Background: The first commercial direct automated immunoassay specific for 25-OH vitamin D(3) (25-OH-D(3)) was recently introduced for use on Roche Diagnostics immunoassay analyzers. We assessed the analytical performance of the Elecsys 25-OH-D(3) assay on a Roche Modular E 170 analyzer.
Methods: The Elecsys 25-OH-D(3) assay is a direct electrochemiluminescence immunoassay for human serum or plasma. It is a competitive assay in which the binding protein of vitamin D is inactivated during incubation. The assay employs a polyclonal antibody directed against 25-OH vitamin D(3). We compared the 25-OH-D(3) assay to assays performed with RIA, HPLC, and liquid chromatography-tandem mass spectrometry (LC-MS/MS).
Results: At concentrations of 48, 76, and 124 nmol/L, within-run CVs were 5.1%, 3.1%, and 7.1% and total CVs were 12.1%, 7.4%, and 10.6%, respectively. A comparison of Elecsys 25-OH vitamin D(3) with RIA yielded the regression equation: Elecsys = 1.114 x RIA - 6.15 (S(y|x) = 15.7 nmol/L; n = 163). The corresponding equation with HPLC was: Elecsys = 1.077 x HPLC + 5.442 (S(y|x) = 13.9 nmol/L; n = 67) and with LC-MS/MS: Elecsys = 0.887 x LC-MS/MS + 5.046 (S(y|x) =12.4 nmol/L; n = 64). Contrary to LC-MS/MS, with the cutoff of 50 nmol/L (deficiency vs normal), approximately 10% of samples were misclassified as normal with RIA and Elecsys. Plasma samples were observed to have markedly higher concentrations than serum samples.
Conclusions: The Elecsys concentrations of 25-OH-D(3) were in good overall agreement with those determined with LC-MS/MS and RIA. However, large between-method variation was observed in individual patient samples. Use of serum rather than plasma is preferred owing to the higher results observed with plasma samples.