The authors recently demonstrated that simultaneous assessment of myocardial perfusion (MP) and wall motion (WM) by myocardial contrast echocardiography (MCE) is feasible and accurate when incorporated into a clinical stress echocardiography (SE) service. However, it is unknown whether the incremental prognostic value of MP beyond WM, previously shown in research studies, is reproducible when MCE is performed in the clinical arena.
In this prospective study, MCE was performed by multiple operators during routine clinical SE, whose results were classified as normal WM and MP, abnormal WM and MP, or normal WM but abnormal MP. Patients were followed for the prospectively determined combined primary outcome of all-cause mortality, nonfatal myocardial infarction, and late revascularization. Cox regression analyses were performed to identify predictors of outcome.
Of 220 patients undergoing simultaneous MCE during SE, 197 patients (90%) with interpretable WM and MP images were available for follow-up at a mean time period of 17 ± 7 months. There were 35 events (six deaths, six myocardial infarctions, and 23 revascularizations). Among prognostic clinical variables, resting left ventricular function, and WM and MP data, abnormal MP at peak stress was the only independent predictor of primary outcome (hazard ratio, 4.41; 95% confidence interval, 1.37–14.20; P = .02). Sequential Cox regression models showed that abnormal MP also carried incremental prognostic value over clinical variables, resting left ventricular function and abnormal WM.
In keeping with previous research studies, this prospective study demonstrates the incremental prognostic benefit of MP assessment beyond WM when MCE is incorporated into a clinical SE service.
Inducible myocardial ischemia carries clinical and prognostic importance in patients with known or suspected coronary artery disease (CAD). Several imaging techniques can determine the presence and extent of ischemia, one of which is stress echocardiography (SE). The hallmark of ischemia during SE is a new wall motion (WM) abnormality. However, because decreased myocardial perfusion (MP) precedes systolic dysfunction in the ischemic cascade, SE is considered less sensitive than techniques such as single-photon emission computed tomography for the detection of CAD and multivessel disease, both of which have prognostic implications.
Myocardial contrast echocardiography (MCE) is an ultrasound-based technique that exploits the acoustic properties of gas-filled microbubbles. Because these microbubbles remain exclusively within the intravascular space, their presence within myocardial capillaries allows the accurate determination of MP. Contrast microbubbles are approved for clinical use for improved accuracy of WM assessment during rest echocardiography and SE. However, MCE also permits the simultaneous assessment of MP and WM, which is advantageous first because MP defects precede WM abnormalities in the ischemic cascade and second because MP assessment can uncover subtle WM abnormalities through subendocardial defects.
Numerous studies have evaluated both the diagnostic superiority of MCE and prognostic benefit of MP over WM alone. However, there are fewer data on the role of MCE in clinical practice. The lack of implementation of MCE clinically is due to the perception that MCE may lack feasibility, requires a prohibitive level of expertise, and is too cumbersome for day-to-day practice. We recently reported the findings from the Incorporation of Myocardial Perfusion Assessment into Clinical Testing with Stress Echocardiography study, a single-center, prospective study of 220 patients representing our experience of incorporating MCE into clinical practice. In that study, we established criteria for our echocardiography laboratory for determining who should undergo MCE, based in part on previous research that has demonstrated that MCE has the most incremental value when images are analyzed at low work level. Our study, a prospective report of our clinical experience, demonstrated that MCE has excellent feasibility when performed by multiple operators within the time constraints of a clinical SE service. We also found that MCE has superior sensitivity for detection of left anterior descending CAD and multivessel disease and detects a larger ischemic burden compared with WM alone. The present study has now followed these same 220 patients of the Incorporation of Myocardial Perfusion Assessment into Clinical Testing with Stress Echocardiography study to determine whether the superior diagnostic value of MCE translates into improved prognostic impact of MCE over WM alone when incorporated into a routine clinical SE service.
This study involved follow-up of a recently studied patient cohort, for whom the criteria for performing MCE have recently been described. In brief, we performed MCE in consecutive adult patients referred for SE for the evaluation of CAD receiving pharmacologic stress and consecutive patients undergoing treadmill exercise in whom the cardiologist was not certain that a high workload or target heart rate (THR) would be achieved (and thus it was anticipated that additional MP data would be helpful in addition to WM). This was a judgement made at the time of the stress echocardiographic study on the basis of the patient’s general level of fitness and limitations to exercise, such as orthopedic or respiratory disorders or significant obesity.
