Dosimetric and biologic intercomparison between electron and proton FLASH beams

Abstract

Background and purpose: The FLASH effect has been validated in different preclinical experiments with electrons (eFLASH) and protons (pFLASH) operating at a mean dose rate above 40 Gy/s. However, no systematic intercomparison of the FLASH effect produced by e vs. pFLASH has yet been performed and constitutes the aim of the present study.

Materials and methods: The electron eRT6/Oriatron/CHUV/5.5 MeV and proton Gantry1/PSI/170 MeV were used to deliver conventional (0.1 Gy/s eCONV and pCONV) and FLASH (≥100 Gy/s eFLASH and pFLASH) irradiation. Protons were delivered in transmission. Dosimetric and biologic intercomparisons were performed with previously validated models.

Results: Doses measured at Gantry1 were in agreement (± 2.5%) with reference dosimeters calibrated at CHUV/IRA. The neurocognitive capacity of e and pFLASH irradiated mice was indistinguishable from the control while both e and pCONV irradiated cohorts showed cognitive decrements. Complete tumor response was obtained with the two beams and was similar between e and pFLASH vs. e and pCONV. Tumor rejection was similar indicating that T-cell memory response is beam-type and dose-rate independent.

Conclusion: Despite major differences in the temporal microstructure, this study shows that dosimetric standards can be established. The sparing of brain function and tumor control produced by the two beams were similar, suggesting that the most important physical parameter driving the FLASH effect is the overall time of exposure which should be in the range of hundreds of milliseconds for WBI in mice. In addition, we observed that immunological memory response is similar between electron and proton beams and is independent off the dose rate.

Figure 1:

Results of dose measurements (target dose 10 Gy) using PSI dosimeters, at dose rates of 0.1 Gy/s and 110 Gy/s.

Figure 2:

Results of the dosimetric intercomparison between PSI and IRA dosimeters, at dose rates of 0.1 Gy/s and 110 Gy/s. The dose measured with IRA dosimeters (Alanine, TLD, EBT3 IRA) is reported as Co-60 absorbed dose to water (A) and with experimentally determined corrections for beam quality applied (B). Error bars represent the combined standard uncertainty (k=1).

Figure 3:

Novel Object Recognition Test: Animals exposed to 10 Gy with both eFLASH and pFLASH, have statistically indistinguishable recognition ratios relative to controls indicating a preference for the novel object, whereas mice irradiated with eCONV and pCONV showed impairment compared to controls. Mean ± SEM (n = 10–12 per group); p-values were compared against CONV and derived from One-way ANOVA followed by Tukey’s correction for multiple comparisons. * p < 0.05, **p < 0.01**p < 0.001.

Figure 4:

e/pFLASH and e/pCONV are equipotent in curing animals and generate a similar immunological memory response against GL261 cell line. GL261 glioblastoma (GBM) tumors were irradiated at 20 Gy with e/pFLASH or e/pCONV after subcutaneous engraftment into immunocompetent C57BL/6J female mice (A and C). Cured immunocompetent C57BL/6J female mice were rechallenged with 5 × 10⁶ cells implanted in the opposite flank (B and D). Tumor growth delay was followed by caliper measurement 3 times per week. Results are given in individual values. Statistical analysis of tumor growth curves was performed using Mann-Whitney test. ns (0.12), * (0.0332), ** (0.0021), *** (0.0002), **** (<0.0001).