Dose, dose rate and split dose impacts murine skin responses following photon FLASH irradiation

Preclinical evidence has demonstrated the potential of FLASH radiotherapy (FLASH-RT) to spare normal tissues compared to conventional (CONV) exposures. Most FLASH studies have used ultra-high dose rate (>40 Gy/sec) electrons and protons whilst comparatively few studies have reported photon FLASH responses. Given the widespread use of photons clinically, there is a need to characterise the FLASH effect using photons. In this study, we applied a novel photon FLASH system (FLASH-SARRP, Xstrahl) to investigate the effects of dose, dose rate and split dose on murine skin toxicity.

Skin toxicity was assessed at CONV (3.2 Gy/min) and FLASH (72 Gy/s) dose rates using the SARRP or FLASH-SARRP. CONV responses were investigated at a dose of 20.2 Gy and FLASH responses at doses of 18.1, 21.3 & 25.8 Gy. Comparative studies were conducted using a split dose exposure with an average dose rate of 2.8 Gy/s. Skin toxicity on the hind leg of C57BL/6 mice was visually scored and histopathological analysis performed at 8-12 weeks. Tumour growth delay was also assessed using a melanoma (B16-F10) xenograft model irradiated at FLASH and CONV dose rates.

Skin toxicity was delayed for FLASH exposures and tissue analysis showed hyperplasia and significant fibrosis deposition (p < 0.01) in CONV mice compared to FLASH. Tissue recovery was observed for both dose rates from 8 weeks post RT. A dose dependent relationship for FLASH sparing was observed, while a split dose exposure resulted in loss of sparing. FLASH was equally effective for tumour control in comparison to CONV exposures (p = 0.99).

These results demonstrate it is feasible to deliver photon FLASH exposures with sparing consistent with observations from previous studies using proton and electron beams. Dose, average dose rate and beam structure are key parameters that modulate radiobiological responses to photon FLASH.