Epidemiology

The diverse clinical manifestations have made the true incidence of myocarditis difficult to determine. The clinical course of acute myocarditis can be insidious, with limited inflammation and cardiac dysfunction, or it can be overwhelming, leading to severe cardiac injury and cardiac failure. As a clinical entity, myocarditis is an uncommon occurrence in children. At Texas Children’s Hospital, Houston, between 1954 and 1977, myocarditis represented 0.3% of the 14,322 patients seen by the Cardiology Service.

Because not all cases of myocarditis are recognized clinically, a much higher incidence is recorded in autopsy series. An autopsy incidence of 1.15% was found from 4343 studies performed between 1954 and 1977 at Texas Children’s Hospital. This rate is considerably lower than the incidence of 6.83% reported by Saphir and Simon in 1944 for 1420 autopsies performed on children. In Saphir’s series, 32 of 97 cases had or probably had rheumatic carditis, whereas only two cases occurred in the Texas Children’s Hospital series. The discrepancy is even more pronounced when these observations are compared with those of Burch and colleagues, who demonstrated evidence of interstitial myocarditis in the hearts of 29 of 50 infants and who showed evidence of interstitial myocarditis in the hearts of 29 of 50 infants and young children undergoing routine postmortem studies.

Recently Freedman et al. reported an estimated prevalence of myocarditis of 0.5 cases per 10,000 emergency center visits to a single pediatric Canadian center. Some of the discrepancies between the clinical and autopsy series may be explained by the fact that the manifestations of myocarditis are subclinical in many cases and may be recognized only by changes on electrocardiogram (ECG) or perhaps not at all. In many instances, myocarditis is only one component of a generalized illness, and the cardiac dysfunction, if mild, may be overlooked.

Etiologies

Myocarditis may occur with many common infectious illnesses that affect infants and children ( Box 28.1 ). In most cases of myocarditis, the etiologic agent is never identified, however. In the United States and Western Europe, viruses are the most common causes of acute myocarditis. Myocarditis generally is a sporadic disease, but epidemics have been reported. Most epidemics have been caused by coxsackievirus group B and have affected infants in the newborn period. Gear and Measroch were the first to identify coxsackievirus B in association with myocarditis after an epidemic occurred in a nursery in a maternity home in southern Rhodesia.

The association between virus infection and the development of myocardial disease also was made by Grist and Bell who presented comprehensive serologic data correlating enterovirus infection with acute viral myocarditis. In the World Health Organization report during the 10-year period from 1975–85, the coxsackieviruses B represented the most frequent inflammatory agents in cardiovascular disease (34.6/1000), followed by influenza B virus (17.4/1000), influenza A virus (11.7/1000), coxsackievirus A (9.1 per 1000), and cytomegalovirus (CMV) (8/1000). Karjalainen and colleagues prospectively examined 104 conscripts during the 1978 influenza A virus (H1N1) epidemic in Sweden. The incidence of myocarditis was 9% of the 67 verified cases of influenza virus infection. Randolph and colleagues reported that of 838 children with pandemic H1N1 admitted to a pediatric intensive care unit, 1.4% were diagnosed with myocarditis. Although H1N1-related acute myocarditis was uncommon, it was found to be an independent risk factor for death.

