The LaryngoscopeLippincott Williams & Wilkins, Inc.© 2006 The American Laryngological,Rhinological and Otological Society, Inc.

Intratympanic Dexamethasone for SuddenSensorineural Hearing Loss After Failure ofSystemic Therapy

David S. Haynes, MD; Matthew O’Malley, MD; Seth Cohen, MD; Kenneth Watford, NP;Robert F. Labadie, MD, PhD

Objective: Intratympanic steroids are increas-ingly used in the treatment of inner ear disorders,especially in patients with sudden sensorineuralhearing loss (SNHL) who have failed systemic ther-apy. We reviewed our experience with intratympanicsteroids in the treatment of patients with suddenSNHL to determine overall success, morbidity, andprognostic factors. Hypothesis: Intratympanic ste-roids have minimal morbidity and the potential tohave a positive effect on hearing recovery in patientswith sudden SNHL who have failed systemic therapy.Study Design: The authors conducted a retrospectivereview. Methods: Patients presenting with suddenSNHL defined as a rapid decline in hearing over 3days or less affecting 3 or more frequencies by 30 dBor greater who underwent intratympanic steroidstherapy (24 mg/mL dexamethasone) were reviewed.Excluded were patients with Meniere disease, retro-cochlear disease, autoimmune HL, trauma, fluctuat-ing HL, radiation-induced HL, noise-induced HL, orany other identifiable etiology for sudden HL. Pa-tients who showed signs of fluctuation of hearingafter injection were excluded. Pretreatment and post-treatment audiometric evaluations including pure-tone average (PTA) and speech reception threshold(SRT) were analyzed. Patient variables as they re-lated to recovery were studied and included patient

age, time to onset of therapy, status of the contralat-eral ear, presence of diabetes, severity of HL, andpresence of associated symptoms (tinnitus, vertigo). A20-dB gain in PTA or a 20% improvement in SDS wasconsidered significant. Results: Forty patients fit thecriteria for inclusion in the study. The mean age of thepatients was 54.8 years with a range from 17 to 84years of age. Overall, 40% (n � 16) showed any im-provement in PTA or SDS. Fourteen (35%) men and 26(65%) women were included. Using the criteria of20-dB improvement in PTA or 20% improvement inSDS for success, 27.5% (n � 11) showed improvement.The mean number of days from onset of symptoms tointratympanic therapy was 40 days with a range of 7days to 310 days. A statistically significant differencewas noted in those patients who received earlier in-jection (P � .0008, rank sum test). No patient receivingintratympanic dexamethasone after 36 days recov-ered hearing using 20-dB PTA decrease or a 20% in-crease in discrimination as criteria for recovery.Twelve percent (n � 5) of patients in the study haddiabetes with 20% recovering after intratympanicdexamethasone (not significantly different from non-diabetics at 28.6%, Fisher exact test, P � 1.0). Compar-ison to other studies that used differing steroid type,concentration, dosing schedule, inclusion criteria,and criteria for success revealed, in many instances, asimilar overall recovery rate. Conclusions: Difficultyin proving efficacy of a single modality is present inall studies on SNHL secondary to multiple treatmentprotocols, variable rates of recovery, and a high rateof spontaneous recovery. Forty percent of patientsshowed some improvement in SDS or PTA after treat-ment failure. When criteria of 20-dB PTA or 20% isconsidered to define improvement, the recovery ratewas 27.5%. Modest improvement is seen with the cur-rent protocol of a single intratympanic steroid injec-

From Vanderbilt University Medical Center/The Otology Group ofVanderbilt, Nashville, Tennessee, U.S.A.

Editor’s Note: This Manuscript was accepted for publication August30, 2006.

Send correspondence to David S. Haynes, MD, Associate Professor,Department of Otolaryngology/The Otology Group of Vanderbilt, 1215 21stAvenue South, 7209 MCE South Tower, Nashville, TN 37232, U.S.A.E-mail: david.haynes@vanderbilt.edu

DOI: 10.1097/01.mlg.0000245058.11866.15

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

3

tion of 24 mg/mL dexamethasone in patients whofailed systemic therapy. Dramatic hearing recoveryin treatment failures was rarely encountered. No pa-tient showed significant benefit from intratympanicsteroids after 36 days when using this protocol foridiopathic sudden SNHL. If patients injected after 6weeks are excluded from the study, the improvementrate increases from 26.9% to 39.3%. Earlier intratym-panic injection had a significant impact on hearingrecovery, although with any therapeutic interventionfor sudden SNHL, early success may be attributed tonatural history. If we further exclude seven patientstreated with intratympanic steroids within 2 weeks ofthe onset of symptoms (i.e., study only those patientstreated with intratympanic dexamethasone between2 and 6 weeks after onset of symptoms), still, 26%improved by 20 dB or 20% SDS. The recovery ratesafter initial systemic failure are higher than would beexpected in this treatment failure group given ourcontrol group (9.1%) and literature review. Thesefindings indicate a positive effect from steroid perfu-sion in this patient population. Key Words: Suddensensorineural HL, intratympanic therapy, dexameth-asone, steroid perfusion.

Laryngoscope, 117:3–15, 2007

INTRODUCTIONIdiopathic sudden sensorineural hearing loss (SNHL)

is defined as a decline in hearing over 3 days or lessaffecting 3 or more frequencies by 30 dB or greater with noidentifiable etiology.1 Sudden SNHL affects between 5and 20 persons per 100,000 year or approximately 4,000new cases annually in the United States.2 The HL isnearly always unilateral and is commonly associated withtinnitus and aural fullness. The true incidence of suddenSNHL is probably underestimated because many who re-cover hearing early (within the first few days) are unlikelyto seek medical therapy.

Although this disorder is not one of the more commonetiologies for HL, disproportionate interest in suddenSNHL exists most likely because it is one of the fewreversible (sensorineural) hearing losses encountered byclinicians.3 Another potential reason for high interestlevel is that sudden SNHL is encountered by all otolaryn-gologists in all areas of the country and treated as a trueemergency often without the timeframe to allow for ter-tiary referral.

The etiology, natural history, and treatment of thisdisorder have been subjects of debate for many years. Theactual number of patients recovering spontaneously fromsudden SNHL without having sought medical attention isnot known. The high rate of spontaneous recovery, up to65%,4 also confounds reviews as to the therapeutic efficacyof any single agent or therapeutic intervention. Any pro-posed therapeutic intervention must improve on the 65%recovery rate that would be seen if no intervention wasoffered. The treatment of patients with sudden SNHLremains varied among otologic centers with no standardprotocol universally accepted. With no specific etiology,and a short timeline for effective therapy realized, a tech-nique termed “shotgun” therapy is often used. This ther-apy entails multiple therapeutic agents geared toward thehypothetical etiologies given at once, because the narrow

therapeutic window prevents trials with each agent sin-gly. This commonly used technique prevents the determi-nation of which, if any, of the agents were effective inrestoring hearing. Notwithstanding the timeframe inwhich maximum recovery may occur, from several days5,6

to possibly several months,7,8 also leads to errors in deter-mining treatment efficacy versus natural history. Despitehigh reported spontaneous recovery rates, it is our expe-rience, and that of others, that hearing recovery is poor inthose patients who have failed systemic therapy.5,9,10

Multiple treatment protocols and agents have beenproposed to treat SNHL. Steroids, antiviral agents, anti-coagulants, vasodilators, antiinflammatory agents, andothers have been proposed as therapeutic agents to treatsudden SNHL, most of which propose some benefit in thetreatment of sudden SNHL. The most accepted currenttreatment of sudden SNHL is systemic steroids. Althoughproven to be effective in randomized, double-blind,placebo-controlled trials,1,11 other studies have questionedthe efficacy of systemic steroids in the treatment of sud-den SNHL.2,4,8

Both short-term and long-term complications fromsystemic steroids are well known to otolaryngologists,leading to the continued investigation into directed ther-apy for inner ear disorders, including sudden SNHL. In-tratympanic steroids offer the potential for directed ther-apy of a high concentration to the inner ear withavoidance of systemic side effects. Like most proposedtherapies, the efficacy of intratympanic steroid therapy inthe treatment of sudden HL has yet to be determined,although several reports have demonstrated efficacy espe-cially after treatment failures.9,12–21,25–27

Intratympanic TherapyItoh was the first to report on the use of intratym-

panic steroids for inner ear disease when he treated pa-tients for Meniere disease in 1991.22 The first report onthe use of intratympanic steroids for sudden SNHL was bySilverstein in 1996.12 Other authors have also describedthe use of intratympanic steroids in the treatment of sud-den SNHL.9,12–21,23–27 Although the efficacy has not beendefinitively proven, intratympanic steroids as a therapeu-tic option for sudden HL is increasingly used in the UnitedStates. The variability that exists in treatment protocolsfor sudden SNHL also applies to protocols that involveintratympanic steroids. The use of intratympanic steroidshas evolved into 3 main protocols for treatment of suddenSNHL:

● As an initial or primary treatment for suddenSNHL without systemic steroids;

● As adjunctive treatment given concomitantly withsystemic steroids for sudden SNHL; and

● As “salvage therapy” after failure of systemic ste-roids for sudden SNHL.

