American Heart Journal
Volume 148, Issue 5 , Pages 739-742, November 2004

Evidence-based decision limits for cardiac troponin: Low-level elevation and prognosis

  • David A. Morrow, MD, MPH

      Affiliations

    • TIMI Study Group and Cardiovascular Division, Department of Medicine, Brigham and Women's Hospital, Boston, Mass, USA
    • Corresponding Author InformationReprint requests: David A. Morrow, MD, MPH, Cardiovascular Division, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA.

Article Outline

 

Applied in conjunction with a careful history, physical examination, and interpretation of the electrocardiogram, biomarkers of myocardial injury are a cornerstone of the diagnostic and prognostic assessment of patients presenting with possible acute coronary syndrome (ACS). Cardiac troponin offers clinical sensitivity and specificity that is superior to other available biomarkers of myocardial necrosis and thus is the preferred biomarker for diagnosis and risk assessment in this setting.1, 2 Despite these advantages, integration of troponin into clinical practice has come with substantial growing pains. In particular, uncertainty regarding cut-points and the appropriate clinical response to seemingly frequent abnormal troponin results in patients ultimately determined not to have ACS has created frustration for practitioners. Guidance from expert panels1, 3 and the continued accumulation of clinical data4, 5 have provided a basis for more consistent reporting of decision limits and reassurance regarding the prognostic importance of troponin in patients with suspected ACS. Nevertheless, questions regarding optimal decision limits and the clinical relevance of low-level increases in troponin have not been completely resolved. The report from Pham6 provides additional data addressing these two closely tied issues.

Back to Article Outline

Establishing decision limits for cardiac troponin 

Initially, diagnostic cut-points for myocardial infarction (MI) for each troponin assay were derived by receiver operating characteristic comparison against CK-MB. Subsequent recognition that increases in troponin among patients with normal levels of CK-MB are associated with a higher risk of recurrent events led to the recommendation7 and common reporting of two decision limits for cardiac troponin I (cTnI): a “diagnostic” limit for the definitive diagnosis of MI based on comparison to CK-MB and a lower limit “suggestive” of myocardial injury, or “minor myocardial damage” important to prognosis (Figure 1). Although the practice of reporting two cut-points is still followed in some centers, present guidelines1 developed by a committee from the European Society of Cardiology and American College of Cardiology (ESC/ACC) advocate a single decision limit for diagnosis and risk assessment based on the 99th percentile for each assay in a reference population. This recommendation is grounded on the principle that any reliably detected circulating troponin is abnormal and indicative of myocardial injury.8 To minimize the potential for analytic false-positives at very low concentrations of troponin, these guidelines also require adequate analytic performance (coefficient of variation ≤10%) at the decision limit (Figure 2).

  • View full-size image.
  • Figure 1. 

    Evolution of diagnostic criteria for MI, using cardiac troponin. See text for details. Current recommendations are for a single decision limit at the 99th percentile or, if there is not acceptable analytic precision at this concentration, at the level of the 10% CV. MMD, Minor myocardial damage.

  • View full-size image.
  • Figure 2. 

    Example of a troponin assay with a CV of 10% at 0.1 ng/mL. Analytic performance of each troponin assay is characterized by repeated measurements on the same sample. CV is a measure of the precision of the assay at a given concentration. If the CV were 20% instead of 10% at this concentration, 67% of values from the same sample would fall within the range of 0.08 to 0.12 ng/mL and 95% within the range of 0.06 to 0.14 ng/mL.

These ESC/ACC recommendations have provided a template toward more uniform clinical reporting of troponin results. However, several challenges (anticipated by the committee) have emerged during implementation. First, at the time of this writing, at most one assay achieves the desired level of precision at the 99th percentile.5 Thus, functionally for most available assays, the ESC/ACC decision limit is determined not by the 99th percentile but rather the level of 10% coefficient of variation (CV). Second, because manufacturers of the multiple available assays for cTnI use antibody combinations targeting different epitopes, the values generated for the same sample will usually differ between assays. Clinicians must be cognizant of intermethod variation and should not generalize decision limits established for one assay to others. Third, the ESC/ACC-specified cut-point originates from thoughtful expert consensus and not direct clinical evaluation. Thus, the clinical efficacy of decision limits for risk stratification should also be supported by evaluation in peer-reviewed clinical studies documenting the relation with outcomes.7 Moreover, this evidence-based approach to implementation of decision limits has assumed greater importance as newer assays with lower limits of detection (LLD) have been and continue to be developed.

