Effect of race on the measurement of angiogenic factors for prediction and diagnosis of pre‐eclampsia

Abstract

Objective To examine the effect of self‐declared race on serum placental growth factor (PlGF) and sFlt‐1/PlGF ratio and the impact on pre‐eclampsia (PE) prediction. Design Prospective observational study. Setting Two UK maternity hospitals. Population 29 035 women with singleton pregnancies attending a routine 35+0 to 36+6 weeks' gestation hospital visit, including 654 (2.3%) who subsequently developed PE. Methods The predictive performance of PlGF and sFlt‐1/PlGF for PE in minority racial groups (versus white) was examined. Main outcome measure Delivery with PE. Results Compared with white women, mean PlGF was higher and sFlt‐1/PlGF ratio lower in black, South Asian, East Asian and mixed race women. In white women at a PlGF concentration cut‐off corresponding to a screen‐positive rate (SPR) of 10%, detection rates (DRs) were 49.1% for PE at any time and 72.3% for PE within 2 weeks after screening. In black women, at the same PlGF concentration cut‐off for white women, the SPR was 5.5%, and DRs 33.6% and 55.0%, respectively; the number of PE cases was too small to evaluate screening performance in other racial groups. Using a fixed cut‐off in sFlt‐1/PlGF ratio to identify women at risk of developing PE, similarly diagnostically disadvantaged black women. Bias was overcome by adjusting metabolite concentrations for maternal characteristics and use of the competing risks model to estimate patient‐specific risks. Conclusion Screening for PE with fixed cut‐offs in PlGF or sFlt‐1/PlGF diagnostically disadvantages black women. It is essential that measured levels of PlGF be adjusted for race as well as other maternal characteristics.

A Comprehensive and Bias-Free Machine Learning Approach for Risk Prediction of Preeclampsia with Severe Features in a Nulliparous Study Cohort

“…However, after correcting for this imbalance, the bias still persisted. We then hypothesize that this bias might come from a difference in the distribution of values for the top placental analytes, as suggested in another study [40]. We did observe signi cant differences in the distribution of top predictive features (P<0.001), such as BMI and PLGF (V1, V2).…”

Incidence of pre‐eclampsia: effect of deprivation

“…Similarly, race is not included in NICE guidance for the use of angiogenic markers in the assessment of women with suspected PE [19][20][21][22] . Given the higher levels of angiogenic markers in black women, they are disadvantaged by the use of fixed cut-offs, particularly when PlGF is used alone (rather than as part of the sFlt-1/PlGF ratio) 16 .…”

“…However, the relationship between race and PE is not accounted for by deprivation. Previous work suggests that the relationship is at least partially accounted for by biology; black women have significantly higher serum levels of angiogenic markers, including placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1) 16 . Thus, while some component of the relationship between race and PE may be modifiable, based on the care sought or received, it is essential that race is included both in screening algorithms for PE (as in the Fetal Medicine Foundation competing-risks model 3,5 ) and when interpreting the results of biochemical testing for PE diagnosis.…”

Prediction of hypertensive disorders after screening at 36 weeks' gestation: comparison of angiogenic markers with competing‐risks model

“…We used automated machines to provide accurate measurement, within 40 min of sampling, of maternal serum concentration of PlGF and sFlt‐1. Also, biomarker values were expressed as MoMs after adjustment for maternal factors (such as ethnicity 21 ) and reagents used that affect the measurements. We used Bayes' theorem to combine the prior risk from maternal factors with MoM values of biomarkers to estimate patient‐specific risk and the performance of predicting delivery with PE at different timepoints after assessment.…”