MicroRNA-21 (miRNA-21) has recently been shown to be a promising prognostic tumor biomarker. However, few studies have not supported this idea and have shown inconsistent data. Thus, we conducted a meta-analysis to elucidate the predictive value of miRNA-21 in gliomas. The relevant studies were identified by performing online search in PubMed, EMBASE and Web of Science databases up to Apr 2016. This meta-analysis study included seven eligible studies, consisting of 1121 gliomas and 533 glioblastoma multiforme (GBM) patients. Heterogeneity between studies was assessed using Egger's and Begg's test. Hazard ratios (HRs) with 95 % confidence intervals (CIs) for overall survival (OS), which compared the expression levels of miRNA-21 in patients with gliomas, were extracted and estimated. Our analysis revealed that the high expression of miRNA-21 is associated with the worse OS in gliomas. Further subgroup analysis indicated that increased expression of miRNA-21 was also associated with OS in GBM patients. Moreover, we observed a correlation between miRNA-21 expression and the World Health Organization defined gliomas grading system (WHO grade). Besides, high miRNA-21 expression was significantly correlated with lowered OS both in the Asian group and non-Asian group. In the cut-off subgroup analysis, both mean cut off value and median cut off value were significantly associated with OS. The expression level of miRNA-21 was not high in low KPS (Karnofsky score) group. miRNA-21 appears to be a promising biomarker for predicting the progression of patients with gliomas or GBM. However, due to the limited sample size, further prospective or retrospective multi-center well designed studies with adequate sample size should be conducted to verify its definite prognostic value.
Keywords: Glioblastomas; Glioma; mir-21.