| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Reviews |
1 From the Department of Radiology, Evanston Northwestern Healthcare, 2650 Ridge Ave, Evanston, IL 60201; and Feinberg School of Medicine, Northwestern University, Chicago, Ill. Received September 29, 2003; revision requested November 11; revision received December 8; accepted January 21, 2004. Supported in part by NIH 5R01HL060708 and 8R01EB002079. Address correspondence to the author (e-mail: redelman@enh.org).
 |
ABSTRACT |
|---|
© RSNA, 2004
Index terms: Gadolinium • Iron • Magnetic resonance (MR), contrast media • Manganese • Myocardium, MR, 511.121411, 511.121412, 511.121413, 511.121416, 511.12143 • Reviews
|
INTRODUCTION |
|---|
EDITOR’S NOTE: This review is not intended to provide in-depth coverage of any topic related to contrast-enhanced MR imaging of the heart. Rather, as noted in the title, it is an overview of the literature and is intended to inform our readers of ongoing activity in this area of imaging.
Anthony V. Proto, MD, Editor
Cardiac disease remains the leading cause of death in the United States, with an estimated 1.5 million myocardial infarctions and 500 000 deaths per year attributed to coronary artery disease (CAD) (1). Despite the high prevalence of heart disease and technical limitations of alternative cardiac imaging tests, magnetic resonance (MR) imaging has played only a minor role in the cardiac work-up. There is ongoing growth in the cardiac applications of MR imaging, for instance, to evaluate myocardial viability and cardiac masses and to delineate pulmonary venous anatomy in preparation for radiofrequency (RF) ablation of arrhythmogenic foci. For certain applications, such as the evaluation of myocardial viability, MR imaging is proving to be the most accurate test, whereas in other applications (eg, detection of ventricular thrombi), it can be a helpful adjunct to standard tests such as echocardiography. Contrast agents play a key role in many of these newer applications (2–5).
This overview of the published literature considers the principles and utility of cardiac MR imaging, with an emphasis on the use of contrast media, including contrast agents used experimentally and in humans, methods for contrast agent administration and data acquisition, as well as promising new developments. A literature search covered the period from 1988 to the present with use of multiple databases, including BIOSIS, CAB, DDFU, EMBASE, MEDLINE, and SciSearch. Search terms included "heart," "cardiac," "myocardium," "magnetic resonance," and "contrast media." Only articles and scientific proceedings with at least an abstract in English were included. In addition, several scientific abstracts were included that concerned key topics where no full publications were found. Information about the type of contrast agent, dose, use of a power injector, rate of injection, imaging methods, and clinical applications was compiled.
|
GENERAL CONSIDERATIONS FOR CONTRAST-ENHANCED CARDIAC MR IMAGING |
|---|
|
CATEGORIES OF CONTRAST AGENTS FOR CARDIAC MR IMAGING |
|---|
|
|
Intravascular blood pool agents include very small coated iron particles (14,15). Examples of iron-based blood pool agents are ultrasmall superparamagnetic iron oxide or monocrystalline iron oxide (16), NC100150 (Clariscan; Amersham Pharmaceuticals, Princeton, NJ) (17), ferumoxtran-10 (Combidex; Advanced Magnetics, Cambridge, Mass) (18), and ferumoxytol (Advanced Magnetics) (19). By virtue of the prolonged enhancement of the blood pool (eg, ferumoxytol has a half-life in the blood of 14 hours), such agents may prove useful for MR angiography and cardiac cine imaging (20). In the heart, they may prove of particular value for non–breath-hold navigator-imaging methods, since the imaging times are lengthy (eg, 10–15 minutes). In addition to their benefits for MR angiography, some particulate blood pool agents can be taken up by macrophages within unstable plaques (21). This property could eventually prove useful for studies in vulnerable patients at risk for coronary plaque rupture.
Among the gadolinium chelates with mixed properties of blood pool and extracellular agents are those bound to albumin or other components of plasma (eg, MS-325; EPIX Medical, Cambridge, Mass), which thereby prolong the residence in the intravascular space. Moreover, such binding slows down the rotation of the molecule and thereby increases T1 relaxivity. One can also improve T1 relaxivity and half-life in the blood by increasing the size of the molecule and the number of gadolinium atoms. For instance, SH L 643A (Gadomer-17; Schering Pharmaceuticals, Berlin, Germany), which consists of a polymer of 24-gadolinium cascades with a molecular weight of 35 kDa, extravasates very slowly out of the intravascular space (22). Blood pool agents enhance the normal myocardium to a lesser degree than do extracellular agents because the fractional distribution volume (only about 5%–10% of the myocardial volume consists of blood) is several times smaller. This might indicate reduced utility for the evaluation of myocardial perfusion. On the other hand, in an animal model of myocardial ischemia, a cascade polymer was found to prolong the temporal window of dynamic contrast-enhanced MR imaging for the differentiation of ischemic and normal myocardium (23). The sensitivity of blood pool agents for myocardial infarction, and therefore, their potential value for evaluation of myocardial viability, is unknown.
Porphyrin derivatives labeled with gadolinium, such as bis-gadolinium mesoporphyrin (Gadophrin-2; Schering, Berlin, Germany), have an affinity for necrotic tissue, including infarcts. Authors of some studies have found that enhancement with bis-gadolinium mesoporphyrin accurately represents the infarct zone as defined with histochemical staining, whereas the area enhanced by gadopentetate dimeglumine is substantially larger. This result suggests that the periinfarct zone, and not just the infarct, enhances with gadolinium chelates (24,25). However, the eventual utility of porphyrin derivatives appears limited because of their toxicity.
