Comment on standardization for the assay values of serum vitamin B ₁₂ as measured by automated analysis: Measurements using SRM 1955 as a reference material

In cooperation with the Japan Committee for Vitamin Laboratory Standards, the Committee on Nutrition of the Japan Society of Clinical Chemistry has issued a statement on the standardization in Japan of serum vitamin B₁₂ measurements (sum of methycobalamin, adenosylcobalamin, hydroxocobalamin, and cyanocobalamin). The intra-laboratory and inter-method variations presented in the automated assays (i.e., Access, Centaur, and Elecsys) confounded the diagnosis of hypovitaminosis B₁₂, and required standardization from the clinical and nutritional laboratories. Based on the study conducted by the World Health Organization (WHO), the National Institute of Standards and Technology (NIST) developed a Standard Reference Material (SRM 1955). The SRM 1955 was composed of three levels, the assigned values of which were 217 pg/mL for level 1, 488 pg/mL for level 2, and 474 pg/mL for level 3. Using values by Access related to the assigned value of SRM 1955 as a temporary comparison value, we investigated the possibility of standardization for the assay values of serum vitamin B₁₂ as measured by the automated assays; this became necessary because the isotopic assay, Quantaphase, the established comparison method worldwide, is not sanctioned for use in Japan. When the observed values of SRM 1955 and 46 patient sera by Centaur and Elecsys were compared with the comparison values by Access, the values by SRM 1955 (levels 2 and 3) were plotted within a 95% prediction interval of the regression line. Because the regression equation was not applicable to the lower level of 217 pg/mL, the observed value of level 1 determined by Elecsys was plotted outside the lower 95% prediction limit. In addition, the normalized residuals obtained from level 2 by Centaur and level 3 by Elecsys were both within ± 3.0 enable us to conclude that the SRM 1955 had a physicochemical characterization similar to that of native serum. In the measurements of the 46 patient sera by Centaur and Elecsys, we corrected their observed values for the assigned value of SRM 1955, and compared those corrected values with the comparison values by Access. The biases between the comparison values and the values before and after correction were calculated. After the correction for SRM 1955, the biases were reduced in Centaur but increased in Elecsys. In the measurement by Centaur, the magnitudes of biases still present in the corrected values were at a level high enough to produce a misleading evaluation of hypovitaminosis B₁₂. Based on the above observations, we concluded that a traceability chain to the comparison method should be established in the automated assays by using SRM 1955 as a reference material. We requested the preparation of an additional level of SRM 1955 having an assigned value of around 1000 pg/mL.