Personalised treatment schedules for metastatic prostate cancer — A set of novel mathematical biomarkers

Dynamic approaches to drug scheduling, such as adaptive therapy, enable individual-level personalisation of the dosing schedule to delay patient relapse. Promising clinical results in prostate cancer indicate the potential of these protocols, but demonstrate broad heterogeneity in patient response. This naturally leads to the question: why does this heterogeneity occur, and is a ‘one-size-fits-all' protocol best for all patients?

Using deep reinforcement learning, we obtain personalised and clinically feasible treatment protocols based on individual patient dynamics. We can subsequently rationalise these findings through a mathematical tumour model, and propose new mathematical biomarkers that can identify the best responders from a clinical dataset after only the first treatment cycle. Overall, I will highlight the importance of personalised treatment schedules that explicitly account for patient heterogeneity, and the power of mathematical models to capture, analyse and facilitate this personalisation.

Kit Gallagher completed his PhD at the University of Oxford under the supervision of Prof. Philip Maini, using population modelling, statistical inference, and deep learning to improve treatment scheduling and personalisation for metastatic prostate cancer. Throughout this, he has worked with Prof. Alexander Anderson at the Moffitt Cancer Center (Tampa, Florida), integrating clinical data into mathematical modelling and informing experimental & clinical trial design. Currently, he is starting a postdoctoral fellowship in Prof. Ignacio Vazquez-Garcia’s lab at Mass General Research Institute and Harvard Medical School, applying time series modelling to longitudinal genomic datasets.

ONLINE BBB: Link ZIH-Colloquia