Ó³»­´«Ã½

Glioblastoma multiforme (GBM) shows profound inter- and intratumoral heterogeneity, but differences across ancestries remain understudied. We assemble the largest cohort of GBMs from East Asian patients (EAS-GBM) and perform genomic and transcriptomic analyses. Transcriptomic clustering reveals distinct EAS-GBM molecular subtypes-proliferative (PL), neurosynaptic (NS), metabolic (MB), and immunomodulatory (IM)-based on activated biological pathways. We also identify protein markers enabling subtype classification in pathology. Comparative analysis with the predominantly European-ancestry The Cancer Genome Atlas (TCGA) GBM dataset (EUR-GBM) reveals similar main drivers, though EAS-GBMs lack the epidermal growth factor receptor (EGFR)-defined classical EUR-GBM subtype. EAS PL-GBMs display CCNE1/CCND2 overexpression, with myelocytomatosis viral oncogene (MYC) signaling activation. The IM-GBM cluster exhibits an immunotherapy-responsive expression profile, corroborated by improved overall survival under immune checkpoint blockade in an external dataset. Together, our data reveal distinctive genomic features of EAS-GBMs that may inform patient stratification for targeted therapy and drug development tailored to the EAS population.