CAR T cells from code to clinic: Framing modeling approaches with current translational research goals

Chimeric antigen receptor (CAR) T cell therapy has transformed treatment of resistant cancers, yet it faces major limitations such as variable patient responses, relapse, and safety issues. Optimizing CAR T cell treatment regimens and genetic design could offer solutions to some of these problems, but the resulting design space is often too large to explore experimentally alone. Mathematical modeling-spanning ordinary differential equation (ODE), agent-based, statistical, and machine learning methods-provides a framework to link mechanistic understanding of CAR T cells to predictive design and clinical decision-making. Here, we present a perspective that organizes modeling approaches around key translational challenges in CAR T therapy, evaluating the strengths, limitations, and data gaps of current approaches. We outline a roadmap for clinicians and modelers alike, highlighting advances in machine learning, multiscale modeling, control theory, and alternative cell types as emerging opportunities in model-guided design of next-generation cell therapies.