The percentage of free prostate-specific antigen does not predict extracapsular disease in patients with clinically localized prostate cancer before radical prostatectomy.

Abstract

To determine whether the percentage of free/total prostate-specific antigen (f/tPSA) can predict the pathological features in patients with clinically localized prostate cancer before radical prostatectomy. Univariate and multivariate logistic regression was used to analyse data from 171 untreated patients who underwent radical prostatectomy. Variables included the total PSA (tPSA), fPSA, f/tPSA, biopsy Gleason score, clinical stage and patient age. In 115 patients with pathologically organ-confined tumours ( pT2N0) the mean (SD) tPSA value was 6.9 (5.6) ng/mL; in 56 patients with extracapsular disease ( pT3pN0/N+) it was 10.2 (7.6) ng/mL; the respective f/tPSA values were 14.9 (8.1)% and 14.2 (12.9)%. In the univariate and multivariate analysis, tPSA and biopsy Gleason score were highly significant in predicting extracapsular disease (P < 0.001 and 0.002) but the f/tPSA was not (P = 0.18). There was no significant difference between the mean f/tPSA and final Gleason scores. The f/tPSA does not predict extracapsular disease in patients with clinically localized prostate cancer before radical prostatectomy. Knowing the f/tPSA provides no significant additional information in predicting extracapsular disease when the biopsy Gleason score and tPSA are known.