Profound hearing loss in the cat following the single co-administration of kanamycin and ethacrynic acid

Hearing Research, 70 (1993) 205-215 0 1993 Elsevier Science Publishers B.V. All rights reserved

HEARES

205 037%5955/93/$06.00

01997

Profound hearing loss in the cat following the single co-administration of kanamycin and ethacrynic acid Shi-Ang

Xu, Robert

Department

K. Shepherd

of Otolaryngology,

(Received

4 January

*, Yin Chen

The Uni~,ersity of Melbourne,

1993; Revision

received

’

and Graeme

Purkdle.

M. Clark

Victoria 3%52, Australia

8 May 1993; Accepted

30 June

1993)

Co-administration of kanamycin (KA) with the loop diuretic ethacrynic acid (EA) has previously been shown to produce a rapid and profound hearing loss in guinea pigs. In the present study we describe a modified technique for developing a profound hearing loss in cats. By monitoring the animal’s hearing status during the intravenous infusion of EA the technique minimizes the effects of individual variability to the drug regime. Seven cats received a subcutaneous injection of KA (300 mg/kg) followed by intravenous infusion of EA (1 mg/min). Click-evoked auditory brainstem responses (ABRs) were recorded to monitor the animal’s hearing during the infusion. When the ABR thresholds rose rapidly to levels in excess of 90 dB SPL the infusion of EA was stopped. This occurred at EA doses of IO-25 mg/kg, indicating considerable individual variability to the deafening procedure. However. there was a strong negative correlation (r = -0.93) between the EA dose and body weight which accounted for much of this variability. Subsequent ABR monitoring showed that this profound hearing loss was both bilateral and permanent. Significantly, blood urea and creatinine levels, monitored for periods of up to three days after the procedure, remained within the normal range. Furthermore, there was no clinical evidence of renal dysfunction as indicated by weight loss or oliguria. Cochlear histopathology, examined after a two months to three year survival period, showed an absence of all inner and outer hair cells in the majority of cochleas. The extent of loss of spiral ganglion cells was dependent on their distance from the round window and the period of survival follow,ing the deafening procedure. Clearly. the degeneration of spiral ganglion cells continued for several years following the initial insult. Finally, we observed no evidence of renal histopathology. In conclusion, the co-administration of KA and EA produces a profound hearing loss in cats without evidence of renal impairment. Monitoring the animal’s hearing status during the procedure ensures that the dose of EA can be optimised for individual animals. Moreover. it may be possible to adapt this procedure to produce animal models with controlled high frequency hearing losses. Ototoxicity;

Nephrotoxicity:

Hearing

loss; Ethacrynic

acid; Kanamycin;

Introduction During the past decade cochlear implantation has become an important means of treating patients with profound sensorineural hearing losses. Experimental studies have played a significant role in the development and improvement of these devices. Often it is necessary to use profoundly deaf animals in these studies in order to adequately model the clinical condition. To develop a reliable deafening technique, the pathophysiological response should show minimal variation within a species, and the lesion generated should be bilaterally symmetrical to allow the contralateral ear to be used as a control. In addition, to ensure a uniform cochlear pathology, there should be minimal mechanical trauma to the structure of the cochlea.

* Corresponding author. (Present address) Department of Otolaryngology, The University of Melbourne, 32 Gisborne Street. East Melbourne 3002, Australia. Fax: + 61 3 663-1958. ’ Present address: Department of Otolaryngology, Shandong Medical University, Jinan, China.

Cochlear

implant

Finally, since many ototoxic drugs are nephrotoxic, a safe deafening procedure should also ensure minimal adverse effects on the kidneys. Several techniques have been used to develop partially or totally deaf animal models. These include: (i) use of intense noise; (ii) mechanical destruction of the inner ear; (iii) perilymphatic infusion of ototoxic drugs; and (iv) chronic administration of ototoxic drugs. However, most of these deafening techniques have disadvantages that make them unsuitable for use as an animal model for cochlear implant research. High intensity noise results in only a partial and highly variable hearing loss among individual animals (Bredberg, 1973; Cody and Robertson, 1983). Mechanical destruction of part of the basilar membrane or osseous spiral lamina can induce restricted cochlear lesions, resulting in a partial hearing loss and a severe ganglion cell loss adjacent to the lesion (Simmons, 1967). In addition to the uncontrolled nature of the site of trauma, a pathological response of this type can significantly complicate the interpretation of the pathology associated with cochlear implantation. This technique also results in new bone formation, making subsequent electrode insertion into the Scala tympani difficult (Leake-Jones et

