RBE with Non-Poisson Distribution of Radiation Induced Strand Breaks

TL;DR

This paper introduces a non-Poisson statistical model based on the negative binomial distribution to better describe the clustering of lethal DNA lesions caused by high-LET radiation, improving understanding of RBE variations.

Contribution

It presents a mechanistically grounded non-Poisson model for lethal lesion distribution, enhancing predictions of biological effects of different ion types and LET values.

Findings

Model aligns with experimental data at medium and high LET.

Predictions are closer to Microdosimetric-Kinetic model for helium and carbon ions.

Clustering of lethal events explains RBE saturation at high LET.

Abstract

Postulating that increasing linear energy transfer (LET) causes non-random clustering of lethal lesions to deviate from the Poisson distribution, we employ a non-Poisson approach as a more flexible alternative that accounts for overdispersion of lethal lesions. Using non-homologous end-joining (NHEJ) pathway of double-strand break repair, a customized negative binomial (NB) distribution is used to describe the distribution of lethal events in a cell nucleus. The proposed model provides a novel, mechanistically based explanation for the measured values of the biological relevant quantities, such as model parameters and relative biological effectiveness (RBE) of the surviving cells, for various light ion types and LET values. The estimated quantities are compared with the predictions of several mechanism-inspired models and experimental data at medium and high LET values. The results examined are closer to the Microdosimetric-Kinetic model predictions for helium and carbon ions but progressively lower than trends predicted by the Local Effect Model and the Repair Misrepair-Fixation model in the large LET region. The results support the view that the limitation in the increase in RBE at high LET can be accounted for entirely, or, in large part, by clustering of lethal events to cause deviation from the Poisson distribution.