Back to Article Outline

Low-level elevation 

The term “low-level” elevation has been applied as a descriptive term referring to quantitatively minor increases in the concentration of troponin but has not been defined in relation to specific analytic criteria (eg, LLD, 99th percentile, or 10% CV) and thus has been used variably by different investigators.6, 9, 10 Since introduction of the ESC/ACC definition, many clinicians have had concerns regarding the significance of concentrations of troponin above this decision limit yet below manufacturer-recommended cut-points for MI (which often are still based on comparison to CK-MB). Strong and consistent data from at least 6 studies, including clinical trials and unselected patients with suspected ACS, demonstrate that levels of troponin in this range are independently associated with the risk of death and recurrent ischemic events.6, 9, 10, 11, 12, 13 Collectively, these data provide robust support for the prognostic importance of quantitatively minor elevation above the ESC/ACC decision limit in patients with a clinical history suggestive of ACS.

Emerging data also suggest that in patients with a high clinical probability of ACS increases in troponin below the ESC/ACC decision limit are associated with greater risk of recurrent cardiac events. For example, in the TACTICS-TIMI 18 study, patients with a baseline cTnI >99th percentile (0.1 ng/mL) but below the ESC/ACC decision limit (0.4 ng/mL, 10% CV) were at more than 3-fold higher risk of death or MI (P < .001). Ten percent of patients fell into this range of low-level elevation and would be “misclassified” as low risk if the decision limit of 0.4 ng/mL were applied. Similar findings were evident with cTnT and have been reproduced with the use of another cTnI assay in two separate clinical trials9, 14, 15 as well as in unselected patients with suspected ACS.10 In a cohort of patients with possible ACS stratified according to peak cTnI—negative (less than LLD), low (greater than or equal to LLD to less than ESC/ACC decision limit), intermediate (ESC/ACC decision limit to greater than the manufacturer's suggested MI limit), and high (greater than or equal to suggested MI limit)—6-month mortality rate increased in a graded fashion across categories of cTnI (hazard ratios, 2.5 [95% CI, 1.4 to 4.4] for low, 3.9 [95% CI, 2.3 to 6.8] for intermediate, and 6.1 [95% CI, 4.2 to 8.7] for high concentrations of cTnI).10 In this study, patients with low-level elevation were less likely to have angiographically significant coronary disease (defined as ≥70% stenosis). Nevertheless, their higher risk of death/MI supports the significance of low-level elevation for risk assessment. One may speculate that this finding relates both to incomplete sensitivity of angiography for confirming ACS and to the prognostic importance of myocardial injury from causes other than acute atherothrombosis.

Back to Article Outline

Increases in patients without ACS 

The introduction of troponin and subsequent analytic advances have substantially extended the ability to detect previously unrecognized myocardial injury and thus have facilitated identification of patients with ACS at higher risk for recurrent events as well as concurrently increasing detection of acute and/or chronic myocardial injury in a variety of conditions in which it was not previously appreciated.16, 17 Although the antibodies used in assays for troponin are highly specific to the cardiac isoform, the tissue specificity of cardiac troponin should not be confused with specificity for the mechanism of injury. The diagnosis of MI is established only when there is both biochemical evidence of myocardial necrosis and clinical evidence that the cause is myocardial ischemia. Therefore, clinicians must integrate the available clinical data to discriminate myocardial injury caused by an acute coronary event from other causes of ischemia such as pulmonary embolism or an alternative cause such as myocarditis.1

Recognition of the diverse circumstances in which myocardial injury can occur has raised appropriate questions regarding the implications of elevated troponin outside of ACS. Pham6 evaluated the relation between cTnI and outcomes at 1 year in a varied group of patients without definite ACS. In a cohort selected on the basis of cTnI ≤3.0 ng/mL (2- to 3-fold above the 10% CV) and the absence of angina with electrocardiographic changes or clinical diagnosis of MI at presentation, patients with a peak troponin level ≥1.0 ng/mL (ESC/ACC cut-point) were at significantly higher risk for cardiac death or MI (adjusted HR, 3.4; 95% CI, 1.3 to 9.4). Moreover, each 1-ng/mL increase in cTnI above the LLD (0.4 ng/mL) was associated with an ∼2-fold higher risk of cardiac death or MI. These findings are notable in several regards. First, in this diverse population without ACS, detectable circulating troponin was a strong indicator of poorer long-term outcomes. Second, this association was manifest at concentrations of troponin immediately above the ESC/ACC decision limit. These findings add to accumulating evidence establishing the prognostic implications of myocardial injury in a variety of conditions, including pulmonary embolism, blunt chest trauma, sepsis, heart failure, and cardiomyopathy.3 Although such data speak to the higher risk of these patients, the appropriate therapeutic responses have not been defined. In a few cases, such as fibrinolysis for pulmonary embolism,18 specific treatments have been proposed for patients with elevated troponin; however, strategies using troponin to guide therapy have not yet been prospectively studied outside of the setting of ACS. A strategy of more intensive therapy tied to higher global risk may be reasonable.

