Autoimmune inner ear disease: Steroid and cytotoxic drug therapy. (Original Article).

Abstract

The goal of this study was to assess the effects of immunosuppressive therapy on hearing in patients with presumed autoimmune sensorineural hearing loss (AISNHL) and a Western blot assay positive for a 68 kD inner ear antigen. To achieve this objective, we conducted a retrospective review of 39 such patients who were treated with either a steroid alone or with a steroid followed by a cytotoxic agent. Pure-tone average (PTA) at 500 Hz, 1 kHz, 2 kHz, and 3 kHz and speech discrimination scores (SDS) were used as objective measures of outcome. At the completion of treatment, 23 of the 39 patients (59.0%) exhibited a positive response to therapy. The steroid-only responders (n = 6) tended to demonstrate a greater improvement in PTA (14.8 vs 4.5 dB), while the cytotoxic-agent responders (n = 17) had a significantly greater improvement in SDS (26.2 vs 6.9%; p<0.Ol). We conclude that most patients with AISNHL benefit from immunosuppressive therapy and that cytotoxic medications appear to improve SDS, even in some pa tients who have not responded to corticosteroid therapy.

Introduction

Autoimmune inner ear disease is most often characterized by a bilateral (BPSNHL) or unilateral progressive sensorineural hearing loss (UPSNHL) that occurs over a period of weeks to months, with or without vestibular symptoms and tinnitus. This disease can also cause sudden sensorineural hearing loss (SSNHL). (1) Both pure-tone thresholds and speech discrimination scores (SDS) can be affected detrimentally. A predilection toward females has been reported, and the disease appears to have a higher incidence in patients who have known autoimmune disease. (2)

The diagnosis of autoimmune sensorineural hearing loss (AISNHL) is ascertained by the history, clinical findings, response to immunosuppressive medication, and an immunologic evaluation of the patient's serum. Among the diagnostic tests that have been advocated are the lymphocyte transformation assay, the Western blot analysis for inner ear antibodies, and, most recently, supporting-cell immunofluorescence. (3-5) The latter has been reported to be the most sensitive.(5,6) An association between AISNHL and the presence of human leukocytic antigen (HLA) markers has also been proposed. (7) Although some controversy surrounds the interpretation of the Western blot assay for cochlear autoantibodies, the literature referenced elsewhere in this article, as well as our own clinical experience, suggests that it is one of the better tools available for establishing the diagnosis.

Spontaneous improvement or resolution of untreated sensorineural hearing loss does not appear to occur often, at least in patients who experience a progressive loss. The most common outcome is a steady deterioration of both pure-tone average (PTA) and SDS values. However, most patients with suspected AISNHL confirmed by a positive Western blot assay do benefit from immunosuppressive medications, which are the mainstay of the treatment of AISNHL, just as they are for other immune-response-mediated diseases. Patients who experience an objective response to treatment tend to enjoy subjective improvement, as well.

McCabe first reported the benefits of using a steroid and cyclophosphamide in l979. (8) Seven years later, Hughes et al advocated a steroid trial in cases of suspected AISNHL; they also recommended that cytotoxic agents be reserved for refractory cases. (2) Harris and Ryan endorsed an initial 1-month steroid trial, which could be continued for an additional 1 to 2 months for patients who responded to treatment, as long as their hearing remained stable. (9) Sismanis et al administered methotrexate to 23 patients with immune-response-mediated cochleovestibular disorders and found that 16 (69.6%) responded favorably; side effects were minor and reversible. (10)

Whereas most authors agree that a steroid trial is warranted for patients with suspected AISNHL, the role of cytotoxic therapy has not been defined fully. The patients in the present study were treated with a steroid first. In time, those who were unresponsive and those who did respond but whose hearing deteriorated as the steroid was tapered and discontinued were switched to a cytotoxic drug.

Throughout the course of the treatment of these patients, our diagnostic and therapeutic strategies continued to evolve as paradigms were tested and modified. Today, we use both a steroid and a cytotoxic drug as an initial treatment for a high percentage of our patients, as we will report in a future publication.

Patients and methods

We studied a group of 62 consecutive patients--32 males and 30 females, aged 5 to 87 years (mean: 50)--who were seen at the practice of the senior author (R.T.S.) for evaluation of suspected autoimmune inner ear disease and whose Western blot assays were positive for a 68 kD inner ear antigen.

