Comparison of three methods for total homocysteine plasma determination

Hyperhomocysteinemia is a well established risk factor for atherothrombotic disease, and the request for homocysteine determinations and the number of laboratories that need to perform this assay to assess individual risk profile is increasing. Different methods to evaluate homocysteine plasma levels are at present available. In the present study three methods, an in-house high-pressure liquid chromatographic (HPLC) method (considered as reference method) and two commercial immunoassays, an enzyme-linked immunoassay (EIA) and an automated fluorescence polarization immunoassay (FPIA), were used to measure homocysteine plasma levels in 100 samples. The median of homocysteine plasma levels obtained by HPLC was 9.0 micromol/L (range 4.2-23.0); the median of values obtained by EIA and FPIA were 10.6 micromol/L (range 3.3-21.5) and 9.6 micromol/L (4.8-20.2), respectively. The FPIA method showed the lowest within-run and between-run coefficients of variation (3.6% and 4.1%, respectively). There was a significant correlation between EIA and HPLC (r=0.81; p<0.0001), and between FPIA and HPLC (r=0.85; p<0.0001). The Bland-Altman analysis showed that FPIA agreed best with HPLC; EIA displayed a relatively wide scatter of difference data points. The present results indicate that the technological characteristics of the FPIA assay make this method suitable for the determination of Hcy in clinical laboratories.