FLASH radiotherapy (FLASH-RT) delivers radiation at ultra-high dose rates (UHDR) and has been shown to spare normal tissue while maintaining tumor control (FLASH effect). This could be due to a reduction in radiation-induced DNA damage in normal tissue. Consequently, plasmid assays have been proposed as a way to evaluate these potential differences. However, experimental results have been varied. Track-structure Monte Carlo (MC) simulations may offer a way to disentangle these differences. In this work, we propose a MC model of plasmid DNA damage at UHDR using TOPAS-nBio. Approach. The radiolysis of plasmids (pUC19, 50 μg/ml) in an oxygenated (21%) aqueous solution containing DMSO (0.01 - 100 mM) were modeled in-silico using TOPAS-nBio. 100 Gy was deposited in the solution by 225 kVp X-rays, delivered in a single pulse at conventional (CONV) dose rates (0.1 Gy/s) or UHDR (2 x 107Gy/s). Two models were evaluated, model 1 in which there was no DNA repair, and model 2 in which oxygen competition was introduced in the form of WR-1065 to induce chemical repair. These models were compared against published experimental data. Main results. At CONV dose rates, the model reproduced published experimental Single Strand Break (SSB) yields across a range of scavenging capacities, with statistical uncertainties within 2 % (one standard deviation). At low scavenging capacities, there was a 54.7% reduction in SSBs and 73.5% reduction in Double Strand Breaks at UHDR compared to CONV. At biologically relevant scavenging capacities this difference was within the statistical uncertainty, and there were no observed differences in chemical repair by WR-1065 between UHDR and CONV. Significance. These results suggest that the reduction in DNA damage observed experimentally at low DNA concentrations and low scavenging capacities is due to the intertrack effect, with no difference predicted at low DNA concentrations and cell-like scavenging capacities.
Masilela et al. (Tue,) studied this question.