Prognostic Implications of Procedural vs Spontaneous MI
Prognostic Implications of Procedural vs Spontaneous MI
The results of the present study indicate that among a broad population of unselected patients undergoing PCI in contemporary practice, those who experience a procedural MI differ from those who present with spontaneous MI in several ways. First, we observed that peak cardiac biomarker levels were lower after procedural MI as compared with spontaneous MI. This observation suggests that patients experiencing procedural MI generally experience a lesser degree of myocardial necrosis when compared with patients with spontaneous MI. Moreover, we found that in risk-adjusted models, spontaneous MI was independently associated with both death and cardiovascular death at 12 months, whereas procedural MI was not. These findings indicate that among unselected patients undergoing PCI, the prognostic impact of procedural MI is less than that of spontaneous MI.
Procedural MI can range from a small elevation of cardiac enzymes to a large Q-wave MI, be asymptomatic or symptomatic, and is typically more common than spontaneous MI during follow-up of clinical trials of coronary revascularization. The frequency of procedural MI varies depending on the population studied, marker tested, and threshold definition used, with frequencies as high as 50% in some series. In an analysis of 20 studies, the frequency of procedural MI varied from 0 to 29% when CK-MB was used as the basis for end point determination but was several fold higher, with a frequency of 14% to 48% when troponin was used. Because these frequencies far exceed those of spontaneous MI in patients followed up after PCI, the potential exists for the results of clinical trials to be driven primarily by the frequency of procedural MIs.
Although it may be argued that a given level of biomarker elevation, whether it occurs spontaneously or as the result of a procedure, should have similar prognostic significance, in the setting of a clinical trial, the prognostic impact of procedural vs spontaneous MI may differ for a number of reasons. First, in clinical trials, spontaneous MIs are detected only in patients who seek medical care because of chest pain or other concerning symptoms (unless new Q waves are detected on electrocardiogram [ECG]). In contrast, in many clinical trials, the occurrence of procedural MI is evaluated on all patients undergoing revascularization procedure regardless of symptom status. Because many spontaneous MIs are clinically silent, there is thus a bias toward enhanced detection of procedural MIs. Second, as noted in the current study, peak biomarker levels, which can serve as a crude estimate of the extent of myocardial necrosis, differed substantially for procedural and spontaneous MIs in our PCI population. Specifically, patients with spontaneous MI had a greater extent of myocardial necrosis, as demonstrated by higher peak cardiac biomarker levels. Third, in patients with a procedural MI, there may be a balance between the adverse effects of loss of myocardium and the beneficial effect of reestablished flow after PCI, contributing, in part, to better prognosis when compared with spontaneous MI. Given these important differences between spontaneous and procedural MI, it is not surprising that their prognostic impact differs as well. Indeed, in our risk-adjusted analyses, we found that the prognostic impact of a procedural MI was intermediate between that of no MI and spontaneous MI—a finding that is concordant with several other studies. Moreover, for similar biomarker levels, the prognosis of procedural and spontaneous MI was somewhat similar, although we are likely underpowered for such subgroups. In addition, the strength of this analysis is the fact that both types of MI were followed for the same period, thereby reducing lead-in bias.
Composite end points are often used in clinical trials to enhance statistical power and to avoid challenges in dealing with competing risks. Myocardial infarction is itself a composite end point that includes both spontaneous and procedural MIs. Although both procedural and spontaneous MIs result in myonecrosis, the pathophysiology of these elevations may not be the same. Typically, spontaneous MI is the result of plaque rupture or supply-demand mismatch in the setting of angiographically severe coronary artery stenosis. In contrast, procedural MI is most commonly related to angiographic complications such as side-branch occlusion or distal embolization of atherosclerotic debris or thrombus.
