Kit Gallagher

Now time for @kitcgallagher.bsky.social on cool work with @mathonco.bsky.social on clinically applicable mathematical biomarkers using a Lotka-Volterra model #UKMathBio

When implementing Adaptive Therapy (AT) treatment schedules clinically, patients are monitored (and treatment schedules updated) at discrete time intervals. This leads to a trade-off; more frequent clinical appointments allow greater control of the tumor at a higher average size (leading to longer time to progression, or TTP), but are more costly and logistically challenging to implement. We can derive the optimal threshold size for treatment to be used in an AT protocol, and plot this in white over different time intervals between appointments. This is superimposed over the TTP for all different treatment protocols (given by the coloured background) - super-critical protocols on one side of the optimal line leads to chaotic outcomes while sub-optimal protocols achieve a lower TTP. This pattern is tiled and inverted to create a 'chess board', with the highest TTP at the center of each tile, reflecting the optimal game that clinicians are trying to play against the cancer tumor.

As featured in the #mathonco newsletter, this inspired an artwork exploring the trade-off between appointment frequency and time to progression (background color). The optimal strategy line separates stable and chaotic outcomes, and we tile this pattern with the highest TTP at each tile's center

Deriving Optimal Treatment Timing for Adaptive Therapy: Matching the Model to the Tumor Dynamics Adaptive therapy (AT) protocols have been introduced to combat drug-resistance in cancer, and are characterized by breaks in maximum tolerated dose treatment (the current standard of care in most clin...

Chasing Perfection - previous work to optimize adaptive therapy schedules for cancer rely on monitoring the patient continuously. We find that accounting for discrete clinical appointments across multiple tumor models motivates patient-specific personalization in optimal tx!

doi.org/10.1101/2025...

Great news for the promotion and dissemination of open science, and hopefully institutions that benefit from this service will also be able to support this long-term initiative financially!

#HiSciSky! I'm in the final year of my PhD applying math modeling and deep learning to improve treatment scheduling in late-stage cancers. I'm starting to look for #postdocs in #mathonco and cancer #evolution if you know anyone who's hiring!

Sounds like what we need is something that makes it easier to generate quantitative, interpretable metrics that describe observable biological differences in tissue, without being a black box and requiring unreasonable amounts of training data...

www.muspan.co.uk

doi.org/10.1101/2024...