ESTRO 2025 - Abstract Book

S3147

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

Results: the results of the first 840 treatment fractions in 42 patients are reported. The median age was 64 years (range 33 78). prostate bed PTV dose prescription (Dp) >95% was 26.5% and 98.8% for scheduled and adapted plans, respectively. The 107% Dp was <2% in 78.5% and 82.1% of scheduled and adapted plans respectively. Concerning the rectum the Dmean <38Gy was maintained in 60% of the scheduled and 85% of the adapted plan. The bladder V40.8Gy was <55% in 91% of scheduled fractions and 93% in the adapted fractions. Pelvic lymph nodes were treated in 17 patients with a PTV pelvis coverage >95% in 67% of the scheduled fractions and 92% in the adapted fractions. In scheduled and adapted plans, the bowel Dmax <50Gy was 74% and 80%, respectively. Conclusion: adapted plans showed higher PTV coverage and better OARs sparing. These dosimetric parameters will be matched with clinical data to confirm the safety of adaptive sRT in PC.

Keywords: adaptive radiotherapy, prostate cancer

2633

Proffered Paper Quantitative and automatic plan-of-the-day assessment to facilitate adaptive radiotherapy in cervical cancer Sarah A Mason 1,2 , Lei Wang 1,2 , Sophie E Alexander 1,3 , Susan Lalondrelle 1,2 , Helen McNair 1,2 , Emma J Harris 1,2 1 Joint Department of Physics, Institute of Cancer Research, London, United Kingdom. 2 Radiotherapy and Imaging, Royal Marsden NHS Foundation Trust, Sutton, United Kingdom. 3 Radiotherapy and Imaging, Royal Marsden NHS Foundation Trust, London, United Kingdom Purpose/Objective: To improve ease of implementation and workflow efficiency when utilising plan-of-the-day (POTD) selection for treating locally advanced cervical cancer (LACC), we developed a segmentation-based POTD assessment tool for CBCT-guided radiotherapy (RT). This tool overcomes several technical and resource barriers to widespread adoption of POTD selection 1 by: (1) incorporating patient-specific prior knowledge to improve segmentation accuracy within an nnUNet 2 , (2) automatically selecting the optimal plan using a novel quantitative standard operating procedure (qSOP) developed in this work, and (3) providing transparent and human-interpretable results. Material/Methods: The planning CT[i], corresponding structure set[ii], and manually contoured CBCTs[iii] (n=226) from 39 LACC patients treated with either POTD (n=11) or non-adaptive RT (n=28) were used to develop U-Seg3, an algorithm incorporating deep learning 2 and deformable image registration techniques to segment the low-risk clinical target volume (LR-CTV), high-risk CTV (HR-CTV), bladder, rectum, and bowel bag. A single-channel input model (iii only, U Seg1) 3 was also developed as a benchmark to see whether the inclusion of patient-specific information in U-Seg3 provided significant benefit. Contoured CBCTs from the POTD patients were (a) reserved for U-Seg3 validation/testing, (b) audited to determine optimal and acceptable plans, and (c) used to empirically derived a qSOP (structure and thresholds) that maximised classification accuracy. Results: The median [interquartile range (IQR)] DSC between manual and U-Seg3 contours was 0.83 [0.80], 0.78 [0.13], 0.94 [0.05], 0.86[0.09], and 0.90 [0.05] for the LR-CTV, HR-CTV, bladder, rectum, and bowel. These were significantly higher than U-Seg1 DSCs in all structures but bladder.