Variation in prostate specific antigen results from 2 different assay platforms: clinical impact on 2304 patients undergoing prostate cancer screening

Abstract

Purpose: Urologists must contend with fluctuating prostate specific antigen (PSA) results from different assay platforms when deciding whether prostate biopsy is indicated. The contribution of cross-platform variation to these fluctuations is not well understood. To address this question we performed a cross-sectional study comparing 2 popular PSA assays during prostate cancer screening.

Materials and methods: The Baylor College of Medicine Prostate Cancer Screening Program is a community outreach program providing free screening to Houston residents. From September 18th to 23, 2000, 2,304 patients underwent digital rectal examination and parallel serum PSA assay by the Hybritech Access (Beckman Coulter, Inc., Chaska, Minnesota) and Centaur (Bayer Diagnostics, Tarrytown, New York) systems. The Wilcoxon signed ranks test was used to compare results.

Results: Median PSA was low for the 2 assays (Centaur 0.99 ng/ml and Access 1.09 ng/ml) and Access results were significantly higher (1.23 times) than those obtained with Centaur (p <0.001). Frozen serum from 50 patients with PSA greater than 2.5 were then re-assayed using a third methodology (third Generation Immulite, Diagnostic Products Corp., Los Angeles, California). These data were consistent with Centaur results and significantly lower than Access results (p <0.001). Using a cutoff of PSA greater than 4.0 ng/ml 55 of the 288 patients (19%) with PSA greater than 2.5 ng/ml on the 2 platforms would have been candidates for prostate biopsy based on Access but not on Centaur data.

Conclusions: In a large screening population the Access system measured consistently higher PSA than the Centaur system. These findings provide a basis for interpreting PSA results obtained from 2 commonly used clinical assays.