TITLE: Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease: A Review of Diagnostic Accuracy

Transcription

1 TITLE: Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease: A Review of Diagnostic Accuracy DATE: 25 February 2009 CONTEXT AND POLICY ISSUES: X-ray angiography is considered the reference standard, also known as the gold standard, for the diagnosis of coronary artery disease (CAD). This procedure, however, is not without risk. Major adverse events like death, myocardial infarction, or stroke occur in 0.2% to 0.3% of patients, and minor complications, usually related to insertion of peripheral venous catheters, occur in 1% to 2% of patients. 1 The procedure is painful, costly, and exposes the patient to ionizing radiation. 1 Thus, there is considerable interest in developing non-invasive diagnostic tests such as cardiac magnetic resonance imaging (MRI). Numerous methods can be used to detect CAD using cardiac MRI. These include visualization of the coronary artery lumen, ischemic myocardial injury, or detection of the effect of ischemia on wall motion, perfusion, or coronary blood flow. 2 Two and 3-dimentional images can be generated. Visualization of coronary arteries using cardiac MRI is technically challenging due to the small size of the vessels and the motion that occurs during the cardiac and respiratory cycle. 2 Scanning must be synchronized with diastole when cardiac movement is reduced. Patients must either hold their breath or other computerized methods must be used to compensate for diaphragm movement. 3 At present, resolution of cardiac MRI images are lower than those of X-ray angiography and there is difficulty visualizing distal coronary arteries. 1 It is not always possible to obtain a good quality MRI image and in some cases the scan cannot provide any diagnostic information. Technical advances continue to be made with cardiac MRI and now include higher field strength (3-Tesla), whole-heart coronary scans, faster scan times (parallel imaging), quantitative measures of blood flow (phase-contrast imaging), and higher signal intensity without the use of contrast agents (steady-state with free-precession method). 1 Information on the diagnostic accuracy of cardiac MRI and its impact on patient outcomes is required for planning future investment in MRI technology. Disclaimer: The Health Technology Inquiry Service (HTIS) is an information service for those involved in planning and providing health care in Canada. HTIS responses are based on a limited literature search and are not comprehensive, systematic reviews. The intent is to provide a list of sources and a summary of the best evidence on the topic that CADTH could identify using all reasonable efforts within the time allowed. HTIS responses should be considered along with other types of information and health care considerations. The information included in this response is not intended to replace professional medical advice, nor should it be construed as a recommendation for or against the use of a particular health technology. Readers are also cautioned that a lack of good quality evidence does not necessarily mean a lack of effectiveness particularly in the case of new and emerging health technologies, for which little information can be found, but which may in future prove to be effective. While CADTH has taken care in the preparation of the report to ensure that its contents are accurate, complete and up to date, CADTH does not make any guarantee to that effect. CADTH is not liable for any loss or damages resulting from use of the information in the report. Copyright: This report contains CADTH copyright material. It may be copied and used for non-commercial purposes, provided that attribution is given to CADTH. Links: This report may contain links to other information on available on the websites of third parties on the Internet. CADTH does not have control over the content of such sites. Use of third party sites is governed by the owners own terms and conditions.

