Evaluation of intensity-modulated electron FLASH radiotherapy in a clinical setting using veterinary cases.

Abstract

The increased normal tissue tolerance for FLASH radiotherapy (FLASH-RT), as compared to conventional radiotherapy, was first observed in ultra-high dose rate electron beams. Initial clinical trials in companion animals have revealed a high risk of developing osteoradionecrosis following high-dose single-fraction electron FLASH-RT, which may be related to inhomogeneities in the dose distribution. In the current study, we aim to evaluate the possibilities of intensity-modulated electron FLASH-RT in a clinical setting to ensure a homogeneous dose distribution in future veterinary and human clinical trials. Our beam model in the treatment planning system electronRT (.decimal, LLC, Sanford, FL, USA) was based on a 10-MeV electron beam from a clinical linear accelerator used to treat veterinary patients with FLASH-RT in a clinical setting. In electronRT, the beam can be intensity-modulated using tungsten island blocks in the electron block cutout, and range-modulated using a customized bolus with variable thickness. Modulations were first validated in a heterogeneous phantom by comparing measured and calculated dose distributions. To evaluate the impact of intensity modulation in superficial single-fraction FLASH-RT, a treatment planning study was conducted, including eight canine cancer patient cases with simulated tumors in the head-and-neck region. For each case, treatment plans with and without intensity modulation were created for a uniform bolus and a range-modulating bolus. Treatment plans were evaluated using a target dose homogeneity index (HI), a conformity index (CI), the near-maximum dose outside the target ( ), and the near-minimum dose to the target ( ). By adding intensity modulation to plans with a uniform bolus, the HI could be improved (p=0.017). The combination of a range-modulating bolus and intensity modulation provided a further significant improvement of the HI as compared to using intensity modulation in combination with a uniform bolus (p=0.036). The range-modulating bolus also improved the CI compared to using a uniform bolus, both with an open beam (p=0.046) and with intensity modulation (p=0.018), as well as increased the (p=0.036 with open beam and p=0.05 with intensity modulation) and reduced the median (not significant). By using intensity-modulated electron FLASH-RT in combination with range-modulating bolus, the target dose homogeneity and conformity in canine patients with simulated tumors in complex areas in the head-and-neck region could be improved. By utilizing this technique, we hope to decrease the dose outside the target volume and avoid hot spots in future clinical electron FLASH-RT studies, thereby reducing the risk of radiation-induced toxicity.