Author + information
- Received February 7, 2012
- Revision received June 4, 2012
- Accepted June 26, 2012
- Published online September 18, 2012.
- Marit Kristine Smedsrud, MD⁎,†,
- Sebastian Sarvari, MD⁎,†,
- Kristina H. Haugaa, MD, PhD⁎,†,
- Ola Gjesdal, MD, PhD⁎,
- Stein Ørn, MD, PhD‡,
- Lars Aaberge, MD, PhD⁎,
- Otto A. Smiseth, MD, PhD⁎ and
- Thor Edvardsen, MD, PhD⁎,†,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Thor Edvardsen, Department of Cardiology, Oslo University Hospital, RIkshospitalet, Postboks 4950 Nydalen, N-0424 Oslo, Norway
Objectives This study sought to investigate whether the duration of left ventricular (LV) early systolic lengthening could accurately identify patients with significant coronary artery disease (CAD).
Background Ischemic myocardium with reduced active force will lengthen when LV pressure rises during early systole before onset of systolic shortening.
Methods We included 88 patients with suspected CAD referred to elective diagnostic coronary angiography. Two of these patients were excluded from the study due to evidence of previous myocardial infarction on contrast-enhanced magnetic resonance imaging. Speckle tracking echocardiography was performed before coronary angiography and at follow-up scheduled 1 year after revascularization, and global longitudinal strain and duration of average LV early systolic lengthening were recorded.
Results Forty-three of 86 patients had significant CAD. The duration of early systolic lengthening was significantly prolonged in patients with significant CAD compared with patients without significant coronary artery stenoses (76 ± 37 ms vs. 38 ± 23 ms, p < 0.001). Correspondingly, global systolic strain was significantly lower in patients with CAD (−17.7 ± 3.0% vs. −19.5 ± 2.6%, p = 0.003). Prolonged duration of early systolic lengthening showed the best accuracy in detecting CAD, with an area under the receiver-operating characteristic curve of 0.83. The area under the curve for global strain was 0.68. At 1-year follow-up, the duration of early systolic lengthening was significantly reduced (64 ± 37 ms vs. 76 ± 37 ms, p = 0.041) in the patients treated with revascularization.
Conclusions Duration of myocardial early systolic lengthening was prolonged in patients with significant CAD; this might be a useful parameter to identify patients who might benefit from reperfusion therapy.
Prospective identification of patients with ischemic heart disease who may benefit from revascularization remains a clinical challenge. The management of suspected coronary artery disease (CAD) poses a central question: in which patients should coronary angiography be performed? Several published series, including the National Heart, Lung, and Blood Institute-sponsored CASS (Coronary Artery Surgery Study) and the WISE (Women's Ischemia Syndrome Evaluation) study, reported that up to one-half of patients who underwent coronary angiography were found to have normal or nonobstructed coronary arteries (1,2). In our laboratory, approximately 15% to 20% of the patients undergoing a clinically indicated elective coronary angiogram have normal or nearly normal appearing coronary arteries with <50% stenosis in any segments.
Two-dimensional speckle tracking echocardiography (2D-STE) is a novel method for assessment of left ventricular (LV) function and ischemic changes (3–5). Several studies showed that significant coronary stenoses might cause persistently impaired longitudinal LV function at rest, which could be detected by myocardial strain imaging (6–8).
When LV pressure rises during early systole, ischemic myocardium will tend to lengthen before onset of systolic shortening due to its reduced ability to generate active force. However, it has not yet been assessed whether the presence of initial stretch provides diagnostic information in patients with suspected CAD. The object of this study was to investigate whether the duration of LV early systolic lengthening measured by 2D-STE at rest could accurately identify patients with significant CAD.
This study was conducted in a single tertiary coronary care center, using a prospective design. A total of 88 consecutive outpatients referred to elective diagnostic coronary angiography because of stable chest pain and/or increased risk profile were included. All patients were tested by exercise electrocardiography (ECG) before coronary angiography, and 77% of them had objective signs of ischemia from exercise ECG. Exclusion criteria were acute coronary syndrome, a history of myocardial infarction (MI) and/or any evidence of scar by late enhancement on contrast-enhanced magnetic resonance imaging (CE-MRI), severe valvular disease, previous heart surgery, atrial fibrillation, and bundle branch block with QRS >120 ms. The study was approved by the Regional Committee for Medical Research (REK Sør, Oslo, Norway), and all subjects gave written informed consent.
