Genetic Characteristics in Children with Cochlear Implants and the Corresponding Auditory Performance

Transcription

1 The Laryngoscope VC 2011 The American Laryngological, Rhinological and Otological Society, Inc. Genetic Characteristics in Children with Cochlear Implants and the Corresponding Auditory Performance Chen-Chi Wu, MD, PhD; Tien-Chen Liu, MD, PhD; Shih-Hao Wang, MD; Chuan-Jen Hsu, MD; Che-Ming Wu, MD Objectives/Hypothesis: To explore the genetic characteristics of children with cochlear implants (CIs) and to correlate the auditory performance after implantation to the genetic diagnosis of children with CIs. Study Design: Prospective cohort study. Methods: Mutations of four common deafness-associated genes, GJB2, SLC26A4, the mitochondrial 12S rrna gene, and OTOF, were screened in 743 unrelated children with idiopathic sensorineural hearing impairment, including 180 and 563 children with and without CIs, respectively. The allele frequencies and audiologic features were compared between both groups. The Categories of Auditory Performance (CAP) scores at 3 years after implantation were then analyzed according to the genotypes. Results: A definitive genetic diagnosis was made in 37 (20.6%) of the 180 CI children. A significant difference in allele frequencies between CI and non-ci children was found in GJB2 mutations (chi-square test, P <.01), but not in SLC26A4 mutations, mitochondrial 12S rrna mutations, or OTOF mutations (all P >.05). Further analysis revealed that the difference might have resulted from distinct audiological features in each group. Among the 110 CI children who had received more than 3 years of rehabilitation after implantation, the 35 children with mutations had better CAP scores than the 75 children without mutations. Conclusions: A significant prevalence of genetic mutations was identified in children with CIs, suggesting the need for routine genetic assessments. The frequencies of common deafness-associated mutations were different between children with and without CIs. The presence of genetic mutations was associated with an excellent long-term auditory performance outcome after implantation. Key Words: Cochlear implant, auditory performance, GJB2, SLC26A4, OTOF. Level of Evidence: 1b. Laryngoscope, 121: , 2011 INTRODUCTION Approximately 1 in 1,000 children suffer from severe to profound sensorineural hearing impairment (SNHI). 1 For these children, the technological development of multichannel cochlear implantation as well as improvement in implant surgery has made cochlear implants (CIs) an accepted and standard treatment. The benefits of cochlear implantation for spoken language, reading skills, and cognitive development have been From the Departments of Otolaryngology (C-C.W., T-C.L., S-H.W., C-J.H.); and Medical Genetics (C-C.W.), National Taiwan University Hospital, Taipei, Taiwan;, Department of Otolaryngology (C-M.W.), Chang Gung Memorial Hospital, Chang Gung University, Taoyuan, Taiwan. Editor s Note: This Manuscript was accepted for publication January 10, The authors have no financial relationships relevant to this article to disclose. The authors have no conflicts of interest to disclose. Send correspondence to Chuan-Jen Hsu, MD, Department of Otolaryngology, National Taiwan University Hospital, 7 Chung-Shan S. Road, Taipei, 100, Taiwan. cjhsu@ntu.edu.tw or Che-Ming Wu, MD, Department of Otolaryngology, Chang Gung Memorial Hospital, Chang Gung University, 5 Fu-Shin Street, Kueishan, Taoyuan, 333, Taiwan. bobwu506@hotmail.com DOI: /lary clearly demonstrated. 2 However, the variability of outcomes following implantation remains high. Many factors seem to contribute to the successful use of CIs, including the age of implantation, duration of implant use, residual hearing, primary mode of communication before operation, presence of inner ear malformations (IEMs), parent child interactions, and socioeconomic status. 2 4 Recently, the outcomes of CIs have also been correlated to genetic factors. Among children with SNHI, more than 50% of cases are attributable to a genetic etiology. 5 Genetic diagnosis provides direct clues about the pathogenesis of hearing loss and thus might be a major determining factor of the outcome with CIs. The performance after implantation has been reported in patients with different types of genetic deafness, including GJB2 (or Cx26) (Gene ID: 2706) mutations, 6 SLC26A4 (or PDS) (Gene ID: 5172) mutations, 4 mitochondrial mutations, 7 OTOF (Gene ID: 9381) mutations, 8 etc. However, most of these study series suffered from the limitation of a small case number. To optimize the clinical use of genetic diagnosis in predicting the outcome of CIs, two basic questions must be clarified: Which genetic mutations should be included in the examination panel for CI candidates? and How good is the auditory performance after implantation in 1287

