Newborn Hearing Screening — A Silent Revolution
http://www.100md.com
《新英格兰医药杂志》
Hearing loss is an etiologically heterogeneous trait with many known genetic and environmental causes.1 Historically, some environmental causes of hearing loss, such as rubella embryopathy, have been epidemic in nature, establishing that the incidence of congenital deafness can vary widely at various times and among populations. Other important environmental causes of hearing loss include prematurity, prenatal and postnatal infections, head trauma, subarachnoid hemorrhage, and pharmacologic ototoxicity. Genetic causes account for at least 50 to 60 percent of childhood hearing loss in developed countries and can be classified according to the pattern of inheritance, the presence (syndromic) or absence (nonsyndromic) of distinctive clinical features, or the identification of the causal mutation. Because hearing is so critical to the normal development and acquisition of language, we review prelingual hearing loss, which is either present at birth or begins before the age of five years, when language has normally been acquired.
Prevalence of Hearing Loss in Newborns
The initiation of screening of newborns for hearing defects in this country can largely be attributed to the career-long efforts of the audiologist Marion Downs. As long ago as 1964,2 she showed that severe-to-profound hearing loss could reliably be detected by behavioral hearing screening of neonates. She found such losses in 17 of 17,000 infants,3 a figure identical to contemporary estimates for severe-to-profound bilateral hearing loss. For every such infant, one or two are born with lesser but clinically significant degrees of hearing loss.4
Before the implementation of universal newborn hearing screening — beginning in Rhode Island in 1989,5 Hawaii in 1990,6 and Colorado in 19937,8 — two important developments were required. The first was the application of objective noninvasive physiological tests for hearing loss that could be administered by nonprofessional personnel.9 The second was the demonstration that the early detection of hearing loss influences the educational outcome of affected infants.10 When these prerequisites were met, the endorsement of universal newborn hearing screening by the National Institutes of Health Consensus Development Conference in 1993 led to the gradual spread of programs throughout the nation during the next 10 years.
Two complementary screening techniques are now in widespread use: the automated auditory brain-stem response measures average neural response to a large number of repeated sound signals of the same pitch and intensity, whereas measurement of spontaneous or sound-induced otoacoustic emissions detects sound produced by movements of outer hair cells of the cochlea. Both methods have acceptable sensitivities and specificities and are often used together in two-stage screening protocols. However, in patients with auditory neuropathy, the hair-cell response — measured by the otoacoustic emissions — may be completely normal, whereas the auditory brain-stem response is abnormal because of asynchrony in the transmission of neural signals. Thus, two-stage screening protocols in which the auditory brain-stem response is used only to confirm abnormal otoacoustic-emissions responses may fail to detect this form of hearing loss. Differences in diagnostic criteria, completeness of follow-up, and screening protocols have contributed to the variation in the reported incidence of deafness at birth. In England, where compliance with confirmatory testing is high, permanent childhood hearing loss is defined as a bilateral sensorineural loss of 40 dB or more, and the reported incidence is 1.33 per 1000 newborns. In the United States, a sensorineural loss of 35 dB or more in either ear is typically the threshold used to identify patients who are referred for confirmatory testing, and because 30 to 40 percent of detected losses are unilateral, 1.86 per 1000 seems to be a reasonable overall estimate for the incidence at birth (Figure 1).11,12,13 Nonetheless, estimates vary and some may be higher,14 reflecting true etiologic, as well as diagnostic, differences. The prevalence of permanent sensorineural hearing loss continues to increase during childhood and reaches a rate of about 2.7 per 1000 children before the age of five years and 3.5 per 1000 during adolescence. For every 10 infants with permanent hearing loss at birth, similar losses develop in another 5 to 9 children before the age of nine years.12
Figure 1. Estimates of Causes of Deafness at Birth and at Four Years in the United States.
The incidence of deafness at birth in the United States, and its prevalence at four years of age, were obtained by adjusting estimates from the United Kingdom11,12 (where, in contrast with the United States, follow-up is nearly complete) to include unilateral hearing loss. The overall proportion of genetic cases at four years of age was estimated by sentinel phenotype analysis.13 Estimates for specific causes were obtained from previously published data (as documented in the Supplementary Appendix, available with the full text of this article at www.nejm.org). No studies were available in which universal newborn testing was performed for more than one specific cause of deafness in the same population sample. CMV denotes cytomegalovirus, mtA1555G the mitochondrial A1555G mutation, and EVA enlargement of the vestibular aqueduct.
When universal programs to screen newborns for hearing defects were first implemented, the test failure rates ranged from 2 to 4 percent. Among newborns failing the screening tests, 85 to 90 percent were determined later to have normal hearing; this was considered to be an acceptable performance standard for well-established programs. However, the high proportion of infants with normal hearing who failed screening led to criticism that unwarranted parental anxiety elicited by the test failure would outweigh the benefits of the program. To minimize such anxiety, the euphemism "refer" was adopted to characterize failed screening tests. With time, the testing characteristics of these programs have improved dramatically, and many hospitals now have a failure rate of less than 0.5 percent, only about half of which is due to infants who actually have normal hearing.
With the advent of newborn screening, the average age at which hearing loss is confirmed has dropped from 24 to 30 months to 2 to 3 months.15 Infants in whom remediation is begun within six months are able to maintain language and social and emotional development that is commensurate with their physical development, in striking contrast to those whose hearing loss is first detected after six months of age. This observation does not depend on the mode of communication: it applies to children who are identified early as having hearing loss and who speak, sign, or use both modes of communication.16
Given the improvements in the test characteristics of these screening programs, the term "refer" should be replaced with the word "fail" to describe the results of a screening test that indicates a hearing loss. The health care professional responsible for the screening program should meet with the parents of infants who fail the screening test to review its importance, reinforcing the need for prompt audiologic confirmation and emphasizing the potential for the normal development of language with prompt and appropriate intervention. In hospitals in which the personnel and facilities are available, audiologic confirmation before discharge is feasible for infants who fail the screening test. The initiation of programs that facilitate the immediate confirmation of a hearing loss would also permit prompt initiation of genetic evaluation, counseling, and testing and would serve as model programs for the delivery of these services.
Early hearing detection and intervention programs have been established in every state in the union, are mandated in at least 39 states, and provide audiologic screening for nearly 93 percent of all newborn infants.17,18 In addition to identifying infants who will benefit from early intervention, these programs should be able to provide valuable epidemiologic data on secular trends in the incidence of genetic and environmental causes of hearing loss, as well as variation in specific forms of hearing loss between populations. An extraordinarily successful newborn screening program has been implemented in Poland; as of the summer of 2004, 99 percent of infants were screened before leaving the hospital.19 Although newborn screening programs are under way in other European countries, none has had success similar to that in Poland. Better hearing for persons of all nations is an achievable, important goal, given that a disabling hearing impairment affects about 4.2 percent of the world's population, with two thirds of such persons living in developing countries.
Infection is an important contributing cause of deafness worldwide. Hearing disorders related to the acquired immunodeficiency syndrome, which include both sensorineural and conductive losses, overwhelm South Africa, and congenital rubella infection, often accompanied by hearing loss, is the most common birth defect in India.20
The success of newborn screening in developed countries has had enormous personal, societal, and economic benefit. Although finding the resources to implement solutions in developing countries is a major problem, standardizing screening and intervention programs for the detection and treatment of newborns remains an important goal.
