Author + information
- Received January 9, 2009
- Revision received March 23, 2009
- Accepted March 28, 2009
- Published online September 1, 2009.
- Sahar S. Abdelmoneim, MD, MSc⁎,‡,
- Mathieu Bernier, MD⁎,
- Christopher G. Scott, MS†,
- Abhijeet Dhoble, MD, MPH⁎,
- Sue Ann C. Ness, RN⁎,
- Mary E. Hagen, RDCS⁎,
- Stuart Moir, MD⁎,
- Robert B. McCully, MD⁎,
- Patricia A. Pellikka, MD⁎ and
- Sharon L. Mulvagh, MD⁎,⁎ ()
Reprint requests and correspondence:
Dr. Sharon L. Mulvagh, Division of Cardiovascular Diseases, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905
Objectives We evaluated the short- and long-term safety of contrast agents during stress echocardiography (SE).
Background Concerns about contrast agent safety led to revised recommendations for product use in the U.S.
Methods We studied 26,774 patients who underwent SE between November 1, 2003, and December 31, 2007. The 10,792 patients who comprised the contrast cohort received second-generation perfluorocarbon-based agents for left ventricular opacification during SE. The noncontrast cohort comprised 15,982 patients who had their first SE in the same period but without contrast agents. Short-term (≤72 h and ≤30 days) and long-term (up to 4.5 years) end points were death and myocardial infarction (MI). Cox regression models were used. Immediate contrast agent-related adverse effects were also reported.
Results The contrast cohort had older patients (mean [SD] age, 65.8 [12.1] years vs. 62.6 [14.1] years; p < 0.001), a higher percentage of males (57.4% vs. 52.8%, p < 0.001), and higher-risk patients compared with the noncontrast cohort. In addition, dobutamine SE patients had greater cardiac risk than exercise SE patients. Abnormal SE findings in patients who received contrast agents were more frequent (32.4% vs. 27.9%, p < 0.001). The 2 cohorts had no statistical difference in the incidence of short-term events (death and MI). Within 72 h, 1 patient in the contrast cohort and 2 patients in the noncontrast cohort died (p = 0.54); 3 in the contrast cohort and 7 in the noncontrast cohort had MI (p = 0.92). Within 30 days, 37 patients (0.34%) in the contrast cohort and 57 patients (0.36%) in the noncontrast cohort died (p = 0.85); 17 patients (0.16%) in the contrast cohort and 16 patients (0.10%) in the noncontrast cohort had MI (p = 0.19). Adjusted hazard ratios were not different between cohorts for death (0.99; 95% confidence interval: 0.88 to 1.11) or MI (0.99; 95% confidence interval: 0.80 to 1.22).
Conclusions The use of contrast agents during SE was not associated with an increased short-term or long-term risk of death or MI.
Opacification of the left ventricular cavity enhances endocardial border detection and improves the assessment of ventricular chamber dimensions and regional systolic function (1). A growing body of international literature has demonstrated the safety of contrast agents (2,3). Recent concerns about the safety of ultrasonographic contrast agents led to changes in the recommendations for product use in the U.S. The U.S. Food and Drug Administration (FDA) emphasized the risks of serious cardiopulmonary reactions and required that warnings be added to the labeling of ultrasonographic contrast agents. This action followed a post-marketing report of 4 deaths within 30 min of contrast agent administration, which were not clearly attributable to contrast agent injection. Subsequently, on May 12, 2008, revised labeling changes, including a relaxation of previous warnings, were implemented by the FDA (4).
In the present study, we evaluated the safety of contrast agents used during stress echocardiography (SE) with exercise or dobutamine in a high-volume echocardiography laboratory by comparing 2 cohorts: 1 with patients who underwent contrast SE and a reference cohort of patients who underwent SE without contrast agents. The end points were death and myocardial infarction (MI). We hypothesized that the use of contrast agents would not result in additional harm to patients undergoing stress testing.
