- Email: [email protected]
Comparison of 2- and 3-Dimensional Exercise Echocardiography for the Detection of Coronary Artery Disease
Comparison of 2- and 3-Dimensional Exercise Echocardiography for the Detection of Coronary Artery Disease Jesús Peteiro, MD, Pablo Piñon, MD, Ruth Perez, MD, Lorenzo Monserrat, MD, Diego Perez, MD, and Alfonso Castro-Beiras, MD, A Coruña, Spain
Background: Although 3-dimensional echocardiography (3DE) has been applied during dobutamine echocardiography it has not been used during exercise echocardiography. We sought to compare feasibility and accuracy of 3DE and 2-dimensional echocardiography (2DE) during exercise echocardiography. Methods: 100 patients underwent peak and postexercise (PEx) 2DE and 3DE on separate days. Coronary artery disease was detected in 58 patients. A quality score was calculated by assigning 0 to 3 points to each wall (apicoseptal, posterolateral, anterior, inferior). Results: Feasibility of peak 2DE, peak 3DE, PEx 2DE, and PEx 3DE was 99%, 92%, 100%, and 95%, respectively (2DE at peak or PEx vs peak 3DE, P < .05). Agreement between 2DE and 3DE was 82% at peak ( ⴝ 0.62) and 78% at PEx ( ⴝ 0.55). Quality score less than 2 was seen in 4% of the walls with peak 2DE, in none with PEx 2DE, in 18% by peak 3DE, and
in 14% by PEx 3DE. The mean quality score was lower with 3DE at peak and at PEx (2.4 ⴞ 0.9 vs 2.9 ⴞ 0.3; and 2.5 ⴞ 0.8 vs 3.0 ⴞ 0.1, both P < .0001). Acquisition time was shorter with 3DE at peak and PEx (22 ⴞ 8 vs 43 ⴞ 14 seconds; and 15 ⴞ 5 vs 31 ⴞ 14 seconds, both P < .0001). Sensitivity of peak 2DE, peak 3DE, PEx 2DE, and PEx 3DE was 84%, 78%, 71%, and 58%, respectively (P < .05 vs peak 3DE and peak 2DE). Specificity was 76%, 73%, 93%, and 88%, respectively. Accuracy for peak 2DE was 81% (area under the curve [AUC] 0.81, 95% confidence interval [CI] ⴝ 0.71-0.91); for peak 3DE was 76% (AUC 0.76, 95% CI ⴝ 0.65-0.86); for PEx 2DE was 80% (AUC 0.84, 95% CI ⴝ 0.75-0.92); and for PEx 3DE was 71% (AUC 0.73, 95% CI ⴝ 0.62-0.83). Conclusions: Three-dimensional echocardiography during exercise is comparable with 2DE in terms of sensitivity and specificity but feasibility is lower. (J Am Soc Echocardiogr 2007;20:959-967.)
O
METHODS
ne of the possible applications of 3-dimensional echocardiography (3DE) is stress echocardiography. Although 3DE has been applied during dobutamine stress echocardiography1,2 it has not been used during exercise echocardiography (EE). However, EE is more safe3,4 and sensitive than pharmacologic stress echocardiography.5,6 The possibility of acquiring 3DE data sets of the entire left ventricle (LV) in only 4 cardiac cycles is truly exciting as one of the main limitations of EE is the narrow time window for acquisition, particularly at postexercise (PEx).7
From the Unit of Echocardiography and Department of Cardiology, Juan Canalejo Hospital, A Coruña University. Supported by the Spanish Network of Cardiovascular Studies (RECAVA). Reprint requests: Jesús Peteiro, MD, Unit of Echocardiography and Department of Cardiology, Juan Canalejo Hospital, P/Ronda 5-4° izda.15011-A Coruña University, A Coruña, Spain (E-mail: [email protected]). 0894-7317/$32.00 Copyright 2007 by the American Society of Echocardiography. doi:10.1016/j.echo.2007.01.034
Patient Population We selected consecutively 84 patients who were referred for coronary angiography, after exclusions, and 16 patients with very low pretest probability of coronary artery disease (CAD) referred for EE. Exclusion criteria for patients referred to angiography were as follows: acute ST elevation myocardial infarction (MI), unstable angina, heart failure, significant valvular heart disease, cardiomyopathy, atrial fibrillation/flutter, and LV bundle branch block or pacemaker ectopic rhythm. Conditions to include the 16 patients with very low pretest probability of CAD were to be asymptomatic or have atypical chest pain, to have normal resting electrocardiography (ECG) findings, absence of diabetes mellitus, and pretest probability of CAD less than 5%.8 These 16 patients were considered as having no CAD. Therefore, the final study group comprised 100 patients. Table 1 shows the clinical characteristics of the study group and medications at the time of the EE. Informed written consent was obtained from all patients.
959
Journal of the American Society of Echocardiography August 2007
960 Peteiro et al
Table 1 Clinical characteristics Male Age ⫾ 1SD (range) Diabetes mellitus, % Hypertension, % Hypercholesterolemia, % Smoking, % Familial history of CAD, % LV hypertrophy (by echocardiography) Previous myocardial infarction, % Non-S-T elevation S-T elevation Previous percutaneous revascularization, % Previous coronary bypass, % Medications, % Nitrates, % Calcium channel blockers, % Beta-blockers, % Angiotensin-converting enzyme inhibitors
67 62 ⫾ 11 (35-84) 18 52 50 24 25 54 20 4 19 3 52 8 30 48
CAD, Coronary artery disease; LV, left ventricular.
EE Patients underwent exercise 2-dimensional echocardiography (2DE) and exercise 3DE on separate days within 1 week in a random order. The 2DE images were obtained with a Sonos 7500 (Phillips Medical Systems, Andover, Mass) or a Vivid 5 (General Electric) and 3DE images with a Sonos 7500. The medications were the same during the two tests. Heart rate, blood pressure, and a 12-lead ECG were obtained at baseline, at each stage of the exercise protocol, and during the PEx period. A treadmill exercise test (Bruce protocol in 96 patients, and modified Bruce in 4) was performed until exhaustion or until the patient reached an end point. End points included ST segment change greater than 2.5 mm, significant arrhythmia, severe hypertension (systolic blood pressure ⬎ 240 mm Hg or diastolic blood pressure ⬎ 110 mm Hg), severe hypotensive response (decrease ⬎ 20 mm Hg from baseline), or limiting symptoms. The exercise ECG was considered positive in case of horizontal or downsloping ST segment depression of at least 1 mm at 80 milliseconds after the J point or in case of S-T elevation in leads without q wave. 2DE The 2DE method using harmonic imaging was performed at baseline, peak exercise,9,10 and immediately after exercise. Apical long-axis, 4-chamber, and 2-chamber views, and parasternal long- and short-axis views were obtained at rest. The same order was followed at peak and PEx with the use of a continuous acquisition imaging system. Peak imaging was performed with the patient still exercising, when signs of exhaustion were present, end point was achieved, or both. The patient was asked to walk quickly rather than run, to decrease body and respiratory movements. The transducer was firmly positioned on the cardiac apex to obtain the apical views by applying slight pressure to the patient’s back with the left hand, so maintaining the patient between the transducer and the
Table 2 Clinical, electrocardiographic, hemodynamic, and echocardiographic data during 2- and 3-dimensional echocardiographic exercise testing 2DE
Angina, % Positive ECG, % Peak heart rate, beats/min Peak systolic blood pressure, mm Hg Peak heart rate ⫻ systolic blood pressure ⫻ 1000 Maximal age-predicted heart rate, % MET Resting wall-motion score index Peak wall-motion score index Postexercise wall-motion score index Resting ejection fraction Peak ejection fraction Postexercise ejection fraction
3DE
28 26 131 ⫾ 24 164 ⫾ 27
28 22 129 ⫾ 25 163 ⫾ 25
22.5 ⫾ 5.9
22.0 ⫾ 5.8
87 ⫾ 13
85 ⫾ 13
8.8 1.05 1.29 1.18
⫾ ⫾ ⫾ ⫾
2.8 0.16 0.35 0.27
59 ⫾ 6 59 ⫾ 12 64 ⫾ 11
9.0 1.05 1.23 1.15
⫾ ⫾ ⫾ ⫾
2.9 0.16 0.31† 0.25*
59 ⫾ 6 61 ⫾ 11* 63 ⫾ 11
ECG, Electrocardiography; MET, metabolic equivalent; 2DE, 2-dimensional echocardiography; 3DE, 3-dimensional echocardiography. *P ⬍ .05. †P ⬍ .01.