2 children with a specific genetic mutation? In other words, if children with CIs reveal a different spectrum of genetic mutations from other SNHI children, then the examination panel should be adjusted accordingly to achieve an accurate and effective genetic assessment. Likewise, if the auditory performance of a specific genetic mutation is known, then the benefits of CIs could be predicted with precision based on the genetic diagnosis. To address these questions, we have investigated the prevalence of mutations in common deafnessassociated genes in a large cohort composed of children with CIs (hereinafter, CI children) and those without (non-ci children). The genetic characteristics of the CI children were then analyzed by comparison with those of the non-ci children, and the auditory performance after implantation was correlated to the genotypes of the CI children. METHODS Participant Recruitment and Clinical Evaluation From 2002 to 2010, a total of 743 unrelated children with idiopathic SNHI were enrolled in this study. All children were Han Chinese in ethnicity and came from families whose native language was Mandarin. For each child, a comprehensive family history, past history, physical examination, neurologic examination, audiologic results, and temporal bone imaging results were obtained and analyzed. Audiologic results were assessed and characterized in terms of two parameters: hearing levels and audiogram shapes. 9 The hearing level of the better ear of each subject, calculated by a four-tone average (0.5 k, 1 k, 2 k, and 4 k Hz), was labeled as mild (2040 dbhl), moderate (4170 dbhl), severe (7195 dbhl), or profound (>95 dbhl) hearing loss (GENDEAF: and served as a proxy for residual hearing. The temporal bone imaging results were obtained using high-resolution computed tomography (HRCT) and/or magnetic resonance imaging (MRI), and IEMs were determined according to the criteria in the literature. 9 Cochlear Implantation and Rehabilitation Of the 743 children enrolled, 180 children underwent cochlear implantation. The selection criteria for cochlear implantation were: 1) hearing loss was bilaterally severe to profound, and at least one ear had to be profoundly deaf; or 2) poor development or improvement of speech perception and production after using a properly fitted hearing aid for at least 6 months. Auditory neuropathy (AN)/auditory dyssynchrony (AD) was not an excluding criterion because it had been demonstrated that cochlear implantation offered benefit in children with AN/AD. 10 The side with poorer hearing was selected for cochlear implantation, and none of the 180 children received bilateral CIs. All of the 180 children were implanted with Nucleus 24 or Nucleus Freedom, and each patient received verbal education in mainstream schools or rehabilitation facilities after implantation. Genetic Examination All 743 children underwent mutation screening for four genes commonly associated with hearing impairment in the Taiwanese population: GJB2, 9 SLC26A4, 9 the mitochondrial 12S rrna gene, 9 and OTOF. 11 Mutation screening included both exons of GJB2, all the 21 exons of SLC26A4, the whole mitochondrial 12S rrna gene, and exon 41 of OTOF, which encompassed the common p.e1700q mutation in the Taiwanese population. For those who carried only one mutant GJB2 allele variant, the DNA samples were further studied for mutations in the coding region of the GJB6 (or Cx30) gene. Analyses of Mutation Frequencies, Genotypes, and Phenotypes To investigate the genetic characteristics in CI children, the frequencies of mutant alleles of each gene in the CI children were compared