All the study examinations were performed using a Vivid 7 scanner (GE Vingmed Ultrasound, Horten, Norway) with a phased-array transducer. Echocardiography was performed immediately before coronary angiography and at follow-up scheduled 1 year later. Patients without significant coronary artery stenosis did not undergo a follow-up examination. Three consecutive heart cycles from the 3 standard apical planes (4-chamber, 2-chamber, and long-axis) were obtained by conventional 2-dimensional grayscale echocardiography. Loops were digitally stored and later analyzed offline using EchoPac version 7.0.0 (GE Vingmed, Horten, Norway). All analyses were performed with drift compensation turned on.
LV volumes were traced manually at end-diastole and end-systole at apical 4- and 2-chamber views, and left ventricular ejection fraction (LVEF) was derived from the modified biplane Simpson's method. Diastolic septum diameter was measured by M-mode in the parasternal long-axis view.
Longitudinal strain was measured by 2D-STE using a 16 LV segment model (9). In each of the apical planes, a region of interest was manually drawn, and tracking of deformation was automatically performed by the software. The average frame rate was 76 ± 16 s−1. For each segment, the peak negative systolic strain value (representing maximum segmental systolic shortening), peak positive early strain (representing maximum segmental systolic lengthening), duration of early systolic lengthening, and post-systolic shortening were recorded by fully automatic software. Values for all analyzed segments were then averaged to obtain an average value. End-systole was defined by aortic valve closure in the apical long-axis view. The duration of LV early systolic lengthening was defined as time from onset of the Q wave on ECG (onset of the R wave if the Q wave was absent) to maximum myocardial systolic lengthening.
Post-systolic shortening was defined as peak negative strain during diastole minus peak negative systolic strain (10). When the minimum segment length occurred within systole, post-systolic shortening was zero by definition. Hence, post-systolic shortening could only take negative values.
Based on a standardized model of myocardial perfusion territories (9), regional strain was calculated as the average value of the segments belonging to each perfusion territory in the left anterior descending, left circumflex, and/or right coronary arteries, respectively.
The echocardiographic data were analyzed blinded to all clinical information.
Magnetic resonance imaging
Cardiac CE-MRI was performed at the follow-up scheduled 1 year after revascularization, using either a 1.5-T unit (Magnetom Sonata, Siemens, Erlingen, Germany) or a 3-T unit (Philips Medical Systems, Best, the Netherlands). The CE-MRI images were obtained by late enhancement 10 to 20 min after intravenous injection of 0.1 to 0.2 mmol/kg gadopentetate dimenglumine (Magnevist, Schering, Berlin, Germany) in multiple short-axis slices covering the entire left ventricle.
Coronary angiography and revascularization
Coronary angiography was performed on clinical indication by standard (Judkins) technique, using digital imaging acquisition and storage. Percutaneous coronary intervention (PCI) was performed at the discretion of the operator. Successful complete revascularization was defined as either post-procedural Thrombolysis In Myocardial Infarction flow grade 3 in all epicardial vessels or bypass surgery.
All continuous data are expressed as mean ± SD. Differences between groups were analyzed by independent Student's t-test. Changes from baseline to follow-up were analyzed with paired samples Student's t-test. Chi-square tests were used to compare discrete data between groups; in those cases where the expected cell count was <5, Fisher's exact test was used.
