International Journal of Cardiology 168 (2013) 3807–3811

Contents lists available at ScienceDirect

International Journal of Cardiology

j ourna l homepage: www.e lsev ie r .com/ locate / i j ca rd

Indications, safety and image quality of cardiovascular magnetic resonance:Experience in N5,000 North American patients

Judy M. Luu a,c, Neil G. Filipchuk c, Matthias G. Friedrich b,⁎,1

a Department of Cardiac Sciences and Radiology, University of Calgary, Calgary, Canadab Department of Cardiac Sciences and Radiology, Montreal Heart Institute, Montreal, Canadac Department of Cardiac Sciences, University of Calgary, Calgary, Canada

⁎ Corresponding author at: Stephenson Cardiovasculacular Institute of Alberta, Department of Cardiac ScienceCalgary and Montreal Heart Institute, Université de MMontreal QC H1T 3C8. Tel.: +1 514 376 3330.

E-mail address: mgwfriedrich@gmail.com (M.G. Frie1 Matthias G. Friedrich is a board member and cons

Imaging Inc.

0167-5273/$ – see front matter © 2013 Elsevier Irelandhttp://dx.doi.org/10.1016/j.ijcard.2013.06.032

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 8 February 2013Accepted 20 June 2013Available online 15 July 2013

Keywords:Cardiovascular magnetic resonanceIndicationsSafetyImaging qualityComplications

Background: CMR offers accurate assessment of structure and function with high resolution. Although the useof CMR has been well established in Europe, information is lacking for the extent of this emerging modality inNorth America.Objectives: This study aimed to summarize indications, safety, image quality, extent of contrast use and extentof stress tests performed in a high-volume CMR centre.Methods: Consecutive patients scanned from July 2005 to November 2010 were included, with duplicates andresearch subjects removed. Original clinical referrals were categorized into 10 main indications.Results: Retrospective analysis was performed on 6463 patients (mean ± SD age = 50 ± 17). The mostcommon clinical indications were non-ischemic cardiomyopathies (28%), including myocarditis (18%), coronaryartery disease (17%), ARVD and/or other RV disease (12%), and congenital heart disease (11%). Gadolinium-based

contrast was given to 89.5% of patients as part of their CMR protocol. Of 10.9% (703/6463) of patients thatunderwent stress CMR, adenosinewas administeredmost commonly. Of 703 patients, 1 (0.14%) suffered ventric-ular tachycardia during adenosine stress, and transient, asymptomatic AV block was occasionally observed. Mod-erate to severe complications after contrast agent administration occurred in 9 (0.16%)of 5782 contrast-enhancedstudies, characterized by nausea and vomiting in 6 (0.12%) and by symptoms of acute systemic allergic reaction in2 (0.04%). Image quality was good (82.0%), moderate but diagnostic (16.6%) and poor in 1.4% of cases.Conclusion: In the high-volume CMR centre, main clinical indications were for myocarditis/cardiomyopathies,coronary artery disease and RV-related queries. CMR showed an excellent safety profile and high image qualityin 99% of cases.

© 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Cardiovascularmagnetic resonance (CMR) is widely accepted as thenon-invasive gold standard for quantitative imaging of the heart.However, due to various reasons such as conflicts between radiologistsand cardiologists, reluctance of imagers to perform quantitative evalua-tion procedures and the need for specific hardware and software set-tings, it is not as widely used as could be expected. Therefore, large-scale data on its clinical use, including data on cost-efficiency, safetyand workflow efficiency, are scarce. Pilot data from the EuroCMR regis-try have provided important data on indications, safety and imagequality from various sites in Germany. However, North American data

r MR Centre, Libin Cardiovas-s and Radiology, University ofontreal, 5000 Rue Belanger,

drich).ultant of Circle Cardiovascular

Ltd. All rights reserved.

are still lacking. Therefore, we sought to report relevant data from asingle, high-volume CMR-dedicated facility in North America on clinicalindications, safety, imaging quality, extent of contrast use and extent ofstress tests performed.

2. Methods

2.1. Institution

The Stephenson Cardiovascular MR Centre is a dedicated centre for cardiovascularMRI at the Libin Cardiovascular Institute of Alberta in the Foothills Medical Centre andis under the administration of a governmental health care provider (Alberta HealthServices) and an academic institution (University of Calgary), through the Division ofCardiology, Diagnostic Imaging as well as the University Departments of CardiacSciences and Radiology.

