Author + information
- Received January 8, 2004
- Revision received March 4, 2004
- Accepted March 11, 2004
- Published online July 7, 2004.
- Taha Taher, MD*,
- Yuling Fu, MD*,
- Galen S Wagner, MD†,
- Shaun G Goodman, MD‡,
- Claudio Fresco, MD§,
- Christopher B Granger, MD†,
- Lars Wallentin, MD∥,
- Frans Van de Werf, MD¶,
- Freek Verheugt, MD∥ and
- Paul W Armstrong, MD*,* ()
- ↵*Reprint requests and correspondence:
Dr. Paul W. Armstrong, University of Alberta, 2-51 Medical Sciences Building, Edmonton, Alberta T6G 2H7 Canada.
Objectives The investigators undertook a systematic, comprehensive analysis of the therapeutic response and clinical outcomes of reperfusion therapy for acute ST-segment elevation myocardial infarction (STEMI) in 5,470 patients from the Assessment of the Safety and Efficacy of a New Thrombolytic Regimen (ASSENT)-3 trial.
Background Prompt effective reperfusion therapy for acute STEMI may attenuate major myocardial necrosis.
Methods We prospectively collected sequential electrocardiographs and clinical data. Aborted myocardial infarction (MI) was defined as maximal creatine kinase ≤2× upper limit of normal coupled with typical evolutionary electrocardiographic changes.
Results Of the patients, 727 (13.3%) had an aborted MI, with the highest frequency (25%) occurring in patients treated <1 h after symptom onset. As compared with MI patients, patients with aborted MI more often had complete ST-segment resolution at 60 min (56.3% vs. 30.2%, p < 0.001) and 180 min (61.5% vs. 53%, p < 0.001); they also had smaller infarct sizes based on QRS score at discharge (2.37 vs. 4.62, p <0 .001). Mortality in aborted MI patients compared with those who had true MI was 3.9% versus 4.6% at 30-day and 7.0% versus 7.4% at 1-year. The baseline-adjusted mortality was significantly lower in patients with aborted MI (odds ratio [OR] 0.76, 95% confidence interval [CI] 0.63 to 0.92, p = 0.005 for 30-day and OR 0.70, 95% CI 0.50 to 0.98, p = 0.035 for one year). A very low-risk subset was identified with ≥70% ST-segment resolution at 60 min whose 30-day and 1-year mortality was 1.0% and 2.7%, respectively, compared with 5.9% and 9.3% in aborted MI patients with <70% ST-segment resolution at 60 min (all p ≤ 0.002).
Conclusions Prompt fibrinolytic treatment improved the likelihood of aborted MI. The subgroup with complete 60-min ST-segment resolution had the best clinical outcomes.
The outcome of patients with ST-segment elevation myocardial infarction (STEMI) is strongly influenced by the time from symptom onset to successful reperfusion. Some patients who receive prompt reperfusion are known to avoid myocardial necrosis. This population of “aborted myocardial infarction” patients was first characterized by Lamfers et al. (1)in 42 patients who had improved clinical outcomes. Given the increased emphasis on early therapy and improved outcomes (2), we undertook the current study to comprehensively analyze aborted MI patients from the Assessment of the Safety and Efficacy of a New Thrombolytic Regimen (ASSENT)-3 trial population. This trial was a prospective, randomized study evaluating three different antithrombotic strategies in conjunction with tenecteplase in STEMI (3). This large population provided a unique opportunity to evaluate patients with aborted myocardial infarction (MI) and undertake a systematic analysis of their therapeutic response and outcome. Specifically, we asked: 1) What is the frequency of patients escaping myocardial necrosis (i.e., aborted MI)? 2) Do such individuals have particular identifying features? 3) What are their clinical outcomes?
The specific entry criteria of the ASSENT-3 trial have been defined previously (3); briefly, patients were 18 years of age or older, with symptom onset <6 h before randomization, and ST-segment elevation of at least 0.1 mV in two or more limb leads or at least 0.2 mV in two or more contiguous precordial leads or presumed new left bundle branch block. Plasma samples for creatine kinase (CK) and the CK-MB fraction were collected on admission and subsequently during the hospitalization according to local hospital practice. For confirmation of the entry acute MI as well as any recurrent ischemia or reinfarction during hospitalization, peak CK/CK-MB and/or troponin I/T were used and recorded in the case report form as multiples exceeding the upper limits of normal.