Manganese ions differ from other paramagnetic contrast agents since they pass through voltage-dependent calcium channels and localize intracellularly within the myocyte. Manganese is rapidly taken up by normal myocardium, with less uptake and retention by ischemic or infarcted myocardium (26,27). The uptake pattern of manganese is analogous to that of thallium 201 (201Tl) used for stress nuclear studies. Unfortunately, free manganese is toxic. To minimize any potential toxicity, the manganese can be chelated (eg, mangafodipir trisodium). Free manganese is slowly released from the chelate and taken up by the myocardium (28). However, this agent must be given at several times the usual clinical dose to produce adequate myocardial contrast enhancement in vivo. Alternatively, manganese can be mixed with calcium ions (EVP-1001; EagleVision Pharmaceuticals, Exton, Pa) to improve tolerance. This agent, which is entering phase II clinical trials, has produced substantial myocardial contrast enhancement and, in an animal model, detection of myocardial ischemia at tolerable doses (29).
With further refinement of molecular imaging methods, there will be still greater specificity for particular anatomic structures, receptors, or cellular processes (30). For instance, ultrasmall iron particles linked to antibodies directed against cardiac myosin help detect infarcted tissue (31). The function of myocardial calcium channels could potentially be studied by using MR contrast agents that change in signal intensity on binding with calcium (32).
|
METHODS FOR CONTRAST AGENT ADMINISTRATION |
|---|
The techniques for contrast agent infusion varied widely, depending on the investigator and application. There were few reports about the choice of hand versus power injection, but the ratio was approximately 2:1. In most cases, the contrast agent was administered as a bolus, followed by a saline flush. Among 103 studies in which the dose of contrast agent was reported, less than 0.05 mmol of the contrast agent per kilogram of body weight was used in 10%, 0.05 mmol/kg was used in 18%, 0.1 mmol/kg was used in 42%, 0.15 mmol/kg was used in 2%, and 0.2 mmol/kg was used in 21% of cases. Miscellaneous infusion protocols in the remaining cases included repeated boluses of 0.1 mmol/kg, a bolus of 0.1 mmol/kg followed by a slow infusion of either 0.002 or 0.004 mmol/kg, and non–weight-based boluses of 20–40 mL. The rate of infusion was mentioned in a minority of reports and was mostly in the range of 3–5 mL/sec. A slow prolonged infusion was applied in a few instances to quantify myocardial perfusion.
|
ROLE OF CONTRAST AGENTS FOR EVALUATION OF CARDIAC ANATOMY AND FUNCTION |
|---|
Ventricular Function
Cine images shows the motion of the heart and blood over multiple phases of the cardiac cycle (40–42). Cine imaging generally uses some form of a multishot ("segmented") gradient echo. Use of a small number of phase-encoding steps in each breath hold (encompassing a small portion of the cardiac cycle, eg, 20–60 msec) minimizes blurring from cardiac motion. Accumulation of multiple segments over a typical breath-hold period of 10–20 heart beats results in a complete cine examination (43). Balanced steady-state free precession sequences (eg, true FISP [fast imaging with steady-state precession], FIESTA [fast imaging employing steady-state acquisition], or balanced FFE [fast field echo]) give the best image quality, particularly if the repetition and echo times are kept very short and there has been adequate shimming of the static magnetic field to the region of the heart (44). Projection reconstruction techniques that involve radial sampling of k-space also offer great potential for the fast acquisition of cine images with high temporal resolution (45).
Although contrast agents are not required for cine imaging, image quality generally improves after a contrast agent is administered. Cine gradient-echo acquisitions, which are inherently T1-weighted, show a marked enhancement of the blood pool after administration of a T1-active contrast agent. The enhanced blood pool appears only slightly brighter with a cine steady-state free precession acquisition, where tissue contrast depends primarily on the T2/T1 ratio. The modest amount of T1-weighting is nonetheless sufficient to help detect delayed enhancement of infarcts (46). In addition, iron oxide agents distort the local magnetic field homogeneity, so that image quality with cine imaging techniques that are sensitive to off-resonance effects (spiral imaging and, to a lesser extent, steady-state free precession) may be adversely affected.
Myocardial Tagging
Measurement of myocardial strain previously required invasive placement of markers on the heart. It is now possible to use RF presaturation (eg, spatial modulation of magnetization) (47) and harmonic phase imaging (48) to observe local myocardial motion and measure mechanical properties such as strain. The persistence of the tags is T1-dependent. Because paramagnetic contrast agents shorten the T1 relaxation time of the myocardium and hence reduce the persistence of the tags, tagging measurements should be performed prior to the administration of a contrast agent.
Perfusion MR Imaging
One can assess myocardial perfusion by using a variety of means. For instance, one can use an intrinsic contrast mechanism such as blood oxygen level–dependent, or BOLD, rather than an exogenous contrast agent (49). On the basis of changes in tissue oxygenation manifested on T2*-weighted images, BOLD MR imaging can depict a stenotic coronary artery (50) and, along with measurement of wall thickening, may help differentiate scar from viable myocardium and may help identify patients who would be suitable candidates for revascularization (51). However, these studies are preliminary and have substantial technical limitations, such as low contrast-to-noise ratios that are inherent to the BOLD technique.
