Conventionally Fractionated FLASH Treatment Planning for Head and Neck Cancer using Transmission Beam Proton Therapy

Abstract

Preclinical research suggests ultrahigh dose rate irradiation (FLASH irradiation, e.g. ≥ 40Gy/s) can reduce normal tissue toxicity while maintaining tumor control. Although current evidence is for high dose in a single fraction, a benefit for conventionally fractionated treatments has not been excluded. We therefore investigated if ultrahigh dose rate 244MeV proton transmission beams (TBs) can be used to make acceptable plans for conventionally fractionated head and neck cancer (HNC) treatment, and how much of the dose could be delivered at FLASH-rates (≥ 40Gy/s). TB plans have the Bragg peak behind the target, mostly outside the body. Although they cannot exploit the Bragg peak, they have a sharp penumbra and planning and delivery are simplified. Plan quality was compared to intensity modulated proton therapy (IMPT) and volumetric modulated arc therapy (VMAT), and dose rate distributions are calculated for each beam. Photon VMAT plans, 3-field IMPT plans and 10-field proton TB plans, delivering 70/54.25Gy in 35 fractions to the boost/elective volume respectively, were made for 10 HNC patients. Optimization objective settings were based on knowledge-based planning models for VMAT and IMPT (also used for TB). To achieve higher spot peak dose rates (SPDRs), the TB plans were split into three subplans, based on spot monitor units (MUs) of 10-20, 20-40 and >40, each of which was delivered with different gantry current. Targeted intraoperative radiotherapy dose rate distribution and irradiation times were evaluated for both non-split and split TB plans. TB plan quality was compared to IMPT and VMAT. OAR sparing in TB plans was on average comparable to IMPT: mean oral cavity/body dose were 4.1/2.5Gy higher (9.1/2.2Gy lower than VMAT), other OAR doses (salivary glands, larynx, pharynx) were within +1.1 and -1.9Gy. For a spot dose contribution threshold of 2cGy/0.2cGy the average percentage of dose delivered at FLASH rates was 66%/37% for the split plans and 18%/10% for the non-split plans. The mean dose-averaged dose rate (DADR) was 61/39Gy/s for the split plans and 27/18Gy/s for non-split plans for thresholds of 2cGy/0.2cGy, respectively. The total beam on irradiation time was on average ∼51% less for split plans (1.9s vs 3.8s), but overall time including scanning was ∼16% higher (8.9s vs 7.6s). Conventionally fractionated single-energy TB proton plans can achieve comparable OAR sparing to IMPT and better than VMAT, with total irradiation times <10s. If a FLASH effect exists at low dose/fraction, then this will further enhance TB proton therapy. Higher proportions of FLASH dose rates were obtained for those areas irradiated with a higher dose. Splitting TB plans based on spot MUs into subplans delivered with different gantry currents is one way to increase the proportion of dose delivered at FLASH dose rates.