Applied Radiology

Dr. Heasley is a third-year Radiology Resident at the Johns Hopkins Hospital, Baltimore, MD. He received his MD from the University of Virginia School of Medicine in 1999, and will begin a Fellowship in Neuroradiology at Johns Hopkins Hospital in 2004. Dr. Bluemke is an Associate Professor and Clinical Director of MRI in the Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD.

Chronic myocardial ischemia is a major cause of cardiac injury, often leading to permanent dysfunction and mortality. However, in some cases, injured myocardium remains "viable": it will recover function if blood flow to it is improved. Identification of viable myocardium has prognostic and therapeutic implications in the management of chronic ischemic heart disease. Recent advances in imaging sequences for contrast-enhanced cardiac magnetic resonance (MR) imaging now allow for identification of the characteristic delayed enhancement in scar tissue with high resolution. This offers several advantages in assessment of viability over stress echocardiographic and nuclear medicine techniques currently in use. These include the fact that MR imaging can be performed at rest and its ability to distinguish subendocardial from transmural scar.

Diseases of the heart are the leading cause of mortality in the United States. ¹ Of these, chronic ischemic heart disease is the most common, and is responsible for more deaths annually than is acute myocardial infarction. Cardiomyopathy results when chronic myocardial ischemia leads to ventricular injury and dysfunction. The prognosis of ischemic cardiomyopathy is poor, with 5-year survival rates of only 50% to 60%. ²

Fortunately, in some patients, injured myocardium will recover function if blood flow to it is improved. This tissue is referred to as viable myocardium. Noninvasive methods of identifying viable myocardium are valuable, as they allow clinicians to target revascularization efforts for maximum benefit to the patient. ³ Recent advances in MR imaging research have enabled the radiologist to distinguish viable tissue from nonviable tissue with a speed and precision unmatched by other modalities. This article discusses the clinical assessment of myocardial viability, with emphasis on contrast-enhanced cardiac MR imaging techniques.

Hibernating myocardium

Chronic ischemia of the myocardium may lead to either tissue necrosis and scar formation or, if less severe, to induction of myocardial "hibernation." Hibernation is a self-protecting strategy of myocytes in which the function of a muscle segment is down-regulated until it matches the diminished blood supply. ⁴ Histologically, decreases in cellular contractile apparatus and glycogen stores have been documented in biopsies of hibernating myocardium. ⁵ This decreased contractility results in diminished ejection fraction and cardiac function. In effect, hibernating myocardium may simulate infarcted tissue, with the crucial distinction that it can regain some or all of its lost activity if revascularized. Its existence was inferred from the reversal of cardiac wall-motion abnormalities after coronary artery bypass graft surgery. Viability assessment began when areas of functional improvement were predicted preoperatively by findings on positron emission tomography (PET). ^6,7 Hibernation is distinct from myocardial "stunning," which refers to transient myocardial dysfunction after acute ischemia is relieved, though the two may coexist. ⁴

Extensive research has shown the therapeutic and prognostic value of identifying hibernating, but viable, myocardium. Patients with substantial regions of viability fare better with revascularization than with medical therapy alone. ^8-11 Viable tissue portends better survival in the perioperative period and later, greater exercise tolerance following revascularization, and improved measures of left ventricular function. ^9,11-13 The degree of recovery is greater when larger zones of viable myocardium are detected. ^14,15

MR identification of viable myocardium

Viability has been identified by evidence of myocardial metabolism on PET, ¹⁶ by delayed fill-in of myocardial perfusion defects on single positron emission computed tomography (SPECT) imaging, ¹⁷ or by evidence of preserved contractile reserve as assessed by dobutamine stress echocardiography. ¹⁸ These will be discussed briefly below.

