Full text of DOI:10.1097/00005537-200204000-00018

The Laryngoscope
Lippincott Williams & Wilkins, Inc., Philadelphia
© 2002 The American Laryngological,
Rhinological and Otological Society, Inc.
Intentional Ablation of Vestibular Function
Using Commercially Available Topical
Gentamicin-Betamethasone Eardrops in
Patients With Meniere’s Disease: Further
Evidence for Topical Eardrop Ototoxicity
Daniel M. Kaplan, MD; Sukhminderjit S. Hehar, MB, BCh, FRCS; Manohar L. Bance, MB, BCh, FRCSC;
John A. Rutka, MD, FRCSC
Objective: To determine whether the controver-
sial findings of suspected ototoxicity from commer-
cially available gentamicin sulfate and betametha-
sone sodium phosphate eardrops can be used in a
therapeutic fashion to ablate (or attenuate) vestibu-
lar function in patients with unilateral Meniere’s dis-
ease. Study Design: Prospective case review. Methods:
At a tertiary care dizziness unit at the University
Health Network, Toronto General Hospital, Univer-
sity of Toronto (Toronto, Ontario, Canada), adults
with unilateral Meniere’s disease undergoing intra-
tympanic ablation therapy were studied. After inser-
tion of a tympanostomy tube with the patient under
local anesthesia, patients instilled gentamicin con-
taining eardrops three times daily until they became
vertiginous for longer than 24 hours and then for an
additional 2 days longer or for 1 month, whichever
came first. Electronystagmographic caloric test re-
sponses were measured before treatment using bi-
thermal water caloric and after treatment using
air caloric tests. Main outcome measures included
clinical titration of drops to the onset of pro-
longed vertigo. As well, post-treatment findings
on electronystagmography and audiometry were
compared with pretreatment testing. Results:
Twenty patients were available for review. Fif-
teen patients had a significant reduction in ca-
loric test responses compared with pretreatment
values; among them, 10 patients had absent air
caloric test responses on the treated side. In 10
patients hearing worsened according to the 1995
American Academy of Otolaryngology—Head and
Neck Surgery Committee on Hearing and Equilibrium
Guidelines for reporting in Meniere’s disease. Con-
clusions: Topical gentamicin-betamethasone ear-
drops can pass through a tube into the middle ear,
where they may prove primarily vestibulo-ototoxic
patients with Meniere’s disease. The study further
confirms clinical observations that gentamicin-
containing eardrops might prove ototoxic, especially
in noninflamed ears with a tympanic membrane de-
fect. Key Words: Gentamicin ablation, intratympanic
gentamicin, topical gentamicin, Meniere’s disease,
ototoxicity.
Laryngoscope, 112:689–695, 2002
INTRODUCTION
Eardrops containing gentamicin or other related ami-
noglycosides are commonly prescribed by primary care
physicians and otolaryngologists for treatment of external
otitis, chronic suppurative otitis media, and discharging
tympanostomy tubes. Although topical aminoglycoside
ototoxicity appears to be infrequent, there has been a
genuine concern that this might develop if eardrops were
to reach the middle ear in the presence of a tympanic
membrane (TM) perforation or defect.1 Until recently, at-
tention had been primarily focused on possible cochlear
ototoxicity, based on a small number of largely anecdotal
case reports describing patients with sensorineural hear-
ing loss following application of eardrops.2 For this reason,
cautionary warnings have been provided for most ototopi-
cal agents, advising against their use in the presence of a
TM perforation. Despite this concern, as of 1992, most
otolaryngologists in the United States felt comfortable
From the Department of Otolaryngology, University Health Net-
work, Toronto General Hospital, University of Toronto, Toronto, Ontario,
Canada.
Supported by fellowships from the Thomas Wickham Jones Founda-
tion (S.S.H.) and the Canadian International Scientific Exchange Program
(D.M.K.).
Editor’s Note: This Manuscript was accepted for publication Novem-
ber 26, 2001.