Assessment of hearing. Hearing loss was confirmed in 56 of the 62 patients (90.3%); tinnitus was present in 44 of these patients (71.0%), vertigo in 30(48.4%), Meniere's syndrome in seven (11.3%), and Raynaud's disease in two (3.2%). Thirteen patients (21.0%) exhibited positive responses to at least one other autoimmune laboratory test--antinuclear antibody (n = 9), rheumatoid factor (n = 3), antimicrosomal antibody (n = 2), and anticardiolipin antibody (n = 1).

Of the 56 patients with confirmed hearing loss, 26 had BPSNHL (46.4%), 15 had UPSNHL (26.8%), and 15 had SSNHL. Of the six patients who did not have hearing loss (and who thus did not undergo drug treatment), three had tinnitus and fluctuating vertigo, two had fluctuating vertigo only, and one had unilateral tinnitus.

Patients were suspected of having AISNHL if they demonstrated at least a 30-dB loss at three or more frequencies between 250 and 8,000 Hz (with or without a decrease in SDS) in addition to a subjective hearing loss in the affected ear(s) at the time of the initial audiogram. AISNHL was considered to be active if hearing at any frequency between 250 and 8,000 Hz deteriorated by at least 10 dB within 3 months or if there was a decrease in SDS of at least 20% within 3 months. SSNHL was considered to be active if its onset had occurred within the previous 1 year.

Of the group of 62 patients, 14 were found to have an inactive hearing loss, and thus they were not treated with immunosuppressive medications; four others were lost to followup before their active or inactive status could be determined. Of the 44 patients who were identified as having an active hearing loss, 39 agreed to undergo treatment with immunosuppressive therapy (the remaining five patients declined treatment).

Among the group of 39 patients, 20 (51.3%) had BPSNHL, 11(28.2%) had UPSNHL, and eight (20.5%) had SSNHL.

BLA A, B, and C typing data on 44 patients were available for analysis, and HLA DR data on 36 patients. The presence or absence of four specific antigens--HLA B35, CW4, CW7, and DR4--was determined so that we could analyze the proposed association with AISNHL (the presence of B35, CW4, and CW7 and the absence of DR4). (7) HLA B35, CW4, and CW7 were present in 13 (29.5%), 14 (31.8%), and 22 (50.0%) of the 44 patients, respectively. HLA DR4 was absent in 29 of the 36 patients (80.6%).

Finally, all patients had a cerebellopontine angle tumor ruled out by gadolinium-enhanced magnetic resonance imaging of the brain and internal auditory canals.

Drug treatment. Initially, all 39 patients received an oral steroid; 34 received approximately 1 mg/kg/day of prednisone for 2 to 12 weeks, four patients received an equivalent dosage of dexamethasone for 2 weeks, and one patient received an equivalent dosage of methylprednisolone for 2 weeks.

After the initial steroid trial, cytotoxic treatment was offered to those patients who did not improve and to those who did respond initially but whose improvement was not sustained as the steroid was tapered and discontinued. A total of 27 patients elected to receive a cytotoxic agent. Of these, 21 were started on 7.5 to 15 mg/wk of oral methotrexate for at least 6 months unless toxicity or a lack of response necessitated an earlier termination of treatment; the remaining six were started on oral cyclophosphamide at 100 mg twice a day.

Those who did not respond to methotrexate or who relapsed during therapy were offered the opportunity to switch to oral cyclophosphamide or oral azathioprine, both at 100 mg twice a day. Those who did not respond to or relapsed during initial cyclophosphamide treatment were offered either methotrexate or azathioprine at the previously specified dosages. The duration of cyclophosphamide and azathioprine treatment was determined by the audiologic response.

By the completion of the study, 12 of the 39 patients (30.8%) had received a steroid only, and 27 (69.2%) had received a steroid followed by a cytotoxic agent. Ten patients received more than one cytotoxic drug, but none received more than one such agent at a time. Every effort was made to wean all patients off their steroid before or during cytotoxic therapy.

Potential medication side effects were reviewed with all patients prior to the start of cytotoxic therapy, and women of child-bearing age were informed that these medications are teratogenic. All patients underwent periodic complete blood cell count measurements and liver and renal function studies during treatment with a cytotoxic agent.

