| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Departments of Radiology and Cardiology, Beneficencia Portuguesa Hospital, Sao Paulo, Brazil (R.C.C., C.A.M.C., L.A.G.G., D.J.R., J.M.d.G., S.S.L.); Cardiac MRI Unit, Massachusetts General Hospital, Harvard Medical School, Boston, Mass (R.C.C., T.J.B.); and MGH Institute for Technology Assessment, Massachusetts General Hospital, Harvard Medical School, Boston, Mass (U.S.). Received July 11, 2005; revision requested August 31; revision received September 29; accepted October 18; final version accepted November 23. Address correspondence to R.C.C. (e-mail: rcury{at}partners.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
70% stenosis) in patients suspected of having or known to have coronary artery disease (CAD), with coronary angiography serving as the reference standard. Materials and Methods: The committee on human research approved the study protocol, and all participants gave written informed consent. This study was HIPAA compliant. Forty-seven patients (38 men and nine women; mean age, 63 years ± 5.3 [standard deviation]) scheduled for coronary angiography were prospectively enrolled: 33 were suspected of having CAD (group A) and 14 had experienced a previous myocardial infarction and were suspected of having new lesions (group B). The MR imaging protocol included cine function, gadolinium-enhanced stress and rest first-pass perfusion MR imaging, and delayed-enhancement MR imaging. Myocardial ischemia was defined as a segment with perfusion deficit at stress first-pass perfusion MR imaging and no hyperenhancement at delayed-enhancement imaging. Myocardial infarction was defined as an area with hyperenhancement at delayed-enhancement imaging.
Results: One patient was excluded from analysis because of poor-quality MR images. Coronary angiography depicted significant stenosis in 30 of 46 patients (65%). In a per-vessel analysis (n = 138), stress first-pass perfusion MR imaging and delayed-enhancement imaging yielded sensitivity of 0.87, specificity of 0.89, and accuracy of 0.88, when compared with coronary angiography. The diagnostic accuracy of stress first-pass perfusion MR imaging and delayed-enhancement imaging was slightly better than that of stress and rest first-pass perfusion MR imaging in the entire population (0.88 vs 0.85), in group A (0.86 vs 0.82), and in group B (0.93 vs 0.90).
Conclusion: Stress first-pass perfusion MR imaging followed by delayed-enhancement imaging is an accurate method to depict significant coronary stenosis in patients suspected of having or known to have CAD.
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Myocardial perfusion can be assessed with first-pass perfusion magnetic resonance (MR) imaging at rest and during pharmacologic vasodilatation. Preliminary studies have shown promising results (1–4). In studies with larger populations, researchers assessed inducible ischemia by combining the information provided by rest and stress first-pass perfusion MR imaging and demonstrated accurate detection (range, 85%–88%) of clinically significant CAD as defined at coronary angiography (stenosis diameter 50%, 70%, or 75%, depending on the study) (5–7). However, some patients could have had previous myocardial infarction that was either silent or clinically missed. Thus, the perfusion deficits identified could have included areas of myocardial infarction. Contrast material–enhanced MR imaging can be used to characterize tissue injury after myocardial infarction by means of delayed-enhancement imaging, with excellent histopathologic comparison (8–11), and may aid in the assessment of ischemia when combined with first-pass perfusion MR imaging.
We hypothesized that by using a combined MR imaging approach with stress first-pass perfusion MR imaging followed by delayed-enhancement MR imaging, it is possible to detect significant coronary stenosis (70% diameter stenosis) in patients clinically suspected of having or known to have CAD. Thus, the purpose of our study was to prospectively determine the accuracy of this combined MR imaging approach for depicting significant coronary artery stenosis in patients suspected of having or known to have CAD, with coronary angiography serving as the reference standard.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Exclusion criteria were hemodynamic instability, acute coronary syndromes, severe hypertension, atrial fibrillation, known severe aortic stenosis, and asthma. Antianginal medication and caffeinated beverages or food were withdrawn 24 hours before the examination. All patients underwent electrocardiography before and after MR imaging for comparison and to rule out any possible side effects after pharmacologically induced hyperemia.
MR Imaging Examination
All patients were examined with a 1.5-T imager (Signa CV/i; GE Medical Systems, Milwaukee, Wis), and a four-element phased-array cardiac coil was used for signal reception. After pulse sequences were used to localize the heart, vasodilatation was induced by using dipyridamole (0.56 mg per kilogram of body weight delivered intravenously over 4 minutes). Approximately 2 minutes later, breath-hold first-pass perfusion MR imaging was performed by using a hybrid gradient echo-planar imaging pulse sequence, which has been described elsewhere (12), during a bolus injection of 0.1 mmol/kg of gadopentetate dimeglumine (Schering, Berlin, Germany) performed with an infusion pump (Medrad, Indianola, Pa) at 5 mL/sec, followed by a 20-mL saline flush. This pulse sequence yields five to eight sections (depending on the heart rate) in the short-axis view covering the entire left ventricle every other heart beat (repetition time msec/echo time msec, 6.7/1.4; echo train length, four; flip angle, 20°; matrix, 128 x 128; bandwidth, 125 kHz; field of view, 34 x 34 cm; section thickness, 8 mm; saturation pulse, 90°). During the study, electrocardiographic results, heart rate, and respiratory curve were monitored continuously, and blood pressure was measured at 4-minute intervals.
