Abstract
Background:
In patients with myocardial infarction (MI), microvascular obstruction (MVO) determined by cardiac magnetic resonance imaging (CMR) is associated with left ventricular (LV) remodeling and worse prognosis.
Methods and Results:
In 71 patients with ST-segment elevation MI (STEMI) treated by primary percutaneous coronary intervention (PCI), speckle tracking echocardiography (STE) and CMR were performed early after PCI. All patients underwent CMR at 6 months after hospital discharge to assess the occurrence of LV remodeling. The values of 3-dimensional (3D)-circumferential strain (CS), area change ratio (ACR), and 2-dimensional (2D)-CS were significantly different for the transmural extent of infarct, whereas the values of 3D- and 2D- longitudinal strain (LS) were not significantly different. In transmural infarct segments, the values of 3D-CS and ACR were significantly lower in segments with MVO than in those without MVO. At 6-month follow-up, LV remodeling was observed in 22 patients. In multivariable logistic regression models, global 3D-CS and ACR were significant determinants of LV remodeling rather than the number of MVO segments.
Conclusions:
Regional 3D-CS and ACR reflected the transmural extent of infarct and were significantly associated with the presence of MVO. In addition, global 3D-CS and ACR were preferable to the extent of MVO in the prediction of LV remodeling.
Even in this era of primary percutaneous coronary intervention (PCI) for acute myocardial infarction (AMI), adverse left ventricular (LV) remodeling, which occurs in some patients after successful revascularization and optimal medical treatment, is associated with worse long-term clinical outcome.1
Although infarct size is a major determinant of LV remodeling and a significant predictor of outcome,2,3
recent studies have reported that microvascular obstruction (MVO) is also associated with LV remodeling and relates to prognosis after AMI.4–6
MVO refers to the phenomenon of no-reflow at the tissue level despite restoration of epicardial coronary artery patency. MVO or no-reflow has been assessed by coronary angiography, electrocardiography, contrast echocardiography, and contrast-enhanced magnetic resonance imaging (CE-MRI).7–9
CE-MRI is considered the clinical standard for evaluations of both the presence and extent of MVO and the spatial and transmural extent of the myocardial infarct zone.
Recent studies have shown that myocardial deformation assessed by 2-dimensional (2D) speckle tracking echocardiography (STE) relates to infarct size and the transmural extent of infarct10,11
and is significantly altered by MVO.12
Recently developed 3-dimensional (3D) echocardiography allows assessment of myocardial deformation with 3D-STE from 3D full-volume data. However, there are limited data on the utility and accuracy of 3D strain analysis compared with 2D strain analysis, especially in the acute phase of AMI. Therefore, this study sought to evaluate both the association between 3D strain and the presence of MVO and the predictive value for LV remodeling following ST-elevation MI (STEMI) treated with primary PCI.
Methods
Study Population
The study group prospectively included 75 consecutive patients with a first STEMI. Patients with STEMI who underwent primary PCI within 24 h of the onset of symptoms were enrolled. STEMI was defined as (1) typical chest pain lasting >30 min with presentation within 24 h of symptom onset; (2) ST-segment elevation >0.1 mV within 2 contiguous leads on initial ECG; and (3) elevated creatine kinase-MB (CK-MB) isoenzymes. The exclusion criteria included (1) PCI >24 h of symptom onset; (2) previous MI; (3) previous PCI or coronary artery bypass graft surgery; (4) presence of atrial fibrillation or frequent extrasystoles; (5) hemodynamic or respiratory instability; (6) renal insufficiency (creatinine clearance <40 mL/min), or other contraindications to CE-MRI. All patients underwent comprehensive Doppler echocardiographic examinations and CMR in the acute post-PCI phase. The study was approved by the institutional review board, and all patients enrolled gave informed consent.