Air-conduction (250 to 8,000 Hz) and bone-conduction thresholds (250 to 4,000 Hz) were measured during the initial visit, at least bimonthly during treatment, and after the completion of treatment in all patients. SDS values were obtained at the same time. The PTA at 500 Hz, 1 kHz, 2kHz, and 3kHz was calculated for all patients prior to therapy, during any response period, and post-therapy. Subjective improvement in hearing and alleviation of vertigo and tinnitus were also recorded for each patient.

Assessment of response. Three criteria were used to assess each patient's response to drug treatment:

* Criterion I. An improvement of 10 dB or more in at least three frequencies between 250 and 8,000 Hz.

* Criterion II. An improvement of 15 dB or more in at least two frequencies between 250 and 8,000 Hz.

* Criterion III. An improvement of at least 20% in SDS.

To be considered a responder, a patient had to meet one of these three objectives in at least one ear for (1) at least 2 consecutive weeks at some point following the completion of steroid therapy or (2) a minimum of 2 months following the cessation of cytotoxic therapy. The result of each patient's initial audiogram was used as the baseline from which any improvement or decline in pure-tone threshold or SDS could be measured. A paired two-tailed Student's t test was used to detect any statistically significant differences (p<0.05) in pre- and post-treatment PTA and SDS values.

Control group. To establish a basis for comparison, we reviewed our records for the earliest available PTA and SDS data on the 39 patients who eventually underwent drug treatment. Adequate pretreatment audiometric data were available on 15 of these patients (22 ears), and they essentially served as their own controls during an extended period of time prior to the initiation of treatment. Ten of these patients had had audiometric data measured at 3-month intervals, three at 6-month intervals, and two at 1-year intervals.

The duration of evaluation of these patients prior to the start of this study ranged from 6 weeks to 15 years (mean: 2.7 yr). The duration of the documented hearing loss was greater than 3 years in six ears and less than 3 years in 16 ears. Pretreatment audiometric data were available on seven patients who were awaiting their Western blot assay results. Four patients had been referred from outside otolaryngologists without having undergone previous immunosuppressive therapy. Of these four referrals, one patient had received conservative medical treatment for suspected Meniere's disease, one was presumed to have cochlear otosclerosis and was treated with auditory rehabilitation, and two had undergone auditory rehabilitation only.

During the early 1980s, we had treated two of these 15 patients for suspected Meniere's disease; we administered conservative medical therapy followed by endolymphatic sac decompression. Both of these patients subsequently underwent Western blot analysis when it became available. One of our patients had undergone surgical treatment for bilateral chronic otitis media, and another had undergone a congenital atresia repair with resection of cholesteatoma prior to Western blot analysis. In addition to the early audiometric data on these 15 patients who underwent treatment, we also obtained adequate early data on four of the five eligible patients who declined treatment.

Results

Overall, a positive response was achieved by 23 of the 39 patients (59.0%)--six of 12 (50.0%) in the steroid-only group and 17 of 27 (63.0%) in the cytotoxic medication group (figure 1; table 1). Of these 23 responders, 10 (43.5%) had BPSNHL, nine (39.1%) had UPSNHL, and four (17.4%) had SSNHL.

Among the 27 patients who went on to cytotoxic therapy, 21 (77.8%) had failed steroid therapy outright; the other six patients (22.2%) had responded initially, but their hearing later deteriorated as the steroid was tapered and discontinued. Therefore, a total of 12 of the 39 patients (30.8%) did exhibit some initial response to steroid therapy, although the response was sustained in only six (15.4%).

Prior to the initiation of treatment, the mean hearing deficit had been more severe (i.e., higher PTA and lower SDS values) among those who ultimately required cytotoxic therapy than in those who received a steroid only (table 2). Following cytotoxic therapy, the 17 patients who responded experienced a modest improvement in mean PTA (4.5 dB) and a more pronounced improvement in mean SDS (26.2%) (figures 2 and 3). Conversely, the six responders in the steroid-only group had a substantial improvement in mean PTA (14.8 dB) and a small improvement in SDS (6.9%).

Subjective assessments of hearing improvement were available for 21 of the 23 patients who responded to treatment. Subjective improvement was reported by 19 of these 21 responders (90.5%). Alleviation of symptoms was reported by 10 of the 20 patients (50.0%) who had pretreatment vertigo and by seven of the 34 (20.6%) with pretreatment tinnitus.

There was no report of subjective improvement by any of the 16 patients who did not respond to either therapy. Their mean pre- and post-treatment PTA values were 64.8 and 67.6 dB, respectively, and their corresponding SDS values were 49.6 and 47.5%.

