Agent-Based Models of the Tumor Microenvironment: PredictingFlow-Mediated Invasion and Cancer Viability in Response toInterstitial Flow, Chemotherapy, and Stromal Cell Density

Therapeutic delivery within the tumor is attenuated by static pressure and the permeability of the tissue, which is mediated by the extent of ECM remodeling. Additionally, cellular signaling from stromal fibroblasts limits the efficacy of treatment. This phenomenon is of particular relevance at the tumor border, where there is a transitional region from predominantly cancer cells to predominantly stromal cells. These factors promote cancer growth and select for a potentially metastatic subpopulation. Furthermore, interstitial fluid flow at the tumor border affects the migration characteristics of cancer cells. Interstitial fluid flow has been implicated in an increase in cancer cell invasion, which is the major effector of cancer spread and decreased patient viability. In brain cancer, two different mechanisms have been implicated in this increased invasion, and this invasion is mediated by both CXCR4 and CD44. Therefore, TME complexity necessitates utilization of robust models to elucidate the effects of the TME on cancer development and progression. The current work defines two novel agent-based models that describe and predict cancer specific outcomes within in vitro TME mimetic systems. Our models indicate that brain cancer cell invasion is increased in the presence of interstitial flow, and this increased invasion is driven by two specific mechanisms, CXCR4-CXCL12 autologous chemotaxis and CD44 mechanotransduction. Additionally, in vitro and in silico models of the tumor border transition zone predict that regional viability within the breast tumor border is affected by both fibroblast signaling and transport mechanisms, and this affect could be selecting for important resistant subpopulations.