al., 1982). Direct application of aminoglycosides via perilymphatic infusion may produce a rapid and severe disruption of the organ of Corti (Leake-Jones et al. 1980, 1982; Sutton and Miller 1983; Shepherd and Clark, 1984). Such a model is of use in certain acute studies where it is necessary to eliminate electrophonic activation of hair cells by the electrical stimulus (e.g. Brown et al., 1992). However, the morphological changes induced by perilymphatic infusion of aminoglycosides are generally so variable and often so severe that comparison between chronically stimulated and deafened control cochleas is difficult (Duckert, 1983; Sutton and Miller, 1983). Moreover, this procedure can result in extensive fibrous tissue and new bone growth within the Scala tympani, making subsequent cochlear implantation difficult (Sutton and Miller, 1983). Both Leake-Jones et al. (1980, 1982) and Duckert (1983) concluded that the animal model produced by perilymphatic infusion was unsuitable for studies of chronic intracochlear electrical stimulation. Long-term systemic injection of aminoglycoside antibiotics can cause a profound hearing loss (Hawkins, 1959; Ylikoski et al., 1974; Leake-Jones et al., 1982; Shepherd and Clark, 1985; Harrison et al., 1991). Because there are large differences in the sensitivity to these drugs within individuals of a species, considerable variability in cochlear pathology is commonly observed (Ylikoski et al., 1974, Shepherd and Clark, 1985). Moreover, since the dose of aminoglycoside required to induce a profound hearing loss can frequently lead to renal failure, both hearing loss and kidney function must be closely monitored throughout the procedure. Evidence of renal impairment requires the deafening procedure to be discontinued (e.g. Leake and Hradek, 1988). Finally, this procedure is also time consuming. Hawkins (1959), for example, reported that it took at least 44 days to profoundly deafen cats and guinea pigs by daily injection of kanamycin (KA). Loop diuretics, such as ethacrynic acid (EA) and furosemide, have an ototoxic effect causing a reversible hearing loss by temporarily decreasing the endocochlear potential and potassium concentration in the endolymph (Syka and Melichar, 1981). However, on the basis of early clinical observations, West et al. (1973) and Brummett and Fox (1982) showed that a single co-administration of ISA with the diuretic EA could permanently deafen guinea pigs. This interaction was evident provided the EA was administered within four hours of the KA. The function between hair cell loss and the dose of EA was very steep - i.e. above some threshold dose, a small increment in EA could cause extensive hair cell loss. In contrast, animals exhibited a more graded hearing loss with an increase in the dose of KA (Brummett and Fox, 1982). Although the complicated interaction between aminoglycosides and loop diuretics remains somewhat unclear (Russell

et al., 1979; Orsulakova and Schacht, 1981; Tran Ba Huy et al., 1983; Yamane et al.. 19881, this technique provides an appropriate way to rapidly create deal animal models. In a preliminary study, we applied the deafening technique developed for guinea pigs (400 mg/ kg KA and 40 mg/ kg EA; Brummett and Fox, 1982) to four adult cats. All the animals exhibited profound hearing losses, however, they also showed evidence of acute renal dysfunction. Only two of the four animals survived the procedure (Shepherd et al., 1988). These preliminary results indicated the need to modify Brummett’s technique for use in the cat. Moreover, the evidence of kidney pathology indicated the need to monitor renal function in these studies. An abstract describing our preliminary physiological findings has been presented previously (Xu et al., 1990).

Materials and Methods Animals

Seven normal hearing cats were used in this study. At the time of the deafening procedure their ages ranged from one to seven months and their weight varied from 0.5 to 2.9 kg. Previous studies have demonstrated that the cat cochlea is both physiologically and morphologically mature at one month of age (Pujol and Marty, 1970; Fernandez and Hinojosa, 1974; Romand and Romand, 1986; Walsh and McGee, 1987). The animals had normal external ears and otoscopically normal tympanic membranes. ABRS

The animals were anaesthetized with Saffan (Alphaxalone, 9 mg/ kg) and maintained with halothane and methoxyfhrrane using a closed circuit anaesthetic machine (Komesaroff, Medical Developments Pty, Ltd.). For each animal, normal hearing status was verified by recording click-evoked ABRs from both ears in a sound attenuated and electrically shielded room. Rarefaction clicks were produced under computer control by generating electrical rectangular pulses of 0.1 ms duration. The acoustic stimuli were presented in a sound field from a Richard Allen DT-20 loudspeaker placed 0.1 m from the ipsifateral pinna while the contralateral ear was plugged with an ear mould compound (Otoform). Calibration of the acoustic stimulus was determined by replacing the animal with a half-inch condenser microphone (Bruel and Kjaer type 4134) positioned in approximately the same location as the ipsilateral pinna. The output of the microphone was fed into a measuring amplifier (Bruel and Kjaer 2615) and the peak equivalent sound pressure level (pe

207

SPL, rc 20 mPa) determined using the technique of Burkard (1984). The ma~mum sound pressure level at the ipsilateral pirma was 92 dB pe SPL. ABRs were recorded using subcutaneous needle electrodes (positive vertex; negative neck; ground abdomen). The scalp recorded potentials were amplified by 100 dB and filtered by a SO Hz notch filter and a band pass filter (Krohn-Hite 3750; high pass 150 Hz, 24 dB/octave; low pass 3 kHz, 6 dB/octave). ABRs were sampled at a rate of 10 kHz by a lo-bit analog-to-digital converter. Clicks were presented at 33 per second and the responses averaged over 500 trials. When the ABR threshold was berow 37 dB pe SPL, the animal’s hearing status was regarded as normal. Deafening procedure

