Int J Radiat Biol. 2026 Jan 12:1-14. doi: 10.1080/09553002.2025.2606998. Online ahead of print.
ABSTRACT
BACKGROUND: When the same energy is delivered to a cellular target, DNA damage and the resulting cellular response may vary depending on the density and distribution pattern of the energy delivered to the critical volume of each cell. DNA damage can be quantitated based on the pattern of dose distribution over the sub-micrometer volumes in nucleus. DNA double-strand breaks (DSBs) are considered the most critical events for cellular effects. Local effect model (LEM), DNA damage model (DDM), and Giant LOop Binary LEsion (GLOBLE) model have been used to predict cell survival under radiation exposure.
PURPOSE: This study aims to implement computational modeling for prediction of cell survival under radiation exposure, by quantitating radiation events on cellular targets, such as local energy deposition and DSB production, in a unified frame. The conceptual bases of LEM, DDM, and GLOBLE model were adopted to derive parameters for radiation events.
METHODS: The physics models of Geant4-DNA were used to simulate the interactions of X-rays and alpha particles with bio-matter. Cell nucleus was modeled to be a collection of sub-volumes. Statistical variation of energy deposition to individual sub-volumes was analyzed to count DSB production and DSB multiplicity. Cell surviving fractions (SFs) were calculated by LEM based on the distribution of local doses to sub-volumes and by DDM and GLOBLE model based on the DSB production and their potential interactions in sub-volumes. Model parameters were derived by fitting the models to experimental data for rat diencephalon (RD) cells and rat gliosarcoma (RG) cells.
RESULTS AND CONCLUSIONS: The overkill effect was reflected in the models based on LEM and DDM by employing threshold local dose and threshold number of DSBs in sub-volumes, respectively. Results suggest that the number of sub-volumes impacted with DSBs rather than the DSB multiplicity within individual sub-volumes would be better parameter to predict cell killing effect, which complies with the GLOBLE model.
PMID:41525142 | DOI:10.1080/09553002.2025.2606998