2 RESEARCH QUESTION: What is the diagnostic accuracy of cardiac magnetic resonance imaging (MRI) for the diagnosis and evaluation of patients with coronary artery disease? METHODS: A limited literature search was conducted on key health technology assessment resources, including PubMed, The Cochrane Library (Issue 1, 2009), University of York Centre for Reviews and Dissemination (CRD) databases, ECRI, EuroScan, international health technology assessment (HTA) agencies, and a focused Internet search. An initial search was conducted for HTAs, systematic reviews, and meta-analyses, published between 2004 and February A second search was conducted between January 2007 and February 2009 to update the results of the reviews; articles are limited to English publications only. No filters were applied to limit the retrieval by study type. Studies were included if the patients enrolled had known or suspected CAD and underwent both cardiac MRI and X-ray angiography. From the second search, only prospective clinical trials were selected. HTIS reports are organized so that the higher quality evidence is presented first. Therefore, HTAs, systematic reviews, and meta-analyses are presented first. These are followed by randomized controlled trials and observational studies. SUMMARY OF FINDINGS: This search identified two HTAs, one systematic review, three meta-analyses, no randomized controlled trials, and seventeen observational studies that addressed cardiac MRI for patients with CAD. Appendix 1 provides a brief guide to interpreting statistics used in diagnostic studies. Appendix 2 and appendix 3 provide summary details of the included studies. Health technology assessments The search identified two HTAs. 3,4 The first was commissioned by the Agency for Healthcare Research and Quality and the Centers for Medicare and Medicaid Services in the US and was published in They conducted an update of systematic reviews evaluating cardiac MRI and computed tomography (CT) for the diagnosis of CAD in native coronary arteries only (excluding patients who had undergone coronary bypass). The authors identified 28 studies in 980 patients that compared cardiac MRI to X-ray angiography. Patient inclusion criteria was not adequately reported. Sensitivity ranged from 68% to 96% and specificity from 57% to 100%. Limitations included the exclusion of a number of patients with poor quality scans and the inability of cardiac MRI to visualize the entire coronary tree. The authors stated that there was no direct data to support substitution of cardiac MRI for X-ray angiography. They also stated that prospective studies to determine the impact on treatment plans and patient outcomes were required. 3 The second report was published in 2005 and is somewhat dated therefore is only described briefly. The authors summarized published systematic reviews evaluating MRI or CT for a variety of clinical conditions. 4 Three systematic reviews provided information on the diagnosis of Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 2

3 CAD using cardiac MRI (no further patient information was provided). All reports concluded that the technology was promising but further technical advances and additional study were required before cardiac MRI could replace X-ray angiography. 4 Systematic reviews and meta-analyses Summary data of the included systematic review and three meta-analyses are summarized in Appendix 2, Tables 1 to 4. 2,5-7 These reviews compared the diagnostic accuracy of cardiac MRI to X-ray angiography in patients with known or suspected CAD. The report by Nandular et al. 5 published in 2007, included 37 studies (2,191 patients) and evaluated two cardiac MRI imaging techniques. Stress wall imaging cardiac MRI showed an overall sensitivity of 83% (95% confidence interval (CI) 79% to 88%) and a specificity of 86% (95% CI 81% to 91%). Stress perfusion imaging cardiac MRI had a sensitivity of 91% (95% CI 88% to 94%) and a specificity of 81% (77% to 85%). The positive and negative likelihood ratios were similar with each technique and showed strong diagnostic evidence. Approximately 1% to 4% of scans were unsuccessful and did not provide any diagnostic information. The authors concluded that stress perfusion cardiac MRI had good sensitivity and specificity in the diagnosis if CAD among patients with a high prevalence of disease. Limited data was available for a low prevalence patient group. There was no information provided on the definition of high versus low prevalence. The systematic review and meta-analysis by Beanlands et al evaluated cardiac MRI, positron emission tomography and multidetector computed tomography for the diagnosis and management of ischemic heart disease (Appendix 2, Table 2). The authors evaluated several cardiac MRI techniques including angiography, stress wall motion, stress perfusion and myocardial viability imaging. The report also presented Canadian consensus guidelines on the use of these technologies. Using cardiac MRI angiography (28 studies, 999 patients), the specificity and sensitivity for detecting vessel stenosis varied from 73% to 91% (CI not reported) depending on the scan method, however, only 86% of coronary segment images were assessable. 2 The positive and negative predictive value was 65% and 90% respectively if only interpretable segments were included in the analysis. The positive and negative predictive values dropped substantially to 37% and 85% if all segments (assessable and un-assessable) were analyzed. The positive predictive value for determining bypass graft patency was 95% and the negative predictive value was 44% (reduced by metallic clip artifacts). 2 Stress wall motion (893 patients) and stress perfusion imaging (647 patients) techniques showed sensitivity and specificity values 84%. 2 Myocardial viability was assessed by two techniques. Dobutamine stress imaging (401 patients) showed a pooled sensitivity and specificity >90%. The values for late gladoliuium enhancement (357 patients) were lower (81% and 83% respectively) for predicting recovery of left ventricular function. The authors concluded that future diagnostic performance studies need to consistently report details of the patient population and selection methods as well as conduct blinded evaluation of imaging results. The authors stated that additional study is required on the impact of imaging on patient management and outcomes. 2 The meta-analysis by Schuijf et al included 28 studies in 903 patients evaluated for CAD by cardiac MRI. Data from all studies were pooled and showed a sensitivity of 72% (95% CI Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 3