Contrast-enhanced MR imaging identifies viable tissue by taking advantage of an unusual feature of myocardial enhancement after infarction. Beginning several minutes after gadolinium-chelate contrast injection, contrast agent that has accumulated in the normal myocardium begins to "wash out." However, the wash-out of gadolinium-chelate from nonviable myocardium is delayed. Thus, there is progressive relative enhancement of chronically infarcted myocardium. ¹⁹ This is true regardless of infarct age, extent of resting wall motion, or reperfusion status. ²⁰ The mechanism of this delayed enhancement is not entirely clear, but it appears to be due to differing volumes of distribution. In chronic myocardial infarction (older than 4 weeks), there is progressive replacement of myocardium by fibrous scar, which has increased interstitial space compared with normal myocardium. ²¹ Gadolinium-chelate contrast agents are extracellular. Presumably they diffuse more slowly through tissues with larger interstitial volumes. This may explain the slower wash-out of these agents from areas of scar than from areas of intact myocardium. The delayed enhancement, which can persist for up to 40 minutes, forms the basis for distinguishing infarct from viable tissue using MR imaging. ^22,23

Clinically, if a myocardial segment is dysfunctional on wall-motion imaging but does not display delayed enhancement, it is viable. ²⁴ The sensitivity of nonenhancement for viability in the setting of chronic infarction was 98% in one study. ²⁵ However, early research on the utility of contrast-enhanced MR imaging in assessing viability was muddied both by conventional MR pulse sequences, which resulted in poor visualization of enhancement, and by the observation that some cardiac segments with nontransmural delayed enhancement also show improved function after revascularization. ^18,26

The central role of MR imaging for infarct delineation has been clarified by the introduction of electrocardiogram-segmented fast inversion recovery gradient-echo pulse sequences. Similar sequences have been commercially distributed for several years, though they were not initially applied to cardiac imaging. The resolution and contrast-to-noise ratio they provide for infarct imaging is superior to spin-echo, steady-state free precession (SSFP), or short-T1 inversion recovery (STIR) sequences. ²⁷ One important aspect of these sequences is the use of a segmented k-space approach, so-called because several lines (segments) of k-space are acquired during each cardiac cycle, rather than a single line. When combined with short TEs (2 msec) and TRs (<10 msec), this allows high-resolution, breath-hold images of the heart to be obtained. ²⁸ Also important is the selection of an inversion time, TI, that nulls normal myocardium. This is achieved by performing a preliminary sequence using a range of inversion times and then choosing the optimal parameter for a particular patient from among the range of resulting images. Myocardial viability imaging with MR is often termed myocardial delayed enhancement (MDE) imaging. This term reflects the improved delineation of nonviable segments at 10 to 40 minutes following gadolinium-chelate contrast injection.

Kim and colleagues ²⁹ used MDE imaging to prospectively assess viability in a series of 42 patients with ischemic cardiomyopathy. Cardiac wall motion abnormalities were evaluated by cine MR imaging, and the degree of delayed transmural enhancement in each segment was stratified (eg, none, 1% to 25%, 75% to 100%). Using post-revascularization improvement in function as the gold standard, these investigators found that lesser transmural enhancement in a dysfunctional segment predicted increasing likelihood of improved wall motion after revascularization.

A particular strength of this study was that the positive predictive value for recovery increased to 88% when analysis was limited to severely dysfunctional segments. Such segments are typically the most difficult to evaluate with other modalities. ³⁰ The findings were also in concordance with myocardial biopsy studies that showed that the percentage of transmural scar correlates with degree of post-revascularization improvement in function. ³¹

Gerber et al ³² further clarified the question regarding post-revascularization "recovery" of hyper-enhancing (and therefore nonviable) segments on MDE imaging. Using MR tagging techniques, they found evidence of "tethering" of the edges of infarcted tissue to myocardial segments that regained function after revascularization.

First-pass myocardial perfusion is often measured as a way of assessing coronary artery disease, but it can be misleading in delineating chronic myocardial infarction with MR imaging. ¹⁸ Areas of decreased signal during the first minute of perfusion imaging may represent either scar/fibrosis or rest occlusion of a stenotic coronary artery segment. ¹⁸ First-pass perfusion imaging is useful in acute infarction, as there may be areas of decreased perfusion in the central infarct core due to microvascular obstruction by necrotic and/or inflammatory cells (sometimes called the "no-reflow" phenomenon). ^33,34 Such areas, if present, represent tissue that will not recover function and is, therefore, nonviable.

