A RNA sequencing-based six-gene signature for survival prediction in patients with glioblastoma

Glioblastoma (GBM) is an aggressive tumor of the central nervous system that has poor prognosis despite extensive therapy. Therefore, it is essential to identify a gene expression-based signature for predicting GBM prognosis. The RNA sequencing data of GBM patients from the Chinese Glioma Genome Atlas (CGGA) and The Cancer Genome Atlas (TCGA) databases were employed in our study. The univariate and multivariate regression models were utilized to assess the relative contribution of each gene to survival prediction in both cohorts, and the common genes in two cohorts were identified as a final prognostic model. A prognostic risk score was calculated based on the prognostic gene signature. This prognostic signature stratified the patients into the low- and high-risk groups. Multivariate regression and stratification analyses were implemented to determine whether the gene signature was an independent prognostic factor. We identified a 6-gene signature through univariate and multivariate regression models. This prognostic signature stratified the patients into the low- and high-risk groups, implying improved and poor outcomes respectively. Multivariate regression and stratification analyses demonstrated that the predictive value of the 6-gene signature was independent of other clinical factors. This study highlights the significant implications of having a gene signature as a prognostic predictor in GBM, and its potential application in personalized therapy.