Standard 12-lead electrocardiograms (ECGs) were collected at baseline, at 60 and 180 min after treatment, and at discharge. The ST-segment measurements were evaluated centrally at the ECG core laboratories (Canadian VIGOUR Centre, University of Alberta, Edmonton, Canadian Heart Research Centre, Toronto, Canada, and Duke Clinical Research Institute, Durham, North Carolina) without knowledge of treatment or outcome. ST-segment elevation was measured manually at the J-point using a hand-held caliper. The sum of ST-segment elevation in leads V1to V6, I, and aVL was added to the sum of ST-segment depression in leads II, III, and aVF for anterior MI. For inferior MI, the sum of ST-segment elevation in leads II, III, and aVF (and I, aVL, V5, and V6, if present) were added to the sum of ST-segment depression in leads V1to V4. The ST-segment resolution at 60 min and 180 min was classified according to the Schroder et al. method (4): complete (i.e., resolution of the initial ST-segment elevation ≥70%), partial (i.e., ST-segment resolution <70% to 30%), and none (i.e., ST-segment resolution <30%). QRS scores were calculated (5)on both baseline and discharge ECGs.
For this study, sequential ECGs of patients with maximal CK ≤2× upper normal were reanalyzed by three independent reviewers to examine the evolution in QRS and ST-T waves. Aborted MI was defined as maximal CK ≤2× upper limit of normal combined with typical evolutionary ECG changes. Patients who had neither a rise in enzyme levels nor evolutionary changes in the QRS and ST-T waves were separately classified as masquerading MI.
Descriptive statistics were summarized as medians with 25th and 75th percentiles for continuous variables, and the Mann-Whitney Utest was used. For categorical variables, data were summarized in percentages, the Fisher exact test was used for dichotomous variables, and the chi-square test was used for the rest of the categorical variables. The crude mortality rate was adjusted for key baseline clinical characteristics (Table 1) using multivariate logistic regression. The statistical analysis was performed using SPSS Version 11.0 (SPSS Inc., Chicago, Illinois).
In Figure 1, the derivation of our patient population is presented. There were 6,095 patients enrolled in the ASSENT-3 trial. Patients excluded from our study consisted of 100 patients (1.6%) who died in the first 24 h, 509 patients (8.4%) without available maximal CK data, and 16 patients (0.3%) without an available baseline ECG. The remaining 5,470 patients formed the study population, of whom 838 (15.3%) patients had maximal CK ≤2× upper limit of normal. When this group was analyzed, 111 (2%) patients had no evolutionary ECG changes and were characterized as “masquerading MI.” Thus, a final cohort of 727 (13.3%) patients fulfilled the criteria of “aborted MI.”
Demographic and clinical characteristics are shown in Table 1. Compared with true MI, patients with aborted MI were more likely to be older, female, and have a history of hypertension and diabetes. They also had a higher frequency of previous MI, coronary artery bypass grafting (CABG), percutaneous coronary intervention (PCI), and chronic aspirin use and were less likely to be current smokers. Aborted MI patients had lower sum of ST-segment deviation on their ECG at the time of admission. Time from symptom onset to treatment was shorter for aborted MI patients largely because of their faster presentation to the hospital. The occurrence of aborted MI according to hours from symptom onset to treatment is shown in Figure 2, where aborted MI was highest within the first hour (25%) but decreased thereafter (by 3 h, it was 10%).
Patients with aborted MI were more likely to have a non-urgent PCI or CABG (urgent or non-urgent) and less likely to develop in-hospital complications such as congestive heart failure and major arrhythmias (Table 2). Although there were no differences in the crude mortality rates at 30 days and 1 year between the aborted MI and true MI patients, mortality was significantly lower in aborted MI patients as compared with true MI (odds ratio [OR] 0.76, 95% confidence interval [CI] 0.63 to 0.92, p = 0.005 for 30-day mortality; OR 0.70, 95% CI 0.50 to 0.98, p = 0.035 for one year) after adjusting for baseline characteristics. There were no differences in treatment assignments between aborted MI and true MI groups.
