Author + information
- Received April 28, 2003
- Revision received July 24, 2003
- Accepted July 28, 2003
- Published online January 7, 2004.
- Martin Penicka, MD*,
- Jozef Bartunek, MD, PhD*,
- William Wijns, MD, PhD*,
- Ilse De Wolf, RN*,
- Guy R Heyndrickx, MD, PhD*,
- Herbert De Raedt, MD*,
- Emanuele Barbato, MD* and
- Bernard De Bruyne, MD, PhD*,* ()
- ↵*Reprint requests and correspondence:
Dr. Bernard De Bruyne, Cardiovascular Center, OLV Ziekenhuis, Moorselbaan 164, 9300 Aalst, Belgium.
Objectives We tested the hypothesis that the tissue Doppler imaging (TDI)-derived positive preejection velocity (+VIC) can predict the recovery of contractile function after revascularization in patients with a recent myocardial infarction.
Background In experimental studies, the presence and extent of TDI-derived +VIC correlated with the extent of viable myocardium.
Methods Forty-three patients with a large myocardial infarction and an occluded left anterior descending (n = 38) or dominant right coronary (n = 5) artery were selected. The median duration of occlusion was 24 h. Longitudinal myocardial velocities were recorded at rest by pulsed-wave TDI echocardiography 6 ± 2 h after revascularization. Functional recovery was defined as an increase in segmental chordal shortening ≥10% at three-month follow-up left ventricular angiogram as compared with baseline.
Results A good quality TDI signal was obtained in 309 of 324 analyzed segments (95.4%). Severe dysfunction was present in 198 segments of which 126 (64%) showed recovery at three-month follow-up. Sampling of all dysfunctional segments lasted 11 ± 4 min per patient. Sensitivity, specificity, and accuracy of the +VIC to predict segmental recovery were 91%, 71%, and 84%, respectively. The percentage of segments that were dysfunctional at angiography but showed a +VIC correlated with improvement of both global left ventricular ejection fraction (r = 0.60, p = 0.001) and wall motion score index (r = −0.78, p < 0.0001) at follow-up.
Conclusions Assessment of +VIC by pulsed-wave TDI is a simple and accurate method that predicts recovery of contractile function after revascularization in patients with a recent myocardial infarction.
The extent of dysfunctional, but viable, myocardium is a major prognostic determinant in patients with coronary artery disease and left ventricular (LV) dysfunction (1). Therefore, the assessment of viability in dysfunctional segments has important therapeutic implications. Yet, the current methods are expensive, time consuming, or operator-dependent. Tissue Doppler imaging (TDI) allows assessment of myocardial velocities with an excellent temporal resolution (2–4). Normal myocardial velocity pattern during the preejection period (i.e., between the onset of the ventricular activation and aortic valve opening) consists of a positive wave often followed by a small negative wave reflecting the brief inward and outward wall motion during this phase (2–5). Recent experimental data (4)suggest that analysis of preejection velocities is highly sensitive to blood supply. Furthermore, in an animal model of acute myocardial infarction (MI) (6), persistence of a positive velocity in reperfused dysfunctional segments indicated nontransmural necrosis. The aim of the present study is to determine whether TDI-derived preejection velocities can predict recovery of contractile function in humans.
