1118 - Split Field IMPT Optimization to Maximize OAR Dose Rate and Implications for the Minimal FLASH Effect

P. Ramesh¹, Q. Lyu², D. Ruan³, and K. Sheng²; ¹University of California, Los Angeles, Los Angeles, CA, ²University of California, San Francisco, San Francisco, CA, ³Department of Radiation Oncology, University of California, Los Angeles, Los Angeles, CA

Purpose/Objective(s): Current multi-field optimization methods with the proton spot-scanning technique can achieve superior dosimetry compared to single-field plans but may fail to deliver ultra-high dose rates for FLASH normal tissue sparing due to long gantry switching times during delivery. To eliminate the gantry switching time for regions where a FLASH effect is desired, we propose the spatially separated beam orientation optimization (SSBOO) to penalize dose overlap between beams. Materials/

Methods: A voxel-repetition term, consisting of a sigmoid function, is added to our existing BOO framework to binarize the dose contribution in each voxel from each beam. The number of beams delivering dose to OAR voxels is constrained to be under one. Using the average dose rate definition, we investigate what the biological effectiveness of FLASH radiotherapy would have to be for one beam or two beams using the voxel-repetition term (2BM-VR) to surpass the dosimetry of 2-4 beams without the term (nBM-NVR). A bilateral head-and-neck, pancreas, and prostate case were tested.
Results: 2BM-VR increased the percentage of critical OAR volumes receiving greater than 40 Gy/s by 92.3% in the head-and-neck case, 60.3% in the prostate case, and 55.1% in the pancreas case compared to 2BM-NVR. For 1BM and 2BM-VR to outperform the OAR doses of [2BM-NVR, 3BM-NVR, 4BM-NVR], the FLASH dose rate would have to produce a [27.2%, 34.0%, 27.8%] and [28.7%, 29.4%, 22.5%] reduction in dose for the head-and-neck case, [11.1%, 25.2%, 16.3%] and [49.5%, 47.5%, 60.6%] reduction for the pancreas case, and [35.1%, 45.3%, 45.6%] and [37.8%, 50.3%, 44.7%] reduction for the prostate case.
Conclusion: The 2BM-VR SSBOO plans are able to achieve normal tissue sparing effects similar to those of single field plans for head-and-neck and prostate cases, if the biological effectiveness of FLASH radiotherapy is ~25-40%.

Organ	Dose (Gy)	Volume Receiving Dose (%)				FLASH Dose Rate Coverage (%)
		1BM	2BM-NVR	2BM-VR	3BM-NVR	1BM	2BM-NVR	2BM-VR	3BM-NVR
Glnd-Sub-R	15	90.8%	43.7%	20.4%	43.0%	100%	19.4%	100%	1.8%
Parotid-L	15	76.1%	20.1%	100%	19.1%	100%	4.8%	100%	2.0%
Constrictors	50	0.6%	0.5%	0.3%	0.4%	100%	6.8%	100%	5.1%
Rectum	20	4.3%	2.9%	3.4%	2.6%	100%	0%	27.8%	0%
Bladder	20	1.8%	1.6%	2.0%	1.4%	100%	1.3%	67.2%	0%
Penile Bulb	20	15.0%	13.0%	14.8%	0.1%	100%	11.6%	98.7%	0%
Stomach	20	1.3%	1.0%	0.9%	0.9%	100%	0%	46.3%	0%
Sm Bowel	20	2.5%	2.3%	0.9%	1.8%	100%	0%	61.7%	0%
Kidney-L	12	17.5%	19.8%	13.9%	<0.1%	100%	0%	41.0%	0%