Update on topical ototoxicity in chronic suppurative otitis media.

I define chronic suppurative otitis media (CSOM) as a persistent or recurrent bacterial infection of the middle ear that is characterized by a tympanic membrane perforation, hearing loss, and persistent or recurrent otorrhea. Uncomplicated CSOM can be managed quite safely medically because it is primarily a mucosal problem. CSOM accompanied by cholesteatoma is more difficult and is usually managed by surgery; left untreated, it can be fatal.

The primary reason we treat CSOM is to prevent complications. The most feared complication is an otogenic intracranial brain abscess. Nunez and Browning reported that the annual risk of such an abscess in active CSOM is only 1 in 10,000. (1) However, the lifetime risk in an individual aged 30 years is 1 in 200. Nunez and Browning also reported that this risk is three times higher in men and that 5% of these abscesses occur in the immediate postoperative period following mastoidectomy.

Ototoxicity in CSOM

I define ototoxicity as the tendency of certain substances to cause functional impairment and cellular damage to tissues of the external, middle, and especially the inner ear, regardless of whether the offending agent is applied systemically or topically.

The major recognized groups of systemic ototoxic drugs identified in clinical practice are the aminoglycosides, traditional time-honored medications such as the quinines and salicylates, chemotherapeutic agents (cisplatinum in particular), and diuretics. One interesting facet of ototoxicity is that we often fail to realize that when a drug gets into the inner ear, it often stays there. This might he one reason that one of the risk factors for systemic gentamicin toxicity is a second treatment course.

There are three major groups of topical ototoxic drugs: the aminoglycosides (e.g., streptomycin, neomycin, and gentarnicin), the antifungals (e.g., gentian violet), and the antiseptics (e.g., alcohol and chlorhexidine). Whichever agent is put into the middle ear space will most likely reach the inner ear eventually. Given time, most topical substances can pass through the round window membrane and into the inner ear. For this reason, one should never use a chlorhexidine preparation during ear surgery if, for example, there is a hole in the tympanic membrane.

I have no doubt that ototopical preparations are effective in treating CSOM, but practice patterns vary and continue to evolve. Some 15 years ago, these patterns were drastically different in the United Kingdom and in the United States. In 1988, a survey of general practitioners in the U.K. revealed that 66% did not prescribe a topical agent for otorrhea in a patient whose eardrum was perforated. (2) Moreover, 85% said they would not use a topical agent in a patient with post-tympanostomy tube otorrhea (PTTO). The primary reason for this reluctance was the risk of ototoxicity and its complications.

The opposite was true among otolaryngologists in the United States, as Lundy and Graham reported in 1993. (3) They found that 84% of otolaryngologists used an ototopical agent in the presence of a draining perforation, and 94% did so in a patient with PTTO. Most of these physicians said that they had no compunction about using drops because they were perfectly safe, even those drops that contained an aminoglycoside.

We have since come to learn more about topical ototoxicity. In 1999, the National Formulary passed by the British Parliament limited the duration of therapy with a topical agent containing an aminoglycoside to 7 days because of mounting evidence that these preparations are ototoxic.

Although ototoxicity from topical aminoglycosides is relatively uncommon considering the large number of patients who are treated with them every year, the risk is not negligible and is probably higher than we previously estimated. There are many reasons why topical ototoxicity does not occur as often as one might expect. For example, drops might not get through the membrane of a very small perforation. Other reasons have to do with round window membrane permeability, the status of eustachian tube function, mechanical barriers in the round window niche (e.g., mucosal webs), and an individual patient's inherent susceptibility to aminoglycosides.

Studies of topical gentamicin ototoxicity

New information obtained from gentamicin studies has since prompted otolaryngologists in North America to rethink our practice patterns and become much more cautious about using topical aminoglycosides in patients with perforated eardrums. Over the past decade, we have been using topical gentamicin in very high doses to treat Meniere's disease. The aim was to cure vertigo by inducing a "chemical labyrinthectomy." According to Blakley (4) and Gustafson and Pensak, (5) intratympanic gentamicin, especially in high doses, reportedly controlled vertigo in 80 to 100% of patients with Meniere's disease.

Initially, our preferred technique in Toronto for gentamicin ablation was to place the tubing of a butterfly catheter through a myringotomy hole in the eardrum. We would then inject 1.5 ml of solution that contained approximately 25 mg/ml of gentamicin three times a day for 3 days. Within a week, patients would typically become profoundly vertiginous. They did not become vertiginous during the actual treatment; the reaction was delayed. We later realized the reason for this. Following absorption of the drops through the round window membrane, gentamicin has to travel from the basal turn of the cochlea and through all its turns before it reaches the neuroepithelium of the otolithic organs and the semicircular canals in the vestibule where it does its damage. This takes time.