left hand, to avoid movement. Finally, the transducer was positioned in the parasternal region to obtain the parasternal views. Image acquisition was performed online and stored on disk. The 2DE analysis was performed on a digital quadscreen display system. 3DE A full matrix-array X3 transducer (2-4 MHz) with secondharmonic imaging was used. The 3DE images were acquired from the apical window at rest, peak exercise, and in the immediate PEx period. The 4- and 2-chamber apical views were displayed on the screen to help the operator to position the sample with all targets being covered by the imaging volume. Once the imaging volume was visualized, data sets of LV were obtained. Four conical subvolumes of 20 ⫻ 80 degrees were scanned during 4 consecutive cardiac cycles without moving the probe. ECG gating was used and the 4 subvolumes were acquired during 4 consecutive cardiac cycles. The 4 subvolumes were then automatically integrated in a full volume. Patients were taught to hold their breath before the start of the testing and they were asked to do it again during each full-volume acquisition (usually 4 seconds at rest and 2 seconds at exercise). One full volume was acquired at rest whereas 2 to 3 were acquired at peak exercise and in the immediate PEx period. The best full volume without motion and/or respiratory artefacts at peak and PEx was chosen for comparison with the resting one. Regional LV wall motion was analyzed using cropped planes. At least 3 short-axis views were produced by cropping the long axis of the LV from the apex to the base. At least two parallel long-axis, 4-chamber, and 2-chamber apical views were obtained for each patient. Therefore, various long-axis,
Journal of the American Society of Echocardiography Volume 20 Number 8
Peteiro et al 961
Table 3 Intraobserver and interobserver agreement of peak and postexercise 2- and 3-dimensional echocardiography 2DE
Intraobserver, % (k) Intraobserver, % (k)
3DE
Peak
Postexercise
Peak
Postexercise
92 (0.83) 88 (0.75)
92 (0.85) 88 (0.74)
93 (0.93) 88 (0.63)
96 (0.96) 87 (0.65)
2DE, Two-dimensional echocardiography; 3DE, 3-dimensional echocardiography.
Table 4 Sensitivity, specificity, and accuracy of peak and postexercise 2- and 3-dimensional echocardiography for the prediction of left anterior descending coronary artery and combined left circumflex/right coronary arteries stenoses Peak
LAD sensitivity, % LAD specificity, % LAD accuracy, % LCx/RCA sensitivity, % LCx/RCA specificity, % LCx/RCA accuracy, %
Figure 1 Sensitivity, specificity, and accuracy of 2- and 3-dimensional echocardiography at peak and postexercise for prediction of coronary artery disease.
4-chamber, and 2-chamber apical views and short-axis views were used for comparison of rest and peak exercise and rest and PEx images. LV Wall-motion Analysis The LV was divided into 16 segments.11 Each of the 16 segments was assigned to one of the 3 coronary artery territories, as recommended by the American Society of Echocardiography.12 The development of new regional dysfunction or worsening from hypokinesia in at least one segment was considered an ischemic response. The persistence of regional baseline dysfunction or worsening from akinesia to dyskinesia was considered as infarction without ischemia, except for the case of isolated hypokinesia of the basal inferior or basal septal segments.13 A positive EE result for CAD was defined when there was ischemia or infarction.14-16 A wall-motion score index at rest, peak, and PEx was calculated, with normal wallmotion scoring 1, hypokinetic scoring 2, akinetic scoring 3, and dyskinetic scoring 4. LV ejection fraction was assessed visually.17 Resting images were compared with peak and PEx images either by 2DE and 3DE in separate days from disk-stored images. Analysis was blinded to clinical characteristics, and clinical and ECG response to exercise. The quality of the studies was assessed by a score that assigned 0 to 3 points to each wall (apicoseptal, posterolateral, anterior, inferior) according to the visualization of
Postexercise
2DE
3DE
2DE
3DE
89 80 84 52 74 64
84 73 78 43 88 66
67 93 81 38 92 67
60 87 75 32 94 65
LAD, Left anterior descending coronary artery; LCx, left circumflex coronary artery; RCA, right coronary artery; 2DE, 2-dimensional echocardiography; 3DE, 3-dimensional echocardiography.
myocardial thickness and motion1 as followed: 0, very bad visualization or not acquired; 1, visualization of less than 70% of the extension of the wall; 2, visualization of 70% to 95% of the extension of the wall; and 3, visualization of 95% to 100% of the extension of the wall. Coronary Angiography Coronary angiography was performed within 2 weeks of the EE studies. A luminal narrowing of 70% or more by visual estimation in a coronary artery, major branch, or bypass, or 50 or greater in the left main coronary artery were considered significant. Of the 84 patients submitted to angiography, 22 patients had one-vessel disease and 36 had multivessel CAD. Of the remaining 26 patients, 11 had normal coronary arteries and 15 had noncritical disease. Left anterior descending coronary artery (LAD) disease was found in 45 patients, left circumflex coronary artery (LCx) disease in 36 patients, right coronary artery (RCA) disease in 32 patients, and left main CAD in 7 patients. Statistical Analysis Continuous variables are expressed as mean ⫾1 SD and compared by use of paired Student t test. Sensitivity was defined as the number of true positive tests divided by the number of patients or territories irrigated by a stenotic vessel. Specificity was defined as the number of true negative tests divided by the number of patients or territories irrigated by a nonstenotic vessel. Accuracy was defined as the total number of true positive and true negative tests divided by the total number of patients or
962 Peteiro et al
Journal of the American Society of Echocardiography August 2007
Figure 2 Example of 64-year-old man with normal angiography and normal peak exercise 3-dimensional echocardiography results. Resting (left), peak exercise (right), diastolic (D), and systolic (S) images. Cropped 4-chamber (A) and 2-chamber (B) apical views. Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. Note increase in regional function in all cropped views. (See supplementary material on line: Videos 1 and 2 for apical views and Video 3 for short-axis views.) territories. A value of P less than .05 was considered significant for comparisons. Values of area under the curve (AUC) and 95% confidence interval (CI) are given from receiver operator characteristic curves. The agreement between the test results obtained by 2DE and 3DE, and the agreement between observers were tested by measuring the coefficient of variation in 25 randomized patients. A value of 0.45 or greater was considered to be good agreement, and a value of greater than 0.75 was considered to be excellent agreement.
