The drug delivery system. The cannula was composed of Micro-Renathane tubing (MRE025; Braintree Scientific Inc, Braintree, Massachusetts) and the plastic portion of a 24-gauge angiocatheter with an injection cap (MX492; Medex Inc, Dublin, Ohio) screwed onto the angiocatheter to allow for serial substance administration.
Auditory thresholds at 2, 8, and 16 kHz (1024 responses, 512 bins, 25 microseconds) for animals treated with ciprofloxacin hydrochloride, 0.3%/dexamethasone, 0.1%, at baseline and 13 and 28 days after the initiation of treatment. SPL indicates sound pressure level. The error bars indicate the standard deviation of the mean.
Auditory thresholds at 2, 8, and 16 kHz (1024 responses, 512 bins, 25 microseconds) for animals treated with ciprofloxacin hydrochloride, 1.0%/dexamethasone, 0.3%, at baseline and 13 and 28 days after the initiation of treatment. SPL indicates sound pressure level. The error bars indicate the standard deviation of the mean.
Auditory thresholds at 2, 8, and 16 kHz (1024 responses, 512 bins, 25 microseconds) for animals treated with ciprofloxacin hydrochloride, 0.3%, at baseline and 13 and 28 days after the initiation of treatment. SPL indicates sound pressure level. The error bars indicate the standard deviation of the mean.
Auditory thresholds at 2, 8, and 16 kHz (1024 responses, 512 bins, 25 microseconds) for animals treated with the vehicle control at baseline and 13 and 28 days after the initiation of treatment. SPL indicates sound pressure level. The error bars indicate the standard deviation of the mean.
Auditory thresholds at 2, 8, and 16 kHz (1024 responses, 512 bins, 25 microseconds) for animals treated with the isotonic sodium chloride solution control at baseline and 13 and 28 days after the initiation of treatment. SPL indicates sound pressure level. The error bars indicate the standard deviation of the mean.
Auditory thresholds at 2, 8, and 16 kHz (1024 responses, 512 bins, 25 microseconds) for animals treated with the neomycin sulfate, 10%, positive control at baseline and 13 and 28 days after the initiation of treatment. SPL indicates sound pressure level. The error bars indicate the standard deviation of the mean.
Lemke LE, McGee DH, Prieskorn DM, Wall GM, Dolan DF, Altschuler RA, Miller JM. Safety of Ciprofloxacin and Dexamethasone in the Guinea Pig Middle Ear. Arch Otolaryngol Head Neck Surg. 2009;135(6):575-580. doi:10.1001/archoto.2009.30
To investigate the ototoxic potential of ciprofloxacin hydrochloride, 0.3%, plus dexamethasone, 0.1%, after administration to the guinea pig middle ear.
Fifty guinea pigs were randomly assigned to 4 test groups of 10 animals each and 2 control groups of 5 animals each. The 4 test groups were treated twice daily for 4 weeks with 10 μL of (1) ciprofloxacin hydrochloride, 0.3%, plus dexamethasone, 0.1%; (2) ciprofloxacin hydrochloride, 1.0%, plus dexamethasone, 0.3%; (3) ciprofloxacin hydrochloride, 0.3%, or (4) vehicle. The positive and negative control groups were treated with neomycin sulfate, 10%, or isotonic sodium chloride solution, respectively.
Academic research laboratory.
Study animals were implanted with a drug delivery cannula to the middle ear, terminating in the round window niche for direct delivery to the round window membrane.
Main Outcome Measures
Auditory brainstem responses were collected at baseline and following 2 and 4 weeks of dosing. At the termination of the study, inner ear tissues were evaluated microscopically.
No biologically relevant hearing losses were observed after either 2 or 4 weeks of treatment with vehicle, ciprofloxacin alone, or combinations of ciprofloxacin plus dexamethasone. Examination of the organ of Corti revealed normal hair cell counts in all animals that received isotonic sodium chloride solution, vehicle, ciprofloxacin, or combinations of ciprofloxacin and dexamethasone. Conversely, the neomycin sulfate positive control group demonstrated a significant elevation in hearing threshold and profound hair cell loss (P <.001, P = .02, and P <.001 at 2, 8, and 16 kHz, respectively).
