Ototoxicity

Rita M. Schuman
Gregory J. Matz

Drug-induced inner ear damage is a common finding in present-day medical practice. In many developing countries, where drugs such as the aminoglycosides are frequently prescribed to treat pneumonia, diarrhea, and tuberculosis, the incidence of ototoxicity is high (1). Physicians in practice need to recognize that ototoxic drugs can cause significant auditory and in many instances, poorly recognized, vestibular toxicity. Physicians therefore need to be cognizant of the many categories of drugs that produce ototoxicity.

Early examples of drug ototoxicity are arsenic, the salicylates, and quinine. Salicylates, for example, administered in doses in excess of 2,700 mg a day, once commonly used to treat arthritis, were found to cause a transient flat, bilateral sensorineural hearing loss and tinnitus. There has never been a case of permanent hearing loss following salicylate use in therapeutic drug dosing; however, most patients experience complete reversal within 2 to 3 days. Later, in the 1960s, thalidomide, a well-known drug used at that time and now known to cause amelia and phocomelia, was also discovered to cause aplasia of the inner ear.

The introduction of the first aminoglycoside, streptomycin, in 1944 by Waxman, who was awarded the Nobel prize for this discovery, heralded a new era of antibiotic therapy for the treatment of tuberculosis. Unfortunately, Hinshaw and Feldman at the Mayo Clinic described a significant number of patients with vestibular toxicity from this drug (2). A few years later, an analog of streptomycin, dihydrostreptomycin, was used in clinical practice with the hopes of reducing the streptomycin ototoxicity. Dihydrostreptomycin, however, was also shown to have an unacceptably high incidence of cochlear toxicity and was subsequently withdrawn from the market. Likewise, other early aminoglycosides, such as kanamycin and neomycin had unacceptably high rates of cochlear toxicity when used systemically and therefore are rarely used in that manner today. Later, a newer aminoglycoside, gentamicin, was shown to have about a 3% incidence of vestibular injury (3). Subsequent aminoglycosides such as netilmicin, tobramycin, and amikacin were developed to reduce this incidence of toxicity. In fact, netilmicin has been found to be the least ototoxic of all of the aminoglycosides available (4).

Other considerations must include the cancer chemotherapeutic agents, such as cisplatin, which has been found to result in a moderate level of ototoxicity with resultant permanent bilateral hearing loss. Clinicians are also faced with a sporadic low incidence of ototoxicity with drugs such as vancomycin and the macrolides. Most studies in the literature regarding the ototoxicity of the macrolides have been found to be reversible. The mechanism by which these drugs are toxic is unknown. Finally, numerous case reports have also indicated that hydrocodone in combination with acetaminophen can cause a rapidly progressive sensorineural hearing loss. (5). The mechanism of toxicity at this time is unknown.

Ototoxicty of Ototopical Antibiotics
It is well known that systemic administrations of aminoglycosides can cause both cochlear and vestibular toxicity. This naturally leads to the question of whether these drugs, which are used extensively to treat ear infections through topical administration to the middle ear, can cause ototoxicity. Animal data have been quite uniform in that almost all of the aminoglycoside antibiotics used in the middle ear as topical otic preparations are ototoxic (6). The use of aminoglycoside ototopical drops confined to the external auditory canal, however, presents little, if any, risk of ototoxicity.

Current review of the literature reveals documentation of a total of 54 cases of gentamicin vestibular toxicity from ototopical use in the middle ear or open mastoid cavity (7) (Table 148.1). In addition, 24 of these patients developed an associated auditory toxicity. A review of the literature in the above cited study also included 11 patients who experienced auditory toxicity from the topical use of neomycin-polymyxin-based eardrops. It was therefore recommended that when possible,
topical antibiotic preparations free of potential ototoxic side effects should be used in preference to ototopical preparations that have had the potential for ototoxic injury if the middle ear or mastoid are open (17). Aminoglycoside-containing antibiotic topical drops are not FDA approved for use in the middle ear or open mastoid cavity. Indeed, current labels contain warnings against the use of these drugs if the tympanic membrane is not intact. Although the evidence suggests that otologic damage from topical preparations with ototoxic potential is infrequent, the evidence also indicates that they offer no advantage over nonototoxic agents (17). If these ototoxic agents are considered, potentially ototoxic, antibiotic preparations should be used only in acutely infected ears and use should be discontinued shortly after the infection has resolved. Finally, if the clinician must use potentially ototoxic antibiotics in the middle ear or mastoid space, the patient or parents should be warned of the risk of ototoxicity (17).

