Development and validation of a glioma prognostic model based on telomere-related genes and immune infiltration analysis.

Abstract

Background: Gliomas are the most prevalent primary brain tumors, and patients typically exhibit poor prognoses. Increasing evidence suggests that telomere maintenance mechanisms play a crucial role in glioma development. However, the prognostic value of telomere-related genes in glioma remains uncertain. This study aimed to construct a prognostic model of telomere-related genes and further elucidate the potential association between the two.

Methods: We acquired RNA-seq data for low-grade glioma (LGG) and glioblastoma (GBM), along with corresponding clinical information from The Cancer Genome Atlas (TCGA) database, and normal brain tissue data from the Genotype-Tissue Expression (GTEX) database for differential analysis. Telomere-related genes were obtained from TelNet. Initially, we conducted a differential analysis on TCGA and GTEX data to identify differentially expressed telomere-related genes, followed by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses on these genes. Subsequently, univariate Cox analysis and log-rank tests were employed to obtain prognosis-related genes. Least absolute shrinkage and selection operator (LASSO) regression analysis and multivariate Cox regression analysis were sequentially utilized to construct prognostic models. The model's robustness was demonstrated using receiver operating characteristic (ROC) curve analysis, and multivariate Cox regression of risk scores for clinical characteristics and prognostic models were calculated to assess independent prognostic factors. The aforementioned results were validated using the Chinese Glioma Genome Atlas (CGGA) dataset. Finally, the CIBERSORT algorithm analyzed differences in immune cell infiltration levels between high- and low-risk groups, and candidate genes were validated in the Human Protein Atlas (HPA) database.

Results: Differential analysis yielded 496 differentially expressed telomere-related genes. GO and KEGG pathway analyses indicated that these genes were primarily involved in telomere-related biological processes and pathways. Subsequently, a prognostic model comprising ten telomere-related genes was constructed through univariate Cox regression analysis, log-rank test, LASSO regression analysis, and multivariate Cox regression analysis. Patients were stratified into high-risk and low-risk groups based on risk scores. Kaplan-Meier (K-M) survival analysis revealed worse outcomes in the high-risk group compared to the low-risk group, and establishing that this prognostic model was a significant independent prognostic factor for glioma patients. Lastly, immune infiltration analysis was conducted, uncovering notable differences in the proportion of multiple immune cell infiltrations between high- and low-risk groups, and eight candidate genes were verified in the HPA database.

Conclusions: This study successfully constructed a prognostic model of telomere-related genes, which can more accurately predict glioma patient prognosis, offer potential targets and a theoretical basis for glioma treatment, and serve as a reference for immunotherapy through immune infiltration analysis.