Univariate and multivariate logistic regression analyses (enter model) were performed to evaluate the association between the presence of significant CAD, gender, systolic blood pressure, diastolic blood pressure, age, LV hypertrophy, diabetes, and the following echocardiographic variables: LVEF, peak positive early strain, duration of early systolic lengthening, and peak systolic strain. All variables found significant on univariate logistic regression were considered for multiple regression analysis. Highly intercorrelated variables were identified by Spearman's correlation coefficient, and only 1 of these variables was entered into the multiple regression model. In addition, odds ratio and 95% confidence intervals (CIs) were calculated. By comparing the c-statistics for 2 prediction models, 1 with duration of early systolic lengthening and 1 with peak systolic strain, the increments in diagnostic utility were quantified (a c-statistic of 1 indicates perfect discrimination, whereas a value of 0.5 indicates no discrimination).
Receiver-operating characteristic (ROC) curve analysis was performed to determine the accuracy of duration of early systolic lengthening and peak systolic longitudinal strain to detect significant CAD, with an area under the curve (AUC) value of 0.50 indicating no accuracy and a value of 1.00 indicating maximal accuracy. The statistical significance of the difference between the areas under the ROC curves was tested using Analyse-it Software (Analyse-it Software Ltd., Leeds, United Kingdom) for Microsoft Excel (11). The optimal cutoffs were defined as the value of the ROC curve that was closest to the upper left corner. The reliability of the optimal cutoff value separating patients with clean coronary arteries from patients with significant CAD was validated using bootstrap resampling (n = 1,000) (12), and a 95% CI based on bootstrap percentiles was presented. The statistical software package R version 2.11.1 (Foundation for Statistical Computing, Vienna, Austria) was used for bootstrap analysis.
Global longitudinal strain and duration of early systolic lengthening were analyzed repeatedly by 2 independent observers on 10 randomly selected patients, and results were expressed as Cronbach's alpha.
All p values are 2-tailed, and a significance level of 0.05 was used. With exception of the comparison of the areas under the ROC curves and the bootstrap analysis, all statistical analyses were performed on SPSS version 16.0 (SPSS Inc., Chicago, Illinois).
In total, 88 patients were included in the study. Forty-five patients had indications for revascularization and were included in the CAD group. In 2 of these patients, evidence of MI by late enhancement was seen on CE-MRI at follow-up, and those 2 patients were excluded from further analyses. The clinical data according to the study groups with the remaining patients are shown in Table 1. The patients with significant CAD were older and more frequently male. They also had more symptoms as assessed by New York Heart Association functional class.
Angiographic findings and revascularization
Angiographic data are presented in Table 2. Of the 86 patients, there were 43 with significant CAD defined as ≥50% diameter stenosis in any coronary artery. Nineteen (44%) patients had single-vessel coronary artery disease, whereas 24 (56%) patients had coronary disease in ≥2 vessels. PCI was performed in 31 patients (72%), whereas 10 patients (23%) underwent bypass surgery. In 2 of the patients with single-vessel disease, the coronary artery stenosis was not suitable for revascularization. Of the 43 patients without significant CAD, there were 28 with clean coronary arteries.
Baseline echocardiographic parameters are presented in Table 3. Patients with CAD had significantly increased peak positive early strain and reduced systolic contraction compared with those without significant coronary artery stenoses. Correspondingly, the duration of early systolic lengthening was significantly prolonged in patients with CAD (Fig. 1). Excluding patients with total or subtotal coronary occlusions from the analyses (n = 10) did not significantly influence the results (duration of early systolic lengthening in patients with significant CAD, but without coronary occlusion, vs. patients without significant CAD: 76 ± 36 ms vs. 38 ± 23 ms, p < 0.001). Consequently, the prolonged duration of early systolic lengthening could not be attributed to the subgroup of CAD patients with total or subtotal coronary occlusions. There were no differences between the patients with and without significant CAD regarding LVEF, post-systolic strain, or the ratio of post-systolic strain to maximal global longitudinal strain.
The possible predictors of CAD found significant on univariate logistic regression are shown in Table 4. LV hypertrophy, diastolic blood pressure, diabetes, and LVEF were not significant predictors of CAD. In a multivariate analysis, duration of early systolic lengthening was a strong and independent predictor of significant CAD (p < 0.001). A close relationship was observed between peak positive early strain and duration of early systolic lengthening and, therefore, peak positive early strain was excluded from the multivariate analysis. However, when exchanging duration of early systolic lengthening for peak positive early strain, peak positive early strain was an independent predictor of significant CAD (p = 0.002). Furthermore, diagnostic utility of the models with duration of early systolic lengthening and peak negative systolic strain, respectively, was assessed by calculating the c-statistic in the multivariate model including each of the parameters. The c-statistic was 0.91 for duration of early systolic lengthening and 0.79 for peak negative systolic strain (p = 0.002).
