Reversing Sudden Deafness

Some people go to bed with perfect hearing and by the next morning are deaf in one ear. Others feel their hearing decrease over minutes or hours, like air leaking from a punctured tire. Still others experience a sudden pop, and their hearing is gone.

What would you do?

One rather common scenario is that your ears have been blocked during a head cold. But, a week after your cold is gone, and one ear has opened, the other remains blocked. You assume the problem is earwax. When you still have trouble hearing after a week or two, you buy an ear-cleaning kit at the drugstore. This does not work, so, after a few days, you call your doctor’s office, and the nurse recommends an over-the-counter decongestant, an antihistamine, or both.

Two weeks later, your ear is still blocked, so you call the doctor’s office again. The nurse checks with the doctor, who prescribes an antibiotic. Two weeks later, you call again: “My ear is still blocked.” This earns you an appointment with the doctor, who, discovering your ear appears normal, refers you to an ear, nose, and throat specialist.

A month later, when you finally see the specialist, a hearing test shows a severe nerve-caused deafness in the affected ear. Unfortunately, two months have passed since the beginning of what is called sudden sensorineural hearing loss (SSNHL), and it is now too late for you to regain any hearing by means of the most common and effective treatment, steroid therapy.

The confusion that underlies this all-too-common, unfortunate, and usually unnecessary outcome is the result of the different possible causes of deafness and the fact that sudden hearing loss originating in nerve damage is a special case. Time is of the essence in saving hearing, and researchers are working hard to prolong the precious open window of opportunity for effective treatment.

Blocked Hearing—But Why?

Three basic, related processes must occur for a person to hear. First, sound vibrations traveling through the air must reach the eardrum and be carried through the middle ear bones to the fluid of the inner ear. This is sound conduction. Second, in the inner ear, sensory hair cells of the cochlea must convert the sound energy in a way that triggers signals in the auditory nerve. This is signal transduction. (As an aside: Because of their complex shape, these hearing structures and the associated balance, or vestibular, organs are collectively known as the labyrinth, which is in a portion of the skull base called the temporal bone.) Finally, the signals must be carried by nerves to the auditory centers of the brain. This is neural transmission.

Although hearing can be compromised by certain changes in any of these three processes, hearing loss is broadly divided into two categories: conductive and sensorineural. Conductive hearing loss denotes conditions in which the inner ear and auditory nerve are working normally, but sound vibrations simply are not reaching them. A few common causes are blockage of the ear canal, perforations of the eardrum, accumulation of fluid in the middle ear, and damage to the middle ear bones. The other category, sensorineural hearing loss, denotes conditions in which sound vibrations do reach the inner ear but then fail to reach the brain as nerve signals. This involves some failure of signal transduction by hair cells in the cochlea or failure of neural transmission by the auditory nerve.

Conductive hearing loss is often correctable. The unfortunate person in our opening scenario assumed conductive hearing loss was the problem, especially because the hearing loss occurred in conjunction with a cold, and sought remedies such as an earwax kit. On the other hand, conventional wisdom is that sensorineural hearing loss is not correctable, because it originates in nerve damage of some kind. In fact, however, many cases of sensorineural hearing loss are reversible: for example, those caused by mild acoustic trauma (noise injury), some types of drug reactions, Ménière’s disease, autoimmune disease of the inner ear, and some cases of post-meningitis deafness. Many cases with no identifiable cause also resolve themselves over time.

True, cases of reversible sensorineural hearing loss are not the most common forms of deafness. But they are disproportionately important to study. They are natural experiments—valuable windows into the workings of the inner ear in health and disease. To grasp why these cases of reversible sensorineural hearing loss hold such opportunity for research progress, consider an analogy. Imagine you are an auto mechanic and the only cars that come into your shop are wrecks that were totaled in accidents—hopelessly beyond repair. You would never get a car back on the road. Your morale as a repairman would be dismal, you would have no satisfied customers, and you would not learn much about how cars worked or how to fix them. This is exactly the situation facing the physician encountering most cases of sensorineural hearing loss. Most cases are either congenital losses of hearing, caused by damage in the uterus or a genetic malfunction, or are the result of progressive degenerative changes with aging. Although physicians can offer these patients hearing aids to compensate for some of their hearing loss, at present physicians have no way to return their hearing to normal. But, in contrast, cases of reversible sensorineural hearing loss are analogous to cars that are driven into the auto repair shop running roughly, stalling, or making worrisome noises, but that still offer some hope of repair. Customers tend to drive away satisfied with the repair job, the mechanic feels good, and with each case he learns more and more about how to fix cars.

