Differential vulnerability of inner and outer hair cell systems to chronic mild hypoxia and glutamate ototoxicity: Insights into the cause of auditory neuropathy

Abstract

OBJECTIVE: To describe the effects of long-term mild hypoxia and of glutamate poisoning on the functional properties of the cochlea. METHODS: Outer hair cell activity was monitored using otoacoustic emissions and cochlear microphonics, and inner hair cell/cochlear afferent function was measured using neural responses (cochlear action potentials or auditory brainstem responses [ABRs]). RESULTS: In contrast to the effects of acute anoxia, in which all aspects of cochlear function are simultaneously lost, mild, long-term hypoxia results in a clear differential effect on outer versus inner hair cell systems. During a 2-hour period of mild hypoxia, ABR amplitude and threshold deteriorate significantly, whereas outer hair cell function, as reflected by otoacoustic emissions, shows little or no change. A similar dissociation between inner and outer hair cell function is observed during instillation of glutamate (1-10 mM), where the cochlear microphonic and the otoacoustic emissions are unchanged, whereas cochlear action potential amplitudes are reduced. CONCLUSION: These studies demonstrate a difference in vulnerability of inner and outer hair cell systems. The inner hair cell/cochlear afferent system is vulnerable to long-term, mild hypoxia; this may be an etiologic factor in hearing loss of cochlear origin, particularly in high-risk birth infants with auditory neuropathy.

Headnote

Abstract

Headnote

Objective: To describe the effects of long-term mild hypoxia and of glutamate poisoning on the functional properties of the cochlea.

Methods: Outer hair cell activity was monitored using otoacoustic emissions and cochlear microphonics, and inner hair cell/cochlear afferent function was measured using neural responses (cochlear action potentials or auditory brainstem responses [ABRs]).

Results: In contrast to the effects of acute anoxia, in which all aspects of cochlear function are simultaneously lost, mild, long-term hypoxia results in a clear differential effect on outer versus inner hair cell systems. During a 2-hour period of mild hypoxia, ABR amplitude and threshold deteriorate significantly, whereas outer hair cell function, as reflected by otoacoustic emissions, shows little or no change. A similar dissociation between inner and outer hair cell function is observed during instillation of glutamate (1-10 mM), where the cochlear microphonic and the otoacoustic emissions are unchanged, whereas cochlear action potential amplitudes are reduced.

Conclusion: These studies demonstrate a difference in vulnerability of inner and outer hair cell systems. The inner hair cell/cochlear afferent system is vulnerable to long-term, mild hypoxia; this may be an etiologic factor in hearing loss of cochlear origin, particularly in high-risk birth infants with auditory neuropathy.

Sommaire

Objectif. Decrire les effets a long terme de Phypoxie legere et de Vempoisonnement au glutamate sur les proprietes de la cochlee.

Methode: L'activite des cellules ciliees externe a ete monitoree avec les emissions otoacoustiques et la microphonie cochleaire pendant que les cellules ciliees internes/systemes afferent de la cochlee ont ete evaluees par les reponses nerveuses (potentiels d'action cochleaires et potentiels evoques du front cerebral).

Resultats: Contrairement aux effets de Panoxie aigue, durant laquelle toutes les fonctions cochleaires sont affect6es simultanement, dans le cas de Phypoxie legere on note une difference entre les effets sur les cellules ciliees internes et externes. Durant une periode de 2 heures d'hypoxie legere, l'amplitude et le seuil du PETC se sont deteriores de facon significative pendant que la fonction des cellules externes etait maintenue. Une difference similaire est observee durant l'instillation de glutamate (1-10 mM) soit une diminution de l'amplitude des potentiels d'action cochleaire pendant que la microphonie cochleaire et les emissions otoacoustiques demeurent inchangees.

Conclusion: Ces etudes montrent une difference de vulnerabilite entre les systemes de cellules ciliees internes et externes. Les cellules ciliees internes/systeme afferent de la cochlee, est vulnerable a Phypoxie legere a long terme. Cela peut etre un facteur etiologique dans la surdite d'origine cochleaire, particulierement chez les enfants a haut rique de neuropathie.

