External validation of University of California, San Francisco, Cancer of the Prostate Risk Assessment score.

OBJECTIVES In 2005, the University of California, San Francisco, proposed the Cancer of the Prostate Risk Assessment (UCSF-CAPRA) score to predict the risk of biochemical recurrence (BR) after radical prostatectomy. This study provides external validation and a modified version of the model using a large cohort of men treated with radical prostatectomy at a high-volume, tertiary referral center. METHODS From 1984 to 2006, 6737 men underwent radical prostatectomy at our institution for clinical Stage T1c-T3a prostate cancer with available follow-up information and no neoadjuvant or adjuvant therapy before BR. The BR-free survival was estimated using the Kaplan-Meier method and compared by UCSF-CAPRA score using the log-rank statistic. Performance of the UCSF-CAPRA was evaluated using Cox proportional hazards regression analysis and Harrell's concordance (c) index and compared with the Kattan nomogram. The UCSF-CAPRA score and final pathologic findings were assessed by odds ratios. RESULTS The 5-year BR-free survival rate was 83.1% overall and decreased from 94.4% for men with a UCSF-CAPRA score of 1 or less to 25.8% for those with a score of 7 or more (P <0.0001). The hazards ratio approximately doubled for each UCSF-CAPRA point until a score of 4, when the hazards ratio increased at a slower rate. The c-index of the UCSF-CAPRA and Kattan nomogram was 0.76 and 0.78, respectively. A greater UCSF-CAPRA score correlated with the final pathologic findings. CONCLUSIONS The UCSF-CAPRA performed well in this tertiary, referral-based cohort with a c-index similar to that of the Kattan nomogram. It remains an effective prognostic instrument for predicting the risk of biochemical recurrence after radical prostatectomy.