Outcomes of Intratympanic Gentamicin Injection to Treat ：庆大霉素鼓室内注射治疗的结果.doc

Outcomes of Intratympanic Gentamicin Injection to Treat Meniere DiseaseLeh-Kiong Huon MD,1 Te-Yung Fang MD,1 Pa-Chun Wang MD, MSc1,2,3Affiliations/Institutions1Department of Otolaryngology, Cathay General Hospital, Taipei, Taiwan2Fu Jen Catholic University School of Medicine, New Taipei City, Taiwan3Department of Public Health, College of Public Health, China Medical University, Taichung, TaiwanRunning Title: Intratympanic Gentamicin for Meniere DiseaseWord Count: 2892 WordsConflict of Interest: NoneFinancial Support: NoneEvidence Level: IcAddress correspondence and reprint requests to:Pa-Chun Wang MD, MSc.Department of Otolaryngology, Cathay General Hospital280 Sec 4 Jen-Ai Rd, 106 Taipei, TaiwanE-mail address: drtony.twTel.: +886-2-27082121 ext 3333; Fax: +886-2-66362836ABSTRACTObjective: To investigate the outcomes of intratympanic (IT) gentamicin sulfate injection to treat intractable Meniere disease (MD) using evidence-based methods.Data Source: Data were retrieved from a MEDLINE search (January 1995 through January 2011).Study Selection: Selected were English-language articles presenting outcomes of IT gentamicin injection to treat MD that had a prospective study design and an evidence level of II or higher.Data Extraction: The database was searched using Boolean combinations of the keywords intratympanic, gentamicin, and Meniere disease.Data Synthesis: Dichotomous outcomes (success rate and total deafness) were expressed using a Mantel-Haenszel fixed-effects model. Pretreatment and posttreatment pure-tone audiometry results and word discrimination scores were summarized using data synthesis techniques.Conclusions: A critical literature appraisal and meta-analysis shows that IT gentamicin injection can control vertigo in patients with MD who have limited hearing injury. The quality of research in this field has improved remarkably over the years. However, a large-scale randomized controlled trial is warranted to confirm the predictive factors for the effectiveness of IT gentamicin injection in treating MD.Key Words: adult, antibiotics, ear, evidence-based medicine, hearing loss, vertigoINTRODUCTIONMeniere disease (MD) is an idiopathic inner ear condition characterized by tinnitus, fluctuating hearing impairment, and episodic spinning vertigo. Various treatment options exist, including a salt restriction diet, medication, and shunt or labyrinthine ablation surgical procedures.Intratympanic (IT) gentamicin sulfate injection for the treatment of MD has recently gained increasing popularity, partially because of its convenience, safety, and efficacy. Injections can be administered in an ambulatory setting, with selective labyrinthine ototoxic effects. Several systematic reviews1-4 have proved the clinical efficacy of IT gentamicin injection to treat MD.Several modalities of IT gentamicin injection have been proposed to treat MD, including fixed, titration, or continuous protocols based on single, daily, or weekly schedules. However, because of difficulties in conducting randomized controlled trials, evidence-based consensus on the most appropriate IT gentamicin injection protocol for the treatment of MD has not been established. Furthermore, previous systematic reviews1-4 failed to identify patient characteristics or therapeutic variables that may confound the outcomes of IT gentamicin injection.Meta-analysis represents a powerful tool to pool data across studies and to increase the chance of detecting statistically significant treatment effects. The objectives of this study were as follows: (1) to identify and review evidence-based studies on IT gentamicin injection to treat intractable MD, (2) to pool the data and perform meta-analysis to increase the statistical power of treatment efficacy, and (3) to explore patient characteristics (sex, age, disease severity, and disease duration) and therapeutic variables (total number of injections and cumulative dose) that may affect the outcomes of IT gentamicin injection to treat MD.MATERIALS AND METHODSLiterature SearchThe MEDLINE database (January 1995 through January 2011) was searched using the following Boolean combinations of keywords: intratympanic and Meniere disease, gentamicin and Meniere disease, and intratympanic and gentamicin and Meniere disease. The search was supplemented by identifying additional listings using the feature to link with related articles from each of the original articles. Duplicate results were excluded. The quality of articles was assessed through independent cross-checking by 2 of us (L.K.H. and P.C.W.).Article Selection Criteria and Data ExtractionOur systematic review of current literature was based on articles published in peer-reviewed journals up to and including January 2011. Only original articles were considered. To identify published studies for inclusion in this analysis, study