SPECIAL PAPER IN CELEBRATION OF PROF. YANG'S 50 YEARS CAREER IN MEDICINE

Transcription

1 JOURNAL OF OTOLOGY SPECIAL PAPER IN CELEBRATION OF PROF. YANG'S 50 YEARS CAREER IN MEDICINE Diagnosis and Treatment of Auditory Neuropathy and Related Research JI Fei, YANG Shiming Introduction Deafness is one of the most common otologic diseases and a major disease that greatly impacts the Chinese population. From the Second National Sample Survey of Disabled Persons, it is estimated that there are 27,800, 000 hearing disabled persons in China, about 2.14% of the population. Hearing disability continues to increase at a rate of 30,000 hearing impaired newborns per year and is at the top of the six major disabilities. In the clinic, different types of hearing loss are treated differently based upon our understanding of their pathogenic mechanisms and hearing-speech loss characteristics, but the recently noticed auditory neuropathy (AN) lacks thorough research and standard effective management [1]. AN is a clinically unique disease characterized by deterioration of temporal processing and intact outer hair cell (OHC) functions [2-4]. From the literature, AN may account for up to 10% of cases of severe or profound hearing loss [5]. Studies have shown that lesions in AN may involve inner hair cells (IHCs), synapses and auditory nerve trunk, but the main challenge now is the lack of clear localizing diagnosis which leads to uncertainties in its intervention [6]. Discovery and naming of AN In the 1980s, a few researchers noticed ABR waves and thresholds that were not explainable by pure tone audiometric results in a small number of patients. For example, Worthington et al reported 4 cases with reasonable audiograms but no ABRs or significantly inappropriate ABR thresholds [7]. Krause et al reported 7 cases in which audiometry showed normal hearing or moderate hearing loss with no history of brainstem injury, but ABR waves III and V were missing [8]. These cases generated little interests at the time. In 1991, Starr et al [2] reported one case in which ABRs were missing with moderate low frequency hearing loss, deteriorated speech discrimination and absent acoustic reflex, but normal CM and OAEs. Psychophysical studies showed impaired timing cue perception but unaffected intensity and frequency cue perceptions. The authors believed that the lesion locations in such cases might include the auditory nerve, synapses, IHCs and, to a small degree, OHCs. In 1992, GU Rui et al reported 16 cases of bilateral sensorineural hearing loss with unmatched speech discrimination and pure tone thresholds, and absent or abnormal ABRs. Except for the two cases in which cerebral atrophy or brainstem tumor were believed to the underlying etiology, there was no evidence of neurological pathologies in the rest 14 cases [9]. Starr et al conducted follow up studies on similar cases and reported 10 cases of seemingly intact cochlear function but deteriorated temporal auditory processing in In 8 of these cases (80% ), there was evidence of auditory nerve pathology [1]. Based on these works, Starr et al believed that the pathological process in these patients was similar to that in peripheral sensory neuropathy, including loss of afferent nerve fibers, demyelination and partial demyelination in surviving nerve fibers, and for the first time proposed the concept of AN in describing this group of patients [1]. The term fitted well in the clinical observations at the time. Increasing clinical evidence has shown that, among patients diagnosed with AN, the rate of accompanying neurological pathologies is significantly lower than that in Starr s initial report [5]. Researchers have later realized that lesion sites in these patients are not limited to the auditory nerve, but lesions at IHCs, synapses and the auditory nerve trunk can all lead to symptoms in AN and hair cells and synapses are probably more likely to be involved. This adds increased controversy in the naming of this type of disease. Rapin and Gravel believe that, Affiliation: Department of Otolaryngology Head and Neck Surgery, PLA General Hospital, Beijing PLA Institute of Otolaryngology 86 Corresponding authors: YANG Shi-ming, yangsm301@263.net