Key words: auditory brainstem response, auditory neuropathy, cochlear hypoxia, cochlear microphonics, glutamate ototoxicity, hair cell degeneration, otoacoustic emissions

The aim of the present study was to define the effects of mild, long-term cochlear hypoxia on inner hair cell (IHC) versus outer hair cell (OHC) systems. We are studying the effects of chronic cochlear hypoxia for two reasons. First, dysfunction arising from such a condition appears to be very different from that found with ischemia, anoxia, or acute hypoxic insults. Second, mild chronic cochlear hypoxia may be a relatively common etiology for hearing disorders, particularly in relation to auditory neuropathy in high-risk birth infants.1,2

There have been a number of histologic, biochemical, and electrophysiologic studies of the inner ear in which the effects of ischemia or anoxia have been evaluated. These include (seminal) electrophysiologic,3-6 biochemical,7,8 and histologic investigations9-11; more recent studies have extended these original observations.12,13

Various electrophysiologic measures have been used to monitor different aspects of cochlear function, for example, the endocochlear potential, which indicates strial function, the cochlear microphonic (CM), which mainly reflects OHC activity, and neural potentials, which monitor IHC transduction and cochlear afferent neuronal activity. At the onset of acute anoxia or hypoxia, all of these potentials change rapidly and apparently simultaneously (within tens of seconds), and there is little evidence of differential effects on the OHCs compared with IHC/cochlear afferent function. However, one study14 in which cochlear action potentials (CAPs) and CM were monitored during transient ischemia did report that the neural potentials were reduced a few tens of seconds before the CM. Furthermore, it was also reported that the CAP recovered from ischemia later than CM. There has been a similar report of neural potentials (from the IHC/cochlear afferent system) that lag behind in recovery compared with CM (OHC system) after transient anoxia.3 We suggest that these observations indicate that the IHC/cochlear afferents are more vulnerable than the OHC system.

In the present study, we show that mild, long-term hypoxia reveals a difference in vulnerability between IHC and OHC systems (i.e., not all cochlear mechanisms are simultaneously or equally affected). Our first evidence that the IHC/cochlear afferent system was primarily affected in mild, chronic hypoxia came from histologic studies15,16 that showed IHC swelling and blebbing (cytoplasmic extrusion). These studies also showed that OHCs were relatively unaffected by the hypoxic episode, at least as seen using scanning electron microscopy (SEM). These morphologic changes were a confirmation of other studies in which hypoxia or postmortem changes (anoxia) resulted in abnormalities to the IHC/cochlear afferent system region preceding damage to OHCs.9-11,13,17,18 In the present study, we have investigated the functional aspects of this differential damaging effect during mild, chronic hypoxia by monitoring OHC activity with otoacoustic emission (OAE) recording and the IHC/cochlear afferent system by neural evoked potential recording.

Cochlear perfusion with glutamate is known to have a blocking effect on the synaptic connections of IHC with cochlear afferents and will theoretically have a similar effect on signals from the IHC and OHC systems as that produced by chronic, mild hypoxia. We confirm that this is the case, providing further evidence to support our hypothesis that long-term, mild hypoxia has deleterious effects on the IHC/cochlear afferent system. Materials and Method

Cochlear Hypoxia Experiment

Five adult chinchillas (Chinchilla laniger), weighing between 400 and 600 g and free from middle ear infection, were used. All procedures were performed in accordance with the guidelines of the Canadian Council on Animal Care and with the approval of the Faculty of Medicine Animal Care Committee of the University of Toronto and the local committee at the Hospital for Sick Children.

All experiments were carried out in a sound-attenuating booth. Animals were anaesthetized with ketamine (15 mg/kg, IM) and xylazine (25 mg/kg, IM). To prevent airway obstruction by secretions, atropine (0.1 mg/kg, IM) was administered. Half doses of ketamine and xylazine were given for maintenance for anaesthesia. All animals were tracheotomized and had a small tube inserted into the trachea. To produce systemic hypoxia, a dead space was added to the respiratory tidal volume by connecting the tracheal tube to a 20-mL syringe with a small hole at the distal end (as in previous studies,"6).

Otoacoustic Emissions

Both transient evoked otoacoustic emissions (TEOAEs) and distortion product otoacoustic emissions (DPOAEs) were recorded using the IL088/92 system (Otodynamics, London, UK). Transient evoked otoacoustic emission and DPOAEs were measured at 10minute intervals before and during the period of experimental hypoxia. The parameters used for TEOAE recording were as follows. Stimuli were 80-psec clicks at 80 dB SPL (peak equivalent) presented in the nonlinear mode (three condensation clicks of equal amplitude presented in sequence with one inverted polarity click at three times the amplitude). Acoustic signal responses (0.1-7.5 kHz bandpass) were time averaged (20-msec time window; 260 sweeps).