with those in the non-ci children. Audiologic features, including audiogram shapes and hearing levels, were then compared between the two groups of children. The genotypes of the CI children and the associated phenotypes were also analyzed. Evaluation of Auditory Performance in Children with CI The auditory performance in children with cochlear implantation was evaluated using the Categories of Auditory Performance (CAP). The CAP provides an ordinal scale of auditory receptive ability composed of eight categoric ratings that range from no awareness of environment to the use of telephone with known users, and is intended to reflect the real life progress of children in audition. The CAP scores were recorded during follow-up for every CI child in the present cohort. The CAP scores of the CI children were then correlated to their genotypes. Children with developmental difficulties were excluded first because additional developmental disabilities might affect language performance. 12 Based on our previous study, there was significant difference in CAP scores between 1 year and 3 years after implantation, but no significant difference was found between 3 years and 5 years after implantation. 13 Accordingly, the CAP score at 3 years was adopted as a parameter for the long-term outcome of cochlear implantation, and only children who received auditory verbal (re)habilitation for more than 3 years after implantation were included in the analysis. RESULTS Demographic Characteristics of the Subjects The demographic characteristics of the 743 subjects, inclusive of 180 CI children and 563 non-ci children, are summarized in Table I. In general, CI children and non-ci children showed an equal distribution in age (Student t-test, P >.05) and gender (chi-square test, P >.05). Forty-seven (26.1%) of the 180 CI children had at least another family member affected by SNHI (i.e., familial cases), whereas 148 (26.3%) of the 563 non-ci children were familial cases, with no difference showing between both groups (chi-square test, P >.05). Ten (5.6%) CI children and 32 (5.7%) non-ci children, respectively, were diagnosed with syndromic hearing loss. Common syndromes in the present series included Pendred syndrome (presence of goiter on physical examination þ deafness, n ¼ 8), branchio-oto-renal syndrome (n ¼ 8), Alport syndrome (pathological proof of renal biopsy þ deafness, n ¼ 4), Waardenburg syndrome (n ¼ 2), and Usher syndrome (retinitis pigmentosa þ deafness, n ¼ 2). Of note, the prevalence of Pendred 1288

3 TABLE I. Demographic Characteristics of CI Children and Non-CI Children. Variables CI Children (n ¼ 180) Non-CI Children (n ¼ 563) P-Value Age, mean 6 SD, years >.05* Male, n (%) 84 (46.7) 289 (51.3) >.05 Familial cases, n (%) 47 (26.1) 148 (26.3) >.05 Clinical diagnosis, n (%) >.05 Syndromic 10 (5.6) 32 (5.7) Nonsyndromic with IEMs 55 (30.6) 108 (19.2) Nonsyndromic without IEMs 115 (63.9) 423 (75.1) *Student t-test. Chi-square test. Chi-square test for heterogeneity in distribution of clinical diagnosis, df ¼ 2. IEM ¼ inner ear malformation; CI ¼ cochlear implant. syndrome might be underestimated because the present cohort was composed mainly of children younger than 10 years of age, whereas goiter of Pendred syndrome usually does not develop until the second decade of life. The remaining majority in both groups were clinically classified as having nonsyndromic hearing loss. Inner ear malformations were identified in 55 nonsyndromic CI children and 108 nonsyndromic non-ci children, respectively. The distributions of clinical diagnosis did not differ between CI and non-ci children in terms of syndromic deafness, nonsyndromic deafness with IEMs, and nonsyndromic deafness without IEMs (chi-square test for heterogeneity, df ¼ 2, P >.05). Genetic Mutations in CI and Non-CI Children The allele frequencies of common deafness-associated mutations in the CI children and non-ci children are shown in Table II. More common mutations included p.v37i and c.235delc of GJB2, as well as c.919 2A>G of SLC26A4. Mutations or deletions of GJB6 were not detected in children with 1 mutated GJB2 allele. With regard the mitochondrial 12S rrna gene, only the m.1555a>g mutation was identified in the present cohort. Eleven of the 743 children segregated the m.961deltþc(n) variant, but recent evidence has indicated that this variant is merely a polymorphism. 