After the ABR, KA (kanamycin monosulfate, Sigma) dissolved in normal saline was subcutaneously injected at a dose of 300 mg/ kg. Within 30-60 min following the injection of KA, EA (Ethacrynate sodium, MSD) dissolved in normal saline was intravenously infused via a slow infusion apparatus at a rate of 1 mg/minute. Unilateral click-evoked ABRs were continuously recorded during the EA infusion. When the ABR threshold exceeded 92 dB pe SPL, the EA infusion was stopped. The hearing status of the contralateral ear was then determined by recording ABRs. After recovery from the anaesthesia, each animal was returned to its quarters and placed on a normal diet. The hearing status of each animal was again monitored one and seven days after the procedure, and immediately before sacrifice. Monitoring of renal function

Renal function was monitored in the first four cats by collecting blood samples and mcasurin~ serum creatinine and urea concentrations. These results were compared with normal data for this species (Mitruka and Rawnsley, 1977). Blood samples were collected prior to and immediately after the deafening procedure, and at variable periods of up to three days following the procedure. After observing normal renal function in the first four animals this monitoring was not continued.

In order to examine the effects of the cochlear lesion as a function of survival time, animals were sacrificed at periods from two months to three years following the deafening procedure. Each animal was sacrificed with an overdose of sodium pentobarbital and then perfused intra-arterially with a 0.1 M phosphate buffered normal saline solution

followed by a fixative containing 1% paraformaldehydc and 1% glutara~dehyde in 0.1 M phosphate buffer. Both cochleas were collected and immersed in a phosphate buffered 2.5% glutaraldehyde solution. Both kidneys were also removed and immersed in 10% neutral buffered formalin for subsequent histology. Following the completion of fixation each cochlea was decalcified in 4% EDTA, embedded in Spurr’s resin and sectioned at a thickness of 3 pm in the horizontal plane. Sections were collected every 125 km and stained with haematoxylin and eosin. The histological examination of each cochlea was made by two independent observers examining all sections serially. Each cochlea was divided into six regions: lower basal (LB), upper basal (UB), lower middle (LM), upper middle (UMI, lower apical (LA) and upper apicat &JAI turns. The histol~)~ical examination included an evaluation of hair celt survival, the presence or absence of the supporting cells of the organ of Corti (pillar and Deiters’ cells), an estimation of the survival of dendrites (peripheral processes) within the osseous spiral lamina, and the survival of the spiral ganglion cells within Rosenthal’s canal. The degree of dendrite and spiral ganglion cell survival was divided into one of four broad categories (O-25%; 25-50%~; 50-75%; 7.5-100% of normal). The assesments made by the two observers were in close agreement. Gross examination of the kidneys was conducted and tissue samples from the kidneys were embedded in paraffin, sectioned at a thickness of 3 pm and stained with haematoxylin and eosin for microscopic examination. Both cortex and medulla were examined. The care and use of the animals reported on in this study were approved by the Animal Experimentation Ethics Committee of the University of Melbourne (‘Studies of Pediatric Auditory Prosthesis Implants’ NIH Contract NOl-DC-7-23421, and were performed in accordance with the principles of the National Health and Medical Research Council of Australia.

Results Hearing loss

The co-administration of KA and EA resulted in a severe hearing loss which developed soon after commencing the intravenous infusion of EA (Fig. If. Foilowing the subcutaneous injection of KA, and for a short period of time after the initiation of the intravenous infusion of EA, there was little change in the click-evoked ABR (Figs. 1 and 2). However, when the dose of EA was within a range of 10-2.5 mg/kg, the auditory thresholds suddenly rose dramatically to levels beyond the limits of our equipment (92 dB pe SPL; Fig. 2). Subsequent monitoring showed no evidence of

TABLE

I

Summary of age, weight and EA Cat

Age (weeks)

+EA

dose for each animal in this study a Weight

EA dose

(kg)

bvz/kg)

Kl K2 K3 K4

6 18 19 17

0.6 1.8 2.4

25.1

1.X

22.9

K.5

18

14.7

K6

30

2.5 2.9

K7

4

0.5

25.6

16.4 I s.0

10.0

* Note: All animals received 300 mg/kg KA.

I

I ’

0

II

I

1

1

2

III

a recovery in hearing for any animal. Immediately prior to sacrifice we confirmed that all ABR thresholds were above 92 dB pe SPL.

IV I

I

3 4 5 Time (msec)

I

I

6

’

,

8

7

0

Relationship between EA dose and body weight

Fig. 1. ABRs evoked from cat K2 in response to a 77 dB pe SPL acoustic click. ABRs were recorded prior (0 min) and periodically following co-administration of KA and EA. The intravenous infusion of EA commenced 30 min after a single S.C. injection of KA (300 mg/kg). Fifty minutes after the KA administration the amplitude of the ABR exhibited a reduction. Within a further 5 min the ABR was completely absent. Fifty six minutes following the KA administration the intensity of the clicks was increased to 92 dB pe SPL, however, this also failed to evoke a response. This hearing loss was permanent. The four waves of the cat ABR are illustrated in the figure. Vertexpositive potentials are plotted upward. Time (in minutes) following the KA administration is illustrated to the right of each response.

