First FLASH Investigations Using a 35 MeV Electron Beam From the Duke/TUNL High Intensity Gamma-ray Source

Purpose: An interest in FLASH radiotherapy has been reawakened due to its noted ability to spare normal tissue, equal tumor control compared to conventional irradiation methods and technological advancement allowing for ultra-high dose rates required for FLASH radiotherapy to be more accessible compared to previous decades. The underlying biological mechanism of the FLASH effect are unknown and developing an in vitro model to study it has proven difficult. This work aims to combine two unique technologies, an organotypic rat brain slice model which models the in-vivo micro-environment in an in vitro setting and a linear accelerator capable of delivering variable FLASH pulses to design experiments which will facility the study of the FLASH effect.Methods: The experiments utilize a 35 MeV electron beam provided by Triangle Universities Nuclear Laboratory’s (TUNL) High Intensity Gamma-ray Source linac (HIGS). The beam can supply electron pulses with a temporal width of 1.2 s or 100 ns and work was performed with Gafchromic EBT3 and EBT-XD film to accurately determine the dose and dose rates of each pulse. Experiments were performed over 5 sessions to establish the use and effectiveness of the HIGS linac and biological rat brain model. A 2D translational stage was developed and targeting procedures were developed to ensure accurate targeting of each well containing an organotypic rat brain slice in a 12 well plate. Each rat brain was shot with a yellow fluorescent protein marker and seeded with 4T1 cancer cells tagged with mCherry and firefly luciferase. An imaging analysis workflow was developed to effectively capture and segment mCherry signal and determine the 4T1 proliferation four to five days after irradiation. These were compared to a final firefly luciferase readout. Each experiment was followed by a conventional irradiation as a control group. Monte Carlo model using TOPAS was created to simulate the HIGS linac dose profiles. Results: The HIGS linac can provide a mean dose rate up to 100 Gy/s and an instantaneous dose rate up to 100 MGy/s. The repeatability of the pulse dose was found to be within 4-5% of the average dose for a given experiment. Targeting was repeatable and dose superposition was confirmed. Well targeting quality assurance procedures of the translational stage allowed for consistent targeting of the pulse to each well. Yellow fluorescent protein bleed through in the mCherry signal was effectively filtered out and mCherry analysis reflects the end readout of firefly Luciferase. A gamma analysis between simulated and measured dose demonstrates a passing rate of 99.4% when using a criteria of 2%/2mm and threshold of 10%. Conclusions: FLASH capable dose rates can be supplied by the HIGS linac and is amongst the highest instantaneous dose rates currently available. The HIGS linac and organotypic rat brain model can be combined to irradiate and measure radiation effects to 4T1 cancer cell growth. There is qualitative data to support the observation of the FLASH effect in the rat brain model and the mCherry and firefly luciferase analysis agreement demonstrates the capabilities of the model to measure radiation effects to cancer cells in the 1-10 Gy range. Future work will be to quantitively measure the neuron health of the brain slices and DNA damage differences between FLASH and conventional irradiation.