ESTRO 2025 - Abstract Book

S3272

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2025

treatment CBCT images was performed by a second senior oncologist, not part of the online treatments, in a gold standard setting i.e. with all diagnostic information available and without time-pressure from the patient laying on the treatment table. These CTVs were compared to the CTV used at adaptation (ART-CTV). Intrafraction motion for the CTV in LARC patients was investigated in a separate cohort with pre- and post-fraction CBCTs from 15 prospectively recruited LARC patients, a total 146 CBCTs. Both delineation uncertainties and intrafraction motion were assessed by dice similarity coefficient (DSC) and Hausdorff distance (HD) between the sets of CTVs in addition to geometric margin expansion. Results: In the five first patients included in the TNT-RECORD study, five planning CTs and 25 CBCTs were evaluated. The adapted plan was chosen at all fractions for each patient. The CTVs defined in the planning CT were prone to less delineation uncertainty with median (range) DSC of 0.96 (range: 0.93-0.99) and 95 th -percentile HD of 5.5 (0.8 5.8) mm, as compared to delineation uncertainty for the ART-CTVs with median DSC of 0.90 (0.86-0.92) and 95 th percentile HD of 8.0 (range 5.3-10.1). Intrafraction motion management resulted in significantly less difference between contours as compared to delineation uncertainties on ART-CBCTs, with a median DSC and 95 th percentile HD of 0.94 (0.87-0.99) (p<0.0001) and 4.2 (1.1-18.5) mm (p<0.0001) (Fig.2). A geometric margin of up to 6 mm was required for the online ART-CTVs to cover 99% of the re-contoured CTV in more than 90% of the scans, whereas smaller margins of 4 mm were required to compensate for intrafraction motion.

Conclusion: Delineation uncertainties in online CBCT-based ART were more pronounced than the effect of intrafraction motion for the CTV in short-course RT for LARC patients. Keywords: Ethos, inter-observer variability, short-course RT