14 Other mutations of the mitochondrial 12S rrna gene, such as the m.1494c>t mutation, were not identified in the present cohort. The allele frequencies of GJB2 mutations, SLC26A4 mutations, the m.1555a>g mutation, and the OTOF p.e1700q mutation in CI children were 16.4%, 14.2%, 0.6%, and 1.1%, respectively, whereas those in non-ci children were 27.7%, 12.8%, 1.4%, and 0.8%, respectively. There was no difference in the frequencies of SLC26A4 mutations, mitochondrial 12S rrna mutations, and OTOF mutations between the two groups (chi-square test or Fisher s exact test, all P >.05). In contrast, the frequencies of GJB2 mutations were significantly higher in non-ci children than in CI children (chi-square test, P <.01). Specifically, the difference seemed attributable to the higher allele frequency of TABLE II. Comparison of Allele Frequencies of Common Mutations between CI Children and Non-CI Children. Genes/Variant Alleles Mutant Alleles No. (%) in CI Children* Mutant Alleles No. (%) in Non-CI Children P-Value GJB2 p.v37i 36 (10.0) 243 (21.6) <.01 c.235delc 19 (5.3) 57 (5.1) >.05 Others 4 (1.1) 12 (1.1) >.05 Total 59 (16.4) 312 (27.7) <.01 SLC26A4 c.919 2A>G 37 (10.3) 111 (9.9) >.05 Others 14 (3.9) 33 (2.9) >.05 Total 51 (14.2) 144 (12.8) >.05 Mito. 12S rrna m.1555a>g 1 (0.6) 8 (1.4) >.05 OTOF p.e1700q 4 (1.1) 9 (0.8) >.05 *360 GJB2, SLC26A4, and OTOF alleles, but 180 mitochondrial 12S rrna alleles GJB2, SLC26A4, and OTOF alleles, but 563 mitochondrial 12S rrna alleles. Chi-square test. Fisher s exact test. CI ¼ cochlear implant. p.v37i in non-ci children than in CI children (21.6% vs. 10.0%, chi-square test, P <.01). Comparison of Audiological Features between CI and Non-CI Children To elucidate the cause of the difference in mutation frequencies, we further analyzed the audiological features between CI and non-ci children. As Table III shows, the chi-square test for heterogeneity revealed that the distributions of audiogram shapes and hearing TABLE III. Comparison of Audiologic Features between CI Children and Non-CI Children. Variable CI Children (n ¼ 180) Non-CI Children (n ¼ 563) P-Value Audiogram shapes n (%) <.01* Sloping 58 (32.2) 262 (46.5) Flat type 122 (67.8) 259 (46.0) Low tone loss 0 (0) 13 (2.3) Midtone loss 0 (0) 11 (2.0) High tone loss 0 (0) 18 (3.2) Hearing levels, n (%) <.01 Mild (2040 dbhl) 0 (0) 99 (17.6) Moderate (4170 dbhl) 2 (1.1) 220 (39.1) Severe (7195 dbhl) 23 (12.8) 154 (27.4) Profound (>95 dbhl) 155 (86.1) 90 (16.0) *Chi-square test for heterogeneity in distribution of audiogram shapes between CI children and non-ci children, df ¼ 4, P <.01. Chi-square test for heterogeneity in distribution of hearing levels between CI children and non-ci children, df ¼ 3, p <.01. CI ¼ cochlear implant. 1289

4 TABLE IV. Genotypes of CI Children with Common Mutations and the Associated Phenotypes. Genotype No. of Patient Phenotype Highlights GJB2 c.235delc/c.235delc 6 None with IEM p.v37i/p.v37i 4 None with IEM, all with profound SNHI p.v37i/p.r143w 1 No IEM, profound SNHI c.235delc/c.176_191del16 1 No IEM c.235delc/c.299_300delat 1 No IEM p.v37i/wt 28* One with acquired SNHI, three were syndromic, nine with IEM c.235delc/wt 5 None with IEM p.v63l/wt 1 No IEM SLC26A4 c.919 2A>G/c.919 2A>G 12 Eight with EVA, four with EVA and incomplete partition of cochlea, one with goiter c.919 2A>G/p.S448L 2 One with EVA, one with EVA and incomplete partition of cochlea c.919 2A>G/p.K77I 1 EVA and incomplete partition of cochlea c.919 2A>G/c.1001þ5G>C 1 EVA and incomplete partition of cochlea c.919 2A>G/p.A387V 1 EVA c.919 2A>G/p.H723R 2 One with EVA, one with EVA and incomplete partition of cochlea p.t410m/p.t410m 1 EVA and incomplete partition of cochlea p.h723r/p.h723r 1 EVA c.919 2A>G/wt 6 Five with EVA, one with EVA and incomplete partition of cochlea, one with goiter c.916_917insg/wt 1 EVA p.t721m/wt 1 EVA and incomplete partition of cochlea p.h723r/wt 1 EVA and incomplete partition of cochlea Mito. 12S rrna homoplasmic m.1555a>g 1 No IEM, without exposure history to aminoglycosides OTOF p.e1700q / p.e1700q 2 Both with profound SNHI compatible with auditory