RESULTS All the patients completed both EE studies without complications. Exercise testing data were not different in 2DE and 3DE exercise testing (Table 2). Resting wall-motion abnormalities (WMA) were detected in 8 patients by the two techniques (3 of
them had previous MI), in 4 patients by 2DE alone, and in 5 patients by 3DE alone (baseline agreement 91%, ⫽ 0.59). Feasibility of 3DE and 2DE Feasibility of peak 2DE, peak 3DE, PEx 2DE, and PEx 3DE was 99%, 92%, 100%, and 95%, respectively (P ⬍ .05 between peak 2DE and peak 3DE; P ⬍ .01 between post-2DE and peak 3DE). Both peak 2DE and 3DE imaging were not possible in one patient because of premature interruption of exercise, whereas 3DE was unfeasible or unreadable at peak or at PEx in 7 and 5 patients, respectively. Quality scores were qualified as 0 in these patients. Inadequate quality of images (score 0-1 in any wall) was seen in only 4% of the walls with peak 2DE and in none with PEx 2DE, but they were seen in 18% of the walls at peak 3DE and in 14% of the walls at PEx 3DE. These scores lower than 2 were more fre-
Journal of the American Society of Echocardiography Volume 20 Number 8
Peteiro et al 963
Figure 3 Example of 75-year-old-man with 3-vessel coronary artery disease (left anterior descending [90%], right [95%], and left circumflex [90%] coronary arteries) and abnormal peak exercise 3-dimensional echocardiography results. Resting (left), peak exercise (right), diastolic (D), and systolic (S) images. Cropped 4-chamber (A) and 2-chamber (B) apical views. Note regional dysfunction at peak exercise in apicoseptal segments (A) (arrows) and exercise-induced dysfunction in anterior and posterior walls (B) (arrows). Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. See exerciseinduced wall-motion abnormalities and dilation at apical level (C) (arrows). (See supplementary material on line: Videos 4 and 5 for apical views and Video 6 for short-axis views.)
quently seen in the inferior and anterior walls in contrast to the septoapical and posterolateral territories (75% vs 25%). This was caused by inadequate match of subvolumes at peak or post-EE or bad acoustic wall definition. The mean quality score (mean score of the 4 walls) was lower with 3DE at baseline (2.9 ⫾ 0.2 vs 3.0 ⫾ 0.1, P ⬍ .05), peak (2.4 ⫾ 0.9 vs 2.9 ⫾ 0.3, P ⬍ .0001), and PEx (2.5 ⫾ 0.8 vs 3.0 ⫾ 0.1) imaging. Total imaging optimization and acquisition time in 25 randomized patients was lower with 3DE at rest (10 ⫾ 4 vs 30 ⫾ 8 seconds, P ⬍ .0001), peak (22 ⫾ 8 vs 43 ⫾ 14 seconds, P ⬍ .0001), and PEx (15 ⫾ 5 vs 31 ⫾ 14 seconds, P ⬍ .0001).
Interobserver and Intraobserver Agreement Interobserver and intraobserver agreement (positive or negative test) of peak and PEx 2DE and 3DE is shown in Table 3. Agreement Between Studies Agreement between 2DE and 3DE at peak exercise was 82% ( ⫽ 0.62), whereas at PEx it was 78% ( ⫽ 0.55). There was a trend to better agreement in the detection of WMA in LAD and LCx territories than in RCA territory at peak imaging (LAD 84%, ⫽ 0.67; LCx 86%, ⫽ 0.67; RCA 74%, ⫽ 0.27), whereas agreement was similar at PEx imaging for the differ-
964 Peteiro et al
Journal of the American Society of Echocardiography August 2007
Figure 4 Example of 76-year-old woman with proximal left anterior descending (LAD) (99%) and right (70%) coronary artery stenosis and abnormal peak exercise 3-dimensional echocardiography results (ischemia in LAD territory). Resting (left), peak exercise (right), diastolic (D), and systolic (S) images. Cropped 4-chamber (A) and 2-chamber (B) apical views. Note regional dysfunction at peak exercise in apicoseptal segments (A) (arrows) and exercise-induced dysfunction in anterior wall and apex (B) (arrows). Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. Note septal dyssynergia at peak exercise (C) (arrow). (See supplementary material on line: Videos 7, 8, and 9 for apical views, Video 10 for short-axis apical view, and Video 11 for short-axis, mid, and basal views.)
ent territories (LAD 78%, ⫽ 0.52; LCx 87%, ⫽ 0.55; RCA 89%, ⫽ 0.53). Sensitivity, Specificity, and Accuracy of 3DE and 2DE at Peak and PEx All 16 patients with low pretest probability considered to have no CAD had normal EE results by either peak and PEx 2DE, whereas false-positive results were seen in two patients by peak and in one patient by PEx 3DE, respectively. Of the 26 patients submitted to angiography who had no CAD, false-positive results were seen in 10 patients by peak 2DE, in 3 by PEx 2DE, in 8 by peak 3DE, and in 4 by PEx 3DE. Sensitivity, specificity, and accuracy of 3DE and 2DE at peak and PEx for the detection of CAD in a binary manner and according to the different arteries in-
volved are shown in Figure 1 and Table 4, respectively. The higher sensitivity was achieved with peak imaging (either 2DE or 3DE). Specificity was slightly higher with PEx 2DE. Accuracy was similar: 81% by peak 2DE (AUC 0.81, 95% CI ⫽ 0.71-0.91); 76% by peak 3DE (AUC 0.76, 95% CI ⫽ 0.65-0.86); 80% by PEx 2DE (AUC 0.84, 95% CI ⫽ 0.75-0.92); and 71% by PEx 3DE (AUC 0.73, 95% CI ⫽ 0.62-0.83). There was a trend to higher sensitivity of each approach in patients who were not on beta-blocker therapy (peak 2DE 89% vs 75%; peak 3DE 83% vs 68%; PEx 2DE 76% vs 62%; and PEx 3DE 66% vs 45%). Sensitivity in 36 patients with multivessel CAD was 97% for peak 2DE, 88% for peak 3DE, 83% for PEx 2DE, and 68% for PEx 3DE (P ⬍ .001 vs peak 2DE, P ⬍ .1 vs peak 3DE). An example of normal
Journal of the American Society of Echocardiography Volume 20 Number 8
Peteiro et al 965
Figure 5 Example of 48-year-old man with right coronary artery (RCA) and left circumflex coronary artery (LCx) stenoses (both 99%) and abnormal peak exercise 3-dimensional echocardiography results (ischemia in RCA/LCx territories). Cropped 4-chamber (A) and 2-chamber (B) apical views. Note regional dysfunction at peak exercise in basal septal segment (A) (arrow) and exercise-induced dysfunction in posterior wall (B) (arrows). Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. Note exercise-induced abnormalities in posterior septum in short-axis views (arrows). (See supplementary material on line: Videos 12 and 13 for apical views and Video 14 for short-axis views.)
peak exercise 3DE is showed in Figure 2, and Videos 1 to 3 whereas Figures 3 to 5 and Videos 4 to 14 are examples of ischemic responses by peak 3DE.
DISCUSSION The main findings of this study were that volumetric 3DE is as accurate as 2DE during EE, but feasibility and imaging quality were lower than those of 2DE. Although 2DE and 3DE by acquisition of a full LV volume have been compared during pharmacologic stress with dobutamine in the same study they have not during exercise testing.
