Extensive effort has been put on miRNA expression signatures in epithelial ovarian cancer (EOC). Unfortunately, consistent conclusion rarely yielded from diverse studies, mainly due to the high inter-lab variability and small sample sizes. To overcome above limitations, an integrated analysis of miRNA expression signature was performed by employing Robust Rank Aggregation (RRA) method. Diagnostic analysis, Kaplan-Meier survival curves and pathway enrichment analysis were used to investigate the clinical values and biological functions of meta-signature miRNAs. A total of 519 EOC and 248 noncancerous samples were included. Seven mostly dysregulated miRNAs were identified by RRA method and two miRNAs (miR-200a-3p and miR-200c-3p) remained statistically significant after Bonferroni-correction. Diagnostic meta-analysis showed reliable diagnostic capacity of miR-200a-3p (with a pooled sensitivity of 0.84 and specificity of 0.83) and miR-200c-3p (with a pooled sensitivity of 0.75 and specificity of 0.66) for EOC. Pathway enrichment analysis and expression correlation analysis suggested miR-200a/c might contribute EOC progression by affecting cellular adhesion process. Kaplan-Meier survival analysis based on two independent cohorts revealed a strong association between miR-200a/c and overall survival in EOC patients. miR-200a/c was identified as the mostly dysregulated miRNAs in EOC and might be novel diagnostic and prognostic biomarkers for patients with EOC.
Keywords: Robust Rank Aggregation; epithelial ovarian cancer (EOC); miRNA-200a/c.