SSNHL, whether reversible or not, affects an estimated 1 in 5,000 Americans a year. In metropolitan Boston, where I practice, a population of approximately 1.5 million produces approximately 300 cases of SSNHL each year. The peak incidence is in the age range of 45 to 55 years, although SSNHL can occur at any age, and men and women are equally affected. No single test is available for this condition. The physician must make an educated clinical diagnosis by considering the signs and symptoms present and eliminating other possible explanations for them. Physicians generally agree on the signs and symptoms to look for, because these have been well described in the medical literature since SSNHL was first characterized in 1944 by a Dutch otolaryngologist named DeKleyn.¹

SSNHL refers to a new onset of deafness in one ear that occurs in less than 72 hours. The hearing loss can vary from mild to profound. Although the hearing loss seems to strike suddenly—as when a patient experiences a sudden pop and hearing is gone—in actuality, as patients look back, they often recall fluctuating ear symptoms: tinnitus (noises in the ear), transient hearing loss, or dizziness for a week or two before the SSNHL really hit. Virtually all patients report loud tinnitus (ringing, roaring, screeching, buzzing) and intense ear fullness or pressure in the affected ear. About half experience transient dizziness, disequilibrium, or vertigo during the first few days of their illness. Because these symptoms can be indicative of other disorders, a physician should perform other tests. For example, brain imaging is performed to rule out tumors, stroke, or demyelinating (loss of nerve fiber insulation) disease such as multiple sclerosis. If left untreated, 25 to 30 percent of patients with SSNHL will achieve some degree of spontaneous improvement, although not necessarily to normal. If treated promptly, however, the chances of some improvement rise to half or more of cases.²

Debating the Causes

Many mechanisms have been proposed to explain SSNHL. The most popular explanations invoke either impaired blood flow to the inner ear (with varieties such as vascular insufficiency, embolism, or thrombosis) or viral infection. In fact, both of these mechanisms probably do occur, but the evidence supporting them is still circumstantial; no direct cause-effect relationship has been pinned down.

Impaired Blood Flow to the Inner-Ear

Blood is delivered to both the vestibular (balance) and cochlear (hearing) divisions of the inner ear at the end of a branching tree of blood vessels. As a result, only single tiny arterioles actually carry red blood cells and oxygen to the inner ear. If one of these terminal arterioles is blocked, everything beyond the blockage is deprived of oxygen. Tissue injury resulting from oxygen deprivation is called ischemia, whereas tissue death resulting from oxygen deprivation is called infarction. Both ischemia and infarction are proposed as part of the vascular theory. Three types of circumstantial evidence suggest that some cases of SSNHL result from vascular blockage:

• Sudden onset is suggestive of infarction (which often occurs suddenly).
• Case reports exist of sudden deafness in association with known systemic vascular diseases.
• Animal experiments have demonstrated that sudden deafness and related cochlear changes can be caused by vascular occlusion, or blocking.

The instantaneous onset of deafness in many cases of SSNHL is similar to the onset of neurologic events such as a stroke or transient ischemic attack that are precipitated by blood clots or similar causes. In experiments, deliberately blocking the artery to the inner ear of guinea pigs to cut off the oxygen supply led to loss of cochlear electrical activity after 60 seconds and permanently depressed cochlear functioning after half an hour. Sudden deafness has been reported also in human medical conditions associated with abnormal blood vessels, abnormal blood viscosity, or abnormal blood clotting, as, for example, in sickle cell disease. Examination of the temporal bones of guinea pigs after experimental blockage of labyrinthine veins or arteries shows loss or death and degeneration of several different cell types, scarring of the inner ear, and eventual new bone formation within the cochlea. This new bone formation, called labyrinthine ossification, is a hallmark of ischemic injury, although it has been observed only rarely in human SSNHL temporal bones.

