Endomyocardial Fibrosis (EMF): Comprehensive Disease Research Report

1. Disease Information

Overview. Endomyocardial fibrosis (EMF) is a rare, insidious restrictive cardiomyopathy characterized by dense fibrous scarring of the ventricular endocardium — predominantly the inflow tracts and apices of the right and/or left ventricle — that obliterates ventricular cavity volume, tethers the atrioventricular (AV) valve subvalvular apparatus (papillary muscles and chordae tendineae) causing regurgitation, and produces severe diastolic dysfunction with markedly preserved systolic function. It is the most common cause of restrictive cardiomyopathy worldwide and is endemic to poor tropical and subtropical regions within roughly 15° of the equator (StatPearls, NBK513293; PMC4239813).

"Subendocardial fibrosis of the apices and inflow tracts of the right ventricle, left ventricle, or both defines the disease... This restrictive scarring prevents ventricular filling, and tethering of the papillary muscles leads to valvular regurgitation." — Bukhman, Ziegler & Parry, PLoS Negl Trop Dis 2008 (PMID: 18301727)

Key identifiers:

Synonyms: Davies' disease/Davies disease, tropical endomyocardial fibrosis, African endomyocardial fibrosis, endomyocardial sclerosis, obscure African cardiomyopathy, eosinophilic endomyocardial disease (when linked to Löffler/hypereosinophilic pathophysiology) (GARD; StatPearls NBK513293).

Evidence base: Information is derived almost entirely from aggregated disease-level resources — hospital case series, autopsy series, and a small number of population-based echocardiographic screening studies (notably the 2008 Mozambique study) — rather than large-scale EHR/biobank data, reflecting both the rarity of formal cohorts and the resource-limited settings where EMF is endemic.

Sources: - Endomyocardial Fibrosis - StatPearls - Endomyocardial Fibrosis: Still a Mystery after 60 Years - PLOS NTD (PMID: 18301727) - Endomyocardial fibrosis: A form of endemic restrictive cardiomyopathy - PMC - Orphanet: Tropical endomyocardial fibrosis - GARD - Endomyocardial fibrosis - OMIM 226000 - Endocardial Fibroelastosis

2. Etiology

EMF has no single confirmed cause; the current model is multifactorial, requiring convergence of infectious/immune stimuli, malnutrition, environmental exposures, and host genetic susceptibility in the context of poverty.

"No single proposed factor can explain the occurrence of EMF worldwide." — cdt.amegroups.org review (Cardiovasc Diagn Ther)

Disease causal factors — historically proposed hypotheses

1. Helminth/parasite–eosinophilia hypothesis: Filariasis, schistosomiasis, and other chronic helminthic infections drive sustained hypereosinophilia; eosinophil granule protein-mediated cardiotoxicity (see Mechanism section) is the leading mechanistic model. Ive & Brockington (Nigeria) found filariasis in ~100% of 42 angiographic EMF cases vs. 44% of 115 controls (cited in PMID: 18301727), though the hypothesis fails to explain the absence of EMF in other high-helminth-burden regions (e.g., parts of Southeast Asia).
2. Löffler endocarditis / hypereosinophilic syndrome (HES) equivalence: Histological and echocardiographic comparisons (Brockington & Olsen 1975; Davies 1983) found the fibrotic end-stage of Löffler endocarditis indistinguishable from EMF, suggesting a shared final common pathway of eosinophil-mediated endomyocardial injury regardless of the trigger for eosinophilia (idiopathic, parasitic, or clonal/neoplastic).
3. Nutritional/toxic hypotheses: Cassava-based diets combined with severe protein deprivation were causally tested — feeding uncooked cassava to Cercopithecus aethiops (African green monkeys) produced EMF-like cardiac lesions (vs. no lesions on a banana-diet control), supporting a cassava/protein-deficiency mechanism (linked to cyanogenic glycoside/cerium toxicity). The competing "serotonin–plantain" hypothesis was tested by feeding plantains to guinea pigs, rats, and Patas monkeys but failed to reproduce EMF lesions and was abandoned by the 1970s (PMID: 18301727).
4. Geochemical/toxin exposure: Cerium and thorium in monazite-rich soils (e.g., coastal Kerala, India) have been speculatively linked to regional clustering, without confirmatory studies.
5. Autoimmunity: Elevated immunoglobulins and circulating anti-myosin antibodies (against actin, tropomyosin, and HSP-70) found in 53.6% of EMF patients vs. 10% of controls (see Genetic/Molecular section) support an autoimmune amplification loop, possibly triggered by molecular mimicry after infection (PMID: 20422043).
6. Malaria/immune dysregulation: Migration-associated changes in anti-malarial and anti-heart antibody titers were noted among Rwanda–Burundi migrant populations developing EMF in Uganda, though Plasmodium species distribution does not match EMF geography.

Risk factors

Genetic risk factors: - Familial clustering and ethnic-group concentration strongly suggest heritable susceptibility (PMID: 757895, familial EMF in Zambia). - The only formal genetic association study to date found HLA-B*58 associated with EMF in Mozambique (p=0.03) and HLA-A*02:02 in Uganda (p=0.005) (Beaton et al., Glob Cardiol Sci Pract 2014, PMID: 25780800). No genome-wide association study has yet been performed or validated these findings. - In the eosinophilic (Löffler-variant) end of the spectrum, the FIP1L1-PDGFRA fusion gene (constitutively active tyrosine kinase from an interstitial 4q12 deletion) drives clonal hypereosinophilia with cardiac (Löffler endocarditis/EMF-pattern) involvement in a subset of chronic eosinophilic leukemia patients — a somatic, acquired lesion rather than germline (PMC12082641, PMC10484160, PMC10217393).

Environmental risk factors: - Extreme poverty, rural residence, subsistence farming, going barefoot, and cassava-based diets (Uganda case-control data cited in PMC12701864/PMID: 41399600 and PMID: 18301727). - Chronic helminthic (filarial, schistosomal) and malarial infection burden. - Magnesium and protein-calorie malnutrition. - Geography: equatorial low-lying humid tropical zones — coastal Tanzania/Mozambique, southern Nigeria, Uganda, Kerala (India), Guangxi Province (China), Bahia/Colombia (South America).

