History

The French pathologist Corvisart described “carditis” in 1806 as an important clinical syndrome, which most commonly was acute and fatal, but could develop into more “chronic organic disease.” The term carditis was later refined to myocarditis, first introduced by Sobernheim in 1837, to describe the myocardial inflammation presumed to be the cause of most nonvalvular cardiac dysfunction. By the latter part of the nineteenth century it was increasingly recognized that primary myocardial inflammation was responsible for only a small subset of nonvalvular cardiac dysfunction, because coronary disease and hypertensive heart disease were far more common causes. In contrast to the previous nondiscriminant use of the term, attempts at more accurate cardiac diagnosis of myocarditis in the twentieth century led to a marked diminishment in its recognition.

The emergence of endomyocardial biopsy (EMB) in the 1960s as a diagnostic tool allowed clinicians to delineate cellular inflammation of the myocardium, and fueled the hope that this heterogeneous group of disorders could now be classified into histopathologic subsets with distinct therapies and outcomes. The Dallas criteria were developed in 1986 by leading cardiac pathologists to standardize the histologic definition of lymphocytic myocarditis (LM), the most commonly observed form of cellular inflammation. They defined borderline myocarditis as mononuclear cell infiltrates without myocyte necrosis ( Fig. 28.1 ) and myocarditis as cellular infiltration with myocyte necrosis ( Fig. 28.2 ). Despite the wide acceptance of these histologic criteria for the pathologic assessment of myocarditis, significant variation in interpretation remained in the practical application. A decade later, the World Health Organization task force defined inflammatory cardiomyopathy as “myocarditis in association with myocardial dysfunction. Myocarditis is an inflammatory disease of the myocardium and is diagnosed by established histologic, immunologic, and immunohistochemical criteria.” This task force expanded the definition by adding both quantitative and immunohistochemical criteria for detection of cellular infiltrates and expression of human leukocyte antigen (HLA) class II molecules.

Histopathologic appearance of borderline myocarditis by Dallas criteria (lymphocytic infiltrates without myocyte necrosis) with routine staining with hematoxylin and eosin (H&E) under (A) low power (100×) and (B) high power (350×).

Histopathologic appearance of myocarditis by Dallas criteria (lymphocytic infiltrates with myocyte necrosis) with routine staining with hematoxylin and eosin under (A) low power (100×) and (B) high power (350×).

The observed histologic similarity of LM to cardiac allograft rejection led to the hypothesis that a therapeutic strategy of immunosuppression would improve clinical outcomes. When the Myocarditis Treatment Trial (MTT) demonstrated no benefit, the absence of a histologically guided therapy led to a diminished role for EMB, and native cardiac biopsy is not commonly performed in the United States. Efforts continue to improve the diagnostic utility of EMB with the addition of molecular diagnostics to detect viral nucleic acids and routine immunohistochemistry to improve the specificity, but this is only performed at selected centers. The histologic and clinical diversity of myocarditis remains a challenge in terms of the development of targeted therapeutics.

Viral Etiologies

The most common infectious agents initiating myocarditis in North America are viral pathogens ( Table 28.1 ), although the dominant viral species continues to evolve over time. Enteroviruses were first described in clinical and serologic studies decades ago and remain a major cause of myocarditis in infants and children. In adults, adenovirus, influenza A and B, and herpesviruses also have been implicated as important viral pathogens, whereas hepatitis C has been implicated in Asia, but less commonly reported in case series in the United States. Over the past decade, influenza and erythroviruses, such as parvovirus B19, emerged as two important pathogens associated with myocarditis. The changing viral milieu represents a considerable challenge for developing therapeutic efforts targeting specific viral pathogens. Effective viral therapies can diminish viral pathogenesis. For example, human immunodeficiency virus (HIV) infection has been associated with myocarditis and dilated cardiomyopathy (DCM) ; however, the incidence of HIV-associated cardiomyopathy has been diminished with more aggressive antiviral therapies against HIV.