Remember, proponents of the vascular theory are seeking a smoking gun: a direct linkage between some vascular event such as ischemia or infarction and definite changes associated with hearing loss. Changes in the temporal bones, such as labyrinthine ossification, if consistently found in patients with SSNHL, would be such a smoking gun. Saumil N. Merchant, M.D., and his colleagues at my institution, the Massachusetts Eye and Ear Infirmary, recently reviewed specimens of temporal bone in the Infirmary’s collection from 17 cases of SSNHL.³ Only one specimen showed any evidence of new bone formation—the vascular theorists’ hoped-for smoking gun. Another strike against the vascular theory is the relationship of SSNHL prognosis to the site of cochlear injury. The cochlear artery runs from the base of cochlea, where high-frequency sounds are detected, to the apex, where low-frequency sounds are detected. Since there is no collateral blood supply to the cochlear apex, blockage of the cochlear artery should cause the most severe damage to low-frequency hearing. But clinical reports show exactly the opposite: SSNHL affecting low-frequency hearing of the cochlear apex actually has a better prognosis than SSNHL affecting high-frequency hearing in the cochlear base. Another difficulty is that if the labyrinthine artery itself were affected by some vascular event, both auditory and balance functions should be impaired, but only a few patients with SSNHL experience severe or sustained vertigo.

Viral Infection

For the other chief explanation of SSNHL, the viral theory, four types of direct and indirect evidence are seen:

• SSNHL seems to occur quite frequently when people have colds.
• Blood tests of people with SSNHL sometimes show an active viral infection.
• Post-mortem examination of the temporal bones of humans with SSNHL has shown changes similar to those observed in known viral causes of deafness.
• Experiments with animals show that viruses can penetrate the inner ear.

As with the theory that impaired blood flow of the inner-ear causes SSNHL, the evidence for the viral theory is primarily circumstantial and sometimes contradictory.

Some reports maintain that a cold or other upper respiratory illness preceded the onset of SSNHL in as many as 40 percent of cases. Unfortunately, these reports lack corresponding data on the comparative frequency of upper respiratory illness in a matched control population. What about the evidence of blood examinations? In response to a virus, the immune system produces a temporary increase in the level of antibodies against the specific virus, and many case reports on patients with SSNHL show that they experience a brief, sharp rise in antibody levels against common viruses such as herpes, flu, mumps, or rubella.

Researchers have studied damage to human temporal bone cells and tissues of people who had SSNHL. Because the temporal bones are not available for study until after a person dies, often decades after the onset of SSNHL, the microscopic findings reflect a lifetime of changes—not just those arising immediately from the SSNHL. Even so, such studies are useful and have shown extensive degeneration of cochlear structures, including hair cells and supporting cells. The myelin sheath that insulates certain nerve cells has unraveled in these patients, although their spiral ganglion cells (nerve cell bodies) remain relatively intact. ^{4, 5} These changes are similar to those observed in the temporal bones in cases of known causes of viral deafness. For example, guinea pigs with systemic and certain viral infections of the inner ear have viral particles in the spiral ganglion and experience inflammatory reactions that confirm the virus has penetrated the inner ear.

One case in Merchant’s study of the temporal bones of patients with SSNHL was unique, because that particular patient had developed his sudden deafness and died only nine days later of an unrelated medical condition. Although tragic for this patient, the quickness of his death was a boon to SSNHL research. Here was an unprecedented opportunity to study SSNHL tissue damage immediately after the onset of the hearing loss, not decades later. The patient’s temporal bone showed extensive swelling, distortion, and damage. But the blood vessels and auditory neurons looked normal, and no signs were seen of infection or inflammation—bad for both the vascular and viral theories.

