Circulating miRNAs as novel diagnostic biomarkers in hepatocellular carcinoma detection: a meta-analysis based on 24 articles

The diagnostic value and suitability of circulating miRNAs for the detection of hepatocellular carcinoma have been inconsistent in the literature. A meta-analysis is used to systematically evaluate the diagnostic value of circulating miRNAs. Eligible studies were selected and the heterogeneity was assessed by subgroup analysis, meta-regression, and publication bias. After strictly and comprehensive screening, the source methods, internal reference and the cut-off values of the included miRNAs were first listed. Circulating miRNAs demonstrated a relatively good diagnostic value in hepatocellular carcinoma, In the subgroup analysis, diagnosis odds ratio showed a higher accuracy with multiple miRNAs than with a single miRNA as well as with serum types than plasma types. In addition, although miRNAs have many expression patterns, the high frequency expression miRNAs (miR-21, miR-199 and miR-122) might be more specific for the diagnosis of hepatocellular carcinoma. The sources of heterogeneity might be related to the number of miRNAs and the specimen types in meta-regression. Furthermore, it's surprised that the pooled studies were first demonstrated publication bias (P < 0.05). In conclusion, multiple miRNAs in serum have a better diagnostic value, and the publication bias was stable. To validate the potential applicability of miRNAs in the diagnosis of hepatocellular carcinoma, more rigorous studies are needed to confirm these conclusions.