Background
Two-dimensional speckle-tracking applied to dobutamine stress echocardiography (DSE) may aid in the detection of coronary artery disease (CAD). The aim of this study was to determine the value of strain, strain rate, and postsystolic strain index (PSI) measured by speckle-tracking during DSE in the evaluation of the presence, extent, and severity of myocardial ischemia.
Methods
Fifty patients 63 ± 7 years of age with intermediate probability of CAD were prospectively recruited. All patients underwent DSE, quantitative positron emission tomographic perfusion imaging, and invasive angiography. Regional peak systolic longitudinal strain, strain rate, and PSI were measured at rest, at a dobutamine dose of 20 μg/kg/min, at peak stress, and at early recovery (1 min after stress). Obstructive CAD was defined as >75% stenosis or 40% to 75% stenosis combined with either fractional flow reserve < 0.80 or abnormal findings on myocardial perfusion positron emission tomography.
Results
Obstructive CAD was detected in 22 patients and in 36 of 150 coronary arteries. Strain analyses showed the highest reproducibility at rest, at a dobutamine dose of 20 μg/kg/min, and at early recovery. Increased PSI and reduced strain during early recovery were the strongest predictors of obstructive CAD and were associated with the extent, localization, and depth of myocardial ischemia by positron emission tomography. On vessel-based analysis, strain, PSI, and visual analysis of wall motion provided comparable diagnostic accuracy, whereas the combination of strain or PSI with visual analysis provided incremental value over visual analysis alone.
Conclusions
Assessment of systolic or postsystolic strain by speckle-tracking echocardiography during early recovery after DSE can help in the detection of hemodynamically significant coronary stenosis compared with visual wall motion analysis alone.
Dobutamine stress echocardiography (DSE) is a well-established imaging modality in the detection of coronary artery disease (CAD) on the basis of regional wall motion abnormalities induced by myocardial ischemia. Numerous studies have shown the high accuracy and prognostic value of DSE for obstructive CAD. However, wall motion analysis during DSE is subjective, and considerable expertise is required to achieve the published levels of accuracy.
Quantitative analysis of myocardial deformation by strain and strain rate (SR) imaging may help in the detection of ischemic wall motion abnormalities during DSE. Two-dimensional (2D) speckle-tracking enables the quantification of myocardial deformation on frame-to-frame tracking of ultrasonic speckles in grayscale images with automated software. In addition to ease of application and angle independence, the strength of speckle-tracking compared with Doppler tissue imaging is that strain can be measured in any direction within the imaging plane. Despite a lower frame rate compared with Doppler techniques, experimental validation has shown good agreement between speckle-tracking strain and sonomicrometry at high heart rates during dobutamine stress. Furthermore, experimental and clinical studies have indicated potential value for 2D speckle-tracking strain and SR during DSE in the detection of CAD. In addition to systolic deformation, postsystolic shortening has been proposed as a sensitive marker of myocardial ischemia. However, previous studies have been limited by the use of an anatomic angiographic definition of CAD (i.e., >50% or >75% stenosis) or small sample size. Therefore, there is a need for studies validating regional systolic and postsystolic strain measured by speckle-tracking during different phases of DSE for the detection of functionally significant CAD and severity of myocardial ischemia. We hypothesized that speckle-tracking can provide incremental diagnostic information and increase the sensitivity to conventional wall motion analysis during DSE in the evaluation of patients with suspected CAD.
We prospectively compared the value of systolic strain, SR, postsystolic strain, and visual wall motion analysis in the detection of myocardial ischemia during DSE in patients with intermediate pretest probability of CAD. All patients underwent invasive coronary angiography with fractional flow reserve (FFR) and positron emission tomographic (PET) myocardial perfusion imaging to detect significant CAD. Furthermore, the extent and severity of myocardial ischemia were assessed using quantitative PET perfusion. Obstructive CAD was defined as >75% stenosis or in the presence of intermediate stenosis (40%-75%) as FFR ≤ 0.80 or in the absence of FFR measurement as abnormal myocardial perfusion by PET.
Methods
Study Population and Design
We prospectively recruited 52 patients who were referred for investigation of stable chest pain and had intermediate pretest likelihood of obstructive CAD on the basis of the type of symptoms, age, sex, and findings on exercise electrocardiography from 2009 to 2013. The referring centers were informed of the exclusion criteria of the study, which were age <30 or >75 years, low or high pretest probability of CAD, pregnancy, acute coronary syndrome, known diagnosis of CAD, ejection fraction <35%, asthma, significant valvular disease, congenital heart disease, cardiomyopathy, severe hypertension, recent (<6 months) cerebral ischemic attack, active cancer, persistent atrial fibrillation, and atrioventricular block. The study was performed according to the Declaration of Helsinki and was approved by the local ethics committee, and each patient gave written informed consent.
