Assessment of Oncological Outcomes After Radical Prostatectomy According to Preoperative and Postoperative Cancer of the Prostate Risk Assessment Scores: Results from a Large, Two-center Experience

Abstract

BackgroundAmong prostate cancer (PCa) patients undergoing radical prostatectomy (RP) and with virtually identical unfavorable pathological characteristics, those deemed at low risk (LR) preoperatively had better oncological outcomes than those with intermediate (IR) or high risk (HR) preoperatively. ObjectiveTo examine if this phenomenon still applies when preoperative Cancer of the Prostate Risk Assessment (CAPRA) scores are compared to postoperative scores (CAPRA-S) in RP patients. Design, setting, and participantsWe evaluated 10 290 PCa patients who underwent RP at tertiary care centers in Hamburg (Germany) and Milan (Italy) during 1991–2011. Outcome measurements and statistical analysisAll patients were classified as CAPRA/CAPRA-S LR, IR, or HR (≤2, 3–5, and ≥6 points, respectively). Kaplan-Meier and Cox regression analyses were performed to assess the risk of biochemical recurrence (BCR) and metastatic disease (metD). Results and limitationsOverall, 59.6%, 29.3%, and 11.1% patients were CAPRA-S LR, IR, and HR respectively. For CAPRA-S LR patients, 5-yr BCR-free and metD-free rates for preoperative CAPRA LR versus IR/HR patients were 93.1% versus 85.7% (p<0.001) and 99.7% versus 98.9% (p=0.017), respectively. For CAPRA-S IR/HR patients, the corresponding rates were 69.9% versus 57.2% and 98.2 versus 93.7% (both p<0.001). On multivariable Cox regression analyses, a combination of CAPRA-S and preoperative CAPRA resulted in increases in predictive accuracy for BCR (from 76.0% to 78.3%) and metD (from 82.0% to 84.0%). Specifically, biopsy Gleason patterns and the percentage of positive cores added information to the CAPRA-S score. Long-term follow-up is needed to discern clinical metD differences between preoperative CAPRA risk groups. ConclusionsCAPRA-S LR patients in the CAPRA IR/HR group had higher BCR/metD risk. Conversely, CAPRA-S IR/HR patients in the CAPRA LR group had lower BCR/metD risk. Future tools should incorporate better tumor and Gleason quantification to optimize prediction. Patient summaryWe demonstrated that among 10290 European patients with prostate cancer who underwent radical prostatectomy, those deemed at low risk preoperatively had better oncological outcomes than their peers with intermediate or high risk, despite virtually identical unfavorable pathological characteristics.