Abstract
PURPOSE: Consistency in delineation of pelvic lymph node regions for prostate cancer elective nodal radiation therapy is still challenging despite current guidelines. The aim of this study was to evaluate the interobserver variability in elective lymph node delineation in the PEACE V - STORM randomized phase 2 trial for oligorecurrent nodal prostate cancer. METHODS AND MATERIALS: Twenty-three centers were asked to delineate the elective pelvic nodal clinical target volume (CTV) of a postoperative oligorecurrent nodal prostate cancer benchmark case using a modified Radiation Therapy Oncology Group (RTOG) 2009 template (upper limit at the L4/L5 interspace). Overall, intersection and overflow volumes, Dice coefficient, Hausdorff distance, and count maps merged with computed tomography images were analyzed. RESULTS: The mean volume including the 23 nodal CTVs was 430.4 ± 64.1 cm(3), larger than the modified RTOG 2009 CTV reference volume (386.1 cm(3)). The intersection common volume between the modified reference RTOG 2009 and the 23 nodal CTVs was estimated at 83.9%, whereas the overflow volume was 23.4%, mainly located at the level of the presacral and the upper limit of the L4/L5 interspace. The mean Dice coefficient was 0.79 ± 0.02, whereas the mean Hausdorff distance was 27 ± 4.4 mm. CONCLUSIONS: In salvage radiation therapy treatment of oligorecurrent nodal prostate cancer, variations in elective lymph node volume delineation were mainly observed in the presacral and common iliac areas. Routine implementation and diffusion of available contouring guidelines together with a constant evaluation and evidence-based updating are expected to further decrease the existing variability in pelvic node contouring.