The study protocol included DSE, PET perfusion imaging during adenosine stress, and invasive coronary angiography, which was performed on average 5 ± 3 weeks after DSE. One patient was excluded because of a poor echocardiographic window and one because of known CAD. Thus, the final study group consisted of 50 patients, whose clinical characteristics are presented in Table 1 .
| Variable | Value |
|---|---|
| Men | 26 (52%) |
| Age (y) | 63 ± 7 |
| Patients with CAD | 22 (44%) |
| One-vessel disease | 12 (24%) |
| Two-vessel disease | 6 (12%) |
| Three-vessel disease | 4 (8%) |
| Obstructive coronary stenoses | 36 (24%) |
| Stenosis > 75% | 19 (52%) |
| FFR < 0.8 and 40%–75% stenosis | 5 (14%) |
| Ischemic PET and 40%–75% stenosis | 12 (33%) |
| Location of stenosis | |
| LM | 1 (3%) |
| LAD | 14 (39%) |
| Proximal or middle | 10 (71%) |
| Distal | 4 (29%) |
| LCX | 9 (25%) |
| Proximal or middle | 7 (78%) |
| Distal | 2 (22%) |
| RCA | 11 (31%) |
| Proximal or middle | 7 (64%) |
| Distal | 4 (36%) |
| Risk factors | |
| Hypertension | 29 (58%) |
| Hypercholesterolemia | 31 (62%) |
| Diabetes | 7 (14%) |
| Current smoker/previous smoker | 4 (8%)/8 (16%) |
| Family history of CAD | 7 (14%) |
| Medication | |
| Aspirin | 38 (76%) |
| β-blocker | 33 (66%) |
| Statin | 40 (80%) |
| ACE inhibitor/ARB | 23 (46%) |
| Calcium channel blocker | 7 (14%) |
| Long-acting nitrate | 2 (4%) |
Dobutamine Stress Echocardiography
DSE was performed using a standard staged protocol. Dobutamine was infused through a peripheral infusion line intravenously with a mechanical pump starting at dose of 10 μg/kg/min. The dose was increased at 3-min intervals to 20, 30, and 40 μg/kg/min with intravenous atropine up to 2 mg given if necessary to augment the heart rate response. Blood pressure and electrocardiogram were monitored continuously. Criteria for terminating the test were achieving a target heart rate response of 85% of the age-predicted maximum, development of wall motion abnormality, angina pectoris, severe ischemic electrocardiographic changes, systolic blood pressure >240 mm Hg, abnormal blood pressure reaction during stress, or significant arrhythmia. Beta-blockers were withdrawn for 2 days and long-acting nitrates the morning of the study.
Image Acquisition
Images were obtained by three experienced echocardiographers using the GE Vivid 7 and M4S transducer (GE Vingmed Ultrasound AS, Horten, Norway), with patients in the lateral decubitus position. Standard 2D grayscale images of three standard apical views (four-chamber, two-chamber, and apical long-axis) and parasternal long-axis and parasternal short-axis views at the level of mitral valve, papillary muscles, and apex were acquired at rest, at a dobutamine dose of 20 μg/kg/min, at peak stress, and at recovery 1 min after stress. Per protocol, a cine image of one representative cardiac cycle per stage and view was digitally stored for later offline analysis. To optimize speckle-tracking at high heart rate, images were individually optimized for left ventricular analysis, and the frame rate was increased to achieve at target of 60 to 90 frames/sec without compromising endocardial border detection. The average frame rate was 80 ± 8 frames/sec, and it was maintained the same throughout the study.
Image Analysis
Wall motion was analyzed visually by consensus of at least two experienced echocardiographers blinded to the results of other investigations in all apical and short-axis views at all stages of the test using a 16-segment model, including six (anteroseptal, anterior, lateral, posterior, inferior, and inferoseptal) basal and mid segments and four (anterior, lateral, posterior, and septal) apical segments, as recommended by the European and American societies of echocardiography. Visual analysis of wall motion was a composite of online and offline analyses, and any wall motion abnormality induced by dobutamine was considered a positive result.
