ESTRO 2025 - Abstract Book

S3092

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

Material/Methods: Ten patients with primary cancer of the distal esophagus or gastroesophageal junction who received neoadjuvant CRT (41.4Gy in 23 fractions) were retrospectively included. For each patient, an oART workflow was simulated based on their daily CBCT scans. Herein new target and organ-at-risk (OAR) contours were created by the oART software and adapted by a clinician. The PTV margins in the adapted plans compared to non-adaptive plans were reduced from 5;7;10mm to 3;5;5mm (AP;LR;CC). The adapted contours were subsequently registered to the planning CT and associated treatment plan to calculate the clinically delivered (non-adaptive) dose. Fraction-equivalent (FE) dosimetric values for the target and OARs were compared between the planned, non-adaptive, and adaptive plans. Additionally, the proportion of fractions in which the non-adaptive and adaptive plans violated the target coverage goal (V95% < 98%) was determined. Statistical significance was determined with Wilcoxon signed-rank tests ( p <0.05). Results: The simulated oART workflow was feasible in all 230 fractions. Compared to delivered doses, the adaptive plan significantly decreased target coverage violations (13.9% vs 0.4% of fractions; p=0.002) and median hotspot dose (p=0.002), despite reduced PTV margins. The adaptive plans significantly reduced the median mean heart dose by 11% ( p =0.037), and the median heart V30Gy by 32% (FE 4.3% vs. 7.3%; p =0.002). Regarding the lungs, similar significant improvements were observed between adaptive and delivered plans, with a 12% reduction in median mean lung dose ( p =0.002), and a 39% and 22% average reduction in median lungs V20Gy and V10Gy (2.6% vs. 4.2%; p=0.002, and 19.3% vs. 24.6%; p =0.002), respectively.