RESUMO
BACKGROUND: The accurate measurement of 25-hydoxy vitamin D (25OH-D) in serum has been a challenge for many years. We developed a liquid chromatography tandem mass spectrometry (LC Tandem MS) assay for the quantitative determination of 25OH-D(2) and 25OH-D(3) in serum. The new method was compared with two widely used commercially available immunoassays. METHODS: Sample preparation involved protein precipitation with acetonitrile containing deuterated forms of the target species as internal standards. An API 5000 mass spectrometer coupled with a photoionization source was used for quantitation. The performance of the new LC Tandem MS assay was compared with a radioimmunoassay (RIA, Diasorin) and a chemiluminescence immunoassay (ECLIA, Roche Diagnostics), analysing serum obtained from 152 individuals. RESULTS: Using 100 µl of serum, the LC Tandem MS assay had a limit of quantitation of 1.3 nmol/L for both 25OH-D(2) and 25OH-D(3) with a linear response between 1.3 and 625 nmol/L and accuracy of between 95 and 124%. Intra- and inter-assay precision were ≤7% and ≤4%, respectively. Measurement of 25OH-D levels in 152 serum samples gave run averages of 71, 56 and 62 nmol/L for LC Tandem MS, ECLIA and RIA, respectively. Correlations between the various methods were: LC Tandem MS vs. RIA: r=0.931; LC Tandem MS vs. ECLIA: r=0.784; RIA vs. ECLIA: r=0.787. The LC Tandem MS method had a positive proportional bias of 26% over the RIA, whereas the ECLIA showed variable differences. CONCLUSION: The new LC Tandem MS assay is accurate and precise at physiologically relevant 25OH-D concentrations, and compares favourably with the RIA. In contrast, the ECLIA shows variable bias with the other assays tested.