A multicenter study on the precision and accuracy of homogeneous assays for LDL-cholesterol: comparison with a beta-quantification method using fresh serum obtained from non-diseased and diseased subjects.

BACKGROUND: Homogeneous assays for low-density lipoprotein-cholesterol (LDL-C) have good precision and are pretreatment-free procedures. However, their accuracies have been questioned, especially in diseased subjects. In this study, we aimed to verify whether LDL-C levels determined by homogeneous assays [LDL-C (H)] agree with those determined by a beta-quantification method [LDL-C (BQ)] in fresh clinical samples. METHODS: We determined LDL-C levels in 49 non-diseased and 124 diseased subjects whose triglyceride (TG) levels were less than 11.29 mmol/L (1000 mg/dL) using 12 homogeneous assays and a BQ method simultaneously. RESULTS: In total, 30.6% of non-diseased subjects and 46.0% of diseased subjects were in the postprandial state. The maximum inter- and intra-assay CVs were 1.8% and 1.5%, and 8 reagents had a CV of 1.0% or less. The mean bias ranged from -0.5% to 1.8% for non-diseased subjects and from -0.7% to 1.6% for diseased subjects. For non-diseased subjects, all but one reagent achieved the National Cholesterol Education Program (NCEP) total error requirement in more than 90% of samples. However, for diseased subjects, the number of reagents that met this requirement was low. With some reagents, LDL-C (H) was higher than LDL-C (BQ), especially in subjects with hypertriglyceridemia. While for other reagents, the difference between the two methods was not associated with hypertriglyceridemia except for type I (n = 2) and type III hyperlipidemia (n = 1). Postprandial sampling was not the main factor for discordant results. CONCLUSIONS: LDL-C (H) agrees with LDL-C (BQ) in non-diseased subjects, but exhibits positive bias for subjects with hypertriglyceridemia in diseased subjects for some reagents.