Physicians turn to cardiac magnetic resonance (CMR) to help guide their treatment plans for patients with coronary artery disease (CAD). A meta-analysis published online Jan. 29 in the Journal of the American College of Cardiology showed that CMR also may help them assess future risk.
Existing studies on the prognostic value of CMR typically are small or have a limited number of events. Hamza El Aidi, MD, of University Medical Center Utrecht, the Netherlands, and colleagues conducted the meta-analysis through a systematic literature review in an effort to overcome these shortcomings.
They performed a literature search of MEDLINE and EMBASE on Feb. 25, 2013. They focused on CMR imaging findings, hard event outcomes and study populations.
The imaging findings included left ventricular ejection fraction (LVEF), wall motion abnormalities (WMA) at rest or after administration of pharmacological stress, myocardial perfusion at rest or after administration of pharmacological stress, early and late microvascular obstruction, presence and extent of late gadolinium enhancement, presence of edema and presence of intramyocardial hemorrhage.
Outcomes were all-cause mortality, cardiac death, cardiac transplantation or MI and major adverse cardiac events (MACE). Patients were divided into either having a recent MI or suspected or known CAD. They defined an independent prognostic CMR finding as a finding based on at least three studies and a total of more than 1,000 patients.
They identified 56 studies—27 involving patients with a recent MI and 29 with CAD patients—for a total of 25,497 patients. Findings in the recent MI group never reached the 1,000 patient threshold for hard events. LVEF was the sole predictor independently associated with MACE.
“Even though CMR may be used as a diagnostic tool in patients after a recent myocardial infarction, current literature does not support the use of CMR for prognostication,” they wrote. “Although CMR is the reference standard for LVEF, other more readily available and less expensive imaging modalities such as echocardiography are probably more suitable for this aim in clinical practice.”
In the CAD group, El Aidi and colleagues determined that WMA, inducible perfusion defects, LVEF and infarct size were independent predictors of hard events while inducible perfusion defects were associated with MACE. “These results indicate that both inducible CMR as well as infarct size measurements are important in the prediction of future cardiovascular events,” they wrote.
Although the authors described the quality of the individual studies as good, they listed several challenges using the studies in aggregate: differences in classification and reporting of the findings; too many variables in multivariable analyses; and how authors in the original studies defined MACE as a combination of endpoints.
In an accompanying editorial, Ilan Gottlieb, MD, PhD, and Gabriel Camargo, MD, of the National Institute of Cardiology in Rio de Janeiro, Brazil, pointed out that the 1,000 patient threshold was arbitrary but that the results showed the need for further CMR studies involving acute MI.
“Curiously, both WMA and perfusion assessments appear to have similar diagnostic accuracies, but, as El Aidi et al. have shown, perfusion abnormalities appear to have greater prognostic impact (adjusted HR 3.02-7.77) than WMA (adjusted HR 1.87-2.99),” they wrote. “The reason for this is not completely clear, and worthwhile of further scientific investigation, as this is one of the most important messages from their work.”