The results of this preclinical study support the safety of ciprofloxacin hydrochloride, 0.3%, plus dexamethasone, 0.1%, for clinical use in the open middle ear cavity.
Otitis media (OM) represented an important cause of childhood mortality prior to the advent of antibiotics and is second only to the common cold as the most frequently reported inflammatory disease of childhood.1 Although administration of systemic antibiotics considerably reduced the mortality associated with OM, the rate of complications and adverse effects with this route of therapy remains high.2 One complication of systemic antibiotics, the development of antimicrobial resistance, has raised global concerns regarding the best practices for use of these agents.3 By using a local approach to therapy, the risk for the development of drug resistance can be decreased.4 This benefit is realized by limiting the opportunity to have a negative impact on normal systemic flora, thereby decreasing the probability of opportunistic infection, while ensuring that the antibiotic concentration at the site of disease is sufficient to readily eradicate the pathogen.
Both ciprofloxacin hydrochloride and dexamethasone are pharmaceutical agents with a long history of safe and effective use after systemic administration. However, the concentrations achieved in the middle ear following systemic therapy are limited. Research in animals prior to this study had suggested that both ciprofloxacin5- 7 and dexamethasone8 could be safely used in topical therapy for OM with an open tympanic membrane. This study investigated the ototoxic potential of the combination of ciprofloxacin hydrochloride, 0.3%, and dexamethasone, 0.1%, as well as 3-fold higher concentrations, ciprofloxacin hydrochloride, 1%, and dexamethasone, 0.3%, when applied to the round window niche of the guinea pig middle ear for 28 consecutive days. The results of this study were filed in support of the regulatory review and approval of CIPRODEX Sterile Otic Suspension (Alcon Laboratories Inc, Fort Worth, Texas; trademark licensed from Bayer AG) in the United States.
Specific pathogen-free young adult (225-368 g) male and female National Institutes of Health pigmented guinea pigs from Elm Hill Breeding Laboratory, Chelmsford, Massachusetts, were used for the otic studies.9 Animal husbandry was in accordance with all applicable guidelines Good Laboratory Practice (GLP) in US Department of Agriculture and Association for Assessment and Accreditation of Laboratory Animal Care International accredited facilities. All animal procedures were reviewed and approved by the University of Michigan committee on use and care of animals.
The test drugs administered in this study are shown in Table 1. All test and control drugs were maintained at room temperature throughout the study. The vehicle used in this study contained benzalkonium chloride, sodium acetate, acetic acid, sodium chloride, glycerin, edetate disodium, and tyloxapol in proprietary concentrations. Neomycin sulfate (10%, in the presence of benzalkonium chloride, 0.006%, mannitol, 4.0%, sodium acetate, 0.03%, acetic acid, 0.04%, and edetate disodium, 0.05%) served as the positive control, and the negative control was isotonic sodium chloride solution. The test and control drugs were applied to 1 ear in 10 μL doses twice daily for 28 days.
The drug delivery system (Figure 1) was developed in the Cochlear Signaling and Tissue Engineering Laboratory at the University of Michigan. The cannula was composed of Micro-Renathane tubing (MRE025; Braintree Scientific Inc, Braintree, Massachusetts), a polymer-based elastomer, and the plastic portion of a 24-gauge angiocatheter with an injection cap (MX492; Medex Inc, Dublin, Ohio) screwed onto the angiocatheter to allow for serial substance administration.
The surgery was performed aseptically under anesthesia on study day 1. A small hole was created in the left bulla through which a cannula was inserted into the middle ear, such that it would terminate in the niche above the round window membrane. The cannula was secured at the bulla defect and was directed postauricularly to the top of the head where it was affixed to the skull via small anchor screws and methyl methacrylate (Figure 1).
