Modeling beam timing effects in FLASH proton therapy with TOPAS‑nBio Monte Carlo radiochemistry

Purpose: FLASH proton radiotherapy delivered at ultra-high dose rates (UHDR; ≥40 Gy/s) has been reported to reduce normal-tissue toxicity without compromising tumor control; however, the underlying radiochemical mechanisms remain unclear. This study aims to investigate how beam temporal structure modulates radical dynamics during proton irradiation and compares FLASH with conventional delivery. Methods: Monte Carlo radiochemical simulations were performed using TOPAS-nBio with the Independent Reaction Time (IRT) approach, in which reaction times between radical pairs are stochastically sampled based on diffusion-controlled kinetics. A monoenergetic 100 MeV proton beam with an isotropic distribution was modeled in liquid water with a fixed total dose of 10 Gy. Two delivery regimes were compared: conventional (1000 pulses; 1 Gy/s) and FLASH (10 pulses; 100 Gy/s), with a pulse width of 10 µs and pulse interval of 0.01 s. Sensitivity analyses were conducted by varying pulse interval (1 s to 1 µs) and pulse width (10 ps to 10 ms) while maintaining constant total dose. Results: Time-dependent G-values (molecules/100 eV) for hydrated electrons (e⁻aq), hydroxyl radicals (•OH), and hydrogen peroxide (H₂O₂) were evaluated. Under FLASH, the maximum G-values of •OH and e⁻aq decreased to 1.47 and 1.34, compared with 2.00 and 1.99 under conventional irradiation (reductions of 26.5% and 32.5%), while H₂O₂ yield increased. Extremely short pulse intervals (~1 µs) promoted inter-pulse radical accumulation, whereas ultrashort pulse widths accelerated early-phase radical kinetics. Conclusion: These findings indicate that beam temporal structure significantly modulates radiochemical evolution and suppresses reactive radical yield under FLASH relative to conventional delivery.