Limitations of Existing Screening Programs
Existing universal screening programs to identify hearing defects in newborns in the United States still do not enjoy the extraordinarily high follow-up rates for positive test results that characterize most metabolic screening programs for newborns. Another limitation is that some forms of early-onset hearing loss are not apparent at birth. To address this issue, the Joint Committee on Infant Hearing has identified a series of 10 risk indicators that should prompt the continued monitoring of hearing status, even if the results of newborn screening are normal (Table 1).21 Most of the risk factors are highly relevant features of the clinical, family, or medical history, but it is important to recognize that many, and perhaps most, cases of late-onset hearing loss may not have any of these risk factors. Testing protocols should also be standardized to permit the comparison and aggregation of data from different sites and to avert the failure to identify infants with specific forms of hearing loss, such as auditory neuropathy.22 Finally, most screening programs have lacked an etiologic focus, which may compromise meaningful interpretation of the results of early intervention.
Table 1. Risk Indicators for Audiologic Monitoring for Progressive or Delayed-Onset Sensorineural Hearing Loss, Conductive Hearing Loss, or Both, in Infants (29 Days through 2 Years of Age) with Normal Hearing on Newborn Screening.
Genetic Causes of Syndromic Hearing Loss
More than 300 forms of syndromic hearing loss, in which distinctive associated clinical features are a constant (or at least an occasional) feature, have been characterized.23 In many cases several genes that can cause the same phenotype or a closely related one have been identified. In other cases, observed variation in the severity or clinical findings can be attributed to different mutations of the same gene, resulting in a genotype–phenotype correlation. Some of the main recognized forms of syndromic deafness are summarized in Table 2.
Table 2. Selected Forms of Syndromic Hearing Loss.
Among syndromic entities, Pendred's syndrome is a relatively common form of autosomal recessive deafness with an onset in infancy or early childhood. Although the specific step in the transport of iodide across the thyrocyte that is defective in Pendred's syndrome has now been identified, the resulting thyromegaly that defines the syndrome may not be apparent until adolescence or adult life, thus complicating attempts to anticipate the hearing loss from syndromic features.24 Structural abnormalities of the cochlea, ranging from Mondini malformations (the cochlea lacks the normal two and a half turns) to enlargement of the vestibular aqueduct, are almost invariably found; the latter can occur with deafness in the absence of thyroid disease.25 Patients with Pendred's syndrome typically carry two SLC26A4 mutations, but 61 percent of persons with nonsyndromic enlargement of the vestibular aqueduct are found to carry a single mutation,26 implying a risk ratio in excess of 30 for the 1.7 percent of the population who are heterozygous for SLC26A4. These observations suggest that some cases of nonsyndromic enlargement of the vestibular aqueduct result from an interaction between a single SLC26A4 mutation and a second mutation involving another gene.27 In a recent study, this disorder was found in 20.8 percent of 810 children with sensorineural hearing loss. Virtually all the children became symptomatic after birth, and the mean age at referral was 5.8 years.28 It seems reasonable to assume that at least a third, or 7 percent, of these children had prelingual deafness and would have benefited from early detection and intervention.
Genetic Causes of Nonsyndromic Hearing Loss
Most cases of genetic hearing loss are nonsyndromic (not associated with distinctive clinical features). The search for specific genes for nonsyndromic as well as syndromic forms of deafness has met with astonishing success, resulting in the identification of more than 110 chromosomal loci and at least 65 genes (available at http://webhost.ua.ac.be/hhh).
Nonsyndromic forms of hearing loss can also be distinguished by their pattern of inheritance and often by the age at onset, progression, and audiologic characteristics of the hearing loss or by associated otologic findings such as vestibular dysfunction. Four terms are used to distinguish different forms of nonsyndromic hearing loss. When new loci are mapped, those showing dominant transmission are designated by DFNA, followed by an accession number. DFNB refers to recessive forms, DFN to X-linked genes, and DFNM to genes that modify the expression of other genetic forms of hearing loss. When the causal genes are identified, their names are often used to supplement or replace the DFN terms. Examples of forms of dominant and recessive nonsyndromic hearing loss for which the genes have been identified are shown in Table 3 and Table 4, respectively.
Table 3. Cloned Autosomal Dominant Hearing-Loss Genes.
Table 4. Cloned Autosomal Recessive Hearing-Loss Genes.
Because of the degree of heterogeneity of hearing-loss genes, it came as a surprise that mutations in a single gene, GJB2, account for 30 to 50 percent of all cases of profound nonsyndromic hearing loss in many populations.29 GJB2 encodes connexin 26, a hexameric gap-junction protein widely expressed in supporting cells and connective tissues of the cochlea. The connexin hexamers on the surface of adjacent cells bind together to form intercellular channels purportedly allowing recycling of potassium ions from hair cells to the stria vascularis (Figure 2), where they are actively pumped back into the cochlear endolymph.30 The maintenance of a high endocochlear potential by potassium recycling is of critical importance for sound perception.
Figure 2. View of the Outer, Middle, and Inner Ear with a Cross-Sectional View of the Cochlear Duct and a View of Hair Cells.
The genes related to deafness and the locations of the products they encode are shown. The Hereditary Hearing Loss Homepage (http://webhost.ua.ac.be/hhh) gives an interactive view of the expression of these genes.
Although deafness caused by mutation of GJB2 was first thought to be profound in degree and congenital in onset, it is now known to show considerable phenotypic variation,31,32 including well-documented cases in which affected infants showed no hearing loss on screening. Because all such cases are retrospective, it is not clear whether the hearing in these infants was actually normal at birth or whether they had subclinical losses. Establishing the full range of expressivity of GJB2 mutations is an important research priority. Although more than 100 mutations involving GJB2 have been identified, a single variant, designated 35delG, accounts for up to 70 percent of all pathologic mutations in many populations.33
Most cases of genetic deafness result from mutations involving a single gene, but a small and growing number are being identified in which hearing loss is determined by mutations in two independent genes. For example, DFNB1 can result from two mutations involving GJB2, two mutations in the closely linked GJB6 gene, or a combination of mutations involving both genes (the combination of mutations accounts for about 8 percent of DFNB1 cases).33 GJB6 is a gene with sequence similarity to GJB2 and is also expressed in the cochlea; its product, connexin 30, can form heteromeric gap-junction channels with connexin 26 subunits, thus explaining the observed cases of digenic transmission. Digenic transmission also provides an explanation for a substantial proportion of deaf persons who are heterozygous for mutations in GJB2, in apparent contradiction to the rules of recessive inheritance.
The mating pattern of deaf persons who communicate by sign language is characterized by a very high frequency of marriages among the deaf (assortative mating), often extending over many generations, which brings together rare deafness genes of all types with a much greater frequency than would be observed if marriages occurred at random with respect to deafness. As the frequency of deafness in the ancestors of deaf persons increases, there is a progressive increase in the frequency of persons who by chance are heterozygous for DFNB1 but are deaf owing to some other genetic cause. Finally, rare DFNB1 mutations have been identified that show a dominant pattern of transmission and thus represent another explanation for deafness in persons who are heterozygous for GJB2. Most of these persons, however, have distinctive dermatologic findings and would be properly classified as having a syndromic form of deafness.
