Spatial Transcriptomics Reveals Immune Remodeling in Recurrent Oligodendroglioma 2259684

Abstract

Abstract Introduction Oligodendrogliomas frequently recur despite standard therapy, but how the immune microenvironment evolves during recurrence remains poorly understood. Methods Visium Spatial Transcriptomics data from primary and matched recurrent oligodendrogliomas from 14 patients were processed using SpaceRanger and deconvoluted with SpaCET’s low-grade glioma model to estimate cell-type proportions. Cell-type fractions were normalized using Centered Log-Ratio transformation and clustered (k = 5, Euclidean distance) to define reproducible spatial niches. Immune subset distributions were compared between primary and recurrent tumors to assess niche-specific remodeling, with propensity score matching applied to control for baseline clinical and molecular differences between treated and untreated patients. Results Five spatial niches were identified, including immune-enriched perivascular and infiltrative regions. Recurrent tumors showed marked loss of adaptive immunity, with downregulation of CD4+ T-helper subsets (Th1, Th2, Th17, Tfh; log₂FC ≈ —3) and total CD4+ T cells (—2.9). This loss, most pronounced after radiation and/or chemotherapy, indicates therapy-associated collapse of adaptive immune niches. Macrophage and microglial populations were preserved or increased, reflecting a shift toward a myeloid-dominant, immunosuppressive microenvironment. Immune-rich perivascular regions exhibited depletion of T-helper diversity and enhanced myeloid signaling, consistent with immune exclusion and vascular remodeling. These findings persisted after propensity score matching, with treated tumors remaining enriched for immunosuppressive niches (p < 0.05). Validation by Xenium spatial transcriptomics and CODEX multiplex imaging will be included in the final presentation. Conclusion Recurrent oligodendrogliomas exhibit collapse of adaptive immune niches and expansion of myeloid-dominant regions, defining immune remodeling as a hallmark of recurrence and myeloid-driven immunosuppression as a potential therapeutic target. Funding Source n/a Topic Categories Computational and Systems Immunology (COMP)