To test whether the drift compensation had any influence on our results, we selected 5 random patients for an assessment of duration of early systolic lengthening with and without drift compensation. Duration of early systolic lengthening with drift compensation versus without drift compensation was exactly the same (140 ± 89 ms vs. 140 ± 89 ms). Consequently, the drift compensation did not seem to affect the results in this early phase of systole.
Receiver-operating characteristic analysis
Figure 2 illustrates the results of the ROC curve analysis for duration of early systolic lengthening and peak longitudinal strain in the detection of significant CAD. Duration of early systolic lengthening identified patients with significant coronary artery stenosis with a specificity of 86% and a sensitivity of 74%, with an optimal cutoff value of 58 ms. In addition, the optimal cutoff value for peak systolic longitudinal strain was −17.4, yielding a specificity of 81% and a sensitivity of 51%. The AUC for duration of early systolic lengthening was significantly greater than the AUC for peak systolic longitudinal strain (0.83; 95% CI: 0.75 to 0.92 vs. 0.68; 95% CI: 0.56 to 0.79; p = 0.022). The positive and negative predictive values using the optimal cutoff value for duration of early systolic lengthening were 0.84 (95% CI: 0.72 to 0.93) and 0.77 (95% CI: 0.67 to 0.84), respectively, whereas the corresponding values for peak systolic longitudinal strain were 0.73 (95% CI: 0.57 to 0.86) and 0.63 (95% CI: 0.54 to 0.69).
To explore how to separate pathological early systolic lengthening from lengthening occurring in patients without CAD, we compared patients with clean coronary arteries with patients with significant coronary artery stenoses. As determined by ROC curve analysis, a duration of early systolic lengthening of >48 ms (95% CI by bootstrapping: 38 to 58 ms) separated a patient with significant CAD from a patient with clean coronary arteries with a specificity of 82%.
Regional myocardial systolic function
Duration of early systolic lengthening was significantly increased in segments related to the significant stenoses compared with remote myocardial segments (83 ± 42 ms vs. 31 ± 28 ms, p < 0.001), whereas systolic contraction was reduced (−17.9 ± 3.5% vs. −20.1 ± 2.9%, p = 0.015). As determined by ROC analysis, duration of early systolic lengthening could discriminate between segments related to significant stenoses versus segments with normal perfusion with an AUC of 0.88 (95% CI: 0.79 to 0.98). The AUC for systolic longitudinal strain was 0.67 (95% CI: 0.52 to 0.82).
Follow-up echocardiography was performed 12 ± 2 months after revascularization. As demonstrated in Table 5, a persistent impairment of LV systolic function was found in the patients with significant CAD. However, the duration of early systolic lengthening was significantly reduced after intervention. Comparing the patients treated with PCI with those who underwent bypass surgery, there were no differences regarding duration of early systolic lengthening (66 ± 42 ms vs. 62 ± 30 ms, p = 0.823) at follow-up.
Infarction on CE-MRI
Evidence of MI by late enhancement was seen in 2 patients. These patients were excluded from further analysis due to the exclusion criteria.
Feasibility and reliability
Myocardial strain could be assessed in 89% of the myocardial segments. Remaining segments were excluded because of reverberations or poor image quality. Reliability analyses of global longitudinal strain revealed Cronbach's alpha values of 0.96 for intraobserver variability and 0.92 for interobserver variability, and 0.99 for intraobserver variability and 0.93 for interobserver variability for duration of early systolic lengthening.
