ESTRO 36 Abstract Book

S564 ESTRO 36 2017 _______________________________________________________________________________________________

Severity Index (DSI). Rosenbeks 8-points-penetration- aspiration-scale (PAS, Rosenbek 1996) was used to determine the severity. The Functional Oral Intake Scale (FOIS) classified by Crary (Crary et al., 2005) was used to assess the oral food intake of the patients. Results A paired t-test showed a significant change in volume of five OARs: superior and middle pharyngeal constrictor muscle, cricopharyngeal muscle, proximal esophagus and transglottic larynx (p≤0.055). There was a significant increase in volume in four OARs and a significant decrease in the proximal esophagus. The linear regression analysis of the volume changes of the respective OARs and the applied dose showed no significant correlation. The binary logistic regression showed a significantly (p=0.015) 1.5 times higher risk to suffer from dysphagia when the dose is increased by steps of 1 Gy. No association could be found for dose-dependent dysphonia. The results of the FEES showed conspicuous PAS levels in all five subjects and the FOIS scale reached grade 3 to grade 6. Also, dysphonia ranged from a mild to a high degree. The evaluation of the ADI-D questionnaire correlated with the degree "rather conspicuous" in the Bauer and Rosanowski Scale. The Voice Handicap Index showed one patient with a moderate and four with no subjective voice disorder. Apart from problems caused by xerostomia only minor discomfort in respect of tumor-specific symptoms was observed based on the EORTC QLQ-H&N35 questionnaire. Conclusion In this study, a tendency of increase in the volume of OARs under RT with distinct clinical symptoms (dysphagia and dysphonia) was detected, also after more than one year post-RT. MRI use for RT is essential for an optimal protection of the OARs in terms of adaptive radiotherapy (ART) and is expected to improve treatment as MR-guided radiotherapy (MRgRT) . EP-1029 Improved interobserver reproducibility in nasopharyngeal tumor delineation using a reference GTV A. Lapierre 1 , F. Lorchel 1 , P. Boisselier 2 , B. Fleury 3 , F. Craighero 4 , M. Deberne 1 , G. Gascou 5 , M. Rousset 6 , V. Favrel 1 , J.B. Pialat 7 , J.M. Ardiet 1 , P. Céruse 8 , O. Chapet 1 1 Centre Hospitalier Lyon Sud, Service de Radiothérapie, Pierre Benite, France 2 CRLC Val d'Aurelle, Service de Radiothérapie, Montpellier, France 3 Centre Marie Curie, Service de Radiothérapie, Valence, France 4 Centre Hospitalier Universitaire Croix Rousse, Service de Radiologie, Lyon, France 5 Centre Hospitalier Universitaire Saint Eloi, Service de Radiologie, Montpellier, France 6 Centre Hospitalier Lyon Sud, Service de Radiologie, Pierre Benite, France 7 Centre Hospitalier Universitaire Edouard Herriot, Service de Radiologie, Lyon, France 8 Centre Hospitalier Universitaire Croix Rousse, Service de chirurgie ORL, Lyon, France Purpose or Objective Standard treatment for nasopharyngeal cancer (NPC) relies on concurrent chemoradiotherapy, using intensity modulated radiotherapy (IMRT). Adequate tumor volume definition is essential for precision radiotherapy such as IMRT. However, tumor volume delineation reproducibility on CT scans has been shown to be variable among radiation oncologists. The main goal of this study was to assess the reproducibility of radiologists in defining nasopharyngeal tumor volumes, and whether a common gross tumor volume (GTV) delineated by specialized radiologists would improve reproducibility between radiation oncologists in defining high risk tumor clinical target volumes (CTV) in NPC cases. Material and Methods

Ten patients treated at our institution over the last 5 years for nasopharyngeal tumors were selected for the study. In the first part of the study, five experienced radiation oncologists were asked to independently delineate tumor GTV and high risk tumor CTV. Meanwhile, three radiologists independently delineated the tumor GTV on the diagnosis MRI, after which, the radiation oncologist were asked to delineate the high risk tumor CTV again, using the reference GTV delineated by the radiologists. Results The intraclass correlation coefficient (ICC) for GTV delineation was 0.914 for radiologists, and 0.754 for radiation oncologists. Use of a common GTV increased the generalized conformity index (CI gen ) from 0.44 to 0.49 for CTV delineation by the radiation oncologists.

Conclusion Interobserver variations tend to be lower between radiologists than radiation oncologists in the definition of NPC tumor volumes, and that use a single reference GTV improves reproducibility between radiation oncologists. EP-1030 External beam radiation therapy for locoregionally recurrent differentiated thyroid carcinoma C.Y. Kim 1 , N.K. Lee 1 , K.Y. Jung 2 , S.K. Baek 2 1 Korea University Anam Hospital, Radiation Oncology, Seoul, Korea Republic of 2 Korea University Anam Hospital, Otolaryngology–Head and Neck Surgery, Seoul, Korea Republic of Purpose or Objective the purpose of this study was to evaluate the treatment outcomes of external beam radiation therapy in patients with locoregionally recurrent differentiated thyroid carcinoma. Material and Methods