The value of screening tests in the detection of prostate cancer. Part II: Retrospective analysis of free/total prostate-specific analysis ratio, age-specific reference ranges, and PSA density.

OBJECTIVES The ratio between free and total prostate-specific antigen (PSA) in serum (F/T ratio) was shown to improve the specificity of total serum PSA for the detection of prostate carcinoma in selected populations. In this study, the value of the F/T ratio for screening of prostate cancer was compared with that of age-specific reference ranges for PSA and PSA density (PSAD) by a simulation experiment. METHODS In 1726 men between 55 and 76 years old, 67 prostate carcinomas were detected by application of digital rectal examination (DRE), transrectal ultrasonography (TRUS), and total serum PSA. A serum PSA of 4.0 ng/mL or more, an abnormal DRE, or an abnormal TRUS were the indications to perform 308 biopsies. A simulation was performed in which an F/T ratio of 0.20 (ProStatus PSA Free/Total), age-specific PSA reference ranges, and a PSAD of 0.12 ng/mL/cc were used to study their capability to increase the specificity of total serum PSA in predicting prostate biopsy results. RESULTS Using age-specific PSA reference ranges and DRE as indicators for biopsy, a reduction of 37% (113) of biopsies would have been obtained with a loss of detected cancers of 12% (11). For the use of PSAD and DRE, these numbers were 28% (96) and 11% (7), respectively. Application of a serum PSA of 4.0 ng/mL or more and an F/T ratio of 0.20 or less and an abnormal DRE as indicators for biopsy would reduce the number of biopsies by 37% (112) and the number of detected cancers by 11% (7). The biopsy to prostate cancer ratio of these simulations varied between 3.3 and 3.6. Minimal loss of cancer detection of 3% (2) with a reduction in the number of biopsies of 17% (53) is obtained when TRUS is omitted from the screening protocol. Selecting men by a total serum PSA value of 2.0 ng/mL for further diagnostic workup by TRUS and DRE would have reduced the number of biopsies by 30% (102), and the number of cancers detected by 6% (4). CONCLUSIONS The most cost-effective protocol for screening prostate carcinoma appears to be prescreening by total serum PSA. The F/T ratio might be used to detect carcinomas in the PSA range below 4.0 ng/mL, but the best threshold remains to be assessed.