Ebstein Anomaly (Including the Form with MYH7- and TPM1-Related Non-Compaction) and NKX2-5-Related Heart Disease

Malformation of the tricuspid valve and right ventricle with heterogeneous genetic aetiology, notably the autosomal dominant subgroup with left ventricular non-compaction caused by sarcomeric variants; and the allelic spectrum of the transcription factor NKX2-5, whose hallmark is atrial septal defect with progressive atrioventricular conduction disease.

Literature and regulatory information updated: 31 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Definitions and scope

Ebstein anomaly is a malformation of the tricuspid valve and right ventricle in which the septal and posterior (also called inferior) leaflets fail to separate from the ventricular wall during development and remain attached below the anatomical annulus, displaced towards the apex, while the anterior leaflet arises at the level of the annulus, enlarged and “sail-like” [25, 31]. The anatomical consequence is the division of the right heart into a morphological atrium, an “atrialised” ventricular portion functionally integrated with the atrium, and a residual functional right ventricle; the usual haemodynamic consequence is tricuspid regurgitation [31]. OMIM describes the anomaly as a displacement of variable severity of the valve into the right ventricle, with leaflets that are sometimes dysplastic, and stresses that the spectrum ranges from severe fetal and neonatal disease to virtually asymptomatic survival into adulthood; about 20% of patients have extracardiac anomalies in the context of chromosomal or Mendelian disorders, and the non-syndromic form may be sporadic or familial [1].

NKX2-5-related heart disease is the set of phenotypes caused by heterozygous variants in the gene for the transcription factor NKX2-5 (NK2 homeobox 5, human homologue of Drosophila tinman). OMIM groups it into six phenotypic entries, of which atrial septal defect 7 with or without atrioventricular conduction defects (ASD7) is the most frequent; the others are ventricular septal defect 3, tetralogy of Fallot, conotruncal malformations, hypoplastic left heart syndrome 2 and congenital non-goitrous hypothyroidism 5 [13]. In November 2024, ClinGen classified as definitive the relationship between NKX2-5 and an entity designated “congenital heart disease, conduction disease and cardiomyopathy” with autosomal dominant transmission [6].

The two entities share common ground. Ebstein anomaly and other tricuspid anomalies are part of the NKX2-5 spectrum [16, 30], and Ebstein anomaly is the congenital heart defect most often associated with left ventricular non-compaction, an association that, in families with MYH7 and TPM1 variants, is transmitted as an autosomal dominant trait with variable penetrance [14, 38]. Ebstein anomaly is therefore described below as a malformation of heterogeneous aetiology, with a table of genetic causes and an in-depth discussion of the sarcomeric subgroup, and NKX2-5 as an allelic spectrum. Isolated left ventricular non-compaction, progressive cardiac conduction disease, the general approach to congenital heart disease, non-syndromic tetralogy of Fallot, hypertrophic and dilated cardiomyopathies, 1p36 deletion, FLNA-related disorders, CHD4-related disorders and NONO-related X-linked intellectual disability belong to the differential diagnosis.

1.2. Historical milestones

Wilhelm Ebstein described the malformation in 1866, in an autopsy specimen; Helen Taussig characterised it clinically in 1947, and the first diagnosis in a living patient was published in 1949 [27]. Familial aggregation was documented from the 1960s onwards: affected siblings in 1966, a boy and a maternal uncle in 1968, a father and son with persistent atrial standstill in 1985 and two sisters with prenatal diagnosis, daughters of consanguineous parents, in 1992 [1]. The observation that Ebstein anomaly is not confined to the right heart took shape with the description, in 2004, of non-compacted left ventricular myocardium in three patients, and with the 2005 Mayo Clinic series, in which 39% of patients had left-sided lesions, of the myocardium or the valves [39, 40].

The molecular link between the phenotype and the sarcomere began with Budde and colleagues, who in 2007 mapped non-compaction to chromosome 14q by genome-wide linkage analysis in a three-generation German family and identified the MYH7 variant p.(Arg281Thr), present in carriers with non-compaction and, in some of them, with Ebstein anomaly and atrial septal defect [41]. In 2011, Postma and colleagues found MYH7 variants in 8 of 141 probands with Ebstein anomaly (6%), concentrated in patients with non-compaction [7]. TPM1 was the second gene, first in a de novo case (Kelle and colleagues, 2016) and then in a family with segregation (Nijak and colleagues, 2018) [8, 9].

For NKX2-5, the milestone is the 1998 article by Schott and colleagues, who mapped to 5q35 a dominant locus for cardiac malformations with atrioventricular conduction abnormalities and identified three NKX2-5 variants in four families, in which 27 of the 33 affected individuals had ostium secundum atrial septal defect and all had atrioventricular block of variable degree [29, 34]. The clinical description of families with this combination was earlier: three families were published in 1966 and 1967 and father-to-son transmission was demonstrated in 1970 [4]. In 1999, Benson and colleagues extended the spectrum to ventricular septal defect, tetralogy of Fallot, double-outlet right ventricle and Ebstein anomaly, and showed that atrioventricular block was the main manifestation in almost a quarter of carriers [30]. In 2006, the identification of NKX2-5 variants in patients with thyroid dysgenesis added an extracardiac phenotype [42].

2. Molecular Genetics and Pathophysiology

2.1. Morphogenesis of the tricuspid valve and compact myocardium

The leaflets of the atrioventricular valves form from the endocardial cushions of the atrioventricular canal, which result from endocardial-to-mesenchymal transition between the fourth and eighth weeks; the leaflets then elongate and separate from the myocardial wall by delamination, which in the septal leaflet of the tricuspid valve probably occurs through integrin-dependent apoptosis (anoikis) [3]. Failure of delamination of the septal and posterior leaflets is the proposed morphological mechanism of Ebstein anomaly [3, 25]. Compaction of the ventricular trabeculae proceeds in parallel, around the eighth week, under the control of the same signalling pathways (TGF-β/Smad, canonical and non-canonical Wnt, Notch, BMP and FGF), and epicardium-derived cells provide signals both to the atrioventricular valves and to the compact myocardium [3, 38]. This overlap explains, as a hypothesis, how often Ebstein anomaly and non-compaction coexist; the introduction to inherited cardiomyopathies in the same book notes that non-compaction cardiomyopathy may occur as part of a syndrome, in a congenital heart defect such as Ebstein anomaly or as an isolated cardiomyopathy, with overlap with the dilated and hypertrophic forms [3, 43]. The non-compaction chapter of Clinical Cardiogenetics notes that, in mice, inactivation of myocardiogenesis genes, including Nkx2.5 and Bmp10, causes congenital non-compaction and that, in humans, only NKX2.5 had described variants among those genes [38].

The right ventricle in Ebstein anomaly is not merely a volume-overloaded ventricle: the North American guidelines describe the disease as both valvular and myopathic, and surgical reviews record a variable degree of right ventricular myopathy with a reduced number of sarcomeric units [25, 28]. The pathophysiology results from systolic regurgitation from the functional ventricle into the atrialised portion and the atrium, which dilate; an atrial septal defect, when present, allows left-to-right shunting or, especially on exertion, right-to-left shunting, with cyanosis, and systemic output may remain chronically low [31].

2.2. Genetic architecture of Ebstein anomaly

Ebstein anomaly is genetically heterogeneous. In an Italian series of 44 consecutive patients (two centres, 1997–2009), 12 (27%) had syndromic forms, including CHARGE syndrome in two and VACTERL association, Noonan syndrome, Kabuki syndrome, Holt-Oram syndrome and Cornelia de Lange syndrome in one each; array comparative genomic hybridisation in 10 of the 12 syndromic patients identified an interstitial deletion of about 4 Mb at 8p23.1 and a 1pter→1p36.32 deletion with an Xp22.32 duplication, and the karyotype showed an 18q deletion; no GATA4 or NKX2.5 variants were found in the 28 non-syndromic patients tested [1, 12]. The authors concluded that 1p36 deletion and 8p23.1 deletion (where GATA4 is located) are the most frequent chromosomal imbalances in Ebstein anomaly [12].

Copy number variation (CNV) studies reinforce the heterogeneity. In 47 cases of isolated Ebstein anomaly from New York State, 35 candidate CNVs were identified in 24 cases (51%) and rare variants in cardiomyopathy genes in 11 (23%), with two CNVs near HEY1 (Notch pathway) in three unrelated cases [44]. In 60 cases among 2,891,076 live births in California counties (1991–2010), 38 candidate CNVs arose in 28 cases (46%), five of them near myocardial development genes (NODAL, PDLIM5, SIX1, ASF1A, FGF12) [45]. The proportion of candidate CNVs is not equivalent to diagnostic yield: most of these changes have no established pathogenicity.

In a population-based case-control study from the United States (National Birth Defects Prevention Study, 1997–2011; 135 cases and 11,829 controls, excluding known syndromes and chromosomal disorders), a family history of congenital heart disease in first-degree relatives was reported by 5% of case mothers and by 1% of control mothers, with an adjusted odds ratio (OR) of 4.1 (95% confidence interval [95% CI] 1.8–9.5); maternal exposure to second-hand smoke at home had an OR of 2.2 (1.1–4.4), but maternal smoking was not associated (OR 1.3; 0.8–2.1) [46]. These data suggest a hereditary component even in the non-syndromic form, without quantifying it.

AD, autosomal dominant; AR, autosomal recessive; AVB, atrioventricular block; ASD, atrial septal defect; VSD, ventricular septal defect; LV, left ventricle.

2.3. The sarcomeric subgroup: MYH7 and TPM1

MYH7 encodes β-myosin heavy chain, the predominant myosin in the human ventricle and in slow type I skeletal muscle fibres [64]. In the non-compaction chapter of Clinical Cardiogenetics, MYH7 is the most frequent gene in isolated non-compaction, accounting for up to 21% of cases, and sarcomeric variants as a whole explain about 30%; the MYH7 variants associated with non-compaction cluster in the ATPase active site of the N-terminal head, a conserved region needed for force generation, and it is in this region that the variants associated with non-compaction with or without Ebstein anomaly arise, whereas about 30% of MYH7 variants lie in the C-terminal rod, more associated with skeletal myopathies [38]. The review by Vermeer and colleagues counted at least nine distinct MYH7 variants in Ebstein anomaly with non-compaction, five in the head and four in the rod [3].

In the study by Postma and colleagues, MYH7 analysis in 141 unrelated probands of Western European ancestry identified heterozygous variants in 8 (6%): seven distinct variants, five novel and two previously described in hypertrophic cardiomyopathy, all missense except for a three-base-pair deletion, one of them de novo [7]. Non-compaction was present in 6 of the 8 probands with a variant according to the authors (OMIM records 7 and one uncertain case) and in none of the 133 without a variant; in families with a variant, other carriers had cardiomyopathy or malformations such as secundum atrial septal defect, ventricular septal defect, bicuspid aortic valve, coarctation and pulmonary artery stenosis or hypoplasia [7, 47]. The authors’ practical conclusion is that MYH7 variants are found predominantly in Ebstein anomaly with non-compaction, and that it is this subgroup that warrants genetic testing and family assessment [7]. The systematic review by Thareja and colleagues (58 articles up to February 2022) tabulated the variants described in the association: MYH7 p.(Arg281Thr) (family), p.(Leu301Gln), p.(Met362Arg) (family) (these two are published designations, with no corresponding ClinVar record), a maternally inherited splicing variant in a family in which only the proband had the full association, TPM1 p.(Asp159Asn) and p.(Leu113Val), KLHL26 p.(Arg237Cys) (family), two NONO variants in patients with an extracardiac syndrome and chromosomal gains at 15q11.2 and 1q44 in a case negative on a 17-gene panel [3].

ClinGen classifies the relationship between MYH7 and congenital heart disease as “limited” (February 2024), whereas the relationships with hypertrophic cardiomyopathy, dilated cardiomyopathy 1S and skeletal myopathy are definitive [64]. This difference has interpretive consequences: a MYH7 variant in a patient with Ebstein anomaly should be assessed by the criteria applicable to cardiomyopathy, and isolated Ebstein anomaly, without non-compaction, has a low probability of being explained by MYH7 [7]. In the GeneReviews chapters, MYH7 accounts for 33% of genetic hypertrophic cardiomyopathy and 4% of dilated cardiomyopathy, with non-compaction among the allelic disorders [19, 20]. The dilated cardiomyopathy gene table in Clinical Cardiogenetics assigns to MYH7 the phenotypes dilated, hypertrophic and non-compaction cardiomyopathy and myopathy, with early onset and conduction system disease, and to TPM1 the phenotypes dilated, hypertrophic and non-compaction cardiomyopathy, both with autosomal dominant transmission [65]. In skeletal muscle, MYH7 variants cause myosin storage myopathy and Laing distal myopathy, both autosomal dominant, and have been described in scapuloperoneal and limb-girdle weakness, congenital fibre-type disproportion and multiminicore disease; cardiomyopathy may coexist with myosin storage myopathy [66]. Therefore, in a carrier of a MYH7 variant identified through Ebstein anomaly, the history and examination should include distal and proximal muscle strength.

TPM1 encodes striated muscle α-tropomyosin, which polymerises along the actin filament and, with the troponin complex, mediates calcium-dependent regulation of cross-bridge formation [67]. It is a definitive cause of hypertrophic cardiomyopathy (less than 3% of cases) and a moderate cause of dilated cardiomyopathy (less than 1%), and ClinGen records no evidence of haploinsufficiency, so an isolated deletion of the gene should be interpreted with caution [19, 20, 67]. OMIM includes non-compaction in entry 611878 (LVNC9) and notes that some patients have Ebstein anomaly and others mitral regurgitation [18]. Nijak and colleagues recommended testing MYH7 and TPM1 whenever non-compaction and Ebstein anomaly coexist [9].

The review by Thareja and colleagues writes the Nijak variant as p.L131V; the original article and ClinVar give p.(Leu113Val). VUS, variant of uncertain significance.

2.4. NKX2-5: structure, function and mechanism

NKX2-5 is located at 5q35.1 and encodes a homeodomain transcription factor, required for cardiac and splenic development, which activates target genes such as NPPA in cooperation with GATA4 and interacts with TBX5 [6, 72]. Beyond morphogenesis, it governs septation and the maturation and lifelong maintenance of atrioventricular node function, which explains the coexistence of structural defects and conduction disease that becomes manifest long after birth [29, 34]. In the formation of the conduction system, temporal repression of NKX2-5 in the areas of the sinus node, atrioventricular node and peripheral system is required for their specification [16].

The predominant mechanism is haploinsufficiency. Of Schott’s three founding variants, two reduced binding to target DNA and the third could increase it, so the mechanism is not limited to loss of protein [29]. ClinGen dosage curation assigns a haploinsufficiency score of 3 (sufficient evidence) and a triplosensitivity score of 0; a 5q35 deletion removing NKX2-5 is therefore read as causal, whereas a duplication has no established dosage effect [6]. OMIM records a del(5)(q35.1q35.3) deletion in a girl with atrial septal defect, patent arterial duct and ventricular non-compaction, and four other cases of deletions of the region with atrial septal defect or cardiomyopathy [13]. Some missense variants in the homeodomain have a dominant-negative effect on the normal allele, such as those found in thyroid dysgenesis [42].

2.5. NKX2-5 variant spectrum and correlations

The review by Ellesøe and colleagues gathered 60 NKX2-5 variants in 208 patients, including a new Danish family [16]. Variants were significantly more frequent in probands from families with congenital heart disease (18 of 198; 9.1%) than in non-familial cases (17 of 1037; 1.6%; p = 7.1 × 10⁻⁹), and 94% of carriers with atrial septal defect and conduction disease were familial cases [16]. In a screen of 474 patients with congenital heart disease, McElhinney and colleagues found variants in 3 of 71 patients with secundum atrial septal defect (4%), two of them without a conduction defect, and in 3% of the extended set of 608 patients [4, 13]. A 2026 report cites a proportion of 8–12% of NKX2-5 variants in familial atrial septal defects [78]. NKX2-5 also appears as a possible cause of dilated cardiomyopathy in the GeneReviews gene table, with “limited” ClinGen evidence for this isolated relationship [19]; the corresponding table in Clinical Cardiogenetics associates it with dilated cardiomyopathy, congenital heart defects and congenital hypothyroidism, with arrhythmias and sudden death, with autosomal dominant transmission [65]. In bicuspid aortic valve, although more than 50 variants distributed across the functional domains have been described in atrial septal defect, ventricular septal defect, hypoplastic left heart and tetralogy, only two have been described in patients with a bicuspid valve (p.(Arg25Cys) in a female patient with trisomy 21 and p.(Lys192*) (published designation, with no corresponding ClinVar record) in a three-generation family, with almost complete loss of transcriptional activity), and NKX2-5 is not unequivocally recognised as a bicuspid valve gene [79].

The most cited variant is NM_004387.4:c.73C>T, p.(Arg25Cys), normalised in VariantValidator (NC_000005.10:g.173235011G>A) [68]. It has been described in patients with tetralogy, interrupted aortic arch, common arterial trunk, hypoplastic left heart and thyroid dysgenesis, but OMIM considers it low-penetrance, Ellesøe and colleagues note that it has been suggested to be non-causal given the number of unaffected carriers, and ClinVar records conflicting classifications (assessment of March 2026) [13, 16, 80]. A variant of this kind does not establish the diagnosis on its own.

The Icelandic variant NP_004378.1:p.(Phe145Leu) is the example of a founder effect: in a genome-wide association study with whole-genome sequencing (424 dilated cardiomyopathy cases and 337,689 controls), it was carried by 1 in every 7100 Icelanders, was associated with dilated cardiomyopathy, heart failure, sudden death, high-grade atrioventricular block and atrial septal defect, and had a higher penetrance of serious heart disease than the FLNC variant identified in the same study [26]. In ClinVar, the same protein change corresponds to three different alleles (NM_004387.4:c.433T>C, pathogenic; c.435C>G, conflicting; c.435C>A, VUS), and the Icelandic study designates the variant by protein notation, so only this is used here [81].

Ellesøe’s Danish family had a one-nucleotide deletion at position 112 of exon 1, described as “112delG”, with an abnormal sequence from amino acid 37 and termination at 175 [16]. The description NM_004387.4:c.112del, normalised in VariantValidator as p.(Glu38ArgfsTer138), predicts termination at codon 175, consistent with the original description [68]. Recent truncating variants, such as c.300del, p.(Asp101ThrfsTer75), in a three-generation Canadian family, and c.604_605del, p.(Leu202GlyfsTer49), in a Chinese family, confirmed the same core phenotype [68, 82, 83].

Genotype-phenotype correlations are limited. In a functional assay in yeast, variants in the third helix of the homeodomain, which confers DNA-binding specificity, were associated with ventricular or atrioventricular septal defects, with partial to complete loss of function [13]. In Ellesøe’s review, none of the 3 patients with hypoplastic left heart or the 13 with tetralogy had conduction disease, in contrast to 112 of the 145 with atrial septal defect [16]. Penetrance is incomplete: Goldmuntz and colleagues found asymptomatic carriers in families of patients with tetralogy, and Bosi and colleagues suggested that a prolonged PR interval may be the only manifestation [4, 13].

