When Random Beats Routine: Stochastic Hormone-Sensitive Cancer Deprivation Therapy

Deprivation therapies for hormone-sensitive cancers, such as prostate, breast, and ovarian malignancies, embody an evolutionary game in which therapy suppresses drug-sensitive cells but leaves resistant subclones unchecked. The ultimate threat to the patient is not the size of the tumor per se but the metabolic load imposed by the tumor ecosystem on the body. In contrast to intuition, we show that pseudo-randomized therapy schedules, a feed-forward strategy without cell number monitoring, can paradoxically prolong survival by sustaining sensitive cell populations and minimizing metabolic load, an example of Parrondo's paradox. By contrast, systematic therapy schedules drive sensitive cells to extinction and accelerate resistant takeover. We propose that therapy optimization should be cast as a Pareto problem: maximizing time to metabolic overload while minimizing drug burden. This metabolic perspective reframes adaptive therapy as an eco-evolutionary control problem, highlighting how noise, randomness, and evolutionary games intersect in cancer treatment.