Part 3: Acquired Cardiovascular Disease

INTRODUCTION

Kawasaki disease, an acute febrile illness of childhood, was described in Japan in 1967 by Dr. Tomasaku Kawasaki as the mucocutaneous lymph node syndrome. A vasculitis of medium-sized vessels, Kawasaki disease has a predilection for the coronary arteries, and 20% to 25% of untreated children develop coronary artery aneurysms. Fortunately, a regimen of intravenous immunoglobulin and aspirin reduces the incidence of coronary artery abnormalities to less than 5%. Nonetheless, Kawasaki disease is the leading cause of acquired heart disease in children in North America and Japan. Furthermore, the etiology of Kawasaki disease remains elusive, impeding the development of targeted therapy and diagnostic testing.

EPIDEMIOLOGY AND SPECIFIC POPULATIONS AT RISK

The Japanese Ministry of Health collects epidemiologic data on Kawasaki disease every 2 years with nationwide surveys and has done so since 1970. In total, approximately 300,000 cases of Kawasaki disease were reported in Japan between 1970 and 2012. In addition to endemic disease, 3 epidemics of Kawasaki disease occurred in Japan in 1979, 1982, and 1986, with an incidence of 196.1 cases per 100,000 children less than 5 years of age in 1982, the most severe of the epidemics. Interestingly, there have been no recent epidemics, but surveys reveal a steady increase in the incidence of Kawasaki disease. The incidence of Kawasaki disease in Japan in 2012 was 264.8 per 100,000 in children younger than 5 years of age, the highest reported incidence to date.

In contrast to Japan, rates of Kawasaki disease in the United States have been stable for approximately the past 2 decades. The hospitalization rates for Kawasaki disease in the United States were 19 per 100,000 children less than 5 years of age in 2009, 19.6 per 100,000 children in 2003, and 17.5 per 100,000 children in 1997. Children of Asian and Pacific Islander heritage have consistently higher hospitalization rates in the United States, a finding consistent with the hypothesis that Kawasaki disease results from unidentified environmental triggers in genetically susceptible hosts. Furthermore, siblings of patients with Kawasaki disease may be at higher risk for developing the disease, and parents of affected children in Japan are twice as likely to have a history of Kawasaki disease themselves.

RISK FACTORS FOR POOR OUTCOMES

Risk factors for poor coronary artery outcomes have been identified in a number of studies. A study using the Centers for Disease Control and Prevention (CDC) passive surveillance system for Kawasaki disease found that coronary artery aneurysms were significantly associated with age less than 1 year or greater than 9 years, male sex, Asian and Pacific Islander race, and Hispanic ethnicity. A number of investigators have created predictive models to identify risk factors for poor clinical outcome. Young age, male sex, and a number of laboratory parameters, including neutrophilia, thrombocytopenia, hyponatremia, elevated C-reactive protein, and transaminitis, have been associated with poor response to intravenous immunoglobulin or the development of coronary artery aneurysms. Increased diameter of the coronary arteries on baseline echocardiogram may also predict the subsequent development of coronary artery aneurysms. In short, younger, sicker children with worse systemic inflammation are at higher risk.

ETIOLOGY

Despite years of extensive research, the etiology of Kawasaki disease remains unknown. Many aspects of the disease implicate an infectious trigger in the pathogenesis. Kawasaki disease is a disease of childhood, rarely seen in adults or infants less than 2 months of age. Discrete seasonal peaks with increased incidence in selected geographic areas suggest vector borne transmission. Many of the clinical features of Kawasaki disease resemble other childhood infections with exanthems (see below). These observations led to the investigation of many infectious agents as the causative factor, including the Epstein-Barr virus, adenovirus, human coronavirus NL63, human bocavirus, Yersinia pseudotuberculosis, herpesviruses, and others. However, no single etiologic agent has emerged as the clear culprit.

Similarities between Kawasaki disease and toxin-mediated diseases, such as toxic shock syndrome and streptococcal toxic shock syndrome, suggest that the features of Kawasaki disease are mediated by superantigens. Superantigens are proteins that can bind to the major histocompatibility complex on antigen-presenting cells and to T-cell receptors at sites other than conventional peptide-binding clefts, enabling simultaneous activation of many T cells. Binding of the T cell occurs at the variable region of the β chain (V β region). The resultant massive release of cytokines likely produces many of the features of toxic shock syndrome, streptococcal toxic shock syndrome, and other toxin-mediated syndromes.

Detection of superantigen activity is best determined by V β skewing, which is the finding of a disproportionate number of T cells that express a particular family of V β T-cell receptor genes. However, studies of T cells and attempts at isolating superantigen-producing bacteria in Kawasaki disease patients have yielded conflicting results, as some studies have found evidence of V β skewing in the peripheral blood, the intestinal wall mucosa, and coronary arteries, whereas other studies have shown an absence of V β skewing. A study of serologic responses to superantigens revealed an increase in immunoglobulin (Ig) M antibodies to multiple toxins, indicating that more than 1 superantigen could be involved in the pathogenesis of Kawasaki disease. Despite the wealth of data regarding superantigens in Kawasaki disease, their role remains controversial.

Studies of other immunologic mechanisms at work in Kawasaki disease, including oligoclonal IgA plasma cells, toll-like receptors and adaptor proteins, CD25+/CD4+ regulatory T cells, and genetic variations in interleukin-4 and chemokine (C-C motif) receptor 5, have yielded intriguing but inconclusive results.

GENETICS

Genome-wide association studies have identified susceptibility genes to Kawasaki disease. For example, polymorphisms in the inositol 1,4,5-triphosphate 3-kinase (ITPKC) gene have been associated with increased risk for Kawasaki disease as well as for development of coronary artery aneurysms, and have been described in Japanese and US populations. ITPKC is a negative regulator of the Ca⁺⁺/NFAT signal pathway, and polymorphisms lead to sustained activation of T cells and resultant cytokine expression. This mechanism lends validity to the use of cyclosporine (a downregulator of NFAT) as a possible treatment for Kawasaki disease (see Treatment section). Other candidate genes for Kawasaki disease as identified by genome-wide association studies include CASP3, BLK, and FCGR2A. Lastly, associations of single nucleotide polymorphisms in the human leukocyte antigen (HLA) class II region (HLA-DQB2 and HLA-DOB) with Kawasaki disease have been reported.

CLINICAL MANIFESTATIONS

Classic clinical criteria, as well as supportive clinical and laboratory findings, are summarized in Table 482-1. In brief, the epidemiologic case definition of Kawasaki disease includes at least 4 days of fever (3 days in expert hands), together with 4 or 5 principal clinical criteria. Those with fewer than 4 principal clinical criteria may meet the case definition if they have coronary artery disease. Incomplete Kawasaki disease, in which fewer than 4 principal clinical criteria are evident, is associated with a rate of coronary aneurysms that is similar to that of complete disease. Because the diagnosis of incomplete Kawasaki disease is particularly challenging, the American Heart Association recommendations, endorsed by the American Academy of Pediatrics, provide an algorithm for evaluation and treatment of suspected incomplete Kawasaki disease (Fig. 482-1).

Kawasaki disease is a vasculitis, and its features indicate systemic inflammation, likely incited in part by tumor necrosis factor and interleukin-1. Fever is the hallmark of Kawasaki disease. Children with Kawasaki disease tend to have daily, high fevers (> 39°C) that are of abrupt onset. Without intravenous immunoglobulin treatment, the average duration of fever is 11 to 12 days, with rare patients remaining febrile for as long as 3 to 4 weeks. Patients who receive appropriate therapy tend to defervesce within 2 to 3 days of intravenous immunoglobulin administration. The clinical features that accompany fever in Kawasaki disease patients may appear and subside at differing times over the course of illness, highlighting the importance of a complete medical history.

More than 90% of children with Kawasaki disease develop bilateral, marked conjunctival injection in a limbal sparing pattern. Anterior uveitis, detected by a slit lamp examination, also occurs in the acute phase of the disease. Patients with Kawasaki disease develop mucosal changes in their oropharynx; diffuse erythema of the oropharynx, a strawberry tongue, and cracked, reddened lips are characteristic of Kawasaki disease, whereas exudative tonsillitis or discrete oral lesions suggest other etiologies. Typically, the rash is erythematous and begins in the perineal area with some desquamation; morbilliform and erythema multiforme lesions can also occur. Vesicular or bullous lesions suggest other diagnoses. Extremity changes in the acute phase of Kawasaki disease include erythema of the palms and soles, as well as edema of the hands and feet. In the convalescent phase of Kawasaki disease, usually 2 weeks after fever onset, periungual peeling occurs on the fingers, followed by the toes. The least common manifestation of Kawasaki disease is cervical lymphadenopathy, specifically a single large (> 1.5 cm) unilateral node or, by computed tomography or ultrasound, a cluster of nodes, that is usually nontender, without erythema or warmth. Diffuse lymphadenopathy is more likely explained by another diagnosis.

A variety of signs and symptoms may support the diagnosis of Kawasaki disease. Patients with Kawasaki disease are usually irritable, perhaps reflecting central nervous system inflammation. Children with Kawasaki disease may complain of arthralgias; reported rates of frank arthritis range from 7.5% to 31% in Kawasaki disease patients.Although large joints of the lower extremities are most likely to be involved, small-joint arthritis has been reported as well. Gastrointestinal complaints are also common and include abdominal pain, nausea, and vomiting. Hydrops of the gallbladder can present with abdominal pain or jaundice or can be clinically silent. Possible etiologies for acalculous distension of the gallbladder include vasculitis, serositis, or compression of the cystic duct by inflamed mesenteric lymph nodes.

DIAGNOSIS

As fever and rash are the most common features of Kawasaki disease, the differential diagnosis is dominated by infectious diseases of childhood (Table 482-2). Common viral infections such as adenovirus, enterovirus, and Epstein-Barr virus can mimic Kawasaki disease. As the rash of Kawasaki disease is frequently morbilliform, measles is a differential diagnosis but is less commonly encountered in developed countries due to widespread vaccination. Diseases with exanthems such as scarlet fever, Rocky Mountain spotted fever, leptospirosis, staphylococcal scalded skin syndrome, and toxic shock syndrome should be considered when evaluating a child for Kawasaki disease. The toxin-mediated shock syndromes are characterized by hemodynamic instability that is also seen in Kawasaki disease shock syndrome, a recently described entity whereby children present with signs of Kawasaki disease as well as prominent hypotension, requiring vasoactive medications and/or transfer to a higher level of care. Drug hypersensitivity syndromes, including Stevens-Johnson syndrome, are also possible. In some children who appear at first to have Kawasaki disease with fever, rash, joint pain, and coronary artery changes may ultimately evolve into systemic-onset juvenile idiopathic arthritis. Rarely, mercury poisoning or acrodynia, with its swollen, erythematous hands and feet, may mimic Kawasaki disease. Although cervical chain lymphadenopathy is the least common manifestation of Kawasaki disease, it can be the presenting symptom in lymph node–first Kawasaki disease. In this scenario, children are treated for presumed bacterial lymphadenitis but do not defervesce with appropriate antimicrobials and typically develop other signs of Kawasaki disease during their clinical course.

Complicating the diagnosis of Kawasaki disease is the finding that up to one-third of patients have confirmed intercurrent illnesses. Additionally, patients with Kawasaki disease frequently present with nonspecific symptoms. Thus, the examining physician should not discount the possibility of Kawasaki disease in a patient with nonspecific complaints if stigmata of this disease are present.

Importantly, any child < 6 months of age with unexplained fevers for 7 or more days and signs of systemic inflammation on laboratory studies should be evaluated with an echocardiogram, as incomplete Kawasaki disease is more likely in infants who are also at higher risk for poor coronary artery outcomes. Early identification and treatment of these infants are critical to decrease long-term morbidity from coronary artery aneurysms.

LABORATORY STUDIES

Laboratory studies in patients with Kawasaki disease may demonstrate leukocytosis, anemia, and, in the later stages of the acute phase, a pronounced thrombocytosis with platelet counts from 500,000 to 1,000,000/mm. Thrombocytopenia in the early phase of Kawasaki disease may be a sign of consumption or evidence of an activated endothelium with platelet adherence, and has been identified as a poor prognostic factor. Inflammatory markers are nearly always elevated in Kawasaki disease, with the erythrocyte sedimentation rate usually greater than or equal to 40 mm/h and the C-reactive protein usually greater than or equal to 3.0 mg/dL. Elevation of the erythrocyte sedimentation rate may lag behind that of C-reactive protein, so both should be measured at diagnosis. Assessment of systemic inflammation following treatment with intravenous immunoglobulin should include C-reactive protein only, as the erythrocyte sedimentation rate will be elevated due to the intravenous immunoglobulin and should not be interpreted as a sign of worsening inflammation. Further supportive laboratory findings include transaminitis, hyperbilirubinemia, elevated plasma gamma-glutamyl transpeptidase, hypoalbuminemia, and sterile pyuria (see Table 482-1).

