Contrast magnetic resonance imaging in the assessment of myocardial viability in patients with stable coronary artery disease and left ventricular dysfunction.

BACKGROUND The utility of contrast MRI for assessing myocardial viability in stable coronary artery disease (CAD) with left ventricular dysfunction is uncertain. We therefore performed cine and contrast MRI in 24 stable patients with CAD and regional contractile abnormalities and compared MRI findings with rest-redistribution 201Tl imaging and dobutamine echocardiography. METHODS AND RESULTS Delayed MRI contrast enhancement patterns were examined from 3 to 15 minutes after injection of 0.1 mmol/kg IV gadolinium diethylenetriamine pentaacetic acid (Gd-DTPA). Comparable MRI and 201Tl basal and midventricular short-axis images were subdivided into 6 segments. Segments judged nonviable by quantitative and qualitative assessment of 201Tl scans showed persistent, systematically greater MRI contrast signal intensity than segments judged viable (P</=0.002). Delayed contrast hyperenhancement also occurred in segments judged nonviable by dobutamine echocardiography (P</=0.03). The presence or absence of hyperenhancement correlated most closely with nonviability and viability, respectively, in segments that were akinetic or dyskinetic under resting conditions (83% concordance with 201Tl in both cases). In segments with resting hypokinesis, 58% of segments showing hyperenhancement were judged viable by 201Tl and may have represented an admixture of scar tissue and viable myocardium. CONCLUSIONS Delayed (by 3 to 15 minutes) hyperenhancement of Gd-DTPA contrast-enhanced MRI images occurs frequently in dysfunctional areas of the left ventricle in patients with stable CAD. Hyperenhancement is associated with nonviability by rest-redistribution 201Tl scintigraphy and dobutamine echocardiography, particularly in regions exhibiting resting akinesis/dyskinesis. The absence of hyperenhancement correlates with radionuclide and echocardiographic determinations of viability, regardless of resting contractile function.