ESTRO 2025 - Abstract Book

S2933

Physics - Dose prediction, optimisation and applications of photon and electron planning

ESTRO 2025

4384

Digital Poster Human Verses Machine (HuVeM): dosimetric analysis of clinical decision to replan Head and Neck patients on treatment Hannah Price 1 , Priya Narga-Martin 1 , Kevin Chiu 2 1 Radiotherapy Physics, Mount Vernon Cancer Centre, London, United Kingdom. 2 Clinical Oncology, Mount Vernon Cancer Centre, London, United Kingdom Purpose/Objective: H&N patients often encounter set-up difficulties during treatment. This can have a negative dosimetric impact if not addressed. Multi-disciplinary meetings are arranged weekly to discuss patients encountering these issues. These patients were recruited to HuVeM (NCT06035653), and were assigned to Cohort A or B depending on outcome: replan or not replan respectively. This study aims to retrospectively analyse the decision to replan patients partially through treatment, and to assess time saving capabilities for clinical volumes using DIR. Material/Methods: Weekly CBCT recalculations were performed on 12 patients from cohort A, totalling 6 dose calculations each representing 5 fractions. Two scenarios were calculated: firstly, where the original plan was calculated initially, then the remainder on the replan. Secondly, the original plan calculated on every CBCT: to simulate a ‘no replan’ decision. VelocityTM v4.1 (Varian Medical Systems) “Accumulation: ACTIVE Dosimetry” was used to accumulate dose using deformable registration of CBCTs to original planning CT. Technique included RIR focussed on oral cavity, then CBCT Corrected Multi Pass Deformable registration to entire image. ROI reduction in DIR was implemented on a case by case basis. Difficulties arose with large superficial contour change: as such, nodal CTVs were excluded in this study. DIR between plan CT and replan CT was performed using Deformable Multi Pass registration to deform planning CTVs onto replan CT. Dice Similarity Coefficient (DSC) was calculated to analyse conformity of CTVs. Results: 4 patients were excluded due to unsuitable DIR solutions for dose accumulation. V65Gy and V61.75Gy were calculated for CTVp6500, and V54Gy and V65Gy for CTVp5400 for 8 patients for the treatment plan and the two aforementioned scenarios. It was anticipated that the replanned dose would have improved dose distribution compared to the no replan dose, however this was not always the case (see figure 1). Ultimately there was no statistical significance found (p<0.05) between any metrics and therefore, no significant change in dose between the two scenarios.