CNSC-98. SYNAPTIC REMODELING DRIVES IDH-MUTANT ASTROCYTOMA PROGRESSION AND EVOLUTION

Abstract IDH-mutant astrocytomas often progress from low-grade to higher-grade tumors, through incompletely defined mechanisms. The neuronal microenvironments through direct and paracrine-mediated signaling drive glioblastoma progression; therefore, we hypothesized its role in IDH-mutant astrocytoma evolution. In this study, we integrate sixteen patient-matched newly diagnosed and recurrent astrocytomas (CNS WHO grades 2–4), which were interrogated using spatial RNA sequencing, multiplexed immunofluorescence, and three-dimensional co-cultures comprising mouse cortical neurospheres (cNS) and patient-derived astrocytoma organoids (GliA; n = 17). Harmony-integrated UMAP of 10,189 spatial transcriptomes identified ten gene-expression programs across 8 IDH-mutant astrocytomas. CNS WHO grade 2 and grade 3 tumors shared overlapping transcriptomic profiles, whereas grade 4 tumors formed a distinct cluster. A 28-gene IDH-mutant glioma neural connectivity score, characterized by TSP1, NDRG2, and Cx43, correlated with overall survival. Immunofluorescent staining confirmed overexpression of TSP1, NDRG2, and Cx43 in cortex-infiltrative and dense tumor regions. Functional interrogation using novel in vitro co-culture model systems revealed a grade-specific increase in Ki67⁺ proliferating tumor cells throughout the tumor, not restricted to areas of glioma-neuron integration. This was quantified across 7 patient-matched GliA–cNS co-cultures (n = 172 regions) and GliA monocultures (n = 193 regions). Proliferation in GliA co-cultures decayed radially with increasing distance from the neuron–tumor interface, demonstrating that astrocytomas are most proliferative in close proximity to neurons. Furthermore, GliA-cNS co-culture demonstrated no increase in neuronal firing rates and synchrony, as quantified across 22 GliA–cNS co-cultures from 5 patients and 20 cNS monocultures. IDH-mutant astrocytomas display uniform, grade-dependent synaptic remodeling, revealing potential therapeutic targets and treatments for patients.