PURPOSE: Benign prostatic hyperplasia (BPH) was shown to be associated with high concentrations of urinary prostate specific antigen (PSA). We investigated the serum-to-urinary PSA ratio in patients undergoing prostate biopsy to assess its efficacy in enhancing serum PSA specificity in the detection of prostate carcinoma. MATERIALS AND METHODS: From November 1995 through January 1996 consecutive patients undergoing prostate biopsy were prospectively included in the study. Serum and urine PSA levels were measured at our laboratory with the Tandem-R assay. Samples were drawn 24 hours before prostate biopsy and at a distance from prostatic manipulation or ejaculation. RESULTS: We studied 73 patients with BPH and 57 with prostate cancer. Differences between BPH and prostate cancer were statistically significant considering serum PSA or serum-to-urinary PSA ratios. In the 50 patients with a serum PSA of 4.0 to 10.0 ng./ml. (35 with BPH and 15 with prostate cancer) the differences between prostate cancer and BPH were still significant only when considering serum-to-urinary PSA ratio. Receiver operating characteristic curves showed that serum-to-urinary PSA ratio was a better predictor of prostate cancer than serum PSA. CONCLUSIONS: Our results suggest that the serum-to-urinary PSA ratio may be useful in distinguishing BPH from prostate cancer, particularly in the diagnostic gray zone of serum PSA between 4.0 and 10.0 ng./ml.
PURPOSE:Benign prostatic hyperplasia (BPH) was shown to be associated with high concentrations of urinary prostate specific antigen (PSA). We investigated the serum-to-urinary PSA ratio in patients undergoing prostate biopsy to assess its efficacy in enhancing serum PSA specificity in the detection of prostate carcinoma. MATERIALS AND METHODS: From November 1995 through January 1996 consecutive patients undergoing prostate biopsy were prospectively included in the study. Serum and urine PSA levels were measured at our laboratory with the Tandem-R assay. Samples were drawn 24 hours before prostate biopsy and at a distance from prostatic manipulation or ejaculation. RESULTS: We studied 73 patients with BPH and 57 with prostate cancer. Differences between BPH and prostate cancer were statistically significant considering serum PSA or serum-to-urinary PSA ratios. In the 50 patients with a serum PSA of 4.0 to 10.0 ng./ml. (35 with BPH and 15 with prostate cancer) the differences between prostate cancer and BPH were still significant only when considering serum-to-urinary PSA ratio. Receiver operating characteristic curves showed that serum-to-urinary PSA ratio was a better predictor of prostate cancer than serum PSA. CONCLUSIONS: Our results suggest that the serum-to-urinary PSA ratio may be useful in distinguishing BPH from prostate cancer, particularly in the diagnostic gray zone of serum PSA between 4.0 and 10.0 ng./ml.