Effect of Overdispersion of Lethal Lesions on Cell Survival Curves

TL;DR

This study investigates how overdispersion of DNA damage affects cell survival curves after hadron exposure, using a negative binomial model to better fit experimental data and understand dose-response behaviors.

Contribution

It introduces a negative binomial distribution-based model accounting for overdispersion in lethal lesions, improving fit to experimental survival data at various doses and LET values.

Findings

Overdispersion causes survival curves to straighten at high doses.

The NB model fits low and intermediate dose data better than existing models.

The approach offers insights into microscopic mechanisms affecting radiobiological responses.

Abstract

We explore the effects of overdispersed DNA lesion distribution on the shapes of cell surviving curves of mammalian cells exposed to hadrons at various doses. To provide a theoretical framework in resolving discrepancies between experimental data and Linear-quadratic (LQ)model predictions, we employ a non-Poisson distribution of lethal lesions together with non-homologous end-joining (NHEJ) pathway of double-strand break (DSB) repair. A negative binomial (NB)distribution is used to study the effect of the overdispersion on the shapes and possible reduction of dose-response curvature at high doses. The error distribution is customized to include an adjustable parameter so that the overdispersion parameter of NB is not constant but depends on the mean of the distribution. The trends in predicted cell survival responses are compared with the experimental data in low and high dose regions at various LET values for proton, helium, and carbon ions. The cell survival responses calculated by the present method reveal straightening of survival curves at high doses. This suggests that the overdispersion causes the cell survival dose-response to approximate log-linear behaviour at high doses. Comparison of the cell survival predictions with the Particle Irradiation Data Ensemble (PIDE) shows that the NB model provides better fits to the experimental data following low and intermediate doses. Whereas the model predictions are not validated at tiny and very high doses, nonetheless, the presented approach provides insight into underlying microscopic mechanisms which may help to improve the radiobiological responses along the dose-response curves and resolve discrepancies between experimental data and current cell survival models.