Some patients do not recover from left ventricular (LV) systolic dysfunction after acute myocardial infarction (AMI) despite being treated successfully with primary angioplasty and receiving optimal medical therapy according to current guidelines. Indeed, the presence and extent of postischemic viable myocardium is an important determinant of functional recovery after AMI, and its quantification has prognostic value in this setting. This is particularly true in anterior AMI, in which the area exposed to ischemia is large. Several tools can assess directly or indirectly the stunned myocardium after AMI, including dobutamine stress echocardiography, nuclear cardiac imaging, and magnetic resonance imaging (MRI). However, nuclear cardiac imaging and MRI are not easily available, are expensive, and involve radiation and gadolinium exposure. Dobutamine stress echocardiography is a subjective test with currently very few indications in the acute phase of AMI. Myocardial contrast echocardiography is also a useful technique providing regional and transmural information, which is a critical determinant of prognosis after AMI. However, this technique requires advanced training and expertise and an intravenous infusion of a microbubble contrast agent. Myocardial deformation imaging by two-dimensional strain (2DS) allows the evaluation of regional and global LV systolic function at rest with high reproducibility and a shallow learning curve, without the limits of Doppler tissue imaging due to its angle dependency. This is of particular importance when assessing viability in anterior AMI, in which the apex is always involved. Some studies have used the quantitative method of 2DS to assess myocardial recovery after AMI, using a global or regional approach. However, with regard to the qualitative evaluation of 2DS and when focusing on the segmental strain pattern aiming to predict recovery and prognosis, no available data currently exist. One study using Doppler tissue imaging before reperfusion showed a positive correlation between longitudinal systolic lengthening (SL) and transmural extent of necrosis in anterior AMI, but such an evaluation has never been performed by 2DS. In a preliminary study involving 21 patients with AMI compared with 21 patients with typical takotsubo syndrome, we found an independent relationship between SL and segmental recovery. The strain pattern analysis by 2DS may offer a new diagnostic capability in patients with AMI, particularly by studying in impaired segments the presence of SL, which identifies a passive myocardium. However, whether this specific analysis could contribute to the important topic of myocardial viability and prognosis in patients with AMI is unclear. Therefore, our objective was to estimate the diagnostic value of the longitudinal strain pattern in segments with wall motion abnormalities (WMAs) to predict recovery after anterior AMI and test its link with in-hospital events, as well as with coronary microvascular impairment (CMI), as already shown in terms of diagnostic and prognostic value in this setting.

Methods

Population Studied

Forty-nine consecutive patients with anterior AMI who underwent successful primary angioplasty at our institution <12 hours after symptom onset and comprehensive transthoracic Doppler echocardiography <24 hours of admission and at follow-up, including 2DS analysis, were prospectively included in the study. All patients were in normal sinus rhythm. Significant valvular disease (at least moderate regurgitation or stenosis), left bundle branch block, and poor echogenicity were exclusion criteria. The diagnosis of AMI was based on chest pain lasting >30 min, ST-segment elevation > 2 mm in at least two contiguous precordial electrocardiographic leads, and increase in serum troponin T. Diagnostic coronary angiography using the radial or femoral approach and coronary angioplasty were performed by standard techniques. During the procedure, glycoprotein IIb/IIIa inhibitors, thrombectomy (Export; Medtronic, Inc, Minneapolis, MN), and intracoronary vasodilators (adenosine and verapamil) were used at the discretion of the interventional cardiologist. Successful angioplasty, required for inclusion in the study, was defined as a final angiographic Thrombolysis In Myocardial Infarction flow grade of 3 with a residual stenosis <30% in the left anterior descending coronary artery (LAD). All patients received aspirin (250–500 mg), a P2Y ₁₂ inhibitor (a loading dose of clopidogrel, prasugrel, or ticagrelor) and intravenous bolus of heparin (5,000 U) before angiography. After the procedure, all patients received medical therapy according to current guidelines for ST-segment elevation myocardial infarction. All patients gave informed consent for the protocol.

