The Delphi definition for selective fetal growth restriction may not improve detection of pathologic growth discordance in monochorionic twins

Abstract

A consensus definition of selective fetal growth restriction in monochorionic diamniotic twins was recently proposed following a Delphi procedure involving an international panel of experts. The new definition augments the traditional definition with additional sonographic criteria. We sought to determine whether the augmentations of the "Delphi definition" identified additional morbidity and mortality compared with a traditional definition. Furthermore, we sought to determine the benefit of each definition in identifying pathologic growth restriction relative to uncomplicated monochorionic diamniotic twins. This was a retrospective analysis of unselected monochorionic diamniotic twins that underwent fortnightly ultrasound surveillance at a single center between 2011 and 2020. Patients with concomitant twin-to-twin transfusion syndrome, twin anemia polycythemia sequence, or twin reversed arterial perfusion sequence at the time of diagnosis of selective fetal growth restriction were excluded. The diagnosis of selective fetal growth restriction using the Delphi definition required either an estimated fetal weight of <3rd percentile or presence of 2 of 4 observations in the smaller twin: (1) estimated fetal weight of <10th percentile, (2) estimated fetal weight discordance of >25% compared with the larger twin, (3) abdominal circumference of <10th percentile, (4) umbilical artery pulsatility index of >95th percentile. Diagnosis using the traditional definition required an estimated fetal weight of <10th percentile and an estimated fetal weight discordance of >25%. To determine the efficacy of the augmentations in the Delphi definition, 3 groups were compared: group I, uncomplicated monochorionic diamniotic twins; group II, twins with selective fetal growth restriction using the traditional definition (and therefore the Delphi definition); and group III, twins with selective fetal growth restriction solely using the Delphi definition. Demographic characteristics, subsequent development of twin-to-twin transfusion syndrome or twin anemia polycythemia sequence, pregnancy outcomes, and neonatal outcomes were compared. There were 325 patients with monochorionic diamniotic twins that met inclusion criteria. Of these, 213 (66%; group I) were uncomplicated, 37 (11%; group II) met the traditional definition for selective fetal growth restriction, and 112 (35%) met the Delphi definition for selective fetal growth restriction with 75 (67%) meeting solely the Delphi definition (group III). Demographic characteristics were similar between groups. Patients in group II delivered earlier than uncomplicated twins (32.1 vs 35.7 weeks of gestation; P<.01) and patients in group III (32.1 vs 35.6 weeks of gestation; P<.01). Furthermore, they were more likely to have critical umbilical artery Doppler abnormalities (38% vs 4%; P<.01) and be delivered for deteriorating fetal status (30% vs 5%; P<.01) than those in group III. Overall, survival was lower in group II than groups I and III (89% vs 96% and 100%, respectively; P=.04). Moreover, composite neonatal morbidity and mortality were greater in group II (30%) than either group I (6%; P<.01) or group III (9%; P<.01). The rates of composite neonatal morbidity and mortality were similar between groups I and III (P=.28). The addition of abdominal circumference and umbilical artery pulsatility index thresholds and isolated estimated fetal weight of <3%, as proposed by the Delphi definition, increased the diagnosis of selective fetal growth restriction; however, there was no added benefit in the identification of growth discordant pregnancies at risk of adverse outcomes. Prospective analysis of monochorionic diamniotic twins is required to contextualize these findings.