Echocardiography
Standard echocardiographic examinations and 3D image acquisition were performed in the acute phase (median 6 days after the onset of AMI) using an Artida ultrasound system (Toshiba Medical Systems Co., Tokyo, Japan). 2D- and 3D-STE were performed using dedicated wall motion tracking software (2D and 3D Wall Motion Tracking, Toshiba Medical Systems Co.). For 2D-STE, 3 parasternal short-axis views at the basal, mid, and apical levels and apical 4-chamber, 2-chamber, and long-axis views were acquired at high frame rate with a PST-25BT probe. 2D-longitudinal strain (2D-LS) was computed from each apical view after the initial manual tracing of the endocardial border and the subsequent automatic tracing of the endocardial and epicardial borders. 2D-circumferential strain (2D-CS) was computed from each parasternal short-axis view. Both global and regional strain values were obtained. Regional strain was measured as the peak systolic strain value of each of 16 LV segments. Global strain was the average of the peak systolic global strain values of cross-sectional views (LS: apical views, CS: short-axis views) calculated automatically by software. Because this software allowed layer-specific strain analysis, LS of the endomyocardial layer and CS of the mid-myocardial layer were measured. For 3D-STE, full-volume ECG-gated 3D data sets were acquired from the apical position with a PST-25SX probe during breath hold over 6 cardiac cycles to ensure an optimal volume rate (median 28.8 vps). The software automatically generated a 3D myocardial contour after manual outlining of the endocardial border (on apical 4- and 2-chamber views at end-diastole) was completed. After these tracings, the myocardial border was manually adjusted on the short-axis images. Next, the 3D endocardial surface was automatically reconstructed, and the 3D strain parameters were calculated by tracking 3D template volumes frame by frame throughout a cardiac cycle.13
We evaluated regional and global 3D strain parameters: 3D-LS, 3D-CS, and area change ratio (ACR), which represent the changes in the endocardial surface area (Figure 1).14
All strain parameters were determined as peak systolic strain.
The CMR examinations were performed in the acute phase (median 5 days after the onset of MI) and in the chronic phase (6 months after MI) with a 1.5T clinical scanner (Magnetom Avanto, Siemens, Erlangen, Germany) using a dedicated 6-element phase-array cardiac coil. The ECG-gated images were acquired during repeated breath holds of varying duration depending on the heart rate. Cine images were obtained using steady-state free-precession cine MRI in the 2-chamber, 4-chamber, and short-axis views. In the cardiac short-axis direction, the LV was completely encompassed by contiguous 8-mm thick slices with a 2-mm interslice gap. Late gadolinium enhancement (LGE) images were acquired in the same orientation as the cine images using a segmented inversion-recovery gradient-echo pulse sequence 5–10 min after injection of 0.2 mmol/kg gadopentetate dimeglumine (Magnevist, Bayer, Berlin, Germany). Analysis of the scans was performed using Argus viewing software (Siemens). LV end-diastolic volume (EDV), LV end-systolic volume (ESV), and LV ejection fraction (EF) were obtained from the short-axis cine-MR images. The transmural extent of infarct and presence of MVO were visually assessed on the LGE images with the 16-segment LV model. MI was defined as a hyper-intense area on the LGE images. In segments with LGE, transmural extent of LGE >50% was defined as a transmural infarct, and transmural extent of LGE ≤50% was defined as a non-transmural infarct. MVO was defined as the presence of hypoenhancement within the hyperenhanced area on the LGE images (Figure 1).
Follow-up
All patients underwent CMR at 6 months after hospital discharge to assess LV remodeling, which was defined as a >5% increase in LVEDV in the chronic phase compared with that in the CMR study in the acute phase.15
Statistical Analysis
Results are expressed as number (%), mean±SD, or median and interquartile range if the variables were not normally distributed. Comparisons of data were performed using Student’s t-test for unpaired continuous variables, the Mann-Whitney U test for continuous variables that were not normally distributed, and chi-square
tests for categorical variables. The diagnostic performance of strain parameters was assessed by receiver-operating characteristic (ROC) curve. The cutoff value of the strain parameter was chosen to maximize the value of (sensitivity+specificity−1). The predictors of LV remodeling were assessed with multivariable logistic regression analysis in which variables that showed significant univariate association were included. Strain parameters and the number of transmural segments were entered into the regression model as binary variables using the cutoff values determined from the ROC analysis. The presence of MVO was entered into the model because the cutoff value of the number of segments with MVO was 1. Intra- and interobserver variability was assessed by 2 independent operators blinded to previous measurements in 5 random patients (80 strain values). Reproducibility was calculated as the absolute difference divided by the mean of the 2 measurements.
A P value <0.05 was considered statistically significant. Statistical analysis was performed using commercially available software: JMP 10 (SAS Institute, Cary, NC, USA). Comparisons of the area under the curve (AUC) were performed with Analyse-it (Analyse-it Software Ltd., Leeds, UK).
Results
Because 4 patients were excluded through loss to follow-up, finally 71 patients were analyzed. Their clinical characteristics are summarized in
Table 1. Among 1,136 LV segments, 1,082 (95.2%) were suitable for assessment by both CMR and STE.
Table 1.