A summary of the age, weight and the EA dose for each animal in this study is presented in Table I. We found a highly statistically significant (P = 0.002) correlation between body weight and the EA dose (in mg/ kg) required to deafen these animals (r = -0.93; Fig. 3). This finding indicates that much of the individual variability observed in the present data are a result of significant variations in the size of the animals studied. Animal S general condition

I’l’l’l’l’l’l loo

0 K4 - OK5 80-D K2

3

All animals made an uneventful recovery from anaesthesia and survived the deafening procedure. There was no evidence of loss of balance or ataxia. Each animal was placed on a normal diet and all showed normal appetite. No animal displayed clinical evidence of renal insufficiency such as weight loss or oliguria following the deafening procedure.

‘I’I_ EA: 22.9 mg/kg EAz 14.?mg/kg EA: 16.4 mg/kg

S-i 2

60-

3 ; 40rm E

a

20-

3

20 .

z .P

o-, 0

I lo

I

I 20

*

1.1.1. 30 40 50 Time (minutes)

11 60

18 70

1 80

J

r = -0.93

+

-

.

e 6

l.O-

B %

Fig. 2. Time course for the development of the hearing loss of three representative cats is indicated by the elevation of the ABR threshold following KA/EA administration. KA was injeizted at 0 min in each example. Initiation of the EA infusion (1 mg/min) is indicated by the solid symbols. ABR thresholds remained stable for the first 15 to 20 min of the EA infusion, afterwhich they exhibited a rapid increase and were completely eliminated within a few minutes. At the end of the deafening procedure, thresholds were above 92 dB pe SPL as indicated by upwards arrows.

0.01, , , ,

(

,

,

.

IO

,

,

,

1

30

&se

of I% bng/kg)

Fig. 3. The relationship between animal body weight and the dose of EA (in mg/kg) required to produce a profound hearing loss. This relationship exhibited a highly significant correlation (r = - 0.931, indicating that much of the individual variability in the EA dose was a result of large variations in body weight.

200

Renal function and histology

Blood samples, taken both before and after the deafening procedure were collected from four of the cats tK2-K5) in order to establish serum creatinine and urea levels (Table II). While both the creatinine and urea levels were slightly elevated at one and three days after the procedure, they nevertheless remained within the normal range (Mitruka and Rawnsley, 1977). On the basis of these results, blood samples were not routinely collected from animals deafened later in the study (Kl, K6 and K7). Gross examination of the kidneys indicated a normal appearance. They exhibited a smooth and shiny external cortical surface without any evidence of fibrosis. Microscopic examination found no evidence of degeneration or necrosis of epithelial cells in the glomeruli and convoluted tubules. Moreover, there was no evidence of fibrosis in either cortex or medulla. Histopatholoby of the cochleu The hist~)pathological Table III.

results

are summarized

in

Hair cells Of the 14 cochleas examined, nine showed complete hair cell loss throughout all cochlear turns. The remaining five cochleas exhibited a few isolated inner and/or outer hair cells in the upper apical turn. It was interesting to note that this included both long-term survival animals (2.5 and 3.25 years; Table III). Supporting cells of the organ of Corti Preservation of the supporting cells (pillar and Deiters’ cells) of the organ of Corti was related to their position within the cochlea; supporting cells were always absent in the lower basal turn and frequently absent in the upper basal turn. The presence of supporting cells increased in an apical direction (Table III; Figs. 4 and 51, and were always preserved in the apical turn - even following post-deafening survival periods in excess of three years. A clear transition between the TABLE

II

Range of crcatinine (mmol/l):

Creatinine Urea Cats tested

and urea

level in serum

of four deafened

Predeafening

Immediately after deafening

1 day after deafening

3 days after deafening

0.07-0.09 7.0 -9.1 4

0.07-0.10 7.6 -9.7 4

0.0x- 0.09 9.5 -10.9 2

0.0’) 9.9 1

Notes: Normal range is 0.04-0.23 mmol/l for creatinine mmol/l for urea from Mitruka and Rawnsley (1977).

