Author + information
- Received May 16, 2008
- Revision received August 14, 2008
- Accepted August 25, 2008
- Published online January 6, 2009.
- Melda S. Dolan, MD⁎,⁎ (, )
- Simil S. Gala, MD⁎,
- Saritha Dodla, MD†,
- Sahar S. Abdelmoneim, MD‡,
- Feng Xie, MD†,
- David Cloutier, MS†,
- Michelle Bierig, MPH, RDCS, RDMS⁎,
- Sharon L. Mulvagh, MD, FACC, FRCP‡,
- Thomas R. Porter, MD, FACC, FASE† and
- Arthur J. Labovitz, MD, FACC, FACP, FCCP, FASE⁎
- ↵⁎Reprint requests and correspondence:
Dr. Melda S. Dolan, Division of Cardiology, Saint Louis University, 3635 Vista Avenue, 14th Floor, St. Louis, Missouri 63110-0250
Objectives The authors sought to define the risks versus benefits of ultrasound contrast agents in patients undergoing stress echocardiography.
Background The Food and Drug Administration recently placed a “black box” warning on the ultrasound contrast agents Definity (Bristol-Myers Squibb Medical Imaging, Billerica, Massachusetts) and Optison (GE Healthcare, Princeton, New Jersey) after their use was temporally related to 4 deaths. The safety of contrast has not been systematically evaluated.
Methods We retrospectively analyzed 42,408 patients at 3 different institutions who had baseline suboptimal images and/or underwent myocardial perfusion imaging and received contrast agents; 18,749 of these underwent stress echocardiography. The outcomes (death and myocardial infarction [MI]) within 30 min, 24 h, and during long-term follow-up were recorded.
Results No deaths or MIs were observed within 30 min; 1 death and 5 nonfatal MIs were observed within 24 h. This was not different from a matched cohort of 15,989 patients not receiving contrast. At 1 h and at 30 days after contrast administration, no significant differences in death rates or MIs were observed between patients who did and did not receive contrast during their stress echocardiogram. Endocardial border visualization in patients with suboptimal images resulted in comparable sensitivity (81% vs. 73%, p = NS) and diagnostic accuracy (82% vs. 77%, p = NS) for wall motion analysis compared with patients with optimal image quality. At long-term follow-up, abnormal wall motion and/or myocardial perfusion predicted adverse outcomes (20.6%) when compared with patients with normal studies (3.7%).
Conclusions Despite recent warnings regarding echocardiographic contrast, our findings indicate it is a safe and useful diagnostic tool in assessment of patients suspected of having coronary artery disease.
Transthoracic echocardiography has long been recognized as a safe, noninvasive, and reproducible tool that has been used to gain valuable information about cardiac structure and function. Unfortunately, up to 30% of studies are deemed technically difficult due to poor image quality, making these studies very challenging to interpret (1). Patients with large body habitus, chest wall deformities, and severe chronic lung disease are at higher risk of having suboptimal image quality (1–3). Furthermore, bedside echocardiograms are frequently ordered to identify the cause of hemodynamic instability and respiratory failure in critically ill patients. Since these patients cannot be turned on their side, the imaging windows are suboptimal and the diagnostic information obtained is limited (1–4).
In 1997, the Food and Drug Administration (FDA) approved the use of ultrasound contrast agents (UCAs) to improve the diagnostic accuracy of echocardiography after reviewing safety and efficacy data. Before approval, studies showed no significant changes in systemic or pulmonary hemodynamics, myocardial contractility, or regional myocardial blood flow (5). Phase III studies subsequently confirmed the absence of safety concerns and led to the approval of UCAs (6,7). The approved indication was for left ventricular (LV) endocardial border delineation.