On typical days, up to nine CMR scans are performed during one 8-h shift, 5 days aweek. Reports are issued by board-certified cardiologists and radiologists (senior reader),often supported by training fellows (primary reader). In the presence of another reader,scans were presented and reports discussed on a daily basis between the primary reader,the senior reader of the day and other group members.

The clinical environment in the Foothills Medical Centre represents a tertiary carecentre with 650 beds and a strong academic interest and historically characterized by astrong nuclear cardiology presence and a focus on acute coronary interventions.

Fig. 1. Proportion of indications for CMR scans from July 2005 to November 2010(n = 6463).

3808 J.M. Luu et al. / International Journal of Cardiology 168 (2013) 3807–3811

2.2. Protocols

All CMR scans were performed using standardized protocols (Table 1), in agreementwith recommendations published by the Society for Cardiovascular Magnetic Resonance[1] and, more recently, following the recommendations of the Canadian Society forCardiovascular MR (www.canscmr.org). Patients were referred by cardiologists orinternists from local hospitals and private practices, mostly in the Calgary area.

2.3. Reports

All reports were issued using a custom-designed reporting template based oncommercially available data management software (FileMaker Pro Advanced v. 9 to11, FileMaker Inc., Santa Clara, CA), including key images and quantitative data,which are stored within a searchable database of the software.

2.4. Data acquisition

A retrospective analysis was performed on 6463 consecutive CMR studies scannedfrom July 2005 to November 2010 at the centre. A blanket research ethics approval wasgranted to utilize only patient data included in the database for descriptive analysis.For the analysis, FileMaker data were exported to Microsoft Excel (Microsoft, Redmond,USA).

Research scans were categorized separately and were not included in the primaryanalysis. Similar to the EuroCMR Registry [2], 10 categories were defined to representthe primary CMR clinical indications. CAD/Ischemia composed of indications for anyconditions related to coronary artery disease, ischemia or assessment of viability. Thecategory of right ventricular (RV) composed of indications for arrhythmic rightventricular dysplasia, along with any queries for right ventricular dysfunction. Cardiacmass included indications for thrombus formation. Other miscellaneous indicationsincluded arrhythmia, general LV function assessment and others (see Fig. 1).

The clinical scans were classified into 10 diagnostic categories, based on the clinicalindication assigned at the time of the scan by the clinical CMR reader.

Severe complications were defined as death, resuscitation or any condition thatrequired the patient be monitored at the hospital as a result of the CMR protocol,such as arrhythmia and significant allergic reaction. Complications were defined asmild to moderate if they did not meet the criteria for being severe and includedtransient shortness of breath, chest pain or problems with intravenous lines [2].

The readers assigned the image quality during the report generation. Images thatdid not allow for adequate answering of the diagnostic question were graded aspoor. Moderate image quality was defined as limited but sufficient for the diagnosticpurposes of the scan. The image quality was deemed good if not limited by artefactsor other problems.

2.5. Patients

Scans were not performed in the presence of standard contraindications (implantedpacemaker or defibrillator, implanted insulin or other pumps, cerebrovascular clips

Table 1CMR protocols used for the various indications.

Indications Protocol

Acute coronary syndrome including acute MI Acute MISuspected myocardial ischemia Adenosine stress

perfusionSuspected or known ARVC/D ARVC/ARVDSuspected or known congenital heart disease Congenital⁎Rule out or rule in coronary anomaly orproximal coronary artery stenosis

Coronary angiography

Rule out or rule in inducible myocardial ischemia Dobutamine stressfunction

Assessment of myocardial viability and scarring,suspected or known cardiomyopathy

Viability (fibrosis)

Suspected acute or chronic myocarditis,rule out active myocarditis

Inflammation

Assessment of volumes, mass, and globaland/or regional LV function.

Ventricular functionand morphology

Rule out or rule in pericardial disease PericardiumPulmonary vein anatomy pre or post atrialablation procedures

Pulmonary venography

Suspected or known valvular disease Valvular†Suspected or known aortic disease 3D aortographySuspected or known myocardial iron overload T2⁎

⁎ The protocols for questions related to congenital heart disease are tailored to thelesion and the individual clinical scenario and therefore vary substantially.