Forty-three patients were selected prospectively using the following criteria: 1) documented acute or subacute MI with severe hypokinesia or akinesia (segmental chordal shortening <20% as assessed by biplane LV angiogram) of a large myocardial area supplied by an occluded left anterior descending artery (LAD) (n = 38) or dominant right coronary artery (RCA) (n = 5); 2) successful percutaneous transluminal coronary revascularization of the occluded artery; 3) sustained patency of this artery at follow-up angiogram; 4) absence of other significant coronary stenoses requiring a revascularization procedure; 5) patient's agreement to undergo a control LV and coronary angiogram after three months. Acute MI was defined as ischemic chest pain for at least 30 min with onset ≤12 h, along with electrocardiographic criteria of ST-segment elevation MI. Subacute MI was defined as documented infarction (chest pain, electrocardiogram [ECG] changes, and laboratory markers) with time interval between onset and revascularization from 12 h to six months. In patients with subacute infarction, the decision to perform left heart catheterization and revascularization was guided by symptoms. Patients with old MI (>6 months), suboptimal baseline and/or follow-up angiograms, atrial fibrillation, bundle branch block, LV hypertrophy, or more than mild valvular disease were excluded. Poor two-dimensional image quality at echocardiography was not considered as an exclusion criterion. The presence and the number of Q waves were assessed using the Selvester method (7). The study protocol was approved by the Medical Ethical Committee of the OLV Ziekenhuis Aalst, and informed consent was obtained from all patients.
Coronary and LV angiography were performed at baseline and repeated three months later. Biplane LV angiography was acquired in the right anterior oblique (30°) projection (right anterior oblique artery [RAO]) and in the left anterior oblique (60°) plus 20° cranial angulation projection (left anterior oblique artery [LAO]). Baseline and follow-up LV angiograms were analyzed off-line by the same operator. Ectopic and two beats following an ectopic beat were excluded from analysis. Left ventricular volumes and ejection fraction were assessed from biplane LV angiograms by the area-length method (8,9).
Regional function was assessed using the centerline method (10). Briefly, end-diastolic and end-systolic LV endocardial contours were traced manually. A centerline was computed midway between the two contours. Wall motion was measured along 100 chords constructed perpendicularly to the centerline. Segmental chordal shortening (%) was averaged in eight predefined segments (10–12): in the RAO projection: anterobasal (chords 1 to 16), anterolateral (chords 17 to 32), apical (chords 33 to 48), inferior (chords 49 to 64), inferobasal (chords 65 to 80); in LAO projection: midposterolateral (chords 31 to 46), midseptal (chords 68 to 84), and basal septal (chords 84 to 100). The anterobasal, anterolateral, apical, basal, and midseptal segments were considered to belong to the perfusion territory of the LAD (10,12,13). The inferior, inferobasal, and midposterolateral segments were considered to belong to the territory of the RCA. In each segment, the regional function was graded as follows: normokinesis and mild hypokinesis = score 1 (segmental chordal shortening ≥20%), severe hypokinesis = 2 (segmental chordal shortening ≥10 < 20%), akinesis = 3 (segmental chordal shortening ≥0 < 10%), and dyskinesis = 4 (segmental chordal shortening <0%). A wall motion score index was calculated as the sum of scores divided by the total number of analyzed segments (14).
Definition of viability
Dysfunctional segments were defined as viable if they exhibited an increase in segmental chordal shortening ≥10% U between baseline and follow-up as assessed by LV angiogram. Intra- and interobserver variability for measurement of segmental chordal shortening, determined in 20 study patients (100 segments), were 3 ± 2% U and 4 ± 2% U, respectively.
Pulsed-wave TDI echocardiography was performed within 24 h of revascularization using a commercially available ultrasound system with TDI capabilities (Acuson Sequoia C256, Mountain View, California). To determine the preejection period, aortic flow was recorded by pulsed-wave Doppler at the level of LV outflow tract at the beginning and at the end of each examination. Preejection was defined as a time interval between the onset of QRS complex and the aortic flow (Fig. 1). From the apical two-chamber view and the apical long-axis view, longitudinal myocardial velocities were assessed in the same eight predefined myocardial segments as described on the biplane LV angiogram. The correspondence between the myocardial segments defined on the biplane LV angiogram and on the two apical echocardiographic views is illustrated in Figure 2(15,16). A sample volume of 5 mm was positioned in the center of each myocardial segment parallel to the analyzed vector of regional motion. Gains and filters were adjusted to obtain an optimal tissue signal. Sweep speed was set up at 150 mm/s. Heart rate and blood pressure were monitored during each examination. All studies were saved on S-VHS videotapes and analyzed off-line. The average value from three beats was taken for each measurement. Figure 3shows a representative TDI tracing from an akinetic anterior wall in two patients with large anterior infarction, one with and one without recovery. The extent of dysfunctional, but viable, myocardium was calculated as the ratio of the total number of dysfunctional segments with preserved positive preejection velocity (+VIC) at TDI divided by the total number of segments (n = 8) analyzed at LV angiography (17).