Send Correspondence to John A. Rutka, MD, FRCSC, University
Health Network, Toronto General Hospital, 200 Elizabeth Street EN7-222,
Toronto, Ontario, Canada M5G 2C4. E-mail:jrutka@home.com
Laryngoscope 112: April 2002
Kaplan et al.: Ablation of Vestibular Function
689

prescribing aminoglycoside-containing drops for a drain-
ing perforation or for infective tympanostomy tubes.3
The development of vestibular ototoxicity associated
with systemic gentamicin administration has been a well-
recognized phenomenon since its introduction in the
1960s. Paradoxically, this property of gentamicin has been
used therapeutically since the late 1970s for the topical
ablation of vestibular function in the treatment of unilat-
eral Meniere’s disease. Different treatment protocols in
Meniere’s disease have been introduced subsequently,
with vertigo control rates generally in the order of 80% to
100%.4–16
The suspicion that the prolonged clinical use of com-
mercially available gentamicin-containing eardrops might
result in inadvertent vestibular ototoxicity has been doc-
umented recently. In 1999, Bath et al.17 updated a study
of 16 patients with suspected topical ototoxicity from Ga-
rasone eardrops (1 mL contains 3 mg gentamicin sulfate
and 1 mg betamethasone sodium phosphate (Schering,
Canada Inc., Pointe-Claire, Quebec, Canada). Patients ei-
ther had perforated tympanic membranes or tympanos-
tomy tubes in place. All had applied the drops for more
than 7 days before developing a vestibular loss. The study
also reported on one patient with unilateral Meniere’s
disease who underwent deliberate vestibular ablation
with the same commercially available drops. The patient
had normal pretreatment electronystagmographic (ENG)
caloric test responses and achieved an absent response
with air caloric tests after treatment.
As a result of these findings, we were encouraged to
further investigate the ototoxic potential of a common,
commercially available gentamicin-containing ototopical
preparation in a select group of patients with incapacitat-
ing unilateral Meniere’s disease. The objectives of the
present study were threefold: 1) to determine whether
commercially available gentamicin containing eardrops
could access the middle ear through a conventional tym-
panostomy tube, 2) to determine whether these drops
could result in either vestibular or cochlear ototoxicity,
and 3) to serve as a foundation for future studies to assess
the potential therapeutic role of commercially available
gentamicin-containing eardrops for patients with incapac-
itating Meniere’s disease.
Audiovestibular Testing
A pretreatment audiogram and electronystagmogram with
bithermal water caloric testing was obtained before the intended
procedure. These two tests were repeated, typically, 1 to 4 months
after treatment. Audiometry included pure-tone testing, speech
reception thresholds, and speech discrimination scores. The pure-
tone average (PTA) was calculated using frequencies of 0.5, 1.0,
2.0, and 3.0 kHz. Change in hearing was defined by the 1995
AAO-HNS CHE criteria, which considered a change of 10 dB or
more or a change in word recognition score (speech discrimina-
tion) of 15% or more to be clinically significant.18
Pretreatment bithermal ENG caloric test responses were
performed, and the excitability difference (ED) between the af-
fected and nonaffected sides calculated. Post-treatment caloric
test responses were measured using ENG air caloric tests (ears
were stimulated alternately with air at 5°C for 30 seconds) (model
NCA-105, ICS Caloric, Addison, IL), because of the tympanos-
tomy tube,18 and the ED was calculated.19 In our center, an ED of
0% to 15% was considered within normal limits, and an ED of
16% to 25% was considered a mild, 26% to 50% a moderate, 51%
to 90% a severe, and greater than 90% a profound caloric reduc-
tion.20 No response after treatment implied no discernible re-
sponse to air caloric testing above any spontaneous nystagmus in
the caloric test position.20 A vestibular ototoxic effect was consid-
ered when there was a deterioration in one or more ED groupings.
Treatment Protocol
A tympanostomy tube (Sheppard Grommet Vent Tube,
1.14-mm internal diameter, Smith & Nephew Inc., Bartlett, TN)
was inserted into the posteroinferior quadrant of the TM under
topical anesthesia (phenol or eutotic mixture of local anesthetics
[EMLA], Astra Pharma Inc., Mississaugua, Ontario, Canada).
Topical gentamicin-betamethasone eardrops were prescribed,
and patients were instructed to instill three or four drops (ap-
proximately 0.1–0.15 mL) three times daily, followed by alternat-
ing tragal pressure against the ear canal to “pump” the drops
while lying with the affected ear upright for 15 minutes. The
patients were instructed to continue until prolonged vertigo or
imbalance occurred for longer than 24 hours and then to continue
drop instillation for another 2 days or for a maximum of 1 month,
whichever came first.
RESULTS
Between April 1999 and November 2000, 23 patients
in total were treated with intratympanic commercially
available gentamicin-betamethasone ototopical drops.
Three patients were not available for post-treatment
follow-up. However, these three patients were asymptom-
atic for further vertigo when contacted by means of a
telephone interview 2 to 8 months after treatment.
