Does year of radical prostatectomy independently predict outcome in prostate cancer?

Objectives. To examine how the biochemical outcomes after radical prostatectomy (RP) have changed in the prostate-specific antigen (PSA) era when controlling for the effects of other prognostic variables. Since the beginning of the PSA era, the presentation, treatment, and therapeutic outcomes of prostate cancer have evolved. Methods. We reviewed the Columbia University Comprehensive Urologic Oncology Database and identified 1319 patients who had undergone RP without adjuvant therapy, performed by three surgeons, between 1988 and 2003 (minimal follow-up of 12 months). Univariate Cox proportional hazards models were used to determine which variables affect the hazard of biochemical failure (BCF), defined as a PSA level of 0.2 ng/mL or greater on at least two occasions, after RP. Multivariate analysis, controlling for the effects of other prognostic variables, was used to determine the effect that the year of surgery had on hazard of BCF. Results. Univariate analysis confirmed the importance of the year of surgery, preoperative PSA level, pathologic stage, Gleason sum, and surgical margin status in estimating the hazard of BCF (P <0.001). Age at surgery did not have a statistically significant effect. A multivariate Cox model showed that the year of surgery had a highly significant impact on the hazard of BCF even when controlling for PSA, stage, Gleason sum, and surgical margin status. Conclusions. Patients undergoing RP in more recent years are at significantly less risk of BCF compared with patients who underwent surgery earlier in the PSA era, even when controlling for the effects of other prognostic variables. The factors that account for this change in outcomes over time have yet to be identified.