Chronic and Recurrent Otitis Media – Genetic Predisposition

A complex trait in which susceptibility to recurrent acute otitis media, chronic otitis media with effusion and chronic suppurative otitis media results from the combination of common variants of small effect, rare variants of incomplete penetrance and environmental factors, alongside a large group of syndromes and inborn errors of immunity in which otitis media is a frequent manifestation.

Literature and regulatory information updated: 29 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Definitions and scope

Otitis media (OM) denotes inflammation of the middle ear and encompasses a spectrum of conditions that follow one another or coexist in the same child. Acute otitis media (AOM) is the acute infectious episode, with bulging and erythema of the tympanic membrane, otalgia and fever. Recurrent acute otitis media (rAOM) is defined as three or more distinct episodes in six months or four or more in twelve months; the 2013 American Academy of Pediatrics (AAP) guideline requires the episodes to be well documented and separate and, for the twelve-month criterion, at least one to have occurred in the preceding six months [2, 40, 46]. Otitis media with effusion (OME) is the presence of fluid in the middle ear without signs of acute inflammation; when it persists for three months or more it is termed chronic otitis media with effusion (COME) [2]. Chronic suppurative otitis media (CSOM) is characterised by tympanic membrane perforation with otorrhoea and hearing loss [29]. In genetic research, the otitis-prone phenotype was operationalised, for example in the Galveston (Texas) cohort, as a first episode before six months, three or more AOM episodes in six months, four or more in twelve months, six or more by six years, or tympanostomy tube placement for recurrent or persistent OM; the comparison group included children with zero to two episodes by two years [1].

This article covers two realities with distinct mechanisms and implications. The first is non-syndromic predisposition, a complex trait with an estimated heritability of between 40% and 70%, in which common variants of small effect, rare variants of incomplete penetrance and environmental exposures combine [2, 8, 47]. This is the OMIM 166760 entity, whose braces indicate a susceptibility locus rather than a Mendelian disease [7, 16]. The second is the class of syndromes and inborn errors of immunity in which OM is a frequent manifestation, where recognition of the underlying disease modifies investigation, treatment and counselling; these conditions are presented here with the information that concerns the middle ear.

1.2. Historical overview

Familial aggregation of OM was documented in cohort studies before the molecular era. In the prospective Boston cohort, followed from birth to seven years (877 children observed for at least one year), a history of recurrent AOM in a sibling was, alongside male sex, early first episode and absence of breastfeeding, an independent risk factor for AOM and for prolonged effusion [36]. Twin studies subsequently quantified the genetic component: Casselbrant and colleagues (Pittsburgh, recruitment between 1982 and 1995) estimated a heritability of 0.73 for time with effusion in the first two years [33]; Rovers and colleagues, in a population cohort of twins born in England and Wales in 1994, obtained values between 0.49 and 0.71 between two and four years of age [48]; Kvestad and colleagues, in 4247 Norwegian twin pairs, attributed 72% of the variance in boys and 61% in girls to additive genetic effects [34].

The first mapping phase relied on linkage analyses in families with COME or rAOM, which pointed to regions on 10q, 19q and 3p [49], and on candidate gene studies chosen for biological plausibility or from mouse models, with associations in about 21 genes (among them FBXO11, TLR4 and TNF), few of which were replicated [2]. Mouse models with spontaneous OM, such as the dominant Junbo mutant (gene Evi1) and mice with loss of Eya4, Tlr4, p73, MyD88, Fas, E2f4, Plg or Fbxo11, provided candidate pathways [47, 50]. Genome-wide association studies (GWAS) began in 2012 with the Australian Raine cohort [51] and extended to North American family cohorts [52], the Finnish cohort [53], the first AOM GWAS with genome-wide significance, at the FNDC1 gene [10], and biobanks with tens of thousands of cases [9, 11].

The landmark of the rare-variant era was the description, in 2015, of a duplication in A2ML1 co-segregating with OM in an Indigenous Filipino community with a prevalence close to 50%, with replication in North American children of European and Hispanic ancestry [1]. This was followed by FUT2 in 2018 [8], SPINK5 [54], CDHR3 [55], SLPI [56] and a set of 24 genes in families with OM [12]. In parallel, the genetic explanation of OM in syndromes advanced with the identification, in mouse models, of DYRK1A triplication as the determinant of OM in trisomy 21 [14] and with the population association of TBX1, a gene in the 22q11.2 region, with ear infections and myringotomy [9].

1.3. Nomenclature

OMIM records a single phenotype, “{Otitis media, susceptibility to}” (166760), mapped to A2ML1 at 12p13.31 with autosomal dominant inheritance and molecular evidence level 3 [7]. Entry 166760, created in 1987 with the title “Otitis media, chronic/recurrent” (COME/ROM), uses the number symbol because there is evidence that variation in A2ML1 confers susceptibility, and brings together familial aggregation, twin studies, linkage to 10q26.3, 19q13.42–q13.43 and 3p25.3, and the Eya4 mouse model [57]. The OMIM entry for FUT2 does not list OM among its phenotypes, which reflects the associative nature of the available data [19, 58]. Orphanet has no entity for susceptibility to OM; the associated syndromes have their own codes [20, 24]. In the English-language literature, “glue ear” denotes OME, and “chronic otitis media” may refer to COME or CSOM depending on the study, which makes it necessary to read each cohort's case definition before comparing results [2, 11].

2. Molecular Genetics and Pathophysiology

2.1. Pathophysiology: Eustachian tube, mucosa and microbiome

The middle ear is a cavity lined by respiratory mucosa that communicates with the nasopharynx through the Eustachian tube. OM results from the interaction between the pathogenicity of the agents (mainly Streptococcus pneumoniae, non-typeable Haemophilus influenzae and respiratory viruses), host immune status, the anatomy and function of the tube, mucociliary clearance, mucus production and environmental exposures [59, 60]. The genes identified in animal models and in human studies fall into three groups: innate immunity signalling pathways and the mucosal inflammatory response; cellular function, such as mucociliary transport, mucin production and mucosal metaplasia; and development of the tube, cavitation and middle ear homeostasis [59]. Studies of families with rare variants added a fourth dimension, the composition of the middle ear and nasopharyngeal microbiome, which differs according to genotype at FUT2, SPINK5, CDHR3 and SLPI [8, 54–56].

In the syndromes, the mechanisms are largely anatomical. In 22q11.2 deletion, smaller sinuses and ostia, narrow nasal passages, Eustachian tube anomalies, submucous cleft and velopharyngeal insufficiency increase susceptibility to ear infections; palatal dysfunction, present in more than half of individuals, allows nasal regurgitation and reflux of milk into the tube [32]. In cleft palate, OME is almost constant and its frequency increases with the extent of the cleft [30, 61]. In motile ciliopathies, failure of mucociliary clearance sustains the effusion [5]. In humoral immunodeficiencies, the lack of antibodies against capsular antigens explains the recurrence of OM due to S. pneumoniae and H. influenzae [6].

2.2. Heritability and genetic architecture

Twin estimates measure the proportion of the variance of a phenotype, in a given population and period, attributable to genetic differences, and do not translate into individual risk. Differences between studies reflect the case definition (prospectively measured time with effusion, symptom score, recalled history), age and design: the Finnish family estimates, obtained by questionnaire in operated children and their parents, are lower than the twin estimates [35]. In the British study, the shared-environment component fell from 0.41 at two years to 0.16 at four, and chronic blockage proved more heritable than acute infection, indicating that the genetic component is distributed differently across the phases of the disease [48].

The architecture emerging from molecular studies combines common loci with odds ratios (OR) between 1.05 and 1.2, identified in GWAS, with rare variants of larger effect but incomplete penetrance, identified in families and in founder populations [1, 9]. In the multi-ethnic families studied by Santos-Cortez and colleagues, intrafamilial heterogeneity was marked: affected members of the same family carried different combinations of common and rare variants, which led the authors to describe the individual phenotype as a complex trait [8].

2.3. Genes with rare variants in families

The A2ML1, FUT2, SPINK5, CDHR3, SLPI and EYA4 variants were normalised in VariantValidator (GRCh38) with the MANE Select transcript, whose reference sequence matched the authors' nomenclature [68]. The BSN, FBXO11 and PLG variants match the MANE transcripts (NM_003458.4, NM_001190274.2, NM_000301.5); the CHD7 variant was published on an alternative transcript (NM_001316690.1) and corresponds to NM_017780.4:c.8488G>A p.(Ala2830Thr) on MANE [67, 68].

2.4. A2ML1

A2ML1 encodes alpha-2-macroglobulin-like 1, a secreted protease inhibitor of the alpha-macroglobulin family that captures proteases of all four catalytic classes by a bait-and-trap mechanism: cleavage of the bait region triggers a conformational change and covalent capture of the protease through a thioester bond [16]. The protein is 41% identical to alpha-2-macroglobulin (A2M), a marker of vascular permeability of the middle ear mucosa during infection [1]. In the mouse, A2ML1 immunoreactivity is located in the epithelium of the middle ear mucosa and is virtually absent from the inner ear [1, 7].

In the Indigenous Filipino community studied (about 250 people, of Negrito and Indo-Malay ancestry, with intermarriage), otoscopy of 175 members identified current or previous OM in 82 (46.9%) [1]. Exome sequencing of two second cousins with CSOM and Sanger sequencing of 51 members of a 134-person pedigree showed that only the duplication NM_144670.6:c.2478_2485dup p.(Ser829TrpfsTer9) (published as c.2478_2485dupGGCTAAAT) co-segregates with the disease, with a two-point LOD of 7.5 under an autosomal dominant model with 95% penetrance and 5% phenocopies [1]. The duplication shifts the reading frame, is predicted to cause mRNA degradation by nonsense-mediated decay and loss of the thioester-binding and receptor-binding domains; it was absent from 61,109 Exome Aggregation Consortium (ExAC) samples, from 1385 exomes of the University of Washington Center for Mendelian Genomics and from 100 Southeast Asian genomes [1].

The same duplication was found in three of 123 otitis-prone children followed from birth in Galveston (two homozygotes, one of European and one of Hispanic ancestry, and one heterozygote of European ancestry) and in none of 118 children without recurrent OM; all three had severe early-onset OM, with tympanostomy tubes before six months, and share with the Filipino individuals the haplotype on which the variant lies [1]. Seven other variants were identified (three nonsense and four missense), each in a single proband and with only nominal significance [1]. One missense variant, NM_144670.6:c.887T>C p.(Val296Ala), which segregated in a Minnesota family, was present in 96 ExAC alleles (European frequency 0.0013) and was considered by the authors probably not causal, illustrating the need to filter by population frequency [1].

In the Filipino pedigree, 37 carriers had the full spectrum of OM, aged between three months and 58 years, and nine carriers (median 18 years) had no evidence of disease; one non-carrier woman had healed OM, documenting phenocopies [1]. In a multivariate analysis of the same community, A2ML1 genotype was the strongest predictor of OM (OR 3.7; 95% confidence interval, CI, 1.3–10.8), with no association with sex, body mass index, breastfeeding, tobacco exposure or diving [69]. Larson and colleagues described 16 additional variants in Filipino, North American, Finnish and Pakistani individuals, found an independent association of gingivitis and A2ML1 variants with OM, and observed lower salivary expression of A2ML1 in carriers, with overexpression of RND3, suggesting a role in the differentiation of middle ear keratinocytes [62]. The CRISPR A2ml1 knockout mouse develops spontaneous OM (OR 11; 95% CI 1.1–573.6, in 29 animals assessed) without craniofacial changes, with overexpression of Dsp in the middle ear [70].

