Speaker
Description
Prostate cancer is commonly treated with androgen deprivation therapy, yet resistance frequently emerges during long-term treatment. Intermittent androgen deprivation therapy (IADT), in which treatment is periodically suspended and reintroduced, has been proposed to reduce treatment burden while potentially delaying resistance. Determining when resistance is likely to arise remains a central challenge for optimizing treatment schedules in individual patients, given the variability in prostate-specific antigen (PSA) dynamics across patients.
Here we develop a mathematical model describing the interaction between treatment-sensitive and resistant tumor cell populations and their relationship to PSA dynamics during IADT. To enable patient-specific inference from clinical data, we incorporate Bayesian inference to estimate PSA dynamics from longitudinal measurements and update tumor composition estimates.
Using longitudinal PSA measurements from patients undergoing IADT \cite{bruchovsky2006final}, we examine how early PSA trajectories reflect tumor dynamics and assess the probability of resistance during subsequent treatment cycles. Our results suggest that patient-specific PSA patterns contain signals about tumor composition and the likelihood of resistance under treatment cycling. The proposed framework provides a quantitative approach for interpreting PSA dynamics under intermittent therapy and provides a basis for predicting resistance in patient-specific strategies.
Bibliography
@article{bruchovsky2006final,
title={Final results of the Canadian prospective phase II trial of intermittent androgen suppression for men in biochemical recurrence after radiotherapy for locally advanced prostate cancer: clinical parameters},
author={Bruchovsky, Nicholas and Klotz, Laurence and Crook, Juanita and Malone, Shawn and Ludgate, Charles and Morris, W James and Gleave, Martin E and Goldenberg, S Larry},
journal={Cancer},
volume={107},
number={2},
pages={389--395},
year={2006},
publisher={Wiley Online Library}
}