Because of these observations, Merchant and his colleagues hypothesized that SSNHL might not require viral penetration or a vascular event at all. Might SSNHL arise instead as a result of abnormal activation of cellular stress pathways? The origins and mechanisms of such stress pathways are complex, but the end result can be the synthesis of proteins that upset the natural balance of cells and tissues, or even cause what is called cellular apoptosis, or programmed cell death. Biochemical analysis of parts of the cochlea appears to support this hypothesis.⁶

Treatment

A computer search of the National Library of Medicine turns up thousands of publications on SSNHL during the past 40 years. The many uncontrolled trials, retrospective reviews, and studies of sets of cases propose a multitude of treatments, but only 15 English-language publications report randomized controlled trials of initial therapy of untreated SSNHL. Justifications for the various proposed treatments are based either on some presumed cause of SSNHL (including the two most popular theories, viral and vascular) or on previous experience of the treating physician. Virtually all published treatments, including so-called “shotgun therapy” of multiple simultaneous treatments, achieve success rates of 40 to 75 percent in uncontrolled trials. Unfortunately, there is no satisfactory way to judge the statistical significance of these results because the studies made no effort to define an adequate sample size, use random assignment to treatment, and control or standardize the time of treatment initiation. All these shortcomings increase the risk that the studies overestimate the positive effects of a particular treatment. But the small number of studies adhering to more stringent methods do offer valuable insights.

Most treatments studied in randomized controlled trials fall into one of three categories: corticosteroid treatment, antiviral therapy, or treatment of vascular insufficiency. The justification for steroid treatment is a presumed inflammation in the inner ear. Such inflammation might arise from a viral infection, an autoimmune mechanism, or even as a sequel to chemical changes associated with tissue injury or death from a lack of oxygen. At the same time, though, one would also predict that steroids would be effective for treating the cellular stress pathways that Merchant and his colleagues think may be involved. In other words, steroid therapy is not specific; it may be beneficial in cases of differing causes. Antiviral therapy, by contrast, specifically addresses the viral theory of SSNHL. The third category of treatments, addressing supposed vascular insufficiency, has been attempted with therapies such as anticoagulants specific to that hypothesized cause.

The first, and still most frequently cited, randomized controlled trial of SSNHL therapy, by William Wilson, M.D., Frederick Byl, M.D., and Nan Laird, Ph.D., was published in 1980. In that trial, 67 patients with SSNHL at the Massachusetts Eye and Ear Infirmary and at a Kaiser Permanente Hospital in southern California were randomly placed in two groups that received either oral steroids or a placebo, and results from the two sites were pooled. A third group of 52 patients, who did not qualify for the randomized study for various reasons, were followed as an untreated control group. The studies used strict measurements of hearing, giving the results statistical significance despite the relatively small sample size. The study showed a strong effect of treatment with steroids and a clear relationship between the degree of hearing loss with which patients began and the success of treatment. Depending on the initial degree of hearing loss, some patients had excellent hearing recovery regardless of treatment, some responded positively to the steroids, and some showed little or no hearing improvement at all. Those in the “steroideffective zone” recovered more than twice as well as patients who received the placebo. Overall, 61 percent of the patients treated with steroids showed hearing improvement, compared with only 32 percent in the placebo group. This study has driven the widespread (although not universal) adoption of oral steroid therapy as the standard for treating SSNHL.

What about treatments based specifically on either the viral theory or the vascular theory? Four randomized controlled trials compared the use of steroids alone with the use of steroids plus antivirals, but none showed an added benefit from the antiviral medications. Seven randomized controlled trials have tried various treatments for vascular insufficiency. In one study with a steroid treatment comparison group, the vascular treatment was no more effective than treatment with oral steroids, whereas the other six studies showed no benefit from any of the treatments compared with a placebo. Other studies have tried filtering patients’ blood to remove clotting factors, using magnesium supplements, and giving patients ginkgo biloba extract. Although these treatments all had some positive effects, the results were no better than with steroids, and the studies did not have all the necessary controls.

As a result, in the 25 years since Wilson, Byl, and Laird reported their landmark study of steroid therapy, the conventional wisdom has remained that primary therapy for SSNHL is best achieved with oral steroids, that a window of opportunity exists during which steroids may be beneficial, and that little chance of hearing recovery remains after four weeks of deafness or after completion of primary therapy.

In Search of Better, Faster, and Safer

So things seemed to stand, until recently. Then several studies were published that challenged these long-held assumptions, suggesting that steroids given by injection into the ear, called “intratympanic steroid injection,” could achieve a higher concentration of the drug in the inner ear than could oral administration and that this might salvage more hearing than oral steroid therapy. The use of the injection was first studied in guinea pigs by Lorne Parnes, M.D., and his colleagues, who later showed the efficacy of this approach in some cases of SSNHL in humans.⁷ The steroid, injected into the middle ear in a simple office procedure, will partially diffuse into the inner ear fluids. Parnes reported some hearing improvement in 6 of 12 patients with SSNHL treated this way. Since then, a handful of other uncontrolled studies on intratympanic steroid injection for SSNHL has suggested that treatment is beneficial in one-half to three-fourths of cases, similar to the results of using oral steroids.