A2ML1 was also proposed as a Noonan syndrome gene, but the ClinGen RASopathy expert panel classified this association as disputed in 2018, and the variants identified in individuals tested for RASopathy were inherited from unaffected parents where testable; carriers of OM-associated variants do not have the cardinal features of Noonan syndrome [1, 16]. The protein is also the p170 autoantigen of paraneoplastic pemphigus, an acquired disease unrelated to germline variation [16].

2.5. FUT2 and secretor status

FUT2 encodes secretor fucosyltransferase, a Golgi enzyme that adds fucose in alpha-1,2 linkage to the terminal galactose of epithelial and secreted glycoproteins and glycolipids and forms the H antigen, the precursor of the soluble A, B and Lewis b antigens; it determines secretor status and does not contribute to the erythrocyte H antigen, which depends on FUT1 [58]. About 20% of people, in most populations studied, are non-secretors through homozygosity or compound heterozygosity for null alleles; in Europeans, the usual null allele is NM_000511.6:c.461G>A p.(Trp154Ter) (rs601338, historically 428G>A or W143X), with an allele frequency of 0.44–0.47, virtually absent in Koreans and Japanese, in whom other null alleles are responsible [58, 68].

In 609 multi-ethnic families and isolated cases with OM, the variant NM_000511.6:c.604C>T p.(Arg202Ter) co-segregated with OM in a Filipino pedigree (LOD 4.0), the rare variant c.412C>T p.(Arg138Cys) was associated with recurrent or chronic OM in children of European ancestry (p=1.2×10⁻⁵) and was over-transmitted in trios (p=0.01), and c.461G>A was also over-transmitted in trios (p=0.01) and associated with a different and more diverse middle ear microbiota [8]. The variants p.(Ala104Val) (c.311C>T), p.(Arg138Cys), p.(Trp154Ter) (c.461G>A) and p.(Arg202Ter) reduced A antigen in transfected cells, and Fut2 increased transiently in the murine middle ear after inoculation with non-typeable H. influenzae [8]. In c.461G>A carriers, saliva showed downregulation of FN1, KMT2D, MUC16 and NBPF20, and the middle ear a greater load of potential otopathogens [63].

The direction of the FUT2 effect differs between designs. In the 23andMe GWAS (46,936 cases and 74,874 controls of European ancestry, self-reported childhood ear infections), the risk allele of the strongest signal, rs681343 (in almost complete linkage disequilibrium with rs601338), corresponds to the functional secretor allele, with OR 1.11 (95% CI 1.10–1.12; p=3.51×10⁻³⁰), consistent with the over-representation of secretors in upper respiratory infections [9]. In the family studies, by contrast, it was the non-secretor allele that was over-transmitted [8]. Possible explanations include differences in phenotype (self-report of any ear infection in the general population vs operated recurrent or chronic OM), population and pathogens, and the sources consulted do not resolve the discrepancy; FUT2 genotype alone has no established individual predictive value [8, 9, 58].

2.6. Other genes and variants

In a Filipino pedigree with 81 affected and 53 unaffected members, which co-segregated two A2ML1 variants, a common FUT2 variant and a rare SPINK5 variant, NM_006846.4:c.1682A>G p.(Glu561Gly) (LOD 4.09), eight new SPINK5 variants were subsequently found in 12 families and individuals with OM; the protein is strongly localised in the skin of the external auditory canal and weakly in the middle ear mucosa and tympanic membrane, and is increased in cholesteatoma [54]. CDHR3 is expressed only in the ciliated epithelial cells of the middle ear and decreases after AOM; the splice variant c.1653+3G>A was found in nine families from the USA and Finland, and the common variant p.(Cys529Tyr), known to increase rhinovirus C binding to the epithelium, was associated with absence of effusion at surgery and with differences in microbiota [55]. A SLPI splice variant, c.394+1G>T, co-segregates with OM in a large pedigree (LOD 4.59) and causes alternative splicing detectable in saliva [56].

The largest family study to date analysed the exomes of 287 individuals from 243 families and identified 33 variants in 24 genes co-segregating with OM in 28 families, 18 of them in 10 genes classified as pathogenic or likely pathogenic; 81 additional variants in 21 of these genes were found in 83 unrelated probands, and the common EYA4 variant c.829G>A was associated with OM in a case–control cohort; 22 of the 24 genes form a network enriched in signalling pathways, Th1, Th2 and Th17 lymphocyte differentiation and viral infections [12]. PLG, a GWAS signal, was not associated with OM in 314 families analysed by transmission disequilibrium test (TDT), and its expression in the middle ear is very low [9, 71]. In 214 Finnish probands, blood group O was protective against rAOM and group A was associated with a higher risk of COME [72].

Candidate gene studies of innate immunity gave inconsistent results. In 348 children with two or more AOM episodes, IL6, TNFA and TLR4 genotypes were associated with the otitis-prone condition, and TNFA genotypes with a lower anticapsular response to pneumococcal conjugate vaccine [73]; the TLR4 association in Finland was not replicated in British and North American cohorts [74]. The review by Mittal and colleagues summarises this set of associations in cytokine and innate immunity genes [75].

2.7. Genome-wide association studies

The GWAS pattern shows small individual effects and an overlap between immune and developmental pathways. In 23andMe, bacterial infections, including ear infections, were associated mainly with HLA class II variation, with the strongest signal at residue Gln96 of HLA-DRB1 [9]. TBX1, a gene in the 22q11.2 region whose haploinsufficiency in the mouse disrupts middle and outer ear development, was associated with ear infections and myringotomy (OR 1.17) [9] and again with mucoid OM in FinnGen [11], linking general-population susceptibility to the mechanism of OM in 22q11.2 deletion. In FinnGen, annexin A13 is expressed at the apical pole of the Eustachian tube epithelium and GAS2L2 participates in ciliated epithelium function, so the authors relate the two loci to tubal and mucosal dysfunction [11]. Common limitations are the heterogeneity of case definitions (diagnostic codes, self-report, surgical criteria), the predominance of European ancestry and inconsistent replication across populations [9, 53].

2.8. Animal models

Mouse mutants with spontaneous OM have provided candidate genes and mechanisms. The Junbo mutant, with a missense variant in the zinc finger domain of Evi1 (MECOM), develops CSOM with otorrhoea after spontaneous postnatal AOM, with severity dependent on microbiological status and air quality [50]. The mouse lacking Splunc1 (BPIFA1), an abundant secreted protein of the nasal, oral and respiratory mucosae, has a higher frequency of OM with accumulation of neutrophils and mucus and remodelling of the middle ear wall [77]. The 50 Eya4-null mice studied by Depreux and colleagues (2008) had severe hearing loss, hypervascularisation and retraction of the tympanic membrane and middle ear effusion, with an abnormal middle ear cavity and Eustachian tube dysmorphology, whereas none of the 50 controls had these changes; the authors proposed that genes regulating the anatomy of the middle ear and the tube contribute to human susceptibility [57]. Other models include Tlr4, p73, MyD88, Fas, E2f4, Plg and Fbxo11 [47] and the A2ml1 knockout mouse [70].

2.9. Genetic basis of syndromic otitis media

In trisomy 21, Tateossian and colleagues (eLife, 2025) showed in mouse models that OM is associated with the extra copy of a region of murine chromosome 16 homologous to human chromosome 21, refined to 12 genes that include Dyrk1a; DYRK1A overexpression increased interleukins 6 and 17 (IL-6, IL-17) in middle ear fluid, vascular endothelial growth factor (VEGF) signalling and vascular permeability, and reduced cilia, and in samples from children with Down syndrome DYRK1A was the most overexpressed of the 12 genes [14]. The original study was read through the commentary published in the same journal. In 22q11.2 deletion, the predominant mechanism is anatomical and the frequency of infections does not correlate with the T-lymphocyte count; TBX1 haploinsufficiency and the population association of this gene with ear infections suggest a contribution of middle ear development [9, 32]. Conversely, mice overexpressing four genes of the region, including TBX1, developed chronic OM, sensorineural hearing loss and malformations of the middle and inner ear resembling the Mondini dysplasia described in velocardiofacial syndrome (Funke and colleagues, 2001) [78]. In the monogenic forms of immunodeficiency and PCD, the defect is that of the underlying disease [5, 6].

2.10. Genes shared with other diseases

Several susceptibility genes are also Mendelian disease genes, which has implications for report interpretation. EYA4 is associated in OMIM with autosomal dominant deafness 10 (DFNA10, 601316) and dilated cardiomyopathy 1J (605362) [66]. FBXO11, a candidate from a mouse model, causes an autosomal dominant neurodevelopmental disorder (OMIM 618089), with cleft palate or bifid uvula in 3 of 20 patients [47, 79]. PLG in compound heterozygosity or homozygosity causes type 1 plasminogen deficiency (OMIM 217090), with ligneous pseudomembranes that can involve the middle ear, and the variant NM_000301.5:c.988A>G p.(Lys330Glu) causes hereditary angioedema type 4 [80]. CHD7 is the CHARGE syndrome gene, in which hearing loss has a conductive component due to ossicular defects and OME [67, 81]. In these cases, a heterozygous variant found in the context of OM does not establish the diagnosis of the corresponding Mendelian disease, and the converse also applies [16, 79].

3. Clinical Spectrum and Manifestations by System

3.1. Natural history of otitis media in the general population

OM is concentrated in the first years of life. In the Boston cohort, 62% of children had had at least one AOM episode by 12 months and 83% by three years; three or more episodes occurred in 17% by 12 months and in 46% by three years, with peak incidence in the second half of the first year; effusion persisted for weeks to months after each episode [36]. The 2025 Italian consensus reports that more than 60% of children under three years have at least one episode and about 24% three or more [40]. In England, between 2003 and 2019 (1,500,686 children, 274,008 episodes in primary care records), the incidence of AOM was 8286.7 per 100,000 person-years before two years, 7951.8 between two and four years and 2184.4 between five and 17 years [82]. OME is the most frequent sequela of AOM, and its chronic form is associated with fluctuating conductive hearing loss; CSOM, with perforation and otorrhoea, predominates in low- and middle-income countries and was associated, in the community studies that measured it, with disabling hearing loss in 62% of cases [29].

Documented environmental and individual risk factors include male sex, a history of recurrent AOM in siblings, early first episode and absence of breastfeeding [36], and, for rAOM, allergies, upper respiratory infections, chronic nasal obstruction and snoring [40]. The 2013 AAP guideline associates recurrence with the winter season, male sex and passive exposure to tobacco smoke, reports that half of children under two years treated for AOM have a new episode within the following six months, and points to symptoms lasting more than 10 days as a possible predictor of recurrence [46]. Global estimates of mortality from complications (21,000 deaths per year in 2005) and of OM-related hearing loss (30.82 per 10,000) concern mainly the regions with the poorest access to care [27].

3.2. Phenotype associated with susceptibility variants

In the Filipino community carrying the A2ML1 duplication, the clinical spectrum was complete, from AOM and OME to chronic otitis media (COM) with perforation and tympanic scarring, with onset in the first year of life and persistence of COM into adulthood, especially in carriers [69]. The otoscopic criteria used were: COM, perforation with defined margins, thickened mucosa and mucopurulent discharge; AOM, hyperaemic membrane with or without perforation; OME, intact, dull or retracted, non-hyperaemic membrane with fluid; and healed OM, membrane with a scar or healed perforation [1]. None of the individuals had craniofacial deformities, heart disease, immune deficiency or co-segregating intellectual disability [1]. In the three North American carrier children, the disease began early and required tubes before six months [1].

