Mitophagy-Oxidative Stress Molecular Subtypes Define an Immunosuppressive Ecosystem and Vulnerabilities in...

Glioblastoma (GBM) is an aggressive brain tumour with an immunosuppressive environment and poor prognosis; however, the roles of mitophagy and oxidative stress in its prognosis remain underexplored. This study analysed multi-omics data from TCGA-GBM (n = 168) and two GEO cohorts (GSE43378, n = 50; GSE147352, n = 85), compiling 603 mitophagy and oxidative stress-related genes. Unsupervised consensus clustering identified molecular subtypes, and prognostic genes were found via univariate Cox regression, leading to a refined gene signature using LASSO. The tumour immune microenvironment was characterized with ESTIMATE, CIBERSORT, and ssGSEA, while immunotherapy response was predicted using TIDE and IPS. Drug sensitivity was evaluated with GDSC and CTRP data, and a clinical nomogram integrating the gene signature and clinical variables was constructed and validated. Two molecular subtypes, C1 and C2, showed different prognoses (HR = 0.64, p = 0.011) and immune profiles. A 7-gene signature indicated high-risk patients had worse survival (p < 0.001) and an immunosuppressive environment. The risk score correlated with anti-PD-1 response (p < 0.05) and predicted therapy sensitivity. The signature was an independent prognostic factor (HR = 3.37, p < 0.001), and the nomogram accurately predicted survival at 1, 3, and 5 years. An innovative prognostic signature from mitophagy and oxidative stress genes stratifies GBM patients, characterizes the immunosuppressive microenvironment, and identifies therapeutic vulnerabilities for personalized treatment.