CNSC-15. TARGETING GLIOMA-NEURONAL CIRCUIT REMODELING TO OVERCOME IMMUNOSUPPRESSION: FUNCTIONAL AND THERAPEUTIC INSIGHTS

Abstract Neuronal activity profoundly influences glioblastoma (GBM) cell proliferation and invasion, while GBM actively remodels the surrounding neuronal circuitry. We recently demonstrated that intratumoral regions with heightened neuronal connectivity exhibit prominent regional immunosuppression. Our preclinical investigations revealed that tumor-secreted Thrombospondin-1 (TSP1) enhances glioma-neuronal circuit remodeling, contributing to an immunosuppressive tumor microenvironment (TME) characterized by increased anti-inflammatory tumor-associated macrophages (TAMs) and decreased CD8+ T cells. Notably, this immunosuppressive TME was alleviated either by eliminating tumor-derived TSP1 or by treatment with perampanel (PER), which inhibits glutamatergic excitatory signaling. However, the therapeutic implications of these findings remain to be fully elucidated. In this study, we conducted functional validation by isolating TAMs from mice bearing TSP1-WT or KO tumors and co-culturing them with carboxyfluorescein succinimidyl ester (CFSE)-labeled T cells. Strikingly, the suppressive effect of TAMs on T cell proliferation was significantly alleviated in the TSP1-KO group (P = 0.03). Moreover, a similar trend was observed in TAMs from PER-treated tumor-bearing mice (P = 0.06). Building on this functional evidence, we explored combination therapeutic strategies. Specifically, we tested the combination of anti-EGFRvIII CAR-T cells and anti-PD1 antibody, with or without PER, in mice bearing EGFRvIII-expressing SB28 tumors. Complete tumor eradication occurred more frequently in the PER-treated group compared to controls (44% vs. 22%). Furthermore, in brain-infiltrating leukocytes isolated from five “cured” mice, persisting CAR-T cells were detected in all three PER-treated mice but only one of the two controls. Notably, a CD62L+CD44+ memory phenotype, associated with long-term tumor remission after treatment, was observed exclusively in the PER-treated group. In summary, our findings highlight the intricate crosstalk between glioma, neurons, and immune cells as a pivotal therapeutic vulnerability. Targeting neuronal activity may represent a promising strategy to modulate the immunosuppressive TME in GBM and enhance the efficacy of immunotherapeutic modalities, such as CAR-T cell therapy.