The audiological profile of 109 members of the same community showed that A2ML1 and FUT2 variants are associated with high-frequency hearing loss at 4000 Hz, that flat tympanograms predominate at younger ages, and that COM perforations are associated with severe to profound loss at all frequencies [37]. In the Pakistani family with BSN variants, the 12 affected members had CSOM with otorrhoea for at least two to three years and mild to profound hearing loss, without vertigo, mastoiditis or cholesteatoma in the two individuals examined in detail [67]. In carriers of CDHR3 p.(Cys529Tyr), surgery more often showed absence of effusion [55].

3.3. Otitis media in syndromes and inborn errors of immunity

The frequencies in this table come from series with different sizes, ages and definitions and are not directly comparable between syndromes. The reference books (Smith, Oxford Desk Reference) give percentages without an explicit denominator for several syndromes [89, 90]. In 22q11.2 deletion, the cohort of 55 individuals aged nine years or older used a stringent criterion (more than three episodes per year treated with antibiotics in the previous two years) and found recurrent OM in 25%, sinusitis in 27%, bronchitis in 7% and pneumonia in 4%, with 40% as healthy as their peers [32].

3.4. Sequelae and complications

Conductive hearing loss associated with OME is the most common sequela and, when bilateral and prolonged, can affect speech, language and learning; in cleft lip and palate, mean thresholds were 35.71 dB HL (decibels hearing level), and in Down syndrome 36.7 dB HL with bilateral OME versus 28.3 dB HL with at least one aerated ear [30, 31]. Acute complications (mastoiditis, meningitis, intracranial abscess, sinus thrombosis, facial palsy) are rare with or without antibiotics in high-income countries [39]. CSOM leads to chronic perforation and hearing loss, which, in community studies, affected 50–78% of participants [29]. Increased SPINK5 expression in cholesteatoma raises the hypothesis of a role for this gene in chronic destructive disease [54].

4. Clinical Recognition and Red Flags

4.1. When to suspect predisposition

Most children with rAOM or COME have no identifiable underlying disease, and non-syndromic genetic predisposition is, in current clinical practice, an inference from the personal and family history rather than a laboratory diagnosis [2, 12]. Features suggestive of increased susceptibility are a first episode before six months, rAOM, the need for tympanostomy tubes in the first year, persistence of OM beyond preschool age, and a history of recurrent OM in siblings or parents [1, 36]. In communities with intermarriage and a high prevalence of OM, such as the Indigenous Filipino community carrying the A2ML1 variant, familial aggregation is the rule [69].

4.2. Signs pointing to an underlying disease

4.3. At-risk children according to the guidelines

The American Academy of Otolaryngology–Head and Neck Surgery Foundation (AAO-HNSF) guidelines introduced the concept of the child at risk of speech, language or learning problems due to underlying sensory, physical, cognitive or behavioural factors. For these children, the 2016 OME guideline recommends evaluating for OME at the time of diagnosis of the at-risk condition and at 12–18 months if the diagnosis was earlier, and performing hearing testing for OME of any duration; in children not at risk and without symptoms, it recommends against screening for OME [38]. The 2022 tympanostomy tube guideline asks that it be determined whether the child with rAOM or OME is at risk (key action statement 8, recommendation) and allows tubes in at-risk children with unilateral or bilateral OME that is likely to persist, documented by a flat (type B) tympanogram or by effusion of three months or longer (statement 9, option) [41]. The list of risk factors in the 2022 update has eight categories and added that of intellectual disability, learning disorder and attention-deficit/hyperactivity disorder; the panel assigned high confidence to the evidence in trisomy 21, cleft palate and permanent hearing loss, and medium confidence in the remaining categories [41]. In 22q11.2 deletion, humoral evaluation should be brought forward in the presence of recurrent, unusual, refractory or severe infections [32].

Source: Rosenfeld and colleagues (2022), Table 2 [41].

The guideline justifies statement 9 by the low probability of spontaneous resolution: in trials, OME already present for three months resolved in only 19% of ears after a further three months, 25% at six and 31% at twelve, compared with 75–90% resolution at three months for effusion following an AOM episode; in preschool children with a type B tympanogram at screening, normalisation was 20% at three months and 28% at six [41]. A type B tympanogram with an equivalent canal volume between 0.3 and 0.9 cm³ is associated with effusion in 85–100% of ears, and type A with effusion in 3%; a low volume points to cerumen, probe contact or a stenotic canal, frequent in trisomy 21, and a high volume to perforation or a patent tube [41]. In children with severe cognitive delay who cannot cooperate with behavioural testing, hearing loss approaches 50% when auditory brainstem responses under sedation are used [41]. There are no randomised trials of tubes in at-risk children, who are excluded from trials; in an observational study in which about half of the sample was at risk, the likelihood of caregivers reporting marked improvement in speech and language after tubes was 5.1 times higher in at-risk children (95% CI 2.4–10.8), and 3.5 times higher for learning and school performance (95% CI 1.8–7.1), with no difference in other outcomes [41].

4.4. Emergency signs

Signs of severe systemic illness, suspected sepsis or acute complications (mastoiditis, meningitis, intracranial abscess, venous sinus thrombosis, facial palsy) require immediate hospital referral [39]. In children with humoral immunodeficiency, OM may coexist with invasive infections, and about 60% of boys with X-linked agammaglobulinaemia are recognised only when they develop a severe infection [6]. There is no newborn screening directed at susceptibility to OM; newborn screening for T lymphopenia by quantification of T-cell receptor excision circles (TREC), where available, identifies some of the children with 22q11.2 deletion and significant lymphopenia [32].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Diagnosis of otitis media

The diagnosis of OM is clinical and otoscopic. The AAO-HNSF OME guideline recommends documenting effusion by pneumatic otoscopy, performing pneumatic otoscopy in the presence of otalgia or hearing loss, and obtaining tympanometry when the diagnosis remains uncertain (strong recommendations) [38]. Age-appropriate hearing testing is mandatory in OME lasting three months or longer, before surgery, and in OME of any duration in at-risk children [38, 41]. Otoscopy does not reliably distinguish viral from bacterial infection, which limits therapeutic decision-making [40]. The classification of the tympanic membrane and middle ear used in genetic studies (AOM, OME, COM with perforation, healed OM) follows the same otoscopic criteria [1].

5.2. Aetiological investigation

5.3. Genetic testing in non-syndromic susceptibility

Genetic testing for susceptibility to OM remains in the research setting. Variants in A2ML1, FUT2, SPINK5, CDHR3, SLPI and other genes have incomplete penetrance, also occur in carriers without disease and, within the same family, affected members may carry different combinations of variants [1, 8, 12]. The authors of the 24-gene study anticipate the future use of genetic screening to identify children who would benefit from a personalised approach, but present no clinical validation [12]. In a founder population, adding A2ML1/FUT2 genotype to otoscopy improved the validity of screening to rule out moderate to profound hearing loss, which cannot be extrapolated to populations without the founder variant [37]. Population frequency is decisive: the A2ML1 variant p.(Val296Ala), present in 96 European ExAC alleles, was considered non-causal despite segregating in a family [1]. GWAS results (OR of 1.05 to 1.2 per allele) have no individual predictive utility [9].

When an exome performed for another reason identifies a variant in one of these genes, the report should state the gene, the transcript with version, the Human Genome Variation Society (HGVS) nomenclature, the zygosity and the classification according to the American College of Medical Genetics and Genomics and Association for Molecular Pathology (ACMG/AMP) criteria, which were designed for Mendelian diseases and apply poorly to reduced-penetrance susceptibility factors; an A2ML1 variant does not establish Noonan syndrome, and a heterozygous PLG, FBXO11, EYA4 or CHD7 variant does not establish the Mendelian diseases of those genes without a concordant phenotype [16, 79, 80]. A negative test does not exclude predisposition, given the heterogeneity and the polygenic contribution [8].

5.4. Algorithm

5.5. Prenatal and preimplantation diagnosis

The sources consulted do not describe prenatal or preimplantation diagnosis for non-syndromic susceptibility to OM, and the incomplete penetrance and heterogeneity of the variants would not support a predictive interpretation. In the associated syndromes, prenatal diagnosis follows the rules of the underlying disease; cleft lip, for example, can be detected by ultrasound from 18 weeks [124].

6. Differential Diagnosis

The differential diagnosis of recurrent or chronic OM in a child with suspected predisposition is not between diseases that mimic OM, but between the causes that favour it. The aim is not to leave unrecognised an immunodeficiency, a ciliopathy, a palatal anomaly or a syndrome that would change management. The table brings together the main entities (AD, autosomal dominant; AR, autosomal recessive).

The most relevant distinction in practice is between the child with isolated OM, even if recurrent, and the child with OM as part of a pattern of sinopulmonary infections. In X-linked agammaglobulinaemia, OM is the most frequent infection before diagnosis, but it is accompanied by pneumonia, sinusitis and conjunctivitis, and the agents (S. pneumoniae and H. influenzae) may continue to cause OM even on gammaglobulin [6]. In PCD, persistent mucoid otorrhoea after tubes and the coexistence of rhinosinusitis and wet cough from birth are the distinguishing elements [5]; among the patients described in the OMIM entry for PCD type 1, a nine-year-old boy with DNAI1 variants presented in early childhood with chronic sinusitis, serous otitis and recurrent bronchitis [125]. Humoral immunodeficiencies and PCD share the possibility of low nasal nitric oxide, so this test does not separate them [5].

In 22q11.2 deletion, the frequency of infections does not depend on the T-lymphocyte count, and the anatomical mechanism (palate, tube, milk reflux) explains most OM; however, a fraction of patients progress to IgG deficiency resembling common variable immunodeficiency, which justifies periodic humoral assessments [32]. In syndromes with cleft palate, OME is almost constant and its severity increases with the extent of the cleft [61]. In the child with OM and conjunctivitis or gingivitis with thick membranes, plasminogen deficiency is the diagnosis to exclude; ear involvement in ligneous conjunctivitis is described in the literature cited by OMIM [80, 126].

7. Treatment and Risk Management Strategies

7.1. Principles

There is no treatment directed at the genotype of susceptibility to OM, and no registered trial evaluates therapy guided by A2ML1, FUT2 or other gene variants [12, 127]. Management of the child with predisposition follows the general AOM, rAOM and OME guidelines, with three adjustments: recognising and treating the underlying condition when present; identifying the child at risk of speech and language impairment, in whom the thresholds for audiological surveillance and surgery are lower; and weighing prevention (vaccination, control of risk factors, tubes) on the basis of the observed recurrence [38, 40, 41]. The antibiotics used in OM are medicines in routine paediatric use; the doses given follow the guidelines, which are the reference when they diverge from one another.

7.2. Acute otitis media episode

European and North American guidelines converge on analgesia, watchful waiting in children without risk factors and amoxicillin as first choice, but diverge on dose. The National Institute for Health and Care Excellence (NICE), in guideline NG91 (2018, amended in 2022), recommends considering no antibiotic prescription or a delayed prescription in most children, because antibiotics make little difference to symptoms and the frequency of complications, and reserves immediate prescription for children who are systemically very unwell or at high risk of complications, also allowing it in children with otorrhoea or under two years of age with bilateral infection [39]. The 2025 Italian intersociety consensus recommends watchful waiting for 48–72 hours in children over 24 months without comorbidities, risk factors or severity criteria (moderate evidence, strong recommendation) [40]. The 2023 Cochrane meta-analysis (13 trials, 3401 children) showed that antibiotics do not reduce pain at 24 hours (risk ratio, RR, 0.89; 95% CI 0.78–1.01), reduce it slightly at two to three days (RR 0.71; number needed to treat for an additional beneficial outcome, NNTB, 20) and increase adverse effects such as vomiting, diarrhoea or rash (RR 1.38; number needed to treat for an additional harmful outcome, NNTH, 14) [128].

