doi: 10.1161/01.CIR.0000135581.79161.35
(Circulation. 2004;110:104-106.)
© 2004 American Heart Association, Inc.
Editorial |
Acute Coronary Syndrome Biomarkers
The Need for More Adequate Reporting
From the Mayo Clinic, Rochester, Minn (A.S.J.), and University of Heidelberg, Heidelberg, Germany (H.K.).
Correspondence to Allan S. Jaffe, MD, Mayo Clinic, Department of Laboratory Medicine & Pathology, 200 First St SW, 16th Floor, Rochester, MN 55905. E-mail jaffe.allan{at}mayo.edu
Key Words: Editorials • coronary disease • troponin
Improvements in our understanding of the pathophysiology of atherosclerosis and acute coronary syndromes (ACS) have led to an "explosion" in the development of assays of blood biomarkers to characterize these processes and to predict prognosis. These research assays are performed in a single-center research setting and often use plasma specimens collected in highly selected study populations on which a multitude of biomarkers has been tested in the past. All of these novel markers seem to add prognostic information to established risk indices of ACS. How can so many different analytes all be so predictive?
One reason may be that the diagnostic tools under investigation are optimized with respect to their characteristics and discriminator limits to gain optimal predictive power in the very specific study cohort. Should these tools and their decision limits be applied to a chest pain population with different characteristics, however, specificity may become a significant problem markedly affecting accuracy. Furthermore, application of these tools in a routine setting not allowing such careful sample preparation or optimization procedures may markedly reduce sensitivity.
Another reason, and we shall concentrate on this point here, is that the presently available markers are used in a less than optimal manner. However, if the lack of optimal use of the present makers is not made explicit, it may distort both the importance of the new marker and how to use established markers. Many recent reports have, in our estimation, done this but have failed to acknowledge it overtly in the manuscript, perhaps leading readers to miss this important consideration. The informed but busy clinician may be less able to discern these issues. As investigators who have worked over many years to establish reasonable criteria for the use of troponin makers, we have become concerned that the lack of clarity about these issues is not only leading to confusion about the new markers but has also further confounded clinicians about how to use the troponin assays on which we now rely. Accordingly, the purpose of this editorial is to sensitize authors, readers, and reviewers to these important issues.
Use of Inappropriately High Cut-Off Values for Troponin
Initially, some companies were reluctant to use very low cut-off values because they worried that clinicians might lose faith in the assays if they started to see a large increase in the number of elevations detected compared with creatine kinase-MB. Thus, at first companies pushed the use of values like the "myocardial infarction cut-off or the receiver-operator curve–determined value which was, when data driven, often the level that equated troponin values with creatine kinase-MB values, which at the time was the gold standard. In addition, the initial troponin assays lacked the sensitivity to use very low cut off values. However, assays have improved and, spurred by the European Society of Cardiology/American College of Cardiology redefinition of acute myocardial infarction guidelines,1 many groups began to use lower values. It has been clear since 1992 that any detectable level of troponin has prognostic significance in patients with acute coronary syndromes.2–8 These values were therefore used in studies that attempted to assess the impact of therapy.5,6 In a recent analysis, the Fast Revascularization during InStability in Coronary artery disease (FRISC) and Global Utilization of Streptokinase and tPA for Occluded arteries (GUSTO) groups quantitated the effects of differences in the cut-off values for cardiac troponin T in patients with ACS based on serial samples.8 The sensitivity of troponin T for detection of death at 30 days varied from 0.92 to 0.85 to 0.73 when the values for the 99th percentile, the 10% coefficient of variability value, and the receiver-operator curve value were used, respectively. For death plus myocardial infarction, the values were 0.89, 0.81, and 0.66, respectively, and for death at 1 year, the values were 0.87, 0.79, and 0.68, respectively. The corresponding odds ratios for these values for death or myocardial infarction at 30 days were 3.1, 2.6, and 2.2. Although the data for cardiac troponin I assay vary, the principles expounded are similar.3,5
Thus, when studies evaluate a new marker in patients with ACS using the receiver-operator curve value, they reduce the sensitivity of troponin. This approach alone might allow a putative marker to appear to provide additional prognostic significance in instances when, had troponin been used optimally, it may not have had a significant additive benefit.9–11 At times it is difficult to discern that a higher cut-off value is being used. We believe that when such high values are used, the authors, reviewers, and editors have a responsibility to make sure that there is an adequate discussion of why that cut-off value was used and what the implications for using a more sensitive and more predictive cut-off value for troponin might have been. In most cases, the data to answer these questions are already available within the data set.