In recent years, there have been rapid advances in biomarker assessment of MI, moving toward biomarkers with greater specificity for cardiac myonecrosis such as CK-MB and troponins. In addition, there has been a shift from assays using enzymatic activity to highly specific immunoassays. The result of these changes has been that progressively smaller concentrations of cardiac biomarkers can now be detected in the peripheral circulation. With the increasing use of high-sensitivity troponin assays, this frequency is likely to increase exponentially. This higher frequency of MI qualification will apply to both procedural MI and spontaneous MI because many events that were previously classified as unstable angina will now qualify as MI based on sensitive markers. The results and interpretation of clinical trials will therefore change depending on the definition of MI used. Although the prognostic significance of smaller biomarker elevations after PCI is still uncertain, some studies have shown that procedural MIs with larger biomarker elevation (such as CK-MB ≥50 ng/mL) are independently associated with increased 1-year mortality. Similarly, although the universal MI criteria have the same cutoff threshold for both CK-MB and troponins, other studies have shown a much higher threshold (~7 times the CK-MB threshold) for troponin that correlated with similar outcomes to that for CK-MB. Thus, it is likely that procedural MI is prognostically important, but at a higher threshold of biomarker elevation than is currently recommended. These factors should be considered in component definition of MI in clinical trials.
The definition of procedural MI used in this study was based on biomarkers alone without taking into consideration symptoms or ECG findings, which are not uniformly assessed after PCI in contemporary practice. The recently published third Universal Definition of MI states that "MI associated with PCI is arbitrarily defined by elevation of cTn values >5× the 99th percentile upper reference limit in patients with normal baseline values or a rise of cTn values >20% if the baseline values are elevated and are stable or falling." In addition to biomarker criteria, the revised Universal Definition also requires the presence of one additional criterion (symptoms, ECG changes, angiographic complications, or imaging evidence of new loss of viable myocardium). Although we reported angiographic complications in the study, these reflected site reported events only and were likely an underestimate of the true frequency of such complications. In addition, the study assessed only 1-year outcomes, and it is not clear if procedural MI will be prognostically important over a longer term or for other outcomes such as heart failure. Moreover, the spontaneous MI cohort in our study was defined based on a population who presented with an MI at baseline and underwent an index PCI procedure, and hence, the results may not be generalizable to other cohorts. In addition, although a propensity score approach was used to account for baseline differences, these statistical techniques will not account for individual differences and for unmeasured confounders.
Discussion
The results of the present study indicate that among a broad population of unselected patients undergoing PCI in contemporary practice, those who experience a procedural MI differ from those who present with spontaneous MI in several ways. First, we observed that peak cardiac biomarker levels were lower after procedural MI as compared with spontaneous MI. This observation suggests that patients experiencing procedural MI generally experience a lesser degree of myocardial necrosis when compared with patients with spontaneous MI. Moreover, we found that in risk-adjusted models, spontaneous MI was independently associated with both death and cardiovascular death at 12 months, whereas procedural MI was not. These findings indicate that among unselected patients undergoing PCI, the prognostic impact of procedural MI is less than that of spontaneous MI.
Prognostic Impact of Procedural vs Spontaneous MI
Procedural MI can range from a small elevation of cardiac enzymes to a large Q-wave MI, be asymptomatic or symptomatic, and is typically more common than spontaneous MI during follow-up of clinical trials of coronary revascularization. The frequency of procedural MI varies depending on the population studied, marker tested, and threshold definition used, with frequencies as high as 50% in some series. In an analysis of 20 studies, the frequency of procedural MI varied from 0 to 29% when CK-MB was used as the basis for end point determination but was several fold higher, with a frequency of 14% to 48% when troponin was used. Because these frequencies far exceed those of spontaneous MI in patients followed up after PCI, the potential exists for the results of clinical trials to be driven primarily by the frequency of procedural MIs.
