- Email: [email protected]
Noninvasive Imaging Techniques for the Diagnosis of Myocarditis
Heart Failure Clin 1 (2005) 377 – 389
Noninvasive Imaging Techniques for the Diagnosis of Myocarditis G. William Dec, MDa,b,T a
Massachusetts General Hospital, Boston, MA, USA b Harvard Medical School, Boston, MA, USA
The clinical manifestations of active myocarditis are highly varied and are generally not specific enough to establish a diagnosis with certainty. Clinicians have increasingly relied during the past two decades on right (or left) ventricular endomyocardial biopsy for histologic confirmation of suspected inflammatory heart disease. Mason and coworkers [1] were among the first investigators to demonstrate evidence for myocarditis using right ventricular endomyocardial biopsy in a small group of patients with previously diagnosed idiopathic dilated cardiomyopathy. Although endomyocardial biopsy has become the international standard for establishing the diagnosis, the exact histologic criteria remain controversial. To provide more uniform criteria for establishing the pathologic diagnosis, a panel of cardiac pathologists developed a classification known as the Dallas criteria [2]. In this schema, myocarditis is characterized by an inflammatory cellular infiltrate within the myocardium with necrosis or degeneration of adjacent myocytes (or both), a pattern that is not typically seen in ischemic injury. The inflammatory infiltrate is typically lymphocytic but may also include eosinophilic, neutrophilic, granulomatous, or mixed cellularity. The amount of inflammation may be mild, moderate, or severe and its distribution may be focal, confluent, or diffuse, respectively. Despite the widespread adoption of this histopathologic classification, some clinicians believe that the definition
T Harvard Medical School, 55 Fruit Street, Bigelow 800, Mailstop 817, Boston, MA 02114. E-mail address: [email protected]
is too restrictive and have proposed alternatives based on clinicopathologic criteria or immunohistochemical findings [3,4]. Sampling error remains the most crucial limitation to the diagnostic accuracy of the endomyocardial biopsy. Chow and coworkers [5] provided clear evidence for the insensitivity of right ventricular biopsy in detecting myocarditis. Likewise, Hauck and coworkers [6] analyzed hearts from autopsy specimens in which myocarditis was determined to have contributed directly to death. Ten biopsies from the apex and septum of both ventricles were evaluated for myocarditis using the Dallas criteria. When only five right ventricular samples from each heart were evaluated (which is the most common clinical sampling rate), the diagnosis of myocarditis could not be established in 55% of the hearts (Fig. 1). In a similar postmortem study of 14 hearts, 17.2 samples per heart were required to diagnose myocarditis correctly in more than 80% of cases [5]. Dec and coworkers [7] examined the role of repeat right and left ventricular biopsy among patients who were strongly suspected of having myocarditis clinically but whose initial right ventricular biopsy failed to provide histologic confirmation. Biopsy of both ventricles detected an additional 15% incidence of myocarditis. A positive endomyocardial biopsy unequivocally establishes the diagnosis; however, the absence of conclusive histologic confirmation does not exclude consideration of myocarditis in most clinical settings. Given the focal or multifocal nature of this disease process, it is not surprising that substantial sampling error exists. Clinicians are increasingly reluctant to recommend routine biopsy even when myocarditis is clinically
1551-7136/05/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.hfc.2005.06.007
heartfailure.theclinics.com
378
dec
Fig. 1. Sensitivity of endomyocardial biopsy for detecting myocarditis in a postmortem study of 38 hearts with proven myocarditis. Each section demonstrates the rate of detection of each additional endomyocardial biopsy specimen. One biopsy detected myocarditis in 18% of specimens. Light gray shading: five biopsy specimens detected myocarditis in only 43% of specimens. White shading: 17 biopsies confirmed myocarditis in 82% of explanted hearts. (From Dec GW. Clinical myocarditis. In: Cooper L, editor. Myocarditis: bench to bedside. Totowa (NJ): Humana Press; 2003. p. 261. Copyright Mayo Clinic Foundation; with permission.)
strongly suspected. The contemporary role of endomyocardial biopsy in the diagnosis and management of myocarditis has recently been reviewed [8]. Given the cost of biopsy, its invasiveness, and the relatively low histologic rate of detection for myocarditis or other treatable forms of acute left ventricular dysfunction (eg, cardiac sarcoidosis), sensitive and specific noninvasive screening methodologies are urgently needed for the diagnostic evaluation of this population. Serum biomarkers including creatine kinase and the troponins are increasingly recognized as having predictive utility as screening tools. Additional prospective confirmatory studies are needed, however, to verify the predictive accuracy of these biomarkers. Noninvasive imaging techniques including echocardiography, radionuclide imaging, and MRI have been recently studied and may provide additional diagnostic information.
Echocardiographic findings in myocarditis Echocardiography is currently recommended as the most important initial diagnostic study in patients with heart failure and frequently provides a substantial contribution to the differential diagnosis of disease etiology. Several studies have evaluated transthoracic echocardiography for diagnosing myocarditis [9,10]. Pinamonti and coworkers [9] retrospectively analyzed echocardiographic findings among 42 patients with
biopsy-proven myocarditis. Clinical presentations included heart failure (63%); high-grade atrioventricular block (17%); chest pain despite normal coronary anatomy (15%); and supraventricular arrhythmias (5%). Left ventricular dysfunction was commonly observed (69%), particularly among patients with congestive heart failure (present in 88%). Left ventricular cavity enlargement was frequently minimal or absent, consistent with other forms of acute dilated cardiomyopathy. Right ventricular dysfunction was present in only 23% of this cohort. Not surprisingly, patients who presented with chest pain or heart block almost always had preserved ventricular size and function. Segmental wall motion abnormalities were observed in 64% of patients and included hypokinetic, akinetic, or frankly dyskinetic regions. Reversible left ventricular hypertrophy was noted in 15% of patients and resolved over several months. Echocardiographic findings were quite varied and relatively nonspecific. More recently, Pinamonti [10] reviewed the diagnostic use of echocardiography in diagnosing patients with chronic heart failure. It was concluded that ‘‘no useful echocardiographic findings can identify patients with genetic dilated cardiomyopathy or affected by myocarditis from other cases of idiopathic dilated cardiomyopathy’’ [10]. This statement is not meant to imply that routine echocardiography has no role in the evaluation and management of myocarditis patients. Serial studies have been shown to
noninvasive imaging techniques
be useful in assessing the response to treatment of several forms of myocarditis. Resolution of marked concentric left ventricular hypertrophy in eosinophilic myocarditis following corticosteroid treatment has been reported [11]. Similarly, echocardiography has provided a useful method of following-up chlamydial myocarditis [12]. It has also been used to evaluate prognosis in childhood myocarditis. Carvalho and coworkers [13] have described the use of the ratio of end-diastolic posterior wall thickness to cavity dimension in the pediatric population. Relative preservation of posterior wall thickness (ratio > 0.17) was predictive of better prognosis and higher likelihood of recovery. Mendes and coworkers [14] examined the process of ventricular remodeling among patients with active myocarditis enrolled in the Multicenter Myocarditis Treatment Trial. Compared with normal controls, myocarditis patients demonstrated larger mean left ventricular volumes (81 ± 29 mL/m2 versus 50 ± 8 mL/m2; P = .001). Chamber dilatation occurred primarily along the mid-cavity diameter, which measured 5.3 ± 0.8 cm among myocarditis patients compared with 4.2 ± 0.4 cm in controls (P=.001). Active myocarditis was associated with early left ventricular remodeling and the development of a more spherical chamber. Multivariable modeling of echocardiographic parameters demonstrated only baseline left ventricular ejection fraction (but not degree of left ventricular dilatation or sphericity) to be an independent predictor of longterm survival. Although anatomic features on echocardiography (ie, chamber dimensions, ejection fraction, wall motion abnormalities) are insufficient to differentiate myocarditis from other forms of cardiomyopathy, ultrasonic tissue characterization may be of more use. Myocarditis is associated with inflammation, myocardial edema, and fibrosis of varying degrees of severity. Ultrasound for myocardial tissue characterization is based on the hypothesis that quantifiable connections exist between the tissue structures and their acoustical properties. Transmission and reflection of ultrasound in tissue depends on tissue density, elasticity, and acoustical impedance. Changes in one or more of these factors lead to different ultrasonic backscatter and altered image texture. A systematic relationship exists between tissue structure and echocardiographic texture. Lieback and coworkers [15] evaluated mean gray scale values (indicative of average brightness) in 52 patients with biopsy-proven myocarditis, 12 patients with persistent myocarditis, 9 patients with healed myocarditis who lacked fibrosis, and 17 patients with healed myocarditis and fibrosis. Echocardiographic features were com-
379
pared with normal controls and a cohort with dilated cardiomyopathy. Tissue characterization was highly effective in differentiating myocarditis from healthy control myocardium with sensitivity and specificity values of 100% and 90%, respectively [16]. Ultrasonic tissue characterization could not accurately differentiate, however, between idiopathic dilated cardiomyopathy and active myocarditis. More recent techniques have been able better to characterize tissue changes in acute myocarditis and to evaluate changes in these parameters over time [16]. Cardiac tissue Doppler imaging has been shown successfully to detect biopsy-verified myocarditis in a former Olympic cyclist with recurrent ventricular tachycardia [17]. M-mode Doppler tissue echocardiography has also been shown to detect recurrent giant cell myocarditis in transplant recipients [18]. An abnormal myocardial velocity gradient across the left ventricular posterior wall during isovolumic relaxation and a reduction during rapid ventricular filling indicative of impaired myocardial relaxation have been demonstrated during the active phase of the disease (Fig. 2). These abnormal findings resolved completely after augmented immunosuppressive therapy. These limited case reports suggest that newer echocardiographic imaging modalities may possess sufficient resolution effectively to screen for inflammatory diseases of the myocardium. Additional validation studies are required to determine their predictive accuracy.
