Speaker
Description
Introduction
Conventional radiotherapy (CONV) delivers a radiation treatment dose in order of minutes, resulting in dose rate of about 2 Gy/min. However, multiple recent preclinical studies demonstrated substantial healthy tissue sparing effect if the ultra-high dose rate, FLASH (> 40 Gy/s), is employed instead. At the same time, radiobiological effect on tumor remains unchanged. However, precise explanation of FLASH effect is still unclear, and dose measurements at high beam currents remains challenging.
The aims of this study are therefore to develop a reliable dosimetry setup and employ it to investigate cellular mechanisms involved in FLASH effect during irradiations with carbon ions. Carbon ions are gaining interest in radiotherapy as they offer superior treatment possibilities and a wide range of Linear Energy Transfer (LET) suitable for mechanistic investigations.
Methodology
In this study, we adapted a passively scattered carbon ion beamline at GANIL to deliver FLASH (40-60 Gy/s) and CONV (1.5-2.5 Gy/min) dose rates in the same conditions. The dose rate and the radiation field of 1.8 x 1.8 cm$^2$ is monitored with secondary electron detector, X-ray detector, multi-strip ionization chamber DOSION, and gafchromic films, as well as verified in GATE Monte Carlo environment.
The setup was employed to irradiate two cell lines (in monolayers): lung cancer cells (A549) and healthy fibroblasts (AG01522), at LET of either 30 or 80 keV/µm, and in either normoxia (21% pO$_2$) or hypoxia (1% pO$_2$) conditions. Biological endpoints included cell survival, cell cycle arrest, DNA damage, and Reactive Oxygen Species (ROS) in mitochondria.
Results
Radiochromic films and Monte Carlo simulations have shown a radiation fields of at least 90% homogeneity. Dose read-out from films, DOSION, X-ray detector, and secondary electron detector were found consistent within 90%. Additionally, we were able to observe beam spatial (~0.1 mm) and time (~0.02 ms) structures with DOSION. LET values obtained in GATE were in good agreement with independent calculations in SRIM software.
Regarding cell responses, we observed a reduced cell death of fibroblast in FLASH in hypoxia. The effect was dependent on dose and LET, up to around 50% (isoeffect at 15% survival). The mechanistic model of radiologic oxygen depletion (one of FLASH explanation hypotheses) was not able to explain observed differences. Changing the dose rate did not change the response of lung carcinoma, as expected. Other endpoints will also be discussed.
Conclusions
This preliminary study validates the sparing effect of carbon FLASH in vitro in AG01522 fibroblasts as well as unaltered response of tumor A549 cells, for different doses, LET, and oxygen concentrations. We also demonstrated a reliable dosimetry solution for preclinical $^{12}$C-FLASH studies. We plan to pursuit this research to potentially unravel new treatment solutions with heavy ions and FLASH.
