TY - JOUR
AU - Antonelli, Patrick, J
AB - Abstract Background This study examined whether the use of quinolone ear drops increased the risk of perforation with intact tympanic membranes and acute otitis externa (AOE). Methods This was a retrospective cohort study using Medicaid clinical encounter and pharmacy billing records from 1999 through 2010. Children and adults had to have 24 months continuous enrollment in Medicaid prior to the first antibiotic ear drop dispensing (index date), and they had to maintain their enrollment for at least 18 months thereafter. Included ear drops were ofloxacin, ciprofloxacin plus hydrocortisone, ciprofloxacin plus dexamethasone, and neomycin plus hydrocortisone. Tympanic membrane perforation (TMP) was identified as 2 inpatient or outpatient encounters associated with TMP diagnosis at least 30 days apart. A Cox regression model adjusting for patient demographics, calendar year, and the number of ear drop prescriptions was used to compare TMP risk between quinolone and neomycin-exposed patients. Results A total of 94 333 patients entered the study cohort. Use of quinolone ear drops was associated with increased risk for TMP compared with neomycin plus hydrocortisone, with an adjusted hazard ratio of 2.26 (95% confidence interval [CI], 1.34–3.83). Adjusted hazard ratios were 2.53 (95% CI, 1.27–5.05) for ofloxacin, 2.24 (95% CI, 1.03–4.85) for ciprofloxacin plus hydrocortisone, and 2.30 (95% CI, 1.09–4.87) for ciprofloxacin plus dexamethasone. Sensitivity analyses were consistent with the primary analysis. Conclusions Use of quinolone ear drops to treat AOE is associated with a previously unreported increased risk of developing TMPs. Selection of otic preparations to treat self-limited conditions with intact tympanic membranes should consider TMP risk. quinolone, acute otitis externa, tympanic membrane perforation; antibiotic The literature is replete with studies linking quinolones to soft tissue damage, such as tendon rupture [1, 2], retinal detachment [3], and aortic aneurysm or dissection [4–7]. Topical delivery of quinolones, which may be as high as 6000 µg/mL [8], may result in tissue exposure 1000 times that of systemic administration, raising concern about soft tissue damage with otic and ophthalmic preparations [9]. Topical application of quinolones to the eye has been linked with increased risk of corneal perforation [10]. We found in animal experiments that quinolone ear drops interfered with the healing of perforated tympanic membranes (TMs) [11]. In a large population-based study, we recently reported a 2-fold increased risk of persistent tympanic membrane perforation (TMP) after tympanostomy tube (TT) placement and otic exposure to quinolones compared to neomycin plus hydrocortisone (HC) [12]. Persistent TMPs frequently require surgical repair to prevent or treat otitis media [13–15] and ameliorate hearing loss [16], which may not improve with surgery [17]. As quinolone ear drops may lead to loss of TM collagen, causing permanent TMPs after TT, concerns are raised that similar effects could impact intact TMs. We recently discovered that quinolone ear drops can cause TMPs in intact, healthy rat TMs [18], but no study to date has examined whether the tissue-damaging effects of otic quinolones can lead to perforation of intact TMs in humans. Using the MarketScan commercial claims data, which provide a representative sample of privately insured patients in the United States, we estimated that 2.6% of beneficiaries received at least 1 prescription for quinolone topical treatment in 2016. Otic quinolone exposure is, thus, common. To evaluate the impact of topical quinolones on intact human TMs, we examined treatment outcomes of acute otitis externa (AOE), an uncomplicated ear infection that is not known to cause eardrum perforations [19, 20], resulting in approximately 2.4 million US healthcare visits annually [21, 22]. Ototopical antibiotics, such as quinolone ear drops, are the recommended therapy for AOE [19]. Given the widespread use of otic quinolones, high incidence of AOE, and growing concerns about the adverse effects of quinolones on soft tissues, as well as the clinical significance of TMPs, we aimed to investigate whether quinolone ear drop use increases the risk of TMPs with uncomplicated AOE. METHODS Data Source and Study Population The study cohort was established from Medicaid Analytical eXtract files for the period 1999–2010, made available for research by the Centers for Medicare and Medicaid Services (see Supplementary Section I for details on states included in the analysis) [23]. This timeframe was used to ensure a balance of exposure groups, as quinolone use has risen in recent years [24]. The dataset provides information on patients’ Medicaid enrollment and medical inpatient/outpatient encounters with detail on diagnoses and procedures, as well as outpatient-filled prescription drugs. Patients aged <65 years entered the cohort at the first dispensing (index date) of ear drops within 3 days following their first diagnosis of AOE. We chose this window to ensure that the ear drops were related to AOE treatment [25]. Identification of AOE was through outpatient encounter records with International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes 380.10 and 380.12. Participants had to have at least 24 months of continuous Medicaid enrollment before the index date to ensure full capture of exclusionary diagnoses and they had to maintain their enrollment status for at least 18 months thereafter. To focus the analysis on