ESTRO 2025 - Abstract Book

S1017

Clinical – Head & neck

ESTRO 2025

2285

Digital Poster The effect of evolving clinical practice in head and neck radiotherapy: auto-planning and new CTV guidelines Eva Onjukka 1,2 , Jeehong Lee 1 , Emmy Dalqvist 1 , Erik Lampa 3 , Ingmar Lax 4,2 , Claes Mercke 4,2 , Signe Friesland 4,2 , Anna Embring 5,2 1 Department of Nuclear Medicine and Medical Physics, Karolinska University Hospital, Stockholm, Sweden. 2 Department of Oncology Pathology, Karolinska Institutet, Stockholm, Sweden. 3 -, Epistat AB, Uppsala, Sweden. 4 Department of Head, Neck, Lung and Skin Tumors, Karolinska University Hospital, Stockholm, Sweden. 5 Department of Radiotherapy, Karolinska University Hospital, Stockholm, Sweden Purpose/Objective: We used a local quality registry for radiotherapy of head and neck cancer (HNC) to evaluate how the risk of late side effects has changed over time, specifically after the introduction of knowledge-based auto-planning and after the implementation of international guidelines for CTV definition [1]. We considered recorded outcome and planned dose for ten organs-at-risk (OAR). Material/Methods: At follow-up, side effects of patients treated for HNC are recorded in a local quality registry. Late side effects are evaluated using a modified RTOG scale. Patients treated in the period of 2017-2022 for oropharyngeal- and oral cavity cancers were considered for this study, excluding patients with treatment adaption and re-treated patients. This period was chosen to capture the introduction of knowledge-based auto-planning around June 2019, and the implementation of the new CTV definition around May 2020. As a varying number of repeat follow-up scores had been recorded in the registry, with varying follow-up time, the score closest in time to 2 years post radiotherapy was selected for this analysis. The mean dose (Dmean) and near maximum dose (D2%) were extracted for each OAR. The change in OAR dose and risk of side effects were tested using multivariate ANOVA and a multivariate rank based test, respectively, both adjusted for multiple testing. Results: In total, 503 patients were included in the analysis, distributed over: 1) January 2017 – May 2019, 2) June 2019 – April 2020 and 3) May 2020 – Dec 2022. Figure 1A shows the difference in Dmean comparing each of the time periods: 2 1 (introduction of auto-planning), 3-1 (the new CTV definition and auto-planning compared to before either change), and 3-2 (the implementation of the new CTV, exclusively). The dose was significantly reduced in 2-1 and 3-1 for the ipsilateral parotid gland. The brainstem had a reduced D2%, but this may be a result of missing delineations in cases with low dose in period 1. While mean values did not change significantly for the oral cavity, a scatter plot of D2% revealed a benefit for some patients from the new CTV definition (Figure 2). No statistically significant effect on the risk of toxicity was observed in this analysis (Figure 1B).