Genomic landscape and evolutionary dynamics of <i>RAS</i> alterations in glioma: A large-scale clinicopathologic analysis.

Abstract

e14077 Background: Gliomas are molecularly heterogeneous. RAS alterations ( KRAS , NRAS , HRAS ) are discussed in the 2021 WHO under a number of glioma subtypes, but their prevalence and clinical significance remains unclear. Preclinical data suggest RAS/MAPK pathway activation is a potentially actionable vulnerability. We utilized a large, clinically sequenced dataset to characterize the incidence, hotspot distribution, and longitudinal evolution of RAS -altered gliomas. Methods: We retrospectively analyzed a cohort of 1,804 patients (pts) with glioma (2,100 samples) who underwent MSK-IMPACT targeted sequencing at Memorial Sloan Kettering Cancer Center. The prevalence of KRAS/NRAS/HRAS alterations was assessed across the full cohort. Detailed genomic analysis was performed on 80 RAS -altered pts (105 samples: 99 tumors, 6 CSF), including mutations, amplifications, fusions, and co-occurring alterations at diagnosis and recurrence. Results: RAS alterations were identified in 80/1,804pts (4.4%), including KRAS (3.4%), NRAS (0.7%), and HRAS (0.3%). Histologies included glioblastoma (37.5%), oligodendroglioma (27.5%), astrocytoma (18.8%), pilocytic astrocytoma (5.0%), and other tumor types (11.2%). Among KRAS -mutant tumors, the most frequent hotspots involved codon G12 (G12D/R/A/V/C; n=20), followed by Q61H/E/L (n=7) and Q61K (n=4). Common co-alterations included TERT (54%), IDH1 (43%), CDKN2A (30%), and TP53 (8%). Longitudinal sequencing (n=19 pts) revealed KRAS alterations persistence in 32% (n=6). Notably, acquired RAS alterations at recurrence were identified in 37% (n=7/19), involving KRAS (n=3), NRAS (n=2), and multi-gene RAS alterations (n=2): KRAS/HRAS and KRAS/NRAS . Two pts with oligodendrogliomas who acquired NRAS mutations demonstrated rapid clinical progression within 12 months. Conclusions: In this large institutional cohort, RAS family alterations were observed across diverse glioma histologies. The acquired RAS mutations at recurrence may suggests these alterations can drive tumor evolution and treatment resistance. These findings underscore the importance of longitudinal genomic profiling and support the evaluation of RAS/MAPK -targeted therapies in selected glioma pts.