Dynamic first-pass imaging.—Dynamic first-pass contrast-enhanced MR imaging is the most practical method for evaluating myocardial perfusion (52–56). From a theoretical point of view, the use of an intravascular contrast agent simplifies the modeling of tissue perfusion since one does not need to model tissue extravasation, but such agents are not yet available for routine clinical use. Nonetheless, accurate measurements are also possible with use of extravascular agents that are widely available, even though 30%–50% of the agent leaks out of the vascular bed during the first pass (57). Close correlation exists between microsphere and MR imaging measures of myocardial perfusion in animal studies with gadopentetate dimeglumine as a contrast agent (58). In patients, there is good correlation between the perfusion reserve with MR imaging and the coronary flow reserve with Doppler ultrasonography (r = 0.80) (59). For quantification of myocardial perfusion, a low dose of contrast agent (eg, 0.025–0.05 mmol/kg) may improve accuracy, whereas a dose of 0.1–0.2 mmol/kg gives better myocardial enhancement and image quality. High infusion rates (eg, 3–5 mL/sec) are typical, followed by a saline flush. Perfusion MR imaging, with use of a pharmacologic agent to induce vasodilatation, is sensitive to flow variations over a wide range of low- and high-flow states (60).
Pulse sequences for dynamic first-pass imaging.—With interleaved gradient-echo echo-planar imaging (61) (essentially a gradient-echo pulse sequence that uses an abbreviated echo-planar read-out of the signal) or an inversion-recovery–prepared fast gradient-echo sequence (snapshot fast low-angle shot or turbo fast low-angle shot) (62,63), one can distinguish perfused tissue in the endocardial and epicardial layers of the myocardium. Recently, the saturation-recovery true fast imaging with steady-state precession sequence proved to give superior image quality, since it makes the most efficient use of the available magnetization (64).
Quantification of perfusion.—The quantification of tissue perfusion derives from the Kety model (65). The procedure is mathematically complex and involves a number of assumptions, including parameters such as water exchange rates, contrast agent distribution, and the relationship between signal intensity and the concentration of contrast agent. Qualitative measures such as the maximum myocardial enhancement, transit time, or upslope of myocardial enhancement (ie, slope of the initial portion of the signal intensity versus time curve after gadolinium administration) (66) are easier to implement in clinical practice. Even simple visual evaluation of the images can be sufficient. For instance, in one study of 104 patients without prior myocardial infarction in which a receiver operating characteristic analysis of myocardial perfusion was used, MR imaging had a 90% sensitivity for depicting at least one coronary artery with significant stenosis and an 85% specificity in identification of patients with significant coronary artery stenoses (67). The authors found that stress enhancement at dynamic MR imaging correlated more closely with quantitative coronary angiography results than did stress enhancement at single photon emission computed tomography (SPECT). Because of the reliance on visual analysis, this study used a relatively large contrast agent dose of 0.075 mmol/kg infused at 4 mL/sec to improve the contrast-to-noise ratio.
Resting perfusion.—Imaging of resting perfusion is only moderately sensitive in CAD. For instance, a study of 12 patients with significant stenoses of major epicardial vessels demonstrated a sensitivity of only 77% (68). Lesser stenoses can prove undetectable at rest. Nonetheless, resting perfusion studies may have some uses. For instance, underperfused regions of myocardium in resting studies correlate with increased amounts of nonviable tissue, as detected with myocardial delayed-enhancement MR imaging (69). Early after acute myocardial infarction, the resting perfusion deficit correlates with the long-term severity of left ventricular functional abnormality, and one finds higher perfusion in segments that have a residual contractile reserve at stress echocardiography (70).
Stress perfusion.—To detect significant stenoses of the epicardial coronary arteries, it is helpful to stress the patient so that a measure of the coronary flow reserve, or similar measures such as the myocardial perfusion reserve index, can be obtained. Though technically possible, it is awkward to perform a physiologic exercise stress test within the confined bore of an MR imager. Instead, a pharmaceutical agent, such as dipyridamole (typical dose, 0.56 mg per kilogram of body weight) or shorter-acting adenosine (typical dose, 140 µg/kg/min), may be administered to induce coronary vasodilatation. Side-effects of dipyridamole may necessitate administration of aminophylline, whereas most side-effects of adenosine are rapidly terminated when the infusion is discontinued. The safety profile and more consistent course of action make adenosine the preferred agent for stress perfusion MR imaging.
An abnormal perfusion reserve at MR imaging helps distinguish patients with CAD from normal subjects (71). In a study of 34 patients with a stenosis of an epicardial coronary artery of at least 75%, a cutoff value of 1.5 for MR imaging perfusion reserve helped differentiate normal from ischemic myocardial segments (72). This cutoff value yielded high sensitivity (90%), specificity (83%), and diagnostic accuracy (87%) for CAD, with excellent inter- and intraobserver agreement. The perfusion reserve was determined by means of a linear fit of the upslope of the signal intensity–time curves in perfusion studies performed at rest and after dipyridamole infusion.