Baseline Characteristics of the STEMI Patients (n=71)
| Variable |
|
| Age, years |
63.8±12.9 |
| Male sex, n (%) |
52 (73.2) |
| Peak CK, IU/L |
1,586 [972–3,428] |
| Peak CK-MB, IU/L |
152 [94–249] |
| Onset-to-balloon time, h |
3.5 [2–6.9] |
| Hypertension, n (%) |
44 (62) |
| Dyslipidemia, n (%) |
36 (51) |
| Diabetes mellitus, n (%) |
18 (25) |
| Smoking, n (%) |
24 (34) |
| Family history of CAD, n (%) |
2 (3) |
| Medications at discharge |
| ACEI/ARB, n (%) |
55 (77) |
| β-blocker, n (%) |
48 (68) |
| Aspirin, n (%) |
70 (99) |
| Thienopyridine, n (%) |
68 (96) |
| Statin, n (%) |
69 (97) |
| Angiography, culprit vessel |
| LAD, n (%) |
33 (46.5) |
| LCx, n (%) |
9 (12.7) |
| RCA, n (%) |
29 (40.8) |
| CMR variables |
| EF, % |
46±10 |
| EDV, mL |
117±35 |
| ESV, mL |
65±30 |
ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; CAD, coronary artery disease; CK, creatine kinase; CMR, cardiac magnetic resonance imaging; EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume; LAD, left anterior descending coronary artery; LCx, left circumflex coronary artery; RCA, right coronary artery; STEMI, ST-segment elevation myocardial infarction.
Based on the CMR findings, 202 (19%) segments were classified as segments with transmural infarct, 161 (15%) segments were classified as segments with non-transmural infarct, and 719 (66.5%) segments showed no LGE. The 3D strain values were significantly different between the no-infarct, non-transmural infarct, and transmural infarct segments (P<0.001), except for the 3D-LS values of no-infarct and non-transmural infarct segments (P=0.22) (Table 2).
Table 2.
Regional Strain Values According to Transmural Extent of Infarct
| Segment |
No infarct
(n=719) |
Non-transmural infarct
(n=161) |
Transmural infarct
(n=202) |
P value ANOVA |
| 3D-CS |
−27.8±10.9* |
−21.9±10.0† |
−14.8±10.9‡ |
<0.0001 |
| 3D-LS |
−10.9±6.2* |
−10.0±6.0 |
−8.0±5.8‡ |
<0.0001 |
| ACR |
−36.2±12.5* |
−30.2±12.0† |
−22.2±13.1‡ |
<0.0001 |
| 2D-CS |
−14.3±5.7* |
−12.3±5.4† |
−9.7±5.5‡ |
<0.0001 |
| 2D-LS |
−12.6±5.2* |
−10.9±4.8† |
−9.7±5.5 |
<0.0001 |
*P<0.01 vs. transmural infarct, †P<0.01 vs. no-infarct, ‡P<0.01 vs. non-transmural infarct. ACR, area change ratio; CS, circumferential strain; LS, longitudinal strain.
The 2D strain parameters showed similar trends to those of the 3D strain parameters. The values of 2D-CS were significantly different for infarct transmurality (P<0.001). The values of 2D-LS were significantly lower in non-transmural and transmural segments than in non-infarct segments (P<0.001), but there was no significant difference between the non-transmural and transmural infarct segments (P=0.10).
MVO and Segmental Strain
MVOs were observed in 11 of 150 (7%) segments with non-transmural infarct and in 111 of 202 (55%) segments with transmural infarct. In non-transmural infarct segments, none of the 3D and 2D strain parameters were significantly different between segments with and without MVO. In the transmural infarct segments, the values of 3D-CS and ACR were significantly lower in segments with MVO than in those without MVO (CS, −12.3±9.9 vs. −17.9±11.3, P<0.0001; ACR, −19.3±11.7 vs. −25.8±13.9, P<0.0001, respectively) (Figure 2). However, 3D-LS did not significantly differ between the segments with and without MVO (−7.7±5.3 vs. −8.4±6.3, P=0.38). The values of 3D-CS and ACR were significantly different between non-transmural infarct segments, transmural infarct segments without MVO, and transmural infarct segments with MVO (P<0.0001). The values of 3D-LS in non-transmural infarct segments were significantly higher than those in transmural infarct segments with and without MVO. The value of 2D-CS in segments with MVO was significantly lower than that in segments without MVO (−8.7±5.2 vs. −11.0±5.5, P=0.01), whereas 2D-LS showed no significant difference for the presence of MVO (−9.1±4.8 vs. −10.4±6.2, P=0.10). As with the 3D strain parameters, the value of 2D-CS was significantly different between non-transmural infarct segments, transmural infarct segments without MVO, and transmural infarct segments with MVO (P<0.0001), whereas the values of 2D-LS in non-transmural infarct segments were significantly higher than those in transmural infarct segments with and without MVO. 3D-CS (AUC 0.80, 95% confidence interval (CI) 0.76–0.84), ACR (AUC 0.79, 95% CI 0.75–0.83), and 2D-CS (AUC 0.75, 95% CI 0.70–0.80) showed fair accuracy for predicting the presence of MVO; however, both 3D-LS (AUC 0.64, 95% CI 0.58–0.69) and 2D-LS (AUC 0.64, 95% CI 0.59–0.70) showed poor accuracy for predicting the presence of MVO (Figure 3). Moreover, the accuracy of 3D-CS was better, but not significantly so, than that of 2D-CS (P=0.06), and ACR was not significantly different compared with 2D-CS (P=0.15).
