Identification of Anomalous Coronary Arteries and Their Anatomic Course by Magnetic Resonance Coronary Angiography
Background Anomalous coronary arteries are a rare but recognized cause of myocardial ischemia and sudden death. Identification currently requires x-ray angiography, which may have difficulty defining the three-dimensional course of the anomalous vessel. Magnetic resonance coronary angiography (MRCA) has been shown to image coronary artery anatomy noninvasively. We hypothesize that MRCA may be useful in the identification of anomalous coronary arteries and their anatomic course.
Methods and Results Sixteen patients (9 men, 7 women, age 44 to 81 years) with anomalous aortic origins of the coronary arteries by conventional x-ray angiography underwent MRCA. Multiple images of the major epicardial coronary arteries were obtained by use of a breathhold, fat-suppressed, segmented–k space, gradient-echo technique by investigators blinded to all patient data. Anomalous coronary artery pathology, by x-ray angiography, included right-sided left main coronary artery (n=3), right-sided left circumflex artery (n=6), separate left-sided left anterior descending and left circumflex arteries (n=2), left-sided right coronary artery (n=4), and an anteriorly displaced right coronary artery (n=1). MRCA correctly identified the anomalous coronary vessel(s) in 14 of 15 patients. In 1 patient, the anomalous vessel was incorrectly identified, and in 2 patients the course of the anomalous vessel was not clearly seen; one of these was a nondominant, anomalous right coronary artery.
Conclusions MRCA is a useful technique for the noninvasive identification of anomalous coronary arteries and their anatomic course.
Congenital anomalous origins of the coronary arteries are a rare but well-described cause of myocardial ischemia and sudden death.1 2 3 4 5 In adults referred for conventional x-ray coronary angiography, anomalous aortic origins of the coronary arteries are noted in 0.6% to 1.2% of patients.2 4 Although the majority of coronary artery anomalies are not thought to be hemodynamically significant, origin of a coronary artery from the contralateral side with subsequent passage between the aorta and pulmonary artery has the potential to impair myocardial perfusion.3 5 The ability to identify coronary artery anomalies and define their anatomic course would be an important component of any noninvasive coronary imaging technique. It would be useful to further evaluate patients in whom the results of conventional angiography are uncertain or to study young patients with chest pain or syncope in whom a coronary anomaly is suspected.
The current diagnostic method of choice for detecting coronary artery anomalies is conventional x-ray coronary angiography.6 However, x-ray angiography provides only a two-dimensional view of a vessel’s complex three-dimensional path, so the anatomic course of the anomalous vessel with respect to the aorta and pulmonary artery may be difficult to discern. In addition, the anomalous vessel may be erroneously overlooked or assumed to be occluded if not selectively engaged.6
Magnetic resonance (MR) is a noninvasive, three-dimensional imaging technique that has recently been shown to image extensive portions of the epicardial coronary arteries and detect coronary artery pathology.7 Previous MR studies of patients with anomalous coronary arteries have involved few patients and have used MR to confirm the suspected origin and path of the anomalous vessel in an unblinded fashion.8 9 10 In this study, we applied MR coronary angiography (MRCA) to the blinded, noninvasive identification of anomalous coronary arteries and definition of their anatomic course with respect to the aorta and pulmonary artery.
Sixteen patients (9 men, 7 women, age 44 to 81 years) previously diagnosed by conventional x-ray angiography as having an anomalous aortic origin of one or more coronary arteries were studied. Patients were excluded if they had nonsinus rhythm, an aortic valve prosthesis, or contraindications to MR, including a pacemaker, implantable defibrillator, intracranial clips, or intraocular/intra-auricular devices. Informed consent was obtained from all subjects under a protocol approved by the Beth Israel Hospital Investigational Review Board.