For DPOAE (2f1-f2) measurement, f2 test stimulus ranged from 0.5 kHz to 8 kHz, presented at 60 dB SPL; f1 and f2 amplitudes were equal, and the f1:f2 ratio was 1.22; results were plotted as DPgrams. At 10minute intervals, a DPOAE input/output (I/O) function was measured for stimuli between 35 and 80 dB SPL, with f2 = 8 kHz.

Auditory Brainstem Responses

Auditory brainstem responses were recorded with needle electrodes (vertex-bulla configuration). Stimuli were 0 to 80 dB SPL tone bursts (SigGen, Tucker-Davis, FL). The responses were amplified (x 10^sup 3^), filtered (150 Hz3 kHz), and time averaged (n = 250; BioSig, TuckerDavis). Thresholds of ABR (PS equivalent) were determined to within 5 dB by visual inspection of averaged waveforms. Before and after hypoxia, ABR thresholds were determined at six frequencies (0.5, 1, 2, 4, 8, and 16 kHz) to produce ABR audiograms. During the hypoxic period, ABR to an 8-kHz stimulus was measured at 10-minute intervals.

Scanning Electron Microscopy

After the final (posthypoxic) ABR and OAE recordings, the animals were sacrificed, and the cochleas were prepared for SEM observation. Cochleas were quickly removed and perfused with fixative (phosphate buffered 1% glutaraldehyde/4% formaldehyde), postfixed with 1% OSO^sub 4^, and dissected in 70% alcohol. After dissection, specimens were prepared for SEM using standard procedures.

Glutamate Infusion Experiment

Seven adult chinchillas weighing between 400 and 600 g and free from middle ear infection were used. Anaesthetic procedures were as described above. The bulla was exposed ventrally, and a 200-(mu)m hole was drilled into the basal turn (scala tympani) of the cochlea. Cochlear action potential and CM were recorded in response to 8-kHz tone bursts at 10 to 100 dB SPL (BioSig, Tucker-Davis) via a closed system transducer (Etymotic, ER-2, Elk Grove Village, IL). The electrical potentials were recorded using a silver ball electrode placed near the basal turn of the cochlea. The experimental set-up is illustrated in Figure 1. A glass capillary was carefully inserted into the scala tympani. Distortion product otoacoustic emissions, CAP, and CM were measured during successive cochlear perfusions with artificial perilymph (AP), AP with 0.1-mM glutamate, AP with 1-mM glutamate, and AP with 10mM glutamate (all at a perfusion rate of 30 (mu)L/min). Results

Cochlear Hypoxia Experiment

Figures 2 to 4 show results from three subjects, which are representative of the range of results obtained from five animals. The upper plots of each figure show the time course of ABR threshold changes compared to amplitude change of DPOAE (left panel; right-hand scale) and TEOAE (right panel; right-hand scale) during and after the hypoxic period (indicated by bar). During the hypoxic period, DPOAE I/O functions were repeatedly measured. For clarity, in the time course plots, we only show DPOAE amplitudes to stimuli presented at 60 dB SPL. First, note that in all animals, there are no substantial ABR threshold shifts (> 5 dB) or OAE changes for 1 to 2 hours after hypoxia onset. After this period, ABR threshold (to 8 kHz) becomes gradually elevated at a rate of between 0.5 dB/min (see Fig. 2) to > 2 dB/min (see Figs. 3 and 4). During this period of ABR threshold change, TEOAE and DPOAE amplitudes for the subjects of Figures 2 and 3 show little change. For the animal shown in Figure 4, both TEOAE and DPOAE amplitude reductions lag behind ABR changes by about 30 minutes.

The lower right-hand graphs of Figures 2 to 4 show ABR audiograms (ABR thresholds to tone frequencies of 0.5-16 kHz) before and after the hypoxic period. For all subjects, there is a general pattern of threshold shift across all frequencies measured. Although there is a tendency for greater threshold elevations at 4 to 8 kHz compared with 0.5 to 1 kHz, there is no obvious high-frequency loss. The centre plot shows DPOAE amplitudes from 1 to 8 kHz (DPgram) before and after hypoxia, and the lower left panel shows the DPOAE I/O function (f2 = 8 kHz). For two of the subjects shown (see Figs. 2 and 3) and the two others not illustrated here, there is little change in the DPgram or I/O curves after the hypoxic period. The animal of Figure 4 has larger changes in these DPOAE measures. Note that the DPgram shows amplitude changes at all frequencies tested.