At 6-month follow-up, 22 of 68 patients (32%) showed LV remodeling. The characteristics of the patients with and without LV remodeling can be compared in
Table 3. Patients with LV remodeling showed significantly lower 3D global strain values than those without LV remodeling, except for 3D global LS, whereas the 2D global strain values were not significantly different. The number of segments with MVO was higher in the patients with LV remodeling than in those without LV remodeling, although the number of transmural infarct segments was not significantly different. Global 3D-CS and ACR showed fair accuracy for predicting LV remodeling, with an AUC of 0.73 (95% CI 0.60–0.86), sensitivity of 84%, and specificity of 71% for a cutoff value of −23% for global 3D-CS, and AUC of 0.72, sensitivity of 84%, and specificity of 58% for a cutoff value of −31% for global ACR. In contrast, global 3D-LS showed poor accuracy for predicting LV remodeling, with an AUC of 0.67, sensitivity of 47%, and specificity of 83% for a cutoff value of −7.5%. 3D global strain parameters and the presence of segments with MVO were determined to be significant predictors of LV remodeling in the univariate logistic regression analysis (Table 4). Stepwise logistic regression analyses were then performed for each 3D strain parameter because the global 3D strain parameters strongly correlated with each other (global 3D-CS vs. global 3D-LS, r=0.73, P<0.0001; global 3D-CS vs. global ACR, r=0.97, P<0.0001; and global 3D-LS vs. global ACR, r=0.85, P<0.0001). Stepwise logistic regression models including each 3D strain parameter, the presence of MVO, and the use of β-blockers revealed that the global strain parameters were significant variables, whereas the presence of MVO was not statistically significant.
Table 3.
STEMI Patients With and Without LV Remodeling
| Variable |
LV remodeling |
P value |
| Yes (n=19) |
No (n=49) |
| Age, years |
62.2±14.3 |
63.8±12.5 |
0.65 |
| Male, n (%) |
16 (84) |
34 (69) |
0.35 |
| Peak CK, IU/L |
2,597 [1,550–4,532] |
1,522 [931–2,999] |
0.04 |
| Culprit vessel; LAD, n (%) |
11 (57) |
22 (49) |
0.42 |
| Onset-to-balloon time, h |
6 [3–9] |
3 [2–4.8] |
0.05 |
| ACEI/ARB, n (%) |
16 (84) |
38 (78) |
0.53 |
| β-blocker, n (%) |
18 (94) |
29 (59) |
0.002 |
| Statin, n (%) |
18 (94) |
48 (97) |
0.50 |
| EF (acute phase), % |
43.6±10.4 |
45.7±9.6 |
0.43 |
| EDV (acute phase), mL |
109.2±37.0 |
121.7±35.1 |
0.19 |
| ESV (acute phase), mL |
63.1±32.4 |
67.7±28.9 |
0.56 |
| EF (follow-up) |
47.9±9.7 |
52.9±10.4 |
0.073 |
| EDV (follow-up) |
134.9±41.9 |
105.4±34.1 |
0.004 |
| ESV (follow-up) |
71.6±33.4 |
51.9±24.0 |
0.009 |
| DT, ms |
213±43 |
197±47 |
0.19 |
| E/e’ |
10.4±4.2 |
9.7±2.7 |
0.43 |
| Global 3D-CS, % |
−20.0±5.2 |
−25.1±7.3 |
0.008 |
| Global ACR, % |
−27.6±6.6 |
−33.2±8.7 |
0.014 |
| Global 3D-LS, % |
−8.6±2.7 |
−10.2±3.0 |
0.051 |
| Global 2D-CS, % |
−13.3±2.8 |
−12.7±3.4 |
0.55 |
| Global 2D-LS, % |
−10.5±2.8 |
−11.6±3.8 |
0.28 |
| No. of transmural infarct segments |
3.6±2.8 |
2.7±2.5 |
0.21 |
| No. of segments with MVO |
2.6±2.2 |
1.5±2.1 |
0.056 |
DT, mitral deceleration time; MVO, microvascular obstruction. Other abbreviations as in Tables 1,2.
