Author + information
- Harvey White, DSc∗ ()
- ↵∗Reprint requests and correspondence:
Prof. Harvey White, Green Lane Cardiovascular Service, Auckland City Hospital, Private Bag 92024, Victoria Street West, Auckland 1142, New Zealand.
An expert consensus group from the Society for Cardiovascular Angiography and Interventions (SCAI) propose “a new definition of clinically relevant [myocardial infarction] MI after [percutaneous coronary intervention] PCI.” “Clinically relevant MI” is defined as “an event associated with a worsened prognosis.” This is a shift from previous definitions that did not consider prognosis, including the old World Health Organization definition.
Isolated biomarker elevations (>10×) are recommended in the SCAI definition with creatine kinase–myocardial band, which is increasingly no longer available, as the preferred biomarker using the local laboratory upper limit of normal (ULN). Whereas the universal definition recommends the use of the more sensitive and specific troponins >5× the 99th percentile upper reference limit (URL) and requires associated ischemic features, including symptoms of ischemia or associated angiographic complications. There is also emphasis on the importance of a stable baseline in the universal definition.
Different definitions will be challenging for clinicians and also for the interpretation of trials and registries. More information will be available in the next few years to inform future modifications of the definition of periprocedural MI. Meanwhile, the metrics for both should be reported.
The First Consensus Redefinition of Myocardial Infarction (MI) was published in 2000 following the formation of a Joint European Society of Cardiology/American College of Cardiology Committee (17). Elevation of biomarkers was essential for the diagnosis, and it was pathologically based and not based on prognosis. In 2007, the American Heart Association and the World Heart Federation joined the Universal Definition Consensus Group, and the classification of 5 types of MI was developed. This definition identified a type 4a MI as being associated with percutaneous coronary intervention (PCI). In this issue of the Journal, an expert consensus group from the Society for Cardiovascular Angiography and Interventions (SCAI) propose “a new definition of MI after PCI or type 4a MI” (1). They present an excellent review of our current understanding of the significance of type 4a MI and agree with the logic upon which the definition proposed by the universal definition of MI was based (2), but nevertheless propose a new definition different from the universal definition of MI published in 2012. Their major argument for doing this is to have a practical proposal, which is commendable. However, as with the universal definition, criticisms can be directed at this consensus proposal, which by definition is an opinion or position reached by a group as a whole, and there will be different or outlier views and considerations. In this paper, some of these issues and considerations will be discussed.
What are the differences between the definitions?
The differences between the definitions are tabulated in Table 1. The universal definition of MI (2) defines type 4a MI as MI associated with PCI with prolonged ischemic chest discomfort, electrocardiogram (ECG) changes, or angiographic decreased flow, embolism or loss of a side branch, or imaging evidence of a new wall motion abnormality or loss of viability plus an increase in biomarkers from a normal baseline value to a >5× the 99th percentile URL. Biomarkers alone are not sufficient.
To make the definition more clinically relevant, the biomarker cutpoint for MI was raised from >3× in the 2007 universal definition to >5× the 99th percentile URL (3). The increase was chosen arbitrarily to accommodate the increasing sensitivity of troponin assays. The addition of ischemic factors and/or angiographic complications was a consensus based on clinical judgment and the societal implications of the label of periprocedural MI.
Biomarkers may increase with successful procedures and optimal stent implantation (4). Thus, isolated biomarker elevation should not be used to define MI. Biomarker elevations have been shown to occur more frequently and be associated with increased mortality in unsuccessful stent procedures associated with coronary complications compared with successful stent procedures with similar biomarker elevations (5). Unsuccessful procedures were defined as >50% coronary stenosis, TIMI (Thrombolysis In Myocardial Infarction) flow grade <3, final National Heart, Lung, and Blood Institute dissection grade ≥D, urgent target vessel revascularization or stent thrombosis <24 h. Hence, the universal definition added criteria to define unsuccessful procedures.
