S2913
RTT- RTT operational practice and workflow innovations
ESTRO 2026
Assessment of margin robustness in online adaptive bladder radiotherapy Pauline Hinault, Jessica Prunaretty, Dorian Trauchessec, Aurelie Morel, Nicolas Bizot, Morgan Michalet, Olivier Riou, David Azria, Norbert Ailleres Radiotherapy, Cancer Institute, Montpellier, France Purpose/Objective: Bladder radiotherapy is challenged by variations in bladder size, shape, and position during and between sessions (Deers-Ribers, 2014), which may cause target underdosing and/or organs at risk (OARs) overdosing (Astrom, 2022). At Montpellier Cancer Institute, the Ethos system (Varian) uses artificial intelligence (AI) for real-time online adaptive radiotherapy (oART).This study aimed to assess the benefits of this adaptive workflow and the robustness of the selected margins. Material/Methods: This retrospective study included 10 bladder cancer patients treated with 20 fractions of 2.75 Gy adaptive radiotherapy. PTVs, for the bladder, were defined with asymmetric margins: 8 mm in anterior, posterior and superior directions and 5 mm in left, right and inferior directions. 5 mm was used for lymph nodes. At each session, targets and OARs were automatically generated on the daily CBCT (CBCT1) using Ethos AI and reviewed by the radiation oncologist. A synthetic CT (sCT) was generated by deforming the planning CT to match CBCT1 anatomy. Two plans were computed on the sCT: the scheduled plan, recalculated from the initial plan, and the adaptive plan, optimized using updated contours. A final CBCT (CBCT2) was acquired before delivery to confirm anatomy.For each session, bladder and OARs were redefined on CBCT2, a delivered plan was obtained by recalculating the adaptive plan on this dataset and then compared with the scheduled and adaptive plans. Dosimetric parameters were analyzed: V95% for PTV55, Dmean for the rectum, V25.8 Gy and V45.8 Gy for the bowel. Results: In the analysis of all delivered plans (Figure 1), the median V95% for PTV55 was 89.7%, 96.2%, and 93.4% for the planned, adapted, and delivered plans, respectively. The dosimetric evaluation showed improved target coverage with the adapted and delivered plans compared with the initial scheduled plans. For OARs, similar results were observed across all plan types.Adapted plans were then stratified according to the bladder volume ratio: session bladder volume (Vol_S) divided by planning bladder volume (Vol_P) (Figure 2). The median mean rectal dose was higher when the bladder volume during treatment was smaller than at simulation, 15.1 Gy vs. 11 Gy when larger. For the bowel, dose–volume values increased
Conclusion: Automated CBCT-based dose calculations provide efficient daily and accumulated dose assessments, showing robust target coverage but organ of interest overdosages. Plan adaptation improved interfraction dose stability, confirming our adaptive strategy and supporting CBCT-based delivered dose monitoring as a practical tool for data-driven decision-making in offline adaptive radiotherapy. References: [1] Thing RS, et al. Evaluation of CBCT based dose calculation in the thorax… Physica Medica 2022;103:157–65. doi.org/10.1016/j.ejmp.2022.10.012.[2] Thomsen SN, et al. Daily CBCT-based dose calculations for enhancing the safety of dose-escalation in lung cancer radiotherapy. Radiotherapy and Oncology 2024;200:110506. doi.org/10.1016/j.radonc.2024.110506.[3] Callens D, et al. Is full-automation in radiotherapy treatment planning ready for take off? Radiotherapy and Oncology 2024;201:110546. doi.org/10.1016/J.RADONC.2024.110546.[4] Parker W, et al. ICRU Report 50, ICRU; 1994.[5] Bradley JD, et al. RTOG 0617: a randomised, two-by- two factorial … Lancet Oncol 2015;16:187–99.
doi.org/10.1016/S1470-2045(14)71207-0. Keywords: CBCT dose calculations, adaptive radiotherapy
Digital Poster Highlight 1921
Made with FlippingBook - Share PDF online