LAUSR

PURPOSE The combined use of Bragg peak and shoot-through beams has previously been shown to increase the normal tissue volume receiving FLASH dose rates while maintaining dose conformality compared to conventional IMPT methods. However, the fixed beam optimization method has not considered the effects of beam orientation on the dose and dose rates. To maximize the proton FLASH effect, here, we incorporate dose rate objectives into our beam orientation optimization (BOO) framework. METHODS From our previously developed group-sparsity dose objectives, we add upper and lower dose rate terms using a surrogate dose-averaged dose rate (DADR) definition and solve using the fast-iterative shrinking threshold algorithm. We compare the dosimetry for three head-and-neck cases between four techniques: 1) spread-out Bragg peak IMPT (BP) 2) dose rate optimization using Bragg peak beams only (BP-DR), 3) dose rate optimization using shoot-through beams only (ST-DR), and 4) dose rate optimization using combined BP and ST (BPST-DR), with the goal of sparing OARs without loss of tumor coverage and maintaining high dose rate within a 10mm region of interest (ROI) surrounding the CTV. RESULTS For [BP, BP-DR, ST-DR, BPST-DR], CTV HI and Dmax were found to be on average [0.886, 0.867, 0.687, 0.936] and [107%, 109%, 135%, 101%] of prescription, respectively. Although ST-DR plans were not able to meet dosimetric standards, BPST-DR was able to match or improve either maximum or mean dose in the right submandibular gland, left and right parotids, constrictors, larynx, and spinal cord compared to BP plans. Volume of ROIs receiving greater than 40 Gy/s ( V γ 0 ) ${{\rm{V}}_{{{\gamma}}0}})$ was [51.0%, 91.4%, 95.5%, 92.1%] on average. CONCLUSIONS The dose rate techniques, particularly BPST-DR, were able to significantly increase dose rate without compromising physical dose compared with BP. Our algorithm efficiently selects beams that are optimal for both dose and dose rate. This article is protected by copyright. All rights reserved.