Protective factors: No specific genetic or environmental protective factors have been formally identified in the literature; declining incidence in some hospital series has been attributed non-specifically to "improving healthcare and living standards" (PMC4239813) — i.e., socioeconomic/nutritional/parasite-control improvement rather than a defined protective exposure or allele.

Gene–environment interactions: The prevailing model is that HLA-conferred immune-response variability modulates the intensity/character of the host response (Th2-skewed, eosinophil/mast-cell-driven inflammation) to a chronic antigenic trigger (helminth, malarial, or nutritional/toxic) that is itself environmentally determined by poverty and geography — i.e., genetic susceptibility determines who among the exposed develops fibrotic disease (PMID: 25780800).

Sources: - Endomyocardial Fibrosis: Still a Mystery after 60 Years (PMID: 18301727) - Genetic susceptibility to endomyocardial fibrosis (PMID: 25780800) - Endomyocardial fibrosis: familial and other cases from northern Zambia (PMID: 757895) - A Narrative Review on Endomyocardial Fibrosis (PMC12701864 / PMID: 41399600) - Loeffler endocarditis revealing chronic eosinophilic leukaemia with FIP1L1-PDGFRA rearrangement (PMC12082641)

3. Phenotypes

EMF phenotypes span cardiac structural/functional signs, systemic congestive symptoms, hematologic/laboratory abnormalities, and constitutional findings from chronic malnutrition. Frequencies below are drawn from hospital case series (Iroegbu et al., Cardiovasc Diagn Ther; Mozambique population study PMID: 18596273) and should be treated as approximate/series-specific.

Cardiac signs and symptoms

Systemic/constitutional signs

• Exophthalmos, central cyanosis, lip/gum hyperpigmentation (distinctive but non-specific findings reported in African case series)
• Cachexia, growth stunting, malnutrition (chronic disease)
• Clubbing
• Testicular atrophy and sexual dysfunction in males (advanced chronic disease)

Laboratory abnormalities

• Eosinophilia during the acute/inflammatory phase (variable; not present once fibrotic stage is reached)
• Hypoalbuminemia in chronic phase
• Elevated NT-proBNP/BNP and high-sensitivity troponin (disease-progression/prognostic markers, not diagnostic)
• Elevated plasma cytokines (see Mechanism section): TNF-α, IL-4, IL-10
• Circulating anti-myosin (anti-actin, anti-tropomyosin, anti-HSP70) IgG/IgM autoantibodies in a disease-activity-correlated subset (PMID: 20422043)

Phenotype characteristics

• Age of onset: Bimodal — childhood peak (~first decade of life; "more than half of reported EMF cases originating in sub-Saharan Africa," per PMC12701864) and a secondary adult peak in women of childbearing age.
• Severity/progression: Ranges from asymptomatic/subclinical (in the Mozambique population screen, only 22.7% of 211 EMF-positive subjects were symptomatic, PMID: 18596273) to end-stage NYHA class III/IV heart failure. In symptomatic hospital cohorts, 62–98% present in NYHA class III/IV (cdt.amegroups.org review; Iroegbu et al.).
• Course: Chronic and progressive once fibrotic; historically an acute febrile/eosinophilic myocarditic phase (Davies stage 1, up to ~5 months) precedes a subacute thrombotic stage (Davies stage 2, starting ~10 months) and finally the irreversible fibrotic stage (Davies stage 3, over years).
• Ventricular distribution: Biventricular ~50–55%, isolated LV ~28–40%, isolated RV ~10–28% (varies by series; NEJM Mozambique study: biventricular 55.5%, right-sided 28.0%, presumably left-sided the remainder) (PMID: 18596273).

Quality of life impact

Formal disease-specific quality-of-life instruments (EQ-5D, SF-36) have not been reported for EMF specifically. Functional impact is inferred from NYHA class distributions — the majority of clinically ascertained (hospital-based) patients present in NYHA III/IV, i.e., marked-to-severe limitation of ordinary activity — and from the socioeconomic/constitutional burden (growth failure, cachexia, sexual dysfunction) documented in pediatric and adult case series.

*HPO term suggestions are provisional and should be verified against the ontology (label match) before curation, per standard practice.

Sources: - A population study of endomyocardial fibrosis in a rural area of Mozambique (PMID: 18596273) - Endomyocardial fibrosis - Iroegbu (Cardiovasc Diagn Ther) - Endomyocardial Fibrosis - StatPearls

4. Genetic/Molecular Information

Causal genes: EMF (the idiopathic/tropical form) is not a single-gene Mendelian disorder — no causal gene has been established, and there is no dedicated OMIM phenotype entry for it (distinguishing it from OMIM:226000 endocardial fibroelastosis, a different, largely infantile/congenital disease with heterogeneous — sometimes monogenic ciliopathy-related — causes).

Associated genetic/genomic findings: - HLA-B*58 (Mozambique) and HLA-A*02:02 (Uganda) — population-specific susceptibility alleles, PMID: 25780800. These require replication and are not diagnostic markers. - FIP1L1-PDGFRA fusion (interstitial 4q12 deletion producing a constitutively active PDGFRA tyrosine kinase) — a somatic driver of clonal hypereosinophilic syndrome/chronic eosinophilic leukemia, a recognized cause of the Löffler-endocarditis/EMF phenotype in a subset of patients, and clinically actionable because these patients respond to imatinib (tyrosine kinase inhibitor) (PMC12082641, PMC10484160). This is a somatic, not germline, genetic lesion, relevant to a specific EMF-associated etiologic subset rather than tropical/idiopathic EMF as a whole. - No pathogenic germline variant, chromosomal abnormality, or copy-number variant has been established as causal for classic tropical/nutritional EMF.

Autoantibody/molecular immune findings: - IgG antibodies against myocardial proteins of 35 kDa (actin), 42 kDa (tropomyosin), and 70 kDa (HSP-70) detected in 53.6% of 56 Mozambican EMF patients vs. 10% of 10 controls (p<0.05); IgM antibodies in 19.6% vs. 0%. Antibody reactivity correlated with disease activity (mean 19.6±3.7 antibodies in active disease vs. 7.1±3.3 in remission) (PMID: 20422043).

