Translating FLASH to the clinic: treatment planning system for a FLASH-compatible dose delivery using a novel x-ray UHDR machine

Objective. To develop and validate a treatment planning system (TPS) for a novel x-ray ultra-high dose rate (UHDR) system and compare its performance with conventional volumetric modulated arc therapy (VMAT). Approach. A TPS was developed for a novel x-ray UHDR system featuring stationary beamlines and a Scanning Pencil-beam High-speed Intensity-modulated x-ray source (SPHINX). We studied an exemplary case using 16 beam angles, 1D scanning perpendicular to the bore axis, and a linear accelerator operating at 12 MeV electron energy with 1 mA average beam current. Treatment plans were generated using various system configurations and compared with clinical VMAT plans for lung (PTV: 239 cm^3), brain (PTV: 372 cm^3), and head-and-neck (PTV: 55 cm^3) cases. Configurations included different beamlet widths for lung cancer, coplanar and conical beam arrangements for brain cancer, and different beamlet spacings for head-and-neck cancer. Dose distributions in terms of target conformity and homogeneity indices, local dose rates (LDR), local irradiation time (LIT), and organ-at-risk (OAR) sparing were compared. A validation workflow combining Monte Carlo simulations with TPS-generated treatment parameters was developed and tested on a lung case to ensure TPS dose calculation accuracy. Main results. TPS-generated plans achieved comparable target coverage to VMAT while delivering significantly higher dose rates (>12.5 Gy/s vs. 0.03 Gy/s) and ultra-short LITs. However, UHDR plans in general showed increased OAR mean doses. Conical beam arrangements in the brain case yielded higher maximum LDRs at isocenter of up to 49 Gy/s with LITs as low as 37 ms, but increased integral dose. The head-and-neck case demonstrated high LDRs (62 Gy/s to 98% of PTV) within 20 ms, with minimal differences between beamlet spacing configurations. The lung case validation workflow demonstrated >98% 3D gamma-index pass rate using 3%/3 mm threshold, confirming plan accuracy. Significance. The developed UHDR-TPS enables treatment planning for a novel x-ray UHDR system. In this preliminary study, it achieved plan quality comparable to VMAT for the specific lung, brain, and head-and-neck cancer cases studied, while delivering significantly higher dose rates and shorter irradiation times. Further optimization of the UHDR delivery with x-rays is needed, however, to decrease OAR doses and validate these initial findings.