populations had to include patients undergoing IT gentamicin injection for MD; publications had to address outcomes of IT gentamicin injection to treat MD. We excluded letters to editors, case series (<10 cases), review articles, and publications available only in abstract format. Studies had to have been conducted on humans. NonEnglish-language publications were excluded.Observations and OutcomesThe selected articles were screened for the following: study design (prospective or retrospective clinical trial, controlled, or randomized), population size, patient demographics (sex and age), clinical characteristics (disease classification, hearing impairment, tinnitus, aural fullness, vestibular dysfunction, disease duration, disease severity or stage, functional level, or subjective symptom survey results), treatment protocol (fixed or titration, total number of injections, and cumulative dose), and follow-up period.Using the American Academy of OtolaryngologyHead and Neck Surgery, Inc5 guidelines for the diagnosis and evaluation of therapy in MD (AAO-HNS guidelines), the following were described: vertigo control (class A-F), pure-tone audiometry (PTA) results (mean at 250, 1000, 2000, and 3000 Hz), word discrimination scores (WDS) percentage change, functional level (range, 1-6), vestibular dysfunction (percentage deficit on the caloric test), and subjective symptom survey results (tinnitus, aural fullness, and Tumarkin crisis drop attack).We defined substantial vertigo control as class A or B using the AAO-HNS guidelines.5 Hearing changes of 10 dB in PTA results and 15% in WDS were considered clinically significant. Change in tinnitus and aural fullness (better, same, or worse) was recorded.Study HypothesesThere were 4 study hypotheses: (1) Intratympanic gentamicin injection can control vertigo in patients with MD. (2) Intratympanic gentamicin injection can cause further hearing damage in patients with MD. (3) Intratympanic gentamicin injection treatment protocol and cumulative dose can affect vertigo control and hearing outcomes. (4) Confounders of IT gentamicin injection include clinical characteristics, such as disease classification, disease duration, disease severity or stage, and functional level.Data AnalysisData were extracted and systemically reviewed. Dichotomous (substantial control of ITG, total deafness) outcomes were expressed using Mantel-Haenszel fixed-effects model, and the odds ratio was calculated. If the 95% CI did not include 1.00, it was considered significant at p< 0.05. The results for the overall outcomes (pretreatment and posttreatment means SDs and the effect mean SD) of IT gentamicin injection were summarized. Data synthesis techniques were used for PTA results and WDS.Cochran Q test for determining heterogeneity was also applied. Sample population differences between studies were considered significant at p< 0.05.RESULTSStudy CharacteristicsThe MEDLINE search revealed 14 original articles6-19 dealing with the use of IT gentamicin injection for the treatment of MD. The article selection algorithm is shown in Figure 1. Two studies15,18 used control groups, and their results were graded as level I evidence; in terms of research quality, the remaining 12 articles6-14,16,17,19 would be graded as containing level II evidence.The participants in the 14 studies (sample size range, 11-127) totaled 599 patients with MD who received IT gentamicin injection. Study characteristics are given in Table 1. In terms of clinical characteristics, the populations were heterogeneous across studies (p< 0.05).Treatment ProtocolAmong the reviewed articles, IT gentamicin injection was administered using fixed protocols in 7 studies.7-10,15-17 Intratympanic gentamicin injection was administered using titration protocols in 7 studies.6,11-14,18,19The studies were further classified according to their protocol for meta-analysis.2 The results showed that no study had a multiple daily protocol, 1 study had a weekly protocol, 6 studies had low-dose and high-dose (1-2 injections and fixed) protocols, no study had a continuous protocol with microcatheter, no study had a titration daily protocol, and 7 studies had titration weekly protocols.Among studies, the mean number of IT gentamicin injections was 2.1 (1.1), with a mean cumulative dose of 27.7 (20.5) mg (fixed range, 7.4-90.8 mg; titration range, 8.8-32.4 mg). Eight studies had more than 2.1 injections; 11 studies had a cumulative dose exceeding 13.4 mg. The treatment protocols are summarized in Table 1.Patient Assessments and Outcomes MeasuresBased on objective assessments by patients, PTA was performed in 11 studies,7-11,13,14,16-19 WDS in 6 studies,7,10,14,16,17,19 and caloric test in 12 studies.6-14,16,17,19 Among other physiological or imaging tests, electrocochleography was included in 1 study,8 rotating chair in 2 studies,13,14 auditory brainstem-evoked potentials in no study, vestibular