3. Clinical Spectrum and Manifestations by System

3.1. Anatomy and associated lesions in Ebstein anomaly

The pathognomonic features are rotational and apical displacement of the leaflets with failure of delamination, mainly of the septal and inferior leaflets (anterior displacement is limited by the ventriculo-infundibular fold), the large, “sail-like” anterior leaflet with abnormal chordae, the division of the right ventricle into a proximal atrialised portion, dilated, non-trabeculated and often thinned, and a distal functional ventricle, the anatomical annulus that is always dilated, and variable degrees of infundibular obstruction by the redundant tissue of the anterior leaflet [25]. The most frequent associated anomalies are atrial-level shunting (secundum atrial septal defect or patent foramen ovale) and accessory pathways, manifest or concealed, including Mahaim-type pathways; in one series, atrial septal defect or patent foramen ovale was present in 50%, pulmonary stenosis or atresia in 4% and ventricular septal defect in 3%, and conotruncal defects and hypoplastic left heart occurred in less than 0.5% [25, 31]. An Ebstein-like malformation of the systemic tricuspid valve occurs in one third of patients with congenitally corrected transposition of the great arteries, but it is an entity distinct from classic Ebstein anomaly [25, 31].

The left heart is also affected. In the consecutive Mayo Clinic series (106 patients examined by echocardiography between 2001 and 2003, 72% with severe forms), 17.9% had non-compaction-like changes in the left ventricular myocardium, 7% systolic dysfunction, 36% (34 of 95) diastolic dysfunction and 4% dilatation; 15% had mitral valve prolapse, 8% bicuspid aortic valve, 4% mitral dysplasia and 8% ventricular septal defect, and 21% had Wolff-Parkinson-White syndrome; overall, 39% had left-sided lesions of the myocardium or valves [40].

3.2. Fetal life

Ebstein anomaly and tricuspid dysplasia diagnosed in fetal life carry high perinatal mortality. In a retrospective study from 23 centres in the United States and Canada (2005–2011), 243 fetuses were diagnosed at 27 ± 6 weeks; there were 15 terminations of pregnancy (6%), 41 fetal deaths (17%, at a median of 30 weeks) and, of the 176 live births, 56 (32%) died before discharge, giving a perinatal mortality of 45% (97 of 215) [56]. The independent predictors at diagnosis were gestational age below 32 weeks (OR 8.6; 95% CI 3.5–21.0), the z-score of the tricuspid annulus diameter (OR 1.3 per unit; 1.1–1.5), pulmonary regurgitation, which indicates circular shunt physiology (OR 2.9; 1.4–6.2), and pericardial effusion (OR 2.5; 1.1–6.0); in those diagnosed before 32 weeks, mortality was 54% [56]. In circular shunting, blood regurgitated through the tricuspid valve passes from the right to the left atrium through the atrial septal defect, goes on to the aorta and returns to the pulmonary artery through the arterial duct and to the right ventricle through the incompetent pulmonary valve, with hypoxaemia and reduced systemic output [25]. Eleven per cent of the fetuses in the multicentre series had a genetic anomaly or syndrome (trisomy 21 in 11, CHARGE syndrome in 2 and 1p36 deletion in 2), and less than one third underwent amniocentesis, with an abnormal result in 20% of those performed [56]. Among live births, a low tricuspid regurgitation jet velocity and absence of antegrade flow across the pulmonary valve (OR 4.5; 1.3–14.2) were associated with neonatal mortality [84].

The 2024 consensus of the American Association for Thoracic Surgery (AATS) describes fetal disease as a spectrum. At the mild end, a minority among prenatally diagnosed cases, apical displacement of the septal leaflet is slight and tricuspid regurgitation minimal, with no consequences before or after birth, but a lesion identified in the first or second trimester may progress to severe forms and warrants surveillance until the end of gestation [36]. At the severe end, which is the majority of prenatal cases, spontaneous fetal death affects 15–20% of continuing pregnancies and neonatal death 25–32% in contemporary series [36]. Cardiomegaly results from volume overload of the right-sided chambers by tricuspid regurgitation and, when massive and early, reduces lung growth (low lung weights at autopsy and reduced lung volumes on fetal ultrasound and magnetic resonance imaging). Functional pulmonary atresia arises when the right ventricle cannot generate the systemic pressure needed to open the pulmonary valve, with retrograde or bidirectional flow in the arterial duct; pulmonary regurgitation with a patent duct “steals” output from the fetal and placental circulation and worsens right ventricular overload [36]. The left ventricle, which would have to sustain the combined output on its own, often has low output due to adverse ventricular-ventricular interaction (dyssynchrony, reduced circumferential contractility and altered diastolic filling), which can reduce cerebral oxygen delivery, with possible neurodevelopmental consequences; the right atrium is also dysfunctional, and atrial flutter, ectopic atrial tachycardia and atrioventricular re-entrant tachycardia may arise, decompensating an already fragile circulation [36]. Fetal hydrops also carries a maternal risk of mirror syndrome [36]. Fetal growth restriction, hydrops, cardiomegaly and a reduced observed/expected lung-to-head ratio and total fetal lung volume on magnetic resonance imaging are potential indicators of poor prognosis [36].

3.3. Neonatal period and infancy

The symptomatic newborn is usually cyanosed; with the physiological rise in pulmonary vascular resistance and a patent arterial duct, the functional right ventricle may fail to generate antegrade flow, creating functional pulmonary atresia, sometimes compounded by anatomical atresia; massive cardiomegaly (a “wall-to-wall” heart) can cause pulmonary hypoplasia through a mass effect, and the dilated right ventricle can compress the left [25]. A cardiothoracic ratio above 0.65 indicates a poor prognosis in the symptomatic newborn [25]. In the 2024 AATS consensus, newborns with severe cardiomegaly, prematurity (less than 30–32 weeks), intrauterine growth restriction, anatomical or functional pulmonary valve atresia, circular shunting, left ventricular dysfunction or cardiogenic shock are considered at high risk of morbidity and mortality (class I, level B-NR) [36]. Unstable newborns with circular shunting usually present in the first hours of life, in cardiogenic shock [36]. Infants present more often with heart failure [25].

The neonatal surgical outcomes cited by the consensus show the severity of this group: in the Society of Thoracic Surgeons database, overall mortality of neonatal surgical treatment was 27.4% and that of primary repair 39.9%; a more recent series of critically ill newborns and infants recorded a mortality of 37.5% in the first year of follow-up, with about two thirds of those operated on undergoing a Starnes procedure; and the European Congenital Heart Surgeons Association described mortality above 50% in the newborn and infant subgroup [36]. Patients who survive beyond the first year generally do well, although they may need further surgery [36].

3.4. Child, adolescent and adult

In the older child and the adult, presentation ranges from an incidentally detected murmur to exercise intolerance, arrhythmias and cyanosis, and some patients with mild forms remain asymptomatic for decades and are diagnosed in the sixth or seventh decade [25, 28, 31]. The typical complications described by the 2020 European guideline are high-grade tricuspid regurgitation, right ventricular dysfunction and failure, liver cirrhosis, brain abscesses, paradoxical and pulmonary embolism, tachyarrhythmias, sudden death and infective endocarditis; the main symptoms are arrhythmias (atrioventricular re-entrant tachycardia is the most common), dyspnoea, fatigue, exercise intolerance, chest pain and peripheral or central cyanosis [31].

Arrhythmias account for much of the morbidity. Wolff-Parkinson-White-type accessory pathways are present in 10–20% of cases, usually around the tricuspid orifice, and about one third of adults with ventricular pre-excitation have multiple pathways, associated with a high risk of sudden death, especially when atrial fibrillation with rapid conduction occurs [25, 28]. The 2022 international consensus on genetic testing in cardiology records that Ebstein anomaly is the defect most associated with Wolff-Parkinson-White syndrome and that at least 10% of patients have one or more accessory pathways [33]. Supraventricular tachycardia predominates in younger patients and atrial fibrillation or flutter in older ones [25]. The 2025 AATS consensus places electrocardiographic pre-excitation at about 15–20% of patients and stresses that the delta wave may be masked by delayed intra-atrial conduction, decremental conduction of the accessory pathway and slow propagation of pre-excitation through the diseased ventricular myocardium; in unoperated children, supraventricular tachycardia is more common than ectopic atrial tachycardia, atrial flutter and atrial fibrillation, and ventricular arrhythmia is rare [35]. In unoperated adults, atrial and ventricular arrhythmias become more frequent and are associated with a worse course, whereas atrioventricular re-entrant tachycardia and pre-excitation are associated with a better course, probably because they lead to earlier assessment; the risk of arrhythmia and of sudden death, attributed to ventricular arrhythmia, increases with age and is greatest in older unoperated adults, and the onset of atrial arrhythmia often precedes ventricular arrhythmia and death [35]. The substrate combines right atrial and ventricular dilatation, right ventricular myopathy resulting from failure of delamination and atrioventricular accessory pathways [35].

The left ventricle weighs more on prognosis than the right. According to the 2025 AATS consensus, although right ventricular dysfunction is frequent, left ventricular dysfunction has a greater influence on outcomes; contributing factors are interventricular septal dyskinesia, intraventricular dyssynchrony, non-compaction and left ventricular fibrosis [35]. Because of septal dyskinesia, M-mode and two-dimensional echocardiography may underestimate systolic dysfunction, and myocardial deformation (strain) provides a more accurate assessment [35]. Severe cardiomegaly, with a cardiothoracic ratio above 0.65 on chest radiography, reflects a greatly dilated right atrium and atrialised ventricle, deteriorating haemodynamics and possible pulmonary hypoplasia, and stretching of the atrial and ventricular myocardium favours arrhythmias [35]. Measured exercise capacity is lower than that reported by patients, who adapt to chronic limitation; children cannot always describe it, and resting oxygen saturation predicts exercise intolerance [35].

3.5. Population-based natural history

The national registries of Denmark and Sweden made it possible to study operated and unoperated patients without the referral bias of surgical series. In the mortality cohort (530 patients born between 1970 and 2017, 5300 matched controls, median follow-up of 11 years), 46% had isolated Ebstein anomaly, 11% had only an associated atrial septal defect, 52% were diagnosed in the first year of life and 43% underwent surgery [27]. Cumulative mortality was 16% at 20 years and 20% at 40 years; at 35 years it was 9% in the isolated form, 16% with atrial septal defect and 39% in complex forms [27]. In patients diagnosed before one year of age, mortality was 16%, 22% and 28% at 1, 5 and 30 years from diagnosis, versus 3% and 10% at 5 and 30 years in those diagnosed after 10 years of age; the subgroup with presumably mild disease (diagnosis after 10 years of age, with or without atrial septal defect) had a mortality of 11% at 35 years, versus 4% in controls (hazard ratio [HR] 6.0; 95% CI 2.7–13.6), and the subgroup with presumably severe disease had an HR of 36.2 (15.5–84.4) [27]. Deaths were cardiac in 85% of unoperated and 86% of operated patients; sudden death was the most frequent cardiac cause in unoperated patients (16%) and heart failure in operated patients (20%) [27]. Mortality improved in those diagnosed from 2000 onwards, partly through detection of milder forms [27].

In the morbidity cohort from the same registries (794 patients born between 1930 and 2017, 7940 controls, median follow-up of 33 years), 56% had the isolated form and 27.5% an associated atrial septal defect; 43% underwent surgery, but 88% retained the native tricuspid valve at the end of follow-up [32]. Compared with controls, arrhythmia occurred in 35.9% (HR 19.7; 16.5–23.5), heart failure in 15.0% (HR 12.0; 9.3–15.5), infective endocarditis in 2.5% (HR 26.8; 11.8–60.8) and ischaemic stroke in 4.0% (HR 5.0; 3.3–7.5) [32]. The cumulative lifetime incidence of supraventricular arrhythmia was about 70% and that of ventricular pre-excitation 19% in patients with the isolated form or with atrial septal defect; ventricular arrhythmia reached about 6% at 60 years, with onset from the third decade [32]. The onset of supraventricular arrhythmia multiplied the risk of ischaemic stroke (HR 22.6; 11.1–45.9) and was followed by heart failure in 18% at ten years [32]. Heart failure by 60 years affected about 6% of isolated forms, 16% of forms with atrial septal defect and 40% of complex forms, versus about 2% in the general population; in the isolated form it was negligible before the fourth decade, and an associated atrial septal defect increased the burden of arrhythmia and heart failure [32].

3.6. Ebstein anomaly with left ventricular non-compaction

In the 2022 systematic review, non-compaction was present in 15–29% of patients with Ebstein anomaly in paediatric or mixed studies and in less than 5% when only adults were considered [3]. In Houston, 10 of 61 infants with Ebstein anomaly (16.4%) had non-compaction, were diagnosed earlier (9 of 10 at birth) and had more progressive left ventricular dysfunction; in Warsaw, 4 of 84 adults (4.8%) had the association, and only these had cardiac arrest or ventricular arrhythmia; in the Japanese series from Toyama, children with non-compaction and congenital heart disease had a lower ejection fraction and worse prognosis than controls with ventricular septal defect [3]. In Brazil, analysis of 259 postnatal hearts from the Heart Institute of the University of São Paulo (InCor) showed that the prevalence of non-compaction in the 23 hearts with Ebstein anomaly was 28.6%, 9.5% or 0% depending on the Chin, Jenni or Petersen criteria, illustrating how the figures depend on the criterion used [3]. Conversely, in non-compaction studies Ebstein anomaly is one of the most frequent congenital heart defects: 11 of 130 patients (8%) in the series compiled in Clinical Cardiogenetics, alongside atrial septal defect (22 of 135; 16%), ventricular septal defect (23 of 218; 11%) and patent arterial duct (16 of 182; 9%), and 78% of 46 children with non-compaction had congenital heart disease [38].

Presentation depends on age: cyanosis in the prenatal period and in infancy, heart failure and murmur in the child, exertional dyspnoea, palpitations, chest pain and arrhythmias in the adolescent and adult, and an incidental finding in the elderly; exertional dyspnoea and the systolic murmur of tricuspid regurgitation are the most reported signs [3]. In a study of cardioverter-defibrillators in congenital heart disease and non-compaction, patients with Ebstein anomaly had the highest rate of lead-related and procedure-related complications [3].

3.7. NKX2-5-related heart disease

The core phenotype is ostium secundum atrial septal defect with atrioventricular block of nodal location, narrow QRS and progressive course; Clinical Cardiogenetics also attributes to it atrial fibrillation, occasional cardiomyopathy, frequent congenital heart disease and a risk of sudden death, with no extracardiac manifestations [34]. In Schott’s families, 27 of the 33 affected individuals had atrial septal defect, 8 had other structural defects (ventricular septal defect, tetralogy, subaortic stenosis, left ventricular hypertrophy, pulmonary atresia and redundant fenestrated mitral leaflets), all those with structural heart disease had an atrioventricular conduction defect, and 14 had a pacemaker [4]. In the review of 208 carriers, atrial septal defect was present in 70% and the combination with conduction disease or arrhythmia in 54%; there were 11 patients with cardiomyopathy (non-compaction, hypertrophy or ventricular dilatation), and the other defects included four tricuspid anomalies (atresia or Ebstein anomaly), interrupted aortic arch, common arterial trunk, congenitally corrected transposition, coarctation, double-outlet right ventricle, anomalous pulmonary venous drainage and heterotaxy [16]. In one of the families of Watanabe and colleagues, the atrial septal defect was of the sinus venosus type in three of four operated patients, and one carrier had polysplenia, a symmetrical midline liver and intestinal malrotation [4].

Atrioventricular block is postnatal and may appear years after surgery: in Benson’s study it was the main manifestation in almost a quarter of carriers, some without a structural defect [30]. A prolonged PR interval may be the only expression of the genotype in relatives [4]. Atrial fibrillation may be late (in one family, 28 years after atrial septal defect surgery) or the only manifestation, and in a three-generation family studied by Gutierrez-Roelens and colleagues one carrier had unexplained ventricular tachycardia on Holter monitoring [4].

Sudden death is the complication that justifies lifelong surveillance. In Ellesøe’s review, 19 sudden deaths before 50 years of age occurred in 9 families with atrial septal defect and conduction disease (corresponding to 8% of confirmed or obligate carriers and 15% including possible carriers), two of them in patients with a functioning pacemaker, which suggests tachyarrhythmia as the mechanism; four of those nine families (44%) also had cardiomyopathy, and there were no sudden deaths in non-familial cases [16]. Reports from 2026 document non-sustained ventricular tachycardia with presyncope in a woman aged 42 years with a truncating variant and almost preserved ventricular function, and aborted ventricular fibrillation at 17 years of age, four years after closure of a simple atrial septal defect, in a mother whose carrier son developed complete block at 7 years and non-sustained ventricular tachycardia at 10 [78, 85]. The Icelandic variant p.(Phe145Leu) is associated with dilated cardiomyopathy and heart failure, in addition to high-grade block and atrial septal defect [26].

Congenital non-goitrous hypothyroidism 5 is the recognised extracardiac manifestation. Congenital hypothyroidism affects 1 in 3000–4000 newborns and results from thyroid dysgenesis in 80–85%; in NKX2-5 screening of 241 patients with dysgenesis, three heterozygous missense variants (R25C, A119S and R161P, historical notation; on the MANE Select transcript NM_004387.4:c.73C>T p.(Arg25Cys), c.355G>T p.(Ala119Ser) and c.482G>C p.(Arg161Pro)) were found in four patients, with reduced DNA binding or transactivation and a dominant-negative effect [42, 68]. The current ClinVar classification is benign or likely benign for p.(Ala119Ser), uncertain significance for p.(Arg161Pro) and conflicting for p.(Arg25Cys), so the functional data from this study are not sufficient to assign causality to any of the three [80, 86]. The two R25C carriers with thyroid ectopy or athyreosis had no heart disease [13]. The GeneGuidance curation describes the thyroid phenotype as usually mild, with most carriers euthyroid and a minority with hypothyroidism and a hypoplastic thyroid [6].

4. Clinical Recognition and Warning Signs

4.1. When to suspect Ebstein anomaly

In fetal life, Ebstein anomaly is suspected when obstetric ultrasound shows cardiomegaly with an enlarged right atrium and severe tricuspid regurgitation; the signs of high risk of intrauterine death and of postnatal morbidity and mortality listed by the 2024 AATS consensus are severe cardiomegaly early in gestation, bidirectional or retrograde shunting in the arterial duct, anatomical or physiological pulmonary valve atresia, circular shunting (pulmonary regurgitation), left ventricular dysfunction, low right ventricular systolic pressure (below 20–25 mmHg, estimated from a low-velocity tricuspid regurgitation jet) and hydrops (class I, level B-NR) [36]. Pulmonary regurgitation and pericardial effusion are the signs that most increase the risk of death in the North American multicentre study [56]. In the newborn, cyanosis with massive radiological cardiomegaly and a “box-shaped” heart should prompt urgent echocardiography [25].