Twenty patients in total were available for review (11
men and 9 women). Mean age was 50.1 years. The right
ear was affected in 9 patients, and the left in 11. Patients
had symptoms of Meniere’s disease for a duration of 1 to
13 years. Three patients included in the present study had
previously undergone intratympanic gentamicin therapy
that had failed to control their vertigo and were included
to further assess the effects in the present study on au-
diovestibular function. Two of the three patients (patients
11 and 18) had received treatment in the same ear using
a fixed-treatment protocol with concentrated intratym-
panic gentamicin (24 mg gentamicin/mL injected 11/2 cc
three times daily for 4 days) 4 to 6 years previously. The
third patient (patient 17) had undergone the same treat-
PATIENTS AND METHODS
All patients were seen in the Department of Otolaryngology,
University Health Network, University of Toronto (Toronto, On-
tario, Canada) by the senior authors (J.A.R. and M.L.B.). All had
definite unilateral Meniere’s disease as defined by the 1995
American Academy of Otolaryngology—Head and Neck Surgery
Committee on Hearing and Equilibrium (AAO-HNS CHE) Guide-
lines for the Diagnosis and Evaluation of Therapy in Meniere’s
Disease.18 Most patients initially received medical therapy that
could have included dietary salt restriction, diuretics, and beta-
histine dihydrochloride (Serc, Solvay Pharma, Scarborough, On-
tario, Canada) for a minimum of 6 months and had achieved
limited or no control of their vertiginous attacks. Treatment with
topical intratympanic, commercially available gentamicin-
betamethasone eardrops was initiated between April 1999 and
November 2000.
Laryngoscope 112: April 2002
690
Kaplan et al.: Ablation of Vestibular Function

ment with these same drops 6 months earlier without
success. This patient did not initially develop symptoms
suggestive of ototoxicity after 3 weeks and subsequently
discontinued treatment. She continued to demonstrate a
mild vestibular loss on post-treatment ENG air caloric
testing with no change in ED groupings or her hearing. No
patient in the present study had previously undergone
other surgery for Meniere’s disease.
The majority of patients developed symptoms in the
first 2 weeks after starting treatment. Duration of treat-
ment was 9 to 28 days (mean duration, 15 days). Patients
recalled their symptoms as being compatible with vertigo,
imbalance, lightheadedness, or nausea, or a combination
of these. Post-treatment clinical assessments, as well,
sought signs (not listed) that specified completion of
therapy.10
Fig. 1. Caloric test responses before and after treatment in 20
patients with Meniere’s disease who were treated with a Garasone
ablation protocol through a tympanostomy tube.
Table I lists the patients and the observed ototoxic
effects following the instillation of commercially available
gentamicin-betamethasone eardrops.
patients realized a significant decrease in ENG caloric test
responses on the affected side, as reflected by increasing
excitability differences or an absent response after treat-
ment. Three patients demonstrated no or minimal change
in ED, and in two patients (patients 6 and 12) there was
an apparent improvement.
Vestibulotoxicity
Table I and Figure 1 list and demonstrate, respec-
tively, the comparative caloric ENG results before and
after treatment. Before treatment, six patients had an ED
greater than 50% and no patient had an absent caloric
response. After treatment, 5 patients had an ED greater
than 50% and 10 had absent caloric responses. In total, 15
Cochlear Ototoxicity
Pretreatment hearing acuity was variable, ranging
from normal to no measurable hearing. Ten patients had
TABLE I.
Patients’ Pre- and Posttreatment Audiometry and ENG Caloric Test Results.
Pretreatment
Posttreatment
Patient
No.
ENG Calorics*
ENG Calorics*
Age/Sex
ED (%)
PTA
SDS%
ED (%)
PTA
SDS%
Ototoxic Effects
1
2
63/F
75/M
51/F
38/F
30/M
56/M
39/M
47/F
61/M
44/M
62/M
30/F
58/M
49/F
71/M
54/M
22/F
34/M
69/F
72/F
30
68
0
53
100
40
80
15
13
60
15
40
67
57
60
75
50
70
53
80
38
40
50
CNT
CNT
72
Absent‡
Absent
90
55
100
22
100
30
15
100
13
30
22
55
80
95
68
53
70
78
55
45
50
16
CNT
92
Vest. loss
Vest. loss
3
Vest. loss & hearing improvement
Vest. & hearing loss
Vest. & hearing loss
Improved vest. function
Vest. & hearing loss
Vest. loss
4
35
45
51
22
19
29
36
75
64
75
20
28
18
22
44
80
9
12
Absent
86
14
5
100
96
100
96
6
4
7
40
Absent
Absent
Absent
Absent
52
CNT
96
8
88
9
72
92
Vest. loss
10
11
12
13
14
15
16
17
18
19
20
CNT
32
88
Vest. loss & hearing improvement
Hearing im provement
Improved vest. function & hearing loss
Vest. & hearing loss
Vest. & hearing loss
Hearing improved
64
CNT
CNT
36
5
CNT
CNT
16
Absent
66
8
40
32
12
72
0
Vest. & hearing loss
Hearing loss
32
18
12
72
Absent
Absent
9
80
Vest. & hearing loss
Vest. & hearing loss
Vest. Loss
72
25
10
0
*Pretreatment ENG calorics were performed with alternate bithermal water irrigation.
