Histone acetylation risk model predicts prognosis and guides therapy selection in glioblastoma: implications for chemotherapy and anti-CTLA-4 immunotherapy

Abstract

Glioblastoma is characterized by high aggressiveness, frequent recurrence, and poor prognosis. Histone acetylation-associated genes have been implicated in its occurrence and development, yet their predictive ability in glioblastoma prognosis remains unclear. This study constructs a histone acetylation risk model using Cox and LASSO regression analyses to evaluate glioblastoma prognosis. We assessed the model’s prognostic ability with univariate and multivariate Cox regression analyses. Additionally, immune infiltration was evaluated using ESTIMATE and TIMER algorithms, and the SubMAP algorithm was utilized to predict responses to CTLA4 inhibitor. Multiple drug databases were applied to assess drug sensitivity in high- and low-risk groups. Our results indicate that the histone acetylation risk model is independent and reliable in predicting prognosis. Low-risk patients showed higher immune activity and longer overall survival, suggesting anti-CTLA4 immunotherapy suitability, while high-risk patients might benefit more from chemotherapy. This model could guide personalized therapy selection for glioblastoma patients.