The interfering component in cardiac troponin I immunoassays: need for further experimental evidence.

To the Editor: I read with great interest the report by Eriksson et al. (1 ), discussing the existence of a frequently occurring blood component that interferes with immunoassays measuring cardiac troponin I (cTnI) by use of antibodies against epitopes in the central part of the molecule. As a main indicator of the assay accuracy, the authors studied the analytical recovery of a purified preparation of human cardiac ternary TnI-TnT-TnC complex (from HyTest Ltd., Turku, Finland) added to different samples obtained from healthy individuals and from patients with acute coronary syndrome, using assays that recognize different cTnI epitopes. By definition, recovery refers to the ability of an analytical method to measure an analyte correctly when a known amount of it is added to authentic samples (2 ). Estimation of recovery can therefore be an effective means of obtaining accuracy information because it tests whether the method can measure the analyte in the presence of all other compounds contained in the matrices of authentic samples. Unfortunately, there were some pitfalls in the recovery experiments performed by Eriksson et al. (1 ) as the criterion for demonstrating the presence of a circulating inhibitory component. They did not consider the characteristics of the material they add to the biological fluids. Addition of pure analyte, such as the native, tissue-derived troponin complex, in a complex matrix in vitro is obviously an artificial approach. One cannot be sure that the physical or chemical state of the cTnI in the HyTest preparation, e.g., its solubility, is the same as that for cTnI in vivo. The purification procedure can lead to partial alteration of cTnI structure with the consequence that the probability of significant effects by the complex matrix becomes high, which might affect the immunoassay reaction (3, 4). Recently we performed recovery experiments using a similar material to test the reactivity to various cTnI forms in a newly introduced cTnI immunoassay (5 ). Although the cTnI recovery was good [mean (SE), 92.2 (10.1)%; four dilutions] when the stock solution of the material was serially diluted with appropriate buffer, the recovery decreased [73.7 (7.2)%; four dilutions] when the material was diluted with a commercially available cTnI-free diluent (cat. no. 8TFS; HyTest). In this case, we simply considered that the physicochemical properties of the material might differ significantly according to the different diluents used. As in other studies (6 ), we have also seen problems in the stability of the material. After thawing, we noticed a significant loss in the specific activity within a few hours. This could be an additional confounding factor. In their discussion, Eriksson et al. (1 ) recognized that the “inhibitory” effect was “somewhat more prominent with tissue-derived ternary troponin complex (calibration material) than with endogenous cTnI forms and more evident in EDTA plasma than in serum”; thus, they indirectly considered the possible influence of the type of added cTnI and of the sample matrix on the obtained results (1 ). Another limitation of their study was that the authors support their conclusions by claiming higher sensitivity in the early phase of infarction by use of a cTnI immunoassay supplementing a monoclonal antibody with an epitope in the N-terminal region of cTnI and one with an epitope in the C-terminal region to the midfragment cTnI antibodies. Apart from the anecdotal report of only two cases showing a different behavior, the use for comparison of three cTnI assays that previously have shown only moderate performance for precision and sensitivity at low-range cTnI concentrations should be considered inadequate (7, 8). I agree that a comparison study may be the only practical way to definitively assess the accuracy of the different analytical approaches evaluated for cTnI measurement, but method-comparison studies should compare results obtained with the new proposed analytical approach with those obtained with a referencequality immunoassay for which, in addition to the use of antibodies against the central part of cTnI, excellent performance for analytical sensitivity has been documented (9, 10). As also reported in the Information for Authors of this journal (11 ), it is desirable to test 100–200 different samples from patients with acute coronary syndrome who have been selected to present a range of cTnI values that includes those likely to be encountered in routine application, i.e., samples collected during development of minor and major myocardial injuries, comprising very early and late phases after symptom onset. Difference plots are recommended to show dependence of the differences on cTnI concentrations. In conclusion, as one of the coauthors of the IFCC recommendations on the use of antibodies that “preferably recognize epitopes that are located in the stable part of the cTnI molecule and are not affected by complex formation and other in vivo modifications” (12 ), I concur with Eriksson et al. (1 ) on the importance of obtaining detailed and incontrovertible information on the possible occurrence in the blood of an interfering factor that may negatively affect cTnI measurements obtained with many commercial assays, whose manufacturers have endorsed the above-mentioned recommendation (13 ). However, more and possibly definitive experimental evidence should be collected that permit identification and isolation of this component, which should then be used to directly test the interference in cTnI immunoassays.