Role in revascularization.—Perfusion MR imaging demonstrates the effectiveness of coronary interventions. In a study of 35 patients with single- and multivessel CAD imaged within 24 hours of coronary revascularization, a myocardial perfusion reserve index (upslope index) was calculated from resting and stress perfusion MR imaging (73). The perfusion reserve index improved but did not normalize after successful revascularization. The improvement was greater in patients receiving stents than in those only undergoing angioplasty.
Comparisons with PET and SPECT.—The coronary flow reserve determined at contrast-enhanced MR imaging correlates well with that at nitrogen 13 ammonia positron emission tomography (PET) (74). Flow reserve values at MR imaging are lower than those at PET, in part, owing to a low extraction fraction for extracellular agents such as gadopentetate dimeglumine.
Findings of several studies have confirmed a sensitivity and specificity of stress perfusion MR imaging equivalent or superior to those of SPECT. In the literature, sensitivity and specificity values of MR imaging range 64%–92% and 71%–100%, respectively (75–80). MR imaging would therefore appear to be a reasonable alternative to SPECT for the evaluation of patients suspected of having CAD, with the additional advantages of better depiction of wall motion, cardiac morphology, and myocardial viability. Nonetheless, MR imaging has yet to make much impact into routine clinical practice, in part because SPECT or dobutamine stress echocardiography can generally provide the needed diagnostic information, and because of a lack of the large-scale multicenter trials that would validate any potential superiority of MR imaging.
Microvascular Integrity
Whereas angiographic methods help evaluate the epicardial coronary arteries, events at the level of the microvasculature (arterioles, capillaries, and venules) are beyond the spatial resolution of standard imaging methods. For instance, one can recanalize an occluded coronary artery, yet within the capillary there may be a persistently diminished blood flow because the microvasculature remains plugged by red blood cell stasis (81). Other possible causes of microvascular obstruction include myocardial edema and endothelial cell damage from free radical formation. The "no-reflow" phenomenon, which indicates lack of reperfusion from microvascular impairment at the core of a reperfused infarct, presents as a subendocardial region of persistent hypoenhancement (82,83). Since flow at the core is very low but not zero, the hypoenhanced regions eventually become hyperenhanced.
In a study of 22 patients with acute myocardial infarction, contrast-enhanced imaging performed a few minutes after contrast agent infusion showed subendocardial hypoenhancement inside hyperenhancing myocardium in nearly half of the patients, which is consistent with no reflow and microvascular obstruction (84). Microvascular obstruction may indicate a worsened prognosis (85) and may predict a larger amount of adverse left ventricular remodeling (86). With respect to the prediction of adverse left ventricular remodeling, the optimal time for assessment of microvascular obstruction was 1 day after coronary angioplasty (87). Microvascular obstruction is also associated with an increased incidence of intramyocardial hemorrhage (88) and is more common after angioplasty than thrombolysis with or without angioplasty (89).
Syndrome X.—Syndrome X refers to a heterogeneous group of patients with anginal symptoms but normal epicardial coronary arteries at angiography and no evidence of coronary artery spasm (90). Although the origin is unknown, one possibility is an ischemic origin related to endothelial dysfunction at the microvascular level. Patients with syndrome X demonstrate a reduced myocardial perfusion reserve at contrast-enhanced MR imaging (91,92). Moreover, a study of 24 patients with syndrome X demonstrated delayed subendocardial gadolinium enhancement in two-thirds of the subjects (93). More interesting, with successful relief of symptoms by using ß-blockers, the delayed contrast enhancement disappeared in most of the patients. The investigators suggested that the enhancement might have been caused by alterations in cell membrane permeability related to the frequent ischemic episodes rather than to the irreversible consequences of infarction.
Myocardial Viability
Infarcts occur when there has been prolonged occlusion (eg, >30–40 minutes) of an epicardial coronary artery without sufficient collateral blood supply to the affected myocardium. The measurement of the amount of scar tissue helps to prognosticate whether patients will functionally improve in response to coronary revascularization therapy. Morphology (eg, wall thickness) alone is inadequate. Standard methods for making this distinction include the use of reinjection 201Tl imaging and measurement of wall-motion and systolic thickening with echocardiography.
Dobutamine stress testing.—Viability can be assessed without the need for contrast media by performing cine MR imaging during a low-dose infusion (5–10 µg/kg/min) of dobutamine, which is comparable to dobutamine stress echocardiography. Improved left ventricular wall thickening with stress in a segment that functions poorly at rest indicates viability (94). A comparison among dobutamine stress transesophageal echocardiography, dobutamine cine MR imaging, and fluorine 18 (18F) fluorodeoxyglucose PET in 43 patients with chronic infarction and wall-motion abnormalities resulted in a respective sensitivity and specificity of 77% and 81% and 94% and 100% for echocardiography and MR imaging (95). Criteria for viability were a resting wall thickness greater than 5.5 mm or wall thickening of at least 1 mm with stress. Dobutamine stress MR imaging can also be used to detect ischemic myocardium (96).
In patients with nondiagnostic echocardiograms, dobutamine stress cardiac MR imaging may have prognostic value (97). The presence of inducible ischemia or left ventricular ejection fraction under 40% was associated with future myocardial infarction or cardiac death, independent of the presence of risk factors for coronary arteriosclerosis.