†Posttreatment ENG calorics were performed with air caloric method.
‡An ED of 100%. CNT-could not be tested; Vest ϭ vestibular.
ENG ϭ electronystagmography; ED ϭ excitibility difference; PTA ϭ pure tone average (0.5, 1.0, 2.0, 3.0 KHz); SDS% ϭ speech discriminations score.
Laryngoscope 112: April 2002
Kaplan et al.: Ablation of Vestibular Function
691

worsening of hearing (in patient 5, this occurred mainly in
the high-frequency range greater than 4 kHz). In four
patients (patients 3, 10, 11, and 15) hearing improved,
whereas in the remaining six patients no change was
noted. In the small subset of three patients who had been
previously treated with topical gentamicin, the PTA wors-
ened in one patient (patient 18) and speech discrimination
worsened in another patient (patient 17) and improved in
one patient (patient 11). Average follow-up was 7 months.
Except for ototoxicity no other complications were
encountered.
Meniere’s disease should be reported under the vertigo
control rate, functional level, and hearing status, the most
important aspect of reporting being the requirement of a
minimum 2-year follow-up. Nevertheless, our goal in the
present study was to primarily demonstrate in a prospec-
tive fashion that commercially available gentamicin-beta-
methasone–containing eardrops can prove ototoxic if used
in a prolonged fashion when a tympanostomy tube is
present.
In the past, reports of suspected ototopical gentami-
cin ototoxicity have been largely criticized for being anec-
dotal and retrospective. Because pretreatment caloric
function was largely unknown, there was legitimate con-
cern about whether the suspected changes in vestibular
function actually arose from the use of these drops. Hav-
ing performed both pretreatment and post-treatment au-
diometric and caloric testing in the present study, we
think we have overcome some of these criticisms.
In the present study we evaluated whether a com-
monly prescribed commercially available ototopical prep-
aration containing gentamicin and betamethasone could
penetrate the middle ear through a tympanostomy tube,
be absorbed into the inner ear, and subsequently result in
ototoxicity if used in a prolonged clinical fashion. Findings
supportive for an ototoxic event related to this interven-
tion included a clinical history of prolonged vertigo lasting
days during treatment, post-treatment signs of vestibular
deafferentation (e.g., the presence of nystagmus after
head shake or a positive result on head-thrust or Hal-
magyi maneuver),10 and changes in the pretreatment and
post-treatment audiometry and ENG caloric testing.
DISCUSSION
Following preliminary reports from Beck and
Schmidt4 and Schmidt and Beck,5 numerous authors have
reported on the therapeutic use of intratympanic genta-
micin in the treatment of unilateral Meniere’s disease.6–16
As shown in Table II, treatment protocols have varied
with respect to frequency, concentration of gentamicin
(6–40 mg/mL), total dosage and duration (i.e., 1–12 doses
of 6–40 mg gentamicin delivered three times daily to once
every 2 weeks), type of delivery system used (e.g., tran-
stympanic injections, injections through a tympanostomy
tube or ventilation tube, catheter injections), and how the
end point of therapy was determined (fixed vs. titration
protocol).
It should be clearly stated that our study was not
intended to report results of this protocol for symptom
control in Meniere’s disease. According to the 1995 AAO-
HNS CHE Guidelines for the Diagnosis and Evaluation of
Therapy in Meniere’s Disease,18 results of treatment for
TABLE II.
Summary of Pre- and Posttreatment Using ENG Caloric Responses in Studies on Intratympanic Gentamicin Therapy in Unilateral
Meniere’s Disease.
No. of
Method of
Dose per
Treatment
Frequency of
Treatment Sp.
No. of
Treatments
Pretreatment Caloric
Responses
Posttreatment
ENG-Caloric Responses
Study
Patients Protocol Application
6
7
Youssef & Poe
*
37
93
22
T
T
F
II
II
30 mg?
12–20 mg
26.7 mg
Weekly
Daily
1–8
1–4
12
Not recorded
Not recorded
Average ED 48%
Timing-? 23% had decreased
responses- 0–4°/sec
Kaaslinen et al
*
Timing-? Absent IW-39%
Positive IW only - 25%
8
McFeely et al
*
TT with
3 day (4 days)
Timing-? Average ED 92%
catheter or
polyethylene
tube
Absent IW Ϫ75%
9
Minor
*
34
32
68
90
T
T
T
F
II
8–16 mg
5–8 mg
Weekly
1–6
(1 to ?)
