Longitudinal evaluation of prostate-specific antigen levels in a case-control study.

Background: Over the last period--late 1970 to early 1990--the incidence of prostate carcinoma has nearly doubled, even though many more patients die with prostate cancer rather than of it. This finding, together with the slow growth of this tumor and the absence of a controlled trial, makes early diagnosis for this pathology quite questionable. On the other hand, it is well known that prostatic carcinoma is curable as long as it is intracapsular and that there is an ever increasing encouragement for early detection in all diseases. Since prostatic pathology increases its incidence as age advances, the first step in early diagnosis is to be able to discriminate between healthy, benign prostatic hyperplasia and cancer cases with a well-accepted and easily understandable method. The problem is to find the best method to do it.

Methods: We measured serum prostate-specific antigen (PSA) levels in 435 men participating in an epidemiological study, at first in 1987 and again in 1992. Men with PSA levels above 4 ng/ml (in 1992) were invited to undergo other diagnostic tests (digital rectal examination, transrectal ultrasonography, biopsy) in sequence on the basis of the results of the previous tests. The pathological findings from biopsies were the reference test to determine the presence or absence of prostate cancer (2.5% of the population).

Results: We divided PSA concentrations into three categories, according to the most used cutoffs ( < 4, 4-10, > 10 ng/ml); in the meantime, we took into account the change rate in PSA concentration in time, defined as delta PSA. By the comparison between PSA categories and delta PSA, we found out that the first one does not discriminate between benign and malignant pathologies, while the use of delta PSA strongly discriminates them (p < 0.001).

Conclusion: On the basis of our results, we think that delta PSA might be the best parameter to indicate the presence of prostate cancer cases in an asymptomatic population.