Abstract 809: Spatially resolved multiomics profiling of glioblastoma reveals molecular signatures of neuropathology and immuno-oncology architecture using CosMx SMI.

Abstract Glioblastoma (GBM) is a highly heterogeneous and aggressive brain tumor with limited therapeutic success, underscoring the need for advanced approaches to decode its complex molecular landscape. Here, we employed the CosMx® Spatial Molecular Imager to perform same-cell spatial multiomics profiling on serial sections from a human GBM tissue block, integrating the whole-transcriptome RNA panel (∼19,000 genes) with two high-plex protein panels encompassing 64 neuropathology or immuno-oncology markers. This comprehensive dataset enables correlative mapping of transcriptomic and proteomic features across tumor and peri-tumoral regions. We identified five GBM tumor cell subtypes - astrocyte-like, OPC-like, mesenchymal-like, stem-like, and proliferating cells - each exhibiting distinct spatial localization and transcriptional signatures. Concurrent protein profiling uncovered regionally enriched phospho-Tau variants (e.g., p-Tau S214, S396, S404), whose accumulation marked hypoxic and epithelial-mesenchymal transition (EMT) active tumor microenvironments. Spatial regression analyses linked elevated neighborhood p-Tau scores to coordinated upregulation of hypoxia-responsive genes (e.g., HIF1A, VEGFA) and metabolic stress pathways, revealing a localized molecular response to cellular stress. Integration of the immuno-oncology protein panel highlighted spatially restricted immune checkpoint expression, including PD-L1, B7-H3, and TIM-3, delineating immune-evasive niches within distinct tumor neighborhoods. Tumor niches driven by divergent subtype GBM cells adopted unique immune evasion strategies, suggesting microenvironment-specific mechanisms of immune suppression. Overall, this study demonstrates the power of spatially resolved, same-cell multiomics to dissect GBM’s transcriptional and proteomic heterogeneity at unprecedented resolution. The findings establish a mechanistic link between phospho-Tau-associated stress adaptation, metabolic remodeling, and immune modulation in GBM. This scalable framework offers a blueprint for future multi-sample spatial oncology studies aimed at uncovering therapeutic vulnerabilities, stratifying tumors by microenvironmental features, and identifying biomarkers predictive of treatment response. Citation Format: Yi Cui, Claire Williams, Chia-Ying Lee, Chi Phan, Sierra Mckinzie, Shanshan He, Ashley Heck, Kimberley Young, Lidan Wu, John Lyssand, PRAJAN DIVAKAR, Joseph M. Beechem. Spatially resolved multiomics profiling of glioblastoma reveals molecular signatures of neuropathology and immuno-oncology architecture using CosMx SMI abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 809.