This retrospective cohort study was approved by the Mayo Clinic Institutional Review Board. Only patients who approved the use of their medical records for research were included in the cohorts.
Identification of study cohorts
Using the Mayo Clinic Stress Echocardiography Database, we identified all patients who underwent their first SE between November 1, 2003, and December 31, 2007. Patients who received contrast agents during SE comprised the contrast cohort. The remaining patients who had SE in the same period but who did not receive contrast agents comprised the reference cohort (the noncontrast cohort). A total of 3.6% of patients did not provide consent to use their medical records and thus were excluded from study cohorts.
The 2 contrast agents used were second-generation perfluorocarbon-based agents: Optison (GE Healthcare, Princeton, New Jersey) and Definity (Lantheus Medical Imaging, North Billerica, Massachusetts). Contrast agents were administered for left ventricular opacification according to the American Society of Echocardiography Task Force on Standards and Guidelines for the Use of Ultrasonic Contrast in Echocardiography (5). The contrast agents were administered in accordance with the package inserts.
Patients in both groups had SE with dobutamine or exercise. Dobutamine SE was performed according to a standard protocol as previously described (6). For exercise SE, the symptom-limited Bruce protocol was performed according to a standard protocol (7). Commercially available ultrasound systems were used: SONOS 7500 or iE33 (Philips Medical Systems, Andover, Massachusetts), ACUSON Sequoia 512 (Siemens Medical Solutions, Mountain View, California), and Vivid 7 (GE Healthcare, Princeton, New Jersey). All studies were recorded on videotape and digitized for storage. Regional wall motion abnormalities were interpreted with the use of videotaped and digitized images according to the standard 16-segment model. The 12-lead electrocardiography and blood pressure were monitored continuously.
Data collection and outcome measurements
Patient demographic and SE data were prospectively recorded. The primary outcome of this study was the short-term safety of the contrast agents used. End points were death (due to any cause or to a cardiac event) and MI within 72 h and 30 days after SE. The secondary outcome was the long-term safety of contrast agent use and the time to the end point (death or MI) occurring up to 4.5 years. Arrhythmias, including supraventricular tachycardia, atrial fibrillation, and ventricular tachycardia (VT), were also reported. Sustained VT was defined as a regular wide QRS complex and a tachycardia of more than 100 beats/min that lasted for 30 s or more or caused hemodynamic collapse and that required medical or electrical therapy. Nonsustained VT was defined as VT for 3 or more beats but with a duration of less than 30 s. In addition, we prospectively documented adverse effects related to the use of contrast agents in 2007.
Deaths were identified by first searching the SE Database and the Mayo Clinic Death Database for the same patients. Then, a thorough chart review was conducted to confirm the death and the date of occurrence in relation to contrast agent administration. Similar measures were taken to identify MI. We used the International Classification of Diseases-Ninth Revision (ICD-9) code. Chart reviews were conducted of all patients reported to have MI after SE.
For categorical variables, numbers of patients and proportions of the cohorts are presented; differences between cohorts were tested by using the Pearson chi-square test or Fisher exact test, as appropriate. For continuous variables, mean and standard deviation are presented; differences between groups were tested by using the 2-sample t test. Survival was estimated using the Kaplan-Meier method. Multivariate Cox proportional hazards regression models were used to test for differences in long-term outcomes after adjusting for known confounders. We reported the annualized rates of combined events in both cohorts. In addition, comparison of long-term outcomes between patients receiving contrast and those not receiving contrast was performed by propensity matching analysis. A logistic regression model was fit including all relevant variables and propensity for receiving contrast as calculated from this model. Each patient receiving contrast was matched on propensity score to patients not receiving contrast. To ensure that a good match was made, the maximal allowable difference in propensity score was 0.25 standard deviations. In all analyses, significance was set at a 2-sided p value of ≤0.05. All analyses were performed with SAS version 9.1 software (SAS Institute, Cary, North Carolina).