ELECTROCARDIOGRAPHY AND PLAIN RADIOGRAPHS OF THE CHEST

Electrocardiography in patients with Kawasaki disease may reveal sinus tachycardia, prolongation of the PR and QT intervals, or nonspecific ST- and T-wave changes. Plain radiographs of the chest tend to be unrevealing, although rare children have pulmonary infiltrates or effusions. Cardiomegaly is rarely present in the acute phase.

ECHOCARDIOGRAPHY

Echocardiography is the cornerstone of diagnosis of coronary artery abnormalities in children with Kawasaki disease. Taken together with Doppler imaging, echocardiographic imaging is also used to assess mitral and, more rarely, aortic regurgitation, left ventricular function, and pericardial effusion.

Coronary artery dimensions are generally described in terms of their absolute diameter, as well as in standard deviation units (z-scores) adjusted for body surface area. The widely used 1984 Japanese Ministry of Health criteria for coronary artery aneurysms include any of the following: an internal lumen diameter of greater than 3 mm in children less than 5 years of age or greater than 4 mm in children 5 years of age or older, an internal lumen diameter that is greater than or equal to 1.5 times that of an adjacent segment, or a coronary lumen that is clearly irregular. Many natural history studies of coronary aneurysms in the Japanese literature are based on these definitions. In a more recent revision of the Japanese Ministry of Health criteria, the size of an aneurysm is classified according to internal lumen dimensions: small, ≤ 4 mm; medium, > 4 to ≤ 8 mm; and giant, > 8 mm. For children age ≥ 5 years, aneurysm size can also be determined by the ratio of the internal diameter compared to an adjacent segment: small, 1.5×; medium, 1.5× to 4×; and giant, > 4×.

Coronary artery dimensions are related to body size, and use of z-scores has demonstrated that the Japanese Ministry of Health criteria underestimate the prevalence of coronary artery dilation in Kawasaki disease patients. Aneurysms are considered small when z-scores are ≥ 2.5 to < 5, medium if z-scores are ≥ 5 to < 10, and large if z-scores are either ≥ 10. Coronary artery z-scores are considered in the American Heart Association criteria for treatment of suspected incomplete Kawasaki disease (see below), as well as for choice of antithrombotic agents. There are 3 coronary artery z-score calculators in common use in North America. At larger coronary dimensions, these calculators can vary considerably in their calculated z-scores, potentially affecting therapeutic decisions about the need for anticoagulation. Moreover, the z-score calculators do not apply to all arterial segments (eg, the distal right coronary, distal left anterior descending, and the posterior descending arteries). For this reason, classification of coronary aneurysms continues to rely on both z-scores and Japanese Ministry of Health criteria.

At the time of presentation, within the first 10 days of illness, children with Kawasaki disease have enlarged coronary artery dimensions, adjusted for body surface area, compared to afebrile normal children. In addition to coronary artery enlargement, arterial walls may appear bright, and the arteries may not taper normally. Abnormalities on echocardiography form an important criterion for intravenous immunoglobulin treatment of children with suspected Kawasaki disease with incomplete criteria (see Fig. 482-1). Echocardiography should be performed at the time of diagnosis, at 2 weeks, and then approximately 5 to 6 weeks after illness onset. Children with documented coronary artery abnormalities or persistent fever require more frequent imaging surveillance; more aggressive anti-inflammatory therapy is initiated in those with expanding aneurysms. In addition, echocardiographic surveillance as often as twice weekly may be prudent for detection of coronary thrombosis in children with giant aneurysms who are early in the disease.

TREATMENT

Standard Therapy: Intravenous Immunoglobulin and Aspirin

Once the diagnosis of Kawasaki disease has been established, intravenous immunoglobulin (IVIG) should be administered at a dose of 2 g/kg over 8 to 12 hours. Treatment with IVIG should be instituted no later than day 10, and ideally by day 7, of illness; the first day of fever in Kawasaki disease is considered day 1. Administration of IVIG beyond day 10 of illness should be reserved for children with persistent fever or those with coronary artery lesions and evidence of ongoing systemic inflammation on laboratory testing. In general, IVIG is well tolerated. Use of premedications (diphenhydramine and acetaminophen) and using a slow rate of administration decrease infusion reactions. IVIG-related hemolysis is dependent on dose and recipient blood type (non-O is protected).

Aspirin was the first agent to be used in the treatment of Kawasaki disease, and subsequent studies evaluating the efficacy of IVIG and other treatments have included aspirin in the treatment regimen. However, the use of aspirin alone does not decrease the frequency of coronary artery lesions. Aspirin is recommended in medium (30–50 mg/kg/d divided into 4 doses) or high doses (80–100 mg/kg/d divided into 4 doses) for anti-inflammatory effects followed by low-dose aspirin (3–5 mg/kg/d) for antiplatelet effects. There is practice variation among experts in the field with respect to the duration of medium- to high-dose aspirin use. Most commonly, low-dose aspirin is instituted after the patient has been afebrile for 48 hours, but some centers continue higher dose aspirin until the child is both afebrile and beyond day 14 of illness, whichever is longer. Most children with Kawasaki disease are maintained on low-dose aspirin therapy until 6 weeks after illness onset, at which point aspirin therapy may be discontinued if the child’s echocardiogram is normal. Patients with coronary artery abnormalities on follow-up echocardiogram may be maintained on low-dose aspirin indefinitely, and annual influenza vaccinations are recommended in such patients. Reye syndrome has been reported in patients who received high-dose aspirin therapy for Kawasaki disease; the risks conferred by daily low-dose aspirin as chronic therapy are uncertain, although likely very low. Aspirin should be temporarily discontinued in children who have been exposed to varicella or influenza; other antiplatelet agents, such as clopidogrel, can be used in children with coronary lesions. Lastly, administration of live vaccines, including the measles and varicella immunizations, should be delayed for 11 months after IVIG treatment, as it may render the vaccines less immunogenic.

Rescue Therapy and Primary Intensification

Rescue Therapy

Two strategies have been implemented to decrease the incidence of coronary artery aneurysms in children with Kawasaki disease. One strategy is to administer “rescue therapy” to the approximately 15% to 20% of patients who have persistent or recrudescent fever after a single dose of IVIG (so-called IVIG resistance). Importantly, the subset of patients with IVIG resistance is at increased risk for developing coronary artery abnormalities. Although specific recommendations for patients with persistent or recrudescent fever after 1 dose of IVIG are hampered by lack of prospective trials, additional IVIG, at a dose of 2 g/kg, is generally administered. Infliximab, a chimeric monoclonal antibody to tumor necrosis factor-α, has been used in the IVIG-resistant population and is reported to shorten fever duration and hospital days. However, there is no evidence to demonstrate that use of infliximab in IVIG resistance reduces the prevalence of coronary aneurysms. Other treatments include intravenous methylprednisolone, oral corticosteroids, and cyclosporine. In severe cases, cyclophosphamide and plasma exchange can be considered.

Primary Intensification

As coronary artery aneurysms develop dynamically with peak coronary artery dimensions typically in the first 6 weeks of illness, the need to aggressively treat coronary artery inflammation early in the course of disease seems paramount. Accordingly, intensification of primary treatment (ie, treatment administered with the first dose of IVIG) has also been implemented as a strategy to improve coronary artery outcomes. Unfortunately, North American trials of primary intensification have not yielded positive results. Adding a single dose of methylprednisolone (30 mg/kg) to initial IVIG treatment did not lead to improvement of coronary artery aneurysms in a multicenter trial. Similarly, a trial of infliximab administered with the initial IVIG dose did not lead to decreased IVIG resistance, the primary end point. In contrast, the RAISE trial (Randomized Controlled Trial to Assess Immunoglobulin plus Steroid Efficacy for Kawasaki Disease) in Japan used corticosteroids (2 mg/kg/d) in a high-risk cohort and demonstrated improved coronary artery outcomes as well as decreased IVIG resistance. It seems likely that North American trials of more intensive primary therapy were negative, in part, because they included unselected Kawasaki disease patients, whereas the Japanese RAISE trial enrolled only children at high risk for aneurysms. Currently in North America, both rescue therapies and approaches to primary intensification remain center dependent and subject to significant variation.

COMPLICATIONS

Virtually all serious morbidity and mortality derive from the cardiac effects of Kawasaki disease. The most important complication of Kawasaki disease is coronary artery aneurysms, which may lead to myocardial ischemia, infarction, and sudden death. Prior to the development of an effective treatment for Kawasaki disease, 20% to 25% of affected children developed coronary artery aneurysms, with a 2% mortality rate. Fortunately, a regimen of IVIG and aspirin reduces the rate of coronary lesions to less than 5%, with only 1% of children developing giant aneurysms. The mortality rate in recent studies is less than 0.5%.

Stenosis

Coronary aneurysms created by the intense inflammatory process in the acute phase of Kawasaki disease evolve in their shape and dimension over years after illness onset. Regression to normal lumen diameter occurs via myofibroblastic proliferation in approximately half of coronary artery aneurysms; regression is most likely in patients with aneurysms less than 8 mm in internal diameter and with a fusiform morphology, in distal vessels, and in young children. Stenotic lesions, usually occurring at the proximal and distal ends of aneurysms, form in the chronic phase and are associated with myocardial ischemia. Rarely, stenotic lesions can give rise to new aneurysm formation years after disease resolution secondary to hemodynamic factors such as sheer stress. Stenotic lesions are most likely to occur at the points of entrance to or exit from giant aneurysms; patients with small aneurysms are at much lower risk to develop stenosis.

Myocardial Infarction

Myocardial infarction occurs most frequently in children with giant aneurysms. The combination of sluggish blood flow through an enlarged lumen, enhanced platelet activation due to shear stress at the proximal and distal sites of the aneurysm, and marked thrombocytosis significantly increases the risk for thrombus formation in the subacute phase. Although the relative risk of myocardial infarctions is highest in the first year, deaths from myocardial infarction continue to occur late after the onset of illness. Myocardial infarctions may be accompanied by symptoms of pallor, crying, vomiting, and abdominal pain. Children older than 4 years may complain of chest pain. Of note, myocardial infarctions are silent in more than one-third of patients with Kawasaki disease. Rupture of coronary aneurysms is a very rare complication occurring in rapidly expanding, inflamed coronary arteries, usually within the first 6 weeks of illness.

Other Cardiac Involvement

Kawasaki disease affects not only the coronary arteries, but also cardiac muscle and heart valves. Children with Kawasaki disease may have depressed left ventricular function by echocardiography in the acute phase of the disease, and cardiac biopsies as well as nuclear imaging studies have shown that myocarditis is extremely common. Indeed, myocarditis and arrhythmias are the leading causes of death in the first week of Kawasaki disease. Mitral regurgitation secondary to valvulitis occurs in approximately 1 in 4 children during the acute phase; late mitral regurgitation results from papillary muscle dysfunction in children with ischemic heart disease. Aortic root dilation and mild aortic regurgitation occur in Kawasaki disease but are usually not progressive.

Kawasaki disease may produce long-term systemic abnormalities, including diminished reactivity and greater stiffness of the peripheral arteries; increased intimal medial thickness of the carotid arteries; and dyslipidemia. These findings are most pronounced in patients with coronary aneurysms. Data are conflicting on the long-term effects of acute Kawasaki disease on arterial health among patients who never had coronary involvement. Some studies report endothelial dysfunction in these patients, whereas others found normal arterial reactivity and no evidence of arterial stiffness. Two studies have shown that myocardial flow reserve is altered in patients with Kawasaki disease regardless of coronary artery status. Lastly, an unfavorable lipid profile can be found, even among patients in whom coronary involvement has not been detected in any stage of illness.

LONG-TERM MANAGEMENT

The prognosis for children with Kawasaki disease is determined by the extent of coronary artery involvement and consequent risk of myocardial ischemia. For this reason, recommendations for frequency of follow-up and types of testing, exercise, and medications for children after Kawasaki disease are stratified according to their coronary artery status (Table 482-3). Because even patients who never had coronary abnormalities may be at higher risk for later atherosclerotic coronary artery disease, preventive cardiology assessment and counseling are of paramount importance. All children should have a lipid profile 1 year after Kawasaki disease onset. In addition, children and their families should be encouraged to follow a heart-healthy diet, exercise regularly, and avoid smoking.