3DE During Dobutamine Two different approaches have been used by matrixarray transducer technology during dobutamine stress: the volumetric approach and the multiplane imaging approach that allows the simultaneous visualization and acquisition of 2 to 3 LV planes. The volumetric approach resulted marginally superior to 2DE in terms of sensitivity (89% vs 79%) in one study where exercise single photon emission computed tomography was the reference standard,2 and it was similar in another one (86% vs 86%) that used coronary angiography.1 With the use of multiplane imaging, Eroglu et al18 demonstrated identical sensitivity (93%) and specificity (75%). In all these studies
966 Peteiro et al
the acquisition time was lower with 3DE. Although real-time multiplane imaging has higher temporal resolution than volumetric 3DE, the number of views than can be obtained by volumetric 3DE is greater. 3DE During Exercise The 3DE method may theoretically be more useful for exercise than for pharmacologic stress because of the narrow acquisition time window, particularly at PEx. Actually, we and others19 demonstrated lower acquisition time with volumetric 3DE than with 2DE during exercise. Advantages of exercise 3DE according to the results of this study would be lower acquisition time and the possibility of image multiple parallel planes in each view. Uncertainty regarding the presence/absence of WMA could be resolved by reviewing the parallel or orthogonal available views. The major disadvantages come from an inadequate matching of the subvolumes and from the low frame rate. The inadequate merging of the subvolumes is usually produced by body or breathing movements, or the eventual presence of an artefactual ECG. It is important to teach patients how to hold their breath during the approximately 2 seconds of the acquisition at peak or PEx while avoiding previous inspiration. As the respiratory movements may be suppressed by asking patients to hold their breath, it is more difficult to avoid body movements during exercise. An artefactual ECG or the development of arrhythmia prevents data blocks acquisition. The low frame rate may underdiagnose minor hypokinesia or tardokinesia. Improvement in imaging quality (spatial and temporal resolution) and acquisition of the full volume in only one cycle may constitute requirements for the future application of this technique during exercise. Limitations Exercise 3DE was less feasible than exercise 2DE, therefore, any comparison of accuracy between both techniques should be taken cautiously. Peak exercise is a highly dependent imaging technique modality. However, its major advantage is the relative lack of hurry for imaging acquisition in comparison with PEx where a limited time window exists. Apical 2DE views are particularly easy to acquire and their quality was the same of those of PEx 2DE in a previous report.10 Therefore, a fullvolume 3DE acquisition from the apical view is theoretically feasible during peak exercise. Most of our patients had adequate acoustic window with 2DE. Therefore, these results may not be applied to the total population of patients referred for EE. A significant percentage of them were taking beta-blockers at the time of the tests. Thus, global sensitivity for the prediction of CAD may be infraestimated. There was a trend to higher sensitivity by
Journal of the American Society of Echocardiography August 2007
each of the approaches in patients without betablockers. Patients with recent S-T elevation MI were not included and consequently the percentage of patients with resting WMA was small. This fact also explains why sensitivity is lower than in studies that include a significant percentage of patients with MI and resting WMA. Particularly weak sensitivity was obtained for the RCA and LCx. Several factors may explain this finding. First, sensitivity of stress echocardiography is in general lower for these arteries as compared with the LAD, because the LAD territory is nearest to the transducer from each view. Second, lower imaging quality in the inferior wall may explain low sensitivity for the RCA, particularly with 3DE. Third, as above referred, 30% of the patients were on beta-blockers. We did not use contrast agents in this study. With the use of contrast agents, the quality of 3DE may be enhanced to be comparable with peak and PEx 2DE.20 Finally, patients with atrial fibrillation/flutter, LV bundle branch block, or pacemaker ectopic rhythm were not included because these conditions may prevent adequate match of the subvolumes with 3DE. If included, the feasibility of 3DE may be even lower. Conclusions Currently, 3DE is not superior to 2DE during EE, mainly because the imaging quality and feasibility are lower. However, peak 3DE is similar to peak 2DE and PEx 3DE is similar to PEx 2DE in terms of sensitivity and specificity.
REFERENCES 1. Matsumura Y, Hozumi T, Arai K, Sugioka K, Ujino K, Takemoto Y, et al. Non-invasive assessment of myocardial ischemia using new real-time three-dimensional dobutamine stress echocardiography: comparison with conventional two-dimensional methods. Eur Heart J 2005;26:1625-32. 2. Ahmad M, Xie T, McCulloch M, Abreo G, Runge M. Realtime three-dimensional dobutamine stress echocardiography in assessment of ischemia: comparison with two-dimensional dobutamine stress echocardiography. J Am Coll Cardiol 2001;37:1303-9. 3. Varga A, Garcia MA, Picano E; International Stress Echo Complication Registry. Safety of stress echocardiography (from the international stress echo complication registry). Am J Cardiol 2006;98:541-3. 4. Rodríguez García MA, Iglesias-Garriz I, Corral Fernández F, Garrote Coloma C, Alonso-Orcajo N, Branco L, et al. Evaluation of the safety of stress echocardiography in Spain and Portugal. Rev Esp Cardiol 2001;54:941-8. 5. Beleslin BD, Ostojic M, Stepanovic J, Djordjevic-Dikic A, Stoljkovic S, Nedeljkovic G, et al. Stress echocardiography in the detection of myocardial ischemia: head-to-head comparison of exercise, dobutamine, and dipyridamole tests. Circulation 1994;90:1168-76. 6. Dagianti A, Penco M, Agati L, Sciomer S, Dagianti A, Rosanio S, et al. Stress echocardiography: comparison of exercise,
Journal of the American Society of Echocardiography Volume 20 Number 8
7.
8.
9.
10.
11.
12.
13.
14.
dipyridamole and dobutamine in detecting and predicting the extent of coronary artery disease. J Am Coll Cardiol 1995;26:18-25. Picano E. Exercise echocardiography. In: Picano E, editor. Stress echocardiography. 4th ed. Berlin: Springer-Verlag; 2003. p. 103-14. Marrugat J, Solanas P, D’Agostino R, Sullivan L, Ordovas J, Cordón F, et al. Coronary risk estimation in Spain using a calibrated Framingham function. Rev Esp Cardiol 2003;56: 253-61. Peteiro J, Monserrat L, Perez R, Vazquez E, Vazquez JM, Castro-Beiras A. Accuracy of peak treadmill exercise echocardiography to detect multivessel coronary artery disease: comparison with post-exercise echocardiography. Eur J Echocardiogr 2003;4:182-90. Peteiro J, Garrido I, Monserrat L, Aldama G, Calvino R, Castro-Beiras A. Comparison of peak and postexercise treadmill echocardiography with the use of continuous harmonic imaging acquisition. J Am Soc Echocardiogr 2004;17: 1044-9. Bourdillon PD, Broderik TM, Sawada SG, Armstrong WJ, Ryan T, Dillon JC, et al. Regional wall motion index for infarct and noninfart regions after reperfusion in acute myocardial infarction: comparison with global wall motion index. J Am Soc Echocardiogr 1989;9:398-407. Schiller NB, Shah PM, Crawford M, DeMaria A, Devareaux R, Feigenbaum H, et al. Recommendations for quantification of the left ventricle by two dimensional echocardiography. J Am Soc Echocardiogr 1989;2:358-67. Hoffmann R, Lethen H, Marwick T, Rambaldi R, Fioretti P, Pingitore A, et al. Standardized guidelines for the interpretation of dobutamine echocardiography reduce interinstitutional variance in interpretation. Am J Cardiol 1998;82: 1520-4. Roger VL, Pellikka PA, Oh JK, Bailey KR, Tajik AJ. Identification of multivessel disease by exercise echocardiography. J Am Coll Cardiol 1994;24:109-14.
Peteiro et al 967
15. Quiñones MA, Verani MS, Haichin RM, Mahmariam JJ, Suarez J, Zoghbi WA. Exercise echocardiography versus 201 Tl single-photon emission computed tomography in evaluation of coronary artery disease: analysis of 292 patients. Circulation 1992;85:1026-31. 16. Ryan T, Vasey CG, Presti CF, O’Donnell JA, Feigenbaum H, Armstrong WF. Exercise echocardiography: detection of coronary artery disease in patients with normal left ventricular wall motion at rest. J Am Coll Cardiol 1988;11:993-9. 17. Stamm RB, Carabello BA, Mayers DL, Martin RP. Twodimensional echocardiographic measurement of left ventricular ejection fraction: prospective analysis of what constitutes an adequate determination. Am Heart J 1982;104:136-44. 18. Eroglu E, D’hooge J, Herbots L, Thijs D, Dubois C, Sinnaeve P, et al. Comparison of real-time tri-plane and conventional 2D dobutamine stress echocardiography for the assessment of coronary artery disease. Eur Heart J 2006;27:1719-24. 19. Zwas DR, Takuma S, Mullis-Jansson S, Fard A, Chaudhry H, Wu H, et al. Feasibility of real-time 3-dimensional treadmill stress echocardiography. J Am Soc Echocardiogr 1999;12: 285-9. 20. Takeuchi M, Otani S, Weinert L, Spencer KT, Lang RM. Comparison of contrast-enhanced real-time live 3-dimensional dobutamine stress echocardiography with contrast 2dimensional echocardiography for detecting stress-induced wall-motion abnormalities. J Am Soc Echocardiogr 2006;19: 294-9.
