Author: Xiaoda Cong, Rohan Deraniyagala, Xuanfeng Ding, Xiaoqiang Li, Jian Liang, Peilin Liu, Craig Stevens, Xiangkun Xu, Weili Zheng 👨🔬
Affiliation: Corewell Health William Beaumont University Hospital, Corewellhealth William Beaumont University Hospital, William Beaumont University Hospital, Corewellhealth William Beaumont Hospital, Department of Radiation Oncology, Corewell Health William Beaumont University Hospital 🌍
Purpose:
Commission a step-and-shoot arc therapy(SPArc-step&shoot) for treating head-neck cancer patients as a desired interim milestone toward full dynamic treatment.
Methods:
An in-house developed spot and energy layer sparsity optimization algorithm was implemented in a clinical treatment plan system via scripting. The SPArc-step&shoot algorithm iterative selects the higher MU weighted energy layer and spots to ensure:(1) the efficiency of the treatment delivery of SPArc-step&shoot therapy and (2) superior plan quality while meeting the minimum-MU per spot machine requirement (0.02MU per spot). At meanwhile, an in-house developed Dynamic SPArc therapy simulator was used to simulate the treatment and calculate the delivery time. During the treatment course, a machine-learning based synthetic CT platform was implemented to monitor the dose robustness coverage. In June 2024, a head-neck cancer patient with parotid gland malignancy was treated using SPArc-step&shoot (66Gy in 33 fx) with nine static fields distributed at 20-degree intervals. Comparative plans, including SFO-IMPT and SPArc-Dynamic, were generated. Dose metrics, delivery times, and adaptive planning were evaluated.
Results:
SPArc-step&shoot and SPArc-Dynamic showed similar target coverage and OARs sparing, and the plan quality is superior to the three-field SFO-IMPT in the brain stem, oral cavity, and spinal cord sparing. (Figure 1). The simulated total dynamic treatment delivery time is 15.9mins, 6.32mins, and 4.31mins for SPArc-step&shoot, SFO-IMPT and SPArc-Dynamic, respectively(Table 1). The actual recorded average treatment delivery time for SPArc-step&shoot in 33 fx is 16.7 ± 1.56 mins. QA-CT and synthetic CT showed a similar target coverage degradation and perturbation, and a replan was initiated(Figure 2).
Conclusion:
The SPArc-step&shoot therapy was successfully implemented in the clinical settings, and 1st patient was successfully treated between June and August 2024. The synthetic CT platform serves a critical role in the daily monitoring process as SPArc-Dynamic might be more sensitive to the patient geometry changes in HNC treatment.