Consistency and diagnostic accuracy of 4 assays in the detection of the total and free prostate-specific antigen.

Abstract

The lack of interchangeability among prostate-specific antigen (PSA) assays causes difficulties in clinical interpretation. The currently available mainstream assays for PSA were based on Western populations, but it is not clear whether these assays yield different results in prostate cancer (PCa) screening in Chinese populations. A total of 163 men with a total PSA (tPSA) level of 2-10 µg/L scheduled for prostate biopsy were enrolled in this study. The levels of the tPSA and free PSA (fPSA) were detected by the Beckman (using the Hybritech calibration), Roche, Abbott, and Mindray (using the World Health Organization's calibration). Methodological comparison were performed according to EP9-A3 of the Clinical and Laboratory Standards Institute, USA. With pathological diagnosis as the gold standard, the predictive accuracy of the biomarkers was quantified as the area under the receiver operating characteristic curve (AUC) for each of the four methods. A total of 32 PCa patients and 131 patients with benign prostate disease were included in this study. The tPSA levels detected by the Roche, Abbott, and Mindray showed good consistency with Beckman but not of the fPSA levels. Compared to the Beckman tPSA values, the measured values were 1.1% higher for Roche, 2.1% higher for Abbott, and 6.7% higher for Mindray. The fPSA levels measured by Roche, Abbott, and Mindray were 5.4% lower, 22.1% higher, and 4.6% higher than Beckman, respectively. When the tPSA was 4 ng/mL, the diagnostic performance (sensitivity, specificity, positive predictive value, negative predictive value, and missed diagnosis rate) was similar among these 4 assays. When the %fPSA was 16%, Abbott had the highest missed diagnosis rate (37.50%), while Mindray had a sensitivity of 81.25%, the highest negative predictive value (93.88%), and the lowest missed diagnosis rate (18.75%). When the tPSA level is 2-10 ng/mL, the Mindray, like the Roche and Abbott, has good consistency with the Beckman in detecting tPSA, which makes it possible to relieve the pressure of clinical interpretation. However, 4 assays using the same %fPSA cut-off may lead to diverse missed diagnosis rates for PCa screening in China.