† Protocols vary slightly with respect to slice positioning, depending on the valve tobe assessed.

or other MRI-incompatible material) or if patients were unstable or unable to lieflat for at least 20 min. Intracoronary stents, even if just recently implanted, did notlead to exclusion. Furthermore, patients with atrial fibrillation or tachycardia werealso accepted.

2.6. Workflow

The workflow in the centre has been continuously modified for efficient use ofscanner times, ensuring a large number of scans despite offering five shifts per weekonly. This was possible by the “cardiac only” environment and may be difficult toreproduce in other sites with competing specialties. However, some important changessuitable for all imaging environments included having well-defined protocols that theentire team (administrative front desk, nurses and readers) understood, allowing foroptimal scheduling of the scanner.

Patients were asked to arrive at least 45 min prior, which ensured that missedappointments could be re-scheduled with the next outpatient or inpatient immediately.Having the patient arrive early also allowed for the nurses to prep and start IV sites, sothat as soon as the patient in the scanner was finished, the next one would be ready.

Finally, a 7-day pre-screening telephone call to the patient was also implementedthat greatly increased efficiency. The pre-screening ensured that all necessary conflictssuch as obtaining x-rays for metal orbits or unknown MRI contraindications could beresolved well in advance.

2.7. Cardiovascular MRI

All scans were performed using a 1.5-T MRI system (Avanto®; Siemens MedicalSolutions, Erlangen, Germany), using the body coil (images used for signal intensityquantification) and the surface coil (all other sequences). All scans were ECG triggered.

During the scans, communication was maintained through an intercom device,while patients were monitored with a video system, a 3-lead ECG and a respiratorymonitor. For pharmacologic studies, blood pressure and heart rate were also moni-tored during the scan using a separate monitor (Magscreen; Schiller Medical SA,Wissembourg, France).

Depending on the indication and the pertinent CMR protocol, patients werescheduled in time slots of 30 min (LV function, pericardial disease and great vessels),

3809J.M. Luu et al. / International Journal of Cardiology 168 (2013) 3807–3811

45 min (non-ischemic CMP, myocarditis and first-pass perfusion with viability),60 min (RV disease, cardiac or para-cardiac mass) or 90 min (congenital heartdisease). This time included the time needed to switch patients and set them up inthe scanner room (ECG, coil, noise protection and drug infusion).

Patient preparation included a final check for contraindications, establishing one ortwo (adenosine protocol) intravenous line(s), taking blood pressure values, and – ifdeemed of value – a resting ECGwas completed before the actual exam in a roomadjacentto the scanner room.

Depending on the protocol indicated, a surface or body coil was used. As contrastagents, gadolinium-based contrast agents were used, mostly as Gd-DTPA, typically ina complex with diethylenetriaminepentacetate, Gd-DTPA (Magnevist, Berlex Canada,Pointe-Claire, QC) and was determined based on the indicated protocol. A list of theprotocols is included in Table 1.

Pharmacological agents used included adenosine (140 μg/kg/min) for inducingmyocardial hyperemia or dobutamine (10 μg/kg/min for each 3-min interval, up to40 μg/kg/min, plus 0.25 mg atropine if the response to dobutamine was inadequate)for functional stress scans.

All CMR protocols included the quantitative assessment of LV function, typicallyusing six long axis views and three short axis views [3] followed by indication-specific protocols. CMR imageswere analyzed immediately using dedicatedworkstations(Mac Pro, double-screen setup) and certified software (cmr42, versions 3.1 through 3.4;Circle Cardiovascular Imaging Inc., Calgary, Alberta, Canada).

2.8. Reporting

The evaluation of the images was typically started during scanning, with functionalimages being transferred right after their acquisition. Reporting was completed duringscan time.

Fig. 2. Observed image quality for CMR scans from July 2005 to November 2010(n = 6463).

2.9. Statistical analysis

Considering the primary purpose of this study was for a descriptive nature, noformal hypothesis was tested. One-sample T-tests were used to determine whetherthere was a significant increase in proportions of scanned indications over theyears. Logistic regression analysis was used to determine which variable(s) wouldbe significant predictors of image quality. Significance was achieved at p b 0.05.Values were charted as frequencies and/or percentages. All statistical analysis wasperformed using SPSS (PASW) version 18.0 for Mac OS X.