The primary reason for the use of intratympanic ste-roids without systemic steroids is in patients who cannottolerate systemic steroids or those at greater risk for com-plications from systemic steroids (e.g., diabetics).13,15,16

The majority of the literature concerning the use of intra-

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

4

tympanic steroids in the treatment of sudden SNHL hasreported the experience in treatment after failure of sys-temic therapy.9,12–21,25–27 Two studies, however, havestudied the effects of intratympanic steroids as a primaryor first-line agent for patients with sudden SNHL23,24

used adjunctively with systemic steroids.There are several advantages of intratympanic ste-

roids as a treatment for sudden SNHL (Table I). Theprocedure is well tolerated and relatively easy to perform.As an office-based procedure done under local (topical)anesthesia, there is an avoidance of general anesthesia.Most patients understand the concept of intratympanicinjection and readily accept the proposed therapy. Unlikesystemic therapies, intratympanic therapy allows for theselection of the affected ear to be treated. In addition toglucose intolerance and avascular necrosis of the hip,other less severe side effects of systemic steroids such asinsomnia, irritability, gastritis, and mood changes maypotentially be avoided with topical therapy. The primarydisadvantage of intratympanic steroids is the lack ofproven efficacy and/or superiority over systemic steroids.Other potential disadvantages include pain, tympanicmembrane perforation, acute otitis media, otorrhea, ver-tigo, and the potential for further HL.

Described techniques for steroid perfusion of the mid-dle ear for sudden SNHL differ in many aspects, includingthe type of steroid used. Dexamethasone is the most com-mon steroid used for intratympanic use14,15,18–21,24,25 fol-lowed by methylprednisolone.9,13,15,16,17,23,26,27 Reports inthe literature also differ in the strength of the solution(2–4 mg/mL14,20 to 25 mg/mL dexamethasone15; 32 mg/mL23 to 62.5 mg/mL methylprednisolone).9,16,17 Theamount injected into the middle ear in most studies is be-tween 0.3 and 0.5 mL, the approximate volume of the middle

ear space. Techniques also differ in method of delivery: tran-stympanic needle injection,9,13,15,19,20,17,24–27 deliverythrough a myringotomy,13,14 delivery through a myringot-omy with a tube,12,23 delivery with a wick placed in a myr-ingotomy (Micromedics, Eagan, MN),18,21 and deliverythrough an implantable pump (Round Window m-Cath; Du-rect Corp., Cupertino, CA) to deliver the steroid as a constantinfusion.16,17,21

The length of time and number of injections in whichpatients are treated with intratympanic steroids also dif-fers ranging from a single day to weekly transtympanicinjections9,12,14,15,18,20,23 to multiple weeks with self-administered steroid drops18,21 to transtympanic injec-tions given several times per week19,23,25,26 or to an im-plantable pump.16,17,21 Reported complications are rareand include pain,13 vertigo,13,16,17,20,21 otitis media,13 tym-panic membrane perforation,9,21 acne,20 dysgeusia,21

chronic otitis media,21 and further HL.16,21

This study was undertaken to review the experiencewith 24 mg/mL intratympanic dexamethasone given at asingle time point in the treatment of patients with idio-pathic sudden SNHL that have failed systemic steroidtherapy. Special attention was given to evaluating thesafety of the procedure and correlation with improvementin hearing related to age, time to onset of therapy, priortherapy, diabetes, severity of loss, and status of the oppo-site ear.

MATERIALS AND METHODS

Inclusion CriteriaA retrospective review of the patients undergoing intratym-

panic steroid injection from January 1, 2000, to July 30, 2005,were reviewed, yielding 312 procedures in 195 patients (Table II).These records were further reviewed to determine which patientsunderwent intratympanic steroid therapy for sudden SNHL. Ex-

TABLE I.Transtympanic Dexamethasone Perfusion for Sudden

Sensorineural Hearing Loss.

Advantages

Office-based procedure

Easily administered

Rapid administration after diagnosis

Relatively painless

Use in patients in which systemic steroids may becontraindicated

(example: immunocompromised patients, HIV, tuberculosis,diabetes)

Use in patients in which use of systemic steroids are declined

Ability to direct therapy to the affected ear

A high concentration of medication can be delivered directly tothe (affected) ear

Side effects/complications are uncommon

Disadvantages/complications

Tympanic membrane perforation

Pain

Otitis media

Vertigo (usually temporary)

Hearing loss

TABLE II.Inclusion and Exclusion Criteria.

● Sudden, unilateral sensorineural hearing loss of at least 30 dBover three frequencies developing within 72 hours

● An audiogram was performed pretreatment and at least oneposttherapy audiogram was performed

● Underwent intratympanic injection with 24 mg/mLdexamethasone at a single time period

● No evidence of retrocochlear disease evident on magneticresonance imaging

● No history of otologic surgery

● No history of Meniere disease, autoimmune hearing loss,radiation-induced hearing loss, or other potential etiology forsensorineural hearing loss

● No history of acoustic trauma or barotrauma

● No history of genetic sensorineural hearing loss or known innerear anomaly

● No history of fluctuation of hearing before or after intratympanictherapy

● Failed systemic steroid trial or did not receive steroid trial (i.e.,patient refused, diabetes)

● No evidence of acute otitis media or chronic otitis media onexamination

● Failed systemic steroids

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

5

cluded from this group were those patients who did not meet theinclusion criteria outlined in Table I, patients with incorrectcoding, incomplete records, inadequate follow up, or inadequateaudiometric analysis. Patients undergoing multiple injections orpatients receiving dexamethasone at any dose other than 24mg/mL were excluded. All patients with fluctuating HL or Me-niere disease were excluded. Eighty-five patients underwent in-tratympanic steroids for sudden SNHL for diagnoses other thanMeniere disease. Excluded from this group were patients withfluctuating HL (8) head trauma (3), autoimmune HL (3),radiation-induced HL (1), noise trauma (1) barotrauma (1), laby-rinthitis (1), congenital HL (1), delayed perilymph fistula afterstapedectomy (1), and herpes zoster oticus (1). These exclusionsleft 64 patients with idiopathic sudden SNHL who underwentintratympanic dexamethasone perfusion and available for study.Six patients had had inadequate audiometric data, 4 patientsfailed to follow up, 4 patients had intratympanic steroids startedconcomitantly with oral steroids, and 2 had HL less than 30 dBand were excluded. Seven patients did not receive systemic ste-roids (diabetes, n � 4; or refused, n � 3) and were excluded. Onepatient failed a 5-day course of steroids and was injected on thefifth day and was excluded. Forty patients were available to studywith idiopathic sudden SNHL who failed systemic therapy andunderwent intratympanic dexamethasone as salvage therapy af-ter at least 7 days of therapy.

We also present a control group of patients who failed sys-temic therapy, did not receive intratympanic therapy, and had atleast one follow-up audiogram after determining failure of sys-temic steroids. Excluded from this group were patients who hadMeniere disease, autoimmune HL, or any other identifiable hear-ing cause for HL. Any patient with fluctuating HL before steroidtherapy was excluded.

Audiometric DataPatients were all evaluated using standardized methods for

pure-tone threshold audiometry and speech intelligibility by cer-tified audiologists pre- and postinjection (Grason-Stadler GSImodel 16 or 61). Pure-tone average (PTA) was calculated as anaverage of the threshold measured at 0.5, 1.0, and 2.0 KHz.Speech intelligibility (SDS) was tested by calculating the percentcorrect of a phonetically balanced, monosyllabic word list (North-western University, NU-6).

