Dosimetric Impact of Dose Rate Fall-Off in FLASH Proton Pencil Beam Scanning Treatment.

PURPOSE/OBJECTIVE(S) The lifetime of hydroxyl radicals, which cause radiobiological damage, is estimated to be a few microseconds. This time is much shorter than the actual spot-to-spot scanning time in proton pencil beam scanning (PBS) delivery (> 10 milliseconds). This mismatch in timescales becomes problematic in the case of FLASH treatment using PBS. Given that the FLASH effect is a spatially local and temporally finite phenomenon, it will be more accurate to evaluate it based on the spatial dose rate in Gy/sec/ cm3. Thus, delivering high doses at FLASH dose rates using PBS could be harmful to healthy tissues irradiated by the non-FLASH dose rate. In this study, we evaluate the extent of the healthy tissues that can be exposed to non-FLASH dose as well as the quantification of the received non-FLASH doses. MATERIALS/METHODS The PBS dose rate was determined by the proton beam energy, the beamline transmission efficiency and the cyclotron beam current. Its spatial distribution is governed by pencil beam's Gaussian distribution. Our investigation was carried out using simulations and experimental measurements. Two independent but complementary measurement setups were used: (i) A radiographic film sandwiched in solid water, along the beam axis, was irradiated to 15 Gy with a FLASH proton beam of 245 MeV and the spatial dose distribution was evaluated. (ii) A volumetric plastic scintillator was irradiated by the same beam and a high-speed camera was used to record the spatial distributions of dose and dose rate at the imaging rate of 100,000 fps. The absolute dose was measured at 2cm water depth using a plane-parallel ionization chamber to calibrate the imaging data from (i) and (ii) which were used to validate the simulations. RESULTS Based on our simulation which validated using the film and scintillator imaging data, the dose rate is highest at the cyclotron maximum beam energy (250 MeV, ∼100% transmission), can achieve a maximum of 480 Gy/sec/cm3 at the central region of the Bragg peak. Then, as result of limited beamline efficiency, it drops to about 10% (47 Gy/sec/cm3) for the next highest beam energy (249 MeV) and exponentially decreases to 0.5 Gy/sec/cm3 for 70MeV. The dose rate values shift from FLASH to non-FLASH magnitudes depending on the location within the spatial dose distribution. The proportion of volume irradiated by a non-FLASH dose rate can be up to 40% of the total irradiated volume. Therefore, for a treatment using PBS, considerable amount of healthy tissue could receive non-FLASH dose. The received non-FLASH dose could be up to full prescription dose if the treatment is delivered in beam transmission mode and up to 30% of the prescription dose if the treatment is delivered using Bragg peaks. CONCLUSION Our results show that the Gaussian distribution of the proton beam can limit the FLASH benefit. In order to fully utilize the proton pencil beam for FLASH treatment, a robust solution to FLASH dose rate fall-off is needed.