4 69% to 75%) and specificity of 87% (95%CI 86% to 88%). Approximately 83% of coronary segment images were assessable. The fourth meta-analysis, published in 2004 by Danias et al. 7, included 39 studies in >1,500 patients. Sensitivity and specificity of cardiac MRI for detecting CAD were 88% (95% CI 82% to 92%) and 56% (95% CI 43% to 68%) respectively. Since two more recent meta-analyses were identified, the Schuijf et al. 6 and Danias et al. 7 reports have been described only briefly. Further details can be found in Appendix 2 Tables 3 and 4. Observational studies A total of 17 prospective studies were identified that compared cardiac MRI to angiography (Appendix 3 Table 5) Studies included higher and lower field strength magnetic resonance machines (3 Tesla and 1.5 Tesla) and different scan techniques (stress perfusion imaging, dobutamine stress imaging, and cardiac MRI angiography). These studies tended to be small with a median of 37 patients (range 18 to 141). The patients enrolled consisted of those preselected to undergo X-ray angiography and had a higher prevalence of CAD (median 58%). In the 11 studies assessing stress perfusion imaging, the sensitivity ranged from 64% to 100% (median 89%) and the specificity from 66% to 100% (median 78%) Dobutamine stress imaging was assessed in two studies. In the first study, no patients had CAD detected on angiography and all cardiac MRI scans were negative. 20 The second study has specificity of 80% and sensitivity of 86%. 19 Using cardiac MRI angiography, the sensitivity ranged from 75% to 88% and specificity ranged from 57% to 98% For all scan methods, the diagnostic accuracy of cardiac MRI compared to angiography ranged from 72% to 96 %, with a median value of 85%. Limitations Reporting on the patient selection process and demographics were poor. Generally, the studies included in the reviews and summarized in Appendix 3, were small and enrolled a select higher risk population, the majority of which were male. No studies were identified in patients with a low risk of CAD, such as those attending an emergency department. Most observational studies were single centre studies. As a result, the generalizability of study findings may be limited. Among the observational studies summarized, the majority stated the physicians reading the cardiac MRI scans were blinded to the patient s clinical history and X-ray angiography results. It was not clear however, if the clinician interpreting the X-ray angiography images were also blinded. In most studies, the sensitivity and specificity was calculated excluding the patients with poor quality scans. Had these patients been included in the analysis, the diagnostic accuracy of cardiac MRI would have been lower. In addition, assessor training and experience differences between and within studies may impact the assessors accuracy and thus, the sensitivity, specificity, etc. Different definitions of coronary stenosis were used when interpreting X-ray angiography and this may explain some of the variation in the sensitivity and specificity reported. Cardiac MRI technology in still developing and some scan procedures evaluated in older studies may no longer be relevant. No studies evaluated the impact of cardiac MRI on clinical management or patient outcomes. Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 4

5 CONCLUSIONS AND IMPLICATIONS FOR DECISION OR POLICY MAKING: The sensitivity and specificity of cardiac MRI compared to X-ray angiography varies depending on the type of scan performed but generally, values are above 80%.This means that approximately 2 out of 10 patients with the disease would not be diagnosed. Obtaining good quality images with cardiac MRI is not always possible. Any estimates of specificity and sensitivity that excluded patients with poor quality scans would over-inflate the diagnostic accuracy of MRI. Studies evaluating the effect of cardiac MRI on clinical management and patient outcomes are lacking therefore the impact of cardiac MRI on the care of patients with CAD is not known. Additional large, multicentre studies in patients with varying levels of CAD risk are required to assess the diagnostic accuracy of cardiac MRI and its value compared to X-ray angiography. PREPARED BY: Gaetanne Murphy, Research Officer, BSc Pharm Charlene Argáez, Information Specialist, MLIS Health Technology Inquiry Service htis@cadth.ca Tel: Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 5