Two-Dimensional Speckle-Tracking Strain Analysis
Speckle-tracking strain was measured using commercially available software (EchoPAC PC version 10; GE Vingmed Ultrasound AS) by an analyst blinded to the other imaging results of the patient. The endocardial borders were traced at the end-systolic frame in three apical views. The software then automatically tracked myocardial motion and rejected poorly tracked segments. In the presence of poor tracking, the observer readjusted the endocardial trace. Segments rejected both by software and by the analyst were excluded. If more than two segments were excluded from one coronary territory, the vessel was excluded from the study. The numbers of coronary arteries excluded from analysis of strain, SR, and postsystolic strain were four (3%), three (2%), and four (3%) at rest; three (2%), three (2%), and three (2%) at a dobutamine dose of 20 μg/kg/min; three (2%), three (2%), and three (2%) at peak stress; and four (3%), five (3%), and six (4%) at recovery. Numeric and graphical displays of average deformation parameters of each segment were automatically generated.
Strain is defined as the amount of myocardial deformation and calculated as ( L − L 0)/ L 0, where L is the length of the object after deformation, and L 0 is the basal length of the object. We measured peak longitudinal strain during systole. SR is the myocardial deformation rate and is expressed per second, and we measured peak systolic longitudinal SR. Postsystolic strain is peak systolic strain subtracted from peak strain. We measured postsystolic strain index (PSI), which is postsystolic strain divided by peak strain. Software automatically approximated the end of systole, and this was confirmed by visual assessment of aortic valve closure in apical long-axis view. Strain, SR, and PSI were measured in each of the 16 left ventricular segments. Average strain values were calculated for the whole left ventricle (global) and for the myocardial regions subtended by major coronary arteries (left anterior descending, left circumflex, or right coronary artery). The coronary artery territories were defined using the standard template of coronary anatomy. In the presence of obstructive stenosis, the risk area was defined according to the location of the stenosis. Stenosis in the left main coronary artery was considered to affect both the left anterior descending and left circumflex coronary artery regions.
Reproducibility of Speckle-Tracking
To assess reproducibility, global strain, SR, and PSI were repeatedly analyzed twice on 2 consecutive days in 20 patients (50% with CAD) and once by another independent investigator. Measurements were made from the same designated cardiac cycles. Readers were blinded to previous measurements.
PET Perfusion Imaging
Patients were instructed to refrain from the intake of products containing caffeine or xanthine 24 hours before the scan. Patients underwent H 2 15 O positron emission tomography/computed tomography on a Discovery VCT scanner (GE Medical Systems, Waukesha, WI), as previously described. H 2 15 O (900–1,100 MBq) was injected as an intravenous bolus over 1 sec at an infusion rate of 10 mL/min. Dynamic acquisition lasting 4 min, 40 sec was performed (14 × 5, 3 × 10, 3 × 20, and 4 × 30 sec) in 2D mode. Adenosine infusion at a rate of 140 μg/kg was started 120 sec before injection and continued until the end of the scan. To correct for photon attenuation and scatter, a single low-dose computed tomographic scan was performed. Regional myocardial blood flow was analyzed using Carimas software (Turku PET Centre, Turku, Finland), and on the basis of our validation study, stress myocardial blood flow <2.5 mL/g/min was used as a threshold for abnormal stress perfusion. Furthermore, stress myocardial blood flow was categorized as mildly reduced (2.0-2.5 mL/g/min) or severely reduced (<2.0 mL/g/min). Perfusion defect covering >10% of the myocardium was considered as extensive ischemia.
Invasive Coronary Angiography and FFR
Coronary angiography was performed on a Siemens Axiom Artis coronary angiographic system (Siemens, Erlangen, Germany). Quantitative analysis was performed using software with an automated edge detection system (Quantcore; Siemens) by an experienced reader who was blinded to the results of the other imaging modalities. Seventeen standard segments were analyzed. FFR measurement was performed using the ComboMap pressure/flow instrument and 0.014-inch BrightWire pressure guidewires (Volcano Corporation, San Diego, CA) in the presence of intermediate stenosis of 40% to 75% always when feasible. The pressure was measured distally to the lesion during maximal hyperemia induced by 18-μg intracoronary boluses of adenosine with simultaneous measurement of aortic pressure through the catheter. FFR was calculated as the ratio of mean distal pressure to mean aortic pressure.
