ESTRO 2023 - Abstract Book

S458

Sunday 14 May 2023

ESTRO 2023

Universitaire Vaudois (CHUV), Otolaryngology - Head and Neck Surgery, Lausanne, Switzerland; 9 Centre Léon Bérard, Radiation Oncology, Lyon, France; 10 Institut Jules Bordet, Radiophysics and MRI Physics Laboratory, EORTC RTQA Chair, Brussels, Belgium; 11 University Hospital Zürich, Radiation Oncology, EORTC RTQA Chair, Zürich, Switzerland Purpose or Objective The EORTC 1420 ‘Best of’ trial aims to assess swallowing function after trans-oral surgery versus radiotherapy in head and neck cancer (RT). RT participating sites had to complete a benchmark case (BC) procedure by delineating and planning a specific patient case according to study protocol. Our objective was to analyse interobserver variability (IOV) in contouring and planning in the complete BC dataset. Materials and Methods A total of 43 sites participated in the BC procedure. Each site’s treatment plan was sent for expert review for protocol compliance and approval. For each site, the submitted structure set (Sub-SS) was compared to a set of gold standard structures (GS-SS) created by consensus between the RTQA experts. IOV was evaluated by measuring mean conformity index (CI) and maximum Hausdorff Distance (HD) between Sub-SS and GS-SS. Dose-volume parameters for targets and OAR were extracted from the approved plan using both Sub-SS and GS-SS and compared. Results Forty-two of the 43 sites (97.67%) completed the procedure. Due to delineations 32 sites had to resubmit Sub-SS at least once before approval, while only 9 had to resubmit RT plans. Lowest CI indexes and highest HD were measured in OAR not usually delineated, such as the swallowing structures (Fig. 1). Sites successfully created acceptable target volumes coverages, only 5/42 sites having minor deviations attributable to treatment planning system differences. The OAR most affected by delineation IOV, when applying site doses to GS-SS, were the Brainstem PRV and Cervical Esophagus where we observed a dose difference of >4 Gy (Table 1). While within OAR protocol limits, this variation resulted in unacceptable PTV coverages with GS-SS (6/42 sites).

Organs at Risk/Target Volumes Mean Dose (Gy) Mean dGS*** (Gy) Mean diff. Brainstem PRV 18.37 26.88 -8.51 Brainstem 13.64 16.51 -2.87 Cervical Esophagus 14.35 18.51 -4.16 Parotid Left 5.24 5.25 -0.01 Parotid Right 24.34 24.63 -0.30 CPI* 51.83 48.17 3.65 Submandibular Left 7.50 8.18 -0.68 Submandibular Right 55.31 54.94 0.38 Oral Cavity - PTV 27.05 29.08 -2.03 Larynx 22.12 22.04 -1.08 Mandible 65.93 66.16 -0.24 PCM** 33.42 33.20 0.22 Spinal Cord PRV 35.1 38.56 -3.45