Stratification of pregnancy care based on risk of pre-eclampsia derived from biophysical and biochemical markers at 19-24 weeks' gestation.

Abstract

We have proposed previously that all pregnant women should have assessment of risk for pre-eclampsia (PE) at 20 and 36 weeks' gestation and that the 20-week assessment should be used to define subgroups requiring additional monitoring and reassessment at 28 and 32 weeks. The objective of this study was to examine the potential improvement in screening at 19-24 weeks' gestation for PE with delivery at < 28, < 32, < 36 and ≥ 36 weeks' gestation by the addition of serum placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1) to the combination of maternal demographic characteristics and medical history, uterine artery pulsatility index (UtA-PI) and mean arterial pressure (MAP). This was a prospective, non-intervention study in women attending for an ultrasound scan at 19-24 weeks as part of routine pregnancy care. Patient-specific risks of delivery with PE at < 36 weeks' gestation were calculated using the competing-risks model to combine the prior distribution of gestational age at delivery with PE, obtained from maternal characteristics and medical history, with multiples of the median values of UtA-PI, MAP, PlGF and sFlt-1. Different risk cut-offs were used to vary the proportion of the population stratified into each of four risk categories (very high risk, high risk, intermediate risk and low risk) with the intention of detecting about 80%, 85%, 90% and 95% of cases of delivery with PE at < 28, < 32 and < 36 weeks' gestation. The performance of screening was assessed by plotting the detection rate against the screen-positive rate and calculating the areas under these curves, and by the proportion stratified into a given group for fixed detection rates. Model-based estimates of screening performance for these various combinations of markers were also produced. In the study population of 37 886 singleton pregnancies, there were 1130 (3.0%) that subsequently developed PE, including 160 (0.4%) that delivered at < 36 weeks' gestation. In both the modeled and empirical results, there was incremental improvement in the performance of screening with the addition of PlGF and sFlt-1 to the combination of maternal factors, UtA-PI and MAP. If the objective of screening was to identify about 90% of cases of PE with delivery at < 28, < 32 and < 36 weeks and the method of screening was a combination of maternal factors, UtA-PI and MAP, the respective screen-positive rates would be 3.1%, 8.5% and 19.1%. The respective values for screening by maternal factors, UtA-PI, MAP and PlGF were 0.2%, 0.7% and 10.6%, and for screening by maternal factors, UtA-PI, MAP, PlGF and sFlt-1 they were 0.1%, 0.4% and 9.5%. The empirical results were consistent with the modeled results. There was good agreement between the predicted risk and the observed incidence of PE at < 36 weeks' gestation for all three strategies of screening. Prediction of PE at ≥ 36 weeks was poor for all three screening methods, with the detection rate, at a 10% screen-positive rate, ranging from 33.2% to 38.4%. The performance of screening at 19-24 weeks' gestation for PE with delivery at < 28, < 32 and < 36 weeks' gestation achieved by a combination of maternal demographic characteristics and medical history, UtA-PI and MAP is improved by the addition of serum PlGF and sFlt-1. The performance of screening for PE at ≥ 36 weeks' gestation is poor irrespective of the method of screening at 19-24 weeks. © 2021 International Society of Ultrasound in Obstetrics and Gynecology.