potential risks of topical treatments, we excluded patients who had systemic neomycin or quinolone treatment within 3 days after AOE diagnosis. To reduce the chance for existing tympanic perforation at index date, we excluded patients with any inner, middle, or external ear surgery, otitis media, Eustachian tube dysfunction, mastoiditis, malignant otitis externa, or diagnosis of tympanic perforation during the 24-month look-back period. We also excluded patients with disease states that might predispose to TMP based on inpatient or outpatient encounter diagnoses during the baseline or follow-up (see detail in Supplementary Table 1). Participants were followed until the end of Medicaid enrollment, which was per study inclusion criteria a minimum of 18 months, or until they switched between the 2 ear drop classes or had a diagnosis of otitis media, malignant otitis externa, Eustachian tube dysfunction, mastoiditis, or the study outcome, whichever occurred first. For patients who were not censored, we terminated follow-up at a maximum of 24 months or the end of 2010, whichever occurred first. Study Outcomes Occurrence of TMP was defined as the presence of 2 inpatient or outpatient encounters at least 30 days apart with a diagnosis of TMP using ICD-9-CM codes 384.20, 384.21, or 384.23–384.25. In a sensitivity analysis, the outcome was defined as 2 inpatient or outpatient TMP diagnoses at least 6 months apart or 1 encounter claim for eardrum repair surgery (tympanoplasty) using the Common Procedural Terminology codes 69631–69633, 69635–69637, or 69641–69646. Finally, to capture adverse sequalae of TMP, we followed all TMP cases with at least 1 year of continuous enrollment to ascertain medical encounters indicating tympanoplasty, conductive or mixed hearing loss (ICD-9-CM codes 389.0x and 389.2x), or suppurative otitis media (ICD-9-CM code 382.xx). Exposure Definition Patients were considered exposed to antibiotic ear drops and maintained their exposure status throughout the follow-up period if they had a pharmacy claim for quinolones (ofloxacin, ciprofloxacin + HC, or ciprofloxacin + dexamethasone) or neomycin + HC ear drops within 3 days of AOE diagnosis. Otic neomycin + HC preparations were considered the reference group because they are the only alternative topical antimicrobial agent commonly used in AOE. Because we expected irreversible effects on collagenous tissue and to avoid exposure misclassification, we terminated follow-up when patients received a subsequent ear drop from the comparator class (ie, switched from quinolone to neomycin ear drops or vice versa). Covariates Covariates included in the comparison of quinolone vs neomycin + HC risk for tympanic perforations were age, sex, race/ethnicity, and calendar year at study entry. Because multiple courses of ear drops may be used during follow-up, we also included the total number of ear drop dispensings as a time-varying covariate. Statistical Analyses We compared the distributions of demographic and clinical characteristics between the 2 exposure groups and calculated the incidence rate of TMP. A Cox regression model was used to estimate the hazard ratios (HRs) of TMP comparing exposure to quinolones and neomycin + HC and their corresponding 95% confidence intervals (CIs), adjusted for covariates. The TMP risk for different quinolone-containing ear drop formulations was compared to neomycin + HC by using a categorical exposure variable that identified the initial antibiotic dispensation as neomycin + HC, ofloxacin, ciprofloxacin + HC, or ciprofloxacin plus dexamethasone. We further examined the role of age as an effect modifier by stratifying the analysis to patients <18 years and ≥18 years old, because there is evidence linking age to TMPs and clinical action involving surgical repair (ie, tympanoplasty) [26, 27], which may lead to delays in identification of TMPs in children using claims data. We conducted a sensitivity analysis to evaluate the effect of potential outcome misclassification by restricting the analysis to persistent TMP, defined as 2 inpatient or outpatient TMP diagnoses at least 6 months apart or 1 encounter claim for tympanoplasty. Data management and analyses were performed with SAS version 9.4 software (SAS Institute, Cary, North Carolina). The study was approved by the privacy and institutional review boards from the Centers for Medicare and Medicaid Services and our institution. RESULTS We identified 229 262 patients initiated on study ear drops within 3 days after the AOE diagnosis within our age criteria. After applying all exclusion criteria, 43 653 were initiated on quinolones with or without corticosteroids and 50 680 were initiated on neomycin + HC (Figure 1). Compared to patients exposed to neomycin + HC, those exposed to quinolones were more likely to be children (Table 1). The use of quinolone ear drops increased in more recent years, while neomycin + HC ear drops use slightly decreased after 2003. Exposure groups had the same average number of ear drops dispensed during follow-up. Table 1. Patient Characteristics, by Antibiotic Exposure Characteristic Quinolones (n = 43 653) Neomycin (n = 50 680) Age  Children (<18 y) 35 851 (82.13) 32 753 (64.63)  Adults (≥18 y) 7802 (17.87) 17 927 (35.37) Sex  Female 24 870 (56.97) 30 271 (59.73)  Male 18 783 (43.03) 20 409 (40.27) Race/ethnicity  White 24 868 (56.97) 27 454 (54.17)  Other 18 785 (43.03) 23 226 (45.83) Calendar year  2001 3337 (7.64) 6643 (13.11)  2002 4560 (10.45) 6655 (13.13)  2003 5223 (11.96) 6841 (13.50)  2004 5737 (13.14) 6403 (12.63)  