UC Riverside

Chimeric antigen receptor (CAR) T-cell based immunotherapy has shown its potential in treating blood cancers, and its application to solid tumors is currently being extensively investigated. For glioma brain tumors, various CAR T-cell targets include IL13R$\alpha$2, EGFRvIII, HER2, EphA2, GD2, B7-H3, and chlorotoxin. In this work, we built a multiscale model to help us understand the treatment at different levels and provide insights into the dynamics of interaction between CAR T-cells and glioma cells. In our first model, we focus on extending the work of Kuznetsov et al. (1994) by considering the binding of multiple CAR T-cells to a single glioma cell, and the dynamics of these multi-cellular conjugates. Our model more accurately describes experimentally observed CAR T-cell killing assay data than the models that do not consider multi-cellular conjugates. Moreover, we derive conditions in the CAR T-cell expansion rate that determine treatment success or failure. Finally, we show that our model captures distinct CAR T-cell killing dynamics from low to high antigen receptor densities in patient-derived brain tumor cells. In our second model, we mainly investigate the antigen heterogeneity of high-grade glioma and explore the treatment result in 3D by an agent-based and PDE hybrid model. We apply the model to our experimental data to study treatment schedules, dosage, and injection location, which helps us to optimize the therapy.