S1942
Physics - Dose prediction/calculation, optimisation and applications for photon and electron planning
ESTRO 2026
Digital Poster 3614
Australia. The CT-based reference plan was recalculated using bulk density overrides instead of HU-derived electron densities (ED). The number of structure overrides and their assigned densities were varied across 5 separate calculations (ED-plans) ranging from a single structure ED override (O1) to multiple structure, population-based, ED overrides (O1-4) and patient specific ED overrides (O5), see table. All plans were calculated using a 2mm grid and 0.5% statistical uncertainty and evaluated by two CNS Radiation Oncologists (RO) for clinical acceptability. Dosimetric equivalency to the CT-plan was evaluated based on DVH metrics and dose subtraction plots.
heart-sparing biaxially rotational dynamic- radiation therapy using customized dynamic trajectories in combination with gimbal for esophageal cancer Kanako Nakatsu 1 , Shinya Hiraoka 1 , Yuka Ono 1 , Hideaki Hirashima 1 , Ryo Narukami 1 , Katsuyuki Sakanaka 1 , Mitsuhiro Nakamura 2 , Takashi Mizowaki 1 1 Department of Radiation Oncology and Image- Applied Therapy, Graduate school of medicine, Kyoto university, Kyoto, Japan. 2 Department of Advanced Medical Physics, Graduate school of medicine, Kyoto university, Kyoto, Japan Purpose/Objective: Radiotherapy for esophageal cancer is required to reduce heart dose to prevent cardiac toxicity. OXRAY features an O-ring gantry in combination with a gimbal head, enabling an expanded irradiation field. It also delivers Biaxially Rotational Dynamic Radiation Therapy (BROAD-RT), providing customized dynamic trajectories for non-coplanar VMAT. This study aimed to evaluate whether customized dynamic trajectories in BROAD-RT contribute to reducing heart dose compared with coplanar VMAT (COVMAT). Material/Methods: This study included fifteen patients with locally advanced thoracic esophageal cancer who underwent chemoradiotherapy at our institution in 2024. For BROAD-RT, two customized trajectories were created using the Dijkstra algorithm: one optimized to avoid beam entry through the heart (Good Arc) and another intentionally designed to maximize beam entry through the heart (Bad Arc). For each patient, three plans—Good Arc, Bad Arc, and COVMAT—were generated (total = 45 plans), with a prescribed dose of 50.4 Gy in 28 fractions to 50% of the PTV. Dose– volume parameters for the PTV and OARs were compared using the Friedman test followed by pairwise Wilcoxon signed-rank tests with Bonferroni correction. The estimated delivery time for each plan was also measured and compared using the same statistical approach.
Results: Compared to the CT-plan, apart from ED-plan O1 (water alone), all other ED-plans were considered clinically acceptable by the RO. No significant differences were detected in reported DVH metrics, including GTV D100, CTV D95, PTV D95, PTV D2, and Dmax of organs of interest (brainstem, optic chiasm and optic nerves). By defining air, bone and soft tissue as the minimum number of density structures with population-based EDs, the plans generated were deemed dosimetrically equivalent (variation in DVH<1%) to the CT- plan, which uses voxelised, patient-specific EDs. Reviewing the dose subtraction plots however, provided specific geometric information which highlighted the need for additional scrutiny of structure numbers and applied EDs, to ensure accurate dosimetry when targets or organs of interest are close to tolerance and fall within or near to varying tissue density boundaries. Conclusion: Bulk density overrides with minimum tissue definitions of bone, air and soft tissue were found to be an equivalent surrogate for CT-based dose calculations for brain treatments on the MR-Linac. Extra consideration to the dosimetric accuracy of the calculation should be given when target or organ of interest volumes are near tolerance or fall within areas of varying tissue density. Keywords: electron density, MR-Linac, Brain
Results:
Made with FlippingBook - Share PDF online