Mathematical Biology Seminar Jiyeon Park, U of U Mathematics, Wednesday, Sept. 7, 2022 3:05pm in LCB 225 STOCHASTIC ADAPTIVE CHEMOTHERAPY CONTROL OF COMPETITIVE RELEASE IN TUMORS Abstract: Adaptive chemotherapy seeks to manage chemoresistance by delaying the competitive release of a resistant sub-population, and to manage cancer by maintaining a tolerable tumor size rather than seeking a cure. Models typically follow interactions between infinite populations of sensitive (S) and resistant (R) cell types to derive a chemotherapy dosing strategy C(t) that maintains the balance of the competing sub-populations. Our models generalize to include healthy (H) cells, and finite population sizes. With finite population size, stochastic fluctuations lead to escape of resistant cell populations that are predicted to be controlled in the deterministic case. We test adaptive schedules from the deterministic models on a ?nite-cell (N = 10,000 50,000) stochastic frequency-dependent Moran process model. We quantify the stochastic fluctuations and variance (using principal component coordinates) associated with the evolutionary cycle for multiple rounds of adaptive chemotherapy, and show that the accumulated stochastic error over multiple rounds follows power-law scaling. This accumulates variability and can lead to stochastic escape which occurs more quickly with a smaller total number of cells. Moreover, we compare these adaptive schedules to standard approaches, such as low-dose metronomic (LDM) and maximum tolerated dose (MTD) schedules, finding that adaptive therapy provides more durable control than MTD even when we include the effects of ?nite population size. Although low-dimensional, this simplified model elucidates how well applying adaptive chemotherapy schedules for multiple rounds performs in a stochastic environment. Increasing stochastic error over rounds can erode the effectiveness of adaptive therapy.