Cytokine/molecular profiling (plasma, n=27 EMF patients vs. 38 controls, Bossa et al. 2014, PLoS ONE, DOI: 10.1371/journal.pone.0108984, PMCID: PMC4193862): | Cytokine | EMF patients | Controls | p-value | % positive in EMF | |---|---|---|---|---| | TNF-α | 2.77 ± 4.64 pg/mL | 0.94 ± 0.24 pg/mL | 0.006 | 77.7% | | IL-4 | 4.51 ± 7.79 pg/mL | 1.22 ± 0.87 pg/mL | 0.001 | 88.8% | | IL-10 | 4.11 ± 5.27 pg/mL | 0.99 ± 0.89 pg/mL | 0.0001 | 92.6% | | IL-6, IFN-γ, IL-2 | Not significantly different | — | — | — |

Interpretation: "a mixed pro- and anti-inflammatory/Th2 circulating cytokine profile" consistent with a persistent inflammatory stimulus with compensatory anti-inflammatory (Th2/IL-10) upregulation, possibly residual from prior helminthic infection.

Epigenetic information: No EMF-specific DNA methylation, histone modification, or chromatin studies were identified in the literature search — this is an open gap.

Functional consequences: Because no causal germline gene/variant is established, LOSS_OF_FUNCTION/GAIN_OF_FUNCTION functional-impact categorization does not apply to a variant in the way it would for a monogenic disease; the FIP1L1-PDGFRA subset is the clearest example of a gain-of-function somatic lesion (constitutive kinase activation) driving eosinophil-mediated cardiotoxicity in a specific EMF-associated etiology.

Suggested HGNC/gene annotations (for the eosinophilic/Löffler-variant subtype only): PDGFRA (hgnc:8803), FIP1L1 (hgnc:26845) — verify via HGNC before curation.

Sources: - Genetic susceptibility to endomyocardial fibrosis (PMID: 25780800) - Presence of Circulating Anti-Myosin Antibodies in Endomyocardial Fibrosis (PMID: 20422043) - Plasma Cytokine Profile in Tropical Endomyocardial Fibrosis (PMC4193862) - Loeffler endocarditis revealing chronic eosinophilic leukaemia with FIP1L1-PDGFRA rearrangement (PMC12082641)

5. Environmental Information

Environmental factors: - Cassava (manioc) consumption combined with severe protein deprivation — experimentally reproduced EMF-like cardiac lesions in African green monkeys (Cercopithecus aethiops) fed uncooked cassava vs. banana-fed controls. - Cerium/thorium exposure from monazite-rich soils (speculative, regional correlation only, e.g., coastal Kerala). - Magnesium deficiency.

Lifestyle/socioeconomic factors: - Extreme poverty; subsistence farming; going barefoot (a marker of poverty and of soil-transmitted helminth exposure) — Ugandan case-control study found associations between EMF and "markers of poverty such as farming, lack of shoes, and cassava-based diets" (PMID: 18301727). - Rural residence in low-lying, humid, equatorial regions. - Chronic malnutrition/protein-calorie deficiency.

Infectious agents implicated (none proven definitively causal): - Helminths: filaria, Schistosoma spp. - Plasmodium spp. (malaria) — via immune dysregulation/antibody cross-reactivity hypothesis rather than direct cardiac invasion. - Coxsackievirus (proposed as a possible triggering acute myocarditic insult in some hypotheses).

Suggested ECTO exposure terms (to verify via OAK before curation): exposure to cassava/cyanogenic glycosides, exposure to helminth antigens, exposure to Plasmodium falciparum antigens, dietary protein deficiency exposure.

Sources: As cited in Sections 2 and 3 above (PMID: 18301727; PMC12701864/PMID: 41399600).

6. Mechanism / Pathophysiology

Causal chain overview (from trigger to clinical manifestation)

Chronic antigenic/toxic stimulus (helminth infection, malaria, cassava/protein malnutrition)
   → (in genetically susceptible hosts, e.g., HLA-B*58/HLA-A*02:02)
Sustained eosinophilia / Th2-skewed immune activation (elevated IL-4, IL-10, TNF-α)
   → Eosinophil degranulation in endocardium: release of eosinophil cationic protein (ECP),
     major basic protein (MBP), eosinophil-derived neurotoxin, reactive oxygen species
   → Endothelial and myocyte injury (necrosis) — Davies Stage 1: acute eosinophilic
     (necrotic) myocarditis/endocarditis (up to ~5 months)
   → ECP-mediated activation of coagulation factors + MBP-mediated platelet activation
     → Mural thrombus formation at ventricular apex and beneath posterior mitral leaflet
       — Davies Stage 2: thrombotic stage (from ~10 months, over several years)
   → Organization of thrombus + fibroblast activation/excess collagen and ECM deposition
     → Dense acellular fibrocollagenous endocardial scar — Davies Stage 3: fibrotic
       (healed) stage
   → Endocardial fibrosis obliterates ventricular apex/inflow tract, tethers papillary
     muscles/chordae to the ventricular wall
   → Restrictive diastolic physiology + AV valve regurgitation (mitral and/or tricuspid)
   → Atrial dilation (compensatory) → atrial fibrillation, further thromboembolic risk
   → Congestive heart failure, pulmonary hypertension (left-sided disease),
     systemic venous hypertension/ascites/hepatomegaly (right-sided disease)

Autoimmune amplification (anti-myosin/actin/tropomyosin/HSP-70 antibodies) may perpetuate myocardial injury independent of ongoing eosinophilic infiltration, particularly in chronic/relapsing disease.

Molecular pathways

• Th2/eosinophil-driven inflammatory pathway: IL-4/IL-10-skewed cytokine milieu with TNF-α co-elevation (PMC4193862).
• PDGFRA/tyrosine kinase signaling: constitutively activated in the FIP1L1-PDGFRA somatic-fusion subset, driving eosinophil clonal proliferation (relevant to Löffler-variant/EMF overlap).
• Coagulation cascade activation: eosinophil cationic protein directly activates coagulation factors, and MBP stimulates platelet activation — a distinctive eosinophil-to-thrombosis mechanistic link (JIR review, DOI: 10.2147/JIR.S458692, PMCID: PMC10984210).
• Fibrotic/ECM pathway: excessive fibroblast activation and collagen/extracellular-matrix deposition in the endocardial subendothelial layer (mechanistically convergent with the dismech fibrotic_response module pattern: tissue injury → inflammation → mesenchymal/fibroblast activation → excessive ECM → organ dysfunction).

