Download Champions' SITC 2025 Poster #1244
Modeling the Immunosuppressive Tumor Microenvironment for CAR-T Development: A Patient-Derived 3D Co-Culture Platform with CAFs and M2 Macrophages
Poster #1244, presented at SITC 2025
Solid tumors limit CAR-T performance through an immunosuppressive microenvironment, stromal barriers, and inhibitory cytokines. This study builds a translational triculture using PDX-derived organoids from our annotated biobank, co-cultured with cancer-associated fibroblasts and M2-polarized macrophages in 96-well format. We profile non-targeting, targeted, and armored CAR-T constructs, then quantify tumor killing and immune activation with a custom multiparameter flow panel. The model reproduces suppressive features seen in patients and reveals construct and combination effects that simple 2D assays miss, enabling better selection, dosing strategies, and go or no go decisions before in vivo.The triculture generates stable 3D structures and an immunosuppressive context that still permits CAR-T infiltration and function. Targeted and armored CAR-T cells increase T-cell activation and reduce viable tumor cells versus non-targeting controls, with combination regimens further enhancing responses. Flow cytometry distinguishes PDXO, CAF, M2, and T-cell compartments, enabling mechanism-focused readouts and rapid iteration on constructs and combinations.
Learning outcomes:
- Learn how a PDXO, CAF, and M2 triculture recreates a clinically reflective immunosuppressive TME in a scalable 96-well workflow for CAR-T testing.
- See how a custom flow panel separates PDXO, fibroblast, macrophage, and T-cell populations to quantify killing and activation in one assay.
- Understand that targeted and armored CAR-T cells show higher activation and stronger tumor reduction than non-targeting controls in this suppressive context.
- Review example signals from the poster, about three-fold increases in CD69+ T cells and marked drops in CD70+ tumor subsets with targeted and armored CAR-T cells.
- Identify how to extend the platform to combination testing and mechanistic studies that inform construct choice, dosing, and cohort design.
Download the Poster