The present study demonstrated that duration of early systolic lengthening was a strong and independent predictor of significant CAD. An increased duration of early systolic lengthening had excellent ability to identify patients with coronary artery stenosis, and was superior to that of peak systolic longitudinal strain. An average duration of early systolic lengthening of 58 ms revealed optimal sensitivity and specificity for identification of significant CAD. The relationship between increased duration of early systolic lengthening and significant CAD was also found on a segmental level. Increased duration of early systolic lengthening had excellent ability to identify myocardial segments related to significant stenoses. Our findings of an increased duration of early systolic lengthening in patients with significant CAD probably reflected reduced ability to generate active force in ischemic myocardial segments.
Few previous studies addressed the issue of subclinical LV systolic dysfunction in relation to CAD in stable patients. Choi et al. (6) found that resting peak systolic longitudinal strain was significantly reduced in patients with severe CAD. Nuciflora et al. (8) studied patients with suspected CAD referred for coronary evaluation and showed that impaired systolic longitudinal strain was related to the presence of obstructive CAD. Our study supported these observations. However, whether duration of early systolic lengthening provides added value for prediction of the presence of significant CAD had not been previously studied.
In the present study, the average duration of early systolic lengthening was significantly prolonged in patients with CAD. Systolic contraction was also reduced compared with those without significant coronary artery stenosis, but had markedly lower sensitivity to discover CAD. Importantly, the measurements were done with fully automatic software.
Several mechanisms have been proposed for the impaired myocardial function observed in patients with CAD. Repetitive ischemic insults to the left ventricle, which occur with significant coronary stenosis, can reduce systolic longitudinal function, and the presence of severely reduced coronary flow may induce structural remodeling of the myocardium through myocardial fibrosis and hypertrophy (13). Normal myocardium generates sufficient active force to shorten when LV pressure increases during systole, whereas ischemic myocardium with reduced active force tends to passively lengthen during early systole (14). Most likely, this mechanism explained the observed myocardial lengthening during early systole in patients with significant coronary artery stenosis. During left bundle branch block, the activation of the LV lateral wall is delayed, resulting in a passive early systolic stretching, even in the absence of lateral wall ischemia (15). Due to this, bundle branch block was an exclusion criterion in the present study.
One year after revascularization, the duration of early systolic lengthening was significantly reduced, indicating that myocardial function was not entirely normalized after revascularization. Duration of early systolic lengthening also remained elevated relative to patients without coronary artery stenosis, although this variable decreased slightly. The mechanism of this slight decrease was unknown, but might suggest that duration of systolic lengthening is a more sensitive marker of myocardial function than peak negative strain. As shown previously by our group, during the isovolumic contraction phase, when the mitral valve has moved to its final closing position and LV volume is constrained, there is a “tug-of-war” between the different LV segments (16). Therefore, during the isovolumic contraction phase, segments with reduced rate of force development might demonstrate systolic lengthening. However, more studies are needed to determine if this proposed mechanism explains the present findings.
The persistent impairment of LV systolic contraction supported the hypothesis that the myocardium underwent permanent structural remodeling, assumingly because of reduced coronary flow and repetitive ischemic insults. In addition, patients who underwent revascularization of the culprit lesion often had nonsignificant coronary stenoses in other vessels, which implied that they still had widespread atherosclerotic disease after the intervention. This might cause changes in myocardial function other than the perfusion area of the stenotic artery. However, revascularization might provide long-term benefits even in the absence of full contractile recovery by preventing further functional decline, infarctions, progressive LV dilation, and sudden cardiac death.
LVEF is related to LV systolic deformation. In the present study, LVEF failed to identify the patients with significant CAD. Ischemic injury is associated with a regional reduction in myocardial contraction. Impairment of LVEF, however, requires a decrease in several LV segments and might not be present in patients without visible myocardial scars (17,18).
The presence of post-systolic shortening was observed both in acute and chronic ischemia, and to a lesser degree also in normal myocardium (19), and was proposed as a marker of viability (20). In the present study, we found no differences between patients with and without significant CAD with regards to post-systolic shortening.
The results of the present study demonstrated that assessment of duration of early systolic lengthening in patients with suspected CAD might be a useful tool for the identification of significantly diseased coronary arteries. In particular, duration of early systolic lengthening provided incremental value over peak systolic longitudinal strain for the identification of patients with significant CAD. However, whether measurements of early systolic lengthening could contribute to the important issue of myocardial viability still remains a question. Our present study design in patients without symptoms or signs of earlier MI could not include this important question; further studies, focusing on viability, are needed.
