ESTRO 2025 - Abstract Book

S3284

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

ESTRO 2025

Material/Methods: We consider simulated PBS treatments of lung cancer optimized and evaluated using four-dimensional dose computations (4DDCs), considering the interplay effect by explicitly modeling the time structure of the delivered plans. A dataset of synthetic 4DCTs, generated from multiple unique breathing cycles acquired with 4DMRI [2], is used to create patient motion scenarios partitioned into an optimization set and an evaluation set, respectively. The optimization set is used to optimize interplay-robust treatment plans (IPRO) [3] with varying degrees of robustness, increasing with the size and variability of the set of scenarios. The evaluation set is then used to evaluate each treatment plan based on the variability of dosimetric criteria. Results: IPRO plans consistently improved the near-worst-case target coverage (5 th percentile CTV D98) compared to 4DRO, typically without substantial changes to CTV dose homogeneity or OAR dose. For example, in each method's evaluation scenario with the near-worst target coverage, IPRO showed better dose conformality to the CTV and increased sparing of the proximal OAR (Figure). When each plan was normalized by scaling the doses to match 4DRO in target coverage, it was found that the IPRO plan with the more realistic representation of the breathing variability, using 25 scenarios (IPRO-25S), decreased the dose-limiting objectives by an average of 3.5% compared to 4DRO (Table). In contrast, the IPRO plan with the least complex representation, with nine scenarios generated by variations in the frequency and start phase of a single breathing cycle (IPRO-1C-9S), led to a mean decrease of 1.4%.