Differential genes and scoring criteria among immunogenomic clusters of lower-grade gliomas

High incidence of recurrency had been a significant threat among glioma patients. Moreover, the performance of traditional therapies among recurrent gliomas was far from satisfying. Advances in the tumor microenvironment (TME) and immune responses on the brain inspired immunotherapy researches. Nevertheless, verification of classic PD-1/PD-L1 inhibitors failed in phase III clinical trials. Additional gene targets were required for future studies among glioma patients. Immune cell infiltration (ICI) scores, defined based on multiple prognostic genes, were proved as the marker for the sensitivity of immunotherapies in many tumors. However, relevant results were not reported in gliomas. In the study, a retrospective cohort of 495 patients was classified into two ICI score subgroups. High ICI scores were closely related to high tumor mutation burden (TMB) values, indicating a high instability of genes. Furthermore, ICI scores were proved as reliable prognostic predictors. And a predictive model was built based on the ICI scores and multiple clinical features. The model showed its superiority through both internal validation and external validation. The ICI scores and the predictive model showed significant clinical values through decision curve analysis (DCA) since high ICI scores were related to high sensitivity for treatment. The prognostic immune-related gene list provided targets for immunotherapy researches.