Patients who clearly had no mobility, respiratory, or frailty issues and who exercised regularly were deemed likely to reach a high workload and attain the THR, and thus we hypothesized that MCE would provide little extra information over WM in these patients, as it is known that the hard event rate in patients undergoing exercise echocardiography who achieve good workloads and who achieve the THR with no WM abnormalities at rest or stress is <1% per year. However, patients who perhaps seemed less robust or less confident in their exercise capacity—and thus whom we judged might not attain a high workload—were those in whom we felt that additional data on MP, as well as WM, would be of clinical value. A study by Elhendy et al ., which showed that the sensitivity for the detection of flow-limiting CAD at intermediate stress (defined as 65%–75% of the THR) is 84% by MCE versus just 20% by WM ( P < .001), we believed helped justify this approach.
All patients were considered for MCE during SE, irrespective of the quality of the resting echocardiogram. Only patients with known sulfur allergies were not candidates for MCE. Patient demographics and results of SE were collected prospectively when the patient attended for SE. Patients were deemed to have preexisting ischemic heart disease if there was a history of acute myocardial infarction (AMI) or coronary revascularization. Approval to undertake this study was granted by the local institutional review board.
Images of the left ventricle (parasternal long-axis, short-axis, and apical four-, two-, and three-chamber views) were obtained with the patient in the left lateral decubitus position (iE33; Philips Medical Systems, Eindhoven, The Netherlands) at rest and peak stress by two-dimensional echocardiography. A real-time low–mechanical index perfusion setting (0.10) was used for both WM imaging and MCE. Standard symptom-limited treadmill exercise was the preferred stress modality, but in patients unable to exercise, dobutamine echocardiography was performed.
For assessment of ischemia only, dobutamine was infused peripherally in 3-min dose increments, starting from 10 μg/kg/min and increasing to 20, 30, and 40 μg/kg/min. In patients with resting WM abnormalities, both ischemia and viability were assessed, commencing dobutamine at 5 μg/kg/min and increasing to 10, 15, and then 20 μg/kg/min at 5-min intervals. Thereafter, the dose was increased to 30 and 40 μg/kg/min at 3-min intervals, per the ischemia protocol. If no end point was reached, atropine was added to the continuing 40 μg/kg/min dobutamine infusion in bolus doses of 0.3 mg (up to a maximum of 1.2 mg). For ischemia protocol studies, imaging was performed at rest and at low-dose and peak-dose dobutamine, whereas for viability studies, images were acquired at rest and at low-, intermediate-, and peak-dose stages of the dobutamine infusion. Criteria for a diagnostic study were attainment of >85% age-predicted heart rate, chest pain, ST-segment elevation, intolerable side effects of dobutamine, ventricular arrhythmia, severe hypotension or hypertension, and new or worsening WM abnormality.
For the minority of patients unable to exercise and in whom dobutamine was contraindicated, WM and MP assessment was performed at rest and following slow intravenous injection, over 4 min, of dipyridamole (0.56 mg/kg body weight). If there was no change in heart rate or the patient had not experienced any symptoms, a further 0.28 mg/kg dipyridamole was administered over 2 min. All patients were given aminophylline (50–75 mg) on completion of the test.
Resting left ventricular (LV) systolic function was normal if all 17 segments had normal contractility and abnormal if any segment exhibited a WM abnormality. Ejection fraction (EF) was quantitatively assessed using the biplane Simpson method in patients with resting WM abnormalities and visually judged as >50% in patients with normal contractility of all segments. The presence of ischemia was defined as a new or worsening WM abnormality at peak stress, including a biphasic response during viability studies. Inducible ischemia of one or two segments was classified as mild ischemia, whereas ischemia of three or more segments was considered moderate to severe.