The 2013 AAP clinical practice guideline, developed with the American Academy of Family Physicians (AAFP), applies to children aged 6 months to 12 years with uncomplicated AOM who are otherwise healthy; it expressly excludes those with tympanostomy tubes, anatomical abnormalities such as cleft palate, genetic conditions with craniofacial abnormalities such as Down syndrome, immunodeficiencies or cochlear implants, as well as OME without AOM [46]. Its action statements therefore apply to the otitis-prone child without underlying disease, and not to the populations in section 7.5, whose natural history may be different. Each statement has an aggregate evidence quality grade (A to D, and X in exceptional situations) and a strength: a strong recommendation should be followed unless a clear and compelling rationale for an alternative approach is present, a recommendation should be followed with attention to new information and family preferences, and an option corresponds to questionable evidence or to studies showing no clear advantage of one approach [46].

Source: Lieberthal and colleagues (2013), summary of action statements and the profile of each statement [46].

The guideline's initial management table considers severe, in addition to the criteria of statement 3A, a toxic-appearing child, otalgia persisting for more than 48 hours, temperature ≥39 °C in the past 48 hours and uncertain access to follow-up; observation is admissible only in AOM diagnosed with high certainty [46]. One subcommittee member did not endorse statement 3C, considering antibiotic preferable in that group [46]. In the meta-analysis underpinning these choices, the difference in rates of pain, fever or both between antibiotic and control was 13% (number needed to treat, NNT, 8), 15% below two years (NNT 7) and 11% from two years (NNT 10); in unilateral AOM it was 6% (NNT 17) and in bilateral 20% (NNT 5), with 25% (NNT 4) in bilateral AOM below two years [46].

Source: Lieberthal and colleagues (2013), Table 5 and text [46]. In the original Table 5, ceftriaxone appears as “50 mg”; the guideline text specifies the dose of 50 mg/kg, and the AAFP review adds a maximum of 1 g per day; for amoxicillin, amoxicillin with clavulanic acid, the other cephalosporins and clindamycin, neither source defines a maximum dose [46, 129].

The guideline justifies the choice of cephalosporins in penicillin allergy by cross-reactivity lower than historically estimated: the rate of about 10% derived from data from the 1960s and 1970s, reactivity is greater with first-generation cephalosporins and negligible with second- and third-generation ones, and cefdinir, cefuroxime, cefpodoxime and ceftriaxone have chemical structures that make cross-reactivity very unlikely [46]. Macrolides have limited efficacy against Haemophilus influenzae and Streptococcus pneumoniae, and clindamycin is not active against H. influenzae, which is why it is used alone only when penicillin-resistant pneumococcus is suspected [46]. With antibiotic, clinical improvement is expected within 48–72 hours: symptoms may worsen slightly in the first 24 hours, begin to improve over the next 24 and fever falls within 48–72 hours; persistence may reflect another disease, concomitant viral infection or a resistant bacterium, and middle ear fluid is sterile in 42–49% of children with persistent symptoms [46]. In failure on amoxicillin, neither co-trimoxazole nor the erythromycin–sulfisoxazole combination is adequate, because of pneumococcal resistance; after successive failures, bacteriological diagnosis should be attempted by tympanocentesis with Gram stain, culture and susceptibility testing, and levofloxacin or linezolid, without US Food and Drug Administration (FDA) approval for AOM, are considered only after otorhinolaryngology and infectious diseases consultation, especially in the presence of multidrug-resistant serotype 19A [46]. Without tympanocentesis, nasopharyngeal culture provides guidance: the positive predictive value is 22–44% for S. pneumoniae, 50–71% for non-typeable H. influenzae and 17–19% for Moraxella catarrhalis, and the negative predictive value 95–99% for all three bacteria, so a culture negative for pneumococcus removes the need for non-conventional drugs [46].

The duration recommended by the guideline is 10 days below two years and at any age with severe symptoms, 7 days between 2 and 5 years with mild to moderate AOM, and 5 to 7 days from 6 years with mild to moderate symptoms [46]. The AAFP reviews summarise this scheme as 5 to 7 days from two years without severe symptoms, which shortens the duration between 2 and 5 years compared with the original text [129, 130]. The guideline does not require routine reassessment at 10–14 days, but allows it in young children with severe symptoms and in those with rAOM; effusion persists in 60–70% of children two weeks after successful treatment, in 40% at one month and in 10–25% at three months, and, without symptoms, corresponds to OME, which is not treated with antibiotics; ensuring resolution of OME is particularly important in the child with cognitive or developmental delay [46]. The high weight-based dose coincides with that of the Italian consensus, which states that international guidelines promote high-dose amoxicillin, and differs from the fixed age-based doses of NICE [39, 40]; the 14:1 formulation used in the USA makes it possible to reach 90 mg/kg/day of amoxicillin without excess clavulanate, a problem that the Italian consensus solves with supplementary plain amoxicillin because it has the 7:1 formulation [40, 129]. In the trial by Hoberman and colleagues cited by the Italian consensus (520 children under two years), five days of amoxicillin with clavulanic acid were associated with more clinical failure than ten days (34% vs 16%), the basis for the longer duration in younger children [40].

7.3. Prevention of recurrent acute otitis media

Antibiotic prophylaxis

The Cochrane review by Leach and Morris (16 trials, 1483 children at increased risk of OM) showed that continuous antibiotic prophylaxis reduces the occurrence of any AOM (RR 0.62; 95% CI 0.52–0.75) and the number of episodes (incidence rate ratio 0.48; 95% CI 0.37–0.62), with about five children treated to prevent one child having AOM and a mean reduction from three to 1.5 episodes per year; adverse effects did not differ significantly, but with a wide interval (RR 1.99; 95% CI 0.25–15.89) [131]. The Italian consensus, after reviewing the evidence, does not recommend routine prophylaxis in rAOM (moderate evidence, weak recommendation against), because the number of episodes prevented is small, because of adverse effects and because of selection of resistance, and allows it as an alternative in selected cases with frequent recurrences when invasive interventions such as tubes or adenotonsillectomy are being considered (low evidence, weak recommendation); the Koivunen trial (662 children) reported a reduction in recurrences of 18–73% and in effusion duration of 33–41% with prophylactic amoxicillin [40]. NICE NG91 makes no recommendation on the prevention of rAOM [40]. The 2013 AAP/AAFP guideline recommends not prescribing prophylactic antibiotics to reduce the frequency of episodes in rAOM (evidence B, recommendation) and exempts from the assessment only young children whose only alternative would be tubes [46]. The guideline cites the 2006 Cochrane review (16 studies): prophylaxis prevented 1.5 episodes per year and halved the number of episodes while it lasted; in placebo-controlled trials, the reduction was 0.09 episodes per month (about 0.5 to 1.5 per year in 95% of children), with about five children treated for one year to prevent one episode and a possibly greater effect in those who had six or more episodes in the previous year [46, 131]. The benefit ceases when prophylaxis is stopped, a trial by Teele found no difference from placebo in recurrences or in persistence of OME, and prophylaxis is not appropriate for prolonged effusion or for infrequent episodes; the harms weighed are allergic reactions, gastrointestinal consequences and selection of resistance [46]. The AAFP reviews justify the recommendation against prophylaxis by the lack of efficacy and the increase in resistance [129, 130]. Among preventive measures, the guideline recommends pneumococcal conjugate vaccine according to the ACIP, AAP and AAFP schedule (evidence B, strong recommendation), the annual influenza vaccine, recommended in the USA for all children from six months (B, recommendation), exclusive breastfeeding for at least six months (B, recommendation) and avoidance of tobacco smoke (C, recommendation) [46]. In the multivariate analysis cited by the guideline, adjusted for day-care attendance, the risk of non-recurrent OM with exclusive breastfeeding up to six months, compared with no breastfeeding or breastfeeding for less than four months, was 0.61 (95% CI 0.40–0.92); in a systematic review, any formula use in the first six months was associated with more OM (OR 1.78; 95% CI 1.19–2.70), and the pooled OR for formula in the first three months was 2.00 (95% CI 1.40–2.78); these studies are observational, with a predominance of white children, parent-reported feeding and incomplete control for day care and tobacco [46]. The 2019 AAFP review cites a 43% risk reduction with exclusive breastfeeding [130]. The guideline acknowledges that avoiding supine bottle-feeding, reducing pacifier use in the second half of the first year and altering day-care attendance patterns may decrease AOM and rAOM, without making a recommendation because of limited evidence [46]. For xylitol, also without a recommendation, the 2011 Cochrane review cited by the guideline (four studies in healthy children in day care) showed a 25% reduction in the risk of AOM (RR 0.75; 95% CI 0.65–0.88; risk difference −0.07), greater with chewing gum and lozenges than with syrup; chewable forms are not safe below two years, xylitol has to be taken 3 to 5 times a day throughout the respiratory infection season, and it does not treat AOM or act with sporadic use [46]. The Italian consensus does not state the prophylactic dose, and the sources consulted do not provide it.

Vaccination

The 2020 Cochrane review (11 trials, 60,733 children) showed that the 7-valent pneumococcal conjugate vaccine with the CRM197 carrier protein, given in early infancy, reduces all-cause AOM by 6% in low-risk infants (95% CI 4–9% in the trial of 37,868 children) and pneumococcal AOM by 20%, but did not reduce all-cause AOM in high-risk infants (relative reduction −5%) or in older children with a history of OM; the PHiD-CV11 vaccine (conjugated to H. influenzae protein D) reduced all-cause AOM by 34% (95% CI 21–44%) [132]. In England, the incidence of AOM in children fell by 40.3% between the period before the 7-valent vaccine (2003–2005) and the late 13-valent period (2015–2019), with a rate ratio of 0.75 (95% CI 0.73–0.78) [82]. The 2013 AAP guideline, which predated the 13-valent vaccine data, cited a 29% reduction in AOM due to any pneumococcal serotype with the 7-valent vaccine before 24 months, a benefit of 6–7% in all-cause AOM in trials and reductions of up to 40% in otitis visits in observational studies, with no effect in children vaccinated for the first time at older ages, and warned of serotype replacement; influenza vaccine showed 30–55% efficacy in preventing AOM during the respiratory season, and up to two thirds of young children with influenza may have AOM [46]. The population effect is modest for all-cause AOM and uncertain in the otitis-prone child, but vaccination according to the national programme remains recommended; specific rules apply to children with immune deficiency (section 7.5) [32, 132].