Cellular processes

• Eosinophil degranulation and cytotoxicity
• Endothelial injury
• Myocyte necrosis (subendocardial)
• Platelet activation and thrombus organization
• Fibroblast activation and excessive collagen synthesis (myofibroblast-like phenotype implied but not explicitly characterized at single-cell resolution in the literature reviewed)
• Chronic lymphocytic/mononuclear inflammatory infiltration (persists into the fibrotic stage in some series)
• Neovascularization within subendocardial fibrotic tissue

Protein dysfunction / biochemical abnormalities

• Eosinophil cationic protein (ECP) and major basic protein (MBP) act as direct mediators of endothelial and myocardial cytotoxicity and of pathological coagulation activation.
• No structural protein misfolding/aggregation mechanism (distinguishing EMF from, e.g., amyloidosis, another restrictive-cardiomyopathy differential).

Immune system involvement

Central and defining: eosinophil-mediated tissue injury (whether from reactive/secondary eosinophilia due to parasitic infection, idiopathic hypereosinophilia, or clonal/neoplastic hypereosinophilic syndrome with FIP1L1-PDGFRA), compounded by autoimmune anti-myocardial antibody production in a subset.

Tissue damage mechanisms

Eosinophil-granule-protein cytotoxicity → necrosis → thrombosis → fibrotic scarring (a distinctive three-stage, immune-cell-initiated fibrogenesis mechanism, mechanistically related to — but histologically and etiologically distinct from — classic tissue-injury-driven fibrotic_response chains seen in organ fibrosis elsewhere in the KB).

Molecular profiling

• Transcriptomics/proteomics/metabolomics/lipidomics/single-cell/spatial data: No dedicated omics datasets for human EMF cardiac tissue were identified in this search (a notable knowledge gap — EMF is markedly under-studied by modern molecular methods relative to its disease burden, largely due to being endemic in resource-limited settings without omics infrastructure).
• Histopathology (traditional, most detailed available "molecular profiling" surrogate):
• Gross: atrial dilation, apical mural thrombus, reduced ventricular cavity size, AV annular dilation.
• Microscopic: dense acellular fibrocollagenous endocardial thickening; lymphocyte-predominant infiltrate; minimal myocardial (as opposed to endocardial) tissue loss; subendocardial neovascularization; coronary vessel changes (medial sclerosis, intimal proliferation, plexiform lesions) reported in one pediatric surgical/histopathology series (mean endocardial thickness 3,000 ± 1,519 µm, maximum 5,591 µm) (cdt.amegroups.org).

Suggested ontology terms (verify before curation)

• GO (biological process): GO:0030198 extracellular matrix organization; GO:0006954 inflammatory response; GO:0043534 blood vessel endothelial cell migration (angiogenesis-adjacent); collagen biosynthesis/fibril organization terms.
• CL (cell type): CL:0000771 eosinophil; CL:0000057 fibroblast; CL:0000097 mast cell; CL:0000236 (B cell, for autoantibody production context).
• UBERON: UBERON:0002348 endocardium; UBERON:0002080 right ventricle; UBERON:0002084 left ventricle; cardiac papillary muscle and chordae tendineae terms.
• CHEBI: eosinophil cationic protein / RNase 3 and major basic protein are proteins rather than small molecules — represent via UniProt/gene rather than CHEBI.

Sources: - In-Depth Review of Loeffler Endocarditis: What Have We Learned? (PMC10984210) - Endomyocardial fibrosis - Iroegbu (Cardiovasc Diagn Ther) - The cardiotoxicity of eosinophils (PMC2417450) - Plasma Cytokine Profile in Tropical Endomyocardial Fibrosis (PMC4193862)

7. Anatomical Structures Affected

Organ level: - Primary: Heart — right ventricle, left ventricle (either alone or, most commonly, biventricular), atrioventricular valves (mitral, tricuspid), atria (secondary dilation). - Secondary/complication-driven: Liver (congestive hepatomegaly), spleen (splenomegaly), lungs (pulmonary hypertension, pleural effusion, pulmonary congestion in left-sided disease), peritoneum (ascites — notably exudative/lymphocytic, suggesting a degree of peritoneal inflammatory involvement rather than pure transudative congestion), coronary vasculature (secondary sclerotic/proliferative changes reported in some histopathologic series). - Body systems: Cardiovascular (primary); hepatic, pulmonary, and hematologic (thromboembolic) systems secondarily.

Tissue/cell level: - Endocardium (subendothelial layer) — primary site of fibrous deposition. - Myocardium — subendocardial injury; relatively spared compared to endocardium. - Cell populations: eosinophils (infiltrating), fibroblasts (activated, ECM-producing), lymphocytes/mononuclear cells (chronic infiltrate), endothelial cells (injured), platelets (thrombus formation).

Subcellular level: No specific organelle-level pathology (e.g., mitochondrial, ER) has been characterized as central to EMF pathogenesis in the literature reviewed; this contrasts with some other cardiomyopathies (e.g., storage/metabolic cardiomyopathies) and represents a gap.

Localization: - Apex and inflow tract predominate (both ventricles can be affected). - Right ventricle: trabecular cavity obliteration, apical retraction, tricuspid valve tethering. - Left ventricle: apical obliteration (rounded "obliterated apex" morphology on echo), posterior mitral leaflet/chordal involvement (fibrosis characteristically engulfs the posterior mitral leaflet). - Laterality: Right-sided, left-sided, or biventricular — biventricular is the most common pattern (~50–55% in major series).

Sources: As cited above (StatPearls NBK513293; PMC4239813; PMID: 18596273).

8. Temporal Development

Onset: - Typically pediatric/adolescent onset (more than half of reported cases arise in the first decade of life), with a secondary adult-onset peak in women of childbearing age. Onset pattern is classically an insidious acute febrile illness (facial swelling, pruritus, eosinophilia) that may be mistaken for viral myocarditis or acute rheumatic fever, though many cases are only detected later in the fibrotic/chronic stage, or incidentally via echocardiographic screening (subclinical disease).