Table 2Influence of age on image quality in patients scanned for coronary artery andischemic-related queries.

Age groups

b44 years 45–59 years 60–74 years 75+ years

Image qualityGood 88% 81% 78% 69%Moderate 11% 17% 20% 29%Non-diagnostic/poor 1% 1% 2% 2%

3. Results

The most common clinical indications were non-ischemic cardiomy-opathies (28%), including myocarditis (18%), coronary artery disease/ischemic heart disease (17%), arrhythmogenic right-ventriculardysplasia and/or other right ventricular disease (12%) and congenitalheart disease (11%) (Fig. 1).

The mean age of all patients scanned were 50 ± 16.5 years (range,9–100 years). The majority of scans were performed in males, with aratio of 3:2, males (61.3%) to females (38.7%). The largest proportionof scans indicated coronary artery disease, including suspectedmyocardial ischemia in patients ranging from 45 to 74 years of age(76%). Similarly, 47% of patients queried for myocarditis were lessthan 44 years of age.

The assigned time slots of typically 45 to 60 min were sufficient inalmost all cases, except for longer times often given for congenitalheart disease protocols.

A gadolinium-based contrast agent was administered to 5782(89.5%) of 6463 patients. Pharmacologic vasodilation or stress imagingwas done in 703 (10.9%) of 6463 patients, with adenosine being theagent of choice. Image quality was good in 82.0%, moderate in 16.6%and poor in 1.4% of cases (Fig. 2).

Age was the only variable that influenced the quality of images,while the addition of gender and body surface did not (Table 2).Interestingly, of the limited number of patients with poor imagequality, 17 of 94 studies were aborted or not initiated primarily dueto claustrophobia. Of the remaining, more than 60% (50/77) weredue to arrhythmia. Actual incidences of poor image quality thathindered the ability to answer the clinical question were 21% (16/77)of all non-diagnostic images. However, in these instances, assessmentof general LV function was still obtained.

3.1. Complications of CMR

Of 703 patients, 1 (0.14%) suffered transient ventricular tachycardiaduring adenosine stress and transient. Transient, asymptomatic AVblock was not systematically recorded but were observed in at least5% of adenosine scans. Moderate to severe complications after contrastagent administration occurred in 9 (0.16%) of 5782 contrast-enhancedstudies, characterized by nausea and vomiting in 7 patients (0.12%)and by symptoms of acute systemic allergic reaction in 2 patients(0.03%). None of the patients required hospitalization because of thesereactions

4. Discussion

The most frequent clinical indications performed at our centre werethose related to myocarditis and non-ischemic cardiomyopathies,confirming previous experience from the EuroCMR registry pilot data[2]. A large proportion (44%) of the indications for cardiomyopathywere for myocarditis (example seen in Fig. 3), confirming the utilityof CMR for diagnosing myocarditis using the integrated Lake LouiseCriteria [4]. We also observed similar data on image quality andsafety [5]. It should be emphasized that the high image quality wasachieved despite the inclusion of patients with atrial fibrillationand/or tachycardia.

Over the years, there was a consistent trend in clinical indications,with statistically significant increases in queries for myocarditis, great

Fig. 3. Example of a comprehensive CMR protocol done in a patient with acute myocarditis. Still frame of a cine loop in an end-diastolic (A) and end-systolic (B) short axis view.Note the small pericardial effusion close to the inferior myocardium. (C): T2-weighted image in the same slice orientation, showing high signal intensity consistent with myocardialedema, predominantly in the lateral wall. T1-weighted images before (D) and shortly after (E) gadolinium contrast administration. Late gadolinium enhancement (LGE) image ofthe same patient in a short axis (F) and long axis view (G), showing extensive irreversible injury of the myocardium.

3810 J.M. Luu et al. / International Journal of Cardiology 168 (2013) 3807–3811

vessels, and RV-related conditions, reflecting the unique diagnosticutility of CMR in this setting.

Compared to the EuroCMR Registry, our centre performed morecongenital scans (658/6463, 10%), likely because the centre serves alarge referral base without other scanners providing that service andbecause of a close collaboration with the Alberta Children's Hospital, awell-known centre for congenital heart disease management.