Magnetic Resonance Imaging
MRCA was performed by investigators blinded to all patient angiographic and clinical data. Patients were studied in the supine position in a 1.5-T whole-body MR research system (Siemens Magnetom, Siemens Medical Systems or Philips Gyroscan NT, Philips Medical Systems). The Siemens body coil or Philips surface coil (C1) was used as radiofrequency receiver. After scout imaging was performed, an ECG-gated, gradient-echo sequence was used with incremented flip angle series, k-space segmentation, and a fat-saturation prepulse applied before each segment.7 Images were acquired during breathholding, with a typical scan time of 12 to 18 seconds to complete the 120 to 180×256 matrix. Multiple transverse and oblique images were obtained to visualize the proximal origin and path of the major epicardial coronary arteries, using a 3- to 4-mm slice thickness with 1-mm overlap, a 220- to 250-mm field of view, a repetition time of 13 to 14 ms, and an echo time of 7 ms. Total imaging time averaged 50 minutes per patient.
Conventional X-Ray Angiography
All patients had previously undergone standard clinical diagnostic x-ray angiography in multiple views, which had established the diagnosis of one or more anomalous coronary arteries.
MR images were analyzed by two experienced MRCA angiographers (W.J.M., R.R.E.) blinded to all patient data. Conventional angiograms were reviewed by two experienced x-ray angiographers (P.G., A.P.S.) blinded to MRCA results. For both MRCA and x-ray angiography, the following data were recorded: (1) the anomalous coronary artery, (2) its origin, and (3) its path with respect to the aorta and pulmonary artery.
Coronary artery anomalies among the 16 patients, by conventional x-ray angiography, included (Table⇓) right-sided origin of the left main coronary artery (n=3), right-sided origin of the left circumflex artery (n=6), separate left-sided origins of the left anterior descending and left circumflex arteries (n=2), left-sided origin of the right coronary artery (n=4), and an anteriorly displaced origin of the right coronary artery (n=1). MRCA was performed without complication in all subjects. One subject could not hold her breath adequately, resulting in obvious poor image quality, and was excluded from the analysis.
Of the remaining 15 patients, 14 had the correct anomalous vessel(s) identified. All 3 patients with right-sided left main coronary arteries were correctly identified, including 2 in whom the anatomic course was anterior to the aorta and posterior to the pulmonary artery (Fig 1⇓⇓). In the 6 patients with right-sided left circumflex arteries, the anomalous vessel was correctly identified in all 6, but the site of origin and initial path were not clearly seen in 1 (patient 9). The anomalous circumflex coursed posterior to the aorta in all 6 patients (Fig 2⇓). The 2 patients with separate left-sided origin of the left anterior descending and left circumflex arteries were correctly identified. In the 3 patients with left-sided right coronary arteries who could perform breathholding, the anomalous vessel was correctly identified in 2 (Fig 3⇓). In 1 patient (patient 13), the MR images had a linear opacity coursing posterior to the aorta that appeared to originate from the right side and terminate near the left atrioventricular groove. This was interpreted as showing an anomalous right-sided left circumflex artery. Unblinded review of the images did identify the left-sided origin of the right coronary artery coursing initially along the anterior aortic wall and posterior to the pulmonary artery. In 1 other patient with a small and nondominant anomalous right coronary artery (patient 14), the site of origin and initial path were not clearly seen. Patient 15 had originally been diagnosed with a non–coronary-sinus origin on the clinical angiography report. However, the blinded expert review for this study found that the origin was from the anterior right coronary sinus near the junction with the left coronary sinus, which was the MRCA finding. Thus, in all 12 patients in whom the correct vessel was identified and the initial path clearly seen, the anatomic course with respect to the aorta and pulmonary artery was defined correctly.
This blinded study of patients with coronary artery anomalies has shown that MR coronary angiography can identify a wide range of anomalous coronary vessels (in 93%) as well as define their anatomic course with respect to the aorta and pulmonary artery (in 80%). This was done noninvasively in less than 1 hour without contrast agents or ionizing radiation.
