ESTRO 38 Abstract book

S654 ESTRO 38

morbidities and to reduce inevitable radiation related complications such as dysphagia, speech issue and neurological damage. Inspite of all, the prognosis remains uncertain. EP-1181 Structure delineation using a deformable image registration-based contour propagation in HNC B. Ng-Cheng-Hin 1 , D. McQuaid 2 , A. Dunlop 2 , S. Court 2 , I. Petkar 1 , C. Nutting 1 , K. Harrington 1 , S. Bhide 1 , K. Newbold 1 1 The Institute of Cancer Research and Royal Marsden NHS Trust, Head and Neck Radiotherapy and Imaging, Sutton, United Kingdom ; 2 The Institute of Cancer Research and Royal Marsden NHS Trust, Joint Department of Physics, Sutton, United Kingdom Purpose or Objective Delineation of head and neck structures and target volumes on planning CT is an essential but time-consuming process for the radiation oncologist (RO). Repeat delineation as part of adaptive radiotherapy substantially increases clinical workload, therefore there is interest in facilitating this process in a safe and efficient manner. The aim of this study is to assess the benefit of the contour propagation component of a commercial adaptive planning workflow and the potential to save RO delineation time. Material and Methods 5 head and neck cancer patients underwent a pre- treatment and repeat planning CT during treatment as part of the INSIGHT trial (CCR3926). The organs at risk (OARs) and target volume were delineated on the pre- treatment and repeat CT using RayStation 7.0 (RaySearch Laboratories, Stockholm, Sweden). The delineated structures from the pre-treatment scan were propagated to the repeat CT using an intensity-based Deformable Image Registration (DIR). This involved an initial rigid registration followed by DIR of the pre-treatment CT onto the repeat CT. The propagated structures were independently reviewed, modified and compared to the RO-delineated structures on the repeat scan. All contours were delineated by a single RO and reviewed by 2 experienced RO. RO delineation was compared to the DIR propagated group using time factors and qualitative analysis using Dice similarity coefficient (DSC) and mean distance to agreement (MDA). Results The mean time taken by the RO to delineate on the repeat CT was 87.34 minutes (SD 13.32). The mean time to propagate and amend the contours was significantly reduced (Paired t-test, p = 0.0039) to 34.46 minutes (SD 7.02). Compared to RO delineation, the mean MDA were 1.10 mm (SD 1.10) and 0.77 mm (SD 0.38) for unmodified and modified contours respectively. The mean DSC were 0.76 (SD 0.17) and 0.82 (SD 0.13) for the unmodified and modified contours respectively. DSC for the unmodified structures were <0.8 for the GTV and small OARs such as the lens, optic nerves and chiasm. All DSC either improved or remained unchanged following RO amendment.

Conclusion Contour propagation using DIR significantly reduces RO delineation time. Although volumetric overlap did not perform well for small organs and the GTV, the mean MDA for each structure showed a high level of agreement for all structures apart from the unmodified GTV. This suggests that DSC overestimates the errors for small structures. RO modifications led to an improvement in volumetric overlap of the majority of the organs and target volume, but did not match perfectly. This may be accounted for by intra-observer variability or a bias towards making minimal changes to the propagated structures. A larger patient database is required to confirm the findings of this study and further work on the dosimetry impact of RO modification versus no modifications should be made. EP-1182 Selective neck irradiation for oropharynx cancer in relation with human papilloma virus status S. Park 1 4 , D. Oh 1 , J.M. Noh 1 , Y.C. Ahn 1 , Y.H. Ko 2 , M.K. Chung 3 , H.S. Jeong 3 , Y.I. Son 3 , C. Baek 3 1 Samsung Medical Center, Radiation Oncology, Seoul, Korea Republic of ; 2 Samsung Medical Center, Pathology, Seoul, Korea Republic of ; 3 Samsung Medical Center, Otorhinolaryngology-Head and Neck Surgery, Seoul, Korea Republic of ; 4 Keimyung University Dongsan Medical Center, Radiation Oncology, Daegu, Korea Republic of. Purpose or Objective To evaluate the feasibility of selective neck irradiation (SNI) policy during definitive radiation therapy (dRT), in relation with human papilloma virus (HPV) status, for oropharyngeal cancer (OPC). Material and Methods From January 2008 to December 2017, 214 OPC patients received dRT, mainly by Helical Tomotherapy. HPV status was known in 150 patients (HPV+ in 115 and HPV- in 35), and the majority (138/150, 92.0%) received concurrent systemic therapy. Between HPV+ and HPV- patients, there