The development of molecular techniques such as polymerase chain reaction (PCR) has improved the testing of endomyocardial biopsy specimens for potential viral pathogens. A study using PCR identified viral genome in 38% of endomyocardial biopsy specimens from patients diagnosed with acute myocarditis. Of the positive PCR samples, 23% were positive for adenovirus, 14% for enterovirus, and 3% for CMV. Parvovirus B19, influenza A virus, Epstein-Barr virus, herpes simplex virus (HSV), and respiratory syncytial virus were detected in less than 1% of cases. In a recent retrospective study of pediatric myocarditis, viral studies (PCR of blood, myocardium, or serology) were performed in 30 of 58 patients. A viral cause was identified in 17 of 30 patients (56%) and included six parvovirus B19 with influenza coinfection, seven enterovirus, one EBV, and one CMV. In a recent study of children with the diagnosis of clinical myocarditis, Simpson et al. reported that blood PCR was positive at the time of presentation in 43% (9 of 21) for one of four known cardiotropic viruses (four enterovirus, two parvovirus B19, one adenovirus, and two HHV-6). The majority (89%) of the patients with clinical myocarditis and positive blood PCR were younger than 12 months old. In contrast, only 3.5% of healthy control children (four of 114) had a positive blood PCR for any of these four viruses. In one pediatric case series of pediatric cardiac transplant patients, parvovirus B19 genome was detected in 100 of 700 (82.6%) biopsies from 99 patients. The presence of the parvovirus B19 genome did not correlate with rejection score. However, transplant coronary artery disease occurred in 20 patients, with persistent detection (>6 months) of parvovirus B19. In two retrospective studies, parvovirus B19 has been identified as a common cause of viral myocarditis in healthy children. Patients with parvovirus B19 myocarditis often demonstrate persistent myocardial dysfunction requiring medical therapy and transplantation. Given the high prevalence of parvovirus B19 infection in the pediatric population, its pathogenic role in pediatric myocarditis and dilated cardiomyopathy (DCM) is still being investigated.

Investigators also have speculated for decades on the possibility that acute myocarditis is a common forerunner of idiopathic DCM. Evidence supporting this hypothesis was presented first by Orinius and Pernow who found cardiac disease in humans years after an apparent uncomplicated coxsackievirus infection. Subsequently, Bowles and colleagues using a slot-blot hybridization technique provided conclusive evidence for the presence of enterovirus in endomyocardial biopsy samples from patients with DCM. In another study, Bowles and associates detected viral genomes in 20% of 149 patients with the diagnosis of DCM. In these patients, adenovirus was identified in 12% and enterovirus in 8% of DCM cases. In all age groups, adenovirus and enterovirus were the viruses most commonly detected in acute myocarditis and DCM.

In December 2002, the U.S. Department of Defense began mandatory smallpox vaccination for select service members and employees without contraindications to vaccination, and in January 2003, the U.S. Department of Health and Human Services implemented a voluntary civilian smallpox vaccination program. As of June 15, 2003, the Department of Defense identified more than 50 cases of suspected, probable, or confirmed myopericarditis occurring within 30 days of vaccination in these individuals, based on clinical evaluation of symptoms, electrocardiography, cardiac enzyme assays, echocardiography, and the exclusion of ischemic coronary artery disease. Myocarditis occurred in 7.8 per 100,000 primary vaccinees in the U.S. Army, an incidence that was 3.6-fold more than that in unvaccinated individuals.

Pathology

Isolated or idiopathic myocarditis is a rare pathologic entity. The pathologic cardiac findings usually are nonspecific; similar gross and microscopic changes occur regardless of the causative agent. Grossly all four chambers of the heart are enlarged, and the cardiac weight is increased. The heart usually is flabby and pale. In some instances, especially with coxsackievirus B infections, petechial hemorrhages may be seen on the epicardial surfaces; pericardial fluid may be tinged with blood. On cut section, the ventricular muscle walls may be thinned. Occasionally the ventricles are hypertrophied or increased in thickness because of edema. The valves are spared. The endocardial surface usually is unaffected but occasionally may be thickened and appear glistening white. This important observation suggested to some investigators that endocardial fibroelastosis, which manifests as congestive cardiomyopathy, represented a progression from acute viral myocarditis. In a study of 64 hearts of children who had myocarditis or endocardial fibroelastosis, Hutchins and Vie found 18 with endocardial fibroelastosis only, five with myocarditis only, and 41 with features of both diseases. When time from onset of illness to death was 2 weeks or less, only myocarditis was evident. When the time interval was 2 weeks to 4 months, a combined picture was seen, whereas only endocardial fibroelastosis with trivial myocarditis was evident when the time from onset of disease to death was more than 4 months.

