| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorial |
1 From the Department of Medical Radiology, Division of Nuclear Medicine (G.K.v.S.) and MR Center (G.K.v.S., J.S.); and the Department of Internal Medicine, Division of Cardiology (J.S.), University Hospital, Rämistrasse 100, CH-8091 Zurich, Switzerland. Received October 17, 2000; accepted October 18. Address correspondence to G.K.v.S. (e-mail: vonschulthess@dmr.usz.ch).
Index terms: Arteries, grafts and prostheses, 54.4523, 941.1269, 949.1269 • Arteries, MR, 54.4523, 941.129412, 941.129416, 941.12944 • Arteries, stenosis or obstruction, 54.721, 941.721, 949.721 • Blood, flow dynamics • Coronary vessels, MR, 54.12142 • Editorials • Magnetic resonance (MR), vascular studies, 941.12944
In the article by Langerak et al (1) in this issue of Radiology, the authors describe a phase-contrast–interleaved echo-planar imaging sequence to measure blood flow in bypass grafts. The described technique represents another stepping stone in the almost 15-year-old quest for magnetic resonance (MR) imaging to become a major player in cardiac imaging, and, as such, serves as a point of focus to evaluate the current status of MR imaging in this quest.
To be a major player in cardiac imaging, an imaging technology has to either be relatively inexpensive, like echocardiography, or provide substantial and reliable information on coronary artery disease. The most relevant imaging information to be obtained in evaluating this disease is (a) coronary artery morphology, (b) wall motion analysis, (c) myocardial perfusion, and (d) myocardial viability. Workers in the field have long recognized that MR imaging can potentially yield definitive information in all of these areas and have become accustomed to calling MR imaging the "one-stop-shopping" examination for the heart. Why has the road toward reaching this goal been so slow? The answer requires some in-depth discussion.
MR imaging is an inherently slow technique, and imaging of the coronary vessels is, in many respects, the most challenging imaging task. Images of the coronary arteries have to have a spatial resolution in the order of 0.5 x 0.5 x 1 mm with a temporal resolution in the order of 30–60 msec, and these parameters have to occur with reliable and consistent image quality. Until recently, MR imaging was unable to come close to meeting these criteria. Optimizing MR imaging for cardiac imaging has been a constant "navigating" within the "triangle" of speed, spatial resolution, and signal-to-noise ratio (ie, contrast).
Electrocardiographically triggered MR imaging can enable one to evaluate wall motion (2). For this task, spatial resolution is not so relevant, so somewhat imprecise registering of the image sections acquired during different heart cycles is acceptable. Various MR imaging methods have been devised to this end. The established technique is the measurement of the position and thickness of the myocardial wall during various cardiac phases; this provides information on ejection fraction, regional wall thickness changes, and ventricular mass (3). Although MR imaging is an excellent method for this task, echocardiography is adequate for daily clinical practice, relatively inexpensive, and available. More sophisticated approaches, such as myocardial tagging (4,5) and direct motion measurements with motion-sensitive MR sequences, provide information on tissue deformation that is not available by using echocardiography. However, because data analysis with these methods is time-consuming, these techniques are not more useful for the work-up of patients with coronary artery disease compared with echocardiography. The effect of advanced postprocessing approaches on the clinical utility of MR images remains to be proved (6). However, owing to the precision of the data obtained, MR imaging is evolving to become the method of choice in studies to evaluate the effect of cardioactive drugs, because the added precision reduces the number of patients needed to demonstrate the presence of a drug effect (7).
Probably the most exciting and unique aspect of MR imaging is that it allows direct measurement of flow or motion (8). An appropriate MR pulse sequence sensitized to velocity can enable one to measure the speed of blood flow or the motion of the myocardial wall. Because the number of pixels within a vessel cross section also can be readily determined, flow can be calculated.
Velocity-encoded MR imaging has proved to be the most quantitative and versatile noninvasive method to measure blood flow and is superior to Doppler echocardiography. MR flow measurements in the large vessels and even in the renal arteries have proved to be very reliable (9); therefore, it is tempting to extend this work to the measurement of flow in bypass grafts (10) and coronary arteries (11). The added difficulty in using this technique in the coronary vessels and in bypass grafts represents the standard problem in MR imaging of the heart: the issues of temporal and spatial resolution.
In their study, Langerak et al (1) used currently available MR imaging system improvements to make MR imaging work for their task of measuring blood flow in bypass grafts. They used a breath-hold imaging technique and an imaging unit with very fast gradients: It is probable that only systems with gradient strengths of 25–50 mT/m and gradient rise times (ie, slew rates) of 100–200 mT/m/msec work properly for the most important cardiac imaging tasks. They used interleaved echo planar–based MR pulse sequences, which, when used with linear or spiral k-space sampling, have been shown since the 1990s (12) to be promising for improving imaging speed without losing too much signal intensity and while maintaining spatial resolution. Despite the use of the most recent additions in MR imaging hardware and software, the data obtained by Langerak et al (1) still had error margins (ie, 95% CIs) in the 50% range for internal mammary flow measurements, as well as for bypass graft flow measurements, despite the fact that, in general, bypass grafts are more readily imaged because of their larger size and relatively limited motion during a cardiac cycle. Although this large SD may be due partly to physiologic variability, it also reflects technical shortcomings. Such an error margin is inadequate for individual patient assessment and probably not accurate enough even to longitudinally follow a patient’s disease, but it may be useful when statistically summing entire patient groups for study purposes. To our knowledge, few studies have involved the evaluation of noninvasive Doppler echocardiography to measure flow velocity in the coronary arteries (13), whereas intravascular ultrasonography is generally accepted to yield reliable information on coronary artery flow velocity.