2 unless there is evidence of auditory nerve damage, the term "auditory neuropathy" may not be appropriate. They propose the use of a broader definition of "nerve conduction disease" [10]. Others have proposed using the double meaning term "auditory neuropathy/auditory dys-synchronization (or AN/AD)" [4, 11]. At the Internal Conference on Diagnosis and Intervention of Auditory Neuropathy in Italy in 2008, a guide line for differential diagnosis and intervention of AN in children was proposed, which named this class of diseases as"auditory neuropathy spectrum disorder"(or ANSD) [12, 13]. The two mainstream terms at this time are AN and ANSD. In consideration of its broad use in the clinic, AN is used in this report to refer to the condition in this group of patients. Epidemiology and etiology of AN Summarized in Table 1 are epidemiology reports regarding AN in the literature. It shows that the rate of AN is about 7% to 13.4% (average 10%) among children with confirmed diagnosis of severe or profound hearing loss [5, 14-19], with a matched gender ratio [19, 20]. When the average hearing screened newborn population is used, the rate falls blow 1% [17, 20]. Among all age populations or deaf-mute school students, average rate of AN is below 3% [21-24]. These data show that AN rate decreases as patient age increases. From this, it is speculated that AN may be a condition related to hereditary and/or peri-natal factors and may be present in about 10% of high risk children. It may be diagnosed at newborn times and should be suspected in children failing hearing screening or deemed at high risk. Etiologies leading to AN clinical presentations may include neonatal hyperbilirubinemia, degenerative neural diseases, metabolic neural diseases, hereditary motor neuron diseases, demyelination diseases, inflammatory neural diseases, ischemic/hypoxic neuropathy, cerebral edema, cerebral palsy, abnormal release of neurotransmitters, infectious diseases (e.g. parotiditis), autoimmune diseases and development retardation. From the latest literature, neonatal hyperbilirubinemia and hypoxia are the most important risk factors [18]. Gene mutation is another major cause for some AN phenotypes, especially in AN children with non-syndromic pre-lingual deafness. AN clinical audiology characteristics and diagnosis Currently, reported clinical characteristics vary greatly among AN patients. Clinically, compromise of auditory nerve activities are represented by abnormal or absent ABRs, absence or elevated thresholds in acoustic reflex with absent or weak contralateral inhibition, etc. Cochlear function indicators such as otoacoustic emissions (OAEs) and cochlear microphonic (CM) are usually normal. There are also unique psychophysical presentations. Clinical complaints or functional hearing loss in AN patients are reflected mainly through psychophysical Table 1. AN Prevalence in Literature Reports Year Authors Study Subjects AN/Total n Prevalence (%) 1999 Rance et al [20] Children with severe or profound hearing loss 1/ Siniger et al [26] Cheng et al [21] Patients in NIDCD sponsored newborn hearing screening Students in US deaf and mute schools Review 76/ Foerst et al [23] Children with severe SNHL 32/ Vlastarakos et al [24] Children with confirmed hearing loss Meta-analysis Sanyelbhaa [7] Children with severe or profound hearing loss 15/ Berlin et al [8] Children with hearing loss Review > Dowley et al [27] Children with severe SNHL 12/ Foerst et al [23] Average newborns 32/ Dowley et al [27] Average newborns 12/ Kumar et al [29] Adult SNHL patients 1/ Tang et al [30] Students in HK deaf and mute schools 3/ Lotfi et al [31] Students in deaf and mute schools 13/ Lee et al [32] Students in deaf and mute schools 2/

3 testing. Because psychophysical testing requires cooperation by the patient, its results are available only from older children and adults. Among AN patients reported in the literature, pure tone hearing loss ranges from mild to profound, although mild and moderate in most cases. Audiogram patterns are also variable, although dominated by ascending patterns [25]. Speech discrimination disorders are the most difficult for AN patients, typically showing unproportionally poor discrimination contrasted to pure tone thresholds, especially in noise [4, 26]. Middle ear muscle reflex and olivocochlear reflex are important to hearing in noise. These are compromised in AN patients, i.e. loss of efferent inhibition. This may explain the poor speech discrimination in noise. Zeng et al found that AN patients had decreased sensitivity to temporal envelop fluctuation and speculated that speech discrimination difficulties in AN patients might be related to compromised temporal processing [3]. Compromised timing cue perception may affect low frequency discrimination, temporal separation perception, timing integration, temporal modulation perception, forward and backward masking, signal recognition in noise, binaural beat, interaural temporal difference based sound localization, etc. On the other hand, intensity cue perception, such as intensity discrimination and interaural intensity difference based sound localization, is no