SUPPLEMENTARY DATA Supplementary data associated with this article can be found, in the online version, at 10.1016/j.echo. 2007.01.034.
Background: Although 3-dimensional echocardiography (3DE) has been applied during dobutamine echocardiography it has not been used during exercise echocardiography. We sought to compare feasibility and accuracy of 3DE and 2-dimensional echocardiography (2DE) during exercise echocardiography. Methods: 100 patients underwent peak and postexercise (PEx) 2DE and 3DE on separate days. Coronary artery disease was detected in 58 patients. A quality score was calculated by assigning 0 to 3 points to each wall (apicoseptal, posterolateral, anterior, inferior). Results: Feasibility of peak 2DE, peak 3DE, PEx 2DE, and PEx 3DE was 99%, 92%, 100%, and 95%, respectively (2DE at peak or PEx vs peak 3DE, P < .05). Agreement between 2DE and 3DE was 82% at peak ( ⴝ 0.62) and 78% at PEx ( ⴝ 0.55). Quality score less than 2 was seen in 4% of the walls with peak 2DE, in none with PEx 2DE, in 18% by peak 3DE, and
in 14% by PEx 3DE. The mean quality score was lower with 3DE at peak and at PEx (2.4 ⴞ 0.9 vs 2.9 ⴞ 0.3; and 2.5 ⴞ 0.8 vs 3.0 ⴞ 0.1, both P < .0001). Acquisition time was shorter with 3DE at peak and PEx (22 ⴞ 8 vs 43 ⴞ 14 seconds; and 15 ⴞ 5 vs 31 ⴞ 14 seconds, both P < .0001). Sensitivity of peak 2DE, peak 3DE, PEx 2DE, and PEx 3DE was 84%, 78%, 71%, and 58%, respectively (P < .05 vs peak 3DE and peak 2DE). Specificity was 76%, 73%, 93%, and 88%, respectively. Accuracy for peak 2DE was 81% (area under the curve [AUC] 0.81, 95% confidence interval [CI] ⴝ 0.71-0.91); for peak 3DE was 76% (AUC 0.76, 95% CI ⴝ 0.65-0.86); for PEx 2DE was 80% (AUC 0.84, 95% CI ⴝ 0.75-0.92); and for PEx 3DE was 71% (AUC 0.73, 95% CI ⴝ 0.62-0.83). Conclusions: Three-dimensional echocardiography during exercise is comparable with 2DE in terms of sensitivity and specificity but feasibility is lower. (J Am Soc Echocardiogr 2007;20:959-967.)
O
METHODS
ne of the possible applications of 3-dimensional echocardiography (3DE) is stress echocardiography. Although 3DE has been applied during dobutamine stress echocardiography1,2 it has not been used during exercise echocardiography (EE). However, EE is more safe3,4 and sensitive than pharmacologic stress echocardiography.5,6 The possibility of acquiring 3DE data sets of the entire left ventricle (LV) in only 4 cardiac cycles is truly exciting as one of the main limitations of EE is the narrow time window for acquisition, particularly at postexercise (PEx).7
From the Unit of Echocardiography and Department of Cardiology, Juan Canalejo Hospital, A Coruña University. Supported by the Spanish Network of Cardiovascular Studies (RECAVA). Reprint requests: Jesús Peteiro, MD, Unit of Echocardiography and Department of Cardiology, Juan Canalejo Hospital, P/Ronda 5-4° izda.15011-A Coruña University, A Coruña, Spain (E-mail: [email protected]). 0894-7317/$32.00 Copyright 2007 by the American Society of Echocardiography. doi:10.1016/j.echo.2007.01.034
Patient Population We selected consecutively 84 patients who were referred for coronary angiography, after exclusions, and 16 patients with very low pretest probability of coronary artery disease (CAD) referred for EE. Exclusion criteria for patients referred to angiography were as follows: acute ST elevation myocardial infarction (MI), unstable angina, heart failure, significant valvular heart disease, cardiomyopathy, atrial fibrillation/flutter, and LV bundle branch block or pacemaker ectopic rhythm. Conditions to include the 16 patients with very low pretest probability of CAD were to be asymptomatic or have atypical chest pain, to have normal resting electrocardiography (ECG) findings, absence of diabetes mellitus, and pretest probability of CAD less than 5%.8 These 16 patients were considered as having no CAD. Therefore, the final study group comprised 100 patients. Table 1 shows the clinical characteristics of the study group and medications at the time of the EE. Informed written consent was obtained from all patients.
959
Journal of the American Society of Echocardiography August 2007
960 Peteiro et al
Table 1 Clinical characteristics Male Age ⫾ 1SD (range) Diabetes mellitus, % Hypertension, % Hypercholesterolemia, % Smoking, % Familial history of CAD, % LV hypertrophy (by echocardiography) Previous myocardial infarction, % Non-S-T elevation S-T elevation Previous percutaneous revascularization, % Previous coronary bypass, % Medications, % Nitrates, % Calcium channel blockers, % Beta-blockers, % Angiotensin-converting enzyme inhibitors
67 62 ⫾ 11 (35-84) 18 52 50 24 25 54 20 4 19 3 52 8 30 48
CAD, Coronary artery disease; LV, left ventricular.
EE Patients underwent exercise 2-dimensional echocardiography (2DE) and exercise 3DE on separate days within 1 week in a random order. The 2DE images were obtained with a Sonos 7500 (Phillips Medical Systems, Andover, Mass) or a Vivid 5 (General Electric) and 3DE images with a Sonos 7500. The medications were the same during the two tests. Heart rate, blood pressure, and a 12-lead ECG were obtained at baseline, at each stage of the exercise protocol, and during the PEx period. A treadmill exercise test (Bruce protocol in 96 patients, and modified Bruce in 4) was performed until exhaustion or until the patient reached an end point. End points included ST segment change greater than 2.5 mm, significant arrhythmia, severe hypertension (systolic blood pressure ⬎ 240 mm Hg or diastolic blood pressure ⬎ 110 mm Hg), severe hypotensive response (decrease ⬎ 20 mm Hg from baseline), or limiting symptoms. The exercise ECG was considered positive in case of horizontal or downsloping ST segment depression of at least 1 mm at 80 milliseconds after the J point or in case of S-T elevation in leads without q wave. 2DE The 2DE method using harmonic imaging was performed at baseline, peak exercise,9,10 and immediately after exercise. Apical long-axis, 4-chamber, and 2-chamber views, and parasternal long- and short-axis views were obtained at rest. The same order was followed at peak and PEx with the use of a continuous acquisition imaging system. Peak imaging was performed with the patient still exercising, when signs of exhaustion were present, end point was achieved, or both. The patient was asked to walk quickly rather than run, to decrease body and respiratory movements. The transducer was firmly positioned on the cardiac apex to obtain the apical views by applying slight pressure to the patient’s back with the left hand, so maintaining the patient between the transducer and the
Table 2 Clinical, electrocardiographic, hemodynamic, and echocardiographic data during 2- and 3-dimensional echocardiographic exercise testing 2DE
Angina, % Positive ECG, % Peak heart rate, beats/min Peak systolic blood pressure, mm Hg Peak heart rate ⫻ systolic blood pressure ⫻ 1000 Maximal age-predicted heart rate, % MET Resting wall-motion score index Peak wall-motion score index Postexercise wall-motion score index Resting ejection fraction Peak ejection fraction Postexercise ejection fraction
3DE
28 26 131 ⫾ 24 164 ⫾ 27
28 22 129 ⫾ 25 163 ⫾ 25
22.5 ⫾ 5.9
22.0 ⫾ 5.8
87 ⫾ 13
85 ⫾ 13
8.8 1.05 1.29 1.18
⫾ ⫾ ⫾ ⫾
2.8 0.16 0.35 0.27
59 ⫾ 6 59 ⫾ 12 64 ⫾ 11
9.0 1.05 1.23 1.15
⫾ ⫾ ⫾ ⫾
2.9 0.16 0.31† 0.25*
59 ⫾ 6 61 ⫾ 11* 63 ⫾ 11
ECG, Electrocardiography; MET, metabolic equivalent; 2DE, 2-dimensional echocardiography; 3DE, 3-dimensional echocardiography. *P ⬍ .05. †P ⬍ .01.
