NCOG-35. IDENTIFYING PROGNOSTIC FACTORS IN PEDIATRIC DIFFUSE MIDLINE GLIOMAS, H3 K27-ALTERED

Abstract INTRODUCTION Diffuse midline gliomas, H3 K27-altered (DMG) are rare, grade 4 tumors localized to the diencephalon, brainstem, and cervical spine. DMGs, characterized by a K27M missense mutation in histone H3, often cannot be safely resected and predominantly present with poor outcomes in children. Few prognostic factors for DMGs have been identified and/or need to be further elucidated, including adolescent age of onset, re-irradiation status, and tumor molecular profile (i.e., PIK3CA mutation). Herein, we describe the clinical outcomes of DMG patients at our institution. METHODS Pediatric patients (<18 y/o) between 2015 and 2023 with biopsy-proven DMG were reviewed. Demographic, treatment, and tumor characteristics (histologic and molecular) were recorded. Adolescent-onset DMGs were defined as patients ≥10 y/o at diagnosis and early-onset DMGs were defined as patients <10 y/o. Progression free survival (PFS) and overall survival (OS) were calculated according to the Kaplan Meier-method. Multivariate regression analyses were performed. RESULTS Twenty-one patients were included (median age=9 y/o). Early-onset DMGs (n=13) progressed significantly earlier and experienced significantly greater mortality than adolescent-onset DMGs (n=8); PFS and OS at 12 months was 7.7% and 23% for early-onset DMGs and 50% and 87.5% for adolescent-onset DMGs (PFS: p<0.01, OS: p<0.02). Patients that received re-irradiation for progressive disease (n=9, mean=84Gy) had an OS of 75% at 12 months, which was significantly greater than an OS of 23% at 12 months (p<0.05) for patients that received no re-irradiation (n=12, mean=53Gy). PIK3CA mutation status was identified as an independent predictor of survival, with PIK3CA-mutant DMGs (n=7) having a median survival of 4.1 months versus 9.7 months for PIK3CA-WT DMGs. CONCLUSION We describe a pediatric cohort that presents several possible prognostic factors of survival in DMGs. Our findings on adolescent age of onset, if further corroborated, has the potential to inform the design and enrollment of DMG clinical trials.