Molecular subclassification of MGMT-unmethylated high-grade glioma identifies divergent tumor states and therapeutic vulnerabilities

Abstract

MGMT-unmethylated high-grade glioma (HGG) is associated with poor response to temozolomide and limited therapeutic options. However, substantial molecular heterogeneity exists within this clinically defined subgroup and remains insufficiently characterized. Transcriptomic and clinical data from the CGGA and TCGA were analyzed. Unsupervised consensus clustering based on highly variable genes was used to identify molecular subclasses within MGMT-unmethylated HGG. The subclasses were characterized according to survival, genomic alterations, biological processes, immune landscape, immunotherapy response, and predicted drug sensitivity. Two reproducible molecular subclasses were identified across the CGGA and TCGA cohorts. Group 1 exhibited a neuron-like transcriptional profile, whereas Group 2 was characterized by enrichment of inflammatory, extracellular matrix, and stress-response pathways. Group 2 was associated with shorter overall survival in univariable analyses; however, subtype assignment was not independently associated with survival after adjustment for established clinical and molecular factors. Genomic analyses revealed shared core alterations but distinct patterns of mutation enrichment between the two subclasses. Group 2 also showed increased myeloid infiltration and higher expression of immune checkpoint genes, consistent with an immunosuppressive immune microenvironment. Despite these differences, no consistent subtype-specific differences in immunotherapy response were observed. Pharmacogenomic analysis identified histone deacetylase inhibitors as shared candidate agents, while Group 2 showed a broader spectrum of predicted vulnerabilities involving cell-cycle regulation, DNA damage, and signaling pathways. MGMT-unmethylated HGG comprises two biologically distinct molecular subclasses with divergent transcriptional, genomic, immune, and therapeutic features. Although subtype assignment was not independently associated with survival after adjustment for established prognostic factors, the proposed classification provides additional biological resolution within MGMT-unmethylated HGG and may serve as a framework for investigating subtype-associated therapeutic vulnerabilities.

The authors' abstract, as published at the source. Discover Oncology, 2026 · DOI ↗