Multi-omics-guided characterization of neoantigens improves patient stratification and potentiates combinatorial immunotherapy in glioma

The identification of tumor-derived neoantigens that elicit an immune response has led to significant advancements in cancer immunotherapy. We comprehensively investigated the neoantigen atlas in Chinese glioma patients via the integration of whole-exome sequencing, transcriptome, and ribosome profiling. We further proposed MINER (Multi-omics Integration for NEoantigen Recognition) to prioritize neoantigens that integrate HLA-I binding and the TCR response to increase accuracy. In particular, we emphasized the importance of non-canonical open reading frames in neoantigen generation and demonstrated that both gene fusions and somatic mutations contribute to the production of neoantigens complementally. We identified a novel subtype (neo_W) of IDH-wildtype glioma, which was characterized by a higher neoantigen burden and correlated with better survival outcomes. Finally, we experimentally validated therapeutic combinations of immune checkpoint inhibitors that target the RAS-RAF-MEK-ERK signaling pathway for treatment of neo_W glioma patients. Multi-omics-guided characterization of neoantigens provides valuable knowledge for precision therapy of glioma.