Technique

Myocardial delayed enhancement imaging is performed with the patient at rest. Cardiac gating and a cardiac coil are necessary. Table 1 details the generic post-myocardial infarction protocol at our institution. Although it includes first-pass perfusion imaging, this sequence is relevant only when the protocol is performed in the setting of acute infarction, as noted above, and it may be omitted as appropriate. Scout images are obtained in sagittal, axial, and oblique planes. Then, during the injection of 0.1 mmol/kg gadodiamide (Omniscan, Amersham Biosciences, Piscataway, NJ) at a rate of 5 mL/sec, an inversion recovery-prepared fast gradient-echo pulse sequence is performed to assess first-pass perfusion. An additional 0.1 mmol/kg is then administered. The split contrast dose avoids oversaturation during the first-pass perfusion imaging, but achieves a more optimal intravascular concentration for later MDE imaging. Cine images in short and long axis are obtained using SSFP sequences to evaluate morphology and function (eg, ejection fraction, cardiac output, wall thickness, ventricular mass). Myocardial delayed imaging is performed 10 to 20 minutes after contrast administration, using an inversion recovery prepared fast-gradient-echo sequence with blood suppression and nulling of normal myocardium (Figures 1 and 2). The total duration of the imaging protocol can be as brief as 20 minutes under ideal circumstances. It may be longer, particularly when the patient is severely ill or debilitated, and physically less able to cooperate.

Contraindications to MDE imaging are the same as for any MR imaging examination. Patients with implantable defibrillator-pacemakers or orbital metallic fragments should not be imaged. While gadolinium is not nephrotoxic and has a low incidence of side effects, its safety in pregnancy is unknown. ^35,36

MR compared with other modalities

Before comparison is made to MDE imaging, a brief review of the other noninvasive modalities for assessing myocardial viability is in order. All of these have been proven to have prognostic utility in evaluating viability. ^6,10,13

Positron emission tomography exploits the fact that ischemic myo-cardium preferentially metabolizes glucose over fatty acids. In this technique, [ ¹⁸ F]fluorodeoxyglucose is used as a radiotracer, and normal or increased uptake by a segment which demonstrates decreased perfusion (as measured by a PET or SPECT perfusion agent) is interpreted to indicate viability of that tissue. ¹⁶ Positron emission tomography has higher spatial resolution than SPECT, and allows attenuation correction and the quantification of some physiologic parameters. ¹⁷ The positive predictive value of a perfusion-metabolism mismatch for functional improvement after revascularization has been reported to from 76% to 85%. ^6,37 Negative predictive value for the technique has been found to be as high as 92%. ³⁷

Single-positron emission computed tomography studies can employ either thallium-201 or technetium-99m sestamibi uptake to measure myocardial perfusion. Early perfusion defects that fill-in on images obtained hours later are interpreted to represent viable myocardium. Thallium is a potassium congener and enters cells through energy-dependent transport. ¹⁷ Technetium sestamibi uptake relies on active mitochondria. ³⁸ Neither agent will be taken up by dead myocardium. Two different protocols are reported in the literature. In the rest-redistribution technique, cardiac SPECT is performed immediately after injection of radiotracer, again after a delay of 3 to 4 hours (a second dose may be injected just prior to delayed imaging), and possibly again after a 24-hour delay. ³⁹ Alternatively, a stress-redistribution-reinjection technique can be used, with initial injection of radiotracer at peak stress, and delayed images as with the prior technique. ¹⁷ While the delayed fill-in can be assessed qualitatively, some researchers have quantitated it. Delayed uptake of more than 60% of peak activity had a sensitivity of 78% and a specificity of 58% in predicting functional recovery after revascularization, in one study that employed thallium and the rest-redistribution protocol. ⁴⁰ Pooled study results for both techniques have reported positive predictive value of 69% and a negative predictive value of 89% to 92%. ⁴¹ Technetium viability studies offer the possibility of excellent gated wall-motion imaging due to the long half-life of the isotope. ⁴² However, this radiotracer is especially susceptible to attenuation artifacts, especially in the inferior wall. ⁴³

Dobutamine stress echocardiography measures contractile reserve, which is the ability of myocardium to increase its thickening after stimulation with dobutamine (or another inotropic agent). Myocardium that is dysfunctional but demonstrates sufficient contractile reserve is interpreted to be viable. ¹⁸ Sensitivity for post-revascularization improvement has been reported from 74% to 88%, with specificity ranging between 73% and 83%. ⁴⁴ Dobutamine stress examinations may also be performed in conjunction with MR imaging. This is of particular use for patients with poor acoustic windows for echocardiography.