Statistical Analysis and Definitions
Significant coronary stenosis was defined as >75% anatomic stenosis by quantitative coronary angiography or in the presence of intermediate stenosis (40%-75%) as FFR < 0.8. When FFR measurement was not available, the presence of myocardial ischemia detected by PET perfusion confirmed the significance of stenosis. Data are expressed as mean ± SD unless otherwise stated. Normally distributed variables were compared using paired or nonpaired t tests as appropriate. Other variables were compared using Wilcoxon signed-rank tests or Mann-Whitney U tests. Two-tailed P values <.05 were considered to indicate statistical significance. Coefficients of variation (CVs) were calculated to test reproducibility. Receiver operating characteristic (ROC) curves were analyzed to compare the diagnostic performance of strain parameters and wall motion analysis. Comparison of areas under the curve (AUCs) was performed using the method of DeLong et al . Differences in diagnostic accuracy were studied with the McNemar test. When combining visual analysis and speckle-tracking, either ischemic wall motion abnormality or abnormal strain was considered as ischemia, and speckle-tracking was performed only in patients without visual wall motion abnormalities. Statistical analyses were done using SPSS (SPSS, Chicago, IL).
Results
The study group consisted of 50 individuals (26 men) 63 ± 7 years of age without histories of CAD. On the basis of the Framingham risk score, 35% of the study group had low (<10%), 28% had intermediate (10%-20%), and 37% had high (>20%) 10-year cardiovascular risk. Echocardiography did not show myocardial scars of previous infarcts or significant structural heart disease in any of the participants. Ejection fractions were normal in all patients (mean, 68 ± 5.2% at rest and 77 ± 7% at peak stress). There were 36 of a total of 150 vessels and 22 of 50 patients with obstructive stenosis, defined as >75% or intermediate stenosis (40%-75%) with either FFR ≤ 0.80 or PET perfusion defect. Detailed characteristics and angiographic findings of patients are shown in Table 1 . Systemic hemodynamics in response to dobutamine stress are shown in Table 2 .
| Variable | No CAD | CAD | P |
|---|---|---|---|
| Systolic blood pressure at rest (mm Hg) | 136 ± 13 | 146 ± 22 | .08 |
| Systolic blood pressure at stress (mm Hg) | 144 ± 28 | 155 ± 23 | .20 |
| Heart rate (beats/min) | |||
| Rest | 69 ± 18 | 64 ± 9 | .27 |
| 20 μg/kg/min | 90 ± 21 | 80 ± 17 | .08 |
| Peak stress | 143 ± 21 | 137 ± 19 | .51 |
| Recovery phase | 115 ± 20 | 101 ± 28 | .053 |
| Rate pressure product rest (mm Hg/min) | 9,025 ± 1,987 | 9,241 ± 1,981 | .23 |
| Rate pressure product maximum (mm Hg/min) | 20,054 ± 5,579 | 20,502 ± 5,251 | .50 |
Reproducibility of Speckle Tracking
Interobserver CVs for regional strain and SR were 6% and 3% at rest, 6% and 4% at a dobutamine dose of 20 μg/kg/min, 5% and 3% at peak stress, and 8% and 5% at recovery. Intraobserver CVs for strain and SR were 6% and 5% at rest, 6% and 6% at a dobutamine dose of 20 μg/kg/min, 9% and 9% at peak stress, and 7% and 6% at recovery. For PSI, interobserver and intraobserver CVs were 24% and 18% at rest, 19% and 18% at a dobutamine dose of 20 μg/kg/min, 43% and 13% at peak stress, and 14% and 20% at recovery. Thus, assessment of reproducibility revealed that measurements of global strain and SR were repeatable at rest and during DSE. Measurement of PSI showed higher variation at peak stress than at other stages. Bland-Altman analysis of reproducibility is shown in the Supplemental Material .