TechniqueThe correct ear is confirmed for injection by patient re-

sponse and audiometric review. EMLA cream (AstraZeneca, Wil-mington DE) is used for anesthesia by topical application. TheEMLA cream is left on the tympanic membrane for 30 to 45minutes. The cream is removed and the head is placed in position45° toward the unaffected ear. The dexamethasone solution of 24mg/mL (Table III) is checked and warmed to body temperaturebefore injection. Before each procedure, the patient is counseledregarding the risks and expectations of the procedure and in-formed consent obtained. Approximately 0.3 to 0.5 mL of thesolution is injected into the middle ear. No myringotomy,pressure-equalizing tube, or secondary myringotomy is made. Nomiddle ear endoscopy is performed. On completion of the injec-tion, the head is turned toward the affected side and then backaway to the original position. This maneuver is performed in anattempt to maximize exposure of the solution in the middle earspace to the round window membrane. A second injection may beimmediately performed if the first was felt to be inadequate or ifinspection of the middle ear shows a predominantly air-filledmiddle ear space. Up to 3 injections may be given during this timeperiod. The patient is asked to lie in the supine position with the

head turned 45° away from the treated ear for 10 to 15 minutes onaverage. No subsequent injections are given on follow-up visits.

Statistical AnalysisData are presented in numeric and percent form. Categorical

data analysis was performed using �2 techniques or the Fisher exacttest. Comparison between days before treatment and improvementwas performed using the rank sum test. All statistical analysis usedSigmaStat software 2.03 (SPSS Inc., Chicago, IL).

Reporting RecoveryThe criteria used to define a successful recovery after ther-

apy differs in the literature pertaining intratympanic steroids. A20-dB improvement in PTA or a 20% improvement in discrimi-nation was considered a successful therapeutic intervention. Thedata from this study were also applied to the criteria for successas defined in other studies that investigated the use of intratym-panic steroids for sudden SNHL (see Table IV).

TABLE III.Dexamethasone Solution for Otic Injection.

Ingredients Dexamethasone sodium phosphate, powder,USP 120 mg

Starting materials Sterile empty 10 mL vials, dry only, forexample, Abbott brand

Sterilized stainless steel spatula, vial stopperdecapper and crimper, electronic scaleswith printer, serum bottle aluminum seals20 mm

Compounding 1. Use scale in chemo preparation area

2. Don appropriate compounding attire,including gown, mask, hair cover, andgloves; no other personnel should be inwork area when weighing powder

3. Remove stoppers from 10-mL vials

4. Place vial on scale and tare weight

5. With stainless steal spatula; add 120 mgof powder to each vial

6. Replace stoppers and add new aluminumseals; crimp tightly and check seal

7. Label appropriately

Dispensingdirections

1. Dilute each vial with 5 mL of preservative-free sterile saline for a final concentration24 mg/mL

2. Filter before dispensing with 0.22-� mfilter such as Millex GS filter

3. Expiration date of diluted product 24hours

References Requested for use by physicians

Expiration date Vials of powder 6-month expiration date asa result of USP standards

Assign 24-hour expiration date on dilutedproduct

Storage Refrigerate; powder storage directions frommanufacturer

Auxiliary labels For external/otic irrigation use only

NOT FOR INJECTION

Sample labels Dexamethasone sodium phosphate, USP

For otic irrigation use only

Lot # 102700M1 date

Must be filtered with 0.22-� m filter beforeuse

NOT FOR INJECTION

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

6

RESULTS

Patient PopulationAfter inclusion and exclusion criteria were applied,

40 patients were available for study. There were 14 (35%)men and 26 (65%) women. The mean age was 54.8 with arange from 17 to 84 years of age. The mean age for thewomen was 58 years and for the men 48 years. The overallrecovery rate for men was 35.7% and for women was 23%(P � .5, Fisher exact test).

Overall Recovery. Forty patients fit the criteria forinclusion in the study. Overall, 40% (n � 16) showedimprovement in PTA or SDS. For those 37.5% (n � 15)showing an improvement in PTA, the mean gain was 15dB. For the 37.5% (n � 15) showing an improvement inSDS, the mean gain was 31.9% (range, 8% to 88%).Using the criteria of 20-dB improvement in PTA or 20%

improvement in SDS for success, a 27.5% (n � 11) improve-ment was noted (Fig. 1). For these 27.5% who had an im-provement, an average improvement in PTA of 16.9 dB(range, 0–42 dB) and average improvement in SDS of 38.9%(range, 8% to 88%) was noted (Fig. 2). Seven patients (17.5%)showed worse PTA with a mean decrease of 3.8 dB (range,2–7 dB). Five patients (12.5%) showed worse SDS after in-jection with a mean decrease of 16% (range, 8% to 28%).Thirty-five percent of patients had no change in hearingafter intratympanic steroids.

Age Related to Recovery. Recovery as related topatient age was studied. Sixty-three percent of patientswere under 60 years of age and had an overall recoveryrate of 24%. Thirty-seven percent of patients were 60years of age or older and had an overall recovery rate of33.3% (P � .7, Fisher exact test).

TABLE IV.Comparison of Recovery Rates in Sudden Sensorineural Hearing Loss Treated With Intratympanic Steroids.

Author Percent Improved Criteria for Improvement Current Study With Applied Criteria

Silverstein et al.,12 1996 25% 10-dB PTA 32.5%15% SDS

Parnes,13 1999§ 46% 5 normal thresholds 40%One serviceable hearing (57% if including only patients

treated within 6 wks)Kopke et al.,17 2001 (�6 wks)† 0% 10-dB PTA 0%

15% SDSKopke et al.,17 2001 (�6 wks) 83% 10-dB PTA 35% overall

15% SDS (50% if including only patientstreated within 6 wks)

Chandrasekhar,14 2001 73% Increase in SDS, decrease inPTA

40%

Gianoli and Li,15 2001 44% 10-dB PTA 40%10% SDS

Lefebre and Staeker,16 2002 100% �16 dB improvement in PTA 12.5% (overall)50% if treated within 10 days

Gouveris,19 2003 Complete recovery (CR):33.3%

CR: within 10 dB of unaffectedear

CR 2.5%

Partial: 39% NR: less than 10-dBimprovement

Partial 15%

No recovery: (NR) 28.6% NR 82.5%Jackson,18 2002 31% Positive response 40%Ho et al.,20 2004 53% 30-dB PTA 7.5%

(10.3% if only those treated within50 days)

Herr and Marzo,21 2005 53% 10 dB PTA 32.5%20% SDS (33.3% if only those treated within

20 wks)Battista,24 2005‡ 8% full Full within 10 dB baseline 2.5% full

12% partial Partial within 50 dB baseline 10% partial overallSlattery et al.,9 2005 55% 10-dB PTA 40% (42.1% if only those treated

within 3 months)12% SDS

Choung et al.,25 2006 38.2% 10 dB PTA15% SDS

35%(10% if treated after 28 days)

Dallan et al.,26 2006 75% 15 dB PTA 12.5% (5% if treated after 21 days)

Xenellis27 2006 47% 10 dB PTA 15% (5% if treated after 2 days)

†Study divided into two groups based on time to presentation.‡Study involved only profound hearing loss; our data were adjusted accordingly.§Although response rates are similar, those that did respond in Parnes et al. had a significantly greater response than the current study.PTA � pure-tone average; SDS � speech discrimination score; Dex � dexamethasone; MP � methylprednisolone.

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

7

Prior Treatment. All patients entered had receivedsystemic therapy before intratympanic injection. In addi-tion to steroids, 85.4% received antiviral agents with a28.9% recovery rate in this subset. Diuretics were used in27.1% with a 23.1% recovery rate noted in this subset.

Recovery Related to Associated Symptoms. Ver-tigo was present in 37.5% of patients with a recovery rateof 20%. A total of 62.5% of patients did not have symptomsof vertigo and had an overall recovery of 32% (P � .5,Fisher exact test). Tinnitus was present in 65% of patientswith a 23.1% recovery in this group. Tinnitus was absentin 35% of patients with 36% showing recovery in thisgroup (P � .5, Fisher exact test).

Status of the Opposite Ear. A total of 87.5% ofpatients had normal hearing in the contralateral ear. Therecovery rate in this group was 26.5%. Only 12.5% ofpatients had abnormal hearing in the contralateral ear.The recovery rate in this group with abnormal hearing inthe contralateral ear was 33.3% (P � 1.0, Fisher exacttest).

Recovery Related to Time of Onset of Symptoms.The average number of days from onset of symptoms tointratympanic therapy was 40 days overall with a range of7 days to 310 days. For the group (n � 11) that respondedto injection, the median was 14 days. For the group thatdid not respond, the mean was 31 days (P � .008, rank

sum test). No patient receiving intratympanic dexameth-asone after 36 days recovered hearing using 20-dB PTA/20% discrimination as criteria for recovery. If patientsreceiving injections for sudden SNHL after 6 weeks areexcluded, the recovery rate increases to 39.3% (Table V).