Transesophageal echocardiography has also been used to image coronary anomalies. The largest published series included nine coronary anomaly patients, and transesophageal echocardiography was used to confirm the origin and initial course of the anomalous coronary vessel found on angiography.11 The image acquisition and analysis in that study were performed unblinded to the angiographic findings. Thus, the ability of transesophageal echocardiography to identify coronary anomalies in a blinded analysis is unknown. In addition, the technique is semi-invasive. Electron beam computed tomography has recently been used to image coronary arteries noninvasively, but experience with this modality is very limited.12
Previously published work using MR to image coronary artery anomalies has largely been in the form of case reports.8 9 A single, unblinded study of five coronary anomaly patients used a nonbreathhold spin-echo MR imaging sequence to confirm the anatomic course suspected on angiography.10
To the best of our knowledge, this is the first blinded report of a noninvasive imaging technique for the identification and evaluation of coronary artery anomalies. The investigators performing and analyzing the MR studies were blinded so as to avoid bias during both MR image acquisition and MR data interpretation. Investigators were aware, however, that the patient carried a diagnosis of anomalous coronary anatomy. Another approach would have been to include some number of patients without coronary anomalies. However, to investigate a truly representative population for this rare occurrence (prevalence ≈1%), 1000 to 2000 MRCA studies would need to be performed to evaluate a similar number of cases.
The three cases in which a complete, correct diagnosis could not be made point to the current limitations of MR coronary angiography. The largest clinical experience to date for noninvasive coronary imaging used the breathhold MR coronary angiography technique we used in this study.7 However, ongoing improvements in MR hardware and software offer the potential to improve the diagnostic accuracy of MR coronary angiography. Emerging techniques such as selective tagging of aortic blood and subtraction of background tissue could selectively image the proximal epicardial vessels and suppress noncoronary structures,13 which would be particularly advantageous for identifying anomalous coronary origins. New, nonbreathhold methods for respiratory motion compensation (eg, navigators)14 could accommodate a broader range of patients and would allow enhanced spatial and temporal resolution, and fewer registration errors between slices.
MR coronary angiography as a noninvasive imaging technique for identifying coronary artery anomalies has several potential clinical applications. First, it can evaluate the three-dimensional path of anomalous coronary vessels identified by x-ray angiography, particularly when there is uncertainty as to whether an anomalous vessel follows a hemodynamically significant course between the aorta and pulmonary artery. Second, MRCA could distinguish an occluded vessel from an anomalous one in cases in which a vessel cannot be engaged by conventional angiography. Finally, it could potentially be used in young patients with unexplained syncope or chest discomfort in whom a coronary anomaly is part of the differential diagnosis.
In conclusion, MR coronary angiography appears to be a valuable technique for the noninvasive evaluation of coronary artery anomalies. With more widespread clinical experience, it will probably be the noninvasive imaging modality of choice for this diagnosis.
Dr McConnell was supported by an Individual National Research Service Award from the National Heart, Lung, and Blood Institute (HL-09095), Bethesda, Md. Drs Edelman and Manning were supported in part by a grant from the National Institutes of Health (R01-HL-48538), Bethesda, Md. Dr Manning was also supported by the Edward Mallinkrodt, Jr Foundation, St Louis, Mo.
- Received August 9, 1995.
- Revision received September 14, 1995.
- Accepted September 15, 1995.
- Copyright © 1995 by American Heart Association
Cheitlin MD, De Castro DM, McAllister HA. Sudden death as a complication of anomalous left coronary origin from the anterior sinus of Valsalva: a not-so-minor congenital anomaly. Circulation. 1974;50:780-787.
Chaitman BR, Lesperance J, Saltiel J, Bourassa MG. Clinical, angiographic, and hemodynamic findings in patients with anomalous origin of the coronary arteries. Circulation. 1976;53:122-131.
Kimbiris D, Iskandrian AS, Segal BL, Bemis CE. Anomalous aortic origin of coronary arteries. Circulation. 1978;58:606-615.
Levin DC, Fellows KE, Abrams HL. Hemodynamically significant primary anomalies of the coronary arteries: angiographic aspects. Circulation. 1978;58:25-34.
Fernandes F, Alam M, Smith S, Khaja F. The role of transesophageal echocardiography in identifying anomalous coronary arteries. Circulation. 1993;88:2532-2540.
Sapoval MR, Mousseaux E, Desnos M. Anomalous origin of the left coronary artery from the right coronary sinus diagnosed by electron beam computerized tomography. Circulation. 1995;91:2093.