Endomyocardial fibrosis

INTRODUCTION — Endomyocardial fibrosis (EMF) is a disease of rural poverty that is characterized by fibrosis of the apical endocardium of the right ventricle (RV), left ventricle (LV), or both. The clinical manifestations are largely related to the consequences of restrictive ventricular filling, including left and right sided heart failure (HF).

EMF refers to a specific syndrome with characteristic epidemiologic features. The epidemiology, pathophysiology, clinical manifestations, diagnosis, and treatment of EMF are reviewed here. Other cardiomyopathy syndromes with similar pathologies, including hypereosinophilia and/or fibrotic changes of the endocardium are addressed separately. (See "Carcinoid heart disease" and "Hypereosinophilic syndromes: Clinical manifestations, pathophysiology, and diagnosis", section on 'Cardiac disease' and "Treatment and prognosis of myocarditis in adults", section on 'Eosinophilic myocarditis'.)

EPIDEMIOLOGY — Although accurate epidemiologic data are lacking [1,2], EMF is estimated to be the most common form of restrictive cardiomyopathy worldwide.

Most studies of EMF have been conducted in tropical regions where there is a high prevalence of disease such as Uganda [3], Nigeria [4], Ivory Coast, south India [5], Venezuela [6], and Brazil (figure 1) [7]. Advanced disease has been described primarily in poor rural areas from several countries of eastern Africa (including Mozambique and Uganda). Endomyocardial fibrosis (EMF) was first recognized in Uganda during the 1940s and accounts for as much as 20 percent of cardiac cases sent for echocardiography in that country in contemporary series [1,2,8]. An echocardiographic screening study in Mozambique found a population prevalence of 20 percent; however, this study included patients with early, subclinical disease [9]. EMF also occurs in subtropical regions. Within endemic countries, there appears to be a regional variation [10,11].

EMF is primarily a disease of the young, occurring in children, adolescents, and young adults. In Uganda, a bimodal peak at ages 10 and 30 has been observed [12], and a similar pattern was found in Mozambique [9]. The differences between genders in the frequency of disease have been variable [9,12,13].

PATHOLOGY — In EMF, there is fibrosis of the right and/or left apical endocardial surfaces that leads to restrictive physiology [14]. Additionally, tethering of the atrioventricular (AV) valve papillary muscles and/or fibrosis of the respective ventricular inflow tracts leads to significant AV valve regurgitation. Gross pathology reveals ventricular endocardial thickening and fibrosis often with overlying thrombus. The atrium of the affected ventricle is often dramatically enlarged (image 1) with reduced ventricular volumes. Histopathology demonstrates increased type I collagen deposition, subendocardial infarction, fibrosis, and thrombus [11,15]. In many pathology studies, there is a lack of inflammation or eosinophilia, implying that at end-stage disease, the inflammatory process is inactive [16].

PATHOPHYSIOLOGY — The cause of the underlying fibrotic process of endomyocardial fibrosis (EMF) is largely unknown; however, several theories exist and are briefly reviewed.

Eosinophilia — Eosinophilia is the most commonly cited etiologic link in EMF. In support of the eosinophilia theory is the observation that EMF resembles a late stage of Loeffler's endocarditis (eosinophilic myocarditis) (picture 1), a process known to result from sustained eosinophilia in patients with hypereosinophilic syndrome [17,18]. EMF and intraventricular thrombosis have also been observed following a variety of other eosinophilic syndromes including hypersensitivity myocarditis [19], parasitic infections [17,18], eosinophilic leukemia, sarcoma, carcinoma, and lymphoma [20], GM-CSF administration [21], and prolonged drug-induced eosinophilia. (See "Hypereosinophilic syndromes: Clinical manifestations, pathophysiology, and diagnosis", section on 'Cardiac disease' and "Treatment and prognosis of myocarditis in adults", section on 'Eosinophilic myocarditis'.)

Despite the similarities between Loeffler’s endocarditis and EMF, serum and myocardial eosinophilia have not been consistently demonstrated in EMF. Although one study from Uganda found that 60 percent of patients with EMF had at least mild eosinophilia at the time of diagnosis compared with 10 percent of controls [22], in Kerala, India, most with EMF did not have active eosinophilia at the time of diagnosis [23]. Endomyocardial biopsies have not demonstrated eosinophilia in EMF, even in those suspected to have early disease [24]. It is possible that many with EMF have had significant eosinophilia at one time that is not detected by the time of presentation to medical care.

