Abstract

The determination of the total concentration of plasma homocysteine is of interest in a variety of clinical circumstances, especially, in the evaluation of the risk of cardiovascular disease. However, most of the methods available to date, many of them chromatographic, are not well suited for the majority of clinical laboratories. Several automated methods are now or will be, shortly, commercially available. We have compared one of them, the fluorescence polarization immunoassay (FPIA) adapted to the IMx^® analyzer (Abbott Laboratories), with the high-performance liquid chromatography (HPLC) method with fluorescent detection currently used in our laboratory. The results show that the FPIA-IMx^® method is less imprecise and slightly more sensitive than the HPLC. The comparison of 67 clinical plasma specimens indicated that there is a proportional error disagreement between FPIA-IMx^® and HPLC (FPIA=1.19 HPLC + 0.92; confidence region for slope and y-intercept were, respectively, from 1.06 to 1.31 and from−0.06 to 2.32). The nature of this error is not explained by the experiments performed to study the inaccuracy of both methods, which included the investigation of dilution parallelism, analytical recovery and cross-reactivity. The different results of homocysteine concentration obtained with FPIA-IMx^® and HPLC must be taken into account when a change of methodology is under consideration.