Identification of Patients With Prostate Cancer Who Benefit From Immediate Postoperative Radiotherapy: EORTC 22911
Van der Kwast, Theodorus H. × Bolla, Michel Van Poppel, Hendrik Van Cangh, Paul Vekemans, Kris Da Pozzo, Luigi Bosset, Jean-Francois Kurth, Karl H. Schröder, Fritz H. Collette, Laurence #
American Society of Clinical Oncology
Journal of Clinical Oncology vol:25 issue:27 pages:4178-86
PURPOSE: The randomized controlled European Organisation for Research and Treatment of Cancer (EORTC) trial 22911 studied the effect of radiotherapy after prostatectomy in patients with adverse risk factors. Review pathology data of specimens from participants in this trial were analyzed to identify which factors predict increased benefit from adjuvant radiotherapy. PATIENTS AND METHODS: After prostatectomy, 1,005 patients with stage pT3 and/or positive surgical margins were randomly assigned to a wait-and-see (n = 503) and an adjuvant radiotherapy (60 Gy conventional irradiation) arm (n = 502). Pathologic review data were available for 552 patients from 11 participating centers. The interaction between the review pathology characteristics and treatment benefit was assessed by log-rank test for heterogeneity (P < .05). RESULTS: Margin status assessed by review pathology was the strongest predictor of prolonged biochemical disease-free survival with immediate postoperative radiotherapy (heterogeneity, P < .01): by year 5, immediate postoperative irradiation could prevent 291 events/1,000 patients with positive margins versus 88 events/1,000 patients with negative margins. The hazard ratio for immediate irradiation was 0.38 (95% CI, 0.26 to 0.54) and 0.88 (95% CI, 0.53 to 1.46) in the groups with positive and negative margins, respectively. We could not identify a significant impact of the positive margin localization. CONCLUSION: Provided careful pathology of the prostatectomy is performed, our results suggest that immediate postoperative radiotherapy might not be recommended for prostate cancer patients with negative surgical margins. These findings require validation on an independent data set.