Auditory brainstem response (ABR) evaluations10,11 were performed at frequencies of 2, 8, and 16 kHz on the left ear of each animal at pretreatment, again on study day 13 prior to dosing, and on day 28 prior to termination. Only those animals with baseline hearing within reference range were assigned to study. Tone bursts (rise/fall time = 1.0 milliseconds, duration = 10-15 milliseconds) were controlled by a computer using PATT digital attenuators (Wilsonics, San Diego, California) and Best Switch in Town (Wilsonics) tone switches. The acoustic stimuli were delivered through an encased and shielded transducer (Beyer earphone; Beyerdynamics, Farmingdale, New York) within the ear canal. Acoustic calibration was performed in a volume approximating the external ear canal using a 1/8-inch condenser microphone (model 4138; Brüel and Kjaer, Norcross, Georgia). The electroencephalogram (EEG) voltage was measured from electrodes placed at the vertex (plus), below the test ear (minus), and below the nontest ear (ground). It was filtered (300-3 kHz) and amplified (1000 times) using a modified Grass P15 (Astro-Med, West Warwick, Rhode Island) amplifier. The output of the amplifier was further amplified (×100) through an amplifier that was custom-built at Kresge Hearing Research Institute (KHRI) and delivered to the analog-to-digital converter for averaging. The ABR waveforms (1024 responses, 512 bins, 25 microseconds) were analyzed and controlled by software that was custom designed at KHRI.
At the completion of the study, 6 animals from each of the 4 experimental groups and all animals from the control groups were euthanized, and the temporal bones were harvested. The inner ear was immediately perfused with fixative (paraformaldehyde, 4%) through a hole in the apex and round window membrane as well as an opening in the oval window created by pushing the stapes into the vestibulum.
The temporal bones were then fixed for at least 2 hours followed by removal of the otic capsule and tectorial membrane from the cochleae. Each cochlea was then stained with fluorescent-labeled phalloidin. The organ of Corti was microdisected into surface mount preparations and mounted on slides for assessment while maintaining fragment orientation from base to apex. Both inner and outer hair cells for each animal were counted to yield a cytocochleogram mapping the presence or absence of hair cells by location along the cochlear spiral.
The remaining 4 animals from each of the 4 experimental groups were vascularly perfused with paraformaldehyde, 4%, in phosphate buffer, and the temporal bones were removed. The middle and inner ears were immersed in fixative for preservation, then rinsed in buffer, and decalcified in ethylenediamine tetraacetic acid. After decalcification was complete, the bones were embedded in methacrylate. Five- to ten-micron-thick sections through the middle ear were cut using a Leitz (New York, New York) microtome and collected onto glass slides. Slides were stained with an ethanol-based stain containing toluidine blue and basic fuschia, coverslipped with a xylene-based permanent mounting medium, and allowed to dry prior to microscopic analysis. Assessment focused on the mucosa and the bone-mucosal interface.
The means (SDs) for auditory thresholds recorded from each group, frequency, and time point were calculated using Sigma-Stat statistical software (Systat Software Inc, San Jose, California) and are shown in Figures 2, 3, 4, 5, 6, and 7. Statistical comparisons were conducted as unpaired t tests also using Sigma-Stat software.
The distribution of animals to test groups is shown in Table 1. The ABR testing showed a statistically significant threshold shift of less than 20 dB on day 28 at 2 kHz, in both the ciprofloxacin hydrochloride, 0.3%/dexamethasone, 0.1%, group (Figure 2, Table 2) and the vehicle control (Figure 5, Table 2) (P = .05 and P = .03, respectively). However, no statistically significant effects were seen at other frequencies in these groups (P = .50 at both 8 and 16 kHz in the ciprofloxacin hydrochloride, 0.3%/dexamethasone, 0.1%, group, and P = .63 and P = .55 at 8 and 16 kHz, respectively, in the vehicle group) or at any frequency in the ciprofloxacin hydrochloride, 0.3%, group (P > .07 at all frequencies evaluated) (Figure 4) and ciprofloxacin hydrochloride, 1.0%/dexamethasone, 0.3%, group (P > .25 at all frequencies evaluated) (Figure 3). Such low-frequency threshold shifts, in the absence of receptor (hair) cell losses, are indicative of middle ear involvement and consistent with our observations of residue in the ears of most of the animals treated with either the combinations or ciprofloxacin alone. In addition, there was a small increase in the mucosa lining the periosteum and around the ossicles relative to historical untreated ear data. This slight increase in the thickness of the mucosa lining seemed to be