In the child and the adult, the clinical clues are the systolic murmur of tricuspid regurgitation, wide splitting of the first and second heart sounds, multiple systolic clicks, third and fourth heart sounds, cyanosis and hepatomegaly; on the electrocardiogram, right atrial hypertrophy, prolonged PR, right bundle branch block with fragmented QRS, deep Q waves in II, III, aVF and V1–V4, low voltage, ventricular pre-excitation and tachyarrhythmias [31]. A supraventricular tachycardia, especially with pre-excitation, in a young person, a paradoxical embolism or brain abscess in a patient with unexplained mild cyanosis, or right heart failure with hepatic congestion should raise the possibility of undiagnosed Ebstein anomaly [28, 31]. The European guideline notes that unrecognised Ebstein anomaly may underlie long-standing hepatic congestion with liver failure [31].

4.2. When to suspect a genetic cause in Ebstein anomaly

The factors that increase the likelihood of an identifiable genetic cause are coexisting left ventricular non-compaction, a family history of Ebstein anomaly, other congenital heart defects or cardiomyopathy (hypertrophic, dilated or non-compaction), extracardiac anomalies, intellectual disability or developmental delay, dysmorphic features and conduction disease in relatives [7, 9, 12, 33, 46]. Coexisting Ebstein anomaly and non-compaction should prompt testing of MYH7 and TPM1; a family with Ebstein anomaly in several males and mild skeletal signs suggests FLNA [9, 11]. Other allelic FLNA valvular disorders are part of the same assessment: in X-linked myxomatous valvular dystrophy, males have severe polyvalvular disease, from the neonatal period to 40 years of age, with a lifetime risk of valve surgery of about 75%, and heterozygous females have milder manifestations [87]; in otopalatodigital syndrome type 2, affected boys usually die in the perinatal period from pulmonary hypoplasia or heart defects, including tricuspid valve defects [88]. Maternal lithium exposure in the first trimester does not exclude a genetic cause, because the attributable absolute risk is low [24].

4.3. When to suspect NKX2-5-related heart disease

Suspicion arises with a familial atrial septal defect, an atrial septal defect with a prolonged PR interval or atrioventricular block, unexplained atrioventricular block in a young person, or a family history of atrial septal defect, pacemaker implantation at a young age or sudden death [4, 16, 34]. Clinical Cardiogenetics recommends considering a genetic cause and genetic testing in conduction disease with a family history of conduction disease, cardiomyopathy or congenital heart disease, and in patients under 50 years of age with severe, unexplained conduction disease [34]. A repaired atrial septal defect does not close the matter: recent reports recommend routine electrocardiograms even after repair of a simple atrial septal defect, and NKX2-5 testing when progressive conduction disease appears [85].

4.4. Red flags

Urgent signs in Ebstein anomaly are syncope or presyncope with known pre-excitation (possibility of pre-excited atrial fibrillation), new-onset atrial fibrillation or flutter, progressive cyanosis, signs of right heart failure, paradoxical embolism and persistent fever in a patient with a tricuspid prosthesis or residual shunt [28, 31, 32]. In a carrier of an NKX2-5 variant, syncope, presyncope, exercise intolerance and palpitations should prompt immediate assessment of conduction and a search for ventricular arrhythmia, even if the patient has a pacemaker [16, 34, 85]. In pregnancy, signs of heart failure before conception are the only independent predictor of maternal cardiac events in the European registry of pregnancy and cardiac disease [89].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Echocardiography and morphological criteria

The diagnosis of Ebstein anomaly is echocardiographic. The most used criterion is apical displacement of the septal (or posterior) leaflet of at least 8 mm/m² of body surface area relative to the crux of the heart (0.8 cm/m² in adults, according to the European guideline), together with an elongated, redundant anterior leaflet; without these features, tricuspid dysplasia should be considered [25, 31]. Echocardiography describes the anatomy and function of the valve, the tethering of the septal and posterior leaflets, the size of the anterior leaflet, the dimensions and function of the atrium, the atrialised ventricle, the functional right ventricle and the left ventricle, right ventricular outflow tract obstruction and associated lesions [31]. Two-dimensional or three-dimensional transoesophageal echocardiography is useful for surgical planning when transthoracic images are insufficient (class IIa, level B-NR, in the North American guideline) [28].

In the newborn, the Celermajer index, also called the Great Ormond Street score (GOSE), divides the sum of the areas of the right atrium and atrialised ventricle by the sum of the areas of the functional right ventricle and the left-sided chambers, in the four-chamber view at end-diastole [25]. The table summarises the classifications in use.

ACHD AP, anatomical and physiological classification of adult congenital heart disease; NYHA, New York Heart Association.

5.2. Magnetic resonance imaging, electrophysiology and exercise testing

Cardiac magnetic resonance imaging has prognostic value and is useful before and after surgery, because it quantifies the dilated right heart, right ventricular function and tricuspid function without window limitations; the North American guideline gives it class IIa (level B-NR) to define right ventricular anatomy, dimensions and systolic function, and the 2025 AATS consensus recommends it, whenever feasible, to measure the volumes of both ventricles, the tricuspid regurgitant fraction and right ventricular stroke volume (class I, level B-NR) [28, 31, 35]. Magnetic resonance imaging with cine and phase-contrast velocity mapping is the reference for volumes and flows, but applying published values requires caution, because imaging planes (axial or short axis), delineation between the functional and atrialised portions of the right ventricle (valve insertion or leaflets) and the method of quantifying tricuspid regurgitation (direct or indirect) vary between centres, with no reference standard to validate accuracy [35]. In the young child, the benefit of magnetic resonance imaging is weighed against the risk of general anaesthesia, especially if haemodynamics are unstable [35]. Magnetic resonance imaging also assesses diffuse or focal fibrosis, deformation and left ventricular ejection fraction, and late gadolinium enhancement helps distinguish ischaemic fibrosis and infiltrative or inflammatory cardiomyopathies when there is left-sided dysfunction [35]. In the subgroup with non-compaction, magnetic resonance imaging is preferred to characterise the left ventricle [3].

An electrophysiological study, with ablation if needed, is reasonable before tricuspid surgery even without pre-excitation or tachycardia, because surgery subsequently hinders catheter access to right-sided pathways and the slow nodal pathway (class IIa, level B-NR), and is useful in pre-excitation without tachycardia (IIa, B-NR); in the European guideline, symptomatic arrhythmias or pre-excitation on the electrocardiogram indicate an electrophysiological study followed by ablation or surgical treatment of the arrhythmias if surgery is planned (class I, level C) [28, 31]. The 2025 AATS consensus indicates preoperative arrhythmia assessment in all patients, with a history focused on arrhythmias and suggestive symptoms, electrocardiography and ambulatory rhythm monitoring (class I, level B-NR), and a preoperative electrophysiological study in older children and adults with documented supraventricular or ventricular tachycardia, pre-excitation or a history suggestive of arrhythmia (class IIa, level C-LD); it considers it reasonable to extend this to all older candidates for tricuspid surgery, especially with severe disease, because some concealed pathways, nodal re-entry and atrial or ventricular arrhythmias only become manifest after surgery [35]. In children under 15 kg, catheter ablation is controversial and of higher risk, so if the arrhythmia is controlled with drugs, the study may be deferred [35].

Asymptomatic patients should undergo cardiopulmonary exercise testing with measurement of oxygen consumption, to unmask hidden exercise intolerance (class I, level B-NR) [35]. Patients with Ebstein anomaly have lower peak oxygen consumption, shorter exercise duration and greater respiratory effort than healthy individuals, with a limitation that is mainly cardiac and also respiratory; testing makes it possible to follow the trajectory of functional capacity over time, assess risk before pregnancy and measure the change after a surgical or percutaneous intervention [35]. Chest radiography tracks changes in heart size [31]; a cardiothoracic ratio above 0.65 defines severe cardiomegaly [35]. In adults and in patients with reduced left ventricular function who are candidates for a bidirectional cavopulmonary shunt, preoperative cardiac catheterisation is essential to assess haemodynamic eligibility (left ventricular end-diastolic pressure below 12 mmHg, transpulmonary gradient below 10 mmHg and mean pulmonary artery pressure below 18 mmHg) [35].

5.3. Diagnosis of NKX2-5-related heart disease

There is no non-molecular test that confirms the disease: electrocardiography and echocardiography define the phenotype, but the same phenotype may result from TBX5 or other genes; confirmation is the identification of a heterozygous pathogenic NKX2-5 variant or a 5q35 deletion that includes the gene [6]. The initial assessment includes a 12-lead electrocardiogram (first-degree atrioventricular block and its progression, bundle branch or fascicular blocks), Holter monitoring, echocardiography (septal defect, right ventricular volume overload, non-compaction, ventricular function), magnetic resonance imaging when the myocardium needs better characterisation, and thyroid function [6]. In carriers with intermittent symptoms, Holter monitoring and event recorders detect intermittent block, and exercise testing detects chronotropic incompetence or exercise-induced block [34].

5.4. Genetic testing: indications and strategy

The 2022 consensus of the European Heart Rhythm Association, Heart Rhythm Society, Asia Pacific Heart Rhythm Society and Latin American Heart Rhythm Society (EHRA/HRS/APHRS/LAHRS) organises genetic testing in congenital heart disease by category [33]. In newborns and infants who need investigation or procedures, chromosomal microarray or low-coverage sequencing for CNVs should be performed, and exome or genome sequencing may be considered, with low yield in sporadic cases; in patients with extracardiac anomalies (about 20% of congenital heart defects), microarray should be followed by trio exome or genome sequencing if non-diagnostic, with a diagnosis in 25–40%; in familial forms (one or more affected first-degree relatives), exome sequencing identifies monogenic variants in about 10% when there are two affected individuals and in more when there are three or more, and a gene suggested by the type of defect may be analysed first; in sporadic non-syndromic forms, less than 5% have a monogenic variant and routine testing is not justified outside research [33]. The same document recommends targeted genetic testing in conduction disease, isolated or with structural heart disease, when diagnosis is early or there is a family history, and lists NKX2.5 among the high-confidence genes for familial atrial septal defect with or without atrioventricular block [33].

5.5. Test limitations and interpretation

A negative result does not exclude a genetic cause. Congenital heart disease panels (usually 10–40 genes) have a low yield and are being replaced by exome or genome sequencing, which allow reanalysis [33]. ClinGen had no curation of congenital heart disease genes at the time of the 2022 consensus, and interpretation is hampered by isolated patients with no possibility of segregation analysis and by reduced penetrance and variable expressivity within families [33]. In Ebstein anomaly, the relationship with MYH7 is “limited” in ClinGen, and interpretation of MYH7 and TPM1 variants should rest on the coexisting myocardial phenotype [7, 64]. CNV studies identify candidate changes in almost half of isolated cases, but most have no established pathogenicity [44, 45]. For NKX2-5, haploinsufficiency is established and deletions of the gene are causal, but variants such as p.(Arg25Cys) have conflicting classifications and low penetrance, and the same protein change may correspond to nucleotide alleles with different classifications [6, 80, 81].

Recent examples show how frequent variants of uncertain significance are in this area: the MYH7 variant p.(Glu1220del), which segregated in a family with Ebstein anomaly, was classified as of uncertain significance by a ClinVar expert panel in November 2025, and the MYH7 variant p.(Tyr283Cys) from the Mayo Clinic case published in 2026 was classified as a VUS (criteria PM1, PM2 and PP3) despite being absent from population databases and predicted deleterious [39, 70]. The report should state the gene, the transcript with version (NM_004387.4 for NKX2-5, NM_000257.4 for MYH7, NM_001018005.2 for TPM1), the normalised HGVS description, zygosity and classification according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) [33, 68]. A VUS should not guide predictive testing of relatives; segregation in the family and parental testing may reclassify it [33, 39].

5.6. Prenatal diagnosis

Fetal echocardiography allows prenatal diagnosis of complex heart defects, with scans at 14, 19 and 22 weeks in at-risk families, and should be offered to women with congenital heart disease who are planning pregnancy [90]. Once the anomaly is detected, microarray on amniocentesis or chorionic villus samples identifies chromosomal anomalies in 10–15% of fetuses with heart defects, and trio exome sequencing adds 5–12% [33]. In the multicentre series of fetuses with Ebstein anomaly or tricuspid dysplasia, less than one third underwent amniocentesis and 20% of results were abnormal [56]. Once the diagnosis is made, fetal assessment estimates right ventricular systolic pressure from the tricuspid regurgitation jet velocity, looks for antegrade flow across the pulmonary valve, pulmonary regurgitation and the direction of flow in the arterial duct, and evaluates left ventricular function, cardiomegaly and hydrops, which define risk according to the 2024 AATS consensus; the presence of pulmonary regurgitation indicates that the pulmonary valve is patent [36]. When there is a known familial pathogenic variant (MYH7, TPM1, NKX2-5, FLNA), prenatal diagnosis and preimplantation genetic testing are technically possible; their discussion follows the general principles for inherited heart diseases, with counselling on variable penetrance [33, 91]. Chorionic villus sampling is performed between 10 and 13 weeks and amniocentesis from 15 weeks, with an estimated miscarriage risk of 0.5% and 0.3%, respectively; at the time of the Clinical Cardiogenetics chapter (2020), non-invasive prenatal diagnosis by cell-free fetal DNA in maternal plasma was not available for inherited heart diseases [91].

6. Differential Diagnosis

6.1. Morphological differential diagnosis of Ebstein anomaly

Tricuspid valve dysplasia shares the haemodynamic consequence (severe tricuspid regurgitation) and, in fetal life, the same prognosis, but the leaflets are deformed without significant apical displacement; the correlation between echocardiography and pathological anatomy is imperfect, and the fetal multicentre study did not separate them in the analysis [25, 56]. The Ebstein-like malformation of the systemic tricuspid valve in congenitally corrected transposition is another entity, with systemic right ventricle physiology [25, 31]. In the cyanosed newborn with cardiomegaly, the differential diagnosis includes pulmonary atresia with intact ventricular septum (hypoplastic and hypertrophied right ventricle) and tricuspid regurgitation with a normal valve due to cardiomyopathy or premature closure of the arterial duct, exclusion criteria of the fetal multicentre study [56].

6.2. Aetiological and genetic differential diagnosis

Recognising a syndromic form changes counselling and extracardiac surveillance. The 1p36 deletion is the chromosomal cause most associated with Ebstein anomaly and other tricuspid defects, and the Oxford Desk Reference recommends considering it in Ebstein anomaly with dysmorphic features, developmental delay or other congenital anomalies [54, 90]. Lithium exposure should be sought in the history of every case of Ebstein anomaly [90]. When the dominant finding is left ventricular non-compaction in a boy in the first year of life, with neutropenia, muscle weakness or growth faltering, Barth syndrome (TAZ, X-linked) enters the differential diagnosis of the cardiomyopathy, with urinary 3-methylglutaconic acid often elevated (normal in some patients) and the monolysocardiolipin/cardiolipin ratio as screening; confirmation is molecular [92].

AD, autosomal dominant; AR, autosomal recessive; AVB, atrioventricular block; ASD, atrial septal defect; VSD, ventricular septal defect; AVSD, atrioventricular septal defect; ECG, electrocardiogram; LV, left ventricle.

6.3. Differential diagnosis of NKX2-5-related heart disease

The coexistence of congenital heart disease and conduction disease suggests NKX2-5 or TBX5; the latter is invariably associated with skeletal anomalies of the upper limb (Holt-Oram syndrome), sometimes visible only on radiography [34]. In Holt-Oram syndrome, penetrance is 100%, atrial septal defect occurs in 34%, ventricular septal defect in 25% and isolated electrocardiographic changes in 39%, with heart defects (including minimal electrocardiographic changes) in 95% [90]. Familial atrial septal defect without conduction disease may be due to GATA4 [33]. When conduction disease is accompanied by dilated cardiomyopathy, LMNA and DES are the main hypotheses, and pre-excitation or unexplained hypertrophy suggests PRKAG2 [34]. Conduction disease, very common in the elderly, is almost always degenerative; the prevalence of bundle branch block rose from 1% at 50 years to 17% at 80 years in a prospective cohort of 855 men, and that of pacemakers is below 0.5% before 65 years and above 2% after 75 years, which makes a diagnosis of hereditary disease more likely in the young [34].

CHARGE syndrome and trisomy 21 were the most frequent genetic anomalies, together with the 1p36 deletion, in fetuses with Ebstein anomaly or tricuspid dysplasia in the North American multicentre series [56].

7. Treatments and Risk Management Strategies

7.1. Principles

There is no pharmacological therapy targeting the cause of Ebstein anomaly or of NKX2-5-related heart disease; treatment is haemodynamic, surgical, electrophysiological and preventive, and is decided by symptoms, anatomy and the course of ventricular function [6, 28, 31]. The European guideline recommends that surgery be performed by a congenital heart surgeon with specific experience in Ebstein anomaly (class I, level C) and that follow-up take place in specialised adult congenital heart disease centres [31]. The guidelines and consensus documents mainly cover the adult (ESC 2020, AHA/ACC 2018), the newborn and infant (AATS 2024) and the child over one year of age and the adult (AATS 2025) [28, 31, 35, 36]. The two AATS consensus documents were drafted by international panels of paediatric cardiothoracic surgeons and cardiologists (12 surgeons and 2 cardiologists in 2024; 11 surgeons and 6 cardiologists in 2025), based on a literature search in four databases restricted to English-language articles published since 2000 (455 results, 71 on newborns and infants), and the statements were approved by a modified Delphi method, with votes from at least 80% of members and agreement of at least 75% on each statement [35, 36]. The class of recommendation and level of evidence follow the American College of Cardiology and American Heart Association system (classes I, IIa, IIb and III; levels A, B-R, B-NR, C-LD and C-EO), and questions without sufficient evidence for a formal recommendation were left as good practice statements [35, 36]. Because of the rarity of the disease and the absence of randomised trials, most recommendations rest on expert opinion and on non-randomised institutional series of moderate size, which the authors themselves identify as a limitation [35, 36].