The extent of ST-segment resolution according to the MI type is presented in Figure 3. The ST-segment resolution of ≥70% occurred with a higher frequency in the aborted MI population at 60 min (56.3% vs. 30.2%, p < 0.001) and 180 min (61.5% vs. 53%, p < 0.001). By 180 min, there were a similar proportion of patients (14% vs. 14.8%) in both groups with <30% ST-segment resolution. Baseline and discharge QRS scores (mean ± SE) were 2.42 ± 0.05 and 4.62 ± 0.06 for the MI group (n = 3,074); 1.80 ± 0.11 and 2.37 ± 0.12 for the aborted MI group (n = 492); and 1.27 ± 0.24 and 1.58 ± 0.27 for the masquerading MI group (n = 60).
In Figure 4, the effect of early ST-segment resolution on the outcomes within the aborted MI population revealed a statistically significant difference in 30-day mortality (i.e., 1% in the subgroup with ≥70% ST-segment resolution at 60 min vs. 5.9% mortality in those with <70% ST-segment resolution; p = 0.002). This difference increased at one year (i.e., mortality was 2.7% vs. 9.3%, respectively; p = 0.002) and remained statistically significant after adjusting for baseline characteristics (OR 0.28, 95% CI 0.11 to 0.71, p = 0.007 for aborted MI patients with complete ST-segment resolution vs. those without). There were no significant differences in the frequency of angiography, urgent PCI (ischemia-driven PCI before hospital discharge), urgent CABG, or non-urgent CABG in aborted MI patients with <70% or ≥70% ST-segment resolution at 60 min. However, there was a higher rate of non-urgent PCI in the aborted MI group, with ≥70% ST-segment resolution at 60 min (30% vs.18.4% in the remaining aborted MI patients, p = 0.003).
In the course of our analysis, we identified a group we have termed “masquerading MI” that presents without a rise in enzyme levels or typical evolutionary ECG changes but has several distinguishing features (Table 1). Specifically, when compared with patients with true MI, masquerading MI patients were older with more comorbidity (hypertension, diabetes, previous MI, and CABG) and a higher frequency of Killip class >I on admission. They also had a longer time to treatment from symptom onset, mostly because of a delayed presentation to hospital and a much lower ST-segment deviation on baseline ECGs. Their clinical outcomes differed from MI patients in that they had a higher rate of in-hospital major bleeds (excluding intracerebral hemorrhage) and a lower rate of PCI (urgent or non-urgent) and major arrhythmias.
Among the 509 patients without peak CK, we identified the 463 patients with either CK-MB or cardiac troponins available, and the remaining 46 patients did not have enzyme data. The criteria (i.e., ≤2 × the upper limit of normal of CK-MB or cardiac troponins) were used for defining the aborted MI for these 463 patients. When these patients were included in separate analyses, the aborted MI incidence remained nearly identical (i.e., 12.7% vs. 13.3% without these patients), and the other characteristics and clinical outcomes of aborted MI patients also were similar.
The following diagnoses simulating STEMI were identified in the 111 “masquerading MI” patients: early repolarization (n = 26; 23.4%), left bundle branch block (n = 16; 14.4%), right bundle branch block/ventricular tachycardia/Wolff-Parkinson-White syndrome (n = 3; 2.7%), pericarditis (n = 5; 4.5%), previous MI with persistent ST-segment elevation (n = 15; 13.5%), normal ECG (n = 4; 3.6%), and nondiagnostic ST-T changes (n = 42; 37.8%).
Our study of 727 patients with aborted MI is the largest and most comprehensive analysis to date, providing new insights into those STEMI patients treated with fibrinolysis who escape myocardial necrosis as measured by an increase in CK. Specifically, we have identified the incidence of aborted MI by hour of treatment, evolution of ST segments, and further validated the protection from myocardial damage by the change in QRS scores. Finally, we have identified two key patient subgroups; one with ≥70% ST-segment resolution at 60 min and another with “masquerading” MI that have significant clinical implications.
The frequency of aborted MI in our study was 13.3%, and the greatest occurred in those patients treated within 1 h of symptom onset, with a sharp drop off after 3 h. This pattern closely parallels the concept of the “golden hour” as popularized by Boersma et al. (2). Hence, our findings add further evidence to the growing body of data highlighting the major benefit of early treatment on STEMI patients (6).