Spatial resolution

The chief advantage of MR imaging in comparison to other noninvasive methods is its higher spatial resolution, up to 1.5 mm in-plane, which allows for assessment of the percentage of transmural extent of scar. ²⁸ By contrast, when assessed by PET, SPECT, or echocardiography, viability of an individual dysfunctional segment is an "all-or-none" phenomenon. This additional clinical information--small subendocardial versus large subendocardial versus transmural infarct--may change patient management. In addition, the high spatial resolution is of benefit when using MR cine images to quantitatively evaluate cardiac morphology and function.

Accuracy

While the gold standard in viability assessment is improvement of function after revascularization, PET has been used in many clinical studies as a reference modality for assessing myocardialviability. ^18,45,46 A recent study of 31 patients showed that MDE imaging findings not only agreed closely with findings on PET but showed additional areas of scar which PET did not, due to the higher resolution. ⁴⁵ If clinical trials find prognostic or therapeutic significance to this extra level of detail, MR imaging may come to be the preferred modality for assessing viability.

Reproducibility

MR imaging has the potential to reliably produce complete, high-quality examinations regardless of operator or patient body habitus. Approximately 15% to 20% of echocardiography examinations are limited by incomplete visualization of segments, primarily due to dependence on acoustic windows. ³⁰ Attenuation artifacts limit the sensitivity of SPECT imaging, particularly in the inferior wall of the heart and in obese patients. ⁴³

In the studies cited above, no patients were excluded due to poor MR image quality. However, patients who are unable to lie supine or to hold their breath for appropriate periods may be poor candidates for MDE imaging.

Availability

An advantage of SPECT imaging and dobutamine stress echocardiography for evaluating myocardial viability is that the expertise to perform and interpret such studies is widely available. Cardiac MR imaging is still relatively novel in many communities. However, MDE imaging can be performed using hardware and software that has been offered commercially for several years, and the use of intravenous contrast and cardiac gating is not new. Positron emission tomography remains somewhat limited by high cost and lack of broad availability.

Simplicity

Dobutamine stress echocardiography (and in some cases SPECT imaging) derives its information from comparison of stress and rest physiology. The need for either exercise or pharmacologic stress adds time, cost, and complexity to those examinations. Like the SPECT rest-redistribution technique and PET, MDE imaging has the advantage of being performed at rest, and may be substantially faster than those modalities in many instances. The protocol at our institution can be completed in 20 minutes under optimal conditions. It is now also possible to obtain postcontrast three-dimensional whole-heart images in a single breath-hold, further accelerating the examination (Figure 3). ⁴⁷

It is worth noting that much research has been completed and is ongoing with regard to dobutamine stress cardiac MR imaging. It has proven prognostic value in assessing viability, and provides a greater degree of quantitative accuracy than echocardiography. ^48-50 However, like dobutamine stress echocardiography, it adds an additional level of complexity when compared with the MDE technique.

The future

Large prospective studies of MDE imaging, with attention to the prognostic implications of partial-thickness scar, could clarify the clinical significance of this technique in relation to other modalities. ²⁸ Other MR imaging techniques of assessing myocardial viability are under investigation, though they may not have potential of MDE imaging for prompt impact. Necrosis-avid MR imaging contrast agents, metalloporphyrins, are paramagnetic and have high specificity for scarred myocardium. Imaging is performed hours after injection. ⁴² One manganese-based MR contrast agent has been shown to accumulate in normal myocardium but not infarct. ⁵¹ MR imaging using sodium-23 and MR spectroscopy of phosphorus-31 can also distinguish viable from nonviable myocardium. ^28,42 The latter techniques may be primarily of research interest due to low resolution (sodium-23 MRI) or highly specialized equipment requirements (phosphorus-31 MR spectroscopy). ^28,42

The future development of MR imaging of viable myocardium will also be influenced by the concurrent evolution of other cardiac MR imaging applications, such as MR perfusion and MR coronary angiography. These are areas of much research ferment, and the goal of cardiac MR imaging as a comprehensive modality for evaluation of coronary heart disease remains an active pursuit. ⁵²

Conclusion

Myocardial delayed enhancement imaging has many features that make it competitive with existing modalities in the assessment of viable myocardium in chronic ischemic heart disease, including relative simplicity and potential speed. Foremost among these is its uniquely high level of spatial resolution, which may provide information about myocardium that can change patient management.