A preeclampsia risk prediction model based on maternal characteristics and serum markers in twin pregnancy

Aim: To develop preeclampsia (PE) risk prediction model based on maternal characteristics and serum markers at the first and second trimester in the twin pregnancy. Methods: Between January 2005 and September 2017, we retrospectively reviewed medical records of 532 twin pregnant women who underwent maternal serum integrated test and gave birth at a Bundang CHA Medical Center. Maternal serum pregnancy-associated plasma protein A (PAPP-A) was determined at 10(+0) to 13(+6) weeks and the serum alpha-fetoprotein (MSAFP), human chorionic gonadotrophin (hCG), unconjugated estriol (uE(3)) and inhibin A were assayed at 14(+0) to 22(+0 )weeks. We assessed the relationships of maternal characteristics and serum markers by using multiple logistic regression analysis. Results: The study included 35 patients who diagnosed PE and a control group consisting of the other 497 patients in twin pregnancy. There were no significant differences in the maternal age and body mass index (BMI) between two groups. However, the gestational age and placenta weight at delivery were significantly different between groups (p < .001, p = .005, respectively. Among the maternal serum markers, inhibin A value was significantly higher in women with PE compared to those without preeclampsia (p < .001). In addition, we predicted the PE using maternal age, BMI, uE(3), and inhibin A which were achieving an area under the curve of 0.73 overall in twin pregnancy. Conclusion: A risk prediction model of PE which combined maternal age, BMI, uE(3), and inhibin A was better early predictors than any individual marker twin pregnancy.