SUGGESTED READINGS

INTRODUCTION

The myocardium can be affected by genetic, inflammatory, metabolic, infiltrative, ischemic, or primary disease processes, with significant overlap for many of them (Table 483-1). Myocardial diseases may present acutely, with signs of decompensated heart failure: tachypnea, tachycardia, an enlarged heart, soft heart sounds that may have a tictoc rhythm, cardiac gallop cadence, and either no murmurs or systolic murmurs of mitral or tricuspid regurgitation or aortic or pulmonary stenosis/outflow tract obstruction. Alternatively, they may present chronically with failure to thrive, decreased exercise tolerance, and more subtle findings of heart failure. Occasionally, the predominant process may be an arrhythmia, either atrial or ventricular tachycardia/fibrillation, or bradycardia/heart block. These present clinically with palpitations, syncope, or sudden death.

The electrocardiogram and chest x-ray are useful in suggesting heart disease but are rarely diagnostic. In acute presentations, the chest x-ray usually shows diffuse cardiac enlargement, passive pulmonary congestion, and interstitial edema; the electrocardiogram shows normal or reduced QRS voltages (because hypertrophy has not yet had time to develop) and STsegment depressions with Twave inversion in the anterolateral precordial leads. In one form of myocardial disease, left ventricular noncompaction, giant QRS complexes may be seen. In hypertrophic or restrictive cardiomyopathy, ST-segment abnormalities consistent with ischemia or injury may be seen. Atrial enlargement can be seen in cases of restrictive physiology with dilated atria or in dilated cardiomyopathy associated with mitral valve regurgitation. The echocardiogram often shows diffuse cardiac dilatation, decreased contractions, and atrioventricular valve regurgitation in those with dilated cardiomyopathy, cardiac hypertrophy (typically with asymmetric septal hypertrophy), small chamber dimensions, stiff myocardial relaxation due to diastolic dysfunction, hypercontractile systolic function, or dilated atria due to diastolic dysfunction.

Chronic myocardial diseases (cardiomyopathies) present in 1 of 5 configurations as defined in the American College of Cardiology, the American Heart Association, and the World Health Organization classification. Most commonly, the ventricle is dilated and shows decreased contractility (eg, systolic function is impaired). In the second most common configuration, the ventricular wall is thickened or hypertrophic and can cause obstruction to systolic outflow or impair diastolic filling. Least often, there is a restrictive cardiomyopathy with grossly impaired filling that produces a small heart; there is normal systolic function but impaired diastolic function. Left ventricular noncompaction, the fourth configuration, is complex in that there are several different forms, all of which have hypertrabeculated left ventricles. In some cases, the right ventricle is also hypertrabeculated (eg, the biventricular noncompaction form). In some cases, the hypertrabeculated ventricle is associated with ventricular dilation, hypertrophy, dilated and hypertrophic, restrictive with enlarged atria, normal chamber dimensions and function, or associated with early-onset arrhythmias. The final forms of noncompaction are associated with congenital heart disease, most typically right heart–related congenital heart disease. Finally, the fifth classified form is arrhythmogenic cardiomyopathy (previously called arrhythmogenic right ventricular dysplasia or arrhythmogenic right ventricular cardiomyopathy) in which the ventricle (most commonly the right ventricle) is dilated with depressed systolic function and arrhythmias, particularly ventricular arrhythmias.

Dilated cardiomyopathy is considered most often to be primary or due to viral myocarditis; nutritional and toxic factors may be involved in some patients, and chronic tachycardias are uncommon but important causes. In addition, in patients with cancer, chemotherapy- or radiation-induced dilated cardiomyopathy occurs. Many kindreds in which dilated cardiomyopathy is inherited have been reported and, in approximately 30% to 40% of cases, have an identifiable genetic cause, resulting from mutations in genes mostly encoding sarcomeric and cytoskeletal proteins or genes involved in energy metabolism, or as part of several neuromuscular disorders (eg, Duchenne or Becker muscular dystrophy). Hypertrophic cardiomyopathy is frequently genetic, resulting from mutations in genes mostly encoding sarcomeric proteins or genes involved in energy metabolism, or as part of several neuromuscular disorders (eg, Friedreich ataxia). Restrictive cardiomyopathy is often associated with infiltration of the myocardium by cells or unusual chemicals but is also genetically transmitted, with sarcomeric genes or desmin mutations being most common. Arrhythmogenic cardiomyopathy is typically genetic with mutations in genes encoding proteins of the intercalated disk being most prominent. Ventricular noncompaction is considered a developmental disorder. Transcription factors, as well as genes encoding sarcomeric or cytoskeletal proteins or genes involved in energy metabolism have been implicated. Alternatively it may arise as part of several neuromuscular disorders (eg mitochondrial myopathies).

In chronic cardiomyopathies, the chest x-ray may show profound cardiomegaly (dilated type or the noncompaction type with congenital heart disease, particularly when associated with Ebstein anomaly) or a heart of normal size. The lung fields are usually clear. The electrocardiogram (ECG) often shows left ventricular hypertrophy, frequently with left axis deviation, increased left precordial and inferior forces, and STsegment flattening with Twave inversion. The clinical, ECG, and chest x-ray findings are frequently nonspecific, but the echocardiogram distinguishes the 5 cardiomyopathic configurations. In dilated cardiomyopathy, there is marked cardiac enlargement, ventricular wall thinning, and reduced systolic function. In hypertrophic cardiomyopathy, there is thickening of the ventricles that may affect the septum more than the ventricular free walls or may be uniform. Hypertrophy is frequently massive at the time of diagnosis, and systolic function may be hyperdynamic or normal. In restrictive cardiomyopathy, ventricular appearance and systolic function are near normal, but there are markedly dilated atria and restricted ventricular filling. In ventricular noncompaction and arrhythmogenic cardiomyopathy, the heart is either normal or enlarged on chest x-ray. Specific tests for diagnosis depend on the likely etiology of the disease process, but cardiac biopsy is only occasionally useful for making a specific diagnosis.

INFLAMMATORY DISEASES

Inflammatory diseases of the myocardium may be infectious or autoimmune mediated and acute or chronic. Acute infection is the most common inflammatory process that directly affects the heart. Acute myocarditis can occur at any age, including the fetus, and its course ranges from very mild to fulminant, with death in a few days or weeks in some children while others recover fully. Often there is a history of a recent upper respiratory tract infection, and viruses are usually the causative agents. In the 1970s and 1980s, the enteroviruses (coxsackie and echoviruses) were thought to be most common; then, in the 1990s, adenoviruses dominated; and since the early 2000s, parvovirus B19 has predominated. Polymerase chain reaction (PCR) of myocardial tissue specimens can detect viral genomes in patients with acute and chronic myocarditis. Interestingly, adenovirus and coxsackievirus share a cellular receptor (the coxsackie and adenovirus receptor [CAR]) that is important for viral entry into the cell. Many other viruses including parvovirus B19, mumps, cytomegalovirus, Epstein-Barr virus, hepatitis virus, human herpesvirus-6, and varicella can cause myocarditis, as can other pathogens (rickettsiae, Toxoplasma gondii, Mycoplasma pneumoniae, Chlamydia trachomatis, Borrelia burgdorferi). Diphtheria exotoxin causes profound myocardial damage by interfering with mitochondrial energy metabolism and is often associated with heart block, ventricular tachycardia, and fibrillation; mortality is high. Trypanosoma cruzi (Chagas disease) is a common cause of myocarditis in South and Central America; the acute myocarditis can be severe and is the primary cause of death in the acute phase of the illness, although chronic myocardial dysfunction, occurring after a long latent period, is a far more common result of T cruzi infection.

Immune-mediated myocarditis can also occur acutely. Rheumatic myocarditis may occur early in acute rheumatic fever and is almost always accompanied by endocardial (mitral or aortic valve regurgitation) involvement and sometimes by pericarditis. Kawasaki syndrome causes myocardial and pericardial inflammation, although myocardial dysfunction is rarely severe in the absence of significant obstructive coronary arterial lesions. Many other autoimmune diseases may affect the myocardium, including systemic lupus erythematosus (SLE), which may cause acute myocardial or pericardial inflammation, or chronic endocardial disease (Libman-Sacks endocarditis). Maternal SLE is also associated with conduction system damage and heart block in the fetus.

Treatment of acute myocarditis is supportive; bedrest, diuretics, “inotropic” agents, afterload reduction, and occasionally ventilatory support may all be necessary. Viral myocarditis can be associated with ventricular dysrhythmias and conduction abnormalities that may be potentiated by digoxin; if used, it should be given in low dose and with careful cardiac monitoring. Specific therapies include steroids in acute rheumatic (but not viral) myocarditis, gamma globulin and salicylates in Kawasaki syndrome, gamma globulin in viral myocarditis, antitoxin in diphtheria, and antimicrobials in bacterial, parasitic, mycoplasmal, or rickettsial myocarditis. The diagnosis of myocarditis can be inferred by documenting a specific clinical pattern (as in rheumatic myocarditis, Kawasaki syndrome, or diphtheria), by documenting myocardial inflammation by nuclear isotope scans or cardiac magnetic resonance imaging (CMR) with gadolinium demonstrating late gadolinium enhancement (LGE), or by identifying a potentially causative virus in the myocardium. Identifying by culture of the throat or feces with or without association with an appropriate antibody response may or may not have a cause-and-effect relationship with the disease. Proof of a specific virus as the cause requires its isolation from the myocardium or localization by PCR.

Chronic inflammation of the myocardium also occurs, although much less commonly in children compared to adults. Occasionally, viral myocarditis presents with chronic heart failure, dysrhythmias, and cardiomegaly with hypertrophy. Finding inflammatory infiltrates on myocardial biopsy suggests ongoing inflammation and has been suggested by some authors that perhaps this feature separates myocarditis from idiopathic dilated cardiomyopathy. However, many authors now suggest that the latter is a late manifestation of viral infection. Indeed, viral genomes can be isolated from the myocardium of many patients with dilated cardiomyopathy, even when an inflammatory infiltrate is not found on biopsy. This may reflect the patchy nature of the inflammatory process or be associated with a specific virus. For instance, coxsackievirus typically is associated with a robust inflammatory infiltrate while adenovirus more commonly has limited inflammation but significant fibrosis. Viral genomes generally have not been amplified from the myocardium of control patients with structural heart disease, suggesting that this test is specific for persistent viral infection. Cardiomyopathy is believed to result from ongoing inflammation and release of inflammatory cytokines that may lead to myocyte death. It has been shown that, in the case of coxsackievirus, the virus itself has a protein (a protease) that cuts the dystrophin protein (the cause of Duchenne and Becker muscular dystrophy) into fragments and results in dilated cardiomyopathy. Steroids and azathioprine have been used with only occasional success, probably because symptoms occur late in the process when cardiac reserve is exhausted. Intravenous immunoglobulin (IVIG) has shown promise.

Myocarditis causing a dilated cardiomyopathy is a common finding in human immunodeficiency virus (HIV) infection. Although the HIV virus has been isolated from the myocardium at autopsy or localized by PCR, other viral genomes are recovered at least as frequently in HIV-infected children, typically adenovirus or coxsackievirus, suggesting that HIV infection may predispose patients to persistent myocardial infection with other viruses and that it is these other viruses that lead to the dilated cardiomyopathy.

METABOLIC DISORDERS

Myocardial function may be impaired by dietary deficiencies or excesses of various substances or by enzymatic defects that impair energy utilization. Low calcium or magnesium levels may occasionally cause cardiac dilation and heart failure, particularly in premature infants. Hyperreflexia may be seen clinically, and there may be ECG abnormalities including a prolonged QT interval and occasionally atrioventricular block. Severe hypophosphatemia may also cause reversible heart failure. Hypokalemia and hyperkalemia can also cause rhythm abnormalities. Severe hypoglycemia can cause cardiac dilation and heart failure in newborns, particularly in infants born prematurely, in infants with intrauterine growth retardation or severe cyanotic congenital heart disease, or in infants of diabetic mothers. Infants of diabetic mothers can also have an obstructive cardiomyopathy that resembles hypertrophic cardiomyopathy with concentric or asymmetric septal hypertrophy. The heart usually returns to normal over several months.