These findings were supported further by Hastreiter and Miller who found microscopic evidence of myocarditis after transthoracic needle biopsy of the myocardium in a child who had the classic clinical picture of endocardial fibroelastosis, including left ventricular hypertrophy on ECG. Fruhling and associates extended these observations by showing coxsackievirus B3 in the myocardium of 13 of 28 infants with endocardial fibroelastosis. Ni and associates analyzed 29 myocardial samples from patients with autopsy-proven endocardial fibroelastosis using specific PCR for enterovirus, adenovirus, mumps, CMV, parvovirus, influenza, and HSV. In 90% of samples, viral genome was amplified; more than 70% of the samples were positive for mumps viral RNA, whereas 28% were positive for amplified adenovirus. These data suggest that endocardial fibroelastosis also is a sequela of mumps virus infection.

The microscopic picture of acute myocarditis typically shows a focal or diffuse interstitial collection composed predominantly of mononuclear cells, lymphocytes, plasma cells, and eosinophils ( Figs. 28.1 and 28.2 ). Polymorphonuclear leukocytes rarely are seen unless the cause of the carditis is bacterial. Virus particles and inclusion bodies rarely are recognized. In severe infections caused by any agent, but especially coxsackieviruses and diphtheria, a loss of cross-striation in the muscle fibers, edema, and sometimes extensive necrosis of the myocardium occur.

Right ventricular biopsy specimen. The presumed viral myocarditis is characterized by focal mononuclear cell infiltrates (H&E staining, ×160).

(Courtesy Edith Hawkins, MD, Houston, TX.)

Picornavirus myocarditis characterized by interstitial edema, mononuclear cell infiltrates, and focal myofiber disruption (H&E staining, ×400).

(Courtesy Edith Hawkins, MD, Houston, TX.)

Giant cells with or without granulomata are markers for the diagnosis of giant-cell myocarditis. Granulomata have been observed in the myocardium of patients with tuberculosis, syphilis, rheumatoid arthritis, rheumatic heart disease, sarcoidosis, and certain fungal and parasitic infections. Occasionally giant cells have been seen in interstitial myocarditis (idiopathic or Fiedler). In many cases, giant-cell myocarditis occurs, but no cause is found.

Pathogenesis

The pathogenesis of myocarditis in humans was derived largely from experimental models of coxsackievirus infection. Liu and Mason have suggested that myocarditis should be viewed as a continuum that comprises three separate phases: acute viral infection (phase I), autoimmunity (phase II), and DCM (phase III).

Phase I of the disease is triggered by the entry and proliferation in the myocardium of the causative virus. Impairment of left ventricular function in mice with histopathologically graded moderate cellular infiltration after coxsackievirus B3 infection supports the importance of direct viral damage of the myocardium. Phase I concludes with activation of the cellular immune response, which attenuates viral proliferation but also may enhance cardiac injury. Ideally the immune response should downregulate to a resting state when viral proliferation is controlled. If immune activation continues unabated despite elimination of the virus, autoimmune disease may result, initiating phase II of the disease. The continuous activation of T cells long after viral clearance occurs is detrimental to the host because cytokine-mediated and direct T-cell–mediated myocyte injury leads to impairment of contractile function ( Fig. 28.3 ). Long-term remodeling and progression to DCM characterize phase III of the disease.

Schema for pathogenesis of myocarditis. Viral agents attach to cells by means of surface receptors. After a cell is infected, the cell cycle is changed. Direct virus-mediated cytolysis occurs. Cellular effectors of injury (i.e., macrophages, monocytes, and nonspecific cytotoxic T cells) are involved in the primary reaction. Myocytes that survive are altered in their structure. Cytotoxic T cells specifically targeted against the altered myocyte, natural killer (NK) cells, and complement-activated, antibody-mediated cardiocytolysis or antibody-dependent cellular cytotoxicity (ADCC) participate in the secondary reaction.