Nuclear imaging techniques Gallium-67 (67Ga) cardiac scintigraphic imaging has been used to evaluate conditions that result in myocardial inflammation, a key component of myocarditis. Early reports from centers with extensive experience using this qualitative technique demonstrate, however, that 67Ga imaging can be useful as a screening tool and for predicting therapeutic response [19,20]. O’Connell and coworkers [20], who have evaluated this methodology most extensively, reported a sensitivity of 36% but a specificity of 98% for histologic diagnosis of myocarditis. The difficulty of the technique combined with its low specificity has led to its infrequent use. Indium-111 (111In) – labeled monoclonal antibody fragments (directed against heavy chain myosin) bind to cardiac myocytes that have lost the integrity of their sarcolemmal membranes and have exposed their intracellular myosin to the extracellular fluid space [21]. Unlike 67Ga, which detects the extent of myocardial inflammation, antimyosin cardiac uptake reflects the extent of myocyte necrosis. Because both
380
dec
Fig. 2. Echocardiographic findings in giant cell myocarditis. (A) M-mode Doppler tissue echocardiography (DTE) of the left posterior wall with quantification of myocardial velocity gradient (MVG) and mean myocardial velocities (MMV); parasternal long-axis view (top). (B) Corresponding left ventricular filling pattern on Doppler mitral inflow (E-wave deceleration time, 138 millisecond). During isovolumic relaxation (IVR) (A, bottom), MVG is abnormally positive at 2.9 s 1 (normal range, 2.3 to 1.4 s 1). During rapid ventricular filling (RVF), just after the opening of the mitral valve, MVG is significantly reduced to 2.8 s 1 (normal range, 5.9 to 18.1 sec 1). PCG, phonocardiogram. Dotted vertical lines indicate isovolumic relaxation time. (From Palka P, Lange A, Clarke B, et al. Echocardiographic description of recurrent idiopathic giant-cell myocarditis in cardiac allograft. Circulation 2003;108:e1 – 3; with permission.)
elements are present in active myocarditis, the two imaging modalities may provide complementary information. Unfortunately, no studies have directly compared the use of these radionuclide techniques. Published data suggest that antimyosin imaging may have higher negative predictive value than 67Ga scintigraphy [22]. Yasuda and coworkers [22] first reported antimyosin scintigraphic imaging in 28 patients with clinical findings suggestive of acute myocarditis. Most (75%) had unexplained heart failure of less than 6 months’ duration. Four patients presented with ventricular tachyarrhythmias and three patients had chest pain mimicking acute myocardial infarction. All patients underwent right ventricular biopsy and antimyosin imaging within 7 days. The mean left ventricular ejection fraction was 27% ± 2%. Antimyosin imaging was found to be positive in 61% of patients (Fig. 3). Cardiac uptake was generally heterogeneous and diffuse within the left ventricular myocardium. Importantly, none of the patients with a negative antimyosin scan had histologic evidence of myocarditis and all patients with biopsy-proven myocarditis had a positive scan. Similar findings were reported in a small series by Haber and coworkers [23]. Rezkalla and coworkers [24] validated this technique in a murine model of Coxsackie-induced myocarditis. Following intraperitoneal injection of virus,
the acute phase of the disease is characterized by high-grade viremia, viral replication in the myocardium, and the development of an intense mononuclear cellular infiltrate associated with focal myocyte necrosis. Histopathologic changes and the temporal course of the disease mimic human myocarditis. One, 2, 4, and 12 weeks following Coxsackie inoculation, mice underwent 125In-labeled antimyosin Fab antibody injection (25 mCi); the hearts were excised at 48 hours for assessment using gamma scintillation counting. Myocardial radioactive uptake from the infected animals was significantly higher than controls, from 1 week to 12 weeks following inoculation. More importantly, the degree of radiotracer uptake in the hearts correlated closely with histopathologic changes. Autoradiography revealed localization of radioactivity to areas of myocardial necrosis. Matsumori and coworkers [25] and Kishimoto and coworkers [26] confirmed this close correlation between the extent of 111In-labeled antimyosin uptake and the degree of histologically confirmed myocardial necrosis in an encephalomyocarditis murine model. Dec and coworkers [27] subsequently studied the use of antimyosin imaging in a large cohort (N = 82) of patients with clinically-suspected myocarditis. Presenting symptoms included heart failure and dilated cardiomyopathy (92%); chest pain mimicking acute myocardial infarction (6%); or life-threatening
noninvasive imaging techniques
381
Fig. 3. Antimyosin cardiac imaging for detecting active myocarditis. A positive antimyosin image demonstrates diffuse tracer uptake in the cardiac region on both the anterior planar image (upper left) and in all coronal tomographic reconstructions (bottom left). Biopsy confirmed multifocal lymphocytic myocarditis. Antimyosin imaging was repeated after 6 months of immunosuppressive therapy (top right). Biopsy showed healed myocarditis. No antimyosin uptake is visible on either the planar (top right) or tomographic reconstructions (bottom right). (From Dec GW. Clinical myocarditis. In: Cooper L, editor. Myocarditis: bench to bedside. Totowa (NJ): Humana Press; 2003. p. 273. Copyright Mayo Clinic Foundation; with permission.)
ventricular arrhythmias (2%). Most patients presented with symptoms < 12 months’ duration; five patients had a prior bout of biopsy-proven myocarditis with a second episode of acute heart failure. Following injection of 111In-labeled antimyosin antibody fragments (500 mg coupled to 2 mCi), all patients underwent planer and single-photon emission CT cardiac imaging at 48 hours. Right ventricular biopsy was performed within 48 hours of imaging. On the basis of right ventricular histologic features, antimyosin was highly sensitive but only moderately specific for detecting myocardial necrosis. The sensitivity was 83% and specificity 53%; however, the predictive value of a negative scan was 92%. More recently, Margari and coworkers [28] reported that the presence of both a positive antimyosin scan and a nondilated left ventricle (left ventricle end-diastolic diameter 62 mm) was highly predictive of biopsyproven myocarditis. Perhaps more important than the correlation between cardiac antimyosin uptake and histologic findings, improvement in ventricular function within 6 months of diagnosis was evident in 54% of patients with a positive antimyosin scan compared with only 18% of those with a negative scan [29]. Because spontaneous improvement in ventricular function is a recognized feature of acute myocarditis, it is possible that a number of patients who were scan-positive but biopsy-negative may have, in fact, had myocarditis that was not detected on biopsy because of sampling error. Repeat antimyosin imaging among a cohort whose initial scan was positive showed persistent
uptake in 50% and resolution in the remaining patients. No correlation was evident between ongoing myocarditis on repeat biopsy and clinical improvement [29]. Narula and coworkers [30] also evaluated the role of this technique among patients who presented with chest pain mimicking acute myocardial infarction. Following demonstration of normal coronary anatomy, antimyosin imaging was performed. Uptake was global in seven of eight patients with confirmed myocarditis on biopsy; the eighth patient had borderline myocarditis. In contrast, antimyosin uptake is typically segmental among patients with acute myocardial infarction and almost always confined to the territory of the infarct-related vessel. Sarda and coworkers [31] have also confirmed the ability of antimyosin imaging to detect myocarditis in patients presenting with a clinical syndrome of acute myocardial infarction and normal coronary angiography. Antimyosin imaging may, at times, be useful in differentiating unstable coronary syndromes from acute myocarditis. Its high sensitivity but modest specificity suggests that antimyosin scintigraphy may serve as an excellent initial screening tool to determine which patients should undergo further diagnostic assessment for myocarditis, including biopsy.
MRI MRI is rapidly becoming the gold standard for assessing both right and left ventricular anatomy and
382
dec
Box 1. Applications of MRI in cardiovascular diseases for assessing anatomic changes, tissue relaxation properties, perfusion characteristics, and chemical composition Left and right ventricular mass Ventricular wall thickness (hypertrophy, thinning) Left and right ventricular contractile function T1 relaxation time T2 relaxation time Myocardial perfusion imaging Phosphorus-31 spectroscopy Sodium-23 spectroscopy
for measurement of ventricular function (Box 1). Besides providing anatomic and morphologic information, MRI techniques can provide accurate tissue characterization by measuring T1 and T2 relaxation times and spin density [32]. T1 relaxation is defined by the rate of longitudinal relaxation of protons in the magnetic field (Fig. 4). T2 relaxation is related to the
rate at which proton spins fall out of phase (Fig. 5). Increases of both these relaxation times correlate well with increased tissue water content [32]. Because active myocarditis is generally associated with myocyte injury including edema and cellular swelling, assessment of the relaxation times may provide a sensitive measure for detecting myocarditis. Increases in T2 relaxation have also been reported to occur with focal hemorrhage and myocyte necrosis [32]. Prior animal and human studies have demonstrated the reliability of MRI for tissue characterization of cardiac allograft rejection [32]. Aherne and coworkers [33] examined heterotopic cardiac transplantation in immunosuppressed (prednisone, 1 mg/kg/d, and cyclosporine, 25 mg/kg/d) dogs and compared them with animals that also underwent transplantation but did not receive immunosuppressive therapy. Untreated allografts showed a significant increase in T2 relaxation time (60 ± 8 versus 44 ± 6 milliseconds; P < .01) compared with control hearts as early as 1 week following transplantation. More importantly, T2 relaxation times and signal intensity increased in allografts that developed histologic evidence of acute cellular rejection. There was a significant correlation between histologic grading of rejection severity and T2 relaxation times in the cardiac allograft (r value = 0.72).
Fig. 4. Longitudinal relaxation in the B0 field is described by the relaxation constant T1. T1 can be measured by performing an inversion recovery pulse sequence using a 180-degree pulse. As the time to imaging (TI) is delayed further and further after the inversion pulse, the overall magnetization in the B0 field becomes closer to the fully relaxed state, M0. The signal intensity increases as the imaging time increases, obeying the equation: SI = M0* (1 to 2e (TI/T1). (From Koelling TM. The use of magnetic resonance imaging in cardiac transplant rejection. In: Dec GW, Narula J, Ballester M, et al, editors. Cardiac allograft rejection. Boston: Kluwer Academic Publishers; 2001. p. 324; with permission.)
noninvasive imaging techniques
383
Fig. 5. T2 is the time constant that describes the way spins diphase after being focused by a 90-degree excitation pulse. Signal is maximal immediately after the excitation phase, and degrades with time as spins fall out of phase and net magnetization in the transverse plan approaches zero. The magnetization after any echo time, TE, is described by the equation: SI = M0e (TE/T2)). (From Koelling TM. The use of magnetic resonance imaging in cardiac transplant rejection. In: Dec GW, Narula J, Ballester M, et al, editors. Cardiac allograft rejection. Boston: Kluwer Academic Publishers; 2001. p. 325; with permission.)
