First-trimester markers for the prediction of pre-eclampsia in women with a-priori high risk

Abstract

Objective: To investigate the predictive value of the combination of first-trimester serum placental protein 13 (PP13), uterine artery Doppler pulsatility index (PI) and pulse wave analysis (augmentation index at a heart rate of 75 beats per min (AIx-75)), and to evaluate concurrent and contingent strategies using this combination for assessing the risk of pre-eclampsia in high-risk women.

Methods: In this nested case-control study, serum PP13, uterine artery mean PI and AIx-75 were measured at between 11 + 0 and 13 + 6 weeks' gestation in women at increased risk of pre-eclampsia. For each case of pre-eclampsia (n = 42), five matched controls were randomly selected from the study group. Gestation specific multiples of the median (MoMs) were adjusted for body mass index, ethnicity, smoking, age and parity. MoMs were compared between cases and controls using the Wilcoxon rank sum test. Sensitivities and specificities were derived from receiver-operating characteristics curves.

Results: Compared with controls, women who developed pre-eclampsia had lower PP13, higher uterine artery mean PI and higher AIx-75 (P < 0.001). For a 10% false-positive rate, the best detection rate for pre-eclampsia (85.7% (95% CI, 71.5-94.6%)) and pre-eclampsia requiring delivery before 34 weeks (92.9% (95% CI, 66.1-99.8%)) was achieved by concurrent testing with all three markers. The best contingency screening sequences for pre-eclampsia were (AIx-75 --> PP13 --> mean PI) and (PP13 --> AIx-75 --> mean PI), with an 86% detection rate for false-positive rates of 9 and 10%, respectively. These two sequences would require 410 and 414 tests, respectively, compared with 756 tests in concurrent testing.

Conclusion: Combination of first-trimester PP13, uterine artery mean PI and pulse-wave analysis is promising for the prediction of pre-eclampsia in women at increased a-priori risk and may be useful in clinical practice. Contingency screening achieved similar detection rates to concurrent testing, but required almost 50% fewer tests, making it a more cost-effective option.