Recovery Related to Severity of Loss. Recovery ofhearing as related to severity of initial loss was studied.Twelve patients (30%) had HL greater than 90 dB with an8.3% improvement rate noted in this group. Twenty-twopatients (55%) had HL of 90 dB or less and greater than orequal to 50 dB with a 20% recovery rate noted in thisgroup. Six patients (15%) had HL less than 50 dB andgreater than 30 dB with an 33% improvement rate in thisgroup (P � .2 �2) (Fig. 3). Patients with severe lossesgreater than 90 dB had a poorer recovery (8.3%) comparedwith losses that were less than or equal to 90 dB (35.7%recovery) (P � .1, Fisher exact test).

Recovery Related to Diabetes. A total of 12.5% ofpatients in the study had diabetes. Twenty percent ofthose had recovery after intratympanic dexamethasone.The recovery rate in patients without diabetes was 28.6%(Fisher exact test, P � 1.0). Four of the nine patients withdiabetes did not receive systemic steroids and were notincluded in the primary study. The recovery rate was 25%(1 of 4) in this group.

Control Group. A matched group of patients whofailed systemic steroid therapy were identified within thesame time period as the study group. This group included11 patients identified who had sudden SNHL, failed sys-temic steroid therapy, and were available for long-termfollow up. Patients required pretreatment, posttreatment,and at least one more posttreatment audiogram after doc-umentation of systemic steroid failure. There were 7males and 4 females. The average age was 67 years witha range from 61 to 78 years. Average follow up from onsetof symptoms was 271 days with a range from 22 days to1,460 days. Recovery of hearing of 20-dB PTA or 20% SDSwas seen in 9.1% of patients. Eighteen percent had adecline in hearing thresholds by 20 dB or 20% discrimi-nation over time. The one patient who improved had noimprovement in hearing with steroids at 2 weeks afteronset of symptoms and treatment with steroids but had a20% gain in SDS at 6 weeks follow up.

DISCUSSIONIn a double-blind, placebo-controlled study, Wilson et

al. showed a statistically significant benefit with systemic

Fig. 2. Recovery in SDS as a function of time from onset of symp-toms to injection (n � 11).

Fig. 1. Decibel gain (PTA) as a function of time from onset ofsymptoms (n � 11).

TABLE V.Recovery Related to Time to Therapy From Onset of Symptoms.

Days to InjectionNo. of

Patients20% SDS/20-dB PTA Partial None Worse

10 d or less 4 2 (18%) 0 2 0

11–20 d 10 5 (45%) 1 1 3

21–30 d 7 1 (9%) 2 3 1

�30 d 19 3 (27%) 3 9 4

Totals 40 11 6 15 8

SDS � speech discrimination score; PTA � pure-tone average.

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

8

steroids in recovery of hearing in patients with suddenSNHL.1 Other studies have also demonstrated the benefitof systemic steroids in hearing recovery in suddenSNHL.3,5,6,7,11 On the contrary, systemic steroids wereshown to be of little benefit in the treatment of suddenSNHL in several other studies.2,4,8

Effects of Steroids on Cochlear FunctionDexamethasone (9-fluro-11b,17,21-trihydroxy-16a-

methylpregna-1,4-diene-3,20-dione) is a synthetic cortico-steroid used commonly in clinical medicine through oral,parenteral, and topical routes primarily for its antiinflam-matory effect. The exact mechanism in which steroids mayimprove hearing is unknown. The effects of steroids aremediated through receptors found within the cytoplasm.Both glucocorticoid and mineralocorticoid receptors arefound in the inner ear.28 This study and others cited latersuggest that steroids play a significant role in modulatingcochlear function. Multiple studies have shown systemicsteroids to have a positive effect on cochlear function.Other studies have shown steroids to decrease inflamma-tion from labyrinthitis,29 improve cochlear blood flow,30

protect against cochlear ischemia,31 protect against noise-induced HL,32 and regulate inner ear de novo proteinsynthesis.33 Studies have shown the stria vascularis,which maintains Na�/K� secretion necessary for mainte-nance of the endocochlear potential, to be a site for poten-tial pathology in sudden HL.34 Systemic steroids have alsobeen shown to improve stria vascularis function and mor-phology35 and therefore the potential to recover hearingafter sudden SNHL.

Multiple studies have shown that intratympanic ste-roids are safe without evidence of histologic changes orcochlear dysfunction.12,36–39 Intratympanic steroids havebeen shown to increase cochlear blood flow,12,36 preventaminoglycoside toxicity,40 prevent drill-induced noiseloss,37 and improve ion homeostasis necessary for cochlearfunction.39 Intratympanic steroids were also shown tohave a protective effect on stria vascularis changes afterotitis media.41 In a study of patients with tinnitus, intra-tympanic dexamethasone was found to have no adverseaffect on cochlear function as measured by otoacousticemission.38 Other studies suggest that intratympanic ste-roids may not be beneficial in the treatment of HL. Thepotential for intratympanic steroids to cause decreasedcochlear function has been suggested.42 Intratympanicsteroids have been shown to lead to round window inflam-

mation.43 Yang and colleagues noted that intratympanicsteroids were ineffective at preventing immune-mediatedlabyrinthitis after induction of keyhole limpit hemocyanin(KLH) and therefore may be ineffective in treating suddenSNHL.44

Cochlear PharmacokineticsSteroids delivered intratympanically can achieve

high concentrations in perilymph, higher than when ad-ministered by either intravenous or oral routes.13,45,46 Thepharmacokinetics of topically applied steroids within thecochlea have been studied. Using the markers trimethyl-phenylammonium (TMPA)47 and horseradish peroxi-dase,48 nonuniform distribution was noted with concen-tration of the markers near the round window (basal turn)higher than that of the apical turns. Salt found that sub-stances can reach the vestibule through extracellular com-munication between the scalae across the spiral ligamentas opposed to longitudinal flow through the helicotre-ma.49,50 These studies and that of Parnes13 suggest non-linear flow and interscalar communication of topicallyapplied substances through the spiral ligament.13,49,50 Us-ing the data of Parnes13 and Bachman,46 Plontke, usingthe Washington University Cochlear Fluids Simulatormodel (the Washington University cochlear Fluids Simu-lator, a public domain program available online at http://oto.wustl.edu/cochlea/),47 studied the effects of single ap-plication and continuous delivery of intratympanicsteroids on perilymph concentration. They found that rel-ative distribution of drugs in the inner ear is unlikely to beaffected by the application protocol. This finding is felt tobe secondary to the clearance of the drug from the peril-ymph, i.e., a drug that is rapidly cleared will not progressalong the scalar fluid. A steady state will be establishedwithin a matter of hours that will remain unchanged byfurther application. Application protocols did have amarked affect on drug levels achieved in the perilymphwith continuous delivery resulting in higher perilymphlevels than single application.48

Studies on Intratympanic Steroids for SuddenSensorineural Hearing Loss

The first report of intratympanic steroids in thetreatment of sudden SNHL was by Silverstein in 199612

followed by Parnes in 1999.13 Several other reports havebeen published since this initial report, most since20019,14–21,23–27 (Table VI). Most studies have shown thebenefits of intratympanic steroids in the treatment of sud-den HL in patients who had failed previous systemic ther-apy. Only 2 papers have studied the effects on intratym-panic steroids as a primary or initial therapy usedadjunctively with systemic steroids.23,24 Both of thesestudies reported that the addition of intratympanic ste-roids did not have a significant effect on the hearingrecovery in sudden SNHL. Lauterman23 compared pa-tients who received intratympanic steroids with systemicsteroids with patients who received no intratympanic ste-roids and found no benefit to the addition of intratympanicsteroids in hearing recovery. Battista24 noted minimalimprovement after intratympanic steroids for sudden

Fig. 3. Recovery rate related to severity of initial hearing loss basedon 20dB PTA/ 20% SDS.

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

9

TAB

LEV

I.S

umm

ary

ofS

tud

ies

Pub

lishe

dto

Dat

eon

Intr

atym

pan

icS

tero

ids

for

Sud

den

Sen

sorin

eura

lHea

ring

Loss

.