Deficiencies of vitamins (eg, thiamine, causing beriberi), trace metals (eg, selenium, causing Keshan disease), amino acids (eg, taurine), or cofactors (eg, carnitine) can cause cardiac dilation and heart failure. These deficiencies can either be nutritional or, rarely, due to an inborn error of metabolism. For example, carnitine deficiency may be secondary to a variety of chronic illnesses, including mitochondrial myopathies, or may be a primary process. Treatment in all cases is directed toward redressing the specific deficiency. Hormonal abnormalities can also cause myocardial dysfunction. Hypothyroidism not only causes bradycardia, low cardiac output, and pericardial effusion but also can cause myocardial degeneration with ST and Twave changes. Pheochromocytoma may produce left ventricular hypertrophy and heart failure from sustained or paroxysmal hypertension; in addition, high catecholamine blood levels can cause subendocardial hemorrhages and myocardial degeneration that further impair cardiac function.

Other causes of myocardial dysfunction include the toxic effects of drugs (chloroquine, ipecac, cocaine, and doxorubicin, the last of which may cause severe, longlasting myocardial damage), radiation (particularly mediastinal radiotherapy for lymphoma), metabolic byproducts (uremic cardiomyopathy, which is rapidly reversible with hemodialysis), and severe anemia.

Genetic metabolic disorders are increasingly recognized as a cause of severe myocardial disease in early infancy. These disorders generally produce cardiomyopathy by interfering with energy metabolism. Specific defects may affect transport of substrate or cofactors into mitochondria (eg, carnitine-palmitoyl transferase-1 deficiency) or directly affect mitochondrial enzymes involved in β-oxidation of fatty acids (eg, long-chain and very-long-chain acyl dehydrogenase deficiency) or electron transport. These defects can be associated with dilated or hypertrophic cardiomyopathy and produce profound, but episodic cardiac failure. Although rare, diagnosis of these disorders is important because familial incidence is frequent.

INFILTRATIVE DISEASES

Many substances can infiltrate the myocytes or interstitium and impair cardiac function. Abnormal deposition of glycogen (especially type II glycogen storage, or Pompe disease) or glycolipid (Fabry disease) in the myocytes causes massive ventricular hypertrophy and obstruction with or without preserved systolic function. Hemosiderin may be deposited in the interstitium, causing myocardial fibrosis and subsequent systolic and diastolic dysfunction. This may be primary but is more commonly secondary to repeated blood transfusions in thalassemia, sickle cell anemia, and aplastic anemia. Mucopolysaccharidoses can produce myocardial degeneration. Hurler and Hunter syndromes also cause intimal thickening of the coronary arteries and, along with Morquio and Scheie syndromes, can cause valvar regurgitation. Other syndromes associated with myocardial infiltration and fibrosis include cystinosis, amyloidosis, and sarcoidosis. Infiltration may also be caused by neoplastic processes. The infiltration may be generalized, as in leukemia and lymphoma, or localized.

MYOCARDIAL ISCHEMIA AND ARRHYTHMIA

Myocardial ischemia is rare in childhood. It may be due to embolism, particularly in the perinatal period when venous thrombi easily cross the foramen ovale, but is more commonly caused by coronary artery abnormalities. Most frequent among these are congenital defects in which the coronary arteries communicate with another vascular bed producing ischemia through a vascular steal syndrome: anomalous origin of the left or right coronary artery from the pulmonary artery, arteriovenous fistula, and arteriocameral fistula. Coronary insufficiency also results from inborn errors of metabolism (eg, homozygous type 2 hyperlipidemia, homocystinuria), idiopathic calcification of the coronary arteries, and vasculitis disorders. The most common vasculitis affecting the coronary arteries occurs in Kawasaki syndrome; coronary arterial aneurysms occur in up to 20% of patients, although this incidence can be greatly reduced by early administration of gamma-globulin. Other vasculitis disorders, including juvenile rheumatoid arthritis, polyarteritis nodosa, and systemic lupus erythematosus, may also affect coronary arteries. Obliterative coronary arteritis has been associated with transplant rejection and myocarditis due to parvovirus B19, which causes endothelial inflammation and associated coronary narrowing. All of these diseases can produce myocardial ischemia, the extent of which varies with the size and number of arteries involved and the amount of collateral formation. There may be cardiac dilation, arrhythmias, and often mitral regurgitation due to damage to the papillary muscles.

Chronic atrial or ventricular tachycardias are associated with a dilated cardiomyopathy that, in many cases, can be cured if the arrhythmia is controlled. Myocardial ischemia has been considered to be a likely cause but remains an unproven mechanism. It is now known that some gene defects, such as mutations in the cardiac sodium channel gene SCN5A, lamin A/C, desmin, and the desmosomal genes, cause an arrhythmic form of cardiomyopathy.

PRIMARY CARDIOMYOPATHY

Diseases that primarily alter the structure of the myocyte can be separated into those known to be associated with a neuromuscular disorder and those without (idiopathic) disorders. The most common neuromuscular disorders affecting myocytes include Friedreich ataxia, which causes cardiac dilation or hypertrophy, or arrhythmias and heart failure in older children and adolescents, and progressive (Duchenne or Becker) muscular dystrophy, which may cause left ventricular hypertrophy and ST and Twave changes. There are many less common diseases, often with abnormalities in mitochondrial structure or function, that affect cardiac as well as skeletal myocytes.

Familial/inherited forms of dilated cardiomyopathy are also common. Inheritance may be autosomal dominant (most common), X-linked, or autosomal recessive. There are also maternally inherited (mitochondrial DNA) forms. Abnormalities in a wide variety of genes, most consistently genes encoding cytoskeletal, sarcomeric, or ion channel proteins, have been identified. These genes can be screened in commercially available laboratories and generally have approximately a 30% positive rate.

Dilated cardiomyopathy occurs in children of all ages. In patients requiring hospital admission for acute decompensated heart failure, treatment is supportive and includes diuretics, “inotropic agents” (eg, milrinone, calcium chloride, vasopressin intravenously; some centers continue to use dopamine, dobutamine, epinephrine, or norepinephrine), and vasodilators. In patients with compensated heart failure, angiotensin-converting enzyme inhibitors (or angiotensin receptor blockers) and β-blocker therapy, with or without diuretics, are commonly the treatment of choice. Some children with acute decompensated or chronic heart failure require cardiac transplantation and may need mechanical circulatory support prior to transplantation (or as a bridge to recovery). Another genetically determined cardiomyopathy, currently known as arrhythmogenic cardiomyopathy, was previously thought to only affect the right ventricle (hence its original name arrhythmogenic right ventricular dysplasia [ARVD] or arrhythmogenic right ventricular cardiomyopathy [ARVC]); it is dilated, hypokinetic, and generates severe arrhythmias, usually prior to the heart muscle becoming abnormal. The genes involved most frequently encode proteins of the desmosome within the intercalated disk. Based on the understanding of this disorder gained from genetics, it is now clear that both ventricles can be affected or just the left ventricle. It is usually inherited as an autosomal dominant trait, although autosomal recessive arrhythmogenic cardiomyopathy associated with skin disease (keratoderma, wooly hair) has also been reported.

Hypertrophic cardiomyopathy, characterized by hypertrophy of the interventricular septum and free wall (most commonly asymmetric septal hypertrophy), most commonly is generally transmitted in an autosomal dominant pattern, with nearly complete penetrance if patients are followed lifelong. Genetic hypertrophic cardiomyopathy is a disease of the sarcomere; in other words, most identified genes responsible for hypertrophic cardiomyopathy encode proteins of the contractile apparatus, the sarcomere. Mutations have been identified in many sarcomere-encoding proteins (eg, β-myosin heavy chain, troponin I, troponin T, α-tropomyosin, myosin binding protein C, the regulatory and essential myosin light chains) or the Z-disk, and routine molecular diagnosis is available and has an approximately 75% rate of mutation identification. Similar-appearing forms of ventricular hypertrophy that appear the same by imaging as the familial sarcomeric forms have been noted, including Pompe disease, Friedreich ataxia, some mitochondrial disorders, and Noonan syndrome.

Restrictive cardiomyopathy is uncommon in childhood and is characterized by severe atrial dilation in the presence of normal ventricular dimensions with normal systolic function and severe diastolic dysfunction. It is generally sporadic and idiopathic, but inherited forms also exist. The inherited forms most commonly follow an autosomal dominant pattern, but autosomal recessive and mitochondrial inheritance are also seen. As a general rule, the prognosis is poor, with a 2-year survival rate of less than 50% and a 5-year survival rate of 25%. The events surrounding the episodes of sudden death typically are arrhythmia related, either ventricular tachyarrhythmias or complete atrioventricular block. Multiple genes have been identified, with troponin I and desmin being most common. In addition, children with sickle cell disease have a high risk of developing restrictive cardiomyopathy. Patients undergoing cancer chemotherapy (with or without radiation) may develop restrictive cardiomyopathy as well. Treatment is difficult but usually involves the use of β-blockers. Cardiac transplantation is relatively common for these children.

Left ventricular noncompaction, a classified form of cardiomyopathy, is characterized by abnormal trabeculations in the left ventricle or in both chambers, most commonly seen at the apex with or without trabeculations at the left ventricular free wall. It may be associated with left ventricular dilation and/or hypertrophy, systolic and/or diastolic dysfunction, atrial enlargement, or various forms of congenital heart disease. Affected individuals are at risk of left or right ventricular failure, or both. Heart failure symptoms can be exercise induced or persistent at rest, but many patients are asymptomatic. Chronically treated patients sometimes present acutely with decompensated heart failure. Other life-threatening risks are ventricular arrhythmias and atrioventricular block, presenting clinically with syncope, and sudden death. Genetic inheritance arises in at least 30% to 50% of patients, and several genes that cause left ventricular noncompaction have been identified. These genes appear to generally encode sarcomeric (contractile apparatus) or cytoskeletal proteins, and disrupted mitochondrial function and metabolic abnormalities have a causal role as well.

Treatments focus on improvement of cardiac efficiency and reduction of mechanical stress in those with systolic dysfunction. Further, arrhythmia therapy and implantation of an automatic implantable cardioverter-defibrillator for prevention of sudden death are mainstays of treatment when deemed necessary and appropriate. Patients with left ventricular noncompaction associated with congenital heart disease commonly require surgical or catheter-based interventions.

SUGGESTED READINGS

INTRODUCTION

Infective endocarditis (IE) is a microbial infection of the endocardium. It is serious and life-threatening, with an increasing incidence in children. Over past few decades, the epidemiology of IE has changed due to increased survival of children with congenital heart disease (CHD). Concurrently, a growing proportion of cases occur in children without structural heart disease.

The clinical presentation of IE in children does not follow the classical presentation described in adults. Fever is the most common finding and may be associated with changing murmur and nonspecific symptoms. The primary site of infection, defense mechanisms of the host, and organism determine the clinical manifestations and rate of progression. IE can involve native or prosthetic cardiac valves, septal defects, mural endocardium, patent ductus arteriosus (PDA), or intravascular devices such as catheters, occlusion devices, patches, and surgically constructed shunts. Blood cultures and echocardiography are of paramount importance to aid in the diagnosis.

Despite advances in modern antimicrobial and surgical treatments strategies, IE continues to cause significant morbidity and mortality. Unfortunately, previous antibiotic prophylaxis strategies have proven ineffective. As the incidence of IE in children continues to rise, it is imperative for the general pediatrician and pediatric specialist to identify children at risk of IE, recognize its clinical presentation, and work together in planning its prevention and treatment.

PATHOGENESIS

IE is established by the interaction of a bloodstream pathogen with damaged endocardium. Intact endothelium is a poor stimulator of coagulation and is resistant to colonization by microorganisms; however, damaged endothelium is a potent inducer of thrombogenesis and provides a nidus to which bacteria adhere. Turbulent flow produced by certain congenital or acquired heart diseases causes sheer stress and damages the endothelium. Thrombogenesis ensues and results in the deposition of sterile clumps of platelets, fibrin, and the formation of nonbacterial thrombotic endocarditis (NBTE). NBTE may also be produced in structurally normal hearts when indwelling intravenous catheters damage the endocardium or valvar endothelium. NBTE can then be converted to an infected vegetation in a patient with transient bacteremia or fungemia. Activities of daily living, such as chewing, toothbrushing, and flossing, account for most bloodstream seeding of an NBTE. Bacteremia may also be caused by entry of organisms at the site of entry of percutaneous catheters, via the catheter lumen, or as a result of direct infection of an indwelling device at the time of placement. Bacteremia, however, does not invariably produce IE. The propensity to adhere to an NBTE depends on the type of microorganism. Gram-positive microorganisms are most commonly responsible for IE in children because they possess specific surface components that facilitate adherence.

