PERFORMANCE OF A DIRECT, IMMUNOSEPARATION BASED LDL-CHOLESTEROL METHOD COMPARED TO FRIEDEWALD CALCULATION AND A POLYVINYL SULFATE PRECIPITATION METHOD

The analytical performance of a direct, immunoseparation based LDL-cho lesterol method (Genzyme Corporation) was evaluated on an ELAN analyse r (Merck), and compared with the performance of routinely used methods (LDL-cholesterol estimated by the Friedewald equation, and LDL-choles terol obtained after polyvinyl sulphate precipitation). Within-day coe fficients of variation (CVs) were 0.79 to 2.51% for immunoseparation b ased LDL-cholesterol; the between-day CVs varied between 2.62 and 3.89 %, i.e. within the recommended National Cholesterol Education Program (NCEP) goal of < 4%. A method comparison study, according to the Natio nal Committee for Clinical Laboratory Standards (NCCLS) EP9-P guidelin es, was performed using fasting normo- and hypertriacylglycerolaemic a s well as cholestatic sera. In fresh normotriacylglycerolaemic sera im munoseparation based LDL-cholesterol (y) and Friedewald LDL-cholestero l (x) values were identical as slope and intercept of the Passing and Bablok regression equation were not significantly different from one a nd zero, respectively (y = 1.006 x -0.107; N = 45). In contrast, immun oseparation based LDL-cholesterol (y) differed significantly from poly vinyl sulphate LDL-cholesterol (x) results (y = 0.922 x +0.234; N = 10 3). Freezing normotriacylglycerolaemic sera (three weeks, -20 degrees C) resulted in a negative bias of -5.8% for the immunoseparation based LDL-cholesterol method, and in a positive bias of +5.3% for the polyv inyl sulphate method, compared to fresh specimens. Immunoseparated LDL -cholesterol was completely recovered up to at least 37.84 mmol/l seru m triacylglycerols. We conclude that the immunoseparation based LDL-ch olesterol method is a practical, not technically demanding technique w ell applicable within routine clinical laboratories. The method shows a markedly improved analytical precision in comparison to current rout ine methods, and hence has potential to decrease total test imprecisio n. The immunoseparation based LDL-cholesterol method produces results identical to those obtained by Friedewald in healthy blood donors, and above all overcomes a major pitfall of the Friedewald equation enabli ng LDL-cholesterol measurements in hypertriacylglycerolaemic sera. Its acceptance should improve the reliability of LDL-cholesterol testing and improve clinical decision making.