2005 5767 (13.21) 6067 (11.97)  2006 5265 (12.06) 4939 (9.75)  2007 5105 (11.69) 4319 (8.52)  2008 4835 (11.08) 4128 (8.15)  2009 3824 (8.76) 4685 (9.24) No. of antibiotic ear drops dispensings during follow-up, mean (SD) 1.16 (0.50) 1.15 (0.61) Follow-up time, d, mean (SD) 580 (236.2) 572 (247.9) Characteristic Quinolones (n = 43 653) Neomycin (n = 50 680) Age  Children (<18 y) 35 851 (82.13) 32 753 (64.63)  Adults (≥18 y) 7802 (17.87) 17 927 (35.37) Sex  Female 24 870 (56.97) 30 271 (59.73)  Male 18 783 (43.03) 20 409 (40.27) Race/ethnicity  White 24 868 (56.97) 27 454 (54.17)  Other 18 785 (43.03) 23 226 (45.83) Calendar year  2001 3337 (7.64) 6643 (13.11)  2002 4560 (10.45) 6655 (13.13)  2003 5223 (11.96) 6841 (13.50)  2004 5737 (13.14) 6403 (12.63)  2005 5767 (13.21) 6067 (11.97)  2006 5265 (12.06) 4939 (9.75)  2007 5105 (11.69) 4319 (8.52)  2008 4835 (11.08) 4128 (8.15)  2009 3824 (8.76) 4685 (9.24) No. of antibiotic ear drops dispensings during follow-up, mean (SD) 1.16 (0.50) 1.15 (0.61) Follow-up time, d, mean (SD) 580 (236.2) 572 (247.9) Data are presented as no. (%) unless otherwise indicated. Abbreviation: SD, standard deviation. Open in new tab Table 1. Patient Characteristics, by Antibiotic Exposure Characteristic Quinolones (n = 43 653) Neomycin (n = 50 680) Age  Children (<18 y) 35 851 (82.13) 32 753 (64.63)  Adults (≥18 y) 7802 (17.87) 17 927 (35.37) Sex  Female 24 870 (56.97) 30 271 (59.73)  Male 18 783 (43.03) 20 409 (40.27) Race/ethnicity  White 24 868 (56.97) 27 454 (54.17)  Other 18 785 (43.03) 23 226 (45.83) Calendar year  2001 3337 (7.64) 6643 (13.11)  2002 4560 (10.45) 6655 (13.13)  2003 5223 (11.96) 6841 (13.50)  2004 5737 (13.14) 6403 (12.63)  2005 5767 (13.21) 6067 (11.97)  2006 5265 (12.06) 4939 (9.75)  2007 5105 (11.69) 4319 (8.52)  2008 4835 (11.08) 4128 (8.15)  2009 3824 (8.76) 4685 (9.24) No. of antibiotic ear drops dispensings during follow-up, mean (SD) 1.16 (0.50) 1.15 (0.61) Follow-up time, d, mean (SD) 580 (236.2) 572 (247.9) Characteristic Quinolones (n = 43 653) Neomycin (n = 50 680) Age  Children (<18 y) 35 851 (82.13) 32 753 (64.63)  Adults (≥18 y) 7802 (17.87) 17 927 (35.37) Sex  Female 24 870 (56.97) 30 271 (59.73)  Male 18 783 (43.03) 20 409 (40.27) Race/ethnicity  White 24 868 (56.97) 27 454 (54.17)  Other 18 785 (43.03) 23 226 (45.83) Calendar year  2001 3337 (7.64) 6643 (13.11)  2002 4560 (10.45) 6655 (13.13)  2003 5223 (11.96) 6841 (13.50)  2004 5737 (13.14) 6403 (12.63)  2005 5767 (13.21) 6067 (11.97)  2006 5265 (12.06) 4939 (9.75)  2007 5105 (11.69) 4319 (8.52)  2008 4835 (11.08) 4128 (8.15)  2009 3824 (8.76) 4685 (9.24) No. of antibiotic ear drops dispensings during follow-up, mean (SD) 1.16 (0.50) 1.15 (0.61) Follow-up time, d, mean (SD) 580 (236.2) 572 (247.9) Data are presented as no. (%) unless otherwise indicated. Abbreviation: SD, standard deviation. Open in new tab Figure 1. Open in new tabDownload slide Flow diagram for study cohort inclusion. Abbreviation: AOE, acute otitis externa. Figure 1. Open in new tabDownload slide Flow diagram for study cohort inclusion. Abbreviation: AOE, acute otitis externa. Overall, 38 cases of TMPs were diagnosed in patients who were exposed to quinolone ear drops for AOE during follow-up (5.5 TMPs per 10 000 patient-years), as compared with 25 cases of TMPs in otic neomycin–exposed patients (3.1 TMPs per 10 000 patient-years). The unadjusted HR for TMP associated with any otic quinolone preparation use in comparison to neomycin + HC ear drops use was 1.73 (95% CI, 1.05–2.87). When we adjusted for demographics and other covariates, the HRs for TMP were 2.26 (95% CI, 1.34–3.83; Table 2) with the use of any otic quinolone preparation. Each year increase in patient age was associated with a 3% increase in the hazard of TMP (95% CI, 2%–5%). The frequency of ear drop prescriptions was also associated with higher risk of TMPs (HR, 1.32 [95% CI, 1.21–1.45]). When comparing individual quinolone preparations against neomycin + HC, all showed an increased HR for TMPs: ofloxacin, 2.53 (95% CI, 1.27–5.05); ciprofloxacin + HC, 2.24 (95% CI, 1.03–4.85); and ciprofloxacin plus dexamethasone, 2.30 (95% CI, 1.09–4.87) (Table 3). The risk of TMP was comparable among children and adults with HRs of 2.19 (95% CI, 1.01– 4.73) and 2.55 (95% CI, 1.25–5.18), respectively (Table 4). The primary results were corroborated by analyses using the more restrictive outcome definition with a 4.59-fold increase in the hazard for persistent TMP (95% CI, 2.07–10.17) associated with exposure to any quinolone (Supplementary Tables 2 and 3). Among 55 TMP cases with 1-year continuous plan enrollment, we found 39 cases with tympanoplasty, suppurative otitis media, or conductive or mixed hearing loss (Table 4, see footnotes). Table 2. Adjusted Risk of Tympanic Membrane Perforation for Patients With Acute Otitis Externa Exposed to Quinolone Versus Neomycin Plus Hydrocortisone Ear Drops Characteristics Hazard Ratio (95% CI) Exposure  Neomycin Reference …  Quinolones 2.26 (1.34–3.83) Age 1.03 (1.02–1.05) Sex  Female Reference …  Male 1.52 (.92–2.50) Race/ethnicity  White Reference …  Other 1.10 (.67–1.81) Calendar year  2001 Reference …  2002 0.56 (.22–1.46)  2003 0.38 (.13–1.09)  2004 0.58 (.23–1.45)  2005 0.49 (.19–1.30)  2006 0.60 (.23–1.54)  2007 0.54 (.20–1.46)  2008 0.68 (.26–1.75)  2009 0.57 (.20–1.66) No. of ear drop dispensings during follow-up 1.32 (1.21–1.45) Characteristics Hazard Ratio (95% CI) Exposure  Neomycin Reference …  Quinolones 2.26 (1.34–3.83) Age 1.03 (1.02–1.05) Sex  Female Reference …  Male 1.52 (.92–2.50) Race/ethnicity  White Reference …  Other 1.10 (.67–1.81) Calendar year  2001 Reference …  2002 0.56 (.22–1.46)  2003 0.38 (.13–1.09)  2004 0.58 (.23–1.45)  2005 0.49 (.19–1.30)  2006 0.60 (.23–1.54)  2007 0.54 (.20–1.46)  2008 0.68 (.26–1.75)  2009 0.57 (.20–1.66) No. of ear drop dispensings during follow-up 1.32 (1.21–1.45) Abbreviation: CI, confidence interval. Open in new tab Table 2. Adjusted Risk of Tympanic Membrane Perforation for Patients With Acute Otitis Externa Exposed to Quinolone Versus Neomycin Plus Hydrocortisone Ear Drops Characteristics Hazard Ratio (95% CI) Exposure  Neomycin Reference …  Quinolones 2.26 (1.34–3.83) Age 1.03 (1.02–1.05) Sex  Female Reference …  Male 1.52 (.92–2.50) Race/ethnicity  White Reference …  Other 1.10 (.67–1.81) Calendar year  2001 Reference …  2002 0.56 (.22–1.46)  2003 0.38 (.13–1.09)  2004 0.58 (.23–1.45)  2005 0.49 (.19–1.30)  2006 0.60 (.23–1.54)  2007 0.54 (.20–1.46)  2008 0.68 (.26–1.75)  2009 0.57 (.20–1.66) No. of ear drop dispensings during follow-up 1.32 (1.21–1.45) Characteristics Hazard Ratio (95% CI) Exposure  Neomycin Reference …  Quinolones 2.26 (1.34–3.83) Age 1.03 (1.02–1.05) Sex  Female Reference …  Male 1.52 (.92–2.50) Race/ethnicity  White Reference …  Other 1.10 (.67–1.81) Calendar year  2001 Reference …  2002 0.56 (.22–1.46)  2003 0.38 (.13–1.09)  2004 0.58 (.23–1.45)  2005 0.49 (.19–1.30)  2006 0.60 (.23–1.54)  2007 0.54 (.20–1.46)  2008 0.68 (.26–1.75)  2009 0.57 (.20–1.66) No. of ear drop dispensings during follow-up 1.32 (1.21–1.45) Abbreviation: CI, confidence interval. Open in new tab Table 3. Association Between Quinolones and Tympanic Membrane Perforations, Stratified by Ear Drops Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Neomycin 50 680 25 79 544 3.1 Reference Quinolone 43 653 38 69 373 5.5 1.73 (1.05–2.87) 2.26 (1.34–3.83) Ofloxacin 13 674 13 21 562 6.0 1.91 (.98–3.73) 2.53 (1.27–5.05) Cipro + HC 11 960 10 19 012 5.3 1.67 (.80–3.47) 2.24 (1.03–4.85) Cipro + Dex 18 019 15 28 799 5.2 1.65 (.87–3.12) 2.30 (1.09–4.87) Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Neomycin 50 680 25 79 544 3.1 Reference Quinolone 43 653 38 69 373 5.5 1.73 (1.05–2.87) 2.26 (1.34–3.83) Ofloxacin 13 674 13 21 562 6.0 1.91 (.98–3.73) 2.53 (1.27–5.05) Cipro + HC 11 960 10 19 012 5.3 1.67 (.80–3.47) 2.24 (1.03–4.85) Cipro + Dex 18 019 15 28 799 5.2 1.65 (.87–3.12) 2.30 (1.09–4.87) Abbreviations: CI, confidence interval; Cipro, ciprofloxacin; dex, dexamethasone; HC, hydrocortisone; HR, hazard ratio; PY, patient-years. aAdjusted for age, sex, race, calendar year of acute otitis externa diagnosis, and number of ear drop prescriptions. Open in new tab Table 3. Association Between Quinolones and Tympanic Membrane Perforations, Stratified by Ear Drops Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Neomycin 50 680 25 79 544 3.1 Reference Quinolone 43 653 38 69 373 5.5 1.73 (1.05–2.87) 2.26 (1.34–3.83) Ofloxacin 13 674 13 21 562 6.0 1.91 (.98–3.73) 2.53 (1.27–5.05) Cipro + HC 11 960 10 19 012 5.3 1.67 (.80–3.47) 2.24 (1.03–4.85) Cipro + Dex 18 019 15 28 799 5.2 1.65 (.87–3.12) 2.30 (1.09–4.87) Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Neomycin 50 680 25 79 544 3.1 Reference Quinolone 43 653 38 69 373 5.5 1.73 (1.05–2.87) 2.26 (1.34–3.83) Ofloxacin 13 674 13 21 562 6.0 1.91 (.98–3.73) 2.53 (1.27–5.05) Cipro + HC 11 960 10 19 012 5.3 1.67 (.80–3.47) 2.24 (1.03–4.85) Cipro + Dex 18 019 15 28 799 5.2 1.65 (.87–3.12) 2.30 (1.09–4.87) Abbreviations: CI, confidence interval; Cipro, ciprofloxacin; dex, dexamethasone; HC, hydrocortisone; HR, hazard ratio; PY, patient-years. aAdjusted for age, sex, race, calendar year of acute otitis externa diagnosis, and number of ear drop prescriptions. Open in new tab Table 4. Association Between Quinolone Ear Drops and Tympanic Membrane Perforations, Stratified by Age Age Group/Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Children (<18 y)b  Neomycin 32 753 10 50 892 2.0 Reference  Quinolones 35 851 22 57 268 3.8 1.96 (.93–4.14) 2.19 (1.01–4.73) Adults (≥18 y)c  Neomycin 17 927 15 28 652 5.2 Reference  Quinolones 7802 16 12 104 13.2 2.48 (1.22–5.00) 2.55 (1.25–5.18) Age Group/Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Children (<18 y)b  Neomycin 32 753 10 50 892 2.0 Reference  Quinolones 35 851 22 57 268 3.8 1.96 (.93–4.14) 2.19 (1.01–4.73) Adults (≥18 y)c  Neomycin 17 927 15 28 652 5.2 Reference  Quinolones 7802 16 12 104 13.2 2.48 (1.22–5.00) 2.55 (1.25–5.18) Abbreviations: CI, confidence interval; HR, hazard ratio; PY, patient-years. aAdjusted for sex, race, calendar year of acute otitis externa diagnosis, and number of ear drop prescriptions. bTwenty-nine children with tympanic membrane perforation (TMP) had at least 1-year continuous enrollment after TMP occurred. Among TMP cases in children within the neomycin group, 1 patient had tympanoplasty and 3 patients had suppurative otitis media within 1 year after TMP. Among TMP cases in children within the quinolone group, 7 patients had tympanoplasty, 7 patients had suppurative otitis media, and 3 patients had new onset of conductive or mixed hearing loss within 1 year after TMP. c Twenty-six adults with TMP had at least 1-year continuous enrollment after TMP occurred. Among TMP cases in adults within the neomycin group, 2 patients had tympanoplasty, 5 patients had suppurative otitis media, and 4 patients had new onset of conductive or mixed hearing loss within 1 year after TMP. Among TMP cases in adults within the quinolone group, 3 patients had tympanoplasty, 5 patients had suppurative otitis media, and 5 patients