Fifteen PSA Assays

During 1995/96, this unit performed a series of evaluations of kits for the determination of prostate specific antigen (PSA). The evaluations included an assessment of the presentation of the kits, ease of use and performance characteristics. This document attempts to present the findings of these evaluations in a manner which allows the reader to compare the kits against each other. All kits were found to be well presented and packaged and all complied reasonably well with IFCC guidelines on packaging and labelling. The most frequent omission was quality control material, available as a separate item in most cases. Radioimmunoassay kits proved to be easier to use than instrument based and enzymimmunoassays. Imprecision was acceptable for all the kits evaluated; CVs ranged from 1.1 to 14.1% for between-assay imprecision, 1.4 to 13.3% for within-assay imprecision. As may be expected, the automated assays gave the better results. Linearity was excellent in all cases and no significant high dose hook effects were seen. Clinical sensitivity for prostate cancer remained consistent throughout the series (83% to 91%). The most striking difference between the kits was the relative recovery rate of free PSA to complexed PSA. Only four of the kits assessed were considered to recover the two forms in equimolar proportions. The remaining kits all favoured the detection of the free form to some degree, with ratios up to 7.02 : 1. Correlation to the Hybritech Tandem-R method was also inconsistent, many kits showing altered correlation coefficients at differing total PSA levels.