In October 2007, the FDA mandated that a “black box” warning be placed on Definity (perflutren lipid microsphere [Bristol-Myers Squibb Medical Imaging, Billerica, Massachusetts]) and Optison (perflutren protein type A microsphere [GE Healthcare, Princeton, New Jersey]) after 11 deaths (4 within 30 min of administration) were temporally related to but not clearly caused by contrast injection. The warning stated that the use of these agents was contraindicated in patients with an unstable cardiopulmonary status, including those with acute coronary syndromes, acute myocardial infarction (MI), and worsening or unstable heart failure, as well as in patients with cardiac shunts, ventricular arrhythmias and prolonged QT interval corrected for heart rate, severe emphysema, and pulmonary emboli (8). As a result of this warning, many institutions stopped using these agents. Despite the decision by the FDA, the overall risk versus benefit profile of UCAs has not been clearly defined. A recent single-center experience described the background mortality of patients undergoing transthoracic contrast echocardiography and demonstrated that it is not significantly higher than that in patients undergoing noncontrast echocardiography during short-term follow-up (9). The FDA recently modified their original warning, but an examination of a multicenter experience regarding the safety of UCAs is needed. This study was designed to examine the short- and long-term safety profiles associated with the use of UCAs during stress echocardiography at 3 medical centers. The incremental value of UCA use in this setting was also assessed.
We retrospectively analyzed 42,408 consecutive patients from 1999 to 2007 at 3 academic medical centers (Saint Louis University, University of Nebraska, and Mayo Clinic Rochester–Minnesota) who underwent echocardiographic evaluation with contrast. Figure 1 summarizes the categorical comparisons performed in these 42,408 patients. A total of 23,659 patients had resting contrast echocardiography, with short-term follow-ups at 30 min and 24 h after contrast use. These patients were then compared with a matched group of 5,900 consecutive patients who underwent echocardiographic examination without contrast (Table 1).
A total of 18,749 patients underwent stress echocardiograms for suspected coronary artery disease (CAD), with 13,056 patients (70%) undergoing dobutamine stress echocardiography (DSE) and 5,693 (30%) undergoing treadmill exercise stress echocardiography (ESE). The indication for UCA use was either for suboptimal images or to assess myocardial perfusion. All patients received UCAs at rest and peak stress, with 5,947 (32%) receiving Optison in bolus doses of 0.2 to 0.3 ml and 12,802 (68%) receiving Definity as either bolus injections of 0.1 ml or as a 3% continuous infusion. Contrast-enhanced images from apical views (2-, 3-, and 4-chamber) were obtained and digitized at rest and maximal stress (≥85% predicted maximum heart rate) or a defined test end point. In a cohort of 6,513 patients who underwent DSE with contrast, safety outcomes were compared with those from a matched group of 6,249 patients who underwent DSE without contrast. Likewise, a subgroup of 4,275 patients who underwent ESE had similar outcomes compared with a matched group of 9,740 consecutive patients who did not receive contrast.
In 4,011 patients who received UCA for suboptimal imaging during DSE, we evaluated whether wall motion analysis using enhanced LV opacification with contrast agents led to improvements in sensitivity, specificity, and diagnostic accuracy by comparing results with those from 1,923 matched patients who had optimal images without UCAs during stress echocardiography. Quantitative angiography was used as the reference standard, using a >50% diameter stenosis in at least 1 major epicardial vessel as the definition of significant CAD.
Short-term follow-up consisted of examination of the patient within 30 min of contrast use and at 24 h. Long-term follow-up (2 to 92 months) was conducted in 6,075 patients. It was performed through review of the electronic medical record and data collection through telephone interviews. Nonfatal MI, as defined by a trend in cardiac specific enzymes according to European Heart/American College of Cardiology/American Heart Association guidelines, and death due to any cause were considered end points (10). If more than 1 event was noted during follow-up, the event occurring first was considered the end point.
In the 6,075 patients who underwent DSE with long-term follow-up, the added benefits of myocardial perfusion imaging (MPI) were analyzed by recording the number of adverse outcomes in patients who had abnormal myocardial perfusion (MP) and/or wall motion (WM) compared with patients with normal studies (although contrast use for MPI is not currently FDA approved).
Stress echocardiography protocols
Dobutamine and Exercise Stress Echocardiography Protocols
Dobutamine was infused intravenously at a starting dose of 5 μg/kg/min followed by increasing doses (10, 20, 30, and 40 μg/kg/min) up to a maximal dose of 50 μg/kg/min, in 3- to 5-min stages (11). Atropine (up to 2 mg) was injected in patients without symptoms or signs of ischemia to achieve the target heart rate, calculated as 220 minus age in years. The end points included achievement of target heart rate (85% of predicted maximal heart rate), maximal dobutamine/atropine doses, ST-segment elevation of ≥2 mm at an interval of 80 ms after the J point in non–Q-wave leads, sustained arrhythmias, severe chest pain, or intolerable adverse effects considered to be due to dobutamine or atropine (2,11). The tests were considered nondiagnostic if the patient failed to achieve the target heart rate without inducible ischemia.