EPIDEMIOLOGY

Before the 1970s, rheumatic heart disease (RHD) was the underlying substrate for IE in 30% to 50% of children. While RHD remains an important substrate for IE globally, its incidence in developed countries has diminished substantially. CHD is the predominant underlying condition for IE in children > 2 years of age. Aortic valve abnormalities, ventricular septal defects, and tetralogy of Fallot (TOF) are common underlying conditions. Corrective surgery or catheter-based intervention in children with a septal defect or PDA eliminates the attributable risk of IE 6 months after intervention if there is no residual defect. However, there is now an increasing proportion (~50%) of children with IE who have undergone previous corrective or palliative surgery for CHD. Postoperative IE is a long-term risk after surgery for complex cyanotic CHD. Those at highest risk for postoperative IE include patients with a conduit, prosthetic material, a residual defect, aortic stenosis, or a previous episode of IE.

It is imperative for the pediatrician to recognize that children with structurally normal hearts can still be at risk of IE. At least 10% to 20% of cases of pediatric IE occur in children without structural heart disease. These cases tend to involve right-sided heart structures and are most likely related to modern intensive care measures and nosocomial infections. The incidence of IE in neonates has increased over the past 2 decades, and more than two-thirds of cases of neonatal IE occur in the absence of CHD. This is the result of trauma to the skin and mucous membranes, vigorous endotracheal suctioning, parenteral hyperalimentation, and placement of umbilical and/or peripheral venous catheters.

Microbiology

The most common pathogens in neonates are Staphylococcus aureus, coagulase-negative staphylococci (CONS), gram-negative bacterial species, and Candida species. After the first year of life, viridans group streptococci (VGS; eg, Streptococcus sanguis, Streptococcus mitis group, Streptococcus mutans) and S aureus are the most frequently isolated organisms in patients with underlying CHD. Importantly, S aureus is the most common cause of acute (rapidly progressive) IE. CONS are also a frequent cause of IE and can cause infection in both native and prosthetic valves, indwelling vascular catheters, and prosthetic materials. Enterococcus IE occurs much less frequently in children than in adults. Gram-negative organisms cause < 10% of IE in children and include the HACEK (Haemophilus species, Aggregatibacter actinomycetemcomitans, Cardiobacterium hominis, Eikenella species, and Kingella species) organisms. The prevalence of culture-negative endocarditis (CNE) is approximately 5% to 10%; the most common causes are recent or current antibiotic therapy or infection caused by a fastidious organism such as a HACEK organism, Bartonella, Tropheryma whipplei, Coxiella burnetii (Q fever), or Brucella.

CLINICAL MANIFESTATIONS

The clinical presentation of IE in children is most often indolent and nonspecific with a prolonged low-grade fever. The classically described extracardiac signs of IE such as Janeway lesions, Osler nodes, Roth spots, and splinter hemorrhages are rare in children and absent in neonates. Fever is initially continuous; however, it may become recurrent and low grade with asymptomatic intervals if the child has received antibiotics. Fever may be associated with nonspecific complaints including fatigue, weakness, arthralgias, myalgias, weight loss, rigors, or diaphoresis. On occasion, children with IE may present acutely ill and appear toxic with high fever. Neonates with IE may not have fever but rather nonspecific signs that may be indistinguishable from septicemia. These signs include hypotension, feeding intolerance, or respiratory distress. Septic embolic events in neonates are common and may lead to the development of satellite infections including meningitis and osteomyelitis.

The physical examination is highly variable due to possible structural changes at the local infection site, embolization from vegetations to distant organs, and circulating immune complexes. Serial auscultation is crucial because it may be difficult to appreciate changes in a patient with a pre-existing murmur. A new or evolving valvar lesion may cause leaflet destruction, resulting in a new or louder murmur or gallop due to congestive heart failure (CHF). Alternatively, attenuation of a continuous murmur or decrease in oxygen saturation in a cyanotic child with a prosthetic systemic-to-pulmonary shunt may reflect obstruction to flow due to graft infection. Heart rate and rhythm must also be closely monitored because heart block may be a sign of a life-threatening complication.

Complications are often due to CHF or systemic embolization of a vegetation. Although embolic events can occur late, the majority occur in the first 2 to 3 weeks after diagnosis and initiation of therapy. Embolization is most commonly due to fungal or S aureus infection. IE involving the left side of the heart, particularly the anterior leaflet of the mitral valve, frequently results in peripheral embolization to the abdominal viscera or brain. Renal involvement, which may also be due to immune complex deposition, can result in glomerulonephritis. Neurologic involvement may result in cerebral infarct, abscess, or meningitis. Embolization from the tricuspid valve may cause septic pulmonary emboli, resulting in breathlessness and scattered, fluffy infiltrates on chest radiographs.

IE simulates a wide variety of disorders including other infectious diseases, malignancy, and connective tissue disease. As such, IE should be part of the differential diagnosis in patients with fever of unknown origin and febrile patients with underlying heart disease.

DIAGNOSIS

Blood cultures should be obtained in any patient with fever of unexplained origin and a pathologic heart murmur, a history of CHD, or a previous episode of IE. The yield from blood cultures is related to the amount of blood obtained. At least 1 to 3 mL in infants and young children and 5 to 7 mL in older children are generally adequate. Three blood cultures should be obtained by separate venipunctures on the first day, and if there is no growth by the second day of incubation, 2 more may be obtained. Cultures must be obtained prior to initiation of antibiotic therapy. Initiation of empiric antibiotics in the absence of blood cultures is strongly discouraged. A single culture is not sufficient and increases the likelihood of isolating a contaminant. Whenever possible, samples from indwelling catheters should not be used. Because anaerobic bacteria rarely cause IE, the emphasis is usually on obtaining aerobic cultures. It is not necessary to obtain cultures at any particular phase of the fever cycle because bacteremia is usually continuous.

The microbiology laboratory and infectious disease specialist should be consulted in all suspected cases of IE. The laboratory will need to incubate cultures for at least 2 weeks if fastidious or unusual organisms are suspected. Serologies and molecular techniques, such as polymerase chain reaction, may help identify fastidious organisms, and cell cultures may help identify intracellular organisms such as Bartonella or Coxiella. In patients who undergo surgery for IE, resected cardiac tissue is helpful as well. Several additional but nonspecific laboratory tests may help support the diagnosis. Patients often have elevated acute-phase reactants, anemia (due to hemolysis or chronic disease), and microscopic hematuria or proteinuria; leukocytosis is generally inconsistent, but patients may have a left shift. CNE is diagnosed if a patient has clinical and/or echocardiographic evidence of IE but blood cultures are repeatedly negative. Conversely, a positive blood culture in a child with CHD or a history of previous IE does not necessarily indicate IE.

Two-dimensional echocardiography is the principal modality for detecting vegetations. Typical echocardiographic findings include the presence of a vegetation, abscess, and new or worsening valvar insufficiency. Children have superior echocardiographic windows compared to adults, and transthoracic echocardiography (TTE) has an excellent reported sensitivity (81–97%) for detecting vegetations. Transesophageal echocardiography (TEE) is more invasive, more expensive, and generally unnecessary in most pediatric cases. However, TEE should be considered when TTE views are limited or if there is concern for prosthetic valve IE or left ventricular outflow tract complications. The absence of vegetations on echocardiography does not rule out the presence of IE. Regardless of technique, imaging should be performed at the same center as any previous echocardiogram to allow for serial comparison. A pediatric cardiologist should be consulted to help interpret results of TTE in the setting of a patient’s overall clinical picture, determine whether TEE is indicated, and help confirm or reject the diagnosis of IE.

IE is readily diagnosable in children with at least 2 positive blood cultures, structural heart disease, and identified vegetations. However, the diagnosis can be challenging when echocardiography is negative or when a child with CHD has bacteremia. The pediatrician should therefore refer to the modified Duke criteria (Tables 484-1 and 484-2) to make the diagnosis of IE.

TREATMENT

Management requires a multidisciplinary team consisting of a pediatric cardiologist, infectious disease specialist, cardiothoracic surgeon, and general pediatrician. Empiric antibiotic treatment in a child with CHD with fever alone should be discouraged. For cases of suspected or confirmed IE, a prolonged course of bactericidal intravenous antibiotics is given for 4 to 8 weeks. In patients who are not acutely ill, antibiotics may be withheld while additional cultures are obtained. The choices of antibiotics and duration of therapy are determined by the specific pathogen, site of infection, underlying anatomy, and presence of prosthetic material. The detailed antibiotic regimens for specific pathogens (and CNE) are described elsewhere (see Baltimore et al, 2015, in Suggested Readings). Serum concentrations of vancomycin and aminoglycosides should be monitored to achieve therapeutic levels and prevent potential side effects such as nephrotoxicity and ototoxicity.

Outpatient treatment can be considered in select patients after initial inpatient management. These patients must be hemodynamically stable, afebrile, have negative blood cultures, and not be at high risk for complications. There should be easy access to a hospital for prompt reevaluation should complications develop. Frequent monitoring by a home health nurse who can assess adherence to drug therapy is also essential.

Blood cultures should be obtained daily until bacteremia resolves; this usually occurs within several days of appropriate treatment. The efficacy of antimicrobial therapy is indicated by the disappearance of fever, sterilization of blood cultures, and normalization of inflammatory markers. Clinical and biologic surveillance (including occasional repeat blood cultures) should continue in the subsequent 6 to 8 weeks after completion of therapy when the risk of relapse is highest. A follow-up echocardiogram should also be performed to establish a new baseline. In most cases, this is all that is required to cure the infection and prevent life-threatening complications. However, there are cases where cardiovascular surgery may be necessary and lifesaving.

Prophylactic surgery and use of anticoagulation are not currently recommended to prevent vegetation embolization. Although it is difficult to predict an individual patient’s risk of embolization, there are several clinical and echocardiographic risk factors that may predispose to potentially life-threatening complications necessitating surgery. These include the type of organism (S aureus or fungus), location and size of a vegetation (Table 484-3), specific anatomy (cyanotic CHD, presence of a prosthetic valve or systemic-to-pulmonary shunt), previous occurrence of IE, poor clinical response to antimicrobial therapy (persistent bacteremia despite treatment for > 7–10 days), or prolonged symptoms (> 3 months). Surgical intervention should also be considered in patients with new heart block, mycotic aneurysm, or CHF. Early surgery improves overall outcomes and can be performed with low morbidity and mortality. The presence of bacteremia or degree of illness does not preclude surgery. However, any decision regarding surgical intervention must be individualized.

Most children with IE are cured with antibiotic therapy and occasionally surgical management. However, overall inpatient mortality remains at least 5% to 10%. This rate is even higher in premature infants and patients with IE caused by S aureus. Specific types of CHD are also associated with poorer outcomes. Children who survive an episode of IE are at increased risk for relapse and thus must be followed closely and educated to report occurrence of fever. The pediatrician therefore plays a critical role in surveillance.

PREVENTION

The prophylactic administration of antibiotics prior to dental procedures has been routine for over 50 years. While guidelines have continued to evolve, published data have failed to prove the effectiveness of antibiotic prophylaxis to prevent IE. There is consistent scientific evidence to suggest that dental procedures are a rare cause of IE, and prophylactic antibiotics prior to such procedures may prevent an exceedingly small number of cases. On the contrary, bacteremia from activities of daily living far exceeds that of dental procedures and is therefore much more likely to cause IE. This has resulted in a significant shift in prevention strategies toward a greater emphasis on oral hygiene and access to dental care for children with CHD. This emphasizes the necessary role the pediatrician to stress the importance of good oral hygiene, prevention of gingival and dental disease, and access to routine dental care for children at risk for IE.