7.2. Fetus

There is no established fetal therapy. Constriction of the arterial duct with transplacental non-steroidal anti-inflammatory drugs (NSAIDs) has been proposed to interrupt circular shunting: in a series from 10 centres (2010–2018), 21 pregnant women with fetuses with Ebstein anomaly or tricuspid dysplasia and pulmonary regurgitation were counselled, 15 received indometacin and then, in most cases, ibuprofen, and 12 of 15 (80%) achieved ductal constriction after a median of 2 days, with improvement in pulmonary regurgitation in all; 11 of 12 (92%) with constriction were born alive, versus 4 of 9 (44%) without treatment or without constriction [93]. A meta-analysis of 21 studies (610 fetuses with circular shunting, only 27 treated with NSAIDs) found ductal constriction in 81% and oligohydramnios in 65% of those treated, with live births in 86% versus 67% and survival to discharge in 89% versus 43% with expectant management, stressing the small numbers and limited safety data [94]. This is experimental, off-label use, which the sources do not allow to be converted into a recommendation or a dosing regimen. The 2024 AATS consensus describes the initial experience with indometacin (a non-selective cyclo-oxygenase inhibitor) in fetuses with significant pulmonary regurgitation, especially at non-viable gestational ages and with severe placental steal, as promising, but stresses the unresolved risks: acute fetal renal vasoconstriction, effects on labour and delivery, and severe constriction or irreversible closure of the arterial duct in utero, which can cause acute decompensation at birth and requires delivery planning with immediate medical and surgical support [36].

Fetal surveillance should be frequent in the third trimester, because these fetuses can decompensate rapidly, and prematurity worsens the neonatal prognosis [56]. The 2024 AATS consensus recommends serial fetal echocardiography every 2–4 weeks before 32 weeks and every 1–2 weeks after 32 weeks, because most high-risk signs appear late (class I, level B-NR); obstetric surveillance may include non-stress cardiotocography, magnetic resonance imaging, biophysical profile and biometry [36]. Any sign of decompensation in a third-trimester fetus should prompt discussion of urgent delivery, if neonatal treatment can improve haemodynamics, and anticipated delivery before 34 weeks indicates maternal corticosteroids for fetal lung maturation [36]. The corticosteroid regimen is not specified by the consensus; the 2022 World Health Organization recommendation for pregnant women with a high likelihood of preterm birth within the next 7 days, between 24 and 34 weeks, is intramuscular dexamethasone or betamethasone for a total of 24 mg in divided doses (in the trials, four doses of 6 mg of dexamethasone 12 hours apart or two doses of 12 mg of betamethasone 24 hours apart), with a single repeat course if the first was given at least 7 days earlier and the risk of birth within the next 7 days persists [95]. Fetuses with high-risk signs should be born in a planned delivery at a centre with intensive care, extracorporeal membrane oxygenation (ECMO) and cardiac surgery (class I, level C-EO), and delivery in most pregnancies with Ebstein anomaly should take place at a tertiary or quaternary centre with cardiology and cardiac surgery [36]. As a good practice statement, the consensus calls for regular and complete fetal and neonatal echocardiography when a severe form is suspected and a multidisciplinary team (paediatric cardiothoracic surgery, paediatric cardiology and paediatric cardiac intensive care) available at the delivery of high-risk newborns [36]. At birth, removal of the low-resistance placental circulation increases the afterload of both ventricles, and the fall in pulmonary resistance, accentuated by mechanical ventilation and pulmonary vasodilators, increases pulmonary venous return and left atrial pressure, which can compromise the obligatory right-to-left atrial shunt on which left ventricular preload depends [36].

7.3. Newborn and infant

The 2024 AATS consensus establishes risk stratification as the initial step and defines a newborn as aged under 1 month and an infant as aged under 1 year [36]. It distinguishes two physiologies: circular shunting, in which pulmonary regurgitation makes blood pass from the aorta to the pulmonary artery through the arterial duct, from there to the right ventricle, to the right atrium through tricuspid regurgitation, to the left atrium through the foramen ovale and back to the aorta, with a variable shunt fraction and haemodynamics ranging from stability to cardiogenic shock; and functional pulmonary atresia, with absent or minimal antegrade flow and at most mild pulmonary regurgitation (grades 1–2 of 4), a subgroup distinct from haemodynamically significant circular shunting [36]. Comfort palliative care refers to support without intervention, and a palliative procedure, surgical or catheter-based, is one that falls short of complete biventricular repair and aims to ensure survival [36].

In the stable newborn, the aim is to avoid neonatal surgery. Those without high-risk signs, and high-risk newborns who stabilise, should be observed in a monitored setting awaiting spontaneous closure of the arterial duct (class I, level C-LD): if there is no anatomical pulmonary atresia, closure increases antegrade pulmonary flow, and the fall in pulmonary pressures may reduce tricuspid and pulmonary regurgitation; supplemental oxygen or pulmonary vasodilators, started after the duct begins to constrict and not before, may favour right ventricular ejection, and excessive invasive monitoring, including umbilical arterial and venous catheters, adds morbidity (aortic and renal artery thrombosis) and should be avoided in clinically well newborns [36]. In newborns with functional pulmonary atresia and normal right ventricular systolic pressure (above 20–25 mmHg, about one third to one half of systemic pressure), medical closure of the duct should be attempted, and may be repeated within the first 2 weeks of life if the first attempt fails (class I, level C-LD); if the pulmonary valve is patent but the duct does not close spontaneously, cyclo-oxygenase inhibitors or surgical closure should be considered, and, if there is doubt about pulmonary valve patency, balloon occlusion of the duct in the catheterisation laboratory tests patency and tolerance of closure [36]. Stable newborns with significant pulmonary regurgitation, at risk of developing circular shunting and with right ventricular systolic pressure above 20–25 mmHg, should undergo an attempt at medical closure of the duct (class I, level C-LD); a long wait may, however, worsen haemodynamics and organ failure, and those with lower right ventricular pressure and a greatly enlarged heart should be considered for the Starnes procedure, which has already been performed successfully in preterm infants [36]. Most children with right ventricular systolic pressure above 20–25 mmHg benefit from biventricular repair [36]. In stable newborns with insufficient antegrade pulmonary flow (dependence on prostaglandin or ventilation, arterial oxygen saturation below 80%), a ductal stent or Blalock-Taussig shunt may be considered as the first intervention, as an alternative to primary biventricular repair (class IIa, level C-LD), deferring definitive repair to 3–5 months; simplified primary biventricular repair is an alternative, with an expected mortality of about 10%, similar to that of the shunt or stent, provided the fragile tricuspid leaflets are not manipulated (“cut and sew”) in the early neonatal period, when pulmonary resistance is still high [36].

In the unstable newborn, the priority is to interrupt circular shunting. Haemodynamically unstable newborns with circular shunting should undergo emergency interruption of the shunt (class I, level B-NR), preferably with the Starnes procedure if an operating theatre and cardiopulmonary bypass are immediately available; if patient or hospital logistics preclude immediate intervention, ligation of the main pulmonary artery with bilateral banding of the pulmonary branches, maintaining oxygenation through the duct, can be life-saving, but requires more definitive palliation within 2–3 days, because of progressive dynamic left ventricular outflow tract obstruction caused by the undecompressed right ventricle; the alternative is ECMO with ductal occlusion, followed by definitive biventricular repair or Starnes palliation [36]. Newborns with circular shunting, haemodynamic instability and low right ventricular systolic pressure (below 20–25 mmHg) should undergo ligation or occlusion of the main pulmonary artery with a Starnes procedure (class I, level C-EO), and newborns in refractory cardiogenic shock (on inotropes, ventilation and prostaglandin for duct-dependent pulmonary flow) should be palliated with the Starnes procedure (class I, level C-LD) [36]. The Starnes procedure, with a predicted mortality of 10–20%, is the most reproducible palliation, improves left ventricular dysfunction and dyssynchrony early (which ductal closure cannot achieve) and has lower mortality than initial neonatal biventricular repair; complete atrioventricular block, described in 22%, can be drastically reduced or eliminated by suturing the ventricular inlet occlusion patch to the tendon of Todaro, leaving the coronary sinus on the atrial or ventricular side [36]. Severely symptomatic newborns with anatomical pulmonary atresia should undergo the Starnes procedure as the first intervention; the exception is a small subgroup with a right ventricle of adequate size and function (estimated systolic pressure above 25–30 mmHg), in whom a simple biventricular repair with a valved right ventricle-to-pulmonary artery conduit may be reasonable [36]. A ductal stent or Blalock-Taussig shunt is less suitable palliation in the unstable patient, because it does not decompress the right ventricle and may cause dynamic left ventricular outflow obstruction and low output; mild to moderate pulmonary stenosis can be treated with open pulmonary valvotomy and simple biventricular repair [36]. The Knott-Craig simplified biventricular repair gave, in a large series of newborns and small infants without anatomical pulmonary atresia, early and late survival of 90%, results seldom reproduced at other centres; the da Silva cone repair in the newborn has been described sporadically, even with circular shunting, and the consensus reserves it for older infants after Starnes palliation, at centres and by surgeons experienced in this technique [36].

Comfort palliative care may be reasonable in newborns with significant associated lesions, such as marked prematurity (less than 32 weeks), genetic syndromes, major medical comorbidities, severe left ventricular dysfunction or sepsis (class II, level C-EO, without subclass in the document) [36]. Exceptions to these criteria are newborns with perimembranous ventricular septal defect, atrioventricular canal or left ventricular non-compaction associated with a large ventricular septal defect, who have been successfully repaired by the biventricular route in the neonatal period or early life [36].

After Starnes palliation, subsequent procedures should pursue staging towards a single-ventricle circulation or conversion to biventricular repair (class I, level C-EO), and all patients operated on with Starnes should be assessed for future biventricular or “one-and-a-half ventricle” repair [36]. Whenever possible, biventricular repair of the symptomatic newborn, with or without prior Starnes, should be deferred to 3–4 months, when a bidirectional cavopulmonary shunt that offloads the right ventricle can safely be added; at 3–5 months, the Knott-Craig simplified tricuspid repair is reproducible and of low mortality, and cone repair can be performed in experienced hands [36]. Patients with Starnes should undergo the second single-ventricle stage at around 3–5 months or be converted to biventricular repair, with takedown of the Starnes, before right ventricular involution [36]. The Starnes procedure modified to facilitate conversion preserves the native pulmonary valve with an intraluminal patch in the pulmonary trunk above the valve and places the fenestrated polytetrafluoroethylene ventricular inlet patch away from the tricuspid leaflets, sutured along the tendon of Todaro; leaving the coronary sinus in the right atrium favours involution of the ventricle (useful on the single-ventricle pathway), and leaving it on the right ventricular side favours its growth for future biventricular repair [36]. If the ventricle appears small when the infant outgrows the original systemic-to-pulmonary shunt, the intraluminal pulmonary patch is fenestrated at the time of the cavopulmonary shunt, to load and “rehabilitate” the ventricle and achieve a one-and-a-half ventricle or biventricular circulation [36]. At conversion, the atrial septal defect is closed with a patch with a central 4 mm fenestration, plication of the atrialised ventricle is often unnecessary in the infant, “one-and-a-half ventricle” repair is an alternative when the right ventricle is borderline, and right reduction atrioplasty is performed routinely when the atrium is dilated [36]. The valve technique may be monocusp repair (with simplified annular reduction, without manipulation of the conduction system area and with ready recourse to the Sebening stitch, which approximates the anterior papillary muscle to the septum), more reproducible in the newborn or for less experienced surgeons, or, preferably, cone repair, more anatomical but with a steeper learning curve and a risk of dehiscence of the fragile leaflets at this age [36]. After Starnes, clinical and imaging surveillance takes place every 2–4 weeks, with routine transthoracic echocardiography and magnetic resonance imaging at 3–4 months to assess right ventricular dimensions and function objectively; the indications, criteria and timing of biventricular conversion remain to be defined [36].

Surgical reviews describe initial medical support (oxygen and prostaglandin in stable patients; intubation, ventilation, prostaglandin, inhaled nitric oxide and inotropes in unstable patients; extracorporeal membrane oxygenation in some), trial withdrawal of prostaglandin in functional pulmonary atresia and its early withdrawal in circular shunting, which in one series reduced mortality once anatomical obstruction had been excluded [25]. Indications for neonatal surgery include anatomical pulmonary atresia, circular shunting, dependence on prostaglandin or ventilation, progressive cyanosis or heart failure and severe forms (Carpentier C or D, GOSE above 3) [25]. The biventricular route (Knott-Craig monocusp valvuloplasty or cone repair) applies when the anterior leaflet is large and mobile and the functional ventricle is reasonable; the single-ventricle route (modified Starnes procedure with a 4 mm fenestrated patch, or Sano right ventricular exclusion, with a Blalock-Taussig shunt, unrestricted atrial septum and reduction atrioplasty) applies when the functional ventricle is very small, the tricuspid regurgitation jet is below 3 m/s and right ventricular systolic pressure is below 30 mmHg [25]. In the Knott-Craig series (32 symptomatic newborns, 60% with pulmonary atresia), early mortality was 28% (40% with atresia versus 8.3% without atresia), fell from 44% to 16% with use of a right ventricle-to-pulmonary artery conduit, and 15-year survival was 74% [25]. Heart transplantation is the option in newborns with associated left ventricular dysfunction or hypoplastic left heart [25].

The Ebstein anomaly sources do not give doses for neonatal supportive drugs; the reference doses are those of the official labels, in the general indications that cover these patients (duct-dependent congenital heart disease; neonatal or perioperative pulmonary hypertension), and not doses validated specifically in Ebstein anomaly. The contraindication of inhaled nitric oxide in newborns dependent on right-to-left shunting is relevant, because some newborns with severe Ebstein anomaly depend on the right-to-left atrial shunt.

Use in Ebstein anomaly outside the named label indications of nitric oxide when there is no documented pulmonary hypertension; µg, microgram; ppm, parts per million.

7.4. Child over one year of age and adult: indications for intervention

The three sources of recommendations converge on the main indications and differ in the strength assigned to intervention in the asymptomatic patient, in the role of isolated atrial septal defect closure and in the class of electrophysiological study and magnetic resonance imaging. The table brings them together; the ESC and AHA/ACC guidelines apply to the adult, and the 2025 AATS consensus to the child over one year of age and the adult.

Sources: ESC 2020 [31]; AHA/ACC 2018 [28]; AATS 2025 [35]. ECG, electrocardiogram; RVEF, right ventricular ejection fraction; LVEF, left ventricular ejection fraction; MRI, magnetic resonance imaging; SaO2, arterial oxygen saturation; RV, right ventricle; LV, left ventricle.

The North American rationale for operating before dysfunction is that deferring surgery until heart failure or right ventricular systolic dysfunction is associated with worse outcomes, and that the myopathic right ventricle of Ebstein anomaly tolerates volume overload less well than other overloaded right ventricles [28]. For the child, a surgical review proposes intervening between 2 and 5 years of age when the probability of durable repair is high, to avoid progressive dilatation of the atrialised ventricle, and lists as indications increasing cardiomegaly, increasing cyanosis, exercise intolerance and fatigue, paradoxical embolism and new arrhythmias [25]. The European guideline identifies the selection of asymptomatic patients with severe tricuspid regurgitation who benefit from surgery as a gap in evidence [31].

7.5. Recommendations of the 2024 and 2025 AATS consensus documents

The two AATS consensus documents are, for the newborn and infant, the only source of graded recommendations and, for the child over one year of age, the only one that grades surgical indications; in the adult they overlap with the ESC and AHA/ACC guidelines. The table reproduces, in paraphrase, all the graded recommendations and the good practice statement of the two documents, with the class and level assigned.

Sources: AATS 2024 [36]; AATS 2025 [35]. Level C without subdivision in the two anticoagulation recommendations, as published. ECG, electrocardiogram; ECMO, extracorporeal membrane oxygenation; RVEF, right ventricular ejection fraction; LVEF, left ventricular ejection fraction; RVSP, right ventricular systolic pressure; RV, right ventricle; LV, left ventricle.

The divergences from the European and North American guidelines are of four kinds. The first is the threshold for intervention in the asymptomatic patient: the AATS consensus accepts surgery (IIa, C-LD) with severe tricuspid regurgitation and only moderate right ventricular dilatation, provided the anatomy allows repair, and proposes an age of 3–5 years in the child; the ESC requires progressive right-sided chamber dilatation or a fall in right ventricular systolic function (IIa, C), and the AHA/ACC progressive dilatation, desaturation, paradoxical embolism or atrial tachyarrhythmias (IIa, B-NR) [28, 31, 35]. The AATS justifies earlier intervention by the durability of cone repair and the better biventricular remodelling when surgery is earlier, but acknowledges that the ideal timing in the asymptomatic patient remains under debate and that some experts advocate repair between 6 and 18 months; an earlier surgical review proposed 2–5 years [25, 35]. In symptomatic patients, the three sources agree on class I, with level of evidence C in the ESC and B-NR in the AHA/ACC and AATS [28, 31, 35].

The second divergence is isolated percutaneous closure of the atrial septal defect. The ESC accepts it after paradoxical embolism (IIa, C) and considers it when cyanosis is the main problem (IIb, C), always after showing that right atrial pressure does not rise and output does not fall; the AATS considers it not recommended with severe tricuspid regurgitation or a severe anatomical form (III, C-EO) and advises avoiding it in cyanosis, which it regards as a classic indication for complete surgical repair, because isolated closure does not correct the regurgitation, may accelerate right ventricular dilatation and lower output on exertion; it accepts it only with mild to moderate tricuspid regurgitation [31, 35]. In practice, the two positions agree in requiring prior haemodynamic assessment and diverge in the cyanotic patient with severe tricuspid regurgitation, in whom the AATS favours surgery.

The third divergence is the strength assigned to investigations. The AATS gives class I to magnetic resonance imaging, which the AHA/ACC classifies as IIa and the ESC describes without a class, and class I to cardiopulmonary testing in all asymptomatic patients; for the electrophysiological study, the ESC assigns class I to electrophysiological assessment and ablation with symptomatic arrhythmias or pre-excitation, whereas the AATS assigns class IIa (C-LD) to the preoperative study in patients with documented tachycardia or pre-excitation, and the AHA/ACC class IIa (B-NR) to the study before surgery even without pre-excitation or tachycardia [28, 31, 35]. The bidirectional cavopulmonary shunt has class IIa (B-NR) in the AATS, which extends it to a right/left atrial pressure ratio above 1.5 and inability to wean from cardiopulmonary bypass, and IIb (B-NR) in the AHA/ACC; the three sources require preserved left ventricular function and non-elevated left-sided pressures [28, 31, 35].

The fourth concerns topics that the adult guidelines do not grade: valve replacement after 60 years of age (IIa in the AATS, whereas the ESC merely prefers repair to replacement when feasible), the proscription of tricuspid surgery in irreversible severe biventricular dysfunction, in which the ESC refers to transplantation, postoperative anticoagulation of defined duration and all treatment of the fetus, newborn and infant, with no equivalent in the ESC and AHA/ACC guidelines [31, 35, 36]. The neonatal thresholds also differ from the 2020 surgical review, which indicated the single-ventricle route with a tricuspid regurgitation jet below 3 m/s and right ventricular systolic pressure below 30 mmHg, whereas the 2024 consensus uses 20–25 mmHg to decide between ductal closure and Starnes and 25–30 mmHg to accept biventricular repair in anatomical pulmonary atresia [25, 36]. Two formal details of the documents should be known by those applying them: the recommendation on comfort care was published as class II without subclass, and intraoperative ablation has class I with the wording of a moderate recommendation (“is reasonable”) [35, 36]. For follow-up, the ESC calls for at least annual review, the AHA/ACC defines intervals by physiological stage and the AATS calls for lifelong follow-up with periodic arrhythmia assessment, without an interval [28, 31, 35].

