Do Ultrasensitive Prostate Specific Antigen Measurements Have a Role in Predicting Long-Term Biochemical Recurrence-Free Survival in Men after Radical Prostatectomy?

PURPOSE: In this study we evaluate an ultrasensitive prostate specific antigen assay in patients with prostate cancer after radical prostatectomy to predict long-term biochemical recurrence-free survival. MATERIALS AND METHODS: A total of 754 men who underwent radical prostatectomy and had an undetectable prostate specific antigen after surgery (less than 0.1 ng/ml) were studied. Prostate specific antigen was measured in banked serum specimens with an ultrasensitive assay (Hybritech® PSA, Beckman Coulter Access® 2) using a cutoff of 0.01 ng/ml. Prostate specific antigen was also measured in 44 men after cystoprostatectomy who had no pathological evidence of prostate cancer with the Hybritech assay and with the Quanterix AccuPSA™ assay. RESULTS: Of the 754 men 17% (131) experienced biochemical recurrence (median 4.0 years). Those men without biochemical recurrence (83%, 623) had a minimum of 5 years of followup (median 11). Prostate specific antigen was less than 0.01 ng/ml in 93.4% of men with no biochemical recurrence, whereas 30.5% of men with biochemical recurrence had a prostate specific antigen of 0.01 ng/ml or greater. On multivariate analysis postoperative prostate specific antigen at a 0.01 ng/ml cutoff, pathological stage and Gleason score, and surgical margins were significant independent predictors of biochemical recurrence risk. Kaplan-Meier estimates for mean biochemical recurrence-free survival were 15.2 years (95% CI 14.9-15.6) for prostate specific antigen less than 0.01 ng/ml and 10.0 years (95% CI 8.4-11.5) for prostate specific antigen 0.01 ng/ml or greater (p <0.0001). Biochemical recurrence-free rates 11 years after surgery were 86.1% (95% CI 83.2-89.0) for prostate specific antigen less than 0.01 ng/ml and 48.9% (95% CI 37.5-60.3) for prostate specific antigen 0.01 ng/ml or greater (p <0.0001). Prostate specific antigen concentrations in 44 men after cystoprostatectomy were all less than 0.03 ng/ml, with 95.4% less than 0.01 ng/ml. CONCLUSIONS: In men with a serum prostate specific antigen less than 0.1 ng/ml after radical prostatectomy a tenfold lower cutoff (0.01 ng/ml) stratified biochemical recurrence-free survival and was a significant independent predictor of biochemical recurrence, as were pathological features. Prostate specific antigen concentrations in men without pathological evidence of prostate cancer suggest that a higher prostate specific antigen concentration (0.03 ng/ml) in the ultrasensitive range may be needed to define the detection threshold.