related to the presence of particulate matter in these ears. Therefore, the single-frequency and non–dose proportional audiometric findings were considered spurious in nature. The ABR findings following 2 and 4 weeks of treatment, therefore, revealed no biologically relevant differences in threshold response (day 28, drug-treated, vs day 28, isotonic sodium chloride solution control). Despite drug application directly to the surface of the round window membrane, which separates the middle and inner ears, the animals in the 4 test groups and in the isotonic sodium chloride solution control demonstrated a normal population of hair cells. Losses observed were within reference range (Table 3), with 1 exception in the ciprofloxacin hydrochloride, 0.3%, group, in which an animal demonstrated hair cell loss consistent with a preexisting condition. Conversely, the positive control group on day 28 (neomycin sulfate, 10%) experienced a significant threshold shift of 40 dB or higher at all 3 frequencies tested (P < .001, .02, and <.001 at 2, 8, and 16 kHz, respectively) (Figure 7) and sustained approximately 100% loss of outer and inner hair cells (Table 3), consistent with the known ototoxic nature of this compound.10- 12
The guinea pig model is well established for its ability to measure ototoxic drug potential and margin of safety when compared with effects in humans. Brummett et al9 provided early evidence that ototoxic effects of compounds applied topically to the ear could be measured effectively through audiometric measurements and histopathologic evaluation of the organ of Corti in the guinea pig model. Cochlear damage was clearly demonstrated after application of neomycin and polymyxin B in concentrations commonly found in commercially available otic drops at the time (neomycin sulfate, 5 mg/mL, and polymyxin B, 10 000 U/mL). Brummett et al9 showed that the toxic effect was dose related and could be duration dependent as well. Despite these findings, a discrepancy existed between findings in animal models and the lack of reported toxic effects after clinical exposure to the products in man.
Several reasons were suggested for the discrepancy of observation of ototoxic effects in animals but not in humans. First, anatomical differences between humans and guinea pigs in the location of the round window membrane and the inner ear space allowed for more drug access to the inner and middle ear membranes in the guinea pig. Second, chronic OM may induce some loss of hearing by virtue of the infection and inflammatory process, thus causing underreporting of hearing loss owing to drug therapy. Finally, Brummett et al9 observed that early hearing loss was first detected at test frequencies higher than those used in conventional clinical audiometry testing. By testing at a variety of frequencies and examining inner and outer hair cell loss, the local toxic effects of these topical agents was confirmed. Therefore, the guinea pig animal model was used in the current study to evaluate combinations of ciprofloxacin and dexamethasone.
In recent years, both ciprofloxacin and dexamethasone have been evaluated independently in animal models to assess their ototoxic potential. Brownlee et al5 conducted a study in which ciprofloxacin was administered directly into the middle ear of guinea pigs at concentrations of 0.75% (0.1 mL, twice a day for 7 days). Prieskorn et al13 performed a similar study that evaluated the effect of ciprofloxacin hydrochloride, 0.2% (10 μL, twice a day for 30 days), administered to the middle ear of guinea pigs. Both studies reported no effect on hair cell histopathologic characteristics or ABR. Claes et al6 reported no effect on auditory function following ototopical administration of ciprofloxacin to guinea pigs. Dohar et al14 administered ciprofloxacin hydrochloride, 0.2% (3 drops, twice a day, for 4 weeks), to cynomolgus monkeys with experimental chronic suppurative OM with tympanostomy tubes in place and reported no effect of ciprofloxacin on ABR or cochlear hair cell histopathologic characteristics. These studies suggest a low potential for ototoxic effects with topically applied ciprofloxacin.
There is also evidence in the literature suggesting that topical otic dexamethasone, used alone, has a low potential for ototoxic effects. Parker and James7 reported no ototoxic effects of dexamethasone or dexamethasone sodium phosphate, following transtympanic administration of 100 μL of a 1% suspension or solution, respectively, to guinea pigs once daily for 5 days. This concentration of dexamethasone is 10-fold higher than is present in the ciprofloxacin hydrochloride, 0.3%/dexamethasone, 0.1%, suspension used in the current study. Shirwany et al8 reported no significant effect on auditory sensitivity following administration of 500 μL of dexamethasone phosphate solution, 0.4%, weekly for 4 weeks in a guinea pig model.
