Background
Myocardial deformation analysis by speckle-tracking echocardiography (STE) has been shown to accurately predict viability in patients with chronic ischemic left ventricular (LV) dysfunction. The aim of this study was to evaluate two-dimensional STE for the prediction of global and segmental LV functional changes after acute myocardial infarction (AMI) in comparison with late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR).
Methods
In 93 patients (mean age, 60 ± 11 years) with first AMIs (55 with ST-segment elevation myocardial infarctions and 38 with non–ST-segment elevation myocardial infarctions) treated with acute percutaneous coronary intervention, global peak longitudinal strain was determined to describe global function by STE, and peak systolic circumferential and longitudinal strain was determined for segmental function analysis. LGE CMR was performed to define the amounts of global and segmental myocardial scar. STE and LGE CMR were performed within 48 hours of AMI. At 6-month follow-up, transthoracic echocardiography was repeated to determine global und segmental LV recovery and adverse LV remodeling (increase in end-systolic volume > 15%).
Results
Accuracy to predict global functional improvement as well as LV remodeling at 6-month follow-up after AMI was similar for STE and LGE CMR (areas under the curve, 0.715 vs 0.729 [ P = .8830] and 0.806 vs 0.824 [ P = .7141], respectively). Peak systolic circumferential strain < −14.2% had sensitivity of 71.6% and specificity of 58.1% to predict segmental functional improvement. Compared with LGE CMR, the predictive accuracy of transmural STE for segmental functional improvement at 6-month follow-up was lower (area under the curve, 0.788 vs 0.668; P = .0001). Predictive accuracy for segmental functional improvement could be improved by analysis of endocardial circumferential strain (area under the curve, 0.700 vs 0.668 for transmural speckle-tracking echocardiographic analysis; P = .0023).
Conclusions
Two-dimensional STE allows the prediction of global functional recovery as well as LV remodeling after AMI with accuracy comparable with that of LGE CMR. Accuracy to predict segmental functional recovery using transmural deformation analysis by two-dimensional STE is inferior compared with LGE CMR but can be improved by a layer-specific analysis of endocardial deformation.
In patients who have had acute myocardial infarctions (AMIs), the extent of necrosis is the crucial parameter determining functional outcomes and prognosis. The spectrum of functional changes after AMI is broad, ranging from potential recovery of segmental as well as global left ventricular (LV) function in less extensive infarcts to LV remodeling with increases of LV volumes in large infarcts. Late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) accurately delineates infarcted from viable myocardium and allows the prediction of LV functional recovery after acute infarction. LGE CMR was described to be superior to other imaging modalities for the prediction of LV remodeling after AMI. Speckle-tracking echocardiography (STE) has been introduced as a novel method for analyzing myocardial deformation. Recently, more detailed layer-specific analysis of myocardial deformation has been described, allowing separate analysis of endocardial and epicardial layer strain. Deformation imaging by STE has been shown to have high agreement with LGE CMR in the analysis of myocardial viability in patients with chronic ischemic LV dysfunction. In this study, we evaluated the ability of myocardial deformation analysis by two-dimensional (2D) STE for the prediction of global and segmental changes in LV function during 6-month follow-up after AMI in comparison with LGE CMR.
Methods
Study Population
We screened 300 consecutive patients for possible inclusion in the study. Patients had to be in sinus rhythm and to have undergone acute percutaneous intervention for first AMI. Revascularization had to be achieved in all perfusion beds. Contraindications to CMR, such as device therapy or severe renal dysfunction, as well as insufficient echocardiographic windows resulted in exclusion from the study. Further exclusion criteria were significant valvular heart disease and previous myocardial infarction. Considering these inclusion and exclusion criteria, 93 patients (mean age, 60 ± 11 years) formed the study group. Fifty-five patients (59%) had ST-segment elevation myocardial infarctions (STEMIs), with a maximal creatinine kinase (CK) level of 1,780 ± 1,541 IU/L, and 38 patients (41%) had non–ST-segment elevation myocardial infarctions (NSTEMIs), with a maximal CK level of 592 ± 698 IU/L. Transthoracic echocardiography was performed for LV wall motion analysis and assessment of LV volumes and ejection fraction. STE and LGE CMR were performed within 48 hours of percutaneous coronary intervention for treatment of AMI. Percutaneous coronary intervention for treatment of AMI was performed according to standard guidelines.