7.6. Surgical techniques and outcomes

The da Silva cone repair delaminates about two thirds of the leaflets, creates a mobile free edge supported by native chordae or neochordae, reconstructs a 360° cone of valve tissue attached to the anatomical annulus, closes the atrial septal defect in a valved or fenestrated manner when there is right-sided dysfunction, and reduces the atrium; adjuncts include leaflet augmentation with pericardium, polytetrafluoroethylene neochordae, the Sebening stitch and an annuloplasty ring or band, partial in the child so as not to compromise growth [25]. In the original series of 40 patients (mean age 17 years), operative mortality was 2.5%, there were 2 reoperations (5%) and no valve replacement or atrioventricular block; in the Mayo Clinic series of 235 cone repairs (median age 15 years), early mortality was 0.4%, early reoperation 5.9% and a bidirectional cavopulmonary shunt was used in 20% [25]. The 2025 AATS consensus considers that the cone repair has revolutionised surgical treatment, with durable results across a wide anatomical range [35]. “Ebstein repair” usually refers to the combination of cone tricuspid repair, total or subtotal closure of the atrial septal defect, plication of the atrialised ventricle and right reduction atrioplasty, with arrhythmia surgery and a bidirectional cavopulmonary shunt when indicated [35]. The cone mobilises all available valve tissue, joins the leaflets side by side and attaches the reconstructed valve to the normal atrioventricular junction, with central flow and 360° leaflet-to-leaflet coaptation, without commissures; when the initial result is at most mild regurgitation without stenosis, few patients need re-repair or replacement in the long term [35]. Earlier techniques produced a monocusp mechanism, with the anterior leaflet closing against the septum: Danielson’s technique was only suitable when the anterior leaflet was mobile, and Carpentier’s, applicable to more patients, had a high incidence of residual or recurrent regurgitation; comparative studies showed, with the cone, greater success and less significant tricuspid regurgitation in the medium term than conventional valvuloplasties, and better short-term valve function than Hetzer’s technique [35]. The cone improves septal dyskinesia, stroke volume and left ventricular ejection fraction and promotes reverse remodelling [35]. In the child, the consensus places the ideal timing between 3 and 5 years, when leaflet tissue already allows delamination and reconstruction with low morbidity and mortality and before right ventricular dilatation and dysfunction progress; progressive dilatation, severe tricuspid regurgitation or increasing cyanosis after a systemic-to-pulmonary shunt justify earlier intervention, in which a cavopulmonary shunt may be added, and the difficulty of reproducing the cone in the infant, because of tissue fragility, may justify a monocusp repair with a posterior cone if needed [35]. Because surgery at this age and in the asymptomatic patient remains controversial, referral to an experienced centre is advised [35].

The bidirectional cavopulmonary shunt reduces right ventricular volume load by 35–45% and allows “one-and-a-half ventricle” repair when the ventricle is greatly dilated or dysfunctional, the repaired valve cannot accommodate all venous return or a fragile repair needs protecting; in a series of 62 Mayo Clinic patients, operative mortality was 1.6% [25]. The 2025 AATS consensus considers the bidirectional cavopulmonary shunt reasonable (IIa, B-NR) when there is severe right ventricular dilatation or severe systolic dysfunction, a right-to-left atrial pressure ratio above 1.5 or inability to wean from cardiopulmonary bypass after repair; the shunt offloads the right ventricle and increases the preload of the compressed left ventricle [35]. It requires the absence of overt signs of heart failure (ascites and oedema), normal left ventricular ejection fraction, repairable anatomy and, in the adult or with reduced left-sided function, catheterisation with left ventricular end-diastolic pressure below 12 mmHg, transpulmonary gradient below 10 mmHg and mean pulmonary artery pressure below 18 mmHg [35]. Relative indications are severe cardiomegaly (cardiothoracic ratio above 70%), right ventricular end-diastolic volume above 250 mL/m² and right ventricular ejection fraction below 25%, and further weight is given to a small, underfilled, “D-shaped” left ventricle, a right/left atrial pressure ratio above 1.5:1 after repair, low output, persistent metabolic acidosis and, rarely, tricuspid stenosis with a mean gradient above 8–10 mmHg [35]. The limitations are restricted access to the right atrium and ventricle for electrophysiological procedures, mild facial oedema and the development of venous collaterals from the upper half of the trunk or pulmonary arteriovenous malformations [35]. Fenestration of the atrial septum facilitates postoperative recovery when correction of the regurgitation further reduces right ventricular function, but provides little preload to the left ventricle, does not offload the right ventricle and, in the adult, should be applied selectively because of the risk of paradoxical embolism, and can later be closed by catheter [35]. In left ventricular dysfunction in high-risk patients, “one-and-a-half ventricle” repair may be considered if the dysfunction results from abnormal septal motion and not from a globally dilated and dysfunctional ventricle; with severe left ventricular dilatation and dysfunction, heart transplantation is the preferred option [35]. Tricuspid surgery is not recommended in irreversible severe biventricular systolic dysfunction, with left ventricular ejection fraction below 25% and right below 25% (III, B-NR): severe isolated right-sided dysfunction still allows surgery with a cavopulmonary shunt or atrial fenestration, but severe left-sided dysfunction with dilatation points to transplant assessment [35]. Mitral regurgitation due to prolapse or a flail leaflet can be corrected at the same operation despite a reduced ejection fraction; ischaemic mitral regurgitation with severe left-sided dysfunction, especially with coronary disease, argues against tricuspid surgery, and left-sided dysfunction attributable to revascularisable coronary disease does not preclude it [35].

Valve replacement is reserved for when repair is impossible or has failed, a bioprosthesis being preferred because of the thrombogenicity of mechanical prostheses in the tricuspid position, avoiding the conduction system and the right coronary artery; in patients over 60 years of age, replacement is preferable to a complex and prolonged repair; percutaneous valve-in-valve replacement is a future option [25]. The 2025 AATS consensus considers replacement reasonable in adults over 60 years (IIa, B-NR): with marked right ventricular or annular dilatation, the complexity of repair, plication and atrial septal defect closure prolongs cardiopulmonary bypass and cross-clamping, and replacement, which can be performed on the beating heart together with reduction atrioplasty, shortens cross-clamping and avoids repeated attempts to wean from bypass that often end in replacement and mechanical support; repair remains preferred in the child and in the young or middle-aged adult [35]. A porcine bioprosthesis, with leaflets more flexible than pericardial ones, is preferred, suited to a low-pressure right atrium (opening force) and a right ventricle with reduced function (closing force); replacement with a cavopulmonary shunt, which reduces transvalvular flow by 30–40%, should be avoided and, if unavoidable, the prosthesis should be slightly undersized for a mean gradient of 5–6 mmHg, because an oversized prosthesis (mean gradient of 1–2 mmHg) has reduced leaflet mobility, more thrombosis and accelerated degeneration [35]. A mechanical prosthesis is avoided when there is severe right ventricular systolic dysfunction, because of the risk of leaflet thrombosis despite adequate anticoagulation; in series of acquired and congenital tricuspid replacement, valve thrombosis was 1.3–2.0% per patient-year, mainly with mechanical prostheses, and in studies dedicated to Ebstein anomaly 10-year survival after bioprosthesis was 92% [35]. The risk factors for postoperative right ventricular failure are an indexed right ventricular end-diastolic volume above 200 mL/m², right ventricular ejection fraction below 40% and age above 50 years [25].

The European guideline records that operative mortality, formerly above 25%, has fallen to less than 6% in specialised centres, that more than 90% of patients operated on by experienced surgeons survive more than 10 years and that late reoperation-free survival in a large series was 86%, 74%, 62% and 46% at 5, 10, 15 and 20 years [31]. In a Chinese cohort of 332 patients (2000–2021; 269 operated on, with 150 repairs, 77 replacements and 112 cavopulmonary shunts, 42 isolated), early surgical mortality was 2.60%, reoperation-free survival was 97.95%, 92.48%, 87.04% and 83.22% at 5, 10, 15 and 20 years, and preoperative hepatic congestion (HR 3.042), Wolff-Parkinson-White syndrome (HR 3.463) and elevated alanine aminotransferase (HR 1.023 per unit) were associated with risk [99].

7.7. Arrhythmias and thromboembolic prevention

Accessory pathways should be treated before surgery whenever possible, by catheter, because tricuspid surgery subsequently hinders access to right-sided pathways and the slow pathway of the atrioventricular node; surgical interruption is reserved for failed ablation [28, 31]. In chronic atrial fibrillation, at least a right atrial maze is performed during surgery, with left-sided lesions (Cox-Maze III) if there is left-sided remodelling, and flutter is treated with cavotricuspid isthmus ablation; fibrillation or flutter recurs in about 25% [25, 28].

The 2025 AATS consensus organises the perioperative arrhythmia approach in four steps. First, assessment of all candidates with history, electrocardiography and ambulatory monitoring (I, B-NR) and an electrophysiological study in those with documented arrhythmia, pre-excitation or a suggestive history (IIa, C-LD); a study with ablation before surgery reduces the risk of arrhythmia after tricuspid surgery, and catheter ablation in Ebstein anomaly, although effective, is more difficult and recurs more often than in other heart defects [35]. Second, intraoperative ablation when the substrate could not be treated by catheter or will become inaccessible (I, B-NR): after the cone, part of the cavotricuspid isthmus, the posterior ventricular septum and the right atrium may be excluded from catheter access, and the septal suture line of the reconstruction may lie above or below the conduction tissue, so the electrophysiologist should know the operative report; if catheter ablation was partially effective shortly before, the recent lesions are visible at surgery and guide additional lesions; multiple or complex right-sided arrhythmias justify a right-sided maze, and valve replacement or prosthetic material over the cavotricuspid isthmus justify intraoperative lesions at that isthmus [35]. Third, surgery for atrial fibrillation and flutter (concomitant maze reasonable in paroxysmal or continuous fibrillation, IIa, B-NR): right-sided maze in the dilated right atrium with paroxysmal fibrillation, cryoablation of the cavotricuspid isthmus in flutter and biatrial maze in continuous fibrillation or with left-sided valve disease, with pulmonary vein isolation or a complete left-sided maze; surgical ablation of all atrial tachyarrhythmias should be applied liberally, a prophylactic maze should be considered in selected cases with greatly dilated atria without documented arrhythmia, and right atrial reduction helps maintain sinus rhythm [35]. After one or more previous percutaneous ablations, classic Cox-maze lesions require caution, because they can cause slow junctional rhythm or block and the need for a permanent pacemaker; the cut-and-sew technique, cryoablation and radiofrequency have all been used successfully, especially at reoperation [35]. Fourth, postoperatively, protocols with class III antiarrhythmics are applied and short-term systemic anticoagulation with warfarin or a direct oral anticoagulant, such as apixaban, is considered; lifelong follow-up with periodic arrhythmia assessment is recommended (I, B-NR), because the risk persists and increases with age, owing to fibrosis secondary to chronic atrial dilatation, surgical scars and myopathy of the atrialised ventricle [35].

Oral anticoagulation is recommended in patients with previous paradoxical embolism or atrial fibrillation and may be considered if there is increased thromboembolic risk or a right-to-left shunt [31]. The authors of the Nordic morbidity study, given the risk of ischaemic stroke after the onset of supraventricular arrhythmia (HR 22.6), suggest considering early anticoagulation, alone or with antiarrhythmic intervention [32]. Infective endocarditis prophylaxis is reserved for high-risk patients [31].

After valve surgery, the 2025 AATS consensus considers systemic anticoagulation reasonable for 3–6 months after bioprosthetic replacement (IIa, C), followed by lifelong low-dose acetylsalicylic acid, and accepts that it may be beneficial for up to 3 months after complex tricuspid repair in selected cases (IIb, C), weighing the theoretical benefit against postoperative bleeding risk [35]. Repair, including the cone, is less thrombogenic than a bioprosthesis, but an annuloplasty ring or band, non-endothelialised patches and frequent atrial fibrillation increase the risk [35]. In the absence of data specific to the tricuspid valve and Ebstein anomaly, the consensus extrapolates from mitral surgery, in which the thromboembolic and stroke risk of a bioprosthesis is concentrated in the first 90–180 days and in which, in a retrospective series of 424 patients, anticoagulation started as early as possible and maintained for at least 3 months reduced events [35]. The 2021 European valvular heart disease guideline is consistent with this extrapolation: a vitamin K antagonist for 3 months for all mitral or tricuspid bioprostheses, and low-dose acetylsalicylic acid defined as 75–100 mg [100]. Neither the consensus nor the congenital heart disease guidelines give an INR (international normalised ratio) target or doses, so the following values come from the labels and the closest analogous condition, with the extrapolation declared.

Ebstein anomaly indications outside the named label indications; INR, international normalised ratio.

7.8. Ebstein anomaly with non-compaction

Treatment of the association is non-specific: heart failure therapy, anticoagulation, ablation, cardioverter-defibrillator, ventricular assist device and transplantation, depending on the dominant phenotype [3]. With left ventricular systolic dysfunction, the 2025 AATS consensus recommends excluding concomitant acquired disease and starting guideline-directed medical therapy for heart failure (class I, level B-NR) [35]. Assessment includes history and physical examination, laboratory tests, electrocardiography, chest radiography, echocardiography, non-invasive imaging and consideration of genetic testing; magnetic resonance imaging with late enhancement distinguishes ischaemic fibrosis and infiltrative or inflammatory cardiomyopathy patterns, and coronary disease is investigated with stress imaging, scintigraphy, magnetic resonance imaging, positron emission tomography or coronary CT angiography, with coronary angiography in candidates for revascularisation [35]. Hypertension, arrhythmias, diastolic dysfunction, valve disease and coronary disease should be treated, and, in left heart failure due to systolic dysfunction or non-compaction, guideline-directed therapy includes an angiotensin receptor-neprilysin inhibitor, a beta-blocker, a sodium-glucose cotransporter 2 inhibitor and an aldosterone antagonist [35]; drugs, doses and titration are those of the heart failure guidelines. In carriers of a sarcomeric variant with hypertrophic or dilated cardiomyopathy, management follows the guidelines for those cardiomyopathies; the cardiac myosin inhibitors mavacamten (Camzyos®) and aficamten (Myqorzo®) are authorised only in adults with symptomatic obstructive hypertrophic cardiomyopathy, with no indication in Ebstein anomaly or non-compaction [67, 103]. The 2026 Mayo Clinic case, with inducible ventricular tachycardia, reduced ejection fraction and a MYH7 variant, was treated with modified cone repair and an epicardial cardioverter-defibrillator implanted at the same operation, an individual decision described by the authors and not a routine recommendation [39].

7.9. NKX2-5-related heart disease

Management has two axes. The first is surgical or percutaneous repair of the atrial septal defect or other defect, according to usual practice for congenital heart disease; the second, which tends to be forgotten, is lifelong rhythm surveillance, which continues after repair, because atrioventricular block is postnatal and progressive [6, 29, 30]. There is no specialty society guideline specific to NKX2-5 nor a gene-specific surveillance protocol [6]. Clinical Cardiogenetics indicates the usual management of conduction disease for carriers: no drug halts progression, and a pacemaker is indicated in symptomatic bradycardia and at high risk of complete block and sudden death [34]. The class I indications listed there are sinus node dysfunction with symptoms attributed to bradycardia, third-degree or second-degree Mobitz II block regardless of symptoms, syncope with bundle branch block and infra-Hisian conduction demonstrated on electrophysiological study, and alternating bundle branch block; class IIa indications are Mobitz I with symptoms or infranodal block and syncope with asymptomatic pauses longer than 6 seconds, and a consensus on inherited arrhythmia syndromes suggests considering a pacemaker (IIa) in bifascicular block, given the progressive nature of the disease [34]. In an atrial septal defect registry cited by the GeneGuidance curation, pacemakers were implanted at a mean age of 32 years, versus 76 years for the national mean of first implantation, mostly many years after treatment of the defect, with a greater need for pacing in patients with surgical or percutaneous closure; these data refer to atrial septal defect in general and not to NKX2-5 carriers [6].

The choice between a pacemaker and an implantable cardioverter-defibrillator in the young carrier with complete block remains open. Ellesøe and colleagues, faced with 19 sudden deaths in families with atrial septal defect and conduction disease, two of them with a functioning pacemaker, suggested considering a preventive cardioverter-defibrillator in carriers, especially with a family history of cardiomyopathy [16]. The 2026 reports point in the same direction in selected patients: implantation of a dual-chamber cardioverter-defibrillator in a female patient with a truncating variant and rapid non-sustained ventricular tachycardia with presyncope, despite not meeting the classic indications, and a proposal to convert to a cardioverter-defibrillator the pacemaker of a child with recurrent non-sustained ventricular tachycardia, son of a mother with ventricular fibrillation [78, 85]. The authors of these reports stress that the usual stratification tools, based on ejection fraction or an ischaemic substrate, may be insufficient in this disease [78]. This is case-series evidence, without comparative studies. Assessment of thyroid function is also recommended, given the association with congenital hypothyroidism, as is avoidance, in patients with manifest conduction disease, of drugs that slow conduction, a recommendation formulated for inherited conduction disease in general [6, 34].

7.10. Regulatory status

Sources: [64, 67, 93, 94, 103, 104]. Authorisation does not equate to marketing or reimbursement in each country.

7.11. Trials and interventional studies

LVEF, left ventricular ejection fraction; RV, right ventricle.

These studies do not allow indirect comparisons: the cell trial is phase I, small and compared with historical controls, and the NSAID studies are retrospective, with selection of treated cases and few patients [93, 94, 105]. The 2022 systematic review found in the North American and European trial registries mainly observational studies of risk stratification and genetics, and no therapeutic trial in the association with non-compaction other than the cell study [3].

7.12. Developments in the last 6 to 12 months and research

Between November 2025 and October 2026, several reports of families with NKX2-5 variants were published, all at the case-series level of evidence: a three-generation Canadian family with a novel truncating variant and conduction system disease treated with a pacemaker; a female patient with non-sustained ventricular tachycardia treated with a dual-chamber cardioverter-defibrillator; a family with aborted ventricular fibrillation in the mother and ventricular tachycardia in the son with a pacemaker; and a Chinese family with a truncating variant co-segregating with atrial septal defect [78, 82, 83, 85, 107]. A 2025 report described implantation of a dual-chamber cardioverter-defibrillator during pregnancy in a carrier with progressive nodal disease, with uncomplicated pregnancy and delivery [108]. In Ebstein anomaly, the Mayo Clinic case published in 2026 added an example of association with non-compaction, inducible ventricular tachycardia and a MYH7 variant of uncertain significance [39], and a 20-year Chinese surgical cohort proposed a nomogram for reoperation risk [99]. In July 2025, the AATS consensus for the child over one year of age and the adult was published, complementing the 2024 consensus for the newborn, with 20 graded recommendations (10 class I, 7 class IIa, 1 class IIb and 2 class III); the 2024 consensus has 13 graded recommendations (11 class I) and one good practice statement [35, 36]. At the preclinical level, a study published in June 2025 corrected, by prime editing, a murine model of congenital heart disease associated with an Nkx2-5 variant [109]. A second family with the FLNA variant NM_001110556.2:c.4660G>A p.(Gly1554Arg), Ebstein anomaly and joint stiffness was described in Japan [49]; in ClinVar, this protein change has a single registered allele, classified as pathogenic or likely pathogenic [110]. None of these studies changes the guideline recommendations.

