First trimester circulating miR-194-5p and miR-1278 improve prediction of preeclampsia

Abstract

ObjectivesTo identify first trimester circulating microRNAs associated with preeclampsia (PE) and assess their predictive value in two independent cohorts MethodsCirculating microRNAs were quantified from plasma samples collected at first trimester of pregnancy in women from Gen3G (discovery; N = 385 normotensives/22 PE) and 3D (replication; N = 260 normotensives/24 PE) prospective birth cohorts. MicroRNAs associated with PE in Gen3G were identified using DESeq2 (p-value ≤ 0.05). Replicated microRNAs (selection criteria: p-value ≤ 0.05 in Gen3G, same direction of association, nominal one-side p-value ≤ 0.1 in 3D) were included in a stepwise logistic regression model to assess their predictive values alone and in combination with PE risk factors. ResultsSeventy-three circulating microRNAs were associated with PE (p-value ≤ 0.05) in Gen3G. Five microRNAs were replicated in 3D and included in a stepwise logistic regression model with PE clinical risk factors (maternal age, body mass index and mean arterial pressure (MAP) at first trimester, parity, and smoking status) and gestational age at first visit. The best model included miR-194-5p, miR-1278, maternal age, MAP at first trimester and parity and results in an area under the curve (AUC) of 0.861 [CI 95 %: 0.787–0.935] in Gen3G. Compared to risk factors only, the addition of microRNAs improves the AUC (from 0.826 to 0.861; p = 0.03). In 3D, the best model reached an AUC of 0.842 [CI 95 %: 0.769–0.914]. ConclusionsCirculating miR-194-5p and miR-1278 enhance early detection of women at risk of PE and offer great potential as predictors of PE in combination with classic risk factors.