ESTRO 2025 - Abstract Book

S3141

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

Keywords: robust dose accumulation, proton therapy, thorax

2431

Digital Poster Dosimetric evaluation of on-couch target volume propagation for cervical cancer in CBCT-based online adaptive radiotherapy Robert Schindhelm, Sonja Wegener, Bülent Polat, Marcus Zimmermann, Gary Razinskas Department of Radiation Oncology, Department of Radiation Oncology, Würzburg, Germany Purpose/Objective: On-couch adaptive radiotherapy enables daily treatment plan adjustments to account for interfractional anatomical changes. The AI-driven Varian Ethos system utilizes cone-beam computed tomography (CBCT) to capture organ at risk (OAR) and target volume (TV) position and shape variations. During adaptation, the AI automatically contours OARs — termed influencer structures — that directly impact the irradiated volume. After physician review, TVs are propagated onto the CBCT to align with contours from the planning CT (pCT). Boolean operations, as defined on the pCT, are then applied. Minimizing patient time on the couch requires accurate initial TV propagation and efficient Boolean operations. This study evaluates dosimetric accuracy of treatment plans optimized using Ethos-propagated TVs. Material/Methods: From February to July 2024, four cervical cancer patients received adaptive radiotherapy, including treatment to the lymph node area, using Ethos V1.1. Three patients received 1.8 Gy in 28 fractions, while one received 25 fractions, totaling 109 fractions. During each session, an adapted treatment plan (initADP) was generated based on TVs initially propagated by Ethos and applied Boolean operations. If modified by the physician, the initADP was discarded, and plan optimization restarted with altered TV contours, resulting in a final adapted plan (finADP). The finADP was used 77 times, while the initADP was retained 32 times. After treatment, both initADP and finADP for each patient and fraction were imported into Eclipse V18.0 and recalculated on the first CBCT of each session, using reviewed influencer structures and physician-modified TVs. The evaluation assessed the dosimetric percentage difference between finADP and initADP, calculated as (finADP-initADP)/finADP, in TVs for D98% and D2%, and also analyzed V15Gy and mean values for rectum and bladder, plus D1ccm for bowel.

Results: