Author + information
- Received April 25, 2011
- Revision received June 27, 2011
- Accepted June 29, 2011
- Published online November 1, 2011.
- Oliver Bruder, MD⁎,
- Steffen Schneider, PhD†,
- Detlev Nothnagel, MD‡,
- Günter Pilz, MD§,
- Massimo Lombardi, MD∥,
- Anil Sinha, MD¶,
- Anja Wagner, MD#,
- Thorsten Dill, MD⁎⁎,
- Herbert Frank, MD††,
- Albert van Rossum, MD‡‡,
- Jürg Schwitter, MD§§,
- Eike Nagel, MD∥∥,
- Jochen Senges, MD†,
- Georg Sabin, MD⁎,
- Udo Sechtem, MD¶¶ and
- Heiko Mahrholdt, MD¶¶,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Heiko Mahrholdt, Department of Cardiology, Robert Bosch Medical Centre, Auerbachstrasse 110, 70376 Stuttgart, Germany
Objectives The purpose of this study was to assess the frequency, manifestation, and severity of acute adverse reactions associated with administration of several gadolinium-based contrast agents to patients who underwent cardiac magnetic resonance (CMR) imaging in the EuroCMR (European Cardiovascular Magnetic Resonance) Registry multinational and multicenter setting.
Background The frequency, manifestations, and severity of acute adverse reactions associated with gadolinium-based contrast agents in the specific setting of cardiovascular magnetic resonance have not been systematically evaluated yet.
Methods This is a multicenter and multinational registry with consecutive enrollment of patients in 45 European centers. During the current observation, 17,767 doses of gadolinium-based contrast agent were administered to 17,767 patients. The mean dose was 25.6 ml (range 5 to 80 ml), which is equivalent to 0.128 mmol/kg (range 0.012 to 0.3 mmol/kg).
Results Thirty acute adverse reactions due to contrast administration occurred (0.17%). All reactions were classified as mild according to the American College of Radiology definition. The most frequent complaints following contrast administration were rashes and hives (9 of 30), followed by nausea (7 of 30), and anxiety (6 of 30). The event rate ranged from 0.06% (linear nonionic agent gadodiamide) to 0.47% (linear ionic agent gadobenate dimeglumine). Interestingly, we also found different event rates between the three main indications for CMR imaging, ranging from 0.1% (risk stratification in suspected coronary artery disease) to 0.42% (viability in known coronary artery disease).
Conclusions The incidence of acute adverse reactions after administration of gadolinium-based contrast in the “off-label” setting of CMR in our population was not different from the incidence in the U.S. Food and Drug Administration–approved general radiology setting. Thus, the off-label use of gadolinium-based contrast in CMR should be regarded as safe concerning the frequency, manifestation, and severity of acute events.
Cardiac magnetic resonance (CMR) is a rapidly emerging noninvasive imaging technique providing high-resolution images of the heart in any desired plane without application of radiation (1). CMR consists of several protocols that can be performed separately, or in various combinations during a single-patient examination. For several of those CMR protocols, such as myocardial perfusion imaging, MR angiography, or noninvasive tissue characterization, gadolinium-based contrast agents are necessary.
Gadolinium-based contrast agents are considered generally safe in radiology practice. The incidence of adverse reactions is relatively low compared with that of iodinated contrast agents used for computed tomography (2–7). Nevertheless, the frequency, manifestation, and severity of acute adverse reactions associated with gadolinium-based contrast agents in the specific setting of CMR have not been systematically evaluated yet. Thus, the use of gadolinium-based contrast agents for CMR imaging is still considered as an “off-label” use in the United States and many European countries.
Consequently, the purpose of this study was to assess the frequency, manifestation, and severity of acute adverse reactions associated with administration of several gadolinium-based contrast agents to patients who underwent CMR imaging in the EuroCMR (European Cardiovascular Magnetic Resonance) Registry multinational and multicenter setting.
Study population and data management
The basis of the current paper is the EuroCMR Registry including 11,040 consecutive patients of the German pilot phase (April 2007 to January 2009) and 8,677 patients of the ongoing European registry (March 2009 to March 2011). All data were prospectively collected by trained personnel with the use of online case record forms provided by the Institut für Herzinfarktforschung Ludwigshafen, University of Heidelberg, Germany via a SSL-secured Internet connection, and stored on a central server (8,9). Each participating center (Online Appendix) appointed a senior cardiologist or radiologist as a local investigator responsible for the data quality of each individual patient entered in the registry. A plausibility check was directly carried out after submitting the data to avoid further queries as far as possible. Data collection and management were approved by the ethics committee.
