| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Institut für Klinische Chemie und Pathobiochemie,
2
Medizinische Klinik I, and
3
Klinik für Thorax-, Herz- und Gefässchirurgie, Universitätsklinikum der RWTH Aachen, 52074 Aachen, Germany.
a Address correspondence to this author at: Institut für Klinische Chemie und Pathobiochemie, Universitätsklinikum der RWTH Aachen, Pauwelsstrasse 30, 52074 Aachen, Germany. Fax 49-0-241-8888-512; e-mail hugo.stiegler{at}post.rwth-aachen.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Serum and heparin-plasma samples were drawn simultaneously from 100 patients (50 patients with acute coronary syndrome and 50 patients after open heart surgery) and measured on three different analytical systems, two for determination of cardiac troponin I (cTnI; Abbott AxSYM and Bayer ACS:Centaur) and one for cardiac troponin T (cTnT; Roche Elecsys cTnT STAT). Serum samples were reanalyzed after a second centrifugation to assess the influence of incomplete serum separation.
Results: Mean results (± 95% confidence interval) in heparin-plasma compared with serum were 101% ± 2% (AxSYM cTnI), 94% ± 3% (ACS:Centaur cTnI), and 99% ± 3% (Elecsys cTnT). Differences >20% were seen in 11% of results on the ACS:Centaur, 9% of results on Elecsys cTnT, and 2% of results on the AxSYM. For the Elecsys cTnT assay, the magnitude of the difference between serum and plasma was independent of the absolute concentration and confined to individual samples, and was reversed by treatment with heparinase. A second centrifugation had no effect on serum results by any of the assays.
Conclusion: The concentrations of troponins measured in heparin-plasma are markedly lower than in serum in some cases.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A recent study by Gerhardt et al. (4) published in this journal demonstrated that results for cardiac troponin T (cTnT) are 15% lower in heparin-plasma than in serum. On the basis of these results, Roche Diagnostics, manufacturer of the cTnT assay, advised customers not to use heparin-plasma in their assay. Many laboratories have relied on heparin-plasma for determination of cTnT, and several clinical studies, including the recently published PRISM study (5), were conducted with heparin-plasma. Because at our central laboratory heparin-plasma as well as serum is used routinely for determination of cardiac troponins, we investigated the influence of incomplete serum separation and of heparin-plasma on cardiac troponin concentrations in three common analytical test systems under routine conditions.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All patients gave informed consent. Results were not reported to the primary care physicians and therefore had no influence on the diagnostic work-up or clinical decisions.
blood sampling and sample handling
The study design is shown in Fig. 1
. Briefly, venous blood was collected after phlebotomy
simultaneously into serum separator tubes (cat. no. 04.1935;
Sarstedt) and lithium heparinate tubes (15 IU/mL of whole
blood; 
25 IU/mL of plasma based on a hematocrit of 0.40; cat. no.
01.1608.001; Sarstedt). Immediately after collection, sampling tubes
were gently mixed by inverting the tubes five to eight times. Within
10–15 min after venipuncture, both tubes were centrifuged at
3000g for 10 min. After centrifugation, both the serum
and the heparin-plasma were divided into three aliquots of 500 µL and
measured immediately on all three analytical systems. After initial
measurements for cTnI and cTnT, the serum aliquots underwent a second
centrifugation at 3000g for 10 min and were remeasured to
assess the influence of incomplete serum separation.
|
For measurement of the activated partial thromboplastin time (aPTT), blood was collected into tubes containing sodium citrate as the anticoagulant (cat. no. 05.1071.001; Sarstedt).
determination of cTnI AND cTnT
All samples were assayed within 1 h after collection. cTnI
was measured by a two-step microparticle enzyme immunoassay on the
AxSYM analyzer (Abbott Laboratories, Abbott Park, IL) as well as by an
acridinium ester-based two-step chemiluminescence immunoassay on the
ACS:Centaur analyzer (Bayer Diagnostics, Tarrytown, NY). cTnT was
measured by a third-generation electrochemiluminescence
immunoassay on the Elecsys 1010 analyzer (Roche Diagnostics,
Mannheim, Germany). All assays were performed according to the
manufacturers’ instructions. Results >50 µg/L on the AxSYM
and ACS:Centaur were retested after a 10-fold internal dilution. No
tested sample was found to be above the measuring range of the Elecsys
assay.
