Peri-procedural myocardial infarction: what is in a definition?

Abstract

Peri-procedural myocardial infarction (PMI) is an important outcome measure in several trials. However, there is no consensus regarding the definition of PMI. There is significant variability in terms of biomarkers used, the threshold values, the requirement for additional supporting evidence, and even the duration until which PMI can be said to occur. This lack of clarity leads to inappropriate comparisons between trials. Besides, the outcome of a trial and the inferences drawn are completely altered by using a different definition. Thus, there is a pressing need to formulate a uniform definition for PMI in future studies.

It is hard to recall a definition that has raised more controversies and concerns than peri-procedural myocardial infarction (PMI). Never before has a simple definition exerted so much impact on outcomes of trials, interpretation of findings, and formulation of guidelines. PMI has been an important component of the outcomes measured in several clinical trials comparing coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI). However, a number of definitions significantly different from each other have been used to define PMI resulting in a complete lack of clarity as to what exactly constitutes a PMI.

How long after a procedure is myocardial infarction (MI) considered to be peri-procedural?

There seems to be no agreement among definitions even regarding something as basic as the duration of PMI. The EXCEL (Evaluation of XIENCE versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization) trial using the Society for Cardiovascular Angiography and Interventions (SCAI) definition considered a period of 72 h after the procedure to define PMI [1, 2]. The ISCHEMIA (International Study Of Comparative Health Effectiveness With Medical And Invasive Approaches) trial used a cutoff of 48 h, while the SYNTAX (TAXUS Drug-Eluting Stent Versus Coronary Artery Bypass Surgery for the Treatment of Narrowed Arteries) trial used a cutoff of 7 days [3, 4]. Both the Third and the Fourth Universal definition of myocardial infarction (UDMI) consider a limit of 48 h to define PMI [5, 6].

Which biomarker should be used to define PMI?

There is even disagreement among different definitions with regard to the preferred biomarker to define PMI. While the EXCEL, SCAI, SYNTAX, and ISCHEMIA trials have preferred creatine kinase-MB (CK-MB), the Third and Fourth UDMI prefer cardiac troponins [1–6].

Should the threshold value of the biomarker defining PMI be the same after PCI and CABG?

This is one of the key questions. Due to interventions on the heart, some degree of release of myocardial enzymes is inevitable which does not constitute an MI. Hence, a threshold has been set by all definitions, only above which PMI is said to occur. This release is bound to be higher with CABG due to greater manipulation of the heart. Therefore, fixing a threshold that is the same for both PCI and CABG inherently puts CABG at a disadvantage.

The SYNTAX trial as well as the EXCEL trial used the same level of increase in CK-MB after PCI and CABG to define PMI. In contrast, the ISCHEMIA trial, and the Third and Fourth UDMI all used a higher threshold for CABG compared to PCI to define PMI (Table 1).

Table 1.

Threshold values of biomarkers used for diagnosing post-procedural MI

Definitions	Criteria	Preferred biomarker	PCI	CABG
			(ULN)
SCAI/EXCEL	Standalone	CK-MB	>10	>10
	Associated with additional criteria	CK-MB	>5	>5
SYNTAX Trial	Associated with additional criteria	CK-MB	>5	>5
ISCHEMIA trial	Standalone	CK–MB	>10	>15
	Associated with additional criteria	CK-MB	>5	>10
Third UDMI	Associated with additional criteria	TROPONIN	>5	>10
Fourth UDMI	Associated with additional criteria	TROPONIN	>5	>10

ULN, upper level of normal. Additional criteria included ECG: new pathological Q-waves; angiographic evidence of graft or native vessel occlusion; loss of viable myocardium or new wall motion abnormality on cardiac imaging

Are standalone values of biomarkers sufficient to define PMI?

There are two important issues with using standalone biomarker values to diagnose PMI.

1. Standalone values overdiagnose the MI rates, especially in CABG

When the PMI rates in the SYNTAXES (Synergy between PCI with Taxus and Cardiac Surgery Extended Survival) trial were re-calculated using different definitions, it was seen that with standalone biomarker values the PMI rates doubled in the PCI group but increased almost 8 times in the CABG arm [7].

2. PMI diagnosed with standalone values may not have clinical significance in CABG

It is generally agreed that for PMI to be clinically relevant, it should have an association with early or late mortality, re-intervention, or hospitalization for heart failure [7]. While in PCI, an isolated rise in cardiac enzymes was associated with increased mortality even at 10 years; no such prognostic relationship was seen in the case of CABG at any time point. The rise in biomarkers with supporting ECG changes however was associated with increased mortality in both PCI and CABG [7].

What are the implications of these multiple definitions?

There are several far-reaching implications of not having a uniform definition for PMI. The most intuitive consequence is that the results of one study cannot be compared with another. Besides, the interpretations of the study findings change completely with a change in definition within the same trial. This was demonstrated by the re-analysis of both the SYNTAXES and the EXCEL trial using different definitions of PMI [7, 8].

The incidence of PMI in the SYNTAXES study at 48 h using the SYNTAX definition was 2.7%, which increased to 3% using the Fourth UDMI, to 5.7% with the SCAI/EXCEL, and 6% with the ISCHEMIA trial definition. This increase was far more obvious in the CABG arm where the PMI rates changed from 2.1% with the Fourth UDMI, to 2.4% with the SYNTAX, 8.8% with the ISCHEMIA, and a massive 16.5% using the EXCEL/SCAI definitions.

An even more important observation was made with the re-analysis of the EXCEL data [8].

The PMI rate, using the EXCEL protocol definition (SCAI), was significantly higher (p = 0.01) in the CABG arm (6.1%) compared to the PCI (3.6%) arm. The same data when analyzed with the Third UDMI revealed that while the PMI rate increased to 4% in the PCI group it reduced to 2.2% in the CABG group, a result significantly in favor of CABG (p = 0.02) [8].

Thus, it is not just the incidence of PMI that became significantly in favor of CABG, but the entire conclusion drawn from the EXCEL trial took a different hue. In a complete reversal of results, re-analysis of EXCEL data using the Third UDMI definition showed that the composite of death, stroke, and MI was significantly in favor of CABG [9].

A simple change in definition changes not only the incidence of PMI but the entire study outcome, interpretation, and inferences drawn from the study. This in turn impacts physician and patient perspective of the treatment and influences guidelines and recommendations. Thus, despite the challenges involved in it, there is a pressing need for the adoption of a uniform definition of PMI. It has even been suggested that in case a uniform definition cannot be fairly formulated, PMI should not be included as an outcome measure in future studies [10]. It seems the devil, indeed, lies in the details of the definition.

Funding

None.

Declarations

Conflict of interest

The author declares no competing interests.

Informed consent

Not applicable as review article.

Ethics approval

The manuscript does not mention any patient or patient details, and hence formal ethical approval is not required.

Informed consent

Not applicable.

Human and animal rights statement

Not applicable.