Wentworth S, McBride J A, Walker W H
Department of Laboratory Medicine, Henderson General Hospital, Hamilton, Ontario, Canada.
Clin Chem. 1994 Apr;40(4):537-40.
We evaluated a chemiluminescence receptor assay for vitamin B12 in serum (Magic Lite; Ciba Corning Diagnostics), in which an acridinium ester label is used with magnetic particle separation. Within- and between-batch precisions were generally acceptable, except at low analyte concentrations. The reference range determined from 104 elective preoperative patients was 120-610 pmol/L, compared with 150-590 pmol/L for our in-house radioligand-binding assay. Magic Lite discriminated between normal and abnormal results as effectively as the in-house method when local reference ranges were applied. Magic Lite demonstrated a negative bias at low analyte concentrations and was unable to detect any vitamin B12 in two B12-deficient patients. Assay accuracy--judged from analytical recovery and comparisons with the in-house method and two other radioassay kits (Quantaphase, Bio-Rad Labs., and Immophase, Ciba Corning Diagnostics)--was poor at low B12 concentrations when the manufacturer's recommended two-point calibration was used. This problem was partially corrected by using a full set of calibrators.
我们评估了一种用于检测血清中维生素B12的化学发光受体分析方法（Magic Lite；汽巴康宁诊断公司），该方法使用吖啶酯标记并结合磁颗粒分离技术。批内和批间精密度总体上可以接受，但在低分析物浓度时除外。从104例择期手术前患者中确定的参考范围为120 - 610 pmol/L，而我们内部的放射配体结合分析方法的参考范围为150 - 590 pmol/L。当应用本地参考范围时，Magic Lite区分正常和异常结果的效果与内部方法一样有效。Magic Lite在低分析物浓度时显示出负偏差，并且无法检测到两名维生素B12缺乏患者的任何维生素B12。根据分析回收率以及与内部方法和另外两种放射分析试剂盒（Quantaphase，伯乐实验室，以及Immophase，汽巴康宁诊断公司）的比较判断，当使用制造商推荐的两点校准时，在低B12浓度下分析准确性较差。通过使用全套校准品，这个问题得到了部分纠正。
