Investigating the FLASH Effect: Uncovering Brain Sparing and Cognitive Preservation Across Varying Dose Rates in Whole-Brain Irradiation of In-Vivo Mouse Models

Abstract

Purpose: FLASH irradiation, characterized by ultra-high dose delivery over short timeintervals, is fundamentally a function of dose rate. It has demonstrated strong potential for clinical applications, reducing normal tissue toxicity and complications compared to conventional irradiation. This phenomenon, known as the FLASH effect, has the potential to expand the therapeutic window and improve the effectiveness of radiation therapy for cancer and other diseases. While the FLASH effect is typically observed at mean dose rates (MDRs) ě 40 Gy/s, few studies have examined its behavior across a broad range of dose rates within the defined FLASH MDR spectrum, which extends from 40 Gy/s to over 2.5 million Gy/s. This study is among the first to investigate the FLASH effect as a function of dose rate across such an extensive range. However, the underlying conditions and mechanisms remain poorly understood. Duke University is uniquely positioned to investigate these mechanisms through the High Intensity Gamma-ray Source (HIGS) linear accelerator at the Triangle Universities Nuclear Laboratory (TUNL). This facility, previously characterized in studies conducted within the Oldham Laboratory, serves as a powerful FLASH radiation source, offering the capability to irradiate samples at exceptionally high dose rates. This study presents the first in-vivo FLASH irradiations conducted at Duke University using a live mouse model to investigate the complexities of the FLASH effect in normal tissue and its potential for cognitive function sparing. The experimental platform outlined in this thesis details key methodologies and demonstrates the feasibility of conducting in- vivo FLASH experiments. This work aims to evaluate independent methods for assessing normal tissue response and cognitive behavioral health across a broad spectrum of dose rates, ranging from conventional to ultra-high. Methods: The experiments utilized a 35 MeV electron beam from the Triangle Universi- ties Nuclear Laboratory’s (TUNL) High Intensity Gamma-ray Source (HIGS) linac, located on Duke University’s west campus. The studies presented in this thesis, designated as HIGS 09–14 within the Oldham Lab, were conducted following IACUC animal work approval and iii spanned two years. Mice were divided into treatment groups based on varying dose rates and subjected to whole-brain irradiation, with dose rates ranging from 2.07E + 07 Gy/s to 0.5 Gy/s. Dosimetry was performed using EBT-XD film, previously calibrated with the Varian 2100X, and analyzed using methods developed by prior students. The film was scanned with an EPSON 11000XL scanner to determine the delivered dose, with placement on the front and back of each mouse allowing for interpolation of brain dose and verification of accurate targeting. Following irradiation, mice were sent to the Duke University Behavior Core for a series of cognitive assessments to evaluate potential post-irradiation effects in healthy mice. A subset of each cohort was sacrificed on day four post-treatment for brain sample collection, which was analyzed using confocal microscopy and cytokine profiling. Cytokine analysis was conducted by Eve Technologies, and statistical significance was assessed using ANOVA tests. Additionally, two conventional irradiation experiments were performed, replicating the total dose delivered to the HIGS-irradiated groups. These mice underwent the same be- havioral tests and brain quartering procedures for further analysis. Results: Mouse cohorts were monitored weekly for weight and phenotypic changes. Notably, mice in the conventionally irradiated, total-dose-matched group experienced sig- nificant weight loss, reaching the ethical endpoint and requiring sacrifice 10 days post- irradiation. All irradiated mice showed hair loss and discoloration. Behavioral testing revealed that mice irradiated at MDR 2.07E + 07 Gy/s demonstrated greater preservation of cognitive and spatial learning abilities 2–3 months post-irradiation compared to those treated at lower dose rates (MDRs 0.15, 0.83, and 56.11 Gy/s). Small but significant differences in anxiety-like behaviors were also observed across treatment groups compared to non-irradiated controls. Cytokine analysis showed that interleukin cytokines and TNFα exhibited similar pat- terns of depression in expression relative to the conventionally irradiated group. Notably, cytokine levels in the MDR 56.11 Gy/s group were reduced relative to NoIR controls, iv whereas levels in the MDR 1.66E + 07 Gy/s group were comparatively higher, suggesting a dose-rate-dependent response. The HIGS linac can deliver a mean dose rate equal to the instantaneous dose rate in an in-vivo mouse model with high repeatability. Newly developed targeting techniques enabled accurate cranial whole-brain irradiation, with Monte Carlo simulations calculating beam divergence per 1 cm to be 0.71 mm in the x-direction and approximately 0.62 mm in the y-direction. Targeting accuracy was confirmed using EBT-XD film, which showed strong agreement between simulation and experimental targeting, as evidenced by the clear outline of mouse phantoms on film. Conclusions: A wide range of dose rates is achievable on the HIGS linac, including some of the highest instantaneous dose rates currently available. The combination of the HIGS linac and an in-vivo live mouse model presents a robust system for investigating the FLASH effect on normal tissue sparing. Moreover, the targeting techniques and mouse handling procedures developed in this work establish a strong foundation for future studies involving live animals, including applications in tumor control research. Qualitative data from this study supports the presence of a FLASH effect in the in-vivo model, with cognitive testing indicating that higher dose rates lead to better preservation of cognitive function. Cytokine analysis is suggestive that dose rates closer to the 40 Gy/s threshold, often used to define FLASH, may offer better long-term outcomes. Future research can expand upon these procedures to enhance data collection by in- corporating post-irradiation brain staining and imaging following behavioral assessments, providing deeper insight into the underlying biological mechanisms. Additionally, further studies are needed to evaluate the long-term side effects of whole-brain irradiation across various dose rates and to identify the optimal dose rate that maximizes tissue sparing while maintaining therapeutic effectiveness.