TMIC-76. Deciphering aging-associated drivers of glioblastoma biology at the tumor core and border

Abstract Glioblastoma (GB), IDH-wildtype, exhibits increased incidence and poorer prognosis in elderly patients, yet the biological underpinnings of this age-related disparity remain poorly understood. Since previous studies have suggested only a limited increase in potential GB-driving genetic alterations in the human brain with age, we hypothesize that non-genetic factors may present aging-dependent contributors to GB pathobiology. In particular, age-associated microenvironmental drivers have been revealed in other cancer types, but their relevance in GB remains underexplored. Moreover, recent GB mouse model studies have suggested that poorer survival in older animals is driven by factors distant from the tumor bulk (“extratumoral”), but this has not been examined in humans due to limited analyses of matched non-tumor brain tissue. To address this gap, we performed single-cell RNA sequencing (scRNA-seq) on formalin-fixed paraffin-embedded tumor samples from 36 GB patients, comprising 18 elderly (65–88, median 75 years) and 18 younger (35–55, median 47 years) individuals. Cohorts were balanced by sex and by the three major epigenetic GB subtypes RTK I, RTK II, mesenchymal, as defined by DNA methylation profiling. Whole-exome sequencing was performed to control for mutational events. For half of the patients, tumor-adjacent normal-appearing cortex was histologically selected, and additionally profiled by scRNA-seq to assess potential age-dependent extratumoral changes. Our spatially stratified approach enabled the detection of diverse malignant cell populations within the tumor core, as well as enrichment of non-malignant cell types often underrepresented in tumor-core analyses, including astrocytes, neurons, and OPCs. Preliminary analyses indicate a region-specific increase in potentially tumor-promoting cellular signatures across immune and glial compartments in older patients, such as more reactive myeloid signatures at the tumor border. Ongoing analyses include integration of these findings with published normal brain datasets and correlation with clinical outcomes. Our study thereby provides a comprehensive framework for understanding aging-associated drivers in human GB.