| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Laboratorio Analisi Chimico Cliniche 1, Azienda Ospedaliera Spedali Civili, 25125 Brescia, Italy
aAuthor for correspondence. Fax 39-030-3995369; e-mail panteghi{at}bshosp.osp.unibs.it.
To the Editor:
Cardiac troponin has been designated as the preferred biomarker for diagnosis of myocardial infarction (MI) (1). Previously published data, however, confirm the large diversity among cardiac troponin assays with respect to important analytical characteristics, including assay standardization, antibody specificity, interferences, and assay imprecision, and underscore the need for improved cardiac troponin assays (2)(3). Results obtained with more recently released next-generation assays show that the newer assays indeed have substantially improved analytical performance (4)(5). The aim of this study was to evaluate one of these next-generation cardiac troponin I (cTnI) assays, performed on the Aio!TM immunoanalyzer (Innotrac Diagnostics Oy), by defining key performance characteristics, including detection limit, linearity on dilution, imprecision, reference interval, and cutoff for MI diagnosis.
The assay is based on "all-in-one" dry chemistry technology, in which all of the reagents are precoated in assay cups, and time-resolved fluorometric detection, with a total analysis time of <20 min (6). An eight-point factory-constructed calibration curve is provided on a bar code with each reagent lot, and the instrument-specific calibration adjustment is performed by running the cups of the appropriate calibration pen. A purified preparation of human cardiac ternary troponin I-troponin T-troponin C complex (HyTest Ltd.) is used as calibration antigen. The antibody configuration of the assay, adding a monoclonal antibody with an epitope in the N-terminal region of cTnI (amino acid residues 20–35) and one with an epitope in the C-terminal region (amino acid residues 185–200) to the mid-fragment cTnI antibodies (epitopes in the region of amino acid residues 35–55 and 80–95), has recently been described, and the potential ramifications of this have been discussed (7)(8).
The Aio! analyzer was handled strictly according to the manufacturer’s instructions. Unless otherwise stated, fresh serum was used as sample. The minimum detectable cTnI concentration was assessed by 20 replicate measurements of the cTnI-free diluent in a single run and defined as the cTnI value corresponding to a signal 3 SD greater than the mean found for this sample (9). In the linearity study, five cTnI-rich serum specimens (native cTnI concentrations of 4.3, 9.2, 27.6, 46.9, and 79.5 µg/L) were serially diluted with serum pools having undetectable cTnI concentrations, i.e., lower than the detection limit of the Aio! assay, or with the instrument buffer solution. The undiluted sample and four separate dilutions (3:4, 1:2, 1:4, and 1:8) were assayed in duplicate in the same analytical run. The curve obtained was tested for linearity as suggested by Burnett (10). After demonstration of linearity, linear regression analysis of the data was performed, and correlation coefficients (r) were calculated. A recovery study was also performed. For the imprecision study, seven serum pools were prepared and stored at –80 °C until used. Two replicates/specimens were analyzed per run, and one run per day for 20 days was performed, using two reagent lots and calibrations (3). Using ANOVA method, we calculated the total CV at different concentrations and used the CV values reported for the seven pools to construct the imprecision profile for the method (3).
To establish reference values, we a priori selected 150 apparently healthy individuals (75 women and 75 men; median age, 60 years; range, 23–89 years), using exclusion criteria as suggested in IFCC recommendations on the theory of reference values (11). Serum cTnI was measured, and the 99th percentile of the value distribution was calculated by nonparametric determination of percentiles. The procedures followed were in accordance with the current revision of the Helsinki Declaration.
The detection limit for the assay was 0.006 µg/L. In the linearity studies, the hypothesis of a linear fit was accepted for all examined samples (P >0.11), and r values were >0.9989 for dilution with a negative serum pool and >0.9978 for dilution with buffer. All mean recoveries were within ± 10% of the native concentrations. The following results for assay imprecision were obtained (total CV and mean cTnI concentration): CV = 14% at 0.019 µg/L; CV = 8.2% at 0.044 µg/L; CV = 6.3% at 0.065 µg/L; CV = 6.3% at 0.086 µg/L; CV = 8.1% at 0.14 µg/L; CV = 6.0% at 0.22 µg/L; and CV = 4.3% at 0.39 µg/L. Because a total CV 
10% at the MI decision limit is recommended, the cTnI concentration corresponding to this analytical imprecision was determined from the intercept of the total CV (y axis) equal to 10% on the imprecision profile curve (Fig. 1
). The lowest cTnI concentration yielding a CV of 10% was 0.036 µg/L. This value is therefore recommended as a de facto cutoff for detection of myocardial necrosis. In the group of apparently healthy individuals, 97% of the cTnI values were <0.010 µg/L, whereas only four individuals (two males and two females) had measurable cTnI concentrations, i.e., 0.010, 0.014, 0.016, and 0.029 µg/L. The calculated 99th percentile of the cTnI value distribution was 0.015 µg/L. The evaluated assay was close to meeting the 10% CV at the 99th percentile value, with 10% CV/99th percentile ratio of 2.4.
|
In conclusion, the Aio! second-generation cTnI assay has improved performance compared with data published earlier for the corresponding first-generation assay (12).
Acknowledgments
This study was funded by A. De Mori Strumenti, national distributor of Innotrac Diagnostics products. To avoid a potential conflict of interest, the company was in no way involved in the collection of data or its analysis or interpretation.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. Panteghini Selection of Antibodies and Epitopes for Cardiac Troponin Immunoassays: Should We Revise Our Evidence-Based Beliefs? Clin. Chem., May 1, 2005; 51(5): 803 - 804. [Full Text] [PDF]
|
|
|
|
|
|
S. Eriksson, T. Ilva, C. Becker, J. Lund, P. Porela, K. Pulkki, L.-M. Voipio-Pulkki, and K. Pettersson Comparison of Cardiac Troponin I Immunoassays Variably Affected by Circulating Autoantibodies Clin. Chem., May 1, 2005; 51(5): 848 - 855. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Eriksson, H. Halenius, K. Pulkki, J. Hellman, and K. Pettersson Negative Interference in Cardiac Troponin I Immunoassays by Circulating Troponin Autoantibodies Clin. Chem., May 1, 2005; 51(5): 839 - 847. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