Steroid. Among the 12 patients who received only a steroid, six had BPSNHL, two had UPSNHL, and four had SSNHL. The duration of steroid therapy ranged from 2 weeks to 3 months (mean: 3 wk).

Among the six patients whose treatment was successful (a total of seven treated ears), the onset of a positive response was observed between 3 and 29 days (mean: 14) after the initiation of treatment. The six patients in the steroid-only group who failed therapy declined further treatment.

Methotrexate. Methotrexate was administered to 21 patients following the failure of steroid therapy and to two patients following steroid and cyclophosphamide therapy. Of these 23 patients, 13 had BPSNHL, eight had UPSNHL, and two had SSNHL. The duration of methotrexate treatment ranged from 1 to 30 months (mean: 7.3).

Among the 14 patients (60.9%) whose treatment was successful (17 ears), the onset of a positive response was seen between 6 days and 6 months (mean: 42 days) after the initiation of methotrexate treatment.

Of the two patients who received methotrexate as a second-line cytotoxic agent, one had responded to cyclophosphamide initially, relapsed when it was discontinued, and then responded to methotrexate. The other patient failed both cytotoxic agents.

Cyclophosphamide. In all, cyclophosphamide was given to 10 patients-six as an initial cytotoxic medication and four as a followup to methotrexate. Among the four patients in the latter group, two had responded initially to methotrexate but were forced to discontinue treatment because of elevated liver enzyme levels, one had responded but then relapsed, and one failed methotrexate outright. Three of these four patients responded subsequently to cyclophosphamide, including two previous methotrexate responders and the one methotrexate failure. Of the 10 cyclophosphamide patients, six had BPSNHL, two had UPSNHL, and two had SSNHL. The duration of cyclophosphamide treatment ranged from 1 to 9 months (mean: 2.7).

Among the five patients (50.0%) whose treatment was successful (seven treated ears), the onset of a positive response was seen between 30 and 90 days (mean: 58) after the initiation of cyclophosphamide treatment.

Azathioprine. Seven patients received azathioprine, none as a first-line cytotoxic medication. One of these patients had previously responded to both methotrexate and cyclophosphamide but was forced to discontinue both drugs because of elevated liver enzyme levels; this patient subsequently responded to azathioprine. Another patient had also been treated previously with methotrexate and then cyclophosphamide but relapsed after an initial response to each; this patient also responded positively to azathioprine. The other five patients had been treated previously with methotrexate only. Two of them had responded initially to methotrexate but relapsed when it was discontinued; both of these patients responded subsequently to azathioprine. The other three patients had failed methotrexate therapy; one responded subsequently to azathioprine. Of the seven azathioprine patients, three had BPSNHL, three had UPSNHL, and one had SSNHL. The duration of azathioprine treatment ranged from 6 weeks to 20 months (mean: 7 mo).

Among the five patients (71.4%) whose treatment was successful (five ears), the onset of a positive response was seen between 30 and 90 days (mean: 51) after the initiation of azathioprine treatment.

Steroid vs cytotoxic therapy. We used the paired two-tailed Student's t test to compare any significant differences in changes in PTA and SDS that were observed during the initial steroid trial with those that were observed following cytotoxic therapy. All ears treated with both a steroid and a cytotoxic agent were included in this analysis. Any significant (according to the criteria established earlier) decrease in PTA or increase in SDS was considered to represent improvement.

During the initial steroid trial, hearing in all ears generally deteriorated; the mean PTA rose 0.85 dB and the mean SDS fell 3.3% for all ears. Conversely, overall hearing improved during cytoxic therapy; the mean PTA fell 1.9 dB and the mean SDS rose 11.1%. The difference in PTA before and after cytotoxic therapy approached but did not reach statistical significance (p = 0.056), while the difference in SDS was statistically significant (p<0.01).

Methotrexate. We also evaluated each cytotoxic medication individually. During steroid therapy, all ears that were eventually treated with methotrexate as the initial cytotoxic agent experienced a mean decrease in PTA of 0.68 dB and a mean decrease in SDS of 1.4%. Following methotrexate therapy, the mean PTA fell 2.6 dB and the mean SDS rose 11.5%. The difference in PTA was not statistically significant (p = 0.27), but the improvement in SDS was (p<0.01).

Cyclophosphamide. During steroid therapy for ears that were eventually treated with cyclophosphamide as a first-line cytotoxic agent, the mean PTA rose 4.4 dB and the mean SDS fell 9.5%. Following cyclophosphamide therapy, the mean PTA fell 1.3 dB (p = 0.13) and the mean SDS rose 10.5% (p<0.01).

