ESTRO 2023 - Abstract Book

S660

Monday 15 May 2023

ESTRO 2023

especially for the extremities, which, through the joints, could experience large movements. In this work, the reproducibility of TMI patients’ extremities was evaluated to find the best positioning and reduce unwanted movements which are not accounted for during treatment. Materials and Methods 88 TMI patients (2013-22) treated with VMAT were analyzed. Patients were positioned on a modular immobilization system with three thermoplastic masks (head-shoulders, thorax-abdomen, and lower-extremities). During treatment, a cone beam computed tomography (CBCT) was performed for each isocenter (7-9 to cover the whole body) to best reposition the patient. All CBCTs were evaluated considering: (i) the CBCT-CT shift that best matched the two series; (ii) the extremities residual shift which would still be needed after the online matching to reposition the patient’s extremities in the original simulated position; (iii) the CBCT-CT agreement on the extremities using a qualitative index (range 1-5). Patients were subdivided according to the extremities immobilization methods: (i) arms either leaning on the frame to maximize patient comfort, or above the body to minimize the lateral field of view and facilitate the plan optimization; (ii) lower extremities, w/ or w/o a personal cushion for feet positioning (cushions were used when discarded from other cranial treatments – figure 1). The Mann-Whitney test was considered (p<0.05 significant).

Results 685 CBCT were analyzed (>7000 features). The overall mean online shift was 5±3 mm. No significant differences were found between the qualitative assessment and the online shift. Four CBCT/CT cases were reported in figure 2. Only 1.4% of cases qualitatively ranked <3. The mean qualitative evaluation improved over the years, from 3.6 in 2013 to 4.5 in 2019-22. Only 0.6% of cases had part of bones outside the PTV (note that CTV-PTV margin was 10 mm for the bones within the extremities). Arms leaning on the frame showed significant better agreement than arms suspended (6±5 mm vs. 9±5 mm, p=0.04). The use of a personal cushion to fix the feet significantly improved the residual agreement than without cushions (4±4 mm vs. 8±5 mm, p<0.01).