Integrating nano- and micrometer-scale energy deposition models for mechanistic prediction of radiation-induced DNA damage and cell survival

TL;DR

This paper introduces an integrated multi-scale modeling framework combining physical DNA damage simulations with biological cell survival predictions, validated against experimental data across various radiation types and energies.

Contribution

It presents a novel combined modeling approach linking nanodosimetric and microdosimetric data to biological outcomes, with innovative damage clustering analysis.

Findings

Excellent agreement with experimental survival data

Effective multi-scale linkage of physical and biological models

Potential for optimized radiotherapy applications

Abstract

We present an integrated modeling framework that combines the Generalized Stochastic Microdosimetric Model (GSM2), used to predict cell survival fractions, with MINAS-TIRITH, a fast and efficient Geant4 DNA-based tool for simulating radiation-induced DNA damage in cell populations. This approach enables the generation of spatially and structurally resolved double-strand break (DSB) distributions, capturing key features such as damage complexity and chromosome specificity. A novel application of the DBSCAN clustering algorithm is introduced to group DSBs at the micrometer scale. This allows the identification of physical aggregates of DNA damage and their association with subnuclear domains, providing a direct link to the cell survival probability as predicted by \gsm. The model was validated using experimental data from HUVEC cells irradiated with 220 kV X-rays and H460 cells exposed to protons over a wide linear energy transfer (LET) range, from approximately 4 keV/{\mu}m to over 20 keV/{\mu}m. Results show excellent agreement between simulations and experimental survival probabilities, making this one of the first consistent multi-scale models to bridge nanodosimetric and microdosimetric representations of radiation with biological outcomes such as cell survival. By incorporating the inherent stochastic nature of radiation-matter interactions, this framework effectively connects the physical properties of the radiation field to the biological response at the cellular level. Its accuracy across various radiation types and energies supports its potential for use in biologically optimized radiotherapy.