ESTRO 2025 - Abstract Book

S3134

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2025

2367

Poster Discussion Decision support for daily adaptive prostate radiotherapy David Tilly 1,2 , Samuel Fransson 3 , Martin Lundmark 2 , Adam Johansson 1,2 , Ulf Isacsson 1,2 , Calin Radu 1 , Anna AK Flejmer 1 , Ariadni Kyriakogiannaki 4 , Nina Tilly 1,2 1 Department of Immunology, Genetics and Pathology, Uppsala University, Uppsala, Sweden. 2 Department of Medical Physics, Uppsala University Hospital, Uppsala, Sweden. 3 Department of Surgical Sciences, Uppsala University, Uppsala, Sweden. 4 Department of Oncology, Uppsala University Hospital, Uppsala, Sweden Purpose/Objective: The MR-Linac and its online adaptive radiotherapy workflow can enable a reduction of the PTV margin for prostate radiotherapy. However, the reduced margin comes at the cost of increased risk of an additional replan due to prostate movement during the online treatment planning. The aim of the study was to provide decision support for treatment with a reduced margin that ensures sufficient target coverage, is simple to use, and does not significantly prolong the treatments. Material/Methods: Images are acquired during MR-Linac fractions for planning I PLAN , verification I VER (before beam on) and post irradiation I POST . The CTV (prostate only) displacement between I PLAN and I VER is denoted as d PLAN_VER (and analogous d VER_POST ). A preliminary analysis suggested that a small d PLAN_VER correlated with a small displacement during irradiation ( d VER_POST ). Thus, d PLAN_VER could serve as a decision variable for when an additional replan is needed. An intra-fraction motion analysis was performed on 200 delivered fractions from 38 prostate cancer patients. The CTV motion d PLAN_VER and d VER_POST was determined by propagating the CTV from I PLAN to I VER and I POST using the Elastix [1] deformable image registration. An expert reviewed and adjusted the contours. The displacements were calculated as the difference in the centre of mass position. The threshold d TH on d PLAN_VER and a reduced PTV margin were chosen as: 1. IF d PLAN_VER ≤ d TH , THEN CTV inside PTV during beam on > 90% of the fractions. 2. IF d PLAN_VER > d TH , THEN CTV inside PTV > 90% of the fractions AFTER replan. 3. No more than 10% of the fractions should require an additional replan. 4. Easy to use in an online setting. We evaluated these criteria by varying the d TH and PTV margin and calculate the probability that the CTV was within the PTV. Results: The choice of d TH =(2, 3, 2) mm (L-R, A-P, S-I) fulfils the 10% risk of replan (criteria 3), see Figure 1. Given this d TH the PTV margin becomes (1.5, 3, 1.5) mm to fulfil the CTV inside PTV (criteria 1), see Figure 2. To also fulfil criteria 2 (results not shown) and avoid having a different PTV margin after replan (criteria 4), the PTV margin becomes (2, 3, 2) mm.