7.13. Pregnancy

The 2025 European guideline on cardiovascular disease and pregnancy places uncomplicated Ebstein anomaly (mild to moderate tricuspid regurgitation, without tricuspid stenosis or accessory pathway) in mWHO 2.0 class II–III (modified World Health Organization classification) and Ebstein anomaly with any complication in class III; both require counselling by the cardio-obstetric team, with care shared with the local hospital in class II–III and delivery at a specialised centre in class III [111]. The mean rates of maternal cardiac events cited for classes II–III and III were 17.7% and 28.9% in one series and 12.8% and 21.1% in another [111]. The risk is low in mild or moderate Ebstein anomaly and high or very high with heart failure before pregnancy or cyanosis due to atrial shunting, and accessory pathways increase arrhythmic risk; fetal risk relates to low maternal output and cyanosis (miscarriage, preterm birth in 20–24%, neonatal death in 3%), there is a risk of postpartum haemorrhage, and the recurrence risk of heart disease is 5% [111]. The recommended monitoring is echocardiography and electrocardiography at baseline and at 28–32 weeks in mild to moderate forms and monthly or bimonthly in severe forms, with investigation for arrhythmia if there are palpitations; severe tricuspid regurgitation with heart failure is generally treated medically, and arrhythmias should be treated promptly [111].

In the ESC Registry of Pregnancy and Cardiac disease (ROPAC) (2007–2018), the 81 pregnancies in women with Ebstein anomaly (1.4% of the registry) had major cardiac events in 9.9% (heart failure 7.4%, supraventricular arrhythmia 3.7%, thromboembolism in 2), with no maternal deaths, versus 15.7% in the rest of the registry; signs of heart failure before pregnancy were the only predictor (OR 21; 95% CI 2.87–158.32), with events in 60% of women who had them versus 6.6% in the others [89]. Caesarean section was performed in 45.7%, preterm birth occurred in 24.7%, neonatal mortality was 2.5% and 4.9% of newborns had congenital heart disease; the degree of tricuspid regurgitation and right-sided dimensions did not change after pregnancy [89]. The authors propose mWHO class II for the uncomplicated form and III when there are risk factors, whereas the 2025 guideline adopted II–III, which reflects a difference of degree in the same direction [89, 111]. The 2020 adult congenital heart disease guideline states that asymptomatic women with good ventricular function tolerate pregnancy well, with a risk of right ventricular failure, arrhythmia and paradoxical embolism, greater with cyanosis, severe arrhythmia and right-sided failure [31]. In a Japanese series (13 patients, 27 pregnancies, 1985–2011), the cardiothoracic ratio rose slightly during pregnancy and postpartum, and outcomes were good in patients in NYHA class I [112]. In a Mayo Clinic series (9 patients, 12 deliveries), neuraxial anaesthesia and analgesia were well tolerated, with no maternal or fetal deaths [113].

For carriers of NKX2-5 variants, the sources consulted provide no pregnancy-specific recommendations beyond the report of cardioverter-defibrillator implantation during gestation [108]. In the subgroup with non-compaction, maternal risk depends on the associated cardiomyopathy or congenital heart disease, according to the mWHO classification [38]. The 2025 AATS consensus includes cardiopulmonary exercise testing in risk assessment before pregnancy [35].

7.14. Lithium in pregnancy

Lithium is the main environmental factor associated with Ebstein anomaly, but the risk was overestimated for decades. The international “lithium babies” registry of the 1970s suggested a 400-fold higher risk of Ebstein anomaly (based on two cases) and a 5-fold higher risk of heart defects, with no control group and retrospective inclusion [24]. Later reviews suggested a risk 10 to 20 times higher than in the general population, and a prospective controlled study of 148 women found a relative risk of 1.2 for all anomalies and 3.5 for cardiac anomalies [63]. In the United States Medicaid cohort (1,325,563 pregnancies, 2000–2010), cardiac malformations occurred in 2.41% of the 663 exposed in the first trimester, versus 1.15% of those unexposed (adjusted relative risk 1.65; 95% CI 1.02–2.68); the risk was dose-dependent (1.11 with 600 mg/day or less, 1.60 with 601–900 mg and 3.22 with more than 900 mg), and right ventricular outflow tract obstruction defects, a group that includes Ebstein anomaly, occurred in 0.60% versus 0.18% (relative risk 2.66; 1.00–7.06), all with doses above 600 mg/day [24]. A meta-analysis of six international cohorts (727 exposed pregnancies) found, with first-trimester exposure, major malformations in 7.4% versus 4.3% (adjusted OR 1.71) and major cardiac malformations in 2.1% versus 1.6% (OR 1.54; 0.64–3.70; not significant) [114]. Another meta-analysis estimated an OR of 1.86 for cardiac anomalies (number needed to harm of 71) and concluded that the risk is low, greater in the first trimester and with higher doses, recommending planning pregnancy during remission and using the lowest therapeutic range [115]. The Oxford Desk Reference recommends detailed fetal cardiac ultrasound in exposed pregnant women [63].

7.15. Exercise, perioperative care and transition

Without residual lesions, patients may lead an active life without restrictions, except for intense static sports; with more than mild tricuspid regurgitation, ventricular dysfunction, shunting, arrhythmias or other complications, they should avoid intense isometric exercise in proportion to severity [31]. In older patients who are candidates for a cavopulmonary shunt, preoperative catheterisation to assess haemodynamics becomes important, especially with long-standing hypertension and left ventricular hypertrophy, because of the risk of diastolic dysfunction and elevated pulmonary pressures [28]; the 2025 AATS consensus considers it essential in the adult and with reduced left ventricular function, with thresholds of left ventricular end-diastolic pressure below 12 mmHg, transpulmonary gradient below 10 mmHg and mean pulmonary artery pressure below 18 mmHg [35]. The same consensus notes that, in children, surgery at 3–5 years facilitates social and school integration, and that the risks of cardiac surgery from 1 year of age have become low [35]. The Oxford Desk Reference recommends storing DNA before open-heart surgery or extracorporeal membrane oxygenation in the newborn with heart disease [90]. In the transition to adulthood, the Nordic cohort justifies lifelong follow-up at specialised centres, because heart failure in the isolated form appears mainly from the fourth decade and in forms with atrial septal defect from adolescence [32].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks by scenario

Counselling begins by distinguishing isolated, sporadic Ebstein anomaly, for which empirical risks are used, from forms with an identified genetic cause or obvious familial aggregation, for which the risk follows the apparent inheritance pattern [90]. The table summarises the available values.

Familial Ebstein anomaly exists but is rare: OMIM records affected siblings, mother-to-child transmission with non-compaction and the family with Ebstein anomaly in the father and four children with the MYH7 variant p.(Glu1220del) [1, 69]. In the North American case-control study, a family history of congenital heart disease was four times more frequent in cases [46].

8.2. Cascade testing and assessment of relatives

When a pathogenic variant is identified in the proband, cascade testing of relatives is recommended; carriers should have a baseline cardiological assessment (symptoms, examination, electrocardiography and echocardiography), with Holter monitoring, an event recorder and exercise testing if there are intermittent symptoms, and periodic reassessment [34]. For NKX2-5, Clinical Cardiogenetics suggests reassessment with symptoms and electrocardiography, for example every 1–3 years depending on age and electrocardiographic changes; non-carriers can be reassured, unless pathogenicity is in doubt, with a baseline electrocardiogram [34]. A prolonged PR interval in apparently healthy first-degree relatives of patients with secundum atrial septal defect is useful in genetic counselling [4]. In the Danish family described by Ellesøe, the lack of guidelines led to rhythm checks every five years [16].

In familial non-compaction, which is observed in 18–71% of adults with the isolated form, most affected relatives are asymptomatic, so cardiological assessment should include all adult relatives regardless of clinical history; when there is a pathogenic variant, predictive testing is offered, and without a variant cardiological assessment is the screening method [38]. In the 2026 Mayo Clinic case, with a MYH7 variant of uncertain significance, the authors recommended screening echocardiography for all first-degree relatives and segregation analysis in the parents, without predictive testing [39]. In Ebstein anomaly without an identified cause, the Oxford Desk Reference recommends cardiovascular examination of both parents, with electrocardiography and echocardiography if there is an abnormality in the history or examination [90]; in conduction disease without a genetic cause and without obvious familial disease, a baseline electrocardiogram is recommended for first-degree relatives and echocardiography if the proband has structural heart disease [34].

8.3. Prenatal diagnosis, preimplantation testing and counselling of minors

Prenatal diagnosis by fetal echocardiography should be offered to women with congenital heart disease and to families with recurrence; miscarriage is more frequent when the mother has heart disease (20% with a maternal proband versus 10% with a paternal proband) [90]. With a known familial variant, analysis of chorionic villi or amniocentesis and preimplantation genetic testing are technical options, whose usefulness should be discussed in the light of incomplete penetrance and variable expressivity, which make it impossible to predict severity from the genotype [7, 16, 33]. Preimplantation testing requires in vitro fertilisation, with the increased perinatal risk attributed to it (prematurity and low birth weight), and a test designed for the family, whose preparation takes months and benefits from the participation of affected and unaffected relatives; in the Netherlands, eligibility criteria include a recurrence risk above 10%, serious disease, a known pathogenic or likely pathogenic variant, female age under 40 years and a body mass index up to 35 [91]. Because the diagnosis is made on one or a few cells, the risk of error was estimated at 2% in the case illustrated in the chapter by Nannenberg and Arens, and confirmatory prenatal diagnosis is offered [91]. Demand is small in inherited heart diseases: between 2001 and 2015, 8 prenatal diagnoses for inherited heart disease were made in the Netherlands, and of the 45 couples with inherited heart disease who had a preimplantation testing consultation between 1995 and 2017 (one of them for non-compaction cardiomyopathy) only 11 proceeded; in published series, the couples had severe phenotypes or a family history of sudden death [91]. The authors attribute the low demand to incomplete penetrance, variable expressivity and the availability of preventive measures, reasons that apply to the sarcomeric subgroup and to NKX2-5 [91]. Women with Ebstein anomaly who are planning pregnancy should be followed at an adult congenital heart disease clinic and receive preconception counselling from the cardio-obstetric team, with particular attention to signs of heart failure and cyanosis [89, 90, 111]. Women taking lithium should plan pregnancy with their psychiatrist, weighing the dose-dependent fetal cardiac risk against the risk of relapse, and undergo fetal echocardiography [24, 63, 115].

In children and adolescents from families with NKX2-5, testing has direct clinical utility, because the carrier needs electrocardiographic surveillance from childhood, and block can appear early, as in the carrier who developed complete block at 7 years of age [34, 85]. In the sarcomeric subgroup, testing of minors follows the logic of the cardiomyopathies: it is useful when the result changes surveillance, and the variant must be classified as pathogenic or likely pathogenic [33, 38].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence of Ebstein anomaly and what each figure measures

Published figures vary with the source, period and case definition, and refer to birth prevalence (cases per live births) rather than population prevalence. The table brings them together.

Ebstein anomaly accounts for less than 1% of congenital heart defects, which affect about 7 to 9 per 1000 live births [25, 31, 90]. In Brazil, the book Síndromes Genéticas: Reconhecer para Tratar cites an overall incidence of congenital heart defects of 8 per 1000 live births, with no data specific to Ebstein anomaly [117]. Heterogeneity between European countries is large (from 0.6 per 100,000 in Spain to 27.2 in Switzerland) and the values come from registries for different years (2005 to 2010), with no analysis of the causes of variation in the sources consulted [15]. The frequencies cited by the societies also diverge: 0.005% of live births (5 per 100,000, about 1 in 20,000) in the AHA/ACC guideline and 1 in 200,000 live births (0.5 per 100,000) in the 2025 AATS consensus, a value ten times lower and lower than most of the population registries in the table [28, 35]. The sources consulted do not include data from Portugal.

In Texas, the prevalence of the non-syndromic form was higher with maternal age above 39 years, residence near the border with Mexico and conception in autumn or winter, and affected newborns had more prematurity and growth restriction [23]. In the national North American case-control study, maternal age was not associated with risk, and the likelihood was lower, without statistical significance, in children of non-Hispanic Black mothers (OR 0.6; 0.3–1.2) [46]. In the Nordic registries there was no difference in mortality by sex [27].

9.2. Epidemiology of NKX2-5-related heart disease and founder effect

Orphanet classifies atrial septal defect with atrioventricular conduction defects syndrome (ORPHA:1479) with a prevalence below 1 per 1,000,000 and 11 cases or families recorded [15]. This figure underestimates carrier frequency: in Iceland, the p.(Phe145Leu) variant is carried by 1 in every 7100 inhabitants, about 14 per 100,000, in a genotype-first estimate of a single founder variant [6, 26]. The two figures measure different things (clinically identified disease versus carriers in a founder population) and are not comparable [6]. In screening series, NKX2-5 variants are found in about 9% of probands with familial congenital heart disease and in less than 2% of non-familial cases, and in 4% of secundum atrial septal defects in a series of 474 patients with congenital heart disease [4, 16]. The Icelandic authors proposed screening for this variant in patients with dilated cardiomyopathy and their relatives, with monitoring of carriers [26].

9.3. Special populations

In Europe, the ROPAC registry included 81 pregnancies in women with Ebstein anomaly, 21% of them resident in emerging countries (versus 40% in the rest of the registry) [89]. The Brazilian InCor series showed that the frequency of non-compaction in Ebstein anomaly depends on the morphological criterion applied [3]. Patients with genetic syndromes, prematurity or serious comorbidities have a worse neonatal prognosis and may be candidates for palliative care according to the AATS consensus [36].

9.4. European reference network and patient associations

The relevant European reference network is ERN GUARD-Heart, dedicated to rare, low-prevalence and complex heart diseases [37]. Associations of patients and families with congenital heart disease include the Adult Congenital Heart Association and Conquering CHD in the United States, the European Congenital Heart Disease Organisation (ECHDO) in Europe, the Children's Heart Federation in the United Kingdom and ACTC — Casa do Coração in Brazil, which supports children and adolescents with heart disease and transplant recipients [90, 118–122].

10. Illustrative Clinical Case

In 2007, Budde and colleagues described the German family DU-11, the first in which ventricular non-compaction was linked to a locus by genome-wide analysis and attributed to a MYH7 variant, with Ebstein anomaly and atrial septal defect in some carriers [41]. A total of 24 members were assessed clinically, with 12-lead electrocardiography and transthoracic echocardiography, and non-compaction was confirmed in nine patients, with two deceased individuals and one child considered partially affected [41]. The pedigree, with chromosome 14 haplotypes and segregation of the variant, is in Figure 1 of the article (https://pmc.ncbi.nlm.nih.gov/articles/PMC2137931/figure/pone-0001362-g001/), and the magnetic resonance image and echocardiogram of a sister of the proband, with trabeculations and deep recesses in the apical half of both ventricles, are in Figure 2 (https://pmc.ncbi.nlm.nih.gov/articles/PMC2137931/figure/pone-0001362-g002/) [41].

The proband (III-4) presented at 39 years of age with exertional dyspnoea; she had Ebstein anomaly, atrial septal defect, an abnormal electrocardiogram and non-compaction [41]. She and her sister III-8, who at presentation at 36 years was in NYHA class I and had isolated non-compaction, deteriorated over three years to cardiac outputs of 2.4 and 2.1 L/min by thermodilution (normal 4.0–10.2 L/min) and were placed on the heart transplant waiting list [41]. The mother (II-3) had Ebstein anomaly, atrial septal defect and atrioventricular block with a pacemaker, and died at 61 years of pacemaker-associated sepsis; a brother of the proband (III-2), not genotyped, had a cardiac death at 26 years and also had Ebstein anomaly and atrial septal defect [41]. In the fourth generation, an asymptomatic 13-year-old boy had Ebstein anomaly with non-compaction, a 7-year-old girl had atrial septal defect with non-compaction, and five asymptomatic children and adolescents had isolated or partial non-compaction with an abnormal electrocardiogram; none had the septal hypertrophy or dilatation typical of hypertrophic or dilated cardiomyopathy [41].

Linkage analysis, assuming dominant transmission with reduced penetrance, gave a maximum LOD score of 4.15 at a marker in the 14ptel–14q12 interval, which contains MYH6 and MYH7 [41]. Sequencing of the 40 exons of MYH7 in the proband identified the variant c.842G>C in exon 10, p.(Arg281Thr) (historically described as R281T), present in all affected and partially affected members and absent in unaffected members and in 184 controls; the variant lies in the myosin head, near switch I of the ATPase site, and modelling predicted loss of a salt bridge with residue 325 and destabilisation of the myosin head [41]. On the MANE Select transcript, the description is NM_000257.4:c.842G>C, NP_000248.2:p.(Arg281Thr) (NC_000014.9:g.23430954C>G) [68]. The absence of the variant and of non-compaction in two of the mother’s siblings with the same haplotype led the authors to propose a de novo variant in II-3, with germline mosaicism in one of the parents as an alternative explanation [41].

The case illustrates the core clinical features of the sarcomeric subgroup: autosomal dominant transmission, variable expressivity within the same family (Ebstein anomaly, atrial septal defect, isolated non-compaction, atrioventricular block), absence of symptoms in childhood with rapid deterioration in the third and fourth decades in some carriers, and premature cardiac death [41]. It also shows why assessment of all first-degree relatives should include echocardiography, even when asymptomatic, and why Ebstein anomaly with non-compaction warrants testing of MYH7 and TPM1 [9, 38]. The report describes the listing of the two sisters for transplantation; current guidelines and consensus documents also describe tricuspid surgery, with cone repair and a bidirectional cavopulmonary shunt, in selected patients with left ventricular dysfunction (some patients with an ejection fraction of 30% have recovered with repair or replacement and a shunt), reserving transplantation for severe biventricular dysfunction [25, 31, 35].