After first-pass perfusion MR imaging during hyperemia, aminophylline was administrated (3 mg/kg delivered intravenously over 2 minutes) to antagonize the dipyridamole vasodilator effect. While the gadopentetate dimeglumine was washing out of the myocardium, left ventricular function was assessed with cine images and use of a steady-state free precession technique (3.5/1.4; matrix, 192 x 192; field of view, 34 x 34 cm; section thickness, 8 mm) in the short-axis view to evaluate for global and regional wall motion abnormalities. Ten minutes after stress perfusion (the time in which aminophylline was administrated and cine images were obtained), a second bolus of gadopentetate dimeglumine was injected (0.1 mmol/kg delivered intravenously at 5 mL/sec), and rest perfusion images were acquired by using the same pulse sequence used for stress perfusion.
Approximately 10 minutes later, delayed-enhancement MR imaging was performed by using an inversion-recovery prepared gated fast gradient-echo pulse sequence. Delayed-enhancement images were acquired and displayed to optimally display normal myocardium (dark) and regions of delayed-enhancement myocardium (bright) with proper selection of the inversion time. We acquired multiple sequences with varying inversion times and then selected the images with the most appropriate inversion time. Imaging parameters were as follows: repetition time msec/echo time msec/inversion time msec, 7.1/3.1/150–300; matrix, 256 x 192; flip angle, 20°; and inversion pulse, 180 degrees. The mean time of the MR imaging examination was 50 minutes ± 10 (Fig 1). All patients underwent the complete MR imaging examination without severe complications. Only two patients had minor events: One experienced dyspnea, and the other experienced chest pain that ceased with administration of aminophylline.
|
Image Analysis
Myocardial segments were assigned to the three major coronary arterial territories according to the American Heart Association standardized myocardial segmentation (13). Two experienced observers (R.C.C., C.A.M.C., with 4 and 10 years of experience in cardiac MR imaging, respectively) who were blinded to the results of coronary angiography and to clinical data qualitatively interpreted by consensus each MR imaging technique individually in different sessions (at least 1 week apart) and then by incorporating all MR imaging data (cine function, stress and rest first-pass perfusion MR imaging, and delayed-enhancement MR imaging). The decision-making algorithm for combined MR imaging techniques was performed by using the "believe the positive" rule, which means that the overall test result, combining test 1 and test 2, was positive if the result of at least one of the tests (test 1 or test 2) was positive. If both test results were negative, the overall test result was negative.
Global left ventricular ejection fraction was calculated by using the Simpson method in the stack of short-axis cine images by one observer (R.C.C.). Cine images were considered abnormal if any degree of wall motion abnormality (hypokinesia, akinesia, or dyskinesia) was present. The criterion used for perfusion defects at first-pass perfusion MR imaging was a persistent delay in the enhancement pattern during the first pass of the contrast medium through the myocardium observed in at least three consecutive temporal images and at least two contiguous sections. Myocardial ischemia was defined as a segment with perfusion deficit at stress first-pass perfusion MR imaging and no hyperenhancement at delayed-enhancement MR imaging. Myocardial infarction was defined as an area with hyperenhancement at delayed-enhancement MR imaging consistent with a coronary distribution. MR imaging and angiographic data were analyzed and stored without knowledge of the findings obtained during the other procedure.
Statistical Analysis
Results are expressed as sensitivity, specificity, and overall accuracy, with 95% confidence intervals (CIs) calculated with the normal approximation method (14), and with the angiographic result serving as the reference standard. All measures of diagnostic accuracy were calculated on a per-vessel basis. Diagnostic accuracy of the combined MR imaging approach using stress first-pass perfusion MR imaging and delayed-enhancement MR imaging was calculated for the overall population, patients with one-vessel disease, patients with two-vessel disease, and patients who previously underwent coronary artery bypass graft surgery. Diagnostic accuracy of single and combined cardiac MR imaging techniques was calculated for the overall population. In addition, diagnostic accuracy was calculated for patients in groups A and B, who underwent two different combined MR imaging approaches: Patients in group A underwent stress first-pass perfusion MR imaging and delayed-enhancement MR imaging. Patients in group B underwent stress and rest first-pass perfusion MR imaging. The Fisher exact test was used to compare (a) the accuracy of the combined MR imaging approach among patients with one-vessel disease and those with two-vessel disease and (b) the accuracy between stress first-pass perfusion MR imaging and delayed-enhancement MR imaging versus stress and rest first-pass perfusion MR imaging for the overall population and in groups A and B. A P value of less than .05 was considered to indicate a significant difference. All statistical analyses were performed by using a dedicated software program (Intercooled Stata 6.0; Stata, College Station, Tex).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
70%) in 30 patients. Eight patients had one-vessel disease, 12 patients had two-vessel disease, three patients had three-vessel disease, and seven patients who had previously undergone coronary artery bypass graft surgery had at least one-vessel disease.