neuropathy *Probably not associated with deafness given the high frequency of the GJB2 p.v37i allele in Taiwanese. IEM ¼ inner ear malformation; SNHI ¼ sensorineural hearing impairment; wt ¼ wild type; EVA ¼ enlarged vestibular aqueduct, CI ¼ cochlear implant. levels in CI children differed from those in non-ci children (both P <.01). Profound hearing loss was far more common in CI children than in non-ci children (86.1% vs. 16.0%, chi-square test, P <.01), whereas more non- CI children demonstrated mild to moderate hearing loss (56.7% vs. 1.1%, chi-square test, P <.01). The flat-type audiogram shape was also more common in CI children (67.8% vs. 46.0%, chi-square test, P <.01), which might have resulted from the homogeneously high hearing levels in CI children. Because recent studies indicated that the p.v37i mutation was associated with a milder phenotype, 15 the difference in audiologic features between the CI children and non-ci children, or more specifically, the milder hearing impairment in the non-ci children, might explain the discrepancy in the distribution of genetic mutations between both groups. Genotypes of the CI Children and the Associated Phenotypes Among the 180 CI children, 13 segregated 2 mutated GJB2 alleles (Table IV). These children included six who were homozygous for c.235delc, four who were homozygous for p.v37i, and three who were compound heterozygous for two different mutations. Twenty-eight children were shown to have one p.v37i allele. The SNHI experienced by these 28 children could not be explained by their heterozygosity, given the high frequency of the GJB2 p.v37i allele in Taiwanese individuals. 9 Moreover, among the 28 patients, 5 patients had 2 mutated SLC26A4 alleles, and 1 patient had 1 mutated SLC26A4 allele, indicating that the hearing impairment in these 6 patients might be attributed to mutations in other genes. Five children were found to have one c.235delc allele, and one child was heterozygous for p.v63l. With regard SLC26A4 mutations, 21 of the 180 CI children were found to have 2 mutated alleles, and 9 were found to have 1 mutated allele. In 9 children with only 1 mutated SLC26A4 allele, 5 children heterozygous for GJB2 c.235delc, and 1 child heterozygous for GJB2 p.v63l, the mutation cosegregated with the phenotype of hearing loss in affected family members, implying that the other allele of SLC26A4 or GJB2 in these heterozygotes might harbor an undetected, hidden mutation, as postulated by Kimberling. 16 Moreover, our previous study revealed that Taiwanese patients with PS or DFNB4 and only one mutated SLC26A4 allele could be considered to have a hidden mutation on the other SLC26A4 allele. 17 Alternatively, p.v63l of GJB2 alone might lead to hearing impairment, because 1290

5 TABLE V. Comparison of CAP at 3 Years after Implantation According to the Genotypes. Genotype No. of Patient Residual Hearing, dbhl Age at Implantation, Years CAP Scores at 3 Years After Implantation P-Value* GJB2 Two mutant alleles >.05 One mutant allele ¼.02 Combined >.05 SLC26A4 Two mutant alleles <.01 One mutant allele >.05 Combined <.01 OTOF p.e1700q/p.e1700q Total children with mutations <.01 The others *Wilcoxon rank sum test as compared to the others group. Children with the p.v37i/wt genotype were excluded because this genotype probably might not be related to deafness given the high frequency of the GJB2 p.v37i allele in Taiwanese. CAP ¼ Categories of Auditory Performance. many GJB2 mutations are inherited in an autosomal dominant manner (The Connexin-deafness homepage: davinci.crg.es/deafness). One CI child was diagnosed as having the m.1555a>g mutation, and two children were homozygous for the OTOF p.e1700q mutation. The phenotype highlights of children with specific genotypes are also summarized in Table IV. IEMs such as enlarged vestibular aqueducts (EVAs) and incomplete partition of the cochlea (or Mondini dysplasia) were specific to children with SLC26A4 mutations. In contrast, children with mutations of GJB2, mitochondrial 12S rrna, or OTOF did not reveal IEMs, except for those with the p.v37i/wt genotype. As mentioned before, this genotype might not be related to deafness in children heterozygous for p.v37i. Both children homozygous for the OTOF