ESTRO 2025 - Abstract Book

S3210

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

ESTRO 2025

1 Institute for Robotics and Cognitive Systems, University of Luebeck, Luebeck, Germany. 2 Department of Radiation Oncology, Heidelberg University, Mannheim, Germany. 3 DKFZ Hector Cancer Institute, University Medical Center Mannheim, Mannheim, Germany. 4 Department of Radiation Oncology, University Medical Center Schleswig Holstein, Kiel, Germany. 5 Department of Radiology, Heidelberg University, Mannheim, Germany. 6 Department of Internal Medicine III (cardiology, angiology, and internal intensive care medicine), University Medical Center Schleswig-Holstein, Kiel, Germany. 7 Department of Radiology and Neuroradiology, University Medical Center Schleswig-Holstein, Kiel, Germany. 8 Hannover Heart Rhythm Center, Hannover Medical School, Hannover, Germany. 9 Cardiology & Internal Intensive Care Medicine, St. Bernward Hospital, Hildesheim, Germany. 10 Department of Radiotherapy, Rambam Health Care Campus, Haifa, Israel. 11 Department of Rhythmology, University Hospital Schleswig-Holstein, Luebeck, Germany. 12 German Center for Cardiovascular Research (DZHK), Partner Site Hamburg/Kiel/Luebeck, Luebeck, Germany. 13 Department of Radiation Oncology, University of Luebeck, Luebeck, Germany. 14 Department of Internal Medicine I, Section for Electrophysiology and Rhythmology,, Heidelberg University, Mannheim, Germany. 15 German Center for Cardiovascular Research (DZHK), Partner Site Heidelberg, Mannheim, Germany. 16 Department of Radiotherapy, Hannover Medical School, Hannover, Germany. 17 Division of Pacing and Electrophysiology, Rambam Health Care Campus, Haifa, Israel Purpose/Objective: Motion management strategies such as gating under breath-hold can reduce breathing-induced motion during STAR for refractory ventricular tachycardia (VT). However, information on heartbeat-induced motion is essential to define an appropriate cardiac internal target volume (ITV) margin. In this study, we introduce a patient- and segment-specific cardiac motion estimation method and cardiac motion data of the clinical target volume (CTV), implantable cardioverter-defibrillator (ICD) lead tips and the American Heart Association (AHA) 17 left ventricle (LV) segments in relation. Material/Methods: Data for 10 patients treated with STAR within the harmonized RAVENTA framework [1] was retrospectively analyzed. The LV was semi-automatically segmented according to the American Heart Association 17-segment model using the in-house software CARDIO-RT [2], and the CTV was predefined. ECG-gated contrast-enhanced breath-hold cardiac CT at end-diastole was automatically non-rigidly registered to end-systole based on voxel intensity, resulting in a 4D displacement field which was then applied to the cardiac substructures for motion estimation. ICD lead tip was segmented, and its motion was evaluated against CTV and segment motion using the Wilcoxon signed rank test and Euclidean distance. Results: The CTVs (center of mass) moved 3.4 ± 1.4 mm in 3D (see Fig. 1 A) and the ICD lead tips moved 4.4 ± 2.6 mm in 3D. For these 10 cases, the maximum motion per patient was observed in basal (segments 1, 3, 5 and 6) and mid-cavity (segments 8, 9 and 10) regions (see Fig. 1 B). The ICD lead tip motion was statistically significantly different from the CTV motion (p = 0.005) with a Euclidean distance of 7.4 mm. The Euclidean distance between ICD lead tip motion and 17 segments motion is shown on the 17-segment bull's-eye map (see Fig. 2). The average Euclidean distance between the ICD lead tip motion and 17-segment motion was 9.1 ± 2.6 mm. The smallest Euclidean distance was observed in segment 3, while the largest Euclidean distance occurred in segment 17. In these 10 patients, the Euclidean distances were generally axisymmetric on the 17-segment model bull's-eye map and decreased from the apical segments to the mid-cavity and basal segments. A lower Euclidean distance between ICD lead tip and segment motion indicates closer alignment of motion patterns.