The addition of ischemic complications raises the bar above isolated biomarker elevations and makes the event more clinically relevant. Raising the biomarker cutpoint by 66% and requiring other factors will significantly decrease the number of MIs.
The SCAI Consensus Group suggests use of the term clinically relevant. The old World Health Organization definition (6), required 2 out of 3 criteria, including symptoms, biomarkers, and ECG changes, and the universal definition, where the diagnosis of MI has been pathologically based requiring myocyte necrosis in the setting of ischemia. Prognosis depends on a vast number of factors, such as ejection fraction, the extent of atherosclerosis, diabetes, the success of PCI, and whether the patient is adherent with evidence-based treatments such as dual antiplatelet therapy and statins.
The proposed SCAI definition does not include ischemic symptoms that require ≥20 min in the universal definition. Ischemic symptoms have always been a basic tenet of the diagnosis of MI, and it should be no different for a type 4a MI.
Defining prognostically important criteria
The prognostic importance of isolated biomarker level elevations with type 4a MIs remains controversial, with several studies showing that only elevation of pre-PCI biomarkers, and not post-PCI levels, affects survival (7).
For a definition to have prognostic importance and to have a clinically relevant mortality risk, this can be achieved by requiring associated complications as in the universal definition, or by raising the cutpoint as suggested in both definitions. If the cutpoint is raised, sensitivity is decreased. If a very large biomarker elevation (≥70× troponin URL) is used, as suggested by SCAI, there will be very few PCI-related events identified, and an opportunity to improve patient outcomes may be lost.
Importance of a stable or falling biomarker baseline
When cardiac biomarkers are rising before PCI due to acute coronary syndromes, the curve of biomarker release may overlap with the curve of biomarker release after an ischemic complication of PCI. Thus, the ability to discriminate between the curves is limited. The last attempt to prognosticate the peak of a biomarker curve required 5 h of serial sampling and was done with total CK (8).
The universal definition states that type 4a MI cannot be assessed when biomarkers are rising or unknown at the time of PCI, and it is impossible to detect whether further increases are due to the procedure or to the initial process causing the elevation (2). The baseline needs to be shown to be stable for troponin levels. It is not adequate to document a stable baseline with CK-MB because of the lower sensitivity and late rise of CK-MB as compared with troponins, and an unstable baseline may not be detected with some rising CK-MB levels being due to a pre-PCI event.
It is not possible, as SCAI suggests, to use a cutpoint if biomarker levels are rising (9). The SCAI proposal to use a 50% rise will be different if levels are slightly abnormal (when it may work) compared with very elevated levels where a 50% rise will never work because it would require too large an increase.
Why use troponins instead of CK-MB?
The universal definition recommends the use of troponins, and the SCAI Consensus Group recommends the use of CK-MB. There are a number of reasons to use troponins rather than CK-MB (9). The variability in CK-MB mass assays is considerable, with 2- to 3-fold variability, and levels are variable in normal individuals, making it difficult to ascertain whether levels are rising or not. The analytical issues with CK-MB assays are much greater than with troponin assays (10). Also, because all individuals have significant circulating levels of CK-MB, higher cutpoints are necessary to define abnormality. In addition, sex-specific cutoff values are required (2). If CK-MB is used, a large number of baseline elevations will be missed that carry most of the prognostic information (1).
Troponins are more sensitive and specific markers of myocyte necrosis than CK-MB, increase earlier, and have nearly absolute specificity for myocardial tissue (11). The use of CK-MB will diagnose substantial numbers of individuals falsely due to its lack of specificity, as indicated by the fact that these patients have no associated adverse outcomes (12,13).
Troponins have been shown to be better for prognosis than CK-MB in acute coronary syndromes, and troponins identify patients at higher risk despite CK-MB levels being normal (14). In an analysis using the universal definition and 10-year follow-up with troponins for the diagnosis, MI was significantly related to long-term mortality, but not if MI was defined by elevated CK-MB (15).