Cause of the High Prevalence of Deafness from Mutations in GJB2
There are several possible causes of the high frequency of GJB2 deafness, including recurrent mutation, population bottlenecks, heterozygous advantage, and founder effects (origin from a common ancestor).34 More recently, it has been suggested that a combination of the increased genetic fitness (i.e., fertility) of the deaf population that began when sign language was introduced in Western countries 400 years ago and the resulting linguistic homogamy (mate selection based on the mode of communication) that occurred at least in part from the establishment of residential schools for the deaf has resulted in a doubling of the frequency of GJB2 deafness in the United States during the past 200 years.35 Computer simulation studies have shown that this mechanism will preferentially amplify the most common recessive gene for deafness in a population, and a similar mechanism may well account for the extraordinary acceleration in human evolution that occurred after the first appearance of genes for speech 150,000 to 200,000 years ago.36
Environmental Causes of Prelingual Hearing Loss
Congenital cytomegalovirus infection has replaced rubella embryopathy as the most prevalent environmental cause of prelingual hearing loss in the United States. A retrospective study, using DNA extracted from blood spots from newborns, showed that 10 percent of infants with congenital hearing loss and 35 percent of those with moderate-to-severe late-onset loss were infected with cytomegalovirus at birth.37 The incidence of the infection in newborns varies with maternal age, parity, and socioeconomic status, ranging from 0.1 to 2 percent, and hearing loss is present at birth in 3.9 percent of all infants with the virus. However, 84 percent of newborns with congenital cytomegalovirus infection lack distinctive clinical findings so that the virus is not recognized as the cause in those with a hearing loss at birth. In children with clinically apparent and subclinical infections, the cumulative prevalence of hearing loss by the age of six years is 36 and 11 percent, respectively.38 The hearing loss can be unilateral, fluctuating, or progressive in nature, and its onset can be delayed for months or even years.39,40 In cases not recognized at birth, it can be difficult, if not impossible, to establish a retrospective diagnosis with certainty because seropositive infants may have been infected postnatally.
Pharmacologic ototoxicity is another important environmental cause of prelingual hearing loss. In the United States, 10 percent of persons attributing their deafness to this cause have mutations involving the mitochondrial 12S ribosomal (rRNA) gene including the A1555G substitution, which is associated with extreme sensitivity to aminoglycoside ototoxicity.41 One case in 20,000 to 40,000 births is a reasonable estimate of the prevalence of the A1555G mutation as a cause of prelingual deafness in the United States. In Spain, this mutation accounts for 15 to 20 percent of cases of familial nonsyndromic hearing loss, and hearing loss develops in many older family members even in the absence of documented exposure to aminoglycosides.42
Improving Detection of Late-Onset Prelingual Hearing Loss
Audiologic screening is available for hearing loss caused by hundreds of different genetic mutations, but it cannot detect forms of deafness that are not expressed at birth. The best way to detect cases of prelingual hearing loss that either are not present at birth or are associated with subclinical hearing losses would be to perform molecular genetic tests on blood spots from all newborns to identify those at risk for the most frequent causes of late-onset loss and to add infants to the group who should receive continued audiologic monitoring. Tests for GJB2 deafness and the mitochondrial A1555G mutation are commercially available, and although SLC26A4 is a large gene with 21 exons, 70 percent of persons who are heterozygous for Pendred's syndrome and 91 percent of those who are homozygous could readily be identified by the screening of a limited number of exons. If screening for these three causes of late-onset hearing loss was performed, together with a test for the presence of cytomegalovirus, we estimate that the follow-up of at-risk infants should result in the presymptomatic detection of nearly 60 percent of all infants in whom late-onset prelingual hearing loss develops, as well as an immediate etiologic diagnosis for at least 40 percent of those with congenital loss.
The use of molecular testing alone to detect genetic hearing loss is not currently feasible because of the many uncertainties that would arise in the interpretation of test results. As an adjunct to audiologic screening, however, such tests would provide a powerful and unprecedented strategy for identifying infants at risk for the development of prelingual hearing loss.
There are many benefits of moving from the mere detection of hearing loss to the identification of its cause.43 These include disease prevention, improved therapy, improved interpretation of the results of early intervention, and the psychological benefits of understanding the true cause of a disease. In general, cochlear implants have been successful in allowing speech to develop, even in children with profound hearing loss.44 In some cases the results have been disappointing, and a precise genetic diagnosis may help to identify infants in whom the procedure will not be as useful. As shown in Figure 2, a growing number of genetic causes for auditory neuropathy are being identified in addition to hyperbilirubinemia, and the improved detection of these genetic causes may lead to the development of more effective treatments. In addition, the successful induction of hair-cell regeneration in guinea pigs45 and gene silencing by RNA interference in the mouse provide support for the hope that specific forms of genetic deafness in humans may someday be addressed by similar forms of gene therapy. Such treatments would probably be disease-specific and require a precise etiologic diagnosis.
Despite the recommendations of professional organizations46 and the Joint Committee on Infant Hearing,21 in most screening programs systematic genetic evaluation and counseling are not a routine part of the approach to infants with confirmed hearing loss, even though specific genetic abnormalities are the most common cause of prelingual hearing loss and specific genetic tests are becoming available for a growing number of deafness genes. In the future, the use of molecular diagnostic DNA chips that are already being developed will permit routine, simultaneous testing for mutations involving many genes. When a specific form of syndromic deafness can be identified clinically, testing could be limited to the relevant set of mutations that are known to cause the syndrome. Positive test results are typically highly accurate, although ambiguities may exist in the interpretation of specific or newly recognized mutations. Negative results may not always rule out the diagnosis of the particular disorder or other genetic causes of deafness. Even if the test procedure involves DNA sequencing of coding regions of a deafness gene, regulatory mutations in noncoding regions cannot be ruled out as a potential cause or risk factor. Furthermore, the interpretation of test results for persons who are apparently heterozygous can often be problematic. When the goal of implementing extremely-low-cost genomic DNA sequencing (the so-called $1,000 genome sequence47) becomes a reality, many but not all of these ambiguities will be resolved. Despite current limitations, the ever-expanding use of diagnostic molecular testing of all infants identified with hearing loss is rapidly becoming the standard of care and represents an important advance in the clinical management of cases of deaf infants.
In summary, the astonishing spread of universal programs to screen newborns for hearing defects throughout the world has truly been a revolution in health care. Although these programs have been successful, they would greatly benefit from the standardization of testing protocols, the immediate confirmation of abnormal screening tests, the introduction of an etiologic focus, and the improved identification of infants at risk for late-onset prelingual hearing loss. By conducting molecular tests on all infants for just four important causes of hearing loss, the most common genetic and environmental causes of congenital deafness could be established shortly after birth, and infants at risk for the most common genetic, environmental, and preventable causes of late-onset prelingual hearing loss could be identified.
Supported by grants (DC03402, to Dr. Morton, and DC04293 and DC006707, to Dr. Nance) from the National Institute on Deafness and Other Communication Disorders for studies of hereditary deafness.
No potential conflict of interest relevant to this article was reported.
We are indebted to Dr. Karl R. White for his very helpful advice and careful reading of the manuscript.
Source Information
From the Departments of Obstetrics, Gynecology, and Reproductive Biology and Pathology, Brigham and Women's Hospital and Harvard Medical School — both in Boston (C.C.M.); and the Department of Human Genetics, Pediatrics, Medicine, and Otolaryngology, Medical College of Virginia, Virginia Commonwealth University, Richmond (W.E.N.).
Address reprint requests to Dr. Nance at the Department of Human Genetics, Sanger Hall, 11th floor, 1101 E. Marshall St., P.O. Box 980033, Richmond, VA 23298-0033, or at nance@mail2.vcu.edu.
References
Nance WE. The genetics of deafness. Ment Retard Dev Disabil Res Rev 2003;9:109-119.
Downs MP, Sterritt GM. Identification audiometry for neonates: a preliminary report. J Aud Res 1964;4:69-80.
Downs MP. The identification of congenital deafness. Trans Am Acad Ophthalmol Otolaryngol 1970;74:1208-1214.
Vohr B. Infants and children with hearing loss -- part 2: overview. Ment Retard Dev Disabil Res Rev 2003;9:218-219.
White KR, Vohr BR, Behrens TR. Universal newborn hearing screening using transient evoked otoacoustic emissions: results of the Rhode Island Hearing Assessment Project. Semin Hearing 1993;14:18-29.