11. Prognosis and Long-Term Follow-up

11.1. Prognosis

The prognosis of Ebstein anomaly depends on age at presentation, anatomical severity, associated lesions and the function of both ventricles. In fetal life, perinatal mortality was 45% in a contemporary multicentre series, with 54% in those diagnosed before 32 weeks [56]. In the newborn, the Celermajer index is associated with early mortality without treatment of 0%, 10%, 44% and 50% in grades 1 to 4 [25]. In fetuses with a severe form, fetal death affects 15–20% of continuing pregnancies and neonatal death 25–32%; in operated newborns, mortality was 27.4% in the Society of Thoracic Surgeons database and 37.5% in the first year in a recent series, and the Starnes procedure has a predicted mortality of 10–20%; those who survive the first year generally do well [36]. In the national registries of Denmark and Sweden, cumulative mortality was 16% at 20 years and 20% at 40 years, with 9% at 35 years in the isolated form, 16% with atrial septal defect and 39% in complex forms, and even presumably mild disease had a mortality six times higher than that of the population [27]. Supraventricular arrhythmia affects about 70% over a lifetime, pre-excitation 19%, and heart failure by 60 years about 6% in the isolated form and 40% in complex forms [32].

The factors of worse prognosis listed by the 2025 AATS consensus for the non-neonatal patient are congestive heart failure, cyanosis, significant left ventricular dysfunction (ejection fraction below 40%), severe cardiomegaly (cardiothoracic ratio above 0.65) and persistent arrhythmia (class I, level B-NR); the risk of arrhythmia and sudden death increases with age, even after surgery, and is greatest in older unoperated adults, and 10-year survival after a tricuspid bioprosthesis was 92% [35]. In the Chinese series, preoperative hepatic congestion and Wolff-Parkinson-White syndrome were associated with the risk of reoperation or intervention [99]. After surgery at specialised centres, more than 90% of patients survive more than 10 years, and late deaths are probably attributable to arrhythmias [31]. In the subgroup with non-compaction, presentation in childhood is associated with a greater risk of heart failure and sudden death, and patients with a genetic variant tend to have a worse prognosis than sporadic forms of non-compaction [3, 38].

In NKX2-5-related heart disease, prognosis depends mainly on conduction disease, cardiomyopathy and ventricular arrhythmias. Sudden death affected 8% to 15% of carriers with familial atrial septal defect and conduction disease in Ellesøe’s review, and the Icelandic variant is associated with high penetrance of serious heart disease [16, 26].

11.2. Surveillance

AVB, atrioventricular block; ECMO, extracorporeal membrane oxygenation; RVSP, right ventricular systolic pressure; ICD, implantable cardioverter-defibrillator; ECG, electrocardiogram; RVEF, right ventricular ejection fraction; TR, tricuspid regurgitation; CPET, cardiopulmonary exercise testing; SaO2, arterial oxygen saturation; CT, computed tomography; VT, ventricular tachycardia; RV, right ventricle; LV, left ventricle.

After surgery, surveillance looks for persistent or new tricuspid regurgitation, the usual complications of prostheses, right or left ventricular failure, residual atrial shunts, arrhythmias and high-grade atrioventricular block; reintervention may be needed for recurrent regurgitation and prosthetic failure [31]. The GeneGuidance curation cites, as the lower limit of what an NKX2-5 carrier needs, the general practice of follow-up every 2–3 years for small atrial septal defects without right ventricular dilatation or pulmonary hypertension, stressing that these are data for atrial septal defect in general and that progressive conduction disease justifies, on clinical judgement, more frequent electrocardiograms [6]. Patients should seek urgent assessment if they have syncope [34].

12. Summary and Final Considerations

Ebstein anomaly is a malformation of the tricuspid valve and right ventricle, defined by failure of delamination and apical displacement of the septal and posterior leaflets, which divides the right heart into an atrialised portion and a residual functional ventricle; it affects about 3.5 per 100,000 births in Europe and accounts for less than 1% of congenital heart defects [15, 25, 31]. The disease is not confined to the right heart: left-sided lesions, including left ventricular non-compaction, are present in about 39% of patients in a reference series [40]. The clinical spectrum ranges from fetal death (perinatal mortality of 45% in diagnosed fetuses) to an incidental finding in the elderly adult, and the Nordic national registries show that even the isolated form has higher mortality and morbidity than the population, with supraventricular arrhythmia in about 70% over a lifetime [27, 32, 56].

The aetiology is heterogeneous. Most cases are sporadic; about a quarter are syndromic, with the 1p36 deletion and the 8p23.1 deletion as the most frequent chromosomal imbalances, and first-trimester lithium increases the risk of right ventricular outflow tract defects modestly and in a dose-dependent manner [12, 24]. The subgroup with left ventricular non-compaction has an autosomal dominant sarcomeric basis, with MYH7 variants in 6% of probands with Ebstein anomaly, concentrated in those with non-compaction, TPM1 variants and, in one family, KLHL26 variants; it is this subgroup that warrants targeted genetic testing and echocardiographic assessment of relatives [7–10]. Syndromic forms warrant microarray and, if negative, exome or genome sequencing [33].

NKX2-5-related heart disease is an autosomal dominant allelic spectrum whose core phenotype is ostium secundum atrial septal defect with progressive, postnatal nodal atrioventricular block, to which are added ventricular septal defect, tetralogy, conotruncal defects, hypoplastic left heart, Ebstein anomaly, cardiomyopathy, ventricular arrhythmias and congenital hypothyroidism [16, 29, 30, 42]. Repair of the defect does not eliminate the risk: sudden death affected 8–15% of carriers with familial atrial septal defect and conduction disease, sometimes with a functioning pacemaker, and recent reports reinforce the need for lifelong electrocardiographic surveillance and investigation for ventricular arrhythmia [16, 78, 85].

Treatment of Ebstein anomaly is surgical and electrophysiological and is decided by symptoms, measured exercise capacity, right ventricular dimensions and function and arrhythmias, with cone repair as the reference technique and ablation of accessory pathways before surgery; the European and North American guidelines and the 2025 AATS consensus converge on symptomatic indications and differ in the threshold for intervention in the asymptomatic patient, which the AATS brings forward to severe tricuspid regurgitation with moderate right ventricular dilatation and, in the child, to 3–5 years, in isolated percutaneous closure of the atrial septal defect, which the AATS advises against with severe tricuspid regurgitation, and in the class assigned to magnetic resonance imaging, electrophysiological study and bidirectional cavopulmonary shunt [28, 31, 35]. Left ventricular dysfunction weighs more on prognosis than right ventricular dysfunction, and irreversible severe biventricular dysfunction points to transplantation rather than tricuspid surgery [35]. In the fetus, serial echocardiography every 2–4 weeks and every 1–2 weeks after 32 weeks identifies the high-risk signs that require planned delivery at a centre with extracorporeal membrane oxygenation and cardiac surgery; in the newborn, risk stratification guides the choice between observation, ductal closure, interruption of circular shunting and the Starnes procedure, and the Starnes no longer commits the patient to the single-ventricle pathway, because conversion to biventricular or “one-and-a-half ventricle” repair is possible in the infant [36]. Pregnancy is well tolerated in the uncomplicated form and high-risk with prior heart failure or cyanosis [89, 111]. There is no targeted pharmacological therapy; cell therapies and fetal ductal constriction remain investigational [93, 94, 105].