left hand, to avoid movement. Finally, the transducer was positioned in the parasternal region to obtain the parasternal views. Image acquisition was performed online and stored on disk. The 2DE analysis was performed on a digital quadscreen display system. 3DE A full matrix-array X3 transducer (2-4 MHz) with secondharmonic imaging was used. The 3DE images were acquired from the apical window at rest, peak exercise, and in the immediate PEx period. The 4- and 2-chamber apical views were displayed on the screen to help the operator to position the sample with all targets being covered by the imaging volume. Once the imaging volume was visualized, data sets of LV were obtained. Four conical subvolumes of 20 ⫻ 80 degrees were scanned during 4 consecutive cardiac cycles without moving the probe. ECG gating was used and the 4 subvolumes were acquired during 4 consecutive cardiac cycles. The 4 subvolumes were then automatically integrated in a full volume. Patients were taught to hold their breath before the start of the testing and they were asked to do it again during each full-volume acquisition (usually 4 seconds at rest and 2 seconds at exercise). One full volume was acquired at rest whereas 2 to 3 were acquired at peak exercise and in the immediate PEx period. The best full volume without motion and/or respiratory artefacts at peak and PEx was chosen for comparison with the resting one. Regional LV wall motion was analyzed using cropped planes. At least 3 short-axis views were produced by cropping the long axis of the LV from the apex to the base. At least two parallel long-axis, 4-chamber, and 2-chamber apical views were obtained for each patient. Therefore, various long-axis,
Journal of the American Society of Echocardiography Volume 20 Number 8
Peteiro et al 961
Table 3 Intraobserver and interobserver agreement of peak and postexercise 2- and 3-dimensional echocardiography 2DE
Intraobserver, % (k) Intraobserver, % (k)
3DE
Peak
Postexercise
Peak
Postexercise
92 (0.83) 88 (0.75)
92 (0.85) 88 (0.74)
93 (0.93) 88 (0.63)
96 (0.96) 87 (0.65)
2DE, Two-dimensional echocardiography; 3DE, 3-dimensional echocardiography.
Table 4 Sensitivity, specificity, and accuracy of peak and postexercise 2- and 3-dimensional echocardiography for the prediction of left anterior descending coronary artery and combined left circumflex/right coronary arteries stenoses Peak
LAD sensitivity, % LAD specificity, % LAD accuracy, % LCx/RCA sensitivity, % LCx/RCA specificity, % LCx/RCA accuracy, %
Figure 1 Sensitivity, specificity, and accuracy of 2- and 3-dimensional echocardiography at peak and postexercise for prediction of coronary artery disease.
4-chamber, and 2-chamber apical views and short-axis views were used for comparison of rest and peak exercise and rest and PEx images. LV Wall-motion Analysis The LV was divided into 16 segments.11 Each of the 16 segments was assigned to one of the 3 coronary artery territories, as recommended by the American Society of Echocardiography.12 The development of new regional dysfunction or worsening from hypokinesia in at least one segment was considered an ischemic response. The persistence of regional baseline dysfunction or worsening from akinesia to dyskinesia was considered as infarction without ischemia, except for the case of isolated hypokinesia of the basal inferior or basal septal segments.13 A positive EE result for CAD was defined when there was ischemia or infarction.14-16 A wall-motion score index at rest, peak, and PEx was calculated, with normal wallmotion scoring 1, hypokinetic scoring 2, akinetic scoring 3, and dyskinetic scoring 4. LV ejection fraction was assessed visually.17 Resting images were compared with peak and PEx images either by 2DE and 3DE in separate days from disk-stored images. Analysis was blinded to clinical characteristics, and clinical and ECG response to exercise. The quality of the studies was assessed by a score that assigned 0 to 3 points to each wall (apicoseptal, posterolateral, anterior, inferior) according to the visualization of
Postexercise
2DE
3DE
2DE
3DE
89 80 84 52 74 64
84 73 78 43 88 66
67 93 81 38 92 67
60 87 75 32 94 65
LAD, Left anterior descending coronary artery; LCx, left circumflex coronary artery; RCA, right coronary artery; 2DE, 2-dimensional echocardiography; 3DE, 3-dimensional echocardiography.
myocardial thickness and motion1 as followed: 0, very bad visualization or not acquired; 1, visualization of less than 70% of the extension of the wall; 2, visualization of 70% to 95% of the extension of the wall; and 3, visualization of 95% to 100% of the extension of the wall. Coronary Angiography Coronary angiography was performed within 2 weeks of the EE studies. A luminal narrowing of 70% or more by visual estimation in a coronary artery, major branch, or bypass, or 50 or greater in the left main coronary artery were considered significant. Of the 84 patients submitted to angiography, 22 patients had one-vessel disease and 36 had multivessel CAD. Of the remaining 26 patients, 11 had normal coronary arteries and 15 had noncritical disease. Left anterior descending coronary artery (LAD) disease was found in 45 patients, left circumflex coronary artery (LCx) disease in 36 patients, right coronary artery (RCA) disease in 32 patients, and left main CAD in 7 patients. Statistical Analysis Continuous variables are expressed as mean ⫾1 SD and compared by use of paired Student t test. Sensitivity was defined as the number of true positive tests divided by the number of patients or territories irrigated by a stenotic vessel. Specificity was defined as the number of true negative tests divided by the number of patients or territories irrigated by a nonstenotic vessel. Accuracy was defined as the total number of true positive and true negative tests divided by the total number of patients or
962 Peteiro et al
Journal of the American Society of Echocardiography August 2007
Figure 2 Example of 64-year-old man with normal angiography and normal peak exercise 3-dimensional echocardiography results. Resting (left), peak exercise (right), diastolic (D), and systolic (S) images. Cropped 4-chamber (A) and 2-chamber (B) apical views. Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. Note increase in regional function in all cropped views. (See supplementary material on line: Videos 1 and 2 for apical views and Video 3 for short-axis views.) territories. A value of P less than .05 was considered significant for comparisons. Values of area under the curve (AUC) and 95% confidence interval (CI) are given from receiver operator characteristic curves. The agreement between the test results obtained by 2DE and 3DE, and the agreement between observers were tested by measuring the coefficient of variation in 25 randomized patients. A value of 0.45 or greater was considered to be good agreement, and a value of greater than 0.75 was considered to be excellent agreement.
RESULTS All the patients completed both EE studies without complications. Exercise testing data were not different in 2DE and 3DE exercise testing (Table 2). Resting wall-motion abnormalities (WMA) were detected in 8 patients by the two techniques (3 of
them had previous MI), in 4 patients by 2DE alone, and in 5 patients by 3DE alone (baseline agreement 91%, ⫽ 0.59). Feasibility of 3DE and 2DE Feasibility of peak 2DE, peak 3DE, PEx 2DE, and PEx 3DE was 99%, 92%, 100%, and 95%, respectively (P ⬍ .05 between peak 2DE and peak 3DE; P ⬍ .01 between post-2DE and peak 3DE). Both peak 2DE and 3DE imaging were not possible in one patient because of premature interruption of exercise, whereas 3DE was unfeasible or unreadable at peak or at PEx in 7 and 5 patients, respectively. Quality scores were qualified as 0 in these patients. Inadequate quality of images (score 0-1 in any wall) was seen in only 4% of the walls with peak 2DE and in none with PEx 2DE, but they were seen in 18% of the walls at peak 3DE and in 14% of the walls at PEx 3DE. These scores lower than 2 were more fre-
Journal of the American Society of Echocardiography Volume 20 Number 8
Peteiro et al 963
Figure 3 Example of 75-year-old-man with 3-vessel coronary artery disease (left anterior descending [90%], right [95%], and left circumflex [90%] coronary arteries) and abnormal peak exercise 3-dimensional echocardiography results. Resting (left), peak exercise (right), diastolic (D), and systolic (S) images. Cropped 4-chamber (A) and 2-chamber (B) apical views. Note regional dysfunction at peak exercise in apicoseptal segments (A) (arrows) and exercise-induced dysfunction in anterior and posterior walls (B) (arrows). Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. See exerciseinduced wall-motion abnormalities and dilation at apical level (C) (arrows). (See supplementary material on line: Videos 4 and 5 for apical views and Video 6 for short-axis views.)
