Speaker
Description
Mice orthotopically implanted with bladder cancer showed improved responses on average to intravesical (inV) T cell therapy (OT1) when preconditioned with inV gemcitabine; however, responses varied among individual mice. We developed a virtual cohort for this combination therapy to expand of the murine population while capturing response variability.
Using parameters identifiable with ultrasound data, we fit the model to the 55 experimental mice to create 55 subject-specific calibrations. Then, we performed a treatment sweep of possible regimens on each calibration, aiming to minimize a cost function that balances the final tumor size with stable tumor growth over the treatment period. 13 regimens were optimal for at least one subject-specific calibration representing an experimental mouse. These 13 were tested on the virtual cohort and evaluated for robustness by determining which regimen produced the smallest final tumor size and the longest overall survival for the most virtual mice. Results led to the experimental testing of 3 regimens, which showed improved response compared to the original regimen. An additional digital twin cohort was included to stratify mice into the 3 regimens in real time based on their initial response from ultrasounds.
Better murine administration of these therapies can improve response in clinical trials for non-muscle-invasive bladder cancer, where patient-derived tumor-infiltrating lymphocytes would be used instead of murine OT1 cells.
