High lipoprotein(a) results in overestimation of BQ-based low-density lipoprotein-cholesterol measurement

Most conventional low-density lipoprotein cholesterol (LDL-C) assays are standardized by the CDC reference procedure of beta-quantification (BQ) method, which involves density-based ultracentrifugation to separate very low-density lipoprotein (VLDL) and LDL. Apolipoprotein (apo) (a) covalently binds to apoB, altering lipoprotein density thereby affecting this separation. We evaluated the impact of lipoprotein(a) [Lp(a)] on LDL-C assay accuracy. The distribution of apo(a) between VLDL and LDL was examined by ultracentrifugation and gel permeation high-performance liquid chromatography (GP-HPLC) analysis in three subjects. LDL-C values were compared between the conventional assays calculated by Sampson's formula (S-LDL-C) and direct homogeneous assay (D-LDL-C) and the GP-HPLC analysis (GP-LDL-C) for 1,560 human samples, and the influence of Lp(a) concentration was analyzed. Apo(a) was detected in the LDL fraction by ultracentrifugation and in both VLDL and LDL fractions by HPLC. Strong correlations were observed between S-LDL-C, D-LDL-C, and GP-LDL-C, but significant differences existed among the methods (P < 0.001). The differences between S-LDL-C or D-LDL-C and GP-LDL-C positively correlated with Lp(a) (r = 0.302 and 0.321, respectively, P < 0.001). When Lp(a) exceeded 30 mg/dl, these differences were significantly larger (n = 139). The difference between S-LDL-C and D-LDL-C was not significantly correlated with Lp(a) (P = 0.774). Conversely, VLDL-C measured by Sampson's method showed a negative correlation with Lp(a) (r = -0.262, P < 0.001) Apo(a) increases VLDL density, causing it to be recovered in the LDL fraction during density-based separation. Consequently, Lp(a) causes concentration-dependent overestimation of BQ-based LDL-C, and underestimation of VLDL-C.