1–8
Canal paresis (ED
Ͼ20- 56%
At 3 months: Canal paresis
90%. Absent IW -22%
Positive IW only- 25%
Silverstein et al 10*†
Injected into
middle ear
gelfoam
Every five days
to monthly
Average ED.30–49% At 1 month: Average ED- 59–
(3 protocols)
76%. Absent bithermal 2%
Absent IW -25%
Atlas & Parnes 11* (1999)
II
13–26 mg
19 mg
weekly
Mean bithermal
response-24°/s
At 3 months: Mean bithermal
response-7°/s Positive IW
only- 16% Absent I W -45%
Kaplan et al12
*
TT with
catheter
3 times a day
12
Caloric weakness-
82% IWϩ only 8%
Absent IW- 1%
2 years IWϩ only 16% Absent
IW- 72%
Murofushi et al13
Harner et al14
19
43
T
T
II
II
15–30% X 1–5
30 mg X 1–4
Daily
Mean bithermal
response 15°/s)
Timing- 1–2 months. Mean
bithermal response
Monthly
ED Ͼ20%- 72%
EDϾ 20%- 94%
*Studies reporting according to AAO-HNS CHE guidelines (1985/1995)
†Gentamicin solution was applied to gelfoam placed against the round membrane.
T ϭ titration; F ϭ fixed, II ϭ intratympanic injection; TT ϭ tympanostomy tube; ED ϭ excitability difference; IW ϭ ice water caloric response.
Laryngoscope 112: April 2002
692
Kaplan et al.: Ablation of Vestibular Function

Regarding laboratory testing, it is well recognized
that the ENG caloric test is somewhat limited in its ability
to completely assess vestibular end organ function (i.e.,
theoretically, it only records function of the horizontal
semicircular canal and even this is usually thought to
reflect its lower frequency range). Nevertheless, its chief
value resides in the localization of a diseased side and its
ability to provide a quantitative measure of vestibular
end-organ function for comparison. For this reason, it still
remains the gold standard for the assessment of vestibu-
lar function.
With regard to laboratory testing, our methodology
for testing vestibular end-organ function in the present
study was somewhat inconsistent because it was based on
a comparison of pretreatment bithermal ENG (water) ca-
loric testing and post-treatment air caloric tests (because
a tympanostomy tube was still in place). We acknowledge
that ENG air caloric tests may not be as quantitatively
precise as bithermal ENG water caloric tests; neverthe-
less, we think it is the relative comparison between sides
that still remains the most important aspect for consider-
ation.19 In general terms, the thermal stimulus from an
air caloric test reaching the middle ear directly through a
tympanostomy tube would be expected to result in a
greater or an artifactually increased caloric test response,
not the decreased response that was usually seen after
treatment. With this in mind, it is quite probable that any
attenuation or ablation in air caloric activity on the
treated side was actually even more pronounced taking
into account this phenomenon. Conversely, it is possible
that the presence of a tympanostomy tube may, in certain
circumstances, have biased the caloric test response in
favor of what appeared to be no effect (i.e., false-negative
result) from treatment in a few patients. This may explain
the results of the two patients who had an apparent im-
provement in post-treatment air caloric ED.
In the future, we think improvements in methodology
would certainly strengthen our position that prolonged
use of ototopical gentamicin-containing eardrops are pri-
marily vestibulo-ototoxic. Improvements could include the
use of a closed-loop irrigation system for caloric testing or
insertion of a pliable tympanostomy tube that could be
easily withdrawn immediately after treatment. Then,
rapid healing of the TM would allow for formal ENG
caloric testing to occur a short time later.
Overall, 75% of treated patients in the present study
had evidence of ED change on ENG air caloric testing
which was highly suggestive that a vestibulo-ototoxic
event had occurred. Although we are relatively certain our
intervention caused this, another consideration would be
whether the decrement in caloric activity seen in the ma-
jority of patients after eardrop use was part of the natural
history or a result of fluctuation of vestibular function
seen in Meniere’s disease. There is no absolute way to
refute this, beyond the statistical unlikelihood that a pro-
nounced change in vestibular function would occur in this
many patients within a 1- to 4-month time span, with few
showing random improvements in function. In the ab-
sence of a known comparator (either from randomization
of patients or from historical controls) and with there
being valid concerns regarding our methodology, any con-
ventional statistical analysis cannot be performed. Never-
theless, the finding of even one patient in the present
study with worsening of caloric activity (or cochlear
thresholds) in the treated ear might be significant if one
takes into consideration the previous estimate for topical
ototoxicity by Roland2 of 1 in 10,000. However, in the
absence of a comparator, which could be provided through
randomization, standard statistical tests such as the
Fisher Exact test or the Student t test for paired or un-
paired data cannot be used. Nevertheless, if established
95% nominal confidence intervals are applied to the data,
the statistical likelihood that the effect witnessed occurred
as a result of the intervention ranges from 0.56 to 0.94. In
other words, the effect of this protocol being responsible
for a change in ED on caloric testing ranges from an
incidence slightly greater than the toss of a coin to an
almost assured event.