CK-MB is now unavailable in an increasing number of hospitals. With CK-MB becoming obsolete, troponin will become the gold standard, and CK-MB will no longer have a role in defining PCI injury and infarction in clinical practice.
The SCAI Consensus Group makes the comment that “the use of ‘high-sensitivity’ troponins cannot presently be recommended for post-PCI myonecrosis.” These assays, which are widely used throughout Europe, South Africa, Australia, and New Zealand, offer improved sensitivity and diagnostic accuracy with which the SCAI authors concur (1), and should be recommended with use of the appropriate cutpoint, relative to contemporary assays (16). A lack of rise in the level of a high-sensitivity troponin at 3 h after PCI may rule out myocyte necrosis related to the procedure and allow discharge at that time.
What should the clinically relevant biomarker cutpoint be?
Recommendation to use the 99th percentile URL with biomarkers (CK-MB or troponins) was made in the 2000 document for the redefinition of MI (17). To enable comparisons and consistency among trials, the Universal Definition Task Force suggested that “no matter what, the investigators should ensure that a trial provides comprehensive data including the 99th percentile URL decision limits of the biomarkers used.” The proposed SCAI definition allows each local laboratory to have different cutpoints even though the assays may be the same. This will lead to different outcomes, confounding assessment of the efficacy of a new treatment according to the serendipity as to which hospital a patient is admitted to. It is a simple task to access what the 99th percentile URL is for each assay and to use that in trials as the decision cutpoint.
Relationship of biomarker elevations with mortality
In a recent analysis from the TRITON (Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel) trial, which required patients with elevated baseline biomarkers to have falling levels and a new 50% URL rise, in addition to symptoms or ECG changes, there was an association between type 4a MI, as defined by 3× CK-MB elevation if there were 2 samples after PCI and >5× CK-MB URL if there was only 1 sample, and a cardiovascular mortality at 180 days of 3.2% (adjusted hazard ratio [HR]: 2.4; 95% confidence interval [CI]: 1.6 to 3.7) (18). However, it may well have been that had troponin been used, an unstable rather than a stable baseline may have been detected.
In an analysis of 10,119 patients undergoing PCI from the EARLY ACS (Early Glycoprotein IIb/IIIa Inhibition in Non–ST-Segment Elevation Acute Coronary Syndrome) and SYNERGY (Superior Yield of the New Strategy of Enoxaparin, Revascularization and Glycoprotein IIb/IIIa Inhibitors) trials, for patients with stable or falling baseline troponin levels, there was a 7% increase in the adjusted HR of 1-year death per 10× upper limit of normal (ULN) increase in troponin levels (HR: 1.07; 95% CI: 1.02 to 1.11, p = 0.004) (19).
In some trials, adjustment has been made for some atherosclerotic factors. In a study of 23,604 patients from 8 clinical trials and 3 registries from Korea, type 4a MI occurred in 7.1% when defined as an increase in CK-MB mass >3× URL in patients with a normal baseline or a ≥50% increase in those with elevated, stable, or falling levels (20). After adjustment for baseline covariates, including coronary artery disease, number of stents, bifurcation, and long lesions, type 4a MI was associated with an increased risk of mortality at mean follow-up of 2.9 years (HR: 1.20; 95% CI: 1.04 to 1.39). Using a cutpoint of CK-MB >5× ULN was also significantly associated with increased mortality (HR: 1.33; 95% CI: 1.03 to 1.71). Side-branch occlusion was identified as the most common association (57.3%) with CK-MB elevation. This study has the major limitation of not having measured troponins to ensure a stable baseline.
There are few studies with long-term follow-up. In the Korean report, mentioned in the preceding text, which followed patients for up to 5 years, the HR for mortality with type 4a MI at 30 days was 1.03 95% CI: (0.42 to 2.51), and this increased in the landmark analysis for 30 days to 1 year to 1.34 (95% CI: 0.88 to 2.03) (20). The reasons for a possible effect on long-term mortality are not defined.