This novel approach for detecting CAD could conveniently be integrated into clinical management. Accordingly, routine screening for subclinical LV systolic dysfunction among patients with suspected CAD might possible refine the traditional clinical assessment and might be useful for selection of further tests. The results of the present study indicated that these principles could be applied on a segmental, as well as a global, level. However, separate segmental analysis might be limited, mainly because the impaired area was not necessarily aligned with the standard segmental division of the left ventricle. The theoretical perfusion area would probably not reflect the real perfusion areas.
Two-dimensional speckle tracking echocardiography measurement has the advantage of being relatively angle independent. It is, however, like all echocardiographic methods, dependent on image quality. In the present study, all efforts were made to obtain high-quality images.
Peak systolic strain has been claimed to differentiate between patients with and without CAD. One major problem with measures of amplitudes, however, is that low image quality will hamper the peak value. Measures of duration will probably be less influenced by image quality.
The present study did not assess the presence of diffuse fibrosis. We were, therefore, unable to evaluate the interaction between diffuse fibrosis and early systolic lengthening. However, our data demonstrated a strong relationship between early systolic lengthening and the ischemic territory on a segmental level. This was in contrast to the anticipated global effects on contraction in patient with diffuse fibrosis. Moreover, in multivariable models, our findings were not compromised when the models were adjusted for hypertension, diabetes mellitus, and New York Heart Association functional class. Therefore, it was unlikely that our findings were compromised by the presence of undetected diffuse fibrosis. However, when the utility of the method is explored in a broader range of patients with cardiomyopathy, assessment of diffuse fibrosis should be considered, for example, by T1 mapping by CE-MRI.
Follow-up angiography was not performed in our study. Although none of the patients reported recurrent angina at follow-up, a progression of CAD might be a confounder.
Larger studies are required to confirm the association between significant CAD and prolonged duration of early systolic lengthening.
Prolonged duration of early systolic lengthening in stable patients with suspected CAD was related to the presence of significant coronary artery stenoses. This novel parameter, assessed by myocardial strain, enabled detection of significant CAD in patients at rest, without the need for a stress protocol.
This work was supported by the Research Council of Norway. All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- area under the curve
- coronary artery disease
- contrast-enhanced magnetic resonance imaging
- confidence interval
- high-grade stenosis lesion
- left anterior descending
- left circumflex
- left main coronary artery
- left ventricular
- left ventricular ejection fraction
- left ventricular end-diastolic volume
- left ventricular end-systolic volume
- myocardial infarction
- NYHA FC
- New York Heart Association Functional Class
- odds ratio
- percutaneous coronary intervention
- right coronary artery
- receiver-operating characteristic
- 2-dimensional speckle tracking echocardiography
- Received February 7, 2012.
- Revision received June 4, 2012.
- Accepted June 26, 2012.
- American College of Cardiology Foundation
- Kemp H.G.,
- Kronmal R.A.,
- Vlietstra R.E.,
- Frye R.L.
- Haugaa K.H.,
- Smedsrud M.K.,
- Steen T.,
- et al.
- Sjoli B.,
- Orn S.,
- Grenne B.,
- Ihlen H.,
- Edvardsen T.,
- Brunvand H.
- Cerqueira M.D.,
- Weissman N.J.,
- Dilsizian V.,
- et al.
- Efron B.,
- Tibshirani R.
- Lyseggen E.,
- Skulstad H.,
- Helle-Valle T.,
- et al.
- Gjesdal O.,
- Remme E.W.,
- Opdahl A.,
- et al.
- Remme E.W.,
- Lyseggen E.,
- Helle-Valle T.,
- et al.
- Vartdal T.,
- Brunvand H.,
- Pettersen E.,
- et al.
- Takayama M.,
- Norris R.M.,
- Brown M.A.,
- Armiger L.C.,
- Rivers J.T.,
- White H.D.