Author: Lawrence F. Bronk, Fada Guan, Xun Jia, Youfang Lai, Miao Qi 👨🔬
Affiliation: Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins University, Johns Hopkins University, Department of Radiation Physics, The University of Texas MD Anderson Cancer Center 🌍
Purpose: DNA double-strand breaks (DSBs) are widely regarded as critical indicators of cellular response to ionizing radiation. This study aims to establish a direct, unbiased and universal correlation between the distribution of DNA damage and cell surviving fraction (SF).
Methods: SF was modeled as a function of DSBs, characterized by a nonlinear relationship with DSBs of various complexities (DSB2, DSB3 and DSB4+). We approximated this relationship using a second-order polynomial function. Microscopic Monte Carlo simulations were used to calculate DSBs yields and complexities as well as the microdosimetric quantity lineal energy. Parameters for lethal DSBs were obtained by fitting experimental SF data of lung cancer H460 cells under irradiations of 137Cs, proton dose-mean lineal energy (yD) ranging between 2.0 to 18.4 keV/μm, and carbon ion yD between 18.6 to 87.9 keV/μm.
Results: Without using linear energy transfer, lineal energy, dose, and radiation type explicitly during fitting, the model’s predictions of SF based on lethal DSBs aligned closely with corresponding experimental data across all values, achieving a root mean square error of 0.04. In particular, for irradiation with yD from 16.0 keV/μm to 36.2 keV/μm, the experimental SF value of carbon ion with higher or similar yD was higher than that of protons at the same dose, which was successfully captured by our model. For protons at 5 Gy with yD of 18.4 keV/μm, the average DSB yield was 13.52 ± 1.66 per Gbp per Gy, with 28.2% ± 1.6% high-complexity damage (DSB3, DSB4+). In contrast, carbon ions at yD of 18.6 keV/μm showed 4% lower DSB yields, 6.7% fewer high-complexity damages, and a more uniform spatial distribution.
Conclusion: The model, based solely on DNA damage, effectively predicts cell SFs across varying doses, microdosimetry parameters, and particle types, highlighting the model’s generalizability and its ability to reflect the underlying radiobiology mechanisms.