Stress cardiac MR (CMR) may prove useful for identifying patients with coronary artery disease (CAD) who are at low risk of adverse events, according to a meta-analysis published online June 14 in Circulation: Cardiovascular Imaging.
Paola Gargiulo, MD, of the Institute of Diagnostic and Nuclear Development in Naples, Italy, and colleagues observed that improvements in CMR’s imaging capabilities and feasibility have made it a potential rival to stress SPECT and stress echocardiography for detecting ischemia in patients with CAD. Studies have shown CMR to offer good diagnostic accuracy for detecting CAD, with some placing it as the winner in comparisons with SPECT.
To date, research exploring CMR’s prognostic value has been limited to somewhat small, single-center studies. By conducting a meta-analysis, Gargiulo et al hoped to determine the predictive value for ischemic events in a stress CMR when either inducible perfusion defects (PDs) or inducible wall motion abnormalities (WMAs) were absent.
The researchers scoured the PubMed and Cochrane databases to find articles published between January 1985 and April 2012 of prospective or retrospective studies of patients with known or suspected CAD who underwent pharmacologic stress CMR to assess ischemia. Clinical outcomes of interest were nonfatal MI or cardiac death with a follow-up of three or more months. The studies identified a negative test as the absence of inducible PD or inducible WMA.
The meta-analysis included 14 studies and 12,178 patients. Seven studies were deemed good quality and seven fair quality.
At a weighted mean follow-up of 25.3 months, the summary negative predictive value was 98.12 percent. The pooled event rate after a negative test was 1.88 percent for an annualized event rate after a negative test of 1.03 percent.
Only eight studies included rates or revascularization and admission for unstable angina. With a weighted mean follow-up of 20.4 months, their summary negative predictive value was 97.17 percent; the pooled event rate after a negative test was 2.83 percent for an annualized event rate after a negative test of 1.73 percent.
“Our analysis revealed that, in subjects evaluated for known or suspected CAD, the absence of inducible ischemia on stress CMR predicts a low risk of cardiovascular events over a short- to mid-term follow up,” Gargiulo et al wrote. “The calculated annualized major event rate after a normal stress CMR was around 1 percent and this is only slightly higher than the background event rate observed in healthy low-risk individuals (less than 1 percent).”
CMR protocols include the use of late gadolinium enhancement (LGE) to detect myocardial necrosis. “A growing body of evidence is pointing to the prognostic value of LGE in several clinical scenarios and suggests the potential for combining this information with those provided by myocardial perfusion and wall motion assessment during a stress CMR study,” they wrote.
They could not perform subgroup analyses because of incomplete clinical data, which stymied evaluation of pre-test risk. “However, meta-regression analysis showed that the percentage of patients with previous revascularization (CABG and/or PCI) was a significant predictor of major cardiac events, increasing the risk of cardiac death/non-fatal MI after normal CMR.”
They acknowledged that they included a limited number of studies with moderate to high heterogeneity, and that the methods for imaging acquisition and interpretation for stress CMR are less standardized than the traditional modalities.