For ESE, Bruce (90%) or modified Bruce (10%) protocols were used. Images were obtained at rest and immediately after peak stress. End points included achievement of target heart rate, horizontal or downsloping ST-segment depression of ≥1.4 mm at an interval of 80 ms after the J point in non–Q-wave leads, sustained arrhythmias, and symptom-limiting chest pain and/or shortness of breath.
Real Time Perfusion
Real time myocardial perfusion was performed using High Definition Imaging 5000 (Philips, Andover, Massachusetts), SONOS 5500 (Philips), iE 33 (Philips), Sequoia 512 (Siemens, Malvern, Pennsylvania), and Vivid 7 (GE Healthcare). All scanners were equipped with low-mechanical index real time pulse sequence schemes. During continuous infusions of microbubbles, intermittent high mechanical index impulses (4 to 40 frames) were administered to deplete capillary microbubbles and allow for the visual assessment of myocardial contrast replenishment (12). The equipment was adjusted to achieve a maximal nonlinear signal from contrast. Mechanical indexes were set to <0.3 and the frame rate to between 20 and 25 Hz. Time gain compensation and 2-dimensional gain settings were adjusted to suppress nonlinear signals from the myocardium before contrast injection and remained unchanged throughout the study.
Contrast-enhanced images from the apical window (2- to 4-chamber) were obtained and digitalized at rest, low dose dobutamine (if necessary), and maximal stress (85% age-predicted maximum heart rate), or a defined test end point. Each contrast bolus injection was followed by a slow 3- to 5-ml saline flush. After bolus injections, a minimum of 15 s of image acquisition was performed after appearance of peak myocardial opacification until the disappearance of contrast from the myocardium. For continuous infusions, ultrasound contrast was infused at a rate of 4 to 8 ml/min until perfusion images were obtained.
Baseline ejection fraction (EF) was visually estimated and/or calculated by single or biplane method of discs using contrast-enhanced 2-dimensional images. Contrast-enhanced WM abnormalities and, if done, MP assessments, were interpreted using a 17-segment model according to the joint recommendations of the American Society of Echocardiography (13). Segmental MP and WM in each of the 3 coronary artery territories were analyzed at rest and during peak stress and classified as normal, fixed, or inducible, as previously described (8,14,15). In this study, patients with either inducible or fixed WM abnormalities or perfusion defects were considered abnormal.
Resting echoes with contrast in the nondiagnostic group versus the noncontrast group with good quality images were compared and differences in characteristics were assessed by unpaired t test. Likewise, groups of patients undergoing DSE with contrast versus without contrast were compared for differences in characteristics and hemodynamic responses. The differences in these variables were assessed by unpaired t test. For continuous variables such as heart rate, Student t tests were used for univariate comparison between contrast and noncontrast DSE. Continuous variables were expressed as means and standard deviations. Categoric variables were expressed as proportions. Calculations of sensitivity and diagnostic accuracy of DSE for detection of coronary artery disease were performed with reference to angiograms as a gold standard.
Kaplan-Meier curves were used to estimate the distribution of time to death or nonfatal MI. The concordance of WM and MP analyses during stress real time-myocardial contrast echocardiography (RT-MCE) was calculated as percent agreement and kappa statistics. Differences between time-to-event curves were compared with log-rank testing. Clinical variables considered for univariate and multivariate predictors analyses were age older than 70 years, diabetes mellitus, hypercholesterolemia, hypertension, cigarette smoking, previous MI, earlier history of coronary artery bypass graft, and percutaneous coronary intervention. Analyzed echocardiographic parameters included EF <50% and WM and MP responses. Multivariate predictors of events were determined by Cox proportional hazards model. A value of p < 0.05 was considered significant. Interobserver variability was defined as the absolute difference between 2 blinded observations divided by the mean and expressed as a percentage.