The numbers observed in our centre may differ from other institu-tions with respect to the number of vascular scans. This is due to thelocal policy that vascular MRI scans are not performed in the cardiacsystem, but using other MRI scanners in the Radiology Department.

Consistent with other recent reports [6], complications occurredvery infrequently during a CMR scan.We did not observe any complica-tions leading to any persisting damage, although we performed stresstests and did not exclude patients with advanced heart failure oracute heart disease. In our experience, therefore, CMR is an extremelysafe procedure.

5. Limitations

The purpose of this study was to illustrate the functionality of ahigh-volume, CMR centre through a summary of descriptive, clinicaldata. Unlike the EuroCMR Registry, our results represent singlecentre, retrospective data. We were not able to reliably track theimpact of CMR imaging on diagnostic decision making or prognosticoutcome. We, however, observed a steady increase of requests, whichwas indirectly shown through the increase in number of patientsscanned initially (approximately 1000) to over 1500 patientsscanned per year. The number of referring physicians doubled from60 initially to well over 150 referring physicians throughout thecommunity), which indicates that referring physicians considerCMR important for patient management. We also excluded patient'sresearch-only indicated scans. This was done to keep the clinicalfocus of the outcomes we were assessing. The exclusion of thesepatients may not have accurately reflected the sheer volume of studies,as the centre does contribute nearly one-fifth of CMR scans to research-related studies.

We were not able to adequately follow-up with our patients withrespect to nephrogenic systemic fibrosis (NSF), an extremely rare yetsevere condition that may develop after Gd injection in patients withacute or severe renal dysfunction. Therefore, although we have notencountered any reports in our patients, we cannot comment onthe incidence of this entity in our population. However, to ensure allpossible means of NSF prevention, our centre strictly follows theguidelines established by the US Food and Drug Administration(FDA 2010) and the American College of Radiology (2010) [7].

6. Conclusion

In the era of multi-modality imaging, CMR is a highly accepted non-invasive tool to evaluate various cardiovascular disease and conditions.We have shown that in a dedicated, high-volume cardiac MR centre,CMR is very safe and can be provided with high efficiency. The mostfrequent clinical indications include cardiomyopathies, includingmyocarditis, coronary artery disease, right-ventricular abnormalitiesand congenital heart disease.

Acknowledgements

The authors sincerely thank Loreen Thon for all her efforts inpatient scanning over the years and Linda Fauth and the researchnursing team for all their support with our studies.

References

[1] Kramer CM, Barkhausen J, Flamm SD, Kim RJ, Nagel E. Society for CardiovascularMagnetic Resonance Board of Trustees Task Force on Standardized Protocols.Standardized cardiovascular magnetic resonance imaging (CMR) protocols, societyfor cardiovascular magnetic resonance: board of trustees task force on standardizedprotocols. J Cardiovasc Magn Reson Jul. 7 2008;10:35.

[2] Bruder O, Schneider S, Nothnagel D, et al. EuroCMR (European cardiovascularmagnetic resonance) registry: results of the German pilot phase. J Am Coll CardiolOct. 6 2009;54:1457–66.

[3] Childs H, Ma L, Ma M, et al. Comparison of long and short axis quantification of leftventricular volume parameters by cardiovascular magnetic resonance, withex-vivo validation. J Cardiovasc Magn Reson Aug. 11 2011;13:40.

3811J.M. Luu et al. / International Journal of Cardiology 168 (2013) 3807–3811

[4] Friedrich MG, Sechtem U, Schulz-Menger J, et al. Cardiovascular magneticresonance in myocarditis: a JACC white paper. J Am Coll Cardiol Apr. 28 2009;53:1475–87.

[5] Bruder O, Schneider S, Nothnagel D, et al. Acute adverse reactions to gadolinium-based contrast agents in CMR multicenter experience with 17,767 patients fromthe EuroCMR registry. JACC Cardiovasc Imaging Nov. 2011;4:1171–6.

[6] Bruder O, Wagner A, Lombardi M, et al. European cardiovascular magneticresonance (EuroCMR) registry—multi national results from57 centers in 15 countries.J Cardiovasc Magn Reson 2013;15:9.

[7] Reiter T, Ritter O, Prince MR, et al. Minimizing risk of nephrogenic systemic fibrosisin cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2012;14(31):429X-14-31.