Delayed enhancement.—Myocardial delayed enhancement, also called "late enhancement," "delayed enhancement MR imaging," "delayed hyperenhancement," and so forth, is the most accurate means to detect myocardial infarction (Fig 1). Delayed uptake of contrast agents within infarcts was first described more than 10 years ago (98), but there have been extensive experimental studies, improved pulse sequences, and recognition of the clinical importance of the method within the past few years (82,99,100). Cellular degradation in the infarcted region results in an increase in the permeability and enlargement of the extravascular space, and hence, an increased distribution volume for the extracellular contrast agent. Moreover, gadolinium chelates wash out of infarcted tissue more slowly than out of healthy myocardium. The net result is that infarcted regions appear bright on delayed contrast-enhanced T1-weighted images. The size and location of the infarcted region, as demonstrated histochemically in animal models, correlate with the size and location of myocardial delayed enhancement.
|
|
Techniques for contrast-enhanced imaging of myocardial viability.—The current technique involves the rapid infusion of a gadolinium chelate (doses are typically in the range of 0.1–0.2 mmol/kg) followed approximately 5–30 minutes later by a high-resolution cardiac-gated T1-weighted pulse sequence (101). Imaging too early (eg, <5 minutes after contrast agent infusion) may result in an underestimation of the infarcted region, whereas imaging too late (eg, after >30 minutes) may result in excessive washout of the contrast agent and a poor SNR. Animal (102) and human (103) study findings of acute myocardial infarction showed a substantial change in the portion of left ventricle that enhanced depending on the timing of the acquisition with respect to the contrast agent infusion, with overestimation of infarct size if imaging was performed too early. However, there is also evidence that the spatial extent of hyperenhancement does not vary substantially over the 5–30-minute period after contrast agent infusion so long as one selects an appropriate inversion time to minimize the signal intensity of viable myocardium (104).
Although rarely used, an alternative approach involves infusing the contrast agent at a constant rate (105,106). The constant infusion maintains an equilibrium concentration of the contrast agent in the blood and enables quantitative measurement of the tissue-blood partition coefficient, 
. This parameter increases in infarcts owing to loss of integrity of the cell membrane. The method eliminates the influence of blood flow, regional perfusion, and renal excretion.
The typical pulse sequence for myocardial delayed enhancement is a multishot inversion-recovery–prepared gradient-echo sequence. T1-weighted images are acquired 5–30 minutes after contrast agent infusion (107). Each breath hold yields one or a few sections. Choice of the appropriate inversion time (approximately 200 msec) to null the signal intensity of normal myocardium is critical for accurate delineation of the infarcted region. The healthy myocardium appears dark, whereas the enhancing myocardium appears bright. However, the optimal inversion time will lengthen over time as the concentration of gadolinium in the blood and myocardium gradually diminishes. Phase-sensitive reconstructions render the technique less sensitive to the choice of inversion time and reduce the variation in apparent infarct size (108). One study suggested the use of a very short inversion time (eg, 50–100 msec) to improve contrast between infarct and blood pool (109). Single-shot versions of the pulse sequence appear promising since they do not have motion sensitivity; alternatively, one can image the entire heart in a single breath hold by using a volumetric acquisition (110).
"Bright is dead" hypothesis.—It is widely accepted that delayed enhancement in the proper location and given the appropriate clinical setting represents infarction, but some controversy remains (111). For instance, there is some contrary evidence that myocardial delayed enhancement in the acute postinfarct phase may cause overestimation in the amount of irreversibly damaged tissue. In one study, 18 patients (1015 segments) were studied within 2 weeks and again 3 months after reperfused acute myocardial infarction (112). Of 324 segments with impaired function initially, 180 improved at follow-up. Of the segments with improved function at follow up, 62% showed transmural enhancement involving more than 50% of each segment. The sensitivity and specificity for myocardial delayed enhancement in predicting functional recovery were only 38% and 88%, respectively. Another study of 10 patients suggested that one should combine myocardial delayed enhancement with wall-motion analysis and first-pass enhancement during dobutamine infusion to maximize sensitivity and, to a lesser extent, specificity (113). Nonetheless, the majority of published study findings indicate that myocardial delayed enhancement accurately depicts the extent of nonviable myocardium.
Comparison with other imaging modalities.—Results of myocardial delayed enhancement correlate well with those of dobutamine stress echocardiography (114,115), and there is excellent agreement with PET as well. Delayed enhancement correlates with areas of decreased flow and metabolism at PET (116). However, MR imaging is slightly more sensitive—in one study, 11% of segments called viable with PET showed delayed enhancement with MR imaging. Another study of 26 patients demonstrated a 96% sensitivity and 84% specificity for myocardial delayed enhancement, with 18F fluorodeoxyglucose PET as the standard (117). In particular, MR imaging is superior in detecting subendocardial infarction (118). There is also good agreement with SPECT (119), but myocardial delayed-enhancement MR imaging has the major advantage of superior spatial resolution by an order of magnitude and the capability to incorporate anatomic and cine imaging in the same imaging session. In a study of 91 patients, SPECT depicted all of the nearly transmural infarctions but missed 47% of subendocardial infarctions that were seen at myocardial delayed-enhancement MR imaging (120). Authors of another study of 20 patients with equivocal stress-rest sestamibi SPECT examination findings found the presence of subendocardial infarction at myocardial delayed-enhancement MR imaging in 40% of the patients (121).