There was also a significant linear relationship between in vivo T2 values and difference in percent water content in the allograft and pretransplant native heart (r = 0.92, P < .001) (Fig. 6). Because myocarditis demonstrates many of the histologic features of acute cardiac allograft rejection, it is not surprising that MRI has also been used as a diagnostic tool in this disease. A preliminary study by Chandraratna and coworkers [34] was the first to report localized myocardial edema in regions of hypokinesis or akinesis on echocardiography. Only two patients were reported and myocarditis was based solely on clinical criteria. Importantly, improvement in ventricular function was associated with resolution of myocardial edema on MRI in this study. Gagliardi and coworkers [35] evaluated MRI and concurrent endomyocardial biopsy findings in 11 consecutive children (age 9 months to 9 years) with clinically suspected myocarditis. Tissue characterization was obtained in regions of interest of the right and left ventricles by using T1 and T2 spin echo sequences [35]. The myocardial-skeletal muscle signal intensity ratio was calculated and accurately identified all patients with histologically verified myocarditis. Although encouraging, these results were obtained in a very small number of patients. Further, myocarditis in children is often associated with a more promi-
nent interstitial edema than is typically observed in adults. The same authors subsequently have reported their experience in 75 consecutive pediatric patients with acute heart failure symptoms, left ventricular systolic dysfunction, and the absence of congenital heart disease [36]. All children underwent concurrent endomyocardial biopsy and cardiac MRI. T2 relaxation sequences correctly identified patients with and without histologic evidence for myocarditis with a high degree of predictive accuracy. Endomyocardial biopsy identified histologic evidence of acute myocarditis in 51 patients; the remaining 24 individuals had idiopathic dilated cardiomyopathy. Compared with biopsy findings, MRI achieved a sensitivity of 100% and a specificity of 90% in this cohort [36]. The 53 children with myocarditis have undergone repeated biopsy and MRI investigation every 6 months to evaluate the efficacy of MRI in identifying persistent, resolving, or resolved myocarditis during a 2-year follow-up [36]. Preliminary data have confirmed the use of MRI signal intensity by T2-weighted relaxation sequences during follow-up because the sensitivity and specificity remain high during the entire evolution of the disease [36]. The diagnostic usefulness of contrast-enhanced cardiac MRI has also been evaluated [37 – 41]. In a
384
dec
Fig. 6. Correlation between percent difference in T2 relaxation times (transplanted hearts minus native hearts) and percent difference in water content (transplanted hearts minus native hearts). (From Aherne T, Tscholakoff D, Finkbeiner W, et al. Magnetic resonance imaging of cardiac transplants: the evaluation of rejection of cardiac allografts with and without immunosuppression. Circulation 1986;74:145 – 56; with permission.)
landmark study, Friedrich and coworkers [37] examined 19 patients with suspected viral myocarditis based on a combination of clinical and laboratory findings. Inclusion criteria included a history of febrile viral illness within 4 weeks of study entry; one or more symptoms (fatigue, malaise, dyspnea, chest pain, fever, or tachycardia); one or more electrocardiographic changes (atrioventricular block, ST segment depression or elevation in three or more leads, recurrent atrial or ventricular tachyarrhythmias); regional or global left ventricular dysfunction by echocardiography and left ventriculography; demonstration of normal coronary anatomy; positive viral serology (Coxsackie B virus, 11 patients; cytomegalovirus, 4 patients; Epstein-Barr virus, 3 patients; and influenza, 1 patient); and a positive cardiac antimyosin scan. Seven of the 19 patients underwent biopsy; myocarditis was detected in four individuals. ECG-triggered, T2-weighted, multislice spin echo sequences were performed in the axial orientation (five acquisitions, slice thickness 6 mm) for controls and suspected myocarditis patients. In addition, ECGtriggered, T1-weighted multislice spin echo images were obtained in the axial and short-axis orientation (four to six acquisitions, slice thickness 6 mm). Repeat imaging was performed beginning 1 minute after injection of 0.1 mmol/kg of gadolinium. Manual regions of interest were drawn for cardiac and skeletal muscle and compared with measurements obtained in
18 volunteers. Myocarditis patients were serially studied on days 2 ± 7, 14 ± 2, 28 ± 4, and 84 ± 10 after symptom onset. The T1-weighted images were found to provide the most useful information. Early following presentation, myocardial enhancement was generally focal in distribution (Fig. 7). Global enhancement became prominent during the later imaging times (Fig. 8). Overall, global left ventricular enhancement was substantially higher in the myocarditis patients than in the controls on days 2 through 28 after symptom onset. The global enhancement index had returned to that observed in controls by the day 84 study (Fig. 9). Unfortunately, the study did not examine the ability of MRI to differentiate viral myocarditis from other causes of acute dilated cardiomyopathy. It did suggest, however, that longitudinal follow-up of the same patient with active disease is feasible and allows for reassessment of persistent myocarditis or recurrent disease. Roditi and coworkers [39] evaluated 20 patients with T1 spin echo cine MR angiography and contrastenhanced spin echo imaging before and after gadolinium enhancement. Four patients had histologically proven myocarditis; eight patients had clinically suspected myocarditis, and eight patients had chronic cardiomyopathy. Focal myocardial enhancement was associated with regional wall motion abnormalities in 10 out of the 12 patients with suspected or proven myocarditis. Using this technique, the myocarditis
noninvasive imaging techniques
385
Fig. 7. T1-weighted MRI cross-sectional views at the midventricular level in a patient with acute myocarditis. Unenhanced view obtained on day 2 after onset of symptoms (left panel). There are small foci of increased signal intensity in the subepicardial parts of the posterior myocardium and in the basal septum, which were more evident (right panel) following administration of gadopentetate dimeglumine (arrows). (From Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1805; with permission.)
cohort demonstrated a mean focal enhancement of 122% (corrected for a background skeletal muscle enhancement), which was significantly different from the mean of 14% observed in the cardiomyopathy (nonmyocarditis) group (Fig. 10). The authors suggest that the findings of focal myocardial enhancement by MRI combined with regional wall motion abnormalities (hypokinesis, akinesis, or dyskinesis) strongly support the diagnosis of myocarditis. The ability of contrast-enhanced MRI techniques to diagnose other forms of inflammatory heart disease, particularly cardiac sarcoidosis, has recently been validated [38]. New contrast MRI techniques using segmented inversion recovery gradient echo pulse sequences provide substantial improvement in contrast between diseased and normal myocardium. These newer techniques may provide up to a 500% improvement
in resolution compared with the protocol used by Frederich and coworkers [37,41]. Mahrholdt and coworkers [41] recently used this new technique to perform MRI-guided biopsy of the right and left ventricles in patients with possible myocarditis. Thirty-two patients were included in the study. Each had clinically suspected myocarditis based on flulike symptoms within 8 weeks of presentation; malaise; chest pain; dyspnea; or tachycardia and electrocardiographic evidence for atrioventricular block, ST segment depression, or unexplained ventricular arrhythmias. Cine and contrast-enhanced ECG-gated MRI was performed using a 1.5-T system, 6-mm thickness slices, and 15-second breathhold technique per slice. Following initial image acquisition, gadoteridol (0.1 mmol/kg) contrast was administered intravenously and repeat imaging was performed 10 minutes later. Left ventricular biopsy was gen-
Fig. 8. T1-weighed MRI obtained before (left panel) and after enhancement (right panel) in the same patient as shown in Fig. 7 at 14 days following presentation. More diffuse enhancement of myocardium after gadopentetate dimeglumine, including the apical part of the septum and visible areas of the right ventricle. (From Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1805; with permission.)
386
dec
Fig. 9. Mean ± SE of the global relative MRI signal enhancement of cardiac tissue compared with skeletal muscle tissue on days 2, 7, 14, 28, and 84 after onset of acute myocarditis. There was a significant difference in the mean results from days 2 through 28 compared with control subjects. (From Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1807; with permission.)
Fig. 10. Myocardial enhancement of T1-weighted spin echo MRI in patients with known or suspected myocarditis (Myocarditis) and those without suspected myocarditis (Not Myocarditis), both uncorrected and corrected for background skeletal muscle enhancement. Box, 25th to 75th percentiles around the median (line); whiskers, 10th to 90th percentiles; x, individual measures with outliners circled. (From Roditi GH, Hartnell GG, Cohen MC. MRI changes in myocarditis: evaluation with spin-echo, cine MR angiography and contrast enhanced spin echo imaging. Clin Radiol 2000;55:756; with permission.)
noninvasive imaging techniques
387
Fig. 11. Cardiac MRI findings and histopathologic findings of typical patients in whom biopsies were obtained from areas of MRI contrast enhancement. Top 3 panels show cases (patients 6, 7, and 14) of active myocarditis with myocyte damage and infiltration of macrophages; bottom panel illustrates a patient (patient 18) without active myocarditis who was diagnosed with hypertrophic cardiomyopathy. SAX, short-axis orientation; LAX, long-axis orientation. (From Mahrholdt H, Goedecke C, Wagner A, et al. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation 2004;109:1253; with permission.)
erally performed from the region showing the most marked contrast enhancement (Fig. 11). Right ventricular septal biopsies were performed for patients with an inaccessible anterobasilar left ventricular location (N = 7); prominent septal enhancement (N = 2); or those who lacked any evidence of focal enhancement (N = 4). Contrast enhancement was present in 28 patients (88%) and was usually seen with one or several distinct regions in the myocardium. Regional enhancement was most frequently observed in the lateral wall. For the 21 patients in whom biopsies were obtained directly from the region of contrast enhancement, histopathology revealed active myocarditis in 19 patients. Conversely, in the remaining patients (N = 11) in whom biopsy was not taken from the region of contrast-enhancement, active myocarditis was detected in only one case. At follow-up, the area of contrast enhancement decreased from 9% ± 11% to 3% ± 4% and the mean left ventricular ejection fraction improved from 47% ± 19% to 60% ± 10% [41]. Overall, biopsy of those specific myocar-
dial regions enhanced by MRI resulted in positive and negative predictive values for detecting myocarditis on biopsy, 71% and 100%, respectively.
