BACKGROUND: Several methods for the identification of patients with prostate carcinoma have been proposed to enhance the clinical usefulness of prostate specific antigen (PSA). However, it remains unclear which method is superior in practical use. The authors attempted prospectively to identify the most powerful method with which to detect prostate carcinoma, especially among patients with intermediate PSA levels. METHODS: Between October 1997 and August 1999, systematic sextant biopsies were performed on 281 patients, including 147 with PSA levels between 4.1 ng/mL and 10.0 ng/mL. The clinical values of PSA, the free PSA to total PSA ratio (free/total PSA ratio), alpha-1-antichymotrypsin-PSA complex (PSA-ACT), the calculated derivatives, PSA density (PSAD), and PSA density of the transition zone (PSATZD) for the detection of prostate carcinoma were compared by using receiver operating characteristic (ROC) curves and logistic regression analyses. RESULTS: According to ROC curve analysis, PSATZD had the greatest area under the curve in the overall patient population and in patients with intermediate PSA levels. In patients with intermediate PSA levels, at the sensitivity of 90%, PSATZD would have prevented unnecessary biopsies in 68 of 117 patients who were without prostate carcinoma, whereas PSA, free/total PSA ratio, and PSA-ACT would have prevented unnecessary biopsies in 25, 28, and 25 patients, respectively. Stepwise logistic regression analysis showed that PSATZD and findings on digital rectal examination were significant independent predictors. CONCLUSIONS: PSATZD had the most useful validity in the differentiation between prostate carcinoma and benign prostatic enlargement in the overall patient population and in patients with intermediate PSA levels. Copyright 2000 American Cancer Society.
BACKGROUND: Several methods for the identification of patients with prostate carcinoma have been proposed to enhance the clinical usefulness of prostate specific antigen (PSA). However, it remains unclear which method is superior in practical use. The authors attempted prospectively to identify the most powerful method with which to detect prostate carcinoma, especially among patients with intermediate PSA levels. METHODS: Between October 1997 and August 1999, systematic sextant biopsies were performed on 281 patients, including 147 with PSA levels between 4.1 ng/mL and 10.0 ng/mL. The clinical values of PSA, the free PSA to total PSA ratio (free/total PSA ratio), alpha-1-antichymotrypsin-PSA complex (PSA-ACT), the calculated derivatives, PSA density (PSAD), and PSA density of the transition zone (PSATZD) for the detection of prostate carcinoma were compared by using receiver operating characteristic (ROC) curves and logistic regression analyses. RESULTS: According to ROC curve analysis, PSATZD had the greatest area under the curve in the overall patient population and in patients with intermediate PSA levels. In patients with intermediate PSA levels, at the sensitivity of 90%, PSATZD would have prevented unnecessary biopsies in 68 of 117 patients who were without prostate carcinoma, whereas PSA, free/total PSA ratio, and PSA-ACT would have prevented unnecessary biopsies in 25, 28, and 25 patients, respectively. Stepwise logistic regression analysis showed that PSATZD and findings on digital rectal examination were significant independent predictors. CONCLUSIONS: PSATZD had the most useful validity in the differentiation between prostate carcinoma and benign prostatic enlargement in the overall patient population and in patients with intermediate PSA levels. Copyright 2000 American Cancer Society.