Speaker
Description
Cancer-associated fibroblasts (CAFs) are key components of the tumor microenvironment (TME) and exhibit highly heterogeneous phenotypes that can either promote or suppress tumor growth. Our previous ODE-based mathematical model quantitatively described the interactions among cancer cells, T cells, and CAFs, demonstrating that CAF composition critically determines treatment outcomes. The model predicted that certain CAF configurations could render monotherapies as effective as combination treatments, highlighting CAF composition as a potential biomarker for therapy selection.
To address the limitation of non-spatial mean-field modeling, we developed a hybrid agent-based model (ABM–PDE–ODE) that explicitly represents individual cells and diffusible factors within the TME. In this model, cancer cells, T cells, and CAFs behave as discrete agents interacting through spatial gradients of oxygen and cytokines described by PDEs. Each T cell independently solves ODE-based binding kinetics for PD-1/PD-L1 interactions, determining its activation or exhaustion state.
This hybrid framework enables spatiotemporal simulation of CAF-driven immunosuppressive niches and immune infiltration dynamics, capturing emergent behaviors that cannot be resolved by ODE models alone. Together, these models provide a multiscale platform to explore how CAF heterogeneity shapes treatment response, supporting the development of CAF-informed, spatially guided therapeutic strategies for precision oncology.
