Background
Left ventricular remodeling (LVr) is still common after ST-segment elevation myocardial infarction (STEMI). Early predictors of remodeling are being investigated. The aims of this study were to evaluate the prognostic value of speckle-tracking echocardiography for the prediction of LVr 3 months after primary percutaneous coronary intervention in patients with STEMI and to analyze the relationship between values of peak longitudinal strain of particular LV segments and relative changes of their subvolumes.
Methods
Patients with first STEMI were enrolled. Baseline enzymes were collected, and electrocardiography and echocardiography (transthoracic echocardiography, speckle-tracking echocardiography, and three-dimensional studies) were preformed. Three months after myocardial infarction, two-dimensional and three-dimensional ultrasonographic studies were done.
Results
Sixty-six patients were divided into two groups: 44 patients without LVr and 22 patients with LVr. Among 31 patients with anterior wall STEMI, the rate of LVr was 42%. On the basis of assessments of baseline and follow-up myocardial wall contractility, 1,041 segments were analyzed. All segments were divided into normal ( n = 842), reversibly dysfunctional ( n = 68), and irreversibly dysfunctional ( n = 131). Receiver operating characteristic curve analysis showed that global longitudinal strain predicted LVr with an optimal cutoff value of −12.5% (area under the curve, 0.77). In multivariate analysis, diabetes mellitus (odds ratio, 4.61; 95% confidence interval, 1.19–18.02) and global longitudinal strain (odds ratio, 1.19; 95% confidence interval, 1.04–1.37) were determinants of LVr. Positive correlations were found between peak longitudinal strain and changes in subvolumes for all segments ( R = 0.11, P = .005) and for those irreversibly dysfunctional ( R = 0.22, P = .04).
Conclusions
In patients with STEMI treated by primary percutaneous coronary intervention, the frequency of LVr during 3-month follow-up was high and mainly affected the population with anterior wall myocardial infarction. The results of this study show the clinical value of global longitudinal strain measured by speckle-tracking echocardiography in the prediction of LVr. A moderate correlation was found between the value of peak longitudinal strain and changes in subvolumes attributed to irreversibly dysfunctional segments.
Outcomes in the treatment of ST-segment elevation myocardial infarction (STEMI) have significantly improved in recent years. This is due mainly to the introduction of reperfusion therapy and modern pharmacotherapy. Nevertheless, left ventricular (LV) remodeling (LVr) is still commonly present and, among those with anterior wall STEMI, affects 30% to 35% of patients. From the clinical point of view, this is a dynamic phenomenon, and it begins in the acute phase of myocardial ischemia. There is a change in myocardial wall structure, wall thinning, elongation, and progression toward hypertrophy and dilatation. Changes in LV geometry may lead to heart failure and life-threatening arrhythmias, thus increasing the mortality rate.
Early predictors of remodeling are still being investigated. It is important to estimate which of the data collected during hospitalization may help identify patients with high probability of remodeling. Traditional echocardiography is widely used, but its value for remodeling prediction is low. Perfusion echocardiography, which is recommended as a noninvasive tool to assess myocardial viability in the region of infarction, is still of limited availability and requires expensive contrast media and extreme echocardiographer experience.
To assess LV function, it seems more valuable to analyze particular segments rather than to assess the whole myocardial wall. Quantification can also help a good deal, and speckle-tracking echocardiography (STE) provides such possibilities. STE allows the measurement of peak longitudinal strain (LS) and peak radial strain independently of ultrasonographic beam angle. To date, angle independence has been possible only with tagged magnetic resonance imaging.
The aims of our study were as follows: (1) to evaluate the prognostic value of STE for the prediction of LVr 3 months after primary percutaneous coronary intervention (pPCI) in patients with STEMI and (2) to analyze the relationship between value of LS of particular LV segments and relative changes of their subvolumes.
Methods
In this prospective study, we enrolled patients diagnosed with first STEMI. The criteria for enrollment were as follows: STEMI with onset of chest pain <12 hours before pPCI, culprit artery closure (Thrombolysis In Myocardial Infarction [TIMI] flow grade 0) and restored blood flow after pPCI (TIMI flow grade 3), age 18 to 80 years, and provision of informed consent. The exclusion criteria were previous myocardial infarction (MI) or coronary artery bypass grafting, thrombolytic therapy during STEMI, significant valvular dysfunction, hypertrophic cardiomyopathy, other than sinus heart rhythm, and poor echocardiographic conditions to analyze the results of STE.
All patients received therapy according to European Society of Cardiology guidelines. They received loading doses of aspirin, 600 mg of clopidogrel, and 100 IU/kg of heparin (maximum 5,000 IU). The use of abciximab was optional and left to the decision of the invasive cardiologist.
Markers of Necrosis
Upon admission and after 6, 12, and 24 hours, cardiac creatine kinase and troponin I were collected using the immuneinhibition and immunoenzymatic quantitative methods.
Electrocardiography
Twelve-lead electrocardiography was performed directly before and 60 min after pPCI of the infarct-related artery. In the first electrocardiographic study, we analyzed the heart rate, the maximum elevation of the ST segment from a single lead, and the sum of elevations of ST segments in all leads. In the second study, only the sum of ST-segment elevations was measured.
