Correlation of prostate-specific antigen before prostate cancer detection and clinicopathologic features: evaluation of mass screening populations

OBJECTIVES: Although prostate-specific antigen (PSA) has become the reference standard for prostate cancer diagnosis, few reports have examined the long-term changes in PSA values before the diagnosis of prostate cancer in a large number of subjects. We investigated serial PSA levels and related values before prostate cancer diagnosis in a mass screening population and analyzed the values in an attempt to discover some values useful in clinical diagnostic science. METHODS: We performed mass screening for prostate cancer in 9671 subjects from 1986 to 1998. The initial screening method was measurement of prostatic acid phosphatase from 1986 to 1991 and measurement of PSA from 1992 to 1998. As a result, 303 cases of prostate cancer were diagnosed. For all the cases diagnosed before 1991, we measured the serum PSA value in preserved frozen serum. RESULTS: The prostate cancer detection rate was 3.1% among all subjects observed during the 13-year period. By measurement of the PSA level using frozen serum during the pre-PSA era, we found that 62% of patients demonstrated a PSA abnormality for more than 1 year (average 2.8) before prostate cancer diagnosis. Prostate cancer that was diagnosed within 1 year after a PSA value became abnormal was not associated with bone metastasis. Concerning the relationship between PSA velocity (PSAV) and clinical stage, the proportion of Stage B cancer was 86% in the subjects whose PSAV level before diagnosis was 0.18 ng/mL/yr or less and it was only 29% in those with PSAV levels of 4.5 ng/mL/yr or more. Only 3 (3.5%) of 86 patients with prostate cancer with PSAV levels of 4.4 ng/mL/yr or less had bone metastasis, and 2 of those 3 patients had poorly differentiated adenocarcinoma. CONCLUSIONS: Although a total of 62% of patients had an abnormal PSA level more than 1 year before prostate cancer diagnosis, no patients with prostate cancer who were diagnosed within 1 year after the PSA level became abnormal had bone metastasis. Among patients who have undergone mass screening twice or more, a clinically useful indicator of the lack of bone metastasis would be a period between the detection of PSA levels of 4.1 ng/mL or more but not more than 10 ng/mL and prostate cancer diagnosis of less than 1 year and a diagnosis of well or moderately differentiated adenocarcinoma or a PSAV of 4.4 ng/mL/yr or less and a cancer diagnosis of well or moderately differentiated adenocarcinoma.