different in AN patients from normal individuals [27]. Among existing diagnostic criteria for AN, the only widely accepted mandatory one is "absent or abnormal ABRs", which represents disruption to synchronicity in activities along the auditory pathway. All patients included in AN studies so far show this presentation. Abnormal ABRs in AN usually refer to inappropriate ABR morphology or thresholds not anticipated based upon pure tone thresholds. This is probably the most distinct clinical feature in AN. Another necessary presentation is "intact OHC function", as represented by existence of OAEs or CM or both. It is widely accepted that OAEs originate from OHCs, but whether CM is solely from OHCs or may contain contributions from IHCs is still being debated. Because of this, when OAEs are not present, CM may not be the definite indication of "intact OHC function". Some believe that the linearity of CM input/output (I/O) curve reflects the presence of OHCs [28] and that perfect linearity represents loss of OHC compression functions and only contributions from IHCs. On the other hand, non-linearity in CM I/O curves indicates intact OHC function, although this does not necessarily provide information on IHCs. Used together with ABR results, OAEs can be an important indicator of AN. However, linearity measurement has not been quantified and its relations with OHC function have not been confirmed by laboratory or clinical data, leaving rooms for debate over if AN can be diagnosed based on CM linearity data. Simultaneous presence of the two indicators discussed above is a necessary requirement (inclusion criteria) for differentiation from other sensorineural hearing loss and diagnosis of AN. Almost all AN subjects in the reported literature have been identified using these two conditions [5]. Combined testing of ABRs, CM and OAEs provides information regarding diseases locations along the auditory pathways including OHCs, auditory nerve and brainstem, although there still lacks means to identify diseases at IHCs and pre- or post-synaptic locations. Electrophysiological characteristics of AN The CM, summiting potential (SP), compound action potential (CAP) and central auditory evoked potentials (CAEPs) in AN patients all have distinct features that can be used to differentiate sub-types of AN and facilitate clinical diagnosis and intervention. Starr [29] and Santarelli [30] found that click induced CM in AN children is abnormally large in amplitude. SHI Wei et al [28] concluded that non-linearity in CM I/O curve indicates intact OHCs but is insufficient to determine the status of IHCs. Round window electrocochleography (RW ECoG) uses transtympanic electrode to record cochlear responses and provides increased details in information regarding the cochlea and VIIIth cranial nerve. Studying the SP and AP in RW ECoG in conjunction with observation of abnormal positive potential (APP) [30] or dentritic potential (DP) [31] may help locating the AN disease site. O'Leary et al were the first to report SP with delayed latencies and increased amplitudes (i.e. APP) on RW ECoG in some children with profound hearing loss [32]. Santarelli [30] and McMahon et al [31] later also recorded APP in some AN children. Santarelli divided RW ECoG findings in AN patients into three groups: 1) presence of both SP and CAP, indicating postsynaptic lesions at proximal auditory nerve endings; 2) presence of SP with absence of CAP, indicating pre-synaptic lesions at IHCs; and 3) increased SP amplitude with delayed latencies (or APP) and no CAP, representing pre-synaptic lesions, perhaps from compromised neurotransmitter release process [30]. McMahon et al [31] believed that APP represented pre-synaptic damage, but the mechanisms may be similar to that causing increased SP amplitude in patients with endolymphatic hydrops (i.e. bias shift of IHC cilia toward an "off" position). In their study, SP latency was normal, followed by a long negative potential, in some AN patients, which was considered as a DP indicating post-synaptic damage. In general, presence of APP appears to indicate pre-synaptic damage or damage to 88