Diagnostic Performance of Stress First-Pass Perfusion MR Imaging and Delayed-Enhancement MR Imaging
The accuracy of the combined MR imaging approach (Table 1) to detect significant stenosis as defined at coronary angiography on a per-vessel basis was as follows: overall population, 0.88; patients with one-vessel disease, 0.96; patients with two-vessel disease, 0.75; and patients who had previously undergone coronary artery bypass graft surgery, 0.90. Three patients had three-vessel disease, and the combined MR imaging approach could depict eight of nine vessels (correlation, 89%) with significant stenosis. Diagnostic accuracy in evaluating perfusion deficits on a per-vessel basis was better predicted in patients with one-vessel disease than in patients with two-vessel disease (P = .04).
|
|
|
|
|
|
|
|
|
All 14 patients with a history of myocardial infarction (group B) demonstrated hyperenhancement at delayed-enhancement MR imaging in at least one coronary artery territory, confirming the presence and location of previous myocardial infarction. All myocardial segments with delayed hyperenhancement at MR imaging demonstrated wall motion abnormalities on cine images.
Five patients, two in group A and three in group B, had nonsignificant stenosis at coronary angiography and demonstrated delayed hyperenhancement associated with perfusion deficits and wall motion abnormalities. These results were considered false-positive at MR imaging because coronary angiography was the reference standard.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
In our study, first-pass perfusion MR imaging was combined with delayed-enhancement MR imaging to determine whether areas of perfusion deficits originate in viable or infarcted myocardium. This distinction is fundamental for differentiation of areas of viable myocardium supplied by an artery having a significant stenosis (inducible ischemia) from areas of infarcted myocardium affected by large-artery occlusion or microvascular obstruction (myocardial infarction) (15). A combined MR imaging protocol has been studied before in specific populations, namely emergency department patients suspected of having acute coronary syndromes (17) and high-risk patients with acute coronary syndrome (18,19).
The major findings of our study include the following: (a) the presented cardiac MR imaging approach, combining stress first-pass perfusion MR imaging and delayed-enhancement MR imaging, had a diagnostic accuracy that was at least similar to or slightly better than that of combined stress and rest first-pass perfusion MR imaging for depicting significant coronary stenosis, not only for the entire population but also for groups A and B; (b) comparison of stress first-pass perfusion MR imaging with delayed-enhancement MR imaging can help characterize perfusion deficits and differentiate areas of inducible ischemia from areas of previous myocardial infarction in an unselected patient population; and (c) seven of 32 patients suspected of having CAD (group A) demonstrated unsuspected subendocardial myocardial infarction at delayed-enhancement MR imaging.
In the entire population we analyzed, the combinations of (a) cine function, stress first-pass perfusion MR imaging, and delayed-enhancement MR imaging; (b) stress first-pass perfusion MR imaging and delayed-enhancement MR imaging; and (c) stress first-pass perfusion MR imaging and cine function were the most accurate techniques for use in the identification of patients with significant stenosis. Thus, it seems reasonable to use a multimodality approach, including at least stress first-pass perfusion MR imaging and delayed-enhancement MR imaging, to detect significant stenosis. In groups A (patients suspected of having CAD) and B (patients with known CAD and new onset of angina), the combined approach of stress first-pass perfusion MR imaging and delayed-enhancement MR imaging had slightly better accuracy than did rest and stress perfusion. This approach also maximized sensitivity in group B (1.00). If one would like to maximize specificity, rest and stress perfusion would be the best technique in this group.
Our results demonstrated that the diagnostic accuracy in evaluating perfusion deficits in a per-vessel-basis analysis was better predicted in patients with one-vessel disease than in those with two-vessel disease (P = .04). This may represent a misclassification between anatomic variation of coronary distribution and the myocardial segments affected or, less likely, balanced ischemia in patients with two-vessel disease. However, it is important to notice, as was reported by Ishida et al (7), that if an analysis would be performed on a per-patient basis instead of on a per-vessel basis, the accuracies for detecting at least one perfusion deficit would have been better (a) in patients with three-vessel disease than in patients with two-vessel disease and (b) in patients with two-vessel disease than in patients with one-vessel disease.
