Adaptive therapy in bone metastatic prostate cancer - a window through agent-based models
Adaptive therapies in cancer, pioneered by my colleague at Moffitt’s IMO Bob Gatenby, work under the assumption that tumors already contain cells that are resistant to our clinical treatments. So rather than using those treatments in the conventional way, in a continuous fashion, it is better to alternate treatment and treatment vacations in an adaptive fashion, to ensure that there are always enough treatment-sensitive cells to keep the resistant cells in check.
Mathematical models, specially agent-based ones, are good theoretical frameworks in which to study adaptive therapies and, in this blog I have presented some ideas related to them previously [here, here, here and here]. These (and others) models tend to ignore the ecosystem that cancers inhabit which, in my view, is likely to affect if adaptive therapies work and how.
This bone metastatic agent-based model was modified so to include two types of tumor cells: TGF-Beta producing (which also requires that cytokine for survival) and TGF-Beta neutral. Cells that can use TGF-Beta proliferate faster (as they can use TGF-Beta) but are sensitive to a treatment that targets them and spares the TGF-Beta neutral cells. How does this tumor look like when a) we don’t treat it, b) we use continuous treatment or c) we use an adaptive therapy?
Lessons? this is very preliminary but clearly the impact of changing the window of control for the adaptive therapy. A key finding is that, under these very idealized conditions, adaptive therapies can control the tumor for an indefinite amount of time. The example that did the best (59-60K cells, bottom-right panel) is also the narrowest window and the one closes to the simulation’s carrying capacity so whether this Is generalizable will require further investigation.