A biophysical model of cell evolution after cytotoxic treatments: Damage, repair and cell response

• Published in: Journal of theoretical biology 2016-01, Vol.389, p.146-158
• Main Authors: Tomezak, M., Abbadie, C., Lartigau, E., Cleri, F.
• Format: Article
• Language: English
• Subjects: Agent-based modelling; Algorithms; Bystander Effect; Calibration; Cancer radiotherapy; Cell Cycle; Cell Movement; Cell population modelling; Cell Proliferation; Cell Survival; Cell survival curves; Computer Simulation; Diffusion; DNA Breaks, Double-Stranded; DNA Repair; Dose-Response Relationship, Radiation; Drug Therapy - methods; Fibroblasts - drug effects; Fibroblasts - metabolism; Fibroblasts - radiation effects; Humans; Infant; Life Sciences; Male; Models, Biological; Monte Carlo Method; Neoplasms - drug therapy; Neoplasms - radiotherapy; Physics; Poisson Distribution; Probability; Radiotherapy - methods; Skin - metabolism
• ISSN: 0022-5193; 1095-8541
• DOI: 10.1016/j.jtbi.2015.10.017

We present a theoretical agent-based model of cell evolution under the action of cytotoxic treatments, such as radiotherapy or chemotherapy. The major features of cell cycle and proliferation, cell damage and repair, and chemical diffusion are included. Cell evolution is based on a discrete Markov chain, with cells stepping along a sequence of discrete internal states from ‘normal’ to ‘inactive’. Probabilistic laws are introduced for each type of event a cell can undergo during its life: duplication, arrest, senescence, damage, reparation, or death. We adjust the model parameters on a series of cell irradiation experiments, carried out in a clinical LINAC, in which the damage and repair kinetics of single- and double-strand breaks are followed. Two showcase applications of the model are then presented. In the first one, we reconstruct the cell survival curves from a number of published low- and high-dose irradiation experiments. We reobtain a very good description of the data without assuming the well-known linear-quadratic model, but instead including a variable DSB repair probability. The repair capability of the model spontaneously saturates to an exponential decay at increasingly high doses. As a second test, we attempt to simulate the two extreme possibilities of the so-called ‘bystander’ effect in radiotherapy: the ‘local’ effect versus a ‘global’ effect, respectively activated by the short-range or long-range diffusion of some factor, presumably secreted by the irradiated cells. Even with an oversimplified simulation, we could demonstrate a sizeable difference in the proliferation rate of non-irradiated cells, the proliferation acceleration being much larger for the global than the local effect, for relatively small fractions of irradiated cells in the colony. •Agent-based model of cell population damage by radio- and chemotherapy.•Includes cell damage, repair, healthy and tumoral cells, chemical diffusion, cell motility, and more.•Simulation of experimental cell survival curves and of bystander effect in radiotherapy.