Progression — Davies three-stage model: 1. Acute (necrotic) stage — up to ~5 months; eosinophilic myocarditis with subendocardial necrosis; may present as fulminant heart failure/cardiogenic shock or be entirely asymptomatic/missed. 2. Thrombotic (intermediate/subacute) stage — beginning ~10 months post-onset, lasting several years; mural thrombus formation at the apex and behind the posterior mitral leaflet. 3. Fibrotic (chronic/healed) stage — the stage at which most patients present clinically; endocardium replaced by dense collagenous scar; restrictive physiology and valvular regurgitation dominate the clinical picture. This stage is essentially irreversible.

Rate/course: Variable — can be relatively indolent (subclinical disease detected on population screening, as in 77% of the Mozambique EMF-positive cohort) or rapidly progressive to severe heart failure and death. Once in the fibrotic stage, the disease course is chronic and progressive, without spontaneous remission; medical therapy does not appreciably alter the underlying fibrotic process (StatPearls NBK513293; a randomized placebo-controlled trial of prednisolone in Uganda found no significant benefit in preventing ascites reaccumulation, PMCID: PMC4678569 — see Treatment section).

Critical periods: The acute eosinophilic/necrotic stage represents the theoretical intervention window before irreversible fibrosis sets in (rationale for anti-eosinophilic/immunosuppressive therapy trials), but this stage is rarely captured clinically because of its nonspecific presentation and the resource constraints of endemic settings.

Sources: - Endomyocardial Fibrosis: Diagnosis and Management (Dove Press / JVD) - A population study of endomyocardial fibrosis in a rural area of Mozambique (PMID: 18596273) - The safety and efficacy of prednisolone... (PMC4678569)

9. Inheritance and Population

Epidemiology: - Prevalence: The only rigorous population-based echocardiographic screening study (rural Mozambique, n=1,063, all ages, PMID: 18596273) found an overall prevalence of 19.8% (211/1,063; 95% CI 17.4–22.2), highest in ages 10–19 (28.1%), and higher in males than females (23.0% vs. 17.5%, p=0.03) — a strikingly high figure reflecting substantial subclinical/mild disease burden not captured by hospital-based series. Note this is markedly higher than clinically ascertained hospital prevalence figures and reflects a broad echocardiographic case definition including mild disease. - Hospital-based series report EMF as accounting for up to ~20% of heart-failure/echocardiography referrals in endemic African centers (e.g., Kampala) and as the 4th most common cause of adult cardiac disease in some equatorial African nations. - One review cites a global burden estimate of ~12 million affected persons, predominantly in sub-Saharan Africa (cdt.amegroups.org) — this figure should be treated cautiously given the absence of large-scale multinational surveillance; it likely derives from extrapolation of regional prevalence data (such as the Mozambique 19.8% figure) rather than direct enumeration. - Historical literature documents >2,400 published cases worldwide, ~50% from sub-Saharan Africa and ~25% from Uganda alone (PMID: 18301727) — though this reflects publication/ascertainment bias rather than true incidence. - Hospital-series incidence appears to be declining over recent decades, plausibly linked to improved nutrition, parasite control, and healthcare access, though this has not been rigorously quantified prospectively.

Inheritance pattern: Not Mendelian — EMF is a complex/multifactorial disease. No autosomal dominant/recessive/X-linked/mitochondrial pattern has been established. Familial clustering (PMID: 757895) and HLA associations (PMID: 25780800) support polygenic/complex susceptibility rather than single-gene inheritance. Penetrance, expressivity, anticipation, germline mosaicism, and founder-effect concepts are therefore not directly applicable in the Mendelian sense; however, the HLA-B*58 and HLA-A*02:02 associations function analogously to susceptibility-locus "carrier frequency" concepts and would require population-specific allele-frequency data (not identified in this search) to quantify.

Population demographics: - Geographic distribution: Endemic — sub-Saharan Africa (Uganda, Mozambique, Nigeria, Cameroon, Congo, Malawi, Zambia most represented), South Asia (Kerala, India), East Asia (Guangxi Province, China), South America (Bahia, Brazil; Colombia). Rare sporadic cases reported in non-endemic/Western populations (e.g., a Western European case report, PMC7319822). - Sex ratio: Roughly equal in childhood-onset disease; adult-onset disease reported to affect women roughly twice as often as men in some series (though the Mozambique population screen found higher male prevalence overall — sex-ratio findings are series-dependent and possibly stage/age-dependent). - Age distribution: Bimodal — childhood/adolescent peak and adult (childbearing-age women) peak. - Socioeconomic gradient: Strongly associated with poverty; a recognized "neglected disease of poverty."

Sources: - A population study of endomyocardial fibrosis in a rural area of Mozambique (PMID: 18596273) - Endomyocardial Fibrosis: Still a Mystery after 60 Years (PMID: 18301727) - Endomyocardial fibrosis - Iroegbu (Cardiovasc Diagn Ther) - Idiopathic endomyocardial fibrosis in a Western European: a case report (PMC7319822)

10. Diagnostics

Clinical laboratory tests: No definitive/diagnostic blood test exists. Eosinophilia may be present in the acute inflammatory phase but is often absent by the fibrotic stage. Hypoalbuminemia is common in chronic disease. Elevated NT-proBNP/BNP and high-sensitivity troponin correlate with disease severity/progression and prognosis but are non-specific.

Electrocardiography: Low-voltage QRS, nonspecific ST-/T-wave abnormalities, AV block, bundle branch block, left/biatrial enlargement patterns.

Chest radiography: Cardiomegaly, atrial enlargement, pulmonary vascular congestion, occasional endomyocardial calcification, pleural/pericardial effusion.

Echocardiography — the primary diagnostic modality. Key structural findings: apical cavity obliteration (right and/or left ventricle), "mushroom sign" apical distortion, dense endocardial echogenicity, mural thrombus/spontaneous contrast, AV valve tethering with regurgitation, biatrial enlargement, restrictive (dip-and-plateau) diastolic filling pattern (short deceleration time, shortened isovolumic relaxation time). Left-ventriculography analog: apical obliteration; M-mode may show a distinctive "M-shaped" septal motion pattern.

