Predictors of time to biochemical recurrence in a radical prostatectomy cohort within the PSA-era

INTRODUCTION: We sought to determine predictors for early and late biochemical recurrence following radical prostatectomy among localized prostate cancer patients. METHODS: The study included localized prostate cancer patients treated with radical prostatectomy (RP) at the University of Southern California from 1988 to 2008. Competing risks regression models were used to determine risk factors associated with earlier or late biochemical recurrence, defined using the median time to biochemical recurrence in this population (2.9 years after radical prostatectomy). RESULTS: The cohort for this study included 2262 localized prostate cancer (pT2-3N0M0) patients who did not receive neoadjuvant or adjuvant therapies. Of these patients, 188 experienced biochemical recurrence and a subset continued to clinical recurrence, either within (n=19, 10%) or following (n=13, 7%) 2.9 years after RP. Multivariable stepwise competing risks analysis showed Gleason score ≥7, positive surgical margin status, and ≥pT3a stage to be associated with biochemical recurrence within 2.9 years following surgery. Predictors of biochemical recurrence after 2.9 years were Gleason score 7 (4+3), preoperative prostate-specific antigen (PSA) level, and ≥pT3a stage. CONCLUSIONS: Higher stage was associated with biochemical recurrence at any time following radical prostatectomy. Particular attention may need to be made to patients with stage ≥pT3a, higher preoperative PSA, and Gleason 7 prostate cancer with primary high-grade patterns when considering longer followup after RP.