had new onset of conductive or mixed hearing loss within 1 year after TMP. Open in new tab Table 4. Association Between Quinolone Ear Drops and Tympanic Membrane Perforations, Stratified by Age Age Group/Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Children (<18 y)b  Neomycin 32 753 10 50 892 2.0 Reference  Quinolones 35 851 22 57 268 3.8 1.96 (.93–4.14) 2.19 (1.01–4.73) Adults (≥18 y)c  Neomycin 17 927 15 28 652 5.2 Reference  Quinolones 7802 16 12 104 13.2 2.48 (1.22–5.00) 2.55 (1.25–5.18) Age Group/Exposure Group No. of Patients No. of Cases Person-time, y Incidence per 10 000 PY Unadjusted HR (95% CI) Adjusted HRa (95% CI) Children (<18 y)b  Neomycin 32 753 10 50 892 2.0 Reference  Quinolones 35 851 22 57 268 3.8 1.96 (.93–4.14) 2.19 (1.01–4.73) Adults (≥18 y)c  Neomycin 17 927 15 28 652 5.2 Reference  Quinolones 7802 16 12 104 13.2 2.48 (1.22–5.00) 2.55 (1.25–5.18) Abbreviations: CI, confidence interval; HR, hazard ratio; PY, patient-years. aAdjusted for sex, race, calendar year of acute otitis externa diagnosis, and number of ear drop prescriptions. bTwenty-nine children with tympanic membrane perforation (TMP) had at least 1-year continuous enrollment after TMP occurred. Among TMP cases in children within the neomycin group, 1 patient had tympanoplasty and 3 patients had suppurative otitis media within 1 year after TMP. Among TMP cases in children within the quinolone group, 7 patients had tympanoplasty, 7 patients had suppurative otitis media, and 3 patients had new onset of conductive or mixed hearing loss within 1 year after TMP. c Twenty-six adults with TMP had at least 1-year continuous enrollment after TMP occurred. Among TMP cases in adults within the neomycin group, 2 patients had tympanoplasty, 5 patients had suppurative otitis media, and 4 patients had new onset of conductive or mixed hearing loss within 1 year after TMP. Among TMP cases in adults within the quinolone group, 3 patients had tympanoplasty, 5 patients had suppurative otitis media, and 5 patients had new onset of conductive or mixed hearing loss within 1 year after TMP. Open in new tab Discussion In this cohort of 94 333 AOE patients enrolled in the Medicaid programs of 29 states, we found the use of quinolone ear drops was associated with a higher risk of TMP compared to neomycin ear drops. This was true for all studied quinolone otic preparations. Using a more stringent definition of persistent TMP with repeated diagnoses during medical encounters at least 6 months apart or surgical repair, as well as subgroup analyses in children and adults further supported these findings. Several biological mechanisms have been proposed by which quinolone ear drops might exert their effects on the TM. For example, quinolones increase apoptosis of extracellular matrix–producing cells [28], increase degradation of extracellular matrix via enhanced matrix metalloproteinase activity [29], and decrease fibroblast viability [30, 31] and levels of intact collagen [31]. High concentrations of quinolones in ear drops may reach the TM in the treatment of AOE, potentially compromising levels of intact collagen and TM fibroblast viability. These findings have been borne out in TM fibroblast tissue culture [31, 32] and an animal model of TMP [33]. Pertinent to this new study in patients with AOE, we have recently expanded our previous animal model and found that otic quinolone exposure for 10 days causes TMPs in intact rat TMs [18]. Our findings are also consistent with our earlier study in humans where exposure to quinolone ear drops was associated with increased risk of perforations requiring tympanoplasty among children who received quinolone ear drops after insertion of TTs [12]. A similar increased rate of corneal perforation has been reported with ophthalmic application of quinolones [10]. Previous reports from both human and animal studies have demonstrated synergism between quinolones and steroids and risk of soft tissue complications. This was first reported for Achilles tendon ruptures with systemic administration. In our previous study investigating risk of permanent TMPs after insertion of tympanic tubes, we found a trend toward an increased risk for quinolone/steroid combinations products compared with quinolone monotherapy, though not statistically significant. Unfortunately, similar sample size constraints (with only a small number of cases in each exposure group) preclude formal conclusions in this study on spontaneous perforation following treatment of AOE. It should be noted that both our present study and our prior study of TMPs after TTs employed neomycin + HC as the comparator group. While use of an active comparator achieved more balanced exposure groups, neomycin preparations have long been known to have adverse effects on the TM [34, 35]. Exposure of TM fibroblasts to neomycin + HC in tissue culture leads to significant cytotoxicity and loss of collagen [31]. Thus, the net effect of quinolone ear drops on TMP may be higher than the relative increase that our findings suggest. The crude incidence rate of TMP reported among children and adults with AOE (5.5 TMPs per 10 000 patient-years over follow-up after quinolone exposure, or considering an average follow-up of over 1.5 years, about 1 case per 1000 treated AOE patients) was roughly a third of the incidence rate we found with quinolone ear drops after TT: 17 TMPs per 10 000 person-years. Even though this TMP risk specific to AOE may appear small, this finding is nevertheless clinically remarkable. Children receiving TTs have Eustachian tube