Ototoxicity of Systemic Drugs

Table 148.2 lists the major classes of drugs that cause ototoxicity—the aminoglycoside antibiotics, the macrolides, loop diuretics, cisplatin, and the salicylates. These drugs are listed because they are commonly seen in otolaryngology consultation practice. There are currently no meta-analysis studies that evaluate ototoxicity for these drugs. Included in the bibliography are two reviews that include ototoxic evaluations for gentamicin and cisplatin in a large cohort of patients. Omitted for the sake of brevity in Table 148.2 are the low-incidence ototoxic drugs, such as chloroquine, which is rarely used in clinical practice in the United States.

For various reasons, the incidence of aminoglycoside ototoxicity in neonates and children is lower than adults (23). In children, it can be useful to obtain pretreatment audiograms to rule out preexisting hearing loss in patients who are to receive a course of aminoglycoside antibiotics. In the United States, that drug is usually gentamicin.

Genetics of Otoxicity

It is well known that aminoglycosides are some of the most common ototoxic drugs causing acquired hearing loss. It was observed that many patients were developing hearing loss, despite the low dosages of aminoglycosides administered. It was also noted that certain families had an exceptionally high number of members with similar findings of aminoglycoside ototoxicity. Based on these observations and the ongoing research regarding the pathophysiology of hearing loss, it has been proposed that certain individuals may have a genetic predisposition or susceptibility to the ototoxic effects of certain drugs and in particular, the aminoglycosides (24).
Recent advances have identified that certain mutations in mitochondrial DNA are found to be associated with a number of hearing disorders, including ototoxicity. Mitochondrial DNA is a double-stranded molecule forming a closed circle. Replication and transcription occurs within the mitochondria, ultimately forming proteins involved with ATP synthesis and electron transport. This specific type of DNA is transmitted exclusively by the maternal line, equally affecting both male and female offspring.

In the early 1990s it was first discovered that a mutation at position 1555 in the nucleotides of the mitochondrial 12S ribosomal RNA was responsible for aminoglycoside toxicity in several Chinese families (25). It was also cited as a cause for a number of cases of nonsyndromic deafness in patients with no previous aminoglycoside exposure. Since that discovery, similar research has been conducted on numerous other families, as well as on sporadic patients with documented sensorineural hearing loss following the administration of intravenous aminoglycosides. These subsequent studies confirmed that these patients also had identical nucleotide mutations of their mitochondrial DNA. It has been proposed that the specific mutation creates another binding site for the aminoglycosides, thus increasing the patient's sensitivity to ototoxicity (26). Most of this work was conducted on an international basis, where severe infections such as tuberculosis more often require widespread use of intravenous aminoglycosides.

A large quantity of research continues in this area. As more becomes known about the genetics of hearing loss and the specific mutations that predispose patients to the ototoxic effects of some drugs, it may be possible to develop molecular tests to identify these patients prior to treatment. With that information, it may be possible to reduce the number of patients suffering from the toxicities of these antibiotics.

Chemoprevention of Ototoxicity
In some instances, it may be necessary to use ototoxic drugs in order to effectively treat patients. In light of this fact, it is necessary to develop mechanisms by which it is possible to protect the inner ear from the toxicities of both the ototoxic intravenous antibiotics and the chemotherapeutic agents such as cisplatin. Some of the agents that have been proposed and studied include iron chelators (deferoxime) (27,28), antioxidants including L-N-acetyl cysteine (29), vitamin E, alpha-tocopherol (30,31,32), as well as the salicylates (33,34).