Diagnostic scoring systems (Mocumbi criteria): - Definite diagnosis requires 2 major criteria, or 1 major + 2 minor criteria. - Major criteria: obliteration of the RV or LV apex; thrombi or spontaneous contrast without severe global ventricular dysfunction; retraction of the RV apex; AV valve dysfunction from adhesion of the valve apparatus to the ventricular wall. - Minor criteria: restrictive mitral/tricuspid inflow pattern; pulmonary valve diastolic opening; enlarged atrium with normal-sized ventricle. - A quantitative severity score (weighted per criterion) stratifies disease as mild (<8), moderate (8–15), severe (>15) in the original Mocumbi formulation; a related pediatric grading (4–6 mild, 7–9 moderate, 10–12 severe) has also been reported in a separate series (cdt.amegroups.org), indicating some variation in scoring implementations across studies — the exact cut-points should be verified against the primary source before formal curation.

Cardiac MRI: More sensitive than echocardiography for detecting intracardiac thrombus and for early/subclinical disease; late gadolinium enhancement (LGE) shows continuous subendocardial enhancement from subvalvular regions to the apex ("double V" / "three-layered" sign), correlating with histopathologic fibrosis. LGE-quantified fibrosis volume has been reported as an independent predictor of mortality (PMC12701864/PMID: 41399600). MRI is valuable for preoperative planning and treatment-response monitoring.

Myocardial contrast echocardiography (MCE): Adjunctive tool for apical obliteration/thrombus detection when conventional imaging is limited.

Cardiac catheterization: Rarely required now; shows a classic restrictive "dip-and-plateau" ventricular pressure pattern; angiography demonstrates apical cavity obliteration.

Endomyocardial biopsy: Can demonstrate subendocardial fibrosis and thrombus, but limited utility due to patchy distribution of fibrosis and procedural risk (risk of thrombus dislodgement/embolization in a fibrotic, thrombus-laden ventricle).

Genetic testing: Not part of the standard diagnostic pathway for classic tropical/idiopathic EMF (no established causal gene). For suspected hypereosinophilic-syndrome-driven (Löffler/eosinophilic) EMF, FIP1L1-PDGFRA fusion testing (FISH or RT-PCR) is clinically actionable, as fusion-positive patients respond to imatinib.

Clinical criteria / differential diagnosis: Key differentials include viral myocarditis (acute stage), cardiac amyloidosis, cardiac sarcoidosis, dilated cardiomyopathy, left ventricular noncompaction, carcinoid heart disease, anthracycline cardiotoxicity, constrictive pericarditis, and radiation-induced cardiomyopathy — all part of the broader restrictive-cardiomyopathy/heart-failure-with-preserved-EF differential.

Screening: No formal national/international newborn or population screening program exists. The single major population-based echocardiographic prevalence survey (Mozambique, PMID: 18596273) demonstrates the feasibility and yield of community echocardiographic screening in endemic areas but has not been scaled into a routine screening program.

Sources: - Endomyocardial Fibrosis - StatPearls - A Narrative Review on Endomyocardial Fibrosis (PMC12701864 / PMID: 41399600) - Endomyocardial fibrosis - Iroegbu (Cardiovasc Diagn Ther) - Left ventricle endomyocardial fibrosis: a case report (PMC10422788)

11. Outcome/Prognosis

Survival/mortality (untreated/medically managed): - Historical Ugandan autopsy series (1959–1969): average survival ~2 years after symptom onset. - Broadly cited figure: "75% mortality within 2 years" / "one-third to one-half of patients with advanced disease dying within 2 years" with medical management alone (figures vary by series and disease-stage-at-presentation). - Atrial fibrillation is associated with worse prognosis.

Surgical outcomes: - Operative (30-day) mortality: ~15–21.7% across major surgical series; one series reported 21.7% 30-day mortality plus 13% late mortality within the first 2 postoperative years. - Life-table survival including operative mortality: ~67% at 2 years, ~55–68% at up to 17 years in selected surgical cohorts; a more recent Mozambican surgical series reported ~76.5% 5-year survival with 70.9% of operated patients showing functional improvement. - Recurrence: Fibrosis recurrence requiring reoperation in ~4–18.8% of surgical patients across series; EMF appearing in the previously unaffected contralateral ventricle in ~8.8% in one series. - Surgery is explicitly regarded as palliative — it corrects structural/valvular consequences but does not alter the underlying fibrotic disease process, and recurrence is well documented.

Morbidity/functional outcomes: - The majority of clinically ascertained (hospital-referred) patients present in NYHA class III/IV (62–98% across cited series). - Postoperative functional improvement is achievable in a substantial subset (e.g., 70.9% improved in one series; younger/less advanced patients achieving NYHA I–II postoperatively in a pediatric series), but a meaningful minority show no improvement or clinical deterioration. - Complications: heart failure, atrial fibrillation, AV block, thromboembolism (stroke, pulmonary embolism), progressive valvular dysfunction, pulmonary hypertension, infective endocarditis susceptibility, pericardial effusion, sudden cardiac death.

Prognostic factors/biomarkers: Advanced diastolic dysfunction, severe atrial enlargement, biventricular involvement, extensive fibrosis/thrombus burden on cardiac MRI, elevated NT-proBNP, and pulmonary hypertension are cited as predictors of poor outcome; LGE-quantified fibrosis volume on cardiac MRI independently predicts mortality (PMC12701864/PMID: 41399600).

Sources: - Endomyocardial fibrosis: Early and late results of surgery in 20 patients - Surgery for endomyocardial fibrosis revisited (Eur J Cardiothorac Surg) - Endomyocardial fibrosis - Iroegbu (Cardiovasc Diagn Ther) - A Narrative Review on Endomyocardial Fibrosis (PMC12701864 / PMID: 41399600)

12. Treatment

Pharmacotherapy (symptomatic/supportive; no disease-modifying drug established): - Diuretics (loop diuretics — furosemide, torasemide) for congestive symptoms (NCIT candidate: Pharmacotherapy NCIT:C15986; specific class terms to be verified). - ACE inhibitors and beta-blockers — standard heart-failure adjuncts, though restrictive physiology limits their hemodynamic benefit relative to dilated cardiomyopathy. - Anticoagulation (warfarin; direct oral anticoagulants limited by cost/access in endemic settings) for documented intracardiac thrombus/atrial fibrillation. - Corticosteroids (prednisolone): Tested in a double-blind, randomized, placebo-controlled trial in Uganda (n=35; 1 mg/kg/day, max 60 mg) for prevention of ascites reaccumulation in EMF: primary outcome (progression to grade 3 ascites) occurred in 60% of prednisolone-treated vs. 86% of placebo-treated patients (RR 0.70, 95% CI 0.43–1.11, p=0.12) — not statistically significant, though the drug was safe (PMCID: PMC4678569). This is the only identified randomized controlled trial of a disease-directed medical therapy in EMF and represents a key piece of negative-evidence for immunosuppressive intervention at the (typically late) disease stage studied. - Rate/rhythm control (beta-blockers, digoxin) for atrial fibrillation.