4 the sensory cells, whereas obvious DP may indicate post-synaptic damage or damage to the auditory nerve. Therefore, RW ECoG wave analysis may help formulating audiological intervention strategies for AN patients. The 14 AN children in McMahon's report all received cochlear implant [31]. Electrically evoked ABR (EABR) waves were poorly differentiated in children with SP + DP, showing that neural synchronization was not improved by electrical stimulation. EABRs, however, were normal in children withapp. Gibson and Sanli [33] reported similar findings. Auditory evoked potentials originating from the thalamus and cerebral cortex may provide hearing capabilities in AN patients. The P1-N1-P2 components in CAEPs arise from primary auditory cortex in the transverse temporal gyrus. The important CAEP mismatch negativity (MMN) comes from the planum temporale outside the auditory cortex and from the lateral posterior temporal gyrus and can be viewed as an indicator of pre-cortical hearing discrimination. MMN in AN children indicates retained coding of auditory information at a cortical level. Such CAEP testing may provide objective means in evaluating auditory threshold and speech discrimination capability in AN patients, especially in children who are unable to cooperate or use hearing aids or cochlear implants. Current research data support the notion that presence of CAEPs (including MMN) is correlated to speech discrimination in AN children. Rance et al [34] reported CAEP testing in 18 AN patients, showing that normal CAEPs were correlated to better speech discrimination than those with poor CAEPs. Speech discrimination scores in AN children with positive MMN also seem to be superior over those without. Pearce et al [35] applied CAEP testing in 2 infants diagnosed with AN. Hearing aids were chosen for the one with positive CAEPs and cochlear implant for the one in whom CAEPs could not be induced with or without assistance. Despite the above mentioned electrophysiology tests, the sensitivity and specificity of these tests remain to be further studied for hopefully their utility as an objective means in evaluating hearing and speech discrimination capability in young AN children and in predicting intervention outcomes. Pathophysiological mechanisms of AN Possible lesion sites in AN include IHCs, IHC ribbon synapses and auditory nerve trunk. Damage at these locations may lead to two neurophysiology consequences [27] : 1) loss of synchronization of neural discharges from auditory nerve demyelination and/or synaptic dysfunction between the auditory nerve and IHCs; and 2) decrease in neural discharges due to loss of receptors or axons (Figure 1). 89 Figure 1. Left: Schematic representation of connection of afferent auditory nerve fibers to an IHC via synapses. Upper right: Temporal responses of a number of auditory nerve fibers to a stimulation signal. Black arrows - normal responses; gray arrows - abnormal nerve fiber responses. Bottom right: Gross auditory nerve response to the transient stimulus. Abnormal nerve fiber responses include: a) greatly variable discharge latencies, i.e. poor synchronicity; and b) reduced number of responding nerve fibers, i.e. decreased neural discharges [6]. The accuracy of auditory temporal coding is maintained by multiple physiological mechanisms. IHCs are the cells that provide sound information to the central nervous system. Each IHC synapses with multiple independent afferent nerve fibers and each synapse is able to release glutamate to activate rapid depolarization. Glutamate released in the process quickly opens the post-synaptic AMPA glutamate receptor. The excitatory post-synaptic currents (EPSCs) at the nerve ending show rapid rising and falling, allowing fast repetitive discharges in short time. Similar nerve sizes mean similar transduction velocity, leading to optimal synchronization in discharges. As it approaches the node of Ranvier, signal conduction assumes a saltatory (jumping) pattern, greatly increasing the speed of conduction. Under normal circumstances, all nerve fibers fire in a very short period of time with in a very coherent manner. If neurotransmitter release is disrupted by IHC lesions, or if the number or sensitivity of receptors is compromised by synaptic lesions, or if conduction velocity is impeded by nerve fiber lesions, then the discharge synchronicity and number of discharging nerve fibers will be impaired and timing pattern and magnitude of large scale discharge altered.

5 Particular pathological changes that can potentially disrupt or lower neural conduction synchronicity include mechanical changes in hair cells, metabolic changes within hair cells, compromised release and re-uptake of neurotransmitters, decrease in receptor sensitivity and impaired discharge initiation at nerve endings, as well as pathologies involving neurotransmitters, axons and myelin sheath. At this time, the best evidence to support existence of AN is presence of accompanying peripheral or cranial nerve neuropathy. If such neuropathy is absent and there is clear auditory responses to electrical stimulation, nerve ending-ihc or synaptic lesions are speculated as the etiology. Synaptic lesions may involve temporal processing changes, discharge