MCE was performed in the apical views using low-power real-time and triggered (end-systolic) imaging. Machine settings for optimal perfusion imaging have recently been described in this patient cohort. All patients received intravenous infusion of SonoVue (0.8–1.0 mL/min; Bracco, Milan, Italy) using the VueJect infusion syringe pump (BR-INF 100; Bracco). Additionally, for patients performing exercise, an intravenous bolus injection (0.5 mL) of contrast was given 10 sec before terminating exercise. Nonstandard apical views (e.g., bringing the lateral wall into the sector field) were used, if required, to overcome attenuation artefacts.
Flash echocardiography at a high mechanical index (0.5–1.0) was performed to clear microbubbles in the myocardium and permit evaluation of the rate of microbubble replenishment. This was followed by real-time and end-systolic triggered image acquisition digitally up to 15 cardiac cycles at a mechanical index of 0.1 in each view. Because patients with normal resting LV function must by definition have normal MP, in these patients, MCE was performed at rest in only one view (apical four chamber) to optimize settings (flash frames, flash mechanical index, gain, and time gain compensation settings) for performing MCE at peak stress, which was done immediately after WM cine loops had been acquired. All cardiologists performing SE and MCE were experienced with conventional SE, so it was possible to obtain WM cine loops within 30 to 45 sec (following a bolus injection of SonoVue while patients were on the treadmill), and because the SonoVue infusion was reconnected to the patient upon returning to the echocardiography couch, MCE was commenced immediately thereafter, and image acquisition was completed by 90 sec.
Inducible ischemia was deemed present when subendocardial or transmural defects appeared that replenished beyond 1 sec after flash during stress. Worsening of resting perfusion (defined as patchy contrast uptake at rest becoming absent beyond 1 sec after flash) was also deemed indicative of ischemia. As for WM, mild ischemia was defined as one or two segments with inducible ischemia, whereas moderate to severe ischemia was defined as three or more segments with inducible ischemia. This was performed on a segmental basis using the same 17-segment model used for WM analysis.
Importantly, in our clinical stress echocardiographic protocol, WM remains the cornerstone for the assessment of ischemia, and MP findings were used to assist in reporting WM in patients in whom WM was deemed equivocal (i.e., MP status was used to establish the presence or absence of ischemia). For example, if WM was equivocal but MP was normal, both WM and MP were called normal in that segment (and vice versa). However, if WM was unequivocally normal yet MP was abnormal, the segment was classified as having abnormal MP only. On this basis, the results of SE were classified into one of three categories: normal WM and normal MP, normal WM but abnormal MP, and abnormal WM and abnormal MP. All myocardial contrast echocardiographic data were interpreted by the performing cardiologist together with the expert reader (R.S.).
Patients listed for revascularization after the initial stress or myocardial contrast echocardiographic procedure usually had this performed within 6 to 8 weeks. Thus, early revascularization (ER) was defined as percutaneous coronary intervention or coronary artery bypass grafting occurring before 3 months, whereas late revascularization (LR) included any patients undergoing revascularization after 3 months, because these procedures were likely due to recurrence or persistence of symptoms or a new symptom of angina rather than the test result. Patients were followed for prospectively defined primary outcomes, which were a composite of all-cause mortality, nonfatal AMI, and LR (>3 months after SE). Hard events were death and AMI only. The UK national mortality database and hospital and general practitioner databases were searched to obtain information about the occurrence and timing of mortality and AMI. Mortality was classified as the only event in patients who had AMI or LR, and similarly AMI was classified as the only event even if the patient had LR.
Categorical variables are expressed as percentages and continuous variables as mean ± SD. Cox regression analysis was performed to assess the prognostic impact of the clinical variables (as listed in Table 1 ), resting LV function, and ischemia by WM versus ischemia by MP for the prediction of events. Only variables with P values < .10 in the univariate analysis were entered into the multivariate model. Kaplan-Meier survival curves were constructed showing patient survival plotted against time and were compared using the log-rank score test. To evaluate the incremental benefit of MP over clinical variables, resting LV function, and abnormal WM data, a stepwise model was created, and the incremental prognostic value of the added variables was determined by comparing the global χ 2 values obtained at each step.
|Age (y)||66 ± 11|
|Current smoker||48 (24%)|
|Prior AMI||87 (44%)|
|Prior revascularization||92 (47%)|
|EF > 50%||157 (80%)|
To establish interobserver agreement, 19 stress echocardiographic studies with MCE were randomly selected and interpreted by two experienced cardiologists as normal versus abnormal and compared against the expert reader. For all tests, P value < .05 was considered to indicate statistical significance. All statistical analyses were performed with SPSS version 19.0 (SPSS, Inc, Chicago, IL).