Prognostic implications of myocardial delayed enhancement.—Findings of several studies have shown that the amount of delayed transmural enhancement predicts the degree of functional recovery after acute myocardial infarction (122,123). Extensive transmural myocardial delayed enhancement is highly predictive of a lack of functional improvement after revascularization; conversely, absence of myocardial delayed enhancement correlates with a likelihood of functional recovery (124). In a study of 32 patients with acute myocardial infarction, subjects were imaged within 2 weeks and again after 1 year by using T2-weighted contrast-enhanced first-pass and myocardial delayed-enhancement MR imaging (125). The presence of a high-grade perfusion deficit or delayed enhancement in the early-phase MR imaging study is highly predictive of scar formation and lack of functional recovery at 1 year. Increased signal intensity on T2-weighted images indicates myocardial edema but does not always indicate infarction (126), that distinction requires the additional use of myocardial delayed enhancement. In another study of acute myocardial infarction, the investigators imaged 20 patients within 1 week of the acute event and 7 months later (127). Enhancement patterns correlated with regional circumferential shortening strain (a measure of myocardial function) as determined with the harmonic phase imaging technique. Absence of myocardial delayed enhancement had a positive predictive value of 77% for functional recovery, whereas presence of myocardial delayed enhancement had a negative predictive value of 66%. The authors concluded that, compared with the lack of early hypoenhancement, lack of delayed hyperenhancement is more accurate in predicting functional improvement in dysfunctional segments. The early hypoenhanced regions, corresponding to regions with microvascular obstruction, resulted in substantial underestimation of the amount of irreversibly injured myocardium after acute myocardial ischemia. In the chronic setting, myocardial delayed enhancement correlates inversely with circumferential shortening (128).
History of myocardial infarction greatly increases the mortality rate compared with that of the general population. Although the electrocardiogram is a useful indicator of prior infarction, most infarcts are not associated with the formation of Q waves. MR imaging may therefore be useful in detection of unsuspected myocardial infarcts. In a study of 82 subjects, myocardial delayed enhancement helped predict the presence, location, and transmural extent of healed Q-wave and non–Q-wave myocardial infarction (129).
Myocardial delayed enhancement and myocardial stunning.—Transient hypoperfusion can cause myocardial stunning, which is associated with wall-motion abnormalities in the clinical setting of suspected acute myocardial infarction. Myocardial delayed enhancement in combination with cine MR imaging helps differentiate wall-motion abnormalities of myocardial stunning, which are reversible, from those of myocardial infarction, which are often irreversible depending on the severity of the injury (130). Either condition may cause a wall-motion abnormality, but delayed enhancement occurs only with infarcts. Lack of delayed enhancement indicates stunning rather than infarction and a high likelihood that left ventricular function will fully recover (131). In a study of 30 patients imaged approximately 1 week and 13 weeks after a reperfused myocardial infarction, the likelihood of functional improvement of segments without hyperenhancement was 3, 14, and 20 times higher than that of segments with 26%–50%, 51%–75%, and more than 75% hyperenhancement, respectively. The likelihood of complete functional recovery of segments without hyperenhancement was 3.8, 11.1, and 50.0 times higher than that of segments with 26%–50%, 51%–75%, and more than 75% hyperenhancement, respectively (132). Thus, functional improvement of stunned myocardium is predicted with myocardial delayed-enhancement MR imaging. However, in situations where an intermediate level of myocardial delayed enhancement is present (likely representing a mix of viable and nonviable tissue), the predictive value is less certain.
Triage of patients with chest pain.—Patients who come to the emergency room with chest pain often represent a diagnostic dilemma, since a substantial number of patients with a noncardiac cause of pain can usually be sent home, whereas those with acute coronary syndrome require observation or treatment. One study of 161 patients with more than 30 minutes of chest pain but a nondiagnostic electrocardiogram found that the combination of myocardial delayed enhancement, cine, and perfusion MR imaging was the strongest predictor of CAD and added diagnostic value over clinical parameters (133).
Diagnosis of ventricular aneurysms.—In addition to impairment of left ventricular function and arrhythmias, complications of myocardial infarction include true and false aneurysms. True aneurysms, which are composed of pericardium adherent to underlying epicardium and scar tissue from infarcted myocardium, do not usually rupture. However, false aneurysms, which consist of pericardium that contains a ruptured left ventricle, may enlarge over time and require surgical resection (134). MR imaging can help make this distinction based on morphologic criteria (eg, a wide mouth and anterior location for true aneurysm) (135). A case report demonstrated delayed enhancement of scarred myocardium within the aneurysm wall, which indicated a true aneurysm (136). The wall of a false aneurysm, which does not contain myocardial scar tissue, should not (at least in theory) demonstrate a comparable pattern of enhancement.
Right ventricular infarction.—Right ventricular infarction occurs commonly in patients with ostial right coronary artery occlusion, but the diagnosis is elusive. Delayed enhancement MR imaging is likely the best technique for the noninvasive identification of patients with this syndrome (137).