dysfunction, middle ear inflammation or infection, and an iatrogenic injury to the TM, which are all known risk factors for TMP [36]. These factors play no role in AOE and, although there is little published on the natural history of AOE, no association has been made between AOE and the development of TMP [19]. Given that we excluded patients with risk factors for TMP at study entry and censored patients when risk factors emerged during follow-up, the reported TMP incidence rate in a population without risk factors is noteworthy. The higher proportion of children in the quinolone-treated group might raise concern about a higher risk for TMP in the quinolone group, as a variety of risk factors, such as Eustachian tube dysfunction, otitis media, TTs, are more common in children. However, we employed stringent exclusion criteria based on a 24-month look-back period to minimize the presence of other risk factors and the potential for TMP unrelated to the otic quinolone exposure. We furthermore found similar rates of TMP following otic quinolone treatment in adults. In fact, the rate of TMP after otic quinolone exposure increased with advancing patient age. While neomycin poses its own problems in patients with TTs due to its inner ear toxicity [37], our results may change risk-benefit assessment regarding treatment approaches for AOE. In the usual situation with an intact TM, the risk of TMP with ototopical quinolones might outweigh their benefit, favoring alternative therapies for the initial treatment. When AOE patients have a known or suspected TMP, quinolones would likely remain the favored initial treatment [19] because of the risk of sensorineural hearing loss associated with aminoglycoside preparations, such as neomycin [37]. The finding that ototopical quinolone preparations can adversely affect intact TMs calls for more cautious utilization of these products, especially when quinolones may be considered for off-label indications, such as posttraumatic TMPs and cerumen impaction. Several study strengths and limitations deserve discussion. First, the study cohort was drawn from a nationwide sample of publicly insured beneficiaries, and the large size of exposed groups allowed for additional stratified evaluation in both pediatric and adult populations with good precision. This study was based on Medicaid beneficiaries; therefore, results might not be generalizable to commercially insured patients. Second, the richness of the longitudinal data facilitated exclusion of patients with a broad range of TMP risk factors in a 2-year look-back period and therefore reduced the risk of confounding. Use of neomycin + HC as an active comparator further alleviated concerns about confounding by indication. Those precautions notwithstanding, given the nature of the data, the severity of the AOE infection may not be completely reflected in the database. Specifically, even though we strictly removed patients whose diagnoses might suggest presence of TMP at study entry (or risk factors thereof), patients with TMP that was not documented in clinical encounters could have been channeled to quinolone ear drops because of the known inner ear toxicity of neomycin. The low probability for preexisting TMP, considering the generally benign nature of AOE and our stringent exclusion criteria, and the biological plausibility of the observed quinolone effect may alleviate these concerns. Third, use of claims-based algorithms to identify TMPs has not been validated (eg, through records review). However, our identification of the study outcome and sensitivity analyses minimized the potential for ascertainment bias. Many persons live with TMPs that are asymptomatic (eg, no hearing loss or otorrhea) [38]. Such TMPs may fail to be included among encounter diagnoses or the diagnosis may fall below the top 2 diagnoses, which are captured in Medicaid data for a clinic visit. However, while reduced sensitivity in capturing TMPs might affect the presented incidence estimates, it would not change relative comparisons between exposure groups and, thus, have no effect on the HRs. Because there was no evidence suggesting any detrimental effects of otic quinolone preparations, it is unlikely the diagnoses were made differentially between exposure groups. We also performed a sensitivity analysis by assuming that 2 diagnoses of TMPs 6 months apart or a tympanoplasty surgery procedure would establish satisfactory validity of persistent TMPs [39]. Requiring a more stringent period between 2 TMP diagnoses would be expected to underestimate the occurrence of TMPs overall, but reduce noise through invalid diagnoses and thus focus the evaluation of exposure effects on true persistent TMPs. While CIs became wide, we noted an increase in the quinolone effect to a >4-fold risk for TMP when compared to neomycin exposure. Though not likely, ICD-9 TMP codes, procedure codes for tympanoplasties, and pharmacy claims for ear drops are not ear specific, and thus patients could have a TMP that did not occur in the ear with ear drop exposure. Such measurement error would be expected to be similar between quinolone and neomycin groups and bias the HRs toward a null effect. Finally, due to limitations inherent in administrative databases, exposure misclassification cannot be ruled out. For example, exposure to ear drops was based on filled prescriptions rather than actual use of medications, but it would be expected that the acuity of AOE resulted at least initially in high utilization rates. If not used, we would expect that such misclassification