and _5.0-

cats

1I .h

absence and presence of the supporting cells was not observed. Instead, there usually existed a small transitional region along the basilar membrane where the supporting cells of the organ of Corti had collapsed onto the basilar membrane (e.g. LM turn Figs. 4 and 5). Comparison between the short-term surviving animals (KI-K5) which were sacrificed within six months of the deafening procedure, with the long-term surviving animals (K6 and K7), indicated that the extent of degeneration of the supporting cells increased with the length of the survival period (Table III). Finally, the degree of supporting cell survival showed close bitatera1 symmetry within individual animals. Dendrites The degree of survival of dendrites was related to their location within the cochlea and the durati~)n of the survival period following the deafening procedure. While the majority of cochleas exhibited an apicalward increase in dendrite survival (e.g. compare LB and LA turns Fig. 41, some cochleas clearly showed maximum loss within the upper basal or middle turns (e.g. K3, Table III). Animals with a post-deafening survival period of less than six months exhibited a similar degree of dendrite survival (Kl-K5. Table III). A far greater loss of dendrites was evident in the case of the two long-term surviving cats (K6 and K7, Table III). This was observed even in regions where the supporting cells tif the organ of Corti were present (Table III; Fig. 5). Finally, the survival of dendrites generally exhibited close bilateral symmetry. Spiral ganglion celfs In general, the survival of spiral ganglion cells reflected the pattern of dendrite survival. Spiral ganglion cells increased in population apicalward from the basal turn (Figs. 4 and 5; Table III), although - again consistent with the dendrites - some cochleas exhibited greater ganglion cell loss in the upper basal or middle turns (e.g. Kl, Table III), and we occasionally observed quite high ganglion cell survival in the extreme base of the cochlea. Comparison between the short-term (KlK5) and long-term surviving animals 1K6 and K71, showed an increased spiral ganglion cell loss as a function of the post-deafening survival time (compare Figs. 4 and 5). The relatively large differences in ganglion cell survival between these short and longer term survival animals clearly indicates that the process of ganglion cell degeneration continues for many years after the initial insult. It was of interest to note that the degeneration of spiral ganglion cells in the short-term surviving animals (KI-K5) appeared to be more severe in the peripheral aspect of Rosenthal’s canal than in the area closer to the modiolus (e.g. LM turn, Fig. 4). Ganglion cell

210

longer-term survival animals (K6 and K7) where only small numbers of ganglion cells remained (e.g. Fig. 5).

Stria vascular-is There was no obvious evidence of pathology of the

stria vascularis at the light microscope level. The thickness of the stria vascularis was in the normal range (Russell et al., 1979). The marginal cells, intermediate cells and basal cells were opposed to one another and there was no evidence of enlarged intercellular space (Fig. 4 and 5).

TABLE III Summary of cochlear histopathology --Turns

Hair cells

Cat Kl LB UB LM UM LA UA Cat K2 LB UB LM UM LA UA Cat K3 LB UB LM LM LA UA Cat K4 LB UB LM UM LA UA Cat K5 LB UB LM UM LA UA Cat K6 LB UB LM UM LA UA Cat K7 LB UB LM UM LA UA

Age deafened: 6 weeks

-

-

Supporting cells

-

-

Age deafened: 4 months _ _ _ _ _ _ _ _

Dendrite Survival

Time of sacrifice following deafening: 2 months

+ f + + +

***

-/+ + + + +

**** *** *** *** ****

Spiral Ganglion cell Survival

*** *** *** *** **** ****

Time of sacrifice following deafening: 3 months *** -/+ ** **** *** + + *** *** + + *** *** + + **** **** + + **** *** + + Age deafened: 4.5 months Time of sacrifice following deafening: 4 months **** ** ** ** _ _ _ _ ** -/+ ** -/+ ** ** + + *** *** + + *** **** _ + + -/+ Age deafened: 4 months Time of sacrifice following deafening: 4 months **** *** **** *** _ _ + -/+ **** *** _ --/+ -/+ **** *** + **** **** _ _ + + *** N N + N Age deafened: 4 months Time of sacrifice following deafening: 6 months **** *** -- / + -/+ **** *** + + **** *** _ _ + + **** **** + + **** **** + + **** + + N -/+ Age deafened: 7 months Time of sacrifice following deafening: 2 years and 6 months * * _ _ _ _ * * _ _ _ _ * * _ _ -/+ * _ + -/+ * ** ** + + ** ** + + -/+ -/+ Age deafened: 4 weeks Time of sacrifice following deafening: 3 years and 3 months * _ _ _ _ _ * _ _ * _ * _ _ * + + * * + + -/+

**** **** *** *** **** **** **** **** **** i*** **** **** **** *** *** *** **** **** **** **** **** **** **** N **** **** **** **** **** **** * i * * ** ** * * * * * *

**** c** *** *** **** ****

**** **** **** a*** **** **a*

**** *** *** *** **** ****

**** **** **** **** **** ****

**** **** **** **** ****

N * 1. * 1; ** ***

% * * * * *

Notes: Both cochleas of each cat are presented. Results from the left cochlea are presented on the left side of each column. L lower, U upper, B basal, M middle, A apical, - absent, + present, - /+ present in some sections of that region, * l-25%, * * 25-50%, * ** 50-75%, * * * * 75-100%. N not available.

Fig. 4. Composite photomicrograph of the mid-modiolar section of the right cochlea of cat K3, which was sacrificed four months after deafening, showing lower apical (LA), lower middle (LM), and lower basal (LB) turns. Hair cells were absent in all but the most apical region of the cochlea where they were observed only occasionally. Within the basal turn, the organ of Corti was completely absent. In the LM turn, the support cells of the organ of Corti were present although they had collapsed onto the basilar membrane (arrow). More apicalward the support cells of the organ of Corti appeared morphologically normal. The dendrites and spiral ganglion perikaryon typically showed an increase in survival rate apicalward. Moreover. there appeared to be more extensive spiral ganglion cell loss in the peripheral aspect of the Rosenthal’s canal than in the area closer to the modiolus (double arrow, LM turn). bar = 100 pm

213

Tectorial membrane In the cats which survived for less than six months, the tectorial membrane showed normal position and shape, whereas in the cats which survived for two-three years (K6 and K7) the tectorial membrane in both the basal and middle turns was either not present or was detached from the spiral limbus (LB and LM turns, Fig. 5).

Discussion Profoundly detrf anitnal model The major aim of this study was to provide a safe and effective technique to create profoundly deaf animal models. Monitoring both the cochlear microphonic and hair cell loss, Brummett et al. (1979) demonstrated that the slope of the EA dose-response curve was significantly steeper than that of KA. Based on Brummett’s results we maintained the dose of KA at 300 mg/ kg and varied the dose of EA to levels that were sufficient to just deafen each animal. Our results showed that when the dose of EA reached lo-25 mg/kg, the animal exhibited a rapid and profound hearing loss as monitored by ABR’s. This hearing loss was permanent and bilaterally symmetrical. Moreover, the animals showed no evidence of nephrotoxicity following the procedure. In addition to its safety and effectiveness, this procedure can produce a hearing loss in a short period of time - typically less than two hours. Furthermore, the post-deafening management of the animal requires no special procedures - for example, there was no need to restrict the animal’s intake of food or water. These factors contribute to minimize any stress on the animals. This technique has been used in our Department to create deaf animal models for long-term electrical stimulation studies (e.g. Matsushima et al., 1991). However, when the same technique was used to deafen macaque monkeys (Shepherd, et al., 1994), these animals only exhibited a moderate high frequency hearing loss. Moreover, the majority of monkeys showed evidence of severe acute renal failure that required careful management. This observation highlights the varia-

tions in the response of different species to nephroand ototoxic drugs. While we monitored the KA/EA induced hearing loss using non-frequency specific stimuli. our histological results, together with results following KA/EA co-administration in neonatal kittens and adult monkeys using tonal stimuli (Shepherd et al.. 1993; Shepherd et al., 1994) suggests that the lesion progresses apicalward from the base of the cochlea. This is consistent with the pathophysiological response induced following long-term systemic injection of aminoglycosides (Hawkins, 1959; Stebbins et al., 1969; Ylikoski, 1974; Shepherd and Clark, 1985). In addition, unlike neomycin and dihydrostreptomycin, the KA/EA induced hearing loss appears to be stable over periods of at least several months (Shepherd et al., 1993; Shepherd et al., 19941, and is consistent with early clinical reports of KA/EA induced hearing loss (Mathog and Klein, 1969; Johnson and Hamilton, 1970). Given the nature of the hearing loss, together with its long-term stability, it may be possible to reliably produce animal models with specific high frequency lesions by monitoring the animals hearing status during EA infusion using high frequency tones. Although there was considerable variation in the dose of EA required to deafen the animals in the present study, we found a highly statistically significant correlation between body weight and EA dose (Fig. 3). It would appear that much of the individual variability observed in the present data were a result of the large variation in size of the animals studied. Furthermore, the results indicate that the optimum dose of EA is a more complex relationship than the ratio of the weight of EA to the animal’s body weight. It would appear that the total amount of EA administered is also an important factor.

Cochlear histology The most striking feature of the cochlear histology was the complete absence of hair cells in the majority of the cochleas. Another important feature was the absence of supporting cells of the organ of Corti in the lower basal turn. Their preservation increased in an apical direction, and they were always present in the apical turn. The transitional region along the basilar

Fig. 5. Composite photomicrograph of the mid-modiolar section of the right cochlea of cat K6, which was sacrificed two and half years after deafening. As in Fig. 4. hair cells were absent in all but the most apical region of the cochlea where they were only occasionally observed. Within the basal turn the organ of Corti was absent and the tectorial membrane was detached from the spiral limbus (arrow head). In the LM turn. the organ of Corti had collapsed onto the basilar membrane and appeared to be in the process of degeneration (arrow). In the apical turn. the supporting cells were present and the tectorial membrane was in its normal position. The dendrites and spiral ganglion perikaryon exhibited an increase in survival rate apicalward. This cochlea exhibited a more extensive neural loss than K3 (Fig. 4). bar = 100 pm.

