Abstract
OBJECTIVES: To find the imaging mortality predictors in patients with previous myocardial infarction (MI), symptomatic heart failure (HF), and reduced left ventricle (LV) ejection fraction (EF). METHODS: for the study 39 patients were selected prospectively with prior MI, symptomatic HF, and LVEF ≤40%. All patients underwent transthoracic echocardiography (TTE), single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI), 18F-FDG positron emission tomography (FDG PET). 31 patients underwent cardiovascular magnetic resonance (CMR) with late gadolinium enhancement (LGE). Patients were divided into two groups: 1 group - cardiac death; 2 group - no cardiac death. Myocardial scars were assessed on a 5-point-scale. Follow-up data was obtained. RESULTS: Imaging features disclosed significant difference (p < 0.05) of defect score (CMR and SPECT-PET), LV end-diastolic diameter (EDD) (TTE), LVEDD index (CMR), LV global longitudinal strain (CMR) and LV global circumferential strain (CMR) between the groups. Predictors of cardiac death were: LVEDD index (TTE) and LV global longitudinal strain. The cut-off values to predict cardiac death were: defect score (CMR) 25 (AUC, 79.5%; OR 1.8, 95% CI 1.2-2.7), SPECT-PET defect score 22 (AUC, 73.9%; OR 0.5, 95% CI 0.3-0.7), LVEDD (TTE) 58 mm (AUC, 88.4%; OR 23.6, 95% CI 2.6-217.7), LVEDDi 30 mm/m(2) (TTE) (AUC, 73.6%; OR 22.0, 95% CI 1.9-251.5), LVEDDi 33.6 mm/m(2) (CMR) (AUC, 73.6%; OR 22.0, 95% CI 1.9-251.5), LV global longitudinal strain -13.4 (AUC, 87.8%; OR 2.1, 95% CI 1.2-3.7) and LV global circumferential strain -16.3 (AUC, 76.1%; OR 1.9, 95% CI 1.2-3.0). CONCLUSIONS: Imaging features, such as defect score (CMR) >25, SPECT-PET defect score >22, LVEDD (TTE) >58 mm, LVEDDi (TTE) >30 mm/m(2), LVEDDi (CMR) >33.6 mm/m(2), LV global longitudinal strain -13.4 and LV global circumferential strain -16.3, may increase sensitivity and specificity of FDG PET and LGE CMR predicting of late mortality.