Abstract
Glioblastoma (GBM) remains one of the most fatal brain tumors, in part because its profound spatial heterogeneity undermines the effectiveness of targeted therapies. Despite extensive pursuit, EGFR and VEGF inhibitor monotherapies have shown limited clinical benefit, yielding only brief remissions. We hypothesized that this therapeutic failure stems from region-specific variations in drug target expression within the tumor microenvironment, limiting the reach of monotherapies. To investigate this, we performed spatial transcriptomic profiling on 17 regions from 8 human GBM specimens, mapping gene expression to distinct anatomical compartments defined by the Ivy Glioblastoma Atlas Project: Core Tumor (CT), Core Tumor with Pseudopalisading Necrosis (CTpan), Core Tumor with Microvascular Proliferation (CTmvp), and Leading Edge (LE). Our analysis revealed striking spatial compartmentalization of therapeutic targets: EGFR expression was largely confined to the proliferative CT; VEGF ligands (VEGFA, VEGFB) were enriched within the hypoxic, necrotic CTpan; and VEGF receptors (FLT1, FLT4, KDR) were mostly localized to CTmvp, consistent with their roles in angiogenesis. Notably, none of these targets were expressed at the infiltrative Leading Edge, a compartment strongly implicated in tumor recurrence. These findings provide mechanistic insight into the limited efficacy of EGFR and VEGF inhibitor monotherapies in GBM: the spatial segregation of targets prevents uniform drug action across all tumor regions. By leveraging spatial transcriptomics, our study highlights the inadequacy of one-size-fits-all regimens and underscores the need for spatially informed, multi-agent strategies tailored to the tumor’s regional biology. These results support a shift toward treatment approaches that better align with the evolving landscape of tumor heterogeneity in glioblastoma.