Johnson JL, Kuntz NL, Sia CC, White KR, Johnson RL. Newborn hearing screening in Hawaii. Hawaii Med J 1997;56:352-355.
Downs MP. Universal newborn hearing screening -- the Colorado story. Int J Pediatr Otorhinolaryngol 1995;32:257-259.
Mehl AL, Thomson V. The Colorado newborn hearing screening project, 1992-1999: on the threshold of effective population-based universal newborn hearing screening. Pediatrics 2002;109:E7-E7.
Hayes D. Screening methods: current status. Ment Retard Dev Disabil Res Rev 2003;9:65-72.
Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Language of early- and later-identified children with hearing loss. Pediatrics 1998;102:1161-1171.
Kennedy C, McCann D. Universal neonatal hearing screening moving from evidence to practice. Arch Dis Child Fetal Neonatal Ed 2004;89:F378-F383.
Fortnum HM, Summerfield AQ, Marshall DH, Davis AC, Bamford JM. Prevalence of permanent childhood hearing impairment in the United Kingdom and implications for universal neonatal hearing screening: questionnaire based ascertainment study. BMJ 2001;323:536-540.
Nance WE, Lim BG, Dodson KM. Importance of congenital cytomegalovirus infections as a cause for pre-lingual hearing loss. J Clin Virol 2006;35:221-225.
Massachusetts Universal Newborn Hearing Screening Program. 2004 annual report. (Accessed April 18, 2006, at http://www.mass.gov/dph/fch/unhsp/hear_screen_04_report.pdf.)
Harrison M, Roush J, Wallace J. Trends in age of identification and intervention in infants with hearing loss. Ear Hear 2003;24:89-95.
Yoshinaga-Itano C. Early intervention after universal neonatal hearing screening: impact on outcomes. Ment Retard Dev Disabil Res Rev 2003;9:252-266.
White KR. The current status of EHDI programs in the United States. Ment Retard Dev Disabil Res Rev 2003;9:79-88.
National Center for Hearing Assessment and Management (NCHAM). State UNHS summary statistics. (Accessed April 18, 2006, at http://www.infanthearing.org/status/unhsstate.html.)
Plenary session V: EHDI — a global perspective. In: 3rd Annual National EHDI Meeting, Washington, D.C., February 18–20, 2004. (Accessed April 18, 2006, at http://www.infanthearing.org/meeting/ehdi2004/presentations/index.html#PSV.)
Gray RF. Causes of deafness in schools for the deaf in Madras. Int J Pediatr Otorhinolaryngol 1989;18:97-106.
Joint Committee on Infant Hearing. 2000 Position statement. (Accessed May 1, 2006, at http://www.cdc.gov/ncbddd/ehdi/documents/jcihyr2000.pdf.)
Johnson JL, White KR, Widen JE, et al. A multicenter evaluation of how many infants with permanent hearing loss pass a two-stage otoacoustic emissions/automated auditory brainstem response newborn hearing screening protocol. Pediatrics 2005;116:663-672.
Toriello HV, Reardon W, Gorlin RJ. Hereditary hearing loss and its syndromes. Oxford, England: Oxford University Press, 2004.
Reardon W, Coffey R, Phelps PD, et al. Pendred syndrome -- 100 years of underascertainment? QJM 1997;90:443-447.
Campbell C, Cucci RA, Prasad S, et al. Pendred syndrome, DFNB4, and PDS/SLC26A4 identification of eight novel mutations and possible genotype-phenotype correlations. Hum Mutat 2001;17:403-411.
Pryor SP, Madeo AC, Reynolds JC, et al. SLC26A4/PDS genotype-phenotype correlation in hearing loss with enlargement of the vestibular aqueduct (EVA): evidence that Pendred syndrome and non-syndromic EVA are distinct clinical and genetic entities. J Med Genet 2005;42:159-165.
Hulander M, Kiernan AE, Blomqvist SR, et al. Lack of pendrin expression leads to deafness and expansion of the endolymphatic compartment in inner ears of Foxi1 null mutant mice. Development 2003;130:2013-2025.
Preciado DA, Lim LH, Cohen AP, et al. A diagnostic paradigm for childhood idiopathic sensorineural hearing loss. Otolaryngol Head Neck Surg 2004;131:804-809.
Denoyelle F, Marlin S, Weil D, et al. Clinical features of the prevalent form of childhood deafness, DFNB1, due to a connexin-26 gene defect: implications for genetic counselling. Lancet 1999;353:1298-1303.
Forge A, Becker D, Casalotti S, Edwards J, Marziano N, Nevill G. Gap junctions in the inner ear: comparison of distribution patterns in different vertebrates and assessment of connexin composition in mammals. J Comp Neurol 2003;467:207-231.
Murgia A, Orzan E, Polli R, et al. Cx26 deafness: mutation analysis and clinical variability. J Med Genet 1999;36:829-832.
Cryns K, Orzan E, Murgia A, et al. A genotype-phenotype correlation for GJB2 (connexin 26) deafness. J Med Genet 2004;41:147-154.
Pandya A, Arnos KS, Xia XJ, et al. Frequency and distribution of GJB2 (connexin 26) and GJB6 (connexin 30) mutations in a large North American repository of deaf probands. Genet Med 2003;5:295-303.
Van Laer L, Coucke P, Mueller RF, et al. A common founder for the 35delG GJB2 gene mutation in connexin 26 hearing impairment. J Med Genet 2001;38:515-518.
Nance WE, Liu XZ, Pandya A. Relation between choice of partner and high frequency of connexin-26 deafness. Lancet 2000;356:500-501.
Nance WE, Kearsey MJ. Relevance of connexin deafness (DFNB1) to human evolution. Am J Hum Genet 2004;74:1081-1087.
Barbi M, Binda S, Caroppo S, Ambrosetti U, Corbetta C, Sergi P. A wider role for congenital cytomegalovirus infection in sensorineural hearing loss. Pediatr Infect Dis J 2003;22:39-42.
Peckham CS, Stark O, Dudgeon JA, Martin JA, Hawkins G. Congenital cytomegalovirus infection: a cause of sensorineural hearing loss. Arch Dis Child 1987;62:1233-1237.
Fowler KB, McCollister FP, Dahle AJ, Boppana S, Britt WJ, Pass RF. Progressive and fluctuating sensorineural hearing loss in children with asymptomatic congenital cytomegalovirus infection. J Pediatr 1997;130:624-630.
Fowler KB, Dahle AJ, Boppana SB, Pass RF. Newborn hearing screening: will children with hearing loss caused by congenital cytomegalovirus infection be missed? J Pediatr 1999;135:60-64.
Fischel-Ghodsian N, Prezant TR, Chaltraw WE, et al. Mitochondrial gene mutation is a significant predisposing factor in aminoglycoside ototoxicity. Am J Otolaryngol 1997;18:173-178.
del Castillo FJ, Rodriguez-Ballesteros M, Martin Y, et al. Heteroplasmy for the 1555A>G mutation in the mitochondrial 12S rRNA gene in six Spanish families with non-syndromic hearing loss. J Med Genet 2003;40:632-636.
Schimmenti LA, Martinez A, Fox M, et al. Genetic testing as part of the early hearing detection and intervention (EHDI) process. Genet Med 2004;6:521-525.
Geers AE. Speech, language, and reading skills after early cochlear implantation. Arch Otolaryngol Head Neck Surg 2004;130:634-638.
Izumikawa M, Minoda R, Kawamoto K, et al. Auditory hair cell replacement and hearing improvement by Atoh1 gene therapy in deaf mammals. Nat Med 2005;11:271-276.