Conventional Echocardiography

Comprehensive transthoracic Doppler echocardiography was performed <24 hours after angiography and at 6-month follow-up, and all echocardiograms were digitized online and stored on a workstation (EchoPAC 7 version 108 for PC; GE Medical Systems, Waukesha, WI) for subsequent offline analysis by two observers blinded to patient data. LV end-diastolic volume and end-systolic volume were measured from the apical four- and two-chamber view and LV ejection fraction (LVEF) was calculated using the modified biplane Simpson rule. Left atrial volume index was measured according to the area-length method and LV mass according to the American Society of Echocardiography formula. Wall motion score index (WMSI) was measured using the 16-segment four-point scaling model from the apical four-, two-, and three-chamber views. WMSI was derived as the sum of all scores divided by the number of segments visualized. Segment scores were as follows: 1 = normal, 2 = hypokinesia, 3 = akinesia, and 4 = dyskinesia. Conventional Doppler parameters were also measured according to a standardized examination: early (E) and late (A), diastolic transmitral flow velocity, and deceleration time of E, average of the septal and lateral annular mitral early diastolic (e′), late diastolic (a′), and systolic (Sa) pulsed-wave tissue Doppler velocity, and the E/e′ ratio. Pulmonary artery systolic pressure was calculated using the modified Bernoulli equation from tricuspid regurgitant peak jet velocity and estimated right atrial pressure (from respiratory variation of inferior vena).

Myocardial Deformation Imaging by 2DS

From the apical long-axis and four- and two-chamber views, LV global longitudinal strain (GLS) by 2DS was quantified as previously described. Briefly, the left ventricle was divided into six segments in each apical view, and tracking quality was validated for each segment. Careful attention was paid to cover all the thickness of the myocardium for tracking. Then myocardial motion was analyzed by speckle-tracking within the region of interest. The automated algorithm provided peak systolic longitudinal strain for each LV segment (total, 18 segments). LAD strain was also measured and was defined as the mean of peak systolic longitudinal strain values from the 11 segments assigned to the LAD territory (see Figure 1 ).The corresponding strain curves were stored for segmental strain analysis in segments with WMAs, focusing, in addition to peak systolic segmental strain, on the strain pattern, measuring the duration and amplitude of longitudinal SL, as well as the occurrence of postsystolic shortening. The duration of SL (when present) was measured, as previously described, as the time interval from end-diastole (peak R in the QRS complex) to the return to zero strain and was reported as a percentage of systolic duration (SL % duration). Systolic duration was defined as from the peak R in the QRS complex of the electrocardiogram to aortic valve closure, which was determined automatically by the software and from appropriate pulsed-wave Doppler images. Furthermore, the systolic lengthening-to-shortening ratio was also measured for all concerned segments as previously described. The postsystolic shortening index was measured and defined as (peak strain − peak systolic strain)/peak strain and expressed as a percentage. Peak strain was measured when further deformation occurred after aortic valve closure. All analyzed images were recorded with a frame rate of ≥60 frames/sec. Images were obtained during breath-hold and saved in cine-loop format from three consecutive beats. All echocardiographic measurements were performed offline (EchoPac) in random order by two independent observers without knowledge of the clinical status of the patient. Figure 1 displays an example of longitudinal strain curve of a segment with WMA and the corresponding measurements cited above and by comparison a superimposed strain curve of a segment without WMA (“normal curve”).

Schematic representation of strain curves. ( Left ) The 18-segment model for strain analysis, from the four-chamber (4C), two-chamber (2C), and three-chamber (3C) apical views. Red crosses indicate the segments assigned to the LAD territory (LAD strain). ( Right ) Representative longitudinal strain curves in a segment with WMA ( green curve ) and in a remote segment without WMA ( yellow curve ). The y axis represents the strain value as a percentage and the x axis the duration of a cardiac cycle from R to R on the electrocardiogram in milliseconds. For the segment with WMA, A indicates the duration of SL during the systolic phase in milliseconds and B the duration of systole in milliseconds, from peak R on the electrocardiogram to aortic valve closure (AVC). Thus, SL in % duration of systole = A/B × 100. L designates the amplitude of lengthening and S the amplitude of total shortening (all in percentage) and L/S their ratio. There is also postsystolic shortening for this segment, whereas the peak strain for the remote segment occurs at AVC.

