ESTRO 2025 - Abstract Book

S1325

Clinical - Lung

ESTRO 2025

1586

Digital Poster Development of a Breathing Score for Risk Assessment of Local Failure in Lung and Liver SBRT Hermann Krümpelmann 1 , Cordula Petersen 1 , Laura Büttgen 1 , Lukas Wimmert 1 , René Werner 2 , Tobias Gauer 1 1 Radiotherapy and Radiooncology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. 2 Center for Biomedical Artificial Intelligence, Applied Health Informatics and Digital Medicine, Hamburg, Germany Purpose/Objective: Respiratory pattern variations can impact stereotactic body radiotherapy (SBRT) outcomes for lung and liver metastases. This study aimed to develop a breathing score for local failure risk based on respiratory patterns to guide 4D treatment planning/delivery adjustments of safety margins, target repositioning, and gated irradiation. Material/Methods: We retrospectively analyzed clinical, dosimetric, and respiratory data from 147 patients treated with SBRT for lung and/or liver metastases, in total 164 lesions (118 lung and 46 liver), with a mean follow-up of 16 months. Lesions were treated with larger than 95% BED surrounding the ITV in 5 to 10 fractions. For each patient, up to six breathing curves were recorded: one planning CT and five dose fraction curves. Over 140 statistical metrics were extracted from the breathing curves, including intersession variability and global variables. In a training and test data approach with 82 randomly assigned lesions each, significant and independent variables for local control were determined and correlation matrices and principal-component-analysis were used. A breathing score was calculated based on 4 variables by z-score standardization, weighting by effect size, and summation. The breathing score was calculated using the first three breathing curves, allowing potential treatment modifications in later sessions. Factors associated with local failure were analyzed using Cox regression models. The multivariate analysis included the breathing score group, fraction dose, number of fractions, Karnofsky Performance Status, age at irradiation, lesion location and chemotherapy before, during and after SBRT. Results: Local failure occurred in 30 out of 164 metastases (18%). In multivariable Cox regression, the breathing score group was independently associated with local failure as seen in Table 1. High-risk patients exhibited specific breathing patterns: a decrease in breathing frequency between sessions, less consistency in breathing cycles, greater irregularity and unpredictability in their breathing patterns, and occasional extreme variations in breath depths. Conversely, low-risk patients demonstrated an increase in breathing frequency between sessions, more consistent breathing cycles, more regular breathing patterns, and consistent breath depths.