Electrocardiographic Imaging of Myocardial Ischemia, Infarction and Scar: Correlation with SPECT, MRI and Arrhythmias

Abstract: This thesis focuses on the development and testing of electrocardiographic (ECG) methods to diagnose and characterize myocardial ischemia, infarction and fibrosis (scar). The goal is to provide improved decision support to physicians regarding the optimal selection of patients for two costly, invasive therapies – percutaneous coronary intervention (PCI) for patients with acute myocardial infarction (MI) and implantable cardioverter defibrillators (ICDs) for patients with heart failure and reduced left ventricular ejection fraction (LVEF). The ECG methods evaluated in this thesis are compared to the gold standard imaging techniques of myocardial perfusion single photon emission computed tomography (SPECT) for ischemia and contrast-enhanced magnetic resonance imaging (MRI) for myocardial scar. Additionally, ECG methods to identify and quantify myocardial scar are evaluated as a novel tool to identify vulnerability to ventricular arrhythmias. Myocardial infarction from acute coronary occlusion is a leading cause of morbidity and mortality. Reperfusion therapy with thrombolytics or PCI reduces mortality; however, the therapies come with significant risk and must be given within a narrow time frame. Knowledge of the location, size and acuteness of ischemic myocardium at risk for infarction could help predict the potential for myocardial salvage and impact patient triage and choice of reperfusion therapy. Following infarction, non-salvaged myocardium undergoes remodeling, resulting in chronic scar. In addition, myocardial scars can develop in nonischemic cardiomyopathies. While recent clinical trials have shown that ICDs reduce mortality in patients with low LVEF, the annual rate of appropriate ICD shocks is only 5%. Since myocardial scar forms the substrate required for precipitating and propagating reentrant ventricular tachyarrhythmias, detecting and quantifying scar in patients may result in improved risk stratification. Studies I and II assessed ECG algorithms to quantify the location, size and acuteness of myocardial ischemia. Study I demonstrated that the three-dimensional ST-segment vector derived from the ECG can be graphically projected onto a model of the left ventricle to determine ischemia location and ST-vector magnitude correlates with ischemia size measured by myocardial perfusion SPECT. Study II demonstrated that the Anderson-Wilkins ECG-acuteness score is superior to time-from-symptom onset to PCI for predicting myocardial salvage assessed by myocardial perfusion SPECT and contrast-enhanced MRI. In addition, computer simulation showed that the ECG sign of salvageable myocardium, tall T-waves, could only be produced by shortened action potential duration, which has previously been associated with preconditioning. Studies III and IV evaluated novel Selvester QRS scoring systems for use in the presence or absence of ECG confounders to quantify myocardial scar and predict ventricular arrhythmias. Study III demonstrated that QRS scoring accurately identifies and quantifies scar when validated against contrast-enhanced MRI. Furthermore, higher QRS-score scar size is associated with ventricular arrhythmia susceptibility, as defined by increased frequency of monomorphic ventricular tachycardia induction during electrophysiologic programmed stimulation. Study IV extended the application of QRS scoring to 797 patients with heart failure and reduced LVEF who received ICDs for the primary prevention of sudden arrhythmic death. Increasing QRS score was independently associated with increased ICD shocks and patients with a QRS score of zero (no scar) had a significantly reduced risk of ICD shocks. The predictive value of QRS-score scar size was complementary to that of LVEF and other clinical characteristics.