Our results show that the origin and the proximal course of the coronary arteries are well depicted with 3D-MRCA, but the visualization of stenoses is less reliable. A 10 min 3D-MRCA protocol can be used to screen for CAA. This MRCA protocol was designed for the detection of coronary anomalies and not for coronary stenosis detection.
Comparison to the literature
The number of interpretable 3D-MRCA scans is comparable to the 84% reported by Kim . Of all 335 technically satisfactory scans, the RCA originated from the left sinus in three cases (0.90%), one RCA originated higher from the ventral aorta, and no LAD originated from the right sinus, which is comparable to the x-ray coronary angiography prevalence as reported by Angelini; 0.92% for RCA and 0.15% for LAD [5, 17]. The MDCTA incidence of CAA is 2.5% as reported by Schmitt including the RCA originating outside its sinus in 0.40% and the left coronary arteries (both LAD and LCA) in 0.74% (a LAD originating outside its sinus with a normal LCA was not observed in their study). Although occurring infrequently, physical exercise is suggested to be highly related to SCD when the LAD has an anomalous origin from the right sinus and runs between the aortic root and the pulmonary trunk [6, 19]. One study reports that an anomalous origin of the RCA from the left sinus, as found in three of our subjects, occasionally gives major symptoms such as chest-pain or syncope and is the cause of 13% of CAA related SCD . An unusually high anterior origin of the RCA from the ascending aorta (“high take-off position”) and continuing down at a steep angle to the aorta as seen in one of our male non-athletes, has been described earlier by Schmitt and found in two (0.11%) of 1758 MDCTA screened patients suspected for CAD .
Myocardial bridging of the LAD was observed in seven of our cases, which was only confirmed in one person by MDCTA as the other subjects did not undergo MDCTA. The actual prevalence of asymptomatic myocardial bridging varies widely between x-ray coronary angiographic studies (0.5–2.5%) [20, 21], MDCTA (17–58%) [4, 14, 22] and autopsy series (15–85%), and most commonly affects the LAD [4, 21]. Recent MDCTA results are comparable to autopsy series which is the gold standard for detecting myocardial bridging .
Origin rotation variation of the RCA and the LCA with normal coronary origin from the sinus of Valsalva, caused by rotation of the aortic root between 45 and 90°, have been described earlier although only found in 0.22% .
A 10 min 3D-MRCA protocol can be used as a non-invasive screening tool to exclude CAA in asymptomatic athletes under 35 years of age as they have low risk of CAD [3, 24]. This MRCA protocol, with limited resolution and coverage within the time-limit, was designed for the detection of coronary anomalies and not for coronary stenosis detection. Consequently, older athletes and athletes with cardiovascular complaints should either undergo dedicated full-coverage 3D-MRCA or prospectively triggered MDCTA . Figure 3 illustrates the limited coverage shortcoming of this 3D-MRCA protocol: a soft plaque with a severe stenosis was found in the mid-segment of the RCA on MDCTA and this segment was outside the coverage of the 3D-MRCA. A low dose prospectively triggered MDCTA provides high resolution images which can confirm CAA and visualize the extent of associated intramural course and lumen stenosis, and rule out additional stenosis as it covers the entire coronary artery tree.
Conventional exercise stress testing can be an insufficient stress in a well-trained athlete. Additional testing in an athlete with CAA and/or asymptomatic stenosis can involve exercise testing until physical exhaustion using the super Bruce tread-mill protocol instead of the standard Bruce tread-mill protocol, to detect local ischemia or arrhythmias . If positive, a x-ray coronary angiography or scintigraphy can be performed.
Myocardial bridges in the LAD are a common finding with intramyocardial segments often showing mild-to-moderate luminal narrowing at rest, which is higher during end-systolic phase, but rarely causes angina or infarction through intermittent phase-dependent vessel compression . As a prospective MDCTA is usually only acquired in the end-diastolic phase of the cardiac cycle it will not demonstrate possible narrowing during systole . If detected, myocardial bridging in symptomatic patients can be treated with β-blockers, nitrates or calcium antagonists and, in extremely selected cases, percutaneous coronary intervention or coronary artery bypass grafting .
Although 93% of the 3D-MRCA scans were technically satisfactory, β-blockade could further improve image quality as it lowers heart-rate and heart rate variability and increases the end-diastolic resting period significantly . However, little improvement may be expected in athletes with a regular heart rate below 60 beats per minute.
The main limitation of this study is the lack of validation with MDCTA or X-ray coronary angiography in subjects without a coronary artery anomaly. This study was not set up as a side-by-side comparison to MDCTA or X-ray coronary angiography as previous studies focusing on these comparisons have been already performed and the amount of radiation exposure was considered unethical in these healthy subjects [27, 28]. We discussed the results individually and left the people the choice for further diagnostic testing. As a consequence, test characteristics such as sensitivity and specificity cannot be presented. This study focuses only on the detection of proximal CAA in asymptomatic subjects.
1.5 Tesla MR scanners using five-element phased-array cardiac-coils are the current standard for cardiac imaging. SNR advantages of 3.0 Tesla can be used to improve coronary artery image resolution and shorten the scan time, but 3.0 Tesla is currently hampered by increased susceptibility artifacts. Future optimization of 3.0 Tesla techniques will increase clinical use . Increasing numbers of channels in multiple-channel coils can improve SNR and parallel imaging performance, however, reconstruction time and increased computer memory requirement become inevitable problems which need to be solved by advanced software still in experimental stage .