ESTRO 2021 Abstract Book

S1538

ESTRO 2021

DR-100% (total dose divided by total field duration) and DR-95% (95% dose in a point divided by shortest time to deliver 95% dose) were determined from treatment log files in the region receiving at least 95% dose. In vivo dosimetry was performed for all FLASH mice with alanine in the beam entrance and a fiber coupled ZnSe:O scintillating crystal with 50kHz sampling rate right behind the mouse leg (Fig 1A).

Results Dose inhomogeneities at the entrance caused by the relatively large spot spacing were washed out at the mouse leg depth (Fig 1B). Here, the dose was homogenous near the mouse foot target, but increased 10-20% near the brass shielding. The MU scaling for FLASH determined by ionization chamber agreed with calorimetry within 0.1%. The dose rate for CONV was 0.4 Gy/s. The mean (range) dose rate across all FLASH mice was 80Gy/s (69-90Gy/s) for DR-100% and 197Gy/s (153- 240Gy/s) for DR-95%. The alanine in vivo FLASH dose differed up to 12% from the planned dose (Fig 2A), but agreed better with the scintillator in vivo doses (Fig 2B).

Conclusion An experimental setup to demonstrate normal tissue sparing with proton PBS FLASH was established and thoroughly characterized. Daily FLASH MU scaling with an Advanced Markus chamber agreed well with calorimetry. For FLASH, the same dose variations relative to the planned dose were observed with two independent in vivo dosimetry systems (alanine and scintillators), indicating that the dose variations reflected the actual delivered dose.