Use of Insensitive Assays
All troponin assays are not the same. If one uses an assay incapable of measuring the low levels known to impart prognostic significance, a novel marker might well appear to be predictive, whereas if a more sensitive assay were used, that effect might be attenuated or even obliterated. The International Federation of Clinical Chemistry and Laboratory Medicine has recently published a nearly complete tabulation of assays and their performances at lower levels.12 If an assay lacks precision, using it at lower levels incurs the risk of having false-positive results. This is the rationale for the use of the 10% coefficient of variability level rather than the 99th percentile.13 Nonetheless, many assays, including some point-of-care assays, are either less sensitive or less precise, resulting in the suggestion that higher values be used.14–16 This biases the results in a similar fashion to a high cut-off value. We again believe the authors, reviewers, and editors must insist on a discussion of the adequacy of the assay and, if it is less sensitive than others, the potential influence on the results.
Use of Only the Initial Sample for Analysis
This is another way in which troponin measurements can be used less than optimally.7,17,18 Often, it is hard to know how many samples were included for troponin measurements. In some studies, it is even unclear if all the samples acquired were analyzed for troponin. Frequently, how many samples have been obtained/evaluated is stated only in the methods and thus can easily be missed. The data are clear and have shown for many years that the predictive accuracy of troponin measurements requires more than 1 measurement.1,3,8 Thus, markers that rise earlier than troponin might be more predictive than a solitary admission troponin value but not more predictive if an additional troponin value is included in the analysis. In some situations where immediate events are the outcomes of interest, such a strategy might be reasonable. However, when the end points are not immediate, comparing new markers to what is now standard, ie, serial troponin measurements, is preferred. If this is not done, we believe that authors, reviewers, and editors must assure that the rationale for not doing it and the potential impact of additional sampling should at least be discussed. Thinking about this issue may be particularly important if the time to presentation of different groups varies. In that situation, one might report differences that are predicated based on the time to presentation, rather than on the markers themselves.
Finally, we have been concerned by the paucity of data provided about sample handling when samples from a study have been saved for months or even years. If the sample handling has not been pristine, troponin values may go down.19,20 Although there can be degradation of other analytes as well, there are situations where values can go up and perhaps situations where, depending on the diluent and preservation, one subset rises and another subset falls.21 This could lead to both false-positive and false-negative results. Again, we should make sure that authors address these critical issues.
To facilitate these recommendations for all articles involving biomarkers, we would suggest consideration of the following whenever a paper concerns biochemical markers:
- Journals might consider appointing members of their staffs to specifically scrutinize manuscripts in this area.
- The timing and number of samples used in any analysis of blood biomarkers should be clear and there should be a discussion concerning the adequacy of the timing and numbers of samples used in the study.
- Troponin sampling and cut off values that are the basis for assessments of prognosis in patients with ACS should be used in the most optimal manner.
- The characteristics, including the sensitivity and precision, of the all assays used should be reported. Such reporting should be facilitated by the publication of the International Federation of Clinical Chemistry and Laboratory Medicine paper on the sensitivity and low-end precision of troponin assays.14
- The appropriate cut-off values for the analytes being evaluated and all covariates should be listed. If alternative cut-off values are used, the rationale for their use and their effect on the data should be discussed.