Recent research has demonstrated that administration of aminoglycosides causes the formation of an iron complex that is involved in the generation of free radicals, resulting in hair cell death and subsequent hearing loss (28). Based on this discovery, attempts have been made to use deferoxamine, an iron chelator, to help attenuate these toxic effects. Animal studies have been promising, but considerations must be taken so as to not alter the serum concentrations of the drugs, and a better understanding is needed of the side effects of administering iron chelators to patients and the potential of altering serum iron levels (27).
Cisplatin, a common chemotherapeutic agent used in head and neck cancer, is well known to cause bilateral, irreversible sensorineural hearing loss. Evidence suggests that glutathione reduction secondary to free radical production ultimately causes hair cell damage. Various chemoprotectants have been shown to exert antioxidant properties that ultimately reduce the ototoxic effects of cisplatin. Recent studies with vitamin E (31), L-N-Acetyl cysteine (29), and sodium thiosulfate (35) confirm this theory. Most of the research, however, has thus far been with animals. Further human studies must be done to truly know if these advances will be clinically significant and will ultimately reduce the ototoxic effects.

Summary

One of the authors of this chapter (GJM) has written previously about the use of high-frequency audiometry (8 to 12 Hz) as a predictor of drug-induced ototoxicity (36). Although conventional audiometry may still have a role in monitoring patients exposed to ototoxic medications, high frequency testing is often problematic. It is an extremely difficult test to do for all practical purposes and is often not done clinically for that reason. Few centers perform pretreatment conventional audiograms from 0.25 Hz to 8 Hz when the two most common ototoxic drugs, gentamicin and cisplatin, are given. The authors are not aware of any outcome studies that demonstrate that pretreatment and post-treatment audiograms reduce the incidence of predicted ototoxicity. Some centers have found, however, that it may be beneficial to perform one pretreatment audiogram followed by serial audiograms, in addition to closely monitoring the serum drug levels of the ototoxic medications being administered. The use of vestibular testing both pretreatment and post-treatment for patients receiving long-term gentamicin is also difficult to do in the clinical setting. This is an important factor because gentamicin is mostly a vestibular toxic drug. Some centers do perform electronystagmography, rotational testing, and platform posturography in working up possible vestibular symptoms and have found these tests to be helpful.
It is now well known that the aminoglycoside antibiotics act synergistically with some drugs, thus increasing the incidence of ototoxicity. For example, the use of aminoglycoside antibiotics with loop diuretics can produce an unexpectedly high incidence of ototoxicity. This has been extensively documented in human case reports as well as in animal studies. Ethacrynic acid, an ototoxic loop diuretic, has been shown to increase the permeability of the stria vascularis, facilitating the diffusion of the aminoglycoside into the endolymph. Finally, it has been found that diuretics given prior to the administration of aminoglycosides are less damaging than if done in the reverse (37). Most recently noted is a similar response to aminoglycoside antibiotics and the use of metronidazole (38).
It is unclear at this time if antiviral and protease inhibitors are responsible for the anecdotal reporting of neurosensory hearing loss in patients with human immunodeficiency virus (39). Prospective studies are needed to confirm whether nucleoside analog reverse transciptase inhibitor or antiviral agents cause hearing loss in this patient population. The use of chemoprevention measures as described in animal studies show promise, but so far no prospective clinical trials have been performed and the authors are not aware of any medical centers with protocols to address this issue at this time.
The two most common ototoxic drugs given today in clinical practice are gentamicin and cisplatin. The patients selected in these groups are different. Gentamicin is normally monitored not by audiograms but by serum peak and trough levels. When gentamicin has to be given for long-term therapy (i.e., osteomyelitis), consideration has to be given to genetic testing to see if a patient is going to be more susceptible to ototoxic injury, thus giving the clinician the opportunity to obtain informed consent from the patient. Likewise, accurate dosing during chemotherapy has reduced the incidence of ototoxicity. Further research is necessary to determine if any of the chemopreventative agents will be successful in further animal and ultimately human trials to reduce the unfortunate toxicities of these necessary drugs.