quently seen in the inferior and anterior walls in contrast to the septoapical and posterolateral territories (75% vs 25%). This was caused by inadequate match of subvolumes at peak or post-EE or bad acoustic wall definition. The mean quality score (mean score of the 4 walls) was lower with 3DE at baseline (2.9 ⫾ 0.2 vs 3.0 ⫾ 0.1, P ⬍ .05), peak (2.4 ⫾ 0.9 vs 2.9 ⫾ 0.3, P ⬍ .0001), and PEx (2.5 ⫾ 0.8 vs 3.0 ⫾ 0.1) imaging. Total imaging optimization and acquisition time in 25 randomized patients was lower with 3DE at rest (10 ⫾ 4 vs 30 ⫾ 8 seconds, P ⬍ .0001), peak (22 ⫾ 8 vs 43 ⫾ 14 seconds, P ⬍ .0001), and PEx (15 ⫾ 5 vs 31 ⫾ 14 seconds, P ⬍ .0001).
Interobserver and Intraobserver Agreement Interobserver and intraobserver agreement (positive or negative test) of peak and PEx 2DE and 3DE is shown in Table 3. Agreement Between Studies Agreement between 2DE and 3DE at peak exercise was 82% ( ⫽ 0.62), whereas at PEx it was 78% ( ⫽ 0.55). There was a trend to better agreement in the detection of WMA in LAD and LCx territories than in RCA territory at peak imaging (LAD 84%, ⫽ 0.67; LCx 86%, ⫽ 0.67; RCA 74%, ⫽ 0.27), whereas agreement was similar at PEx imaging for the differ-
964 Peteiro et al
Journal of the American Society of Echocardiography August 2007
Figure 4 Example of 76-year-old woman with proximal left anterior descending (LAD) (99%) and right (70%) coronary artery stenosis and abnormal peak exercise 3-dimensional echocardiography results (ischemia in LAD territory). Resting (left), peak exercise (right), diastolic (D), and systolic (S) images. Cropped 4-chamber (A) and 2-chamber (B) apical views. Note regional dysfunction at peak exercise in apicoseptal segments (A) (arrows) and exercise-induced dysfunction in anterior wall and apex (B) (arrows). Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. Note septal dyssynergia at peak exercise (C) (arrow). (See supplementary material on line: Videos 7, 8, and 9 for apical views, Video 10 for short-axis apical view, and Video 11 for short-axis, mid, and basal views.)
ent territories (LAD 78%, ⫽ 0.52; LCx 87%, ⫽ 0.55; RCA 89%, ⫽ 0.53). Sensitivity, Specificity, and Accuracy of 3DE and 2DE at Peak and PEx All 16 patients with low pretest probability considered to have no CAD had normal EE results by either peak and PEx 2DE, whereas false-positive results were seen in two patients by peak and in one patient by PEx 3DE, respectively. Of the 26 patients submitted to angiography who had no CAD, false-positive results were seen in 10 patients by peak 2DE, in 3 by PEx 2DE, in 8 by peak 3DE, and in 4 by PEx 3DE. Sensitivity, specificity, and accuracy of 3DE and 2DE at peak and PEx for the detection of CAD in a binary manner and according to the different arteries in-
volved are shown in Figure 1 and Table 4, respectively. The higher sensitivity was achieved with peak imaging (either 2DE or 3DE). Specificity was slightly higher with PEx 2DE. Accuracy was similar: 81% by peak 2DE (AUC 0.81, 95% CI ⫽ 0.71-0.91); 76% by peak 3DE (AUC 0.76, 95% CI ⫽ 0.65-0.86); 80% by PEx 2DE (AUC 0.84, 95% CI ⫽ 0.75-0.92); and 71% by PEx 3DE (AUC 0.73, 95% CI ⫽ 0.62-0.83). There was a trend to higher sensitivity of each approach in patients who were not on beta-blocker therapy (peak 2DE 89% vs 75%; peak 3DE 83% vs 68%; PEx 2DE 76% vs 62%; and PEx 3DE 66% vs 45%). Sensitivity in 36 patients with multivessel CAD was 97% for peak 2DE, 88% for peak 3DE, 83% for PEx 2DE, and 68% for PEx 3DE (P ⬍ .001 vs peak 2DE, P ⬍ .1 vs peak 3DE). An example of normal
Journal of the American Society of Echocardiography Volume 20 Number 8
Peteiro et al 965
Figure 5 Example of 48-year-old man with right coronary artery (RCA) and left circumflex coronary artery (LCx) stenoses (both 99%) and abnormal peak exercise 3-dimensional echocardiography results (ischemia in RCA/LCx territories). Cropped 4-chamber (A) and 2-chamber (B) apical views. Note regional dysfunction at peak exercise in basal septal segment (A) (arrow) and exercise-induced dysfunction in posterior wall (B) (arrows). Cropped short-axis views at apical (C), midventricular (D), and basal (E) levels. Note exercise-induced abnormalities in posterior septum in short-axis views (arrows). (See supplementary material on line: Videos 12 and 13 for apical views and Video 14 for short-axis views.)
peak exercise 3DE is showed in Figure 2, and Videos 1 to 3 whereas Figures 3 to 5 and Videos 4 to 14 are examples of ischemic responses by peak 3DE.
DISCUSSION The main findings of this study were that volumetric 3DE is as accurate as 2DE during EE, but feasibility and imaging quality were lower than those of 2DE. Although 2DE and 3DE by acquisition of a full LV volume have been compared during pharmacologic stress with dobutamine in the same study they have not during exercise testing.
