Prostate-specific antigen levels in 1695 men without evidence of prostate cancer. Findings of the American Cancer Society National Prostate Cancer Detection Project

Abstract

The American Cancer Society National Prostate Cancer Detection Project is a prospective, multidisciplinary, and multicenter trial to assess the potential for early detection of prostate cancer by transrectal ultrasonography (TRUS), digital rectal examination (DRE), and serum prostate-specific antigen assay (PSA). By November 1990, 2805 men between the ages of 55 and 70 years with no known signs or symptoms of prostate cancer were enrolled in the study, which is planned to run for 5 years. Annual TRUS, DRE, and PSA tests were done on these subjects, and biopsies were recommended for suspicious lesions when detected. To study the performance of PSA testing in presumed normal subjects, all men were eliminated who had (1) prostate cancer detected on their initial examinations and proven by biopsy or (2) cancer detected during the year or subsequent examinations. Additionally, all men with TRUS or DRE findings that were interpreted as suspicious for cancer but who are being followed and have not yet had biopsies done were removed from this series. This left a unique, extensively screened group of 1695 men who were free of prostate cancer, as far as could be determined. Analyses of the PSA levels in this large population in the appropriate age range for increasing risk of prostate cancer revealed several important findings. First, there was a direct relationship between serum PSA levels and estimated prostate volume for both the currently available monoclonal and polyclonal PSA assays. Individuals with benign prostatic hyperplasia and larger gland volume have a higher normal limit of PSA than men with normal gland volume. Second, analyses showed no relationship between age and PSA levels or between symptoms of prostatism and PSA levels independent of gland enlargement. It was concluded that volume-adjusted upper limits of normal PSA can be determined for different levels of specificity desired. This information may be applicable to the use of PSA in men not already suspected of having prostate cancer and may increase its effectiveness as a tool for early detection.