25-hydroxy-Vitamin D status: limitations in comparison and clinical interpretation of serum-levels across different assay methods.

Background: Over the last decade, clinical interest to evaluate human 25-hydroxy-vitamin D (25[OH]D) serum levels has increased exponentially. In the present study, four chemiluminescence immunoassays (CLIA), one radioimmunoassy (RIA), and one high performance liquid chromatography (HPLC) method were compared and also with the liquid chromatography-tandem mass spectrometry (LC-MS/MS) method in view of 25(OH)D serum level determination.
Methods: For the method comparison, blood samples from 133 consecutive patients were prospectively collected. All participants gave written informed consent for their blood samples to be used in this study. They came to the Department of Nuclear Medicine of the Central Hospital Steyr (Austria) for osteodensidometric measurement as part of their preventive medical check-up. Pearson's correlation coefficients, Bland-Altman plots, and paired t-tests were calculated. Assay-specific reference ranges were considered using blood samples from persons with normal parathormone, calcium, and total-protein values (n = 97).
Results: The highest correlation was between the HPLC and the LC-MS/MS method (r = 0.96). The lowest correlation was between the cobas Vitamin D3 assay (Roche) and any of the evaluated assays (r = 0.46 - 0.63). Bland-Altman plots revealed a big negative mean bias in three assays (cobas Vitamin D3 assay [Roche]: -22.8; DiaSorin LIAISON [25[OH]D total CLIA [Diasorin]: -18.4; Diasorin 25[OH]D125 I RIA [Diasorin]: -23.8 [nmol/L]) and a much smaller positive mean bias in the other assays (ClinRep complete 25[OH]D2/D3 HPLC kit [Recipe]: 2.7; ADVIA Centaur Vitamin D total assay [Siemens]: 8.2; IDS total vitamin D assay [Immunodiagnostic Systems]: 12.1 [nmol/L]) compared to the LC-MS/MS method. Meanwhile, the manufacturer has withdrawn the cobas Vitamin D3 assay from the market. Conclusions: Poor antibody specificity with cross-reactivity to other vitamin D metabolites, incomplete extraction of the 25(OH)D analyte from the vitamin D-binding protein (DBP), and confounding matrix substances such as lipids could be potential reasons for the unacceptable performance of the cobas Vitamin D3 assay (Roche) and also the significant differences in the 25(OH)D determination between various assays. Standardization and harmonization of 25(OH)D measurements are therefore urgently needed. The widespread introduction of well standardized assays in clinical laboratories is the challenge in the next years. (Clin. Lab. 2014;60:1541-1550. DOI: 10.7754/Clin.Lab.2014.131114)