To test the previously reported hypothesis, that the urinary/total serum (U/S) prostate-specific antigen (PSA) ratio improves the detection of prostate cancer, by evaluating the clinical usefulness of the U/S PSA ratio and comparing it with the free/total (F/T) serum PSA ratio.

A total of 165 patients undergoing transrectal ultrasound-guided prostate biopsy were prospectively included in this multicenter study. In all patients, PSA was measured from preoperative serum and 12-hour urine specimens in a centralized laboratory.

Prostate cancer was identified in 83 of 165 patients. The differences between patients with and without prostate cancer were statistically significant (P <0.001) when considering the total PSA value (median, 10.2 ng/mL and 6.6 ng/mL respectively), F/T serum PSA ratio (0.11 and 0.18), and U/S PSA ratio (1.2 and 4.2). In the group of 79 patients with a PSA level between 4 and 10 ng/mL, receiver operating characteristic curves showed that the U/S PSA ratio was associated with a larger area under the curve (0.63; 95% confidence interval, 0.51 to 0.73) than the total PSA value (0.55; 95% confidence interval, 0.43 to 0.66) or F/T serum PSA ratio (0.60; 95% confidence interval, 0.49 to 0.71). The U/S PSA ratio did not correlate with patient age or prostate volume.

Our results confirmed that the U/S PSA ratio may be a useful test in prostate cancer detection when the total serum PSA level is between 4 and 10 ng/mL. The F/T serum PSA ratio and U/S PSA ratio did not correlate. This suggests that these two tests could complement each other.