(From Maisch B, Trostel-Soeder R, Stechemesser E, et al. Diagnostic relevance of humoral and cell-mediated immune reactions in patients with acute viral myocarditis. Clin Exp Immunol . 48:533;1982.)

Numerous effector cells and molecules work in concert to restrict this initial spread of an infectious focus. The responding cells include natural killer cells, natural killer/T cells, and γδ T cells. Several lines of evidence suggest that mediators of the innate immune system, such as tumor necrosis factor (TNF) and nitric oxide, play an important role in the pathogenesis of viral myocarditis. Elevated levels of TNF have been reported in patients with viral myocarditis, and TNF mRNA and protein are consistently upregulated in the hearts of these patients. In mice, the exogenous administration of TNF aggravates myocarditis, and the neutralization of TNF by antibodies or soluble receptors attenuates the disease.

More recent studies also have shown that TNF and nitric oxide are beneficial to the host by virtue of their antiviral effects. Mice with defective TNF or nitric oxide expression have increased myocardial injury, a significant increase in viral titers in the heart, and significantly higher mortality rates after infection with encephalomyocarditis virus or coxsackievirus B3. Although the prevailing notion has been that production of cytokine in the heart during viral infection is detrimental, the host-pathogen relationship is changed fundamentally when the host is unable to produce molecules such as TNF or nitric oxide.

An important component of the innate immune system uses pattern recognition receptors, such as the Toll-like receptors (TLRs), to recognize pathogen-associated molecular patterns present in microbes. The role of TLRs in the pathogenesis of viral myocarditis is still evolving. However, recent studies suggest that cardiac inflammation during viral infection depends on TLRs. The viral genome replicates using the positive-strand RNA as its template, resulting in the formation of dsRNA intermediates. Accordingly both single-strand RNA and dsRNA are present in virally infected cells. TLR3 and TLR7/8 signaling are activated by double-stranded RNA (dsRNA) and single-stranded RNA, respectively. Thus viral infection can activate innate immune signaling in the heart through myeloid differentiation factor 88 (MyD88)-dependent (TLR 7/8) and MyD88-independent pathways (TLR3). Fairweather and colleagues reported that mice with defective TLR4 signaling had decreased coxsackievirus B3 replication and less severe myocarditis 12 days after infection compared with wild-type mice. The presence of TLR4 also was associated with increased production of interleukin-1β (IL-1β) and IL-18 and increased viral replication in the heart.

In a similar study, Fuse and associates reported that mice deficient in MyD88, an adapter protein involved in TLR signaling (except TLR3), also had less myocarditis and attenuated viral replication in the heart after infection with coxsackievirus B3. Coxsackievirus B3–infected, MyD88-deficient mice had significantly higher levels of interferon-β (IFN-β) but reduced expression of the coxsackievirus-adenovirus receptor in the heart. The enhanced IFN expression and lower expression of the coxsackie-adenovirus receptor in the heart could explain the attenuation of disease in the MyD88-deficient mice. Infection of TLR3-deficient mice with encephalomyocarditis virus (EMCV), a positive single-strand RNA virus, resulted in earlier mortality in TLR3-deficient mice that was associated with increased viral replication and myocardial injury when compared with wild-type mice. Similar observations have been reported for coxsackievirus group B serotype 3 in TLR3-deficient mice. Gorbea et al. screened TLR3 in patients diagnosed with enteroviral myocarditis or DCM and identified a rare variant in one patient as well as a significantly increased occurrence of a common polymorphism compared with controls. Expression of either variant resulted in significantly reduced TLR3-mediated signaling after stimulation with synthetic double-stranded RNA. Furthermore coxsackievirus B3 infection of cell lines expressing mutated TLR3 abrogated activation of the type I IFN pathway, leading to increased viral replication.