During the study period, 29,759 patients underwent SE; for 26,774 of these patients, the SE was their first at Mayo Clinic, Rochester, Minnesota (Fig. 1). The contrast cohort included 10,792 patients (4,276 had exercise SE and 6,516 had dobutamine SE; 86% received Definity contrast agent, and 14% received Optison contrast agent). The noncontrast cohort included 15,982 patients (9,738 had exercise SE and 6,244 had dobutamine SE) who had their first SE in the same period but did not receive contrast agents. Patients in the contrast cohort were at greater risk of coronary artery disease than those in the noncontrast cohort, and more patients in the dobutamine SE group had greater cardiac risk compared with patients who underwent exercise SE (Tables 1 and 2).⇓ Compared with the noncontrast cohort, the contrast cohort included more patients with abnormal SE for regional wall motion abnormalities (32.4% vs. 27.9%; p < 0.001).
Short-term safety of contrast SE
Even though it included higher-risk patients, the contrast cohort had percentages of deaths and MIs that were comparable to those of the noncontrast cohort (Table 3). Within 72 h after SE, 1 patient in the contrast cohort and 2 patients in the noncontrast cohort died (p = 0.54). The patient in the contrast cohort who died was a 76-year-old woman with diabetes mellitus who underwent dobutamine SE in a preoperative evaluation for repair of a mycotic aortic aneurysm. SE results were normal, with no reported adverse effects from contrast agents. The patient died of cardiac and multiorgan failures on postoperative day 2 (48 h after SE). Within 72 h after SE, 3 patients in the contrast cohort and 7 in the noncontrast cohort had MI (p = 0.92). All 3 patients in the contrast cohort had dobutamine SE (1 received Definity; 2 received Optison).
Within 30 days after SE, 37 patients (0.34%) in the contrast cohort died and 57 patients (0.36%) in the noncontrast cohort died (p = 0.85). Of the deaths due to cardiac causes, 10 were in the contrast cohort and 11 were in the noncontrast cohort (p = 0.23). Within 30 days after SE, MI was reported in 17 patients (0.16%) in the contrast cohort and in 16 patients (0.10%) in the noncontrast cohort (p = 0.19).
Arrhythmias and contrast SE
The overall incidence of arrhythmias during SE was low in both cohorts. Supraventricular tachycardias and atrial fibrillation occurred in 30 patients (0.28%) in the contrast cohort and in 34 patients (0.21%) in the noncontrast cohort (p = 0.31). Ventricular tachycardia occurred in 11 patients (0.10%) in the contrast cohort and in 7 patients (0.04%) in the noncontrast cohort (p = 0.09). In the contrast cohort, 2 patients had sustained VT, 2 had ventricular fibrillation, and 7 had nonsustained VT. In the noncontrast cohort, 2 patients had sustained VT and 5 had nonsustained VT.
There was no significant difference between the proportion of patients in each cohort who had ventricular arrhythmias: with dobutamine SE, 10 (0.15%) in the contrast cohort and 6 (0.09%) in the noncontrast cohort (p = 0.46); and with exercise SE, 1 (0.02%) in the contrast cohort and 1 (0.01%) in the noncontrast cohort (p = 0.52).
Long-term safety of contrast SE
Of the 10,792 patients in the contrast cohort, 841 patients had an event during follow-up (734 who had dobutamine SE and 107 who had exercise SE): 658 patients died (590 who had dobutamine SE and 68 who had exercise SE) and 183 had MI (144 who had dobutamine SE and 39 who had exercise SE). Of the 15,982 patients in the noncontrast cohort, 1,247 patients had an event during follow-up (988 who had dobutamine SE and 259 who had exercise SE): 978 died (803 who had dobutamine SE and 175 who had exercise SE), and 269 had MI (185 who had dobutamine SE and 84 who had exercise SE). For patients who had dobutamine SE, the annualized event rate for the combined events of death and MI in the contrast cohort was 9.6% (95% confidence interval [CI]: 8.9% to 10.3%) compared with 10.9% in the noncontrast cohort (95% CI: 10.2% to 11.6%). For patients who had exercise SE, the annualized event rate for combined events of death and MI in the contrast cohort was 2.2% (95% CI: 1.8% to 2.6%) compared with 1.9% (95% CI: 1.6% to 2.1%) in the noncontrast cohort.
