S2148
Physics - Inter-fraction motion management and daily adaptive radiotherapy
ESTRO 2026
Conclusion: MRgART achieved superior target coverage and dose accuracy compared to rigid IGRT in liver SBRT, supporting its clinical value—especially for patients with multiple intrahepatic targets. Keywords: dose accumulation,adaptive radiotherapy
adaptive radiotherapy (MRgART) and conventional IGRT for liver SBRT assessing potential gains in target dose accuracy, tumor control probability (TCP), and normal tissue sparing (NTCP). Material/Methods: This retrospective study included fourteen patients with hepatocellular carcinoma treated on an MR-Linac with a total prescription dose of 40 Gy in five fractions to the PTV. The clinical plan guided by the MR (TPMRg), was adapted daily according to the adapt-to-shape (ATS) workflow, in which target and OARs contours were reviewed and the treatment plan re-optimized before each fraction based on daily anatomical changes, using the synthetic CT (sCT)generated from the corresponding MR image. A second, conventional IGRT plan (TPIGRT) was simulated on the same sCYs by applying rigid position registration only and by recalculating the clinical plan at the shifted isocenter. Fractional doses from both TPMRg and TPIGRT were accumulated for analysis. To evaluate the dosimetric and biological benefits of daily adaptation, dosimetric parameters PTV Dmean, V100%, and CI; liver Dmean and D700cc; and radiobiological indices (TCP, NTCP) were compared between TPMRg and TPIGRT. Statistical comparisons were performed using paired t- tests. Results: A total of 70 fractions were analyzed. All ART fractions achieved V100% > 95%, whereas 8/70 IGRT fractions failed. The mean target V100% was 94.07% for ART and 83.71% for IGRT; after dose accumulation, PTV V100% was 86.25% vs 78.98%. The conformity index (CI) for was 0.85 (0.76–0.90) in the ART group and 0.73 (0.54–0.88) in the IGRT group, while TCP values were 0.9704 (0.9576–0.9769) and 0.9497 (0.8890–0.9747), respectively. Patients with multiple intrahepatic targets showed rigid registration errors causing decreased CI, PTV coverage. Significant differences were observed between ART and IGRT in PTV CI, V100%, whereas no significant differences were found in liver dosimetry. TCP was significantly higher in the ART group (p < 0.05), with all cases exceeding the 95% clinical target, while NTCP remained below the 5% safety threshold in both groups, with no significant difference (p > 0.05)(Fig. 1).
Digital Poster 3043
democratizing adaptive radiotherapy: an automated workflow for CBCT-based dose assessment on conventional c-arm linacs Hoi Hin Lau 1,2 , Claudia Katharina Renkamp 1 , Sebastian Klüter 1 , David Neugebauer 1 , Rachel Hanley 1 , Laila König 1 , Vania Batista 1 1 Radiation Oncology, Heidelberg University Hospital, Heidelberg, Germany. 2 Medical Faculty Heidelberg, Heidelberg University, Heidelberg, Germany Purpose/Objective: Adaptive radiotherapy (ART) is typically implemented using integrated systems that provide streamlined workflows. However, dedicated machines for online ART (OART) are expensive and often inaccessible for many clinics. To enable wider adoption of ART on standard linacs, this study assessed the feasibility of an automated workflow for CBCT-based ART, using CBCT images from conventional C-arm linac and a standard treatment planning system (TPS). Material/Methods: CBCT images were acquired during registration on Elekta Synergy/Versa HD linacs. Corrected CBCT (cCBCT) algorithm available in TPS RayStation (Version 2024B, RaySearch) converted these scans into CT-like images. Validation and feasibility were evaluated using ten retrospective thoracic patients, each undergoing at least one re-planning during treatment. Validation focused on three aspects: (a) cCBCT image quality, (b) feasibility of automated structure propagation and segmentation, and (c) accuracy of dose calculation on cCBCT, with rescanned CTs serving as approximate gold standards. For the feasibility investigation, a standardised workflow was established using in-house developed scripts. Volumetric and geometric changes in the target volume and organs-of-interest between the baseline CT and cCBCT were analyzed. The baseline plan was then recalculated on the cCBCT, and dosimetric parameters of the CTV and organs-of- interest were collected. Any manual corrections and workflow errors were documented and analyzed. Results: cCBCT Hounsfield unit (HU) values showed good agreement with planning CT (Figure 1a). In the heart, the median HU difference was 2.83%, and noise (standard deviation of HU) increased from 27.5 HU on reference CT to 50.6 HU on cCBCT, indicating the need
Figure 1. Radiobiological parameters comparing MRgART and IGRT for liver SBRT.
Made with FlippingBook - Share PDF online