Inter-observer comparison of target delineation for MRI-assisted cervical cancer brachytherapy: Application of the GYN GEC-ESTRO recommendations

Background and purpose: To investigate the inter-observer variation of target contouring when using the GYN GEC-ESTRO recommendations for MR image-guided brachytherapy (IGBT) for cervical cancer. Materials and methods: Nineteen cervical cancer patients, treated by radiotherapy at the Institut Gustave Roussy (IGR) in France (n = 9) or at the Medical University of Vienna (AKH) in Austria (n = 10) were included in this study. IGBT was used for all patients. Two radiation oncologists, one from IGR and the other from AKH, outlined the target volumes on MRI at the time of brachytherapy according to the GYN GEC-ESTRO recommendations. The absolute, common and encompassing volumes and their conformity indices (Cls) were assessed for the GTV, HR CTV and IR CTV. D90 and D100 for each volume were assessed. Visual evaluation was made to assess the reasons for the most frequent inter-observer differences. Results: The mean volumes of GTV and HR CTV did not differ significantly between the observers, p > 0.05. Significant differences were observed only for the mean volumes of the IR CTV of both centres, p < 0.05. CIs ranged from 0.5 to 0.7. DVH-parameter analyses did not reveal any statistical differences, except for the D100 for the GTV at AKH, and the D90 for the IR CTV at IGR, p < 0.05. Underlying reasons for inter-observer differences included image contrast adjustment and neglecting to consider anatomical borders. Conclusions: The results of this inter-observer study show that the application of the GYN GEC-ESTRO recommendations for IGBT contouring at two different institutions with two different traditions for applicators, CTV assessment, MR image acquisition and dose prescription is feasible, and it produces acceptable inter-observer variability. (C) 2008 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 91 (2009) 166-172