p.e1700q mutation showed profound SNHI and revealed audiological features compatible with auditory neuropathy. Comparison of CAP at 3 Years after Implantation According to Different Genotypes To investigate the roles of genetic diagnoses in predicting the auditory performance outcome with CIs, CAP scores at 3 years were compared according to the genotypes. Of the 180 CI children, 110 received rehabilitation for more than 3 years and were included in the analysis. Among them, a total of 35 children were diagnosed as having at least 1 genetic mutation, and no mutations were detected in the remaining 75 children (Table V). Mean CAP scores before implantation in the 35 and 75 children were 1.8 þ 2.0 and 1.6 þ 2.0, respectively, showing no difference between both groups (Wilcoxon rank sum test, P >.05). At 3 years after implantation, children with 2 mutated SLC26A4 alleles (n ¼ 18) demonstrated a better CAP score than children with no detected mutation (n ¼ 75) (Wilcoxon rank sum test, P <.01). Even if children with only one mutated SLC26A4 allele (n ¼ 4) were included, the children with SLC26A4 mutations (n ¼ 22) still revealed a significantly better CAP score (Wilcoxon rank sum test, P <.01). Children with GJB2 (n ¼ 12, including nine with two mutated alleles, two heterozygous for c.235delc, and one heterozygous for p.v63l) or OTOF (n ¼ 1) mutations also showed excellent CAP scores at 3 years after implantation, although the difference was not significant when compared with those without mutations, owing to the limited case numbers. In total, children with genetic mutations (n ¼ 35) had better CAP scores than those without mutations (n ¼ 75) at 3 years after implantation (Wilcoxon rank sum test, P <.01), even though both groups had similar residual hearing levels and had received implantation at a similar age. DISCUSSION In the present study, a definitive genetic diagnosis could be made in 37 (20.6%) of the 180 CI children, including 13 with 2 mutated GJB2 alleles, 21 with 2 mutated SLC26A4 alleles, 1 homoplasmic for m.1555a>g mutation, and 2 with 2 mutated OTOF alleles. In another 15 (8.3%) children, including 6 with 1 mutated GJB2 allele and 9 with 1 mutated SLC26A4 allele, SNHI probably also resulted from a genetic etiology owing to a second, as yet undetected, pathologic mutation. Sorted together, genetic factors contributed to SNHI in quite a few children with CIs. In addition, we identified that frequencies of common deafness-associated mutations were different between children with CIs and those without, and the difference seemed to arise from the distinct audiologic features in each group. Furthermore, the present study revealed that mutations of four common genes (i.e., GJB2, SLC26A4, mitochondrial 12S rrna, and OTOF) were associated with an excellent long-term auditory performance outcome after implantation. 1291

6 In addition to confirming the significant prevalence of genetic mutations in children with CIs, a clinical implication that the present study carries is to delineate the genetic characteristics in this subgroup of SNHI children. The genetic basis of idiopathic SNHI is particularly heterogeneous, making the efficient molecular diagnosis of individual patients challenging. To date, more than 100 genes are associated with deafness, and at least 46 genes have been identified to cause nonsyndromic hereditary hearing impairment (The Hereditary Hearing Loss Homepage, 18 Although mutations in certain genes have been shown to be much more prevalent than other genes in many populations, comprehensive screening of mutations in these genes remains a laborious task. To streamline the genetic screening for deafness-associated mutations, several high-throughput strategies like microarray technology, 19 the Invader assay, 20 the SNaPshot multiplex assays, 21 and massively parallel sequencing, 22 have been developed, and protocols for genetic testing have been established. 