Circulating microRNAs as biomarkers for diagnosis and prediction of preeclampsia: A systematic review and meta-analysis

Abstract

ObjectiveWe carried out a meta-analysis to quantitatively summarize the overall diagnostic and predictive effects of circulating microRNAs in diagnosis and prediction of preeclampsia, respectively. Study designWe screened selected databases and systematically retrieved articles until September 20th, 2019 for analysis. After literature screening and data extraction, we firstly conducted quality assessment according to QUADAS-2 score system. And then the pooled diagnostic and predictive parameters were calculated using a bivariate random-effect meta-analysis model. We used threshold effect analysis and subgroup analysis to identify the sources of heterogeneity. The clinical utility was validated through the Fagan’s Nomogram. Sensitivity analysis was performed to assess the reliability of each included study, and we evaluated publication bias with the Deeks’ funnel plot asymmetry test. ResultsThe meta-analysis included 8 articles comprising 273 preeclampsia patients and 343 normal pregnancies. Pooled results of diagnostic values of 5 articles indicated a sensitivity of 0.88 (95 %CI: 0.80−0.93), specificity of 0.87 (95 %CI: 0.78−0.92) and diagnostic odds ratio of 50.24 (95 %CI: 21.28–118.62). The pooled sensitivity, specificity, DOR of circulating microRNAs for predicting preeclampsia of asymptomatic pregnancies were 0.61 (95 %CI: 0.55−0.68), 0.78 (95 %CI: 0.72−0.83) and 5.7 (95 %CI: 3.7–8.7) across other 3 articles. Subgroup analysis revealed that non-plasma specimen type, non-U6 reference gene and non-Asian had better diagnostic value while due to limited data, we did not conduct a subgroup analysis of predictive value. ConclusionsCirculating miRNAs distinguish patients with preeclampsia from controls with relatively high diagnostic and predictive accuracy. Then we conclude that circulating miRNAs could be a useful screening tool to diagnose and predict preeclampsia. However, its utility should be judged with caution and large-sample prospective studies are warranted to explore if its implementation improves maternal and neonatal outcomes.