Aut

hor

Num

ber

ofP

atie

nts

Ste

roid

Typ

e

Tim

eC

ours

eFr

omH

earin

gLo

ssN

umb

erof

Inje

ctio

nsS

tud

yTy

pe

Per

cent

Imp

rove

dC

riter

iafo

rIm

pro

vem

ent

Silv

erst

ein

etal

.,12

1996

8D

exN

AU

pto

thre

etim

esa

wee

kfo

r3–

4w

eeks

Sal

vage

25%

10-d

BP

TA

(var

ious

dos

es)

15%

SD

S

Par

nes,

13

1999

13*

MP

40m

g/m

Lor

2d–6

wee

ks2–

29S

alva

gean

dp

rimar

y46

%Fi

veno

rmal

thre

shol

ds

Dex

One

serv

icea

ble

hear

ing

Kop

keet

al.,1

720

01†

3M

P62

.5m

g/m

L�

6w

ksC

athe

ter

for

14d

ays

Sal

vage

0%10

-dB

PTA

15%

SD

S

Kop

keet

al.,1

720

016

MP

62.5

mg/

mL

6w

eeks

orle

ssC

athe

ter

for

14d

ays

Sal

vage

83%

10-d

BP

TA

15%

SD

S

Gia

noli

and

Li,1

520

0123

Dex

25m

g/m

L72

wee

ksFo

ur(0

.4–0

.6m

L)ov

er10

–14

day

sS

alva

ge44

%10

-dB

PTA

MP

125

mg/

mL

10%

SD

S

Cha

ndra

sekh

ar,

14

2001

11§

Dex

2–4

mg/

mL

33d

ays

1–15

Sal

vage

73%

Incr

ease

inS

DS

,d

ecre

ase

inP

TA

Lefe

bre

and

Sta

eker

,16

2002

6M

P62

.5m

g/m

LA

pp

roxi

mat

ely

10d

ays

Cat

hete

r8–

10d

ays

Sal

vage

100%

�16

-dB

imp

rove

men

tin

PTA

Jack

son,

18

2002

19D

ex4–

24m

g/m

LN

AM

icro

Wic

kN

A31

%N

A

Thre

ed

rop

s3

times

/day

for

2–4

wee

ks

Gou

veris

,19

2003

a21

Dex

8m

g/m

L5.

45d

ays

2.7

aver

age

Sal

vage

Com

ple

tere

cove

ry(C

R)

33.3

%C

R:

with

in10

dB

ofun

affe

cted

ear

(1–7

)P

artia

l39.

1%P

artia

l:�

10-d

Bim

pro

vem

ent

Eve

ryse

cond

day

No

reco

very

(NR

)28

.6%

NR

:le

ssth

an10

-dB

imp

rove

men

t

Gou

veris

,20

03b

10D

ex8

mg/

mL

5.45

day

s2.

7av

erag

eS

alva

geC

R0%

See

a

(1–7

)P

60%

NR

40%

Gou

veris

,20

03c

9D

ex8

mg/

mL

5.45

day

s2.

7av

erag

eS

alva

geC

R0%

See

a

(1–7

)P

31%

NR

55.5

%

Ho

etal

.,20

2004

15D

ex4

mg/

mL

19.7

-day

aver

age

Thre

e(0

.4–0

.7m

L)ov

er3

wks

Sal

vage

53%

30-d

BP

TA

Laut

erm

anet

al.,2

320

0513

MP

32m

g/m

L2–

3d

ays

Five

over

5d

ays

Initi

al15

.3%

full

reco

very

Full,

par

tial,

none

,re

cove

rysc

ale;

noch

ange

from

cont

rol

15.3

%p

artia

l

69%

none

(con

tinue

s)

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

10

SNHL in his study of 25 patients all with profound(�90-dB PTA) HL.

Silverstein, in a retrospective review of 46 patientstreated with transtympanic steroids for a variety of disor-ders, had 8 patients with the diagnosis of sudden SNHL.One patient had improvement in speech reception thresh-olds from 110 dB to 85 dB and another from 75% to 65%SRT.12 Parnes et al. treated 37 patients with a variety ofinner ear disorders; 13 of these patients had suddenSNHL. Patients were treated with intratympanic methyl-prednisolone (9 patients) and intratympanic dexametha-sone (4 patients). All patients presented within 6 weeks ofonset of symptoms. Six of the 13 patients showed signifi-cant improvement in hearing thresholds, with 5 progress-ing from a severe or profound loss to relatively normalthresholds. No correlation between outcome and time oftreatment after HL was noted.13 Chandrasekhar treated10 patients who had variety of inner ear disorders with 2to 4 mg/mL dexamethasone intratympanically. The timeinterval between onset of loss and treatment averaged 33days. Most patients had failed medical therapy; however,some were treated primarily. Overall improvement wasnoted with a mean improvement of 9-dB PTA and 15.8%discrimination. Improvement was noted in all patientswith diabetes (3) and Meniere disease (2). Patients withlong intervals to treatment, downsloping audiogram, andsurgical trauma to the inner ear did not show recoverywith intratympanic steroids.14

Lefebre and Staeker treated 6 patients with suddenSNHL with methylprednisolone infused with a microcath-eter for 8 to 10 days. All patients had failed systemictherapy. All 6 patients showed improvement in hearingthresholds with an average of 16.25- to 25-dB improve-ment in thresholds.16 Kopke et al. reported the results ofintratympanic steroids delivered through an implantedmicrocatheter (62.4 mg/mL methylprednisolone at 10 �L/hour over 14 days) to treat patients with sudden SNHL.All of the patients treated in Kopke’s study had failed a2-week course of oral steroid therapy. Four of the 6 pa-tients had sudden SNHL. Five of 6 patients treated within6 weeks improved their hearing, with 4 returning to base-line hearing levels. None of the patients (3 of 3) whoinitiated treatment 6 weeks or more after the onset of the HLshowed improvement.17 Gianoli and Li in 2001 performed aprospective study on patients treated with intratympanicsteroids (dexamethasone or methylprednisolone) after treat-ment failure with systemic steroids for a minimum of 1week. A change of greater than or equal to 10 dB in the PTAor speech reception threshold or 10% in speech discrimina-tion was considered a positive response. A 44% response ratewas noted in these prior treatment failures with the averageimprovement of 15.2 dB and 21% SDS.15 Silverstein et al.examined 19 patients with sudden SNHL treated with in-tratympanic dexamethasone delivered through a MicroWickfor 2 to 4 weeks. Five (31%) patients had a positive response.The average gain in hearing was 45 dB and 39% discrimi-nation. Patients treated earlier had better results, although1 patient responded more than 1 year after the onset ofsymptoms.18

Gouveris treated 40 patients with intratympanic ste-roids in a prospective study of patients who had failed

TAB

LEV

I.(C

ontin

ued

)

Aut

hor

Num

ber

ofP

atie

nts

Ste

roid

Typ

e

Tim

eC

ours

eFr

omH

earin

gLo

ssN

umb

erof

Inje

ctio

nsS

tud

yTy

pe

Per

cent

Imp

rove

dC

riter

iafo

rIm

pro

vem

ent

Her

ran

dM

arzo

,21

2005

17D

ex10

mg/

mL

6.3

wee

ksP

ump

,M

icro

Wic

kS

alva

ge53

%10

-dB

PTA

MP

62.5

mg/

mL

(2–2

0w

ks)

8–14

day

s20

%S

DS

Bat

tista

,24

2005

‡25

Dex

24m

g/m

L28

day

sFo

urov

er14

day

sIn

itial

8%fu

llFu

ll:w

ithin

10d

Bb

asel

ine

12%

par

tial

Par

tial:

with

in50

dB

bas

elin

e

Sla

tter

yet

al.,9

2005

20M

P62

.5m

g/m

LU

pto

3m

onth

sFo

urin

ject

ions

over

2w

eeks

Sal

vage

55%

10-d

BP

TA

12%

SD

S

Cho

ung

etal

.,25

2006

33D

ex5

mg/

mL

28d

ays

2in

ject

ions

per

wee

kfo

r2

wee

ksS

alva

ge38

.2%

10d

BP

TA15

%S

DS

Dal

lan

etal

.,26

2006

8M

P40

mg/

mL

21.5

day

sS

ingl

ein

ject

ion

Sal

vage

75%

15d

BP

TA

Xen

ellis

etal

.,27

2006

19M

P40

mg/

mL

21d

ays

4in

ject

ions

over

15d

ays

Sal

vage

47%

10d

BP

TA

Hay

nes

etal

.40

Dex

40d

ays

Sin

gle

inje

ctio

nS

alva

ge26

.7%

20-d

BP

TA

24m

g/m

L40

day

s20

%S

DS

Not

e.G

ouve

risab

c;au

thor

rep

orte

dre

sults

sep

arat

ely:

aha

dhe

arin

gth

resh

old

s�

30d

Bb

ut�

80d

B;

bha

dhe

arin

gth

resh

old

s�

80d

B;

cha

dp

red

omin

antly

high

freq

uenc

yhe

arin

glo

ss.