American College of Medical Genetics. Genetics evaluation guidelines for the etiologic diagnosis of congenital hearing loss: Genetic Evaluation of Congenital Hearing Loss Expert Panel -- ACMG statement. Genet Med 2002;4:162-171.
Service RF. Gene sequencing: the race for the $1000 genome. Science 2006;311:1544-1546.(Cynthia C. Morton, Ph.D.,)
Prevalence of Hearing Loss in Newborns
The initiation of screening of newborns for hearing defects in this country can largely be attributed to the career-long efforts of the audiologist Marion Downs. As long ago as 1964,2 she showed that severe-to-profound hearing loss could reliably be detected by behavioral hearing screening of neonates. She found such losses in 17 of 17,000 infants,3 a figure identical to contemporary estimates for severe-to-profound bilateral hearing loss. For every such infant, one or two are born with lesser but clinically significant degrees of hearing loss.4
Before the implementation of universal newborn hearing screening — beginning in Rhode Island in 1989,5 Hawaii in 1990,6 and Colorado in 19937,8 — two important developments were required. The first was the application of objective noninvasive physiological tests for hearing loss that could be administered by nonprofessional personnel.9 The second was the demonstration that the early detection of hearing loss influences the educational outcome of affected infants.10 When these prerequisites were met, the endorsement of universal newborn hearing screening by the National Institutes of Health Consensus Development Conference in 1993 led to the gradual spread of programs throughout the nation during the next 10 years.
Two complementary screening techniques are now in widespread use: the automated auditory brain-stem response measures average neural response to a large number of repeated sound signals of the same pitch and intensity, whereas measurement of spontaneous or sound-induced otoacoustic emissions detects sound produced by movements of outer hair cells of the cochlea. Both methods have acceptable sensitivities and specificities and are often used together in two-stage screening protocols. However, in patients with auditory neuropathy, the hair-cell response — measured by the otoacoustic emissions — may be completely normal, whereas the auditory brain-stem response is abnormal because of asynchrony in the transmission of neural signals. Thus, two-stage screening protocols in which the auditory brain-stem response is used only to confirm abnormal otoacoustic-emissions responses may fail to detect this form of hearing loss. Differences in diagnostic criteria, completeness of follow-up, and screening protocols have contributed to the variation in the reported incidence of deafness at birth. In England, where compliance with confirmatory testing is high, permanent childhood hearing loss is defined as a bilateral sensorineural loss of 40 dB or more, and the reported incidence is 1.33 per 1000 newborns. In the United States, a sensorineural loss of 35 dB or more in either ear is typically the threshold used to identify patients who are referred for confirmatory testing, and because 30 to 40 percent of detected losses are unilateral, 1.86 per 1000 seems to be a reasonable overall estimate for the incidence at birth (Figure 1).11,12,13 Nonetheless, estimates vary and some may be higher,14 reflecting true etiologic, as well as diagnostic, differences. The prevalence of permanent sensorineural hearing loss continues to increase during childhood and reaches a rate of about 2.7 per 1000 children before the age of five years and 3.5 per 1000 during adolescence. For every 10 infants with permanent hearing loss at birth, similar losses develop in another 5 to 9 children before the age of nine years.12
Figure 1. Estimates of Causes of Deafness at Birth and at Four Years in the United States.
The incidence of deafness at birth in the United States, and its prevalence at four years of age, were obtained by adjusting estimates from the United Kingdom11,12 (where, in contrast with the United States, follow-up is nearly complete) to include unilateral hearing loss. The overall proportion of genetic cases at four years of age was estimated by sentinel phenotype analysis.13 Estimates for specific causes were obtained from previously published data (as documented in the Supplementary Appendix, available with the full text of this article at www.nejm.org). No studies were available in which universal newborn testing was performed for more than one specific cause of deafness in the same population sample. CMV denotes cytomegalovirus, mtA1555G the mitochondrial A1555G mutation, and EVA enlargement of the vestibular aqueduct.
When universal programs to screen newborns for hearing defects were first implemented, the test failure rates ranged from 2 to 4 percent. Among newborns failing the screening tests, 85 to 90 percent were determined later to have normal hearing; this was considered to be an acceptable performance standard for well-established programs. However, the high proportion of infants with normal hearing who failed screening led to criticism that unwarranted parental anxiety elicited by the test failure would outweigh the benefits of the program. To minimize such anxiety, the euphemism "refer" was adopted to characterize failed screening tests. With time, the testing characteristics of these programs have improved dramatically, and many hospitals now have a failure rate of less than 0.5 percent, only about half of which is due to infants who actually have normal hearing.
With the advent of newborn screening, the average age at which hearing loss is confirmed has dropped from 24 to 30 months to 2 to 3 months.15 Infants in whom remediation is begun within six months are able to maintain language and social and emotional development that is commensurate with their physical development, in striking contrast to those whose hearing loss is first detected after six months of age. This observation does not depend on the mode of communication: it applies to children who are identified early as having hearing loss and who speak, sign, or use both modes of communication.16
Given the improvements in the test characteristics of these screening programs, the term "refer" should be replaced with the word "fail" to describe the results of a screening test that indicates a hearing loss. The health care professional responsible for the screening program should meet with the parents of infants who fail the screening test to review its importance, reinforcing the need for prompt audiologic confirmation and emphasizing the potential for the normal development of language with prompt and appropriate intervention. In hospitals in which the personnel and facilities are available, audiologic confirmation before discharge is feasible for infants who fail the screening test. The initiation of programs that facilitate the immediate confirmation of a hearing loss would also permit prompt initiation of genetic evaluation, counseling, and testing and would serve as model programs for the delivery of these services.
Early hearing detection and intervention programs have been established in every state in the union, are mandated in at least 39 states, and provide audiologic screening for nearly 93 percent of all newborn infants.17,18 In addition to identifying infants who will benefit from early intervention, these programs should be able to provide valuable epidemiologic data on secular trends in the incidence of genetic and environmental causes of hearing loss, as well as variation in specific forms of hearing loss between populations. An extraordinarily successful newborn screening program has been implemented in Poland; as of the summer of 2004, 99 percent of infants were screened before leaving the hospital.19 Although newborn screening programs are under way in other European countries, none has had success similar to that in Poland. Better hearing for persons of all nations is an achievable, important goal, given that a disabling hearing impairment affects about 4.2 percent of the world's population, with two thirds of such persons living in developing countries.
Infection is an important contributing cause of deafness worldwide. Hearing disorders related to the acquired immunodeficiency syndrome, which include both sensorineural and conductive losses, overwhelm South Africa, and congenital rubella infection, often accompanied by hearing loss, is the most common birth defect in India.20
The success of newborn screening in developed countries has had enormous personal, societal, and economic benefit. Although finding the resources to implement solutions in developing countries is a major problem, standardizing screening and intervention programs for the detection and treatment of newborns remains an important goal.
Limitations of Existing Screening Programs
Existing universal screening programs to identify hearing defects in newborns in the United States still do not enjoy the extraordinarily high follow-up rates for positive test results that characterize most metabolic screening programs for newborns. Another limitation is that some forms of early-onset hearing loss are not apparent at birth. To address this issue, the Joint Committee on Infant Hearing has identified a series of 10 risk indicators that should prompt the continued monitoring of hearing status, even if the results of newborn screening are normal (Table 1).21 Most of the risk factors are highly relevant features of the clinical, family, or medical history, but it is important to recognize that many, and perhaps most, cases of late-onset hearing loss may not have any of these risk factors. Testing protocols should also be standardized to permit the comparison and aggregation of data from different sites and to avert the failure to identify infants with specific forms of hearing loss, such as auditory neuropathy.22 Finally, most screening programs have lacked an etiologic focus, which may compromise meaningful interpretation of the results of early intervention.
