ESTRO 2025 - Abstract Book

S3928

Radiobiology - Normal tissue radiobiology

ESTRO 2025

Purpose/Objective: The precise treatment delivery conditions needed to achieve the normal tissue sparing effect of ultra-high dose rate radiotherapy (FLASH-RT) remains underexplored. Our research aims to systematically define the key experimental parameters necessary for administering pulsed-electron FLASH beams, including the temporal beam structure, fractionated dose delivery regimes and the impact of tissue oxygenation. Our goal is to provide clear guidance for translating FLASH-RT to the clinic. Material/Methods: The abdomen of C57BL/6 mice was irradiated with pulsed electron beams, using the Mobetron FLASH irradiator. The beam’s physical parameters were varied by adjusting the pulse repetition frequency and dose per pulse, spanning from conventional dose rate radiotherapy (CONV-RT) to FLASH-RT regimens. The impact of tissue oxygenation on the FLASH effect was examined by performing dose escalation irradiations in FLASH- and CONV-RT modes under various anesthesia conditions, including isoflurane carried by air or oxygen and ketamine. The sparing effect of fractionated FLASH-RT was assessed by varying the number of fractions and dose per fraction. Symptoms of intestinal toxicity following irradiation were evaluated for each configuration of experimental parameters. Results: Increasing the average dose rate while maintaining a constant dose per pulse and total dose reduced the radiotoxicity in intestine, achieving the maximum sparing effect at a minimum dose rate of 100 Gy/s. Conversely, reducing the average dose rate increased side effects and decreased overall survival independent of the dose per pulse (Figure 1).

The impact of anesthesia revealed difference on normal tissue complication probability (NTCP) between CONV-RT and FLASH-RT following irradiation with ketamine or isoflurane carried by medical air, maintaining a dose modifying factor (DMF) of 1.17-1.19. Isoflurane carried by oxygen shifted the FLASH-RT NTCP curve towards CONV-RT curve, reducing the DMF towards 1. Examining fractionated RT using 10 fractions with increasing doses resulted in overlapping CONV-RT and FLASH-RT NTCP curves followed by a DMF of 1 (Figure 2), indicating the loss of the FLASH effect within this setup.