Clinical Utility of Next-generation Sequencing in the Aetiological Diagnosis of Sensorineural Hearing Loss in a Childhood Hearing Loss Unit

Abstract

Introduction

Sensorineural hearing loss (SNL) is the most prevalent sensory deficit in our environment. The next generation genomic sequencing (NGS) allows to obtain an etiological diagnosis in a high percentage of patients. Our pilot study shows the results of the systematic application of NGS in a Childhood Hearing Loss Unit, as well as its implications in the clinical management of patients and their families.

Material and method

We included 27 patients diagnosed with SNH between 2014 and 2017, in which an environmental cause was ruled out. The genetic test consisted of a panel of genes analyzed by NGS (OTOgenics™ panel). This panel has been designed to include genes associated with sensorineural or mixed hearing loss, early onset or late, syndromic and non-syndromic, regardless of their inheritance pattern.

Results

A genetic diagnosis was obtained in 56% (15/27) of the patients (62% in the case of bilateral HNS). 5/27 (19%) presented pathogenic variants in the GJB2 gene and the rest variants pathogenic and/or probably pathogenic in other genes associated with isolated HNS (PR2 × 2, TECTA and STRC), with syndromic HNS (CHD7, GATA3, COL4A5, MITF and SOX10) or with syndromic and non-syndromic HNS (BSND, ACTG1 and CDH23).

Discussion

The etiological diagnosis of SNL is a challenge in clinical practice. Our series demonstrates that it is possible to implement genetic diagnosis in the daily routine and that this information has prognostic and therapeutic implications.

Resumen

Introducción

La hipoacusia neurosensorial (HNS) es el déficit sensorial más prevalente en nuestro medio. La secuenciación genómica de nueva generación (NGS) permite obtener un diagnóstico etiológico en un alto porcentaje de pacientes. Nuestro estudio piloto muestra los resultados de la aplicación sistemática de la NGS en una Unidad de Hipoacusia Infantil, así como sus implicaciones en el manejo clínico de los pacientes y sus familiares.

Material y método

Se incluyeron 27 pacientes diagnosticados de HNS entre 2014 y 2017 en los que se descartó una causa ambiental. El test genético consistió en un panel de genes analizados mediante NGS (panel OTOgenics™). Este panel ha sido diseñado para incluir genes asociados con hipoacusia neurosensorial o mixta, de inicio precoz o tardío, sindrómica y no sindrómica, independientemente de su patrón de herencia.

Resultados

Se obtuvo un diagnóstico genético en el 56% (15/27) de los pacientes (62% en el caso de las HNS bilaterales). 5/27 (19%) presentaron variantes patogénicas en el genGJB2 y el resto variantes patogénicas y/o probablemente patogénicas en otros genes asociados con HNS aislada (PR2 × 2, TECTA y STRC), con HNS sindrómicas (CHD7, GATA3, COL4A5, MITF y SOX10) o con HNS sindrómicas y no sindrómicas (BSND, ACTG1 y CDH23).

Discusión

El diagnóstico etiológico de la HNS supone un desafío en la práctica clínica. Nuestra serie demuestra que es posible implementar el diagnóstico genético en la rutina asistencial y que esta información tiene implicaciones pronósticas y terapéuticas.

Introduction

Hearing loss affects approximately 30% of the population at some point in their lives.¹ Early-onset sensorineural hearing loss (SNL) is one of the most common hereditary diseases in our environment.² Approximately 2/3 of childhood SNL is of genetic aetiology, usually non-syndromic and with no family history. In the era of precision medicine, attempting to obtain an aetiological diagnosis of early-onset SNL is imperative and genetic testing demonstrates the highest diagnostic performance.^2,3 The prevalence of the condition and the implementation of early detection programmes for hearing loss, together with the recent exponential development of DNA sequencing techniques, give the specialty of ENT the potential to lead in this new way of practicing preventive, predictive, personalised and participatory medicine.

