Abstract
BACKGROUND: Glioblastoma (GBM) remains largely incurable, in part because the highly immunosuppressive tumor microenvironment (TME) poses a major barrier to conventional therapies, including to chimeric antigen receptor (CAR)-T cell therapy. Compared with conventional αβ T cells, γδ T cells bridge innate and adaptive immunity, amplifying anti-tumor responses through cytokine secretion and cross-talk with other immune populations, so arming γδ T cells with a B7–H3-targeting CAR could combine their inherent tumor-homing and immunomodulatory capabilities with precise antigen-specific cytotoxicity for therapeutic synergy. METHODS: CD276 (B7–H3) expression and prognostic relevance were analyzed using bulk and single-cell RNA-seq data from TCGA and CGGA cohorts. Human γδ T cells and conventional αβ T cells were expanded from healthy donor PBMCs and engineered with a second-generation B7–H3 CAR. CAR expression and phenotype were assessed by flow cytometry. Antitumor activity against B7–H3(+) glioma cell lines (U87, U251) was evaluated using cytotoxicity and cytokine-release assays. Therapeutic efficacy was tested in NSG mice bearing glioma xenografts, with tumor growth, survival, tumor infiltration, apoptosis, and checkpoint expression assessed by immunofluorescence. RESULTS: B7–H3 CAR-γδ T cells exhibited superior antitumor functionality compared with B7–H3 CAR-αβ T cells in GBM models. In vitro, B7–H3 CAR-γδ T cells displayed enhanced, sustained cytotoxicity against GBM cell lines and secreted significantly higher levels of key effector cytokines (IFN-γ, TNF-α) than B7–H3 CAR-αβ T cells, indicating a polyfunctional and exhaustion-resistant phenotype. In GBM mouse models, a single dose of B7–H3 CAR-γδ T cells mediated robust tumor control and significantly prolonged survival. Mechanistic studies traced this superior efficacy to enhanced tumor infiltration and more potent induction of tumor cell apoptosis. CONCLUSIONS: B7–H3 CAR-γδ T cells therefore represent a promising new approach to GBM immunotherapy which, through ongoing platform optimization and clinical exploration, could offer new therapeutic hope to patients with glioma. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12967-026-07994-6.