The most recent recommendations from the American Heart Association (AHA) and European Society of Cardiology now limit the use of prophylactic antibiotics to patients with the greatest risk of morbidity or mortality. Prophylaxis is now only recommended for children with a prosthetic valve or prosthetic material, certain types of CHD (unrepaired cyanotic CHD, repaired CHD with prosthetic material or device during the first 6 months after the procedure, and repaired CHD with residual defects at the site or adjacent to the site of a prosthetic patch or prosthetic device), or a previous case of IE. It is also recommended in heart transplant recipients with cardiac valvulopathy. Prophylaxis consists of a single dose of amoxicillin (50 mg/kg; maximum, 2 g) 30 to 60 minutes prior to all dental procedures that involve manipulation of gingival tissue, the periapical region of teeth, or perforation of the oral mucosa. The 2007 AHA guidelines contain alternative regimens. Prophylaxis is also reasonable for these patients prior to procedures on respiratory tract or infected skin, skin structures, or musculoskeletal tissue. Prophylaxis is not recommended prior to genitourinary or gastrointestinal tract procedures. To date, no study has demonstrated an increased rate of IE as a result of these guidelines.

SUGGESTED READINGS

INTRODUCTION

The pericardium serves as the external protective layer of the heart, providing a barrier to trauma, malignancy, inflammation, and infection. In addition, a thin layer of fluid within the pericardial space lubricates the heart, thereby reducing the energy expenditure and shear stress encountered with cardiac motion. A basic understanding of the anatomy and physiology of the pericardium and pericardial space is necessary to recognize and properly diagnose disease processes affecting this important structure.

ANATOMY

The pericardium is a fibrous lining surrounding the heart, consisting of 2 layers: the parietal (outer) layer and visceral (inner) layer. The visceral layer comes into direct contact with the epicardial surface, whereas the parietal layer contains collagenous fibers and serves as an outer protective layer. The pericardium completely envelops the atria and ventricles and terminates just superior to the semilunar valves; it also extends to surround the vena cavae and pulmonary veins near their respective entrances into the heart. Pericardial fluid is a serous, low-viscosity fluid found between the visceral pericardium and the epicardial surface of the heart. In children, approximately 10 mL of fluid is normally found within the pericardial space, with larger volumes (20–30 mL) encountered in adults.

PATHOLOGIC CONDITIONS

Pericardial Effusion

Pericardial effusion is defined as the excess accumulation of fluid within the pericardial space. It is frequently, but not universally, a manifestation of pericardial inflammation. The most common cause of pericardial effusion in children is pericarditis, which in turn has a variety of etiologies. The causes of pericardial effusion and pericarditis can broadly be categorized into infectious, inflammatory/autoimmune, traumatic, toxic, and idiopathic (Table 485-1). Infectious etiologies are the most common, most often in conjunction with viral infections associated with pericardial inflammation or as part of a broader cardiac inflammatory process (ie, myocarditis). Bacterial seeding of the pericardial space occurs less frequently; however, tuberculosis is a common cause of pericarditis and pericardial effusion in developing countries. Many autoimmune processes may have an associated pericarditis, such as systemic lupus erythematosus, in which pericardial effusion may be a presenting sign of the disease. Trauma is a common cause of pericardial effusion, via contusion of the heart or direct perforation of the pericardium by penetrating thoracic injury. Several environmental toxins and drugs can be associated with pericardial effusion; in pediatric practice, this is often associated with chemotherapeutic agents or chest irradiation in the setting of treatment for malignancy. Finally, interruption or obstruction of the lymphatic drainage in the chest, usually encountered in the postoperative setting after cardiac surgery in children, may manifest as pericardial effusion with a high triglyceride content (chylopericardium).

The hemodynamic effects of pericardial effusion, as well as the presence and nature of symptoms, depend on a number of factors. The size of the fluid collection is the most important determinant of symptoms. In addition, the rate at which a pericardial effusion accumulates has an effect on symptoms. Relatively small accumulations of fluid that develop rapidly (eg, hemopericardium occurring in the setting of trauma) may be life-threatening; in contrast, in other etiologies with a more indolent clinical course, gradual accumulation of pericardial fluid allows time for the pericardium to stretch and therefore adapt to the anatomic and physiologic changes brought about by the presence of sizeable fluid collections.

Pericardial effusions exert their adverse hemodynamic effects by impeding cardiac filling in diastole, which in turn lowers stroke volume and cardiac output. With the accumulation of sufficient pericardial fluid, intrapericardial pressure exceeds atrial pressure, leading to collapse of the atrial cavity, impairment of cardiac filling, and limitation of cardiac output, a scenario known as cardiac tamponade. If unrecognized or left untreated, tamponade may quickly progress to circulatory collapse and death.

The initial physiologic response to pericardial effusion is an increase in heart rate, allowing for preservation of cardiac output in the setting of lower stroke volume. Hypotension is a late and ominous finding and may herald impending shock and circulatory collapse. In addition to tachycardia, other physical examination findings in patients with significant pericardial effusions include narrow pulse pressure, jugular venous pulsations, and tachypnea. Hepatomegaly may be present in large effusions that have enlarged gradually. Cardiac auscultation will reveal muffled or distant heart tones. A pericardial friction rub, resembling the crackling sound create by rubbing a tuft of hair between 2 fingers, will often be present in small pericardial effusions but absent in large collections.

In a patient with suspected pericardial effusion, rapid clinical assessment is necessary to determine whether there is evidence of existing or impending tamponade. An important physical examination finding frequently used to assess this risk is known as pulsus paradoxus. Normally, systolic blood pressure varies by a few mm Hg throughout the respiratory cycle, with the blood pressure decreasing during inspiration due to increased intrathoracic pressure. In a hemodynamically significant pericardial effusion, this normal respiratory variation in blood pressure is exaggerated, with systolic blood pressure decreasing by more than 10 mm Hg during inspiration. Pulsus paradoxus is caused by limitation of left ventricular filling in the setting of a significant pericardial fluid collection. Pulsus paradoxus is demonstrated using a manual blood pressure cuff; the patient should be supine and breathing spontaneously. The blood pressure cuff is inflated, and with deflation, auscultation is used to detect and measure the systolic blood pressure at which the Korotkoff sound is first heard. The cuff is then deflated further; the systolic blood pressure is then detected and measured when the Korotkoff sound is heard continuously during inspiration and expiration. The difference between these 2 systolic blood pressures is the magnitude of the pulsus paradoxus. The presence of pulsus paradoxus can also be inferred through palpation of the pulses (in which an exaggerated change in the amplitude of the pulses can be detected with normal respiration) or in patients with invasive arterial catheters (in which the magnitude of the blood pressure change can be observed on the cardiorespiratory monitor).

With a significant pericardial effusion, the chest radiograph will usually exhibit an enlarged cardiac silhouette, classically referred to as a “water bottle” silhouette (Fig. 485-1), although in cases of rapid accumulation, the increase may not be dramatic. Electrocardiography may show sinus tachycardia with low QRS voltage amplitudes (Fig. 485-2); in severe cases, electrical alternans, the presence of variation in QRS amplitude from beat to beat, can be observed. If there is associated pericarditis or myocarditis, there may be ST-segment and T-wave abnormalities; these changes are more diffusely distributed through the precordial and limb leads compared to the ST-segment and T-wave changes observed with myocardial ischemia.

Echocardiography can rapidly delineate the size and extent of pericardial effusion, with several findings associated with cardiac tamponade (Fig. 485-3). These include right atrial free wall collapse in diastole, compression of the right ventricular outflow tract, and a > 20% variation in mitral inflow Doppler velocities with respiration (ultrasonographic correlate of tamponade). However, tamponade is a clinical diagnosis that cannot solely be made by imaging. Certain ultrasound characteristics of the pericardial fluid may give insight into the etiology of the effusion; for example, the presence of fibrinous strands within the fluid implies an inflammatory or infectious origin. It is important to recognize that a small rim of pericardial fluid is often present along the inferior border of the heart in normal healthy individuals and does not represent a pathologic effusion.

For patients with symptomatic pericardial effusion, impending cardiac tamponade, or frank tamponade, urgent drainage of the fluid (pericardiocentesis) by percutaneous means may be lifesaving. Surgical drainage of pericardial effusion is usually limited to patients with bacterial pericarditis or in patients with chronic, recurrent effusions (pericardial window). In the presence of tamponade, temporizing medical therapy includes oxygen, fluid resuscitation to augment preload, and avoidance of agents that decrease systemic vascular resistance. Diuretics, sometimes used in small effusions without hemodynamic compromise, should be avoided if there is concern for tamponade. Pericardiocentesis is usually performed with real-time echocardiographic guidance. A long needle is advanced into the pericardial space from the subxiphoid region, with brisk flow of fluid from the needle confirming entry into the pericardial space. Frank blood from the needle may indicate hemopericardium or entry into heart itself. After acute drainage of fluid, a catheter is usually advanced over a wire into the pericardial space to allow for continued drainage while the inciting process for the effusion resolves. Pericardial fluid should be sent for cell count analysis, Gram stain, cytology, and culture for diagnostic purposes. Testing of pericardial fluid via polymerase chain reaction (PCR) methods is available and can rapidly identify many viral etiologies of pericarditis.

Pericarditis

Pericarditis, or inflammation of the pericardium, is the most common cause of pericardial effusion in children. Pericarditis has numerous etiologies, with the most common including viral pathogens (see Table 485-1). Certain viral infections, such as enteroviruses (coxsackievirus), adenovirus, influenza, and parvovirus, have been associated. It is presumed that most cases of idiopathic pericarditis are related to viral infection, regardless of whether a prodromal illness or evidence of viral infection on laboratory testing is present. Bacterial (purulent) pericarditis is much less common and is associated with Staphylococcus aureus, Streptococcus pyogenes, and Neisseria meningitidis organisms or mycobacterial infection (Mycobacterium tuberculosis). Frequently, there is a history of prior or concurrent bacterial infection by the same agent at another site, such as osteomyelitis, meningitis, or pneumonia. Unlike viral pericarditis, bacterial pericarditis does not resolve spontaneously and requires intravenous antibiotic therapy with pericardial drainage. The course may be fulminant, and prolonged courses of antibiotics and pericardial drainage are usually required.

The most common symptom of pericarditis is chest pain, which is usually sharp in nature and localized to the left mid-sternal region. Older patients may describe a sensation of heaviness in the chest, often relieved by leaning forward and exacerbated by inspiration; frequently, there is associated fever. In younger patients, a history of irritability and feeding intolerance is common. Auscultation may reveal a pericardial friction rub, the hallmark physical examination finding in pericarditis.

Diagnostic testing for suspected pericarditis usually involves an electrocardiogram (ECG), echocardiogram, and laboratory assessment. The ECG in pericarditis is notable for sinus tachycardia, depression of the PR segment, and ST-segment elevation with T-wave abnormalities; these are often diffuse throughout all the precordial leads. The ECG findings often overlap those found in myocarditis, although arrhythmias (common in myocarditis) are usually absent in pericarditis. Serum inflammatory markers, such as the erythrocyte sedimentation rate and C-reactive protein, are usually elevated, albeit in nonspecific fashion. Serum troponin is a sensitive marker for cardiac injury. In pericarditis, serum troponin is normal or only minimally elevated; this is in contrast to myocarditis, myocardial trauma, or myocardial ischemia. An echocardiogram is indicated to assess for pericardial effusion, atrioventricular valve regurgitation, and determination of ventricular systolic function. In pericarditis, a pericardial effusion is usually present, and importantly, ventricular function is preserved. The presence of impaired ventricular function, significant atrioventricular or semilunar valvar regurgitation, and/or cardiac chamber dilation is suggestive of an alternative diagnosis, such as myocarditis or cardiomyopathy. In the middle of the clinicopathologic spectrum between pericarditis and myocarditis is an increasingly recognized entity that has been termed “myopericarditis.” Myopericarditis, which is frequently observed in adolescent males, is a distinct syndrome that presents with chest pain, ECG abnormalities, normal ventricular systolic function on echocardiography, and elevation of cardiac enzymes (troponin and creatine kinase-MB fraction). At times, discrimination among pericarditis, myopericarditis, and myocarditis is difficult, and there is an increasing role for cardiac magnetic resonance imaging (MRI) in establishing a diagnosis in uncertain cases.

Idiopathic or viral-associated pericarditis is most often a self-limited process, with therapy mainly being supportive. This is in contrast to bacterial pericarditis, in which patients are systemically ill and often bacteremic. The long-term prognosis for viral-associated pericarditis is favorable, although recurrence may occur in as much as approximately 30% of patients. Drainage of any associated pericardial effusion is indicated for the circumstances described earlier in this chapter. Nonsteroidal anti-inflammatory agents (ibuprofen, naproxen, indomethacin) are often useful at alleviating symptoms. Corticosteroids (prednisone, methylprednisolone) have been used in refractory cases. There has been increasing study of the microtubule assembly inhibitor colchicine in the primary treatment of pericarditis, which in recent studies has been demonstrated to be superior to nonsteroidal anti-inflammatory drugs alone in effecting the resolution of symptoms. Colchicine also has value in the prevention of recurrent pericarditis. There is limited experience with the use of colchicine in infants and very young children.

