Evaluation of an ELISA test for determination of the serum transferrin receptor. Demonstration of discordance between results obtained with two methods

We undertook this study to evaluate a recently introduced ELISA kit for det ermining serum transferrin receptor (TfR) concentration (TfR(R), Ramco Labo ratories, Inc.), to produce reference values for healthy adults, and to com pare the results with another commercially available reagent system. The me an (SD) recovery of added TfR was 88% (6%). In dilution studies, the ratio between the measured and expected values was 0.98 (0.11). The intra-assay a nd interassay coefficients of variation were from 5% to 7% and from 6% to 9 % in a physiological and a supraphysiological concentration range, respecti vely, and from 13% to 16% in a subnormal concentration range. In healthy ad ults between 20 and 60 years of age, we observed no age- or sex-related dif ferences in TfR values. Thus, the same reference interval, 3.0-8.2 mg l(-1) , may be used for this population. The correlation between the results obta ined with the Ramco TfR(R) test and the Amgen Diagnostics Clinigen(TM) test was satisfactory (r = 0.79). The Ramco TfR test produced higher values (Tf R(R) = 0.40 (-0.45 - 1.25) + 1.46 (1.16 - 1.75)* Clinigen(TM)). The number of samples that fell within the same concentration interval with both metho ds (low, normal or high in relation to the respective reference interval) w as only 45% (27/60). The Ramco TfR(R) lest had fewer values falsely suggest ing iron deficiency than the Clinigen(TM) test. Serum TfR methods need to b e uniformly standardized.