Significant reduction of the bias among commercial immunoassays for lipoprotein(a) after use of a uniform calibrator.

Despite the increasing interest in the measurements of lipoprotein(a) (Lp(a) in serum or plasma, at present there is no effective standardization for Lp(a) assays; the main problems to solve are represented either by the lack of a suitable primary standard or by the absence of a reliable and widely available reference method. A first step is hence the uniformity of calibration of different immunoassays. We calibrated three commercial immunoassays for Lp(a) (enzyme linked immunosorbent assay (ELISA), latex-enhanced immunonephelometric assay (LINA), and immunonephelometric assay (INA) with either proprietary standards or purified Lp(a) material obtained with a rapid and simple procedure. Final results of purified Lp(a) calibration were reported in terms of protein Lp(a) mass whereas we were able to quantify the exact protein concentration of our purified lipoprotein. The uniformity of the calibration of the different assays led to a significant improvement of regression slopes (from 1.88 to 0.90 ELISA vs. LINA, from 1.45 to 0.95 ELISA vs. INA and from 1.27 to 0.96 INA vs. LINA) and correlation coefficients (from 0.990 to 0.994 ELISA vs. LINA, from 0.987 to 0.990 ELISA vs. INA and from 0.985 to 0.987 INA vs. LINA). Furthermore, the significant differences among Lp(a) values obtained after calibration with proprietary standards were minimized, becoming non-significant in two out of three cases. In conclusion, we demonstrated that a better agreement of Lp(a) values obtained with different commercial assays could be simply reached by uniformity of calibration and by employing standards with values accurately measured.