EXTH-11. PEDIATRIC GLIOMA IMMUNE PROFILING IDENTIFIES TIM3 AS A THERAPEUTIC TARGET IN BRAF-FUSION PILOCYTIC ASTROCYTOMA

Abstract Despite being the leading cause of childhood mortality, pediatric gliomas have been relatively understudied. Repurposing of adult brain tumor immunotherapies in pediatric patients has not been successful, highlighting the need for pediatric-specific immunotherapies. Pilocytic astrocytomas (PA) are the most common pediatric glioma. While surgical resection of PA is the first-line treatment, the ten-year progression-free survival rate for patients with residual tumors is <50%. Notably, some PAs undergo spontaneous regression after partial resection, suggesting potential baseline tumor immunoreactivity. Whole transcriptome sequencing performed on a commercial platform of pediatric gliomas (n=250) showed that MAPK-driven gliomas have higher interferon signatures compared to other molecular subgroups. Single-cell sequencing identified a unique activated and cytotoxic microglia population designated MG-Act in BRAF-fused MAPK-activated PA (n=13; p < 0.05), but not in pediatric high-grade gliomas (n=5) or normal brain (n=3). TIM3 was expressed on the luminal side of endothelial cells and MG-Act, but not in the surrounding normal brain based on sequential multiplex imaging. Flow cytometry showed that TIM3 intensity is upregulated on immune cells in the PA tumor microenvironment (TME) relative to matched peripheral blood immune cells (n=6; p < 0.01). Anti-TIM3 reprogrammed fresh ex vivo immune cells from the TME of human PAs (n=3) to a pro-inflammatory cytotoxic phenotype. In a genetically engineered murine model of MAPK-driven low-grade gliomas, median survival (MS) was 252 days in mice treated with anti-TIM3 (n=21; p<0.01) relative to IgG control (MS: 110 days; n=20) or anti-PD-1 (MS: 134 days; n=20). The therapeutic activity of anti-TIM3 was abrogated in the CX3CR1 microglia knockout background. ScRNA sequencing data during the therapeutic window of anti-TIM3 in wild-type mice demonstrated enrichment of the MG-Act population within low-grade gliomas at two different longitudinal time points but not in IgG-treated controls. Together, these data indicate that anti-TIM3 should be considered for clinical trials in pediatric low-grade MAPK-driven gliomas.