Potential serum biomarkers for predicting premature rupture of fetal membranes in the first trimester

Objective: To invest an accurate and timely method for predicting the occurrence of premature rupture of the fetal membranes (PROM) in the first trimester. Design Prospective cohort study from gestational week 5 to 14. Setting Shunyi District Maternal and Child Health Hospital in Beijing, P. R. China. Population or Sample We collected serum samples and clinical information from 41 pregnant women who developed PROM later in gestation and 106 healthy pregnant women as a control group for the cross-sectional study. Methods Untargeted metabolomics was applied to characterize metabolite profiles related to PROM in early pregnancy. Logistic regression analysis was adjusted to analyze a PROM prediction model in the first trimester. A WISH amniotic cell viability assay was applied to explore the underlying mechanisms involved in PROM. Main Outcome Measures Serum differential metabolites in PROM cohorts comparing to healthy control. Results Compared with healthy controls, 13 serum metabolites were identified. The prediction model comprising four compounds (Cer 40:0; O2, sphingosine, isohexanal and PC O-38:4) had moderate accuracy to predict PROM events with the maximum area under the curve of a receiver operating characteristics curve of approximately 0.70. Of these four compounds, Cer 40:0; O2 with an 1.81-fold change between PROM and healthy control serum samples was defined as a potential biomarker and inhibited the viability of WISH cells. Conclusions Thirteen serum differential metabolites were identified in these women with PROM. The prediction model comprising four compounds might help predict PROM events in the first trimester.