An early diagnosis of hearing loss has great impact on the patient’s clinical progress and life, and enables early intervention or monitoring.⁴ Obtaining an aetiological diagnosis has multiple advantages: 1) it can provide information on the prognosis of the hearing loss, thus individualising the follow-up; 2) it enables identification of hidden syndromes, anticipating their manifestations, either preventing them or mitigating their consequences, improving the quality of life of patients; 3) it allows the avoidance of unnecessary diagnostic tests, containing medical expense and minimising the risks for the patient; 4) it provides information on the different reproductive options, all of them with the common objective of avoiding the condition being passed down to future descendents,^5–7 and 5) in certain cases, fundamentally syndromic, it facilitates the selection of the most appropriate treatment. Furthermore, in the near future, a genetic diagnosis will be essential to include patients in clinical trials of gene, drug or cell therapy, targeting their specific molecular abnormality.⁸

Genetic diagnosis of SNL has been a great challenge, given the extreme genetic heterogeneity of the disease.^9–11 The new technologies of DNA sequencing (next generation sequencing [NGS]), which allow simultaneous analysis of hundreds of genes in time and cost compatible with healthcare practice, have become the new standard for achieving an aetiological diagnosis of hereditary hearing loss.^12,13

Both the 2014 guideline for the "Clinical Evaluation and Etiological Diagnosis of Hearing Loss" by the American College of Medical Genetics and Genomics (ACMG),² and the 2015 Spanish recommendations for the "Aetiological Diagnosis of Childhood Deafness", by the Commission for the Early Detection of Hearing Loss (CODEPEH),¹⁴ recommend incorporating early genetic diagnosis into protocols for the diagnosis of deafness. This will consist either of a sequential study in which the GJB2 and GJB6 genes are analysed first, and in negative cases, in a second stage analysing a gene panel using NGS, or directly analysing a gene panel in cases where this is considered to be indicated.

The purpose of this article is to show the first results of the implementation of NGS in clinical practice in a Childhood Hearing Unit, with special emphasis on its potential utility for patient management

Material and Method

Patients

Twenty-seven patients diagnosed with SNL or mixed hearing loss in the Children’s Hearing Loss Unit of the Hospital Universitario Central de Asturias were included between 2014 and 2017. Of the patients, 56% (15/27) were diagnosed with SNL after they had failed neonatal screening. The remaining 44% (12/27) were subsequently referred to the Unit with hearing loss detected at a later age or because they were born in another autonomous community. An environmental cause for their hearing loss was not identified in any of the patients. They were all assessed by a clinical genetics’ specialist, on suspicion of a hereditary aetiology.

The ACMG classification² was followed to label the hearing loss.

This study was approved by the Research Ethics Committee of the Principality of Asturias (research project #75/14).

Genetic Panel

After obtaining 4 ml peripheral blood in EDTA tubes and extracting genomic DNA, a gene panel was sequenced using NGS (OTOgenics™ panel, IMOMA, Spain). This panel is designed to include genes associated with SNL or mixed hearing loss, early or late onset, syndromic or non-syndromic, regardless of their inheritance pattern. The methodology used in the panel has been previously published by Cabanillas et al.¹⁵ The first version of the panel used included 165 genes: 76 genes involved in syndromic SNL, 62 genes in non-syndromic SNL and 27 genes related to both types. The second version included 184 genes: 84 associated with syndromic SNL, 73 related to non-syndromic SNL and 27 genes described in both types (Table 1).

Table 1.

Genes Included in the Panel Version 2.

Genes Consistently Associated in the Scientific-medical Literature With Sensorineural or Mixed Hearing Loss
ABHD12	BSND	CLRN1	DIAPH1	GJB3	KCNJ10	MSRB3	NARS2	PEX1	SERAC1	STRC	USH1G
ACTB	CABP2	COCH	DNMT1	GJB6	KCNQ1	MT-CO1	NDP	PEX2	SERPINB6	SYNE4	USH2A
ACTG1	CACNA1D	COL2A1	EDN3	GPSM2	KCNQ4	MT-RNR1	NLRP3	PEX3	SIX1	TBC1D24	WFS1
ADGRV1	CCDC50	COL4A3	EDNRB	GRHL2	LARS2	MT-TH	OPA1	PEX5	SLC17A8	TECTA
AIFM1	CDH23	COL4A4	ESPN	GRXCR1	LHFPL5	MT-TK	OSBPL2	PEX26	SLC19A2	TIMM8A
ALMS1	CEACAM16	COL4A5	ESRRB	HARS2	LHX3	MT-TL1	OTOA	POU3F4	SLC26A4	TJP2
ANKH	CHD7	COL9A1	EYA1	HGF	LOXHD1	MT-TS1	OTOF	POU4F3	SLC33A1	TMC1
AP1S1	CIB2	COL11A1	EYA4	HOXA1	LRP2	MYH9	OTOG	PRPS1	SLC52A2	TMIE
ATP1A3	CISD2	COL11A2	FGF3	HOXB1	LRTOMT	MYH14	OTOGL	PTPN11	SLC52A3	TMPRSS3
ATP6V1B1	CLCNKA	DDX11	FGFR3	HSD17B4	MARVELD2	MYO3A	P2RX2	PTPRQ	SLITRK6	TPRN
BCAP31	CLCNKB	DFNA5	GATA3	ILDR1	MASP1	MYO6	PAX3	RAF1	SMPX	TRIOBP
BCS1L	CLDN14	DFNB59	GIPC3	KARS	MIR96	MYO7A	PCDH15	RDX	SNAI2	TSPEAR
BRAF	CLPP	DIABLO	GJB2	KCNE1	MITF	MYO15A	PDZD7	RMND1	SOX10	USH1C

