Definition and preoperative predictors of persistently elevated prostate-specific antigen after radical prostatectomy: results from the Shared Equal Access Regional Cancer Hospital (SEARCH) database.

OBJECTIVES

To define a level of persistently elevated prostate-specific antigen (PSA) after radical prostatectomy (RP) that equates with high-risk for disease progression, and to identify preoperative predictors of PSA persistence among men from the Shared Equal Access Regional Cancer Hospital (SEARCH) database.

PATIENTS AND METHODS

A total of 901 men treated with RP between 2001 and 2008 were separated into groups based upon PSA nadir within 6 months after RP. We explored the association between nadir groups and time to biochemical recurrence (BCR) using multivariate Cox proportional hazards and determined the preoperative predictors of PSA persistence using logistic regression.

RESULTS

Relative to men with undetectable PSA levels, those with a PSA nadir of 0.03 (hazard ratio [HR] 3.88, P < 0.001), 0.04 (HR 4.87, P < 0.001), 0.05-0.09 (HR 12.69, P < 0.001), 0.1-0.19 (HR 13.17, P < 0.001), and 0.2 ng/mL (HR 13.23, P < 0.001) were at increased risk of BCR while men with a nadir of 0.01 (HR 1.36, P = 0.400) and 0.02 (HR 1.64, P = 0.180) were not. Using the PSA persistence definition of a PSA nadir > or = 0.03 ng/mL, 230 men (26%) had persistence. The independent preoperative predictors of PSA persistence were higher body mass index (BMI, P = 0.002), pathological Gleason score (relative to 2-6: 4 + 3-10, P = 0.001) and preoperative PSA level (P < 0.001).

CONCLUSIONS

Men with a PSA nadir > or = 0.03 ng/mL after RP were at higher risk for BCR. Using a PSA persistence definition of a PSA nadir > or = 0.03 ng/mL, persistence was predicted by known factors associated with aggressive disease (tumour grade, PSA level and BMI). Validation of the present definition in different populations using later end-points remains necessary to assess its prognostic usefulness.