Nonsyndromic Genetic Hearing Loss, Pendred Syndrome and Mitochondrial Ototoxicity

Literature and regulatory information updated: 6 October 2026.

DFNB1 (GJB2/GJB6), STRC, OTOF, SLC26A4 (Pendred syndrome and DFNB4), MYO7A and other DFNA/DFNB/DFNX loci, mitochondrial hearing loss (MT-RNR1, MT-TS1) and aminoglycoside susceptibility; newborn hearing screening, aetiological algorithm, cochlear implantation and gene therapy for OTOF-related hearing loss

0. Initial Reference Table

1. Introduction

Hearing loss is not itself a diagnosis: it is the clinical expression of a structural or functional change in the auditory pathway whose cause must be identified [1]. Severe or profound hearing loss affects approximately one in a thousand children at birth or during the prelingual period, and another two to three in a thousand have moderate or progressive hearing loss requiring amplification [2]. In countries with developed healthcare systems, genetic causes predominate: the Oxford Desk Reference estimates that at least 60% of severe childhood hearing loss is hereditary, and GeneReviews attributes approximately 65% of prelingual hearing loss to genetic causes (80% in some series), with the remaining 35% due to environmental causes, most commonly congenital cytomegalovirus infection (cCMV) [1, 2]. Within genetic prelingual hearing loss, approximately 70% is nonsyndromic, meaning no visible external ear malformations or involvement of other organs, although middle or inner ear anomalies may coexist [1].

Genetic heterogeneity defines the field. More than 150 nonsyndromic hearing loss genes have been described, including more than 80 autosomal recessive genes, more than 60 autosomal dominant genes, six X-linked genes and two mitochondrial genes with a principal role [1]. Despite this diversity, a small subset explains many diagnoses: GJB2 is the most common genetic cause of severe to profound congenital hearing loss in European and Asian populations, STRC is the second cause of hereditary hearing loss and the leading cause of mild to moderate hearing loss, and SLC26A4 is the third cause when Pendred syndrome and nonsyndromic hearing loss with an enlarged vestibular aqueduct are considered one spectrum [1, 4, 5].

The history of hearing loss genetics reflects the development of medical genetics itself. In 1896, Vaughan Pendred described in the Lancet two siblings from an Irish family with deaf-mutism and goitre, giving his name to the syndrome linking hearing loss with impaired iodine organification [28]. The first pathogenic variant associated with nonsyndromic hearing loss was identified in the mitochondrial genome: in 1993, Prezant and colleagues sequenced mitochondrial DNA from families with maternally inherited aminoglycoside-induced hearing loss and a large Arab-Israeli family with hearing loss without antibiotic exposure, identifying the m.1555A>G substitution in the 12S ribosomal RNA gene (MT-RNR1) [29]. In 1995, Weil and colleagues demonstrated that the myosin VIIA gene (MYO7A) causes Usher syndrome type 1B, opening the way to understanding sensory cell hair bundles [30]. In 1997, Kelsell and colleagues identified GJB2 variants, encoding connexin 26, in families with nonsyndromic sensorineural hearing loss [31], and Everett and colleagues cloned the Pendred syndrome gene (PDS, now SLC26A4), encoding the anion transporter pendrin [32]. These discoveries were followed by identification of OTOF (otoferlin) as the cause of DFNB9 by Yasunaga and colleagues in 1999 [33], STRC (stereocilin) at the DFNB16 locus by Verpy and colleagues in 2001 [34], and the large deletion involving GJB6 (connexin 30) at the DFNB1 locus described by del Castillo and colleagues in 2002 [35].

Three recent changes have altered clinical practice. The first was widespread universal newborn hearing screening and diagnostic and early-intervention timing targets [15]. The second was incorporation of next-generation sequencing into aetiological evaluation, now considered standard care in investigating anyone with hearing loss [1, 14]. The third was the arrival of gene therapy: after initial Chinese trials in 2024, the FDA approved the first gene therapy for a genetic hearing loss form, OTOF-related hearing loss, in April 2026 [11, 36, 37]. In parallel, MT-RNR1 pharmacogenetics became the subject of formal guidelines recommending aminoglycoside avoidance in carriers of risk variants [10, 16].

Counselling should distinguish “deaf” (total or near-total hearing loss), “hard of hearing” (partial hearing loss) and the Deaf community, which shares its own culture and sign language and often does not regard deafness as a disease requiring treatment [1]. This perspective shapes discussion of communication goals, habilitation options and reproductive decisions with families.

2. Molecular Genetics and Pathophysiology

2.1. Physiological foundations and nomenclature

Hearing loss is classified as conductive (external ear or middle ear ossicular abnormalities), sensorineural (cochlear or auditory nerve dysfunction, potentially with vestibular dysfunction), mixed or central (cochlear nerve, brainstem or cortical lesion) [1]. In sensorineural hearing loss, pathology lies between the cochlear hair cells and auditory brain areas; most cases result from hair cell dysfunction and a minority from auditory nerve pathology or the synapse between inner hair cells and the nerve (auditory neuropathy) [2]. Outer hair cells act as the cochlear amplifier and generate otoacoustic emissions (OAEs); inner hair cells perform mechanoelectrical transduction and transmit signals to the auditory nerve through ribbon synapses, whose function is assessed by auditory brainstem responses (ABRs) [23].

Nonsyndromic hearing loss loci use the prefix DFN (DeaFNess), followed by A (autosomal dominant), B (autosomal recessive) or X (X-linked) and a number reflecting mapping order; gene-based naming (for example, “OTOF-related hearing loss”) is preferred because a locus may contain more than one gene, as with GJB2 and GJB6 at DFNB1 [1, 3].

2.2. Distribution by inheritance pattern

In genetic prelingual nonsyndromic hearing loss, approximately 80% of cases are autosomal recessive, 19% autosomal dominant and fewer than 1% X-linked or mitochondrial [1]. The Oxford Desk Reference describes a similar proportion (80–85% AR), with heterogeneity across more than 60 AD loci and more than 90 AR loci [2]. Generally, AR hearing loss is prelingual and severe to profound, whereas AD hearing loss is postlingual, progressive and predominantly high-frequency; exceptions are clinically informative and summarised in the following tables [1]. AD inheritance is most frequent in postlingual hearing loss [1]. An Icelandic study and a US series of 2022 showed that approximately 2% of nonsyndromic AD hearing loss results from de novo variants [1].

2.3. DFNB1 locus: GJB2 (connexin 26) and GJB6 (connexin 30)

GJB2 (13q12.11) has a single coding exon and encodes connexin 26 [2, 3]. Six connexins assemble around a central 2.3 nm pore to form a connexon (hemichannel); connexons in adjacent cells dock to form gap junctions, permitting direct exchange of ions and small molecules between cells [3]. Connexin 26 forms functional junctions with itself and with connexins 32, 46 and 50 [3]. In the cochlea, loss of coupling between supporting cells compromises potassium recycling required for sensory transduction; recessive hearing loss results from loss of function [3]. The inner ear is structurally normal and there is no vestibular pathology [2].

GJB6, adjacent to GJB2 at DFNB1, encodes connexin 30, coexpressed with connexin 26 in the inner ear [2]. The 342 kb deletion involving GJB6—del(GJB6-D13S1830)—is the second most common DFNB1 alteration in Mediterranean populations; the GJB2 coding region remains intact, and hearing loss results from removal of a distant GJB2 regulatory element in or near GJB6 [2, 35]. Sensorineural hearing loss occurs in individuals homozygous or compound heterozygous for GJB2 variants, homozygous for the GJB6 deletion, and double heterozygous for a GJB2 variant and a GJB6 deletion [2]. GeneReviews currently describes approximately 1% of individuals with DFNB1 hearing loss as compound heterozygotes for a GJB2 variant and an upstream deletion (covering GJB6 and parts of CRYL1, or only parts of CRYL1), removing GJB2 regulatory regions on the same allele and abolishing expression; rarely, the deletion includes GJB2 itself [3]. Upstream or downstream noncoding variants and one case of mosaic maternal uniparental disomy for c.235delC have also been described [3].

Some GJB2 variants have a dominant effect. Fewer than 30 variants cause AD hearing loss (DFNA3A), often associated with skin manifestations: p.Arg75Gln and p.Arg75Trp cause fully penetrant hearing loss and incompletely penetrant palmoplantar keratoderma [3]. Gain-of-function variants cause skin disorders with hearing loss—keratitis–ichthyosis–deafness (KID) syndrome, hystrix-like ichthyosis with deafness, Vohwinkel syndrome, Bart–Pumphrey syndrome and palmoplantar keratoderma with deafness—whereas loss-of-function variants cause nonsyndromic hearing loss [3, 38]. In KID syndrome, abnormal connexin 26 hemichannel opening compromises keratinocyte integrity; p.Ala88Val hemichannels are insensitive to carbon dioxide and associated with central apnoea, and p.Gly45Glu and p.Ala88Val have been associated with childhood death [38]. GJB6 missense variants cause Clouston hidrotic ectodermal dysplasia, whose defining triad (alopecia, nail dystrophy and palmoplantar keratoderma) does not include hearing loss [2, 39].

2.4. STRC (stereocilin) and the contiguous STRC-CATSPER2 deletion

STRC (15q15.3) encodes stereocilin, an extracellular structural protein of outer hair cells forming horizontal links between adjacent stereocilia and linking the hair bundle to the tectorial membrane, allowing sound damping, amplification of quiet sounds and frequency tuning [4]. Because inner hair cells remain functional, stereocilin loss impairs rather than abolishes hearing, explaining the typical mild to moderate DFNB16 hearing loss [4]. The region lies within a segmental duplication with a pseudogene (STRCP1) sharing 98.9% homology across the sequence and more than 99% in the coding region, complicating sequencing [40]. Approximately 80% of pathogenic alleles are copy-number variants (CNVs), and the most prevalent genotype is the biallelic contiguous deletion covering STRC and CATSPER2; loss of CATSPER2 on both alleles causes male infertility through impaired sperm motility (deafness–infertility syndrome, OMIM 611102) [4, 41].

2.5. OTOF (otoferlin): an auditory synaptopathy

OTOF (2p23.3) encodes otoferlin, a protein of the FER-1 family with six C2 domains (C2A to C2F), a transmembrane domain and two Fer domains [23]. Otoferlin is the calcium sensor for exocytosis at the ribbon synapses of inner hair cells: it tethers glutamatergic synaptic vesicles to the membrane, triggers their fusion and replenishes the pool of available vesicles, processes indispensable for the sustained and temporally precise release of glutamate on which sound localisation and speech comprehension depend [23]. Loss of function impairs exocytosis and transmission to the auditory nerve while outer hair cells remain intact; therefore, OAEs are present and ABRs are absent or severely abnormal, and speech discrimination is severely reduced in all patients [23]. The correct term is auditory synaptopathy (or prelingual non-syndromic auditory neuropathy spectrum disorder, ANSD) [23]. Variants should be annotated on the cochlea-specific transcript NM_001287489.2 [23]. The Oxford Desk Reference also cites PJVK (pejvakin) as a cause of non-syndromic auditory neuropathy [2]. According to the French PNDS, biallelic OTOF variants account for approximately half of hearing loss in the auditory neuropathy spectrum; among the dozens of variants described, p.(Gln829Ter) is recurrent in some populations [20].

2.6. SLC26A4 (pendrin): Pendred syndrome and DFNB4

SLC26A4 (7q22.3) encodes pendrin, a non-specific anion exchanger that exchanges anions (chloride, iodide) for bases (bicarbonate, hydroxyl) across the apical membrane of epithelial cells of the inner ear, thyroid, kidney and airways [5]. In the inner ear, absence or reduction of pendrin function acidifies the endolymph, leads to prenatal formation of an enlarged vestibular aqueduct (EVA) and, through mechanisms not yet clarified, to degeneration of cochlear sensory cells [5]. In the thyroid, it reduces iodide efflux into the follicular lumen and impairs iodine organification (thyroid dyshormonogenesis type 2B) [5, 42]. The mechanism is loss of function and there are no clinically relevant genotype–phenotype correlations [5]. More than 600 pathogenic variants have been described, mostly single-nucleotide variants or small insertions and deletions; intragenic deletions and duplications are rare [5]. The French PNDS counted, in June 2018, 535 pathogenic variants recorded in HGMD: 354 missense or nonsense, 75 at canonical splicing sites, small deletions, insertions and delins, six large deletions and two complex rearrangements; the same document attributes practically all cases of Pendred syndrome to SLC26A4 [7].

A substantial proportion of patients with EVA have only one identified pathogenic variant. Among patients of Chinese, Korean and Japanese ancestry with EVA, 67–90% have two variants (M2 genotype) and 8–21% one (M1); among patients of Northern European ancestry, approximately 25% are M2, 25% M1 and 50% have no identifiable variant (M0) [5]. Many M1 patients carry in trans a haplotype termed CEVA (Caucasian EVA), comprising 12 single-nucleotide variants across a 0.89 Mb region extending from upstream of PRKAR2B to SLC26A4 intron 3, present in 3% of the non-Finnish European population; because it does not always segregate with the phenotype and has no demonstrated functional effect, it is classified as a variant of uncertain significance and cannot establish the diagnosis [5]. Digenic inheritance involving FOXI1, KCNJ10 or EPHA2 variants has been proposed but not confirmed by recent studies [1, 5].

2.7. Other autosomal recessive hearing loss genes

Source: adapted from GeneReviews, Genetic Hearing Loss Overview, Table 3, and the indicated sources.

2.8. Autosomal dominant hearing loss genes

Source: adapted from GeneReviews, Genetic Hearing Loss Overview, Table 4, and the indicated sources.

2.9. X-linked hearing loss genes (DFNX)

Six genes are associated with X-linked non-syndromic hearing loss: AIFM1, COL4A6, GPRASP2, POU3F4, PRPS1 and SMPX [1]. POU3F4 (DFNX2, OMIM 304400) is the most common cause and produces mixed hearing loss with a conductive component due to stapes fixation; CT shows incomplete partition type III of the cochlea with a dilated internal auditory canal, an image considered pathognomonic, and the communication between cerebrospinal fluid and perilymph explains the risk of perilymphatic gusher and total hearing loss in stapes surgery [1, 2, 57]. PRPS1 gives rise to a spectrum ranging from isolated hearing loss (DFNX1) to syndromic forms of phosphoribosylpyrophosphate synthetase deficiency and superactivity [1, 58]. AIFM1 is associated with X-linked auditory neuropathy, isolated or with other neuropathies [1]: the entity is X-linked deafness type 5 (DFNX5, formerly designated AUNX1), and the other disorders of the same gene are X-linked Charcot-Marie-Tooth disease type 4, combined oxidative phosphorylation deficiency type 6 and spondyloepimetaphyseal dysplasia with hypomyelinating leukodystrophy [1].

2.10. Mitochondrial genome: MT-RNR1 and MT-TS1

Mitochondrial DNA (mtDNA) is transmitted exclusively by the mother. Homoplasmy means a single mtDNA type in a cell, tissue or individual—the Oxford Desk Reference specifically uses mitochondria containing only m.1555A>G mtDNA as an example—whereas heteroplasmy means a mixture of mutant and normal mtDNA in proportions varying between cells and tissues [59]. The mitochondrial bottleneck in the female germline explains variation in the mutation burden transmitted to children by a heteroplasmic mother [59].

MT-RNR1 encodes 12S ribosomal RNA, the mitochondrial homologue of bacterial 16S ribosomal RNA, to which aminoglycosides bind [10]. m.1555A>G and m.1494C>T lie in the mitochondrial ribosome's decoding centre and alter the conformation of the aminoglycoside-binding loop, making it resemble the bacterial loop; this increases aminoglycoside binding to the mitochondrial ribosome, reduces mitochondrial protein synthesis, causes amino acid misincorporation and destroys hair cells [8]. m.1095T>C also has sufficient evidence for the drug–gene interaction [10]. The mechanism of hearing loss without aminoglycoside exposure is unknown; penetrance is modulated by nuclear factors and additional MT-RNR1 variants [8].

MT-TS1 encodes the serine (UCN) transfer RNA. The variants m.7445A>G, m.7505T>C and m.7511T>C reduce the amount of tRNA-Ser(UCN) through a processing or stability defect; m.7445A>G places a non-cleavable cytosine at the endonucleolytic processing junction [8]. The three bases m.7443–7445 also encode the stop codon of MT-CO1, which becomes a sense codon and extends the protein by three amino acids [8]. m.7471dupC (formerly m.7472insC) causes isolated progressive hearing loss or hearing loss with ataxia, dysarthria and myoclonus [2, 8, 60]. In families with two or more affected members in generations linked through the maternal line, mitochondrial variants were found in up to 30%, but in only 1% of sporadic early-onset non-syndromic hearing loss; the age at onset is highly variable [2, 60].