- At times, data will not be available to clarify all of these issues. If so, the implications of alternative approaches in sampling and cut-off values should be included in the discussion section.
- Information on the characteristics of the novel assays used and their variability, precision, and normal values should be included and commented on when important for the proper use of the marker. When preserved samples are used, data concerning their reliability over time should be included.
- Whenever possible, the incremental value of a given approach over and above other analytes should be calculated.
This is an important responsibility for both authors and journals. Unless we begin to pay attention to these important considerations, there will be continued confusion about the value of newer markers and increased difficulty in deciding how best to use markers such as troponin on which we now rely so heavily.
Footnotes
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
References
1. Alpert JS, Thygesen K, Antman E, et al. Myocardial infarction redefined: a consensus document of The Joint European Society of Cardiology/Am College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol. 2000; 36: 959–969.
2. Hamm CW, Ravkilde J, Gerhardt W, et al. The prognostic value of serum troponin T in unstable angina. N Engl J Med. 1992; 327: 146–150.[Abstract]
3. Ottani F, Galvani M, Nicolini FA, et al. Elevated cardiac troponin levels predict the risk of adverse outcome in patients with acute coronary syndromes. Am Heart J. 2000; 140: 917–927.[CrossRef][Medline] [Order article via Infotrieve]
4. Hamm CW. Cardiac biomarkers for rapid evaluation of chest pain. Circulation. 2001; 104: 1454–1456.
5. Antman EM, Tanasijevic MJ, Thompson B, et al. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. N Engl J Med. 1996; 335: 1342–1349.
6. Aviles RJ, Askari AT, Lindahl B, et al. Troponin T levels in patients with acute coronary syndromes, with or without renal dysfunction. N Engl J Med. 2002; 346: 2047–2052.
7. Morrow DA, Cannon CP, Rifai N, et al. Ability of minor elevations of troponins I and T to predict benefit from an early invasive strategy in patients with unstable angina and non-ST elevation myocardial infarction: results from a randomized trial. JAMA. 2001; 286: 2405–2412.
8. James S, Armstrong P, Califf R, et al. Troponin T levels and risk of 30-day outcomes in patients with the acute coronary syndrome: prospective verification in the GUSTO-IV trial. Am J Med. 2003; 115: 178–184.[CrossRef][Medline] [Order article via Infotrieve]
9. Brennan M, Penn MS, Van Lente F, et al. Prognostic value of myeloperoxidase in patients with chest pain. N Engl J Med. 2003; 349: 1595–1604.
10. Heeschen C, Dimmeler S, Hamm CW, et al. Soluble CD40 ligand in acute coronary syndromes. N Engl J Med. 2003; 348: 1104–1111.
11. Newby LK, Storrow AB, Gibler WB, et al. Bedside multimarker testing for risk stratification in chest pain units: the chest pain evaluation by creatine kinase-MB, myoglobin, and troponin I (CHECKMATE) study. Circulation. 2001; 103: 1832–1837.
12. Panteghini M, Pagani F, Yeo KTJ, et al. Evaluation of the imprecision at low-range concentrations of the assays for cardiac troponin determination. Clin Chem. 2004; 50: 327–330.
13. Jaffe A. Caveat emptor. Am J Med. 2003; 115: 241–244.[CrossRef][Medline] [Order article via Infotrieve]
14. McCord J, Nowak RM, McCullough PA, et al. Ninety-minute exclusion of acute myocardial infarction by use of quantitative point-of-care testing of myoglobin and troponin I. Circulation. 2001; 104: 1483–1488.
15. de Lemos JA, Morrow MA, Sabatine MS, et al, Association between plasma levels of monocyte chemoattractant protein-1 and long-term clinical outcomes in patients with acute coronary syndromes. Circulation. 2003; 107: 690–695.
16. Kwong RY, Schussheim AE, Rekhraj S. et al. Detecting acute coronary syndrome in the emergency department with cardiac magnetic resonance imaging. Circulation. 2003; 107: 531–537.