Woodruff and Woodruff, using a murine model, were the first to show a role for T lymphocytes in the pathogenesis of viral myocarditis. In this study, depletion of T lymphocytes using antithymocyte serum or thymectomy and irradiation led to a decrease in mortality rates and in the inflammatory infiltrate after coxsackievirus B3 infection. Huber and associates, using BALB/c mice infected with coxsackievirus B3, showed that cytolytic T cells were the agents responsible for the major part of myocardial cell injury. In addition, proinflammatory mediators, such as TNF, released by infiltration cells also adversely affect cardiac function.

Opavsky and colleagues defined the specific contributions of T-cell subsets (CD4 and CD8) and the T-cell receptor β chain to the pathogenesis of viral myocarditis. When CD4 ^−/− or CD8 ^−/− mice were exposed to CVB3, loss of CD8 ^+/+ immune cells did not affect survival significantly, but viral proliferation was attenuated. In contrast, CD4 ^−/− mice showed a trend toward an improvement in survival and a small but significant decrease in the inflammatory infiltrate at 14 days after infection. Mice deficient in CD4/CD8 immune cells and T-cell receptor β had the best outcome in terms of decreased mortality. A marked decrease in inflammatory infiltrate was noted in CD4/CD8 double-knockout mice. Although no significant change occurred in viral titers, a marked decrease in myocardial TNF mRNA 4 days after infection was seen in CD4/CD8 double-knockout mice. These same investigators have shown that the T-cell receptor–associated tyrosine kinase p56 ^lck is crucial for coxsack­ievirus B3 proliferation in the heart and activation of T cells to target the heart. Mice deficient in p56 ^lck were protected against the development of myocarditis, providing further support for the hypothesis that T-cell activation during viral myocarditis contributes to increased inflammation and myocyte destruction in the host.

T-regulatory (T-reg) cells also are important in modulating the inflammatory response and preventing the development of autoimmunity through the production cytokines like IL-10 and tumor growth factor-β (TGF-β). T-regs express CD4, but they also express the α-subunit of the IL-2 receptor, CD25. They also are high expressors of the transcription factor FoxP3. Li and associates have shown that the allograft of M2 (antiinflammatory) macrophages led to improvement of virus-induced myocarditis, which was associated with enhanced levels of T-regs. Similarly, Huber and colleagues have reported decreased viral titers and inflammation after adoptive transfer of a CD4 ⁺ CD25 ⁺ regulatory–like T-cell population into a mouse model of coxsackievirus B3 infection. Thus modulating the immune response may be critical for the prevention of chronic virus replication.

The ongoing injury that persists may be considered an autoimmune process. At least in murine models, it is clear that both cellular and humoral autoimmunity are involved in the progression to chronic heart disease. The recently identified T helper 17 (T _H 17) subset has been implicated in the onset of chronic myocarditis. These cells secrete high levels of IL-17 and have been implicated in the production of autoantibodies. In a mouse model of coxsackievirus B3 myocarditis these T _H 17 cells contribute to chronic myocarditis through persistent inflammatory signaling involving the secretion of IL-17. Consistent with this, there is significant T _H 17 expansion approximately 2 weeks after coxsackievirus B3 infection in mice. IL-17 and its various isotypes can induce expression of TNF and T _H 2 responses, which in turn leads to a prolonged inflammatory milieu that might foster the production of autoantibodies.