3DE During Dobutamine Two different approaches have been used by matrixarray transducer technology during dobutamine stress: the volumetric approach and the multiplane imaging approach that allows the simultaneous visualization and acquisition of 2 to 3 LV planes. The volumetric approach resulted marginally superior to 2DE in terms of sensitivity (89% vs 79%) in one study where exercise single photon emission computed tomography was the reference standard,2 and it was similar in another one (86% vs 86%) that used coronary angiography.1 With the use of multiplane imaging, Eroglu et al18 demonstrated identical sensitivity (93%) and specificity (75%). In all these studies
966 Peteiro et al
the acquisition time was lower with 3DE. Although real-time multiplane imaging has higher temporal resolution than volumetric 3DE, the number of views than can be obtained by volumetric 3DE is greater. 3DE During Exercise The 3DE method may theoretically be more useful for exercise than for pharmacologic stress because of the narrow acquisition time window, particularly at PEx. Actually, we and others19 demonstrated lower acquisition time with volumetric 3DE than with 2DE during exercise. Advantages of exercise 3DE according to the results of this study would be lower acquisition time and the possibility of image multiple parallel planes in each view. Uncertainty regarding the presence/absence of WMA could be resolved by reviewing the parallel or orthogonal available views. The major disadvantages come from an inadequate matching of the subvolumes and from the low frame rate. The inadequate merging of the subvolumes is usually produced by body or breathing movements, or the eventual presence of an artefactual ECG. It is important to teach patients how to hold their breath during the approximately 2 seconds of the acquisition at peak or PEx while avoiding previous inspiration. As the respiratory movements may be suppressed by asking patients to hold their breath, it is more difficult to avoid body movements during exercise. An artefactual ECG or the development of arrhythmia prevents data blocks acquisition. The low frame rate may underdiagnose minor hypokinesia or tardokinesia. Improvement in imaging quality (spatial and temporal resolution) and acquisition of the full volume in only one cycle may constitute requirements for the future application of this technique during exercise. Limitations Exercise 3DE was less feasible than exercise 2DE, therefore, any comparison of accuracy between both techniques should be taken cautiously. Peak exercise is a highly dependent imaging technique modality. However, its major advantage is the relative lack of hurry for imaging acquisition in comparison with PEx where a limited time window exists. Apical 2DE views are particularly easy to acquire and their quality was the same of those of PEx 2DE in a previous report.10 Therefore, a fullvolume 3DE acquisition from the apical view is theoretically feasible during peak exercise. Most of our patients had adequate acoustic window with 2DE. Therefore, these results may not be applied to the total population of patients referred for EE. A significant percentage of them were taking beta-blockers at the time of the tests. Thus, global sensitivity for the prediction of CAD may be infraestimated. There was a trend to higher sensitivity by
Journal of the American Society of Echocardiography August 2007
each of the approaches in patients without betablockers. Patients with recent S-T elevation MI were not included and consequently the percentage of patients with resting WMA was small. This fact also explains why sensitivity is lower than in studies that include a significant percentage of patients with MI and resting WMA. Particularly weak sensitivity was obtained for the RCA and LCx. Several factors may explain this finding. First, sensitivity of stress echocardiography is in general lower for these arteries as compared with the LAD, because the LAD territory is nearest to the transducer from each view. Second, lower imaging quality in the inferior wall may explain low sensitivity for the RCA, particularly with 3DE. Third, as above referred, 30% of the patients were on beta-blockers. We did not use contrast agents in this study. With the use of contrast agents, the quality of 3DE may be enhanced to be comparable with peak and PEx 2DE.20 Finally, patients with atrial fibrillation/flutter, LV bundle branch block, or pacemaker ectopic rhythm were not included because these conditions may prevent adequate match of the subvolumes with 3DE. If included, the feasibility of 3DE may be even lower. Conclusions Currently, 3DE is not superior to 2DE during EE, mainly because the imaging quality and feasibility are lower. However, peak 3DE is similar to peak 2DE and PEx 3DE is similar to PEx 2DE in terms of sensitivity and specificity.
REFERENCES 1. Matsumura Y, Hozumi T, Arai K, Sugioka K, Ujino K, Takemoto Y, et al. Non-invasive assessment of myocardial ischemia using new real-time three-dimensional dobutamine stress echocardiography: comparison with conventional two-dimensional methods. Eur Heart J 2005;26:1625-32. 2. Ahmad M, Xie T, McCulloch M, Abreo G, Runge M. Realtime three-dimensional dobutamine stress echocardiography in assessment of ischemia: comparison with two-dimensional dobutamine stress echocardiography. J Am Coll Cardiol 2001;37:1303-9. 3. Varga A, Garcia MA, Picano E; International Stress Echo Complication Registry. Safety of stress echocardiography (from the international stress echo complication registry). Am J Cardiol 2006;98:541-3. 4. Rodríguez García MA, Iglesias-Garriz I, Corral Fernández F, Garrote Coloma C, Alonso-Orcajo N, Branco L, et al. Evaluation of the safety of stress echocardiography in Spain and Portugal. Rev Esp Cardiol 2001;54:941-8. 5. Beleslin BD, Ostojic M, Stepanovic J, Djordjevic-Dikic A, Stoljkovic S, Nedeljkovic G, et al. Stress echocardiography in the detection of myocardial ischemia: head-to-head comparison of exercise, dobutamine, and dipyridamole tests. Circulation 1994;90:1168-76. 6. Dagianti A, Penco M, Agati L, Sciomer S, Dagianti A, Rosanio S, et al. Stress echocardiography: comparison of exercise,
Journal of the American Society of Echocardiography Volume 20 Number 8
7.
8.
9.
10.
11.
12.
13.
14.
dipyridamole and dobutamine in detecting and predicting the extent of coronary artery disease. J Am Coll Cardiol 1995;26:18-25. Picano E. Exercise echocardiography. In: Picano E, editor. Stress echocardiography. 4th ed. Berlin: Springer-Verlag; 2003. p. 103-14. Marrugat J, Solanas P, D’Agostino R, Sullivan L, Ordovas J, Cordón F, et al. Coronary risk estimation in Spain using a calibrated Framingham function. Rev Esp Cardiol 2003;56: 253-61. Peteiro J, Monserrat L, Perez R, Vazquez E, Vazquez JM, Castro-Beiras A. Accuracy of peak treadmill exercise echocardiography to detect multivessel coronary artery disease: comparison with post-exercise echocardiography. Eur J Echocardiogr 2003;4:182-90. Peteiro J, Garrido I, Monserrat L, Aldama G, Calvino R, Castro-Beiras A. Comparison of peak and postexercise treadmill echocardiography with the use of continuous harmonic imaging acquisition. J Am Soc Echocardiogr 2004;17: 1044-9. Bourdillon PD, Broderik TM, Sawada SG, Armstrong WJ, Ryan T, Dillon JC, et al. Regional wall motion index for infarct and noninfart regions after reperfusion in acute myocardial infarction: comparison with global wall motion index. J Am Soc Echocardiogr 1989;9:398-407. Schiller NB, Shah PM, Crawford M, DeMaria A, Devareaux R, Feigenbaum H, et al. Recommendations for quantification of the left ventricle by two dimensional echocardiography. J Am Soc Echocardiogr 1989;2:358-67. Hoffmann R, Lethen H, Marwick T, Rambaldi R, Fioretti P, Pingitore A, et al. Standardized guidelines for the interpretation of dobutamine echocardiography reduce interinstitutional variance in interpretation. Am J Cardiol 1998;82: 1520-4. Roger VL, Pellikka PA, Oh JK, Bailey KR, Tajik AJ. Identification of multivessel disease by exercise echocardiography. J Am Coll Cardiol 1994;24:109-14.
Peteiro et al 967
15. Quiñones MA, Verani MS, Haichin RM, Mahmariam JJ, Suarez J, Zoghbi WA. Exercise echocardiography versus 201 Tl single-photon emission computed tomography in evaluation of coronary artery disease: analysis of 292 patients. Circulation 1992;85:1026-31. 16. Ryan T, Vasey CG, Presti CF, O’Donnell JA, Feigenbaum H, Armstrong WF. Exercise echocardiography: detection of coronary artery disease in patients with normal left ventricular wall motion at rest. J Am Coll Cardiol 1988;11:993-9. 17. Stamm RB, Carabello BA, Mayers DL, Martin RP. Twodimensional echocardiographic measurement of left ventricular ejection fraction: prospective analysis of what constitutes an adequate determination. Am Heart J 1982;104:136-44. 18. Eroglu E, D’hooge J, Herbots L, Thijs D, Dubois C, Sinnaeve P, et al. Comparison of real-time tri-plane and conventional 2D dobutamine stress echocardiography for the assessment of coronary artery disease. Eur Heart J 2006;27:1719-24. 19. Zwas DR, Takuma S, Mullis-Jansson S, Fard A, Chaudhry H, Wu H, et al. Feasibility of real-time 3-dimensional treadmill stress echocardiography. J Am Soc Echocardiogr 1999;12: 285-9. 20. Takeuchi M, Otani S, Weinert L, Spencer KT, Lang RM. Comparison of contrast-enhanced real-time live 3-dimensional dobutamine stress echocardiography with contrast 2dimensional echocardiography for detecting stress-induced wall-motion abnormalities. J Am Soc Echocardiogr 2006;19: 294-9.
SUPPLEMENTARY DATA Supplementary data associated with this article can be found, in the online version, at 10.1016/j.echo. 2007.01.034.