RESUMEN
OBJECTIVE: To investigate the potential value of uterine artery pulsatility index (UtA-PI) and serum levels of the angiogenic placental growth factor (PlGF) and the antiangiogenic factor soluble fms-like tyrosine kinase-1 (sFlt-1) in the prediction of adverse perinatal outcome in small-for-gestational-age (SGA) and non-SGA neonates at 35-37 weeks' gestation. METHODS: This was a prospective observational study of 19 209 singleton pregnancies attending for a routine hospital visit at 35 + 0 to 36 + 6 weeks' gestation. This visit included recording of maternal demographic characteristics and medical history, sonographic estimation of fetal weight, color Doppler ultrasound for measurement of mean UtA-PI, and measurement of serum concentrations of PlGF and sFlt-1. Multivariable logistic regression analysis was carried out to determine which of the factors from maternal or pregnancy characteristics and measurements of UtA-PI, PlGF and sFlt-1 provided a significant contribution in the prediction of each of four adverse outcome measures: first, stillbirth; second, Cesarean delivery for suspected fetal compromise in labor; third, neonatal death or hypoxic ischemic encephalopathy Grade 2 or 3; and, fourth, admission to the neonatal unit (NNU) for ≥ 48 h. Predicted probabilities from logistic regression analysis were used to construct receiver-operating characteristics curves to assess the performance of screening for these adverse outcomes. RESULTS: First, 83% of stillbirths, 82% of Cesarean sections for presumed fetal compromise in labor, 91% of cases of neonatal death or hypoxic ischemic encephalopathy and 86% of NNU admissions for ≥ 48 h occurred in pregnancies with a non-SGA neonate. Second, UtA-PI > 95th percentile, sFlt-1 > 95th percentile and PlGF < 5th percentile were associated with increased risk of Cesarean delivery for suspected fetal compromise in labor and NNU admission for ≥ 48 h; the number of stillbirths and cases of neonatal death or hypoxic ischemic encephalopathy was too small to demonstrate significance in the observed differences from cases without these adverse outcomes. Third, multivariable logistic regression analysis demonstrated that, in the prediction of Cesarean delivery for suspected fetal compromise in labor, there was no significant contribution from biomarkers; the prediction of NNU admission for ≥ 48 h by maternal demographic characteristics and medical history was only marginally improved by the addition of sFlt-1 or PlGF. Fourth, for each biomarker, the detection rate of adverse outcome was higher in SGA than in non-SGA neonates, but this increase was accompanied by an increase in false-positive rate. Fifth, the relative risk of UtA-PI > 95th , sFlt-1 > 95th and PlGF < 5th percentiles for most adverse outcomes was < 2.5 in both SGA and non-SGA neonates. CONCLUSIONS: In pregnancies undergoing routine antenatal assessment at 35-37 weeks' gestation, measurements of UtA-PI, sFlt-1 or PlGF provide poor prediction of adverse perinatal outcome in both SGA and non-SGA fetuses. Copyright © 2019 ISUOG. Published by John Wiley & Sons Ltd.