The clinical cutoff values for acute myocardial infarction, as stated by the manufacturers in their respective reagent packages, are 2.0 µg/L for the Abbott AxSYM cTnI assay, 1.5 µg/L for the Bayer ACS:Centaur cTnI assay, and 0.1 µg/L for the Roche Elecsys cTnT assay.
analytical performance
Three different concentrations for serum as well as for plasma
were prepared by diluting patient serum or plasma samples containing
high cardiac troponin concentrations with cTnI- and cTnT-negative serum
or plasma pools. Aliquots of each concentration were frozen at
-70 °C and were thawed immediately before analysis. For
determination of within-run imprecision, all three concentrations were
analyzed 21 times in one analytical run. For determination of
between-day imprecision, all three concentrations were analyzed in 11
analytical runs on 11 different days. Measurements for evaluation of
analytical performance were carried out with the same assay lot and
without calibration changes. The total CV was calculated for
each concentration. The analytical sensitivity, or lower detection
limit, of the tested assays, as stated by the manufacturers in their
respective reagent packages was 0.3 µg/L for the Abbott AxSYM cTnI
assay, 0.15 µg/L for the Bayer ACS:Centaur cTnI assay, and 0.01
µg/L for the Roche Elecsys cTnT assay.
reversal of heparin binding
If results were >20% lower in heparin-plasma than in serum,
samples were remeasured after pretreatment with Hepzyme (Dade Behring).
Hepzyme is a lyophilized preparation of purified bacterial heparinase 1
(EC 4.2.2.7) that cleaves the heparin molecule at multiple sites
(6). Heparin degradation was carried out according to the
manufacturer’s instruction. Briefly, a sample of frozen (-70 °C)
heparin-plasma was brought to 37 °C in a water bath. After complete
thawing, the sample was recentrifuged at 3000g for 10 min,
and 1 mL of the plasma was sequentially neutralized by dissolving
lyophilized purified bacterial heparinase 1 in the sample.
Samples were incubated at room temperature for 15 min, and measurements
were carried out immediately after incubation.
measurement of aPTT
To evaluate the influence of therapeutic or bolus doses of heparin
on troponin concentrations in heparin-plasma, the aPTT was measured and
compared with the difference in results for heparin-plasma. The aPTT
was determined automatically on an Amelung AMGA CS-400 coagulometer
(Sigma Diagnostics) using Dade Actin FS activated PTT reagent (Dade
Behring).
statistical analysis
The correlation between serum and plasma as well as the
correlation between the two serum measurements was evaluated by the
statistical procedure developed by Passing and Bablock (7).
The percentage difference was evaluated by the Bland-Altman method
(8), as modified by Pollock et al. (9). This
procedure involves plotting the mean of the results for serum and
heparin-plasma on the x axis against the percentage
difference of both results (% relative difference) on the y
axis.
Spearman correlation coefficients for comparisons of aPTT values with differences in results between heparin-plasma and serum were calculated by a commercially available computer program (SPSS 7.5 for Windows; SPSS).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Generally, at comparable relative concentrations the
imprecision of the Roche Elecsys cTnT assay was twofold lower
than the imprecision of the Abbott AxSYM cTnI method or the Bayer
ACS:Centaur cTnI method.
|
A total of 100 serum and corresponding plasma samples were analyzed. None of the tested samples was below the detection limit of any assay. The initial serum (plasma) results were 0.31–648.05 µg/L (0.30–648.55 µg/L) for the AxSYM, 0.15–591.33 µg/L (0.15–590.35 µg/L) for the ACS:Centaur, and 0.010–18.530 µg/L (0.010–13.350 µg/L) for the Elecsys assay.
Comparison of results for serum and heparin-plasma by Passing-Bablock regression analysis revealed the following: AxSYM plasma cTnI = 0.999 (serum) + 0.000 µg/L (r = 0.998); ACS:Centaur plasma cTnI = 0.958 (serum) - 0.017 µg/L (r = 0.949); Elecsys plasma cTnT = 0.996 (serum) + 0.002 µg/L (r = 0.984). The 95% confidence intervals for the slope and intercept were as follows: 0.988–1.016 and -0.026 to 0.052 for the AxSYM; 0.924–0.985 and -0.110 to 0.019 for the ACS:Centaur; and 0.961–1.024 and -0.002 to 0.012 for the Elecsys. Mean results (± 95% confidence interval) in heparin-plasma compared with serum were 101% ± 2% for the AxSYM; 94% ± 3% for the ACS:Centaur; and 99% ± 3% for the Elecsys.