For all its apparent promise, however, the actual success rate (and rate of side effects and complications) is not known for intratympanic steroid injections as compared with oral treatment, because no studies have been published with control populations treated with conventional oral dosing. To produce the evidence and answers needed, a national multicenter clinical trial sponsored by the National Institutes of Health (NIH) is now underway.

The study brings together eight clinical centers and a data management center to perform the direct comparison of oral versus intratympanic steroid administration for primary treatment of SSNHL. Enrollment of patients began in December 2004 and will continue for approximately three years until 250 subjects are enrolled nationwide.

Part of the excitement about the potential of intratympanic steroid treatment arises from the discovery that some patients responded to the treatment six weeks after the onset of their SSNHL, when the treatment window for oral steroids is believed closed. In addition, some apparently responsive patients had already tried oral steroids without hearing improvement. The rationale for intratympanic steroid injections seems obvious: Instead of diluting a large quantity of oral steroid through the patient’s bloodstream and hoping that an adequate dose reaches the inner ear, why not put a little bit of steroid right where it’s needed? In theory, the systemic risk of the steroids will be much lower if only a small amount is injected versus a larger amount ingested orally. Perhaps a dose in the inner ear would cause even more recovery in patients who respond to oral steroids or would be effective in more patients.

Logical though the idea may be, it is well to remember that steroids are not chicken soup—if some is good, more is not necessarily better. This effect can be seen in cell cultures, for example. If small amounts of certain steroid drugs are added to the nutrients given to cell cultures, the cell growth is dramatically increased. But, if a higher dose of steroids is added, the cells die. Although I have not read published reports of patients becoming deaf after intratympanic steroid injections, a few such cases have been reported anecdotally. Until results of the new eight-center study are published, it is impossible to know whether such cases represent a complication of the steroid treatment or the natural history of further ear degeneration seen in the occasional patient with SSNHL. Another concern is that injecting the medication into the ear causes perforations of the eardrum, which may not always heal, and other unanticipated complications may occur.

Despite this uncertainty, the past few annual meetings of the American Academy of Otolaryngology and Head and Neck Surgery have seen a proliferation of courses on how to administer steroids by the intratympanic route. Already, billing codes enable physicians to obtain lucrative payment for this treatment, even though its efficacy has not been shown to be greater than that of oral treatment—for which, by the way, the physician receives no special reimbursement—and consensus has not developed on which steroid to use and at what dose for intratympanic treatment. This is a real dilemma. A promising new treatment is available that may offer real advantages to patients, but the clinical and financial pressures to offer the treatment have outpaced the scientific support for it.

Salvaging Hearing Loss

I began with the common scenario of the patient with sudden hearing loss who delays too long in obtaining an accurate diagnosis, and effective treatment for SSNHL. The window of opportunity for treating SSNHL appears to be no more than two to four weeks. After that, it is too late. Who is to blame for this tragic delay? No one, really. In most cases, the patient would have been correct in initially assuming that the blockage was from the head cold and then the second assessment that it was due to earwax. Likewise, most patients contacting their primary care physician about ear blockage do have earwax or middle ear fluid or infection, for which recommendations of decongestants, antihistamines, or antibiotics are effective. It seems reasonable for patients and doctors to act in this fashion, on the probabilities.

