Multiinstitutional study on target volume delineation variation in breast radiotherapy in the presence of guidelines.

PURPOSE This study aims to determine magnitude, causes and consequences of post-operative breast tumour target volume delineation variation among radiation oncologists in the presence of guidelines. MATERIALS AND METHODS Excision cavities, CTVs and PTVs of eight breast cancer patients were delineated on CT scans by 13 Dutch radiation oncologists (observers) from 12 Dutch institutes participating in the international Young Boost Trial. Delineated volumes and conformity indices were determined. CTV delineation variation (SD) was determined for anatomically relevant regions. Non-parametric statistics were performed to establish effects of observers, patient characteristics and regions on delineation variation. RESULTS Even in the presence of delineation guidelines considerable delineation variation is present (0.24<SD<1.22 cm). Presence of clips or seroma reduced interobserver variation (0.24<SD<0.62 cm). Region-specific analysis showed distinct regions of higher variability per patient. This could not always be ascribed to anatomical features, suggesting interobserver variation is not solely due to lack of image quality. CONCLUSIONS In this study, interobserver delineation variation in breast tumour target volume delineation is larger than, e.g. setup inaccuracies and results from limited reliable visual guidance as well as interpretation differences between observers, despite guidelines. Reduction of delineation variation is essential in view of current developments in planning techniques, particularly for External Partial Breast Irradiation.