Grouped-pair analysis and Bland-Altman difference plots revealed
samples with results >20% lower in heparin-plasma than in serum in
all three assays: 11% of all results on the ACS:Centaur; 9% of all
results on the Elecsys; and 2% of all results on the AxSYM. Results
are shown in detail in Table 2
and Fig. 2
. For the Abbott AxSYM cTnI assay, only two samples with
concentrations near the detection limits were detected. Because
differences in results between heparin-plasma and serum at such low
concentrations might be affected by the higher imprecision of
the analytical system and might also be clinically without any
relevance, we excluded cases with results near the detection limit
(
0.50 µg/L for the AxSYM; 0.40 µg/L for the ACS:Centaur; and
0.05 µg/L for the Elecsys) from further data analysis. Table 3
lists all samples with results >20% lower in heparin-plasma
than in serum of the total of 100 samples tested. As shown, the lower
results were confined to individual samples and mostly to only one test
system.
|
|
|
For the Elecsys cTnT assay, the difference in results affected only samples well above the clinical cutoff value for acute myocardial infarction. We could identify two samples with differences >40% (samples 2 and 8) that were from patients in the acute phase of myocardial infarction (<12 h after onset of chest pain). Of the remaining six samples, four were from the group of patients after open heart surgery, whereas two samples were from patients in the subacute phase of myocardial infarction (>12 h after onset of chest pain).
For the Bayer ACS:Centaur cTnI assay, the difference in results was mainly concentration dependent, with the largest differences in the lower concentration range. All affected samples except one were from patients with suspected or diagnosed acute coronary syndrome.
Values for the aPTT ranged from 24 s to >150 s (median, 45 s). For all three test systems, no correlation could be detected between the aPTT and troponin results in heparin-plasma (corresponding rS = 0.015, -0.091, and -0.154 for the Abbott AxSYM, ACS:Centaur, and Elecsys, respectively), assuming that for the investigated patients there was no additional influence of therapeutic heparin concentrations or bolus doses of heparin on the troponin results in heparin-plasma.
To further clarify, if a lower result was caused by an in vitro heparin
influence, the affected samples were remeasured after pretreatment with
heparinase. The results are shown in Table 4
. After treatment with heparinase, the heparin-plasma/serum
ratios for cTnT, which had been 34–78%, were now 83–101%.
In contrast to cTnT, the cTnI results obtained by the Bayer ACS:Centaur
were worse after treatment with heparinase.
|
Serum results obtained after a second centrifugation were 0.30–650.13 µg/L for the AxSYM, 0.15–595.51 µg/L for the ACS:Centaur, and 0.010–18.340 µg/L for the Elecsys. Comparison of results between the two serum measurements by Passing-Bablock regression analysis revealed the following: AxSYM serum 2 cTnI = 1.009 (serum 1) - 0.004 µg/L (r = 0.998); ACS:Centaur serum 2 cTnI = 0.981 (serum 1) - 0.002 µg/L (r = 1.0); Elecsys serum 2 cTnT = 1.035 (serum 1) + 0.000 µg/L (r = 0.999). The mean results in serum (± 95% confidence interval) after centrifugation compared with the initial serum result were: 101% ± 2% (AxSYM cTnI), 98% ± 1% (ACS:Centaur cTnI), and 103% ± 2% (Elecsys cTnT). None of the results obtained after the second centrifugation were >20% lower than the initial result for the same serum sample.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most patients with cardiac disease are anticoagulated during hospitalization. When serum is collected, full clot retraction from tubes without preservatives can take up to 15 min (2). Clots can continue to form even after the sample has been centrifuged and the serum is placed into the analytical instrument. When this occurs, the instrument can be blocked by fibrinous material, which leads to a flagged sample run. This will delay analysis, because of the additional time needed for recentrifugation and reanalysis of the sample, but will not lead to an erroneous result. On the other hand, small amounts of fibrin can cause nonspecific binding of the antibody, producing falsely increased troponin results without any flagging by the test instrument (1). In our study, we did not detect any increased troponin concentrations attributable to incomplete separation of serum, assuming that the total incidence is <1% of all tested samples.
