Search

Glioblastoma (GBM) is resistant to multimodality therapeutic approaches. A high burden of tumor-specific mutant peptides (neoantigens) correlates with better survival and response to immunotherapies in selected solid tumors but how neoantigens impact clinical outcome in GBM remains unclear. Here, we exploit the similarity between tumor neoantigens and infectious disease-derived immune epitopes and apply a neoantigen fitness model for identifying high-quality neoantigens in a human pan-glioma dataset. We find that the neoantigen quality fitness model stratifies GBM patients with more favorable clinical outcome and, together with CD8+ T lymphocytes tumor infiltration, identifies a GBM subgroup with the longest survival, which displays distinct genomic and transcriptomic features. Conversely, neither tumor neoantigen burden from a quantitative model nor the isolated enrichment of CD8+ T lymphocytes were able to predict survival of GBM patients. This approach may guide optimal stratification of GBM patients for maximum response to immunotherapy. Jing Zhang et al. analyze glioblastoma patient data with a computational model integrating mutation-derived neoantigens similar to pathogen-specific antigens with enrichment of signatures for individual immune cells. Using a fitness model for identifying high-quality neoantigens and T lymphocyte infiltrates in glioblastoma, they uncover a subset of patients with prolonged survival. [ABSTRACT FROM AUTHOR]