Table 4.
Logistic Regression Analysis for LV Remodeling (Stepwise Forward Regression)
| Predictors |
Univariable |
Model 1
Global 3D-CS |
Model 2
Global 3D-LS |
Model 3
Global ACR |
| OR (95% CI) |
P value |
OR (95% CI) |
P value |
OR (95% CI) |
P value |
OR (95% CI) |
P value |
| Use of β-blockers |
12.41
(1.53–100.63) |
0.0016 |
|
NS |
10.98
(1.31–92.15) |
0.027 |
|
NS |
| Onset-to-balloon time |
1.10
(0.97–1.24) |
0.14 |
|
|
|
|
|
|
| Global 3D-CS >−23% |
8.83
(2.68–35.52) |
0.0002 |
8.83
(2.68–35.52) |
0.0002 |
|
|
|
|
| Global 3D-LS >−7.5% |
4.38
(1.36–14.75) |
0.013 |
|
|
3.86
(1.09–13.60) |
0.035 |
|
|
| Global ACR >−31% |
7.2
(2.05–34.04) |
0.0014 |
|
|
|
|
7.2
(2.05–34.04) |
0.0014 |
No. of transmural
infarct segments >1 |
2.91
(0.90–11.40) |
0.07 |
|
|
|
|
|
|
| Presence of MVO |
4.82
(1.49–18.8) |
0.0074 |
|
NS |
|
NS |
|
NS |
CI, confidence interval; OR, odds ratio. Other abbreviations as in Tables 2,3.
Respective intra- and interobserver variabilities were 0.6% and 1.4% for 3D-CS, 3.7% and 4.3% for 3D-LS, 0.1% and 3.7% for ACR, 2.5% and 9.1% for 2D-CS, and 0.5% and 1.4% for 2D-LS.
Discussion
The present study demonstrated that regional 3D strain reflects the transmural extent of infarct, and that the lower strain value in segments with transmural infarct is associated with the presence of MVO in the acute phase of STEMI. Furthermore, 3D global strain may provide significant prognostic information for LV remodeling at 6 months after primary PCI, independent of CMR. The accuracy of 3D STE in predicting the presence of MVO was comparable to that of 2D STE, whereas 3D-STE was preferable to 2D-STE in the prediction of LV remodeling.
This study is one of the few to evaluate the relationship between 3D strain parameters and infarct transmurality and the presence of MVO in the acute phase of STEMI and to prospectively assess the predictive value of 3D strain parameters for LV remodeling. Previous studies have shown a significant association between 3D strain parameters and both segmental and global infarct size in the chronic phase of MI;16,17
however, no studies have focused on 3D-STE in the acute phase of MI. It is important for the management of patients with AMI to assess the extent of transmural infarct and the presence of MVO early after the initial treatment of AMI. Although CE-MRI is considered to be the clinical standard for the assessment of infarction, it is not easily performed, because of its contraindications such as chronic kidney disease and the long examination time that is unsuitable for unstable patients. The present study’s results have valuable implications because echocardiography is widely and easily available and can be performed even at the patient’s bedside. Moreover, in the present study, 3D-STE was not only inferior to 2D-STE in assessing the extent of myocardial injury but preferable to 2D-STE in the prediction of LV remodeling with comparable reproducibility. These findings support the concept that 3D-STE could provide more accurate deformation data than 2D-STE by overcoming the limitation of 2D-STE. 2D-STE has advantages in image quality and frame rate; however, 2D-STE can only evaluate visualized images and has the unique problem of through-plane motion. In addition, 2D-STE requires multiple steps to assess global deformation, whereas 3D-STE can be performed in a single tracking step.
Transmural Extent of Infarct and Strain Parameters
In this study, the values of ACR and 3D- and 2D-CS were significantly different for the transmural extent of infarct, and LS was not sensitive to the transmurality of infarct segments. Our findings that CS was significantly different for infarct transmurality and LS was not significantly different for the transmural extent of infarct can be explained by the directions of the myocardial fibers in the LV wall. Myocardial ischemia is a wavefront phenomenon extending from the endocardium to the epicardium. The inner subendocardium, in which myocardial fibers are oriented longitudinally, is the most vulnerable to ischemia, and then the mid-myocardium, in which myocardial fibers are oriented circumferentially, is subsequently damaged. Therefore, CS seems to better reflect the transmural extent of infarct than LS. ACR is a unique parameter showing the change in unit area on the endocardial surface, which can be measured by 3D-STE but not 2D-STE.14,18
Although ACR has components of both LS and CS, ACR depends more on CS than LS because of its stronger correlation with CS (r=0.94, P<0.0001) than with LS (r=0.71, P<0.0001), as shown in this study.