After the hypoxia experiment, SEM of each cochlea was carried out. Figure 5 shows a typical finding. In this case, we show the 8-kHz region (upper basal turn) from the subject of Figure 3. The IHCs show cytoplasmic protrusions (blebs), and the stereocilia show some swelling and disarrangement; OHCs appear normal. This pattern existed, in various degrees, throughout the cochlea.

Figure 1

Glutamate Infusion Experiment

Representative data from two of the seven subjects (A and B) are presented in Figure 6. All data are in the form of I/O functions. The upper plots show CAP amplitude ((mu)V) to 8-kHz tone burst, the middle panels show CM amplitude ((mu)V) to 8-kHz stimulus, and the lower plot is the DPOAE I/O function curve at 8 kHz. Measurements were made before and after perfusion of artificial perilymph alone and glutamate added at concentrations of 0.1 mM, 1 mM, and 10 mM. In all subjects, perfusion with AP alone caused minimal change to any function measured. In both subjects shown, the experimental procedures up to and including 10-mM glutamate have little effect on either CM or DPOAE amplitude. In contrast, CAP amplitudes are significantly reduced with 1-mM glutamate in subject A and with 10-mM glutamate in subject B.

Discussion

Chronic Mild Hypoxia

We designed these experiments to explore the differential vulnerability of IHC and OHC systems to mild chronic hypoxia. Although there have been numerous studies investigating electrophysiologic and histologic changes in the hypoxic/ischemic cochlea, few have focused on this issue of differential vulnerability of hair cell systems, and none have specifically modelled mild, long-term hypoxic conditions. It is evident from our data that when mild hypoxic conditions eventually cause cochlear dysfunction, the IHC/cochlear afferent system is first affected, perhaps indicating that it is the most vulnerable. We are aware that our method of inducing mild hypoxia will result in a systemic effect. The ABR changes we observe might be central in origin rather than a reflection of cochlear afferent function. However, this is unlikely given that all components of the ABR waveform are reduced together. In other words, we do not observe centrally generated components being more affected than contributions from the periphery.

Figure 2

In studies involving cochlear anoxia or ischemia (including postmortem monitoring), deterioration of all aspects of cochlear function is too rapid to reveal differential vulnerability. An exception to this was the study by Mom et al.,14 in which very rapid monitoring of neural potential (CAP) and OHC-derived signals (CM and OAE) was achieved during transient ischemia. They reported immediate reductions to CAP amplitudes, followed some tens of seconds later by OAE and CM deterioration. That study, in a sense, showed similar results to that of our subject in Figure 4, except that the time scale of our changes is one to two orders of magnitude longer. To our knowledge, no previous studies have reported cochlear conditions (due to hypoxic effects) in which neural function is significantly deteriorated but with no apparent change to OHC function.

It is worth noting here that the chronic hypoxic conditions of our present study probably do not change the endolymphatic potential (EP) significantly. Although we have not directly measured EP, it is clear from the work of others19,20 that OAEs change hand in hand with EP during hypoxia; we equate our lack of OAE change with a stable EP.

Our present study stemmed directly from two previous sets of experiments by our group. In an early histologic study's on the ototoxic effects of deferoxamine mesylate (an iron chelating agent), we noted that IHCs showed cytoplasmic swelling and disorganized stereocilia, but that OHCs appeared normal (by SEM examination). In subsequent studies, we found a similar pattern of damage in animals after many hours of mild hypoxia.15,16 These findings are consistent with reports of others on experimental anoxia or early postmortem effects, which noted that abnormalities were first observed or were more evident at the IHC level.9-11,13,17,18

Figure 3

The second set of observations that formed the background to the present study was related to the ototoxic effects of carboplatin treatment in the chinchilla in which extensive (sometimes total) IHC loss was produced with little or no damage to OHCs.21-24 It is now clear that in certain types of cochlear insult, the IHC system is more vulnerable than the OHCs. This conclusion is in contrast to aminoglycoside poisoning and acoustic trauma, where the OHCs are generally more vulnerable.

