Immunological classification of gliomas based on immunogenomic profiling

BACKGROUND: Gliomas are heterogeneous in the tumor immune microenvironment (TIM). However, a classification of gliomas based on immunogenomic profiling remains lacking. METHODS: We hierarchically clustered gliomas based on the enrichment levels of 28 immune cells in the TIM in five datasets and obtained three clusters: immunity-high, immunity-medium, and immunity-low. RESULTS: Glioblastomas were mainly distributed in immunity-high and immunity-medium, while lower-grade gliomas were distributed in all the three subtypes and predominated in immunity-low. Immunity-low displayed a better survival than other subtypes, indicating a negative correlation between immune infiltration and survival prognosis in gliomas. IDH mutations had a negative correlation with glioma immunity. Immunity-high had higher tumor stemness and epithelial-mesenchymal transition scores and included more high-grade tumors than immunity-low, suggesting that elevated immunity is associated with tumor progression in gliomas. Immunity-high had higher tumor mutation burden and more frequent somatic copy number alterations, suggesting a positive association between tumor immunity and genomic instability in gliomas. CONCLUSIONS: The identification of immune-specific glioma subtypes has potential clinical implications for the immunotherapy of gliomas. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12974-020-02030-w.