All 17,767 patients (90% of 19,717) receiving contrast media were included in the analysis. All patients received the contrast media in compliance with the current American College of Cardiology Foundation/American College of Radiology/Society of Cardiovascular Computed Tomography/Society of Cardiovascular Magnetic Resonance/American Society of Nuclear Cardiology/North American Society of Cardiovascular Imaging/Society for Cardiac Angiography and Interventions/Society of Interventional Radiology appropriateness criteria for CMR imaging (8–10). The completeness of the analysis dataset was higher than 98%. The patient charts of all acute adverse reactions due to contrast media were reviewed centrally and categorized by the project management.
Variables and definitions
All variables assessed were pre-defined and were collected directly from patients and/or from medical records. Variables include demographic data, history, and indications for CMR; procedural parameters; as well as complications (8,9). Complications caused by acute adverse reactions to contrast media (= onset within 60 min after administration) are defined according to the American College of Radiology (11) criteria and can be viewed in Table 1.
Statistical analysis was performed by the Biometrics Department of the Institut für Herzinfarktforschung, Ludwigshafen, Germany, which is chaired by senior statistician Steffen Schneider, PhD. Absolute numbers and percentages were computed to describe the distribution of the different contrast media and the rate of adverse reactions in the patient population. Medians (with ranges) and means (with standard deviation) were calculated to describe the characteristics of contrast media and to characterize the patients with adverse reactions. Categorical parameters were compared by using Fisher exact test and continuous variables by using Wilcoxon rank sum test. However, the low rates of adverse events preclude most significance tests and the absence of pre-specified statistical hypothesis in the registry allows descriptive comparisons in most cases only. All analyses were performed using the SAS statistical package (version 9.2, SAS Institute, Cary, North Carolina).
Acute adverse reactions by severity
During the current observation, 17,767 doses of gadolinium-based contrast agent were administered to 17,767 patients undergoing CMR imaging. The mean dose was 25.6 ml (range 5 to 80 ml), which is equivalent to 0.128 mmol/kg (range 0.012 to 0.3 mmol/kg). During and immediately after the CMR procedure, 30 acute adverse reactions due to contrast administration occurred (0.17%). No relation between acute reactions and the dose administered was detected using Wilcoxon rank sum test (p = 0.07). All reactions were classified as mild (Table 1). There were no moderate or severe reactions or death due to contrast administration. Furthermore, we did not observe an accumulation of events in a single center or a cluster of centers.
Acute adverse reactions by type of contrast media
The EuroCMR Registry individually recorded the adverse reactions for the following contrast media: gadopentetate (e.g., Magnevist), gadoteracid (e.g., Dotarem), gadobenat (e.g., Multihance), gadobutrol (e.g., Gadovist), gadoteridol (e.g., Prohance), and gadodiamid (e.g., Omniscan). All other contrast media were summarized as “others.” Eighteen of the 45 centers exclusively used a single contrast agent for all patients (gadopentetate: n = 5, gadoteracid: n = 1, gadobenat: n = 0, gadobutrol: n = 6, gadoteridol: n = 2, gadodiamid: n = 4). All other centers used at least 2 or more different contrast agents in their clinical routine.
Table 2 displays the adverse reactions categorized by severity (all mild) and the specific contrast media. Note that the event rate ranges from 0.06% to 0.47% between the different agents. However, we did not find a relation between the individual event rates and the specific characteristics of the agents, such as molecular structure or chelate stability (Table 3).
The characteristics of all patients with acute adverse reactions sorted by contrast agent type can be viewed in Table 4. The most frequent complaints following contrast administration were rashes and hives (9 of 30), followed by nausea (7 of 30) and anxiety (6 of 30). Of 30 patients with adverse events, 18 required short-term observation and 11 of them were treated with steroids and/or antihistamines as a precaution. All patients improved during observation and could be discharged later. No one needed to stay as an inpatient due to the adverse contrast reaction.
Acute adverse events by CMR indication
When focusing on the reactions sorted by the initial CMR indication, we found different event rates between the three main indications (8), ranging from 0.1% in the group of basically healthy patients undergoing CMR for risk stratification in suspected coronary artery disease to 0.42% in patients undergoing CMR for detection of myocardial viability in the setting of left ventricular dysfunction and known coronary artery disease (p = 0.006) (Table 5).
Only 30 of our 17,767 CMR patients suffered from gadolinium-related acute adverse reactions. All reactions were classified as mild and only 18 of those 30 patients needed short-term observation before discharge. This is unique in that the present dataset is the first to systematically evaluate the frequency, manifestation, and severity of acute adverse reactions associated with gadolinium-based contrast agents in the specific setting of CMR in a multinational and multicenter setting.
