DEfibrillators To REduce Risk by MagnetIc ResoNance Imaging Evaluation

Detailed Description: The utilization of ICD therapy has resulted in significant reduction in mortality among those at highest risk of sudden cardiac death, such as survivors of cardiac arrest and patients presenting with symptomatic sustained ventricular arrhythmias. Patients with CAD and advanced LV dysfunction (EF \<35%) also benefit from ICD. However, although at high risk, these patients represent only a small percentage of the population who die suddenly. While there are many tests that have been used for stratification of risk for sudden cardiac death, the two that have documented clinical utility are determination of left ventricular ejection fraction and presence of inducibility of ventricular tachycardia during programmed electrical stimulation performed as part of EP testing. The utility of these tests likely result from their ability to select patients who have the requisite substrate allowing for sustained ventricular tachyarrhythmias. It has been shown that ventricular tachycardia occurs more commonly in the setting of larger infarcts. Ejection fraction has been related to infarct size; presumably, the larger the area of infarction, the lower the ejection fraction. Electrophysiologic testing directly establishes the presence of substrate by the actual induction of ventricular tachycardia. A major limitation of electrophysiologic programmed stimulation is the high number of false negative findings. Thus, a significant number of patients without inducible arrhythmias remain at risk. CE-MRI provides functional information (EF, LV Volumes, LV mass, etc), which is routine in the initial evaluation of post-MI patients, and in addition provides detailed geometry of scar tissue. There is a clear association between inducible arrhythmias and scar size which until the development of cardiac MRI, could not be seen in humans. Use of cardiac MRI has demonstrated that although most patients with a large MI were inducible, a small but significant number of patients, who remain at risk, were not inducible. There was also an association between death and infarct size in patients with cardiovascular risk factors but no established CAD. The Center for Medicare Services (CMS) has recently decided in a coverage decision that patients with left ventricular dysfunction, heart failure, and an ejection fraction of \<35% would be eligible to receive an ICD as long as they are enrolled in a prospective registry. Patients with LV ejection fractions greater than 35% or those without heart failure and ejection fractions over 30%, represent a more difficult management dilemma. However, since the majority of out of hospital cardiac arrests occur in patients with EF \>35%, managing these patients is crucial in addressing the epidemiologic problem of sudden cardiac death. The primary objective of this trial is to test the hypothesis that therapy with an ICD combined with medical therapy improves long-term survival compared to medical therapy alone in patients with CAD, infarct mass greater than or equal to 10% of the left ventricle and left ventricular dysfunction who do not have an indication for ICD by either of the following criteria. Patients must have an EF of \>35% or have an EF of 30-35% and must not have inducible ventricular tachycardia or have NYHA Class II or greater heart failure (Target Population). The secondary objective is to test the hypothesis that therapy with an ICD combined with medical therapy improves arrhythmic survival compared to medical therapy alone in patients with CAD, infarct mass greater than or equal to 10% and left ventricular dysfunction who do not have an indication for ICD based on the Target Population described above. Recruitment: All patients who have a history of coronary heart disease (CAD) with documentation of either myocardial infarction (MI) or left ventricular dysfunction (LVD), a preliminary ejection fraction (EF) \> 35% and have previously undergone a contrast-enhanced MRI (CE-MRI) study for clinical diagnostic reasons or as part of the study entry screening procedure may be further evaluated for eligibility for this trial. In addition, patients with an ejection fraction of 30-35% may be eligible for the study if they do not currently have an indication for an ICD based on Target Population criteria described above. These patients may have NYHA Class I heart failure, no non-sustained VT on holter monitor, or if non-sustained VT is present, there is the absence of inducible VT at EP study. The first 1550 patients who are found to have an EF \>30% with NYHA Class I heart failure or 35% by routine clinical evaluation and who also have an MI involving greater than or equal to 10% of total left ventricular mass will be enrolled in the main randomized portion of the trial. These patients will be randomly assigned to one of two groups: ICD therapy in combination with medical therapy (ICD Group) or medical therapy alone (Control Group). Follow-Up: Clinic visits are required every 6 months until the completion of the study. Telephone contact is required every six months to assess vital status and obtain new information regarding medical status and/or medical events. The telephone calls alternate with the clinical visits, so that patient contact will occur every 3 months until the completion of the study. Non-Investigational Registry: The primary objective of the registry sub-study is to test the hypothesis that infarct mass as measured by contrast enhanced Cardiac MRI (CE-MRI) is a better predictor for sudden cardiac death than LV ejection fraction. The registry will examine infarct mass as measured by Cardiac MRI and LV ejection fraction (EF) as predictors for SCD. The purpose of the registry is hypothesis generating and no labeling or other indications are anticipated based on registry findings. Participation in the Registry requires that the patient has undergone a contrast-enhanced cardiac MRI prior to enrollment. In order to ensure consistent infarct mass assessments, the cardiac MRI study submitted to determine eligibility for randomization must meet the following criteria: 1. CE-cardiac MRIs must be obtained using Siemens, General Electric or Phillips equipment. 2. The contrast agent must be gadolinium-based at a dose sufficient to render images of acceptable quality. 3. The CE-cardiac MRI must be available for electronic submission to the MRI Core laboratory for analysis. If techniques for infarct mass measurement or data acquisition change during the course of the trial, the Core laboratory and the Steering Committee may choose to alter some of the above parameters. Once consent to participate in the registry has been obtained, the site will forward the CE-MRI study to the CE-MRI core lab for analysis to determine placement in the appropriate registry, baseline demographics characteristics will be collected on all registry patients. This will aid in statistical analysis to verify the generalizability of the findings. The differences in total survival between these groups will also be compared using the same methodology as for the primary end-point. Patients with and without ICD implants will be compared separately. In addition, blood specimens for genetic sampling and biomarker testing will be obtained on all patients in the registry cohort who agree, to determine if any of SCD substrates exist in the DETERMINE registry population. Study subjects will be contacted by mail by the Endpoint Coordinating Center at Brigham and Women's Hospital in Boston every 6 months to determine vital status and obtain any new information regarding change in medical status or the occurrence of any medical events. This will promote the continued relationship between the study participant and the enrolling center and can also be used to remind the study subject of their next scheduled appointment. Scope and Duration of the trial: * Up to 100 sites to screen a total of 10,000 patients will enroll 1550 patients into the randomized study recruited from a total of 10,000 patients contributing data and CE-MRI images to the registries. * Registry enrollment per site = \~90 subjects * Randomization per site = 1-3 per month (expected randomized enrollment per site is 20 patients or more ) * Estimated enrollment period: 36 months * 24 months of follow-up after the last patient is randomized