Previous reports concerning intratympanic gentami-
cin in the treatment of Meniere’s disease have included
data concerning pretreatment and post-treatment caloric
activity using ENG caloric tests. Table II shows the ENG
caloric findings after treatment in studies involving more
than 15 patients. In general, the timing of the post-
treatment ENG, as well as the method, is not uniformly
reported. For the most part, post-treatment ENG caloric
tests appear to have been usually performed within the
first few months of treatment and were reported as a
change in ED or as the fraction of patients with absent
bithermal or ice-water caloric test responses. Overall, re-
sults have tended to show variable changes in ENG caloric
test responses after treatment.
Results from the present study seemingly demon-
strated variability in the effect of intratympanic gentami-
cin clinically and with respect to a change in caloric ED
among the individuals treated. Although most of our pa-
tients developed convincing evidence of ototoxicity, some
did not. Several factors might explain this, including the
following:
1. Possible differences in the position of the tympa-
nostomy tube, the size of the myringotomy inci-
sion, and failure to use tragal pressure to “pump”
the drops down the ear canal through the tube may
have resulted in different amounts of gentamicin
reaching the middle ear and thus accumulating in
the round window niche.
2. Factors affecting middle ear contact time (e.g., pa-
tency of the eustachian tube, vascularity of the
middle ear mucosa, the presence of inflammation).
3. The anatomy of the round window niche and bar-
riers to the inner ear absorption (i.e., mucosal
webs, thickness of round window membrane, and
different diffusion characteristics across the round
window membrane).
4. Variable rates of metabolism and clearance of gen-
tamicin from the inner ear.
5. Variable susceptibility of an individual to amino-
glycosides (i.e., the genetic predisposition of an
individual).
Laryngoscope 112: April 2002
Kaplan et al.: Ablation of Vestibular Function
693

Whether patients with ears affected by Meniere’s
disease might be more sensitive to the effects of topical
gentamicin compared with so-called normal subjects can-
not be answered at present. Although no specific human
data are available to answer this question, extrapolation
from an animal study by Kimura et al.21 demonstrated
that chinchillas with hydropic ears appeared more sensi-
tive to the effects of aminoglycosides than did their control
animals.
Because the topical drops in the present study con-
tained betamethasone in addition to gentamicin, one
might wish to consider what the contribution of the steroid
component was, if any, in the development of ototoxicity?
Limited data are available regarding the effect of intra-
tympanic steroids, which in some animal studies have
shown them to be possibly cochleotoxic.22–24 To date, there
are no reports that have conclusively demonstrated oto-
toxic effects of intratympanic steroids in humans. In one
series, intratympanic steroids appeared to have been ben-
eficial in patients with presumed immune-mediated hear-
ing loss (i.e., sudden sensorineural hearing loss and in
Cogan’s syndrome).25 However, whether this form of
treatment may be beneficial in endolymphatic hydrops
remains controversial.25,26
In previous studies from our institution, patients
with suspected inadvertent topical ototoxicity from com-
mercially available gentamicin-betamethasone appeared
to develop primary vestibulo-ototoxicity around treatment
day 20 on average (range, 7–56 days of treatment), typi-
cally from the start of therapy for treatment of chronic
suppurative otitis media or otorrhea after tube place-
ment.16,27 In the present study, all patients treated with
unilateral Meniere’s disease had dry ears at the begin-
ning. Symptoms of ototoxicity clinically occurred by his-
tory around day 15 on average. This possibly suggests that
inflamed, hypertrophic middle ear mucosa with purulent
debris and so forth may have afforded some protection to
round window initially, thus delaying the onset of topical
ototoxicity in the two groups. Therefore, we think our
study continues to demonstrate on clinical grounds fur-
ther dangers for ototoxicity with the prolonged use of
commercially available topical gentamicin preparations.
To date, we have found this particular protocol of
administering topical gentamicin into the middle ear to be
convenient for both the patient and the physician. Inser-
tion of a tympanostomy tube is a procedure that is mas-
tered by all otolaryngologists, tolerated well by patients as
an office procedure, and proves to be cost-effective. Once
the tympanostomy tube is inserted, the patient is not
dependent on the physician for treatment, which may be a
significant advantage, especially in underserviced re-
gions. As with most drug therapy, it is suitable for a
compliant patient who understands how to apply the drug
and when to stop its use.