Cardiomyopathy and Myocarditis
Morphologic features and the presence or absence of contrast enhancement can be helpful in the detection and characterization of various types of cardiomyopathy and myocarditis. Focal myocardial enhancement in conjunction with a wall-motion abnormality correlates strongly with myocarditis (138), which may be of viral or other origin. The contrast-enhancement pattern becomes diffuse over a period of days to weeks (139). One case report described a patient who developed a right ventricular pseudoaneurysm owing to a presumed viral myocarditis. Portions of both the right and left ventricles showed transmural contrast enhancement (140). Contrast-enhanced MR imaging can be helpful for other infectious processes as well (141,142). Another kind of myositis, hypereosinophilic syndrome, is characterized by peripheral eosinophilia and multiorgan dysfunction owing to eosinophilic infiltration. Contrast-enhanced MR imaging depicts myocardial involvement (143) and ventricular thrombi that may occur in this condition (Fig 2).
|
|
|
Right ventricular dysplasia.—Although most ventricular tachycardias originate from the left side of the heart, the right ventricle may also be a source. The cause can be right ventricular dysplasia (148), which has a poor prognosis, or right ventricular outflow tract tachycardia, which is more benign but can cause sudden death. The diagnosis is made on the basis of major and minor criteria involving family history, genetic factors, arrhythmias, conduction and repolarization abnormalities, biopsy, and right ventricular morphology and function (eg, marked dilatation or dyskinetic areas).
The origin of arrhythmia is commonly determined with an invasive electrophysiologic study. MR imaging can be helpful in problematic cases, sometimes demonstrating abnormal wall motion in the body or outflow tract of the right ventricle and, less commonly, fatty replacement of the myocardium (149). Although contrast enhancement is not commonly used, a recent report suggests that delayed enhancement of the right ventricular myocardium is more common than was previously appreciated and may have diagnostic value (150).
Atrial arrhythmias.—Atrial fibrillation is commonly associated with arrhythmogenic foci located near the pulmonary vein orifices. Contrast agents are helpful in MR angiography of the pulmonary veins, which is used to guide RF ablation procedures and detect complications such as pulmonary vein stenosis (151).
Sarcoidosis.—In advanced sarcoidosis, echocardiography can help detect septal thinning, systolic and diastolic left ventricular dysfunction, and pericardial effusion. Electrocardiographic changes are frequent. However, diagnosis of early sarcoidosis is more challenging. Endomyocardial biopsy can provide the diagnosis but is invasive and has sampling errors. In a study of 16 patients with cardiac sarcoidosis, contrast enhancement was present in half of the cases and diminished after steroid therapy (152). Another study suggested that MR imaging is specific but has low sensitivity for sarcoid involvement of the myocardium (153).
Amyloidosis.—Amyloidosis is a cause of restrictive cardiomyopathy and can be primary or secondary in origin. Infiltration with fibrillar proteins causes a loss of compliance and impairs diastolic function, and systolic function can be reduced as well. Echocardiography may demonstrate a "sparkling" pattern, whereas both echocardiography and MR imaging can show functional impairment and biventricular hypertrophy (154). Inhomogeneous contrast enhancement may occur but is not specific. Some preliminary data suggest that ventricular contrast enhancement may also occur within hypertrophic regions in certain forms of glycogen storage disease, but this pattern is again nonspecific (155).
Dilated cardiomyopathy.—Delayed enhancement MR imaging may help differentiate the underlying origin of dilated cardiomyopathy. In a study of 90 patients with dilated cardiomyopathy, 59% showed no enhancement, whereas myocardial delayed enhancement occurred in 41% of subjects (156). Although in 13%, subendocardial enhancement was indistinguishable from infarction, in 28%, midwall myocardial enhancement (presumably associated with fibrosis) permitted differentiation from an ischemic origin. The authors suggest that use of coronary catheterization, rather than MR imaging, to decide if left ventricular dysfunction was caused by CAD would have led to the incorrect assignment of a diagnosis of dilated cardiomyopathy in 13% of their patients. The degree of enhancement correlates with the severity of functional abnormality (157). In patients undergoing chemotherapy with cardiotoxic agents such as anthracyclines, which can cause dilated cardiomyopathy, an early (day 3) increase in myocardial contrast enhancement was predictive of future worsening in left ventricular function (158).
Cardiac Masses, Thrombi, and Infections
Although echocardiography and CT are often sufficient for the evaluation of cardiac and paracardiac masses, MR imaging can provide useful information in many instances (159). Echocardiography is an excellent method for evaluation of left-sided structures but is less accurate for the right side of the heart, paracardiac region, lungs, and mediastinum. Contrast-enhanced MR imaging can help distinguish tumors (160–165), which generally enhance, from thrombi. Along with cine and T2-weighted images, the standard perfusion and myocardial delayed-enhancement sequences can be very helpful in the evaluation of the contrast enhancement characteristics of cardiac masses.
Preliminary data suggest that myocardial delayed-enhancement MR imaging may be more accurate in detecting left ventricular thrombi than is transthoracic echocardiography (166). In a study of 24 patients, MR imaging was more sensitive than transesophageal echocardiography in detecting intracardiac thrombi (167). Some thrombi showed delayed contrast enhancement, which characterized them as organized clots. However, care must be taken to use thin sections (eg, 5 mm or thinner) for the detection of small thrombi, and the entire region of interest (eg, left atrial appendage or left ventricular apex) must be included in the imaging volume.