MVO and Strain Parameters
MVO is a marker of severe myocardial damage caused by prolonged ischemia, subsequent reperfusion injury, and coronary microembolization of atherosclerotic debris or thrombi.19
In our data, more than 90% of the MVO occurred in the transmural infarct segments. The present study showed that the presence of MVO was significantly associated with lower CS and ACR values, whereas LS was not significantly different for the presence of MVO in transmural infarct segments. These findings indicate that MVO represents more severe myocardial damage extending over the mid-myocardium. In the non-transmural infarct segments, none of the strain values were significantly different for the presence of MVO. However, the number of non-transmural infarct segments with MVO (n=11) was too small to evaluate the effect of MVO on regional strain.
LV Remodeling
The present study showed that global 3D strain indices, especially 3D-CS and ACR, obtained in the acute phase were significantly associated with LV remodeling. We speculate this is because CS and ACR reflect not only infarct transmurality and MVO but also viable myocardium within the infarct area and outer layer. Hung et al reported the association between CS rate rather than LS rate and LV remodeling in their study of 603 patients with MI, suggesting that preserved circumferential function plays an important role in the prevention of LV dilation.20
In the present study, global 3D-CS and ACR were significantly associated with LV remodeling, rather than the presence of MVO, in the multivariable analysis. However, the data did not show the superiority of 3D-STE over CMR assessment because we did not conduct quantitative assessment of infarct area or the area of MVO. Furthermore, global 2D-CS was not associated with LV remodeling, although segmental 2D-CS reflected the transmural extent of infarct and the presence of MVO. Global 2D-CS was calculated as the mean of 3 parasternal short-axis images. We think that an incorrect image plane that was not representative of the infarct area at the ventricular level may have caused inaccurate global strain values.
In addition, we conducted the same analyses among the branches of the coronary arteries (Table 5A,B). The 3D strain indices and extent of infarct were significantly different in patients with the left anterior descending as the culprit vessels in a comparison of all patients, whereas patients with the right coronary artery (RCA) or left circumflex artery (LCx) as culprit vessel showed no significant difference in strain indices and the extent of infarct. However, the number of patients with the LCx or RCA as the culprit vessel was too small to provide statistical power.
Table 5.
(A) Comparisons of Strain Parameters in STEMI Patients With and Without LV Remodeling According to Coronary Artery Branch, (B) Logistic Regression Analysis for LV Remodeling in STEMI Patients With the LAD as Culprit Vessel
| A |
LAD |
LCx |
RCA |
| Yes (n=11) |
No (n=21) |
P value |
Yes (n=4) |
No (n=5) |
P value |
Yes (n=4) |
No (n=22) |
P value |
| Global 3D-CS, % |
−18.0±3.8 |
−23.4±5.5 |
0.0007 |
−22.7±6.9 |
−21.5±6.1 |
0.78 |
−22.9±5.4 |
−27.5±8.5 |
0.31 |
| Global 3D-LS, % |
−6.9±1.8 |
−8.8±2.8 |
0.051 |
−10.8±3.1 |
−10.0±1.3 |
0.60 |
−10.8±1.2 |
−11.6±3.0 |
0.62 |
| Global ACR, % |
−24.2±4.9 |
−30.9±7.1 |
0.009 |
−32.4±3.4 |
−29.7±3.0 |
0.57 |
−31.9±4.6 |
−36.3±9.9 |
0.40 |
| Global 2D-CS, % |
−13.2±3.1 |
−12.1±3.4 |
0.41 |
−12.2±2.0 |
−11.1±2.8 |
0.58 |
−14.1±2.8 |
−13.7±3.2 |
0.82 |
| Global 2D-LS, % |
−9.3±2.2 |
−9.0±3.1 |
0.80 |
−11.0±3.9 |
−11.2±1.8 |
0.96 |
−13.1±2.1 |
−14.5±2.6 |
0.33 |
| No. of transmural infarct segments |
4.1±2.8 |
2.9±3.1 |
0.27 |
3.0±2.9 |
3.4±1.3 |
0.79 |
2.5±2.6 |
2.3±2.0 |
0.85 |
| No. of segments with MVO |
3.0±2.2 |
1.4±2.2 |
0.068 |
2.2±2.0 |
3.4±1.3 |
0.34 |
1.7±2.2 |
1.1±1.7 |
0.51 |
| B |
|
|
|
|
|
|
|
|
|
| Predictor |
OR (95% CI) |
P value |
|
|
|
|
|
|
| Global 3D-CS >−22% |
11.25 (1.85–68.13) |
0.0032 |
|
|
|
|
|
|
| Global 3D-LS >−7.5% |
6.7 (1.31–34.02) |
0.015 |
|
|
|
|
|
|
| Global ACR >−31% |
9.0 (1.51–53.41) |
0.0073 |
|
|
|
|
|
|
| No. of transmural infarct segments >1 |
10 (1.09–91.98) |
0.013 |
|
|
|
|
|
|
| Presence of MVO |
7.9 (1.35–45.83) |
0.011 |
|
|
|
|
|
|
LV, left ventricular. Other abbreviations as in Tables 1–4.
