Discordance of arterial and venous flow velocity waveforms in severe placenta‐based fetal growth restriction

Abstract

Balancing iatrogenic prematurity against stillbirth due to terminal placental dysfunction poses a crucial dilemma in the management of severe fetal growth restriction (FGR). The role of multivessel Doppler assessment in defining severity and progression is clear, but gestational age at delivery remains the single most important factor when FGR is severe and early onset1. Recently, Diemert and Hecher2 reported FGR cases where venous Doppler indices may have called for delivery, but pregnancy was extended because arterial indices were not so abnormal. Is this discordance a rare curiosity, or a significant caution to the Doppler-based management of FGR? We were prompted by this question to analyze our collaborative database of 604 FGR cases3, 4. A manual of standard imaging and measurement techniques was distributed. Collaborating centers demonstrated concordance with these techniques, including submission of representative images to the principal author (A.A.B.). Inter-center variation for Doppler parameters and primary outcomes was analyzed to test relative uniformity of participating centers. Further detailed analysis of group vs. individual performance statistics and center-by-center contribution to mortality and morbidity was completed to ensure that results could be reliably grouped3. We defined significant discordance between arterial and venous (A/V) Doppler in two ways: Categorical definition: for patients with positive end-diastolic velocity (PEDV) in the umbilical artery (UA), absence or reversal of the ductus venosus a-wave (DV-RAV) was considered to be significantly A/V discordant (analogous to the cases presented by Diemert and Hecher2). When UA Doppler showed absent or reversed end-diastolic velocity (A/REDV), we reasoned that there would be no clinical difference between antegrade DV a-wave and DV-RAV, so categorical definition did not apply; Numerical definition: since the UA and DV pulsatility indices (PI) can be calculated regardless of end-diastolic velocity, significant discordance was also denoted if a Z-score difference > 2 was present when DV-PI Z-score was compared with UA-PI Z-score. Fifty-four of 604 fetuses (8.9%) had significant A/V discordance of the Doppler parameters. Thirty-four of these had major complications and 23 died during the neonatal period (all, chi-square P < 0.0001 compared with FGR fetuses with concordant A/V measures). Since many experienced major morbidity and ultimately died, cases may be recorded in both adverse outcome groups. Intact survivors numbered 17, 4/15 (26.7%) in the PEDV group and 13/39 (33.3%) in the A/REDV group. Outcomes for fetuses stratified by UA end-diastolic velocity are displayed in Table 1. Within both subgroups, UA cord pH and the amniotic fluid index were significantly lower in patients with disproportionally worse DV Doppler parameters, while gestational ages at delivery did not differ. Discordance of DV and UA Doppler parameters is observed in a smaller proportion of fetuses with positive UA end-diastolic velocity compared to fetuses with A/REDV. While disproportional worsening of DV parameters in A/REDV fetuses is a well-described response to worsening placental dysfunction this observation is less well defined in patients with positive UA end-diastolic velocity. Irrespective of UA end-diastolic category, discordant deterioration of the DV Doppler parameters is associated with worse outcome in neonates with preterm FGR. This appears to be related to fetal deterioration rather than gestational age as illustrated by the lower pH and amniotic fluid parameters. A minority of patients with A/V discordance have normal outcome, but these are important because if timing of delivery is determined by DV parameters alone, 30% of patients with PEDV and up to 50% of patients with A/REDV may be delivered earlier than necessary. The TRUFFLE study4 may answer many questions regarding the neurodevelopmental impacts of delivery triggers based on the DV and computerized cardiotocography. However, the cases presented by Diemert and Hecher2 and our analysis suggest that discordance of Doppler deterioration and outcome may be a significant confounder—additional fetal variables may be important. The discriminatory role of amniotic fluid assessment as one of the biophysical parameters suggests the merit of further evaluation of integrated fetal testing in the context of FGR management1. A. A. Baschat*, C. R. Harman*, * Department of Obstetrics, Gynecology and Reproductive Sciences, University of Maryland School of Medicine, Baltimore, MD, USA