Identification of an Immune-Related Prognostic Signature for Glioblastoma by Comprehensive Bioinformatics and Experimental Analyses

Background: Glioblastoma (GBM), which has a poor prognosis, accounts for 31% of all cancers in the brain and central nervous system. There is a paucity of research on prognostic indicators associated with the tumor immune microenvironment in GBM patients. Accurate tools for risk assessment of GBM patients are urgently needed.

Methods: In this study, we used weighted gene co-expression network analysis (WGCNA) and differentially expressed gene (DEG) methods to screen out GBM-related genes among immune-related genes (IRGs). Then, we used survival analysis and Cox regression analysis to identify prognostic genes among the GBM-related genes to further establish a risk signature, which was validated using methods including ROC analysis, stratification analysis, protein expression level validation (HPA), gene expression level validation based on public cohorts, and RT-qPCR. In order to provide clinicians with a useful tool to predict survival, a nomogram based on an assessment of IRGs and clinicopathological features was constructed and further validated using DCA, time-dependent ROC curve, etc. Results: Three immune-related genes were found: PPP4C (p < 0.001, HR = 0.514), C5AR1 (p < 0.001, HR = 1.215), and IL-10 (p < 0.001, HR = 1.047). An immune-related prognostic signature (IPS) was built to calculate risk scores for GBM patients; patients classified into different risk groups had significant differences in survival (p = 0.006). Then, we constructed a nomogram based on an assessment of the IRG-based signature, which was validated as a potential prediction tool for GBM survival rates, showing greater accuracy than the nomogram without the IPS when predicting 1-year (0.35 < Pt < 0.50), 3-year (0.65 < Pt < 0.80), and 5-year (0.65 < Pt < 0.80) survival.

Conclusions: In conclusion, we integrated bioinformatics and experimental approaches to construct an IPS and a nomogram based on IPS for predicting GBM prognosis. The signature showed strong potential for prognostic prediction and could help in developing more precise diagnostic approaches and treatments for GBM.