BACKGROUND: The aim of this study was to assess the diagnostic accuracy of the proportion of prostate-specific antigen (PSA) complexed to alpha(1)-antichymotrypsin (PSA-alpha(1)ACT:PSA ratio) in the differential diagnosis of prostate cancer (CaP) and benign prostatic hyperplasia (BPH) in men with total PSA of 10-30 microg/L. METHODS: We used our immunoassays (ELISAs) for total PSA and PSA-alpha(1)ACT complex to study 146 men. In 123, total PSA was between 10 and 20 microg/L; 66 of these had CaP and 57 BPH. In 23 men, total PSA was between 20 and 30 microg/L; 14 of these had CaP and 9 BPH. We calculated the area under the ROC curves (AUC) for total PSA, PSA-alpha(1)ACT complex, and PSA-alpha(1)ACT:PSA ratio, and determined the cutoff points that gave sensitivities approaching 100%. RESULTS: In the total PSA range between 10 and 20 microg/L, the AUC was significantly higher for the PSA-alpha(1)ACT:PSA ratio (0.850) than for total PSA (0.507) and PSA-alpha(1)ACT complex (0.710; P <0.0001). A cutoff ratio of 0.62 would have permitted diagnosis of all 66 patients with CaP (100% sensitivity) and avoided 19% of unnecessary biopsies (11 of 57 patients). In the total PSA range between 20 and 30 microg/L, the AUC for the PSA-alpha(1)ACT:PSA ratio (0.980; 95% confidence interval, 0.82-0.99) was greater than the AUC for total PSA (0.750; 95% confidence interval, 0.51-0.89; P = 0.042). In this range, a cutoff point of 0.64 would have permitted the correct diagnosis of all 14 patients with CaP and 6 of the 9 with BPH. CONCLUSIONS: The diagnostic accuracy of the PSA-alpha(1)ACT:PSA ratio persists at high total PSA concentrations, increasing the specificity of total PSA. Prospective studies with large numbers of patients are needed to assess whether the ratio of PSA-alpha(1)ACT to total PSA is a useful tool to avoid unnecessary prostatic biopsy in patients with a total PSA >10 microg/L.
BACKGROUND: The aim of this study was to assess the diagnostic accuracy of the proportion of prostate-specific antigen (PSA) complexed to alpha(1)-antichymotrypsin (PSA-alpha(1)ACT:PSA ratio) in the differential diagnosis of prostate cancer (CaP) and benign prostatic hyperplasia (BPH) in men with total PSA of 10-30 microg/L. METHODS: We used our immunoassays (ELISAs) for total PSA and PSA-alpha(1)ACT complex to study 146 men. In 123, total PSA was between 10 and 20 microg/L; 66 of these had CaP and 57 BPH. In 23 men, total PSA was between 20 and 30 microg/L; 14 of these had CaP and 9 BPH. We calculated the area under the ROC curves (AUC) for total PSA, PSA-alpha(1)ACT complex, and PSA-alpha(1)ACT:PSA ratio, and determined the cutoff points that gave sensitivities approaching 100%. RESULTS: In the total PSA range between 10 and 20 microg/L, the AUC was significantly higher for the PSA-alpha(1)ACT:PSA ratio (0.850) than for total PSA (0.507) and PSA-alpha(1)ACT complex (0.710; P <0.0001). A cutoff ratio of 0.62 would have permitted diagnosis of all 66 patients with CaP (100% sensitivity) and avoided 19% of unnecessary biopsies (11 of 57 patients). In the total PSA range between 20 and 30 microg/L, the AUC for the PSA-alpha(1)ACT:PSA ratio (0.980; 95% confidence interval, 0.82-0.99) was greater than the AUC for total PSA (0.750; 95% confidence interval, 0.51-0.89; P = 0.042). In this range, a cutoff point of 0.64 would have permitted the correct diagnosis of all 14 patients with CaP and 6 of the 9 with BPH. CONCLUSIONS: The diagnostic accuracy of the PSA-alpha(1)ACT:PSA ratio persists at high total PSA concentrations, increasing the specificity of total PSA. Prospective studies with large numbers of patients are needed to assess whether the ratio of PSA-alpha(1)ACT to total PSA is a useful tool to avoid unnecessary prostatic biopsy in patients with a total PSA >10 microg/L.