Modeling Cancer Evolution and Identifying Optimal Drug Dosing Schedules

Abstract

Despite the advancement of targeted therapies to treat cancer, the emergence of resistance to treatment is typical. Response to therapy and the emergence of resistance can be modeled quantitatively as a time-heterogeneous stochastic process. This mathematical model fused with pharmacokinetic models can be used to identify optimal treatment administration schedules in combination therapies. Here, we present an integrated computational modeling and experimental approach to identify an optimal dosing schedule in three different applications. First, we developed a predictive model that encompasses tumor heterogeneity and inter-subject pharmacokinetic variability to predict tumor evolution under different dosing schedules of two EGFR tyrosine kinase inhibitors in the treatment of advanced EGFR-mutant non-small cell lung cancer (EGFR-m NSCLC). Second, we expanded our approach by incorporating a three-drug combination of EGFR inhibitors and a MET inhibitor and including drug-tolerant persister cells in our evolution model of EGFR-m NSCLC. Lastly, we identified optimal drug concentrations and doses which maximized differentiation or minimized tumor volume in irradiated mice for the treatment of acute myeloid leukemia. Our proposed dosing schedule of two EGFR inhibitors in the treatment of EGFR-m NSCLC was subsequently confirmed tolerable in an ongoing phase I clinical trial, demonstrating that our rational modeling approach can be used to identify appropriate dosing for combination therapy in the clinical setting.