Nevertheless, many laboratories prefer plasma rather than serum to shorten turnaround time. In addition, the National Academy of Clinical Biochemistry recommends the use of plasma as the specimen of choice for stat analysis of cardiac markers (2), but some analytical systems have the limitation that plasma anticoagulated with heparin, sodium citrate, or EDTA cannot be used because of random or systematic deviations (3). For EDTA tubes, troponin released as a ternary (cTnT-I-C) or binary (cTnI-C) complex will degrade to free subunits because the ionized calcium needed to maintain this complex is removed by chelation of the metal ions (14). Troponin assays that do not exhibit an equimolar response between complexed and free subunits will produce significant biases between serum and EDTA-plasma. Heparin is not believed to disrupt complexes and therefore is unlikely to cause differences in results between serum and plasma (2).
Many laboratories have relied on heparin-plasma for rapid determination of cTnT, as have several clinical studies, including the PRISM trial (5). A recent study by Gerhardt et al. (4) published in this journal demonstrated that results obtained with heparin-plasma are 15% lower than those obtained with serum. On the basis of those results, Roche Diagnostics, manufacturer of the cTnT assay, strongly advised customers not to use heparin-plasma in their assay.
In contrast to the findings by Gerhardt et al. (4), in our study the mean cTnT result for heparin-plasma was 99% of the value obtained for serum. Nevertheless, in 9 of 100 tested samples, the result for heparin-plasma was >20% lower that the result for serum, independent of the absolute cTnT concentration. A possible explanation for these obvious discrepancies might be the use of different blood collection systems with different heparin concentrations. In contrast to the study by Gerhardt et al. (4), we used only one type of lithium heparinate tube, with a total heparin concentration of 15 IU/mL of whole blood (20–30 IU/mL of plasma), whereas Gerhardt et al. used tubes with higher heparin concentrations. The results of both studies thus might not be directly comparable. In addition, a different selection of patients might have influenced the results of both studies. Because Gerhardt et al. (4) included mainly patients in the acute phase of myocardial infarction, the different distribution of cTnT isoforms in the early vs the late phase of myocardial infarction may explain the differences.
An effect of therapeutic heparin concentrations (usually 1 IU/mL in
patients with acute myocardial infarction) seems unlikely, considering
the large amount of heparin (25 IU/mL of plasma) in the sampling
tubes (15). This is also supported by the lack of
correlation between the aPTT as an equivalent for therapeutic heparin
concentrations and loss of troponin recovery in heparin-plasma.
In our study, treatment with heparinase reversed the differences between heparin-plasma and serum seen in the Elecsys cTnT assay but had no influence on measurements on the Bayer ACS:Centaur. This implies, in combination with the concentration-dependent differences for the Bayer test, that in this case the lower result is more likely the effect of higher imprecision than the direct influence of heparin on the cTnI molecule detected by this assay. For cTnT, the lower results obtained for heparin-plasma could be explained by conformational changes of the antigen induced by direct interactions between the positively charged (pI 5.1) cTnT molecule and the negatively charged heparin. Such complexes can interfere with the antibody-antigen interaction, as shown previously by Katrukha et al. (16).
In this study, we demonstrated that the use of heparin as an in vitro anticoagulant at a total concentration of 15 IU/mL of whole blood leads to significant biases in at least two of the investigated assays because of a certain percentage of samples with lower results compared with serum measurements. Therefore, at present we cannot recommend the use of heparin-plasma for determination of cardiac troponins for at least two of the investigated assays. The lower results obtained for heparin-plasma may affect clinical decisions or risk stratification in patients with suspected acute coronary syndrome. No influence of incomplete separation of serum on troponin concentrations could be detected in our patient cohort.
