Impact of Dose, Time, and Increasing Gaps in Dose Delivery on the FLASH Effect in Mouse Skin

Preclinical FLASH animal studies have primarily employed large single doses of radiation. These results suggest that the FLASH normal tissue-sparing effect may most readily be observed in the context of radiosurgery and stereotactic body radiotherapy, whereas most clinical schedules employ small doses per fraction. The present study examines the impact of dose size, dose rate, and overall treatment time on the magnitude of the FLASH sparing effect in mouse skin. Mouse skin contraction was evaluated 55 days after 27 Gy irradiation, delivered at FLASH (127.5 ± 2.5 Gy/s for 0.213 s) and conventional dose rates (CDR, 0.5 ± 0.01 Gy/s for 54 s). In addition, we introduced time gaps in dose delivery. This approach was based on the hypothesis that oxygen depletion is the mechanism of FLASH tissue sparing, with the loss of tissue sparing corresponding to the time for tissue to reoxygenate. Tissue irradiated with proton spot scanning FLASH treatments may experience similar time structures. To evaluate the impact of dose per pulse size, we delivered 27 Gy in 2 pulses of 13.5 Gy, 4 pulses of 6.75 Gy and 8 pulses of 3.375 Gy over the same 54 s time frame to deliver 27 Gy conventional dose rate irradiation. For each of the beam pulses, the instantaneous FLASH dose rate was 127.5 ± 2.5 Gy/s. To evaluate the impact of longer time intervals that would occur, for example, between treatment fields, two FLASH pulses of 13.5 Gy/pulse were administered, each lasting 0.125 s and separated by up to 150 s. To measure skin contraction, two tattoo dots were placed ∼1 cm apart on the hind leg of each mouse. The legs were then irradiated with a 1.6 × 1.2 cm 2 elliptical beam covering the tattoos. The decrease in the distance between the dots was evaluated at four-day intervals from day 4 to day 20 and weekly thereafter. Delivering a dose of 27 Gy in 2, 4, or 8 equally spaced pulses over 54 s, which is the same as the CDR irradiation time for 27 Gy, resulted in no observed difference in the contraction fraction for any spacing. When a single 27 Gy FLASH dose was administered or the dose was delivered in two 13.5 Gy pulses separated by only 0.125 s, a tissue sparing effect was observed compared to CDR. However, the FLASH effect was no longer observed at pulse separations of 0.463 s or longer, corresponding to dose rates of 40 Gy/s or lower. Our data show that administration of two dose pulses of FLASH irradiation does not elicit a tissue-sparing effect when of 0.463 s or longer elapses between pulses. The results indicate that the FLASH effect depends on the total time required to deliver the total dose. This study does not prove oxygen depletion as the mechanism of the FLASH effect, but it is consistent with it, with tissue reoxygenation during a 0.463 s interval or longer between irradiation pulses resulting in a loss of the FLASH tissue effect.