membrane, where the supporting cells had collapsed onto the basilar membrane, were presumably in the process of degeneration (Figs. 4 and 5). There was generally a reasonable correlation between the status of the supporting cells of the organ of Corti and the survival of spiral ganglion cells and their dendrites. This is consistent with Schuknecht’s findings that the extent of loss of cochlear neurons appears to parallel closely the magnitude of injury suffered by the supporting cells (Schuknecht, 1974). There was a general trend of increasing neuronal survival in the apical direction, although in some cochleas the survival of neurons in the extreme base of the cochlea was also relatively high in the absence of supporting cells of the organ of Corti. There was also a correlation between the duration of the deafening period and neural survival, i.e., the longer the post-deafening survival period, the greater the number of neural elements that had degenerated. It is apparent that progressive neural degeneration continued long after the initial deafening procedure. Indeed, the present findings suggest that the degenerative process continues months, and in fact years, following the initial insult. This finding is consistent with Leake and Hradek’s (1988) observations following long-term survival of neomycin deafened cats. Therefore, survival period following the deafening procedure is an important consideration when deaf animal models having specific ganglion cell populations are required. Significantly, there also appears to be a close relationship between post-deafening survival time and the number of spiral ganglion cells in human temporal bone studies (Nadol et al., 19891. We observed more severe degeneration of spiral ganglion cells in the peripheral aspect of Rosenthal’s canal than in the area closer to the modiolus. While such a degenerative pattern has has also been observed by others (H.F. Schuknecht, personal communication), the underlying mechanism is not clear. The present results illustrate a general bilateral symmetry associated with a KA and EA induced hearing loss. This is a common finding in animal studies where aminoglycosides have been administered systemically (Stebbins et al., 1969; Shepherd and Clark, 198.5; Leake and Hradek, 19881, and indicates that the contralateral cochlea in this type of deaf animal model can serve as an appropriate control. Finally, we have demonstrated that the use of 300 mg/ kg of KA and EA at levels of up to 25 mg/ kg can safely deafen cats. In studies where the use of ABR monitoring is not feasible, e.g. deafening neonatal kittens, the use of 300 mg/kg KA and 25 mg/kg EA has safely and routinely induced a profound hearing loss (Shepherd et al., 1993). As we have observed in the present study, the dose of EA could possibly be reduced in larger animals.

Acknowledgements

This study was funded by the National Institutes of Health, USA (NIH Contract NOl-DC-7-2342). Ethacrynic acid (Ethacrynate Sodium) was kindly supplied by Dr. L. McKenzie, Meek Sharp and Dohme (Aust.) Pty Ltd. We gratefully acknowledge the contributions made by Mr. R. Millard for engineering support, Ms. J. McNaughtan and Ms. M. Clarke for histological support, Mr. H. Neo for biochemistry analysis, Ms. D. Cook for animal technical assistance, and three anonymous reviewers for their helpful comments on an earlier draft of this manuscript.

References Bredberg, G. (1973) Experimental pathology of noise-induced hearing loss. Adv. Oto-Rhino-Laryngol. 22, 102-114. Brown, M., Shepherd, R.K., Webster, W.R., Martin, R.L. and Clark, G.M. (1992) Cochleotopic selectivity of a multichannel scala tympani electrode array using the 2-deoxyglucose technique. Hear. Res. 59, 224-240. Brummett, R.E., Brown, R.T. and Himes, D.L. (1979) Quantitative relationships of the ototoxic interaction of kanamycin and ethacrynic acid. Arch. Otolaryngol. 105, 240-246. Brummett, R.E. and Fox, K.E. (1982) Studies of aminoglycoside ototoxicity in animal models. In: A. Whelton and H.C. Ness (Eds.), The aminoglycosides: microbiology, clinical use and toxicological complications. Marcel Dekker, New York, pp. 419-452. Burkard, R. (1984) Sound pressure level measurement and spectral analysis of brief acoustic transients. Electroenceph. Clin. Neurophysiol. 57, 83-91. Cody, A.R. and Robertson, D. (1983) Variabihty of noise-induced damage in the guinea pig cochlea: Electrophysiological and morphological correlates after strictly controlled exposures. Hear. Res. 9, 55-70. Duckert, L.G. (1983) Morphological changes in the normal and neomycin-perfused guinea pig cochlea following chronic prosthetic implantation. Laryngoscope 93, 841-855. Fernandez, C. and Hinojoas, R. (1974) Postnatal development of endocochlear potential and stria vascularis in the cat. Acta Otolaryngol. 78, 173-186. Harrison, R.V., Shirane, M., Fukushima, N. and Mount, R.3. (1991) Morphological changes to the cochlea in an animal model of profound deafness. Acta Otolaryngol. (Stockh.) Suppl. 489, S-11. Hawkins, J.E. (19.59) The ototoxicity of kanamycin. Ann. Otol. Rhinol. Laryngol. 68, 698-715. Johnson, A.H. and Hamilton, C.H. (1970) Kanamycin ototoxicity. Possible potentiation by other drugs. Southern Med. J. 63, Sll513. Leake-Jones, P.A., O’Reilly, B.F. and Vivion, M.C. (1980) Neomycin ototoxicity: ultrastructural surface pathology of the organ of Corti. SEM/1980/III, 427-434. Leake-Jones, P.A., Vivion, M.C., O’Reilly, B.F. and Menenich, M.M. (1982) Deaf animal models for studies of a multichannel cochlear prosthesis. Hear. Res. 8, 225-246. Leake, P.A. and Hradek, G.T. (1988) Cochlear pathology of long term neomycin induced deafness in cats. Hear. Res. 33, 11-34. Mathog, R.H. and Klein, W.J. (1969) Ototoxicity of ethacrynic acid and aminoglycoside antibiotics in uremia. N. Engl. J. Med. 280, 1223-1224.

