Associate Professor University of Southern California, United States
Introduction::
Agent-based models (ABMs) of cellular systems have been used to explore various facets of physiology and disease. Specifically in the field of oncology, ABMs have been used to explore the many different ways in which the tumor is influenced by the microenvironment, such as through hypoxia, angiogenesis, invasion, and interactions with the immune system. It is known that cell-cell interactions strongly influence tumor growth; however, the impact of these interactions is difficult to study using purely experimental tools. Here, we investigate how interactions between distinct populations of cells in the tumor microenvironment influence overall evolution of breast tumors using an experimentally-calibrated ABM.
Materials and Methods:: We developed an ABM to represent the tumor microenvironment, including tumor cells, T cells, and macrophages, building on our previous modeling efforts [1,2]. We first calibrated the model to multiplexed immunofluorescence data from mouse tumors generated using the 4T1 breast cancer cell line. Here, we used a novel approach that applies neural networks to represent tumor images and ABM simulations as two-dimensional points, with the distance between points acting as a quantitative measure of difference between the two (Fig. 1) [3]. This enables a quantitative comparison of tumor images and ABM simulations, where the distance between simulated and experimental images can be minimized using standard parameter-fitting algorithms.
Results, Conclusions, and Discussions:: We used the calibrated model to simulate evolution of cells in the tumor microenvironment. Model calibration revealed distinct parameter regimes for T cell differentiation leading to vastly different tumor trajectories and end states. We systematically investigate the effects of specific cell-cell interactions to determine their role in influencing tumor growth by exploring the range of fitted parameter values. We find that macrophage-mediated T cell differentiation strongly contributes to the predicted tumor dynamics. Such macrophage-T cell interactions are critical in determining the response to treatment and development of drug resistance. Thus, our model provides a framework to predict strategies to control tumor growth.
