Predictors of overall and cancer-specific survival in patients with clinically localized prostate cancer (PC) treated with primary androgen deprivation therapy (PADT): Results from the Prostate Cancer Outcomes Study

Abstract

4610 Background: Use of PADT as sole therapy for clinically localized PC is rapidly increasing, despite little evidence to support this approach. We examined predictors of overall survival (OS) and cancer-specific survival (CSS) using the data available from the Prostate Cancer Outcome Study (PCOS). Methods: PCOS recruited 5,672 patients between 1994 and 1995, within 6 months of PC diagnosis. Vital status was last updated in 12/02. 276 patients had clinically localized PC and received PADT. We examined the relationship between baseline demographic, socio-economic, and tumor factors and OS and CSS using Kaplan-Meier and Cox regression methods implemented in the SUDAAN software. Results: Of 276 patients, 148 patients have died, 35 from PC. The median age at diagnosis was 75 (50–88). The average follow-up of censored subjects was 7.6 yrs (1.1–8.1). 5-year OS was 66% (95% CI: 69–72), while the 5-year CSS was 91% (95% CI: 86–94). In the multivariate analysis (see Table), age, serum PSA at diagnosis, and Gleason score were the only independent predictors of OS. Abnormal DRE was marginally significant. The Gleason score (≥7) was the only independent predictor of CSS (HR 4.0, p = 0.04); PSA ≥ 20 (HR 2.38, p = 0.09) approached significance. This analysis was limited by a smaller number of deaths due to prostate cancer (n = 35). Conclusions: In an analysis that included a broad range of demographic and socio-economic factors, overall survival of patients with clinically localized prostate cancer treated with PADT was predicted by age and measures of tumor biology (Gleason score) and tumor mass (serum PSA). Thus, cancer remains an important contributor to overall mortality in these patients, particularly those with high Gleason score and high serum PSA. These data can be used to inform patients regarding their individual 5-year survival if they choose PADT for the treatment of their clinically localized prostate cancer. [Table: see text] No significant financial relationships to disclose.