Echocardiography
At discharge (4–6 days after the acute phase), two-dimensional and three-dimensional echocardiography were performed, and the results of STE were assessed.
Two-Dimensional Echocardiography
Resting echocardiography was performed using the Vivid 7 system (GE Vingmed Ultrasound AS, Horten, Norway). Three apical scans of the left ventricle in the four-chamber, three-chamber, and two-chamber views according to the guidelines of the American Society of Echocardiography were performed. A 16-segment model of the left ventricle was used for wall motion score, strain, and the assessment of subvolumes. Segments were graded (1 = normokinetic, 2 = hypokinetic, 3 = akinetic, or 4 = dyskinetic) on the basis of subjective assessments of wall motion amplitude and changes of LV thickness at systole. Wall motion score index was defined as the sum of the segment score ratings divided by the number of segments scored.
For analysis, we divided the segments of the left ventricle into regions of interest on the basis of coronary blood supply to the left ventricle.
Segments perfused by the left anterior descending coronary artery, responsible for the anterior wall MI, were those marked in Figure 1 . The remaining seven segments were those in nonanterior locations.
STE
STE is an echocardiographic, non-Doppler method that analyzes the LS of LV segments by assessing the deformation of an object relative to its original length. By this definition, strain is a dimensionless ratio and often is expressed as a percentage.
STE was performed in typical apical views with frame rate of 60 to 90 frames/sec, and strain was automatically measured using EchoPAC version 6.00 (GE Medical Systems, Milwaukee, WI). All measurements were done offline.
The technique of strain measurement requires manually outlining the LV endocardial contour, and afterward, the system automatically generates myocardial contour in the late systolic phase. Manual correction on the basis of the instructions of the program is entered later.
Segments with poor visualization were excluded from further analysis. Patients in whom more than four segments could be analyzed were excluded. The system generated curves of LS for each segment of the left ventricle, from which we estimated peak LS during the cardiac cycle ( Figure 2 ). Global LS (GLS) was calculated as the average of the observed segmental values of peak LS from the three apical views. In cases of anterior STEMI, anterior GLS was assessed ( Figure 1 ).
Three-Dimensional Echocardiography
Real-time three-dimensional transthoracic echocardiography was performed using a volumetric probe (GE Vingmed Ultrasound AS). This imaging allows the registration of a sector with a depth of 30° and a width of 100° in real time. To obtain large volumes, full-volume acquisition with electrocardiographic tagging was performed. With dedicated software, four to seven small real-time subvolumes were acquired from alternate cardiac cycles and combined to provide a larger pyramidal volume and to ensure complete capture of the left ventricle. The frame rate of the volumetric image was 15 to 24 frames/sec. With the use of TomTec software (TomTec, Munich, Germany), three-dimensional LV end-diastolic volume (LVEDV), LV end-systolic volume, and LV ejection fraction (LVEF) were obtained offline. From the curves of subvolume changes for particular segments, local left end-diastolic subvolumes were measured. Subsequently, during follow-up, the relative changes of segment subvolumes were calculated.
Long-Term Follow-Up
Three months after STEMI, two-dimensional and three-dimensional echocardiography were performed. With the result of contractility index of a particular segment in the first study and after 3 months as our basis, we defined the segments as follows: normal segments, reversibly dysfunctional segments, and irreversibly dysfunctional segments. Normal segments showed normal function in both the baseline and 3-month studies, reversible dysfunction indicated abnormal baseline but normal 3-month function, and irreversible dysfunction indicated abnormal baseline and 3-month contractility.
The definition of remodeling, following definitions from other publications, was an LVEDV increase of >20% compared with the echocardiographic study performed at discharge. Measurement of LV end-diastolic subvolumes after 3 months was defined as the relative change adjusted to every segment.
Statistical Analysis
Statistical analysis of data was performed using Statistica version 8.0 (StatSoft Inc., Tulsa, OK). Quantitative variables are presented as mean ± SD and qualitative variables as crude values and percentages. Normal distribution of data was verified with Kolmogorov-Smirnov, Lilliefors, and Shapiro-Wilk tests. The uniformity of variances was tested using Levene’s test.
Bivariate analyses for group comparisons of continuous variables were performed using Student’s t test for paired or unpaired data or the Mann-Whitney U test. Chi-square or Fisher’s exact tests were used to compare categorical variables. Correlation between quantitative data was assessed on the basis of Spearman’s rank coefficient and its statistical significance.
To assess determinants of LVr, as a dependent variable we estimated crude and logistic odds ratios (ORs) with their 95% confidence intervals (CIs). Crude ORs were calculated using χ 2 tests. Logistic ORs were calculated in a multivariate analysis, which was performed in the model of logistic forward stepwise regression. Independent variables were chosen on the basis of the findings of bivariate analyses. Subjectively, those with “borderline significance” ( P < .10) were included. P values < .05 were considered to indicate statistical significance.
