Speaker
Description
The efficacy of CAR-T cell treatment is heavily impacted by low levels of oxygen (hypoxia) in solid tumors. However, CAR-T cells with different affinities show distinct behavior and functionality in hypoxic areas. Thus, we could optimize the composition of CAR-T cells cocktails of different affinities to maximize their efficacy in the heterogeneous microenvironment of melanoma. To do so, we developed a pipeline to digitize tissue images, simulate oxygen distribution, and create a tissue hypoxia map. We then built a hybrid agent-based model that describes the dynamics of cancer cells, vessels, and CAR-T cells within the tissue, including CAR-T cell division, cytotoxicity (depending on the scFv affinity), and exhaustion. This was combined with kinetics of tissue oxygenation, chemokine secretion patterns, and their microenvironmental distribution. Using these outcomes as a computational domain for our model, we studied how different chemokine secretion patterns impact the travelled distance by CAR-T cells, their killing efficacy, and exhaustion. This suggests that the creation of chemokine gradients can improve CAR-T cell efficacy. Finally, because CARs of different affinities have different properties when exposed to lack of oxygen, we identified the optimal mixtures of CARs of specific affinities that together better target tumor cells in hypoxic areas, resulting in better anti-tumor control in melanoma. This model can generate experimentally testable hypotheses.