Azathioprine. The mean PTA rose 0.40 dB and the mean SDS remained unchanged following steroid therapy. Following azathioprine treatment, the mean PTA fell 0.33 dB (p = 0.749) and the mean SDS rose 10.4% (p = 0.061).

Side effects. Sixteen of the 39 treated patients (41.0%) reported a total of 25 medication side effects (table 3). All of these adverse events were resolved by discontinuing the offending agent and by initiating appropriate medical intervention when necessary.

Control group. The pretreatment audiometric analysis is a crude control tool, but it helps define the natural history of AISNHL. We compared the early PTA and SDS values of the 15 patients in the control group with those that were later obtained during audiometric testing administered just prior to the institution of therapy (table 4); we did the same for the four controls who refused treatment, using the data obtained at their final visit (table 5). The early audiograms were also evaluated to determine whether any of our three response criteria had already been met by any of these controls during the pretreatment period.

Of the 15 controls who underwent therapy, seven had BPSNHL and eight had UPSNHL (total: 22 ears). On average, these control patients did not experience any spontaneous improvement. During the pretreatment period, their mean PTA rose 4.7 dB and their mean SDS dropped 5.8%. Only one of the 22 ears (4.5%) spontaneously met any of the response criteria for audiometric improvement, and that improvement lasted only 2 months.

It is interesting that of the six ears that had had a documented hearing loss for at least 3 years, only one (16.7%) responded to immunosuppressive therapy. Conversely, 15 of the other 16 ears (93.8%) met at least one of our response criteria following immunosuppressive therapy. Thus, as McCabe has suggested, there might be a window of opportunity to treat AISNHL before irreversible inner ear damage occurs. (8)

The early audiometric data on the five eligible patients who declined treatment revealed that one had BPSNHL, two had UPSNHL, one had bilateral SSNHL, and one had unilateral SSNHL. Data on four of these patients were adequate to include in our analysis.

The patient with unilateral SSNHL reported a spontaneous resolution of the hearing loss prior to our evaluation, which was confirmed by audiometric testing in our office. However, the original audiogram was not available for our review, and therefore this patient was not included in our final analysis.

The patient with bilateral SSNHL had a flat, moderate loss on the right and a flat, mild loss on the left; her speech discrimination was preserved, and she did not experience any vertigo. Over the course of 1 month after she had declined treatment, this patient experienced a spontaneous and complete resolution of her hearing loss in both ears. Prior to being lost to followup, she had satisfied response criteria I and II in both ears for a period of 2 months.

One of the two patients with UPSNHL had a moderate loss on the right and an SDS of 88%. Over 5 months, her hearing had improved to the point that it was only mild, and her SDS had risen to 100%; even so, this patient met none of the response criteria.

The other patient with UPSNHL had a mild hearing loss and an SDS of 88%. Over the following year, his hearing had progressively deteriorated until it had become moderately severe, and his SDS had fallen to 52%.

The patient with BPSNHL had a moderate loss in both ears; her SDS values on the right and left were 84 and 88%, respectively. No further decline in her hearing was noted over the 3 months prior to her loss to followup.

Discussion

The presence of antibodies to inner ear antigens in patients with sudden or rapidly progressive sensorineural hearing loss supports the concept that an autoimmune response is responsible for inner ear disease in these patients. In 1987, Harris confirmed that guinea pigs immunized with bovine cochlear tissue experienced sensorineural hearing loss. (11) Further evidence has suggested that AISNHL is associated with the presence of an antibody to a 68 kD inner ear protein (isolated by Western blot assay). (9) Although it had been suggested that the cochlear antigen in question is heat shock protein 70 (HSP 70), Trune et al reported that mice inoculated with HSP 70 did not exhibit a shift in auditory brainstem response despite the fact that they mounted an adequate immunologic response. (12) Harris and Ryan found that the sera of 32% of patients with a rapidly progressive sensorineural hearing loss showed a reaction to a 68 kD antigen. (9) Hughes et al found that 86% of their suspected AISNHL patients had sera tha t reacted with a 68 kD antigen, (13) and Moscicki et al reported that 58% of their patients possessed such antibodies. (14)

Nair et al reported that the monoclonal antibody KHRI-3, which binds to an inner-ear-supporting cell antigen, can lead to sensorineural hearing loss in animals that have high KHRI-3 serum titers and in those that receive antibody infused directly into the inner ear. (15,16) Disher et al reported that KHRI-3 precipitates a 68-70kD antigen that binds human antibodies in the sera of suspected AISNHL patients. (17) Disher et al also showed that sera from patients with AISNHL contain antibody to inner-ear-supporting cells, and they hypothesize that this is the same antigen as that defined by KHRI-3.