6 REFERENCES: 1. Bluemke DA, Achenbach S, Budoff M, Gerber TC, Gersh B, Hillis LD, et al. Noninvasive coronary artery imaging: magnetic resonance angiography and multidetector computed tomography angiography: a scientific statement from the american heart association committee on cardiovascular imaging and intervention of the council on cardiovascular radiology and intervention, and the councils on clinical cardiology and cardiovascular disease in the young. Circulation 2008;118(5): Available: (accessed 2009 Feb 6). 2. Beanlands RS, Chow BJ, Dick A, Friedrich MG, Gulenchyn KY, Kiess M, et al. CCS/CAR/CANM/CNCS/CanSCMR joint position statement on advanced noninvasive cardiac imaging using positron emission tomography, magnetic resonance imaging and multidetector computed tomographic angiography in the diagnosis and evaluation of ischemic heart disease--executive summary. Can J Cardiol 2007;23(2): Patel MR, Hurwitz LM, Orlando L, McCrory DC, Sanders GD, Matchar DB, et al. Noninvasive imaging for coronary artery disease: a technology assessment for the Medicare Coverage Advisory Commission. Am Heart J 2007;153(2): Foerster V, Murtagh J, Kentle BC, Wood RJ, Reed MH, Husereau D, et al. CT and MRI for selected clinical disorders: a systematic review of clinical systematic reviews [Technology report no. 59]. Ottawa: Canadian Coordinating Office for Health Technology Assessment; Available: (accessed 2009 Feb 6). 5. Nandalur KR, Dwamena BA, Choudhri AF, Nandalur MR, Carlos RC. Diagnostic performance of stress cardiac magnetic resonance imaging in the detection of coronary artery disease: a meta-analysis. J Am Coll Cardiol 2007;50(14): Schuijf JD, Bax JJ, Shaw LJ, de Roos A., Lamb HJ, van der Wall EE, et al. Metaanalysis of comparative diagnostic performance of magnetic resonance imaging and multislice computed tomography for noninvasive coronary angiography. Am Heart J 2006;151(2): Danias PG, Roussakis A, Ioannidis JP. Diagnostic performance of coronary magnetic resonance angiography as compared against conventional X-ray angiography: a metaanalysis. J Am Coll Cardiol 2004;44(9): Burgstahler C, Kunze M, Gawaz MP, Rasche V, Wöhrle J, Hombach V, et al. Adenosine stress first pass perfusion for the detection of coronary artery disease in patients with aortic stenosis: a feasibility study. Int J Cardiovasc Imaging 2008;24(2): Cheng AS, Pegg TJ, Karamitsos TD, Searle N, Jerosch-Herold M, Choudhury RP, et al. Cardiovascular magnetic resonance perfusion imaging at 3-tesla for the detection of coronary artery disease: a comparison with 1.5-tesla. J Am Coll Cardiol 2007;49(25): Doesch C, Seeger A, Hoevelborn T, Klumpp B, Fenchel M, Kramer U, et al. Adenosine stress cardiac magnetic resonance imaging for the assessment of ischemic heart disease. Clin Res Cardiol 2008;97(12): Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 6