At 6-month follow-up, transthoracic echocardiography was repeated to define potential recovery of global as well as segmental LV function or LV remodeling. Patient characteristics are shown in Table 1 .
| Variable | Value |
|---|---|
| Clinical variables | |
| Age (y) | 60 ± 11 |
| Men | 83 (89 %) |
| Diabetes mellitus | 11 (12%) |
| Hypertension | 52 (56%) |
| Smoking | 41 (44%) |
| Hypercholesterolemia | 34 (37%) |
| Family history | 24 (26%) |
| Peak CK level (IU/L) | 1,296 ± 1,391 |
| Imaging variables | |
| End-diastolic volume at baseline by echocardiography (mL) | 111 ± 35 |
| ESV at baseline by echocardiography (mL) | 51 ± 26 |
| Ejection fraction at baseline by echocardiography (%) | 56 ± 10 |
| Infarct mass by CMR (g) | 21.7 ± 21.5 |
| Total myocardial mass by CMR (g) | 158.6 ± 32.0 |
| Angiographic variables | |
| Culprit lesion | |
| LAD | 43 (46%) |
| LCX | 22 (24%) |
| RCA | 28 (30%) |
| Concomitant medication | |
| Aspirin | 93 (100%) |
| Clopidogrel | 89 (96%) |
| ACE inhibitors/ARBs | 91 (98%) |
| β-receptor blockers | 90 (97%) |
| Statins | 86 (92%) |
Echocardiography
Echocardiography was performed using a Vivid 7 system (GE Vingmed Ultrasound AS, Horten, Norway) equipped with a 2.5-MHz transducer. Parasternal long-axis and short-axis views at the basal, midventricular, and apical levels, as well as three standard apical views (four chamber, two chamber, and long axis) were acquired (frame rate, 56–92 frames/sec). LV ejection fraction was determined by manual tracing of end-systolic and end-diastolic endocardial borders using apical four-chamber and two-chamber views, using the biplane Simpson’s method. The original 16-segment model of the American Society of Echocardiography was used to divide the left ventricle. Segmental wall motion was graded as (1) normokinetic, (2) hypokinetic, or (3) akinetic or dyskinetic. A segment was considered to demonstrate functional improvement at 6-month follow-up if segmental wall motion grading improved by at least one grade. In case endocardial border definition was found to be insufficient for accurate visual analysis of segmental function, contrast enhancement using SonoVue (Bracco, Milan, Italy) was given. In 47 patients (51%), contrast echocardiography was performed. Serial echocardiography allowed the detection of global functional recovery as well as LV remodeling. Global functional recovery was defined as an absolute increase in LV ejection fraction of ≥5% from baseline to 6-month follow-up in patients with LV ejection fractions < 55% immediately after AMI, corresponding to a recent study. LV remodeling was defined as an increase in LV end-systolic volume (ESV) of >15% from baseline to 6-month follow-up, corresponding to previous studies.
Speckle-Tracking Strain Analysis
For circumferential deformation imaging, the three acquired parasternal short-axis views at the basal, midventricular, and apical levels, and for longitudinal deformation imaging the three standard apical views were analyzed considering the 16-segment model. Analysis was performed offline with the aid of a dedicated software package (EchoPAC BT 05.2; GE Vingmed Ultrasound AS). This system allows analysis of peak systolic segmental circumferential strain from short-axis views and peak global and segmental systolic longitudinal strain from apical views on the basis of detection of natural acoustic markers. These markers are acoustic speckles that are equally distributed within the myocardium and can be identified as well as followed frame to frame during several consecutive images. The natural acoustic markers are expected to change their position from frame to frame in accordance with the surrounding tissue motion. The system calculates mean strain values for whole predefined LV segments.
For global deformation analysis, global peak systolic longitudinal strain was averaged automatically from all LV segments using the three acquired apical views.
Considering the parasternal short-axis views, for each LV segment with adequate tracking quality, total wall thickness peak systolic circumferential strain was automatically calculated. In addition to analysis of total wall thickness strain, analysis of peak systolic circumferential strain was performed for an endocardial and an epicardial layer as described and validated previously. The middle (centerline) of the region of interest represents the average strain values for the full wall thickness, and the strain values presented for the endocardial layer and epicardial layer are the values at the inner and outer region-of-interest lines. According to the segmental circumferential strain analysis, total wall thickness peak systolic longitudinal strain as well as endocardial longitudinal and epicardial longitudinal strain were automatically calculated considering the three apical views. In case of disagreement of the observer with automatically defined borders, manual correction could be applied. Systolic circumferential strain considering the total wall thickness as well as of the endocardial and epicardial layers was automatically calculated using a medium degree of spatial and temporal smoothing. These parameters relate to deformation along the curvature of the left ventricle in the parasternal short axis. The analysis system automatically determines tracking quality. In this analysis, only segments with optimal tracking quality were included. Segments with suboptimal tracking quality were dismissed from the analysis. The tracking quality of the remaining segments was controlled visually to ensure adequate automatic tracking. End-systole was defined as aortic valve closure in the apical axis view and transferred to all other views ( Figure 1 ).
