Assay Standardization Bias: Different Prostate Cancer Detection Rates and Clinical Outcomes Resulting from Different Assays for Free and Total Prostate-Specific Antigen

Abstract

Numerous commercial assays are available for measuring total and free prostate-specific antigen (PSA) levels in serum. These assays can be referenced to different laboratory standards, and interassay variability occurs. Patients and physicians might be affected by the variability between PSA assays that results from the use of different PSA standards. We prospectively compared the free and total PSA measurements obtained using two commercially available PSA assays in 103 participants from a prostate cancer screening program in Caracas, Venezuela. We recommended biopsy to men with a total PSA level of 3 to 10 ng/mL and a free/total PSA ratio of 20% or less with either assay. We compared the sensitivity, specificity, and concordance index between the two assays to assess the effects of interassay variability on the cancer detection rate and clinical outcomes. Although the total PSA results were similar between the assays, the free PSA level was significantly greater with one assay. Therefore, the free/total PSA ratio was discordant between the two assays, resulting in different biopsy recommendations and cancer detection rates. Using a free/total PSA ratio of 20% or less as the threshold for biopsy, the differences in assay sensitivity and specificity for detecting prostate cancer are significant. Commercially available assays for PSA and its derivatives are not necessarily interchangeable, and these differences might lead to different clinical outcomes. When using free and total PSA measurements to make clinical decisions, patients and physicians should be aware of the potential standardization bias and which assay is being used.