Abstract 15338: Prognostic Value of Normal Stress Cardiac MRI in Patients with Known or Suspected Coronary Artery Disease: A Meta-Analysis

Abstract

Background. Stress cardiac magnetic resonance (CMR) has been reported to provide prognostic information for risk stratification of patients with known or suspected coronary artery disease, mostly in single-center clinical studies. Aim of this study was to assess in a meta-analysis the prognostic value of normal stress CMR defined as the absence of inducible perfusion defects (PD) and/or the absence of inducible wall motion abnormalities (WMA).

Methods. We searched the PubMed, Cochrane and DARE databases between January 1985 and April 2012 and reviewed bibliographies of the retrieved articles. We included prospective and retrospective cohort studies of subjects who underwent stress CMR for known or suspected CAD, reporting data on primary outcomes of myocardial infarction (MI) and cardiac death with at least 3 months of follow-up. Secondary outcomes of unstable angina and/or revascularization procedures were abstracted when provided. The risk-adjusted relationship between absence of PD and/or absence of WMA and event-free survival was assessed by meta-regression.

Results. Of 446 identified studies, 12 met inclusion criteria, including 11513 patients. The negative predictive value for MI and cardiac death of the absence of myocardial ischemia was 98.19% (95% confidence interval [97.26% to 98.94%] ) over a mean follow-up of 25 months (range 12 to 74 months), resulting in estimate event rate after negative test equal to 1.71% (95% confidence interval [1.06% to 2.74%]) (figure). The corresponding annualized event rate was 1.03% per year. For secondary events, negative stress CMR had annualized event rates of 1.2%. In subgroup analyses, annualized event rates for hard events was 1.02% in absence of PD and 1.33% for studies that considered absence of inducible WMA.

Conclusions. Normal stress CMR has a high negative predictive value for hard and soft cardiac events. Absence of PD and absence of WMA show similar ability to identify low-risk patients.