gms | German Medical Science

Background: Screening for prostate cancer by serum measurement of prostate-specific antigen (PSA) is increasingly employed in many countries but, due to limitations in specificity, leads to a large number of unnecessary biopsies. Free-to-total PSA (%fPSA) has been proposed as an additional marker to reduce unnecessary biopsies. However, most studies on %fPSA were diagnostic studies conducted among symptomatic patients, and thus do not allow to estimate performance in a population-based screening setting. Prospective population-based studies are therefore necessary.

Methods: 32 incident prostate cancer cases, identified during a 2-year-follow-up of a prospective population-based study, and 759 randomly selected controls (matched by age) were included. In addition, 24 prostate cancer cases were recruited prior to initiation of treatment in a clinical study. Analyses of PSA and free PSA were done with stored blood specimens in a blinded fashion.

Results: Various combinations of PSA and %fPSA yielded higher levels of both sensitivity and specificity than PSA alone. Furthermore, some combinations of PSA and %fPSA showed a reduction of 26-53% of unnecessary biopsies while missing only 5-7% of prostate cancer cases compared to the currently used PSA cut-off value. %fPSA significantly differentiated between controls and both groups of cases both within the 4-10 ng/ml PSA range (area under the receiver operating curve, AUROC: 0.73 and 0.78) and for PSA levels below 4 ng/ml (AUROC: 0.84 and 0.89).

Discussion: A combination of %fPSA and PSA might help to increase both sensitivity and specificity of screening for prostate cancer. A substantial reduction of unnecessary biopsies might be possible by considering %fPSA in addition to PSA in screening for prostate cancer. The use of PSA and %fPSA based screening for reduction of prostate cancer mortality remains to be shown.