13. References

1. Online Mendelian Inheritance in Man, OMIM®. Johns Hopkins University. MIM 224700 — Ebstein anomaly. Available from: https://www.omim.org/entry/224700 (edited 8 Jul 2022; accessed 31 Oct 2026).
2. Orphanet. Ebstein malformation of the tricuspid valve (ORPHA:1880). Available from: https://www.orpha.net/en/disease/detail/1880
3. Thareja SK, Frommelt MA, Lincoln J, Lough JW, Mitchell ME, Tomita-Mitchell A. A Systematic Review of Ebstein's Anomaly with Left Ventricular Noncompaction. J Cardiovasc Dev Dis. 2022;9(4):115. doi:10.3390/jcdd9040115. PMID: 35448091. PMC9031964.
4. Online Mendelian Inheritance in Man, OMIM®. MIM 108900 — Atrial septal defect 7 with or without atrioventricular conduction defects; ASD7. Available from: https://www.omim.org/entry/108900 (accessed 31 Oct 2026).
5. Orphanet. Atrial septal defect-atrioventricular conduction defects syndrome (ORPHA:1479). Available from: https://www.orpha.net/en/disease/detail/1479
6. GeneGuidance. Curation record for the NKX2-5 gene (with the primary references cited therein). GeneGuidance database; accessed 31 Oct 2026.
7. Postma AV, van Engelen K, van de Meerakker J, Rahman T, Probst S, Baars MJ, et al. Mutations in the sarcomere gene MYH7 in Ebstein anomaly. Circ Cardiovasc Genet. 2011;4(1):43-50. doi:10.1161/CIRCGENETICS.110.957985. PMID: 21127202.
8. Kelle AM, Bentley SJ, Rohena LO, Cabalka AK, Olson TM. Ebstein anomaly, left ventricular non-compaction, and early onset heart failure associated with a de novo α-tropomyosin gene mutation. Am J Med Genet A. 2016;170(8):2186-90. doi:10.1002/ajmg.a.37745. PMID: 27177193.
9. Nijak A, Alaerts M, Kuiperi C, Corveleyn A, Suys B, Paelinck B, et al. Left ventricular non-compaction with Ebstein anomaly attributed to a TPM1 mutation. Eur J Med Genet. 2018;61(1):8-10. doi:10.1016/j.ejmg.2017.10.003. PMID: 29024827.
10. Samudrala SSK, North LM, Stamm KD, Earing MG, Frommelt MA, Willes R, et al. Novel KLHL26 variant associated with a familial case of Ebstein's anomaly and left ventricular noncompaction. Mol Genet Genomic Med. 2020;8(4):e1152. doi:10.1002/mgg3.1152. PMID: 31985165. PMC7196453.
11. Mercer CL, Andreoletti G, Carroll A, Salmon AP, Temple IK, Ennis S. Familial Ebstein Anomaly: Whole Exome Sequencing Identifies Novel Phenotype Associated With FLNA. Circ Cardiovasc Genet. 2017;10(6):e001683. doi:10.1161/CIRCGENETICS.116.001683. PMID: 29237676.
12. Digilio MC, Bernardini L, Lepri F, Giuffrida MG, Guida V, Baban A, et al. Ebstein anomaly: Genetic heterogeneity and association with microdeletions 1p36 and 8p23.1. Am J Med Genet A. 2011;155A(9):2196-202. doi:10.1002/ajmg.a.34131. PMID: 21815254.
13. Online Mendelian Inheritance in Man, OMIM®. MIM 600584 — NK2 homeobox 5; NKX2-5. Available from: https://www.omim.org/entry/600584 (accessed 31 Oct 2026).
14. Vermeer AM, van Engelen K, Postma AV, Baars MJ, Christiaans I, De Haij S, et al. Ebstein anomaly associated with left ventricular noncompaction: an autosomal dominant condition that can be caused by mutations in MYH7. Am J Med Genet C Semin Med Genet. 2013;163C(3):178-84. doi:10.1002/ajmg.c.31365. PMID: 23794396.
15. Orphanet. Orphadata Science: epidemiology and natural history of ORPHA:1880 and ORPHA:1479 (version of 23 Jun 2026; CC BY 4.0 licence). Available from: https://api.orphadata.com/rd-epidemiology/orphacodes/1880 (accessed 31 Oct 2026).
16. Ellesøe SG, Johansen MM, Bjerre JV, Hjortdal VE, Brunak S, Larsen LA. Familial Atrial Septal Defect and Sudden Cardiac Death: Identification of a Novel NKX2-5 Mutation and a Review of the Literature. Congenit Heart Dis. 2016;11(3):283-90. doi:10.1111/chd.12317. PMID: 26679770. PMC5019245.
17. Online Mendelian Inheritance in Man, OMIM®. MIM 613426 — Cardiomyopathy, dilated, 1S; CMD1S (includes LVNC5). Available from: https://www.omim.org/entry/613426 (accessed 31 Oct 2026).
18. Online Mendelian Inheritance in Man, OMIM®. MIM 611878 — Cardiomyopathy, dilated, 1Y; CMD1Y (includes LVNC9). Available from: https://www.omim.org/entry/611878 (accessed 31 Oct 2026).
19. Hershberger RE, Jordan E. Dilated cardiomyopathy overview. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2007 Jul 27; last update 2021 Jul 29; revised 2024 Dec 12. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1309/
20. Cirino AL, Channaoui N, Ho C. Nonsyndromic hypertrophic cardiomyopathy overview. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2008 Aug 5; last update 2025 Mar 6. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1768/
21. Robertson S, Wade E. FLNA-related otopalatodigital spectrum disorders. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2005 Nov 30; last update 2025 Jun 26. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1393/
22. Battaglia A. 1p36 deletion syndrome (retired chapter, for historical reference only). In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2008 Feb 1; last update 2013 Jun 6. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1191/
23. Lupo PJ, Langlois PH, Mitchell LE. Epidemiology of Ebstein anomaly: prevalence and patterns in Texas, 1999-2005. Am J Med Genet A. 2011;155A(5):1007-14. doi:10.1002/ajmg.a.33883. PMID: 21465650.
24. Patorno E, Huybrechts KF, Bateman BT, Cohen JM, Desai RJ, Mogun H, et al. Lithium Use in Pregnancy and the Risk of Cardiac Malformations. N Engl J Med. 2017;376(23):2245-2254. doi:10.1056/NEJMoa1612222. PMID: 28591541. PMC5667676.
25. Sainathan S, da Fonseca da Silva L, da Silva JP. Ebstein's anomaly: contemporary management strategies. J Thorac Dis. 2020;12(3):1161-1173. doi:10.21037/jtd.2020.01.18. PMID: 32274197. PMC7139083.
26. Sveinbjornsson G, Olafsdottir EF, Thorolfsdottir RB, Davidsson OB, Helgadottir A, Jonasdottir A, et al. Variants in NKX2-5 and FLNC Cause Dilated Cardiomyopathy and Sudden Cardiac Death. Circ Genom Precis Med. 2018;11(8):e002151. doi:10.1161/CIRCGEN.117.002151. PMID: 30354339.
27. Eckerström F, Dellborg M, Hjortdal VE, Eriksson P, Mandalenakis Z. Mortality in Patients With Ebstein Anomaly. J Am Coll Cardiol. 2023;81(25):2420-2430. doi:10.1016/j.jacc.2023.04.037. PMID: 37344044.
28. Stout KK, Daniels CJ, Aboulhosn JA, Bozkurt B, Broberg CS, Colman JM, et al. 2018 AHA/ACC Guideline for the Management of Adults With Congenital Heart Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;139(14):e698-e800. doi:10.1161/CIR.0000000000000603. PMID: 30586767.
29. Schott JJ, Benson DW, Basson CT, Pease W, Silberbach GM, Moak JP, et al. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science. 1998;281(5373):108-11. doi:10.1126/science.281.5373.108. PMID: 9651244.
30. Benson DW, Silberbach GM, Kavanaugh-McHugh A, Cottrill C, Zhang Y, Riggs S, et al. Mutations in the cardiac transcription factor NKX2.5 affect diverse cardiac developmental pathways. J Clin Invest. 1999;104(11):1567-73. doi:10.1172/JCI8154. PMID: 10587520. PMC409866.
31. Baumgartner H, De Backer J, Babu-Narayan SV, Budts W, Chessa M, Diller GP, et al. 2020 ESC Guidelines for the management of adult congenital heart disease. Eur Heart J. 2021;42(6):563-645. doi:10.1093/eurheartj/ehaa554. PMID: 32860028.
32. Eckerström F, Hjortdal VE, Eriksson P, Dellborg M, Mandalenakis Z. Morbidity Burden in Patients With Ebstein Anomaly: The Natural History. J Am Heart Assoc. 2024;13(21):e037127. doi:10.1161/JAHA.124.037127. PMID: 39494589. PMC11935686.
33. Wilde AAM, Semsarian C, Márquez MF, Shamloo AS, Ackerman MJ, Ashley EA, et al. European Heart Rhythm Association (EHRA)/Heart Rhythm Society (HRS)/Asia Pacific Heart Rhythm Society (APHRS)/Latin American Heart Rhythm Society (LAHRS) Expert Consensus Statement on the state of genetic testing for cardiac diseases. Europace. 2022;24(8):1307-1367. doi:10.1093/europace/euac030. PMID: 35373836. PMC9435643.
34. Tadros R, Cadrin-Tourigny J. Hereditary cardiac conduction diseases. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 273–85. doi:10.1007/978-3-030-45457-9_17
35. Chai P, Konstantinov IE, da Fonseca da Silva L, Qureshi M, Wackel P, Knott-Craig C, et al. The American Association for Thoracic Surgery (AATS) 2025 Expert Consensus Document: Management of Ebstein anomaly in children and adults. J Thorac Cardiovasc Surg. 2025;170(1):1-16. doi:10.1016/j.jtcvs.2025.03.014. PMID: 40317284.
36. Konstantinov IE, Chai P, Bacha E, Caldarone CA, Da Silva JP, Da Fonseca Da Silva L, et al. The American Association for Thoracic Surgery (AATS) 2024 expert consensus document: Management of neonates and infants with Ebstein anomaly. J Thorac Cardiovasc Surg. 2024;168(2):311-324. doi:10.1016/j.jtcvs.2024.04.018. PMID: 38685467.
37. European Reference Network for rare, low prevalence and complex diseases of the heart (ERN GUARD-Heart). Available from: https://guardheart.ern-net.eu/ (accessed 31 Oct 2026).
38. Hoedemaekers YM, Germans T. Left ventricular noncompaction. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 115–38. doi:10.1007/978-3-030-45457-9_7
39. Lee AC, Niaz T, Dearani JA, Madhavan M, Sweeney D, Connolly HM, et al. Ebstein anomaly, left ventricular non-compaction, and ventricular tachycardia, associated with a rare MYH7 genetic variant. Int J Cardiol Congenit Heart Dis. 2026;24:100669. doi:10.1016/j.ijcchd.2026.100669. PMID: 42005012. PMC13091117.
40. Attenhofer Jost CH, Connolly HM, O'Leary PW, Warnes CA, Tajik AJ, Seward JB. Left heart lesions in patients with Ebstein anomaly. Mayo Clin Proc. 2005;80(3):361-8. doi:10.4065/80.3.361. PMID: 15757018.
41. Budde BS, Binner P, Waldmüller S, Höhne W, Blankenfeldt W, Hassfeld S, et al. Noncompaction of the ventricular myocardium is associated with a de novo mutation in the beta-myosin heavy chain gene. PLoS One. 2007;2(12):e1362. doi:10.1371/journal.pone.0001362. PMID: 18159245. PMC2137931.
42. Dentice M, Cordeddu V, Rosica A, Ferrara AM, Santarpia L, Salvatore D, et al. Missense mutation in the transcription factor NKX2-5: a novel molecular event in the pathogenesis of thyroid dysgenesis. J Clin Endocrinol Metab. 2006;91(4):1428-33. doi:10.1210/jc.2005-1350. PMID: 16418214.
43. van der Zwaag PA, van den Berg MP. Introduction to hereditary cardiomyopathies. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 53–61.
44. Sicko RJ, Browne ML, Rigler SL, Druschel CM, Liu G, Fan R, et al. Genetic Variants in Isolated Ebstein Anomaly Implicated in Myocardial Development Pathways. PLoS One. 2016;11(10):e0165174. doi:10.1371/journal.pone.0165174. PMID: 27788187. PMC5082909.
45. Giannakou A, Sicko RJ, Zhang W, Romitti P, Browne ML, Caggana M, et al. Copy number variants in Ebstein anomaly. PLoS One. 2017;12(12):e0188168. doi:10.1371/journal.pone.0188168. PMID: 29216221. PMC5720705.
46. Downing KF, Riehle-Colarusso T, Gilboa SM, Lin AE, Oster ME, Tinker SC, et al. Potential risk factors for Ebstein anomaly, National Birth Defects Prevention Study, 1997-2011. Cardiol Young. 2019;29(6):819-827. doi:10.1017/S1047951119000970. PMID: 31159903. PMC6711372.
47. Online Mendelian Inheritance in Man, OMIM®. MIM 160760 — Myosin, heavy chain 7, cardiac muscle, beta; MYH7. Available from: https://www.omim.org/entry/160760 (accessed 31 Oct 2026).
48. Online Mendelian Inheritance in Man, OMIM®. MIM 191010 — Tropomyosin 1; TPM1. Available from: https://www.omim.org/entry/191010 (accessed 31 Oct 2026).
49. Kuroda Y, Nagai K, Kawai Y, Wakamiya T, Naruto T, Kurosawa K. Recurrent FLNA p.Gly1554Arg Variant Associated With Familial Ebstein Anomaly and Joint Stiffness. Am J Med Genet A. 2025;197(5):e63973. doi:10.1002/ajmg.a.63973. PMID: 39704215.
50. Online Mendelian Inheritance in Man, OMIM®. MIM 300017 — Filamin A; FLNA. Available from: https://www.omim.org/entry/300017 (accessed 31 Oct 2026).
51. Online Mendelian Inheritance in Man, OMIM®. MIM 305620 — Frontometaphyseal dysplasia 1; FMD1. Available from: https://www.omim.org/entry/305620 (accessed 31 Oct 2026).
52. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 76–9 (1p36 deletion syndrome).
53. Pierpont ME, Brueckner M, Chung WK, Garg V, Lacro RV, McGuire AL, et al. Genetic Basis for Congenital Heart Disease: Revisited: A Scientific Statement From the American Heart Association. Circulation. 2018;138(21):e653-e711. doi:10.1161/CIR.0000000000000606. PMID: 30571578. PMC6555769.
54. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 660–5 (Deletions and duplications).
55. Online Mendelian Inheritance in Man, OMIM®. MIM 607872 — Chromosome 1p36 deletion syndrome. Available from: https://www.omim.org/entry/607872 (accessed 31 Oct 2026).
56. Freud LR, Escobar-Diaz MC, Kalish BT, Komarlu R, Puchalski MD, Jaeggi ET, et al. Outcomes and Predictors of Perinatal Mortality in Fetuses With Ebstein Anomaly or Tricuspid Valve Dysplasia in the Current Era: A Multicenter Study. Circulation. 2015;132(6):481-9. doi:10.1161/CIRCULATIONAHA.115.015839. PMID: 26059011. PMC7086479.
57. Weiss K, Lachlan K. CHD4 neurodevelopmental disorder. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2020 Sep 3. Available from: https://www.ncbi.nlm.nih.gov/books/NBK561516/
58. Garavelli L, Contrò G. CDK13-related disorder. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2019 Jan 31; last update 2026 Sep 1. Available from: https://www.ncbi.nlm.nih.gov/books/NBK536784/
59. Nabais Sá MJ, Gabriele M, Testa G, de Vries BBA. Gabriele-de Vries syndrome. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2019 May 30. Available from: https://www.ncbi.nlm.nih.gov/books/NBK541730/
60. Li C, Slavotinek A. FREM1 autosomal recessive disorders. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2008 Jul 9; last update 2025 May 1. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1728/
61. Ho SKL, Leung LT, Luk HM, Lo IFM. Strømme syndrome. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2022 Nov 10. Available from: https://www.ncbi.nlm.nih.gov/books/NBK586170/
62. Lippa N, Mulhern M, Ernst Florido M, Earley C, Sands TT. CSNK2B-related neurodevelopmental disorder. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2024 Sep 5. Available from: https://www.ncbi.nlm.nih.gov/books/NBK606532/
63. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 718–21 (Drugs in pregnancy).
64. GeneGuidance. Curation record for the MYH7 gene. GeneGuidance database; accessed 31 Oct 2026.
65. Cowan JR, van Spaendonck-Zwarts KY, Hershberger RE. Dilated cardiomyopathy. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 77–97 (Table 5.2, p. 80–82). doi:10.1007/978-3-030-45457-9_5
66. Schade van Westrum SM, Wahbi K, Bonne G, van der Kooi AJ. Hereditary neuromuscular diseases and cardiac involvement. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 371–85 (Myosin storage myopathy, p. 378). doi:10.1007/978-3-030-45457-9_22
67. GeneGuidance. Curation record for the TPM1 gene. GeneGuidance database; accessed 31 Oct 2026.
68. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-8. doi:10.1002/humu.23348. Online tool https://variantvalidator.org (GRCh38, MANE Select transcripts), accessed 31 Oct 2026.
69. Bettinelli AL, Mulder TJ, Funke BH, Lafferty KA, Longo SA, Niyazov DM. Familial ebstein anomaly, left ventricular hypertrabeculation, and ventricular septal defect associated with a MYH7 mutation. Am J Med Genet A. 2013;161A(12):3187-90. doi:10.1002/ajmg.a.36182. PMID: 23956225.
70. National Center for Biotechnology Information. ClinVar: VCV000042968, NM_000257.4(MYH7):c.3658_3660del (p.Glu1220del). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/42968/ (accessed 31 Oct 2026).
71. National Center for Biotechnology Information. ClinVar: VCV000043414, NM_001018005.2(TPM1):c.337C>G (p.Leu113Val). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/43414/ (accessed 31 Oct 2026).
72. Vanlerberghe C, Petit F. Holt-Oram syndrome. In: Adam MP, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993–2026. Initial posting 2004 Jul 20; last update 2025 Jul 31. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1111/
73. Online Mendelian Inheritance in Man, OMIM®. MIM 614432 — Ventricular septal defect 3; VSD3. Available from: https://www.omim.org/entry/614432 (accessed 31 Oct 2026).
74. Online Mendelian Inheritance in Man, OMIM®. MIM 187500 — Tetralogy of Fallot; TOF. Available from: https://www.omim.org/entry/187500 (accessed 31 Oct 2026).
75. Online Mendelian Inheritance in Man, OMIM®. MIM 217095 — Conotruncal heart malformations; CTHM. Available from: https://www.omim.org/entry/217095 (accessed 31 Oct 2026).
76. Online Mendelian Inheritance in Man, OMIM®. MIM 614435 — Hypoplastic left heart syndrome 2; HLHS2. Available from: https://www.omim.org/entry/614435 (accessed 31 Oct 2026).
77. Online Mendelian Inheritance in Man, OMIM®. MIM 225250 — Hypothyroidism, congenital, nongoitrous, 5; CHNG5. Available from: https://www.omim.org/entry/225250 (accessed 31 Oct 2026).
78. Gallego F, Tadros R, Khairy P. Novel NKX2-5 mutation associated with familial atrial septal defects and atrial and ventricular arrhythmias. HeartRhythm Case Rep. 2026;12(8):962-965. doi:10.1016/j.hrcr.2026.04.025. PMID: 42626645. PMC13492331.
79. Verstraeten A, Roos-Hesselink J, Loeys B. Bicuspid aortic valve. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 345–61.
80. National Center for Biotechnology Information. ClinVar: VCV000009008, NM_004387.4(NKX2-5):c.73C>T (p.Arg25Cys). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/9008/ (accessed 31 Oct 2026).
81. National Center for Biotechnology Information. ClinVar: VCV001515823, VCV002452059 and VCV002582145, NM_004387.4(NKX2-5) p.Phe145Leu. Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/1515823/ (accessed 31 Oct 2026).
82. Srivatsav V, Guerin A, Abdollah H, Archer SL. Pathogenic Variants in NKX2-5 Associated With Familial Atrial Septal Defect and Conduction System Disease. JACC Case Rep. 2026:110619. doi:10.1016/j.jaccas.2026.110619. PMID: 42891680.
83. Liu H, Hua J, Xiong Q, Xiong L, Chen Q. A novel NKX2-5 mutation causing atrial septal defect and high-degree atrioventricular block in a Chinese family. ESC Heart Fail. 2026;13(2):xvaf039. doi:10.1093/eschf/xvaf039. PMID: 41711205. PMC13122609.
84. Freud LR, McElhinney DB, Kalish BT, Escobar-Diaz MC, Komarlu R, Puchalski MD, et al. Risk Factors for Mortality and Circulatory Outcome Among Neonates Prenatally Diagnosed With Ebstein Anomaly or Tricuspid Valve Dysplasia: A Multicenter Study. J Am Heart Assoc. 2020;9(21):e016684. doi:10.1161/JAHA.120.016684. PMID: 33076749. PMC7763426.
85. Pandey A, Bromberg B, Menon D. NKX2-5 mutation in a family with ventricular tachycardia and aborted sudden cardiac death. HeartRhythm Case Rep. 2026;12(10):1317-1323. doi:10.1016/j.hrcr.2026.06.025. PMID: 42923806. PMC13687988.
86. National Center for Biotechnology Information. ClinVar: VCV000009018, NM_004387.4(NKX2-5):c.355G>T (p.Ala119Ser), and VCV000009019, NM_004387.4(NKX2-5):c.482G>C (p.Arg161Pro). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/9018/ (accessed 1 Nov 2026).
87. Pérez Matos AJ, Oomen T, van Tintelen JP. The genetics of mitral valve prolapse. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 435. doi:10.1007/978-3-030-45457-9_25
88. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Otopalatodigital syndrome type II. In: Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford: Oxford University Press; 2018. p. 315.
89. van der Zande JA, Tutarel O, Ramlakhan KP, van der Bosch AE, Bordese R, Zengin E, et al. Pregnancy outcomes in women with Ebstein's anomaly: data from the Registry of Pregnancy And Cardiac disease (ROPAC). Open Heart. 2023;10(2):e002406. doi:10.1136/openhrt-2023-002406. PMID: 37550057. PMC10407418.
90. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 112–5 (Congenital heart disease).
91. Nannenberg EA, Arens Y. Prenatal diagnosis and preimplantation genetic testing for inherited cardiac diseases. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. p. 457–61. doi:10.1007/978-3-030-45457-9_27
92. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 35, section 35.3.8 Cardiolipin remodelling enzyme (TAZ) deficiency: Barth syndrome.
93. Freud LR, Wilkins-Haug LE, Beroukhim RS, LaFranchi T, Phoon CK, Glickstein JS, et al. Effect of In Utero Non-Steroidal Anti-Inflammatory Drug Therapy for Severe Ebstein Anomaly or Tricuspid Valve Dysplasia (NSAID Therapy for Fetal Ebstein anomaly). Am J Cardiol. 2021;141:106-112. doi:10.1016/j.amjcard.2020.11.013. PMID: 33217351.
94. Mustafa HJ, Aghajani F, Bairmani ZA, Khalil A. Transplacental non-steroidal anti-inflammatory drugs versus expectant management in fetal Ebstein anomaly with circular shunt: Systematic review and meta-analysis. Prenat Diagn. 2024;44(6-7):773-782. doi:10.1002/pd.6446. PMID: 37902170.
95. World Health Organization. WHO recommendations on antenatal corticosteroids for improving preterm birth outcomes. Geneva: World Health Organization; 2022. ISBN 978-92-4-005729-6. Available from: https://iris.who.int/handle/10665/363130 (accessed 31 Oct 2026).
96. U.S. Food and Drug Administration. PROSTIN VR PEDIATRIC (alprostadil) sterile solution: prescribing information (version of 17 Aug 2026; NDA 018484). Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=930f8c02-46a8-4cff-a484-660468ab23c3 (openFDA; accessed 31 Oct 2026).
97. U.S. Food and Drug Administration. INOMAX (nitric oxide) gas, for inhalation: prescribing information (version of 12 Dec 2025; NDA 020845). Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=762b51be-1893-4cd1-9511-e645fc420d3a (openFDA; accessed 31 Oct 2026).
98. European Medicines Agency. INOmax 400 ppm mol/mol medicinal gas, compressed: summary of product characteristics (EPAR, EMEA/H/C/000337). Available from: https://www.ema.europa.eu/en/documents/product-information/inomax-epar-product-information_en.pdf (accessed 31 Oct 2026).
99. Zhu Y, Jian Z, Ma R, Xiao Y. Timely surgical intervention and risk stratification in patients with Ebstein anomaly: a 20-year retrospective cohort study. Int J Surg. 2026;112(4):10104-16. doi:10.1097/JS9.0000000000004760. PMID: 41609396. PMC13105781.
100. Vahanian A, Beyersdorf F, Praz F, Milojevic M, Baldus S, Bauersachs J, et al. 2021 ESC/EACTS Guidelines for the management of valvular heart disease. Eur Heart J. 2022;43(7):561-632. doi:10.1093/eurheartj/ehab395.
101. U.S. Food and Drug Administration. Warfarin sodium tablets (Teva Pharmaceuticals USA): prescribing information (version of 17 Jun 2025; ANDA 040616). Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=0cbce382-9c88-4f58-ae0f-532a841e8f95 (openFDA; accessed 31 Oct 2026).
102. U.S. Food and Drug Administration. ELIQUIS (apixaban) tablets: prescribing information (version of 24 Apr 2026; NDA 202155). Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=a454cd24-0c6d-46e8-b1e4-197388606175 (openFDA; accessed 31 Oct 2026).
103. U.S. Food and Drug Administration. MYQORZO (aficamten) tablets: prescribing information (version in force since 16 Dec 2025; NDA 219083). Available from: https://api.fda.gov/drug/label.json?search=openfda.generic_name:aficamten (openFDA; accessed 31 Oct 2026).
104. ClinicalTrials.gov. NCT02914171: Study of Autologous Bone Marrow Derived Mononuclear Cells for Treatment of Ebstein Anomaly. Available from: https://clinicaltrials.gov/study/NCT02914171 (accessed 31 Oct 2026).
105. O'Leary PW, Qureshi MY, Cetta F, Nelson TJ, Holst KA, Dearani JA. Cone Reconstruction for Ebstein Anomaly: Ventricular Remodeling and Preliminary Impact of Stem Cell Therapy. Mayo Clin Proc. 2021;96(12):3053-3061. doi:10.1016/j.mayocp.2021.02.015. PMID: 34479739.
106. ClinicalTrials.gov. NCT05225311: Fetal Ebstein Anomaly and Tricuspid Valve Dysplasia Registry. Available from: https://clinicaltrials.gov/study/NCT05225311 (accessed 31 Oct 2026).
107. National Center for Biotechnology Information. ClinVar: VCV003257530, NM_004387.4(NKX2-5):c.544G>C (p.Val182Leu). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/3257530/ (accessed 31 Oct 2026).
108. Alhmouz MM, Saeid AJ, Jaabari MAL, Bany Ata RH, Ali MB, Abuhalimeh R, et al. NKX2-5 Mutation and ICD Implantation in a Pregnant AV Block Patient. JACC Case Rep. 2025;30(35):105528. doi:10.1016/j.jaccas.2025.105528. PMID: 41203391. PMC12790246.
109. Ko T, Nishijo D, Tsuji-Hosokawa A, Tsuchiya I, Nomura S, Zhang B, et al. Correction of the Murine Model of Congenital Heart Disease Associated With the Nkx2-5 Mutation Using Prime Editing. Circ Genom Precis Med. 2025;18(4):e005194. doi:10.1161/CIRCGEN.124.005194. PMID: 40485469.
110. National Center for Biotechnology Information. ClinVar: VCV000635975, NM_001110556.2(FLNA):c.4660G>A (p.Gly1554Arg). Available from: https://www.ncbi.nlm.nih.gov/clinvar/variation/635975/ (accessed 1 Nov 2026).
111. De Backer J, Haugaa KH, Hasselberg NE, de Hosson M, Brida M, Castelletti S, et al. 2025 ESC Guidelines for the management of cardiovascular disease and pregnancy. Eur Heart J. 2025;46(43):4462-4568. doi:10.1093/eurheartj/ehaf193. PMID: 40878294.
112. Katsuragi S, Kamiya C, Yamanaka K, Neki R, Miyoshi T, Iwanaga N, et al. Risk factors for maternal and fetal outcome in pregnancy complicated by Ebstein anomaly. Am J Obstet Gynecol. 2013;209(5):452.e1-6. doi:10.1016/j.ajog.2013.07.005. PMID: 23860210.
113. Gurrieri C, Sharpe EE, Connolly HM, Rose CH, Arendt KW. Peripartum anesthetic management in patients with Ebstein anomaly: a case series. Proc (Bayl Univ Med Cent). 2023;36(3):346-350. doi:10.1080/08998280.2023.2169561. PMID: 37091760. PMC10120540.
114. Munk-Olsen T, Liu X, Viktorin A, Brown HK, Di Florio A, D'Onofrio BM, et al. Maternal and infant outcomes associated with lithium use in pregnancy: an international collaborative meta-analysis of six cohort studies. Lancet Psychiatry. 2018;5(8):644-652. doi:10.1016/S2215-0366(18)30180-9. PMID: 29929874. PMC6077091.
115. Fornaro M, Maritan E, Ferranti R, Zaninotto L, Miola A, Anastasia A, et al. Lithium Exposure During Pregnancy and the Postpartum Period: A Systematic Review and Meta-Analysis of Safety and Efficacy Outcomes. Am J Psychiatry. 2020;177(1):76-92. doi:10.1176/appi.ajp.2019.19030228. PMID: 31623458.
116. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 947–9 (Multifactorial inheritance).
117. Holanda RP, Ribeiro EM. Cardiopatia congênita. In: Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 155–9.
118. Adult Congenital Heart Association. Available from: https://www.achaheart.org/ (accessed 31 Oct 2026).
119. Conquering CHD. Available from: https://www.conqueringchd.org/ (accessed 31 Oct 2026).
120. European Congenital Heart Disease Organisation (ECHDO). Available from: https://www.echdo.eu/ (accessed 31 Oct 2026).
121. Children's Heart Federation. Available from: https://www.chfed.org.uk/ (accessed 31 Oct 2026).
122. ACTC — Casa do Coração (Associação de Assistência à Criança e ao Adolescente Cardíacos e aos Transplantados do Coração). Available from: https://actc.org.br/ (accessed 31 Oct 2026).