Timothy C. Hain, MD
Last updated: 2/2003, Please read our disclaimer.
While ototoxicity from intravenous aminoglycoside administration is well documented, there is considerable controversy regarding the existence and significance of ototoxicity from topical preparations. The literature reviewed here suggests that ear drops containing aminoglycosides (such as gentamicin or neomycin), can occasionally cause hearing loss if administered over a long period of time to a person with a perforated ear drum. There is also evidence for vestibultoxicity from gentamicin topical drops as well as evidence for toxicity for select non-aminoglycoside preparations. As nontoxic ear drops are currently available (e.g. Ofloxin (ofloxacin) or ciprofloxin containing products), it would seem prudent in the future to use these agents instead of potentially toxic drugs, when there is a perforation.
Wright, Meyerhoff and Halama (1987) examined the effects of polymyxin B with neomycin and found greater toxic effect from polymyxin B in chinchilla and primates. Primates were less affected than rodents.
Morizono (1988) reviewed the ototoxicity of topical ear drops. There is good evidence for ototoxicity in animals of the aminoglycosides and chloramphenicol. Morizono pointed out that there is less evidence for ototoxicity in humans than in animals, and suggested that this might be related to a less exposed round window in humans, a thinner membrane, and other factors such as the presence of infection or effusion.
Podoshin and others (1989) compared patients treated with an otic mixture containing neomycin, polymyxin B and dexamethasone to a control group of 26 getting dexamethasone only. Duration of followup was 1-2 years. Patients treated continuously had a mean hearing loss of 10 DB, compared to 3.6 DB in patients treated intermittently and 1.8 DB in patients treated with long intermissions. Control patients hearing changed by -0.9 DB (improved), and was not related to intermissions in treatment. Comment: this is the only human study available. It suggests that there is a modest long-term risk of hearing decline with COS drops in perforation.
Morizono (1990) compared 10 minutes of exposure of the middle ear of COS, Vosol, and Colymyxin, in Guinea pigs. In this situation one would expect Colymyxin, which contains neomycin, to be the most toxic. Evoked responses were used as the measurement, at 24 hours. Colymyxin was found to be twice as toxic as COS, and Vosol, 4 times. This result would seem to suggest that the Colistin component of Colymyxin, is ototoxic, and also that propylene glycol in Vosol is ototoxic, but little conclusion can be drawn about long term effects.
Halama and others (1991) examined the cochlear ototoxicity of a single application of COS in baboons. They also examined separate comparisons of polymyxin-B and neomycin, which are components of COS. They examined the cochlea post-mortem, at 5 or 6 weeks They found that polymyxin B caused greater damage than did neomycin, and that both caused clear ototoxic damage including complete loss of inner and outer hair cells in the lower basal turn.
Lundy and Graham (1993) surveyed 2235 otolaryngologists regarding their opinions about the use of ototopical agent. 84.1% used these agents in the presence of a perforation, and many also used them during intraoperative packing. 80% indicated that the risks of hearing loss from otitis media was as great as or greater than the risks of ototoxicity of an ototopical preparation. Only 3.4% of respondents had witnessed irreversible inner ear damage caused by ototopicals.
Barlow and others (1994) compared injections of 0.2 ml of COs, GOS and Ofloxacin in guinea pigs into the middle ear, and measured hair cell damage at 7 days. The mean cochlear hair cell damage was 66% for COs, 6.5% for GOS, and 1.1% for Ofloxacin. This study documents toxicity for drug that penetrates the middle ear in animals. It does not address the question of whether there is a difference between human and animal toxicity, or relative vestibulotoxicity.
Linder and others (1995) reviewed 134 patient charts where ear drops were administered for potential ototoxicity. Two cases were found with bilateral profound hearing loss "directly attributable to topical ear drops". Ten other suspected cases were insufficiently documented. Comment: this study suggests an incidence somewhere between 2-10%.
Hui and others (1997) reported two cases of hearing loss attributed to gentamicin, Corticosporin, and sofracort otic drops, administered for for several years, in patients with tubes or perforations.
Wong and Rutka (1997) reported 5 patients with tympanic membrane perforations who they believed sustained ototoxicity from GOS. Case 1 developed vertigo after 3 weeks of treatment. Case 2 also developed unilateral vestibular loss and worsening of hearing. Cases 3, 4 and 5 developed bilateral vestibular loss after bilateral instillation over weeks to months. In a later article, Rutka commented that he had encountered 29 patients with topical gentamicin ototoxicity (Rutka, 2002). COMMENT: These cases are quite relevant and suggest that unilateral or bilateral vestibular loss can be associated with use of gentamicin based ear drops.
Pickett, Shinn and Smith (1997) considered evidence for and against ototoxicity from ear drops in a comprehensive review. They noted abundant animal evidence for ototoxicity of aminoglycosides, solvents such as propylene glycol (found in Vosol). They concluded that from animal data, it is clear that most ear drops that are used for treatment or prophylaxis of infection have ototoxic potential and point out the puzzling relative lack of common ototoxicity in humans. They suggest that either ototoxicity is not recognized as hearing loss is attributed to underlying disease, or because of anatomic differences between animals and humans in the thickness of the round window membrane, there is less susceptibility to topical ototoxicity in humans. They pointed out that there was a lack of well-controlled studies on the use of ear drops in humans.
Ozagar and others (1997) found no cochlear ototoxicity of either ciprofloxacin or gentamicin drops in 20 patients with chronic otitis media. COMMENT: This is hardly surprising as gentamicin is not very cochleotoxic in humans and ciprofloxacin is not ototoxic in animals. A more interesting study would have been to look at vestibular function 1 year after use of gentamicin drops.
Bath and others (1999) documented 16 patients with vestibulotoxicity, possibly related to topical drops containing gentamicin/steroid containing drops. An additional case was presented reporting intentional ablation of vestibular function using commercially available ear drops, in a patient with unilateral Meniere's disease. Two drops, twice a day were instilled through a ventilation tube over 3 weeks.
There are many things we don't know about ototopicals and their toxicity. Why is it that ototoxicity is so much rarer in humans than in animals ? What ototopicals have the highest risk of toxicity ? How do gentamicin drops cause hearing toxicity when given topically, but rarely when given systemically ? What is the true prevalence of genetic susceptibility to aminoglycosides in the population at large and in genetic subgroups ?