would likely affect both exposure groups, which would bias the study toward a null effect as well. Conclusions Our results suggest that use of quinolone ear drops for AOE is associated with an increased risk of TMPs. The risks and benefits of otic quinolones should be carefully considered prior to the treatment of AOE. Therapy duration, volume dispensed, and refills should be limited to what is necessary to ensure clinical cure. Patients should be counseled on risk of TMP and monitored for TMPs in follow-up visits. Supplementary Data Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author. Notes Author contributions. All authors made substantial contributions to study conception. A. G. W. was responsible for data acquisition. X. W., A. G. W., and P. J. A. had full access to all the data in this study, and X. W. takes full responsibility as a guarantor for the integrity of the data and the accuracy of the data analysis. X. W., A. G. W., A. A., and P. J. A. contributed to study design and interpretation of data. X. W. and A. A. drafted the article and X. W., A. G. W., A. A., and P. J. A. provided critical revisions for important intellectual content and approved the final version. Acknowledgments. The authors thank Carl Henriksen for helping with the acquisition of data. Potential conflicts of interest. P. J. A. has received research support from Alcon Laboratories and Otonomy and remuneration for advisory board service from Otonomy. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of potential conflicts of interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. Presented in part: 34th International Conference on Pharmacoepidemiology and Therapeutic Risk Management, Prague, Czech Republic, 22–26 August 2018. References 1. van der Linden PD , Sturkenboom MC , Herings RM , Leufkens HM , Rowlands S , Stricker BH . Increased risk of Achilles tendon rupture with quinolone antibacterial use, especially in elderly patients taking oral corticosteroids . Arch Intern Med 2003 ; 163 : 1801 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 2. Corrao G , Zambon A , Bertù L , et al. Evidence of tendinitis provoked by fluoroquinolone treatment: a case-control study . Drug Saf 2006 ; 29 : 889 – 96 . Google Scholar Crossref Search ADS PubMed WorldCat 3. Etminan M , Forooghian F , Brophy JM , Bird ST , Maberley D . Oral fluoroquinolones and the risk of retinal detachment . JAMA 2012 ; 307 : 1414 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 4. Lee CC , Lee MT , Chen YS , et al. Risk of aortic dissection and aortic aneurysm in patients taking oral fluoroquinolone . JAMA Intern Med 2015 ; 175 : 1839 – 47 . Google Scholar Crossref Search ADS PubMed WorldCat 5. Lee CC , Lee MG , Hsieh R , et al. Oral fluoroquinolone and the risk of aortic dissection . J Am Coll Cardiol 2018 ; 72 : 1369 – 78 . Google Scholar Crossref Search ADS PubMed WorldCat 6. Pasternak B , Inghammar M , Svanström H . Fluoroquinolone use and risk of aortic aneurysm and dissection: nationwide cohort study . BMJ 2018 ; 360 : k678 . Google Scholar Crossref Search ADS PubMed WorldCat 7. Daneman N , Lu H , Redelmeier DA . Fluoroquinolones and collagen associated severe adverse events: a longitudinal cohort study . BMJ Open 2015 ; 5 : e010077 . Google Scholar Crossref Search ADS PubMed WorldCat 8. Ciprofloxacin otic suspension (Otiprio) for acute otitis externa . Med Lett Drugs Ther 2018 ; 60 : 135–6 . WorldCat 9. Begg EJ , Robson RA , Saunders DA , et al. The pharmacokinetics of oral fleroxacin and ciprofloxacin in plasma and sputum during acute and chronic dosing . Br J Clin Pharmacol 2000 ; 49 : 32 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 10. Mallari PL , McCarty DJ , Daniell M , Taylor H . Increased incidence of corneal perforation after topical fluoroquinolone treatment for microbial keratitis . Am J Ophthalmol 2001 ; 131 : 131 – 3 . Google Scholar Crossref Search ADS PubMed WorldCat 11. Dirain CO , Kosko B , Antonelli PJ . Effects of common ear drops on tympanic membrane healing in rats . Otolaryngol Head Neck Surg 2018 ; 158 : 917 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat 12. Alrwisan A , Antonelli PJ , Winterstein AG . Quinolone ear drops after tympanostomy tubes and the risk of eardrum perforation: a retrospective cohort study . Clin Infect Dis 2017 ; 64 : 1052 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 13. Chung JH , Lee SH , Woo SY , Kim SW , Cho YS . Prevalence and associated factors of chronic suppurative otitis media: Data from the Korea National Health and Nutrition Examination Survey, 2009-2012 . Laryngoscope 2016 ; 126 : 2351 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 14. Matthew LH. Middle ear, tympanic membrane, perforations: practice essentials, epidemiology, etiology . 2017 . Available at: https://emedicine.medscape.com/article/858684-overview. Accessed 11 May 2019. 15. Adams ME , El-Kashlan HK . Tympanoplasty and ossiculoplasty. In: Flint PW , Haughey BH , Robbins KT , et al. , eds. Cummings otolaryngology—head and neck surgery E-book . Philadelphia, PA: Elsevier , 2015 : 2177 – 87 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC 16. Lerut B , Pfammatter A , Moons J , Linder T . Functional correlations of tympanic membrane perforation size . Otol Neurotol 2012 ; 33 : 379 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 17. Pfammatter A , Novoa E , Linder T . Can myringoplasty close the air-bone gap? Otol Neurotol 2013 ; 34 : 705 – 10 . Google Scholar Crossref Search ADS PubMed WorldCat 18. Dirain C , Karnani DN , Antonelli PJ . Quinolone ear drops cause perforations in intact rat tympanic membranes. In: American Otological Society Annual Meeting , Austin, TX, 3–5 May 2019 . 