7 11. Fenchel M, Franow A, Martirosian P, Engels M, Kramer U, Stauder NI, et al. 1 M Gdchelate (gadobutrol) for multislice first-pass magnetic resonance myocardial perfusion imaging. Br J Radiol 2007;80(959): Futamatsu H, Wilke N, Klassen C, Shoemaker S, Angiolillo DJ, Siuciak A, et al. Evaluation of cardiac magnetic resonance imaging parameters to detect anatomically and hemodynamically significant coronary artery disease. Am Heart J 2007;154(2): Gebker R, Jahnke C, Paetsch I, Schnackenburg B, Kozerke S, Bornstedt A, et al. MR myocardial perfusion imaging with k-space and time broad-use linear acquisition speedup technique: feasibility study. Radiology 2007;245(3): Available: (accessed 2009 Feb 9). 14. Gebker R, Jahnke C, Paetsch I, Kelle S, Schnackenburg B, Fleck E, et al. Diagnostic performance of myocardial perfusion MR at 3 T in patients with coronary artery disease. Radiology 2008;247(1): Krittayaphong R, Boonyasirinant T, Saiviroonporn P, Nakyen S, Thanapiboonpol P, Yindeengam A, et al. Myocardial perfusion cardiac magnetic resonance for the diagnosis of coronary artery disease: do we need rest images? Int J Cardiovasc Imaging Klein C, Gebker R, Kokocinski T, Dreysse S, Schnackenburg B, Fleck E, et al. Combined magnetic resonance coronary artery imaging, myocardial perfusion and late gadolinium enhancement in patients with suspected coronary artery disease. J Cardiovasc Magn Reson 2008;10(1):45. Available: (accessed 2009 Feb 9). 17. Meyer C, Strach K, Thomas D, Litt H, Nahle CP, Tiemann K, et al. High-resolution myocardial stress perfusion at 3 T in patients with suspected coronary artery disease. Eur Radiol 2008;18(2): Thomas D, Strach K, Meyer C, Naehle CP, Schaare S, Wasmann S, et al. Combined myocardial stress perfusion imaging and myocardial stress tagging for detection of coronary artery disease at 3 Tesla. J Cardiovasc Magn Reson 2008;10(1):59. Available: (accessed 2009 Feb 9). 19. Kelle S, Hamdan A, Schnackenburg B, Köhler U, Klein C, Nagel E, et al. Dobutamine stress cardiovascular magnetic resonance at 3 Tesla. J Cardiovasc Magn Reson 2008;10(1):44. Available: (accessed 2009 Feb 9). 20. Raman SV, Donnally MR, McCarthy B. Dobutamine stress cardiac magnetic resonance imaging to detect myocardial ischemia in women. Prev Cardiol 2008;11(3): McCarthy RM, Deshpande VS, Beohar N, Meyers SN, Shea SM, Green JD, et al. Threedimensional breathhold magnetization-prepared TrueFISP: a pilot study for magnetic resonance imaging of the coronary artery disease. Invest Radiol 2007;42(10): Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 7

8 22. Maintz D, Ozgun M, Hoffmeier A, Quante M, Fischbach R, Manning WJ, et al. Wholeheart coronary magnetic resonance angiography: value for the detection of coronary artery stenoses in comparison to multislice computed tomography angiography. Acta Radiol 2007;48(9): Öncel D, Öncel G, Türkoglu I. Accuracy of MR coronary angiography in the evaluation of coronary artery stenosis. Diagn Interv Radiol 2008;14(3): Available: (accessed 2009 Feb 9). 24. Ozgun M, Rink M, Hoffmeier A, Botnar RM, Heindel W, Fischbach R, et al. Intraindividual comparison of 3D coronary MR angiography and coronary CT angiography. Acad Radiol 2007;14(8): Egger M, vey Smith G, Altman D, editors. Systematic reviews in health care: metaanalysis in context. 2nd ed. Bristol (UK): BMJ Books; Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 8

9 APPENDICES: APPENDIX 1: Guide to Summary Statistics of Diagnostic Tests 25 In order to test the accuracy of an experimental test, comparison to a reference standard test is required. In this case, X-ray angiography is considered the gold standard and is used to determine which patients have or do not have CAD. The results of the experimental test are then compared to the reference test to determine the number of true and false, positive and negative results. The results of the reference standard test must be reliable or the performance of the experimental test will be poorly estimated. Patients with disease (positive angiography) Patients without disease (negative angiography) Positive cardiac MRI True positive False positive Total positive Negative cardiac MRI False negative True negative Total negative Total with disease Total without disease Total patients Sensitivity = number of true positives / total number with the disease (ie. the proportion with the disease that has positive test results) Specificity = number of true negatives / total number without the disease (ie. the proportion without the disease that have negative test results) Diagnostic accuracy = (number of true positive and true negative)/total number of patients Positive predictive value = number of true positives/total positive Negative predictive value = number of true negatives/total negative Positive likelihood ratio is the ratio of the true positive rate to the false positive rate. Negative likelihood ratio is the ratio of the false negative rate to the true negative rate A positive likelihood ratio >10 or a negative likelihood ratio <0.1 may provide convincing diagnostic evidence. Diagnostic evidence is considered strong if the test shows a positive likelihood ratio >5 or a negative likelihood ratio <0.2. Example: Number of patients with disease (positive angiography) Number of patients without disease (negative angiography) Positive cardiac MRI Negative cardiac MRI Sensitivity = 60/67 or 90% Specificity = 28/33 or 85% Diagnostic accuracy = (60+28)/100 or 88% Positive likelihood ratio = (60/67)/(5/33) or 5.9 Negative likelihood ratio = (7/67)/(28/33) or 0.12 Positive predictive value= 60/65 or 92% Negative predictive value = 28/35 or 80% Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 9

