We modeled tumour evolution as an agent-based discrete time branching process that tracks the expansion of diverse clonal lineages as they acquire driver and passenger mutations that alter their proliferation and mutation rates. Clonal proliferation is subject to a spatio-temporal size-dependent penalty to provide characteristic tumour growth patterns. Once the tumour attains a diagnosable size (1 to 4 billion cells), a mitotic phase-specific perturbation is introduced to model anticancer agents. This environmental disruption impacts clonal dynamics, and we observed diverse heterogeneity, genomic instability, and resistance evolutionary paths. Our tool recovers various tumour development rates seen in the clinic, in which genomic instability promotes clonal diversification, leading to a state of invasiveness and prevailing (cross)-resistance.
That the last couple of years prior to diagnosis are essential in the pathogenesis of the tumour, which requires 2 - 6 driver mutations to bypass the effects of anticancer agents. Our simulated clinical trials comparing cytotoxic and targeted drug combinations show that moderate-dose schemes lead to prolonged survival rates, even in the presence of pre-existing drug resistant clones. Therefore, treatments maintaining clonal proportions should be considered as an alternative way for tumour growth control.
With a Cramer von Misses fitting method we are estimating the parameters of our model to replicate the cases of the BIG 1-98 clinical trial. This approach will enable us to identify the genomic and clinical profile of non-responders and establish a potential evolutionary history of their tumours. Moreover, it will simulate multiple drug dosing regimens to identify the optimal schedule that maximises overall survival. The ultimate goal is to determine the impact of genomic instability in therapeutic failure.