Tympanostomy tubes

The 2022 AAO-HNSF tube guideline recommends against placing tubes in children with rAOM without effusion at the time of assessment (key action statement 6) and recommends offering bilateral tubes to children with rAOM and unilateral or bilateral effusion at that time (statement 7) [41]. The 2018 Cochrane review (five trials, 805 children, all predating pneumococcal vaccination, with uncertain or high risk of bias) found more children free of recurrence at six months with tubes than with active monitoring (46% vs 5%; RR 9.49; NNTB 3; low-certainty evidence) and at twelve months (48% vs 34%; RR 1.41; 95% CI 1.00–1.99; NNTB 8), and more children free of recurrence than with antibiotic prophylaxis (60% vs 35%; RR 1.68; NNTB 5; very low-certainty evidence), with the risk of persistent tympanic membrane perforation after tubes [133]. The 2013 AAP guideline had classified tubes in rAOM as an option (evidence B), without requiring effusion, balancing the lower frequency of AOM and the possibility of treating episodes with topical antibiotics, on the one hand, against anaesthetic and surgical risk, tympanic scarring, chronic perforation, cholesteatoma and otorrhoea, on the other [46]. The requirement for effusion at the time of assessment comes from the 2022 AAO-HNSF guideline [41]. The AAP cited three randomised trials: two showed fewer episodes in the six months after tubes, and a third, with placebo, amoxicillin prophylaxis and tube groups followed for two years, found no difference between tubes and placebo; a Cochrane review of two studies estimated 1.5 fewer episodes in the following six months [46]. In a series of 10,000 placements, with no major complications, minor complications (membrane tears, tubes displaced into the middle ear) occurred in 0.016% of ears; in a meta-analysis, tympanosclerosis occurred in 32% of patients and chronic perforation in 2.2% with short-term tubes and in 16.6% with long-term tubes [46]. Adenoidectomy alone, without myringotomy or tubes, did not reduce AOM episodes compared with prophylaxis or placebo and should not be used to prevent AOM [46]. Adjuvant adenoidectomy is an option when there are adenoid symptoms (infection or nasal obstruction) or from four years to reduce recurrences and repeat tube placement [41].

7.4. Otitis media with effusion

In the child not at risk, OME is managed with watchful waiting for three months from onset or diagnosis (strong recommendation); the 2016 guideline recommends against intranasal or systemic corticosteroids, systemic antibiotics and antihistamines or decongestants (strong recommendations) [38]. Tubes are not placed for a single episode of OME lasting less than three months; bilateral tubes are offered for bilateral OME of three months or longer with documented hearing difficulties, and are an option in chronic OME with attributable symptoms (balance, school performance, behaviour, discomfort, quality of life) [41]. In children under four years operated on for OME, tubes without adenoidectomy are recommended unless there is a distinct indication; from four years, tubes, adenoidectomy or both [38]. Non-operated chronic OME is reassessed every three to six months until resolution, detection of significant hearing loss or suspicion of structural abnormality [38, 41]. In the at-risk child, tubes are an option for OME that is likely to persist (type B tympanogram or effusion of three months or longer) [41].

7.5. Populations with an underlying condition

The 2022 AAO-HNSF tube guideline details two of these populations. In trisomy 21, Eustachian tube dysfunction and stenotic ear canals cause persistent OME throughout childhood; in a case series, most children needed two or more sets of tubes, with normal hearing in both ears after surgery in 71%, and long-term complications, such as perforation, increased with the number of tubes; one study obtained good hearing outcomes with surveillance every three months if the canals were stenotic and every six months if they were not, with immediate replacement of non-functioning tubes on recurrence of OME; the guideline mentions hearing assessments every six months from birth to three or four years and annually thereafter, and otomicroscopic examination every three to six months when the canals are stenotic [41]. Hearing aids have been proposed as an alternative to tubes in trisomy 21, without comparative trials [41]. In cleft palate, almost all children have OME in the first year, with conductive hearing loss; the systematic review cited by the guideline does not support routine tube placement at the time of cleft surgery, and the consensus is to place them when clinically indicated (hearing loss and flat tympanogram), maintaining surveillance of OME and hearing throughout childhood, including after palatoplasty [41].

Long-term tubes (for example, T-tubes), which ventilate the middle ear for two to three years or more, are reserved as the first surgery for children with an anticipated need for prolonged ventilation: cleft palate, trisomy 21, stenotic ear canals that hinder or prevent short-term tubes, atrophic or atelectatic tympanic membrane and early extrusion of at least two previous sets of short-term tubes (statement 10) [41]. The price is a residual perforation that may approach 20%, compared with about 2% with short-term tubes, which ventilate for 8 to 18 months on average, as well as more otorrhoea, granulation tissue and myringosclerosis [41]. About 25% of children need a second set of short-term tubes and 8% a third, and craniofacial disease is one of the risk factors for repeat placement identified by systematic review [41].

The doses of co-trimoxazole, immunoglobulin and vaccines in these conditions (velocardiofacial syndrome, primary ciliary dyskinesia, agammaglobulinaemias) follow the recommendations for each of them, and the sources read for this article do not give them for OM.

7.6. Proposed therapeutic implications from genotype

Therapeutic hypotheses derived from genetics remain without clinical validation. Santos-Cortez and colleagues suggested that bacitracin, a component of ear drops used in Europe for AOM with otorrhoea, might not be the preferred antibiotic in carriers of A2ML1 variants, because it inhibits clearance of A2M–protease complexes, which may compensate for A2ML1 dysfunction; this is a mechanistic hypothesis without clinical data [1]. Overexpression of RND3 in A2ML1 carriers led to the proposal of ROCK pathway inhibition [62]. In trisomy 21, the identification of DYRK1A as the determinant of OM in mouse models points to inhibitors of this kinase (leucettinib-21, aristolactam BIII) as a target, with the accessibility of the middle ear favouring local administration; the level of evidence is preclinical [14]. The association between FUT2, SPINK5, CDHR3 and SLPI variants and microbiota changes suggests microbiome interventions, which have not been tested in carriers [8, 54–56].

7.7. Guidelines and regulatory framework compared

The sources consulted identify no medicine authorised by the Food and Drug Administration (FDA) or the European Medicines Agency (EMA) for genetic susceptibility to OM, and registered trials do not include genotype-guided therapy [127]. The table compares the recommendations of US and European guidelines for the interventions that apply to the predisposed child.

7.8. Reference trials and meta-analyses

These results are not directly comparable with one another, because populations, periods (the tube trials predate pneumococcal vaccination), definitions of recurrence and duration of follow-up differ [132, 133]. Relative reductions should be read alongside the absolute rate: in prophylaxis, the 52% reduction in the number of episodes corresponds to about 1.5 episodes prevented per child per year [131]; for the 7-valent vaccine, the 6% reduction in all-cause AOM is small in absolute terms [132]. The absence of reduction in high-risk infants, who are the closest to the otitis-prone child, is the relevant negative result for this article [132]. None of the trials stratified by genotype.

7.9. Developments in the last 6–12 months

The ClinicalTrials.gov search covered studies updated since April 2025 with the terms indicated; a negative search does not prove the absence of studies in other registries.

7.10. Agents and procedures to avoid and special situations

In OME, intranasal or systemic corticosteroids, systemic antibiotics, antihistamines and decongestants are not indicated [38]. Postoperative antibiotic ear drops are not routinely prescribed after tubes, and prophylactic water precautions (earplugs, avoidance of swimming) are not recommended [41]. Long-term tubes should not be the first surgery without a specific reason for prolonged ventilation [41]. In AOM unresponsive to amoxicillin, co-trimoxazole and erythromycin–sulfisoxazole are not adequate, adenoidectomy alone does not prevent AOM, and levofloxacin and linezolid are reserved for documented failures after specialist consultation [46]. In 22q11.2 deletion, tonsillectomy requires assessment of carotid artery position, and live vaccines require immunological criteria; BCG (bacille Calmette-Guérin) vaccine is postponed when the syndrome is suspected [32]. In A2ML1 carriers, the possible disadvantage of bacitracin is only hypothetical [1].

Breastfeeding was associated with a lower risk of AOM and prolonged effusion in the Boston cohort [36], whereas in the Filipino community with the A2ML1 variant there was no association with breastfeeding or with tobacco exposure, illustrating how a strong-effect variant can override environmental factors [69]. In the transition to adulthood, CSOM persists in A2ML1 carriers and in the adults of the Pakistani family with BSN variants, with perforations and hearing loss requiring continued otological follow-up [67, 69]. The sources consulted do not address pregnancy in women with predisposition to OM.

8. Reproductive Options and Family Counselling

8.1. Non-syndromic susceptibility

In the non-syndromic form, counselling is based on the multifactorial model. Having a sibling with recurrent AOM increases the risk of AOM and of prolonged effusion [36], and the heritability estimates (0.73 for time with effusion in Pittsburgh twins; 38.5% for rAOM in Finnish families) indicate a substantial genetic component, but do not convert into a recurrence risk for siblings or offspring [33, 35]. The sources consulted do not provide empirical recurrence risks by degree of relationship. The editorial accompanying the Pittsburgh twin study drew a practical conclusion for primary care: once an otitis-prone child has been identified, the siblings deserve closer observation [57].

In families with rare variants of larger effect, the model approaches autosomal dominant with reduced penetrance: in the Filipino pedigree with A2ML1, 37 carriers were affected and nine were not, and there were affected non-carriers [1]. Variants in several genes (A2ML1, FUT2, SPINK5) may coexist in the same family, and the affected members of a family carry different combinations, which precludes attributing each relative's risk to a single gene [8, 54]. Cascade testing, prenatal diagnosis and preimplantation genetic testing have no indication outside research, and the sources consulted do not describe them for this entity. The useful information for families is early hearing surveillance of siblings and offspring, attention to signs of OM in the first months and control of modifiable factors [36, 38].

8.2. Syndromic and immunological forms

In the syndromes, counselling is that of the underlying disease. In cleft lip, the recurrence risk for siblings is 2.7% when the cleft is unilateral and 5.4% when it is bilateral, 4% with one affected child and 10% with two [138]. 22q11.2 deletion is autosomal dominant, de novo in more than 90% of cases with the recurrent 2.54 Mb deletion [25]. In X-linked agammaglobulinaemia, half of newly diagnosed boys have no family history, and of these 15–20% have de novo variants [18]. PCD is autosomal recessive in most forms, and the status of at-risk siblings can be clarified by clinical examination, chest imaging, electron microscopy and nasal nitric oxide when the familial variant is not known [5, 17]. In Stickler syndrome, the forms due to COL2A1, COL11A1 or COL11A2 are autosomal dominant and those due to COL9A1, COL9A2 or COL9A3 are autosomal recessive [26]. Type 1 plasminogen deficiency is autosomal recessive [80].

8.3. Minors

Predictive testing for susceptibility to OM in minors has no established clinical utility. In the syndromes, diagnostic testing in affected children follows the clinical indication of the underlying disease, and carrier testing in asymptomatic minors, when the underlying disease justifies it, is governed by the general rules of genetic counselling.

9. Special Considerations: Population Diversity and Epidemiology

9.1. What the numbers measure

Global estimates are not comparable with one another without considering what they measure. Monasta and colleagues estimated for 2005 an annual AOM incidence of 10.85% (709 million cases, 51% in children under five years) and a CSOM incidence of 4.76 per thousand (31 million), from 114 studies covering 15 of the 21 World Health Organization regions and modelling for the remainder [27]. The Global Burden of Disease 2021 estimated 391 million new OM episodes at all ages (incidence of 4958.9 per 100,000) and a decline of 16.3% since 1990 [28]. The systematic review by Onifade and colleagues, restricted to community-based CSOM studies between 2004 and March 2025 and dominated by low- and middle-income countries and paediatric populations, arrived at a pooled prevalence of 3.8% (297 million people, 85% in low- and middle-income countries) [29]. These are models and aggregations, not prevalences measured in each country.

9.2. United States and Europe

In the USA, the Boston cohort provided the reference cumulative estimates (83% with at least one AOM episode by three years) [36], and the Pittsburgh, Minnesota, Galveston and Philadelphia cohorts, 23andMe and the Vanderbilt biobank underpin the genetic studies [1, 9, 10, 33, 49, 76]. In Europe, the data come from England (1994 twins; primary care incidence between 2003 and 2019, with a 40.3% decline after the conjugate vaccines), Norway (twins), Finland (Helsinki families; FinnGen with 429,599 participants), the Netherlands (Generation R cohort), Belgium (Down syndrome) and Sweden (cleft lip and palate) [10, 11, 30, 31, 34, 35, 48, 82]. Most GWAS included mainly people of European ancestry, which limits the transfer of results to other populations [9, 10].