10 APPENDIX 2: Summary of Systematic Reviews or Meta-Analyses Table 1: Summary of Nandular Title Author, publication year Design Description of study Diagnostic performance of stress cardiac magnetic resonance imaging in the detection of coronary artery disease: a meta-analysis Nandular et al Meta-analysis Search: MEDLINE and EMBASE Jan 1990 to January 2007, abstracts and full text publications included; no language restrictions; hand search of reference lists. Outcomes Included studies Key results Patient group: known or suspected CAD; excluded healthy volunteers who did not undergo angiography Intervention: stress cardiac MRI for the diagnosis of CAD (native and non-native arteries), with 50% stenosis as the diagnostic threshold. Comparator: catheter-based X-ray angiography Study characteristics: 10 or more patients were evaluated Diagnostic accuracy: sensitivity and specificity at the patient level and coronary territory level Clinical utility: positive and negative likelihood ratio 37 studies (n=2,191 patients). Of these, 14 studies (n=754) used stressinduced wall motion abnormalities imaging and 24 used perfusion imaging (n=1,516). Stress wall imaging Perfusion imaging Sensitivity (95%CI) 83% (79 to 88) 91% (88 to 94) Specificity (95%CI) 86% (81 to 91) 81% (77 to 85) Positive likelihood 5.24 (3.28 to 7.21) 5.10 (3.92 to 6.28) ratio (95% CI) Negative likelihood 0.19 (0.15 to 0.24) 0.11 (0.07 to 0.15) ratio (95% CI) Unsuccessful cardiac MRI 0.7% 4.1% Prevalence of CAD 70.5% 57.4% Author conclusions Stress perfusion cardiac MRI has good sensitivity and specificity in the diagnosis of CAD among patients with a high prevalence of disease. Limited data is available for a low prevalence patient group. CAD=coronary artery disease; CI=confidence interval; MRI=magnetic resonance imaging Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 10

11 Table 2. Summary of Beanlands Title Author, publication year Design Description of study Outcomes Included studies Key results CCS/CAR/CANM/CNCS/CanSCMR joint position statement on advanced noninvasive cardiac imaging using positron emission tomography, magnetic resonance imaging and multidetector computed tomography angiography in the diagnosis and evaluation of ischemic heart disease executive summary Beanlands et al Systematic review, meta-analysis, position statement Search: MEDLINE, EMBASE and Cochrane from 1966 to 2005; websites of evidence-based agencies Update of published meta-analyses Population: patients with known or suspected ischemic heart disease Intervention: Cardiac MRI, PET or MSCT Comparator: angiography Diagnostic accuracy: sensitivity and specificity, positive and negative predictive values Described in results Sensitivity Specificity PPV NPV Cardiac MRI angiography, 1.5-T (28 studies, 999 patients) 72% 87% Interpretable segments: 65% 90% Stress wall motion (8 studies, 893 patients) Stress perfusion (11 studies, 647 patients) Myocardial viability -Dobutamine stress (10 studies, 401 patients): All segments: 85% 37% 90% 84% NR NR 84% 86% NR NR NR NR 91% 94% Author conclusions -Late gadolinium enhancement (13 studies, 357 patients): 81% 83% Any imaging technique needs to be performed and interpreted by clinicians with adequate training and experience. Considering the ongoing development of non-invasive imaging technology, guidelines on the training of physicians is required. Future diagnostic performance studies need to consistently report details of the patient population, selection method, and blinded evaluation of imaging results. Additional study is required on the impact of imaging on patient management and outcomes. 1.5-T=1.5 Tesla field strength; MRI=magnetic resonance imaging; MSCT=multi-slice computed tomography; NPV= negative predictive value; NR=not reported; PET=positron emission tomography; PPV=positive predictive value Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 11

