Defining left ventricular noncompaction using cardiac computed tomography

PURPOSE:: Left ventricular noncompaction (LVNC) is a cardiomyopathy characterized by a distinctive 2-layered appearance of the myocardium because of increased trabeculation and deep intertrabecular recesses. Echocardiography serves as the initial noninvasive diagnostic test. Currently, magnetic resonance imaging (MRI) is increasingly being used to diagnose LVNC because of its improved temporal and spatial resolution. So far, no criteria have been proposed to define pathologic LVNC with the use of computed tomography (CT). MATERIALS AND METHODS:: We analyzed CT images using an American Heart Association 17-segment model in 8 patients previously diagnosed with LVNC by clinical diagnosis, echocardiography, and/or MRI, as well as in 11 patients with nonischemic dilated cardiomyopathy, 11 patients with hypertrophic cardiomyopathy, 10 patients with severe aortic stenosis, 9 patients with severe aortic regurgitation, 10 patients with left ventricular hypertrophy due to essential hypertension, and, additionally, in a control group of 20 patients who had normal CT scans without a history of cardiovascular disease. The distribution of LVNC was assessed by qualitative analysis of 17 myocardial segments for the presence or absence of any degree of noncompaction. Each segment was analyzed in each of the 3 end-diastolic long-axis views for the presence or absence of noncompaction, and the most prominent trabeculation was chosen for measurement. The left ventricular apex was excluded. Thickness of noncompacted and compacted myocardium was measured perpendicular to the compacted myocardium. The ratio of noncompacted to compacted (NC:C) myocardium was calculated for each segment. Receiver operating characteristics were used to generate cutoff values with sensitivity and specificity to distinguish the LVNC group from other groups. RESULTS:: An end-diastolic NC:C ratio >2.3 distinguished pathologic LVNC with 88% sensitivity and 97% specificity; positive and negative predictive values were 78% and 99%, respectively. CONCLUSIONS:: CT using the standard MRI NC:C ratio cutoff >2.3 accurately characterizes pathologic LVNC.