Table 1. Risk Indicators for Audiologic Monitoring for Progressive or Delayed-Onset Sensorineural Hearing Loss, Conductive Hearing Loss, or Both, in Infants (29 Days through 2 Years of Age) with Normal Hearing on Newborn Screening.
Genetic Causes of Syndromic Hearing Loss
More than 300 forms of syndromic hearing loss, in which distinctive associated clinical features are a constant (or at least an occasional) feature, have been characterized.23 In many cases several genes that can cause the same phenotype or a closely related one have been identified. In other cases, observed variation in the severity or clinical findings can be attributed to different mutations of the same gene, resulting in a genotype–phenotype correlation. Some of the main recognized forms of syndromic deafness are summarized in Table 2.
Table 2. Selected Forms of Syndromic Hearing Loss.
Among syndromic entities, Pendred's syndrome is a relatively common form of autosomal recessive deafness with an onset in infancy or early childhood. Although the specific step in the transport of iodide across the thyrocyte that is defective in Pendred's syndrome has now been identified, the resulting thyromegaly that defines the syndrome may not be apparent until adolescence or adult life, thus complicating attempts to anticipate the hearing loss from syndromic features.24 Structural abnormalities of the cochlea, ranging from Mondini malformations (the cochlea lacks the normal two and a half turns) to enlargement of the vestibular aqueduct, are almost invariably found; the latter can occur with deafness in the absence of thyroid disease.25 Patients with Pendred's syndrome typically carry two SLC26A4 mutations, but 61 percent of persons with nonsyndromic enlargement of the vestibular aqueduct are found to carry a single mutation,26 implying a risk ratio in excess of 30 for the 1.7 percent of the population who are heterozygous for SLC26A4. These observations suggest that some cases of nonsyndromic enlargement of the vestibular aqueduct result from an interaction between a single SLC26A4 mutation and a second mutation involving another gene.27 In a recent study, this disorder was found in 20.8 percent of 810 children with sensorineural hearing loss. Virtually all the children became symptomatic after birth, and the mean age at referral was 5.8 years.28 It seems reasonable to assume that at least a third, or 7 percent, of these children had prelingual deafness and would have benefited from early detection and intervention.
Genetic Causes of Nonsyndromic Hearing Loss
Most cases of genetic hearing loss are nonsyndromic (not associated with distinctive clinical features). The search for specific genes for nonsyndromic as well as syndromic forms of deafness has met with astonishing success, resulting in the identification of more than 110 chromosomal loci and at least 65 genes (available at http://webhost.ua.ac.be/hhh).
Nonsyndromic forms of hearing loss can also be distinguished by their pattern of inheritance and often by the age at onset, progression, and audiologic characteristics of the hearing loss or by associated otologic findings such as vestibular dysfunction. Four terms are used to distinguish different forms of nonsyndromic hearing loss. When new loci are mapped, those showing dominant transmission are designated by DFNA, followed by an accession number. DFNB refers to recessive forms, DFN to X-linked genes, and DFNM to genes that modify the expression of other genetic forms of hearing loss. When the causal genes are identified, their names are often used to supplement or replace the DFN terms. Examples of forms of dominant and recessive nonsyndromic hearing loss for which the genes have been identified are shown in Table 3 and Table 4, respectively.
Table 3. Cloned Autosomal Dominant Hearing-Loss Genes.
Table 4. Cloned Autosomal Recessive Hearing-Loss Genes.
Because of the degree of heterogeneity of hearing-loss genes, it came as a surprise that mutations in a single gene, GJB2, account for 30 to 50 percent of all cases of profound nonsyndromic hearing loss in many populations.29 GJB2 encodes connexin 26, a hexameric gap-junction protein widely expressed in supporting cells and connective tissues of the cochlea. The connexin hexamers on the surface of adjacent cells bind together to form intercellular channels purportedly allowing recycling of potassium ions from hair cells to the stria vascularis (Figure 2), where they are actively pumped back into the cochlear endolymph.30 The maintenance of a high endocochlear potential by potassium recycling is of critical importance for sound perception.
Figure 2. View of the Outer, Middle, and Inner Ear with a Cross-Sectional View of the Cochlear Duct and a View of Hair Cells.
The genes related to deafness and the locations of the products they encode are shown. The Hereditary Hearing Loss Homepage (http://webhost.ua.ac.be/hhh) gives an interactive view of the expression of these genes.
Although deafness caused by mutation of GJB2 was first thought to be profound in degree and congenital in onset, it is now known to show considerable phenotypic variation,31,32 including well-documented cases in which affected infants showed no hearing loss on screening. Because all such cases are retrospective, it is not clear whether the hearing in these infants was actually normal at birth or whether they had subclinical losses. Establishing the full range of expressivity of GJB2 mutations is an important research priority. Although more than 100 mutations involving GJB2 have been identified, a single variant, designated 35delG, accounts for up to 70 percent of all pathologic mutations in many populations.33
Most cases of genetic deafness result from mutations involving a single gene, but a small and growing number are being identified in which hearing loss is determined by mutations in two independent genes. For example, DFNB1 can result from two mutations involving GJB2, two mutations in the closely linked GJB6 gene, or a combination of mutations involving both genes (the combination of mutations accounts for about 8 percent of DFNB1 cases).33 GJB6 is a gene with sequence similarity to GJB2 and is also expressed in the cochlea; its product, connexin 30, can form heteromeric gap-junction channels with connexin 26 subunits, thus explaining the observed cases of digenic transmission. Digenic transmission also provides an explanation for a substantial proportion of deaf persons who are heterozygous for mutations in GJB2, in apparent contradiction to the rules of recessive inheritance.
The mating pattern of deaf persons who communicate by sign language is characterized by a very high frequency of marriages among the deaf (assortative mating), often extending over many generations, which brings together rare deafness genes of all types with a much greater frequency than would be observed if marriages occurred at random with respect to deafness. As the frequency of deafness in the ancestors of deaf persons increases, there is a progressive increase in the frequency of persons who by chance are heterozygous for DFNB1 but are deaf owing to some other genetic cause. Finally, rare DFNB1 mutations have been identified that show a dominant pattern of transmission and thus represent another explanation for deafness in persons who are heterozygous for GJB2. Most of these persons, however, have distinctive dermatologic findings and would be properly classified as having a syndromic form of deafness.
Cause of the High Prevalence of Deafness from Mutations in GJB2
There are several possible causes of the high frequency of GJB2 deafness, including recurrent mutation, population bottlenecks, heterozygous advantage, and founder effects (origin from a common ancestor).34 More recently, it has been suggested that a combination of the increased genetic fitness (i.e., fertility) of the deaf population that began when sign language was introduced in Western countries 400 years ago and the resulting linguistic homogamy (mate selection based on the mode of communication) that occurred at least in part from the establishment of residential schools for the deaf has resulted in a doubling of the frequency of GJB2 deafness in the United States during the past 200 years.35 Computer simulation studies have shown that this mechanism will preferentially amplify the most common recessive gene for deafness in a population, and a similar mechanism may well account for the extraordinary acceleration in human evolution that occurred after the first appearance of genes for speech 150,000 to 200,000 years ago.36
Environmental Causes of Prelingual Hearing Loss
Congenital cytomegalovirus infection has replaced rubella embryopathy as the most prevalent environmental cause of prelingual hearing loss in the United States. A retrospective study, using DNA extracted from blood spots from newborns, showed that 10 percent of infants with congenital hearing loss and 35 percent of those with moderate-to-severe late-onset loss were infected with cytomegalovirus at birth.37 The incidence of the infection in newborns varies with maternal age, parity, and socioeconomic status, ranging from 0.1 to 2 percent, and hearing loss is present at birth in 3.9 percent of all infants with the virus. However, 84 percent of newborns with congenital cytomegalovirus infection lack distinctive clinical findings so that the virus is not recognized as the cause in those with a hearing loss at birth. In children with clinically apparent and subclinical infections, the cumulative prevalence of hearing loss by the age of six years is 36 and 11 percent, respectively.38 The hearing loss can be unilateral, fluctuating, or progressive in nature, and its onset can be delayed for months or even years.39,40 In cases not recognized at birth, it can be difficult, if not impossible, to establish a retrospective diagnosis with certainty because seropositive infants may have been infected postnatally.