3. Clinical Spectrum and Manifestations by System

3.1. Audiological characterisation

Hearing loss is characterised by type, onset age (congenital, prelingual, postlingual, adult after 18 years, and presbycusis), severity, affected frequencies, laterality and audiogram configuration (rising with low-frequency predominance, “U-shaped” with mid-frequency predominance, falling with high-frequency predominance, or flat) [1]. Hearing is considered normal when thresholds are less than 15 dB above reference thresholds, defined as the level at which young adults with normal hearing perceive a pure-tone stimulus in 50% of presentations [1]. Severity classifications in the two main sources differ slightly and are presented side by side.

Sources: [1, 2]. Frequencies are classified as low (<500 Hz), middle (501–2000 Hz) and high (>2000 Hz) [1]; conversational speech lies between 45 and 60 dB [2].

The audiometric pattern can guide diagnosis: low-frequency hearing loss is unusual and suggests WFS1 or DIAPH1 variants; “U-shaped” configurations are usually dominant and associated with COL11A2 and TECTA; a steep high-frequency ski-slope configuration suggests TMPRSS3 [2]. Aminoglycoside-induced hearing loss tends to cause marked high-frequency loss [2]. Approximately 50% of hearing loss is associated with peripheral vestibular dysfunction, because the cochlea and vestibular organs (saccule, utricle and semicircular canals) share physiology and embryological origins [1].

3.2. GJB2-related hearing loss (DFNB1)

DFNB1 is the most common genetic cause of severe to profound, non-progressive congenital sensorineural hearing loss in many populations [3]. It can also be mild to moderate, usually stable, in which form it escapes newborn screening [3]. Vestibular function is normal, children sit and walk at expected ages, and apart from hearing loss they are healthy with normal life expectancy [3]. A systematic review of 216 studies with more than 43,000 probands documented progression, asymmetry or imaging abnormalities in 14–19% of patients [61].

Genotype–phenotype correlation is strong and organised into three classes defined by the truncating (T) or non-truncating (NT) nature of each allele [3, 62]:

p.Val37Ile follows a distinct trajectory: more than 40% of newborns homozygous or compound heterozygous with a truncating allele have thresholds below 20 dB and pass OAE screening, but thresholds decline with age, so approximately 60% of individuals aged 40–60 and approximately 80% aged 60–85 have moderate hearing loss [3]. In a series of eight p.Val37Ile homozygotes, one received a cochlear implant because of progression, six used hearing aids and one was untreated [3]. Intrafamilial variability occurs: 67% of sibling pairs have similar thresholds (difference below 20 dB) and 33% differ by more than 30 dB [3]; the Oxford Desk Reference reports that 20% of sibling pairs differ by two or more severity classes [2]. Occasional newborn screening failures have been documented in severe forms: one newborn passing automated ABR (AABR) screening was diagnosed with hearing loss at 15 months, and another child with normal audiometry at five months had severe hearing loss at nine months [3].

3.3. STRC-related hearing loss (DFNB16)

STRC-related hearing loss is congenital, bilateral, symmetrical and mild to moderate, with a pure-tone average of 40–50 dB at diagnosis and a gently sloping audiogram; it does not reach severe to profound levels in childhood or young adulthood [4]. OAEs are usually absent, and computed tomography (CT) and magnetic resonance imaging (MRI) of the temporal bones and brain are normal [4]. Progression is slow, approximately 0.6 dB/year: among 31 patients followed for a mean of 6.1 years, 18 progressed, but more than 96% retained moderate or better thresholds at follow-up completion [4]. Preliminary data suggest increased recurrent benign paroxysmal positional vertigo risk (39% of 64 patients, median age 13 years), requiring replication in other series [4]. Men with contiguous STRC-CATSPER2 deletions on both alleles have an infertility risk from sperm morphological and motility abnormalities; women with the same deletion and men with one deletion and one intragenic variant do not have this risk [4].

3.4. OTOF-related hearing loss (DFNB9)

The typical form is congenital or prelingual, bilateral and usually severe to profound (70 to ≥90 dB), with a flat audiogram, normal OAEs and abnormal ABRs; severe forms commonly progress to profound hearing loss [23]. With age, OAEs decrease or disappear in 20–80% of patients, potentially erasing the electrophysiological signature [23, 63]. Unilateral OTOF-related hearing loss has not been described; unilateral loss with an auditory neuropathy pattern warrants MRI investigation for a cochlear defect [23].

Atypical forms include temperature-sensitive and progressive hearing loss [23]. In the former, hearing ranges from normal to moderate loss (0–55 dB) at baseline temperature and falls to severe (71–90 dB) or profound (>90 dB) loss when body temperature rises by approximately 0.5 °C or more; above 38 °C, deterioration is rapid and recovery occurs within hours of temperature normalisation [23]. Thirteen variants are associated, including p.Ile515Thr, p.Gly541Ser, p.Gly614Glu, p.Glu841Lys, p.Arg1080Pro, p.Arg1607Trp, p.Pro1628Thr, p.Glu1700Gln and p.Glu1804del; all patients have two non-truncating variants or one truncating and one non-truncating variant [23]. The progressive form starts as mild to moderate and can become profound over months or years (for example, p.Ile1573Thr or p.Glu1700Gln homozygosity) [23]. In all forms, speech discrimination, particularly in noise, is worse than pure-tone audiometry would predict [23]. Biallelic truncating variants cause profound prelingual hearing loss, biallelic non-truncating variants milder phenotypes, and their combination an intermediate phenotype [23, 63].

The French PNDS describes two presentation profiles [20]. In the young child without oral language, the bilateral sensorineural hearing loss on behavioural audiometry is confirmed by absent or desynchronised ABRs with present evoked OAEs; the PNDS example shows bilateral profound hearing loss with OAEs present in both ears (signal-to-noise ratio above 6 dB) and ABRs without an identifiable wave up to 100 dB [20]. In the child with less severe involvement, who has sometimes already developed oral language, the pure-tone audiogram shows mild to severe bilateral loss, but intelligibility on speech audiometry is worse than the audiogram would suggest, and the hearing loss may fluctuate or worsen during febrile episodes (temperature-sensitive form) [20]. DFNB9 is non-syndromic: the PNDS describes no associated ophthalmological disease or vestibular involvement, inner ear imaging is normal and vestibular tests are not systematic once the diagnosis has been established [20].

3.5. SLC26A4 spectrum: Pendred syndrome and DFNB4

Hearing. Onset varies (congenital, prelingual or postlingual), and many children pass newborn screening [5]. Hearing loss fluctuates, improving between deteriorations, and often progresses stepwise; by age three, the pure-tone average is approximately 80 dB, and over time almost all patients develop severe to profound bilateral hearing loss [5, 64]. The Oxford Desk Reference describes severe to profound congenital hearing loss with high-frequency predominance and fluctuating progression, emphasising that speech and language are often more developed than the audiogram would predict [2]. A low-frequency air–bone gap may occur with normal tympanometry, reflecting the enlarged aqueduct's “third-window” effect [5]. Hearing deteriorations after head trauma, infection or vestibular episodes have been reported, but trauma is not considered a major risk factor for overall progression [5].

The French PNDS places the onset of hearing loss at birth or in the prelingual period in approximately two-thirds of patients; in the others, hearing loss appears later, sometimes in adulthood, and is then revealed by an episode of sudden hearing loss or by goitre in a deaf patient without a previous aetiological diagnosis [7]. The hearing loss is usually severe and bilateral, may be asymmetrical, and a mixed component is attributed to raised inner ear pressure [7]. When hearing loss is initially unilateral, the normal ear may deteriorate later, sometimes abruptly, and it is this ear that determines the functional prognosis [7]. The document lists as triggers of abrupt worsening head trauma, even minor, pressure and acoustic trauma, upper respiratory tract infections and fever, at any age; neither the type of variant nor the calibre of the aqueduct correlates with the initial severity or with the course [7].

Inner ear. Bilateral EVA is present in all patients and is defined by an endolymphatic duct midpoint diameter ≥1.0 mm or opercular diameter ≥2.0 mm [5]. Cochlear incomplete partition type II (1.5 turns rather than 2.5, with a cystic apex) occurs in 22–74%; the combination of an incomplete cochlea, enlarged vestibule and EVA constitutes the Mondini triad, distinct from the broader use of “Mondini dysplasia” [5]. Modiolar hypoplasia or deficiency is also frequent [5]. The Oxford Desk Reference reports EVA in approximately 80% with an enlarged endolymphatic sac; this difference reflects varying criteria and imaging techniques, while the latest GeneReviews considers EVA universal in patients with biallelic variants [2, 5].

Vestibular function. Vestibular dysfunction affects 4–47% of patients, without correlation with the severity of hearing loss; it manifests as head tilt, vomiting, episodic rotatory vertigo, clumsiness or loss of previously acquired walking, and may precede or accompany the auditory fluctuations [5]. Caloric tests are abnormal in approximately 30%, rotary chair in approximately 25% and cervical vestibular evoked myogenic potentials in approximately 20% [5]. The French PNDS, which groups Pendred syndrome and DFNB4 in the same series because of the scarcity of patients, reports clinical vestibular signs (attacks of vertigo or instability) in one to three-quarters of patients with biallelic variants and dysfunction on testing, mainly canal caloric hyporeflexia, in most, including patients without symptoms; there is no description of otolith function in the syndrome [7]. In patients with EVA of any cause, cVEMPs have increased amplitude and a low threshold at 250 and 500 Hz, a third-window effect similar to that which produces the conductive component [7]. The difference between this range and the 4–47% of the GeneReviews review depends on the series included and on whether symptoms or test abnormalities are counted [5, 7].

Thyroid. The defect in iodine organification may lead to goitre, with or without hypothyroidism [5]. Goitre, multinodular or diffuse, develops gradually and becomes evident from the second decade, especially when the diet is not supplemented with iodine; it is rare in countries with universal salt iodisation and has incomplete penetrance [5, 42]. In the absence of dietary iodine, the risk of goitre increases 1.1-fold per year of age, which corresponds to a 1.6- and 2.6-fold increase over five and ten years [5, 65]. Most patients are euthyroid, with TSH at the upper limit of normal, and only rarely present with congenital hypothyroidism; hypothyroidism of variable severity may appear later [42]. The French PNDS places the onset of thyroid abnormalities mainly around adolescence, with euthyroid or hypothyroid goitre or non-specific signs of hypothyroidism, and estimates that approximately 10% of children with biallelic variants develop the complete syndrome, with no known predictive factors [7]. It considers neonatal goitre and congenital hypothyroidism exceptional in Pendred syndrome, to the point of warranting a search for another cause [7]. Pendred syndrome accounts for approximately 5% of severe to profound hearing loss in childhood [2].

3.6. Mitochondrial hearing loss: MT-RNR1 and MT-TS1

Aminoglycoside ototoxicity. In carriers of MT-RNR1 risk variants, hearing loss develops days to weeks after any amount, including a single dose, of an aminoglycoside (gentamicin, tobramycin, amikacin, kanamycin or streptomycin) [8]. It is bilateral, severe to profound, irreversible but non-progressive, and cochlear in origin through hair cell loss and dysfunction; vestibular symptoms are uncommon [8]. Unlike dose- and plasma-concentration-dependent ototoxicity that can occur in anyone, carrier ototoxicity is dose-independent [8]. Penetrance is considered almost 100% after exposure, although exposed carrier newborns without hearing loss have been described [8]. In the reference series, m.1555A>G was identified in 15% of individuals with hearing loss and previous aminoglycoside exposure [8].

Hearing loss without exposure. m.1555A>G also causes nonsyndromic hearing loss without aminoglycosides, with highly variable severity and onset, from profound congenital loss to late-onset progressive mild to moderate loss with high-frequency predominance; tinnitus is frequent and vestibular symptoms rare [8]. Penetrance without exposure ranges from 0% to 65% between families; in five families with different heteroplasmy levels, penetrance ranged from 10% to 52% and correlated with mutation burden [8]. Estivill and colleagues showed that familial progressive sensorineural hearing loss in Spanish families was predominantly due to m.1555A>G and worsened by aminoglycosides [66]. For m.1494C>T, mean penetrance in eight Chinese families was 31.7% including aminoglycoside-induced loss and 17.5% excluding it [8]. m.961T>G and m.961_962delTinsC(n) have uncertain significance [8, 10].

MT-TS1. Hearing loss associated with m.7445A>G begins in childhood, is variable in severity and characteristically progressive; some families also have palmoplantar keratoderma (scaling, hyperkeratosis and a honeycomb appearance on palms, soles and heels), developing as early as four or five years [2, 8]. Families with m.7444G>A together with m.1555A>G, or isolated m.7443A>G, lack skin changes [8]. m.7444G>A at the MT-CO1/MT-TS1 boundary has been associated with nonsyndromic and aminoglycoside-induced hearing loss [1, 8].

3.7. Autosomal dominant and X-linked forms

DFNA2 (KCNQ4) hearing loss is symmetrical and begins at high frequencies; in young patients it is mild at low frequencies and moderate at high frequencies, and in older patients moderate at low frequencies and severe to profound at high frequencies. Most patients need hearing aids between ages 10 and 40, and by 70 all have severe to profound hearing loss [48]. Speech recognition is relatively good for the pure-tone thresholds [48]. DFNA9 (COCH) begins in adulthood at high frequencies and is associated with vestibular dysfunction and Ménière-like symptoms; unlike sporadic Ménière disease, which predominates at low frequencies, DFNA9 hearing loss predominates at high frequencies [2, 54]. Families with dominant MYO7A variants and mild vestibular symptoms have also been described [2]. In DFNX2 (POU3F4), men have progressive mixed hearing loss, and heterozygous women may have late-onset hearing loss or an inner ear malformation [57].

4. Clinical Recognition and Warning Signs

Nonsyndromic genetic hearing loss is recognised in three settings: abnormal newborn hearing screening, a child with speech delay or school difficulties, and an adolescent or adult with progressive hearing loss or tinnitus [1, 3]. DFNB1 or DFNB16 should be suspected in children aged three to five with stable mild to moderate loss, no systemic findings and a recessive-compatible family history [3, 4].

History should include a pedigree spanning at least three generations, documenting onset, laterality and progression, with targeted questions about sudden death or syncope (Jervell and Lange-Nielsen), goitre (Pendred), renal disease (Alport, branchio-oto-renal), visual impairment (Usher, peroxisomal and mitochondrial disorders) and pigmentation changes (Waardenburg, piebaldism) [2]. Consanguinity with otherwise negative investigations warrants counselling as recessive inheritance [2]. Prenatal infections (CMV, rubella), severe jaundice, meningitis and aminoglycoside exposure should be recorded; hearing loss secondary to jaundice or perinatal factors usually accompanies neurological changes and learning difficulties because the brain is more sensitive to hypoxia and hyperbilirubinaemia than the inner ear [2]. The child's audiograms should be obtained before consultation, and a sign language interpreter may be needed if a parent has profound congenital hearing loss [2].

Delayed motor milestones are frequent in neurologically normal deaf children because of vestibular involvement, prompting exclusion of Usher type 1, Jervell and Lange-Nielsen, Pendred and branchio-oto-renal syndromes through electrocardiography, electroretinography and inner ear imaging [2]. Examination seeks external ear malformations, preauricular tags and pits, dystopia canthorum, iris heterochromia or hypoplasia, lower eyelid coloboma, retinal pigmentation, branchial cysts, goitre, a white forelock, skin pigmentation changes and neurological signs [2]. Optic atrophy should be sought with diabetes (WFS1) or auditory neuropathy (OPA1) [2].

Approximately 20% of children whose only initial manifestation is hearing loss eventually receive a syndromic diagnosis, so “nonsyndromic” remains provisional until molecular diagnosis [1]. In a recent Brazilian study, 10.4% of patients initially classified as nonsyndromic carried pathogenic or likely pathogenic variants in syndromic genes (PEX6, BSND, USH1C, WFS1) [67].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Newborn hearing screening

Universal newborn hearing screening uses physiological methods: OAEs assess outer hair cell responses, and AABRs assess inner hair cells, the auditory nerve, brainstem and brain [1, 3]. In the United States, screening is mandatory in every state, covers more than 98% of newborns and identifies approximately 6000 children with hearing loss annually [1]. In Brazil, Federal Law No. 12.303 of 2 August 2010 made evoked otoacoustic emission testing (“teste da orelhinha”, the newborn hearing test) free and mandatory in hospitals and maternity units [68]; a national survey of speech and language therapists in 2018–2019 showed that practice did not fully follow official protocols, 48.5% reported screening interruptions and most programmes did not meet predefined quality indicators [69].

Screening was designed to detect moderate to profound loss and may miss mild loss; CDC data show that more than 25% of bilateral newborn hearing loss is slight or mild (<40 dB), and lowering screening thresholds would increase false positives [3]. OAE-only screening does not detect most typical OTOF-related cases, identified only when AABRs are used [23]. Children with postnatal-onset SLC26A4 loss, mild GJB2 or STRC loss and temperature-sensitive loss may also pass screening [3–5, 23]. Screening should be completed by one month, diagnosis established by three months and intervention started by six months (“1-3-6” targets); US states already meeting these should seek a “1-2-3” timetable [3, 15]. Early OTOF molecular diagnosis has become more urgent because gene therapy cannot be performed in an already implanted ear [1]. An opinion article published in September 2026 therefore proposes inclusion of AABR in screening recommendations to identify gene therapy candidates; this is an expert proposal rather than a change to JCIH recommendations [15, 70].

In France, systematic newborn hearing screening became a regulatory obligation with the ministerial order of 23 April 2012 (published on 4 May 2012) and was implemented nationally from December 2014; it is organised by the maternity unit paediatrician, on the second day of life, by evoked OAEs or by AABR, and an abnormal test is repeated before discharge, the child being referred to a reference centre if the second also fails [20]. The PNDS stresses that only AABR detects DFNB9 in the maternity unit and that OAE screening gives a reassuring result that delays diagnosis; it also states that a normal AABR screen excludes DFNB9 [20]. This last statement applies to the typical congenital form: temperature-sensitive forms, with normal or slightly altered baseline hearing, and progressive forms of mild onset may pass screening (3.4) [20, 23].

5.2. Audiological confirmation and ENT assessment

An abnormal screen requires immediate confirmation by diagnostic ABR, followed by assessment by an otorhinolaryngologist, who excludes non-permanent causes (otitis media with effusion) and anomalies of the external and middle ear, and looks for syndromic signs [1, 3]. Audiological assessment is adapted to age: visual reinforcement audiometry below 2.5 years, conditioned play audiometry between 2.5 and 5 years and conventional audiometry above 5 years; ABR and auditory steady-state responses (ASSR) are used in young or uncooperative children [1, 3]. Tympanometry and comparison of air and bone conduction distinguish conductive from sensorineural hearing loss [1]. In OTOF, testing of speech discrimination in quiet and in noise is recommended [23]. The terminology and the pathway for identification and management of auditory neuropathy spectrum disorder still follow the 2008 multidisciplinary consensus document, cited by the GeneReviews review as the reference for ANSD [23, 71]. For hearing loss associated with m.3243A>G, the French PNDS for the MELAS spectrum proposes in all patients an age-appropriate assessment with pure-tone air and bone conduction audiometry (with contralateral masking if the hearing loss is asymmetrical), speech audiometry with headphones and, if useful, in noise, free-field speech audiometry in patients with a hearing aid or implant to measure the benefit, impedance audiometry with stapedial reflexes and ABR or ASSR to confirm the diagnosis or document worsening; a speech and language therapy assessment helps to measure the impact of these progressive hearing losses [19].

5.3. Aetiological algorithm

Initial evaluation of anyone with hearing loss should include genetic testing alongside hearing and family histories, examination, audiometry, vestibular assessment, consideration of imaging and targeted additional investigations [1, 14]. Investigation sequence and timing depend on age and habilitation options: a child with severe to profound loss requires CT or MRI before cochlear implantation, often performed alongside molecular testing given its one-to-three-month turnaround [1].

Sources: [1, 2, 14].

Diagnostic yield reaches 60% in children with severe to profound congenital bilateral sensorineural hearing loss, approximately 40% in older children and approximately 20% in adults [1]. In a US series of 1119 patients tested with a comprehensive panel, the most frequent genes were GJB2, STRC and SLC26A4 [50]. A specific diagnosis is established by identifying pathogenic or likely pathogenic variants, equivalent for clinical decisions under ACMG/AMP; a variant of uncertain significance neither confirms nor excludes diagnosis [1].

5.4. Gene-specific laboratory considerations

GJB2. Sequencing detects more than 99% of variants; deletion/duplication analysis detects fewer than 1% [3]. If Sanger sequencing identifies one variant, transition to a full panel is recommended [3]. Apparent homozygosity may result from an undetected deletion or uniparental isodisomy [3].

STRC. Sequencing detects approximately 20% of variants and CNV analysis approximately 80%; diagnoses comprise approximately 30% biallelic variants, 40% biallelic contiguous deletions and 30% compound heterozygosity [4]. Long-range PCR amplification or MLPA eliminates pseudogene contamination, and genome sequencing is more sensitive to deletions than exome sequencing [4].

OTOF. Sequencing detects 98% of variants and deletion/duplication analysis approximately 2%; no deep intronic variants have been described in the NM_001287489.2 transcript [23]. A structural variant was identified by long-read sequencing with nanopore adaptive sampling [23]. Biallelic molecular confirmation is mandatory for gene therapy [11]. The French PNDS requires, to confirm the diagnosis, DNA from the child and both parents, in order to demonstrate that the two variants are on different alleles; the French decision tree runs from congenital sensorineural hearing loss to the clinical genetics clinic and the hearing loss gene panel [20].

Table 5.4a — Initial assessment of DFNB9 according to the French PNDS (2022) [20]

SLC26A4. The diagnosis requires two variants; if only one is found, an intragenic deletion or duplication should be sought, and the perchlorate discharge test may be considered [5]. The French PNDS maintains Sanger sequencing as the reference method, accepts next-generation sequencing panels and requires analysis of the whole coding sequence, and not only of the most frequently affected exons [7]. When a single allele is found, it proposes looking for a large rearrangement in trans by array CGH or MLPA and lists as unconfirmed hypotheses promoter variants at the FOXI1 binding sites, digenic inheritance with FOXI1 or KCNJ10 and a still unknown variant in the CEVA haplotype [7]. Once the diagnosis has been established in the index case, it recommends study of the parents to confirm segregation and study of adult relatives at risk of being carriers or affected [7].

Mitochondrial DNA. After aminoglycoside-associated hearing loss, targeted m.1555A>G, m.1494C>T and m.7445A>C/T/G testing may be the first step; complete mtDNA sequencing is indicated when the panel is negative but suspicion remains [8]. Methods must detect heteroplasmy, and mutation burden may be higher in urinary epithelium or muscle than in blood [8, 59]. Commercial MT-RNR1 tests do not include rare variants, so a “reference” result may rarely conceal a risk variant [10]. A rapid point-of-care genetic test genotyped m.1555A>G in 26 minutes, with 100% sensitivity and specificity in preclinical validation [72].

5.5. Diagnostic criteria for the SLC26A4 spectrum

There are no consensus clinical criteria; diagnosis rests on two pathogenic SLC26A4 variants [5]. Premolecular Pendred syndrome clinical criteria included sensorineural hearing loss, EVA with or without cochlear and vestibular malformations, and an iodine organification defect, with or without goitre [5]. The perchlorate discharge test is the most sensitive clinical method for demonstrating the organification defect but is rarely used; positivity indicates dyshormonogenesis [5, 42]. Thyroid scintigraphy, cervical ultrasound and thyroid function (T3, free T4, TSH) complete assessment; normal function does not exclude diagnosis [2, 5]. Distinguishing Pendred syndrome from DFNB4 is unreliable in childhood because of intrafamilial variability and late goitre development [5].

Initial assessment according to the French PNDS. The PNDS describes three diagnostic circumstances: the finding of bilateral EVA on imaging of a deaf child, the appearance of goitre or hypothyroidism in a deaf child or young adult, and a family history of Pendred syndrome; the diagnosis is confirmed only with two pathogenic SLC26A4 variants, requested by the geneticist on blood or saliva [7]. The algorithm of the document starts from sensorineural hearing loss, passes through imaging with bilateral EVA and gene study and separates three situations: no pathogenic variant, no thyroid surveillance; one or two variants, annual thyroid surveillance; and, among these, Pendred syndrome when goitre or hypothyroidism appears and DFNB4 when the thyroid remains normal [7].

Perchlorate discharge test (PNDS protocol). Baseline thyroid uptake of iodine-123 is measured and, 180 minutes after its administration, the patient takes potassium (or sodium) perchlorate orally, 1 g in adults and 15 mg/kg in children; thyroid activity is measured at 15, 30, 45 and 60 minutes (some protocols extend it to two hours) and the result is expressed as the percentage fall relative to the initial uptake [7]. A fall of less than 10% excludes an organification defect; above 10% the test is usually considered positive, although some centres use thresholds of up to 30% [7]. A normal test does not exclude Pendred syndrome, and in patients with two variants a discharge greater than 30% may predict the onset of hypothyroidism or goitre [7].

Newborn with congenital hypothyroidism. The PNDS provides, under the responsibility of the maternity unit paediatrician in liaison with paediatric endocrinology, for clinical examination (macroglossia, umbilical hernia, jaundice, goitre), hearing screening by OAEs or AEPs with referral to a deafness centre if abnormal, measurement of thyroid hormones and thyroglobulin, ultrasound with lobe volume, iodine-123 scintigraphy and discharge test before or shortly after the start of replacement, molecular confirmation and a search for cardiac and renal anomalies pointing to another cause of congenital hypothyroidism [7].

5.6. Genetic test limitations and report content

A negative hearing loss panel does not exclude a genetic cause when the phenotype is suggestive, because technical coverage varies considerably between laboratories. Critical points are predictable: upstream GJB2 deletions involving GJB6 or CRYL1 are not seen by coding exon sequencing [3, 35]; in STRC, approximately 80% of pathogenic alleles are CNVs within a segmental duplication containing STRCP1, and short-read sequencing without targeted methods (long-range PCR, MLPA or genome sequencing with a dedicated algorithm) misses most diagnoses [4, 40]; OTOA also has a pseudogene (OTOAP1) and many structural alleles [46]; and OTOF multiexon deletions and a structural variant resolved only by long reads have been described [23, 73]. Intronic variants outside canonical splicing positions require panel coverage beyond ±10 base pairs: the founder OTOF variant c.898-18G>A, causing exon 10 loss in a minigene assay and with a 1.5% carrier frequency among Jews of Moroccan origin, escapes a design covering only exon-adjacent nucleotides [74]. For mitochondrial DNA, reports should state whether mtDNA was analysed (many panels and exomes exclude it), the method, heteroplasmy detection threshold and tissue tested, because burden may be higher in urine or muscle than in blood [8, 59]. The SLC26A4 CEVA haplotype and MT-RNR1 variants with uncertain risk (for example, m.961T>G) should not be used to establish diagnosis [5, 10]. Finally, the proportion of variants detectable by a method (for example, more than 99% of GJB2 variants by sequencing) is not its clinical sensitivity: comprehensive panel diagnostic yield is approximately 60% in children with severe to profound congenital loss, 40% in older children and 20% in adults [1, 3].

Reports should include the gene, versioned transcript, HGVS nomenclature, zygosity and, for two variants in the same recessive gene, phase (cis or trans) demonstrated by parental studies or long reads; each variant's ACMG/AMP classification; and technical limitations (CNVs, homologous regions, introns, mtDNA, mosaicism) [1, 14]. MANE v1.5 transcripts are GJB2 NM_004004.6, GJB6 NM_001110219.3, SLC26A4 NM_000441.2, STRC NM_153700.2, MYO7A NM_000260.4, MYO15A NM_016239.4, CDH23 NM_022124.6, TMC1 NM_138691.3, TMPRSS3 NM_001256317.3, KCNQ4 NM_004700.4, COCH NM_004086.3, TECTA NM_005422.4 and POU3F4 NM_000307.5 [22]. OTOF requires particular attention: MANE Select is NM_194248.3 (transcript variant 1), with NM_194323.3 as MANE Plus Clinical, whereas GeneReviews recommends annotation on cochlear transcript NM_001287489.2 (transcript variant 5), also the isoform carried by lunsotogene parvec; because numbering may differ between transcripts in alternative splicing regions, the transcript must be stated before comparison with literature variants or assessment of gene therapy eligibility [11, 22, 23].

Sources: [3, 8, 10, 22, 75].

A variant of uncertain significance does not become pathogenic through partial segregation in a small family or a favourable hearing response to an intervention; segregation studies and functional assays are additional investigations, and reclassification is valid only when sufficient ACMG/AMP criteria are met [1]. Similarly, identifying one pathogenic allele in a recessive gene, as in approximately one-quarter of Northern European EVA patients, confirms neither diagnosis nor eligibility for targeted therapies [5, 11].

6. Differential Diagnosis

6.1. Acquired causes

Environmental causes account for approximately 35% of severe childhood hearing loss: congenital infection (CMV, rubella), perinatal asphyxia, prolonged hyperbilirubinaemia (serum bilirubin >250 μmol/L), meningitis and ototoxic medicines [2]. cCMV birth prevalence is approximately 0.64%; approximately 10% of infected newborns are symptomatic (with hearing loss in approximately 50%), and of the 90% who are asymptomatic, up to 10% develop unilateral or bilateral loss, often variable [1]. Other causes include TORCH infections, bacterial meningitis (Neisseria meningitidis, Haemophilus influenzae, Streptococcus pneumoniae) and, in adults, noise and medication [1]. ANSD also has acquired causes—prematurity, hyperbilirubinaemia, septicaemia, cisplatin, loop diuretics and aminoglycosides—with a prevalence of approximately 5.6 per 1000 in neonatal intensive care units [23]. Aminoglycoside ototoxicity without mitochondrial predisposition depends on dose and peak and trough concentrations [8].

6.2. Syndromic forms mimicking nonsyndromic hearing loss

Two entities complete this differential diagnosis: Perrault syndrome, in which bilateral sensorineural hearing loss is associated in women with ovarian insufficiency, and 10q26 deletion syndrome, in which loss of HMX2 and HMX3 is associated with sensorineural hearing loss and inner ear malformations.

m.3243A>G hearing loss (MIDD and the MELAS spectrum). According to the French PNDS for the MELAS spectrum, the hearing loss of maternally inherited diabetes and deafness is sensorineural, bilateral, progressive and of endocochlear origin; it usually begins in young adulthood (on average between 20 and 40 years), usually precedes the diagnosis of diabetes and worsens with age, apparently faster in men [19]. In approximately half of patients progression occurs in rapid steps and in the other half it is gradual over the years; it first affects the high frequencies and eventually becomes severe at all frequencies [19]. By 40 years of age, 84–95% of patients have bilateral hearing loss, according to a Japanese and a Finnish series cited by the PNDS; the proportion of 11% that progresses to profound hearing loss with an indication for cochlear implantation comes from a study of implantation in MIDD and not from a population-based cohort [19]. Vertigo and vestibular deficit are rare, and the severity of hearing loss does not correlate with heteroplasmy in blood [19]. Pattern macular dystrophy, described in up to 85% of cases of MIDD, diabetes in a non-obese person and a maternal history of diabetes or deafness distinguish this entity from non-syndromic forms [19].

Saudubray notes that hearing loss in inborn errors of metabolism is usually sensorineural, symmetrical and initially high-frequency, and lists the metabolic causes by age at presentation: in the neonatal period and infancy, peroxisomal disorders, AK2 reticular dysgenesis, aminoacyl-tRNA synthetase deficiencies, PRPP synthetase deficiency and riboflavin transporter defects; in childhood, untreated biotinidase deficiency, mucopolysaccharidoses, alpha-mannosidosis, Kearns–Sayre syndrome and thiamine-responsive megaloblastic anaemia with diabetes and deafness; in adolescence, Fabry disease, adult Refsum disease and PRPS1 mutations [58].

6.3. Differential diagnosis of enlarged vestibular aqueduct and auditory neuropathy

EVA and cochlear hypoplasia are not specific to SLC26A4: they occur in CMV infection, in distal renal tubular acidosis with deafness (ATP6V0A4, ATP6V1B1, FOXI1), in branchio-oto-renal syndrome and, occasionally, in Treacher Collins syndrome [5]. Cochlear aplasia, Michel malformation, common cavity and isolated semicircular canal dysplasia do not belong to the SLC26A4 spectrum and do not justify its study [5]. The French PNDS adds to the differential diagnosis of hearing loss with EVA Kabuki syndrome, trisomy 21, deafness with renal tubular acidosis and branchio-oculo-facial syndrome, in which other clinical signs accompany the malformation; conversely, Wolfram and Alström syndromes combine hearing loss and thyroid disease without inner ear malformation and with many other manifestations [7]. No other known entity combines sensorineural hearing loss, bilateral EVA and goitre [7]. The PNDS considers neonatal goitre and congenital hypothyroidism exceptional in Pendred syndrome and requires a search for another cause in that situation, whereas the Oxford Desk Reference includes congenital hypothyroidism with goitre among the presentations to be considered; both converge in requesting temporal bone imaging in the child with hypothyroidism and hearing loss [7, 42]. In auditory neuropathy, the genetic differential diagnosis includes AIFM1 (DFNX5), ATP11A, DIAPH3, PJVK, ROR1, TMEM43 and OPA1; OTOF is found in 41–91% of children with ANSD tested and in up to 8% of congenital non-syndromic hearing loss [23]. Muckle-Wells syndrome combines progressive hearing loss and intermittent fevers, but fever does not cause hearing fluctuation, unlike the temperature-sensitive form of OTOF [23].

The French PNDS notes that the differential diagnosis of DFNB9 comprises the other causes of the auditory neuropathy spectrum, which share the audiological and electrophysiological features: acquired, or genetic syndromic, with involvement of other nerves (especially optic and peripheral), or non-syndromic; the isolated nature of the hearing loss, recessive transmission and the frequency of DFNB9 within this spectrum guide testing, and only molecular analysis confirms the diagnosis [20].

7. Treatments and Risk Management Strategies

7.1. General principles and communication goals

Childhood hearing loss management centres on the family's speech and language goals: some seek spoken language and mainstream education; others, particularly Deaf community families, prefer visual sign language communication or total communication using all available means (sign language, speech, gestures, facial expressions, images) [1]. Explicit discussion of these goals is therefore the first step [1]. The multidisciplinary team includes an ENT specialist, paediatric-experienced audiologist, speech and language therapist, geneticist, genetic counsellor and paediatrician, with a neuropsychologist, teacher of the deaf or other specialists as needed (for example, an ophthalmologist in Usher syndrome) [1]. Auditory deprivation in the first two years is associated with poorer reading, communication and speech production, deficits incompletely corrected by late intervention; early amplification or implantation is effective [1, 3]. Early sign language access provides an initial advantage even to children who later use spoken language [1]. Adult hearing loss is associated with isolation and increased dementia risk, potentially attenuated by treatment [1]. Hearing aid moulds require periodic replacement as the ear canal grows, and education should be adapted to the degree of loss [2].

7.2. Hearing aids and bone conduction devices

Hearing aids, fitted by the audiologist to the degree and frequencies of the loss, are indicated in mild to severe hearing loss [1]. They are the usual treatment of STRC-associated hearing loss and of mild to moderate GJB2 hearing loss, and patients with STRC are expected to continue to benefit from them without the need for an implant, given that inner hair cell function is preserved [3, 4]. In OTOF, hearing aids may be tried, but a systematic review of 32 patients found none with benefit in speech perception [23, 77]. In mitochondrial hearing loss with residual low-frequency hearing, combined electroacoustic stimulation is an option [8]. Implanted bone-conduction devices are used in some forms of conductive or unilateral hearing loss [1]. The two French PNDS (French National Diagnosis and Care Protocols) covering genetic forms in this article set different thresholds for hearing aids: the MELAS spectrum PNDS recommends them in any permanent bilateral hearing loss with a mean threshold above 30 dB, calculated according to the Bureau International d'Audiophonologie (BIAP), with a speech intelligibility threshold (discrimination of 50% of words) in quiet above 30 dB, or with significant degradation of intelligibility in noise, whereas the Pendred syndrome PNDS uses a mean threshold above 40 dB in children [7, 19]. Both accept hearing aids in mild hearing loss, unilateral or bilateral, depending on the impact, prefer hearing-aid specialists experienced in severe hearing loss and ask for use as regular as possible; the MELAS spectrum PNDS adds that fitting should restore stereophony, with bilateral hearing aids from the outset in bilateral hearing loss, and includes in the hearing-aid service selection and fitting, monitoring of efficacy, follow-up and hearing-aid education [7, 19]. No medicine is recommended to prevent or halt the worsening of m.3243A>G hearing loss [19].

7.3. Cochlear implantation

Cochlear implantation is considered for children with severe to profound loss from nine months; those implanted before two years may attain spoken language indistinguishable from hearing peers at school age [1]. The Oxford Desk Reference considers it established treatment in preschool children retaining a loss above 65 dB despite optimal amplification [2]. There is no upper age limit, but outcomes become less predictable with age [3]. Performance may be lower when the auditory nerve itself is affected [1]. Gene-specific considerations are summarised below.

Follow-up is initially frequent after implantation (speech recognition testing, equipment checks and programming, replacement and upgrades), with many tasks subsequently performed by families at home [3, 23].

French HAS criteria. The HAS good-practice sheet on cochlear and auditory brainstem implants, which summarises the 2007 assessment and its 2011 update on bilateral implantation in children (January 2012 version), reserves cochlear implantation for bilateral severe to profound sensorineural hearing loss, accepts unilateral (the most frequent) or bilateral implantation and always requires, before the decision, a hearing-aid trial under the best conditions [18]. In the child with prelingual hearing loss, implantation is performed as early as possible, because early implantation gives better and faster results in language comprehension and production; above 5 years of age, in profound or total non-progressive congenital hearing loss, there is an indication, save in particular cases, only if the child has developed an inclination for oral communication, and a child who already communicates orally may be implanted at any age, as may a young adult with congenital hearing loss [18]. The audiometric criteria in children are profound hearing loss in which the gain of the hearing aid does not allow language development, severe hearing loss with discrimination of 50% or less in age-appropriate speech audiometry tests performed at 60 dB in free field with well-fitted hearing aids, and, in fluctuating hearing loss, these criteria being met several times a month or fluctuations with an impact on language [18]. In adults there is no upper age limit, older people first undergo a psychocognitive assessment, prelingual hearing loss in adults is not, as a rule, an indication for a first implant, and the criterion is discrimination of 50% or less with the Fournier list or equivalent, under the same conditions, or fluctuation with a major impact on communication [18].

Bilateral implantation may be indicated when there is a risk of bilateral cochlear ossification in the short term (bacterial meningitis, bilateral petrous bone fracture), intervening before ossification advances; in the child with bilateral profound hearing loss who meets the criteria for unilateral implantation and in Usher syndrome; and in the adult with a unilateral implant who loses the benefit of the contralateral hearing aid with socio-occupational consequences or, in older people, loss of independence [18]. The aim of the second implant is to improve speech perception, especially in noise, and sound localisation [18]. The decision is taken by a multidisciplinary team that also provides the pre-implant assessment, surgery, programming and rehabilitation, possibly in a network, and the candidate (and the family, in the case of a child) should contact people already implanted, because motivation weighs in the decision; the activity is subject to requirements for technical facilities, team composition and minimum volume [18]. In more than 2000 patients analysed, the complication rate was below 5%; the most serious is bacterial meningitis, and the others (flap problems, electrode migration, facial paresis) are less serious [18]. Reimplantation for device failure adds no complications compared with the first implant and maintains performance [18]. In 2012, electroacoustic stimulation in the same ear and extension of indications to less severe hearing loss were still considered under evaluation [18].

The auditory brainstem implant, which stimulates the cochlear nucleus and not the fibres of the auditory nerve, is reserved for patients with complete hearing deficit in whom the cause precludes a cochlear implant: bilateral petrous bone fracture with ossification, major cochlear malformation, axonal neuropathy, bilateral labyrinthitis ossificans and posterior fossa tumours near the auditory nerve, including unilateral vestibular schwannoma with contralateral deafness and the bilateral vestibular schwannomas of neurofibromatosis type 2 [18]. Hearing restoration is limited (perception of environmental sounds and support for communication), depends on the number of electrodes activated and should be entrusted to experienced teams [18].

The MELAS spectrum PNDS applies these criteria to m.3243A>G hearing loss: it recommends early implantation when hearing loss progresses to severe to profound without sufficient benefit from the hearing aid, requires a multidisciplinary assessment (clinical, subjective and objective audiometry, vestibular function, imaging, speech and language therapy, psychology) in a specialised centre and accepts unilateral or bilateral implantation, simultaneous or sequential, according to hearing, vestibular function and age, with a contralateral hearing aid in unilateral implantation and a second implant if the other ear worsens [19]. In an implanted patient, contrast-enhanced brain CT is preferred in emergencies; MRI is not formally contraindicated when the clinician considers it necessary (for example, suspected true stroke), but should preferably be performed at 1.5 T, with a tight head bandage over the implant to limit displacement, and after confirming the compatibility of the model, even though many recent implants accept 3 T [19].

The French sources and the other guidelines of this article diverge on thresholds and on some indications; the table makes the differences explicit, and in bilateral hearing loss the HAS sheet is the French reference [1–3, 7, 18, 19].

The MELAS spectrum PNDS attributes to the 2012 HAS two criteria for unilateral hearing loss that do not appear in the January 2012 sheet, which is limited to bilateral hearing loss, and the PNDS gives no other source for them [18, 19]. Both PNDS write «below 50%» where the sheet writes «50% or less», and the MELAS spectrum PNDS replaces free field with the «binaural condition» [7, 18, 19]. As for vaccination, the sheet refers to meningococcal vaccination and the more recent Pendred syndrome PNDS to pneumococcal vaccination; the specific schedule follows the vaccination calendar in force for implant candidates [7, 18].

In DFNB9, the French PNDS always proposes a conventional hearing aid as first line, fitted from diagnosis by hearing-aid specialists with paediatric experience, with use as regular as possible [20]. Cochlear implantation follows the 2012 HAS criteria and is proposed to most children with early-diagnosed profound hearing loss whose families have chosen oral communication; in mild to moderate hearing loss it may be indicated when word intelligibility is poor, because of the pure-tone–speech discrepancy, and the indication is confirmed by a multidisciplinary assessment in a specialised centre (clinical, subjective and objective audiometry, vestibular function, imaging, speech and language therapy, psychology) [20] [18]. The implant may be unilateral or bilateral, simultaneous or sequential; in unilateral implantation, a contralateral hearing aid is encouraged and a second implant may be proposed later [20]. Results in DFNB9 are good in perception and expression and depend, as with any implant, on the age at implantation, associated disorders, speech and language therapy and family involvement; speech and language therapy begins with auditory rehabilitation and continues after activation of the implant [20].

7.4. Gene therapy for OTOF-related hearing loss

Lunsotogene parvec-cwha (Otarmeni®, Regeneron). This gene therapy uses dual adeno-associated virus serotype 1 (AAV1) vectors with a hair cell-specific promoter derived from Myo15 regulatory elements, carrying complementary DNA of human OTOF transcript variant 5 (otoferlin isoform e) [11]. Intracochlear infusion produces OTOF expression mainly in inner hair cells and restores synaptic transmission to the auditory nerve, demonstrated in mice [11]. The FDA granted accelerated approval on 23 April 2026 based on improvement in pure-tone average at week 24; continued approval may depend on confirmation of clinical benefit in a confirmatory trial [11, 12]. The European Medicines Agency medicines list consulted in October 2026 contains no European Union marketing authorisation [79]. According to the sponsor, EMA accepted the application for biallelic OTOF-related loss in May 2026 under accelerated assessment, and the sponsor expects a European Commission decision in the first quarter of 2027; acceptance for assessment is neither a CHMP opinion nor an authorisation [13, 80]. US approval was the first gene therapy approval under the FDA Commissioner's National Priority Voucher programme, and the sponsor announced free US supply, a corporate commitment rather than a label condition [81]. The programme has held EU orphan designation since 20 March 2023 (EU/3/23/2771), which does not imply demonstrated benefit [82].

Source: FDA label (BLA 125874, version of 27/04/2026) [11].

In the multicentre single-arm DB-OTO-001 study (CHORD; NCT05788536), 24 paediatric patients with a median age of 2 years (10 months to 16 years) received unilateral (n=10) or bilateral (n=14) treatment [11]. Among 20 assessed at week 24, 16 (80%; 95% CI 56–94%) met the primary endpoint of pure-tone average ≤70 dB HL and 14 (70%; 95% CI 46–88%) had click ABRs ≤90 dB nHL; 9 (45%) achieved ≤45 dB HL, allowing quiet speech to be heard, and 3 (15%) ≤25 dB HL, normal hearing [11]. Among 12 assessed at week 48, all nine responders maintained response and one additional patient responded (10/12, 83%); 8 (67%) reached ≤45 dB HL and 5 (42%) ≤25 dB HL [11]. One patient received rescue cochlear implantation in the treated ear approximately eight months later because of treatment failure [11]. The New England Journal of Medicine publication covering the first 12 participants described 75% responders at week 24, six children able to hear quiet speech without devices and three with normal average hearing; 67 adverse events occurred or worsened, none caused withdrawal and 17 were surgery-related [83]. Speech and language data are limited: a patient treated at ten months improved word recognition with the contralateral implant switched off; one treated at four years without previous spoken or sign language improved speech perception late, at week 72; and one treated at 28 months improved the Auditory Skills Checklist score from 8% to 60% at week 48 [23, 83].

Other gene therapy programmes. The first human AAV-OTOF administration through the round window membrane was performed in two children aged 5 and 8 years: the first recovered complete hearing in the treated ear within a month and the second improved without reaching normality, without serious adverse effects [36]. A unilateral AAV1-hOTOF single-arm trial and a bilateral paediatric trial followed [37, 84]. The expanded Chinese multicentre cohort included 42 patients aged 0.8–32.3 years, receiving one of three doses (9 × 10^11, 1.5 × 10^12 or 2.1 × 10^12 vg per ear); without dose-limiting toxicity or serious adverse events, hearing recovered in 38 of 42 patients (90%) and 43 of 48 ears (90%), with ABR thresholds improving from >97 ± 1 dB to 54 ± 3 dB at one year and 42 ± 5 dB at 2.5 years, and behavioural audiometry from >96 ± 3 dB to 37 ± 5 dB at 2.5 years [85]. At two years, seven of eight assessed participants reached auditory perception category ≥6 (conversation comprehension without lip-reading), and five of eight had speech intelligible to all listeners; an adult without a prior implant had only modest ABR improvement, suggesting that age, auditory deprivation duration and prior auditory experience influence functional outcomes [85]. Readministration was tested after amendment of the same Chinese trial (ChiCTR2200063181): four children aged 2.2–3.4 years, previously treated in one ear and with pre-existing neutralising antibodies (titres 1:135 to 1:3645), received AAV1-hOTOF in the contralateral ear; there was no dose-limiting toxicity at six weeks or serious adverse events (one grade 3 event, neutropenia), and mean second-ear ABR thresholds changed from >95 dB to 43, 63, 80 and 53 dB at 26 weeks; the authors consider these data preliminary [86].

AK-OTOF (AAVAnc80-hOTOF, Akouos/Eli Lilly) is in a recruiting phase 3 trial (NCT05821959; 48 planned participants; registry updated September 2026), without results published in an indexed journal at this review date [87]. SENS-501 (Sensorion), a dual AAV8 system, restored hearing in Otof−/− mice from three weeks, maintaining effect until ten months and showing good tolerance in non-human primates; the phase 1/2 AUDIOGENE trial (NCT06370351), in children aged 6–31 months, is active, not recruiting (12 planned participants), without published clinical results [87, 88]. DB-OTO-001 (NCT05788536) itself continues recruiting, with 36 planned participants and estimated primary completion in 2029 [87]. EH002 (Fudan University, NCT06722170, recruiting) and OTOV101N+OTOV101C (Otovia, NCT05901480, registry status unknown since 2024) are also registered [87]. AK-OTOF and SENS-501 hold European orphan designations (EU/3/21/2477 of 19 July 2021, and EU/3/22/2698 of 11 October 2022) [82]. An international consensus published in 2026 organises candidate selection, surgery, follow-up and rehabilitation [17]. Gene therapy for other genes remains preclinical: in STRC, dual AAV significantly restored cochlear amplification in 50% of treated mice, without human trials [4, 89].

The French PNDS of November 2022 states that there is no specific curative treatment for DFNB9; the statement reflects the date of the document, which predates the published results of the gene therapy trials and the US approval of April 2026, and the PNDS does not yet have a version that incorporates these therapies [20] [11].

7.5. Pendred syndrome: hearing and thyroid management

There is no curative treatment [5]. Hearing habilitation follows general principles, with hearing aids for mild to severe loss and implantation when hearing aid benefit is limited, reassessing implant candidacy at every visit to avoid delay [5, 78]. Goitre and thyroid dysfunction are managed with endocrinology [5]. Reports of hearing deterioration after head trauma lead some professionals to discourage weightlifting and contact sports, but evidence does not show that avoidance reduces overall progression; families should be informed and decide [5]. The Oxford Desk Reference notes that iodine supplementation is preferable to hormone replacement in sodium–iodide symporter deficiency (SLC5A5), a recommendation specific to that dyshormonogenesis form that does not extend to Pendred syndrome [42].

The HAS French PNDS (2021) is, among the sources consulted, the only national guideline dedicated to Pendred syndrome, and its recommendations structure the management described below; the document stresses that there is no specific treatment and that gene therapy remains in the research domain [7].

Auditory habilitation. Conventional hearing aids are used in mild to severe hearing loss, unilateral in unilateral hearing loss and bilateral from the outset in bilateral hearing loss; they are indicated in any child with permanent bilateral hearing loss with a mean threshold above 40 dB and are proposed in mild hearing loss according to the impact, with a hearing-aid specialist experienced with children and severe hearing loss [7]. Cochlear implantation follows the 2012 HAS criteria: bilateral profound hearing loss; severe hearing loss with discrimination below 50% at 60 dB with a well-fitted hearing aid, or insufficient gain for language development; and fluctuating hearing loss with a marked impact on language or communication [7]. The indication results from a multidisciplinary assessment in a specialised centre (clinical, subjective and objective audiometry, vestibular function, imaging, speech and language therapy, psychology); the implant may be unilateral or bilateral, simultaneous or sequential, according to hearing, vestibular function and age (before or after walking), maintaining the contralateral hearing aid in unilateral implantation and proposing the second implant when the contralateral ear worsens [7]. The surgical particularities of the malformation are perilymph leakage, usually controllable, malpositioning or partial insertion of the electrode array when perilymphatic pressure is high, more difficult sealing of the cochleostomy and a greater risk of postoperative vertigo [7]. An audiogram with abrupt bilateral worsening without recovery after a few weeks may justify discussion of urgent implantation [7].

Perioperative implant protocol. To reduce the risks related to the malformation, the PNDS proposes, depending on the case, electrolytes and confirmation of hydration, 30° head-up position, controlled hypotension with normocapnia (30–35 mmHg), hypertonic saline in children, intravenous acetazolamide if necessary, prednisolone and acetyl-leucine, with the doses in the table below [7].

Prevention of fluctuations and sudden hearing loss. The PNDS advises avoiding sports with a risk of head trauma (contact sports) and activities with a risk of barotrauma (diving, parachuting, wind instruments, carrying heavy loads, Valsalva manoeuvres, abdominal straining), and asks for attention to any fall in hearing during fever or upper respiratory tract infection, which warrants urgent examination by an otorhinolaryngologist with an audiogram [7]. Sudden hearing loss, unilateral or bilateral, is treated as a functional emergency, with corticosteroid therapy started immediately at diagnosis; the regimen is not specific to SLC26A4 and combines prednisolone and, in adolescents and adults, mannitol, with the doses in the table; transtympanic corticosteroid injections, not specifically evaluated in these patients, may be proposed under local anaesthesia in adolescents and adults [7]. The US guideline on sudden sensorineural hearing loss in adults, the reference for the dose, proposes prednisone 1 mg/kg/day up to a maximum of 60 mg/day for 7–14 days, followed by a taper, accepts intratympanic corticosteroid therapy as initial or salvage treatment and advises against the routine use of antivirals, thrombolytics, vasodilators, vasoactive substances and antioxidants; the mannitol of the PNDS is not among the treatments it proposes [90].

Speech and language therapy and balance. In prelingual hearing loss, speech and language therapy begins with auditory habilitation and aims at oral language; in severe to profound hearing loss of postlingual course, it works on lip-reading and the cognitive and executive skills of verbal processing; after implantation, it continues after activation [7]. Acute attacks of vertigo are treated with an antivertiginous agent (acetyl-leucine) and an antiemetic (metoclopramide, with caution and as second line in children over one year), vestibular rehabilitation serves chronic symptoms in adults and older children, psychomotor therapy the psychomotor delay of the young child, and no background treatment is recommended [7]. In the European label, metoclopramide is contraindicated below one year and, in children, is authorised only as second line for postoperative and delayed chemotherapy-induced nausea and vomiting, so its use in childhood vertigo is an off-label extension [91].

Thyroid. Overt hypothyroidism (raised TSH with low free T4) is treated with oral levothyroxine from diagnosis and for life, with a TSH target adapted to age and comorbidities; liothyronine, alone or in combination, has no demonstrated benefit [7]. In subclinical hypothyroidism (raised TSH with normal free T4), whose treatment has not been evaluated in this syndrome, the PNDS proposes levothyroxine in children with TSH above 10 mIU/L or with goitre, and in adults with TSH above 10 mIU/L associated with antiperoxidase antibodies, compatible symptoms, a risk background or goitre, so that TSH does not stimulate growth of the gland [7]. Without compression or nodules, goitre is monitored clinically and by ultrasound; thyroidectomy should be avoided, especially in children, and is decided and performed in an experienced centre after multidisciplinary discussion: total in plunging or compressive goitre (dyspnoea, dysphagia, dysphonia, superior vena cava syndrome), total or partial for a nodule suspicious on ultrasound and cytology [7]. Iodine-131 for volume reduction and levothyroxine at suppressive doses have not been evaluated in the syndrome and are not recommended [7]. Nodules follow the recommendations for the general population [7].

Vaccination, therapeutic education and pregnancy. The inner ear malformation increases the risk of bacterial meningitis, so the PNDS recommends pneumococcal vaccination with the 13-valent conjugate vaccine (Prevenar 13®) and, from six years of age, its combination with the 23-valent polysaccharide vaccine (Pneumovax®), referring the schedule to the national vaccination calendar [7]. This sequential schedule has been abandoned in France. In 2023, HAS recommended a single dose of the 20-valent conjugate vaccine (Prevenar 20®) in adults at increased risk of pneumococcal infection, instead of the 13-valent followed by the 23-valent sequence, and in July 2025 it accepted the 21-valent conjugate vaccine (Capvaxive®) under the same conditions [92]. The recommendation validated on 23 July 2026 extended the 20-valent vaccine to the paediatric age group: a 3+1 schedule in infants (2, 4 and 6 months and a booster at 11 months) and a single dose in at-risk children and adolescents from 15 months to 18 years, at least 8 weeks after the last pneumococcal vaccine, with the 13-valent, 15-valent and 23-valent vaccines no longer used; in at-risk adults aged 18 to 64 years and in all adults aged 65 years or over, the 21-valent vaccine became preferred, as a single dose [93]. The CDC recommends, in never-vaccinated adults with a cochlear implant, one dose of 15-, 20- or 21-valent conjugate vaccine, followed by the 23-valent only when the 15-valent is used; from 6 to 18 years, the 15- or 20-valent; below 5 years, the age-appropriate schedule; and all doses at least two weeks before implant surgery [94]. Therapeutic education is adapted to the degree of hearing loss: in mild to moderate or unilateral hearing loss, it emphasises the risk of worsening, prompt seeking of care and avoidance of risk situations; in profound hearing loss with an implant, maintenance of the equipment, permanent use, speech and language therapy and follow-up; and, in all, information on meningitis [7]. Pregnancy has no thyroid contraindication; in women with pre-existing hypothyroidism, the levothyroxine dose is increased by 30–50% as soon as pregnancy is confirmed, with monthly TSH throughout gestation, and even without pre-existing hypothyroidism TSH is measured monthly; goitre is monitored, because it may enlarge, and there are no data on worsening of hearing loss during pregnancy or postpartum [7].

Divergences between the French PNDS and the GeneReviews review. The two sources agree on biallelic molecular diagnosis, on the absence of useful genotype–phenotype correlation and on stepwise auditory habilitation, but diverge on several management points. The guideline is the reference and the differences are made explicit in the table [5, 7].

The levothyroxine doses of the PNDS are initial doses and differ from the US label mainly in children, in whom the label proposes higher per-kilogram doses stepped by age; for the dose increase in pregnancy, the PNDS uses an immediate percentage increase and the label a fixed increase conditioned by TSH [7, 95]. In both cases, titration is guided by TSH.

7.6. Prevention of aminoglycoside ototoxicity

The Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline, published online in 2021, classifies MT-RNR1 phenotypes into three groups and provides strong recommendations [10].

Source: [10].

The guideline defines no heteroplasmy threshold below which administration is safe and applies to all aminoglycosides and intravenous, intramuscular and intraperitoneal routes; inhaled administration and aminoglycoside-containing orthopaedic cement produce relevant systemic exposure and may warrant avoidance, whereas enteric and topical routes (bioavailability <1%) are unlikely to cause hearing loss [10]. It distinguishes two settings: when an aminoglycoside is chosen for cost or preference (for example, infection susceptible to other antibiotics), it should not be used; when potentially indispensable (severe adult or neonatal sepsis, infection with risk of multidrug-resistant Gram-negative bacteria, tularaemia, plague, brucellosis, enterococcal or streptococcal endocarditis), specialist advice should be sought, the patient informed and the shortest possible course used [10].

The UK CERSI-PGx guidance, published in 2026, estimates the combined frequency of m.1555A>G, m.1494C>T and m.1095T>C at approximately 1 in 330 people, recommends avoiding aminoglycosides at any detectable variant level, states that laboratory testing in England is nationally funded and that approximately 20% of aminoglycoside use is predictable, enabling pre-emptive testing; if results are unavailable and urgency is high, treatment should not be delayed [16]. In PALOH, performed in two UK neonatal units in 2020, 751 newborns were recruited, 424 of those receiving antibiotics (80.6%) were successfully tested, three m.1555A>G carriers avoided aminoglycosides, and antibiotic timing was not delayed [72]. Preliminary economic assessment suggests that neonatal point-of-care testing may save costs [16]. In the United States, injectable aminoglycoside labels now warn about ototoxicity risk in carriers of MT-RNR1 variants, particularly m.1555A>G, even with serum concentrations within recommended ranges, and advise considering alternatives with a known variant or maternal ototoxicity history unless infection severity and lack of safe alternatives justify use; labels state that these variants occur in fewer than 1% of the US population and that the proportion of carriers developing ototoxicity is unknown [96]. The label warning does not mandate prior testing; recommending testing before predictable exposures comes from UK guidance rather than a regulatory decision [16, 96].

In pregnant carriers of m.1555A>G or m.1494C>T, aminoglycosides should be considered only without an alternative, and the newborn, who will inherit the variant, should avoid them [8]. Targeted testing of maternal relatives identifies those who should avoid these antibiotics [8]. Noise should be avoided, particularly by carriers with normal hearing [8].

7.7. Agents and circumstances to avoid

Noise is an avoidable cause of hearing loss; no safe level has been established, but OSHA and NIOSH have set a limit of 85 dB over eight hours of work, and the headphone volume limit of the mobile phone can be programmed [1]. In mild to moderate GJB2 hearing loss, avoidance of repeated exposure above 75 dB is recommended [3]; in STRC, of prolonged exposure above 85 dB [4]. In the temperature-sensitive form of OTOF, febrile episodes should be prevented and treated promptly and exercise or environments that raise body temperature avoided; a fall in hearing may be the first sign of an infection [23]. In POU3F4, fenestration or removal of the stapes footplate may cause perilymphatic gusher and total deafness [1]. In the SLC26A4 spectrum, the French PNDS (French National Diagnosis and Care Protocol) contraindicates endolymphatic sac surgery, because of the risk of postoperative hearing deterioration, and advises avoiding contact sports and barotrauma situations, a more restrictive position than that of the GeneReviews review [5, 7]. Relatives of children with permanent hearing loss should have a complete hearing assessment regardless of the newborn screening result [1].

7.8. Surveillance

7.9. Targeted therapies in syndromic differential diagnoses

Some forms presenting as isolated hearing loss have specific treatments: biotin for biotinidase deficiency, high-dose riboflavin supplementation for riboflavin transporter deficiency, calcium and calcitriol or parathyroid hormone for HDR syndrome, and interleukin-1 inhibitors (anakinra, rilonacept, canakinumab) for NLRP3-associated disorders [1].

7.10. Regulatory status, trials and evidence hierarchy

This information has different levels of evidential strength, distinguished in the text: peer-reviewed publications (Chinese trials, CHORD in the New England Journal of Medicine), the approved label (summarising regulatory data analysis), sponsor announcements and reports (European application acceptance, expected decision date, free access), orphan designations and trial registries. Orphan designation or application acceptance does not demonstrate benefit; US authorisation implies neither European authorisation nor reimbursement or availability in Portugal or Brazil, for which no INFARMED, ANVISA or CONITEC lunsotogene parvec decision was identified in the consulted sources.

These studies cannot rank the products. All are single-arm, implicitly compared with the natural history of stable profound loss, reasonable for thresholds but not language outcomes, which depend on age, auditory experience and rehabilitation. Endpoints differ: the US label uses the proportion with behavioural pure-tone average ≤70 dB HL, whereas Chinese trials report mean (± standard error) ABR thresholds and define “recovery” differently; a 42 dB group mean does not mean every patient reaches that threshold, and response proportion does not describe hearing in noise. Denominators change during follow-up (20 evaluable at week 24, 12 at week 48), ages and doses differ, maximum duration is 2.5 years, no controlled trials or negative published results exist, and at least one treatment failure required rescue implantation [11, 83, 85]. Selective publication of favourable results in small series cannot be excluded.

Evidence extrapolation should be distinguished from off-label use. In the United States, the indication covers paediatric and adult patients, but label evidence derives from 24 paediatric patients aged 10 months–16 years, and the medicine has not been studied from age 65 years; treating an adult is therefore on-label use supported by extrapolation. Conversely, an already implanted ear, mild to moderate baseline loss (as in the temperature-sensitive form) and a patient with only one identified variant or a variant of uncertain significance fall outside the approved indication, because the label requires confirmation of two pathogenic or likely pathogenic variants [11]. There are no pregnancy data [11]. Outside the United States, administration outside clinical trials depends on national early-access rules. For aminoglycosides, avoidance in m.1555A>G carriers is a guideline recommendation; unavoidable use remains within approved indications and should be documented with CPIC-recommended risk assessment [10].

8. Reproductive Options and Family Counselling

8.1. Recurrence risks with an established molecular diagnosis

In AR hearing loss, each sibling of a proband with heterozygous parents has a 25% chance of being affected, 50% of being a carrier and 25% of inheriting neither variant; carriers have no increased hearing loss risk [1]. If only one parent is a carrier, a de novo variant or parental mosaicism should be considered; if the proband appears homozygous, an undetected deletion or uniparental isodisomy [1]. Children of an affected individual are obligate carriers unless the partner is a carrier or affected; in populations with high carrier frequency, pseudodominant inheritance is possible and recurrence should be calculated after partner testing [3]. If both parents have GJB2-related hearing loss, all children will be deaf; if their recessive loss involves different genes, risk is below the general population's [1]. GJB2 carrier frequency in Europe, North America and the Mediterranean is approximately 1/50; carrier testing is possible in relatives when proband variants are known and in the population for c.35delG [2, 97]. Extensive allelic heterogeneity requires full GJB2 sequencing in the partner [3].

In AD hearing loss, each child has a 50% risk; if the variant is absent in the parents, sibling risk is approximately 1% because of germline mosaicism [1]. Variable expressivity may make family history apparently negative [1]. In X-linked loss, men transmit the variant to all daughters and no sons; carrier women transmit it to 50% of children, and carrier daughters may or may not have hearing loss [1].

In mitochondrial hearing loss, the father does not transmit the variant; all children of a carrier woman inherit it, but the risk of hearing loss depends on the mutational load and on exposure to aminoglycosides [8]. Up to 85% of patients have no known family history [8]. m.1555A>G is homoplasmic in almost all families [2, 60]. When hearing loss is due to m.3243A>G (MIDD/MELAS, differential diagnosis in section 6.2), the Oxford Desk Reference states that the most common presenting manifestations are diabetes and progressive sensorineural hearing loss, sometimes without other symptoms, that approximately 10% of patients develop cardiomyopathy, that the mutational load in blood poorly reflects that of tissues and may decrease with age, and that all members of the maternal line have an increased risk of diabetes (usually in early adulthood, initially controlled with diet and, as a rule, insulin-dependent soon afterwards), progressive sensorineural hearing loss and cardiomyopathy. The estimated risk to offspring is approximately 15–25% with a low load (<30% in muscle, <10% in blood), in asymptomatic carriers or those with diabetes only, and greater than 60% with a moderate load (>50% in muscle) or with symptoms beyond diabetes and/or hearing loss [60]. The children of affected men are not at risk [60]. When the mother is heteroplasmic, the proportion of mutant mtDNA transmitted to the children may vary greatly [59].

Counselling men with biallelic contiguous STRC-CATSPER2 deletions should include infertility risk and discussion of assisted reproductive techniques [4].

In Pendred syndrome and DFNB4, the parents of an affected child have a 25% risk in each pregnancy, regardless of sex [7]. An affected adult has a low risk of having affected children, unless the partner is affected by the same disease or there is consanguinity; as the PNDS estimates the frequency of heterozygotes in the general population at below 1%, the risk is also low for healthy relatives with unrelated partners, but heterozygote testing may be offered in the family and to the patient's partner [7].

For DFNB9, the French PNDS adds that an affected adult has a very low risk of having affected children, except if the partner has the same disease or there is consanguinity, and that, as the frequency of heterozygotes in the general population is below 1%, the risk for a healthy relative is low if the couple is not related; carrier testing in relatives is discussed case by case [20].

8.2. Empirical risks without an aetiological diagnosis

Sources: [1, 2, 98].

8.3. Prenatal and preimplantation diagnosis

Once familial variants are identified, prenatal diagnosis (for example, CVS) and preimplantation genetic testing are technically possible; the Oxford Desk Reference describes prenatal diagnosis of hearing loss as highly controversial, and GeneReviews recommends discussing differing perspectives and treating the decision as personal [1, 2]. Ideally this discussion occurs before pregnancy, and counselling should be offered to young adults with hearing loss [1]. Mitochondrial prenatal diagnosis is possible but rarely performed: with a homoplasmic mother, the fetus necessarily inherits the variant; mutation burden in villi or amniocytes does not correspond to other tissues, and variant presence predicts neither occurrence, onset age nor severity [8].

In France, prenatal or preimplantation diagnosis of Pendred syndrome is requested from an accredited multidisciplinary prenatal diagnosis centre, which assesses eligibility under article L2131-4 of the Public Health Code and the feasibility of the request, with prior consultation of a geneticist from the reference centre network [7]. The rare situation of a fetus at known risk with suggestive ultrasound signs involves consultation of the parents with a geneticist and a paediatric endocrinologist, an ultrasound search for goitre and associated malformations, information on the possible need for intra-amniotic levothyroxine if the goitre is compressive, a plan for postnatal ENT and endocrinological follow-up, molecular confirmation and psychological support [7].

For DFNB9, the French PNDS accepts prenatal diagnosis only when the molecular diagnosis has been established in an affected relative: the couple is seen by a geneticist and a genetic counsellor, who establish the risk; the request is assessed case by case by a multidisciplinary prenatal diagnosis centre (CPDPN) with a physician expert in the disease, who must attest to the particular severity and incurability of the condition, and the couple is informed of the possibility of termination of pregnancy for medical reasons if the fetus is affected; testing for the variants is performed on chorionic villi or amniotic fluid, depending on gestational age, with psychological support for the couple [20]. Requests for preimplantation genetic testing are also examined case by case [20]. A woman with DFNB9 has no contraindication to pregnancy, and there are no data on worsening of hearing loss during pregnancy or postpartum [20].

8.4. Communication and terminology

Communication with deaf sign language users requires a certified interpreter; visual aids and signalling topic changes facilitate consultation [1]. Neutral terminology is recommended: “chance” rather than “risk”, “deaf” or “hearing” rather than “affected” or “unaffected”, and “deaf” or “with hearing loss” rather than “hearing impaired” [1, 8]. Many deaf people mainly seek information on their hearing loss cause and medical, educational and social services rather than prevention or family planning [8]. Organisations such as the National Deaf Children's Society provide useful support [2]. The relevant European Reference Network is ERN CRANIO, including rare ENT disorders among its four disease groups; Usher syndrome in the differential is covered by ERN EYE [24, 99]. The European Union of the Deaf represents national deaf organisations and EURO-CIU cochlear implant users; Portugal's Federação Portuguesa das Associações de Surdos brings together local associations [25–27]. Referral should respect the family's choice of spoken communication, sign language or both [1].

The French PNDS recommends that disclosure of the diagnosis of Pendred syndrome take place at a dedicated consultation, in a specialised centre or with professionals linked to it, conducted by the geneticist or the physician who requested the analysis, preferably with a psychologist; the characteristics of the syndrome, management, follow-up, transmission and counselling are explained, and the contact details of associations and a written document, also sent to the paediatrician or general practitioner, are provided [7]. The team includes paediatrician or general practitioner, otorhinolaryngologist, endocrinologist or paediatric endocrinologist, geneticist, genetic counsellor, radiologist, nuclear medicine physician, surgeon, hearing-aid specialist, speech and language therapists specialised in deafness, implant programming technicians, vestibular rehabilitation physiotherapists, psychomotor therapist, nurse, psychologist and social worker [7].

In France, the syndrome is managed by the reference centre for genetic deafness (coordinated at the Hôpital Necker-Enfants Malades, with constituent centres at the Pitié-Salpêtrière and in Lille and a network of competence centres), in liaison with the reference centres for rare endocrine diseases of growth and development and for rare diseases of the thyroid and hormone receptors (coordinated in Angers), within the SENSGENE and FIRENDO filières [7]. There is no patient association dedicated to the syndrome; the PNDS refers to associations of deaf people and of cochlear implant users, to the French thyroid patient associations (Vivre sans thyroïde, Association française des malades de la thyroïde) and to the patient representatives of Endo-ERN [7]. Treatments and rehabilitation may be covered at 100% by health insurance as a long-term condition outside the list; hearing aids are reimbursed at 100% up to 20 years of age, with two annual maintenance services and reimbursement of earmoulds, and after that age the Maison départementale des personnes handicapées may fund them through the disability compensation benefit [7]. At school, the choice between bilingual education (French Sign Language and written French) and education in French, with or without cued speech, is the parents' right; the teacher should be aware of the hearing loss and the possibility of fluctuation and seat the child at the front and on the side of the better ear, and technical and human aids (high-frequency microphone, support assistant, coders, interpreters) are formalised in a personalised schooling plan, in mainstream education with or without the support of specialised services or in specialised teaching units [7].

In DFNB9, the French PNDS provides for disclosure of the diagnosis at a dedicated consultation, in a specialised centre or in liaison with it, by the clinical geneticist who prescribed the test, with the recommended presence of a psychologist; the characteristics of the hearing loss, rehabilitation, follow-up, transmission and genetic counselling are explained, according to age, a summary document is given to the parents and the patient, also sent to the paediatrician or general practitioner, and the contact details of associations are provided [20]. There is no association specific to DFNB9; the PNDS refers to associations of parents of deaf children and of cochlear implant users (Surdi.info, SENSGENE, UNANIMES, Génération Cochlée, CISIC) [20]. On the medico-social level, treatments and rehabilitation may be covered at 100% by Social Security, as may conventional hearing aids up to 20 years; the Maison départementale des personnes handicapées (MDPH) assesses compensation, the education allowance for a disabled child (AEEH) covers expenses up to 20 years, the family support and early education services (SAFEP) support deaf children before 3 years and the family education and schooling support services (SSEFS) from that age, and schooling, mainstream or in a specialised centre, is organised in a personalised plan with the support of an assistant or a cued speech (LPC) coder; after 20 years, the disability compensation benefit may fund hearing aids, and in adults recognition of disabled worker status (RQTH) may be requested [20].

9. Special Considerations: Population Diversity and Epidemiology

9.1. GJB2 and the DFNB1 locus

GJB2 variants may explain up to 50% of prelingual non-syndromic AR hearing loss and 10–40% of sporadic cases in European populations [2]. The allelic spectrum is strongly dependent on ancestry: c.35delG accounts for approximately 70% of variants in European, North American and Mediterranean populations, with a carrier frequency of 2–4% in northern Europe; c.167delT accounts for 3–4% in Ashkenazi Jews; p.Val37Ile, a mild variant, is very common in Southeast Asia, where the East Asian population has the highest carrier frequency described (8.3%); c.235delC is the most frequent in Japan; and p.Trp77Ter, p.Trp24Ter and p.Gln124Ter predominate in the Indian subcontinent [2, 3]. Biallelic GJB2 variants contribute practically nothing to genetic hearing loss in sub-Saharan African populations, where the genetic causes remain to be investigated in most countries [1, 3]. The GJB6 deletion is the second most common DFNB1 alteration in Mediterranean populations [2, 35]. In addition to the carrier frequency of 1/50 given in section 8.1, the Oxford Desk Reference reproduces the empirical DFNB1 carrier frequencies from a US laboratory screen of 23,453 people (Lazarin et al., 2013), expressed as 1 in N (95% CI): 1 in 42.6 (39–47) across all populations; 1 in 46.0 (40–54) in Northwestern Europe; 1 in 37.2 (26–67) in Southern Europe; 1 in 21.3 (17–28) in Ashkenazi Jews; 1 in 22.2 (17–34) in East Asia; 1 in 67.6 (48–114) in Hispanics; 1 in 77.6 (40–1627) in the Middle East; 1 in 99.9 (58–372) in South Asia; and 1 in 226.0 (99–∞) in African Americans. The number of people tested for DFNB1 ranged from 388 (Middle East) to 8,735 (Northwestern Europe), which explains the width of the intervals in the smaller groups; at the time, neither ACMG nor ACOG recommended offering DFNB1 carrier screening to any of these populations [97].

Portugal. In a random sample of 502 children from the Geração XXI birth cohort (Porto), carrier frequencies for c.35delG and p.Met34Thr were each 1 in 50 (10 heterozygotes for each and one c.35delG homozygote); among 89 patients with nonsyndromic loss, 7 were c.35delG homozygotes and 5 heterozygotes [100]. In a subsequent study of 147 patients and 360 participants from the same cohort, less common variants represented 12% of patients, p.Val37Ile and p.Val95Met were more frequent in patients, and p.Arg127His and p.Val153Ile behaved as polymorphisms [101]. The consulted sources provided no national Portuguese newborn hearing screening coverage data or population m.1555A>G frequency; this absence reflects source scope and does not establish that such data do not exist.

Brazil and Latin America. Most Latin American studies investigated only GJB2/GJB6, and the hearing loss population remains underdiagnosed [102]. In 90 Brazilian families with presumed AR loss, after excluding c.35delG, del(GJB6-D13S1830), del(GJB6-D13S1854) and m.1555A>G, next-generation sequencing identified the cause in 36.7%, involving 24 genes with the greatest contribution from GJB2, CDH23, MYO15A, OTOF and USH2A [103]. In 99 Brazilian probands previously negative for DFNB1 and m.1555A>G, a 218-gene panel yielded 43–62%, with MYO7A and MYO15A most frequent and 19 novel variants [67]. m.1555A>G was absent from 8974 Brazilian and 1042 Argentine newborns [8].

9.2. OTOF

Biallelic OTOF variants explain 1–2% of congenital AR loss in Japan, where p.Arg1939Gln is present in 56–87.5% of patients; 2–3% in Pakistan and a US laboratory cohort; approximately 3% in Taiwan, where p.Glu1700Gln occurred in 23% of progressive ANSD patients; and 3–8% in Spain, where p.Gln829Ter occurred in 87% of patients with ANSD and two variants [23]. In Korea, 56.5% of OTOF loss is due to p.Arg1939Gln [23]. An Israeli multicentre study identified intronic c.898-18G>A in seven Jewish families of Moroccan origin with sensorineural loss or ANSD; the variant is absent from gnomAD and other studied Jewish groups and has a 1.5% carrier frequency in this ancestry, suggesting a founder effect relevant to carrier screening and gene therapy candidate selection [74]. Temperature-sensitive loss prevalence is unknown [23].

The French PNDS estimates the prevalence of DFNB9 at 3–5/100,000 and the frequency of heterozygotes in the general population at less than 1%; it does not present its own French data [20].

9.3. SLC26A4 and Pendred syndrome

The biallelic genotype (M2) explains 67–90% of patients with EVA of Chinese, Korean and Japanese ancestry and only approximately 25% of Northern European patients, half of whom have no identifiable variant [5]. Thyroid expression depends on iodine intake, goitre being more frequent in countries without universal salt iodisation [5]. Orphanet gives for Pendred syndrome a point prevalence in Europe of 1–9/100,000 (mean value 7/100,000), which does not include DFNB4 [6]. The French PNDS uses the same prevalence class and estimates the frequency of heterozygotes for SLC26A4 variants in the general population at below 1%, without detailing the population of origin; the French series on which the document is based (SLC26A4 screening in French families with Pendred syndrome and thyroid phenotype of 42 patients with Pendred syndrome or non-syndromic EVA) are cited without frequency estimates [6, 7].

9.4. Mitochondrial variants

m.1555A>G has been observed worldwide (Arab-Israeli, Japanese, Mongolian, Zairian, Spanish, Chinese, Turkish, Balinese, Moroccan, Greek, Polish, Tunisian and North American populations) [8]. The combined frequency of the three MT-RNR1 risk variants is approximately 1 in 330 [16]. In one Chinese province, hearing and genetic screening of 58,000 newborns identified mitochondrial variants in 1.8%, but only one had hearing loss [8]. Saudubray reports m.1555A>G, like low-burden m.3243A>G, in approximately 0.2% of Europeans [76]. MT-TS1 variants occur in 0.8–1.1% of deaf probands in the United States and Mongolia and 0.68% in China [8].

9.5. Other genes

STRC carrier frequency is 1.36% in a meta-analysis of normal-hearing individuals and 1.8% in an ethnically diverse paediatric cohort; STRC explained 4.08% of loss without a GJB2 cause and 14.36% of mild to moderate loss [4, 104]. Orphanet records deafness–infertility syndrome with prevalence <1/1,000,000 and three reported families, reflecting the rarity of formally recorded fertility phenotype rather than genotype frequency [4, 6]. p.Asp543Asn in TMC1 has a founder effect in 11 Japanese AD families [45], and p.Pro51Ser is a COCH founder variant [54]. Most African and many Asian and Latin American regions lack systematic population data on hearing loss gene distribution [3, 102].

10. Illustrative Clinical Case

10.1. Temperature-sensitive OTOF-related hearing loss (published case)

Forli and colleagues described a term-born boy assessed at seven years in Pisa who had passed bilateral newborn OAE screening; family hearing loss history was uncertain and both parents were hearing [73]. The first transient loss episode occurred at two years, followed by episodes after exertion or fever without vestibular symptoms. At seven, language was age-appropriate, but he struggled with television and noisy environments. Between episodes he had mild to moderate low- and mid-frequency sensorineural loss (500–1000 Hz), normal tympanometry and absent stapedial reflexes; OAEs were present, click and tone-burst ABRs absent on the left and without an identifiable wave V on the right up to 95 dB SPL, and word recognition was 100% in quiet and 70% at a 0 dB signal-to-noise ratio. At eight, during fever of 38.7 °C, thresholds worsened to moderate to severe loss and word recognition fell to 60% in quiet and 10% at a 0 dB signal-to-noise ratio, recovering within hours after temperature normalisation; CT and MRI showed no malformations [73]. A 37-gene panel identified the paternal pathogenic OTOF (NM_194248.2) missense variant c.2521G>A, p.(Glu841Lys), and IGV coverage inspection showed halved coverage of exons 10–14, corresponding to a heterozygous 7.4 kb deletion, c.(897+1_898−1)_(1579+1_1580−1)del, confirmed by array comparative genomic hybridisation; parental deletion testing was unavailable, so trans phase was inferred rather than demonstrated [73]. The panel also found, heterozygously and unrelated to the phenotype, a likely pathogenic GJB2 variant (c.−22−2A>C) and a TECTA VUS, both inherited from the asymptomatic father [73]. Original figures show afebrile audiograms (figure 1), febrile audiograms (figure 2) and deletion detection through read coverage (figure 3), available openly at https://pmc.ncbi.nlm.nih.gov/articles/PMC9962730/ [73]. The patient began using hearing aids and a wireless school system with reported speech discrimination improvement, and the family was advised to treat fever promptly and avoid intense exercise [73]. Four points are illustrated: OAE/ABR dissociation defines synaptopathy; the temperature-sensitive form requires a non-truncating allele and escapes OAE newborn screening; structural OTOF variants require CNV detection methods, here revealed by panel coverage; and a heterozygous finding in another recessive gene such as GJB2 does not explain the phenotype [23, 73]. Families should know that sudden hearing deterioration may be the first infection sign [23]. With mild to moderate baseline loss, this patient does not meet the approved gene therapy indication requiring >90 dB HL [11].

10.2. Intrafamilial variability in the SLC26A4 spectrum (published case)

Chen and colleagues described four family members, all with sensorineural loss and homozygous for the SLC26A4 splicing variant c.919-2A>G [105]. Manifestations, particularly inner ear morphology, varied, and one had normal cochleovestibular structures, a rare published finding. The family illustrates the need to investigate siblings after an index diagnosis even with normal imaging, and maintain long-term hearing and thyroid surveillance [5, 105]. Another Japanese child with EVA diagnosed at four years and two SLC26A4 variants had a mildly enlarged thyroid, normal function and positive antiperoxidase antibodies, prompting a diagnosis of probable Pendred syndrome with Hashimoto thyroiditis, both carrying hypothyroidism risk [106].

10.3. Newborn carrying m.1555A>G (composite case based on PALOH data)

A term newborn admitted to neonatal intensive care for suspected early-onset sepsis undergoes rapid m.1555A>G testing at admission, returning in fewer than 30 minutes before the first antibiotic dose [72]. The result is positive. Following CPIC and the 2026 UK guidance, the team replaces the aminoglycoside with an equally effective alternative empirical regimen, documents the genotype and alerts the family [10, 16]. Newborn hearing screening is normal. The mother, also a carrier (the variant is usually homoplasmic), is informed that siblings, maternal grandmother and maternal aunts and uncles may be carriers and benefit from targeted testing; the father and children of carrier men are not at risk [8]. The child continues annual audiometry because the variant may cause loss without exposure, and the family is advised to avoid excessive noise and inform all prescribers, including in future pregnancies [8]. In PALOH, all three carriers identified among 751 newborns avoided aminoglycosides without antibiotic initiation delay [72].

11. Prognosis and Long-Term Follow-up

Apart from the hearing loss, people with GJB2-associated hearing loss are healthy and have a normal life expectancy; with early cochlear implantation, children with severe to profound hearing loss can achieve social and educational functioning indistinguishable from that of their hearing peers [3]. STRC-associated hearing loss usually remains moderate or better, with progression of approximately 0.6 dB per year and good response to hearing aids [4]. In typical OTOF-associated hearing loss, cochlear implantation improves speech perception in the great majority, and gene therapy has opened an alternative whose long-term durability is not yet known: the longest data reach 2.5 years of follow-up in the Chinese cohort and week 115 in the DB-OTO-001 study, and data in adults and on long-term language are lacking [11, 23, 85]. In the SLC26A4 spectrum, almost all patients progress to severe to profound hearing loss, implantation gives good results, and goitre and hypothyroidism may appear throughout life, with a 6.7-fold higher risk of goitre after 20 years [5, 65]. The French PNDS adds that the prognosis is mainly auditory and depends on the severity and earliness of the hearing loss, that the course may extend over decades with episodes of sudden hearing loss, tinnitus and vertigo, and that the patient should be informed of the risk of abrupt worsening at any age [7]. Aminoglycoside-induced mitochondrial hearing loss is irreversible but not progressive, whereas hearing loss without exposure may progress; prevention depends on diagnosis before exposure [8]. In progressive AD forms, the course is predictable by gene: in KCNQ4, severe to profound hearing loss at 70 years; in COCH, profound hearing loss at a mean age of 76 years and bilateral vestibular areflexia between 49 and 60 years [48, 54].

Long-term follow-up rests on gene-adapted serial audiometry, recognition and treatment of superimposed otitis, surveillance for systemic manifestations in subsequently recognised syndromic forms, and ophthalmological assessment given visual reliance [1, 2]. Children should enter early-intervention programmes up to age three, with individualised educational plans and specialist deaf children's services, and transition to adulthood should be prepared during adolescence [1]. Outcomes depend less on the gene than intervention timing: late recognition and language exposure are associated with reading, communication and speech production deficits that later intervention may not correct [1].

At transition, which the PNDS places between 16 and 19 years, the first adult otorhinolaryngology visit reviews the history of the hearing loss and the hearing aid, repeats the examination and audiometry, recalls the situations that warrant an urgent visit and the inadvisable behaviours, provides information on social and university support and receives the previous imaging to avoid repetition; endocrinology reviews the goitre, nodules and hypothyroidism and stresses adherence; genetics explains transmission again to the young adult, assesses the risk in a pregnancy and responds to requests for prenatal or preimplantation diagnosis [7]. Follow-up also looks for otitis that worsens hearing loss, educational, social or occupational difficulties and psychological disorders related to the hearing loss [7]. In DFNB9, the French PNDS places transition between 18 and 20 years, with the same content for the first adult otorhinolaryngology and genetics visits, adds information on university adaptations and recognition of disabled worker status, and mentions therapeutic education programmes supporting transition in some hospitals; the particular nature of the cochlear implant makes this stage critical [20].

12. Summary and Final Considerations

Nonsyndromic genetic hearing loss is the most frequent prelingual form in countries with developed healthcare systems, with extreme genetic heterogeneity across more than 150 genes, predominantly autosomal recessive inheritance and diagnoses concentrated in a few genes [1]. GJB2 is the main cause of severe to profound congenital loss, with strong genotype–phenotype correlation and excellent implant outcomes; STRC dominates mild to moderate loss, requires CNV detection and has male fertility implications; SLC26A4 unites Pendred and DFNB4 in a spectrum defined by enlarged vestibular aqueducts, hearing fluctuation and late goitre [3–5].

Modern aetiological investigation combines early congenital CMV testing, first-line panel or genome molecular testing and targeted investigation of syndromic forms mimicking isolated loss, accounting for approximately 20% of children initially classified as nonsyndromic [1, 14]. Newborn screening with 1-3-6 targets remains the entry point but has known limitations: it misses mild and postnatal-onset loss and, when OAE-only, OTOF synaptopathy [15, 23].

Two fields exemplify genomic medicine in hearing loss. First, MT-RNR1 pharmacogenetics: m.1555A>G, m.1494C>T and m.1095T>C, present in approximately 1 in 330 people, confer irreversible loss after a single aminoglycoside dose, and rapid neonatal testing proved feasible without delaying antibiotics [10, 16, 72]. Second, OTOF gene therapy: lunsotogene parvec-cwha became the first approved gene therapy for genetic hearing loss in 2026, with 80% reaching ≤70 dB HL thresholds at week 24, requiring molecular diagnosis before implantation in the ear to be treated [11]. Approval is accelerated and based on single-arm studies; the EU application has been under assessment since May 2026, and contralateral readministration and Chinese 2.5-year outcomes extend but do not replace controlled benefit confirmation [13, 85, 86]. Genetic counselling should respect diverse perspectives, including Deaf community views, and use neutral language [1].

13. References

1. Shearer AE, Hildebrand MS, Odell AM, Smith RJH. Genetic Hearing Loss Overview. 1999 Feb 14 [updated 2023 Apr 6; reviewed 2026 Jun 2]. In: Adam MP, Bick S, Mirzaa GM, Pagon RA, Wallace SE, Amemiya A, editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1434/.
2. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 122-5 (Deafness in early childhood).
3. Smith RJH, Azaiez H, Booth K. GJB2-Related Autosomal Recessive Nonsyndromic Hearing Loss. 1998 Sep 28 [updated 2023 Jul 20]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1272/.
4. Redfield S, Shearer AE. STRC-Related Autosomal Recessive Hearing Loss. 2023 Dec 14. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK598310/.
5. Smith RJH, Azaiez H, Odell AM. SLC26A4-Related Sensorineural Hearing Loss. 1998 Sep 28 [updated 2025 Jan 9; reviewed 2025 Apr 3]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1467/.
6. Orphanet. Pendred syndrome (ORPHA:705); Autosomal recessive non-syndromic genetic deafness (ORPHA:90636); Autosomal dominant non-syndromic genetic deafness (ORPHA:90635); Mitochondrial non-syndromic sensorineural deafness (ORPHA:90641); Mitochondrial non-syndromic sensorineural deafness with susceptibility to aminoglycoside exposure (ORPHA:168609); Deafness-infertility syndrome (ORPHA:94064) [Internet]. Paris: INSERM; and Orphadata, rare disease epidemiology (en_product9_prev) [accessed 2026 Oct 4]. Available at: https://www.orphadata.com/data/xml/en_product9_prev.xml.
7. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Syndrome de Pendred. Centre de référence des surdités génétiques; filières SENSGENE et FIRENDO. Saint-Denis La Plaine: HAS; 2021 Mar 30 [accessed 2026 Nov 6]. Available at: https://www.has-sante.fr/jcms/p_3245873.
8. Usami SI, Nishio SY. Nonsyndromic Hearing Loss and Deafness, Mitochondrial. 2004 Oct 22 [updated 2018 Jun 14]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1422/.
9. Bitner-Glindzicz M, Pembrey M, Duncan A, Heron J, Ring SM, Hall A, et al. Prevalence of mitochondrial 1555A→G mutation in European children. N Engl J Med. 2009;360(6):640-2. doi:10.1056/NEJMc0806396.
10. McDermott JH, Wolf J, Hoshitsuki K, Huddart R, Caudle KE, Whirl-Carrillo M, et al. Clinical Pharmacogenetics Implementation Consortium Guideline for the Use of Aminoglycosides Based on MT-RNR1 Genotype. Clin Pharmacol Ther. 2022;111(2):366-72. doi:10.1002/cpt.2309.
11. U.S. Food and Drug Administration. OTARMENI (lunsotogene parvec-cwha) suspension for intracochlear infusion. Prescribing information. BLA 125874. Tarrytown (NY): Regeneron Pharmaceuticals; 2026 Apr [version of 27/04/2026; accessed through openFDA on 2026 Oct 4]. Available at: https://api.fda.gov/drug/label.json?search=openfda.brand_name:%22OTARMENI%22; CBER product page (STN 125874; approval letter of 23/04/2026 and Summary Basis for Regulatory Action of 22/04/2026) [accessed 2026 Oct 6]: https://www.fda.gov/vaccines-blood-biologics/otarmeni.
12. McGuigan A. Lunsotogene Parvec: First Approval. Drugs. 2026;86(10):1747-51. doi:10.1007/s40265-026-02361-7.
13. Regeneron Pharmaceuticals, Inc. Regeneron reports second quarter 2026 financial and operating results [press release; exhibit 99.1 to Form 8-K]. Tarrytown (NY); 2026 Jul 30. SEC EDGAR, accession 0000872589-26-000023. Available at: https://www.sec.gov/Archives/edgar/data/872589/000087258926000023/exhibit991q22026.htm.
14. Li MM, Tayoun AA, DiStefano M, Pandya A, Rehm HL, Robin NH, et al. Clinical evaluation and etiologic diagnosis of hearing loss: a clinical practice resource of the American College of Medical Genetics and Genomics (ACMG). Genet Med. 2022;24(7):1392-406. doi:10.1016/j.gim.2022.03.018.
15. Joint Committee on Infant Hearing. Year 2019 Position Statement: Principles and Guidelines for Early Hearing Detection and Intervention Programs. J Early Hear Detect Interv. 2019;4(2):1-44. doi:10.15142/fptk-b748.
16. McDermott JH, Hilton S, Dello Russo C, Booth N, Drysdale S, Howard P, et al. MT-RNR1 genotype testing for preventing aminoglycoside-mediated ototoxicity: a guideline developed by the UK Centre of Excellence in Regulatory Science and Innovation in Pharmacogenomics (CERSI-PGx). Br J Clin Pharmacol. 2026;92(9):3002-11. doi:10.1002/bcp.70712.
17. Fan X, Gao Z, Zhong J, Chen Y, Chen X, Landegger LD, et al. International expert consensus on gene therapy for hereditary hearing loss: based on clinical trials. Med. 2026;7(1):100886. doi:10.1016/j.medj.2025.100886.
18. Haute Autorité de Santé. Le traitement de la surdité par implants cochléaires ou du tronc cérébral. Fiche Bon usage des technologies médicales (BUTS), actualisation janvier 2012. Saint-Denis La Plaine: HAS; online 8 Feb 2012, updated 11 Jan 2013 [accessed 2026 Nov 6]. Available at: https://www.has-sante.fr/jcms/r_1438002.
19. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Maladies mitochondriales apparentées au MELAS. Centres de référence CALISSON et CARAMMEL, filière FILNEMUS; text of September 2021. Saint-Denis La Plaine: HAS; 2021 [accessed 2026 Nov 6]. Available at: https://www.has-sante.fr/jcms/p_3289846.
20. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Surdité DFNB9: synaptopathie due aux variations pathogènes du gène OTOF. Centre de référence des surdités génétiques; filière SENSGENE. Saint-Denis La Plaine: HAS; text of November 2022 (published 12 Dec 2022). Available at: https://www.has-sante.fr/jcms/p_3390116
21. ClinGen Actionability Working Group. Prelingual non-syndromic hearing loss and deafness (AC1018) and Pendred syndrome (AC1019): actionability reports, pediatric and adult contexts [Internet]. Clinical Genome Resource [accessed 2026 Oct 4]. Available at: https://actionability.clinicalgenome.org/ac/.
22. National Center for Biotechnology Information; EMBL-EBI. Matched Annotation from NCBI and EMBL-EBI (MANE), version 1.5, file MANE.GRCh38.v1.5.summary; and NCBI Nucleotide (NM_194248.3, transcript variant 1; NM_194323.3, transcript variant 4; NM_001287489.2, transcript variant 5) [accessed 2026 Oct 6]. Available at: https://ftp.ncbi.nlm.nih.gov/refseq/MANE/MANE_human/release_1.5/.
23. Azaiez H, Thorpe RK, Odell AM, Smith RJH. OTOF-Related Hearing Loss. 2008 Feb 29 [updated 2025 Mar 13; reviewed 2026 May 21]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1251/.
24. ERN CRANIO. European Reference Network for rare and/or complex craniofacial anomalies and ear, nose and throat (ENT) disorders [Internet; accessed 2026 Oct 6]. Available at: https://www.ern-cranio.eu.
25. European Union of the Deaf [Internet; accessed 2026 Oct 6]. Available at: https://www.eud.eu.
26. European Association of Cochlear Implant Users (EURO-CIU) [Internet; accessed 2026 Oct 6]. Available at: https://www.eurociu.eu.
27. Federação Portuguesa das Associações de Surdos [Internet; accessed 2026 Oct 6]. Available at: https://www.fpasurdos.pt.
28. Pendred V. Deaf-mutism and goitre. Lancet. 1896;148(3808):532. doi:10.1016/S0140-6736(01)74403-0.
29. Prezant TR, Agapian JV, Bohlman MC, Bu X, Oztas S, Qiu WQ, et al. Mitochondrial ribosomal RNA mutation associated with both antibiotic-induced and non-syndromic deafness. Nat Genet. 1993;4(3):289-94. doi:10.1038/ng0793-289.
30. Weil D, Blanchard S, Kaplan J, Guilford P, Gibson F, Walsh J, et al. Defective myosin VIIA gene responsible for Usher syndrome type 1B. Nature. 1995;374(6517):60-1. doi:10.1038/374060a0.
31. Kelsell DP, Dunlop J, Stevens HP, Lench NJ, Liang JN, Parry G, et al. Connexin 26 mutations in hereditary non-syndromic sensorineural deafness. Nature. 1997;387(6628):80-3. doi:10.1038/387080a0.
32. Everett LA, Glaser B, Beck JC, Idol JR, Buchs A, Heyman M, et al. Pendred syndrome is caused by mutations in a putative sulphate transporter gene (PDS). Nat Genet. 1997;17(4):411-22. doi:10.1038/ng1297-411.
33. Yasunaga S, Grati M, Cohen-Salmon M, El-Amraoui A, Mustapha M, Salem N, et al. A mutation in OTOF, encoding otoferlin, a FER-1-like protein, causes DFNB9, a nonsyndromic form of deafness. Nat Genet. 1999;21(4):363-9. doi:10.1038/7693.
34. Verpy E, Masmoudi S, Zwaenepoel I, Leibovici M, Hutchin TP, Del Castillo I, et al. Mutations in a new gene encoding a protein of the hair bundle cause non-syndromic deafness at the DFNB16 locus. Nat Genet. 2001;29(3):345-9. doi:10.1038/ng726.
35. del Castillo I, Villamar M, Moreno-Pelayo MA, del Castillo FJ, Álvarez A, Tellería D, et al. A deletion involving the connexin 30 gene in nonsyndromic hearing impairment. N Engl J Med. 2002;346(4):243-9. doi:10.1056/NEJMoa012052.
36. Qi J, Tan F, Zhang L, Lu L, Zhang S, Zhai Y, et al. AAV-mediated gene therapy restores hearing in patients with DFNB9 deafness. Adv Sci (Weinh). 2024;11(11):e2306788. doi:10.1002/advs.202306788.
37. Lv J, Wang H, Cheng X, Chen Y, Wang D, Zhang L, et al. AAV1-hOTOF gene therapy for autosomal recessive deafness 9: a single-arm trial. Lancet. 2024;403(10441):2317-25. doi:10.1016/S0140-6736(23)02874-X.
38. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. p. 782-4 (Senter-KID syndrome).
39. Mellerio J, Greenblatt D. Hidrotic Ectodermal Dysplasia 2. 2005 Apr 25 [updated 2020 Oct 15; reviewed 2025 Apr 3]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1200/.
40. Francey LJ, Conlin LK, Kadesch HE, Clark D, Berrodin D, Sun Y, et al. Genome-wide SNP genotyping identifies the stereocilin (STRC) gene as a major contributor to pediatric bilateral sensorineural hearing impairment. Am J Med Genet A. 2012;158A(2):298-308. doi:10.1002/ajmg.a.34391.
41. Nishio SY, Usami SI. Frequency of the STRC-CATSPER2 deletion in STRC-associated hearing loss patients. Sci Rep. 2022;12(1):634. doi:10.1038/s41598-021-04688-5.
42. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 116-7 (Congenital hypothyroidism).
43. Koenekoop RK, Arriaga MA, Gillies N, Trzupek KM, Lentz JJ. Usher Syndrome Type I. 1999 Dec 10 [updated 2026 Jun 9]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1265/.
44. Zhang J, Guan J, Wang H, Yin L, Wang D, Zhao L, et al. Genotype-phenotype correlation analysis of MYO15A variants in autosomal recessive non-syndromic hearing loss. BMC Med Genet. 2019;20(1):60. doi:10.1186/s12881-019-0790-2.
45. Nishio SY, Usami SI. Prevalence and clinical features of autosomal dominant and recessive TMC1-associated hearing loss. Hum Genet. 2022;141(3-4):929-37. doi:10.1007/s00439-021-02364-2.
46. Sugiyama K, Moteki H, Kitajiri SI, Kitano T, Nishio SY, Yamaguchi T, et al. Mid-frequency hearing loss is characteristic clinical feature of OTOA-associated hearing loss. Genes (Basel). 2019;10(9):715. doi:10.3390/genes10090715.
47. Mustapha M, Weil D, Chardenoux S, Elias S, El-Zir E, Beckmann JS, et al. An alpha-tectorin gene defect causes a newly identified autosomal recessive form of sensorineural pre-lingual non-syndromic deafness, DFNB21. Hum Mol Genet. 1999;8(3):409-12. doi:10.1093/hmg/8.3.409.
48. Smith RJH, Hildebrand M. DFNA2 Nonsyndromic Hearing Loss. 2008 Apr 4 [updated 2018 May 10]. In: Adam MP, Bick S, Mirzaa GM, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993-2026. Available at: https://www.ncbi.nlm.nih.gov/books/NBK1209/.
49. Kubisch C, Schroeder BC, Friedrich T, Lütjohann B, El-Amraoui A, Marlin S, et al. KCNQ4, a novel potassium channel expressed in sensory outer hair cells, is mutated in dominant deafness. Cell. 1999;96(3):437-46. doi:10.1016/S0092-8674(00)80556-5.
50. Sloan-Heggen CM, Bierer AO, Shearer AE, Kolbe DL, Nishimura CJ, Frees KL, et al. Comprehensive genetic testing in the clinical evaluation of 1119 patients with hearing loss. Hum Genet. 2016;135(4):441-50. doi:10.1007/s00439-016-1648-8.
51. Hildebrand MS, Morín M, Meyer NC, Mayo F, Modamio-Hoybjor S, Mencía A, et al. DFNA8/12 caused by TECTA mutations is the most identified subtype of nonsyndromic autosomal dominant hearing loss. Hum Mutat. 2011;32(7):825-34. doi:10.1002/humu.21512.
52. Verhoeven K, Van Laer L, Kirschhofer K, Legan PK, Hughes DC, Schatteman I, et al. Mutations in the human alpha-tectorin gene cause autosomal dominant non-syndromic hearing impairment. Nat Genet. 1998;19(1):60-2. doi:10.1038/ng0598-60.
53. Thorpe RK, Walls WD, Corrigan R, Schaefer A, Wang K, Huygen P, et al. AudioGene: refining the natural history of KCNQ4, GSDME, WFS1, and COCH-associated hearing loss. Hum Genet. 2022;141(3-4):877-87. doi:10.1007/s00439-021-02424-7.
54. JanssensdeVarebeke S, Topsakal V, Van Camp G, Van Rompaey V. A systematic review of hearing and vestibular function in carriers of the Pro51Ser mutation in the COCH gene. Eur Arch Otorhinolaryngol. 2019;276(5):1251-62. doi:10.1007/s00405-019-05322-x.
55. Kim BJ, Han JH, Park HR, Kim MY, Kim AR, Oh SH, et al. A clinical guidance to DFNA22 drawn from a Korean cohort study with an autosomal dominant deaf population: a retrospective cohort study. J Gene Med. 2018;20(6):e3019. doi:10.1002/jgm.3019.
56. Shinagawa J, Moteki H, Nishio SY, Ohyama K, Otsuki K, Iwasaki S, et al. Prevalence and clinical features of hearing loss caused by EYA4 variants. Sci Rep. 2020;10(1):3662. doi:10.1038/s41598-020-60259-0.
57. Bernardinelli E, Huber F, Roesch S, Dossena S. Clinical and molecular aspects associated with defects in the transcription factor POU3F4: a review. Biomedicines. 2023;11(6):1695. doi:10.3390/biomedicines11061695.
58. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 1, section 1.5.4 (Deafness) and Table 1.21 (Sensorineural deafness).
59. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 34-5 (Mitochondrial inheritance).
60. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 490-5 (Mitochondrial DNA diseases).
61. Chan DK, Chang KW. GJB2-associated hearing loss: systematic review of worldwide prevalence, genotype, and auditory phenotype. Laryngoscope. 2014;124(2):E34-53. doi:10.1002/lary.24332.
62. Snoeckx RL, Huygen PL, Feldmann D, Marlin S, Denoyelle F, Waligora J, et al. GJB2 mutations and degree of hearing loss: a multicenter study. Am J Hum Genet. 2005;77(6):945-57. doi:10.1086/497996.
63. Thorpe RK, Azaiez H, Wu P, Wang Q, Xu L, Dai P, et al. The natural history of OTOF-related auditory neuropathy spectrum disorders: a multicenter study. Hum Genet. 2022;141(3-4):853-63. doi:10.1007/s00439-021-02340-w.
64. Mey K, Bille M, Rye Rasmussen SH, Tranebjærg L, Cayé-Thomasen P. The natural history of hearing loss in Pendred syndrome and non-syndromic enlarged vestibular aqueduct. Otol Neurotol. 2019;40(3):e178-85. doi:10.1097/MAO.0000000000002140.
65. Madeo AC, Manichaikul A, Reynolds JC, Sarlis NJ, Pryor SP, Shawker TH, et al. Evaluation of the thyroid in patients with hearing loss and enlarged vestibular aqueducts. Arch Otolaryngol Head Neck Surg. 2009;135(7):670-6. doi:10.1001/archoto.2009.66.
66. Estivill X, Govea N, Barceló E, Badenas C, Romero E, Moral L, et al. Familial progressive sensorineural deafness is mainly due to the mtDNA A1555G mutation and is enhanced by treatment of aminoglycosides. Am J Hum Genet. 1998;62(1):27-35. doi:10.1086/301676.
67. Diogo-Cavassana S, Alencar-Coutinho D, Abreu-Oberhuber R, Paramo-Neto ME, Oiticica J, Bento RF, et al. Genetic landscape of hearing loss in Brazilian patients reveals population-specific variants and clinical correlations. Clin Genet. 2026;110(2):210-26. doi:10.1111/cge.70186.
68. Brasil. Lei n.º 12.303, de 2 de agosto de 2010. Dispõe sobre a obrigatoriedade de realização do exame denominado Emissões Otoacústicas Evocadas. Diário Oficial da União. 2010 Aug 3. Available at: https://www.planalto.gov.br/ccivil_03/_ato2007-2010/2010/lei/l12303.htm.
69. Vernier LS, Cazella SC, Levandowski DC. Triagem auditiva neonatal: protocolos, obstáculos e perspectivas de fonoaudiólogos no Brasil – 10 anos da Lei Federal Brasileira 12.303/2010. CoDAS. 2022;34(2):e20200331. doi:10.1590/2317-1782/20212020331.
70. Guggenheim DS, Sie KCY, Lustig LR, Golub JS. Newborn hearing screening in the otoferlin gene therapy era. JAMA Otolaryngol Head Neck Surg. 2026 Sep 17 [advance electronic publication]. doi:10.1001/jamaoto.2026.2695. PMID: 42752727.
71. Hayes D, Sininger YS, co-chairs. Guidelines for identification and management of infants and young children with auditory neuropathy spectrum disorder. Guidelines Development Conference at NHS 2008, Como (Italy). Aurora (CO): Bill Daniels Center for Children's Hearing, The Children's Hospital – Colorado; 2008. Available at: https://www.audiology.org/wp-content/uploads/2021/06/ANSDGuidelines2008.pdf.
72. McDermott JH, Mahaveer A, James RA, Booth N, Turner M, Harvey KE, et al. Rapid point-of-care genotyping to avoid aminoglycoside-induced ototoxicity in neonatal intensive care. JAMA Pediatr. 2022;176(5):486-92. doi:10.1001/jamapediatrics.2022.0187.
73. Forli F, Capobianco S, Berrettini S, Bruschini L, Romano S, Fogli A, et al. Temperature-sensitive auditory neuropathy: report of a novel variant of OTOF gene and review of current literature. Medicina (Kaunas). 2023;59(2):352. doi:10.3390/medicina59020352. PMID: 36837553; PMCID: PMC9962730.
74. Brownstein Z, Kamal L, Mishan-Montefiori S, Hoffman Y, Fine D, Hirsch Y, et al. A multicenter study reveals a novel pathogenic splice-site founder variant in OTOF. Hum Genomics. 2025;19(1):112. doi:10.1186/s40246-025-00819-5. PMID: 41053850; PMCID: PMC12502505.
75. National Center for Biotechnology Information. ClinVar: records NM_004004.6(GJB2):c.35del, c.167del, c.235del and c.109G>A; NM_000441.2(SLC26A4):c.919-2A>G; NM_194248.3(OTOF):c.5816G>A [accessed 2026 Oct 6]. Available at: https://www.ncbi.nlm.nih.gov/clinvar/.
76. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 10 (Disorders of Oxidative Phosphorylation), section 10.3.3 (Frequency of Mutations).
77. Ford CL, Riggs WJ, Quigley T, Keifer OP Jr, Whitton JP, Valayannopoulos V. The natural history, clinical outcomes, and genotype-phenotype relationship of otoferlin-related hearing loss: a systematic, quantitative literature review. Hum Genet. 2023;142(10):1429-49. doi:10.1007/s00439-023-02595-5.
78. Hansen MU, Rye Rasmussen E, Cayé-Thomasen P, Mey K. Cochlear implantation in children with enlarged vestibular aqueduct: a systematic review of surgical implications and outcomes. Ear Hear. 2023;44(3):440-7. doi:10.1097/AUD.0000000000001309.
79. European Medicines Agency. Download medicine data: medicines report [Internet]. Amsterdam: EMA [accessed 2026 Oct 4]. Available at: https://www.ema.europa.eu/en/documents/report/medicines-output-medicines_json-report_en.json.
80. Regeneron Pharmaceuticals, Inc. Quarterly report (Form 10-Q) for the period ended June 30, 2026: development milestones table (Otarmeni — European Commission decision expected in the first quarter of 2027). SEC EDGAR, accession 0000872589-26-000025; 2026 Jul 30. Available at: https://www.sec.gov/Archives/edgar/data/872589/000087258926000025/regn-20260630.htm.
81. Regeneron Pharmaceuticals, Inc. Regeneron reports first quarter 2026 financial and operating results [press release; exhibit 99.1 to Form 8-K]. Tarrytown (NY); 2026 Apr 29. SEC EDGAR, accession 0000872589-26-000014. Available at: https://www.sec.gov/Archives/edgar/data/872589/000087258926000014/exhibit991q12026.htm.
82. European Medicines Agency. Download medicine data: orphan designations report (EU/3/23/2771, AAV1 containing the 5' and 3' portions of human OTOF, 20/03/2023; EU/3/21/2477, AAVAnc80-OTOF, 19/07/2021; EU/3/22/2698, AAV8 containing the otoferlin coding sequence, 11/10/2022) [Internet]. Amsterdam: EMA [accessed 2026 Oct 6]. Available at: https://www.ema.europa.eu/en/documents/report/medicines-output-orphan_designations-json-report_en.json.
83. Valayannopoulos V, Bance M, Carvalho DS, Greinwald JH Jr, Harvey SA, Ishiyama A, et al.; CHORD Study Group. DB-OTO gene therapy for inherited deafness. N Engl J Med. 2026;394(11):1074-83. doi:10.1056/NEJMoa2400521.
84. Wang H, Chen Y, Lv J, Cheng X, Cao Q, Wang D, et al. Bilateral gene therapy in children with autosomal recessive deafness 9: single-arm trial results. Nat Med. 2024;30(7):1898-904. doi:10.1038/s41591-024-03023-5.
85. Jiang L, Cheng X, Lv J, Chen Y, Chen X, Zhai R, et al. Multicentre gene therapy for OTOF-related deafness followed up to 2.5 years. Nature. 2026;653(8116):1170-7. doi:10.1038/s41586-026-10393-y.
86. Fan X, Zhang L, Gao Z, Han L, Yang P, Wang D, et al. Re-administration of AAV-mediated gene therapy for OTOF-related deafness: a single-arm trial. Nat Med. 2026;32(8):2846-56. doi:10.1038/s41591-026-04505-4. PMID: 42362868.
87. ClinicalTrials.gov. NCT05821959 (AAVAnc80-hOTOF, Eli Lilly/Akouos; phase 3, recruiting, 48 planned); NCT05788536 (DB-OTO, Regeneron; phase 1/2, recruiting, 36 planned); NCT06370351 (SENS-501, Sensorion; phase 1/2, active not recruiting, 12 planned); NCT06722170 (EH002; recruiting); NCT05901480 (OTOV101N+OTOV101C; unknown status) [Internet]. Bethesda (MD): National Library of Medicine [accessed through API v2 on 2026 Oct 6]. Available at: https://clinicaltrials.gov/.
88. Boudra R, Olivier G, Tran Van Ba C, Barrot L, Rambeau P, Vaux C, et al. Efficacy and safety of SENS-501, a dual-AAV otoferlin gene therapy, for DFNB9 congenital deafness. Mol Ther Adv. 2026;34(3):201762. doi:10.1016/j.omta.2026.201762.
89. Shubina-Oleinik O, Nist-Lund C, French C, Rockowitz S, Shearer AE, Holt JR. Dual-vector gene therapy restores cochlear amplification and auditory sensitivity in a mouse model of DFNB16 hearing loss. Sci Adv. 2021;7(51):eabi7629. doi:10.1126/sciadv.abi7629.
90. Chandrasekhar SS, Tsai Do BS, Schwartz SR, Bontempo LJ, Faucett EA, Finestone SA, et al. Clinical Practice Guideline: Sudden Hearing Loss (Update). Otolaryngol Head Neck Surg. 2019;161(1_suppl):S1-S45. doi:10.1177/0194599819859885. PMID: 31369359.
91. European Medicines Agency. Metoclopramide-containing medicines: Article 31 referral (EMEA/H/A-31/1321); European Commission decision of 20/12/2013 [Internet]. Amsterdam: EMA [accessed 2026 Nov 6]. Available at: https://www.ema.europa.eu/en/medicines/human/referrals/metoclopramide-containing-medicines.
92. Haute Autorité de Santé. Stratégie de vaccination contre les infections invasives à pneumocoques: place du vaccin CAPVAXIVE chez l'adulte. Recommandation vaccinale. Saint-Denis La Plaine: HAS; 2025 Jul 7. Available at: https://www.has-sante.fr/jcms/p_3634637
93. 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.
94. Centers for Disease Control and Prevention. Pneumococcal vaccine recommendations for people with cochlear implants [Internet]. Atlanta: CDC; 2026 Feb 25 [accessed 2026-11-07]. Available at: https://www.cdc.gov/pneumococcal/vaccines/cochlear-implants.html
95. U.S. Food and Drug Administration. Synthroid (levothyroxine sodium) tablets. Prescribing information, section 2 Dosage and Administration (Tables 1–3). AbbVie Inc.; NDA021402; version of 20/02/2024 [accessed via openFDA on 2026 Nov 6]. Available at: https://api.fda.gov/drug/label.json?search=openfda.brand_name:synthroid.
96. U.S. Food and Drug Administration. Gentamicin sulfate injection. Prescribing information, section Warnings — Risk of ototoxicity due to mitochondrial DNA variants. Fresenius Kabi USA (ANDA062366); version of 31/10/2023 [accessed through openFDA on 2026 Oct 6]; equivalent wording in amikacin (2024) and tobramycin (2023) labels. Available at: https://api.fda.gov/drug/label.json?search=openfda.generic_name:gentamicin+AND+warnings:mitochondrial.
97. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 810-4 (Carrier frequency and carrier testing for autosomal recessive disorders, Tables A6 and A8).
98. Green GE, Scott DA, McDonald JM, Woodworth GG, Sheffield VC, Smith RJ. Carrier rates in the midwestern United States for GJB2 mutations causing inherited deafness. JAMA. 1999;281(23):2211-6. doi:10.1001/jama.281.23.2211.
99. ERN-EYE. European Reference Network for rare eye diseases [Internet; accessed 2026 Oct 6]. Available at: https://www.ern-eye.eu.
100. Dória M, Neto AP, Santos AC, Barros H, Fernandes S, Moura CP. Prevalence of 35delG and Met34Thr GJB2 variants in Portuguese samples. Int J Pediatr Otorhinolaryngol. 2015;79(12):2187-90. doi:10.1016/j.ijporl.2015.09.046.
101. Reis CS, Santos AC, Barros H, Fernandes S, Moura CP. GJB2: frequency of the less common variants in a sample of the Portuguese population. Acta Med Port. 2021;34(9):592-600. doi:10.20344/amp.13781.
102. De Rosa MA, Bernardi MT, Kleppe S, Walz K. Hearing loss: genetic testing, current advances and the situation in Latin America. Genes (Basel). 2024;15(2):178. doi:10.3390/genes15020178.
103. Antunes LN, Dias AMM, Schiavo BC, Mendes BCA, Bertola DR, Lezirovitz K, et al. Genetic heterogeneity in autosomal recessive hearing loss: a survey of Brazilian families. Front Genet. 2024;15:1409306. doi:10.3389/fgene.2024.1409306.
104. Han S, Zhang D, Guo Y, Fu Z, Guan G. Prevalence and characteristics of STRC gene mutations (DFNB16): a systematic review and meta-analysis. Front Genet. 2021;12:707845. doi:10.3389/fgene.2021.707845.
105. Chen YH, Lin WC, Hwang CF, Tsai MH, Yang CH. Variability in inner ear morphology among a family with Pendred syndrome due to a SLC26A4 gene variant. Ann Otol Rhinol Laryngol. 2024;133(9):828-33. doi:10.1177/00034894241261491.
106. Miura A, Nakagawa T, Sogi C, Shima H, Adachi M, Honkura Y, et al. Hearing loss with two pathogenic SLC26A4 variants and positive thyroid autoantibody: a case report. Clin Pediatr Endocrinol. 2024;33(4):219-23. doi:10.1297/cpe.2023-0084.