19. Rosenfeld RM , Schwartz SR , Cannon CR , et al. Clinical practice guideline: acute otitis externa . Otolaryngol Head Neck Surg 2014 ; 150 : S1 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat 20. Rosenfeld RM , Singer M , Wasserman JM , Stinnett SS . Systematic review of topical antimicrobial therapy for acute otitis externa . Otolaryngol Head Neck Surg 2006 ; 134 : S24 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat 21. Piercefield EW , Collier SA , Hlavsa MC , Beach MJ . Estimated burden of acute otitis externa—United States, 2003–2007 (reprinted) . JAMA 2011 ; 306 : 31 – 3 . WorldCat 22. Centers for Disease Control and Prevention . Estimated burden of acute otitis externa—United States, 2003–2007 . MMWR Morb Mortal Wkly Rep 2011 ; 60 : 605 – 9 . PubMed WorldCat 23. Li Y , Zhu Y , Chen C , et al. Internal validation of Medicaid Analytic eXtract (MAX) data capture for comprehensive managed care plan enrollees from 2007 to 2010 . Pharmacoepidemiol Drug Saf 2018 ; 27 : 1067 – 76 . Google Scholar Crossref Search ADS PubMed WorldCat 24. Winterstein AG , Castillo JM , Xu D , Liu W , Antonelli PJ . Ototopical neomycin exposure in children with nonintact tympanic membranes . Laryngoscope 2012 ; 122 : 2529 – 32 . Google Scholar Crossref Search ADS PubMed WorldCat 25. Wang X , Winterstein AG , Li Y , Zhu Y , Antonelli PJ . Use of systemic antibiotics for acute otitis externa: impact of a clinical practice guideline . Otol Neurotol 2018 ; 39 : 1088 – 94 . Google Scholar Crossref Search ADS PubMed WorldCat 26. Orji FT , Agu CC . Determinants of spontaneous healing in traumatic perforations of the tympanic membrane . Clin Otolaryngol 2008 ; 33 : 420 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat 27. Hardman J , Muzaffar J , Nankivell P , Coulson C . Tympanoplasty for chronic tympanic membrane perforation in children: systematic review and meta-analysis . Otol Neurotol 2015 ; 36 : 796 – 804 . Google Scholar Crossref Search ADS PubMed WorldCat 28. Tan Y , Lu K , Deng Y , et al. The effects of levofloxacin on rabbit fibroblast-like synoviocytes in vitro . Toxicol Appl Pharmacol 2012 ; 265 : 175 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 29. Sendzik J , Shakibaei M , Schäfer-Korting M , Stahlmann R . Fluoroquinolones cause changes in extracellular matrix, signalling proteins, metalloproteinases and caspase-3 in cultured human tendon cells . Toxicology 2005 ; 212 : 24 – 36 . Google Scholar Crossref Search ADS PubMed WorldCat 30. Bujor AM , Haines P , Padilla C , et al. Ciprofloxacin has antifibrotic effects in scleroderma fibroblasts via downregulation of Dnmt1 and upregulation of Fli1 . Int J Mol Med 2012 ; 30 : 1473 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 31. Milne-Davies BA , Antonelli PJ , Orobello NC , Dirain CO . Collagen and α-tubulin of mouse tympanic membrane fibroblasts treated with quinolones and aminoglycosides . Otolaryngol Head Neck Surg 2017 ; 156 : 341 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 32. Orobello NC , Dirain CO , Schultz G , Milne-Davies BA , Ng MR , Antonelli PJ . Ciprofloxacin decreases collagen in mouse tympanic membrane fibroblasts . Otolaryngol Head Neck Surg 2016 ; 155 : 127 – 32 . Google Scholar Crossref Search ADS PubMed WorldCat 33. Dirain CO , Kosko B , Antonelli PJ . Effects of common ear drops on tympanic membrane healing in rats . Otolaryngol Head Neck Surg 2018 ; 158 : 917 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat 34. Vassalli L , Harris DM , Gradini R , Applebaum EL . Inflammatory effects of topical antibiotic suspensions containing propylene glycol in chinchilla middle ears . Am J Otolaryngol 1988 ; 9 : 1 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 35. Masaki M , Wright CG , Lee DH , Meyerhoff WL . Effects of otic drops on chinchilla tympanic membrane . Arch Otolaryngol Head Neck Surg 1988 ; 114 : 1007 – 11 . Google Scholar Crossref Search ADS PubMed WorldCat 36. O’Connell Ferster AP , Tanner AM , Karikari K , Roberts C , Wiltz D , Carr MM . Factors related to persisting perforations after ventilation tube insertion . Int J Pediatr Otorhinolaryngol 2016 ; 81 : 29 – 32 . Google Scholar Crossref Search ADS PubMed WorldCat 37. Winterstein AG , Liu W , Xu D , Antonelli PJ . Sensorineural hearing loss associated with neomycin eardrops and nonintact tympanic membranes . Otolaryngol Head Neck Surg 2013 ; 148 : 277 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat 38. Lerut B , Pfammatter A , Moons J , Linder T . Functional correlations of tympanic membrane perforation size . Otol Neurotol 2012 ; 33 : 379 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 39. Alrwisan A , Winterstein AG , Antonelli PJ . Epidemiology of persistent tympanic membrane perforations subsequent to tympanostomy tubes assessed with real world data . Otol Neurotol 2016 ; 37 : 1376 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2019. Published by Oxford University Press for the Infectious Diseases Society of America. All rights reserved. For permissions, e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Risk for Tympanic Membrane Perforation After Quinolone Ear Drops for Acute Otitis Externa
JF - Clinical Infectious Diseases
DO - 10.1093/cid/ciz345
DA - 2020-03-03
UR - https://www.deepdyve.com/lp/oxford-university-press/risk-for-tympanic-membrane-perforation-after-quinolone-ear-drops-for-1bRo2ma0oP
SP - 1103
VL - 70
IS - 6
DP - DeepDyve
ER -