amplitude changes and afferent nerve receptor changes. In addition, gene mutations that have been shown to be related to AN through genetic screening need to be further studied to clarify the locations and pathological mechanisms in AN (see below). The role of basic research in diagnosis and treatment of AN Animal studies at cellular and molecular levels may provide directions for description and detection of AN lesion sites and for development of treatments. Animal models with known lesion locations at the IHC and synapses include: 1) Cav1.3 ion channel knockout mice [36, 37], in which neurotransmitter release from IHCs is blocked, with 90% decrease in Ca 2 + current [38], absent ABRs, but preserved high frequency DPOAEs; 2) Bassoon protein (a structural protein responsible for locating ribbon synapse to IHC active pre-synaptic area) mutation mice, in which numbers of synapses and synaptic release vesicles are reduced [39], with slightly decreased Ca 2 + current, abnormal ABRs, but normal CM and DPOAEs; and 3) OTOF gene (coding the otoferlin protein which is closely related to the release of ribbon synaptic vesicles [40] ) knockout mice, in which the ribbon synapse morphology and Ca 2 + current are both normal, but IHC neurotransmitter release is decreased with absent ABRs but normal OAEs or CM. Animal models with known lesion locations at the synapse or auditory nerve include: 1) Beta-4 (a protein playing an important role in locating and linking the voltage-gated sodium and potassium channels at the node of Ranvier [41] ) gene mutation mice, in which only Wave I can be induced in some animals with progressive latency delay; 2) myelin sheath protein gene (PMP22) defect animals, in which ABR waves show low amplitudes and delayed latencies and abnormal myelin sheath formation and progressive loss of spiral ganglion [42] ; and 3) erbb sensor defect animals, in which ABRs are absent or Vol.7 No.2 grossly abnormal while OAEs are present. Auditory pathway dysfunction related to this gene involves only loss of the spiral ganglion, with no peripheral or cranial nerve neuropathy. These animal models have served to guide clinical treatment for AN, i.e. cochlear implant can be expected to produce good results in patients with presentations corresponding to animal models of IHC and synaptic lesions, whereas it usually fails in those with presentations corresponding to models of auditory nerve or central auditory pathway damages. References [1] Starr A, Picton TW, Sininger Y, et al. Auditory neuropathy. Brain 1996,119 (3): [2] Satrr A, McPherson D, Patterson J, et al. Absence of both auditory evoked potentials and auditory percepts dependent on timing cues. Brain 1991,114: [3] Zeng FG, Oba S, Garde S, et al. Temporal and speech processing deficits in auditory neuropathy. Neuroreport 1999,10: [4] Berlin CI, Hood L, Morlet T, et al. Auditory neuropathy/ dys-synchrony: diagnosis and management. Ment Retard Dev Disabil Res Rev 2003,9: [5] Berlin CI, Hood LJ, Morlet T, et al. Multi-site diagnosis and management of 260 patients with auditory neuropathy/ dys-synchrony (auditory neuropathy spectrum disorder). Int J Audiol 2010,49: [6] Sarr A, Zeng FG, HJ M. Perspective on Auditory Neuropathy: Disorders on Inner Hair Cell, Auditory Nerve, and Their Synapse. In: Allan I. Basbaum, Akimichi Kaneko, Gordon M. Shepherd, Westheimer G, editors. The Senses: A comprehensive Reference, vol. 3. San Diego: Elsevier, 2008, [7] Worthington D, Peters J. Quantifiable hearing and no ABR: paradox or error?. Ear Hear 1980,1: [8] Kraus N, Ozdamar O, Stein L, et al. Absent auditory brain stem response: peripheral hearing loss or brain stem dysfunction?. Laryngoscope 1984,94: [9] Gu Rui, Yu Li-Ming. The Function of Medial Olivocochlear System in Patients with Central Low. Chinese Journal of Otorhinolaryngology Head and Neck Surgery. 1992,27: [10] Rapin I, Gravel JS. Auditory neuropathy: a biologically inappropriate label unless acoustic nerve involvement is documented. J Am Acad Audiol 2006,17: [11] Berlin C, Hood I, Rose K. On renaming auditory neuropathy as auditory dys-synchrony. Audiol Today 2001,13: [12] Sininger Y, Starr A, Petic C, et al. Guidelines for Identification and Management of Infants and Young Children with Auditory Neuropathy Spectrum Disorder. Guidelines Development Conference on the Identification and Management of Infants and Young Children with Auditory Neuropathy. Como, Italy, [13] Roush P, Frymark T, Venediktov R, et al. Audiologic management of auditory neuropathy spectrum disorder in children: a systematic review of the literature. Am J Audiol 2011,20: [14] Sanyelbhaa Talaat H, Kabel AH, Samy H, et al. Prevalence of auditory neuropathy (AN) among infants and young children with severe to profound hearing loss. Int J Pediatr Otorhinolaryngol 2009,73:

