A MODEL OF RADIATION-INDUCED CELL KILLING: INSIGHTS INTO MECHANISMS AND APPLICATIONS FOR HADRONTHERAPY

Cell death is a crucial endpoint in radiation damage, and many theoretical models have been proposed; in this work, a mechanism-based, two-parameter model based on the link between cell death and chromosome aberrations was developed. More specifically, dicentrics, rings and large deletions were assumed to lead to clonogenic inactivation; furthermore, it was assumed that chromosome aberrations derive from µm-scale mis-rejoining of chromatin fragments, which in turn originate from “dirty” DNA double-strand breaks (called “Cluster Lesions”, or CLs). The threshold distance for chromatin fragment rejoining and the average number of CLs per Gy and per cell are the only two (semi-free) model parameters. The model, which was “translated” into a Monte Carlo code called BIANCA (BIophysical ANalysis of Cell death and chromosome Aberrations) simulating survival curves for different radiation types, was tested against experimental data on V79 and AG1522 cells exposed to photons, protons, alpha particles and carbon ions. The very good agreement between simulations and data validated the model and supported the assumptions. In particular, at least for doses up to few Gy, dicentrics, rings and large deletions were found to be lethal not only for X-rays, as already reported by others, but also for other radiation types; furthermore, the derived CL yields suggest that the critical DNA lesions leading to cell inactivation are more complex than “clean” DSBs, and that these lesions lead to chromosome aberrations via mis-rejoining of chromatin fragments at the µm-scale. Following validation, the model was applied to the characterization of the particle- and LET-dependence of cell killing, and to the prediction of cell death at different positions along hadrontherapy Spread-Out Bragg Peaks. The very good agreement between simulations and data suggests that this approach may be applied to predict cell killing by therapeutic beams. This work allowed to shed light on the mechanisms of radiation-induced cell death, to characterize the dependence of cell killing on radiation quality and to predict cell death by hadrontherapy beams. More generally, a mechanism-based tool was developed that in some minutes can predict cell inactivation by different doses of an ion beam of given energy, basing on an experimental photon curve and a single (experimental) survival point for the considered beam energy. The model does not use RBE values, which can be a source of uncertainties.