Of the 220 patients who underwent simultaneous MCE during SE, performed by eight operators, 209 (95%) were available for follow-up at a mean time interval of 17 ± 7 months. There were 12 patients (with follow-up) in whom MP was uninterpretable (eight exercise stress and four dobutamine stress). All 12 patients had normal WM, and thus the clinical SE report had stated no inducible ischemia, but because MP could not be assessed in these patients, they were excluded from subsequent analyses, thus leaving a final sample size of 197 patients. The mean age was 66 ± 11 years, and 73% were men. Almost half of the patients had histories of AMI, and almost half had previously undergone myocardial revascularization. Complete baseline demographics are provided in Table 1 .
SE and MCE
Of the 197 patients, 116 (59%) underwent dobutamine SE, 78 (40%) treadmill exercise SE, and three (1%) dipyridamole SE. Normal WM and normal MP were reported in 111 (56%) patients, normal WM but abnormal MP in 14 (7%) patients, and abnormal WM and abnormal MP in 72 (37%) patients. One patient demonstrated normal MP but abnormal WM. This patient had an LV EF of <35% with suspected significant aortic stenosis and, after SE, had normal arteries on coronary angiography. Thus, this patient was excluded from the analyses.
Following SE and MCE, 83 of 197 patients (42%) proceeded to coronary angiography, and flow-limiting CAD was present in 56 (68%), of whom 31 (54%) underwent ER. Of the 111 patients who had normal WM and MP, 24 (20%) underwent catheterization, but none underwent ER. On the contrary, of the 14 patients with normal WM and abnormal MP, six (42%) underwent cardiac catheterization and three (50%) underwent ER. Finally, of the 72 patients with both abnormal WM and MP, 53 (74%) underwent catheterization and 28 (50%) underwent revascularization. All 31 patients who underwent ER had abnormal MP, and 26 (84%) had abnormal WM also. On the other hand, among patients who did not undergo ER, only 55 of 178 (31%) had ischemia.
There were 35 events (18%) among the 197 patients over a mean follow-up period of 17 ± 7 months. These consisted of six deaths, six nonfatal AMIs, and 23 LR procedures (14 ± 8 months). Of the 23 patients undergoing LR, 17 procedures (74%) were due to progressive dyspnea (five patients) or recurrent chest pain (12 patients). In the remaining six patients, the reason(s) were unclear. However, the fact that LR occurred at a mean time of 14 ± 8 months suggests that the procedures occurred because of recurrent or progressive symptoms. The distribution of events, on the basis of initial results of SE, is outlined in Table 2 .
|WM and MP status||Primary outcome||Death or AMI|
|Normal WM and normal MP||11/111 (10%)||4/111 (4%)|
|Normal WM and abnormal MP||4/14 (29%)||1/14 (7%)|
|Abnormal WM and abnormal MP||20/72 (28%)||7/72 (10%)|
Cox regression analysis was performed to identify independent predictors of the primary outcome, and these results are shown in Table 3 . Univariate predictors of all events included EF < 50%, prior AMI, abnormal WM, and abnormal MP. However, in the multivariate model, the only independent predictor of outcome was abnormal MP (hazard ratio [HR], 3.98; 95% confidence interval [CI], 1.23–12.84; P = .02). We repeated the regression analyses including the 12 patients with inconclusive results on MCE (clinically reported as normal results), and because there were no events among these 12 patients, the overall results were unchanged, and abnormal MP remained the only predictor of the primary outcome (HR, 4.41; 95% CI, 1.37–14.20; P = .01).
|Variable||Univariate analysis||Multivariate analysis|
|HR||95% CI||P||HR||95% CI||P|
|Prior myocardial infarction||2.41||1.21–4.80||.01||1.38||0.60–3.16||.45|
|EF < 50%||2.46||1.22–4.96||.01||1.66||0.75–3.67||.21|