*Pat

ient

sw

ithsu

dd

ense

nsor

ineu

ralh

earin

glo

ssfr

oma

larg

erre

por

tof

37p

atie

nts.

†Stu

dy

div

ided

into

two

grou

ps

bas

edon

time

top

rese

ntat

ion.

‡Stu

dy

only

ente

red

pat

ient

sw

ithp

rofo

und

(�90

dB

)he

arin

glo

ss.

§Not

allp

atie

nts

had

sud

den

sens

orin

eura

lhea

ring

loss

.P

TA�

pur

e-to

neav

erag

e;S

DS

�sp

eech

dis

crim

inat

ion

scor

e;D

ex�

dex

amet

haso

ne;

MP

�m

ethy

lpre

dni

solo

ne.

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

11

systemic therapy. Overall significant improvements inhearing thresholds were noted. Reduced efficacy wasnoted in patients who had profound HL and primarilyhigh frequency loss at presentation.19 Lauterman et al. in2005 reported the results of a prospective, controlled studyin which transtympanic methylprednisolone was used as aprimary treatment modality, not as salvage therapy aftersystemic steroids failure. Twenty-seven patients weretreated with systemic therapy (rheologic agents and sys-temic steroids) with 13 of these undergoing intratympanicsteroids in addition to the systemic therapy. No differencein recovery was noticed between the group treated withintratympanic steroids and the control group. Ho et al. in2005 reported a randomized, controlled study of 39 pa-tients with sudden SNHL in which 29 (74%) failed sys-temic steroids and were randomized into 2 treatmentgroups. Fifteen patients received intratympanic steroidtherapy and 14 were continued on further medical therapy(without steroids). They noted 53% improvement in theintratympanic steroid group as opposed to 7.1% for thenoninjected group using 30-dB gain in PTA as criteria forsuccessful outcome. Age, treatment delay time, and sexdid not affect response to therapy.20

Herr and Marzo reported on 17 patients treated withtranstympanic steroids (dexamethasone initially, methyl-prednisolone later in the study) through a MicroWickand/or round window catheter placement. All patients hadfailed prior systemic therapy with prednisone and weretreated from 2 to 20 weeks after onset of HL. Overall, 53%showed improvement in thresholds after treatment.21 Theaverage improvement was 24.3 dB in those ears thatshowed improvement. Battista in 2005 enrolled 25 pa-tients with profound SNHL. Both systemic and intratym-panic steroids were used concomitantly. Overall poor re-sults were achieved in this population of profound HLpatients with only 12% (3 of 25) achieving a full or partialresponse.24 Slattery et al. reported 20 patients treatedwith methylprednisolone for sudden SNHL that failedsystemic steroids. Fifty-five percent showed clinically sig-nificant (10-dB PTA or 12% discrimination) improvementin hearing. Improvement in tinnitus was also noted.9

In 2006, Dallan et al. treated 8 consecutive patientswith intratympanic methylprednisolone in a prospectivestudy, with 75% improving after a single injection.26

Choung et al. had a 38% improvement following withintratympanic therapy compared to 6.1% improvement ina control group treated with systemic therapy alone.25

Also in 2006 Xenellis showed a 47% improvement in pa-tients with intratympanic therapy following treatmentfailure, while none of the patients in a matched controlimproved over time.27

With the natural history of sudden SNHL suggestinga high recovery rate, it is difficult to determine if anytherapeutic intervention actually improves hearing recov-ery. The natural history of untreated patients with suddenSNHL ranges from recovery rates of 31% to 65%.1,3,4,8 Therange of hearing recovery reported in the literature intreated patients ranges from 35% to 89%.7,11 Several rea-sons may explain the significant differences in reportedrecovery rates between studies; however, the best expla-nation may lie in what is considered a “successful” treat-

ment outcome. Mattox and Simmons reporting a relativelyhigh spontaneous recovery rate of 65% classified goodrecovery as a final PTA of less than 40 dB or a more than50-dB improvement in the initial audiogram.4 Slattery etal. used the criteria described by Wilson et al.1 to describerecovery (recovery of �50% of baseline hearing). Usingthis formula, they reported a hearing recovery rate withsystemic therapy of 35%. Chen and Wilson using thisformula described hearing recovery rates of 55%3 and78%,1 respectively.

On review of the studies published to date in intra-tympanic steroids, it is clear that studies in the literaturealso differ on the definition of “success” for significantimprovement after therapy. No definitive criteria exist todefine recovery in patients with sudden SNHL, especiallysecondary recovery after initial treatment failure. Thecriteria to which the authors define recovery range fromany improvement in PTA or SDS14 to an improvement in10-dB PTA or 10% SDS15 to the criteria described byWilson et al.1 that describes recovery as �50% of theinitial loss.24 A true meta-analysis of the literature is notpossible given the wide variance in treatment protocols,patient data, and reporting of data. However, if we applyour hearing outcome data to the criteria in the currentliterature, similar hearing recovery rates are often found.For example, Gianoli15 reported a success rate of 44%using the criteria of 10-dB PTA/10% SDS; using this cri-teria applied to our data, the success rate rises from 26.7%to 40%. Table IV and Figure 4 illustrate the change in ourreported success rate when different criteria are applied.Our hearing recovery rates are surprisingly similar tomany other studies on intratympanic steroids for suddenSNHL when similar criteria are used to calculate the ratesof improvement despite marked differences in treatmentprotocols, number of injections, steroid type and concen-tration. There are several possible explanations for thesimilar recovery rates between studies. One possible ex-planation is that the variables of steroid type, dose, andnumber of injections may have minimal influence ontreatment outcome and recovery rates. Another possibleexplanation is that the injection had no effect at all andthe similar recovery rates reflect the natural history of

Fig. 4. Percent recovery of patients injected 6wks or less, usingvarious definitions for recovery.

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

12

recovery from idiopathic sudden SNHL. Although possiblein theory, the prognosis for recovery in patients with sud-den HL who have failed systemic therapy in our experi-ence and others is uniformly poor.9,10,15

With the inclusion/exclusion criteria applied, all pa-tients with Meniere disease, autoimmune HL, or fluctuat-ing loss requiring multiple injections were excluded andonly patients with true idiopathic sudden SNHL thatfailed systemic steroids were studied. Although our recov-ery rate is low (26.7%), it is difficult to compare with thereported natural history recovery rates, because our grouphad failed systemic therapy and were an average of 40days out from onset of symptoms. The literature wouldsupport a low chance of further recovery in this group. InHo’s study, only 7.1% of the control group of those whofailed systemic therapy gained further hearing on followup.20 Choung25 and Xenellis27 respectively showed 6.1%and 0% improvement in hearing after failed systemic ther-apy with long-term follow up. Zadeh noted an improve-ment in hearing for those patients seen and treated within3 days of the onset of symptoms as compared with thosetreated beyond 3 days.6 Shaia and Sheehy noted a signif-icant improvement in patients treated within 1 week orless. However, some patients who initiated therapy after 3months had recovery (10%).52 Fuse et al. noted that themajority of patients who recovered completely after treat-ment with oral steroids did so within 7 to 10 days afteradministration of steroids. In long-term follow up of 3months to 2 years, none of the patients with no recovery orpartial recovery recovered to normal hearing levels. Theynoted that patients resistant to steroids with regard toearly outcome continued to have poor hearing recoveryduring long-term follow up.6 Ito et al. noted hearing out-comes in 90 patients treated for sudden HL. Patients withimprovement within 2 weeks were more likely to have abetter outcome. Patients with poor recovery at 2 weeksshowed minimal improvement at 1-month follow up.10

Lefebre reported that 100% of patients treated with ste-roids for sudden SNHL recovered within 7 days.16

Because controversy exists regarding efficacy of sys-temic steroids in treating sudden SSNHL, it will exist aswell regarding the use of intratympanic steroids in thistreatment population. Truly, only one patient recoveredhearing to within 10 dB of the contralateral ear with asecond patient having a significant improvement in dis-crimination of the 40 treatment failures studied. Theseresults are hardly considered dramatic. However, 39% ofpatients recovering 20 dB or 20% SDS (if treated within 6weeks) in this group of treatment failures is higher thanwould be expected given our controls (9.1%), experience,and literature review. If we further exclude 7 patientstreated with intratympanic steroids within 2 weeks of theonset of symptoms (i.e., study only those patients treatedwith intratympanic dexamethasone between 2 and 6weeks after onset of symptoms), still, 26% improved by 20dB or 20% SDS. This recovery is higher than what wouldbe expected by our experience, control group (9.1%), andliterature review. Although this represents one of thelargest series of treatment failures to date treated withsteroid perfusion, the statistical power of the study doesnot support efficacy. Although it lacks statistical power,

the data suggest a trend toward efficacy of steroid perfu-sion in this treatment group.