On the other hand, if this same patient had called the doctor to report some chest pain after shoveling snow, you can be sure he or she would have been summoned urgently to have an electrocardiogram to rule out a heart attack. One consequence of delaying that test could be death of the patient. Not so with hearing loss. At $300 or more for an audiogram, testing every patient with a stuffy ear would be prohibitively expensive. We do know, however, some simple questions that one might ask the patient (even over the telephone during the patient’s first call) and some simple points of the physical examination can reliably diagnose a sensorineural hearing loss without the audiogram. Patients can be asked to speak out loud (for example, count from 1 to 10), and then asked whether they hear their own voice in the “bad ear” or the “good ear.” If the patient’s voice is louder in the bad ear, he has a conductive hearing loss, such as ear wax or middle ear fluid. If the patient’s voice is louder in the good ear, he has nerve damage in the bad ear, probably due to SSNHL. During a physical exam, the physician can look for three different signs that SSNHL is responsible. If the ear canal and ear drum look normal, the ear drum moves normally when puffed with air, and, when a ringing 512 Hz tuning fork is pressed in the middle of the forehead, it is heard louder in the good ear, the physical signs point to SSNHL. If people knew to seek immediate evaluation for ear blockage, and if primary care doctors, their nurses, and the staffs of emergency rooms were more aware of the possibility of SSNHL, more hearing might be salvaged in these patients.

In summary, whatever the true cause of SSNHL and exactly what inner ear cells are affected, some of the cells remain healthy, some are injured by the disease process, and some are killed. At the onset of the SSNHL, the largest group is probably the injured cells. These injured cells are teetering on a fence; as time passes, they fall off on one side or the other—get well or die. These uncommitted cells, the ones on the fence, can be affected by treatment. Patients with moderate-to-severe SSNHL often can regain some hearing if they are treated with oral steroids, although the window of opportunity for this treatment is usually less than four weeks. Oral steroid therapy is the only treatment shown to be effective through rigorous randomized controlled trials, but evidence is building that intratympanic steroid injections may not only be equally effective but also offer fewer risks. The multicenter clinical trial sponsored by the NIH should give a clear comparison of the two treatments in the next three to five years.

Science in Service of Changing Care

One of the most frustrating experiences for ear specialists is seeing patients with SSNHL who suffered delay in diagnosis and missed the window of opportunity for oral steroid treatment. These patients are often desperate to try any measure that might regain some of their lost hearing. Intratympanic steroid treatment might be an answer for some of them. Despite the absence of definitive studies about its benefit and risk, intratympanic steroid delivery is rapidly spreading as a treatment approach for inner ear diseases. Soon, it will become entrenched, even though scientific evidence of its efficacy is, at this point, still lacking.

Only randomized controlled clinical trials such as the multicenter NIH study can yield data on which to base sound medical decision making. The study, therefore, comes at a critical time, when its outcome can have a powerful effect on clinical practice. One less obvious but pivotal outcome of the study will be to publicize this disorder and its treatment, both to physicians and to the public. By conforming to the rigorous methodologic standards required for publication in a leading general medical journal, such as the New England Journal of Medicine, the results should reach an audience far beyond specialists, raising awareness of SSNHL in the medical community. Related publicity and attention could lead to earlier evaluation and diagnosis, and so earlier treatment. The road has been taken and at its end lies the promise of saved hearing, and prolonged healthy functioning, for far more people in decades to come.

---

---

Sidebar: Banking Bones to Save Hearing

An important tool for studying disorders of hearing and balance is post-mortem examination under the microscope of human temporal bones. In principle, this is not new; for centuries, microscopy of the ear has shown the normal anatomy and the pathologic changes arising from diseases and disorders. Early in the 20th century, scientists began collecting and preserving temporal bone and developing and standardizing methods of measuring and counting inner-ear cells.

Since 1960, the Massachusetts Eye and Ear Infirmary has collected detailed medical records from all donors of temporal bone to enable investigators to correlate medical details of donors’ lives with the tissue changes observed in post-mortem examination. Almost 2,000 temporal bones have been collected. Similar efforts are underway at other medical centers as part of a national temporal bone bank system. In 1992, the National Institute of Deafness and Other Communication Disorders began sponsoring a National Temporal Bone and Brain Pathology Resource Registry at the Massachusetts Eye and Ear Infirmary. A consortium of temporal bone laboratories around the United States, the Registry takes background, contact, and medical information from anyone interested in donating their temporal bones and brains for post-mortem study. Today, the Registry has more than 5,000 donor pledges on file. Since 1992, it has procured more than 380 temporal bone specimens. Registry activities have supported more than 300 medical and scientific publications by scientists in the United States. Information about the Registry and about new donor pledges is available at www.tbregistry.org.