Modelling cancer cell dormancy in mice: statistical distributions predict reactivation times and survival dynamics

Abstract

In predicting metastatic potential and improving treatment outcomes in cancer research, it is crucial that we understand the dynamics of cancer cell dormancy and reactivation. In this paper we propose, study, and evaluate a cancer growth model that incorporates cell death, dormancy, reactivation, and proliferation in the secondary sites. Using experimental data on mice, we test various statistical distributions and identify models that represent the asymmetry and variability observed in dormancy durations. Notably, the estimated cancer cell death rate remained consistent across all tested distributions, supporting its biological relevance as a robust parameter for modelling dormancy survival dynamics. The most suitable among the distributions we studied, the most suitable ones exhibit heavy-tails and asymmetric skewness; this aligns with the prolonged and rare dormancy periods expected of cancer cells. Our findings stress the importance of selecting appropriate statistical models for dormancy, both in predicting cancer cell reactivation events, and in informing therapeutic strategies that focus dormancy-driven metastasis.