Physicochemical indication of the FLASH effect from shoot-through proton pencil beam scanning parameters delivered under ultra-high dose rates.

Ultra-high dose rate (UHDR) proton pencil beam scanning (PBS) delivery results in irregular temporal-varying dose accumulation. It is difficult to establish a dose rate standard for the indication of proton PBS FLASH effect. In this work, we adopted a published physicochemical approach and investigated the impact of proton PBS UHDR parameters on the formation and downstream reactions of reactive oxygen species (ROS).From the ROS physicochemical model, the dose-rate dependent alkyl hydroperoxide (ROOH) formation was validated against published lipid peroxide absorbance data and correlated with mice skin damage data. For proton PBS delivery with specified beam current, voxelized temporal dose and ROS accumulation was calculated at the plateau region to simulate a shoot-through FLASH delivery. The ROS were obtained mimicking the irradiation of hypoxic skin. We examine the ROS-volume histogram in relation to the proton PBS delivery parameters.ROOH production clearly indicates sparing effects under UHDR. For PBS deliveries of 10 Gy to a 100 × 100 mmfield at 8 mm depth, the ROOH yield at 500 nA FLASH beam current is equivalent to a 8.78 Gy delivery at 1nA CONV delivery. The yield of ROOH depends strongly on the dose and beam current but has minimal dependency on the field size and spot spacing. Introducing inter-beam intervals of two minutes reduces the FLASH reduction in ROOH, consistent with reduced FLASH effect in murine experiment.The volumetric statistics of the ROOH yield showed consistent indication of FLASH effects in preclinical observations and correlated with the lipid peroxidation damage in tissue. Using simulated ROOH production metrics can potentially indicate the FLASH sparing effect under various PBS delivery parameters. Our simulations indicate that the shoot-through PBS FLASH effect depends mainly on the total dose and the pencil beam current, and is relatively independent of field sizes and spot spacings.