There were no events within 30 min among 23,659 patients who received contrast for suboptimal resting studies and/or MPI. Three nonfatal MIs and 1 death occurred within 24 h. In the matched group of 5,900 consecutive patients with optimal resting studies and no contrast use, no events occurred within 30 min, but 7 nonfatal MIs and 1 death occurred within 24 h (p = NS compared with event rates in groups receiving contrast).
Among 18,749 patients who underwent stress echocardiograms, no deaths or MIs were observed within 30 min; 1 death and 5 nonfatal MIs were observed within 24 h. Among the 5 nonfatal MIs, 3 occurred after use of Definity and 2 after use of Optison. In the first case, a DSE with RT-MCE demonstrated inducible inferolateral ischemia. Catheterization demonstrated >50% stenosis in the left circumflex artery and right posterior descending artery, which were revascularized. Sixteen hours after intervention, the patient had significantly elevated cardiac enzymes. In the second case, a patient exhibited inducible ischemia during a DSE with RT-MCE, but proceeded to cadaveric renal transplant surgery the same day. Post-operative cardiac enzymes demonstrated a non–Q-wave infarction. In the third case, the patient had inducible ischemia during a DSE with RT-MCE and elevated cardiac enzymes 7 h after DSE. The patient subsequently underwent percutaneous coronary intervention of a 90% stenosis in an obtuse marginal vessel. In the fourth case, a patient had ventricular tachycardia 9 min into recovery after DSE, had inducible septal ischemia, and underwent percutaneous transluminal coronary angioplasty to a 100% occluded distal left anterior descending. In the fifth case, the patient had no ischemia or electrocardiogram changes during stress testing, but had persistent chest pain and an elevated troponin level. Coronary angiography demonstrated a 60% obtuse marginal and 50% posterior descending artery lesion.
One death within 24 h was noted in the stress echocardiography subgroup. This patient was transferred from a local hospital due to dyspnea and required continuous vasopressors for hypotension. He developed frequent runs of ventricular tachycardia requiring intravenous antiarrhythmic therapy. The patient eventually developed disseminated intravascular coagulation. In an attempt to determine whether there was myocardial viability, the patient underwent DSE with RT-MCE using intravenous Optison on hospital day 11. The procedure was performed without complication, but within the next 24 h, the patient developed recurring ventricular tachycardia and died 22 h after contrast administration. In all of these cases, there was no reason to believe contrast use contributed to the events.
At 30 days, in a subgroup of 10,788 patients who received contrast for DSE or ESE, 37 (0.34%) patients died and 68 (0.63%) patients had an MI; this compared with a temporally matched cohort of 15,989 patients undergoing stress echocardiography and not receiving contrast agents, in which 62 (0.39%) patients died and 73 (0.46%) had an MI. There were no statistically significant differences between the 2 groups with respect to nonfatal MI or death.
CAD detection with contrast-enhanced LV opacification in patients with suboptimal images
In a subgroup of 4,011 patients who received UCA for suboptimal images, we evaluated whether WM analysis using enhanced LV opacification led to improvements in sensitivity, specificity, and diagnostic accuracy by comparing these values with those from 1,923 patients who had optimal images during stress echocardiography and did not require UCAs. There were no differences between the 2 groups with respect to age, sex, history of MI, or prevalence of resting WM abnormalities (Table 2). There were no significant differences between the groups with respect to hemodynamics and heart rate responses to dobutamine (Table 3). Likewise, there were no statistically significant differences between the group that received contrast and the noncontrast group with respect to the number of affected vessels, stenosis severity, or stenosis distribution (Table 4).
Sensitivity of DSE was slightly higher in patients who received contrast compared with those who did not. However, this did not reach a statistically significant level (81% vs. 73%, p = NS). The diagnostic accuracy of DSE was slightly improved with contrast in the suboptimal image group versus the noncontrast group. However, no statistically significant difference was found (82% vs. 77%, p = NS) (Table 5).