9.3. Founder populations and high-risk groups

The Indigenous Filipino community studied, of Negrito and Indo-Malay ancestry, with intermarriage favoured by centuries of discrimination, has an OM prevalence close to 50% and a founder A2ML1 duplication shared, on a common haplotype, by North American children of European and Hispanic ancestry [1, 69]. FUT2, SPINK5 and SLPI variants were also identified in the same population [8, 54, 56]. In Pakistan, a six-generation consanguineous family with CSOM showed BSN variants, and another family an A2ML1 frameshift variant [62, 67]. The authors of the Pakistani study cite AOM prevalences of 22.9% in febrile children in tertiary care and OME prevalences of 6.9% to 11.5% (up to 27% in another study) in the country [67]. The FUT2 non-secretor allele most common in Europeans (rs601338) has an allele frequency of 0.44–0.47 and is virtually absent in Koreans and Japanese, in whom non-secretor status depends on other alleles [58]. OMIM entry 166760 reports an unusually high frequency of OM in Native Americans and Aboriginal Australians and a comparatively low frequency in people of African ancestry, and a study in Apache Indians in Arizona (Todd, 1987) suggested familial predisposition [57]. The birth prevalence of cleft lip, a risk condition for OM, ranges from about 1 in 500 in Asians and Native Americans to 1 in 1000 in Europeans and 1 in 2500 in people of African ancestry [138].

9.4. Portugal and Brazil

The sources consulted do not include OM incidence data or genetic susceptibility studies in Portugal or Brazil. Absence from the sources does not mean that such data do not exist.

9.5. Special populations

Children with Down syndrome, cleft palate, 22q11.2 deletion, PCD and humoral immunodeficiencies make up the groups with the highest burden of OM and with their own surveillance recommendations [5, 6, 30–32]. In Indigenous communities, the combination of larger-effect variants with limited access to care increases the risk of permanent hearing loss, and the authors recommend multidisciplinary audiological follow-up with screening tympanometry and audiometry [37].

9.6. European networks and associations

The sources consulted do not identify a European Reference Network (ERN) specific to OM. Children with clefts and craniofacial anomalies are followed within ERN CRANIO, which includes rare ear, nose and throat disorders; PCD within ERN-LUNG; immunodeficiencies within ERN RITA; and malformation and intellectual disability syndromes within ERN ITHACA [42–45]. The associations cited by the sources include the Cleft Lip and Palate Association (CLAPA, United Kingdom) and the American Cleft Palate–Craniofacial Association (ACPA, USA) [3].

10. Illustrative Clinical Case

Yousaf and colleagues described in 2024 family PKOM08, from southern Punjab (Pakistan), of Matilay ethnicity, consanguineous and spanning six generations, in which 12 people (four women and eight men) had CSOM [67]. The age at onset was not known, but all affected participants had had otorrhoea for at least two to three years. Two affected members were examined in detail. Individual IV:2, a 72-year-old man, had bilateral CSOM with perforation of both tympanic membranes, without retraction pockets, cholesteatoma or tympanosclerosis, and moderate to profound hearing loss on the right and moderate on the left. The proband V:1, a 30-year-old woman, had bilateral CSOM with a right-sided perforation, tympanosclerosis on that side and semi-opaque membranes, and moderate hearing loss on the right and moderate to mild on the left. Neither had fever, vertigo, tinnitus, upper respiratory infections, mastoiditis or adenoid hypertrophy at the time of assessment [67].

Exome sequencing of the proband (mean coverage 100-fold), filtered for genes associated with OM in humans or in animal models, with allele frequency up to 0.001 and a CADD (Combined Annotation Dependent Depletion) score of at least 10, revealed four heterozygous variants: two in BSN (NM_003458.4:c.1867C>T p.(Pro623Ser), Genome Aggregation Database (gnomAD) frequency 0.000152, CADD 10.46; and c.11015G>A p.(Arg3672Gln), frequency 0.000116, CADD 12.67), one in FBXO11 (NM_001190274.2:c.160C>A p.(Gln54Lys)) and one in CHD7 (published as NM_001316690.1:c.2341G>A p.(Ala781Thr), = NM_017780.4:c.8488G>A p.(Ala2830Thr) on the MANE transcript) [67]. Sanger sequencing showed that four of the eight affected members tested (IV:2, V:1, V:2 and V:6) were compound heterozygous for the two BSN variants, two (V:5 and V:8) homozygous for c.11015G>A and two (V:9 and VI:2) heterozygous for this variant only; the two-point LOD was 1.25, below the conventional threshold for linkage [67]. BSN encodes bassoon, a scaffold protein of the synaptic active zones of photoreceptors and inner hair cells, and a deep intronic variant in this gene had been associated with childhood ear infections in GWAS; the BSN locus is among the 14 ear infection signals in 23andMe [9, 67]. The pedigree, with the genotypes of each member, and the chromatograms are in Figure 1 of the article: Figure 1.

The case illustrates the current pattern of OM genetics: a chronic disease with hearing consequences in adulthood, a large consanguineous family, rare variants in a plausible gene, genotypic heterogeneity among affected members and modest statistical evidence [67]. The incidental findings in FBXO11 and CHD7, genes of Mendelian neurodevelopmental disorders and of CHARGE syndrome, were interpreted by the authors as variants of small individual effect whose combination might increase the risk burden and contribute to heterogeneity, and not as diagnoses, consistent with the absence of a syndromic phenotype [67, 79, 81]. Identification of the variants did not change the management described, and the article reports no therapeutic interventions; management of CSOM with perforation follows the usual otological standards and does not depend on genotype. For comparison, the Filipino pedigree with the A2ML1 duplication, with 37 affected carriers aged between three months and 58 years and nine carriers without disease, is in Figure 1 of Santos-Cortez and colleagues: Figure 1 [1].

11. Prognosis and Long-term Follow-up

11.1. Prognosis

In the general population, the incidence of AOM decreases after five years (from 7951.8 per 100,000 person-years between two and four years to 2184.4 between five and 17 years, in England) [82]. In Down syndrome, the prevalence of OME has a first peak at 12 months (66.7%), a second at 6–7 years (60%) and declines from eight years [31]. In PCD, ear infections become less evident at school age, but hearing loss may become irreversible if they are not treated [5]. In carriers of larger-effect variants, the disease may persist into adulthood: in the Filipino community, COM persisted into adulthood especially in A2ML1 carriers, and perforations were associated with severe to profound hearing loss [37, 69]. In CSOM, 62% of affected people in community studies had disabling hearing loss [29]. The conductive loss associated with bilateral OME (mean thresholds of 35–37 dB HL in children with cleft or with Down syndrome) is reversible with resolution of the effusion or with tubes, but its duration may affect speech and language [30, 31, 38].

11.2. Surveillance

For achondroplasia, the exact frequency of audiometry in the first three years is not defined in the sources consulted, and for 22q11.2 deletion the immunology guideline does not set the interval of the periodic humoral assessments [32, 108].

12. Summary and Concluding Remarks

Recurrent and chronic OM is a complex trait with high heritability, estimated in twins at between 0.49 and 0.73 depending on age and phenotype and, in Finnish families, at 38.5% for rAOM and 22.1% for COME [33, 35, 48]. Its architecture combines common variants of small effect, identified in GWAS (FNDC1, TBX1, FUT2, ABO, HLA class II, CDHR3, ANXA13, among others), with rare variants of incomplete penetrance in families and founder populations, of which the A2ML1 duplication in the Indigenous Filipino community is the example with the strongest evidence and the only one with a phenotype entry in OMIM [1, 7, 9–11]. The proposed mechanisms converge on the integrity and defence of the middle ear mucosa, Eustachian tube function, mucociliary clearance and microbiome composition [8, 11, 59].

For the clinician, the practical question is to distinguish the child with isolated OM, even if recurrent, from the child whose OM signals an underlying condition: cleft palate, 22q11.2 deletion, trisomy 21, PCD, humoral or combined immunodeficiency, other chromosomal and monogenic syndromes, skeletal dysplasias, storage disorders and, rarely, plasminogen deficiency [5, 6, 30–32, 80]. In these, treatment of the underlying disease, closer audiological surveillance and lower thresholds for tubes are the measures with impact [38, 41].

In the non-syndromic form, genetic testing does not yet have clinical utility: the variants have incomplete penetrance, coexist in different combinations within the same family, and the direction of some associations, such as that of FUT2, differs between studies [8, 9]. Management follows the general guidelines: watchful waiting and amoxicillin in AOM, with a divergence in dose between NICE and the Italian consensus; antibiotic prophylaxis only in selected cases; tubes in rAOM only with effusion present and in persistent OME with hearing loss; vaccination according to the national programme, with a modest and uncertain benefit in high-risk children [39–41, 132]. Recent developments (24 genes in families, SLPI, DYRK1A in trisomy 21) extend the genetic map and open therapeutic hypotheses that remain at the preclinical stage [12, 14, 56].