So far, we have described functional differences between OHC and IHC systems resulting from cochlear hypoxia. It is clear from detailed histologic studies after anoxia (e.g., by Billett et al13) that the IHC itself is not the first affected structure but rather it is the associated afferent nerve endings. The IHC stereocilia disruption and cytoplasmic swelling that we observe with SEM are assumed to be subsequent to injury of the synaptic region. This afferent nerve ending damage shares similar characteristics with two other cochlear insults: noise exposure25,26 and glutamate ototoxicity.27,28

Glutamate Ototoxicity

Glutamate is thought to be the major neurotransmitter of the afferent synapse (IHC/cochlear nerve), and many studies have shown that high doses of a glutamate agonist appear to be toxic to the IHC/afferent nerve system.27-30 It is still unclear as to whether this effect is the same as the glutamate excitotoxicity commonly described for central neurons; they are possibly related. Our data here (see Fig. 6) show clearly that some levels of glutamate have a blocking or damaging effect on IHC mediated neural function, whereas OHCs are unaffected. We have used 1- to 10-mM concentrations of glutamate in our perfusate; however, we do not know the actual concentrations at the site of action. Therefore, we do not know if our effective levels are within the physiologic range or whether similar levels could be reached from natural sources. It has also been shown that transient ischemia increases glutamate levels in the gerbil cochlea.31 Perhaps noise exposure or cochlear hypoxia creates an excess of glutamate, which interferes with and perhaps damages the IHC/cochlear afferent junction.

Figure 4

Figure 5

Figure 6

Auditory Neuropathy

We suggest that these studies have important clinical significance. In humans, Starr et al.1 have drawn attention to an entity termed "auditory neuropathy." This hearing disorder is characterized, in part, by abnormal ABR (i.e., elevated threshold and unusual waveform) but almost normal OAEs (or CM). These patients appear to have a dysfunction of IHCs and/or cochlear nerve whereas OHC functions remain normal.12 It has been observed that many patients with auditory neuropathy had a history of high-risk birth.1,33 We have hypothesized elsewhere2 that because such infants have suffered some degree of chronic hypoxia, this could be one natural etiology in this disorder.

The studies that we report here could be described as animal models of auditory neuropathy. In both the hypoxia and the glutamate experiments, the conditions resulting in IHC/cochlear afferent system damage (reduction in neural activity) with normal OHC function replicate many of the characteristic symptoms seen in auditory neuropathy. We further speculate that the fundamental damaging effect on the IHC/cochlear afferents by a wide range of cochlear insults could be ultimately mediated by glutamate (excito?) toxicity. Such suggestions have previously been made by Pujol et al. in relation to ischemic or anoxic conditions during aging.27,28 We suggest that the list might be extended to include chronic hypoxia, desferoxamine toxicity, carboplatin poisoning (and perhaps other neoplastic agents), and noise exposure.

References

References

References

1. Staff A, Picton TW, Singer Y, et al. Auditor,/ncuropathy. Brain 1996; 119:741-753.

2. Harrison RV. An animal model of auditory neuropathy. Ear Hear 198; 19:355-36.

3. Konishi T, Butler RA, Fernandez C. Efm of ana on cochlear potentials. J Aft Sad Am 1961; 33:349-356.

4. Konishi T, Kee E,. Singleton GT. Negative potential in scale media during erly smse of Ann.Ma,(xia. Acta Otolaryngol (Stockh) 1967; 64:107-118.

5. Konishi T. Some observations on negative endr potential during anoxia: Acta Otolaryagol (Stockh) 1979; 87:-506-516.

References

6. Lawrence M, Nuttall AL, Burgio PA. Cochlear potentials and oxygen associated with hypoxia. Ann Otol Rhinol Laryngol 1975; 84:499-512.

7. Thalmann R, Myoshi T, Thalmann I. The influence of ischemia upon the energy reserves of inner ear tissues. Laryngoscope 1972; 82:2249-2272.

8. Thalmann R, Kusakari J, Myoshi T. Dysfunctions of energy releasing and consuming processes of the cochlea. Laryngoscope 1973; 83:1690-1712.

9. Kimura R, Perlman HB. Arterial obstruction of the labyrinth. Ann Otol Rhinol Laryngol 1958; 67:5-40.

10. Fernandez C. Postmortem changes on the vestibular and cochlear receptors (guinea pig). Arch Otolaryngol 1958; 68:460-487.