Pharmacologic ototoxicity is another important environmental cause of prelingual hearing loss. In the United States, 10 percent of persons attributing their deafness to this cause have mutations involving the mitochondrial 12S ribosomal (rRNA) gene including the A1555G substitution, which is associated with extreme sensitivity to aminoglycoside ototoxicity.41 One case in 20,000 to 40,000 births is a reasonable estimate of the prevalence of the A1555G mutation as a cause of prelingual deafness in the United States. In Spain, this mutation accounts for 15 to 20 percent of cases of familial nonsyndromic hearing loss, and hearing loss develops in many older family members even in the absence of documented exposure to aminoglycosides.42
Improving Detection of Late-Onset Prelingual Hearing Loss
Audiologic screening is available for hearing loss caused by hundreds of different genetic mutations, but it cannot detect forms of deafness that are not expressed at birth. The best way to detect cases of prelingual hearing loss that either are not present at birth or are associated with subclinical hearing losses would be to perform molecular genetic tests on blood spots from all newborns to identify those at risk for the most frequent causes of late-onset loss and to add infants to the group who should receive continued audiologic monitoring. Tests for GJB2 deafness and the mitochondrial A1555G mutation are commercially available, and although SLC26A4 is a large gene with 21 exons, 70 percent of persons who are heterozygous for Pendred's syndrome and 91 percent of those who are homozygous could readily be identified by the screening of a limited number of exons. If screening for these three causes of late-onset hearing loss was performed, together with a test for the presence of cytomegalovirus, we estimate that the follow-up of at-risk infants should result in the presymptomatic detection of nearly 60 percent of all infants in whom late-onset prelingual hearing loss develops, as well as an immediate etiologic diagnosis for at least 40 percent of those with congenital loss.
The use of molecular testing alone to detect genetic hearing loss is not currently feasible because of the many uncertainties that would arise in the interpretation of test results. As an adjunct to audiologic screening, however, such tests would provide a powerful and unprecedented strategy for identifying infants at risk for the development of prelingual hearing loss.
There are many benefits of moving from the mere detection of hearing loss to the identification of its cause.43 These include disease prevention, improved therapy, improved interpretation of the results of early intervention, and the psychological benefits of understanding the true cause of a disease. In general, cochlear implants have been successful in allowing speech to develop, even in children with profound hearing loss.44 In some cases the results have been disappointing, and a precise genetic diagnosis may help to identify infants in whom the procedure will not be as useful. As shown in Figure 2, a growing number of genetic causes for auditory neuropathy are being identified in addition to hyperbilirubinemia, and the improved detection of these genetic causes may lead to the development of more effective treatments. In addition, the successful induction of hair-cell regeneration in guinea pigs45 and gene silencing by RNA interference in the mouse provide support for the hope that specific forms of genetic deafness in humans may someday be addressed by similar forms of gene therapy. Such treatments would probably be disease-specific and require a precise etiologic diagnosis.
Despite the recommendations of professional organizations46 and the Joint Committee on Infant Hearing,21 in most screening programs systematic genetic evaluation and counseling are not a routine part of the approach to infants with confirmed hearing loss, even though specific genetic abnormalities are the most common cause of prelingual hearing loss and specific genetic tests are becoming available for a growing number of deafness genes. In the future, the use of molecular diagnostic DNA chips that are already being developed will permit routine, simultaneous testing for mutations involving many genes. When a specific form of syndromic deafness can be identified clinically, testing could be limited to the relevant set of mutations that are known to cause the syndrome. Positive test results are typically highly accurate, although ambiguities may exist in the interpretation of specific or newly recognized mutations. Negative results may not always rule out the diagnosis of the particular disorder or other genetic causes of deafness. Even if the test procedure involves DNA sequencing of coding regions of a deafness gene, regulatory mutations in noncoding regions cannot be ruled out as a potential cause or risk factor. Furthermore, the interpretation of test results for persons who are apparently heterozygous can often be problematic. When the goal of implementing extremely-low-cost genomic DNA sequencing (the so-called $1,000 genome sequence47) becomes a reality, many but not all of these ambiguities will be resolved. Despite current limitations, the ever-expanding use of diagnostic molecular testing of all infants identified with hearing loss is rapidly becoming the standard of care and represents an important advance in the clinical management of cases of deaf infants.
In summary, the astonishing spread of universal programs to screen newborns for hearing defects throughout the world has truly been a revolution in health care. Although these programs have been successful, they would greatly benefit from the standardization of testing protocols, the immediate confirmation of abnormal screening tests, the introduction of an etiologic focus, and the improved identification of infants at risk for late-onset prelingual hearing loss. By conducting molecular tests on all infants for just four important causes of hearing loss, the most common genetic and environmental causes of congenital deafness could be established shortly after birth, and infants at risk for the most common genetic, environmental, and preventable causes of late-onset prelingual hearing loss could be identified.
Supported by grants (DC03402, to Dr. Morton, and DC04293 and DC006707, to Dr. Nance) from the National Institute on Deafness and Other Communication Disorders for studies of hereditary deafness.
No potential conflict of interest relevant to this article was reported.
We are indebted to Dr. Karl R. White for his very helpful advice and careful reading of the manuscript.
Source Information
From the Departments of Obstetrics, Gynecology, and Reproductive Biology and Pathology, Brigham and Women's Hospital and Harvard Medical School — both in Boston (C.C.M.); and the Department of Human Genetics, Pediatrics, Medicine, and Otolaryngology, Medical College of Virginia, Virginia Commonwealth University, Richmond (W.E.N.).
Address reprint requests to Dr. Nance at the Department of Human Genetics, Sanger Hall, 11th floor, 1101 E. Marshall St., P.O. Box 980033, Richmond, VA 23298-0033, or at nance@mail2.vcu.edu.
References
Nance WE. The genetics of deafness. Ment Retard Dev Disabil Res Rev 2003;9:109-119.
Downs MP, Sterritt GM. Identification audiometry for neonates: a preliminary report. J Aud Res 1964;4:69-80.
Downs MP. The identification of congenital deafness. Trans Am Acad Ophthalmol Otolaryngol 1970;74:1208-1214.
Vohr B. Infants and children with hearing loss -- part 2: overview. Ment Retard Dev Disabil Res Rev 2003;9:218-219.
White KR, Vohr BR, Behrens TR. Universal newborn hearing screening using transient evoked otoacoustic emissions: results of the Rhode Island Hearing Assessment Project. Semin Hearing 1993;14:18-29.
Johnson JL, Kuntz NL, Sia CC, White KR, Johnson RL. Newborn hearing screening in Hawaii. Hawaii Med J 1997;56:352-355.
Downs MP. Universal newborn hearing screening -- the Colorado story. Int J Pediatr Otorhinolaryngol 1995;32:257-259.
Mehl AL, Thomson V. The Colorado newborn hearing screening project, 1992-1999: on the threshold of effective population-based universal newborn hearing screening. Pediatrics 2002;109:E7-E7.