Candidate Genes With Preliminary Evidence in Favour of Their Association With Sensorineural or Mixed Hearing Loss
ADCY1	COL4A6	CRYM	ELMOD3	FGFR1	HMX2	MT-TA	SEMA3E	SLC4A11	SLC26A5	TMEM132E	TNC
ATP2B2	COL9A2	DCDC2	EPS8	FGFR2	HMX3	MT-TE	SIX5	SLC9A1	TK2	TMPRSS5	TP63
ATP6V1B2	COL9A3	DIAPH3	FAM65B	FOXI1	HOMER2	MT-TS2
BDP1	COQ6	DSPP	FBLN1	GRXCR2	MT-CO3	PNPT1

The genes that showed pathogenic/likely pathogenic variants in our patients are highlighted.

The clinical interpretation and final classification of the genetic variants was obtained after analysis of the databases (HGMD professional [Qiagen], ClinVar [https://www.ncbi.nlm.nih.gov/clinvar/], OMIM [https://www.omim.org/], PubMed [https://www.ncbi.nlm.nih.gov/pubmed/], 1000 Genomes project [https://www.genome.gov/27528684/1000-genomes-project], ESP [https://esp.gs.washington.edu/drupal/] and ExAC [http://exac.broadinstitute.org/about]) as well as the references available in each case. The variants were classified according to ACMG² criteria as pathogenic (P) (class 5), "likely pathogenic" (LP) (class 4), uncertain significance (class 3), likely benign (class 2) and benign (class 1).

Results

The Patients’ Clinical and Audiological Characteristics

Seventy-four percent (20/27) of the SNL was classified as congenital. The mean age at diagnosis was 2.6 years, with a range of 0–10 years of age

The relevant clinical details are detailed in Table 2; 19/27 (70%) patients had no family history of hearing loss. Among the 8/27 (30%) that did, 1 was first degree, 2 s degree, 4 third degree, and one had several family members with poorly defined hearing impairments in their clinical history.

Table 2.

Characteristics of the Patients.

Cases	FH of Hearing Loss	Other FH of Interest	Perinatal RF	Otological malformations	Associated problems	Type of hearing loss	Result of genetic testing (gene responsible)
1	No	No	No	No	No	Symmetrical bilateral profound prelingual hearing loss	Positive (GJB2/GJB6)
2	No	No	No	Dilated inner ear canals	Hypertelorism	Symmetrical bilateral profound prelingual hearing loss	Positive (SOX10)
3	No	No	No	No	No	Symmetrical bilateral profound prelingual hearing loss	Positive (CDH23)
4	No	No	No	No	No	Symmetrical bilateral severe prelingual SNL	Positive (GATA3)
5	No	No	No	No	No	Symmetrical bilateral severe mixed prelingual hearing loss	Positive (ACTG1)
6	Paternal great-grandmother	Father and paternal grandmother β-thalassaemia minor	No	Detached pinna	Atrophic testes, hypogonadism, micropenis, scrotal hypoplasia, aortic coarctation, aberrant subclavian	Symmetrical bilateral severe prelingual SNL	Positive (CHD7)
7	Several poorly documented	No	No	No	No	Symmetrical bilateral profound prelingual hearing loss	Positive (GJB2)
8	No	No	No	No	No	Symmetrical bilateral profound prelingual hearing loss	Positive (GJB2)
9	Not assessable	Not assessable	Not assessable	No	Bilateral lenticonus, right eye cataract, glomerulosclerosis	Symmetrical bilateral moderate prelingual SNL	Positive (COL4A5)
10	No	No	No	No	Clubfoot	Symmetrical bilateral profound prelingual hearing loss	Positive (BSND)
11	Maternal grandmother, maternal great aunt	Mother and maternal grandfather with strand of white hair	No	No	No	Unilateral profound prelingual SNL	Positive (MITF)
12	No	No	No	No	No	Asymmetrical bilateral moderate-severe prelingual	Positive (GJB2)
13	Maternal great-grandfather	No	No	No	No	Symmetrical bilateral moderate postlingual SNL	Positive (P2RX2)
14	No	Maternal uncle with type I diabetes mellitus	No	No	Diabetes mellitus pending characterisation	Symmetrical bilateral mild-moderate postlingual SNL	Positive (GJB2)
15	Sibling, maternal uncle, paternal cousin	No	No	No	No	Symmetrical bilateral moderate postlingual SNL	Positive (STRC/TECTA)
16	No	No	No	No	No	Symmetrical bilateral moderate prelingual SNL	Negative
17	No	No	No	No	No	Unilateral profound prelingual SNL	Negative
18	Maternal grandfather and maternal uncle	Frequent miscarriages in maternal family	No	No	Clinodactyly 5th finger	Symmetrical bilateral mild-moderate postlingual SNL	Negative
19	No	No	No	No	No	Unilateral severe prelingual SNL	Negative
20	Sibling and maternal great- grandmother	No	No	No	No	Unilateral profound prelingual SNL	Negative
21	No	No	No	No	Clotting factor II deficiency	Asymmetrical bilateral moderate/severe postlingual SNL	Negative
22	No	Altered clotting factor II in maternal family	No	No	Bilateral, alternating, convergent strabismus with occasional nystagmus	Symmetrical bilateral severe prelingual SNL	Negative
23	Mother	No	No	No	No	Unilateral moderate postlingual SNL	Negative
24	No	No	No	No	No	Symmetrical bilateral moderate prelingual SNL	Negative
25	Paternal aunt	No	Foetal distress. Jaundice	Dilated vestibular aqueduct	No	Asymmetrical bilateral profound/mild prelingual SNL	Negative
26	No	No	No	Cochlear incomplete partition type III and dilated inner ear canals	No	Symmetrical bilateral profound prelingual hearing loss	Negative
27	No	Father: postaxial polydactyly in feet and one hand and polyindactyly in 3rd and 4th fingers of contralateral hand. Paternal grandfather: hypospadias	No	Grooved earlobe. Cochlea: single bilateral cystic cavity with associated semi-circular dysplasia	Poly and syndactyly. Hypospadias. Permeable foramen ovale	Symmetrical bilateral profound prelingual hearing loss	Negative

FH: family history; RF: risk factors; SNL: sensorineural hearing loss.

Twenty-two out of twenty-seven (81%) cases had no other family history; 5/27 (19%) had a family history of another relevant disease.

According to the ACMG² classification:

• •

  19/27 (70%) symmetrical bilateral SNL, 3/27 (11%) asymmetrical bilateral and 5/27 (19%) unilateral.

• •

  19/27 (70%) prelingual SNL, 1/27 (4%) prelingual mixed hearing loss; 7/27 (26%) post-lingual SNL.

• •

  Within the symmetrical hearing loss: 16/27 (59%) severe or profound; 3/27 (11%) mild-moderate/moderate. Asymmetrical: 2/27 (7%) moderate in one ear and severe in the contralateral ear and 1/27 (4%) mild in one ear and profound in the contralateral ear.

• •

  Except for those detected post-lingually, which have therefore varied, the remainder have remained stable to date.

In terms of evolution: 9/27 (33%) implanted patients; 12/27 (44%) with adapted hearing devices; 5/27 (19%) in follow-up; 1/27 (4%) not recorded because they are being monitored in a different autonomous community. The decision to implant either uni- or bilaterally was taken on an individual basis in each case, depending on the indications at the time and the patient’s characteristics. The clinical characteristics of the patients are shown in Table 2.