Infectious — Several infections have been implicated in the pathophysiology of EMF, including toxoplasmosis [25], rheumatic fever [26], malaria [27], and helminthic parasites [4,28]. A consistent association with one organism, however, has not been demonstrated. For example, mice infected with plasmodium berghei develop EMF lesions [29], but a study of a series of 47 African children aged 5 to 15 years old with severe and complicated plasmodium falciparum infection produced insufficient evidence to link these two diseases [30]. Also opposing the infectious hypothesis is the observation that there are many tropical countries with similar burdens of malaria and filariasis as Uganda and Nigeria that do not have reported cases of EMF. (See "Diagnostic testing for toxoplasmosis infection" and "Acute rheumatic fever: Clinical manifestations and diagnosis" and "Malaria: Clinical manifestations and diagnosis in nonpregnant adults and children" and "Laboratory tools for diagnosis of malaria" and "Lymphatic filariasis: Epidemiology, clinical manifestations, and diagnosis".)

Environmental exposure — Given the regional differences in the frequency of EMF cases and the lack of unifying infectious etiology, the idea of a geochemical basis for EMF has gained popularity. Alimentation and toxicity have been proposed as causes of EMF [2]; proposed factors include malnutrition and a protein-poor diet (eg, leucine, valine, and tryptophan deficiency), magnesium deficiency, cassava toxicity, high vitamin D, serotonin, cerium and thorium, and traditional medicine plants.

Cassava is consumed extensively in some tropical African and Latin American countries. It contains linamarin, which may liberate cyanide in the gut. A low-protein diet diminishes the detoxification of cyanide, which may affect myocardial cell function. In animal models, cassava intake may induce interstitial fibrosis, intracellular vacuoles, and endocardial thickening in the absence of parasites or eosinophilia [31,32]. The relationship between cassava and EMF warrants further investigation [33].

Immunologic — The presence of anti-myosin autoantibodies has been demonstrated in EMF; however, these antibodies can be detected in other forms of heart disease such as Dressler’s syndrome, rheumatic heart disease, and in patients with post-transplant rejection [34].

Genetic — A familial link has been identified in many studies; however, it is not known whether this is due to an environmental or genetic cause or both [9,35].

CLINICAL PHASES AND MANIFESTATIONS — Three phases have been identified based upon clinical and pathologic manifestations: an initial phase of active recurrent inflammation, followed by a transient progressive form, and an advanced chronic fibrotic phase with restrictive cardiomyopathy [2].

●The initial form appears as a febrile illness, with pancarditis, eosinophilia, dyspnea, and periorbital swelling [2-4]. There is no specific marker for EMF. Eosinophilia may be present, but its frequency is variable among reported series (ranging from 0 to 70 percent), perhaps in part due to delayed diagnosis of early EMF. The ECG is nonspecific, and echocardiography may show wall thickening (with or without echogenic endomyocardial infiltrates), reflecting inflammation and edema and pericardial effusion. Pathologic examination reveals myocardial interstitial edema, eosinophilic infiltration, subendocardial necrosis, and vasculitis.

●The transient progressive form is characterized by chronic HF with edema, ascites, and cardiomegaly. Echocardiographic wall thickness returns to normal, but affected walls may show scars. Signs of restrictive physiology with chamber dilation (especially of the atria) may be seen.

●The advanced form is characterized by chronic HF with ascites, edema, and cardiomegaly accompanied by typical echocardiographic findings, including apical fibrosis and restrictive ventricular filling pattern (see 'Echocardiography' below). The majority of patients have biventricular involvement, followed by the isolated right-sided form, and rarely the isolated left heart form.

The clinical manifestations of progressive and advanced EMF are largely related to the presence of right and/or left HF. Patients may report dyspnea on exertion, paroxysmal nocturnal dyspnea, orthopnea, lower extremity, and abdominal swelling. A history of a febrile illness with or without urticaria is occasionally obtained [36].

In Africans with EMF in particular, there is often dramatic ascites, which may or may not be accompanied by other signs of right-sided HF, such as elevated jugular venous pressure and/or lower extremity edema [37-40]. The high prevalence of malnutrition and hypoalbuminemia may explain the predilection for ascites in this population. In one series, the ascitic fluid protein content was described as exudative in 35 of 47 patients with EMF, and the authors concluded that an inflammatory process might also be contributing [38]. The serum albumin to ascites gradient (which is a more accurate indicator of portal hypertension such as my result from HF) was not reported. (See "Evaluation of adults with ascites", section on 'Determining the cause of the ascites'.)