Summary Accurate differentiation of active inflammatory diseases of the myocardium from idiopathic and chronic forms of dilated cardiomyopathy is clinically important because their long-term outcomes differ. Accurately diagnosing myocardial histopathology may be more relevant in the near future as effective treatments for myocarditis, sarcoid, and other myocardial diseases evolve. The cost and the low sensitivity and specificity for endomyocardial biopsy suggest that noninvasive testing is essential for deciding which patients with unexplained cardiomyopathy should undergo this procedure. Novel biomarkers may be of use in identifying a cohort of such patients. Radionuclide techniques, particularly anti-
388
dec
myosin scintigraphy, also have a high negative predictive value and can be used in some settings to avoid biopsy. Contrast-enhanced MRI seems to be the most promising technique for diagnosing myocardial inflammation and myocyte injury based on small, nonprospective clinical studies. MRI may not only be useful in identifying those patients who should undergo biopsy but may allow a guided approach to the abnormal region of myocardium. It is hoped that this methodology improves the sensitivity of the technique for establishing a correct histologic diagnosis. Serial MRI studies have also shown promise for tracking the natural history of the disease and could, in the future, allow noninvasive reassessment of the myocardial response to therapy. As with any new diagnostic technology, preliminary observations generate challenges and opportunities. The early identification of patients with active myocarditis may more precisely reformulate the approach to the diagnosis and treatment of unexplained cardiomyopathy. Key unanswered questions include the following: 1. Can specific subgroups of patients with myocarditis, who may respond to immunosuppressive or immunomodulatory therapy, be identified by noninvasive imaging studies? 2. Can MRI or other methods differentiate between cardiomyopathies of different etiology (eg, myocarditis versus alcohol-related injury versus familial forms)? 3. Should right ventricular biopsy ultimately be replaced by MRI-guided left ventricular sampling? Research studies that address these vital questions, and prospective studies that compare the use of MRI techniques with other newer imaging techniques (eg, specifically annexin imaging for detecting myocardial apoptosis) in acute cardiomyopathy should pave the way for more effective treatment of viralmediated cardiomyopathies in the decade ahead.
References [1] Mason JW, Billingham JE, Ricci DR. Treatment of acute inflammatory myocarditis assisted by endomyocardial biopsy. Am J Cardiol 1980;45:1037 – 44. [2] Aretz HT, Billingham ME, Edwards WD, et al. Myocarditis: a histopathologic definition and classification. Am J Cardiovasc Pathol 1986;1:1 – 14. [3] Lieberman EB, Hutchins GM, Herskowitz A, et al. Clinicopathologic description of myocarditis. J Am Coll Cardiol 1991;18:1617 – 26. [4] Kuhl U, Seeberg B, Schultheiss HP, et al. Immuno-
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
histological characterization of infiltrating lymphocytes in biopsies of patients with clinically suspected dilated cardiomyopathy. Eur Heart J 1994;15(Suppl C): 62 – 7. Chow LH, Radio SJ, McManus BM. Insensitivity of right ventricular endomyocardial biopsy in the diagnosis of myocarditis. J Am Coll Cardiol 1898;14: 915 – 20. Hauk AJ, Kearney DL, Edwards WD. Evaluation of post-mortem endomyocardial biopsy specimens from 38 patients with lymphocytic myocarditis: implications for the role of sampling error. Mayo Clin Proc 1989; 64:1235 – 45. Dec GW, Fallon JT, Southern JF, et al. Borderline myocarditis: an indication for repeat biopsy. J Am Coll Cardiol 1990;15:283 – 9. Wu LA, Lapeyre AC, Cooper LT. Current role of endomyocardial biopsy in the management of dilated cardiomyopathy and myocarditis. Mayo Clin Proc 2001;76:1030 – 8. Pinamonti B, Alberti E, Cigalotto A, et al. Echocardiographic findings in myocarditis. Am J Cardiol 1988; 62:285 – 91. Pinamonti B. Contribution of echocardiography to the diagnosis of patients with chronic heart failure. Ital Heart J 2000;1:1311 – 6. Adsett M, West MJ, Galbraith A, et al. Eosinophilic heart: marked left ventricular wall thickening and myocardial dysfunction improving on corticosteroid therapy. Echocardiography 2003;20:369 – 74. Schinkel AFL, Bax JJ, van der Wall EE, et al. Echocardiographic follow-up of Chlamydia psittaci myocarditis. Chest 2000;117:1203 – 5. Carvalho JS, Silva CM, Shinebourne EA, et al. Prognostic value of posterior wall thickness in childhood dilated cardiomyopathy and myocarditis. Eur Heart J 1996;17(8):1233 – 8. Mendes LA, Picard MH, Dec GW, et al. Ventricular remodeling in active myocarditis. Am Heart J 1999; 138:303 – 8. Omi W, Nagai H, Takata S, et al. Ultrasonic tissue characterization in acute myocarditis. Circ J 2002;66: 416 – 8. Lieback E, Hardouin I, Meyer R, et al. Clinical value of echocardiographic tissue characterization in the diagnosis of myocarditis. Eur Heart J 1996;17: 135 – 42. Urhausen A, Kindermann M, Bohm M, et al. Diagnosis of myocarditis by cardiac velocity imaging in an Olympic athlete. Circulation 2003;108:e21 – 2. Palka P, Lange A, Clarke B, et al. Echocardiographic description of recurrent idiopathic giant-cell myocarditis in cardiac allograft. Circulation 2003;108:e1 – 3. O’Connell JB, Robinson JA, Kenkin RE, et al. Immunosuppressive therapy in patients with congestive cardiomyopathy and myocardial uptake on gallium-67. Circulation 1981;64:780 – 6. O’Connell JB, Henkin RE, Robinson JA, et al. Gallium-67 imaging in patients with dilated cardiomy-
noninvasive imaging techniques
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
opathy and biopsy proven-myocarditis. Circulation 1984;70:58 – 62. Khaw BA, Beller GA, Haber E, et al. Localization of cardiac myosin-specific antibody in myocardial infarction. J Clin Invest 1976;58:439 – 46. Yasuda T, Palacios IF, Dec GW, et al. Indium 111-monoclonal antimyosin antibody imaging in the diagnosis of acute myocarditis. Circulation 1987;75: 306 – 11. Haber E, Yasuda T, Palacios IF, et al. Antimyosin antibody imaging in the diagnosis of acute myocarditis. Eur Heart J 1987;8(Suppl J):119 – 23. Rezkalla S, Kloner RA, Khaw BA, et al. Detection of experimental myocarditis by monoclonal antimyosin antibody Fab fragments. Am Heart J 1989;117:391 – 5. Matsumori A, Ohkusa T, Mataba Y, et al. Myocardial uptake of antimyosin monoclonal antibody in a murine model of viral myocarditis. Circulation 1989;79:400 – 5. Kishimoto C, Hung GL, Ishibashi M, et al. Natural evolution of cardiac function, cardiac pathology, and antimyosin scans in a murine myocarditis model. J Am Coll Cardiol 1991;17:821 – 7. Dec GW, Palacios I, Yasuda H, et al. Antimyosin antibody cardiac imaging: its role in the diagnosis of myocarditis. J Am Coll Cardiol 1990;16:97 – 104. Margari ZJ, Anastasiou-Nana MI, Terrovitis J, et al. Indium-111 monoclonal antimyosin cardiac scintigraphy in suspected acute myocarditis: evolution and diagnostic impact. Int J Cardiol 2003;90:239 – 45. Dec GW, Narula J. Antimyosin imaging for noninvasive recognition of myocarditis. Quart J Nucl Med 1997;41(Suppl 1):128 – 39. Narula J, Southern JF, Abraham S, et al. Myocarditis simulating myocardial infarction. J Nucl Med 1991;32: 312 – 8. Sarda L, Colin P, Boccara F, et al. Myocarditis in patients with clinical presentation of myocardial infarction and normal coronary angiograms. J Am Coll Cardiol 2001;37:786 – 92.
389
[32] Koelling TM. The use of magnetic resonance imaging in cardiac transplant rejection. In: Dec GW, Narula J, Ballester M, et al, editors. Cardiac allograft rejection. Boston7 Kluwer Academic Publishers; 2001. p. 321 – 38. [33] Aherne T, Tscholakoff D, Finkbeiner W, et al. Magnetic resonance imaging of cardiac transplants: the evaluation of rejection of cardiac allografts with and without immunosuppression. Circulation 1986;74: 145 – 56. [34] Chandraratna PA, Bradley WG, Kortman KE, et al. Detection of acute myocarditis using nuclear magnetic resonance imaging. Am J Med 1987;83:1144 – 6. [35] Gagliardi MG, Bevilacqua M, Di Renzi P, et al. Usefulness of magnetic resonance imaging for diagnosis of acute myocarditis in infants and children, and comparison with endomyocardial biopsy. Am J Cardiol 1991;68:1089 – 91. [36] Gagliardi MG, Polletta B, Di Renzi P. MRI for the diagnosis and follow-up of myocarditis. Circulation 1998;98:458 – 9. [37] Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1802 – 9. [38] Reiker O, Mohrs O, Oberholzer K, et al. Cardiac MRI in suspected myocarditis. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 2002;174:1530 – 6. [39] Roditi GH, Hartnell GG, Cohen MC. MRI changes in myocarditis-evaluation with spin-echo, cine MR angiography and contrast enhanced spin echo imaging. Clin Radiol 2000;55:752 – 8. [40] Schulz-Menger J, Dietz R, Friedrich MG. Longterm follow-up of patients with acute myocarditis by magnetic resonance imaging. MAGMA 2003;16: 17 – 20. [41] Mahrholdt H, Goedecke C, Wagner A, et al. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation 2004;109:1250 – 8.
