S2117
Physics - Inter-fraction motion management and daily adaptive radiotherapy
ESTRO 2026
Proffered Paper 1032 Estimating adaptation rates depending on specified trigger conditions in dose evaluation- based online adaptive proton therapy Thu T.C. Nguyen 1,2 , Lena Nenoff 3,4 , Annabell Eberhardt 5 , Rebecca Bütof 3,5 , Esther G.C. Troost 3,5 , Christian Richter 3,5 , Kristin Stützer 3,4 1 Center for Advanced Systems Understanding, CASUS, Görlitz, Germany. 2 Helmholtz-Zentrum Dresden- Rossendorf, CASUS - Center for Advanced Systems Understanding, Dresden, Germany. 3 OncoRay – National Center for Radiation Research in Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus, TUD Dresden University of Technology, Helmholtz-Zentrum Dresden-Rossendorf, Dresden, Germany. 4 Helmholtz-Zentrum Dresden-Rossendorf, Institute of Radiooncology – OncoRay, Dresden, Germany. 5 Department of Radiotherapy and Radiation Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus, TUD Dresden University of Technology, Dresden, Germany Purpose/Objective: Rather than scheduling online adaptive proton therapy (OAPT) in advance, it may only be performed when required to avoid unnecessary work. We systematically investigated the influence of predefined dose-volume histogram (DVH)-based OAPT trigger conditions on adaptation rates for estimating OAPT-related work before implementation. Material/Methods: Data from ten non-small-cell lung cancer (NSCLC) proton therapy patients included 52 weekly repeat CT images (rCTs) with clinical target volumes (CTV) and organs of interest contours. Robust intensity- modulated proton therapy plans were optimized in RayStation on the planning CT and rCTs by patient- specific template-based adaptive re-optimization. All plans were calculated on their subsequent rCTs, if any, and clinical goal-related DVH parameters (Table1) were extracted.We emulated the decisions and respective adaptive treatment courses for various OAPT trigger conditions, and determined the adaptation rates across all rCTs. Three DVH-based trigger condition types were considered (Figure1): (1) variable value threshold for the DVH parameter, (2) variable maximal degradation threshold relative to the optimized plan, and (3) the union of both. For each DVH parameter, adaptation rates (1)-(3) were determined for value thresholds equal to the clinical goal and degradation thresholds equal to 10% of the clinical goal value (organs) or 3% (CTV), respectively. Further investigated union trigger conditions included combinations of several DVH parameters, where either of them could indicate OAPT necessity.
Results: Estimated adaptation rates (Table1) revealed that D98%(CTV), Dmax(spinal cord) and V20Gy(lungs-GTV) are most likely to trigger adaptations in up to 35%, 25% and 17% of rCTs, respectively, when considering the union trigger condition. Combining all DVH parameters for triggering OAPT results in an adaptation rate of just 62% due to flagging of same rCTs by multiple DVH parameters. The same rate was reached when considering only D98%(CTV), Dmax(spinal cord), V20Gy(lungs-GTV), and Dmean(heart). Only 12% of the rCTs were flagged solely by D98%(CTV).
Conclusion: A first systematic estimation of adaptation rates in an evaluation-based OAPT workflow was performed for variable DVH-based trigger conditions in NSCLC. Multiple DVH parameters should be considered, but as several of them are interdependent, only a few key parameters require special attention when working under time pressure in an online setting. The determined adaptation rates per rCT will be lower in case of daily instead of weekly imaging. The presented approach is applicable to other patient cohorts and even other adaptive radiotherapy modalities for selecting reasonable DVH-based trigger conditions and estimating expectable OAPT-related work. Keywords: dose-triggered OAPT, proton therapy, lung cancer
Made with FlippingBook - Share PDF online