ECG Diagnosis of Acute Myocardial Infarction
Abstract & Commentary
By Michael H. Crawford, MD
Source: Martin TN, et al. ST-segment deviation analysis of the admission 12-lead electrocardiogram as an aid to early diagnosis of acute myocardial infarction with a cardiac magnetic resonance imaging gold standard. J Am Coll Cardiol. 2007;50:1021-1028.
Although ECG is the standard initial screening test for acute coronary syndromes, its value for the diagnosis of acute myocardial infarction (MI) is difficult to determine because of the vagaries of serum biomarkers. Contrast-enhanced magnetic resonance imaging (MRI) is a unique gold standard for the early detection of acute MI, and may be useful for determining the diagnostic accuracy of the ECG. Thus, this study from Glasgow, United Kingdom, is of interest. They enrolled 116 patients seen at one hospital with new-onset chest pain and interpretable ECGs. MI was confirmed in 58 by the presence of delayed hyperenhancement on MRI done a mean of 50 hours after onset of symptoms. The ECG diagnosis of acute ST segment elevation MI (STEMI) followed the usual criteria, but they also included a STEMI-equivalent category with > 1 mm ST depression in > 2 contiguous leads, or one lead that is anatomically contiguous to one lead with ST elevation. Current ECG STEMI criteria detected 50% of acute MIs, with a specificity of 97%. Adding the STEMI- equivalent criteria increased sensitivity to 84% and minimally reduced specificity (93%). Including troponin resulted in 10 false positive MI diagnoses: 6 by troponin elevation alone; 2 by ECG criteria alone; and 2 by both. Two of the 4 falsely positive by ECG met new STEMI-equivalent criteria. Hannan et al concluded that considering ECG ST segment depression, as well as elevation, significantly increases the sensitivity for acute MI diagnosis without a major impact on specificity.
ECG remains the mainstay of the early triage of chest pain patients because of its wide spread availability, low cost, and ability to identify myocardial ischemia or infarction. Imaging may be more accurate, but is not widely available. Serum biomarkers are very sensitive, but may be normal early in the course of acute MI. However, this study showed that the detection of STEMI, where triage to reperfusion strategies is most critical, using standard ECG criteria, was only 50%. This increased to 84% without significant changes in specificity when ST segment depression in 2 anatomically-contiguous leads, or in one lead anatomically contiguous to a lead with ST elevation, was considered. For example, ST segment elevation in lead aVL and ST depression in lead III (negative ST in lead III reflects ST elevation in a non-existent lead near aVL). Two contiguous leads with ST depression (eg, V1 and V2) would reflect ST elevation in 2 nonexistent leads opposite them on the posterior lateral wall. This makes sense from an ECG point of view, and proved reliable in this study. Acceptance of this modification of the ECG criteria for STEMI would increase the number of patients with chest pain triaged to reperfusion, with a specificity rate in the 90% range.
Of course there are some caveats. This study showed that about one quarter of patients presented challenges in ECG interpretation, and were excluded. The challenging conditions included evidence of prior MI, left ventricular hypertrophy, rapid atrial fibrillation, WPW, bundle branch block, right ventricular hypertrophy, and extensive artifacts. The prevalence of infarction was high in this study (50%), which resulted in positive predictive values in the 90% range. If a lower MI prevalence was considered — say 20%, the positive predictive value would fall into the 75-85% range, but this requires further investigation. Contrast-enhanced MRI cannot distinguish the age of an MI, but in this study, all the MRI-diagnosed MIs had elevations and declines in serum biomarkers consistent with acute MI. Also, MRI currently has a spacial resolution of about one cubic centimeter, so smaller MIs would not be detected. Perhaps some MRI-negative, biomarker-positive patients have such small MIs. Hannan et al referred to these as "necrosettes." Presumably, such small MIs would be low-risk events, and reperfusion may not greatly influence prognosis. They may even represent "supply/demand imbalance" events rather than thrombotic occlusions. In this study, all MRI positive patients had a troponin I > 4.4 ng/mL. The average troponin I in those with a negative MRI was 1.4 ng/mL.
ECG mavens have long recognized that ST depression in I, aVL, or V1-4 may represent left main coronary artery thrombosis, so the concept that significant ST depression may represent an acute thrombotic condition, best treated with reperfusion, is not new. This study shows that it could help detect more STEMI than current ECG criteria without unacceptable increases in false positives. It seems worth keeping in mind when triaging chest pain patients.