Feasibility and reproducibility
The quality of the two-dimensional echocardiographic images was graded as “good” or “poor” by consensus of two experienced operators. The time needed to perform TDI assessment of all dysfunctional segments was recorded. To study the learning curve, and the intra- and interobserver variability, the presence or absence and the amplitude of +VIC were assessed in 42 segments from 21 patients (two per patient) during three consecutive weeks (seven patients per week). The second observer was an operator trained in conventional echocardiography but without experience with TDI. Each observer had to place sample volume de novo in the segment under study. Care was taken to sample the middle zones of the segments under study at end-expiratory apnea. Examination by beginner and expert were performed 5 min apart.
Data are expressed as mean or median ± SD. Differences between continuous data were tested by unpaired ttest, as appropriate. To study correlations, the Pearson correlation was used. Differences between proportional (discrete or categorical) data were tested by Fisher exact test. Accuracy of the TDI-derived +VIC to predict recovery of regional contractile function was defined as: [true positive + true negative measurements]/total number of measurements. A p value >0.05 was considered statistically nonsignificant.
Table 1shows the baseline characteristics of the patients. In the majority of them, the culprit lesion was located in the proximal or mid-LAD. Twenty patients had acute and 23 had subacute MI. The median duration of an occlusion of the epicardial artery was 24 h. Twenty-six patients presented with Q waves on admission ECG. There were no significant changes in medication during follow-up including dose of betablockers and angiotensin-converting enzyme inhibitors. A total of 324 segments in 43 patients was analyzed by both TDI echocardiography and LV angiography, of which 198 were dysfunctional (82 severely hypokinetic, 116 akinetic). At follow-up, 126 (64%) showed functional recovery.
Feasibility and reproducibility
Two-dimensional echocardiographic image quality was graded as poor in 11.4% (37/324) of analyzed segments. Nevertheless, a satisfactory TDI signal was obtained in 95.4% (309/324) of segments (p < 0.01). The assessment of all dysfunctional segments lasted 11 ± 4 min per patient.
Interobserver agreement in detection of +VIC was 71% (10/14), 93% (13/14), and 100% (14/14), in the first, second, and third consecutive week, respectively. Interobserver variability for measurement of +VIC was 41.9 ± 39.7%, 19.4 ± 25.3%, and 9.2 ± 7.2%, in the first, second, and third consecutive week, respectively. Intraobserver variability for measurement of +VIC was 6.9 ± 4.3%, 7.5 ± 5.9%, 7.8 ± 5.7%, respectively.
Preejection velocities versus recovery of regional function
A good agreement was observed between +VIC in dysfunctional segments at baseline and recovery of segmental shortening at follow-up (Table 2). The overall sensitivity, specificity, and accuracy were 91%, 71%, and 84%, respectively. The −VIC and Δ VIC showed poor predictive accuracy and were not associated with an improvement in contractile function.
Preejection velocities versus recovery of global function
In patients with a culprit lesion in the LAD, the percentage of dysfunctional, but viable, segments defined at TDI correlated significantly with improvements of both LV ejection fraction and wall motion score index at follow-up (Fig. 4). Moreover, the percentage of dysfunctional, but viable, segments was inversely associated with baseline LV end-diastolic pressure (r = −0.38, p < 0.05).
Systemic hemodynamics and preejection velocities
No significant relationship was observed between +VIC and heart rate, mean blood pressure, LV end-diastolic pressure, or ejection fraction at baseline. Similarly, no relationship was found between the +VIC and the time interval between infarction and revascularization. In patients with acute infarction, there was an inverse correlation between +VIC and the level of creatine kinase (r = −0.67, p < 0.0001). In patients with subacute infarction, the extent of +VIC correlated with number of Q waves on baseline ECG (r = −0.48, p < 0.01).