12 Table 3. Summary of Schuijf Title Meta-analysis of comparative diagnostic performance of magnetic resonance imaging and multislice computed tomography for noninvasive coronary angiography Author, publication Schuijf et al year Design Meta-analysis Description of study Search: MEDLINE search for English articles published between Jan 1990 and Jan 2005; hand search of cardiology and radiology journals; reference lists reviewed. Outcomes Included studies Key results Author conclusions Patient group: known or suspected CAD Intervention: MSCT or cardiac MRI for the diagnosis of CAD, with 50% stenosis as the diagnostic threshold. Comparator: catheter-based X-ray angiography Diagnostic accuracy: sensitivity and specificity Pooled diagnostic accuracy and odds ratio of the different imaging techniques compared to angiography Percent of assessable coronary segments (i.e., scans of diagnostic quality) 1. Cardiac MRI versus angiography: 28 studies (903 patients) 2. MSCT versus angiography: 24 studies (1,300 patients) Cardiac MRI MSCT Sensitivity (95%CI) 72% (69 to 75) 85% (83 to 87) Specificity (95%CI) 87% (86 to 88) 95% (95) % of assessable coronary segments (95% CI) 83% (82 to 84) 4-16 slice MSCT: 87% (86 to 88) 16 slice only: 96% Prevalence of CAD Range 59 to 100%, NR in 10 studies Range 53 to 100%, NR in 4 studies MSCT was significantly more accurate than cardiac MRI in detecting CAD compared to angiography. CAD=coronary artery disease; CI=confidence interval; MRI=magnetic resonance imaging; MSCT=multislice computed tomography; NR=not reported Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 12

13 Table 4. Summary of Danias Title Author, publication year Design Description of study Diagnostic performance of coronary magnetic resonance angiography as compared against conventional X-ray angiography Danias et al., Meta-analysis Search: MEDLINE and EMBASE up to Jan 2004; full text publications in English, French or German; review of reference lists. Outcomes Included studies Key results Author conclusions Patient group: known or suspected CAD; excluded healthy volunteers or patients with coronary grafts Intervention: cardiac MRI for the diagnosis of CAD, with 50% stenosis as the diagnostic threshold. Comparator: catheter-based X-ray angiography Study characteristics: 10 or more patients were evaluated Diagnostic accuracy: sensitivity and specificity at the coronary artery segment level and at the patient level. 39 studies (1,522 patients) Patient level 13 studies (n=607) Segment level 27 studies (n=993 or 4,620 segments) Sensitivity 88% (82 to 92) 73% (69 to 77) (95%CI) Specificity (95%CI) 56% (43 to 68) 86% (80 to 90) Prevalence of Range 17% to 100 %; NR in 8 studies CAD Cardiac MRI has moderately high sensitivity and may have satisfactory negative predictive value for excluding major stenosis in patients with suspected CAD. The data available are insufficient to introduce CARDIAC MRI as a screening tool in patients with low prevalence of CAD. CAD=coronary artery disease; CI=confidence interval; MRI=magnetic resonance imaging; NR=not reported Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 13

14 APPENDIX 3: Observational Studies Table 5: Summary of observational studies comparing cardiac MRI to angiography Study n Population, (prevalence CAD) Burgstahler 20 Patients with aortic stenosis (40%) Cheng Suspected CAD (66%) Doesch Suspected CAD (81%) Fenchel 25 Suspected CAD (72%) Futamatsu 37 Suspected CAD (97%) Gebker Known or suspected CAD (55%) Gebker 101 Known or suspected CAD (69%) Krittayaphong 66 Suspected CAD (58%) Stenosis definition Cardiac MRI technique (image analysis) 70% 1.5-T stress perfusion imaging (2R, B) Sensitivity Specificity Diagnostic accuracy 100% 80% NR 50% 3-T stress perfusion imaging, (2R, B) 98% 76% 90% 1.5-T stress perfusion imaging 90%, p=0.38* 67%, 82%, p=0.73* p=0.33* 75% 1.5-T stress perfusion imaging (2R, 90% 77% 85% B) 70% 1.5-T stress perfusion imaging 82% 91% NR comparing two contrast media (2R,B) 50% 1.5-T stress perfusion imaging 79% 66% NR (1R,B), qualitative visual assessment 50% 1.5-T stress perfusion imaging (2R, 86% 78% 83% B) 50% 3-T stress perfusion imaging (2R, B) 50% 1.5-T perfusion imaging; myocardial perfusion reserve index 1.5-T perfusion imaging; stress/rest scans 1.5-T perfusion imaging; stress scan only (2R, not blinded, angiographer blinded to cardiac MRI results) Per patient: 90% 71% 84% Per vessel: 89% 86% 76% 90% 79% 85% 76% 75% 76% 87% 75% 82% Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 14