Hayes D. Screening methods: current status. Ment Retard Dev Disabil Res Rev 2003;9:65-72.
Yoshinaga-Itano C, Sedey AL, Coulter DK, Mehl AL. Language of early- and later-identified children with hearing loss. Pediatrics 1998;102:1161-1171.
Kennedy C, McCann D. Universal neonatal hearing screening moving from evidence to practice. Arch Dis Child Fetal Neonatal Ed 2004;89:F378-F383.
Fortnum HM, Summerfield AQ, Marshall DH, Davis AC, Bamford JM. Prevalence of permanent childhood hearing impairment in the United Kingdom and implications for universal neonatal hearing screening: questionnaire based ascertainment study. BMJ 2001;323:536-540.
Nance WE, Lim BG, Dodson KM. Importance of congenital cytomegalovirus infections as a cause for pre-lingual hearing loss. J Clin Virol 2006;35:221-225.
Massachusetts Universal Newborn Hearing Screening Program. 2004 annual report. (Accessed April 18, 2006, at http://www.mass.gov/dph/fch/unhsp/hear_screen_04_report.pdf.)
Harrison M, Roush J, Wallace J. Trends in age of identification and intervention in infants with hearing loss. Ear Hear 2003;24:89-95.
Yoshinaga-Itano C. Early intervention after universal neonatal hearing screening: impact on outcomes. Ment Retard Dev Disabil Res Rev 2003;9:252-266.
White KR. The current status of EHDI programs in the United States. Ment Retard Dev Disabil Res Rev 2003;9:79-88.
National Center for Hearing Assessment and Management (NCHAM). State UNHS summary statistics. (Accessed April 18, 2006, at http://www.infanthearing.org/status/unhsstate.html.)
Plenary session V: EHDI — a global perspective. In: 3rd Annual National EHDI Meeting, Washington, D.C., February 18–20, 2004. (Accessed April 18, 2006, at http://www.infanthearing.org/meeting/ehdi2004/presentations/index.html#PSV.)
Gray RF. Causes of deafness in schools for the deaf in Madras. Int J Pediatr Otorhinolaryngol 1989;18:97-106.
Joint Committee on Infant Hearing. 2000 Position statement. (Accessed May 1, 2006, at http://www.cdc.gov/ncbddd/ehdi/documents/jcihyr2000.pdf.)
Johnson JL, White KR, Widen JE, et al. A multicenter evaluation of how many infants with permanent hearing loss pass a two-stage otoacoustic emissions/automated auditory brainstem response newborn hearing screening protocol. Pediatrics 2005;116:663-672.
Toriello HV, Reardon W, Gorlin RJ. Hereditary hearing loss and its syndromes. Oxford, England: Oxford University Press, 2004.
Reardon W, Coffey R, Phelps PD, et al. Pendred syndrome -- 100 years of underascertainment? QJM 1997;90:443-447.
Campbell C, Cucci RA, Prasad S, et al. Pendred syndrome, DFNB4, and PDS/SLC26A4 identification of eight novel mutations and possible genotype-phenotype correlations. Hum Mutat 2001;17:403-411.
Pryor SP, Madeo AC, Reynolds JC, et al. SLC26A4/PDS genotype-phenotype correlation in hearing loss with enlargement of the vestibular aqueduct (EVA): evidence that Pendred syndrome and non-syndromic EVA are distinct clinical and genetic entities. J Med Genet 2005;42:159-165.
Hulander M, Kiernan AE, Blomqvist SR, et al. Lack of pendrin expression leads to deafness and expansion of the endolymphatic compartment in inner ears of Foxi1 null mutant mice. Development 2003;130:2013-2025.
Preciado DA, Lim LH, Cohen AP, et al. A diagnostic paradigm for childhood idiopathic sensorineural hearing loss. Otolaryngol Head Neck Surg 2004;131:804-809.
Denoyelle F, Marlin S, Weil D, et al. Clinical features of the prevalent form of childhood deafness, DFNB1, due to a connexin-26 gene defect: implications for genetic counselling. Lancet 1999;353:1298-1303.
Forge A, Becker D, Casalotti S, Edwards J, Marziano N, Nevill G. Gap junctions in the inner ear: comparison of distribution patterns in different vertebrates and assessment of connexin composition in mammals. J Comp Neurol 2003;467:207-231.
Murgia A, Orzan E, Polli R, et al. Cx26 deafness: mutation analysis and clinical variability. J Med Genet 1999;36:829-832.
Cryns K, Orzan E, Murgia A, et al. A genotype-phenotype correlation for GJB2 (connexin 26) deafness. J Med Genet 2004;41:147-154.
Pandya A, Arnos KS, Xia XJ, et al. Frequency and distribution of GJB2 (connexin 26) and GJB6 (connexin 30) mutations in a large North American repository of deaf probands. Genet Med 2003;5:295-303.
Van Laer L, Coucke P, Mueller RF, et al. A common founder for the 35delG GJB2 gene mutation in connexin 26 hearing impairment. J Med Genet 2001;38:515-518.
Nance WE, Liu XZ, Pandya A. Relation between choice of partner and high frequency of connexin-26 deafness. Lancet 2000;356:500-501.
Nance WE, Kearsey MJ. Relevance of connexin deafness (DFNB1) to human evolution. Am J Hum Genet 2004;74:1081-1087.
Barbi M, Binda S, Caroppo S, Ambrosetti U, Corbetta C, Sergi P. A wider role for congenital cytomegalovirus infection in sensorineural hearing loss. Pediatr Infect Dis J 2003;22:39-42.
Peckham CS, Stark O, Dudgeon JA, Martin JA, Hawkins G. Congenital cytomegalovirus infection: a cause of sensorineural hearing loss. Arch Dis Child 1987;62:1233-1237.
Fowler KB, McCollister FP, Dahle AJ, Boppana S, Britt WJ, Pass RF. Progressive and fluctuating sensorineural hearing loss in children with asymptomatic congenital cytomegalovirus infection. J Pediatr 1997;130:624-630.
Fowler KB, Dahle AJ, Boppana SB, Pass RF. Newborn hearing screening: will children with hearing loss caused by congenital cytomegalovirus infection be missed? J Pediatr 1999;135:60-64.
Fischel-Ghodsian N, Prezant TR, Chaltraw WE, et al. Mitochondrial gene mutation is a significant predisposing factor in aminoglycoside ototoxicity. Am J Otolaryngol 1997;18:173-178.
del Castillo FJ, Rodriguez-Ballesteros M, Martin Y, et al. Heteroplasmy for the 1555A>G mutation in the mitochondrial 12S rRNA gene in six Spanish families with non-syndromic hearing loss. J Med Genet 2003;40:632-636.
Schimmenti LA, Martinez A, Fox M, et al. Genetic testing as part of the early hearing detection and intervention (EHDI) process. Genet Med 2004;6:521-525.
Geers AE. Speech, language, and reading skills after early cochlear implantation. Arch Otolaryngol Head Neck Surg 2004;130:634-638.
Izumikawa M, Minoda R, Kawamoto K, et al. Auditory hair cell replacement and hearing improvement by Atoh1 gene therapy in deaf mammals. Nat Med 2005;11:271-276.
American College of Medical Genetics. Genetics evaluation guidelines for the etiologic diagnosis of congenital hearing loss: Genetic Evaluation of Congenital Hearing Loss Expert Panel -- ACMG statement. Genet Med 2002;4:162-171.
Service RF. Gene sequencing: the race for the $1000 genome. Science 2006;311:1544-1546.(Cynthia C. Morton, Ph.D.,)