To our knowledge, our study is the first to demonstrate the lack of ototoxicity for the combination of ciprofloxacin plus dexamethasone. This study was undertaken as a regulatory requirement prior to the initiation of clinical trials examining the safety and efficacy of the combination of ciprofloxacin hydrochloride, 0.3%, and dexamethasone, 0.1%, in children and adults. Although results of this animal study demonstrated a lack of ototoxicity in a carefully controlled and rigorous model, evidence of safety in humans is also required before a therapy can be concluded to be safe for human use. In phases II and III of the clinical trials, a total of 937 patients were treated with CIPRODEX Sterile Otic Suspension.15- 17 This included 400 pediatric patients with acute OM with tympanostomy tubes and 537 patients with acute otitis externa. The dosage regimen in these trials was 3 to 4 drops administered twice a day for 7 days. No clinically relevant changes in hearing function were observed in 69 pediatric patients (age range, 4-12 years) treated with CIPRODEX Sterile Otic Suspension and tested for audiometric parameters.16,17 Finally, CIPRODEX Sterile Otic Suspension has been demonstrated to be safe and well tolerated by patients in the clinical setting.15- 17
The results of our preclinical guinea pig model demonstrated that the histopathologic and electrophysiologic measures examined in this animal model are predictive of human safety. The results also demonstrate that the product is not toxic when administered to guinea pig middle ears for 28 days at concentrations 3-fold higher than those present in CIPRODEX Sterile Otic Suspension.
Based on the results of histopathologic and electrophysiologic hearing assessments in this 1-month GLP-regulated evaluation in guinea pigs, we concluded that a sterile suspension of ciprofloxacin hydrochloride, 0.3%/dexamethasone, 0.1%, was safe for administration to the open middle ear cavity of humans for the purpose of establishing clinical safety and efficacy. It is important to consider that the dosing procedure used in the guinea pig studies maximized middle ear (and likely inner ear) exposure to the test compounds and that the length of the animal study greatly exceeded the recommended duration of clinical treatment. The observation of drug residue in the middle ear of test animals was believed to be related to the lack of middle ear ventilation in this model, the small size of the guinea pig middle ear, and the shallow angle of the guinea pig eustacian tube, resulting in limited clearance from the middle ear. In some animals, this may have had a transient effect on low-frequency sensitivity but no adverse effect on the inner ear hair cells. Therefore, the lack of ototoxic effects observed following exposure to concentrations of ciprofloxacin and dexamethasone approximately 3-fold higher than those present in CIPRODEX Sterile Otic Suspension in this guinea pig model provided an adequate margin of safety to support initiation of controlled clinical trials to establish human clinical safety.
Correspondence: Leslie E. Lemke, PhD, DABT, Department of Preclinical Sciences, Alcon Research Ltd, Mailbox R9-7, Fort Worth, TX 76134 (Leslie.Lemke@alconlabs.com).
Submitted for Publication: January 24, 2008; final revision received June 16, 2008; accepted July 7, 2008.
Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: McGee, Prieskorn, Wall, Altschuler, and Miller. Acquisition of data: Prieskorn and Dolan. Analysis and interpretation of data: Lemke, Prieskorn, Dolan, Altschuler, and Miller. Drafting of the manuscript: Lemke, Prieskorn, and Dolan. Critical revision of the manuscript for important intellectual content: McGee, Prieskorn, Wall, Altschuler, and Miller. Administrative, technical, and material support: Lemke, McGee, Prieskorn, Wall, Altschuler, and Miller. Study supervision: Lemke, McGee, Prieskorn, Dolan, Altschuler, and Miller.
Financial Disclosure: Drs Lemke, McGee, and Wall are employees of Alcon Research Ltd. Dr Prieskorn was a consultant for Alcon Research Ltd when the study was conducted.
Funding/Support: The study, conducted at the Kresge Hearing Research Institute, University of Michigan, was sponsored by Alcon Research Ltd.
Previous Presentation: This study was presented at the Society of Toxicology Annual Meeting; March 24, 2004; Baltimore, Maryland.
Additional Contributions: Jill Bradstrom, Pamela Buie, BS, Beth Clark, Gary Dootz, Peter Finger, MS, Karin Halsey, BS, Alice Mitchell, BA, and Noel Wys, MS, provided technical assistance. Susan Gardner, PharmD, provided medical writing assistance.