Résumé
The objective of this study was to compare automated immunoenzymatic assay (IEA) with an isotopic reference method for determination of plasma folate, erythrocyte folate, and plasma vitamin B12 levels.
Analytical performance including within-run and between-run imprecision was assessed on the automated IEA on the Access2 (Beckman Coulter, France). Folate and B12 levels determined routinely with a radioimmunoassay (RIA) (Simultrac, Diasorin, Italy) using EDTA samples (n=100) were compared to the Access2 IEA. To evaluate pre-analytical conditions, EDTA and heparinized plasma (n=56) from simultaneous samplings were used to compare folate and B12 levels via RIA and the Access2. In addition, discordant results were assayed by IEA on the Architect i2000 (Abbott, France).
Coefficient of variation for folate and B12 measurements on the Access2 ranged between 1.5% and 4.1%, respectively, for within-run assays, while between-run imprecision ranged from 4.8% to 9.3%, respectively. Comparison of RIA (x) and IEA on the Access2 (y) using EDTA samples showed a good correlation for plasma folate (y = 0.94x + 1.09; r2 = 0.83), acceptable correlation for erythrocyte folate (y = 0.65x + 137; r2 = 0.57) and discrepancy for B12 (y = 0.67x + 211; r2 = 0.37). This discordance for vitamin B12 was not observed when comparing RIA (x) and IEA on the Architect i2000 (y) (y = 0.74x + 40; r2 = 0.92) and disappeared when using heparinized samples for the Access2 (y = 0.85x + 7.05; r2 = 0.90).
Automated IEA appears to be a good alternative to RIA for the determination of folate and vitamin B12 but requires the use of an adequate anticoagulant.