Postpericardiotomy Syndrome

Postpericardiotomy syndrome is a distinct entity encountered in children after cardiac surgery at which the pericardium has been opened. Postpericardiotomy syndrome typically presents with fever beyond the first postoperative week (without evidence of infection elsewhere), new pericardial effusion, chest pain, friction rub, and in some cases, new pleural effusions. It is encountered in all age groups and surgeries, but for unclear reasons, it appears to be more commonly encountered in patients who have undergone atrial surgery (eg, surgical atrial septal defect repair, atrioventricular canal defect repair). It is also common after heart transplantation. The etiology is uncertain but is felt to have an inflammatory component. Patients may present several weeks to months after surgery, so it is important for clinicians to be aware of this diagnosis in patients with a history of cardiac surgery. Treatment includes nonsteroidal anti-inflammatory agents and pericardiocentesis for large fluid collections.

Pericardial Constriction (Constrictive Pericarditis)

Constriction of the pericardium may occur in several chronic processes, leading to a less compliant pericardium, which in turn impedes ventricular filling in diastole. It is rare in children. It is an indolent process that may follow recurrent episodes of pericarditis, or it may be seen as a late complication of chest irradiation for malignancy. In developing countries, untreated M tuberculosis infection is the most common cause of pericardial constriction. Symptoms are often nonspecific and develop gradually, with fatigue and dyspnea being common. Physical examination may reveal hepatomegaly, neck vein distension with jugular venous pulsations, narrow pulse pressure, a gallop or knock, and distant heart sounds. The chest radiograph may show a small heart, with pericardial calcification. The echocardiogram often shows enlarged atria with normal ventricular cavity size, pericardial thickening, abnormal Doppler indices of diastolic function, and a distinctive abnormal motion of the interventricular septum (“septal bounce”). Systolic function is usually preserved.

The physiology of pericardial constriction is similar to restrictive cardiomyopathy, in which abnormalities of diastolic relaxation result in the need for higher cardiac filling pressures to maintain cardiac output. In clinical practice, discriminating between pericardial constriction and restrictive cardiomyopathy is often difficult. Over time, these higher cardiac filling pressures may be transmitted in a retrograde fashion to the pulmonary vascular bed and lead to pulmonary vascular disease and pulmonary hypertension; thus, early recognition and diagnosis are important. Medical therapy is limited; surgical therapy (pericardial stripping) may be effective at relieving symptoms.

Rare Conditions of the Pericardium

Congenital absence of the pericardium is a rare condition in which a large portion or all of the pericardium is absent. Patients with this entity are usually asymptomatic, particularly in those with complete absence of the pericardium. The diagnosis may be discovered during the incidental discovery of cardiomegaly on a chest radiograph performed for other indications. Congenital absence of the pericardium may also come to clinical attention if a portion of the heart (often one of the atrial appendages) herniates through the defect in the pericardium, which may occur on a relapsing and remitting basis. The diagnosis is confirmed by MRI, and treatment is surgical if there is concern for herniation and/or incarceration of the heart through the pericardial defect.

Certain tumors may be located within the pericardium, causing compression of the adjacent cardiac structures, although primary pericardial tumors are rare. Malignancy affecting the pericardium is more commonly related to metastases from solid organ or hematologic malignancies. Pericardial cysts are benign cystic structures within the pericardium that can lead to localized pericardial fluid collections.

SUGGESTED READINGS

INTRODUCTION

Pulmonary hypertension is an abnormal elevation of the pulmonary arterial pressure, which can occur for a number of different reasons. Pulmonary hypertension can be idiopathic or associated with various cardiac, pulmonary, and systemic diseases in infants and children and is often multifactorial in etiology. It is important to identify and treat any known causes of pulmonary hypertension or associated conditions that may exacerbate the disease. Regardless of the etiology, pulmonary hypertension can be associated with significant morbidity and mortality due to right heart failure. While current therapies treat pulmonary hypertension and prolong life, there is no cure to reverse the disease process in many forms of pulmonary hypertension. There have been recent advances in therapeutic options, but there are limited data on use in children.

PATHOGENESIS AND EPIDEMIOLOGY

Pulmonary hypertension is defined as a mean pulmonary artery pressure ≥ 25 mm Hg. It can occur as an isolated process, previously termed “primary pulmonary hypertension,” or can be related to a number of cardiopulmonary or systemic diseases, previously “secondary pulmonary hypertension.” In 1998, the World Health Organization developed a new classification for pulmonary hypertension that categorizes it into 5 major groups based on etiology. This was updated most recently at the World Symposium of Pulmonary Hypertension in Nice in 2013 (Table 486-1). To address the need for a classification system that incorporates more pediatric-specific aspects of pathobiology and related clinical issues, a comprehensive classification of pediatric pulmonary hypertension was proposed at the Panama Pulmonary Vascular Research Institute symposium (PVRI) in 2011, highlighting 10 major categories of pediatric pulmonary vascular disease (Table 486-2).

Pulmonary arterial hypertension, so-called group 1 within the Nice classification, is typically a progressive vasculopathy characterized by a series of changes, including abnormal muscularization of peripheral arteries, medial hypertrophy of the pulmonary arteries, and proliferation of cells lining the arteries called neointima formation. This results in occlusion of the vessel lumen and, ultimately, loss of peripheral arteries. This progressive decrease in the vascular surface area increases the pulmonary vascular resistance and pulmonary artery pressure, causing an increased workload on the right ventricle. In 1958, Heath and Edwards developed a classification scheme characterizing the progression of structural changes from grade 1 through 6 (Fig. 486-1). While the precise molecular mechanisms underlying these vascular changes remain unclear, recent studies implicate abnormalities in endothelial cell function, response to injury, vascular cell signaling, gene expression, mitochondrial function, metabolic pathways, ion channel function, and chronic inflammation as important contributors to progression of disease.

Pulmonary arterial hypertension in children is most commonly associated with congenital heart disease or is idiopathic in nature, with no cause identified. Additional associated causes of pulmonary arterial hypertension include connective tissue diseases such as scleroderma, portal hypertension, human immunodeficiency virus (HIV), schistosomiasis and certain drugs and toxins, although these etiologies are relatively uncommon in the pediatric population. Pulmonary arterial hypertension can be hereditary. A small number of associated genes have been identified in familial cases of pulmonary arterial hypertension and also as de novo mutations in idiopathic cases, most notably bone morphogenetic protein receptor 2 (BMPR2). However, in families with prevalent BMPR2 mutations, the disease penetrance is low, with only about 30% of those carrying mutations developing the disease. The penetrance may be slightly higher in females. Additional genes associated with pulmonary arterial hypertension include ACVRL1 and endoglin (ENG), which are associated with hereditary hemorrhagic telangiectasia, and the more recently described SMAD9, related to BMPR2 signaling, CAV1, related to caveolar structure and nitric oxide signaling, and KCNK3, a potassium channel. Pulmonary arterial hypertension can also be seen in children with multiple congenital anomalies or genetic syndromes, including trisomy 21.

Pulmonary arterial hypertension can be associated with various types of congenital heart disease, which may be related to effects of prolonged high pressure and flow on the pulmonary vasculature. The age of the child and type of cardiac lesion determine the likelihood of developing pulmonary vascular disease with irreversible structural changes. Patients with a large posttricuspid shunt, such as ventricular septal defect (VSD) or patent ductus arteriosus (PDA) are at highest risk. Natural history studies indicate that most children do not develop irreversible vascular changes prior to 2 years of age. Therefore, early surgical repair of defects, often within the first 6 months of life, is now standard of care. Severe pulmonary vascular disease preoperatively remains an important risk factor for acute right heart failure or death and is therefore considered a contraindication for surgical repair. When there is concern for elevated pulmonary vascular resistance (PVR) preoperatively, suggested by low oxygen saturations, lack of evidence for pulmonary overcirculation, or right-to-left shunting across the defect, a cardiac catheterization is often indicated to assess operability. Patients with a fixed indexed PVR above 6 Woods units • m² are generally considered higher risk for repair and may not be surgical candidates. Response to vasodilators may suggest reversibility. A trial of targeted medical therapy for pulmonary hypertension may be used to improve PVR prior to consideration of surgical repair.

Patients with a large shunt lesion who do not undergo timely repair are likely to develop irreversible vascular changes with a progressive increase in PVR, eventually exceeding the systemic vascular resistance and causing right-to-left shunting. This reversal of shunt is termed Eisenmenger syndrome and is associated with systemic desaturation, cyanosis, and erythrocytosis. Ventricular function is often preserved in these patients; however, complications from severe hypoxemia and hyperviscosity may be profound.

Pulmonary vascular disease is also seen in a subset of patients with certain cardiac lesions, such as d-transposition of the great arteries, complete atrioventricular septal defect, and truncus arteriosus, which may persist despite early surgical repair, suggesting there may be genetic or developmental factors contributing to pulmonary hypertension in these patients. Patients with an atrial septal defect (ASD) are much less likely to develop pulmonary vascular disease and do not typically do so until the third to fifth decade of life.

Pulmonary veno-occlusive disease (PVOD) and pulmonary capillary hemangiomatosis (PCH) are capillary or postcapillary vasculopathies thought to be similar to pulmonary arterial disease but designated as 1′ in the Nice classification given that the postcapillary vessels are predominantly affected. Both are associated with gene mutations in EIF2AK4 and are thought to represent a spectrum of postcapillary disease. Persistent pulmonary hypertension of the newborn (PPHN) is also included in the Nice classification as 1′′ as it is recognized as a distinct disease process.

The second group of conditions within the Nice classification includes pulmonary hypertension due to left heart disease (“venous hypertension”). Pulmonary artery pressure can increase secondary to an elevation of pulmonary venous pressure, which occurs due to postcapillary level obstruction or elevation of left atrial pressure. This can result from pulmonary vein stenosis or factors causing elevation of left atrial pressure, including anatomic lesions of left heart inflow or outflow obstruction, such as mitral valve stenosis, or impaired relaxation of the left ventricle (diastolic dysfunction) from cardiomyopathy. In these cases, the pulmonary artery wedge pressure is elevated at catheterization, and the PVR is typically low. However, longstanding pulmonary hypertension of this type can be associated with remodeling of the pulmonary vasculature causing an elevated PVR. This is thought to be due to endothelial injury from shear stress and the response of the arterial bed to persistent high pressure. These changes are often reversible with normalization of pressures by relief of the obstruction or improvement in left ventricular function.

Group 3 describes pulmonary hypertension associated with lung disease. In children, this is typically due to developmental lung disorders that cause a limitation or abnormality of the vascular bed. The most common associated disorders are bronchopulmonary dysplasia (BPD) and congenital diaphragmatic hernia (CDH). CDH is a rare defect in diaphragm development, occurring in about 1 in 2500 live births, in which the abdominal viscera herniate into the chest during fetal life. It is characterized by lung hypoplasia, a reduction of the cross-sectional area of the pulmonary vascular bed, abnormal vasoreactivity of the pulmonary arteries, and left heart diastolic dysfunction. These factors lead to elevations in PVR. Persistence of pulmonary hypertension is associated with higher mortality in infants with CDH. Management strategies include lung protective ventilation, including permissive hypercapnia and high-frequency oscillatory ventilation. Specific medical therapies are used if pulmonary hypertension persists despite adequate lung recruitment and without significant impairment in left ventricular performance. Extracorporeal membrane oxygenation (ECMO) is recommended for infants with severe disease who do not respond to other treatments.

In premature infants, BPD is defined by an oxygen requirement at day 28 of postnatal life or oxygen or ventilator support at 36 weeks of postgestational age. It is characterized by alveolar simplification and abnormal angiogenesis due to a disruption of normal lung development. Pulmonary hypertension predominantly develops in the very preterm and very-low-birth-weight infant and has been reported in approximately 25% of infants with moderate to severe BPD. The severity of pulmonary hypertension is highly variable and can resolve in some infants. However, persistence of severe pulmonary hypertension is associated with high mortality. Screening by echocardiography is recommended in infants requiring significant respiratory support or those with strong risk factors, including oligohydramnios, intrauterine growth restriction, and extreme prematurity. Evaluation and treatment of lung disease, such as, aspiration or structural airway disease, and assessing adequacy of respiratory support are recommended prior to starting pulmonary hypertension specific therapies.