13. References

1. Santos-Cortez RL, Chiong CM, Reyes-Quintos MR, Tantoco ML, Wang X, Acharya A, et al. Rare A2ML1 variants confer susceptibility to otitis media. Nat Genet. 2015;47(8):917-920. doi:10.1038/ng.3347. PMID: 26121085. PMCID: PMC4528370.
2. Rye MS, Blackwell JM, Jamieson SE. Genetic susceptibility to otitis media in childhood. Laryngoscope. 2012;122(3):665-675. doi:10.1002/lary.22506. PMID: 22252840.
3. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 105.
4. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 624.
5. Zariwala MA, Despotes KA, Davis SD. Primary Ciliary Dyskinesia. 2007 Jan 24 [updated 2025 May 22]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993-2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1122/
6. Smith CIE, Berglöf A. X-Linked Agammaglobulinemia. 2001 Apr 5 [updated 2024 Jun 27]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993-2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1453/
7. Online Mendelian Inheritance in Man (OMIM). *610627 ALPHA-2-MACROGLOBULIN-LIKE 1; A2ML1 [Internet]. Baltimore (MD): Johns Hopkins University. Available from: https://omim.org/entry/610627. Accessed 29 Oct 2026.
8. Santos-Cortez RLP, Chiong CM, Frank DN, Ryan AF, Giese APJ, Bootpetch Roberts T, et al. FUT2 Variants Confer Susceptibility to Familial Otitis Media. Am J Hum Genet. 2018;103(5):679-690. doi:10.1016/j.ajhg.2018.09.010. PMID: 30401457. PMCID: PMC6217759.
9. Tian C, Hromatka BS, Kiefer AK, Eriksson N, Noble SM, Tung JY, et al. Genome-wide association and HLA region fine-mapping studies identify susceptibility loci for multiple common infections. Nat Commun. 2017;8(1):599. doi:10.1038/s41467-017-00257-5. PMID: 28928442. PMCID: PMC5605711.
10. van Ingen G, Li J, Goedegebure A, Pandey R, Li YR, March ME, et al. Genome-wide association study for acute otitis media in children identifies FNDC1 as disease contributing gene. Nat Commun. 2016;7:12792. doi:10.1038/ncomms12792. PMID: 27677580. PMCID: PMC5052699.
11. Bizaki-Vallaskangas A, Rämö J, Sliz E, Kivekäs I, Willberg T, Saarentaus E, et al. Genome-wide association study indicates novel associations of annexin A13 to secretory and GAS2L2 with mucous otitis media. Sci Rep. 2024;14(1):18344. doi:10.1038/s41598-024-68781-1. PMID: 39112560. PMCID: PMC11306868.
12. Santos-Cortez RLP, Elling CL, Gomez HZ, Einarsdottir E, Kere J, Mattila PS, et al. Rare and low-frequency variants in families with otitis media. J Mol Med (Berl). 2025;103(5):559-570. doi:10.1007/s00109-025-02537-w. PMID: 40183840.
13. Orphanet. Primary ciliary dyskinesia (ORPHA:244). Available from: https://www.orpha.net/en/disease/detail/244
14. Delmaghani S, El-Amraoui A. How a gene fuels ear infections. Elife. 2025;14:e105612. doi:10.7554/elife.105612. PMID: 39840732. PMCID: PMC11753777.
15. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 293.
16. GeneGuidance. A2ML1 — gene record (clinical curation). GeneGuidance database; accessed 2026-10-29.
17. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 367.
18. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 465.
19. Online Mendelian Inheritance in Man (OMIM). *182100 FUCOSYLTRANSFERASE 2; FUT2 [Internet]. Baltimore (MD): Johns Hopkins University. Available from: https://omim.org/entry/182100. Accessed 29 Oct 2026.
20. Orphanet. 22q11.2 deletion syndrome (ORPHA:567). Available from: https://www.orpha.net/en/disease/detail/567
21. Orphanet. Down syndrome (ORPHA:870). Available from: https://www.orpha.net/en/disease/detail/870
22. Orphanet. X-linked agammaglobulinemia (ORPHA:47). Available from: https://www.orpha.net/en/disease/detail/47
23. Orphanet. Hypoplasminogenemia (ORPHA:722). Available from: https://www.orpha.net/en/disease/detail/722
24. Orphanet. Cleft palate (ORPHA:2014). Available from: https://www.orpha.net/en/disease/detail/2014
25. McDonald-McGinn DM, Hain HS, Emanuel BS, Zackai EH. 22q11.2 Deletion Syndrome. 1999 Sep 23 [updated 2020 Feb 27; revised 2025 May 8]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993-2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1523/
26. Mortier G. Stickler Syndrome. 2000 Jun 9 [updated 2023 Sep 7]. In: Adam MP, Feldman J, Mirzaa GM, et al., editors. GeneReviews®. Seattle (WA): University of Washington, Seattle; 1993-2026. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1302/
27. Monasta L, Ronfani L, Marchetti F, Montico M, Vecchi Brumatti L, Bavcar A, et al. Burden of disease caused by otitis media: systematic review and global estimates. PLoS One. 2012;7(4):e36226. doi:10.1371/journal.pone.0036226. PMID: 22558393. PMCID: PMC3340347.
28. GBD 2021 Upper Respiratory Infections Otitis Media Collaborators. Global, regional, and national burden of upper respiratory infections and otitis media, 1990-2021: a systematic analysis from the Global Burden of Disease Study 2021. Lancet Infect Dis. 2025;25(1):36-51. doi:10.1016/s1473-3099(24)00430-4. PMID: 39265593. PMCID: PMC11680489.
29. Onifade A, Katolo HW, Mookerjee S, Bhutta MF. Epidemiology of Chronic Suppurative Otitis Media: Systematic Review To Estimate Global Prevalence. J Epidemiol Glob Health. 2025;15(1):55. doi:10.1007/s44197-025-00396-9. PMID: 40178730. PMCID: PMC11968643.
30. Flynn T, Möller C, Jönsson R, Lohmander A. The high prevalence of otitis media with effusion in children with cleft lip and palate as compared to children without clefts. Int J Pediatr Otorhinolaryngol. 2009;73(10):1441-1446. doi:10.1016/j.ijporl.2009.07.015. PMID: 19709760.
31. Maris M, Wojciechowski M, Van de Heyning P, Boudewyns A. A cross-sectional analysis of otitis media with effusion in children with Down syndrome. Eur J Pediatr. 2014;173(10):1319-1325. doi:10.1007/s00431-014-2323-5. PMID: 24816633.
32. Mustillo PJ, Sullivan KE, Chinn IK, Notarangelo LD, Haddad E, Davies EG, et al. Clinical Practice Guidelines for the Immunological Management of Chromosome 22q11.2 Deletion Syndrome and Other Defects in Thymic Development. J Clin Immunol. 2023;43(2):247-270. doi:10.1007/s10875-022-01418-y. PMID: 36648576. PMCID: PMC9892161.
33. Casselbrant ML, Mandel EM, Fall PA, Rockette HE, Kurs-Lasky M, Bluestone CD, et al. The heritability of otitis media: a twin and triplet study. JAMA. 1999;282(22):2125-2130. doi:10.1001/jama.282.22.2125. PMID: 10591333.
34. Kvestad E, Kvaerner KJ, Røysamb E, Tambs K, Harris JR, Magnus P. Otitis media: genetic factors and sex differences. Twin Res. 2004;7(3):239-244. doi:10.1375/136905204774200514. PMID: 15193168.
35. Hafrén L, Kentala E, Järvinen TM, Leinonen E, Onkamo P, Kere J, et al. Genetic background and the risk of otitis media. Int J Pediatr Otorhinolaryngol. 2012;76(1):41-44. doi:10.1016/j.ijporl.2011.09.026. PMID: 22018929.
36. Teele DW, Klein JO, Rosner B. Epidemiology of otitis media during the first seven years of life in children in greater Boston: a prospective, cohort study. J Infect Dis. 1989;160(1):83-94. doi:10.1093/infdis/160.1.83. PMID: 2732519.
37. Santos-Cortez RLP, Ong KMC, Carlos-Hiceta A, Tantoco MLC, Yarza TKL, San Agustin ML, et al. Audiologic Measures in an Indigenous Community with A2ML1- and FUT2-Related Otitis Media. Genet Test Mol Biomarkers. 2023;27(1):12-17. doi:10.1089/gtmb.2022.0171. PMID: 36719978. PMCID: PMC9902046.
38. Rosenfeld RM, Shin JJ, Schwartz SR, Coggins R, Gagnon L, Hackell JM, et al. Clinical Practice Guideline: Otitis Media with Effusion (Update). Otolaryngol Head Neck Surg. 2016;154(1 Suppl):S1-S41. doi:10.1177/0194599815623467. PMID: 26832942.
39. National Institute for Health and Care Excellence. Otitis media (acute): antimicrobial prescribing. NICE guideline NG91. London: NICE; 2018 [updated 2022]. Available from: https://www.nice.org.uk/guidance/ng91/chapter/Recommendations
40. Castelli Gattinara G, Bergamini M, Simeone G, Reggiani L, Doria M, Ghiglioni DG, et al. Antibiotic treatment of acute and recurrent otitis media in children: an Italian intersociety Consensus. Ital J Pediatr. 2025;51(1):50. doi:10.1186/s13052-025-01894-z. PMID: 39980042. PMCID: PMC11844117.
41. Rosenfeld RM, Tunkel DE, Schwartz SR, Anne S, Bishop CE, Chelius DC, et al. Clinical Practice Guideline: Tympanostomy Tubes in Children (Update). Otolaryngol Head Neck Surg. 2022;166(1_suppl):S1-S55. doi:10.1177/01945998211065662. PMID: 35138954.
42. European Reference Network for rare and/or complex craniofacial anomalies and ENT disorders (ERN CRANIO). Available from: https://www.ern-cranio.eu/
43. European Reference Network on rare respiratory diseases (ERN-LUNG). Available from: https://ern-lung.eu/
44. European Reference Network on Rare Immunodeficiency, Autoinflammatory and Autoimmune Diseases (ERN RITA). Available from: https://ern-rita.org/
45. European Reference Network on Rare Congenital Malformations and Rare Intellectual Disability (ERN ITHACA). Available from: https://ern-ithaca.eu/
46. Lieberthal AS, Carroll AE, Chonmaitree T, Ganiats TG, Hoberman A, Jackson MA, et al. The diagnosis and management of acute otitis media. Pediatrics. 2013;131(3):e964-99. doi:10.1542/peds.2012-3488. PMID: 23439909.
47. Rye MS, Bhutta MF, Cheeseman MT, Burgner D, Blackwell JM, Brown SD, et al. Unraveling the genetics of otitis media: from mouse to human and back again. Mamm Genome. 2011;22(1-2):66-82. doi:10.1007/s00335-010-9295-1. PMID: 21107580.
48. Rovers M, Haggard M, Gannon M, Koeppen-Schomerus G, Plomin R. Heritability of symptom domains in otitis media: a longitudinal study of 1,373 twin pairs. Am J Epidemiol. 2002;155(10):958-964. doi:10.1093/aje/155.10.958. PMID: 11994236.
49. Daly KA, Brown WM, Segade F, Bowden DW, Keats BJ, Lindgren BR, et al. Chronic and recurrent otitis media: a genome scan for susceptibility loci. Am J Hum Genet. 2004;75(6):988-997. doi:10.1086/426061. PMID: 15514890. PMCID: PMC1225283.
50. Parkinson N, Hardisty-Hughes RE, Tateossian H, Tsai HT, Brooker D, Morse S, et al. Mutation at the Evi1 locus in Junbo mice causes susceptibility to otitis media. PLoS Genet. 2006;2(10):e149. doi:10.1371/journal.pgen.0020149. PMID: 17029558. PMCID: PMC1592239.
51. Rye MS, Warrington NM, Scaman ES, Vijayasekaran S, Coates HL, Anderson D, et al. Genome-wide association study to identify the genetic determinants of otitis media susceptibility in childhood. PLoS One. 2012;7(10):e48215. doi:10.1371/journal.pone.0048215. PMID: 23133572. PMCID: PMC3485007.
52. Allen EK, Chen WM, Weeks DE, Chen F, Hou X, Mattos JL, et al. A genome-wide association study of chronic otitis media with effusion and recurrent otitis media identifies a novel susceptibility locus on chromosome 2. J Assoc Res Otolaryngol. 2013;14(6):791-800. doi:10.1007/s10162-013-0411-2. PMID: 23974705. PMCID: PMC3825021.
53. Einarsdottir E, Hafrén L, Leinonen E, Bhutta MF, Kentala E, Kere J, et al. Genome-wide association analysis reveals variants on chromosome 19 that contribute to childhood risk of chronic otitis media with effusion. Sci Rep. 2016;6:33240. doi:10.1038/srep33240. PMID: 27632927. PMCID: PMC5025747.
54. Frank DN, Giese APJ, Hafren L, Bootpetch TC, Yarza TKL, Steritz MJ, et al. Otitis media susceptibility and shifts in the head and neck microbiome due to SPINK5 variants. J Med Genet. 2021;58(7):442-452. doi:10.1136/jmedgenet-2020-106844. PMID: 32709676. PMCID: PMC8218788.
55. Hirsch SD, Elling CL, Bootpetch TC, Scholes MA, Hafrén L, Streubel SO, et al. The role of CDHR3 in susceptibility to otitis media. J Mol Med (Berl). 2021;99(11):1571-1583. doi:10.1007/s00109-021-02118-7. PMID: 34322716. PMCID: PMC8541908.
56. Elling CL, Ryan AF, Yarza TKL, Ghaffar A, Llanes EGDV, Kofonow JM, et al. A Novel SLPI Splice Variant Confers Susceptibility to Otitis Media in Humans. Int J Mol Sci. 2025;26(4):1411. doi:10.3390/ijms26041411. PMID: 40003878. PMCID: PMC11855725.