23 For all these new technologies and protocols, however, a thorough investigation in the genetic epidemiologic data of the specific target population is a prerequisite. As shown in the present study, the CI children revealed different common mutations compared with the non-ci children. In other words, if genetic examination is performed in children or candidates with CIs, a priority should be placed on these common mutations. Interestingly, among the 180 children, 4 were homozygous for the GJB2 p.v37i mutation and 1 was compound heterozygous for GJB2 p.v37i and p.r143w (Table IV). All the five children had profound SNHI, and the SNHI phenotype cosegregated with the GJB2 genotypes in the families. Although the p.v37i mutation has been linked to a milder phenotype in the literature, 15 more severe hearing loss (>80 dbhl) has been reported in two patients compound heterozygous for p.v37i and p.r143w in a Japanese series. 24 It is likely that p.v37i actually exhibits a wide range of hearing levels modulated by other factors. Despite a milder phenotype in the majority of the cases, a few outliers might develop profound SHNI instead and become candidates for cochlear implantation. Another clinical implication of the present study is to demonstrate that the presence of certain genetic mutations is associated with a good outcome with CIs. Satisfactory auditory performance outcomes have been documented in patients with various forms of hereditary hearing impairment, including GJB2 mutations, SLC26A4 mutations, mitochondrial mutations, OTOF mutations, Usher syndrome type I, DFNA9, and DFNA17. One reasonable explanation is that the pathogenic consequences of these genetic mutations are confined to the inner ear and spare the integrity of the auditory nerve and central auditory pathway, which are essential for the function of the CI. Consequently, the identification of patients with genetic mutations that exclusively involve the inner ear might assist in selecting CI candidates for whom an excellent outcome can be anticipated. To our knowledge, the present study is the largest series in the literature investigating the genetic characteristics and their contribution to the outcome in CI patients. However, some limitations of the study deserve discussion. First, because the present study was originally designed as a prospective cohort for exploring the genetic epidemiology, we did not have detailed, longitudinal records of auditory perception and expression performance for all the CI children. Nevertheless, a significant difference between the children with and without genetic mutations was indeed identified with regard to the long-term outcome after implantation. Second, a growing number of novel factors have recently been reported to influence the outcome with CIs, such as parent child interactions, socioeconomic status, and bilateral cochlear implantation. These factors were not included in the present analysis owing to the limitation of the study design. However, these factors appear unrelated to the genetic diagnosis, and it is thus reasonable to assume that there is no difference concerning these factors between the children with mutations and those without. Third, the power of the present study might also be compromised by the single ethnic background of the studied cohort, because common deafness genetic mutations differ remarkably among various populations. For instance, the c.35delg mutation alone may account for up to 70% of all GJB2 mutations in the Caucasian populations, 15 whereas p.v37i and c.235delc together comprise more than 90% of GJB2 mutations in Han Chinese, as demonstrated in the present study. Multicenter studies on the long-term results with CIs across populations might be warranted to validate the observations of the current study. CONCLUSIONS In conclusion, we have confirmed in the present study a significant prevalence of genetic mutations in children with CIs, suggesting the need for routine genetic assessments. The frequencies of common deafness-associated mutations were different between children with CIs and those without, and the difference might have arisen from the distinct audiologic features in each group. The presence of genetic mutations was associated with an excellent long-term auditory performance outcome after implantation. Acknowledgments This study was supported by research grants from the National Science Council of the Executive Yuan of the Republic of China (NSC B MY3) and National Taiwan University Hospital (NTUH ). We thank all subjects and their parents for participating in the present study. BIBLIOGRAPHY 1. Nadol JB Jr. Hearing loss. N Engl J Med 1993;329: Niparko JK, Tobey EA, Thal DJ, et al. Spoken language development in children following cochlear implantation. JAMA 2010;303: Geers A, Brenner C, Davidson L. Factors associated with development of speech perception skills in children implanted by age five. Ear Hear 2003;24:24S 35S. 1292