Pathogenic and Likely Pathogenic Genetic Variants Identified

A genetic diagnosis was obtained in 56% (15/27) of the patients, specifically, in 62% of the bilateral SNL (13/21) and 40% of the unilateral SNL patients (2/5). Nineteen percent (5/27) had P variants in the DFNB1 locus (genes GJB2 and GJB6) and the remaining P or LP variants in other genes: PR2 × 2 and STRC associated with non-syndromic SNL; CHD7, GATA3, COL4A5, MITF and SOX10 associated with syndromic hearing loss, and BSND, ACTG1 and CDH23 associated with both.

Sixty-seven percent (10/15) of the variants found showed an autosomal recessive (AR) inheritance pattern and 33% (5/15) an autosomal dominant (AD) pattern. Variants considered to cause the hearing loss phenotype were found in 14/19 patients with bilateral hearing loss and in 1/5 with unilateral hearing loss (mutation in MITF causing Waardenburg syndrome). In addition, 3 de novo P/LP variants were identified: ACTG1 (not previously described in the literature), GATA3 and COL4A5.

Genotype-phenotype Correlations

We classified our results according to the identified genetic alteration and its phenotype in 3 groups (Table 3):

• 1)

  Suspected syndromic SNL pre-test: 7% (2/27) of the genetically diagnosed cases (probands 6 and 9).

• 2)

  Unsuspected syndromic SNL pre-test ("hidden syndromes"): 11% (3/27) of the genetically diagnosed cases fit this classification (probands 2, 4 and 11), since the identified genetic alterations are only associated with syndromic SNL. Additionally, another 11% (probands 3, 5 and 10) present genetic alterations that can be associated with both syndromic and non-syndromic SNL. The clinical evolution of the patients, being aware of the potential complications that could develop, will enable us to classify them into one or other group in the future. Hearing loss was the only pre-test clinical manifestation identified in all of the patients.

• 3)

  Non-syndromic isolated SNL: 26% (7/27) of the diagnosed cases (probands 1, 7, 8, 12, 13, 14 and 15).

Table 3.

Genotype-phenotype Correlations.

	Age at Diagnosis of SNL	Phenotype Pre-test	P and LP Genetic Variants Identified	Phenotypes Associated With the Genetic Variants Identified	Inheritance Pattern	Audiological Management
1	Congenital	Symmetrical bilateral isolated profound SNL	Compound heterozygote GJB2 (c.35delG;p.Gly12Valfs*2) and GJB6 (c.-237135_-5094; of (GJB6-D1351854))	Non-syndromic SNL(DFNB1)	AR	Unilateral CI
2	Congenital	Symmetrical bilateral profound SNL with hypertelorism	SOX10 (c.135_154 del;p.Ser45Argfs*15) in heterozygosis	Waardenburg syndrome type 2B	AD	Unilateral CI
				Waardenburg syndrome type 4C
				PCWH syndrome (or Waardenburg syndrome type 4 “plus”)
3	Congenital	Symmetrical bilateral isolated profound SNL	CDH23 (c.4488 G > C;p.Gln1496His) in heterozygosis and exons 11, 12, 13, 13 and 15 duplicated in CDH23	Non-syndromic SNL(DFNB12)	AR	Bilateral CI
				Usher syndrome type 1D (USH1D)
4	Congenital	Symmetrical bilateral severe isolated SNL	GATA3 (c.1018A > C;p.N340 H) in heterozygosis	Barakat syndrome	AD	Bilateral HA
5	Congenital	Symmetrical bilateral mixed severe isolated hearing loss	ACTG1 (c.548 G > A;p.R183Q) in heterozygosis	Non-syndromic SNL(DFNA20/26)	AD	Bilateral HA
				Baraitser-Winter syndrome
6	Congenital	Symmetrical bilateral severe SNL	CHD7 (c.235A > T;p.Lys79*) in heterozygosis	CHARGE syndrome	AD	Bilateral HA
		CHARGE syndrome
7	Congenital	Symmetrical bilateral isolated profound SNL	GJB2 (c.35delG;p.G12Vfs*2) in homozygosis	Non-syndromic SNL(DFNB1)	AR	Bilateral CI
8	Congenital	Symmetrical bilateral isolated profound SNL	GJB2 (c.35delG;p.G12Vfs*2) in homozygosis	Non-syndromic SNL(DFNB1)	AR	Unilateral CI+HA
			KCNE1 (p.Asp76Asn) in heterozygosis	Long QT syndrome	AD
9	Infancy	Symmetrical bilateral moderate SNL	COL4A5 (c.3525_3529dupTGGAC;p.P1177Lfs*124) in hemicygosis	Alport syndrome	AR	Bilateral HA
		Alport syndrome
10	Congenital	Symmetrical bilateral isolated profound SNL	BSND (c.23 G > A;p.R8Q) in homozygosis	Non-syndromic SNL(DFNB73)	AR	Unilateral CI+HA
				Bartter syndrome Type 4
11	Congenital	Unilateral profound isolated SNL	MITF (c.909 G > A;p.T303 T) in heterozygosis	Waardenburg syndrome type 2A	AD	Control
12	Congenital	Asymmetrical bilateral moderate-severe isolated SNL	GJB2 (c.35delG;p.G12Vfs*2) in homozygosis	Non-syndromic SNL(DFNB1)	AR	Bilateral HA
13	Infancy	Symmetrical bilateral moderate isolated SNL	P2RX2 (c.178 G > T;p.Val60Leu) in heterozygosis	Non-syndromic SNL(DFNA41)	AR	Bilateral HA
14	Infancy	Symmetrical bilateral mild-moderate isolated SNL	Compound heterozygote GJB2 (c.35delG;p.G12Vfs*2) and GJB2 (c.101 T > C; p.M34 T)	Non-syndromic SNL(DFNB1)	AR	Control
15	Infancy	Symmetrical bilateral moderate isolated SNL	Deletion of all the exons of the STRC gene in homozygosis (the deletion also includes the CATSPER2 gene)	Non-syndromic SNL(DFNB16)	AR	Bilateral HA
			Heterozygous TECTA	Non-syndromic SNL(DFNA 8/12)	AD