We excluded patients who had Meniere disease, fluc-tuating HL, and autoimmune HL from the study to refinethe study to patients with idiopathic sudden SNHL. Anypatient whose hearing fluctuated and subsequently re-ceived multiple injections was excluded. This exclusioneliminated a number of patients who would have beenclassified as a good response. That is, some patients whoresponded to intratympanic injection whether recoverywas from a true therapeutic intervention or from recoverythrough natural history were more likely to receive asecond injection if hearing fluctuated. Those who did notrespond to initial intratympanic steroids did not receivesubsequent injections. Exclusion of all patients who dem-onstrated evidence of fluctuation of hearing before injec-tion was done so improvement after injection would be lesslikely attributed to upward fluctuation; 87.5% of patientsin the study group had a normal contralateral ear in thecurrent study group, reflective of patients with idiopathicsudden SNHL.

Recovery in the group with diabetes (20%) was sim-ilar to the nondiabetics (28.6%). No complications werenoted in this group; however, one patient did have wors-ening thresholds. Four of 9 (44%) of the patients withdiabetes did not receive systemic steroids before injectionwith one (25%) recovering hearing using 20 dB or 20%discrimination as definition for recovery. These patientswere not included in the primary study but are comparedwith those receiving systemic steroids in Figure 5. Five of9 (56%) received systemic steroids before injection withone (20%) showing recovery using this criteria. Chan-drasekhar noted improvement in 3 of 3 patients withdiabetes treated with intratympanic therapy for suddenSNHL.14 Diabetic patients are felt to have a poorer overallrecovery from sudden SNHL.52,53

Although not statistically significant, some of thedifferences between groups were noteworthy. Patientswith abnormal hearing in the contralateral ear had aslightly better recovery rate (33%) than those with normalhearing in the contralateral ear (26.5%) (P � 1.0, Fisherexact test). Patients with no vertigo had a recovery rate of32% compared with those with vertigo who had a recoveryrate of 20% (P � .5, Fisher exact test). The presence of

Fig. 5. Recovery related to diabetes. Note diabetics that did notreceive systemic steroids were excluded from the primary study.

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

13

vertigo has been shown to be a poor prognostic sign inseveral studies.2,4,21,52,53 Those patients with severe losseshad a poorer recovery (8.3%) compared with losses thatwere less than 90 dB (35.7% recovery) (P � .1, Fisher exacttest). Severe losses have been shown in several studies tohave poorer recovery rates.2,7,11,24

A total of 17.5% of patients (n � 7) had a worse PTA(defined as any drop in the PTA on follow-up testing) afterinjection. Although this seems relatively high, the averagedrop in PTA was only 3.8 dB. A more significant drop wasseen in changes in SDS. Only 5 patients had worsening ofSDS (defined as any drop in SDS on follow-up testing)after injection, but the average drop was 16%. The major-ity of the loss is from one patient who dropped 28% on theSDS scores after injection. This patient developed a tem-porary perforation with otorrhea. If this patient is ex-cluded, the average drop in discrimination in the remain-ing 4 patients is 12%. No other patient had perforation,otitis media, otorrhea pain, or vertigo after intratympanicinjection. It is unclear as to whether these losses can bedirectly attributed to the procedure. Progression of lossmay be seen in up to 15% of patients with sudden SNHL,4which is consistent with our controls (18%). These lossesmost likely fall within the range of natural progression ofdisease. No other complications were noted.

The limitations of this study lie primarily in theretrospective analysis of patient data. This study lacksformal control and should be interpreted as a descriptionof the clinical experience from a single institution in thetreatment of sudden SNHL with a single injection of in-tratympanic steroids. However, the studied group wouldbe expected to have a poor prognosis given the delay totherapy (average over 40 days) and failure of systemictherapy. Our small control group is reflective of the factthat long-term follow up after documented treatment fail-ure in patients is uncommon. Given the low numbers ofour control population, statistical analysis was not possi-ble; only descriptive analysis can be made in comparisonto our treatment group. Other limitations may be the typeof steroid used and the dosing schedule applied. Dexa-methasone has good round window diffusion; however, theprofile may not be as beneficial as methylprednisolone.Parnes showed that methylprednisolone had a higher con-centration and longer duration in perilymph after tran-stympanic administration than hydrocortisone or dexa-methasone.13 Despite the practicality in treating patientswith a single intratympanic injection of steroids, this pro-tocol may not be as optimal as a continuous infusion ormultiple injections.16,17

CONCLUSIONDramatic improvements in PTA or SDS were uncom-

mon in this group of patients treated as salvage patients(failed systemic therapy), with only 1 patient in 40 recov-ering to within 10 dB of the contralateral ear. However,39% of patients recovering 20 dB or 20% SDS (if treatedwithin 6 weeks) in this group of treatment failures ishigher than would be expected given our controls (9.1%),experience, and literature review. Although this repre-sents one of the largest series of treatment failures to datetreated with steroid perfusion, the statistical power of the

study does not support efficacy. Despite failure to reachstatistical significance, the data suggest a trend towardefficacy of steroid perfusion in patients who have failedsystemic steroid therapy. No patient showed benefit basedon our criteria from intratympanic steroids after 36 dayswhen using this protocol for idiopathic sudden SNHL. ANational Institutes of Health-sponsored, prospective trialis being conducted to determine the potential therapeuticefficacy in treating sudden SNHL, and will hopefully fur-ther answer questions regarding this treatment option.

BIBLIOGRAPHY1. Wilson WR, Byl FM, Laird N. The efficacy of steroids in the

treatment of idiopathic sudden hearing loss. A double-blind clinical study. Arch Otolaryngol 1980;106:772–776.

2. Byl FM Jr. Sudden hearing loss: eight years’ experience andsuggested prognostic table. Laryngoscope 1984;94:647–661.

3. Chen CY, Halpin C, Rauch SD. Oral steroid treatment ofsudden sensorineural hearing loss: a ten year retrospectiveanalysis. Otol Neurotol 2003;24:728–733.

4. Mattox DE, Simmons FB. Natural history of sudden sensori-neural hearing loss. Ann Otol Rhinol Laryngol 1977;86:463–480.

5. Zadeh MH, Storper IS, Spitzer JB. Diagnosis and treatmentof sudden-onset sensorineural hearing loss: a study of 51patients. Otolaryngol Head Neck Surg 2003;128:92–98.

6. Fuse T, Aoyagi M, Funakubo T, Sakakibara A, Yoshida S.Short-term outcome and prognosis of acute low-tone sen-sorineural hearing loss by administration of steroid. ORL JOtorhinolaryngol Relat Spec 2002;64:6–10.

7. Slattery WH, Fisher LM, Iqbal Z, Liu N. Oral steroid regi-mens for idiopathic sudden sensorineural hearing loss.Otolaryngol Head Neck Surg 2005;132:5–10.

8. Cinamon U, Bendet E, Kronenberg J. Steroids, carbogen orplacebo for sudden hearing loss: a prospective double-blindstudy. Eur Arch Otorhinolaryngol 2001;258:477–480.

9. SlatteryWH, Fisher LM, Iqbal Z, Friedman RA, Liu N. Intra-tympanic steroid for the treatment of sudden hearing loss.Otolaryngol Head Neck Surg 2005;133:251–259.

10. Ito S, Fuse T, Yokota M, et al. Prognosis is predicted by earlyhearing improvement in patients with idiopathic sudden sen-sorineural hearing loss. Clin Otolaryngol 2002;27:501–504.

11. Moskowitz D, Lee KJ, Smith HW. Steroid use in idiopathicsudden sensorineural hearing loss. Laryngoscope 1984;94:664–666.

12. Silverstein H, Choo D, Rosenberg SI, Kuhn J, Seidman M,Stein I. Intratympanic steroid treatment of inner ear dis-ease and tinnitus (preliminary report). Ear Nose Throat J1996;75:468–471.