Figure 2 presents Kaplan-Meier curves for survival free from death or MI for the dobutamine and exercise SE subgroups. After adjustment for known confounders, no differences in the hazard ratios (HRs) for the events were observed between the contrast and noncontrast cohorts (Table 4). As a secondary analysis; 7,654 contrast patients (71%) were propensity matched with patients not receiving contrast. Analysis in this matched population showed no evidence of an increased risk of death (HR: 0.99, 95% CI: 0.87 to 1.12, p = 0.89) or MI (HR: 0.97, 95% CI: 0.76 to 1.23, p = 0.79) for patients receiving contrast.
Adverse effects related to contrast agents
Contrast agent-related complications in SE studies were consistently documented electronically in the database starting in January 2007. Among the 3,071 patients who underwent contrast SE in 2007 (1,819 had dobutamine SE and 1,252 had exercise SE), contrast agents were suspected to be the cause of adverse reactions in 68 (2.2%), and a definite relation with contrast agent infusion was established in 41 of the 3,071 patients (1.3%). Backache alone was reported in 19 of 3,071 patients (0.62%), headache alone in 16 (0.52%), and both in 4 patients (0.13%). Transient wheezing in the upper airway without urticaria was reported in 1 patient (0.03%) undergoing dobutamine SE. Urticaria or hives (on limbs and thorax) and swelling in the mouth were reported in 1 patient (0.03%) undergoing exercise SE and were resolved within 15 min with intravenous administration of diphenhydramine and subcutaneous administration of 0.5 ml of epinephrine (1:1,000). Other side effects included abdominal pain in 1 patient (0.03%) and throat tingling in 1 patient (0.03%). The adverse reactions thought to be caused by concomitant infusion of contrast agent and dobutamine were chest pain in 24 patients (0.78%), nausea and vomiting in 3 patients (0.10%), and shivering and tremors in 5 patients (0.16%).
To our knowledge, our cohort study is the largest to address the safety of contrast agents used during SE. During short- and long-term follow-up, patients who had undergone SE with contrast agents were not at increased risk of death or MI compared with those who had not received contrast agents. This was evident even though patients who required contrast agents had greater cardiovascular risks. The large number of patients included in the present study allowed adjustments for known risk factors that might confound the association between the use of a contrast agent as a risk factor and the occurrence of cardiovascular events. Finally, contrast agent-related side effects were minor and infrequent. Recently, concerns about the safety of ultrasonographic contrast agents led to changes in the recommendations for product use in clinical practice (4).
In the medical literature, the 2 main concerns about the use of intravenously administered contrast agents in echocardiography are associated mostly with imaging at a high mechanical index. One study reported the possibility of an increased frequency of premature ventricular contractions, which was hypothesized to be related to the nonthermal bioeffects of the ultrasound field (8). However, a subsequent study did not detect any increase in the frequency of premature ventricular contractions during or after imaging at a high mechanical index (9). The other main concern is the possibility of vascular injury and cardiac injury. This concern was addressed in a recent study that reported on the subclinical release of cardiac biomarkers in humans after imaging at a high mechanical index (10). However, another study reported that the use of contrast agent echocardiography is not associated with risk of myocardial damage (11). In a large study, Tsutsui et al. (12) compared 1,486 patients who underwent dobutamine SE with contrast agents at a low mechanical index and 1,012 patients who underwent conventional dobutamine SE without contrast agents; no difference existed between the 2 groups in the incidence of nonsustained VT, sustained VT, or supraventricular tachycardia. In our study, the occurrence of arrhythmias in the contrast cohort was comparable to the occurrence in the noncontrast cohort; this similarity can be explained by the use of imaging at a low mechanical index in most of the patients.