Back to Article Outline

Summary 

As the clinical sensitivity of assays for troponin has increased, the potential implications of quantitatively modest (“low-level”) increases in troponin have attained greater relevance. The ESC/ACC decision limit, set at the 99th percentile in conjunction with acceptable precision, is likely to be optimal in most settings because it will minimize analytic false-positives. However, clinicians should recognize that in patients with a high (pretest) probability of ACS, a troponin level greater than the LLD (with some present assays) identifies those at higher risk for adverse outcomes. Thus, at present, the following clinical strategy appears reasonable. In most settings, the ESC/ACC decision limit should be applied. When a patient with a compelling history suggesting ACS is found to have a concentration of troponin that is detectable but below this decision limit, this result should be recognized as a marker of increased risk, and appropriate therapeutic actions (eg, early invasive treatment) should be considered. In patients with a low or indeterminate probability of ACS, troponin results in this range should prompt consideration of alternative causes, and repeat testing should be used to reduce the probability of an analytic false-positive.

Back to Article Outline

References 

  1. Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction . ESC/ACC Committee for the redefinition of myocardial infarction. J Am Coll Cardiol. 2000;36:959–969
  2. Braunwald E, Antman EM, Beasley JW, et al.  ACC/AHA 2002 guideline update for the management of patients with unstable angina and non-ST-segment elevation myocardial infarction (summary article). J Am Coll Cardiol. 2002;40:1366–1374
  3. Jaffe AS, Ravkilde J, Roberts R, et al.  It's time for a change to a troponin standard. Circulation. 2000;102:1216–1220
  4. Heidenreich PA, Alloggiamento T, Melsop K, et al.  The prognostic value of troponin in patients with non-ST elevation acute coronary syndromes (a meta-analysis). J Am Coll Cardiol. 2001;38:478–485
  5. Apple FS, Quist HE, Doyle PJ, et al.  Plasma 99th percentile reference limits for cardiac troponin and creatine kinase MB mass for use with ESC/ACC consensus recommendations. Clin Chem. 2003;49:1331–1336
  6. Pham M. Prognostic value of low-level troponin-I elevation in patients without definite acute coronary syndromes. Am Heart J 2004;148:776–82
  7. Wu A, Apple F, Gibler W, et al.  National Academy of Clinical Biochemistry Standards of Laboratory Practice (recommendations for the use of cardiac markers in coronary artery diseases). Clin Chem. 1999;45:1104–1121
  8. Morrow DA. Troponins in patients with acute coronary syndromes (biologic, diagnostic, and therapeutic implications). Cardiovasc Toxicol. 2001;1:105–110
  9. Morrow DA, Cannon CP, Rifai N, et al.  Ability of minor elevations of troponin I and T to identify patients with unstable angina and non-ST elevation myocardial infarction who benefit from an early invasive strategy (results from a prospective, randomized trial). JAMA. 2001;286:2405–2412
  10. Kontos MC, Shah R, Fritz LM, et al. Implication of different cardiac troponin I levels for clinical outcomes and prognosis of acute chest pain patients. J Am College Cardiol 2004;43:958–65
  11. Meier MA, Al-Badr WH, Cooper JV, et al.  The new definition of myocardial infarction (diagnostic and prognostic implications in patients with acute coronary syndromes). Arch Intern Med. 2002;162:1585–1589
  12. Trevelyan J, Needham EW, Smith SC, et al.  Sources of diagnostic inaccuracy of conventional versus new diagnostic criteria for myocardial infarction in an unselected UK population with suspected cardiac chest pain, and investigation of independent prognostic variables. Heart. 2003;89:1406–1410
  13. Pell JP, Simpson E, Rodger JC, et al.  Impact of changing diagnostic criteria on incidence, management, and outcome of acute myocardial infarction (retrospective cohort study). BMJ. 2003;326:134–135
  14. Venge P, Lagerqvist B, Diderholm E, et al.  Clinical performance of three cardiac troponin assays in patients with unstable coronary artery disease (a FRISC II substudy). Am J Cardiol. 2002;89:1035–1041
  15. Morrow DA, Rifai N, Sabatine MS, et al.  Evaluation of the AccuTnI cardiac troponin I assay for risk assessment in acute coronary syndromes. Clin Chem. 2003;49:1396–1398
  16. Wright SA, Sawyer DB, Sacks DB, et al.  Elevation of troponin I levels in patients without evidence of myocardial injury. JAMA. 1997;278:2144
  17. Jaffe AS. Elevations in cardiac troponin measurements (false false positives). Cardiovasc Toxicol. 2001;1:87–92
  18. Goldhaber SZ. Cardiac biomarkers in pulmonary embolism. Chest. 2003;123:1782–1784

PII: S0002-8703(04)00289-3

doi:10.1016/j.ahj.2004.05.024

American Heart Journal
Volume 148, Issue 5 , Pages 739-742, November 2004

Article Tools

* Image Usage & Resolution