A Promising Glycolysis- and Immune-Related Prognostic Signature for Glioblastoma.

Background: Glioblastoma multiforme (GBM) is a malignant brain tumor with a poor prognosis. Aerobic glycolysis and an immunosuppressive microenvironment are potentially correlated with progression of GBM. However, the prognostic value of glycolysis-immune-related genes has not been studied in GBM.

Methods: Using GBM-related data downloaded from Chinese Glioma Genome Atlas database, the overlapped differentially expressed genes were identified between the GBM patients with a different glycolysis status and immune score, which had also undergone functional enrichment. Univariate Cox regression analysis and LASSO (least absolute shrinkage and selection operator) Cox regression analysis were used for risk score construction. Multivariate Cox regression analysis and survival analysis determined the independent prognostic factors.

Results: We found 277 overlapped differentially expressed genes between high glycolysis and low glycolysis, a high immune score and low immune score, and a combination of low glycolysis status and a low immune score and high glycolysis status and a high immune score. These were significantly enriched in 301 gene otology terms and 25 Kyoto Encyclopedia of Genes and Genomes pathways. Of these, 8 genes were found to be optimal for building a risk score. The risk score was an independent prognostic factor for GBM patients, and patients with a high score had a worse prognosis. Moreover, between the high- and low-risk GBM patients, 17 types of immune cells were differentially infiltrated, and 5 immune checkpoints were differentially expressed.

Conclusions: The glycolysis-immune-related risk score using CACNG2, CSMD3, GABRA3, KCNIP2, KSR2, PTPRT, TNFRSF12A, and TNR was able to predict the prognosis of GBM patients relatively reliably.