In the developing world, bacterial pathogens remain important causes as myocarditis as a complication of rheumatic fever and diphtheroids is much more prevalent. In Central and South America, the most common infectious agent is the protozoa Trypanosoma cruzi, the causative agent for Chagas’ disease, which is endemic in certain areas and may affect 15 to 20 million people. This disorder is not seen in North America without a travel history to endemic areas. In immunocompromised hosts, other pathogens, such as toxoplasmosis and aspergillus, can cause myocardial inflammation but generally in the setting of a systemic infection and not as isolated myocarditis.

Autoimmune (Nonviral) Etiologies

There are several autoimmune forms of myocardial inflammation for which no infectious agent can be identified. For certain diagnoses, histopathology combined with the clinical setting can point toward targeted treatments. Multinucleated giant cells in the myocardium in the setting of fulminant myocarditis suggests the diagnosis of GCM ( Fig. 28.3 ), a progressive and destructive form of myocarditis with a high mortality rate on conventional therapy but responsive to immunosuppressive therapy. Finding similar multinucleated granulomas in the myocardium, but with few other inflammatory cells in a more compensated patient, may suggest systemic sarcoidosis, a disorder responsive to treatment with corticosteroids.

Histopathologic appearance of giant cell myocarditis (multinucleated giant cell) with routine staining with hematoxylin and eosin under (A) low power (100×) and (B) high power (350×).

Eosinophilic predominance of inflammatory cells in the myocardium with new myocardial dysfunction may suggest an allergic hypersensitivity myocarditis. Several pharmacologic agents, including tricyclic antidepressants, antipsychotics, and cephalosporin, have been implicated as triggering eosinophilic myocarditis. This disorder is generally self-limited, and the myocardium will recover with removal of the triggering agent. For persistent myocarditis in the setting of peripheral eosinophilia, a short course of corticosteroids can be considered. Chronic peripheral eosinophilia without a clear initiating agent should merit consideration of a systemic syndrome such as Churg–Strauss vasculitis or hypereosinophilic syndrome (HES), for which immunosuppressive therapy is generally indicated.

Myocarditis may also be a clinical feature of several systemic autoimmune disorders. Systemic lupus erythematosus (SLE), dermatopolymyositis, and rheumatoid arthritis also have been associated with inflammatory myocarditis. In general, immunosuppressive therapy is directed toward the systemic autoimmune disorder and will result in the recovery of myocardial function as well. EMB is rarely performed, except in cases where the systemic disorder is quiescent and the finding of myocardial inflammation may be the only indication for immunosuppressive therapy.

Pathogenesis in Murine Models

Much of what is known of the basic pathogenesis of viral myocarditis is derived from murine models of inoculation with the cardiotropic coxsackie group B virus or encephalomyocarditis virus into susceptible strains. The viral particles are taken up by myocytes via receptor-mediated endocytosis. Translation of viral proteins and replication of viral particles can result in cell lysis within 3 days ( Fig. 28.4 ) before the initiation of myocardial inflammation. Macrophage activation results in cytokine expression, including interferon-γ and the activation of natural killer (NK) cells, both of which limit viral replication. Infiltrates of antigen-specific T lymphocytes including T-helper (CD4+) cells and cytotoxic T lymphocyte (CD8+) cells are seen within 7 days of infection. Recognition of viral peptides by cytotoxic T cells results in cellular toxicity of virally infected cells. Up to 20% of infiltrating lymphocytes are B cells, and neutralizing antibodies are important in viral clearing although not in cytotoxicity. Viral particles are no longer evident within 15 days of the initial inoculation, although inflammatory infiltrates may persist for 90 days. Viral nucleic acid can be detected after 90 days in only a small percentage of virally infected mice; however, long-term ventricular dilation and remodeling may develop and are associated with myocardial fibrosis. Myocardial inflammation followed by ventricular dilation and fibrosis occurs in transgenic mice overexpressing TNF-α, and in murine models using antimyosin antibodies, demonstrating the role of chronic inflammation in the pathogenesis of nonischemic DCM.

Timeline of progression in murine models from initial viral inoculation to acute myocarditis to subacute myocarditis to chronic myocarditis.