6 [15] Rance G, Beer DE, Cone-Wesson B, et al. Clinical findings for a group of infants and young children with auditory neuropathy. Ear Hear 1999,20: [16] Cheng X, Li L, Brashears S, et al. Connexin 26 variants and auditory neuropathy/dys-synchrony among children in schools for the deaf. Am J Med Genet A 2005,139: [17] Foerst A, Beutner D, Lang-Roth R, et al. Prevalence of auditory neuropathy/synaptopathy in a population of children with profound hearing loss. Int J Pediatr Otorhinolaryngol 2006,70: [18] Vlastarakos PV, Nikolopoulos TP, Tavoulari E, et al. Auditory neuropathy: endocochlear lesion or temporal processing impairment? Implications for diagnosis and management. Int J Pediatr Otorhinolaryngol 2008,72: [19] Sininger Y. Auditory neuropathy in infants and children: Implicatons for early hearing detection and intervention programs. Audiol Today 2002,14: [20] Dowley AC, Whitehouse WP, Mason SM, et al. Auditory neuropathy: unexpectedly common in a screened newborn population. Dev Med Child Neurol 2009,51: [21] Kumar UA, Jayaram MM. Prevalence and audiological characteristics in individuals with auditory neuropathy/auditory dys-synchrony. Int J Audiol 2006,45: [22] Tang TP, McPherson B, Yuen KC, et al. Auditory neuropathy/auditory dys-synchrony in school children with hearing loss: frequency of occurrence. Int J Pediatr Otorhinolaryngol,2004,68: [23] Lotfi Y, Mehrkian S. The prevalence of auditory neuropathy in students with hearing impairment in Tehran, Iran. Arch Iran Med,2007,10: [24] Lee JS, McPherson B, Yuen KC, et al. Screening for auditory neuropathy in a school for hearing impaired children. Int J Pediatr Otorhinolaryngol,2001,61: [25] Ji Fei, Chen Ai-Ting, Zhao Yang et al. An analysis of the Performance-Intensity function of patients with auditory neuropathy. Chinese Scientific Journal of Hearing and Speech Rehabilitation,2010: [26] Rance G, Barker EJ, Sarant JZ, et al. Receptive language and speech production in children with auditory neuropathy/dyssynchrony type hearing loss. Ear Hear,2007,28: [27] Zeng FG, Kong YY, Michalewski HJ, et al. Perceptual consequences of disrupted auditory nerve activity. J Neurophysiol, 2005,93: [28] Shi W, Ji F, Lan L, et al. Characteristics of cochlear microphonics in infants and young children with auditory neuropathy. Acta Otolaryngol,2011,132: [29] Starr A, Sininger Y, Nguyen T, et al. Cochlear receptor (microphonic and summating potentials, otoacoustic emissions) and auditory pathway (auditory brain stem potentials) activity in auditory neuropathy. Ear Hear,2001,22: [30] Santarelli R, Starr A, Michalewski HJ, et al. Neural and receptor cochlear potentials obtained by transtympanic electrocochleography in auditory neuropathy. Clin Neurophysiol,2008, 119: [31] McMahon CM, Patuzzi RB, Gibson WP, et al. Frequency-specific electrocochleography indicates that presynaptic and postsynaptic mechanisms of auditory neuropathy exist. Ear Hear, 2008,29: [32] O'Leary S, Mitchell T, Gibson W, et al. Abnormal positive potentials in round window electrocochleography. Am J Otolaryngol,2000,21: [33] Gibson WP, Sanli H. Auditory neuropathy: an update. Ear Hear,2007,28:102S-106S. [34] Rance G, Cone-Wesson B, Wunderlich J, et al. Speech perception and cortical event related potentials in children with auditory neuropathy. Ear Hear,2002,23: [35] Pearce W, Golding M, Dillon H. Cortical auditory evoked potentials in the assessment of auditory neuropathy: two case studies. J Am Acad Audiol,2007,18: [36] Platzer J, Engel J, A AS-F, et al. Congenital deafness and sinoatrial node dysfunction in mice lacking class D L-type Ca2 + channels. Cell,2000,102: [37] Dou H, Vazquez A, Namkung Y, et al. Null mutation of alpha1d Ca2+ channel gene results in deafness but no vestibular defect in mice. J Assoc Res Otolaryngol,2004,5: [38] Brandt A, Striessnig J, T TM. CaV1.3 channels are essential for development and presynaptic activity of cochlear inner hair cells. J Neurosci,2003,23: [39] Khimich D, Nouvian R, Pujol R, et al. Hair cell synaptic ribbons are esssential for synchronous auditory signaling. Nature 2005; 434: [40] Roux I, Safieddine S, Nouvian R, et al. Otoferlin, defective in a human deafness form, is essential for exocytosis at the auditory ribbon synapse. Cell 2006,127: [41] Bennet V, Baines A. Spectrin and ankyrin-based pathways: metazoan inventions for integrating cells into tissues. Physiol Rev,2001,81: [42] Zhou R, Assouline J, Abbas P, et al. Anatomical and physiological measures of auditory system in mice with peripheral myelin deficiency. Hear Res,1995,88:87-97.\ (Rrceived August 8,2012) 91