Prediction of preterm pre‐eclampsia at midpregnancy using a multivariable screening algorithm

• Published in: Australian & New Zealand journal of obstetrics & gynaecology Vol. 60; no. 5; pp. 675 - 682
• Main Authors: Black, Carin, Rolnik, Daniel Lorber, Al‐Amin, Ahmed, Kane, Stefan C., Stolarek, Caroline, White, Adrienne, Da Silva Costa, Fabricio, Brennecke, Shaun
• Format: Journal Article
• Language: English
• Published: Australia 01.10.2020
• Subjects: Algorithms; Biomarkers; Female; Humans; Infant, Newborn; multivariable algorithm; Placenta Growth Factor; placental growth factor (PlGF); Pre-Eclampsia - diagnosis; prediction; Predictive Value of Tests; Pregnancy; pre‐eclampsia; Prospective Studies; second trimester; Vascular Endothelial Growth Factor Receptor-1; prediction; placental growth factor (PlGF); pre-eclampsia; multivariable algorithm; second trimester
• ISSN: 0004-8666; 1479-828X; 1479-828X
• DOI: 10.1111/ajo.13113

Background Competing risk models used for midpregnancy prediction of preterm pre‐eclampsia have shown detection rates (DR) of 85%, at fixed false‐positive rate (FPR) of 10%. The full algorithm used between 19+0 and 24+6 weeks includes maternal factors, mean arterial pressure (MAP), mean uterine artery pulsatility index (UtAPI), serum placental growth factor (PlGF) level in multiples of the median (MoM), and soluble Fms‐like tyrosine kinase‐1 (sFlt‐1) level in MoM. Aims To assess performance of the Fetal Medicine Foundation (FMF) algorithm at midpregnancy to screen for preterm (<37 weeks) pre‐eclampsia. The outcome measured was preterm pre‐eclampsia. Materials and Methods This is a prospective study including singleton pregnancies at 19–22 weeks gestation. Maternal bloods were collected and analysed using three different immunoassay platforms. Maternal characteristics, medical history, MAP, mean UtAPI, serum PlGF MoM and serum sFlt‐1 MoM were used for risk assessment. DR and FPR were calculated, and receiver operating characteristic curves produced. Results Five hundred and twelve patients were included. Incidence of preterm pre‐eclampsia was 1.6%. Using predicted risk of pre‐eclampsia of one in 60 or more and one in 100 or higher, as given by the FMF predictive algorithm, the combination with the best predictive performance for preterm pre‐eclampsia included maternal factors, MAP, UtAPI and PlGF MoM, giving DRs of 100% and 100%, respectively, and FPRs of 9.3 for all platforms and 12.9–13.5, respectively. Addition of sFlt‐1 to the algorithm did not appear to improve performance. sFlt‐1 MoM and PlGF MoM values obtained on the different platforms performed very similarly. Conclusions Second trimester combined screening for preterm pre‐eclampsia by maternal history, MAP, mean UtAPI and PlGF MoM using the FMF algorithm performed very well in this patient population.