Although it may be argued that a given level of biomarker elevation, whether it occurs spontaneously or as the result of a procedure, should have similar prognostic significance, in the setting of a clinical trial, the prognostic impact of procedural vs spontaneous MI may differ for a number of reasons. First, in clinical trials, spontaneous MIs are detected only in patients who seek medical care because of chest pain or other concerning symptoms (unless new Q waves are detected on electrocardiogram [ECG]). In contrast, in many clinical trials, the occurrence of procedural MI is evaluated on all patients undergoing revascularization procedure regardless of symptom status. Because many spontaneous MIs are clinically silent, there is thus a bias toward enhanced detection of procedural MIs. Second, as noted in the current study, peak biomarker levels, which can serve as a crude estimate of the extent of myocardial necrosis, differed substantially for procedural and spontaneous MIs in our PCI population. Specifically, patients with spontaneous MI had a greater extent of myocardial necrosis, as demonstrated by higher peak cardiac biomarker levels. Third, in patients with a procedural MI, there may be a balance between the adverse effects of loss of myocardium and the beneficial effect of reestablished flow after PCI, contributing, in part, to better prognosis when compared with spontaneous MI. Given these important differences between spontaneous and procedural MI, it is not surprising that their prognostic impact differs as well. Indeed, in our risk-adjusted analyses, we found that the prognostic impact of a procedural MI was intermediate between that of no MI and spontaneous MI—a finding that is concordant with several other studies. Moreover, for similar biomarker levels, the prognosis of procedural and spontaneous MI was somewhat similar, although we are likely underpowered for such subgroups. In addition, the strength of this analysis is the fact that both types of MI were followed for the same period, thereby reducing lead-in bias.
Definition of MI in Clinical Trials
Composite end points are often used in clinical trials to enhance statistical power and to avoid challenges in dealing with competing risks. Myocardial infarction is itself a composite end point that includes both spontaneous and procedural MIs. Although both procedural and spontaneous MIs result in myonecrosis, the pathophysiology of these elevations may not be the same. Typically, spontaneous MI is the result of plaque rupture or supply-demand mismatch in the setting of angiographically severe coronary artery stenosis. In contrast, procedural MI is most commonly related to angiographic complications such as side-branch occlusion or distal embolization of atherosclerotic debris or thrombus.
In recent years, there have been rapid advances in biomarker assessment of MI, moving toward biomarkers with greater specificity for cardiac myonecrosis such as CK-MB and troponins. In addition, there has been a shift from assays using enzymatic activity to highly specific immunoassays. The result of these changes has been that progressively smaller concentrations of cardiac biomarkers can now be detected in the peripheral circulation. With the increasing use of high-sensitivity troponin assays, this frequency is likely to increase exponentially. This higher frequency of MI qualification will apply to both procedural MI and spontaneous MI because many events that were previously classified as unstable angina will now qualify as MI based on sensitive markers. The results and interpretation of clinical trials will therefore change depending on the definition of MI used. Although the prognostic significance of smaller biomarker elevations after PCI is still uncertain, some studies have shown that procedural MIs with larger biomarker elevation (such as CK-MB ≥50 ng/mL) are independently associated with increased 1-year mortality. Similarly, although the universal MI criteria have the same cutoff threshold for both CK-MB and troponins, other studies have shown a much higher threshold (~7 times the CK-MB threshold) for troponin that correlated with similar outcomes to that for CK-MB. Thus, it is likely that procedural MI is prognostically important, but at a higher threshold of biomarker elevation than is currently recommended. These factors should be considered in component definition of MI in clinical trials.
Study Limitations
The definition of procedural MI used in this study was based on biomarkers alone without taking into consideration symptoms or ECG findings, which are not uniformly assessed after PCI in contemporary practice. The recently published third Universal Definition of MI states that "MI associated with PCI is arbitrarily defined by elevation of cTn values >5× the 99th percentile upper reference limit in patients with normal baseline values or a rise of cTn values >20% if the baseline values are elevated and are stable or falling." In addition to biomarker criteria, the revised Universal Definition also requires the presence of one additional criterion (symptoms, ECG changes, angiographic complications, or imaging evidence of new loss of viable myocardium). Although we reported angiographic complications in the study, these reflected site reported events only and were likely an underestimate of the true frequency of such complications. In addition, the study assessed only 1-year outcomes, and it is not clear if procedural MI will be prognostically important over a longer term or for other outcomes such as heart failure. Moreover, the spontaneous MI cohort in our study was defined based on a population who presented with an MI at baseline and underwent an index PCI procedure, and hence, the results may not be generalizable to other cohorts. In addition, although a propensity score approach was used to account for baseline differences, these statistical techniques will not account for individual differences and for unmeasured confounders.
Source...