Comparison of two commercial nephelometric methods for apoprotein A-I and apoprotein B with standardized apoprotein A-I and B radioimmunoassays.

Normotriglyceridemic and hypertriglyceridemic samples were analyzed for apoproteins A-I and B using the Beckman Array System and the Behring Nephelometer, and the nephelometric values were compared to values obtained by highly standardized radioimmunoassays developed at the Northwest Lipid Research Center. Although the means of the apoA-I values obtained by each method were similar, comparison of sample values by least-squares regression analysis revealed large differences (Sy = 20 mg/dl for Beckman, Sy = 18 mg/dl for Behring) (Sy = standard error of the estimate) regardless of whether the comparison included hypertriglyceridemic samples. For normotriglyceridemic samples, there was good agreement between apoB values obtained by the Behring Nephelometer and those obtained by RIA (r = 0.91, m = 1.03, Sy = 12 mg/dl). However, significantly higher apoB values were obtained on hypertriglyceridemic samples by the Behring Nephelometer. ApoB values for normotriglyceridemic samples obtained by the Beckman System and RIA showed fairly good correlation (r = 0.86, m = 0.71, Sy = 14 mg/dl). However, the nephelometric values for normotriglyceridemic samples averaged 29% lower than those obtained by RIA. This difference could largely be accounted for by the low apoprotein B value assigned to the Beckman calibrator. Significantly lower apoprotein B values were obtained on hypertriglyceridemic samples by the Beckman Nephelometer even after correction for calibration differences. Apoprotein values obtained by nephelometric methods may be inaccurate, particularly if the samples are hypertriglyceridemic.