Purpose: The mechanism underlying the tissue sparing effect of FLASH radiation therapy is currently unknown, but the depletion of tissue oxygen at FLASH dose rates has been proposed as a potential candidate. This hypothesis has only been supported by modeling studies, primarily due to the difficulty of measuring oxygen concentration on millisecond timescales during irradiation. Here we use the phosphorescence quenching method to monitor oxygen concentration during proton delivery at conventional and FLASH dose rates.
Methods: Oxygen concentration was measured at rates of up to 3kHz using a soluble phosphorescent oxygen probe and a phosphorometer. Sealed glass vials, containing the experimental solution and probe, were placed in a proton beam and irradiated at conventional (conv) (~1Gy/s) and FLASH (~100Gy/s) dose rates. The g-value of oxygen depletion was measured at multiple dose rates and oxygen concentrations. Experiments were performed using three solutions containing 1,2, and 5% albumin and a fourth designed to emulate the intracellular environment, containing glucose, glycerol, and glutathione (Buffer 3G). Oxygen concentration was continuously monitored before, during, and after irradiation.
Results: We found that FLASH dose rates resulted in lower g-values for oxygen depletion when compared to conventional dose rates by ~15% (range: 7.2-22.5%) for all solutions tested. E.g., for a 30Gy dose to Buffer 3G, the g-value was 16% lower for FLASH than for conventional (gconv=0.72 ± 0.02μM/Gy, gFLASH=0.63 ± 0.02μM/Gy) (all values listed are mean ± S.E.M.). The g-value was higher for Buffer 3G than for the albumin solutions (gconv=0.72 ± 0.02 and 0.58 ± 0.03μM/Gy respectively, gFLASH=0.63 ± 0.02 and 0.54 ± 0.01μM/Gy respectively).
Conclusion: We have demonstrated that ultra-fast measurements of oxygen concentration can be performed during actual proton irradiation. Furthermore, we found that the g-value of oxygen depletion at FLASH dose rates is lower than that of conventional dose rates by ~15%.