17. Baldus S, Heeschen C, Meinertz T, et al. Myeloperoxidase serum levels predict risk in patients with acute coronary syndromes. Circulation. 2003; 108: 1440–1445.
18. Heeschen C, Dimmeler S, Fichtlscherer S, et al. Prognostic value of placental growth factor in patients with acute chest pain. JAMA. 2004; 291: 435–441.
19. Bodor GS, Porter S, Landt Y, et al. Development of monoclonal antibodies for an assay of cardiac troponin-I and preliminary results in suspected cases of myocardial infarction. Clin Chem. 38: 2203–2214.
20. Muller-Bardorff M, Hallermayer K, Schroder A, et al. Improved troponin T ELISA specific for cardiac troponin T isoform: assay development and analytical and clinical validation. Clin Chem. 1997; 43: 458–466.
21. Eisenberg PR, Jaffe AS, Stump DC, et al. Validity of enzyme-linked immunosorbent assays of cross-linked fibrin degradation products as a measure of clot lysis. Circulation. 1990; 84: 1159–1168.
This article has been cited by other articles:
 |
|
 |
|
  J. P. McConnell and A. S. Jaffe The Spin Stops Here: Inhibition of Lipoprotein-Associated Phospholipase A2-- A Promising Target but a Negative Initial Trial? Clin. Chem., January 1, 2009; 55(1): 21 - 23. [Full Text] [PDF]
|
|
|
|
|
|
A. S. Jaffe Key Issues in the Developing Synergism between Cardiovascular Imaging and Biomarkers Clin. Chem., September 1, 2008; 54(9): 1432 - 1442. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 NACB WRITING GROUP MEMBERS, D. A. Morrow, C. P. Cannon, R. L. Jesse, L. K. Newby, J. Ravkilde, A. B. Storrow, A. H.B. Wu, and R. H. Christenson National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines: Clinical Characteristics and Utilization of Biochemical Markers in Acute Coronary Syndromes Circulation, April 3, 2007; 115(13): e356 - e375. [Full Text] [PDF]
|
|
|
|
|
|
NACB WRITING GROUP MEMBERS, D. A. Morrow, C. P. Cannon, R. L. Jesse, L. K. Newby, J. Ravkilde, A. B. Storrow, A. H.B. Wu, R. H. Christenson, NACB COMMITTEE MEMBERS, et al. National Academy of Clinical Biochemistry Laboratory Medicine Practice Guidelines: Clinical Characteristics and Utilization of Biochemical Markers in Acute Coronary Syndromes Clin. Chem., April 1, 2007; 53(4): 552 - 574. [Full Text] [PDF]
|
|
|
|
|
|
V. L. Roger, J. M. Killian, S. A. Weston, A. S. Jaffe, J. Kors, P. J. Santrach, H. Tunstall-Pedoe, and S. J. Jacobsen Redefinition of Myocardial Infarction: Prospective Evaluation in the Community Circulation, August 22, 2006; 114(8): 790 - 797. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Jaffe, L. Babuin, and F. S. Apple Biomarkers in Acute Cardiac Disease: The Present and the Future J. Am. Coll. Cardiol., July 4, 2006; 48(1): 1 - 11. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Giannitsis and H. A. Katus Markers of myocardial ischaemia: reply Eur. Heart J., March 2, 2006; 27(6): 759 - 759. [Full Text] [PDF]
|
|
|
|
|
|
F. S. Apple, A. H.B. Wu, J. Mair, J. Ravkilde, M. Panteghini, J. Tate, F. Pagani, R. H. Christenson, M. Mockel, O. Danne, et al. Future Biomarkers for Detection of Ischemia and Risk Stratification in Acute Coronary Syndrome Clin. Chem., May 1, 2005; 51(5): 810 - 824. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. L. Roger Invited Commentary: Will Heart Attacks Be "Gone with the Century"? Am. J. Epidemiol., December 15, 2004; 160(12): 1147 - 1149. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||