Advanced/targeted therapeutics (for the eosinophilic/Löffler-variant subset specifically): - Imatinib (tyrosine kinase inhibitor) — first-line for FIP1L1-PDGFRA-fusion-positive hypereosinophilic syndrome/Löffler endocarditis; achieves eosinophil normalization and echocardiographic improvement. - Mepolizumab (anti-IL-5 monoclonal antibody) — used as an eosinophil-targeting immunomodulator in HES/Löffler endocarditis, though evidence specific to established fibrotic EMF is limited (most benefit expected in the pre-fibrotic/acute eosinophilic stage). - Interferon-alfa — reported for corticosteroid/imatinib-resistant HES-associated cardiac disease.

Surgical/interventional: - Endocardiectomy (endocardial decortication) ± mitral and/or tricuspid valve repair or replacement, typically via median sternotomy with cardiopulmonary bypass — the mainstay definitive intervention for NYHA III/IV disease. NCIT candidates: Surgical Procedure (NCIT:C15329), Orthopedic Surgical Procedure not applicable; a cardiac-surgery-specific NCIT term should be verified. - Cavopulmonary connection (Fontan-type) procedures have been proposed as beneficial adjuncts for severe right-ventricular EMF in some case reports. - Heart transplantation: Not an established/first-line therapy (StatPearls notes "no established benefit"), but case reports document successful outcomes — e.g., a patient with FIP1L1-PDGFRA-associated EMF alive and asymptomatic 5 years post-transplant, and a case report describing 2-year good graft function with vigilance for possible disease recurrence in the allograft (PMCID: PMC12046388) — an important, still poorly characterized risk given EMF's presumed ongoing systemic (immune/eosinophilic) driver.

Experimental/investigational: No disease-specific investigational agents in active clinical trials for classic tropical EMF were identified; research priorities (per PMID: 18301727) include measuring inflammatory markers (CRP, TNF-α), studying FIP1L1-PDGFRA prevalence in broader EMF cohorts, examining serotonin receptor polymorphisms, and conducting further population-based echocardiographic surveys.

Treatment outcomes: See Prognosis section — surgical endocardiectomy is the most effective available intervention but carries substantial operative mortality (~15–22%) and disease recurrence risk (~4–19%); medical therapy alone does not appear to alter the natural history of established fibrotic disease (per the negative prednisolone RCT).

Treatment strategy/algorithm: Stage-dependent — acute eosinophilic myocarditis phase (if captured) may warrant corticosteroids ± eosinophil-targeted therapy (imatinib if FIP1L1-PDGFRA+, mepolizumab); established fibrotic-stage disease is managed with heart-failure pharmacotherapy and anticoagulation, escalating to endocardiectomy ± valve surgery for NYHA III/IV symptoms refractory to medical therapy; heart transplantation is reserved for exceptional cases.

Sources: - The safety and efficacy of prednisolone in preventing reaccumulation of ascites among EMF patients in Uganda (PMC4678569) - In-Depth Review of Loeffler Endocarditis (PMC10984210) - Case report on heart transplantation in endomyocardial fibrosis (PMC12046388) - Successful Heart Transplantation for Unreversible EMF Related to FIP1L1-PDGFRA CEL - Endomyocardial Fibrosis Treatment & Management (Medscape)

13. Prevention

Primary prevention: No disease-specific primary prevention strategy is established. Given the etiologic hypotheses, plausible (but not formally trial-proven for EMF-incidence reduction) primary-prevention levers include: - Population deworming and helminth/schistosomiasis control programs in endemic regions. - Nutritional interventions addressing protein-calorie and micronutrient (magnesium) deficiency and reducing dependence on inadequately processed cassava. - Malaria control. - Poverty alleviation (the strongest and most consistently identified structural risk factor).

Notably, no clinical trial has directly tested whether these interventions reduce EMF incidence; the rationale is inferential from the epidemiologic/mechanistic associations discussed above.

Secondary prevention (early detection): Community echocardiographic screening, as piloted in the Mozambique population study, could theoretically identify subclinical/mild disease (the ~77% of prevalent cases who were asymptomatic in that study) for closer monitoring, though no screening program has been operationalized at scale, and there is no proven early intervention that alters the natural history once fibrosis is detected.

Tertiary prevention: Standard heart-failure medical management, anticoagulation to prevent thromboembolic complications, and timely surgical referral (endocardiectomy ± valve surgery) before end-stage/refractory heart failure develops.

Immunization: Not applicable — no vaccine-preventable causal agent has been established.

Genetic counseling: Not applicable in the conventional Mendelian sense given the complex/multifactorial and non-Mendelian inheritance pattern; family history (given documented familial clustering) may still warrant clinical/echocardiographic surveillance of relatives in high-risk families, though this is not a formalized guideline recommendation identified in the literature.

Public health interventions: Improved sanitation and vector/parasite control (reducing helminth and malarial burden), nutritional support programs, and expanded access to echocardiography in endemic primary-care settings are the most plausible public-health levers, consistent with the observed decline in hospital-based EMF incidence attributed generally to "improving healthcare and living standards."

Sources: As cited above (PMID: 18301727; PMC12701864/PMID: 41399600; PMC4239813).

14. Other Species / Natural Disease

Taxonomy: Naturally occurring EMF as described here is a human disease; there is no well-established veterinary/naturally occurring analog reported in companion animals or wildlife in the literature reviewed.

Experimental (induced, non-natural) models: - Cercopithecus aethiops (African green monkey; NCBI Taxon ID needed/verify) fed a cassava-based, severe-protein-deficient diet developed cardiac lesions resembling human EMF, while banana-fed controls did not — supporting the cassava/protein-deficiency causal hypothesis (cited in PMID: 18301727). - Plantain-feeding experiments in guinea pigs, rats, and Patas monkeys (testing the serotonin hypothesis) failed to reproduce EMF lesions, effectively refuting that hypothesis (PMID: 18301727).