The ability of continuous variables to predict LVr was verified on the basis of receiver operating characteristic (ROC) curve analysis. Overall accuracy, sensitivity, specificity, and positive and negative predictive values for optimal cutoff points were calculated. Diagnostic accuracy was defined as the ratio of true-positives and true-negatives to total test results. Area under the ROC curve was also estimated.
To evaluate the reliability of echocardiographic results, interobserver and interobserver variability was assessed. Twenty subjects were randomly chosen for that analysis. The coefficient of variability was calculated as the ratio of the standard deviation to the mean for individual measures. Variability was measured for GLS and subvolume assessment by real-time three-dimensional echocardiography. Intraobserver variability was 1.38% (95% CI, 1.16%–1.62%) for GLS and 3.51% (95% CI, 2.9%–4.2%) for subvolumes, and interobserver variability was 1.81% (95% CI, 1.55%–2.13%) and 4.1% (95% CI, 3.3%–4.7%), respectively.
Results
Seventy-one patients satisfied baseline inclusion criteria. Five patients were further excluded: one patient died during follow-up, and four patients were disqualified from STE because of unfavorable anatomic conditions that made echocardiographic visualization imperfect. As a result, 66 patients with STEMI were investigated in the follow-up study. Demographic data, coronary artery diseases risk factors, biochemical indicators, and echocardiographic and electrocardiographic data are presented in Table 1 . Anterior wall STEMI was diagnosed in 31 patients (47%).
| Parameter | Value |
|---|---|
| Men | 78.8% |
| Age (y) | 59.6 ± 10.3 |
| Duration of hospitalization (d) | 5.2 ± 1.1 |
| Diabetes | 21% |
| Hyperlipidemia | 74% |
| Hypertension | 76% |
| Smoking | 59% |
| Renal failure | 1.5% |
| Angina before infarction | 43% |
| Anterior infarction | 47% |
| Maximum troponin (μg/L) | 13.6 ± 10.4 |
| ΣST before pPCI (mm) | 8.3 ± 5.0 |
| ST max before pPCI (mm) | 2.9 ± 1.4 |
| ΣST 60 min after pPCI (mm) | 2.6 ± 3.1 |
| LVEDV (ml) | 96.8 ± 24.2 |
| LVESV (ml) | 48.2 ± 14.8 |
| LVEF (%) | 49.7 ± 9.2 |
| GLS (%) | −14.9 ± 4.6 |
| Anterior GLS (%) | −16.1 ± 6.8 |
| Nonanterior GLS (%) | −13.5 ± 5.4 |
| WMSI | 1.23 ± 0.20 |
Correlation Between the Change in Regional Strain Parameter and Wall Motion Analysis
The study population was divided into two groups on the basis of echocardiographic study results at 3 months: 44 patients (67%) without remodeling and 22 patients (33%) with remodeling. Characteristics of both groups are included in Table 2 . Among 31 patients with anterior wall STEMI, LVr was present in 13 (42%). On the basis of first and follow-up myocardial wall contractility, we analyzed 1,041 segments. The segments were divided into normal ( n = 842), reversibly dysfunctional ( n = 68), and irreversibly dysfunctional ( n = 131). Peak LS was −15.3 ± 8.1 for all segments and −17.4 ± 7.5, −13.4 ± 6.9, and −5.2 ± 6.3 for normal, reversibly dysfunctional, and irreversibly dysfunctional segments, respectively ( P < .05; Figure 3 ).
| Parameter | Patients without LVr ( n = 44) | Patients with LVr ( n = 22) | P |
|---|---|---|---|
| Men | 77% | 81% | .67 |
| Age (y) | 60.2 ± 10.1 | 58.5 ± 10.9 | .52 |
| Duration of hospitalization (d) | 5.1 ± 09 | 5.5 ± 1.1 | .72 |
| Diabetes | 13.6% | 36.4% | .03 |
| Hyperlipidemia | 70.5% | 81.2% | .32 |
| Hypertension | 75.0% | 77.3% | .83 |
| Smoking | 56.8% | 63.6% | .59 |
| Anterior wall MI | 36.6% | 63.6% | .06 |
| Maximum troponin (μg/L) | 11.2 ± 4.5 | 17.5 ± 6.8 | .01 |
| ΣST before PCI (mm) | 7.9 ± 4.6 | 9.4 ± 5.82 | .33 |
| ST max before PCI (mm) | 0.9 ± 0.6 | 1.7 ± 1.4 | .01 |
| ΣST 60 min after PCI (mm) | 2.0 ± 167 | 4.2 ± 5.0 | .02 |
| LVESV (mL) | 47.9 ± 14.5 | 48.8 ± 15.5 | .83 |
| LVEDV (mL) | 99.1 ± 24.1 | 92.0 ± 24.0 | .91 |
| LVEF (%) | 51.2 ± 9.1 | 46.8 ± 8.8 | .06 |
| GLS (%) | −15.9 ± 3.6 | −12.9 ± 5.9 | .01 |
| Anterior GLS (%) | −17.8 ± 5.4 | −12.6 ± 8.3 | .004 |
| Nonanterior GLS (%) | −13.6 ± 4.9 | −13.3 ± 6.3 | .82 |
| WMSI | 1.17 ± 0.17 | 1.36 ± 0.21 | .005 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
