Impact of assay parameters on the accuracy of free PSA test: Source and stability of calibrator, calibration curve fitting, and level of total PSA in the serum

Abstract

The measurement of PSA is recommended for men over 50 years of age for screening of prostate cancer. However, proper differentiation of prostate cancer from benign prostate hyperplasia (BPH) relies on an accurate measurement of free PSA (fPSA) and a correct calculation of percent fPSA. Because of the extremely low concentration of fPSA in the serum, any slight deviation from its true value may produce large errors in percent fPSA calculated. Therefore, we undertook a study examining carefully those parameters of the fPSA assay which might affect the fPSA determination. We found that the integrity of the calibrator, the computer curve-fitting program selected, the source of the calibrator, and the total PSA or fPSA + PSA complexes (tPSA) concentration of the specimen all had an impact on the accuracy of the fPSA value assayed. We found that an examination of the slope of the calibration curve was important to reveal whether the calibrator had or had not been denatured during storage. We also found that the 4-parameter cure fitting program was best suited for plotting the fPSA calibration curve. The calibrator we isolated from LNCaP cells was acceptable for our assay because it had an affinity for the assay antibody very similar to that of serum fPSA. We also determined the effect of tPSA concentration on the fPSA determinations and found that within the concentration range of 4-10 ng/mL the impact on the percent fPSA calculated was not significant. We believe that our assay produces accurate fPSA values when all these assay parameters are well controlled.