Noninvasive Imaging Techniques for the Diagnosis of Myocarditis G. William Dec, MDa,b,T a
Massachusetts General Hospital, Boston, MA, USA b Harvard Medical School, Boston, MA, USA
The clinical manifestations of active myocarditis are highly varied and are generally not specific enough to establish a diagnosis with certainty. Clinicians have increasingly relied during the past two decades on right (or left) ventricular endomyocardial biopsy for histologic confirmation of suspected inflammatory heart disease. Mason and coworkers [1] were among the first investigators to demonstrate evidence for myocarditis using right ventricular endomyocardial biopsy in a small group of patients with previously diagnosed idiopathic dilated cardiomyopathy. Although endomyocardial biopsy has become the international standard for establishing the diagnosis, the exact histologic criteria remain controversial. To provide more uniform criteria for establishing the pathologic diagnosis, a panel of cardiac pathologists developed a classification known as the Dallas criteria [2]. In this schema, myocarditis is characterized by an inflammatory cellular infiltrate within the myocardium with necrosis or degeneration of adjacent myocytes (or both), a pattern that is not typically seen in ischemic injury. The inflammatory infiltrate is typically lymphocytic but may also include eosinophilic, neutrophilic, granulomatous, or mixed cellularity. The amount of inflammation may be mild, moderate, or severe and its distribution may be focal, confluent, or diffuse, respectively. Despite the widespread adoption of this histopathologic classification, some clinicians believe that the definition
T Harvard Medical School, 55 Fruit Street, Bigelow 800, Mailstop 817, Boston, MA 02114. E-mail address: [email protected]
is too restrictive and have proposed alternatives based on clinicopathologic criteria or immunohistochemical findings [3,4]. Sampling error remains the most crucial limitation to the diagnostic accuracy of the endomyocardial biopsy. Chow and coworkers [5] provided clear evidence for the insensitivity of right ventricular biopsy in detecting myocarditis. Likewise, Hauck and coworkers [6] analyzed hearts from autopsy specimens in which myocarditis was determined to have contributed directly to death. Ten biopsies from the apex and septum of both ventricles were evaluated for myocarditis using the Dallas criteria. When only five right ventricular samples from each heart were evaluated (which is the most common clinical sampling rate), the diagnosis of myocarditis could not be established in 55% of the hearts (Fig. 1). In a similar postmortem study of 14 hearts, 17.2 samples per heart were required to diagnose myocarditis correctly in more than 80% of cases [5]. Dec and coworkers [7] examined the role of repeat right and left ventricular biopsy among patients who were strongly suspected of having myocarditis clinically but whose initial right ventricular biopsy failed to provide histologic confirmation. Biopsy of both ventricles detected an additional 15% incidence of myocarditis. A positive endomyocardial biopsy unequivocally establishes the diagnosis; however, the absence of conclusive histologic confirmation does not exclude consideration of myocarditis in most clinical settings. Given the focal or multifocal nature of this disease process, it is not surprising that substantial sampling error exists. Clinicians are increasingly reluctant to recommend routine biopsy even when myocarditis is clinically
1551-7136/05/$ – see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.hfc.2005.06.007
heartfailure.theclinics.com
378
dec
Fig. 1. Sensitivity of endomyocardial biopsy for detecting myocarditis in a postmortem study of 38 hearts with proven myocarditis. Each section demonstrates the rate of detection of each additional endomyocardial biopsy specimen. One biopsy detected myocarditis in 18% of specimens. Light gray shading: five biopsy specimens detected myocarditis in only 43% of specimens. White shading: 17 biopsies confirmed myocarditis in 82% of explanted hearts. (From Dec GW. Clinical myocarditis. In: Cooper L, editor. Myocarditis: bench to bedside. Totowa (NJ): Humana Press; 2003. p. 261. Copyright Mayo Clinic Foundation; with permission.)
strongly suspected. The contemporary role of endomyocardial biopsy in the diagnosis and management of myocarditis has recently been reviewed [8]. Given the cost of biopsy, its invasiveness, and the relatively low histologic rate of detection for myocarditis or other treatable forms of acute left ventricular dysfunction (eg, cardiac sarcoidosis), sensitive and specific noninvasive screening methodologies are urgently needed for the diagnostic evaluation of this population. Serum biomarkers including creatine kinase and the troponins are increasingly recognized as having predictive utility as screening tools. Additional prospective confirmatory studies are needed, however, to verify the predictive accuracy of these biomarkers. Noninvasive imaging techniques including echocardiography, radionuclide imaging, and MRI have been recently studied and may provide additional diagnostic information.
Echocardiographic findings in myocarditis Echocardiography is currently recommended as the most important initial diagnostic study in patients with heart failure and frequently provides a substantial contribution to the differential diagnosis of disease etiology. Several studies have evaluated transthoracic echocardiography for diagnosing myocarditis [9,10]. Pinamonti and coworkers [9] retrospectively analyzed echocardiographic findings among 42 patients with
biopsy-proven myocarditis. Clinical presentations included heart failure (63%); high-grade atrioventricular block (17%); chest pain despite normal coronary anatomy (15%); and supraventricular arrhythmias (5%). Left ventricular dysfunction was commonly observed (69%), particularly among patients with congestive heart failure (present in 88%). Left ventricular cavity enlargement was frequently minimal or absent, consistent with other forms of acute dilated cardiomyopathy. Right ventricular dysfunction was present in only 23% of this cohort. Not surprisingly, patients who presented with chest pain or heart block almost always had preserved ventricular size and function. Segmental wall motion abnormalities were observed in 64% of patients and included hypokinetic, akinetic, or frankly dyskinetic regions. Reversible left ventricular hypertrophy was noted in 15% of patients and resolved over several months. Echocardiographic findings were quite varied and relatively nonspecific. More recently, Pinamonti [10] reviewed the diagnostic use of echocardiography in diagnosing patients with chronic heart failure. It was concluded that ‘‘no useful echocardiographic findings can identify patients with genetic dilated cardiomyopathy or affected by myocarditis from other cases of idiopathic dilated cardiomyopathy’’ [10]. This statement is not meant to imply that routine echocardiography has no role in the evaluation and management of myocarditis patients. Serial studies have been shown to
noninvasive imaging techniques
be useful in assessing the response to treatment of several forms of myocarditis. Resolution of marked concentric left ventricular hypertrophy in eosinophilic myocarditis following corticosteroid treatment has been reported [11]. Similarly, echocardiography has provided a useful method of following-up chlamydial myocarditis [12]. It has also been used to evaluate prognosis in childhood myocarditis. Carvalho and coworkers [13] have described the use of the ratio of end-diastolic posterior wall thickness to cavity dimension in the pediatric population. Relative preservation of posterior wall thickness (ratio > 0.17) was predictive of better prognosis and higher likelihood of recovery. Mendes and coworkers [14] examined the process of ventricular remodeling among patients with active myocarditis enrolled in the Multicenter Myocarditis Treatment Trial. Compared with normal controls, myocarditis patients demonstrated larger mean left ventricular volumes (81 ± 29 mL/m2 versus 50 ± 8 mL/m2; P = .001). Chamber dilatation occurred primarily along the mid-cavity diameter, which measured 5.3 ± 0.8 cm among myocarditis patients compared with 4.2 ± 0.4 cm in controls (P=.001). Active myocarditis was associated with early left ventricular remodeling and the development of a more spherical chamber. Multivariable modeling of echocardiographic parameters demonstrated only baseline left ventricular ejection fraction (but not degree of left ventricular dilatation or sphericity) to be an independent predictor of longterm survival. Although anatomic features on echocardiography (ie, chamber dimensions, ejection fraction, wall motion abnormalities) are insufficient to differentiate myocarditis from other forms of cardiomyopathy, ultrasonic tissue characterization may be of more use. Myocarditis is associated with inflammation, myocardial edema, and fibrosis of varying degrees of severity. Ultrasound for myocardial tissue characterization is based on the hypothesis that quantifiable connections exist between the tissue structures and their acoustical properties. Transmission and reflection of ultrasound in tissue depends on tissue density, elasticity, and acoustical impedance. Changes in one or more of these factors lead to different ultrasonic backscatter and altered image texture. A systematic relationship exists between tissue structure and echocardiographic texture. Lieback and coworkers [15] evaluated mean gray scale values (indicative of average brightness) in 52 patients with biopsy-proven myocarditis, 12 patients with persistent myocarditis, 9 patients with healed myocarditis who lacked fibrosis, and 17 patients with healed myocarditis and fibrosis. Echocardiographic features were com-
379
pared with normal controls and a cohort with dilated cardiomyopathy. Tissue characterization was highly effective in differentiating myocarditis from healthy control myocardium with sensitivity and specificity values of 100% and 90%, respectively [16]. Ultrasonic tissue characterization could not accurately differentiate, however, between idiopathic dilated cardiomyopathy and active myocarditis. More recent techniques have been able better to characterize tissue changes in acute myocarditis and to evaluate changes in these parameters over time [16]. Cardiac tissue Doppler imaging has been shown successfully to detect biopsy-verified myocarditis in a former Olympic cyclist with recurrent ventricular tachycardia [17]. M-mode Doppler tissue echocardiography has also been shown to detect recurrent giant cell myocarditis in transplant recipients [18]. An abnormal myocardial velocity gradient across the left ventricular posterior wall during isovolumic relaxation and a reduction during rapid ventricular filling indicative of impaired myocardial relaxation have been demonstrated during the active phase of the disease (Fig. 2). These abnormal findings resolved completely after augmented immunosuppressive therapy. These limited case reports suggest that newer echocardiographic imaging modalities may possess sufficient resolution effectively to screen for inflammatory diseases of the myocardium. Additional validation studies are required to determine their predictive accuracy.