In the present study we also observed worsening of
hearing in 10 of the patients according to 1995 AAO-HNS
CHE criteria in the short term (patient 5 also developed a
sensorineural hearing loss mostly in the higher frequen-
cies greater than 4 kHz). When interpreting these results,
several considerations must be taken in account. Assum-
ing that some patients had worsening of hearing from
topical therapy, it should be possible to perform sequential
audiograms in the future, to determine whether the ob-
served cochlear ototoxicity was permanent or reversible.
For example, Kaplan et al.13 demonstrated that 7 of 29
(24%) patients with hearing loss 1 month after treatment
following fixed-schedule intratympanic administration of
19 mg/mL concentrated topical gentamicin recovered
hearing to pretreatment levels over time. Moreover,
changes in hearing might also reflect the expected fluctu-
ation in hearing seen in Meniere’s disease and not a direct
drug effect. This latter point might conceivably explain
the improvement in hearing identified in patients 10 and
15. Nonetheless, 9 (45%) of our patients developed both a
hearing loss and a significant caloric test reduction, im-
plying that there was a simultaneous effect of gentamicin
on both the cochlea and the labyrinth.
When a solution for ablation is delivered directly into
the middle ear by either a transtympanic injection or a
catheter route, one can be certain that it reaches the
middle ear. Using our protocol, we initially wondered
whether the drops would actually pass through the tym-
panostomy tube and reach the middle ear. Our findings
most assuredly demonstrate that this method (not dissim-
ilar to instructions for topical use that would be normally
provided to patients for treatment of external and middle
ear infectious or inflammatory diseases) is quite effective
for the delivery of drops into the middle ear, a cautionary
point that should not be taken for granted when topical
gentamicin- or other aminoglycoside-containing drops,
with their recognized potential for ototoxicity, are pre-
scribed for treatment of middle ear sepsis in the presence
of a TM defect.
CONCLUSION
The present study clearly demonstrates that in the
presence of a tympanostomy tube, commercially available
gentamicin-containing eardrops penetrate the middle ear
cavity, where they can be absorbed into the inner ear,
resulting primarily in a vestibulo-ototoxic and, in some
instances, a cochleo-ototoxic effect, or both. We conclude
this particular protocol to be a simple, convenient, and
effective method for applying gentamicin intratympani-
cally for patients with Meniere’s disease. Success in the
treatment of Meniere’s disease according to the 1995
AAO-HNS CHE recommendations (i.e., vertigo control,
preservation of hearing, and so forth) cannot be evaluated
in the present study and must be assessed after a mini-
mum follow-up of 2 years.
Although the gentamicin concentration in the oto-
topical preparation that was used was only 3 mg/mL (ap-
proximately 10 times less than in some of the solutions
used in previous studies), the ototoxic effects were appar-
ent both clinically and quantitatively, as measured by
pretreatment and post-treatment ENG caloric tests. The
comparatively long duration of continued exposure pre-
sumably allowed for the concentration of gentamicin to
occur within the inner ear, which resulted primarily in
vestibulo-ototoxic effect with some patients having co-
chlear ototoxicity despite the small volume received each
time the drops were applied (approximately 0.10–0.15
mL).
Laryngoscope 112: April 2002
694
Kaplan et al.: Ablation of Vestibular Function

in the treatment of Meniere’s disease. Otolaryngol Head
Neck Surg 1999;120:649–655.
12. Atlas JT, Parnes LS. Intratympanic gentamicin titration
therapy for intractable Meniere’s disease. Am J Otol 1999;
20:35–393.
13. Kaplan DM, Chen JM, Nedzelski JM, Shipp DB. Intratym-
panic gentamicin for the treatment of unilateral Meniere’s
disease: a ten-year experience using a predetermined
treatment regimen. Laryngoscope 2000;110:1298–1305.
14. Murofushi T, Halmagyi GM, Yavor RA. Intratympanic gen-
tamicin in Meniere’s disease: result of therapy. Am J Otol
1997;18:52–57.
15. Harner SG, Kasperbauer JL, Facer GW, Beatty CW. Tran-
stympanic gentamicin for Meniere’s syndrome. Laryngo-
scope 1998;108:1446–1449.
16. Blakley BW. Update on intratympanic gentamicin for
Meniere’s disease. Laryngoscope 2000;100:236–240.
17. Bath AP, Walsh RM, Bance ML, Rutka JA. Ototoxicity of
topical gentamicin preparations. Laryngoscope 1999;109:
1088–1093.