Global Strain, SR, and Postsystolic Strain in Patients with CAD
Global longitudinal peak systolic strain, SR, and PSI in patients with and without CAD are shown in Table 3 . Global strain and SR were lower in patients with than those without CAD at recovery. Global PSI was higher in patients with than those without CAD at rest, at a dobutamine dose of 20 μg/kg/min, at peak stress, and at recovery.
| Variable | No CAD | CAD | P |
|---|---|---|---|
| Global strain (%) | |||
| Rest | −19.0 ± 2.5 | −18.3 ± 2.5 | .34 |
| 20 μg/kg/min | −21.0 ± 2.3 | −20.0 ± 2.8 | .21 |
| Peak stress | −18.9 ± 2.8 | −17.9 ± 4.0 | .25 |
| Recovery | −19.8 ± 2.1 | −17.2 ± 4.0 | .01 |
| Global SR (sec −1 ) | |||
| Rest | −1.1 ± 0.2 | −1.0 ± 0.2 | .08 |
| 20 μg/kg/min | −1.7 ± 0.3 | −1.6 ± 0.3 | .40 |
| Peak stress | −2.2 ± 0.4 | −2.1 ± 0.4 | .22 |
| Recovery | −1.9 ± 0.3 | −1.6 ± 0.4 | .02 |
| Global PSI (%) | |||
| Rest | 2.8 ± 2.5 | 4.1 ± 2.7 | .03 |
| 20 μg/kg/min | 3.4 ± 2.4 | 6.4 ± 5.1 | .01 |
| Peak stress | 4.3 ± 3.9 | 6.3 ± 3.8 | .04 |
| Recovery | 3.0 ± 1.8 | 7.7 ± 5.7 | <.001 |
Table 4 shows that two- or three-vessel CAD and extensive myocardial ischemia (>10% of ischemic myocardium by PET) were most common in patients with global PSI in the highest tertile at rest, at a dobutamine dose of 20 μg/kg/min, and at recovery ( Table 4 ). Furthermore, two- or three-vessel CAD and extensive myocardial ischemia were most common in patients with SR in the lowest tertile at peak stress and strain in the lowest tertile at recovery ( Supplemental Material ).
| Variable | Tertile | P (1st vs 2nd) | P (1st vs 3rd) | P (2nd vs 3rd) | ||
|---|---|---|---|---|---|---|
| PSI baseline | 1st (< 2.1%) | 2nd (2.1%–3.3%) | 3rd (> 3.3%) | |||
| Obstructive CAD | 4 (24%) | 8 (50%) | 10 (59%) | .11 | .04 ∗ | .61 |
| Two- or three-vessel CAD | 0 | 4 (50%) | 6 (71%) | .03 ∗ | <.01 † | .52 |
| ≥10% ischemia | 1 (7%) | 4 (27%) | 10 (59%) | .16 | <.01 † | .07 |
| PSI 20 μg/kg/min | 1st (< 2.6%) | 2nd (2.6%–5.0%) | 3rd (> 5.0%) | |||
| Obstructive CAD | 4 (25%) | 7 (51%) | 11 (65%) | .33 | .02 ∗ | .17 |
| Two- or three-vessel CAD | 0 | 4 (24%) | 6 (35%) | .04 ∗ | .01 ∗ | .52 |
| ≥10% ischemia | 1 (7%) | 6 (38%) | 8 (50%) | .050 | .01 ∗ | .48 |
| PSI peak stress | 1st (< 3.0%) | 2nd (3.0%–6.2%) | 3rd (> 6.3%) | |||
| Obstructive CAD | 4 (25%) | 8 (47%) | 10 (59%) | .19 | .049 ∗ | .49 |
| Two- or three-vessel CAD | 2 (13%) | 3 (18%) | 5 (29%) | .68 | .24 | .42 |
| ≥10% ischemia | 2 (13%) | 6 (40%) | 7 (44%) | .10 | .06 | .83 |
| PSI recovery | 1st (< 2.7%) | 2nd (2.7%–4.9%) | 3rd (> 4.9%) | |||
| Obstructive CAD | 3 (19%) | 6 (35%) | 13 (76%) | 0.29 | .001 ∗ | .02 ∗ |
| Two- or three-vessel CAD | 0 | 2 (12%) | 8 (47%) | 0.16 | <.01 ∗ | .02 ∗ |
| ≥10% ischemia | 1 (7%) | 3 (20%) | 11 (69%) | 0.28 | <.001 ‡ | .01 ∗ |
Regional Strain, SR, and Postsystolic Strain in Ischemic Myocardium
Comparisons of regional peak systolic strain, SR, and PSI between ischemic and nonischemic vessel territories at different time points are shown in Table 5 . Peak systolic strain was significantly lower in the myocardial regions supplied by obstructed than nonobstructed coronary arteries at recovery. However, regional peak systolic SR was similar between nonischemic and ischemic regions at all time points. Regional PSI was significantly higher in the ischemic than the nonischemic regions at rest, at a dobutamine dose of 20 μg/kg/min, at peak stress, and at recovery.
