Accuracy of serum ferritin determinations in tissue preparations and human serum.

Introduction of the NBSB 80/602 reference preparation for the calibration of ferritin immunoassays has reduced the inter-assay variability, and represents the first important step towards standardisation of ferritin immunoassays. However, our investigations show that comparison between assay results is still impossible, owing to differences in assay methodology and performance, differences in the specificities of the antibodies and antigens used and other possible interfering substances. We conclude that all kits detect mainly the more basic isoferritins in serum, and that all isoferritins (acidic, intermediate and basic) are systematically underestimated. Since we also showed possible immunological differences between reference or kit standards and serum ferritin, we conclude that the present kits have poor accuracy. To diminish inter-assay variability and to increase the accuracy of serum ferritin determinations, a method is needed that detects all basic, intermediate and acidic isoferritins and measures the true ferritin concentration in serum under normal and pathological circumstances. This reference method can be used to evaluate interference and systematic errors in routine methods. The introduction of a reference method, in combination with the NBSB 80/602 human liver reference preparation, is the second important step towards the accurate standardisation of ferritin immunoassays.