To determine the ideal cutpoint for defining prostate-specific antigen (PSA) recurrence after radical prostatectomy (RP). Although various cutpoints have been used, a recent study suggested that 0.4 ng/mL may be the most appropriate.

A retrospective survey of 358 men undergoing RP at the West Los Angeles Veterans Affairs Medical Center between 1991 and 2001 was undertaken. The 3-year and 5-year risk of PSA recurrence was estimated by Kaplan-Meier analyses using various cutpoints of postoperative PSA to define recurrence: greater than 0.1, greater than 0.2, greater than 0.3, greater than 0.4, and greater than 0.5 ng/mL. The 1 and 3-year risk of PSA progression after a detectable PSA level (PSA rising to a higher cutpoint) was evaluated for each definition of PSA recurrence using Kaplan-Meier analyses. Multivariate analysis using a Cox proportional hazards model was used to determine the clinical variables that were significant independent predictors of PSA recurrence at each cutpoint.

For patients with a detectable postoperative PSA value from 0.11 to 0.2 ng/mL, the 1 and 3-year risk of PSA progression was 64% (95% confidence interval [CI] 46% to 82%) and 93% (95% CI 74% to 99%), respectively. For patients with a PSA value from 0.21 to 0.3 ng/mL, the 1 and 3-year risk of PSA progression was 86% (95% CI 69% to 97%) and 100% (95% CI 87% to 100%), respectively. The use of higher PSA cutpoints to define recurrence resulted in a lower 5-year risk of PSA recurrence. The 5-year risk of PSA recurrence using a greater than 0.1 ng/mL cutpoint resulted in a 43% (95% CI 36% to 50%) risk of recurrence compared with only 23% (95% CI 18% to 30%) for a greater than 0.5 ng/mL cutpoint. In multivariate analysis, PSA and biopsy Gleason score were significant independent predictors of biochemical recurrence, regardless of the definition of PSA recurrence used (P ≤0.002).

PSA and biopsy Gleason score were significant predictors of biochemical failure, regardless of the definition of failure used. However, the definition of PSA recurrence dramatically affected the perceived success of therapy. Patients with a postoperative PSA value greater than 0.2 ng/mL are at very high risk of developing an additional rise in PSA. On the basis of this finding, a PSA value greater than 0.2 ng/mL is an appropriate cutpoint to define PSA recurrence after RP.