THE CORRELATION BETWEEN SERUM PROSTATE-SPECIFIC ANTIGEN AND PROSTATE-CANCER IS NOT INFLUENCED BY THE SERUM TESTOSTERONE CONCENTRATION

Objectives. To determine if the serum testosterone (T) concentration i nfluences the ability of prostate-specific antigen (PSA) to predict pr ostate cancer volume and stage. Methods. One hundred consecutive patie nts with clinically localized prostate cancer who underwent radical pr ostatectomy were examined prospectively. Each patient was evaluated pr eoperatively with a serum PSA, total T, free T, and percent free T. Al l surgical specimens were evaluated using the whole mount, step sectio n technique for Gleason score, tumor volume, and extraprostatic diseas e.Results. Serum total T, free T, and percent free T did not correlate with the serum PSA level (r = .03, .08, and .07, respectively), tumor volume (r = .11,.08, and .11, respectively), prostate weight (r = .00 , -.08, and .11, respectively), or Gleason score (r = .00, .01, and -. 03, respectively). Serum PSA correlated with tumor volume (r = .51, P < 0.0001). Extraprostatic disease was significantly associated with a higher percent free T value (r = .26, P = 0.02) but not with either th e total or the free T level. Linear regression analysis showed that ne ither the total nor the free T concentration was a significant predict or of extraprostatic disease in the presence of PSA (P = 0.30 and 0.24 , respectively); percent free T contributed only slightly to PSA in th e prediction of extraprostatic disease (P = 0.05). However, neither to tal T, free T, nor percent free T was a significant predictor of tumor volume; in essence, the association between PSA and tumor volume was independent of the serum T concentration (P = 0.30, 0.24, and 0.60, re spectively). Conclusions. Serum total T, free T, and percent free T va lues do not enhance the ability of PSA to predict the tumor volume or pathologic stage in patients with clinically localized prostate cancer .