Nuclear imaging techniques Gallium-67 (67Ga) cardiac scintigraphic imaging has been used to evaluate conditions that result in myocardial inflammation, a key component of myocarditis. Early reports from centers with extensive experience using this qualitative technique demonstrate, however, that 67Ga imaging can be useful as a screening tool and for predicting therapeutic response [19,20]. O’Connell and coworkers [20], who have evaluated this methodology most extensively, reported a sensitivity of 36% but a specificity of 98% for histologic diagnosis of myocarditis. The difficulty of the technique combined with its low specificity has led to its infrequent use. Indium-111 (111In) – labeled monoclonal antibody fragments (directed against heavy chain myosin) bind to cardiac myocytes that have lost the integrity of their sarcolemmal membranes and have exposed their intracellular myosin to the extracellular fluid space [21]. Unlike 67Ga, which detects the extent of myocardial inflammation, antimyosin cardiac uptake reflects the extent of myocyte necrosis. Because both
380
dec
Fig. 2. Echocardiographic findings in giant cell myocarditis. (A) M-mode Doppler tissue echocardiography (DTE) of the left posterior wall with quantification of myocardial velocity gradient (MVG) and mean myocardial velocities (MMV); parasternal long-axis view (top). (B) Corresponding left ventricular filling pattern on Doppler mitral inflow (E-wave deceleration time, 138 millisecond). During isovolumic relaxation (IVR) (A, bottom), MVG is abnormally positive at 2.9 s 1 (normal range, 2.3 to 1.4 s 1). During rapid ventricular filling (RVF), just after the opening of the mitral valve, MVG is significantly reduced to 2.8 s 1 (normal range, 5.9 to 18.1 sec 1). PCG, phonocardiogram. Dotted vertical lines indicate isovolumic relaxation time. (From Palka P, Lange A, Clarke B, et al. Echocardiographic description of recurrent idiopathic giant-cell myocarditis in cardiac allograft. Circulation 2003;108:e1 – 3; with permission.)
elements are present in active myocarditis, the two imaging modalities may provide complementary information. Unfortunately, no studies have directly compared the use of these radionuclide techniques. Published data suggest that antimyosin imaging may have higher negative predictive value than 67Ga scintigraphy [22]. Yasuda and coworkers [22] first reported antimyosin scintigraphic imaging in 28 patients with clinical findings suggestive of acute myocarditis. Most (75%) had unexplained heart failure of less than 6 months’ duration. Four patients presented with ventricular tachyarrhythmias and three patients had chest pain mimicking acute myocardial infarction. All patients underwent right ventricular biopsy and antimyosin imaging within 7 days. The mean left ventricular ejection fraction was 27% ± 2%. Antimyosin imaging was found to be positive in 61% of patients (Fig. 3). Cardiac uptake was generally heterogeneous and diffuse within the left ventricular myocardium. Importantly, none of the patients with a negative antimyosin scan had histologic evidence of myocarditis and all patients with biopsy-proven myocarditis had a positive scan. Similar findings were reported in a small series by Haber and coworkers [23]. Rezkalla and coworkers [24] validated this technique in a murine model of Coxsackie-induced myocarditis. Following intraperitoneal injection of virus,
the acute phase of the disease is characterized by high-grade viremia, viral replication in the myocardium, and the development of an intense mononuclear cellular infiltrate associated with focal myocyte necrosis. Histopathologic changes and the temporal course of the disease mimic human myocarditis. One, 2, 4, and 12 weeks following Coxsackie inoculation, mice underwent 125In-labeled antimyosin Fab antibody injection (25 mCi); the hearts were excised at 48 hours for assessment using gamma scintillation counting. Myocardial radioactive uptake from the infected animals was significantly higher than controls, from 1 week to 12 weeks following inoculation. More importantly, the degree of radiotracer uptake in the hearts correlated closely with histopathologic changes. Autoradiography revealed localization of radioactivity to areas of myocardial necrosis. Matsumori and coworkers [25] and Kishimoto and coworkers [26] confirmed this close correlation between the extent of 111In-labeled antimyosin uptake and the degree of histologically confirmed myocardial necrosis in an encephalomyocarditis murine model. Dec and coworkers [27] subsequently studied the use of antimyosin imaging in a large cohort (N = 82) of patients with clinically-suspected myocarditis. Presenting symptoms included heart failure and dilated cardiomyopathy (92%); chest pain mimicking acute myocardial infarction (6%); or life-threatening
noninvasive imaging techniques
381
Fig. 3. Antimyosin cardiac imaging for detecting active myocarditis. A positive antimyosin image demonstrates diffuse tracer uptake in the cardiac region on both the anterior planar image (upper left) and in all coronal tomographic reconstructions (bottom left). Biopsy confirmed multifocal lymphocytic myocarditis. Antimyosin imaging was repeated after 6 months of immunosuppressive therapy (top right). Biopsy showed healed myocarditis. No antimyosin uptake is visible on either the planar (top right) or tomographic reconstructions (bottom right). (From Dec GW. Clinical myocarditis. In: Cooper L, editor. Myocarditis: bench to bedside. Totowa (NJ): Humana Press; 2003. p. 273. Copyright Mayo Clinic Foundation; with permission.)
ventricular arrhythmias (2%). Most patients presented with symptoms < 12 months’ duration; five patients had a prior bout of biopsy-proven myocarditis with a second episode of acute heart failure. Following injection of 111In-labeled antimyosin antibody fragments (500 mg coupled to 2 mCi), all patients underwent planer and single-photon emission CT cardiac imaging at 48 hours. Right ventricular biopsy was performed within 48 hours of imaging. On the basis of right ventricular histologic features, antimyosin was highly sensitive but only moderately specific for detecting myocardial necrosis. The sensitivity was 83% and specificity 53%; however, the predictive value of a negative scan was 92%. More recently, Margari and coworkers [28] reported that the presence of both a positive antimyosin scan and a nondilated left ventricle (left ventricle end-diastolic diameter 62 mm) was highly predictive of biopsyproven myocarditis. Perhaps more important than the correlation between cardiac antimyosin uptake and histologic findings, improvement in ventricular function within 6 months of diagnosis was evident in 54% of patients with a positive antimyosin scan compared with only 18% of those with a negative scan [29]. Because spontaneous improvement in ventricular function is a recognized feature of acute myocarditis, it is possible that a number of patients who were scan-positive but biopsy-negative may have, in fact, had myocarditis that was not detected on biopsy because of sampling error. Repeat antimyosin imaging among a cohort whose initial scan was positive showed persistent
uptake in 50% and resolution in the remaining patients. No correlation was evident between ongoing myocarditis on repeat biopsy and clinical improvement [29]. Narula and coworkers [30] also evaluated the role of this technique among patients who presented with chest pain mimicking acute myocardial infarction. Following demonstration of normal coronary anatomy, antimyosin imaging was performed. Uptake was global in seven of eight patients with confirmed myocarditis on biopsy; the eighth patient had borderline myocarditis. In contrast, antimyosin uptake is typically segmental among patients with acute myocardial infarction and almost always confined to the territory of the infarct-related vessel. Sarda and coworkers [31] have also confirmed the ability of antimyosin imaging to detect myocarditis in patients presenting with a clinical syndrome of acute myocardial infarction and normal coronary angiography. Antimyosin imaging may, at times, be useful in differentiating unstable coronary syndromes from acute myocarditis. Its high sensitivity but modest specificity suggests that antimyosin scintigraphy may serve as an excellent initial screening tool to determine which patients should undergo further diagnostic assessment for myocarditis, including biopsy.
MRI MRI is rapidly becoming the gold standard for assessing both right and left ventricular anatomy and
382
dec
Box 1. Applications of MRI in cardiovascular diseases for assessing anatomic changes, tissue relaxation properties, perfusion characteristics, and chemical composition Left and right ventricular mass Ventricular wall thickness (hypertrophy, thinning) Left and right ventricular contractile function T1 relaxation time T2 relaxation time Myocardial perfusion imaging Phosphorus-31 spectroscopy Sodium-23 spectroscopy
for measurement of ventricular function (Box 1). Besides providing anatomic and morphologic information, MRI techniques can provide accurate tissue characterization by measuring T1 and T2 relaxation times and spin density [32]. T1 relaxation is defined by the rate of longitudinal relaxation of protons in the magnetic field (Fig. 4). T2 relaxation is related to the
rate at which proton spins fall out of phase (Fig. 5). Increases of both these relaxation times correlate well with increased tissue water content [32]. Because active myocarditis is generally associated with myocyte injury including edema and cellular swelling, assessment of the relaxation times may provide a sensitive measure for detecting myocarditis. Increases in T2 relaxation have also been reported to occur with focal hemorrhage and myocyte necrosis [32]. Prior animal and human studies have demonstrated the reliability of MRI for tissue characterization of cardiac allograft rejection [32]. Aherne and coworkers [33] examined heterotopic cardiac transplantation in immunosuppressed (prednisone, 1 mg/kg/d, and cyclosporine, 25 mg/kg/d) dogs and compared them with animals that also underwent transplantation but did not receive immunosuppressive therapy. Untreated allografts showed a significant increase in T2 relaxation time (60 ± 8 versus 44 ± 6 milliseconds; P < .01) compared with control hearts as early as 1 week following transplantation. More importantly, T2 relaxation times and signal intensity increased in allografts that developed histologic evidence of acute cellular rejection. There was a significant correlation between histologic grading of rejection severity and T2 relaxation times in the cardiac allograft (r value = 0.72).
Fig. 4. Longitudinal relaxation in the B0 field is described by the relaxation constant T1. T1 can be measured by performing an inversion recovery pulse sequence using a 180-degree pulse. As the time to imaging (TI) is delayed further and further after the inversion pulse, the overall magnetization in the B0 field becomes closer to the fully relaxed state, M0. The signal intensity increases as the imaging time increases, obeying the equation: SI = M0* (1 to 2e (TI/T1). (From Koelling TM. The use of magnetic resonance imaging in cardiac transplant rejection. In: Dec GW, Narula J, Ballester M, et al, editors. Cardiac allograft rejection. Boston: Kluwer Academic Publishers; 2001. p. 324; with permission.)
noninvasive imaging techniques
383
Fig. 5. T2 is the time constant that describes the way spins diphase after being focused by a 90-degree excitation pulse. Signal is maximal immediately after the excitation phase, and degrades with time as spins fall out of phase and net magnetization in the transverse plan approaches zero. The magnetization after any echo time, TE, is described by the equation: SI = M0e (TE/T2)). (From Koelling TM. The use of magnetic resonance imaging in cardiac transplant rejection. In: Dec GW, Narula J, Ballester M, et al, editors. Cardiac allograft rejection. Boston: Kluwer Academic Publishers; 2001. p. 325; with permission.)