57. Online Mendelian Inheritance in Man (OMIM). #166760 OTITIS MEDIA, SUSCEPTIBILITY TO; OMS [Internet]. Baltimore (MD): Johns Hopkins University. Updated 3 Jul 2018. Available from: https://omim.org/entry/166760. Accessed 29 Oct 2026.
58. GeneGuidance. FUT2 — gene record (clinical curation). GeneGuidance database; accessed 2026-10-29.
59. Geng R, Wang Q, Chen E, Zheng QY. Current Understanding of Host Genetics of Otitis Media. Front Genet. 2019;10:1395. doi:10.3389/fgene.2019.01395. PMID: 32117425. PMCID: PMC7025460.
60. Yu Y, Yuan R, Zhang J, Wang L, Guo T, Shen H, et al. Systems perspectives on pediatric otitis media: environmental exposures, genetic susceptibility, and biomarker-guided interventions. Front Pediatr. 2026;14:1758098. doi:10.3389/fped.2026.1758098. PMID: 42199501. PMCID: PMC13199312.
61. Aulakh A, Khaira A, Husein M, Strychowsky JE, Graham ME. Risk Factors for Otitis Media in Pediatric Patients With Cleft Lip and/or Cleft Palate: Systematic Review and Meta-analysis. Cleft Palate Craniofac J. 2026;63(7):1910-1919. doi:10.1177/10556656251352708. PMID: 40589130.
62. Larson ED, Magno JPM, Steritz MJ, Llanes EGDV, Cardwell J, Pedro M, et al. A2ML1 and otitis media: novel variants, differential expression, and relevant pathways. Hum Mutat. 2019;40(8):1156-1171. doi:10.1002/humu.23769. PMID: 31009165. PMCID: PMC6711784.
63. Elling CL, Scholes MA, Streubel SO, Larson ED, Wine TM, Bootpetch TC, et al. The FUT2 Variant c.461G>A (p.Trp154*) Is Associated With Differentially Expressed Genes and Nasopharyngeal Microbiota Shifts in Patients With Otitis Media. Front Cell Infect Microbiol. 2021;11:798246. doi:10.3389/fcimb.2021.798246. PMID: 35096646. PMCID: PMC8798324.
64. GeneGuidance. SPINK5 — gene record (identifiers). GeneGuidance database; accessed 2026-10-29.
65. GeneGuidance. CDHR3 — gene record (identifiers). GeneGuidance database; accessed 2026-10-29.
66. GeneGuidance. EYA4 — gene record (identifiers and linked conditions). GeneGuidance database; accessed 2026-10-29.
67. Yousaf A, Yousaf S, Shabbir AS, Yousaf R, Riazuddin S, Shaikh RS, et al. Identification of rare missense variants in the BSN gene co-segregating with chronic otitis media in a consanguineous Pakistani family. Mol Genet Genomic Med. 2024;12(9):e2478. doi:10.1002/mgg3.2478. PMID: 39302268. PMCID: PMC11414307.
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):61-68. doi:10.1002/humu.23348. PMID: 28967166. Available from: https://variantvalidator.org.
69. Santos-Cortez RL, Reyes-Quintos MR, Tantoco ML, Abbe I, Llanes EG, Ajami NJ, et al. Genetic and Environmental Determinants of Otitis Media in an Indigenous Filipino Population. Otolaryngol Head Neck Surg. 2016;155(5):856-862. doi:10.1177/0194599816661703. PMID: 27484237. PMCID: PMC5093071.
70. Elling CL, Gomez HZ, Lee NK, Hirsch SD, Santos-Cortez RLP. The A2ml1-Knockout mouse as an animal model for non-syndromic otitis media. Int J Pediatr Otorhinolaryngol. 2024;181:111980. doi:10.1016/j.ijporl.2024.111980. PMID: 38759260. PMCID: PMC11175999.
71. Bootpetch TC, Hafrén L, Elling CL, Baschal EE, Manichaikul AW, Pine HS, et al. Multi-omic studies on missense PLG variants in families with otitis media. Sci Rep. 2020;10(1):15035. doi:10.1038/s41598-020-70498-w. PMID: 32929111. PMCID: PMC7490366.
72. Wiesen BM, Hafrén L, Einarsdottir E, Kere J, Mattila PS, Santos-Cortez RLP. ABO Genotype and Blood Type Are Associated with Otitis Media. Genet Test Mol Biomarkers. 2019;23(11):823-827. doi:10.1089/gtmb.2019.0135. PMID: 31693456. PMCID: PMC6857544.
73. Emonts M, Veenhoven RH, Wiertsema SP, Houwing-Duistermaat JJ, Walraven V, de Groot R, et al. Genetic polymorphisms in immunoresponse genes TNFA, IL6, IL10, and TLR4 are associated with recurrent acute otitis media. Pediatrics. 2007;120(4):814-823. doi:10.1542/peds.2007-0524. PMID: 17908769.
74. Hafrén L, Einarsdottir E, Kentala E, Hammarén-Malmi S, Bhutta MF, MacArthur CJ, et al. Predisposition to Childhood Otitis Media and Genetic Polymorphisms within the Toll-Like Receptor 4 (TLR4) Locus. PLoS One. 2015;10(7):e0132551. doi:10.1371/journal.pone.0132551. PMID: 26177520. PMCID: PMC4503307.
75. Mittal R, Robalino G, Gerring R, Chan B, Yan D, Grati M, et al. Immunity genes and susceptibility to otitis media: a comprehensive review. J Genet Genomics. 2014;41(11):567-581. doi:10.1016/j.jgg.2014.10.003. PMID: 25434680.
76. Jiang L, Kerchberger VE, Shaffer C, Dickson AL, Ormseth MJ, Daniel LL, et al. Genome-wide association analyses of common infections in a large practice-based biobank. BMC Genomics. 2022;23(1):672. doi:10.1186/s12864-022-08888-9. PMID: 36167494. PMCID: PMC9512962.
77. Bartlett JA, Meyerholz DK, Wohlford-Lenane CL, Naumann PW, Salzman NH, McCray PB. Increased susceptibility to otitis media in a Splunc1-deficient mouse model. Dis Model Mech. 2015;8(5):501-508. doi:10.1242/dmm.019646. PMID: 25765466. PMCID: PMC4415896.
78. Online Mendelian Inheritance in Man (OMIM). #188400 DIGEORGE SYNDROME; DGS [Internet]. Baltimore (MD): Johns Hopkins University. Available from: https://omim.org/entry/188400. Accessed 29 Oct 2026.
79. GeneGuidance. FBXO11 — gene record (clinical curation). GeneGuidance database; accessed 2026-10-29.
80. GeneGuidance. PLG — gene record (clinical curation). GeneGuidance database; accessed 2026-10-29.
81. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 143.
82. Mohanty S, Podmore B, Cuñado Moral A, Weiss T, Matthews I, Sarpong E, et al. Incidence of acute otitis media from 2003 to 2019 in children ≤ 17 years in England. BMC Public Health. 2023;23(1):201. doi:10.1186/s12889-023-14982-8. PMID: 36717794. PMCID: PMC9885604.
83. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 529.
84. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 506.
85. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 85.
86. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 578.
87. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 579.
88. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 322.
89. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 1.
90. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 697.
91. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 28.
92. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 36.
93. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 50.
94. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 144.
95. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 104.
96. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 426.
97. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 198.
98. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 202.
99. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 106.
100. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 107.
101. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 230.
102. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 692.
103. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 866.
104. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 522.
105. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 534.
106. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 335.
107. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 246.
108. Spranger JW, Brill PW, Hall C, Nishimura G, Superti-Furga A, Unger S. Bone Dysplasias: An Atlas of Genetic Disorders of Skeletal Development. 4th ed. Oxford University Press; 2018. p. 6.
109. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 534.
110. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 602.
111. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 107.
112. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 564.
113. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 574.
114. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 603.
115. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 638.
116. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 652.
117. Nyhan WL, Hoffmann GF, Al-Aqeel AI, Barshop BA. Atlas of Inherited Metabolic Diseases. 4th ed. Boca Raton: CRC Press; 2020. p. 728.
118. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 760.
119. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 761.
120. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 764.
121. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 782.
122. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 734.
123. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 334.
124. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 147.
125. Online Mendelian Inheritance in Man (OMIM). #244400 CILIARY DYSKINESIA, PRIMARY, 1; CILD1 [Internet]. Baltimore (MD): Johns Hopkins University. Available from: https://omim.org/entry/244400. Accessed 29 Oct 2026.
126. Online Mendelian Inheritance in Man (OMIM). #217090 PLASMINOGEN DEFICIENCY, TYPE I [Internet]. Baltimore (MD): Johns Hopkins University. Available from: https://omim.org/entry/217090. Accessed 29 Oct 2026.
127. ClinicalTrials.gov. Search “otitis media” with the terms recurrent/otitis-prone/genetic, studies updated since 2025-04-01 [Internet]. Bethesda (MD): National Library of Medicine; accessed 2026-10-29. Available from: https://clinicaltrials.gov/api/v2/studies
128. Venekamp RP, Sanders SL, Glasziou PP, Rovers MM. Antibiotics for acute otitis media in children. Cochrane Database Syst Rev. 2023;11:CD000219. doi:10.1002/14651858.cd000219.pub5. PMID: 37965923. PMCID: PMC10646935.
129. Harmes KM, Blackwood RA, Burrows HL, Cooke JM, Harrison RV, Passamani PP. Otitis media: diagnosis and treatment. Am Fam Physician. 2013;88(7):435-40. PMID: 24134083. Available from: https://www.aafp.org/pubs/afp/issues/2013/1001/p435.html
130. Gaddey HL, Wright MT, Nelson TN. Otitis media: rapid evidence review. Am Fam Physician. 2019;100(6):350-6. PMID: 31524361. Available from: https://www.aafp.org/pubs/afp/issues/2019/0915/p350.html
131. Leach AJ, Morris PS. Antibiotics for the prevention of acute and chronic suppurative otitis media in children. Cochrane Database Syst Rev. 2006;(4):CD004401. doi:10.1002/14651858.cd004401.pub2. PMID: 17054203. PMCID: PMC11324013.
132. de Sévaux JL, Venekamp RP, Lutje V, Hak E, Schilder AG, Sanders EA, et al. Pneumococcal conjugate vaccines for preventing acute otitis media in children. Cochrane Database Syst Rev. 2020;11:CD001480. doi:10.1002/14651858.cd001480.pub6. PMID: 33231293. PMCID: PMC8096893.
133. Venekamp RP, Mick P, Schilder AG, Nunez DA. Grommets (ventilation tubes) for recurrent acute otitis media in children. Cochrane Database Syst Rev. 2018;5:CD012017. doi:10.1002/14651858.cd012017.pub2. PMID: 29741289. PMCID: PMC6494623.
134. Centers for Disease Control and Prevention. Summary of risk-based pneumococcal vaccination recommendations [Internet]. Atlanta (GA): CDC; 2026 May 8 [accessed 2026 Nov 7]. Available from: https://www.cdc.gov/pneumococcal/hcp/vaccine-recommendations/risk-indications.html
135. Haute Autorité de Santé. Révision de la stratégie de vaccination contre les infections invasives à pneumocoques: place des vaccins à haute valence Prevenar 20 chez les nourrissons, les enfants et les adultes et Capvaxive chez les adultes. Vaccination recommendation validated on 23 July 2026 (decision no. 2026.0154/DC/SESPEV). [https://has-sante.fr/jcms/p_3601959/fr/](https://has-sante.fr/jcms/p_3601959/fr/). Accessed 7 Nov 2026.
136. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 19.
137. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 667.
138. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 336.
139. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 625.