MiR-106a is an independent prognostic marker in patients with glioblastoma

Background. Very little is known regarding correlation of micro RNA (miR)–106a with clinical outcomes of patients with glioblastoma multiforme (GBM). This study determined whether miR-106a could be used as an independent prognostic biomarker in those patients. Methods. A total of 156 GBM patients were divided into 2 cohorts. In the first cohort, matched fresh frozen and formalin-fixed paraffin-embedded (FFPE) samples were collected from 24 GBM patients, while in the second cohort, only FFPE samples were collected from 132 GBM patients. MiR-106a expression levels were examined by quantitative real-time PCR in the 2 cohorts and further validated by in situ hybridization assay in the second cohort. The correlation between miR-106a expression levels and overall survival was evaluated in the second cohort of 114 GBM patients available for follow-up by a log-rank test and a multivariate Cox proportional hazards model. Results. Our data showed a very good correlation of miR-106a or U6 expression between fresh frozen and FFPE GBM specimens, with Pearson’s correlation coefficients of 0.849 and 0.823, respectively (P , .001). Their expression levels in archival FFPE samples were quite stable for at least 7 years when stored at room temperature. Multivariate analysis revealed that the expression level of miR-106a was an independent and significant predictor of overall survival in GBM patients (P ¼ .011). Conclusions. MiR-106a expression was relatively abundant and stable in a large cohort of archival FFPE GBM specimens and could be used as an independent prognostic biomarker in those patients. Thus, miR-106a can be used to predict prognosis and treatment response in individual GBM patients.