The present study investigated whether TDI-derived myocardial preejection velocities predict recovery of contractile function after revascularization of an infarcted territory. To test this concept in humans, TDI measurements were performed in patients with a recent large MI caused by the total occlusion of one large epicardial coronary artery. In these patients, a marked improvement of contractile function can be expected three months after revascularization. The TDI-derived myocardial preejection velocities recorded at baseline were compared with the changes in LV wall motion at biplane LV angiography obtained at a three-month interval. The data indicate that the resting pattern of the +VIC accurately predicts recovery of both regional and global LV function after revascularization. In addition, assessing preejection velocities is simple and, therefore, highly reproducible and associated with a short learning curve.
Doppler echocardiographic assessment of viability
Several techniques are used to assess myocardial viability in patients with MI. Nuclear or magnetic resonance imaging is expensive, time consuming, and often unavailable for clinical decision-making. Furthermore, low-dose dobutamine echocardiography is highly image- and operator-dependent and requires additional inotropic stimulation. Recently, tissue Doppler analyses of ejection velocities, strain, and strain rate were used to improve the accuracy of the viability assessment during inotropic stimulation (18,19), yet the ejection phase is load- and heart-rate-dependent, which clouds its predictive accuracy (4,6,20,21).
TDI-derived assessment of the preejection phase
Experimental data indicate that the +VIC decreases in parallel with a decrease in blood supply (4). Furthermore, in an animal model of acute MI, Pislaru et al. (6)demonstrated that the preejection velocity pattern is highly dependent on the amount of the viable myocardium. In particular, the persistence of the +VIC in reperfused dysfunctional segments was associated with nontransmural necrosis, while its absence indicated transmural infarction. Corroborating these experimental studies, we observed a close association between the TDI +VICs and the recovery of both segmental myocardial shortening and global LV function at LV angiography. These findings support the hypothesis that dysfunctional, but viable, segments (with sufficient amount of viable myocytes) are able to generate enough force to shorten during the very onset of the isovolumic contraction period, when intraventricular pressure is rapidly increasing but still low, while they fail to overcome the high afterload imposed by the ejection phase. In contrast, nonviable segments (insufficient amount of viable myocytes) are characterized by the absence of motion (absent positive velocity) already during the preejection period, and, hence, they remain dysfunctional throughout whole systole.
Clinical feasibility of TDI-derived assessment of preejection velocities
In the present study, several findings indicate that the assessment of preejection velocities by TDI is highly feasible in the routine clinical practice: 1) preejection velocities were detected in the vast majority of patients regardless of the two-dimensional echocardiographic image quality; 2) the assessment of the preejection velocities is highly reproducible and requires a short learning period; 3) corroborating experimental data (4,6), no relationship was found between preejection velocities and systemic hemodynamics or diastolic load. Of note, the extent of positive velocities correlated with the level of creatine kinase and with the extent of necrosis on the surface ECG, suggesting their dependence on structural changes rather than on loading conditions.
The specific design of the study can not provide mechanistic explanation linking preejectional wall motion to myocardial viability. Hence, this study should be considered observational. In this study, the segmental LV function was analyzed by LV angiography. In this regard, the echocardiography may be a superior tool for functional analysis of the LV segmental function as compared with LV angiography. Also, matching of segments between angiography and echocardiography as performed in the study might have introduced some noise in the data. On the other hand, the use of independent technique (angiography) as a reference method allows avoidance of a validation of a new echocardiographic index by the same technique (echocardiography). Also, quantitative and objective analysis of segmental chordal shortening at LV angiography might be superior to semiquantitative and subjective analysis of segmental wall motion at echocardiography.