More recent studies also have shed light on the mechanisms by which coxsackievirus B3 may contribute directly to the development of myocarditis and DCM. Badorff and colleagues reported that the 2A protease encoded by coxsackievirus B3 cleaves dystrophin in cultured myocytes and in infected mouse hearts, leading to disruption of dystrophin and the dystrophin-associated glycoprotein α-sarcoglycan and β-sarcoglycan complex. Dystrophin provides a structural link between the muscle cytoskeleton and extracellular matrix to maintain muscle integrity. Xiong and associates compared the effects of coxsackievirus B3 infection in dystrophin-deficient (mdx) and wild-type mice. Coxsackievirus B3 infection significantly enhanced sarcolemmal disruption in the mdx mice compared with wild-type mice; the disruption was detectable 2 days postinfection and continued to increase after initial infection. Viral titers were higher in the hearts of mdx mice than in the hearts of wild-type mice, indicating greater viral replication in the absence of dystrophin. The observed differences seemed to be a result of more efficient release of coxsackievirus B3 from dystrophin-deficient myocytes. The expression of wild-type dystrophin in cultured cells decreased the cytopathic effect induced by coxsackievirus B3 and the release of virus from the cell. The expression of a cleavage-resistant mutant of the dystrophin protein inhibited coxsackievirus B3–mediated cytopathic effect and viral release further.

Pathophysiology

Given the extensive interstitial inflammation, muscle cell injury, or both, myocardial contractility is reduced. Consequently the heart enlarges and the end-diastolic volume of the ventricle increases. In the normal heart, an increase in filling volume leads, by the Starling mechanism, to an increased force of contraction, ejection fraction, and cardiac output. In patients with myocarditis, the myocardium is unable to respond in this manner, and cardiac output is reduced. Systemic blood flow may be maintained, however, by use of the cardiac reserve, mediated by the sympathetic nervous system and leading to vasoconstriction of the skin vessels and an increase in heart rate. With progressive disease, the heart may be unable to meet the oxygen demands of the tissues, and the clinical picture of congestive cardiac failure may become evident. In some infants and young children, the presentation is predominantly that of right-sided heart failure.

An appreciation of the disturbance of myocardial function may be gained from the angiographic frames shown in Fig. 28.4 . The left ventricle is dilated considerably, and the outline is irregular in diastole and systole. The ejection fraction is reduced significantly at 35% instead of the normal 60% to 75%.

The (A) end-diastolic and (B) end-systolic images from a left ventriculogram of a patient with idiopathic myocarditis show irregularity of the wall and poor contractility.

Another means of evaluating left ventricular function is the noninvasive technique of cardiac echocardiography. Gutgesell and colleagues established normal standards for children. An example is shown in Fig. 28.5A .

(A) Normal echocardiogram of a 4-year-old child. (B) Echocardiogram of a 4-year-old child with idiopathic myocarditis shows left ventricular dilation and severely reduced shortening fraction. EDD, End-diastolic dimension; ESD, end-systolic dimension; IVS, interventricular septum; LVPW , left ventricular posterior wall; MA, mitral apparatus; MV, mitral valve, %ΔLVD, percent change in left ventricular dimension (shortening fraction).

The normal shortening fraction (i.e., percentage change in ventricular dimensions between end-diastole and end-systole) is 35% ± 4%, regardless of age (range, 28–44%). Fig. 28.5B illustrates the case of a 4-year-old child with idiopathic myocarditis and shows ventricular dilation with markedly reduced motion of the left ventricular posterior wall and septum, leading to a shortening fraction of only 12%. Further assessment of ventricular function can be achieved by measuring systolic time intervals obtained from simultaneous recording of the ECG and the semilunar valve opening and closing points on the echocardiogram.

Clinical Manifestations

The clinical presentation of myocarditis varies with the age of the patient and the virulence of the organism. At one end of the spectrum is a fulminant, rapidly fatal illness and, at the other, no apparent clinical disturbance at all. A newborn especially is susceptible to the severe form of myocarditis usually caused by the coxsackieviruses B, but it also is recognized with rubella and HSV infections and with toxoplasmosis.

In many of these infections, myocarditis is only one component of a generalized illness, often with severe hepatitis and encephalitis. In some instances, however, infections with these organisms may produce only a mild clinical disturbance. In the report by Brightman and colleagues, a nursery epidemic of coxsackievirus B5 infection in preterm infants was recognized only by chance because a virologic survey was in progress at the time in the institution. No cases of myocarditis were documented, and all the infants recovered. Findings included lethargy, failure to gain weight, and, in some infants, evidence of aseptic meningitis.