Myxomas, the most common benign tumors of the heart, are well evaluated with transthoracic and transesophageal echocardiography (168). However, MR imaging can occasionally provide helpful information about the attachment site and the relationship to key anatomic structures (Fig 3). Myxomas often appear bright on spin-echo images and dark on gradient-echo cine images and may show uniform or inhomogeneous contrast enhancement (169).
|
|
|
CORONARY MR ANGIOGRAPHY |
|---|
Noninvasive imaging of the coronary arteries, along with quantification of flow (181), would provide obvious benefits. Challenges in coronary artery angiography include inadequate spatial resolution, SNR, and contrast enhancement (182). Findings of a study in which a breath-hold single-section two-dimensional acquisition was used demonstrated good sensitivity and specificity for proximal CAD (183), but subsequent study findings with two-dimensional acquisitions were less sanguine (184). Better results are obtained with volumetric acquisitions that cover a large portion of the coronary artery tree (185).
Navigator-gating methods eliminate the need for breath holding by electronically monitoring the diaphragm location (186,187). Data accumulate only during a specified portion of the respiratory cycle, and the total duration of data acquisition is typically about 5–15 minutes. Navigator-gating methods improve the SNR and in-plane spatial resolution (eg, 0.5 mm). A recent multicenter trial in which navigator techniques without contrast enhancement were used found a moderately high sensitivity and specificity of coronary MR angiography (188). The sensitivity, specificity, and accuracy for patients with disease of the left main coronary artery or with three-vessel disease were 100%, 85%, and 87%, respectively. The negative predictive values for any CAD and for left main artery or three-vessel disease were 81% and 100%, respectively.
Recently, several groups have applied breath-hold three-dimensional acquisitions in combination with bolus administration of a paramagnetic contrast agent (189–191). Potential benefits include improved SNR, less sensitivity to artifacts from slowly flowing blood, and reduced imaging times. Authors of one study found that contrast-enhanced breath-hold three-dimensional MR angiography was more accurate than respiratory-gated three-dimensional MR angiography for the coronary arteries (192). Extracellular contrast agents are less useful for lengthy navigator-gated acquisitions, since the contrast agent concentration in the blood diminishes so rapidly. The concentration of blood pool agents, on the other hand, is nearly constant for minutes to hours after infusion, and therefore, might be beneficial (193–195). For instance, authors of one study in which breath-hold and navigator-gated sequences were used found that the administration of SH L 643A improved contrast-to-noise ratios in the coronary arteries up to 30 minutes after infusion (196).
Coronary MR angiography can demonstrate coronary artery aneurysms in patients with Kawasaki disease. Wall thinning and delayed contrast enhancement occur less consistently, despite clinical evidence of myocardial infarction, than is typically observed in adult patients with infarction (197).
Contrast-enhanced three-dimensional MR angiography, either with or without cardiac gating, demonstrates the patency of coronary artery bypass grafts (198–201), though the accuracy is currently less than that of CT (202–204).
|
FUTURE DEVELOPMENTS |
|---|
Interventional MR Imaging of the Heart
In the future, contrast-enhanced MR imaging may be used to guide interventions. For instance, MR imaging can be used in experimental systems to monitor, in real time, electromagnetic catheter-directed RF ablations (eg, for treatment of atrial fibrillation) within the heart (210). Contrast media are helpful for improving the SNR of volumetric cardiac images upon which the movements of the catheter tip are mapped. The method has the potential to provide more precise positioning of the site of RF ablation and, of course, it eliminates radiation exposure to the operator.
Several new treatments for heart failure are under development that involve the transplantation of primitive cell types (eg, stem cells or myoblasts) into the damaged tissue, where they can trans-differentiate into functional tissue (211). In a study of seven patients with severe heart failure, cardiac MR imaging was used to plan therapy involving direct injection of myoblasts (harvested from leg muscle and expanded in vitro) into segments with a diastolic wall thickness greater than 4 mm and transmural infarction as indicated with myocardial delayed enhancement (212). Stem cell therapy in the heart might further benefit from the combination of molecular imaging methods with interventional MR imaging techniques. In animal models, mesenchymal stem cells have been labeled with iron-based contrast agents to determine myocardial localization after implantation (213,214). This technology may allow in vivo evaluation of stem cell retention, engraftment, and migration. In another animal study, MR imaging was able to accurately localize the targeted catheter-based implantation of iron-loaded myogenic precursor cells into locally infarcted left ventricular myocardium (215).
|
CONCLUSION |
|---|
The utilization of contrast agents for the heart will undoubtedly expand in the coming years. Promising developments include receptor-targeted contrast agents, imaging of MR-guided interventions that include primitive cell transplants for heart failure, and imaging at field strengths higher than 1.5 T.
|
ESSENTIALS |
|---|
Unlike the pulse sequences used for contrast-enhanced MR imaging of other organs, a cardiac-gated pulse sequence with an inversion recovery magnetization preparation is used to maximize sensitivity for the T1 effects of contrast agents in the heart.
Contrast-enhanced MR imaging is the most accurate imaging modality for determining myocardial viability; in particular, it is more sensitive for subendocardial infarcts than SPECT, PET, or dobutamine stress echocardiography.
Although tissue-specific contrast-agents are not yet approved for cardiac MR imaging, they have the potential to improve the depiction of cardiac morphology, quantify myocardial perfusion, characterize vulnerable plaque, and improve MR angiography of coronary arteries.
|
FOOTNOTES |
|---|
Abbreviations: CAD = coronary artery disease, RF = radiofrequency, SNR = signal-to-noise ratio
|
REFERENCES |
|---|
TOP
ABSTRACT
INTRODUCTION