Contrast agent-related adverse effects in the present study were generally minor. Some of these side effects can be attributed to contrast agent administration, as evidenced by their reversal after decreasing the dose or discontinuing the administration of the contrast agent. Other side effects, including chest pain, arrhythmias, and shortness of breath, can be linked to the sympathomimetic effect of pharmacologic stress. Nevertheless, the safety of stress testing has been demonstrated in several situations, with the overall incidence of side effects being small (13). In addition, the incidence of acute anaphylactoid reaction in our study was 0.03% (in approximately 3 of 10,000 patients), which is in agreement with the results from previous studies (14,15). In 2008, there was considerable debate about the safety of contrast agents in response to the FDA black box warnings. Both single center and large multicenter retrospective analyses have subsequently demonstrated a good safety profile and risk/benefit ratio for ultrasound contrast agents (14–21) (Table 5).
The main limitation of the present study is that data analysis was retrospective; however, data collection and data entry in the SE Database are ongoing processes that are done prospectively. MI and death were selected as the main end points of the study because they can be identified objectively and are easily retrievable from the medical records with high accuracy and precision. Another limitation is the dependence on the ICD-9 codes for the retrieval of the outcome; however, use of these codes allowed an objective and unbiased method of determining the outcome. Contrast agent-related adverse effects were directly entered into the Stress Echocardiography Database as of January 2007. We used these data to avoid underestimating side effects in previous years, which lacked standardized documentation.
The present study has many strengths. The large sample size, the cohort design, and the enormous amount of information about the risk factors and known confounders allowed thorough evaluation of the association between contrast agent use and the events of interest by using multivariate Cox proportional hazards regression analysis.
We believe that the results of the present study are readily generalizable. The data are from a real-life, high-volume echocardiography laboratory for primary care and referral practices. However, the results might not be valid for other populations because of variations in clinical practice, clinical settings, or patient characteristics. Thus, further validation studies are needed.
In summary, the present study showed that the use of contrast agents during SE is not associated with increased short-term or long-term risk of death or MI. The study also provided further evidence of the safety of contrast agent use in echocardiography in general and during SE in particular.
For a slide show presentation of this study, please see the online version of this article.
Dr. Mulvagh has received research grants from Lantheus Medical Imaging, GE Healthcare, and Astellas Pharma Inc.
- Abbreviations and Acronyms
- confidence interval
- U.S. Food and Drug Administration
- hazard ratio
- International Classification of Diseases-Ninth Revision
- myocardial infarction
- stress echocardiography
- ventricular tachycardia
- Received January 9, 2009.
- Revision received March 23, 2009.
- Accepted March 28, 2009.
- American College of Cardiology Foundation
- Cohen J.L.,
- Cheirif J.,
- Segar D.S.,
- et al.
- ↵Online U.S. Food and Drug Administration Postmarket Drug Safety Information for patients and Providers. http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm110260.htm. Accessed November 6, 2008.
- Vancraeynest D.,
- Kefer J.,
- Hanet C.,
- et al.
- Tsutsui J.M.,
- Elhendy A.,
- Xie F.,
- O'Leary E.L.,
- McGrain A.C.,
- Porter T.R.
- Wei K.,
- Mulvagh S.L.,
- Carson L.,
- et al.
- Kusnetzky L.L.,
- Khalid A.,
- Khumri T.M.,
- Moe T.G.,
- Jones P.G.,
- Main M.L.
- Main M.L.,
- Ryan A.C.,
- Davis T.E.,
- Albano M.P.,
- Kusnetzky L.L.,
- Hibberd M.
- Dolan M.S.,
- Gala S.,
- Dodla S.,
- et al.
- Aggeli C.,
- Giannopoulos G.,
- Roussakis G.,
- et al.