CMR in comparison to general radiology use
The overall rate of acute adverse reactions in our 17,767 patients undergoing CMR was 0.17%. This is completely in line with the observations in the setting of the U.S. Food and Drug Administration (FDA)-approved general radiology use of gadolinium contrast, reporting adverse reaction rates between 0.04% and 2.2% (3–7,12–14).
Interestingly, in our cohort, only mild reactions occurred (see definition in Table 1), whereas in the general radiology use, up to 17.2% of reactions were graded as moderate (12) and up to 6.3% of reactions were graded as severe (12).
Differences between groups
Focusing on the individual reaction rates for the different contrast media recorded by the registry, we found a range of 0.06% and 0.47% between the different agents. This finding is in line with the findings of a recent retrospective analysis of the FDA Adverse Event Reporting System (14). As suggested by Prince et al. (14), a possible explanation for this finding could be differences between the molecular structure and chelate stability in the different agents. However, as displayed in Table 3, this does not seem to play a relevant role for acute adverse reactions in our population. Nevertheless, molecular structure and chelate stability may play a role in long-term complications such as nephrogenic systemic fibrosis. Unfortunately, long-term follow-up is not available for this current dataset (see the Study limitations section).
In addition, we found different reaction rates depending on the indication for CMR scanning (Table 5). We believe that this finding is most likely explained by the different burden of disease in the different patient groups. Whereas basically healthy people undergoing CMR for risk stratification for suspected coronary artery disease have the lowest rate of gadolinium-related adverse events, in comparison the sick group of patients presenting with left ventricular dysfunction due to known coronary artery disease has the highest rate of events. This finding may even suggest that in this group some of the often unspecific symptoms such as nausea or anxiety (Tables 1 and 4) that had been interpreted as gadolinium-related symptoms may also be due to the underlying disease (e.g., heart failure). In fact, based on this observation, one may speculate that the reaction rate truly caused by the gadolinium administration could be even lower than currently reported.
The data presented was collected as part of a registry. Thus, it is important to keep in mind that a prospective randomized trial may be the best tool to prove a certain principle, but only a registry can reveal if the results of controlled trials hold true in the multinational multicenter clinical routine (8). Nevertheless, based on the registry structure, we can only report acute reactions; systematic long-term follow-up (e.g., with regard to nephrogenic systemic fibrosis) is not available. However, we do not expect cases of nephrogenic systemic fibrosis in our population, because patients included in the registry were scanned in compliance with our registry protocol (8,9) and the actual American College of Cardiology Foundation/American College of Radiology/Society of Cardiovascular Computed Tomography/Society of Cardiovascular Magnetic Resonance/American Society of Nuclear Cardiology/North American Society of Cardiovascular Imaging/Society for Cardiac Angiography and Interventions/Society of Interventional Radiology appropriateness criteria for CMR imaging (10), ensuring sufficient renal function by blood testing (or reviewing of recent lab results) before the administration of gadolinium.
The current dataset demonstrates the safety of gadolinium in a multinational and multicenter setting of CMR. The rate of acute contrast-related adverse reactions is low and similar to that reported in the literature in a general radiology setting (3–7,12–14). Thus, we believe that despite the fact that there currently is no FDA approval of gadolinium for CMR in the United States and many European countries, the off-label use does not bring any additional risks for the patient. In fact, datasets such as this, when combined with efficacy studies, may facilitate FDA approval for gadolinium in CMR in the future.
The incidence of acute adverse reactions after administration of gadolinium-based contrast in the off-label setting of CMR in our population was not different from the incidence in the FDA-approved general radiology setting. Thus, the off-label use of gadolinium-based contrast in CMR should be regarded as safe concerning the frequency, manifestation, and severity of acute events.
For a list of participating centers, please see the online version of this paper.
The EuroCMR Registry is supported by unrestricted educational grants from Medtronic, Inc. and Siemens Healthcare. Dr. Schwitter has received research support from Bayer-Schering. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac magnetic resonance
- U.S. Food and Drug Administration
- Received April 25, 2011.
- Revision received June 27, 2011.
- Accepted June 29, 2011.
- American College of Cardiology Foundation
- Fuster V.,
- Kim R.J.
- Li A.,
- Wong C.S.,
- Wong M.K.,
- Lee C.M.,
- Au Yeung M.C.
- Bruder O.,
- Schneider S.,
- Nothnagel D.,
- et al.
- Hendel R.C.,
- Patel M.R.,
- Kramer C.M.,
- et al.
- The American College of Radiology