Patients with EMF also often have large pleural and pericardial effusions. Severe atrial enlargement leads to the appearance of cardiomegaly on chest radiography. Atrial fibrillation is common in end-stage disease and predicts a poor prognosis [41].

DIAGNOSIS — The diagnosis of EMF should be suspected in individuals in or from endemic areas with symptoms of HF (particularly ascites) and suggestive echocardiographic findings.

In high-resource as well as endemic countries, evaluation for cardiac involvement (including myocarditis or EMF) is indicated in patients with hypereosinophilia or abnormal eosinophils. Work-up includes obtaining a serum troponin and an echocardiogram. (See "Hypereosinophilic syndromes: Clinical manifestations, pathophysiology, and diagnosis".)

The diagnosis of EMF is made by summing points using the following severity scoring system with major and minor criteria [9]. Agreement by two independent experienced cardiologists is desirable. Definite EMF diagnosis includes two major criteria, or one major criterion with two minor criteria. Scores may range from 0 to 35; patients with eight or less are classified as mild, those with 8 to 15 as moderate, and those with 15 or more as severe. EMF is also classified as biventricular, right-sided, or left-sided ventricular involvement, according to clinical presentation and imaging findings.

●Major criteria

•Endomyocardial plaques >2 mm in thickness – Two points

•Thin (≤1 mm) endomyocardial patches affecting more than one ventricular wall – Three points

•Obliteration of the right or left ventricular apex – Four points

•Thrombi or spontaneous contrast without severe ventricular dysfunction – Four points

•Retraction of the right ventricular apex (right ventricular apical notch) – Four points

•Atrioventricular-valve dysfunction due to adhesion of the valvular apparatus to the ventricular wall – One to four points (more points assigned for more severe atrioventricular regurgitation)

●Minor criteria

•Thin endomyocardial patches localized to one ventricular wall – One point

•Restrictive flow pattern across mitral or tricuspid valves – Two points

•Pulmonary-valve diastolic opening – Two points

•Diffuse thickening of the anterior mitral leaflet – One point

•Enlarged atrium with normal-sized ventricle – Two points

•M-movement of the interventricular septum and flat posterior wall – One point (see 'Echocardiography' below)

•Enhanced density of the moderator or other intraventricular bands – One point

Cardiac imaging findings are nearly identical for EMF seen in the tropics and non-tropical EMF.  

Echocardiography — Although clinically similar to Loffler’s endocarditis, the diagnosis of endomyocardial fibrosis (EMF) should be reserved for patients from endemic regions without a clearly identified cause for sustained eosinophilia with the classic echocardiography features listed below [6,42,43]:

●Apical fibrosis of the right ventricle (RV), left ventricle (LV), or both ventricles (image 2 and figure 2).

●Tethering the atrioventricular (AV) valve papillary muscles, leading to mitral and/or tricuspid regurgitation (image 1).

●Giant atrial enlargement (image 1).

●Restrictive filling pattern on Doppler recordings of mitral valve inflow (waveform 1).

In addition, apical thrombi are often present (image 3). When available, administration of microbubble echocardiographic contrast agent should be considered to identify and define the extent of apical thrombus (see "Contrast echocardiography: Clinical applications"). In EMF, the apex maintains inward systolic contractile motion that may help to differentiate EMF from other causes of apical thrombi associated with an akinetic or dyskinetic apex such as myocardial infarction or Chagas disease [6] (figure 2).

On M-mode echocardiography, a characteristic interventricular septal M-movement or pattern may be identified; this septal bounce results from rapid anterior movement in early diastole, which creates a characteristic M-movement or pattern [44].

An echocardiographic screening study in Mozambique included echocardiographic criteria for the diagnosis and staging of EMF [9]. As the natural history of EMF is not well defined, these criteria will likely aid in defining the stages of this disease and in determining the clinical significance of early EMF.

Cardiac catheterization — Cardiac catheterization is not required for the diagnosis of EMF. However, on hemodynamic studies, a restrictive pattern is observed with diastolic dip and plateau pressure tracings. Depending on the ventricle involved, mitral and tricuspid regurgitation may be demonstrated. Ventricular angiography reveals apical obliteration of the affected ventricle (image 2) [45].

