胶质母细胞瘤中溶瘤单纯疱疹病毒治疗后肿瘤免疫微环境的空间分析

Glioblastoma (GBM) is the most common and aggressive malignant brain tumor in adults, with an extremely low 5-year survival rate of ~5%. Despite established surgery and chemoradiotherapy treatments, outcomes remain poor due to GBM's profound heterogeneity and highly immunosuppressive tumor immune microenvironment (TIME). Emerging immunotherapies offer new opportunities to overcome these barriers. Among them, oncolytic herpes simplex viruses (oHSVs) stand out for their dual ability to selectively lyse tumor cells while stimulating robust anti-tumor immune responses. Preclinical and early clinical studies demonstrate that oHSVs can profoundly remodel immune cell expression within tumors, but these analyses lack spatial insight. Spatial multi-omics technologies now enable precise mapping of such immune transcriptome and proteome changes, answering how immune reorganization shapes the GBM TIME. The present study utilized 10x Genomics Visium spatial multi-omics to comprehensively map oHSV-induced immune responses in human GBM and explore their association with patient outcomes. Brain tumor samples were obtained from a Phase IA dose-escalation oHSV trial for recurrent glioma, with tissue collected both before treatment and at post-treatment resection. Across four patients, a total of seventeen formalin-fixed, paraffin-embedded samples were profiled using the Visium Spatial Gene Expression (Whole Transcriptome) and Visium Immune Profiling (Protein) panels. Spatial multi-omic data were analyzed in Seurat, leveraging a human brain immune reference integrated with our oHSV single-cell RNA-sequencing data to enhance cell type identification for spatial deconvolution. We characterized key immune cell populations, evaluating how their localization, abundance, and expression varied across treatment status (pre- vs. post-oHSV), spatial context (tumor bed vs. periphery), and clinical response (responders vs. non-responders). We also explored the use of custom viral RNA probes to detect viral presence within tissue; while signal detection was limited, these efforts informed future optimization of viral localization methods. By comparing spatial immune architectures between responders and non-responders, we identified trends that illuminate how immune cell organization shapes therapeutic efficacy. Furthermore, the observed concordance between spatial gene and protein expression of key immune markers reinforces the biological validity of these signatures and suggests their possible association with treatment response. Ultimately, elucidating the mechanisms by which oHSV modulates the GBM TIME advances our understanding of its anti-tumor effect and provides a rationale for its continued clinical development.