Among 6,075 patients with long-term follow-up, the mean baseline LV EF was 56 + 10%. There were 352 patients (5.8%) with baseline EF <50%. Contrast agent was administered in a continuous infusion in 1,944 (32%) patients and via bolus injections in 4,131 (68%) patients. Median duration of follow-up was 25 months (range 2 to 92 months, standard deviation 22). A total of 510 (8.8%) patients had events during follow-up. Events occurred at a median of 21 months (range 1 day to 88 months). Nonfatal MIs occurred in 237 patients (3.9%), and death occurred in 273 patients (4.4%).
DSE was interpreted as normal in 4,251 (69.9%) and abnormal in 1,824 (30%) patients. Both WM and MP were abnormal in 1,458 (24%) patients. All patients with abnormal WM also had abnormal MP. There were 668 patients (11%) with normal WM but abnormal MP. Of the 510 events that occurred during long-term follow-up, 133 (3.7%) events occurred in patients with normal WM and MP responses during RT-MCE, 291 (16%) events occurred in patients with abnormal WM and MP, and 86 (13%) events were noted in patients with normal WM but abnormal MP.
Univariate and multivariate predictors of follow-up events are depicted in Table 6. By univariate analysis, age >70 years, previous coronary artery bypass grafting, previous percutaneous coronary intervention, previous MI, resting EF <50%, abnormal WM, and abnormal MP were predictors. By multivariate analysis, the only independent predictors were EF <50% (odds ratio [OR]: 1.4; 95% confidence interval [CI]: 1.0 to 1.8), previous MI (OR: 0.6; 95% CI: 0.4 to 0.9), and abnormal MP (OR: 2.4; 95% CI: 1.0 to 5.9). After adjustment for MP, the analysis of WM no longer had predictive value (Table 6).
By sequential Cox regression, an EF <50% increased the risk for adverse outcomes at long-term follow-up over other clinical and risk factors (chi-square: 11.4; p < 0.05); WM analysis increased this further over clinical variables and EF <50% (chi-square: 21.8; p < 0.01). Abnormal MP increased the predictive value over EF <50% and WM (chi-square: 39.4; p < 0.001), adding significant incremental value in predicting outcomes. Event-free survival in patients with normal WM and MP was significantly better than in those with normal WM and abnormal MP and abnormal WM and MP responses (chi-square: 19.686, p = 0.0001).
The average interobserver agreement for the 3 institutions was 80% for MP (λ = 0.63) and 88% for WM (λ = 0.64) analysis.
The recent FDA mandated “black box” warning reported a total of 11 deaths related to contrast administration: 10 following Definity and 1 following Optison (8). Although 4 occurred during or within 30 min of Definity use, mitigating factors could have explained the deaths.
Before the black box warning, there were multiple studies that demonstrated the safety of these contrast agents (16,17). Recently presented data suggested that contrast use in echocardiography is safe, and the FDA modified the professional labeling for Optison and Definity contrast agents originally announced in November 2007. The warning currently states that it is contraindicated to administer contrast to patients with R→L, bidirectional, or transient R→L cardiac shunts and to patients with hypersensitivity to perflutren. This study was designed to evaluate the safety profile and incremental benefit of UCA use during stress echocardiography in a multicenter setting. As has been the case, serious adverse events (death or nonfatal MI) within 24 h of contrast use could be easily explained by the patient's underlying condition at the time of the study or to induced ischemia during stress echocardiography. This was confirmed by the similar incidence of adverse events in the patients who did not receive contrast during their stress echocardiography.
It is important to note the significant incremental value of contrast that was observed at the 3 participating centers. Enhanced LV opacification (an approved indication) in patients with suboptimal windows resulted in a sensitivity and specificity in detecting coronary artery disease that were equivalent to those in whom image quality was optimal. The ability of stress echocardiography to detect coronary artery disease is clearly related to the ability to visualize all myocardial segments in each coronary artery territory (18). Suboptimal windows when contrast is not used usually results in the study being cancelled and an alternative noninvasive or invasive study being performed. These alternative tests often involve the use of ionizing radiation, with the patient incurring additional risks and costs. In this study, we demonstrated that ultrasound contrast in patients with suboptimal windows has the ability to improve test accuracy to the level seen only in patients with optimal endocardial border delineation.
