Early detection of prostate cancer local recurrence by urinary prostate-specific antigen

Purpose: We assessed the role of urinary prostate-specific antigen (uPSA) in the follow-up of prostate cancer after retropubic radical prostatectomy (RRP) for the early detection of local recurrences. Methods: We recruited 50 patients previously treated for prostate cancer with RRP and who had not experienced a prostate-specific antigen (PSA) recurrence within their first postoperative year into a cross-sectional laboratory assessment and prospective 6-year longitudinal follow-up study. We defined biochemical failure as a serum PSA (sPSA) of 0.3 mu g/L or greater. Patients provided blood samples and a 50-mL sample of first-voided urine. We performed Wilcoxon rank-sum and Fisher exact tests for statistical analysis. Results: The median sPSA was 0.13 mu g/L. The median uPSA was 0.8 mu g/L, and was not significantly different when comparing Gleason scores or pathological stages. Of the 50 patients, 27 initially had a nondetectable sPSA but a detectable uPSA, and 11 patients experienced sPSA failure after 6 years. Six patients had detectable sPSA and uPSA initially. Fifteen patients were negative for both sPSA and uPSA, and 13 remained sPSA-free after 6 years. The odds ratio (OR) of having sPSA failure given a positive uPSA test was 4.5 if sPSA was undetectable, but was reduced to 2.6 if sPSA was detectable. The pooled Mantel-Haenszel OR of 4.2 suggested that a detectable uPSA quadrupled the risk of recurrence, independent of whether sPSA was elevated or not. The sensitivity of uPSA for detecting future sPSA recurrences was 81% and specificity was 45%. Conclusion: Urinary PSA could contribute to an early detection of local recurrences of prostate cancer after a radical prostatectomy.