ESTRO 2025 - Abstract Book

S3167

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

Conclusion: In short, this work quantified and compared the voxel mapping accuracy of four DIR software, and demonstrated the potential for a more convenient DDA quality assurance process by making use of contour-based metrics.

Keywords: Deformable Image Registration, Dose Accumulation

References: 1. Wong, Y. M. et al. Deformable anthropomorphic pelvis phantom for dose accumulation verification. Phys. Med. Biol. 69 , 12NT01 (2024).

3274

Digital Poster Assessing the Need for Repainting Energy Layers in Proton Therapy for Breast Cancer Bojan Štrbac, Truls Andersen, Pietro Pisciotta, Jeffrey Free, John H. Maduro, Johannes A. Langendijk, Stefan Both Department of Radiation Oncology, University Medical Center Groningen, Groningen, Netherlands Purpose/Objective: The interplay between the delivery of proton pencil beams and respiratory motion can significantly impact dose distribution. Motion mitigation strategies, such as breath hold, gating, and layer repainting, aim to reduce these effects. At our institution, layer repainting—rescanning each energy layer multiple times before transitioning to the next—is employed to statistically average motion effects. This study investigates whether layer repainting (5 and 3 rescans) provides a significant dosimetric benefit compared to No repainting in breast cancer proton therapy. These findings are important as eliminating repainting would decrease treatment session duration, increasing patient comfort and throughput. Material/Methods: A 4D dose evaluation was conducted using beam log files from proton therapy deliveries for 10 breast cancer patients. Patient-specific log files, representing the actual time structure of the beam delivery, were analyzed alongside 4D-CT datasets to account for respiratory motion effects. The delivered dose was evaluated for each treatment session and cumulatively across the entire course of treatment. Plans with 5 repaintings , 3 repaintings , and No repainting were compared. Dosimetric metrics, including D98 and D2 for target structures, were assessed

to evaluate target coverage. Statistical analyses included: •

Shapiro-Wilk test : To assess data normality • Wilcoxon signed-rank test : To compare paired differences between repainting methods. A significance threshold of p < 0.05 was applied.

Results: The Shapiro-Wilk test indicated that most dosimetric metrics were normally distributed (p > 0.05). Paired t-tests showed no statistically significant differences between 5 and 3 repaintings at the target level (e.g., D98 low CTV: p = 0.352). Comparisons between No repainting and the two repainting strategies revealed no significant differences in target coverage across the analyzed structures (e.g., Wilcoxon p = 0.75 for 5 vs No). Boxplot analyses indicated comparable interquartile ranges, with slightly higher variability observed for No repainting in the IMN structure, though this was not statistically significant (Figure 1).