Optimizing Clinical Study Designs for CAR-T Cell Therapy: Development of an Efficient Sampling Strategy Through Optimal Experimental Design.

CAR-T cell therapy is a cellular cancer immunotherapy that has impressively improved outcomes in hematological malignancies compared to conventional treatments. Yet, many patients do not respond permanently. Nonlinear mixed-effects modeling can support a better understanding of the unique and not fully understood dose-exposure and exposure-response relationships for CAR-T cell therapy by integrating available knowledge into a quantitative, physiology-motivated framework. However, to unfold its full potential, it requires informative and efficient study designs for clinical data collection. We aimed to develop an optimal experimental design framework informing a robust and feasible clinical CAR-T cell study design based on a published mechanistic model of CAR-T cell kinetics and tumor dynamics. By considering variability between study populations and parameter uncertainty, we (1) identified the minimal population size required to inform model parameters, (2) assessed different sampling strategies, and (3) informed flexible sampling windows instead of fixed sampling timepoints. The optimized study design consisted of 60 patients, three fixed assessments of tumor burden (days 0, 30 and 90), and three feasible CAR-T cell sampling windows (days 2-4, 12-18 and 32-47 after infusion). In stochastic simulation and estimation, the optimized design with sampling windows showed better performance in informing model parameters, including those characterizing heterogeneous outcomes, than designs with fixed sampling timepoints, confirming its efficiency and robustness. This framework shall facilitate future feasible and resource-efficient CAR-T cell clinical data collection, thus showcasing the potential of optimal experimental design to advance the development and optimization of complex cancer immunotherapies in clinical trials and clinical practice.