ESTRO 2025 - Abstract Book

S3107

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

1428

Digital Poster Assessment of the clinical impact of use of bowel bag structure in place of bowel loops for CBCT guided daily

online adaptive radiotherapy Helen Y C Wang, Matthew Bolt Radiotherapy Physics, Royal Surrey County Hospital, Guildford, United Kingdom

Purpose/Objective: Online Adaptive Radiotherapy (oART) has demonstrated reduction in PTV volume by 33.9% and V45Gy of bowel by 18.8% [1] through reduced margins, at the expense of additional contouring and planning time required during the treatment session. Shelley et al. reported an average of 29 minutes from CBCT acquisition to beam-on for Cervical cancer oART, of which 12.3 minutes were spent contouring, 7.5 minutes of those for bowel loops alone [2]. Removing the need to contour bowel loops should offer significant time saving, with the potential for further margin reduction. This study aims to access the feasibility of planning using alternative bowel structures rather than bowel loops through quantifying the dosimetric differences between automated plans generated using these different bowel structures for initial planning. Material/Methods: A retrospective study was completed for 10 Bladder and 10 Cervical cancer patients treated locally with oART. Plan optimisations were completed using the Varian Ethos treatment planning system with the locally developed RT Intent template. Four alternative bowel bag structures namely Bowel Bag (automatically generated by AI-Rad) [3], RCR Bowel Cavity [4], RTOG Bowel Bag [5], RTTQA Bowel Bag [6] were used as optimisation structures. No other user interaction was given during the plan optimisation. The PTV and OAR clinical goals were compared to the plan optimised using bowel loops as the clinical standard. The dosimetry of the plans were compared using Wilcoxon analysis. The bladder goals compared were D2%, D5%, D50%, D95% and D99% for PTV; V26Gy, V39Gy, V43Gy, V48Gy, V52Gy, V56Gy for bowel. The cervical goals compared were D50%, D95%, D99% for PTV; V30Gy, V40Gy for bowel. Results: Table 1 lists the mean dose parameters for bladder and cervical plans. Cervical dosimetry was comparable for plans optimised using all bowel structures with no statistically significant differences. Bladder plans showed statistically significant reduction in bowel loops dose (p<0.05) when optimised using alternative bowel bag structures for all goals except V56Gy as seen in Figure 1. The mean PTV V95% remained unchanged from the plan optimised using bowel loops. All bladder and cervical plans were clinically acceptable, including the automatically generated Bowel Bag structure.