There was also a significant linear relationship between in vivo T2 values and difference in percent water content in the allograft and pretransplant native heart (r = 0.92, P < .001) (Fig. 6). Because myocarditis demonstrates many of the histologic features of acute cardiac allograft rejection, it is not surprising that MRI has also been used as a diagnostic tool in this disease. A preliminary study by Chandraratna and coworkers [34] was the first to report localized myocardial edema in regions of hypokinesis or akinesis on echocardiography. Only two patients were reported and myocarditis was based solely on clinical criteria. Importantly, improvement in ventricular function was associated with resolution of myocardial edema on MRI in this study. Gagliardi and coworkers [35] evaluated MRI and concurrent endomyocardial biopsy findings in 11 consecutive children (age 9 months to 9 years) with clinically suspected myocarditis. Tissue characterization was obtained in regions of interest of the right and left ventricles by using T1 and T2 spin echo sequences [35]. The myocardial-skeletal muscle signal intensity ratio was calculated and accurately identified all patients with histologically verified myocarditis. Although encouraging, these results were obtained in a very small number of patients. Further, myocarditis in children is often associated with a more promi-
nent interstitial edema than is typically observed in adults. The same authors subsequently have reported their experience in 75 consecutive pediatric patients with acute heart failure symptoms, left ventricular systolic dysfunction, and the absence of congenital heart disease [36]. All children underwent concurrent endomyocardial biopsy and cardiac MRI. T2 relaxation sequences correctly identified patients with and without histologic evidence for myocarditis with a high degree of predictive accuracy. Endomyocardial biopsy identified histologic evidence of acute myocarditis in 51 patients; the remaining 24 individuals had idiopathic dilated cardiomyopathy. Compared with biopsy findings, MRI achieved a sensitivity of 100% and a specificity of 90% in this cohort [36]. The 53 children with myocarditis have undergone repeated biopsy and MRI investigation every 6 months to evaluate the efficacy of MRI in identifying persistent, resolving, or resolved myocarditis during a 2-year follow-up [36]. Preliminary data have confirmed the use of MRI signal intensity by T2-weighted relaxation sequences during follow-up because the sensitivity and specificity remain high during the entire evolution of the disease [36]. The diagnostic usefulness of contrast-enhanced cardiac MRI has also been evaluated [37 – 41]. In a
384
dec
Fig. 6. Correlation between percent difference in T2 relaxation times (transplanted hearts minus native hearts) and percent difference in water content (transplanted hearts minus native hearts). (From Aherne T, Tscholakoff D, Finkbeiner W, et al. Magnetic resonance imaging of cardiac transplants: the evaluation of rejection of cardiac allografts with and without immunosuppression. Circulation 1986;74:145 – 56; with permission.)
landmark study, Friedrich and coworkers [37] examined 19 patients with suspected viral myocarditis based on a combination of clinical and laboratory findings. Inclusion criteria included a history of febrile viral illness within 4 weeks of study entry; one or more symptoms (fatigue, malaise, dyspnea, chest pain, fever, or tachycardia); one or more electrocardiographic changes (atrioventricular block, ST segment depression or elevation in three or more leads, recurrent atrial or ventricular tachyarrhythmias); regional or global left ventricular dysfunction by echocardiography and left ventriculography; demonstration of normal coronary anatomy; positive viral serology (Coxsackie B virus, 11 patients; cytomegalovirus, 4 patients; Epstein-Barr virus, 3 patients; and influenza, 1 patient); and a positive cardiac antimyosin scan. Seven of the 19 patients underwent biopsy; myocarditis was detected in four individuals. ECG-triggered, T2-weighted, multislice spin echo sequences were performed in the axial orientation (five acquisitions, slice thickness 6 mm) for controls and suspected myocarditis patients. In addition, ECGtriggered, T1-weighted multislice spin echo images were obtained in the axial and short-axis orientation (four to six acquisitions, slice thickness 6 mm). Repeat imaging was performed beginning 1 minute after injection of 0.1 mmol/kg of gadolinium. Manual regions of interest were drawn for cardiac and skeletal muscle and compared with measurements obtained in
18 volunteers. Myocarditis patients were serially studied on days 2 ± 7, 14 ± 2, 28 ± 4, and 84 ± 10 after symptom onset. The T1-weighted images were found to provide the most useful information. Early following presentation, myocardial enhancement was generally focal in distribution (Fig. 7). Global enhancement became prominent during the later imaging times (Fig. 8). Overall, global left ventricular enhancement was substantially higher in the myocarditis patients than in the controls on days 2 through 28 after symptom onset. The global enhancement index had returned to that observed in controls by the day 84 study (Fig. 9). Unfortunately, the study did not examine the ability of MRI to differentiate viral myocarditis from other causes of acute dilated cardiomyopathy. It did suggest, however, that longitudinal follow-up of the same patient with active disease is feasible and allows for reassessment of persistent myocarditis or recurrent disease. Roditi and coworkers [39] evaluated 20 patients with T1 spin echo cine MR angiography and contrastenhanced spin echo imaging before and after gadolinium enhancement. Four patients had histologically proven myocarditis; eight patients had clinically suspected myocarditis, and eight patients had chronic cardiomyopathy. Focal myocardial enhancement was associated with regional wall motion abnormalities in 10 out of the 12 patients with suspected or proven myocarditis. Using this technique, the myocarditis
noninvasive imaging techniques
385
Fig. 7. T1-weighted MRI cross-sectional views at the midventricular level in a patient with acute myocarditis. Unenhanced view obtained on day 2 after onset of symptoms (left panel). There are small foci of increased signal intensity in the subepicardial parts of the posterior myocardium and in the basal septum, which were more evident (right panel) following administration of gadopentetate dimeglumine (arrows). (From Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1805; with permission.)
cohort demonstrated a mean focal enhancement of 122% (corrected for a background skeletal muscle enhancement), which was significantly different from the mean of 14% observed in the cardiomyopathy (nonmyocarditis) group (Fig. 10). The authors suggest that the findings of focal myocardial enhancement by MRI combined with regional wall motion abnormalities (hypokinesis, akinesis, or dyskinesis) strongly support the diagnosis of myocarditis. The ability of contrast-enhanced MRI techniques to diagnose other forms of inflammatory heart disease, particularly cardiac sarcoidosis, has recently been validated [38]. New contrast MRI techniques using segmented inversion recovery gradient echo pulse sequences provide substantial improvement in contrast between diseased and normal myocardium. These newer techniques may provide up to a 500% improvement
in resolution compared with the protocol used by Frederich and coworkers [37,41]. Mahrholdt and coworkers [41] recently used this new technique to perform MRI-guided biopsy of the right and left ventricles in patients with possible myocarditis. Thirty-two patients were included in the study. Each had clinically suspected myocarditis based on flulike symptoms within 8 weeks of presentation; malaise; chest pain; dyspnea; or tachycardia and electrocardiographic evidence for atrioventricular block, ST segment depression, or unexplained ventricular arrhythmias. Cine and contrast-enhanced ECG-gated MRI was performed using a 1.5-T system, 6-mm thickness slices, and 15-second breathhold technique per slice. Following initial image acquisition, gadoteridol (0.1 mmol/kg) contrast was administered intravenously and repeat imaging was performed 10 minutes later. Left ventricular biopsy was gen-
Fig. 8. T1-weighed MRI obtained before (left panel) and after enhancement (right panel) in the same patient as shown in Fig. 7 at 14 days following presentation. More diffuse enhancement of myocardium after gadopentetate dimeglumine, including the apical part of the septum and visible areas of the right ventricle. (From Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1805; with permission.)
386
dec
Fig. 9. Mean ± SE of the global relative MRI signal enhancement of cardiac tissue compared with skeletal muscle tissue on days 2, 7, 14, 28, and 84 after onset of acute myocarditis. There was a significant difference in the mean results from days 2 through 28 compared with control subjects. (From Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1807; with permission.)
Fig. 10. Myocardial enhancement of T1-weighted spin echo MRI in patients with known or suspected myocarditis (Myocarditis) and those without suspected myocarditis (Not Myocarditis), both uncorrected and corrected for background skeletal muscle enhancement. Box, 25th to 75th percentiles around the median (line); whiskers, 10th to 90th percentiles; x, individual measures with outliners circled. (From Roditi GH, Hartnell GG, Cohen MC. MRI changes in myocarditis: evaluation with spin-echo, cine MR angiography and contrast enhanced spin echo imaging. Clin Radiol 2000;55:756; with permission.)
noninvasive imaging techniques
387
Fig. 11. Cardiac MRI findings and histopathologic findings of typical patients in whom biopsies were obtained from areas of MRI contrast enhancement. Top 3 panels show cases (patients 6, 7, and 14) of active myocarditis with myocyte damage and infiltration of macrophages; bottom panel illustrates a patient (patient 18) without active myocarditis who was diagnosed with hypertrophic cardiomyopathy. SAX, short-axis orientation; LAX, long-axis orientation. (From Mahrholdt H, Goedecke C, Wagner A, et al. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation 2004;109:1253; with permission.)
erally performed from the region showing the most marked contrast enhancement (Fig. 11). Right ventricular septal biopsies were performed for patients with an inaccessible anterobasilar left ventricular location (N = 7); prominent septal enhancement (N = 2); or those who lacked any evidence of focal enhancement (N = 4). Contrast enhancement was present in 28 patients (88%) and was usually seen with one or several distinct regions in the myocardium. Regional enhancement was most frequently observed in the lateral wall. For the 21 patients in whom biopsies were obtained directly from the region of contrast enhancement, histopathology revealed active myocarditis in 19 patients. Conversely, in the remaining patients (N = 11) in whom biopsy was not taken from the region of contrast-enhancement, active myocarditis was detected in only one case. At follow-up, the area of contrast enhancement decreased from 9% ± 11% to 3% ± 4% and the mean left ventricular ejection fraction improved from 47% ± 19% to 60% ± 10% [41]. Overall, biopsy of those specific myocar-
dial regions enhanced by MRI resulted in positive and negative predictive values for detecting myocarditis on biopsy, 71% and 100%, respectively.
