Comprehensive Prognosis Assessment by CMR Imaging After ST-Segment Elevation Myocardial Infarction

Study Questions:

Does cardiac magnetic resonance imaging (MRI) offer incremental prognostic information in patients who present with ST-segment elevation myocardial infarction (STEMI)?


The investigators conducted a predefined observational substudy of patients enrolled in the AIDA STEMI (Abciximab Intracoronary versus Intravenously Drug Application in STEMI) study, which was an open-label, randomized, multicenter trial of 2,065 participants conducted in 22 centers in Germany. The MRI substudy enrolled 795 consecutive patients at eight centers with proven expertise in cardiac MRI. Patients with severe claustrophobia, hemodynamic instability, pacemakers or defibrillators, metallic cerebral or intracranial implants, or allergy to gadolinium or creatinine clearance <30 ml/min were excluded. Subjects underwent contrast-enhanced cardiac MRI 1-10 days after the index event using either a 1.5 or 3.0 Tesla clinical MRI scanner. Infarct size and microvascular obstruction were quantified on late gadolinium enhancement images. Area at risk was quantified on T2-weighted imaging. Myocardial salvage index was defined as area at risk minus the infarct size divided by the area at risk. Cardiac MRI images were interpreted at a blinded core laboratory. The primary endpoint was the composite of death from any cause, reinfarction, or new congestive heart failure within 1 year of the index infarction.


Of the 795 subjects recruited, 33 were excluded due to evidence of prior MI and 23 due to incomplete cardiac MR data, leading to a final analysis set of 738 patients. A total of 52 major adverse cardiac events (MACE) occurred: 20 deaths, 18 reinfarctions, and 14 readmissions for congestive heart failure. Univariate clinical predictors of MACE basis included age, female gender, hypertension, diabetes, multivessel coronary artery disease, Killip classification, TIMI risk score, culprit lesion in the left anterior descending coronary artery, and impaired resolution of ST-segment elevation. Cardiac MRI predictors of MACE on a univariable basis were increased left ventricular (LV) end-systolic volume, infarct size, less myocardial salvage, greater microvascular obstruction, and greater LV systolic dysfunction. Of note, LV end-diastolic volume was not associated with MACE. The combination of LV ejection fraction (LVEF) ≤47%, infarct size ≥19% of the LV, and microvascular obstruction >1.4% of the LV was seen in 71% of those who experienced MACE (37/52). Using stepwise Cox regression analysis only, the (Thrombolysis in Myocardial Infarction) TIMI risk score and microvascular obstruction ≥1.4% of the LV were independent predictors of MACE.


The authors concluded that in this large, multicenter STEMI population, CMR markers of myocardial damage, particularly microvascular obstruction, provided independent and incremental prognostic information.


The key advantages of this study stem from the high quality, multicenter data used. The cardiac MRI protocols used are high quality and comprehensive. The translatability of these practices outside of specialized centers is less clear. An unanswered question from this analysis is: What are the clinical value of these data? Specifically, how could these data be used in a manner that would improve clinical outcomes? In their discussion, the authors suggest that in the future, a comprehensive cardiac MRI might help improve selection of subjects for implantable cardioverter-defibrillator (ICD) therapy. The authors do not identify any criteria which could be used to justify withholding ICDs from subgroups of those who would otherwise qualify. Nor do they suggest criteria which could identify suitably high-risk subgroups of those with relatively preserved LVEF who could conceivably benefit from ICD therapy. Indeed, the only large trial designed to test this hypothesis, DETERMINE (Defibrillators to Reduce Risk by Magnetic Resonance Imaging Evaluation), was abandoned for difficult enrollment and futility. Similarly, the results of the ADMIRE-HF study of MIBG have also not had much impact on patient selection strategy for ICD therapy in the United States. Furthermore, although this was a relatively large study, the low event rate (7%) and absolute number of events (52 MACE in 738 patients) prevented a truly robust analysis of the value of CMR. Consequently, we cannot be certain that the presented analysis for the incremental value of cardiac MR is not substantially optimistic beyond a comprehensive clinical assessment and LVEF.

< Back to Listings