Long-term total mortality data are not available for either the proposed SCAI definition or the universal definition of type 4a MI.
What will be the impact of having different definitions?
If different MI definitions are used, comparisons between trials and trial outcomes, as well as between registries, will be difficult. There will be variations in the assessment of quality, and combining different definitions in meta-analysis may be impossible. Regulatory authorities will be challenged as to what different definitions may mean. Above all, it will lead to confusion among clinicians.
Quality of care indicators
It is imperative that, as in the universal definition, associated complications are required when quality of care indicators are measured, and not as SCAI recommends, just isolated biomarker elevations.
In the ongoing National Institutes of Health ISCHEMIA trial cardiac markers alone will not qualify for the definition of type 4a ‘prognostically significant MI’ and at least one other factor such as new ST-segment elevation or depression, new Q waves, or decreased TIMI flow are required. This trial will give the opportunity to advance the field by performing pre-specified analyses to determine the best definition for type 4a MI, as defined by the correlation with subsequent clinical events.
It is an unresolved question whether post-procedural myonecrosis is a risk marker of atherosclerotic burden and PCI complexity or a risk factor. New definitions of periprocedural MI will of necessity be arbitrary.
The rationale for the SCAI definition has been well articulated by its authors and may be appropriate in an individual trial, but it should not supplant the universal definition of MI. The metrics for both should be available if the SCAI definition is used.
Strict systematic implementation of the universal definition of MI with emphasis on the importance of a stable or falling troponin baseline will enhance specificity and will allow a rigorous assessment of new interventions. Documentation of angiographic complications for the universal definition of MI will be challenging, but important, because it makes the elevation of biomarkers clinically relevant and likely to impact prognosis.
There is a need to have high sensitivity to identify all clinically important type 4a MIs and to differentiate them from an index MI with high specificity. Further modification of the universal definition of MI associated with PCI should await substantial data that may well be forthcoming in the near future.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. White is co-chairman for the TaskForce for the Universal Definiton of Myocardial Infarction; has received research grants from Sanofi-Aventis, Eli Lilly, The Medicines Company, the National Institutes of Health, Pfizer, Roche, Johnson & Johnson, Schering-Plough, Merck Sharpe & Dohme, AstraZeneca, GlaxoSmithKline, Daiichi Sankyo Pharma Development, and Bristol-Myers Squibb; and has served on advisory boards for AstraZeneca, Merck Sharpe & Dohme, Roche, and Regado Biosciences.
- American College of Cardiology Foundation
- Moussa I.D.,
- Klein L.W.,
- Shah B.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- Jaffe A.S.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- White H.D.,
- on behalf of the Joint ESC-ACC-AHA-WHF Task Force for the Redefinition of Myocardial Infarction
- Iakovou I.,
- Mintz G.S.,
- Dangas G.,
- et al.
- Jeremias A.,
- Baim D.S.,
- Ho K.K.,
- et al.
- Mendis S.,
- Thygesen K.,
- Kuulasmaa K.,
- et al.
- Prasad A.,
- Rihal C.S.,
- Lennon R.J.,
- et al.
- Jaffe A.S.,
- Apple F.S.,
- Lindahl B.,
- Mueller C.,
- Katus H.A.
- Jaffe A.S.
- White H.D.
- Lin J.C.,
- Apple F.S.,
- Murakami M.M.,
- Luepker R.V.
- Newby L.K.,
- Goldmann B.U.,
- Ohman E.M.
- Costa F.M.,
- Ferreira J.,
- Aguiar C.,
- et al.
- White H.D.
- The Joint European Society of Cardiology/American College of Cardiology Committee
- Bonaca M.P.,
- Wiviott S.D.,
- Braunwald E.,
- et al.
- Tricoci P.,
- Leonardi S.,
- White J.,
- et al.
- Park D.W.,
- Kim Y.H.,
- Yun S.C.,
- et al.