AD: autosomal dominant; AR: autosomal recessive; SNL: sensorineural hearing loss; CI: cochlear implant; P: pathogenic; HA: hearing aid; LP: Likely pathogenic.

Discussion

By about 4 years of age, neural plasticity is virtually complete. Hence the importance of stimulation as early as possible to achieve adequate development of both cognitive and oral communication.¹⁴ Obtaining an aetiological diagnosis is extremely important, as it has multiple advantages for the patient and their families.¹⁶ The results of our pilot study not only demonstrate that NGS now enables genetic diagnostics to be included in clinical practice, but also provide examples of the benefits deriving from it. Likewise, our series, despite its limited size, faithfully reflects the genetic characteristics of hereditary hearing loss (genetic heterogeneity, variable expressiveness, incomplete penetrance), characteristics that have classically proven a challenge to the conventional sequencing technologies (Sanger sequencing). Moreover, our results provide us with a first approach to the “molecular epidemiology” of our population with early-onset SNL, which is vitally important information for prevention, diagnosis and treatment.

The diagnostic performance of the OTOgenics™ panel in our cohort was 56% (15/27). The 12 potentially genetic cases in which no variant responsible for the hearing loss was identified have no aetiological diagnosis as yet. This could be due to an unidentified environmental cause or to the limitations of the technology used (large deletions/duplications, rearrangements, variants in genomic regions not yet involved in SNL, etc.). The diagnostic performance increases when we assess cases with bilateral SNL exclusively, reaching 59% (13/22).

In our series, the type of inheritance differs from that described in the literature. We identified 67% of cases with AR inheritance and 33% with AD inheritance. Since more than half (3/5) of the AD cases were caused by de novo variants (GATA3, ACTG1 and COL4A5), the fact that we used an agnostic panel capable of identifying unexpected de novo variants explains these results.¹⁵ This information is vitally important, since in the absence of a family history of hearing loss a priori a pattern of AR inheritance could have been suspected, yet it should be borne in mind that in our environment de novo mutations are a relatively frequent cause of SNL

One of the fundamental questions when planning a strategy to approach the aetiological diagnosis of SNL is to establish the percentage of cases that can be explained by mutations in the genes that make up the DFNB1 locus (GJB2 and GJB6). Screening prior to the use of NGS technology will be justified depending on this frequency, by analysing these genes using the conventional methods (Sanger sequencing of GJB2 and specific PCR of large frequent deletions of GJB6). P variants at this locus may be responsible for up to 50% of early-onset SNL cases. However, their prevalence is highly variable depending on the population, ranging from 0% to 60%. Variant P GJB2 c.35delG is the most frequent cause of autosomal recessive SNL in the Caucasian population.^17,18 In our population there is a frequency of 18.5% of cases that are explained by mutations in the DFNB1 locus, which, a priori, would justify pre-testing prior to analysis using a gene panel.¹⁴

