Speaker
Description
Over 75% of cancer patients receive chemo- and/or radiation therapy, but treatment schedules are not optimized for the individual patient. Optimized schedules may lead to improved outcomes. We aim to (1) collect confluence time courses for glioma cells treated over a wide range of chemo- and radiation therapy schedules, (2) calibrate a biology-based math model to the data, and (3) apply optimal control theory to the calibrated model to identify and test theoretically optimal schedules. As it is highly cost-intensive to test all schedules in vivo, we are collecting data in vitro and optimizing in silico. Using C6 glioma cells, we tested 64 unique schedules for chemotherapy (temozolomide), 72 for radiation therapy, and 24 for combination chemoradiation. Using Incucyte live-cell imaging, phase contrast and confluence data were acquired every four hours. Our math model accounts for the temporal change in tumor cell confluence as a function of logistic proliferation with treatment-induced suppression of the proliferation rate. Fitting our model to the 120 unique treatment schedules complete at time of submission yielded a mean (±SD) R2 value of 0.87 (±0.12). Early analyses identified conditions in which treating with half the dose on a different schedule or the same dose with more rest yielded similar outcomes. These initial results suggest a potential to optimize dose and timing for chemo- and/or radiation therapy. Ongoing work seeks to identify and validate the optimal regimens.