15 Study n Population, (prevalence CAD) Klein Known or suspected CAD (48%) Meyer Suspected CAD (60%) Stenosis definition Cardiac MRI technique (image analysis) 50% 1.5-T stress perfusion imaging; stress scans compared to two other scan methods, alone or in combination (2R, B) 70% 3-T stress perfusion imaging (2R, B) Sensitivity Specificity Diagnostic accuracy 87% 88% 88% Per patient: 89% 79% 85% Per vessel: 86% 83% 76% Stress: 93% 84% 88% 60 Known or 18 Women with a 29 Suspected Thomas 50% 3-T stress perfusion imaging suspected compared to stress tagging (2R, B) CAD (47%) Tagging: 64% 100% 83% Kelle Known or suspected CAD (68%) 50% 3-T dobutamine stress imaging (2R, B) 80% 86% 82% Raman 70% 1.5-T dobutamine stress imaging (B) On angiography, no patients had 70% positive nuclear or ECG stress test (0%) stenosis and 8 patients had luminal narrowing <70%in coronary arteries. Cardiac MRI dobutamine stress test was negative for all patients. McCarthy 50% 1.5-T magnetic resonance Segment 57% 72% CAD (52%) angiography (2R, B) analysis: 87% Maintz Known or 50% 1.5-T magnetic resonance Segment 82% 87% suspected CAD (23%) angiography (2R, B) analysis: 88% Öncel Suspected CAD (61%) 50% 1.5-T magnetic resonance angiography (2R, B) Segment analysis: 75% 98% 96% p=0.62 Ozgun Known or suspected CAD (27% ) 50% 1.5-T magnetic resonance angiography (2R, B) Segment analysis: 85% 88% 87% (1 or 2R, B)=Number of reviewers, blinded assessment; 1.5-T=1.5 Tesla field strength; 3-T=3 Tesla field strength; CAD=coronary artery disease; CRM=cardiac magnetic resonance imaging *p value for 3-T cardiac MRI versus 1.5-T cardiac MRI; per patient analysis unless otherwise specified; per segment prevalence. Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 15

16 APPENDIX 4: Other Articles of Interest 1. Hendel RC, Patel M, Kramer CM, Poon M, et al. ACCF/ACR/SCCT/SCMR/ASNC/NASCI/SCAI/SIR 2006 Appropriateness criteria for cardiac computed tomography and cardiac magnetic resonance imaging: A report of the American College of Cardiology Foundation Quality Strategic Directions Committee Appropriateness Criteria Working Group, American College of Radiology, Society of Cardiovascular Computed Tomography, Society for Cardiovascular Magnetic Resonance, American Society of Nuclear Cardiology, North American Society for Cardiac Imaging, Society for Cardiovascular Angiography and Interventions, Society of Interventional Radiology. J Am Coll Cardiol. 2006;48(7): Bluemke DA, Achenback S, Budoff M. et al. Noninvasive coronary artery imaging: Magnetic resonance angiography and multidetector computed tomography angiography: A scientific statement from the American Heart Association Committee on Cardiovascular Imaging and Intervention of the Council on Cardiovascular Radiology and Intervention, and the Councils on Clinical Cardiology and Cardiovascular Disease in the Young. Circulation. 2008;118: Cardiac Magnetic Resonance Imaging (MRI) for Patients with Coronary Artery Disease 16