Methodological effects on the quantitation of serum ferritin by radio- and enzymoimmunoassays

Serum ferritin is normally quantitated by sensitive radio- or enzymoimmunoassays, which are based either on a competitive reaction between the unlabelled and labelled antigen (RIA), or on a non-competitive reaction between the sample and the labelled antibody (IRMA and ELISA). Both methods appear to be satisfactory for the clinical evaluation of serum ferritin. However, the effect of these methods on the quantitation of serum ferritin has never been carefully studied. This paper describes and compares five different immunoassays for the evaluation of serum ferritin: two competitive RIAs, one non-competitive IRMA and two non-competitive ELISAs. The same anti-ferritin antibody and ferritin standards were used for all the assays. The RIAs were found to be less sensitive than either the ELISAs or the IRMA, and all showed a similar degree of cross-reactivity with ferritin extracted from spleen and HeLa cells. A significant difference between the assays was found in the determination of serum ferritin: in fact the RIAs over-estimated serum ferritin by about 50% more than the IRMA, whereas the ELISAs under-estimated serum ferritin by about 15% less than the IRMA.