From Feldman AM, McNamara D. Myocarditis. N Engl J Med . 2000;343[19]:1388–1398.

Clinical Presentation

Acute myocarditis will typically present with a variety of symptoms, including dyspnea, chest pain, palpitations, syncope, and near syncope, and when associated with left ventricular (LV) dysfunction, most commonly presents with signs of heart failure. A presentation with dyspnea or chest pain associated with electrocardiogram abnormalities or an elevation of cardiac enzymes may be the first sign of a possible myocarditis or myopericarditis. The cardiac examination may be unremarkable or may reveal a pericardial rub. Tachycardia, relative hypotension, jugular venous distention, and peripheral edema may suggest more significant hemodynamic compromise. Chest examination may reveal congestion, and decreased breath sounds are suggestive of pleural effusions.

Electrocardiographic findings are generally diffuse and nonspecific. PR depression and global ST elevation may be seen in cases with an associated pericarditis. Low voltage may be evident, particularly in subjects with an associated pericardial effusion. Heart block, ventricular tachycardia, or a new bundle branch block are all suggestive of a more fulminant disorder. Evaluation of biomarkers may reveal an elevation in cardiac troponin and brain natriuretic peptide. More fulminant myocarditis may be associated with elevated liver function tests and a compromise of renal function. Chest x-ray may be unremarkable or show evidence of heart failure and cardiomegaly.

Cardiac Imaging (See also Chapter 32 )

Transthoracic echocardiography remains a critically important tool in the evaluation of myocarditis and frequently provides the first evidence of ventricular impairment. Systolic dysfunction is generally global; however, it may be segmental and can mimic ischemic disease. Pericardial effusion may be present and provide supportive evidence of an inflammatory process. In severe myocarditis, the ventricular walls may be thickened due to edema; however, routine echocardiography is limited in terms of characterization of the myocardial tissue itself. The dimensions of the LV diameter generally demonstrate minimal LV enlargement and remodeling for an acute presentation, although remodeling may be more evident in more chronic insidious forms.

Cardiac magnetic resonance (CMR) imaging has had an increasing role in the diagnostic evaluation of myocarditis given its ability to characterize cardiac tissue and assess for inflammation ( Fig. 28.5 ). The use of gadolinium on T1-weighted images allows for assessment of hyperemia and capillary leak with early enhancement, and more significant tissue injury and necrosis with late gadolinium enhancement (LGE). The anatomic distribution of gadolinium enhancement also assists in diagnosis, because an epicardial distribution of gadolinium enhancement is suggestive of myocarditis, whereas an endocardial predominance is more consistent with ischemic injury. The addition of T2-weighted images provides an assessment of myocardial water, and increased content is likely a marker of ongoing edema and inflammation.

Cardiovascular magnetic resonance midventricular short-axis images from a 21-year-old man with acute myocarditis who presented with acute chest pain, peak troponin I of 15 ng/mL, and angiographically normal-appearing coronary arteries. (A) Myocardial edema (white arrows) in the epicardial portion of the inferolateral wall on T2-weighted images. (B) Necrosis on late gadolinium enhancement images in the same region (white arrowheads) with uptake of contrast due to loss of cell membrane integrity (20 minutes following 0.2 mmol/kg of gadoteridol).

Courtesy Timothy C. Wong, MD, MS, and Erik B. Schelbert, MD, MS.

The diagnostic criteria for myocarditis by CMR were established by a consensus panel, and the “Lake Louise” criteria were published in 2009. These criteria noted that combining data from all three tissue markers, early and LGE on T1-weighted images and myocardial edema by global and regional T2 relaxation times ( Table 28.2 ), increased the utility of CMR and that myocardial inflammation could be predicted with a diagnostic accuracy of 78%. When only gadolinium enhancement on T1 images was used, CMR assessment would still yield an accuracy of 68%. The time required for data acquisition and the need to transport the patient to the imaging facility limits the ability of CMR to assess the most critically ill subjects. In addition, contrast imaging with gadolinium essential to the evaluation of myocardial inflammation cannot be performed in subjects with renal dysfunction.