As described in the review by Kibrick and Benirschke of 25 infants with coxsackievirus B myocarditis, vague symptoms such as lethargy and anorexia may herald the onset of the severe disease, emphasizing that close attention should be paid to all symptoms, especially in a newborn, no matter how nonspecific. Four infants had episodes of vomiting, and fever was documented for more than half of the cases; occasionally, the temperature was subnormal. Cyanosis, respiratory distress, and tachycardia, cardiomegaly, or ECG changes occurred in 19 of 23 infants. Tachypnea (respiratory rate >60/min in a newborn) is an early sign of heart failure in a young infant and should alert the clinician to this diagnosis.

In older infants and children, the manifestations of myocarditis generally are less fulminant than are the manifestations in newborns. An acute and fatal illness has been associated, however, with idiopathic myocarditis and the myocarditis associated with enteroviruses, adenoviruses, mumps, chickenpox, diphtheria, cytomegalovirus, and many of the other causative agents listed in Box 28.1 . Some older children have been reported with acute, substernal chest pain consistent with angina and have ECG changes of acute myocardial infarction. The usual clinical picture is that of an acute or a subacute illness, which often begins with a mild upper respiratory infection and a low-grade fever. Some infants have only vague, nonspecific suggestions of disease (e.g., irritability, periodic episodes of pallor) before the onset of cardiorespiratory symptoms, which begin a few days to 2 weeks after the onset of the initial symptoms. Abdominal pain may be a prominent complaint in some children.

On examination, these infants and children often are anxious and apprehensive, but some appear apathetic and listless. Pallor may be striking, and mild cyanosis may be present. Respirations are rapid and labored, and grunting may be prominent. The pulse is thready, and blood pressure usually is normal or slightly reduced, unless the infant is in profound shock. The precordium is quiet, without a prominent cardiac impulse. Resting tachycardia invariably is present in children who are critically ill with myocarditis. The heart sounds are muffled, and a prominent gallop rhythm usually is heard. Fine and colleagues found the most sensitive clinical sign of myocarditis to be a soft S ₁ at the apex. A prolonged PR interval, which may be a nonspecific finding in many febrile illnesses, also can cause a soft S ₁ , however, without any other evidence of myocarditis.

Almost uniformly, the liver is enlarged; edema is a rare finding. Some infants are less distressed and have signs of only mild congestive cardiac failure, without the signs of peripheral circulatory failure. Other infants have no signs of cardiac compromise, and myocarditis is recognized only as part of a generalized illness by a disturbance in the ECG pattern.

Diagnosis

Myocarditis often is difficult to diagnose, but it should be suspected in any infant or child who presents with congestive heart failure and who has or recently has had a febrile illness. The history should include information regarding travel, exposure to tuberculosis, recent drug ingestion, and illnesses in other family members or schoolmates. A quiet precordium in the presence of a gallop rhythm and decreased intensity or muffling of the heart sounds are findings that strongly suggest the diagnosis. A tachycardia out of proportion to the level of fever also should be viewed with suspicion. A physiologic S ₃ is a common finding in normal healthy children and in children with anemia and fever. An unusually prominent S ₃ suggests a disturbance of ventricular compliance without other evidence of compromised cardiac function and should be investigated further with an echocardiogram, a chest radiograph, and an ECG.

The occurrence of an arrhythmia, especially after a febrile illness, should alert the clinician to look for other signs of myocarditis. Lind and Hulquist detected significant dysrhythmias in five infants with isolated myocarditis. Four of the five infants died, and three of these infants had paroxysmal atrial tachycardia. Paroxysmal atrial tachycardia has been reported in patients with viral myocarditis and has been described in patients with diphtheritic myocarditis. Atrial ectopic tachycardia may mimic sinus tachycardia and, if not carefully evaluated, may be the primary cause for significant myocardial dysfunction. Complete heart block has been described in children in association with acute idiopathic myocarditis and with rubella, coxsackievirus, and respiratory syncytial virus infections. In some instances, complete heart block is permanent; in others, it is temporary.