Summary Accurate differentiation of active inflammatory diseases of the myocardium from idiopathic and chronic forms of dilated cardiomyopathy is clinically important because their long-term outcomes differ. Accurately diagnosing myocardial histopathology may be more relevant in the near future as effective treatments for myocarditis, sarcoid, and other myocardial diseases evolve. The cost and the low sensitivity and specificity for endomyocardial biopsy suggest that noninvasive testing is essential for deciding which patients with unexplained cardiomyopathy should undergo this procedure. Novel biomarkers may be of use in identifying a cohort of such patients. Radionuclide techniques, particularly anti-
388
dec
myosin scintigraphy, also have a high negative predictive value and can be used in some settings to avoid biopsy. Contrast-enhanced MRI seems to be the most promising technique for diagnosing myocardial inflammation and myocyte injury based on small, nonprospective clinical studies. MRI may not only be useful in identifying those patients who should undergo biopsy but may allow a guided approach to the abnormal region of myocardium. It is hoped that this methodology improves the sensitivity of the technique for establishing a correct histologic diagnosis. Serial MRI studies have also shown promise for tracking the natural history of the disease and could, in the future, allow noninvasive reassessment of the myocardial response to therapy. As with any new diagnostic technology, preliminary observations generate challenges and opportunities. The early identification of patients with active myocarditis may more precisely reformulate the approach to the diagnosis and treatment of unexplained cardiomyopathy. Key unanswered questions include the following: 1. Can specific subgroups of patients with myocarditis, who may respond to immunosuppressive or immunomodulatory therapy, be identified by noninvasive imaging studies? 2. Can MRI or other methods differentiate between cardiomyopathies of different etiology (eg, myocarditis versus alcohol-related injury versus familial forms)? 3. Should right ventricular biopsy ultimately be replaced by MRI-guided left ventricular sampling? Research studies that address these vital questions, and prospective studies that compare the use of MRI techniques with other newer imaging techniques (eg, specifically annexin imaging for detecting myocardial apoptosis) in acute cardiomyopathy should pave the way for more effective treatment of viralmediated cardiomyopathies in the decade ahead.
References [1] Mason JW, Billingham JE, Ricci DR. Treatment of acute inflammatory myocarditis assisted by endomyocardial biopsy. Am J Cardiol 1980;45:1037 – 44. [2] Aretz HT, Billingham ME, Edwards WD, et al. Myocarditis: a histopathologic definition and classification. Am J Cardiovasc Pathol 1986;1:1 – 14. [3] Lieberman EB, Hutchins GM, Herskowitz A, et al. Clinicopathologic description of myocarditis. J Am Coll Cardiol 1991;18:1617 – 26. [4] Kuhl U, Seeberg B, Schultheiss HP, et al. Immuno-
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
histological characterization of infiltrating lymphocytes in biopsies of patients with clinically suspected dilated cardiomyopathy. Eur Heart J 1994;15(Suppl C): 62 – 7. Chow LH, Radio SJ, McManus BM. Insensitivity of right ventricular endomyocardial biopsy in the diagnosis of myocarditis. J Am Coll Cardiol 1898;14: 915 – 20. Hauk AJ, Kearney DL, Edwards WD. Evaluation of post-mortem endomyocardial biopsy specimens from 38 patients with lymphocytic myocarditis: implications for the role of sampling error. Mayo Clin Proc 1989; 64:1235 – 45. Dec GW, Fallon JT, Southern JF, et al. Borderline myocarditis: an indication for repeat biopsy. J Am Coll Cardiol 1990;15:283 – 9. Wu LA, Lapeyre AC, Cooper LT. Current role of endomyocardial biopsy in the management of dilated cardiomyopathy and myocarditis. Mayo Clin Proc 2001;76:1030 – 8. Pinamonti B, Alberti E, Cigalotto A, et al. Echocardiographic findings in myocarditis. Am J Cardiol 1988; 62:285 – 91. Pinamonti B. Contribution of echocardiography to the diagnosis of patients with chronic heart failure. Ital Heart J 2000;1:1311 – 6. Adsett M, West MJ, Galbraith A, et al. Eosinophilic heart: marked left ventricular wall thickening and myocardial dysfunction improving on corticosteroid therapy. Echocardiography 2003;20:369 – 74. Schinkel AFL, Bax JJ, van der Wall EE, et al. Echocardiographic follow-up of Chlamydia psittaci myocarditis. Chest 2000;117:1203 – 5. Carvalho JS, Silva CM, Shinebourne EA, et al. Prognostic value of posterior wall thickness in childhood dilated cardiomyopathy and myocarditis. Eur Heart J 1996;17(8):1233 – 8. Mendes LA, Picard MH, Dec GW, et al. Ventricular remodeling in active myocarditis. Am Heart J 1999; 138:303 – 8. Omi W, Nagai H, Takata S, et al. Ultrasonic tissue characterization in acute myocarditis. Circ J 2002;66: 416 – 8. Lieback E, Hardouin I, Meyer R, et al. Clinical value of echocardiographic tissue characterization in the diagnosis of myocarditis. Eur Heart J 1996;17: 135 – 42. Urhausen A, Kindermann M, Bohm M, et al. Diagnosis of myocarditis by cardiac velocity imaging in an Olympic athlete. Circulation 2003;108:e21 – 2. Palka P, Lange A, Clarke B, et al. Echocardiographic description of recurrent idiopathic giant-cell myocarditis in cardiac allograft. Circulation 2003;108:e1 – 3. O’Connell JB, Robinson JA, Kenkin RE, et al. Immunosuppressive therapy in patients with congestive cardiomyopathy and myocardial uptake on gallium-67. Circulation 1981;64:780 – 6. O’Connell JB, Henkin RE, Robinson JA, et al. Gallium-67 imaging in patients with dilated cardiomy-
noninvasive imaging techniques
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
opathy and biopsy proven-myocarditis. Circulation 1984;70:58 – 62. Khaw BA, Beller GA, Haber E, et al. Localization of cardiac myosin-specific antibody in myocardial infarction. J Clin Invest 1976;58:439 – 46. Yasuda T, Palacios IF, Dec GW, et al. Indium 111-monoclonal antimyosin antibody imaging in the diagnosis of acute myocarditis. Circulation 1987;75: 306 – 11. Haber E, Yasuda T, Palacios IF, et al. Antimyosin antibody imaging in the diagnosis of acute myocarditis. Eur Heart J 1987;8(Suppl J):119 – 23. Rezkalla S, Kloner RA, Khaw BA, et al. Detection of experimental myocarditis by monoclonal antimyosin antibody Fab fragments. Am Heart J 1989;117:391 – 5. Matsumori A, Ohkusa T, Mataba Y, et al. Myocardial uptake of antimyosin monoclonal antibody in a murine model of viral myocarditis. Circulation 1989;79:400 – 5. Kishimoto C, Hung GL, Ishibashi M, et al. Natural evolution of cardiac function, cardiac pathology, and antimyosin scans in a murine myocarditis model. J Am Coll Cardiol 1991;17:821 – 7. Dec GW, Palacios I, Yasuda H, et al. Antimyosin antibody cardiac imaging: its role in the diagnosis of myocarditis. J Am Coll Cardiol 1990;16:97 – 104. Margari ZJ, Anastasiou-Nana MI, Terrovitis J, et al. Indium-111 monoclonal antimyosin cardiac scintigraphy in suspected acute myocarditis: evolution and diagnostic impact. Int J Cardiol 2003;90:239 – 45. Dec GW, Narula J. Antimyosin imaging for noninvasive recognition of myocarditis. Quart J Nucl Med 1997;41(Suppl 1):128 – 39. Narula J, Southern JF, Abraham S, et al. Myocarditis simulating myocardial infarction. J Nucl Med 1991;32: 312 – 8. Sarda L, Colin P, Boccara F, et al. Myocarditis in patients with clinical presentation of myocardial infarction and normal coronary angiograms. J Am Coll Cardiol 2001;37:786 – 92.
389
[32] Koelling TM. The use of magnetic resonance imaging in cardiac transplant rejection. In: Dec GW, Narula J, Ballester M, et al, editors. Cardiac allograft rejection. Boston7 Kluwer Academic Publishers; 2001. p. 321 – 38. [33] Aherne T, Tscholakoff D, Finkbeiner W, et al. Magnetic resonance imaging of cardiac transplants: the evaluation of rejection of cardiac allografts with and without immunosuppression. Circulation 1986;74: 145 – 56. [34] Chandraratna PA, Bradley WG, Kortman KE, et al. Detection of acute myocarditis using nuclear magnetic resonance imaging. Am J Med 1987;83:1144 – 6. [35] Gagliardi MG, Bevilacqua M, Di Renzi P, et al. Usefulness of magnetic resonance imaging for diagnosis of acute myocarditis in infants and children, and comparison with endomyocardial biopsy. Am J Cardiol 1991;68:1089 – 91. [36] Gagliardi MG, Polletta B, Di Renzi P. MRI for the diagnosis and follow-up of myocarditis. Circulation 1998;98:458 – 9. [37] Friedrich MG, Strohm O, Schulz-Menger J, et al. Contrast media-enhanced magnetic resonance imaging visualizes myocardial changes in the course of viral myocarditis. Circulation 1998;97:1802 – 9. [38] Reiker O, Mohrs O, Oberholzer K, et al. Cardiac MRI in suspected myocarditis. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 2002;174:1530 – 6. [39] Roditi GH, Hartnell GG, Cohen MC. MRI changes in myocarditis-evaluation with spin-echo, cine MR angiography and contrast enhanced spin echo imaging. Clin Radiol 2000;55:752 – 8. [40] Schulz-Menger J, Dietz R, Friedrich MG. Longterm follow-up of patients with acute myocarditis by magnetic resonance imaging. MAGMA 2003;16: 17 – 20. [41] Mahrholdt H, Goedecke C, Wagner A, et al. Cardiovascular magnetic resonance assessment of human myocarditis: a comparison to histology and molecular pathology. Circulation 2004;109:1250 – 8.