We also demonstrate that patients with aborted MI have different clinical features from true MI patients. Specifically, they are more likely to have had a previous diagnosis or intervention for coronary artery disease and are much more likely to present earlier to the hospital.
The ECG evolutionary changes in our study also provide new insights. Whereas the initial aborted MI study of Lamfers et al. (1)limited ECG analysis to two observations at baseline and 120 min, our study examined four sequential systematically acquired ECGs and demonstrated a high early frequency of ST-segment resolution at 60 min as compared with true MI patients. This difference became less evident at 180 min and was almost equalized by discharge, emphasizing the value of early reperfusion. This utility of the 12-lead ECG reinforces the close correlation noted between 60-min ECGs and with short-term mortality noted in previous studies (7). The QRS evaluation on baseline and discharge provides further corroborative evidence that MI abortion signals the preservation of the myocardium (5,8–10). We also propose a new subclassification of aborted MI patients (i.e., those with maximal CK enzyme levels ≤2× upper limit of normal and ≥70% ST-segment resolution at 60 min). In this group, early reperfusion is associated with minimal myocardial damage and an excellent one-year clinical prognosis. Because this group had a higher rate of mechanical coronary intervention, we cannot exclude this care as a contributing factor to their better outcomes.
Finally, another key MI subset identified in our study was “masquerading MI,” or those who had neither an enzyme rise nor ST-segment evolution. These 111 patients constituted 2% of the ASSENT-3 trial patients with suspected acute MI. Although such patients have been previously identified as an issue that merits consideration (11–13), they have been poorly characterized and continue to emerge as inappropriate inclusions in clinical studies. These patients present with symptoms that are suggestive of STEMI but represent a diagnostic dilemma presumably reflected by a longer time to treatment and less baseline ST-segment elevation versus patients with true MIs. Our data provide clinical clues that should raise suspicion among physicians treating these patients (i.e., these patients are older and have more incidence of diabetes, comorbidity, and ECG confounders).
Although modest in number, the masquerading MI patients had a disproportionately higher bleeding rate and hence are unnecessarily exposed to the risks of fibrinolytic therapy without the anticipated benefits. We urge greater surveillance to avoid the unnecessary inclusion of such patients in future trials: They may benefit from additional diagnostic measures or mechanical intervention rather than fibrinolysis because they have a higher frequency of Killip class >I and worse one-year mortality despite the initial absence of true MI.
In the absence of a standard, various investigators have used different definitions for aborted MI, limiting the definition to CK-MB (14)or to this plus >50% ST-segment resolution (1,15). In the current study, aborted MI was defined as maximal CK ≤2× upper limit of normal with dynamic ECG changes. We do not have access to the specific times of enzyme measurements the ASSENT-3 trial protocol required for the confirmation of the presence of MI through sequential enzymatic measurements. Although it is impossible to exclude a sampling bias, this appears unlikely given the similar frequency of aborted MI in our study versus previous ones (1,14,15).
Our findings should prove useful for future STEMI management and investigations by enhancing the ability to exclude patients with masquerading MI. In conclusion, the concept of aborted MI appears a meaningful one related to improved clinical outcomes, lesser time to treatment data, and evolutionary ECG changes that are indicative of early reperfusion and preservation of myocardium.
☆ Financial support: Boehringer-Ingelheim (Germany), Genentech (South San Francisco, California), Aventis (Bridgewater, New Jersey).
- Assessment of the Safety and Efficacy of a New Thrombolytic Regimen-3 trial
- coronary artery bypass graft
- confidence interval
- creatine kinase-MB fraction
- myocardial infarction
- odds ratio
- percutaneous coronary intervention
- ST-segment elevation myocardial infarction
- Received January 8, 2004.
- Revision received March 4, 2004.
- Accepted March 11, 2004.
- American College of Cardiology Foundation
- Assessment of the Safety and Efficacy of a New Thrombolytic Regimen (ASSENT)-3 Investigators
- Schroder R,
- Wegscheider K,
- Schroder K,
- Dissmann R,
- Meyer-Sabellek W
- de Lemos J.A,
- Antman E.M,
- Giugliano R.P,
- et al.
- Morrow D.A,
- Antman E.M,
- Sayah A,
- et al.
- Rimar D,
- Crystal E,
- Battler A,
- et al.