Cardiovascular magnetic resonance imaging — Cardiovascular magnetic resonance (CMR) imaging with contrast demonstrates myocardial fibrosis (image 3) [46,47]. In a series of 36 patients with EMF, patients with fibrous tissue deposition identified by late gadolinium enhancement cardiac magnetic resonance imaging indexed to body surface area (FT/BSA) of >19 mL/m², had increased mortality rate, worse New York Heart Association functional class, and increased likelihood of surgery [48]. However, the utility of this information in addition to echocardiography is unclear, and this modality is generally unavailable in areas with the highest burden of disease. In early disease where there is suspicion for active inflammation, CMR may be useful in identifying patients who may benefit from steroid therapy. (See "Clinical utility of cardiovascular magnetic resonance imaging", section on 'Late gadolinium enhancement'.)

DIFFERENTIAL DIAGNOSIS — At the early active inflammation stage, differentiation with other acute febrile illness is difficult, except by the presence of hypereosinophilia. Other causes of acute myocarditis should be excluded. In the chronic fibrotic stage various imaging modalities, as mentioned before, are almost identical between Loeffler's and Davies's diseases. In the presence of ascites, other causes of liver disease or portal hypertension should be ruled out. As a restrictive cardiomyopathy, EMF should be differentiated from other cardiomyopathies as sarcoidosis, amyloidosis, infiltrative process, and of unknown origin.

PROGNOSIS — The natural history of EMF is not fully defined. Most present to medical care with end-stage disease and suffer an annual mortality as high as 25 percent despite medical treatment [41,49].

MANAGEMENT — Limited data are available to guide management. Surgical management has led to long-term survival in some patients with EMF [7]; however, this option is unavailable in regions with a high disease burden.

Medical therapy — Available literature is limited to case series that do not fully define treatment regimens [41,49]. A general approach can be extrapolated from the treatment of patients with HF due to restrictive diastolic dysfunction from other causes. (See "Treatment and prognosis of heart failure with preserved ejection fraction" and "Definition and classification of the cardiomyopathies", section on 'Restrictive cardiomyopathy'.)

Diuretics and rate control for atrial fibrillation are currently the mainstays of therapy.

Pleural, pericardial, or ascitic fluid removal may alleviate symptoms, but these often reaccumulate. In patients with suspected acute carditis, prednisone may be of benefit. (See "Treatment and prognosis of myocarditis in adults", section on 'Eosinophilic myocarditis'.)

Standard recommendations for anticoagulation apply to patients with atrial fibrillation (see "Atrial fibrillation in adults: Use of oral anticoagulants"). However, appropriate use and management of anticoagulants are generally difficult in EMF endemic regions.

Surgery — Endomyocardial resection with valve replacement or repair of the mitral valve has gained prominence at many centers, especially in subjects in advanced HF [7,48,50,51]. Tricuspid valve replacement has had poor results. Immediate postoperative mortality is high, ranging from 15 to 30 percent, but surgery offers the possibility of long-term survival [7,51]. Published series have been small, controlled data are lacking, and overall experience is limited. Questions remain about the appropriate timing, perioperative mortality, and long-term prognosis (image 1 and image 4). Cardiac surgery is not routinely available in areas with high EMF prevalence.

A surgical series of 83 patients with EMF from Brazil with mean age 31 years (range 4 to 59) and New York Heart Association functional class III to IV symptoms had a survival probability of 55 percent at 17 years [7]. There were 15 late deaths (six unrelated to EMF) during mean 7.7-year follow-up. Four patients were reoperated on for recurrent fibrosis, five patients were reoperated on for native or prosthetic valve disease, and in six patients, EMF appeared in the other ventricle.

SUMMARY

●Endomyocardial fibrosis (EMF) is a restrictive cardiomyopathy observed in the tropics usually at the end-stage of the disease. It may be indistinguishable from Loeffler’s endocarditis, observed in temperate climates.

●The pathogenesis remains unknown; however, eosinophilia may play a role.

●Echocardiography may show uni- or bilateral ventricular apex obliteration with severely dilated atria and a restrictive filling pattern.

●The prognosis is poor, with a mortality estimated at 25 percent per year.

●Limited data suggest that surgical treatment with endomyocardial resection and valve replacement may be beneficial in patients with advanced apical obliteration and severe heart failure symptoms; however, this option is largely unavailable where the disease is present.