Optimizing Adaptive Hormone Control for Personalized Prostate Cancer Treatment

With the oncologist acting as the "game leader", we employ a Stackelberg game-theoretic model involving multiple populations to study prostate cancer. We refine the drug dosing schedule using an empirical Bayes feed-forward analysis, based on clinical data that reflects each patient's prostate-specific drug response. Our approach quantitatively explores the parameter landscape of this adaptive multipopulation model, aiming to combat drug-resistant prostate cancer by fostering competition among drug-sensitive cell populations. Our results suggest that not only is it is feasible to considerably extend cancer suppression duration through careful optimization, but even transform metastatic prostate cancer into a chronic condition instead of an acute one for most patients, with supporting clinical and analytical evidence.