A significant association between LV filling and LV remodeling was reported in previous studies.15,21
However, our data showed no difference in mitral deceleration time or E/e’ between patients with and without remodeling. We think this is related to the difference in patient characteristics. In our population, no patient for whom LV remodeling had been predicted had a deceleration time ≤130 ms. Finally, the use of β-blockers was significantly higher in patients with LV remodeling in the present study. Patients without remodeling included those with mild cases who had been considered less likely to have LV remodeling because patients without β-blockers have a significantly lower level of CK compared with those with β-blockers (1,338±1,000 vs. 3,132±2,380 IU/L, P=0.0009). Because β-blockers have been shown to have a beneficial effect on prognosis in patients with MI, our data was not considered to show an unfavorable effect of β-blockers on LV remodeling. Furthermore, stepwise logistic regression analyses did not determine the use of β-blockers as a significant predictor.
Study Limitations
This study has several limitations. First, we assessed the transmural extent of infarct visually rather than with a quantitative method, which is more accurate for assessing infarct size and transmurality. Therefore, our method might not have accurately assessed the relationship between the LGE and STE data. Further studies are needed to validate the usefulness of STE in assessing myocardial damage caused by acute coronary syndrome based on comparisons with more objective methods of measuring LGE burden. Second, segments assessed by STE did not completely correspond with those assessed by CMR. There was the potential for 3D-STE segments corresponding to segments classified as no-infarct based on 2D cross-sectional views of CMR to contain infarct area mainly existing in contiguous segments. Third, patients who were unable to undergo CMR or ultrasonic cardiography within 1 week of PCI were excluded. Therefore, patients in an unstable condition because of heart failure or arrhythmia were not included. As a result, the CK level in the present study was relatively low. Moreover, the present study included patients with symptom onset within 12–24 h, who were likely to show LV remodeling. However, there were no significant differences between patients with onset within 12 h and within 12–24 h in CK level, culprit vessel, LV function or presence of MVO. Although onset-to-balloon time is generally a significant prognostic factor, it was not associated with LV remodeling in the present patient population. We think that further study is needed to confirm our findings of the clinical value of STE in patients unsuitable for CMR.
Conclusion
Regional 2D-CS, 3D-CS, and ACR, which might reflect the transmural extent of infarct, are significantly reduced by the presence of MVO in the acute phase of STEMI. The accuracy of 3D-CS and ACR in predicting the presence of MVO is comparable to that of 2D-CS. Global 3D strain parameters were significant predictors of LV remodeling after adjustment for the presence of MVO. Our findings indicated the usefulness of 3D-STE in the risk stratification of LV remodeling early after primary PCI in patients with STEMI.
References
- 1.
Bolognese L, Neskovic AN, Parodi G, Cerisano G, Buonamici P, Santoro GM, et al. Left ventricular remodeling after primary coronary angioplasty: Patterns of left ventricular dilation and long-term prognostic implications. Circulation 2002; 106: 2351–2357.
- 2.
Chareonthaitawee P, Christian TF, Hirose K, Gibbons RJ, Rumberger JA. Relation of initial infarct size to extent of left ventricular remodeling in the year after acute myocardial infarction. J Am Coll Cardiol 1995; 25: 567–573.
- 3.
Tarantini G, Razzolini R, Cacciavillani L, Bilato C, Sarais C, Corbetti F, et al. Influence of transmurality, infarct size, and severe microvascular obstruction on left ventricular remodeling and function after primary coronary angioplasty. Am J Cardiol 2006; 98: 1033–1040.
- 4.
Wu KC, Zerhouni EA, Judd RM, Lugo-Olivieri CH, Barouch LA, Schulman SP, et al. Prognostic significance of microvascular obstruction by magnetic resonance imaging in patients with acute myocardial infarction. Circulation 1998; 97: 765–772.
- 5.
Orn S, Manhenke C, Greve OJ, Larsen AI, Bonarjee VV, Edvardsen T, et al. Microvascular obstruction is a major determinant of infarct healing and subsequent left ventricular remodelling following primary percutaneous coronary intervention. Eur Heart J 2009; 30: 1978–1985.