Coronary Flow Velocity and Coronary Flow Velocity Reserve (CFVR) Assessment

Coronary flow velocity and CFVR assessment was performed as previously described <24 h after angioplasty. Briefly, the mid-distal part of the LAD was studied using a multifrequency transducer (M5S probe), and the artery was visualized by color Doppler flow mapping guidance in the modified parasternal view. For color Doppler echocardiography, the velocity range was defined as 16 to 19 cm/sec. Blood flow velocity was measured by pulsed-wave Doppler echocardiography using a sample volume of 3 to 4 mm, placed on the color signal in the distal LAD, distally to the stent deployment in all cases. The ultrasound beam direction was aligned as closely as possible with the distal LAD flow, and no angle correction was performed. However, the angle was kept as small as possible. A no-reflow pattern of the resting coronary flow velocity was defined as previously described with a deceleration time of diastolic flow velocity ≤ 600 msec and/or systolic flow reversal. CFVR was assessed using intravenous adenosine infusion (140 μg/kg/min over 2 min). CFVR was calculated as the ratio of hyperemic to basal peak diastolic flow velocity. Blood flow velocity measurements were performed offline by an experienced investigator blinded to patient data, by contouring the spectral Doppler signals. Final values of flow velocity represented an average of three cardiac cycles. The baseline coronary flow pattern was available in all cases, whereas CFVR was performed in 37 cases. Coronary microcirculatory impairment was defined as CFVR < 1.7 (see later discussion) or as a no-reflow pattern.

Recovery

Recovery at the segmental level was defined as a normalization of the segmental wall motion (final segmental wall motion score = 1). WMAs were assessed in random order, with independent analysis of baseline and follow-up echocardiographic loops, by two independent observers, blinded to patient data, who have a good experience with analysis of WMAs. The best image quality possible was selected in every patient. Kappa statistics was used to test the intraobserver and interobserver variability of segmental recovery. Recovery at the patient level was defined as an absolute improvement of LVEF of >5%.

In-Hospital Events

In-hospital adverse events were defined as a composite of death, heart failure, and recurrent myocardial infarction. Heart failure was defined as the presence of pulmonary edema, dyspnea, and/or oxygen desaturation requiring drug therapy (diuretics, nitrates) and/or mechanical support. Recurrent myocardial infarction was defined as recurrent clinical symptoms, new electrocardiographic changes, and a significant increase of serum cardiac enzymes or troponin. Two independent investigators blinded to the tests results collected outcome data.

Statistical Analysis

Continuous variables are expressed as mean ± SD and categorical data a percentages. The Kolmogorov-Smirnov test was used to test the normality of a variable. To compare segments with and without recovery, and patients with and without recovery, as well as the changes in some parameters in each subgroup, nonpaired and paired Student’s t tests for continuous variable with normal distributions and nonparametric tests such as the Mann-Whitney and Wilcoxon tests for continuous variables not normally distributed were used according to the variable tested. For categorical variables, χ ² or Fisher exact tests were used. Multivariate logistic regression analysis was used to test the independent correlates of recovery at the segmental and patient levels, as well as of in-hospital events and CMI. Variables with P values < .10 in univariate analysis were included in the multivariate model. Receiver operating characteristic curve analysis was used to determine the best cutoffs identifying recovery at the segmental and patient levels. Linear correlations were tested between acute strain parameters and LV systolic function at follow-up. However, because the number of segments with SL % duration > 40% was associated with tied values, the nonparametric rank correlation of Kendall was more appropriate to test its relationship with LV systolic function at follow-up. The intraobserver and interobserver variability of longitudinal strain parameters was tested in 10 random patients, using the absolute difference divided by the mean of the repeated observations and expressed as a percentage, and using the intraclass correlation coefficient. Statistical analysis was performed using MedCalc version 15.2.1 (MedCalc Software bvba, Ostend, Belgium).