Endomyocardial Biopsy (See Also Chapter 34 )

Although the evaluation of myocardial histology by EMB remains the gold standard for the diagnosis of myocarditis, the diagnostic yield of biopsy remains low. In the large published biopsy series of over 2000 subjects screened for the MTT with EMB, the prevalence of LM was only 10%. Myocardial histology generally does not affect treatment strategies, and although there are notable exceptions, such as GCM, this is far less prevalent than LM and evident in no more than 2% of subjects who had biopsies in published series. The current American Heart Association (AHA) and the European Society of Cardiology guidelines for the indication for EMB are driven by the need to detect histologic diagnoses, such as GCM, which change therapeutic recommendations. These guidelines give the strongest recommendation for EMB in cases of acute fulminant myocarditis (acute myocarditis associated with hemodynamic compromise) and in acute myocarditis associated with ventricular tachycardia or heart block ( Table 28.3 ). Although the enhancement of EMB through molecular diagnostics remains an area of research, the role of therapies guided by molecular diagnostics remains uncertain.

EMB does have defined risks, most notably the risk of cardiac perforation and tamponade, which can occur in up to 1% of subjects. These risks are diminished when performed by an experienced operator. The risks of the procedure are increased by the hemodynamic instability of suspected fulminant myocarditis, and EMB in this scenario should be performed by an experienced operator at a tertiary center with the ability to provide mechanical support if required.

Myocarditis Mimicking Acute Coronary Syndrome

With the management of myocardial infarction and acute coronary syndromes emphasizing early angiography, the syndrome of myocardial infarction with nonobstructive coronary artery disease (MINOCA) is increasingly recognized, and represents 5% to 10% of all subjects presenting with myocardial infarction. While there are multiple potential etiologies for MINOCA including coronary spasm, resolving thrombus, and Takotsubo cardiomyopathy, acute myocarditis is the most common etiology documented, particularly in younger patients. In subjects with MINOCA who undergo CMR, more than one-third typically have imaging results consistent with acute myocarditis. Early myocardial recovery within the index hospitalization is common for cases of MINOCA found to be the result of myocarditis. EMB is used only selectively for more fulminant cases with hemodynamic compromise for which the possible diagnosis of GCM has both prognostic and therapeutic implications.

Medical Therapy

For myocarditis associated with LV dysfunction, therapy is supportive. Treatment with β-receptor antagonists and with either angiotensin-converting enzyme (ACE) inhibitors or angiotensin receptor antagonists remains the mainstay of therapy. Digoxin should be avoided, as it has been shown to worsen injury in animal models. Diuretics can be used for congestion and fluid overload. For subjects with severe LV dysfunction, anticoagulation should be considered and is definitively indicated for those subjects with evidence of LV thrombus or those presenting with a thromboembolic event.

For subjects with suspected myocarditis and LV dysfunction, the management of potential arrhythmias should be tempered by the possibility of recovery. Treatment with an implantable cardioverter defibrillator (ICD) should be deferred, given the potential for resolution. An ICD is indicated for subjects presenting with “aborted sudden death,” in which the placement of an ICD is required for secondary prevention. For subjects presenting with complex ventricular arrhythmias without sudden death, a temporary external defibrillator or life vest can be considered as an alternative to a more permanent device.

Systolic function recovers rapidly within weeks to months in many subjects with recent onset cardiomyopathy, and the role of long-term therapy with ACE inhibitors and β-blockers in subjects who have recovered remains controversial. In subjects who have complete normalization of systolic function after a documented transient episode of myocarditis, medical therapy with ACE inhibitors and β-receptor antagonists can be gradually weaned off and discontinued with careful monitoring for subsequent declines in left ventricular ejection fraction (LVEF). Subjects with persistent abnormalities of either systolic function or remodeling (increased LV diastolic diameter) should be treated long-term with heart failure therapy with either ACE inhibitors, β-receptor antagonists, or both. In subjects whose systolic function has completely recovered, abnormalities of diastolic function may persist for months and may result in persistent symptoms of exertional dyspnea and fluid retention.