Again, however, with most cardiac diseases, Doppler echocardiography yields adequate information, particularly in the evaluation of valvular heart disease, and, therefore, the well-established capability of MR imaging to enable measurement of flow noninvasively is not very frequently used in clinical practice, but rather it is reserved for complex cases of congenital heart disease after surgery.
There is an additional physiologic issue that has to be considered when promoting flow measurements in bypass grafts as a method of evaluating the blood supply to the heart. In the heart, particularly in the presence of bypass grafts, blood supply to the myocardium has many pathways, and, as a result, measurement of flow in the bypass grafts alone inadequately reflects the total myocardial blood flow. To get a complete picture, one must measure flow in the main coronary arteries as well. Alternatively, the blood supply to the myocardium can be assessed with measurements of blood flow in the coronary sinus (14); because of this vessel’s larger size, error margins are substantially smaller. Because coronary artery disease affects the entire coronary vessel system, this approach may enable the detection of the earliest stages of disease—that is, endothelial dysfunction that may precede obstructive disease.
Langerak et al (1) describe a technique that is potentially applicable to several vessels with a single MR examination. Thus, their contribution is a methodologic rather than clinical stepping stone toward the "one-stop-shopping" MR cardiac examination. In fact, the evaluation of adequate blood supply to the myocardium is best performed by directly demonstrating myocardial perfusion. The evaluation of myocardial perfusion has been the mainstay of nuclear medicine: It yields qualitative, relatively low-spatial-resolution myocardial perfusion images during stress and at rest with single-photon–based myocardial perfusion agents and quantitative images that have somewhat better resolution with positron emission tomography–based perfusion imaging. The identification of subendocardial infarction and ischemia is not possible with these methods. Although some data can be obtained by using microbubble-based contrast agents with echocardiography, this technique has not progressed to clinical application. However, results of recent studies of fast MR imaging techniques in patients suggest that MR imaging can depict coronary artery stenoses with high sensitivity and specificity (15), and, moreover, it is able to provide perfusion data on larger portions of the heart (16). Furthermore, MR imaging offers the spatial resolution required to identify nontransmural ischemia (16,17), and the latest hybrid echo-planar techniques appear to be adequate for quantification of regions of subendocardial ischemia throughout the heart (18).
After a myocardial infarction, viability imaging is often necessary to decide the next therapeutic intervention. Here also, nuclear medicine, echocardiography with dobutamine stress, and above all, positron emission tomography yield reliable results, but MR imaging appears to be able to depict viability as well by using the effect of late contrast enhancement as a sign of lack of viability (19). Whether MR perfusion imaging will supersede the nuclear medicine procedures remains to be seen; in many settings, it will be merely a replacement of one noninvasive fairly costly methodology with another noninvasive fairly costly methodology.
The last and in our opinion most relevant issue in heart imaging is the imaging of the coronary arteries. Being able to measure flow in a bypass graft consistently, as demonstrated by Langerak et al (1), is equivalent to being able to image bypass grafts by themselves at a reasonable spatial resolution. MR imaging seems to be capable of depicting at least the proximal portions of bypass grafts adequately. However, despite the use of a variety of methods (20), the spatial resolution and consistency of MR for imaging coronary stenoses have not reached the level of reliability required for clinical utility. If one were able to replace invasive coronary angiography with MR coronary angiography, an expensive invasive method would be replaced by a somewhat less expensive noninvasive method—an achievement that would have immediate clinical importance—all the other capabilities of MR imaging described would be simple add-ons, and, thus, this technique would be a true "one-stop-shopping" examination.
MR imaging is slowly progressing toward better spatial resolution with shorter and shorter imaging times, and the strides made during the past few years have been impressive. Unfortunately, the speed-enhancing hardware modifications to MR imaging units have reached a level at which the physiologic effects on patients due to excessive slew rates have become a real issue, and further technical improvements probably will not continue without the development of even stronger gradients and slew rates. In looking at other imaging technologies, we suggest that modern multisection computed tomography (CT) also is poised to become the method of choice for noninvasive coronary vessel imaging (21). Whether conventional coronary angiography will be replaced by a noninvasive method—be it MR imaging or CT coronary angiography—in the not too distant future is still a matter that requires further investigation.
In summary, MR imaging of the heart, like no other imaging procedure, has the potential to depict the morphologic and functional parameters relevant to the most common cardiac disease, coronary artery disease. To be fully successful, MR imaging has to be capable of enabling noninvasive imaging of the morphologic features of and flow in the coronary vessels, wall motion, perfusion, and viability. Langerak et al (1) have introduced a feature on the path toward making MR imaging the long sought-after "one-stop-shopping" examination. We believe, however, that the key to cardiac MR imaging will not be flow imaging but rather reliable imaging of coronary artery stenoses.
FOOTNOTES
See also the article by Langerak et al (pp. 540–547 ) in this issue.
REFERENCES
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
