Development and validation of an inflammatory response-related prognostic model and immune infiltration analysis in glioblastoma.

Background: Despite receiving standard treatment, the prognosis of glioblastoma (GBM) patients is still poor. Considering the heterogeneity of each patient, it is imperative to identify reliable risk model that can effectively predict the prognosis of each GBM patient to guide the personalized treatment. Methods: Transcriptomic gene expression profiles and corresponding clinical data of GBM patients were downloaded from The Cancer Genome Atlas (TCGA) and Chinese Glioma Genome Atlas (CGGA) databases. Inflammatory response-related genes were extracted from Gene Set Enrichment Analysis (GSEA) website. Univariate Cox regression analysis was used for prognosis-related inflammatory genes (P<0.05). A polygenic prognostic risk model was constructed using least absolute shrinkage and selection operator (LASSO) Cox regression analysis. Validation was performed through CGGA cohort. Overall survival (OS) was compared by Kaplan-Meier analysis. A nomogram was plotted to accurately predict the prognosis for each patient. GSEA was used for the pathway enrichment analysis. The single sample GSEA (ssGSEA) algorithm was implemented to conduct the immune infiltration analysis. The potential role of oncostatin M receptor (OSMR) in GBM was investigated through the in vitro experiment. Results: A prognostic risk model consisting of 4 genes (PTPRN, OSMR, MYD88, and EFEMP2) was developed. GBM patients in the high-risk group had worse OS. The time-dependent ROC curves showed an area under the curve (AUC) of 0.782, 0.765, and 0.784 for 1-, 2-, and 3-year survival in TCGA cohort, while the AUC in the CGGA cohort was 0.589, 0.684, and 0.785 at 1, 2, and 3 years, respectively. The risk score, primary-recurrent-secondary (PRS) type, and isocitrate dehydrogenase (IDH) mutation could predict the prognosis of GBM patients well. The nomogram accurately predicted the 1-, 2-, and 3-year OS for each patient. Immune cell infiltration was associated with the risk score and the model could predict immunotherapy responsiveness. The expression of the prognostic gene was correlated with the sensitivity to antitumor drugs. Interference of OSMR inhibited proliferation and migration and promoted apoptosis of GBM cells. Conclusions: The prognostic model based on 4 inflammatory response-related genes had reliable predictive power to effectively predict clinical outcome in GBM patients and provided the guide for the personalized treatment.