| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Department of Biotechnology, University of Turku, Tykistökatu 6A, FIN-20520 Turku, Finland
aAuthor for correspondence. Fax 358-2-333-8050; e-mail susann.eriksson{at}utu.fi.
To the Editor:
In a recent Letter to the Editor, Panteghini (1) commented on our report of a blood component that interferes with immunoassays measuring cardiac troponin I (cTnI) if antibodies against epitopes in the central part of the molecule are used (2). We agree that the selection of the standard material used for recovery experiments can be of critical importance. The tissue-derived ternary troponin complex (from HyTest Ltd.) may or may not be the final solution to cTnI standardization, but it certainly represents a comparatively well-studied and widely used standard material. From our experience, it closely reflects the immunochemical characteristics of the circulating endogenous cTnI (mostly in complex with troponin C). Still, recognizing the possibility for deviating performance, we did the recovery experiments with both endogenous cTnI from patients with acute coronary syndrome (ACS) and the tissue-derived ternary complex and obtained essentially similar results. Our intentional and systematic approach using the two sources of cTnI is far from the "indirect considerations" implied by Panteghini in his letter.
When it comes to the diluents used, throughout the study we used a single diluent originally optimized to achieve the best stability of the ternary complex and the best sample-like performance as assessed by our in-house techniques. At no point did we use any other diluents provided by, e.g., commercial assay manufacturers. We are well aware of the risks and pitfalls in switching between different standard diluents, and in our experience this is particularly true for cTnI. Our main point was not that the mean recovery was <100% in serum and EDTA-plasma samples compared with our diluent. We particularly emphasized the divergent recoveries in individual samples, such as samples 3 and 10, and in several samples in the other cohorts in which the analytical recoveries were as low as 0.8%, whereas the mean recoveries in these cohorts were 80.0–84.1%. Among our staff, serum and plasma samples obtained from several healthy young individuals repeatedly showed low recoveries. One could, of course, argue that the tissue-derived material is rapidly cleaved in such samples so as to lose its immunoreactivity. However, the assays that use antibodies against mid-fragment epitopes, which typically have excellent stability, were the ones most affected by the inhibitory phenomenon. In contrast, the assay design providing additional sandwiching possibilities by use of epitopes outside the central part of the molecule gave up to 10-fold increases in the detected cTnI.
We agree with Panteghini (1) that the number of clinical patient samples in our study was small; 108 samples from 16 ACS patients were analyzed. However, the results from the recovery experiments done on the two large cohorts added further weight to these observations. Samples from two ACS patients showed dramatically different performance with the new monoclonal antibody (Mabs 1 + 3/Mabs 2 + 4) assay. The agreement among the three assays with antibodies against mid-fragment epitopes was obvious. It was also clear that the differences were not primarily dependent on the admittedly moderate analytical sensitivity of the assays; we observed >10-fold differences in results obtained by our two similarly calibrated and approximately equally sensitive assays. As pointed out in our report, this kind of interference can be made less consequential in terms of classifying patients by counteracting its effect with an improved analytical sensitivity. Thus, if the cTnI of a sample is, e.g., 99% inhibited from recognition by an assay that can still reliably detect the remaining 1%, the patient would be correctly classified as having myocardial damage. We believe that from an analytical point of view this is not acceptable. However, the most critical aspect to consider is how frequently do minor increases, whether within the early hours after chest pain or later, remain undetectable because of nonoptimal selection of cTnI epitopes.
The lack of final identification of the interfering component is indisputably a limitation of our study. We are currently working on this, and in this endeavor the use of recovery experiments, such as those described in our report, and the use of differently designed cTnI assays are indispensable and reliable tools in monitoring the interfering factor. The main objective of our initial report was to describe a novel and common type of interference that has not previously been described in the literature and to draw the attention of clinical chemists and cardiologists to the risk of obtaining false-negative cTnI results. At the same time, a central objective was to provide a solution whereby this interference is alleviated or, preferably, completely obviated. At present, numerous cTnI assays in clinical use are based on antibodies against the stable mid-fragment region, according to the current recommendation. This recommendation may eventually have to be revised to accommodate the novel and common source of interference that we have described.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
