Critical care and vitamin D status assessment: What about immunoassays and calculated free 25OH-D?

Background: Interpretation of 25OH-D measurement during critical care (CC) may be problematic due to variations of binding protein concentrations (albumin, ALB, and vitamin D binding protein, VDBP). Determination of free 25OH-D concentration may thus be relevant in CC patients. The aim of this observational study was to evaluate effects of an acute hemodilution on vitamin D (VD) status. Methods: Blood samples were obtained before (T1) and after a crystalloid load (T2) administered at anesthesia induction for minor surgery. 25OH-D was measured with LC-MS/MS and with 3 immunoassays (IA): DiaSorin Liaison, IDS iSYS and bioMerieux Vidas. VDBP was measured with the R&D Elisa and ALB on Cobas. Free 25OH-D was calculated using published formula. Accuracy of each 25OH-D IA was calculated as the percentage of IA values within 20% of their respective LC-MS/MS values. Performances of the three AI were compared with LC-MC/MS using Bland-Altman analysis. Results: Twenty adults were included. Compared to T1 values, VDBP, ALB and LC-MS/MS values decreased in parallel by a mean of 23% at T2. IA values decreased less significantly (12, 14 and 15% for Liaison, iSYS and Vidas, respectively). IA-based calculated free 25OH-D significantly increased after dilution, while LC-MS/MS-based free values remained stable. At T1 and T2, bias were demonstrable for all IA. After hemodilution, bias would lead to overestimation for the three IA. Accuracy of IA decreased after dilution. Conclusions: Due to matrix effects, compared to LC-MS/MS, IA results were impacted by hemodilution. In CC patients, LC-MS/MS seems to be the best option to measure 25OH-D. Specific LC-MS/MS method should be developed to measure free 25OH-D. (C) 2014 Elsevier B.V. All rights reserved.