Feasibility of Clinically Practical Ultra-High Dose Rate (FLASH) Radiation Delivery by a Reversible Configuration of a Standard Clinical-Use Linear Accelerator

Abstract

Multiple preclinical in-vivo studies of ultra-high dose rate (FLASH) radiation therapy (RT) have shown reduced normal tissue toxicity with equivalent tumor control when compared to conventional dose rate RT, with benefits seen at dose rates ≥40 Gray (Gy) per second. Leveraging these benefits for clinical translation could profoundly impact RT. Extending an institutionally established reversible configuration of a linear accelerator (LINAC) for FLASH, we explored the feasibility of LINAC-based FLASH RT for potential clinical application.A standard, clinical-use treatment delivery system was used. A single decommissioned beam's program circuit board was replaced with a dedicated experimental board, holding control parameters for RF power and gun current. Dose rates at an electron beam energy of ∼16 MeV were maximized at a gun current setting of 11.8V grid voltage. Notably, this setup kept the scattering foil in the beam, maintaining the advantages of the clinical utility of a standard scattered high-energy electron beam, and its inherent achievable field sizes and depth doses. For measurements, gantry and collimator rotations were set to 0° with open primary jaws. Dose rates were measured at central axis using radiographic film, with a 3 cm buildup of solid water and 20 cm downstream. FLASH treatments were delivered in 90 pulses, with 2 independent measurements at 4 source-to-surface distance (SSD) positions, including at the machine head (59 cm SSD). Percentage depth dose (PDD) was measured with a film strip placed between two vertical 5 cm thick solid water phantoms, obtained at 70, 80, and 100 cm SSD using FLASH, and 100 cm SSD with conventional dose rate 16 MeV electrons.FLASH dose rates > 100 Gy/s were obtained using a standard, scattered electron beam at clinically relevant SSD's and depths with a clinical-use LINAC. Measured average dose rates at 100, 80, 70, and 59 cm SSD were 36.82, 59.52, 82.01, and 112.83 Gy/s, respectively. Average doses per pulse measured were 0.21, 0.33, 0.46, and 0.63 Gy/pulse, respectively. FLASH PDD's emulated 16-18 MeV energies, aligning closely with a conventional 16 MeV PDD. At 70, 80, and 100 cm SSD FLASH, and 100 cm SSD conventional 16 MeV, 90% fall-off measured at 5.2, 5.3, 5.5, and 5.2 cm depth, 80% at 5.9, 6.0, 6.1, and 5.7 cm and 50% at 7.0, 7.0, 7.1, and 6.8 cm depth, respectively.Using a readily reversible configuration of a standard LINAC, FLASH dose rates > 100 Gy/s was achievable at clinically applicable SSDs and depths. Balancing clinical feasibility with FLASH, further studies characterizing the FLASH beam at 70 to 90 cm SSD are underway, including development of a cone-less electron field shaping system and solutions for energy modulation. Advantages of our setup, using a standard scattering foil with high energy electrons, bring us a step closer to clinical practicality of FLASH RT delivery using a standard LINAC.