13. Parnes LS, Sun AH, Freeman DJ. Corticosteroid pharmaco-kinetics in the inner ear fluids: an animal study followedby clinical application. Laryngoscope 1999;109:1–17.

14. Chandrasekhar SS. Intratympanic dexamethasone for sud-den sensorineural hearing loss: clinical and laboratoryevaluation. Otol Neurotol 2001;22:18–23.

15. Gianoli GJ, Li JC. Transtympanic steroids for treatment ofsudden hearing loss. Otolaryngol Head Neck Surg 2001;125:142–146.

16. Lefebvre PP, Staecker H. Steroid perfusion of the inner earfor sudden sensorineural hearing loss after failure of con-ventional therapy: a pilot study. Acta Otolaryngol 2002;122:698–702.

17. Kopke RD, Hoffer ME, Wester D, O’Leary MJ, Jackson RL.Targeted topical steroid therapy in sudden sensorineuralhearing loss. Otol Neurotol 2001;22:475–479.

18. Jackson LE, Silverstein H. Chemical perfusion of the innerear. Otolaryngol Clin North Am 2002;35:639–653.

19. Gouveris H, Selivanova O, Mann W. Intratympanic dexa-methasone with hyaluronic acid in the treatment of idio-pathic sudden sensorineural hearing loss after failure of

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

14

intravenous steroid and vasoactive therapy. Eur Arch Oto-rhinolaryngol 2005;262:131–134.

20. Ho GM, Lin HG, Shu MT. Effectiveness of intratympanicdexamethasone injection in sudden deafness patients assalvage treatment. Laryngoscope 2004;114:1184–1189.

21. Herr BD, Marzo SJ. Intratympanic steroid perfusion for re-fractory sudden sensorineural hearing loss. OtolaryngolHead Neck Surg 2005;132:527–531.

22. Itoh A, Sakata E. Treatment of vestibular disorders. ActaOtolaryngol Suppl 1991;481:617–623.

23. Lauterman J, Sudhoff H, Junker R. Transtympanic corticoidtherapy for acute profound loss. Eur Arch Otorhinolaryn-gol 2005:1–9.

24. Battista RA. Intratympanic dexamethasone for profound id-iopathic sudden sensorineural hearing loss. OtolaryngolHead Neck Surg 2005;132:902–905.

25. Choung YH, Park K, Shin YR, Cho MJ. Intratympanic dexa-methasone injection for refractory sudden sensorineuralhearing loss. Laryngoscope 2006;116:747–752.

26. Dallan I, Bruschini P, Nacci A, et al. Transtympanic steroidsas a salvage therapy in sudden hearing loss: preliminaryresults. ORL J Otorhinolaryngol Relat Spec 2006;68:247–252.

27. Xenellis J, Papadimitriou N, Nikolopoulos T, et al. Intratym-panic steroid treatment in idiopathic sudden sensorineuralhearing loss: a control study. Otolaryngol Head Neck Surg2006;134:940–945.

28. Rarey KE, Luttge WG. Presence of type I and type II/IBreceptors for adrenocorticosteroid hormones in the innerear. Hear Res 1989;41:217–221.

29. Stockroos RJ, Albers FW, Schirm J. The etiology of idiopathicsudden sensorineural hearing loss. Experimental herpes sim-plex virus infection of the inner ear. Am J Otol 1998;19:447–452.

30. Nagura M, Iwasaki S, Wu R, et al. Effects of corticosteroid,contrast medium and ATP on focal microcirculatory dis-orders of the cochlea. Eur J Pharmacol 1999;366:47–53.

31. Tabuchi K, Oikawa K, Uemaetomari I, Tsuji S, Wada T, HaraA. Glucocorticoids and dehydroepiandrosterone sulfateameliorate ischemia-induced injury of the cochlea. HearRes 2003;180:51–56.

32. Lamm K, Arnold W. The effect of prednisolone and non-steroidal anti-inflammatory agents on the normal and noise-damaged guinea pig inner ear. Hear Res 1998;115:149–161.

33. Yao X, Buhi WC, Alvarez IM, Curtis LM, Rarey KE. De novosynthesis of glucocorticoid hormone regulated inner earproteins in rats. Hear Res 1995;86:183–188.

34. Lin DW, Trune DR. Breakdown of stria vascularis blood-labyrinth barrier in C3H/lpr autoimmune disease mice.Otolaryngol Head Neck Surg 1997;117:1–8.

35. Trune DR, Wobig RJ, Kempton JB, Hefeneider SH. Steroidtreatment improves cochlear function in the MRL.MpJ-Fas(lpr) autoimmune mouse. Hear Res 1999;137:160–166.

36. Shirwany NA, Seidman MD, Tang W. Effect of transtympanicinjection of steroids on cochlear blood flow, auditory sensi-tivity, and histology in the guinea pig. Am J Otol 1998;19:230–235.

37. El-Hennawi DM, El-Deen MHB, Abou-Halawa AS, NadeemHS, Ahmed MR. Efficacy of intratympanic methylpred-nisolone acetate in treatment of drill-induced sensori-neural hearing loss in guinea pigs. J Laryngol Otol 2005;

119:2–7.38. Yilmaz I, Yilmazer, Erkan AN, Aslan SG, Ozluoglu LN. In-

tratympanic dexamethasone injection effects on transient-evoked otoacoustic emission. Am J Otololaryngol 2005;26:113–117.

39. Fukushima M, Kitahara T, Uno Y, Fuse Y, Doi K, Kubo T.Effects of intratympanic injection of steroids on changes inrat inner ear aquaporin expression. Acta Otolaryngol 2002;122:600–606.

40. Himeno C, Komeda M, Izumikawa M, et al. Intra-cochlearadministration of dexamethasone attenuates aminoglyco-side ototoxicity in the guinea pig. Hear Res 2002;167:61–70.

41. Sone M, Hayashi H, Yamamoto H, Tominaga M, NakashimaT. A comparative study of intratympanic steroid and NOsynthase inhibitor for treatment of cochlear lateral walldamage due to acute otitis media. Eur J Pharmacol 2003;482:313–318.

42. Spandow O, Hellstrom S, Anniko M. Impaired hearing fol-lowing instillation of hydrocortisone into the middle ear.Preliminary report from an animal study. Acta OtolaryngolSuppl 1988;455:90–93.

43. Nordang L, Linder B, Anniko M. Morphologic changes inround window membrane after topical hydrocortisone anddexamethasone treatment. Otol Neurotol 2003;24:339–343.

44. Yang GSY, Song HT, Keithley EM, Harris JP. Intratym-panic immunosuppressives for prevention of immune-mediated sensorineural hearing loss. Am J Otol 2000;21:499 –504.

45. Chandrasekhar SS, Rubinstein RY, Kwartler JA, et al.Dexamethasone pharmacokinetics in the inner ear: com-parison of route of administration and use of facilitatingagents. Otolaryngol Head Neck Surg 2000;122:521–528.

46. Bachmann G, Su J, Zumegen C, Wittekindt C, Michel O.Permeabilitat der runden Fenstermembran furPrednisolon-21-Hydrogensuccinat. Prednisolongehalt derPerilymphe nach lokaler Applikation vs. systemischer In-jektion. HNO 2001;49:538–542.

47. Salt AN, Ma Y. Quantification of solute entry into cochlearperilymph through the round window membrane. Hear Res2001;154:88–97.

48. Saijo S, Kimura RS. Distribution of HRP in the inner earafter injection into the middle ear cavity. Acta Otolaryngol1984;97:593–610.

49. Salt AN, Ohyama K, Thalmann R. Radial communicationbetween the perilymphatic scalae of the cochlea. I: Estima-tion by tracer perfusion. Hear Res 1991;56:29–36.

50. Salt AN, Ohyama K, Thalmann R. Radial communicationbetween the perilymphatic scalae of the cochlea. II: Esti-mation by bolus injection of tracer into the sealed cochlea.Hear Res 1991;56:37–43.

51. Plontke S, Zenner HP. Pharmacokinetic considerations inintratympanic drug delivery to the inner ear. Acta Otorhi-nolaryngol Belg 2002;56:369–370.

52. Shaia FT, Sheehy J. Sudden sensorineural hearing impair-ment: a report of 1220 cases. Laryngoscope 1976;86:389–398.

53. Fukui M, Kitagawa Y, Nakamura N, et al. Idiopathic suddenhearing loss in patients with type 2 diabetes. Diabetes ResClin Pract 2004;63:205–211.

Laryngoscope 117: January 2007 Haynes et al.: Intratympanic Dexamethasone for Sudden SNHL

15