Having become sensitized to the effects of topical gentamicin in the treatment of Meniere's disease, we began to recognize a pattern in some rather interesting cases at our specialized multidisciplinary neurotology clinic at the University of Toronto. By taking a thorough history, performing a proper otoneurologic examination, and analyzing laboratory data, we believe we now have incontrovertible evidence that 29 patients have been referred to us with topical gentamicin ototoxicity. The patients in question had all used commercially available topical gentamicin preparations for an average of 16 days before they became toxic in the presence of a tympanic membrane perforation, and their ear(s) had stopped discharging over the treatment course many days previously. These drops were probably safe over the short term, but over a longer course, they reached the inner ear and the patient became toxic. Some of these patients developed bilateral toxicity when the drops were applied injudiciously to both ears. To illu strate the seriousness of topical gentamicin toxicity, nine of our patients developed oscillopsia and/or ataxia, and seven of them never returned to work. In fact, five of them eventually were unable to walk and became confined to wheelchairs.

In view of topical gentamicin's ability to cause ototoxicity unintentionally, we wondered if perhaps we could use these drops therapeutically to purposely cause ototoxicity in patients with Meniere's disease. Indeed, we tested this theory on 20 patients with unilateral Meniere's disease by placing a ventilation tube into the posteroinferior quadrant of the tympanic membrane, close to where the round window niche would be located. We then gave these patients follow-up instructions to lie on their side and put 5 or 6 drops into the affected ear three times a day. We told them to administer the drops until they became constantly vertiginous for 2 days, and then to stop.

The vast majority of these patients became toxic by about day 12. We compared their pre- and post-treatment electronystagmography and audiogram results. In the patients whom we ablated for Meniere's disease, we found that the excitability-difference curve shifted up. In fact, the findings were almost identical to what we saw when we used a very high concentration of gentamicin initially for a shorter period of time. Therefore, the use of a less concentrated form of topical gentamicin for a longer period of time will yield the same results. Well over 50% of these patients ultimately had no response to ice-water calorics, and another 25% demonstrated significant changes in excitability difference. The reason that the Meniere's patients became toxic before those whom we believe sustained inadvertent toxicity for treatment of middle ear disease resides in the fact that, unlike the latter group, the Meniere's patients had no inflammatory or infective response in the middle ear prior to treatment.

Comments

Dr. Croxson: Both Gilchrist et a1 (6) and Black et a1 (7) described vestibular recovery following gentamicin injury. Have you had any clinical experience with this?

Dr. Rutka: No, but I'm sure it happens. One reason we do not see it often is because a patient might have already recovered by the time he or she is referred to us. Another reason is that most patients with a unilateral vestibular loss compensate for it fairly well if they're young and reasonably motivated. It's also quite likely that the treating physicians not infrequently ascribe dizziness to the effects of the infection itself rather than to the treatment.

The important point is that the drops we use are pretty effective, regardless of whether they're aminoglycosides or fluoroquinolones. But certainly the downside with topical gentamicin is that it is toxic to the vestibular structures when it is used too long, especially after the middle ear inflammatory/infective process has resolved.

References

(1.) Nunez DA, Browning GG. Risks of developing an otogenic intracranial absccss. J Laryngol Otol 1990;104:468-72.

(2.) Bickerton RC, Roberts C, Little JT. Survey of general practitioners' treatment of the discharging ear. Br Med J (Clin Res Ed) 1988;296: 1649-50.

(3.) Lundy LB, Graham MD. Ototoxicity and ototopical medications: A survey of otolaryngologists. Am J Otol 1993;14:141-6.

(4.) Blakley BW. Update on intratympanic gentamicin for Meniere's disease. Laryngoscope 2000;1 10:236-40.

(5.) Gustafson LM, Pensak ML. Inner ear perfusion therapy: An update. Clinical Opinion in Head and Neck Surgery 2000;8:398-402.

(6.) Gilchrist DP, Curthoys IS, Cartwright AD, et al. High acceleration impulsive rotations reveal severe long-term deficits of the horizontal vestibulo-ocular reflex in the guinea pig. Exp Brain Res 1998;123:242-54.

(7.) Black FO, Gianna-Poulin C, Pesznecker SC. Recovery from vestibular ototoxicity. Otol Neurotol 2001 ;22:662-71.

John Rutka, MD, FRCSC 

Dr. Rutka is a staff neurotologist and an associate professor in the Department of Otolaryngology at the University of Toronto. He is also director of the Ear Pathology Research Laboratory and codirector of the University Health Network Centre for Advanced Hearing and Balance Testing and the Multidisciplinary Neurotology Clinic at Toronto Hospital.