11. Bohne BA. Mechanisms of noise damage in the inner ear. In: Henderson D, Hamernik RP, Dosanjh DS, Mills JH, eds. Effects of noise on hearing. New York: Raven Press, 1976: 41-68.

References

12. Kusakari J, Kambayashi J, Kobayashi T, et al. The effect of transient anoxia upon the cochlear potentials. Auris Nasus Larynx 1981; 8:55-64.

13. Billett TE, Thorne PR, Gavin JB. The nature and progression of injury in the organ of Corti during ischemia. Hear Res 1989; 41:189-197.

14. Mom T, Avan P, Romand R, et al. Monitoring of functional changes after transient ischemia in gerbil cochlea. Brain Res 1997; 751:20-30.

15. Shirane M, Harrison RV. The effects of deferoxamine mesylate and hypoxia on the cochlea. Acta Otolaryngol (Stockh) 1987; 104:99-107.

16. Shirane M, Harrison RV. The effects of hypoxia on sensory cells of the cochlea in chinchilla. Scanning Microsc 1987; 1:1175-1183.

References

17. Jordan VM, Pinheiro ML, Chiba K, et al. Postmortem changes in surface preparations of the cochlea. Ann Otol Rhinol Laryngol 1973; 82:111-125.

18. Theopold HM. Postmortem changes of fine structure in the auditory pathways (organ of Corti and nucleus cochlearis of the guinea pig). Laryngol Rhinol 1977; 56:207-215.

19. Rebillard G, Lavigne RM. Effect of reversible hypoxia on the compared time courses of endocochlear potential and 2f1-f2 distortion products. Hear Res 1992; 62:142-148.

20. Rebillard G, Klis JF, Lavigne RM, et al. Changes in 2fl-f2 distortion product otoacoustic emissions following alterations of cochlear metabolism. Br J Audiol 1993; 27:117-121.

21. Wake M, Takeno S, Ibrahim D, Harrison RV. Selective inner hair cell ototoxicity induced by carboplatin. Laryngoscope 1994; 104:488-493.

22. Wake M, Anderson J, Takeno S, et al. Otoacoustic emission amplification after inner hair cell loss. Acta Otolaryngol (Stockh) 1996; 116:374-381.

23. Takeno S, Harrison RV, Ibrahim D, et al. Cochlear function after selective inner hair cell degeneration induced by carboplatin. Hear Res 1994; 75:93-102.

24. Takeno S, Wake M, Mount RJ, Harrison RV. Degeneration of spiral ganglion cells in the chinchilla after inner hair cell

References

loss induced by carboplatin. Audiol Neurootol 1998; 3: 281-290.

25. Spoendlin FL Anatomical chapp-following various noise exposures. In: Henderson D, Haik RP, Dosanjh DS, Mills JH, eds. E of noise on Ices. New York: Raven pres,1976:696-89.

26.Robertson TJ. Functional significance of dendritid, swelling after loud souda vn the guinea pig, . Hear has 1983; 9:263-278.

References

27. Pujol R, Rebillard G, Puel JL, et al. Glutamate neurotoxicity in the cochlea: a possible consequence of ischemic or anoxic conditions occurring in ageing. Acta Otolaryngol Suppl (Stockh) 1991; 476:32-36.

28. Puel JL, Pujol R, Ladrech S, Eybalin M. Alpha-amino-3hydroxy-5-methyl-4-isoxazole propionic acid electrophysiological and neurotoxic effects in the guinea-pig cochlea. Neuroscience 1991; 45:63-72.

References

29. Puel JL, Ladrech S, Chabert R, et al. Electrophysiological evidence for the presence of NMDA receptors in the guinea pig cochlea. Hear Res 1991; 51:255-264.

30, Eybalin M, Norenberg MD, Renard N. GLutamine synthetase and glutamate metabolism in the guinea pig cochlea. Hear Res 1996; 101:93-101.

31. Hakuba N, Gyo K, Yanagihara N, et al. Efflux of glutamate into the perilymph of the cochlea following transient ischemia in the gerbil. Neurosci Lett 1997; 230:69-71.

32. Stein L, Tremblay K, Pasternak J, et al. Brainstem abnormalities in neonates with normal otoacoustic emissions. Semin Hear 1996; 17:197-213.

33. Deltenre P, Mansbach AL, Bozet C, et al. Auditory neuropathy: a report on three cases with early onsets and major neonatal illnesses. Electroencephalogr Clin Neurophysiol 1997; 104:17-22.