P18.15.a Immunocompetent Glioblastoma Organoids for Personalized Immune Profiling

BACKGROUND Glioblastoma (GB) is a highly aggressive brain tumor with poor prognosis and limited response to immunotherapies. Tumor intrinsic features like low mutational burden, and extrinsic factors such as T-cell dysfunction and an immunosuppressive microenvironment, contribute to resistance. To better study immune-tumor interactions in a clinically relevant context, we have developed patient-derived tumor organoids (PDTOs) that support co-culture with autologous and CAR-engineered immune cells, enabling real-time functional analysis. MATERIAL AND METHODS PDTOs were generated from freshly resected GB tissue, preserving tumor-specific architecture and molecular features. Patient-matched PBMCs were isolated via density gradient centrifugation, and immune cells were enriched using negative selection with MACS sorting. T cells were expanded ex vivo and characterized by flow cytometry for activation and exhaustion markers, although not yet tested in co-culture. Cytokine secretion was analyzed via ELISA to complement phenotypic profiling. For functional studies, EGFR/EGFRvIII-retargeted CAR-NK cells were generated via lentiviral transduction of primary PBMC-derived NK cells and expanded ex vivo in the presence of (low levels of) IL-15. These cells were co-cultured with 3D tumor spheroids at defined effector-to-target ratios. Immune infiltration and tumor cell killing were assessed using real-time live-cell imaging and fluorescence-based readouts. RESULTS As proof of concept, we co-cultured EGFR-diploid and EGFR-amplified 3D GB spheroids with naïve and CAR-NK cells targeting EGFR derived from a healthy donor. CAR-NK cells showed enhanced cytotoxicity against EGFR-amplified spheroids, demonstrating receptor-dependent immune activation. These results confirm the feasibility of functional immune co-culture assays and support the ongoing integration of this approach into our organoid-based platforms. CONCLUSION Our organoid-immune cell co-culture system offers a physiologically relevant platform to investigate GB immunobiology. Future applications include drug screening and CRISPR-based perturbation to identify key modulators of immune response, with the goal of informing personalized immunotherapeutic strategies. Observations from the CAR2BRAIN trial (NCT03383978) have shown CD4⁺ and CD8⁺ T cell recruitment following CAR-NK therapy, highlighting the clinical relevance of combinatorial immune approaches and supporting the use of our model to explore such mechanisms ex vivo.