- 6.
Hombach V, Grebe O, Merkle N, Waldenmaier S, Hoher M, Kochs M, et al. Sequelae of acute myocardial infarction regarding cardiac structure and function and their prognostic significance as assessed by magnetic resonance imaging. Eur Heart J 2005; 26: 549–557.
- 7.
Ito H, Maruyama A, Iwakura K, Takiuchi S, Masuyama T, Hori M, et al. Clinical implications of the ‘no reflow’ phenomenon: A predictor of complications and left ventricular remodeling in reperfused anterior wall myocardial infarction. Circulation 1996; 93: 223–228.
- 8.
van’t Hof AW, Liem A, de Boer MJ, Zijlstra F. Clinical value of 12-lead electrocardiogram after successful reperfusion therapy for acute myocardial infarction. Lancet 1997; 350: 615–619.
- 9.
Nijveldt R, Beek AM, Hirsch A, Stoel MG, Hofman MB, Umans VA, et al. Functional recovery after acute myocardial infarction: Comparison between angiography, electrocardiography, and cardiovascular magnetic resonance measures of microvascular injury. J Am Coll Cardiol 2008; 52: 181–189.
- 10.
Becker M, Ocklenburg C, Altiok E, Futing A, Balzer J, Krombach G, et al. Impact of infarct transmurality on layer-specific impairment of myocardial function: A myocardial deformation imaging study. Eur Heart J 2009; 30: 1467–1476.
- 11.
Gjesdal O, Hopp E, Vartdal T, Lunde K, Helle-Valle T, Aakhus S, et al. Global longitudinal strain measured by two-dimensional speckle tracking echocardiography is closely related to myocardial infarct size in chronic ischaemic heart disease. Clin Sci (Lond) 2007; 113: 287–296.
- 12.
Bergerot C, Mewton N, Lacote-Roiron C, Ernande L, Ovize M, Croisille P, et al. Influence of microvascular obstruction on regional myocardial deformation in the acute phase of myocardial infarction: A speckle-tracking echocardiography study. J Am Soc Echocardiogr 2014; 27: 93–100.
- 13.
Seo Y, Ishizu T, Enomoto Y, Sugimori H, Yamamoto M, Machino T, et al. Validation of 3-dimensional speckle tracking imaging to quantify regional myocardial deformation. Circ Cardiovasc Imaging 2009; 2: 451–459.
- 14.
Seo Y, Ishizu T, Enomoto Y, Sugimori H, Aonuma K. Endocardial surface area tracking for assessment of regional LV wall deformation with 3D speckle tracking imaging. JACC Cardiovasc Imaging 2011; 4: 358–365.
- 15.
Temporelli PL, Giannuzzi P, Nicolosi GL, Latini R, Franzosi MG, Gentile F, et al. Doppler-derived mitral deceleration time as a strong prognostic marker of left ventricular remodeling and survival after acute myocardial infarction: Results of the GISSI-3 echo substudy. J Am Coll Cardiol 2004; 43: 1646–1653.
- 16.
Hayat D, Kloeckner M, Nahum J, Ecochard-Dugelay E, Dubois-Rande JL, Jean-Francois D, et al. Comparison of real-time three-dimensional speckle tracking to magnetic resonance imaging in patients with coronary heart disease. Am J Cardiol 2012; 109: 180–186.
- 17.
Thorstensen A, Dalen H, Hala P, Kiss G, D’Hooge J, Torp H, et al. Three-dimensional echocardiography in the evaluation of global and regional function in patients with recent myocardial infarction: A comparison with magnetic resonance imaging. Echocardiography 2013; 30: 682–692.
- 18.
Seo Y, Ishizu T, Atsumi A, Kawamura R, Aonuma K. Three-dimensional speckle tracking echocardiography. Circ J 2014; 78: 1290–1301.
- 19.
Bekkers SC, Yazdani SK, Virmani R, Waltenberger J. Microvascular obstruction: Underlying pathophysiology and clinical diagnosis. J Am Coll Cardiol 2010; 55: 1649–1660.
- 20.
Hung CL, Verma A, Uno H, Shin SH, Bourgoun M, Hassanein AH, et al. Longitudinal and circumferential strain rate, left ventricular remodeling, and prognosis after myocardial infarction. J Am Coll Cardiol 2010; 56: 1812–1822.
- 21.
Cerisano G, Bolognese L, Carrabba N, Buonamici P, Santoro GM, Antoniucci D, et al. Doppler-derived mitral deceleration time: An early strong predictor of left ventricular remodeling after reperfused anterior acute myocardial infarction. Circulation 1999; 99: 230–236.