evoked myogenic potentials in 1 study,9 computed tomography in 2 studies,10,11 and magnetic resonance imaging in 3 studies.7,10,11 All studies used outcomes measures consistent with AAO-HNS guidelines.5Seven articles6,7,10,13,14,16,17 used subjective symptoms as outcomes end points, including the Functional Level Scale5 in 7 studies, the University of California, Los Angeles Dizziness Questionnaire20 in 2 studies, the Dizziness Handicap Inventory21 in 2 studies, and the Vertigo Symptom Scale22 in 1 study. No article used the general health survey.Patient CharacteristicsThe study populations ranged in age from 46.1 to 59 years, with a mean age of 53.4 years. The proportion of male patients in 8 studies6,7,9,11-14,17 that provided this information ranged from 32.1% to 60%.Eight studies6,7,9,11-14,17 reported the disease duration; 11 studies6,7,9-14,16,17,19 reported the type and stage (severity) of MD. The mean disease duration was 78.9 (2.1) months (range, 27.6-117.6 months); 1 patient had had MD for more than 10 years. The mean follow-up time was 49.8 (10.0) months (range, 12.0-63.6 months); 10 studies6,8,10-14,16,17,19 followed up patients for more than 24 months.At baseline, patients in all studies had definitive MD; 8 studies6,9,10,12-14,17,19 provided staging data for 388 patients. Among these, the stages of disease were stage I in 17 patients (4.4%), stage II in 43 patients (11.1%), stage III in 236 patients (60.8%), and stage IV in 92 patients (23.7%).Outcomes of IT Gentamicin InjectionThe degree of vertigo control among 559 patients with MD was as follows: class A in 457 patients (71.4%), class B in 59 patients (16.1%), class C in 15 patients (4.3%), class D in 7 patients (2.4%), class E in 10 patients (2.9%), and class F in 11 patients (2.9%). The overall success rate of IT gentamicin injection in treating MD was 87.5% (95% CI, 81.1%-92.0%; p< 0.001). Based on preestablished criteria for substantial vertigo control, meta-analysis demonstrated that IT gentamicin injection effectively treated MD. The treatment effect is summarized in Figure 2.Among 14 articles, mean (SD) pretreatment and posttreatment data were provided on PTA results in 11 studies,7-11,13,14,16-19 on WDS in 6 studies,7,10,14,16,17,19 and on caloric test results in no study. The weighted mean PTA results changed from 58.5 (14.2) dB to 54.9 (18.5) dB following IT gentamicin injection (overall P=.09). The weighted mean WDS changed from 50.4% (30.4%) to 52.2% (32.4%) following IT gentamicin injection (overall p=0.07). The mean hearing variable changes of 2.8 (1.7) dB (range, 0.9-6.0 dB) and 2.9% (1.6%) (range, 3.5%-11.4%) were not clinically significant. Eleven of 599 patients (1.8%) experienced total deafness.11,14The improvement in tinnitus (among 77 out of 621 patients 12.4%) was significant (P=.02). The improvement in aural fullness (among 48 out of 623 patients 7.7%) was not significant (p=0.06).Quality-of-life questionnaires were used in 7 studies.6,7,10,13,14,16,17 The efficacy of IT gentamicin injection for the treatment of MD as assessed by subjective symptom survey results shows remarkable improvement (Figure 3). Detailed results of the use of IT gentamicin injection in treating MD are given in Table 2.Effect of Treatment ProtocolThe pooled success rates of fixed protocols (87.5% range, 56.3%-97.5%) vs titration protocols (88.2% range, 71.0%-97.1%) were not significantly different (p=0.94). Patients treated with titration protocols of IT gentamicin injection did not have higher rates of vertigo control (Figure 4).The weighted mean PTA changes of fixed protocols (2.2 2.0 dB range, 0.9-8.5 dB) vs titration protocols (4.1 2.9 dB range, 1.0-6.0 dB) were not significantly different (p=0.09). Patients treated with titration protocols of IT gentamicin injection did not have greater deterioration in PTA results (Figure 5).The weighted mean WDS changes of fixed protocols (0% range, 3.5%-11.4%) vs titration protocols (4% range, 4%) were not significantly different (p=0.07). Patients treated with titration protocols of IT gentamicin injection did not have greater reduction in WDS (Figure 6).The pooled tinnitus improvement rates of fixed protocols (15.5% range, 13.3%-18.2%) vs titration protocols (43.5% range, 0.4%-47.9%) were significantly different (p=0.02). Patients treated with titration protocols of IT gentamicin injection had higher tinnitus improvement rates (Figure 7a).The pooled aural fullness improvement rates of fixed protocols (38.5% range, 36.4%-40.0%) vs titration protocols (39.1% range, 0.4%-45.6%) were not significantly different (p=0.06). Patients treated with titration protocols of IT gentamicin injection did not have higher aural fullness improvement rates (Figure 7a).The pooled rates of worse hearing for fixed protocols (17% range, 6.7%-27.3%) vs titration protocols (17.2% range, 0.4%-38.7%) were not significantly different (p=0.81). Patients treated