Speckle-tracking layer-specific analysis of myocardial deformation and evaluation of scar transmurality in chronic ischemic heart disease

BACKGROUND: Identification of the extent of scar transmurality in chronic ischemic heart disease is important because it correlates with viability. The aim of this retrospective study was to evaluate whether layer-specific two-dimensional speckle-tracking echocardiography allows distinction of scar presence and transmurality. METHODS: A total of 70 subjects, 49 with chronic ischemic cardiomyopathy and 21 healthy subjects, underwent two-dimensional speckle-tracking echocardiography and late gadolinium-enhanced cardiac magnetic resonance. Scar extent was determined as the relative amount of hyperenhancement using late gadolinium-enhanced cardiac magnetic resonance in an 18-segment model (0% hyperenhancement = normal; 1%-50% = subendocardial scar; 51%-100% = transmural scar). In the same 18-segment model, peak systolic circumferential strain and longitudinal strain were calculated separately for the endocardial and epicardial layers as well as the full-wall myocardial thickness. RESULTS: All strain parameters showed cutoff values (area under the curve > 0.69) that allowed the discrimination of normal versus scar segments but not of transmural versus subendocardial scars. This was true for all strain parameters analyzed, without differences in efficacy between longitudinal and circumferential strain and subendocardial, subepicardial, and full-wall-thickness strain values. Circumferential and longitudinal strain in normal segments showed transmural and basoapical gradients (greatest values at the subendocardial layer and apex). In segments with scar, transmural gradient was maintained, whereas basoapical gradient was lost because the reduction of strain values in the presence of the scar was greater at the apex. CONCLUSIONS: The two-dimensional speckle-tracking echocardiographic values distinguish scar presence but not transmurality; thus, they are not useful predictors of scar segment viability. It remains unclear why there is a greater strain value reduction in the presence of a scar at the apical level