Comparative biology: No dedicated comparative pathology or evolutionary-conservation literature on EMF mechanisms across species was identified. The eosinophil-mediated cardiotoxicity mechanism (ECP/MBP-driven endothelial injury and coagulation activation) is presumed broadly conserved across mammals based on general eosinophil biology, but this has not been formally studied in the specific context of EMF model development.

Zoonotic potential / transmission: Not applicable — EMF is not an infectious/transmissible disease itself, though proposed infectious co-factors (helminths, Plasmodium) are separately zoonotic/vector-borne in their own right.

Note on a distinct but nomenclature-adjacent condition: Endocardial fibroelastosis (EFE, OMIM:226000) — a different, largely pediatric/congenital disease — has documented animal models (e.g., distention of the immature left ventricle inducing EFE-like lesions, PMC4433646) and, in some human cases, a ciliopathy-gene basis (e.g., in Alström syndrome, PMC8541947). These EFE-specific models and genetic findings should not be conflated with tropical/idiopathic EMF model organism data.

Sources: - Endomyocardial Fibrosis: Still a Mystery after 60 Years (PMID: 18301727) - Distention of the Immature Left Ventricle Triggers Development of Endocardial Fibroelastosis (PMC4433646) (EFE model — distinct disease, included for disambiguation only)

15. Model Organisms

Summary: Model-organism research specific to tropical/idiopathic EMF is sparse and largely historical, reflecting both the disease's endemic-region concentration (limiting research infrastructure) and the field's general stagnation after the 1980s (publication volume "declined dramatically post-1980s," peaking "prior to the diffusion of echocardiography in much of the tropics," per PMID: 18301727).

Genetic/induced models identified: - Cassava/protein-deprivation model (non-human primate): Cercopithecus aethiops fed uncooked cassava under severe protein restriction — produced cardiac histopathology resembling human EMF, plus hepatic changes resembling tropical splenomegaly syndrome, supporting a shared cassava/malnutrition etiology for both conditions. This model did partially recapitulate the human phenotype but has not been followed up with modern molecular characterization. - Serotonin/plantain hypothesis models (guinea pig, rat, Patas monkey): Failed to reproduce EMF lesions — a negative model result that helped rule out the serotonin-metabolite hypothesis. - No genetically engineered (knockout/knock-in/transgenic/conditional/humanized) mouse or other rodent model of EMF was identified in this search — a clear gap, likely attributable to the absence of an established causal gene to target. - FIP1L1-PDGFRA / hypereosinophilic syndrome models: General HES/eosinophilic-cardiotoxicity models (e.g., IL-5 transgenic or eosinophil-adoptive-transfer mouse models used in broader eosinophilic-disease research) exist in the eosinophil biology literature but were not specifically identified as validated EMF models in this search; they represent the most plausible near-term modeling avenue given the shared mechanism with Löffler endocarditis.

Model limitations: No model captures the full multifactorial human EMF phenotype (chronic malnutrition + parasitic/immune exposure + genetic susceptibility + years-long fibrotic evolution); the primate cassava model is the closest histopathologic recapitulation identified but is decades old, ethically and logistically difficult to repeat with modern techniques, and was not molecularly characterized by contemporary standards (no transcriptomic/proteomic follow-up reported).

Applications: Existing models have been used primarily to test/refute specific etiologic hypotheses (cassava/protein deficiency: supported; serotonin/plantain: refuted) rather than to dissect molecular pathogenesis or screen therapeutics — an important direction the 2025 Nature Reviews Cardiology review (Mocumbi et al., DOI: 10.1038/s41569-025-01138-x) explicitly flags as a priority for identifying preclinical biomarkers and novel therapeutic targets, though full-text access to that review's specific model-organism recommendations could not be retrieved in this session (paywalled).

Resources: No dedicated EMF model-organism database or repository was identified (unlike diseases with established genetic models catalogued in MGI/IMPC/ZFIN); this itself is a notable research-infrastructure gap for EMF.

Sources: - Endomyocardial Fibrosis: Still a Mystery after 60 Years (PMID: 18301727) - Endomyocardial fibrosis: recent advances and future therapeutic targets (Nat Rev Cardiol 2025) — abstract/metadata only accessible (paywalled); Mocumbi et al., Nat Rev Cardiol 2025;22(8):564–576.

Summary of Key Knowledge Gaps (for curation planning)

1. No established causal gene — EMF is genetically complex, not monogenic; avoid asserting a causal gene/OMIM entry (OMIM:226000 belongs to the distinct disease EFE, not EMF).
2. HLA associations (HLA-B*58, HLA-A*02:02) are single-study, population-specific, and unreplicated — should be curated as SUSCEPTIBILITY-level evidence with appropriately guarded confidence, not as established risk alleles.
3. The etiologic model remains a multifactorial hypothesis, not a proven causal chain — the eosinophil/Löffler-equivalence mechanism has the strongest histopathologic support, but "no single proposed factor can explain the occurrence of EMF worldwide" per the primary review literature.
4. No RCT-proven disease-modifying medical therapy exists — the one identified RCT (prednisolone for ascites) was negative; this is important, citable negative evidence.
5. FIP1L1-PDGFRA/imatinib pathway is a distinct, actionable, but etiologically separate subset (clonal hypereosinophilic syndrome/Löffler-variant EMF) — should likely be modeled as a related but distinct causal branch/subtype rather than conflated with classic nutritional/parasitic tropical EMF.
6. Modern omics data (transcriptomic, proteomic, single-cell) for human EMF cardiac tissue are essentially absent from the literature surveyed — a genuine research and knowledge-base gap.
7. Global prevalence/burden estimates (e.g., "~12 million") are poorly sourced/extrapolated and should be cited cautiously with the underlying Mozambique 19.8% population-prevalence study (PMID: 18596273) as the strongest primary anchor.
8. Full text of the most current comprehensive review (Mocumbi et al., Nat Rev Cardiol 2025) was paywalled and not fully accessible in this session — recommend independent retrieval for the most up-to-date synthesis before finalizing a knowledge-base entry.