A multi-institutional study to investigate the sparing effect after whole brain electron FLASH in mice: Reproducibility and temporal evolution of functional, electrophysiological, and neurogenic endpoints

Abstract

Background and purpose: Ultra-high dose-rate radiotherapy (FLASH) has been shown to mitigate normal tissue toxicities associated with conventional dose rate radiotherapy (CONV) without compromising tumor killing in preclinical models. A prominent challenge in preclinical radiation research, including FLASH, is validating both the physical dosimetry and the biological effects across multiple institutions.

Materials and methods: We previously demonstrated dosimetric reproducibility of two different electron FLASH devices at separate institutions using standardized phantoms and dosimeters. In this study, tumor-free adult female mice were given 10 Gy whole brain FLASH and CONV irradiation at both institutions and evaluated for the reproducibility and temporal evolution of multiple neurobiological endpoints.

Results: FLASH sparing of behavioral performance on novel object recognition (4 months post-irradiation) and of electrophysiologic long-term potentiation (LTP, 5 months post-irradiation) was reproduced between institutions. Differences between FLASH and CONV on the endpoints of hippocampal neurogenesis (Sox2, doublecortin), neuroinflammation (microglial activation), and electrophysiology (LTP) were not observed at early times (48 h to 2 weeks), but recovery of immature neurons by 3 weeks was greater with FLASH.

Conclusion: In summary, we demonstrated reproducible FLASH sparing effects on the brain between two different beams at two different institutions with validated dosimetry. FLASH sparing effects on the endpoints evaluated manifested at later but not the earliest time points.

Keywords: Electrophysiology; FLASH; Intercomparison; Neurobehavior; Neurogenesis; Neuroinflammation; Radiotherapy.

Fig. 1.. FLASH sparing of performance on the Novel Object Recognition (NOR) test late after 10 Gy single fraction whole brain irradiation is reproduced between institutions.

A) NOR scores from Stanford (4 months post-RT) were scored manually by core facility experts. CONV irradiated animals had significantly lower discrimination index than unirradiated controls, but FLASH irradiated animals were not statistically different from unirradiated controls. (n = 16/group, each datapoint represents an animal). B) NOR scores from CHUV at 3 timepoints (2,6- and 9-months post-RT) exhibit similar effects between cohorts in that a significant decrease in discrimination index was seen in the CONV irradiated animals compared to both unirradiated controls and FLASH irradiated animals. Data were analyzed using one-way ANOVA and the Bonferroni multiple comparisons test (n = 7–13/group at 2 months, n = 5–7/group at 6 months, n = 3–5/group at 9 months, each datapoint represents an animal). *,P≤0.05; **, P≤0.01; ***, P≤0.001;****, P≤0.0001; ns, not significant.

Fig. 2.. FLASH sparing of Long Term Potentiation (LTP) in late (5 month) mice after 10 Gy single fraction whole brain irradiation is reproduced between institutions.

A) The LTP measurement results from cohorts irradiated at Stanford. (left) Following a stable 20 min baseline recording, the slope of the field Excitatory Postsynaptic Potential (fEPSP) as a percentage of baseline shows an immediate increase in potentiation after delivering Theta Burst Stimulation (TBS). The combined slope of the CONV irradiated cohort fails to stabilize unlike the unirradiated control or the FLASH irradiated group. (middle) The mean potentiation 50–60 min post-TBS for each treatment group. The mean potentiation is significantly lower in the CONV group compared to both the unirradiated controls and the FLASH group. (right) Representative traces collected during baseline (black line) and 50–60 min post-TBS (red line) for each group. Scale = 1 mV/5ms. B) The fEPSP slope as percentage of baseline, mean potentiation and electrophysiological traces for the cohorts irradiated at CHUV. The group differences and levels of significance are the same between each institute. Data were analyzed using one-way ANOVA and the Bonferroni multiple comparisons test (n = 10–11 slices/group). ****, P≤0.0001.

Fig. 3.. LTP is not altered acutely (2 weeks) after 10 Gy single fraction whole brain irradiation by either FLASH or CONV.

The fEPSP slope as percentage of baseline is indistinguishable between unirradiated controls, CONV, and FLASH groups irradiated at CHUV. No significant difference between mean potentiation 50–60 min post-TBS was observed in either irradiation group. Data were analyzed using one-way ANOVA and the Bonferroni multiple comparisons test (n = 10–12 slices/group). ns, not significant. This contrasts with the decrement in LTP observed to emerge later after CONV but not FLASH (see Fig. 2).

Fig. 4.. Neuroinflammation in the hippocampus increases acutely (48 h) after 10 Gy single fraction whole brain irradiation by either FLASH or CONV.

Dual Stain of IBA-1 (microglial stain) and CD68 (activated microglial stain) in the hippocampus of mice 48 h after irradiation at Stanford. (left) The volume of overlap of the two stains is plotted with each data point representing a single section. A significant increase in the volume of activated microglia was observed in both the CONV and FLASH cohorts compared to unirradiated controls. (right) Representative images of the CD68 stain alone (top) with DAPI in blue, the IBA-1 stain alone (middle) and the combined stain (bottom) for each of the three treatment groups. Data were analyzed using one-way ANOVA and the Bonferroni multiple comparisons test (n = 5–11/group, each datapoint represents an average of 2–3 sections/animal). **, P≤0.01; ns, not significant. This contrasts with the resolution of neuroinflammation observed to emerge later after FLASH but not CONV.

Fig. 5.. The number of neural stem cells in the hippocampus is unchanged acutely (48 h) after 10 Gy single fraction whole brain irradiation by either FLASH or CONV.

Sox2 stain of the SGZ of the hippocampus of mice 48 h after irradiation of 10 Gy at Stanford. No significant difference between any irradiation group was observed (n = 3/group).

Fig. 6.. The number of immature neurons in the hippocampus is severely depleted acutely (1–2 weeks) after 10 Gy single fraction whole brain irradiation by either FLASH or CONV, but at 3 weeks recovers more with FLASH than CONV.

DCX staining in the hippocampus of mice 1 week, 2 weeks and 3 weeks after irradiation of 10 Gy at Stanford. A strong depletion in DCX+cells was observed in both CONV and FLASH groups to the same significance level at 1 week. At 2 weeks, some DCX+cells regenerated both in FLASH and CONV. However, the ongoing recovery of the FLASH groups reaches a significant difference compared to the CONV group at 3 weeks. Data were analyzed using one-way ANOVA and the Bonferroni multiple comparisons test (n = 4–5/group). *,P≤0.05; **,P≤0.01; ****,P≤0.0001; ns, not significant.