Myocardial perfusion imaging (not yet approved by the FDA) added significant incremental value to DSE, identifying patients at risk for death or nonfatal MI. Although the FDA has not approved the use of ultrasound contrast to assess MP, most ultrasound scanners now offer real time perfusion pulse sequence schemes that allow one to examine myocardial contrast enhancement during stress echocardiography. This confirms single center findings regarding prognostic value (19), and emphasizes the potential for perfusion imaging in this setting. MPI has become especially useful not only in detecting stress-induced ischemia, but also in detecting acute coronary syndromes in the emergency department (20) and predicting patient outcomes after MI (21).
The number of patients analyzed in this study is a great strength. We have included both short- and long-term follow-up as part of our safety analysis. Furthermore, we have provided evidence of the obvious advantages of contrast agent use. We do recognize, however, that this is a retrospective analysis with intrinsic limitations. In addition, the matched cohorts of patients who did not receive contrast were not the same size as the group receiving contrast.
In the current study, patients with suboptimal images in whom contrast was used were compared with patients with good echocardiographic windows in whom no contrast was used. However, technically difficult patients were not studied with and without contrast. This approach was selected so that we could compare 2 groups in which 17 segments were adequately visualized, either with or without contrast.
This study examined, in a large multicenter experience, the safety and incremental value of contrast use in the clinical practice of stress echocardiography. The risks of both short-term and long-term events, defined as nonfatal MI and death, after contrast administration are very low and are no different than in patients not receiving contrast during stress echocardiography. Contrast use in patients with suboptimal images improves feasibility and accuracy of stress echocardiography testing. Failure to use contrast agents in patients with suboptimal images may result in a misdiagnosis and/or additional alternative imaging techniques with greater inherent risks. MPI was shown in this study to provide significant incremental value to DSE in predicting patient outcomes. Based on this study and recent comparisons of resting ultrasound contrast use in hospitalized patients (9), which results in a combined experience of more than 60,000 patients, there is no evidence to suggest a causal relationship between the use of commercially available contrast use and serious adverse events. The findings in this study provide physicians with the confidence to use contrast agents during stress testing when clinically indicated.
Dr. Labovitz was recently appointed as a blinded reader for Bristol-Myers Squibb.
- Abbreviations and Acronyms
- coronary artery disease
- dobutamine stress echocardiography
- ejection fraction
- exercise stress electrocardiography
- Food and Drug Administration
- left ventricular
- myocardial infarction
- myocardial perfusion
- myocardial perfusion imaging
- real time-myocardial contrast echocardiography
- ultrasound contrast agent
- wall motion
- Received May 16, 2008.
- Revision received August 14, 2008.
- Accepted August 25, 2008.
- American College of Cardiology Foundation
- Geleijnse M.L.,
- Fioretti P.M.,
- Roelandt J.R.
- Skyba D.M.,
- Camarano G.,
- Goodman N.C.,
- Price R.J.,
- Skalak T.C.,
- Kaul S.
- Cohen J.L.,
- Cherif J.,
- Segar D.S.,
- et al.
- ↵U.S. FDA prescribing information for Definity approved October 10, 2007. http://www.fda.gov/cder/foi/label/2007/021064s007lbl.pdf. Accessed November 15, 2007.
- Kusnetzky L.L.,
- Khalid A.,
- Khumri T.M.,
- Moe T.G.,
- Jones P.G.,
- Main M.L.
- Fox K.A.,
- Birkhead J.,
- Wilcox R.,
- Knight C.,
- Barth J.,
- British Cardiac Society Working Group
- Armstrong W.F.,
- Pellikka P.A.,
- Ryan T.,
- Crouse L.,
- Zoghbi W.A.
- Cerqueira M.D.,
- Weissman N.J.,
- Dilsizian V.,
- et al.
- Wei K.,
- Jayaweera A.R.,
- Firoozan S.,
- Linka A.,
- Skyba D.M.,
- Kaul S.
- Dolan M.S.,
- Riad K.,
- El-Shafei A.,
- Puri S.,
- et al.
- Tsutsui J.M.,
- Elhendy A.,
- Anderson J.R.,
- Xie F.,
- McGrain A.C.,
- Porter T.R.
- Tong K.L.,
- Kaul S.,
- Wang X.Q.,
- et al.
- Dwivedi G.,
- Janardhanan R.,
- Hayat S.A.,
- Swinburn J.M.,
- Senior R.