Angiogenic biomarkers for prediction of early preeclampsia onset in high-risk women

Tiffany A. Moore Simas, University of Massachusetts Medical School
Sybil L. Crawford, University of Massachusetts Medical School
Susanne Bathgate, George Washington University School of Medicine & Health Sciences
Jing Yan, University of Massachusetts Medical School
Laura Robidoux, University of Massachusetts Medical School
Melissa J. Moore, University of Massachusetts Medical School
Sharon E. Maynard, George Washington University

Document Type Article


Objective: Chronic hypertension, pregestational diabetes mellitus, history of prior preeclampsia and obese nulliparity are maternal conditions associated with increased preeclampsia risk. Whether altered maternal angiogenic factor levels allow for prediction of pending disease is unclear. Our objective was to evaluate angiogenic factors for early preeclampsia prediction in high-risk women.

Methods: Serial serum specimens were collected from 157 women at high preeclampsia risk and 50 low-risk controls between 23 and 36 weeks gestation in 3 windows (23-27.6, 28-31.6, and 32-35.6 weeks) in a two-center observational cohort. Soluble fms-like tyrosine kinase-1 (sFlt1), placental growth factor (PlGF) and soluble endoglin (sEng) were measured by ELISA.

Results: Multivariate parsimonious logistic regression analyses using backward elimination for prediction of early-preeclampsia (diagnosed < 34 weeks) found the best-fitting model included the predictors (1) sFlt1 measured in the second window (28-31.6 weeks) with AUC 0.85, sensitivity 67% and specificity 96% and (2) sFlt1 measured in the first window (23-27.6 weeks) and sEng change between first and second window with AUC 0.91, sensitivity 86% and specificity 96%.

Conclusions: Two-stage sampling screening protocol utilizing sFlt1 and sEng is promising for prediction of preeclampsia diagnosed before 34 weeks. Larger studies are needed to confirm these findings.