Posttreatment prostate-specific antigen nadir highly predictive of distant failure and death from prostate cancer.

PURPOSE: To link posttreatment biochemical profiles to distant failure and cause-specific survival by assessing the relationship between posttreatment prostate-specific antigen (PSA) nadir and PSA doubling time (PSADT) with these outcome measures. METHODS AND MATERIALS: A total of 615 men were treated at the Fox Chase Cancer Center between April 1989 and December 1995 with three-dimensional conformal radiotherapy alone (median dose 73 Gy). The median follow-up was 64 months (range 2-135). Kaplan-Meier methods were used to estimate the rates of biochemical control, freedom from distant metastasis (FDM), and cause-specific survival. Multivariate predictors of outcome were assessed using stepwise Cox regression analysis.
RESULTS: Multivariate analyses demonstrated that the predictors of improved biochemical control were a lower PSA nadir (p <0.0001), lower pretreatment PSA level (p <0.0001), Gleason score of 2-6 (p = 0.001), Stage T1-T2a tumors (p = 0.03), and higher RT dose (p = 0.02). The predictors of improved FDM were a lower PSA nadir (p <0.0001), longer interval to nadir from start of treatment (p = 0.0002), Gleason score of 2-6 (p = 0.005), androgen deprivation for biochemical failure (p = 0.001), and Stage T1-T2a tumors (p = 0.01). The predictors of improved cause-specific survival were a lower PSA nadir (p = 0.006) and longer interval to nadir from the start of treatment (p = 0.03). The 8-year FDM rate was 96%, 89%, and 61% for PSA nadir values of <or=1.0, 1.1-2.0, and >2.0 ng/mL (p <0.0001), respectively. The 8-year cause-specific survival rate was 97%, 96%, and 78% for posttreatment PSA nadir values of <or=1.0, 1.1-2.0, and >2.0 ng/mL (p <0.0001), respectively. For patients with sufficient PSA follow-up for PSADT calculations (n = 136), multivariate analysis of FDM from the time of biochemical failure demonstrated that androgen deprivation (p = 0.001), longer PSADT (p = 0.003), lower PSA nadir (p = 0.02), and longer interval to nadir from start of treatment (p = 0.04) were independent predictors of improved FDM.
CONCLUSION: This is the first study, to our knowledge, to demonstrate the overwhelming predictive power of posttreatment PSA nadir for distant failure and death from prostate cancer. It is also the first study, to our knowledge, to demonstrate a strong association between posttreatment PSADT and distant failure. The results provide new information regarding disease progression as a function of posttreatment PSA profiles (time to achieve nadir from start of treatment, nadir, and PSADT) and the timing of androgen deprivation for biochemical relapse. This study may be used for the early identification of patients at high risk of distant metastasis and who may be directed to applicable systemic treatment clinical trials.