Prospective observational study to determine the accuracy of first-trimester serum biomarkers and uterine artery Dopplers in combination with maternal characteristics and arteriography for the prediction of women at risk of preeclampsia and other adverse pregnancy outcomes

Abstract

Objectives: To assess the efficacy of biomarkers, arteriography and uterine artery Dopplers for predicting hypertensive disease of pregnancy, small for gestational age (SGA) and stillbirth.

Methods: This was a prospective first-trimester study. Ultrasound was used to assess uterine artery Doppler. Maternal arteriography was performed and serum was taken for the measurement of placental growth factor (PlGF), alpha-fetoprotein (AFP), pregnancy-associated plasma protein (PAPP-A) and beta-human chorionic gonadotrophin levels. Logistic regression with stepwise selection was performed to determine multivariate models.

Results: One thousand and forty-five women were left for analysis after exclusions. Fourteen developed preeclampsia, 23 pregnancy induced hypertension, 64 SGA <5th centile, 118 SGA <10th centile and three stillbirth. Systolic blood pressure (SBP) in the aorta (SBPAO) (p = .002) was significantly associated with preeclampsia. Detection rate (DR) was 72% for a false-positive rate (FPR) of 15%, an area under the curve (AUC) of 0.81, 95% CI 0.69–0.93. MAP and maternal weight (p = .001) were significantly associated with PIH. DR 48%, AUC 0.76, 95% CI 0.65–0.86. Low PAPP-A and PlGF were significantly associated with SGA <10th centile (p = .007 and .004, respectively), DR 30%, AUC 0.608, 95% CI 0.54–0.68. SGA <5th centile was significantly associated with low PlGF (p = <.001), DR 57%, AUC 0.73, 95% CI 0.65–0.80.

Conclusions: No association was found between first-trimester biomarkers and preeclampsia/PIH. There was a significant association between low PlGF and PAPP-A and SGA.

Keywords:

• Biomarkers
• preeclampsia
• pregnancy-induced hypertension
• small for gestational age
• uterine artery Doppler

Acknowledgements

The authors would like to acknowledge Jonathan Bestwick for the statistical analysis and Professor Shakila Thangaratinam for her support.

Disclosure statement

The authors report no conflicts of interest.