The proportion of free to total prostate specific antigen: a method of detecting prostate carcinoma.

Abstract

Prostate specific antigen (PSA) is the most useful marker for prostate carcinoma (CaP). However, the sensitivity and specificity for PSA are not sufficient for the diagnosis of organ-confined prostate carcinoma. Recent studies have revealed that anti-PSA antibody identifies both PSA complexed to alpha-1-antichymotrypsin and free PSA, whereas anti-gamma-seminoprotein antibody recognizes free PSA exclusively. To enhance the ability of PSA to detect CaP in patients with total PSA levels of 10 ng/mL or lower, we developed the ratio of gamma-seminoprotein and PSA (free/total PSA index). We measured free/total PSA indices for 285 patients who had serum PSA levels of 10 ng/mL or lower and who were diagnosed pathologically. Of the 285 patients, 228 had no prostate carcinoma (NC) and 57 had CaP. The mean total PSA level for CaP (5.137 +/- 2.483 ng/mL; mean +/- standard deviation) was significantly greater (P < 0.0001) than that for NC (3.251 +/- 2.129). The mean free/total PSA index for CaP (0.774 +/- 0.468 was significantly lower (P < 0.0001) than that for NC (1.563 +/- 0.938). The sensitivity for the free/total PSA index was similar to that for total PSA (78.9% vs. 75.4%). However, the specificity, positive predictive value, and overall accuracy for the free/total PSA index (75.9%, 45%, and 76.5%, respectively) increased by 15-20% compared with those for total PSA (56.6%, 30.3%, and 60.4%, respectively). The free/total PSA index improved the specificity of PSA without impairing the sensitivity in detecting CaP among patients with serum PSA levels of 10 ng/mL or lower.