Layer-specific strain assessment of left ventricular systolic function changes in patients with non-ST-elevation acute coronary syndrome

Abstract

Objective To analyze territorial and global longitudinal layer-specific strain of left ventricle by two-dimensional speckle tracking imaging (2D-STI) in patients with suspected non-ST-elevation acute coronary syndrome (NSTE-ACS) and to explore the value of layer-specific strain parameters for prediction of significant coronary artery stenosis. Methods Seventy-five patients with suspected NSTE-ACS in People′s Hosptial of Subei from September 2016 to January 2017 were enrolled and all patients underwent coronary arteriography (CAG). Among them, there were 24 subjects in control group (coronary artery without stenosis or stenosis rate <50%) and 51 subjects in coronary atherosclerotic heart disease group (coronary heart disease, CHD). According to whether coronary artery occlusion, the CHD group was divided into coronary stenosis group (32 subjects) and coronary occlusion group (19 subjects). Using EchoPAC software, two-dimensional dynamic images were analyzed to obtain left ventricle 18-segment systolic longitudinal layer-specific strain and to calculate the territorial longitudinal strain (TLS) of endocadium, mid-myocardium and epicardium (TLSendo, TLSmid, TLSepi) and left ventricle global longitudinal strain (GLS) of endocadium, mid-myocardium and epicardium (GLSendo, GLSmid, GLSepi). The differences of left ventricle territorial and global longitudinal layer-specific strain parameters among 3 groups were compared by one-way analysis of variance and the differences between two groups were compared by LSD-t test. The receiver operating characteristic (ROC) curve of each parameter was constructed to predict significant coronary stenosis by using the results of CAG as the gold standard. Results Compared with control group and coronary stenosis group, TLSendo, TLSmid, TLSepi and GLSendo, GLSmid, GLSepi all decreased in patients with coronary occlusion, and the differences were statistically significant (coronary occlusion group vs. control group: t values were -5.819, -5.049, -4.845, -5.955, -5.036 and -4.724, respectively, P values were all less than 0.01; coronary occlusion group vs. coronary stenosis group: t values were -2.983, -3.059, -2.903, -2.989, -3.192 and -3.387, respectively, P values were all less than 0.01). And compared with control group, only TLSendo and GLSendo decreased in patients with coronary stenosis, and the differences were statistically significant (t values were -3.981 and -4.164, respectively, P values were all less than 0.01). TLSendo, TLSmid, TLSepi and GLSendo, GLSmid, GLSepi showed a gradient decrease in all 3 groups, but only in the control group the comparison between two of the three layers showed statistically significant differences (TLSepi vs. TLSendo, GLSepi vs. GLSendo: t values were both -10.083, P values were all less than 0.01; TLSepi vs. TLSmid, GLSepi vs. GLSmid: t values were both -4.559, P values were all less than 0.01; TLSmid vs. TLSendo, GLSmid vs. GLSendo: t values were both -5.549, P values were all less than 0.01). The absolute differences between endocardial and epicardial TLS and GLS (∆TLS and ∆GLS) decreased gradually from the control group, to coronary stenosis group and to coronary occlusion group, and the differences were statistically significant (coronary occlusion group vs. control group: t values were 6.915 and 7.489, respectively, P values were all less than 0.01; coronary stenosis group vs. control group: t values were 4.923 and 7.202, respectively, P values were all less than 0.01; ∆TLS of patients in the coronary occlusion group vs. coronary stenosis group: t value was 2.250, P value was less than 0.05), which reflected a pronounced decrease in endocardial function. By ROC curve analysis, GLSendo and TLSendo showed the highest area under the curve in predicting significant coronary artery stenosis, which were better than strain parameters of mid-myocardium, epicardium and the entire wall thickness of the myocardium. Conclusions Left ventricle showed systolic dysfunction in all three layers in suspected NSTE-ACS patients with CHD, especially the endocardium. The longitudinal layer-specific strain parameters by 2D-STI can be used for quantitative evaluation of the territorial and global systolic dysfunction differences of left ventricle in all layers in suspected NSTE-ACS patients with CHD, which can also be used for prediction of significant coronary artery stenosis. Key words: Echocardiography; Two-dimension-speckle tracking imaging; Acute coronary syndrome; Ventricular, left