Matsushima. J.-l., Shepherd, R.K., Seldon, H.L., Xu, S.A. and Clark. G.M. (1991) Electrical stimulation of the auditory nerve in deaf kittens: effects on cochlear nucleus morphology. Hear. Res. 56, 133-142. Mitruka, B.M. and Rawnsley, H.M. (1977) Clinical biochemical and hematological reference values in normal experimental animals. Mason Publishing Inc., New York, pp. 145-148. Nadol, J.B.. Young, Y.S. and Glynn. R.J. (1989) Survival of spiral ganglion cells in profound sensorineural hearing loss: implications for cochlear implantation. Ann. Otol. Rhinol. Laryngol. 98, 411-416 Orsulakova. A. and Schacht, J. (1981) A biochemical mechanism of the ototoxic interaction between neomycin and ethacrynic acid. Acta Otolaryngol. (Stockh.) 93, 43-48. Pujol, R. and Marty, R. (1970) Postnatal maturation in the cochlea of the cat. J. Comp. Neurol. 139, 115-126. Remand, R. and Romand, M.-R. (1986) Perinatal growth of spiral ganglion cells in the kitten. Hear. Res. 21. 161-165. Russell, N.J. Fox. K.E. and Brummett, R.E. (1979) Ototoxic effects of the interaction between kanamycin and ethacrynic acid. Acta Otolaryngol. (Stockh.) 88, 369-381. Schuknecht, H.F. (1984) Disorders of innervation. In: H.F. Schuknecht (Ed.). Pathology of the ear. Harvard University Press, 331-333. Shepherd, R.K. and Clark, G.M. (1984) Progressive hearing loss following neomycin infusion in the cat cochlea. Proc. Aust. Physiol. Pharmacol. Sot. 14. 75~. Shepherd, R.K. and Clark, G.M. (1985) Progressive ototoxicity of neomycin monitored using derived brainstem response audiometry. Hear. Res. 18, 105-110. Shepherd. R.K., Xu, S.A.. Tong, Y.C., Maffi, CL., Clark. G.M., Franz, B.G.K-H., Millard, R.E., Wills, R.A. Kuzma, J.A. and Patrick. J.F. (1988) Studies on Pediatric Auditory Prosthesis Implants. Third Quarterly Progress Report, N.I.H. Contract NOl-DC-7-2342. Department of Otolaryngology, University of Melbourne.

Shepherd, R.K., Martin. R.L. and Clark. G.M. (1993) Onset of ototoxicity in the cat is related to the onset of auditory function. Proc. Aust. Neurosci. Sot. 4, 112. Shepherd, R.K., Xu, S.A. and Clark, G.M. (1994) Partial hearing loss in the macaque following a single co-administration of kanamycin and ethacrynic acid. Hear. Res. (in press). Simmons, F.B. (1967) Permanent intracochlear electrodes in cats. Tissue tolerance and cochlear microphonics. Laryngoscope 77, 171-186. Stebbins, W.C. Miller, J.M., Johnsson, L.G. and Hawkins, J.E. (1969) Ototoxic hearing loss and cochlear pathology in the monkey. Ann. Otol. Rhinol. Laryngol. 78, 1007-1025. Sutton, D. and Miller, J.M. (1983) Cochlear implant effects on the spiral ganglion. Ann. Otol. Rhinol. Laryngol. 92. 53-58. Syka, J. and Melichar. 1. (1984) Comparison of the effects of furosemide and ethacrynic acid upon the cochlear function in the guinea pig. In: Ototoxic side effects of diuretics. Scan. Audiol. Suppl. 14, 63-69. Tran Ba Huy, P., Manuel, C., Meulemans, A., Sterkers, O., Wassef, M. and Amiel, C. (1983) Ethacrynic acid facilitates gentamycin entry into endolymph of the rat. Hear. Res. 11, 191-202. Walsh, E.J. and McGee, J. (1987) Postnatal development of auditory nerve and cochlear nucleus neuronal responses in kittens. Hear. Res. 28, 97-116. West, B.A., Brummett, R.E. and Himes. D.L. (1973) Interaction of kanamycin and ethacrynic acid. Arch. Otolaryngol. 98, 32-37. Xu. S.A.. Shepherd, R.K. and Clark. G.M. (1990) A rapid and permanent hearing loss in cats following co-administration of kanamycin and ethacrynic acid. Proc. Aust. Physiol. Pharmacol. sot. 21, 44. Yamane, H.. Nakai, Y. and Konishi K. (1988) Furosemide-induced alteration of drug pathway to cochlea. Acta Otolaryngol. (Stockh.) Suppl. 447, 28-35. Ylikoski. J. (1974) Correlative studies on the cochlear pathology and hearing loss in guinea-pigs after intoxication with ototoxic antibiotics. Acta Otolaryngol. (Stockh.) Suppl. 326. l-62.