Chest Radiography

Chest radiographs of infants and children who have signs of congestive cardiac failure invariably show cardiomegaly, usually of a severe degree ( Fig. 28.6 ). All four chambers may be enlarged, and evidence of pulmonary venous congestion often is found. Sometimes, especially in newborns, the first sign of illness is acute circulatory collapse, and, in this circumstance, the cardiac size may be normal. The same is true of children who have an arrhythmia rather than congestive heart failure. Other patients may present with Stokes-Adams attacks caused by complete heart block.

Marked cardiomegaly with a mild increase in the pulmonary venous pattern in the upper lobes.

Electrocardiogram

The ECG is an essential diagnostic tool for all patients with suspected myocarditis. The classic ECG pattern in myocarditis is one of diffuse low-voltage QRS complexes (<5 mm total amplitude) with low-amplitude or slightly inverted T waves and a small or absent Q wave in leads V ₅ and V ₆ ( Fig. 28.7 ). The low-voltage signal may be present in the standard leads and the precordial leads. Fig. 28.8 shows the ECG of an infant with acute myocarditis and shows a pattern of acute myocardial ischemia. Fig. 28.9A shows multifocal extrasystoles and severe intraventricular conduction delay in a patient with diphtheritic myocarditis; the ECG of this child returned to normal over the course of 3 months (see Fig. 28.9B ). The ECG from a 5-month-old infant who had mild fever, diarrhea, and vomiting for 3 to 4 days before admission is shown in Fig. 28.10 . The 2 : 1 atrioventricular block was associated with normal QRS complexes. This abnormality persisted in the absence of clinical symptoms for 1 year.

Diffuse low-voltage or QRS complexes with T-wave flattening and 1-mm Q waves in the lateral precordial leads represents the classic pattern in myocarditis.

In addition to low voltage, there is evidence of acute myocardial ischemia with 4- to 5-mm ST-segment elevation dominantly in the middle and lateral precordial leads.

(A) Multifocal premature beats are caused by atrioventricular dissociation and left bundle branch block resulting from diphtheritic myocarditis. (B) A normal electrocardiogram is shown for the same patient 3 months after an episode of myocarditis.

Second-degree atrioventricular block with an effective ventricular rate of 60 beats/min. The blocked P wave is placed on top of the T wave in each cycle.

Karjalainen and colleagues studied the ECGs of 87 conscripts 18 to 30 years old, 28 of whom had myocarditis. The most frequent findings were T-wave changes of reduced amplitude or inversion in the left chest leads. Sinus tachycardia followed by premature ventricular depolarizations was the most common dysrhythmia. Take and colleagues examined serial ECGs of 16 patients with confirmed viral myocarditis. They found four patterns: (1) complete normalization in the presence of severe myocardial damage in the acute stage; (2) “pseudoinfarction” patterns with Q waves and poor R-wave progression; (3) permanent conduction disturbances that might require pacemaker support; and (4) chronic dysrhythmias, predominantly ventricular tachycardia and supraventricular tachycardia.

Although T-wave and ST-segment changes are the most sensitive indices of myocardial ischemia, they also seem to be nonspecific. Prolongation of the PR interval is another nonspecific ECG finding frequently noted in patients with febrile illnesses. Scott and colleagues showed a 1.49% prevalence of these findings in a group of 737 infants and children with respiratory tract infections, but they also found a similar incidence among 108 control children without respiratory infection or other febrile illness. Abt and Vinnecour recorded PR prolongation and T-wave changes in infants and children who were suffering from pneumonia without other signs of myocarditis. The QT interval has been prolonged in cases of acute myocarditis, but it also appears to be a rather nonspecific finding associated with other infectious diseases. A diagnosis of myocarditis cannot be established with certainty on the basis of these nonspecific changes.

Molecular Diagnostic Studies