18. Committee on Hearing and Equilibrium. Committee on Hear-
ing and Equilibrium guidelines for the diagnosis and eval-
uation of therapy in Meniere’s disease. Otolaryngol Head
Neck Surg 1995;113:181–183.
19. Suter CM, Blanchard CL, Cook-Manokey BE. Nystagmus
response to water and air caloric stimulation in clinical
populations. Laryngoscope 1977;87:1074–1078.
20. Asaswavichianginda S, Fujimoto M, Mai M, Rutka JA. Prev-
alence of head-shaking nystagmus in patients according to
the diagnostic classification in a dizziness unit. J Otolary-
yngol 1997;26:20–25.
21. Kimura RS, Lee KS, Nye CL, Trehey JA. Effects of systemic
and lateral semicircular canal administration of aminogly-
cosides on normal and hydropic inner ears. Acta Otolaryn-
gol 1991;111:1021–1030.
In addition, the present study offers further support
regarding the primarily vestibulo-ototoxic nature of com-
mercially available eardrops containing gentamicin when
used in a prolonged fashion. Physicians are again cau-
tioned to frequently assess the need for preparations with
topical gentamicin and, for that matter, all amino-
glycoside-containing drops in the presence of a TM perfo-
ration or defect and to stop the drops immediately when
discharge from the ear ceases or the middle ear becomes
dry, to prevent their further absorption through the round
window membrane into the inner ear.
Acknowledgments
The authors thank the technical staff of the Univer-
sity Health Network Center for Advanced Hearing and
Balance Testing for their role in the laboratory assess-
ment of patients in the present study.
BIBLIOGRAPHY
1. Pickett BP, Shinn JB, Smith MFW. Ear drop ototoxicity:
reality or myth? Am J Otol 1997;18:782–791.
2. Roland PS. Clinical ototoxicity of topical antibiotic drops.
Otolaryngol Head Neck Surg 1994;110:598–602.
3. Lundy LB, Graham MD. Ototoxicity and ototopical medica-
tions: a survey of otolaryngologists. Am J Otol 1993;14:
141–146.
4. Beck C, Schmidt CL. Ten year experience with intratympani-
cally applied streptomicin (gentamicin) in the therapy of
morbus Meniere. Arch Otorhinolaryngol 1978;221:
149–152.
5. Schmidt CL, Beck C. Treatment of morbus Meniere’s with
intratympanically applied gentamycin. Laryngorhinootolo-
gie 1980;59:804–807.
6. Odkvist LM. Middle ear ototoxic treatment for inner ear
disease. Acta Otolaryngol Suppl (Stockh) 1988;457:83–86.
7. Youssef, Poe DS. Intratympanic gentamicin injection for the
treatment of Meniere’s disease. Am J Otol 1998;19:
435–442.
8. Kaasinen S, Pryykko I, Ishizaki H, et al. Intratympanic gen-
tamicin in Meniere’s disease. Acta Otolaryngol (Stockh)
1998;118:294–298.
22. Ikeda K, Morizono T. Effect of ototopic application of a corti-
costeroid preparation on cochlear function. Am J Otol
1991;12:150–153.
23. Spandow O. Are glucocorticoids in ear drops ototoxic? Acta
Otolaryngol Suppl (Stockh) 1992;493:89–91.
24. Spandow O, Anniko M, Hellstrom S. Hydrocortisone applied
into the round window niche causes electrophysiological
dysfunction of the inner ear ORL J Otorhinolaryngol Relat
Spec 1989;51:94–102.
25. Parnes LS, Sun AH, Freeman DJ. Corticosteroid pharmaco-
kinetics in the inner ear fluids: an animal study followed
by clinical application. Laryngoscope 1999;109:1–17.
26. Shea JJ Jr. The role of dexamethasone or streptomycin per-
fusion in the treatment of Meniere’s disease. Otolaryngol
Clin North Am 1997;30:1051–1059.
9. McFeely WJ, Singleton GT, Rodriguezet FJ, Antonelli PJ. In-
tratympanic gentamicin treatment for Meniere’s disease.
Otolaryngol Head Neck Surg 1998;118:589–596.
10. Minor LM. Intratympanic gentamicin for control of vertigo in
Meniere’s disease: vestibular signs that specify completion
of therapy. Am J Otol 1999;20:209–219.
27. Marais J, Rutka JA. Ototoxicity and topical eardrops. Clin
Otolaryngol 1998;23:360–367.
11. Silverstein H, Arruda J, Rosenberg SI, Deems D, Hester TO.
Direct round window membrane application of gentamicin
Laryngoscope 112: April 2002
Kaplan et al.: Ablation of Vestibular Function
695