Seven of 32 patients (22%) without a history of myocardial infarction (group A) according to clinical history and electrocardiographic criteria were identified as having perfusion deficits during hyperemia in areas of positive delayed enhancement. This outcome demonstrates the accuracy and importance of delayed enhancement in detecting unsuspected previous subendocardial myocardial infarction. Only four of these patients demonstrated perfusion defects at rest perfusion. Thus, if delayed-enhancement MR imaging was not used, these areas would be misclassified as segments of myocardial ischemia instead of as fixed perfusion deficits in areas of myocardial infarction. This finding emphasizes that rest perfusion has lower accuracy in the detection of myocardial infarction than does delayed enhancement because rest perfusion has a lower spatial resolution. This misleading information may have implications for therapy and clinical management. As reported by Kim et al (20), segments with more than a 50% degree of transmurality of delayed hyperenhancement are unlikely to recover after revascularization.
In five patients with delayed hyperenhancement, perfusion deficits, and wall motion abnormalities at cardiac MR imaging, including three patients with a history of myocardial infarction (group B), nonsignificant coronary stenosis was demonstrated at coronary angiography. These results were considered false-positive at MR imaging. However, in these patients, the coronary artery supplying an area of myocardial infarction could have opened spontaneously or by thrombolysis; the perfusion deficit depicted by first-pass perfusion MR imaging in that scenario would represent an area of microvascular obstruction (8,21).
Our study had limitations. Comparison with other myocardial perfusion techniques (SPECT, PET) was not performed. However, invasive quantitative coronary angiography remains the cornerstone to therapeutic decisions (revascularization and angioplasty), and SPECT and PET have false-positive and negative results. A larger sample size would be necessary to show whether there is incremental value of stress first-pass perfusion MR imaging and delayed-enhancement MR imaging over stress and rest first-pass perfusion MR imaging. Another limitation of our study was the qualitative analysis. Although MR imaging allows semiquantitative analysis, such analysis is considerably more time consuming and, therefore, is of limited use in clinical practice. We anticipate that quantitative assessment of first-pass perfusion MR imaging probably would improve the results of this protocol.
In conclusion, our study demonstrates that stress first-pass perfusion MR imaging followed by delayed-enhancement MR imaging is an accurate method for detecting significant coronary stenosis in patients suspected of having or known to have CAD. Comparison of first-pass perfusion MR imaging with delayed-enhancement MR imaging seems to be important for characterizing perfusion deficits if they occur in viable or infarcted myocardium. This MR imaging approach may improve visual assessment and may be at least as good as rest and stress perfusion MR imaging only. Further multicenter studies are necessary to establish the diagnostic accuracy of this method in larger patient groups.
|
ADVANCES IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
|
FOOTNOTES |
|---|
Abbreviations: CAD = coronary artery disease • CI = confidence interval
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, R.C.C., C.A.M.C., T.J.B.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, R.C.C., C.A.M.C., L.A.G.G., J.M.d.G., U.S., T.J.B.; clinical studies, R.C.C., C.A.M.C., L.A.G.G., D.J.R., J.M.d.G., S.S.L.; statistical analysis, R.C.C., U.S., T.J.B.; and manuscript editing, R.C.C., C.A.M.C., D.J.R., U.S., S.S.L., T.J.B.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Stangl, V. Witzel, G. Baumann, and K. Stangl Current diagnostic concepts to detect coronary artery disease in women Eur. Heart J., March 2, 2008; 29(6): 707 - 717. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. R. Nandalur, B. A. Dwamena, A. F. Choudhri, M. R. Nandalur, and R. C. Carlos Diagnostic Performance of Stress Cardiac Magnetic Resonance Imaging in the Detection of Coronary Artery Disease: A Meta-Analysis J. Am. Coll. Cardiol., October 2, 2007; 50(14): 1343 - 1353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Bodi, J. Sanchis, M. P. Lopez-Lereu, J. Nunez, L. Mainar, J. V. Monmeneu, O. Husser, E. Dominguez, F. J. Chorro, and A. Llacer Prognostic Value of Dipyridamole Stress Cardiovascular Magnetic Resonance Imaging in Patients With Known or Suspected Coronary Artery Disease J. Am. Coll. Cardiol., September 18, 2007; 50(12): 1174 - 1179. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Gibbons, P. A. Araoz, and E. E. Williamson The Year in Cardiac Imaging J. Am. Coll. Cardiol., September 4, 2007; 50(10): 988 - 1003. [Full Text] [PDF]
|
|
|
|
 |
|
 T. D. Karamitsos, J. B. Selvanayagam, and R. C. Cury Are Two Tests Always Better than One? Radiology, July 1, 2007; 244(1): 319 - 320. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