One of the characteristics of hereditary SNL that adds difficulty to its diagnosis and treatment is its variable expressiveness.^2,4,6,17 In our series we find several examples that reflect this characteristic, starting with the GJB2 gene. Different degrees of hearing loss have been described even with the same genetic alteration, and certain alterations of this gene can also be the cause of DFNA3A (SNL with AD inheritance) and of syndromes of greater or lesser severity.^18,19

In addition to its utility for genetic counselling and family planning, sometimes the benefits of genetic testing extend further and provide us with unexpected information (secondary findings). Thus, proband 8 presented a P variant in heterozygosis in the paternally inherited KCNE1 gene, which has been associated with long QT syndrome. Silent carriers of P mutations carry a risk of cardiac events. Treatment with beta-blockers should be considered, avoiding drugs that prolong the QT interval and correcting possible electrolyte abnormalities.⁶

Hidden syndromes are another example of predictive medicine. Three hidden syndromes were identified (probands 2, 4 and 11, SOX10, GATA3 and MITF genes) and three other genetic alterations that could be associated with both syndromic and non-syndromic SNL (probands 3, 5 and 10, CDH23, ACTG1 and BSND genes). All these genetic diagnoses determine the clinical management of these patients.

The limitations inherent to this technology are an issue that must be considered when assessing the results of NGS-based genetic testing, and whether the necessary precautions have been taken in the test to minimise them. This is the case with the STRC gene, identified as responsible for the SNL of proband 15. The similarity between STRC and STRCP1 (its pseudogene) is exceptional, both are homologous in 98.9% (in exons and introns) and identical in > 99% (only a coding sequence), for this reason, the analysis of this gene represents a great challenge for the standard methodologies of NGS. The methodological difficulties required to avoid the interference that may be generated by pseudogenes could justify the current hypothesis that STRC is one of the genes that contributes most to autosomal recessive HNS20.²⁰

Finally, our results show that genetic testing can provide clinically relevant information in the case of unilateral SNL as well, although in our pilot study its performance is inferior to that obtained with bilateral SNL (40% versus 62%). Unilateral SNL has become more important since the approval of neonatal deafness screening, as it was previously an under-diagnosed problem. It is believed to have a prevalence of 1.7/1000. It is now known that loss of binaurality can have negative implications for language and speech development, behaviour and academic merit. Unilateral SNL has often been associated with inner ear malformations, and therefore imaging tests are always recommended as well as genetic testing.^21–23

The genetic test that we carried out in this pilot study entailed an approximate cost of less than 2000 Euros per patient, including the technical part plus the interpretation and reporting of the results. Approximately 10 h were dedicated to preparing the report which includes: interpretation of the data and its comparison with the previous data published in the databases. Genome sequencing is a cost-effective technique in the paediatric population, increasing the diagnosis rate to 16%–79% and decreasing the cost by 11 %–64% compared to the standard diagnostic route.²⁴

Our pilot study not only demonstrates that it is possible and useful to implement genetic diagnosis in clinical practice, but also that a first agnostic and systematic approach to the genetic bases of our SNL patient population provides clinically relevant advantages. These results should be validated and expanded in larger cohorts where our goal is to achieve an aetiological diagnosis in 100% of patients with early-onset SNL.

Conclusions

• •

  Sequencing a gene panel using NGS technology has made it possible to establish a genetic diagnosis in 56% of the cases evaluated (59% in the case of bilateral SNL), confirming that genetic testing has the highest diagnostic performance for early-onset SNL.

• •

  In our population, the percentage of P variants in the GJB2 and GJB6 genes is close to that which would justify screening prior to NGS sequencing by analysis of these genes with conventional technology.

• •

  The genetic SNL identified in our sample meets the specific characteristics of hereditary deafness (genetic heterogeneity, variable expressivity and incomplete penetrance).

Conflict of Interests

The authors have no conflict of interest to declare.

Acknowledgements

The work undertaken at IMOMA was financed by the María Cristina Masaveu Peterson Foundation (Spain).