7 4. Wu CC, Lee YC, Chen PJ, Hsu CJ. Predominance of genetic diagnosis and imaging results as predictors in determining the speech perception performance outcome after cochlear implantation in children. Arch Pediatr Adolesc Med 2008;162: Smith RJ, Bale JF Jr., White KR. Sensorineural hearing loss in children. Lancet 2005;365: Dahl HH, Wake M, Sarant J, Poulakis Z, Siemering K, Blamey P. Language and speech perception outcomes in hearing-impaired children with and without connexin 26 mutations. Audiol Neurootol 2003;8: Sinnathuray AR, Raut V, Awa A, Magee A, Toner JG. A review of cochlear implantation in mitochondrial sensorineural hearing loss. Otol Neurotol 2003;24: Rouillon I, Marcolla A, Roux I, et al. Results of cochlear implantation in two children with mutations in the OTOF gene. Int J Pediatr Otorhinolaryngol 2006;70: Wu CC, Chen PJ, Chiu YH, Lu YC, Wu MC, Hsu CJ. Prospective mutation screening of three common deafness genes in a large Taiwanese Cohort with idiopathic bilateral sensorineural hearing impairment reveals a difference in the results between families from hospitals and those from rehabilitation facilities. Audiol Neurootol 2008;13: Rance G, Barker EJ. Speech perception in children with auditory neuropathy/dyssynchrony managed with either hearing aids or cochlear implants. Otol Neurotol 2008;29: Chiu YH, Wu CC, Lu YC, et al. Mutations in the OTOF gene in Taiwanese patients with auditory neuropathy. Audiol Neurootol 2010; 15: Meinzen-Derr J, Wiley S, Grether S, Choo DI. Language performance in children with cochlear implants and additional disabilities. Laryngoscope 2010;120: Wu CM, Sun YS, Liu TC. Long-term categorical auditory performance and speech intelligibility in Mandarin-speaking prelingually deaf children with early cochlear implantation in Taiwan. Clin Otolaryngol 2008;33: Kobayashi K, Oguchi T, Asamura K, et al. Genetic features, clinical phenotypes, and prevalence of sensorineural hearing loss associated with the 961delT mitochondrial mutation. Auris Nasus Larynx 2005;32: Snoeckx RL, Huygen PL, Feldmann D, et al. GJB2 mutations and degree of hearing loss: a multicenter study. Am J Hum Genet 2005;77: Kimberling WJ. Estimation of the frequency of occult mutations for an autosomal recessive disease in the presence of genetic heterogeneity: application to genetic hearing loss disorders. Hum Mutat 2005;26: Wu CC, Lu YC, Chen PJ, et al. Phenotypic analyses and mutation screening of the SLC26A4 and FOXI1 genes in 101 Taiwanese families with bilateral nonsyndromic enlarged vestibular aqueduct (DFNB4) or Pendred syndrome. Audiol Neurootol 2010;15: Hilgert N, Smith RJ, Van Camp G. Forty-six genes causing nonsyndromic hearing impairment: which ones should be analyzed in DNA diagnostics? Mutat Res 2009;681: Gardner P, Oitmaa E, Messner A, Hoefsloot L, Metspalu A, Schrijver I. Simultaneous multigene mutation detection in patients with sensorineural hearing loss through a novel diagnostic microarray: a new approach for newborn screening follow-up. Pediatrics 2006;118: Abe S, Yamaguchi T, Usami S. Application of deafness diagnostic screening panel based on deafness mutation/gene database using invader assay. Genet Test 2007;11: Wu CC, Lu YC, Chen PJ, Liu AY, Hwu WL, Hsu CJ. Application of SNaPshot multiplex assays for simultaneous multigene mutation screening in patients with idiopathic sensorineural hearing impairment. Laryngoscope 2009;119: Shearer AE, Deluca AP, Hildebrand MS, et al. Comprehensive genetic testing for hereditary hearing loss using massively parallel sequencing. Proc Natl Acad Sci USA 2010;107: Choi BY, Stewart AK, Nishimura KK, et al. Efficient molecular genetic diagnosis of enlarged vestibular aqueducts in East Asians. Genet Test Mol Biomarkers 2009;13: Tsukada K, Nishio S, Usami S. A large cohort study of GJB2 mutations in Japanese hearing loss patients. Clin Genet 2010;78: