Problems with the serum vitamin B12 assay.

Quality control trials have shown that, in routine practice, serum vitamin B12 estimations vary so much from laboratory to laboratory that serious confusion can result irrespective of whether microbiological or radioassay methods are used. Even experienced centres find the assay too insensitive and non-specific for a low level to be used as the sole criterion of vitamin B12 deficiency; the haemopoietic and biochemical sequelae of vitamin B12 deficiency also correlate poorly with the serum level. These basic difficulties with the assay seem to stem from the pattern of vitamin B12 (or cobalamin) binding in serum. Human beings are unique in having virtually all of their cobalamin attached to an apparently functionless binder, transcobalamin I. It is therefore not surprising that the serum cobalamin is such a poor predictor of cobalamin deficiency. The metabolically important serum binder is transcobalamin II and deficiency of this protein causes a potentially lethal megaloblastic anaemia even though the serum cobalamin level is normal. Tissue cobalamin depletion with normal serum levels also occurs after nitrous oxide inhalation and in certain inborn errors of metabolism.