In this study, the presence of myocardial viability was defined only by the improvement in regional contractile function between baseline and three-month follow-up LV angiograms. It remains the possibility that some of the false positive results (+VIC present, no recovery of contractile function) could be explained by the presence of residual myocardial viability insufficient for recovery of regional contractile function. In this regard, the inclusion of another technique for detection of myocardial viability such as, e.g., fluorodeoxyglucose positron emission tomography, might have improved accuracy of TDI-derived assessment.
In the present study, TDI tended to overestimate the prediction of myocardial viability as far as a recovery of LV function is concerned. This may be related to several factors. First, similar to nuclear techniques, TDI may detect areas of residual viability too small to cause contractile function recovery. However, this small residual viability may account for segment shortening during the preejection period (against low opposite force = low intraventricular pressure). Second, the present study investigated the LV recovery only at three months after revascularization. It remains the possibility that some segments identified as viable by TDI could have shown functional recovery at a longer follow-up. Finally, the tethering of the dysfunctional segment by adjacent normokinetic segments can cause passive motion of the segment during the preejection period resulting in preserved +VIC (4,21,22). To minimize the effect of tethering, a sample volume was always placed in the midportion of each segment under study. Moreover, Edvardsen (4)demonstrated high correlation of preejection velocities with segment shortening by sonomicrometry both at rest and after induction of severe ischemia due to occlusion of the proximal LAD. In contrast, peak systolic velocities during ejection remained positive despite paradoxical lengthening of the segments. Thus, while tethering affects indexes measured during the ejection phase, its effect is probably limited during the preejection period.
It should be noticed that the absolute size of a +VIC is segment-dependent. In basal segments, the absolute size is larger than in apical segments. Yet, in the present study, only the presence or absence of a +VIC was taken into account, not its absolute size.
The present study indicates that the resting pattern (presence or absence) of myocardial +VIC by TDI accurately predicts the recovery of both the regional and global contractile function after revascularization in patients with a large MI. The detection of preejection velocities by TDI is a feasible and highly reproducible technique that is readily available in daily clinical practice. The present results obtained in patients with acute and subacute MI warrant further studies to define the usefulness of TDI-derived preejection velocities in predicting the functional recovery in patients with chronic LV dysfunction.
☆ Dr. Penicka was the recipient of a research grant of the Czech Society of Cardiology.
- left anterior descending coronary artery
- left anterior oblique artery
- left ventricle/ventricular
- myocardial infarction
- right anterior oblique artery
- right coronary artery
- tissue Doppler imaging
- positive preejection velocity
- Received April 28, 2003.
- Revision received July 24, 2003.
- Accepted July 28, 2003.
- American College of Cardiology Foundation
- Gorscan J.,
- Strum D.P.,
- Mandarion W.A.,
- et al.
- Derumeaux G.,
- Ovize M.,
- Loufoua J.,
- et al.
- Edvardsen T.,
- Urheim S.,
- Skulstad H.,
- et al.
- Pislaru C.,
- Bruce J.C.h,
- Belohlavek M.,
- et al.
- Sheehan F.H.,
- Bolson E.L.,
- Dodge H.T.,
- et al.
- Austin W.G.,
- Edwards J.E.,
- Frye R.L.,
- et al.
- Sheehan F.H.,
- Schofer J.,
- Mathey D.G.,
- et al.
- ↵Feigenbaum H. Coronary artery disease. In: Echocardiography. 5th edition. Philadephia, PA: Lea & Febiger, 1994:452
- Schiller N.B.,
- Acquatella H.,
- Ports T.,
- et al.
- Erbel R.,
- Schweizer P.,
- Lambertz H.,
- et al.
- Larrazet F.,
- Pellerin D.,
- Daou D.,
- et al.
- Jamal F.,
- Strotmann J.,
- Weidemann F.,
- et al.
- Urheim S.,
- Edvardsen T.,
- Torp H.,
- et al.
- Edvardsen T.,
- Skulstad H.,
- Aakhus S.,
- et al.