Additional studies are needed to clarify the exact mechanism by which heparin as an in vitro anticoagulant causes such interactions and why they are confined to individual samples. The manufacturers of cardiac troponin assays should be aware of this potential problem and work on the elimination of this interference. Other commercially available assays for determination of cardiac troponins, including bedside tests, should be checked for a possible disturbance by heparin-plasma.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  D. C Gaze and P. O Collinson Multiple molecular forms of circulating cardiac troponin: analytical and clinical significance Ann Clin Biochem, July 1, 2008; 45(4): 349 - 355. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. S. Apple and M. M. Murakami Cardiac Troponin and Creatine Kinase MB Monitoring during In-Hospital Myocardial Reinfarction Clin. Chem., February 1, 2005; 51(2): 460 - 463. [Full Text] [PDF]
|
|
|
|
|
|
R. F. Salamonsen, H.-G. Schneider, M. Bailey, and A. J. Taylor Cardiac Troponin I Concentrations, but Not Electrocardiographic Results, Predict an Extended Hospital Stay after Coronary Artery Bypass Graft Surgery Clin. Chem., January 1, 2005; 51(1): 40 - 46. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. S. Apple and M. M. Murakami Serum 99th Percentile Reference Cutoffs for Seven Cardiac Troponin Assays Clin. Chem., August 1, 2004; 50(8): 1477 - 1479. [Full Text] [PDF]
|
|
|
|
|
|
F. S. Apple, H. E. Quist, P. J. Doyle, A. P. Otto, and M. M. Murakami Plasma 99th Percentile Reference Limits for Cardiac Troponin and Creatine Kinase MB Mass for Use with European Society of Cardiology/American College of Cardiology Consensus Recommendations Clin. Chem., August 1, 2003; 49(8): 1331 - 1336. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Anderwald, G. Brabant, E. Bernroider, R. Horn, A. Brehm, W. Waldhausl, and M. Roden Insulin-Dependent Modulation of Plasma Ghrelin and Leptin Concentrations Is Less Pronounced in Type 2 Diabetic Patients Diabetes, July 1, 2003; 52(7): 1792 - 1798. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Panteghini Acute Coronary Syndrome: Biochemical Strategies in the Troponin Era Chest, October 1, 2002; 122(4): 1428 - 1435. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. J. Kim, O. F. Laterza, K. G. Hock, J. F. Pierson-Perry, D. M. Kaminski, M. Mesguich, F. Braconnier, R. Zimmermann, M. Zaninotto, M. Plebani, et al. Performance of a Revised Cardiac Troponin Method That Minimizes Interferences from Heterophilic Antibodies Clin. Chem., July 1, 2002; 48(7): 1028 - 1034. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Peetz, G. Hafner, and K. J. Lackner Analytical Characteristics of the AxSYM Cardiac Troponin I and Creatine Kinase MB Assays Clin. Chem., July 1, 2002; 48(7): 1110 - 1111. [Full Text] [PDF]
|
|
|
|
|
|
M. Groschl, R. Wagner, J. Dotsch, W. Rascher, and M. Rauh Preanalytical Influences on the Measurement of Ghrelin Clin. Chem., July 1, 2002; 48(7): 1114 - 1116. [Full Text] [PDF]
|
|
|
|
|
|
D. Uettwiller-Geiger, A. H.B. Wu, F. S. Apple, A. W. Jevans, P. Venge, M. D. Olson, C. Darte, D. L. Woodrum, S. Roberts, and S. Chan Multicenter Evaluation of an Automated Assay for Troponin I Clin. Chem., June 1, 2002; 48(6): 869 - 876. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Cerutti, L. Corsini, R. Finotto, and C. Perazzi Comparison of Cardiac Troponin I in Serum and Heparin Plasma with the Dimension RxL Assay Clin. Chem., May 1, 2002; 48(5): 790 - 791. [Full Text] [PDF]
|
|
|
|
|
|
G. P. Armstrong, A. N. Barker, H. Patel, and H. H. Hart Reference Interval for Troponin I on the ACS:Centaur Assay: A Recommendation Based on the Recent Redefinition of Myocardial Infarction Clin. Chem., January 1, 2002; 48(1): 198 - 199. [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Quinn and D. J. Moliterno Troponins in Acute Coronary Syndromes: More TACTICS for an Early Invasive Strategy JAMA, November 21, 2001; 286(19): 2461 - 2462. [Full Text] [PDF]
|
|
|
|
|
|
G. H. White and P. A. Tideman Increased Troponin T in a Patient with Dermatomyositis Clin. Chem., June 1, 2001; 47(6): 1130 - 1131. [Full Text] [PDF]
|
|
|
|
|
|
P. Venge, B. Lindahl, and L. Wallentin New Generation Cardiac Troponin I Assay for the Access Immunoassay System Clin. Chem., May 1, 2001; 47(5): 959 - 961. [Full Text] [PDF]
|
|
|
|
|
|
F. S. Apple and A. H.B. Wu Myocardial Infarction Redefined: Role of Cardiac Troponin Testing Clin. Chem., March 1, 2001; 47(3): 377 - 379. [Full Text] [PDF]
|
|
|
|
|
|
K. F. Buechler, K. K. Nakamura, W. Gerhardt, G. Nordin, A. Isaksson, S. Haglund, E. Gustavsson, M. Muller-Bardorf, and H. A. Katus Diagnostic accuracy of an agarose gel electrophoretic method in multiple sclerosis. Clin. Chem., January 1, 2001; 47(1): 144 - 144. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