While rare, developmental lung diseases such as alveolar capillary dysplasia or surfactant protein abnormalities, can also contribute to severe pulmonary hypertension early in infancy. Infants may present with presumed PPHN that does not resolve within 2 weeks of proper management and often show significant parenchymal lung disease on radiographic imaging. Alveolar capillary dysplasia is a rare, typically fatal disorder usually associated with misalignment of pulmonary veins and severely underdeveloped alveolar capillaries. A genetic association has been found with mutations in the FOXF1 gene. There is no cure, and lung transplantation is the only potential option for long-term survival. Surfactant protein abnormalities that include SPB and SPC deficiencies (ABCA3 and TTF1 disorders) have also been associated with pulmonary hypertension.

Other lung development anomalies known to be associated with pulmonary hypertension include pulmonary interstitial glycogenosis, a rare nonlethal form of ILD characterized by accumulation of glycogen in interstitial cells; pulmonary alveolar proteinosis, a rare disease associated with abnormal accumulation of surfactant within the alveoli; and pulmonary lymphangiectasia, another rare lung disorder characterized by lymphatic dilation in the subpleural, interlobar, perivascular, and peribronchial spaces.

Pulmonary hypertension may also be seen in children with chronic severe diffuse lung disease, including end-stage lung disease associated with cystic fibrosis or idiopathic pulmonary fibrosis. In addition, obstructive sleep apnea may contribute directly to pulmonary hypertension or be a factor of poor control. High-altitude pulmonary hypertension occurs in susceptible individuals due to pulmonary vasoconstriction as a response to decreased ambient oxygen tension with altitude, typically above 3000 m. Patients who are symptomatic are encouraged to move to low altitude as part of their treatment regimen.

Additional rare causes of pulmonary hypertension in children include chronic thromboembolic disease (Nice classification group 4), and systemic disorders including hemolytic anemia, sickle cell disease, and some metabolic disorders in which the mechanism of disease development remains unclear (Nice classification group 5). A high index of suspicion and early evaluation for pulmonary hypertension are reasonable for children who develop cardiorespiratory symptoms and have metabolic disorders, renal disease, or hemolytic anemia. Segmental pulmonary hypertension due to congenital conditions in which parts of the lung are supplied by aortopulmonary collaterals was also included in the most recent version of the classification.

Despite the well-defined classification system, pediatric patients often have multifactorial etiologies of pulmonary hypertension. For example, a patient may have pulmonary hypertension related to concomitant chronic lung disease of prematurity, congenital heart disease, and a genetic syndrome. This highlights the importance of a thorough investigation of all contributing factors when making a diagnosis of pulmonary hypertension.

CLINICAL MANIFESTATIONS

Symptoms of pulmonary hypertension in children are generally nonspecific, most commonly dyspnea on exertion, fatigue, and syncope. As a result, the diagnosis is often delayed as children undergo evaluation and treatment for more common childhood diseases, such as asthma, vasovagal syncope, and seizure disorders. Additional symptoms include chest pain, arrhythmia, poor growth, and those associated with right heart failure, such as facial edema or right upper quadrant pain due to hepatomegaly. Hemoptysis may occur in advanced disease. Presenting symptoms in infants include tachypnea, poor feeding, diaphoresis, failure to thrive, and lethargy.

Vital signs may reveal baseline tachypnea and tachycardia. Cyanosis may be present if there is coexisting lung disease or right-to-left shunting across an atrial or ventricular septal defect. Physical examination findings include a narrowly split or single second heart sound with a loud pulmonic component. There may be a holosystolic murmur of tricuspid valve regurgitation auscultated at the left lower sternal border or a diastolic murmur of pulmonary valve insufficiency. In advanced disease, there may be a palpable second heart sound or right ventricular heave, an S₃ gallop, facial or peripheral edema, hepatomegaly, and jugular venous distension.

DIAGNOSIS

The goals of diagnostic testing are to understand the etiology of pulmonary hypertension and assess disease severity. A standardized approach to diagnostic testing has been recommended to ensure all potential etiologies of pulmonary hypertension are evaluated, which includes echocardiography, electrocardiography, chest radiograph, blood tests, and cardiac catheterization. Pulmonary function tests, polysomnography, chest computed tomography (CT), and exercise testing may also be indicated. Genetic testing may be recommended in familial cases or when PVOD or PCH is suspected.

Many patients undergo chest x-ray and electrocardiogram (ECG) as part of initial screening tests, which can point to the presence of pulmonary hypertension. The chest radiograph may show enlargement of the main pulmonary artery and cardiomegaly due to enlargement of the right ventricle. There also may be a paucity of vascular markings seen in the peripheral lung fields. Pulmonary venous congestion may indicate PVOD/PCH or group 2 pulmonary hypertension. The ECG may show evidence of right ventricular hypertrophy and right atrial enlargement.

Echocardiography is a useful noninvasive screening tool to evaluate for congenital or acquired structural abnormalities and to assess the size and function of both ventricles. The right ventricular (pulmonary arterial) systolic pressure can be assessed from the tricuspid regurgitation jet velocity, the velocity across the PDA, and from the position of the interventricular septum, with flattening or bowing of the septum in systole correlating with elevated right ventricular systolic pressure. In addition, dilation, hypertrophy, and diminished function of the right ventricle may correlate with disease severity and suggest longstanding elevation of pressures.

Laboratory testing is helpful to evaluate for potential associated conditions, such as connective tissue diseases, infection, portal hypertension, hemolytic anemia, or thromboembolic disease. Brain-type natriuretic peptide (BNP), or its cleavage product N-terminal (NT) pro-BNP, has been shown to correlate with severity of pulmonary hypertension in children. Similarly, cardiopulmonary exercise testing and a 6-minute walk test have been shown to indicate disease severity and prognosis and are useful for monitoring patients over time.

Pulmonary function testing should be performed when possible at time of diagnosis. It is useful to rule out airway or lung parenchymal problems that may be primarily causing pulmonary hypertension or contributing to poor control/deterioration (air trapping, consolidations, or atelectasis). Bronchodilator testing should be considered if there is suggestion of obstructive lung disease. A restrictive lung pattern may prompt further evaluation for interstitial lung disease.

High-resolution chest CT with angiography is also recommended on initial assessment of children suspected of having pulmonary hypertension to exclude parenchymal lung disease, thromboembolic pulmonary hypertension, and other process that have a distinctive CT pattern (PVOD/PCH). It is also important to rule out anatomical obstructions like pulmonary vein stenosis or peripheral pulmonary stenosis that may alter treatment strategies.

Polysomnography should be considered in patients with a history suggestive of obstructive sleep apnea or other sleep disorder (daytime sleepiness, loud snoring) or with predisposition for a breathing disorder (midface abnormalities, trisomy 21, adenotonsillar hypertrophy). It should also be considered in patients with poor response to targeted pulmonary hypertension therapy as untreated sleep apnea could contribute to disease.

Ventilation/perfusion lung scan (V/Q scan) remains the gold standard for diagnosis of chronic thromboembolic pulmonary hypertension (CTEPH). The incidence of CTEPH is very low in pediatric patients, but testing should be considered in patients with idiopathic disease, especially if there are hematologic lab abnormalities or suggestion of thrombotic disease on CT scan.

Although noninvasive tests are clinically useful, cardiac catheterization remains the gold standard for diagnosis. Catheterization includes measurement of baseline hemodynamics, including pulmonary arterial pressure, wedge pressure and cardiac output, as well as an evaluation of the severity of systemic to pulmonary shunts, if present. The PVR is calculated from the transpulmonary pressure gradient divided by the pulmonary blood flow, Qp. The PVR and cardiac output measurements are typically indexed to body surface area. Often the ratio of PVR to systemic vascular resistance is used, with a ratio of > 0.3 considered abnormal. Vasodilator testing, typically with inhaled oxygen and nitric oxide, may be performed to assess vasodilator responsiveness, which may be useful for prognostic information and for considerations of repair of congenital heart disease and medical management. In patients with pulmonary arterial hypertension, wedge angiography may show abnormalities of capillary blush or more advanced findings of loss of distal vessels (“pruning”).

Lung biopsy is not performed routinely but is used if there is a suspicion of PVOD/PCH or a severe developmental disorder such as alveolar capillary dysplasia. These patients typically do not respond well to pulmonary vasodilators, and a definitive diagnosis prompts consideration for lung transplantation.

Predictors of severity of pulmonary hypertension in children include clinical evidence of right ventricular failure, growth failure, elevated BNP, and echocardiographic and hemodynamic indicators of severely elevated pulmonary artery pressure and depressed cardiac function (Table 486-3).

PREVENTION

There is no known method for preventing pulmonary hypertension other than to prevent or treat associated conditions. Even in patients with a predisposing genetic mutation who are considered at risk for developing pulmonary arterial hypertension, there are no data to support prophylactic use of medications to prevent onset of disease. In general, prevention focuses on treating reversible factors that may contribute to pulmonary hypertension, such as obstructive sleep apnea, and optimizing respiratory status. Because respiratory infections may have profound consequences in children with pulmonary hypertension, it is recommended that patients be fully immunized, including yearly influenza vaccinations and other routine vaccinations (Bordetella pertussis, respiratory syncytial virus, and pneumococcus) to prevent deterioration due to avoidable infections. Female adolescent patients should have counseling about safe contraception, as pregnancy has a high risk for severe morbidity and mortality for both mother and fetus (30–50%).

TREATMENT

Medical therapy has been revolutionized with the development of pulmonary vasodilators. Nonetheless, although their use has improved symptoms, quality of life, and survival significantly, they do not reverse the underlying disease process. None are approved for use in children by the US Food and Drug Administration (FDA) due to lack of pediatric-specific data, which makes medical management quite challenging.

Prior to the advent of pulmonary vasodilators, the treatment approach was limited to supportive care with diuretics to optimize fluid status and digoxin for right heart failure, although data to support use of digoxin are limited. Supplemental oxygen is frequently used at night, especially in the setting of nighttime hypoxemia.

In adults, calcium channel blockers were shown to be beneficial in those who responded to acute vasodilator testing with a significant drop in pulmonary pressures and were previously considered first-line therapy for this subset of patients. However, most children with pulmonary hypertension are not responsive to acute vasodilator testing and are unlikely to benefit from this therapy. Calcium channel blockers are contraindicated in infants and in patients with low cardiac output.

The use of pulmonary vasodilators in children with pulmonary hypertension is based on data from adult studies and guided by pediatric expert opinion. Over the past 20 years, at least 10 medications have been approved by the FDA for the treatment of pulmonary hypertension in adults. These drugs target 3 main pathways: the nitric oxide pathway, the endothelin pathway, and the prostacyclin pathway. The nitric oxide pathway promotes vasodilation via cyclic guanosine monophosphate–mediated pathways and includes inhaled nitric oxide, phosphodiesterase type 5 inhibitors (sildenafil and tadalafil), and soluble guanylate cyclase stimulators (riociguat). The endothelin receptor antagonists (bosentan, ambrisentan, macitentan) block pulmonary vasoconstriction by the endogenous vasoconstrictor endothelin. Prostacyclin analogs (epoprostenol, treprostinil, iloprost) mimic endogenous prostaglandin I₂ and are the most potent class of vasodilators. Intravenous, subcutaneous, and inhaled prostacyclins are widely used. Oral prostacyclins, such as treprostinil and the prostacyclin receptor analog selexipag, were recently approved for use in adults with limited pediatric experience to date. The generalized treatment approach is guided by risk stratification (Table 486-3), with a trend toward early combination therapy (Fig. 486-2). Because of the complex etiology of disease and evolving treatment algorithms, it is recommended that treatment of children be guided by a specialized pediatric pulmonary hypertension center.

Progression of disease leads to right ventricular failure, recurrent syncope, and death. In end-stage disease, surgical or transcatheter interventions to create a right-to-left shunt may be considered to decompress the right ventricle and increase systemic cardiac output. Interventions include balloon atrial septostomy, stenting of a PDA, if present, or surgical creation of a Potts shunt (connection between the left pulmonary artery and the descending aorta). Lung transplantation may be considered for patients with end-stage disease. Idiopathic pulmonary arterial hypertension is the second most common indication (9%) for lung transplantation according to International Society for Heart and Lung Transplantation Registry. The median survival after transplant is currently 4.9 years, highlighting the need for development of additional therapies to treat pulmonary hypertension.

SUGGESTED READINGS