describe two growth-restricted fetuses, of 33- and 39-year-old primigravidae, which showed severely abnormal DV flow velocity waveforms with intermittent reversed flow during atrial contraction in the presence of otherwise normal Doppler measurements. There were no signs of fetal abnormalities, with normal echocardiography and normal karyotype in both fetuses. On weekly outpatient visits we performed ultrasound examinations including assessment of fetal growth and Doppler parameters and carried out computerized fetal heart-rate analysis. Follow-up showed large variations in DV waveforms ranging from positive to reversed flow during atrial contraction. The current standard in our center would be to deliver a growth-restricted fetus with such a late stage of abnormality in the DV waveform (a-wave signal reaching the baseline or indicating reversed flow). However, we abstained from delivering these fetuses when the DV showed reversed flow because no other signs of fetal hemodynamic compromise could be detected by Doppler measurements in other vessels (Figures 1 and 2). In particular, we compared the pathological PIV in the DV with Doppler measurements in the hepatic vein and inferior vena cava. For example, in Case 1 at 30 weeks' gestation DV-PIV was pathological with a value of 1.19, whereas PIV in the hepatic vein was normal at 3.4. In Case 2 at 27 weeks' gestation, the DV-PIV was pathological at 1.38, while the PIV in the hepatic vein and the inferior vena cava were within the normal ranges (2.9 and 2.8, respectively). Normal measurements in both the hepatic vein and the inferior vena cava suggested that pathological PIVs in the DV were not due to compromised cardiac function but might rather reflect a localized phenomenon in the DV vasculature.
Additionally, the DV Doppler pulsatility showed wide variability ranging from positive to reversed a-wave during the same Doppler examination. Over the course of several weeks we observed a flattening of the fetal growth curve, an increase in umbilical artery resistance and redistribution of the fetal circulation. At 37 and 33 weeks' gestation two neonates with birth weights of 1600 g and 690 g, respectively, were delivered by Cesarean section.
While we strongly believe that flow measurement in the DV is an important indicator of fetal condition, these cases present examples in which DV pathology was not representative of the fetal condition. Caution is warranted when using single Doppler measurements to trigger delivery. A combination of various parameters constitutes the best approach. We hope that prospective randomized trials such as the ongoing TRUFFLE-study (trial of umbilical and fetal flow in Europe) will indicate the best time-point and trigger for delivery4.