McCabe was the first to recognize AISNHL. (8) Although the treatment of suspected AISNHL has not been well defined, steroids and cytotoxic immunosuppressants have been advocated most often. McCabe reported that a response to steroids and cyclophosphamide is diagnostic, although treatment can fail if irreversible cochlear damage has already occurred. Hughes et al reported that 45 of 52 patients (86.5%) with suspected AISNHL experienced either improvement or stabilization of hearing following treatment with steroids, cytotoxic agents, or lymphocytoplasmaphoresis. (2) Moscicki et al reported that 27 of 31 patients (87.1%) who had antibodies to a 68 kD protein improved with steroid treatment. (14) Most recently, Zeitoun et al reported that 25 of 47 patients (53.2%) with suspected AISNHL who had antibodies positive for inner-ear-supporting cells responded to steroid therapy. (5)

In the present study, we did not find a female preponderance to suspected AISNHL, as has been suggested previously. (2) We did find that 21.0% of our patients tested positive on autoimmune laboratory testing, a figure that is consistent with others reported in the literature. (2,9) Our HLA findings that B35, CW4, and CW7 antigens were present in 29.5, 31.8, and 50.0% of tested patients, respectively, and that DR4 was absent in 80.6% of tested patients are similar to those previously reported by Bowman and Nelson, (7) and they further support the proposed association of these antigens with AISNHL.

It is interesting to note that the six patients in the steroid-only group who responded to treatment had, on average, a lesser degree of hearing loss prior to therapy (PTA: 48.1 dB; SDS 69.1%) than did those who responded to cytotoxic therapy (PTA: 64.2 dB; SDS: 47.9%). The steroid responders also tended to experience a greater improvement in mean PTA (14.8 dB) than did those who responded to cytotoxic medications (4.5 dB). Conversely, the cytotoxic responders had a greater improvement in mean SDS (26.2%) than did the steroid responders (6.9%). Improvements in PTA during cytotoxic therapy were not significantly different (p = 0.056) from the changes observed in these patients during their steroid trial. However, the improvement in SDS was significant (p<0.01). This finding suggests that chance is a highly unlikely explanation for the improvement in SDS during cytotoxic therapy.

The mean duration of response according to criterion III (20% improvement in SDS) in the cytotoxic group was 10.4 months. However, no conclusions can be drawn concerning a comparison of SDS response duration between the cytotoxic and steroid groups because of (1) the short interval that passed between the beginning of the steroid trial and the initiation of cytotoxic treatment (~1 mo) and (2) the fact that only one patient in the steroid-only group met criterion III.

Overall, patients tolerated therapy well. Only three cases of a major side effect were observed during this study; one patient experienced hemorrhagic cystitis while taking cyclophosphamide, and one patient each experienced lymphoblastic vasculitis and pancytopenia/sepsis while taking azathioprine. All major side effects resolved upon termination of therapy and appropriate medical intervention. Although methotrexate had to be discontinued in five patients because of a rise in liver enzyme levels, azathioprine appeared to be the most toxic of the four medications used in the study.

The risks of cytotoxic agents require further study in the research realm and careful consideration in clinical practice. Patient compliance with otologic, audiologic, and laboratory followup protocols is essential to ensure the safe use of cytotoxic agents as well as steroids. Additional research into the natural history of AISNHL and its diagnostic and treatment strategies should be encouraged.

References

(1.) Hughes GB, Kinney SE, Barna BP, Calabrese LH. Practical versus theoretical management of autoimmunne inner ear disease. Laryngoscope 1984;94:758-67.

(2.) Hughes GB, Barna BP, Kinney SE, et at. Clinical diagnosis of immune inner-ear disease. Laryngoscope 1988;98:251-3.

(3.) Hughes GB, Barna BP, Kinney SE, et al. Predictive value of laboratory tests in "autoimmune" inner ear disease: Preliminary report. Laryngoscope 1986;96:502-5.

(4.) Harris JP, Sharp PA. Inner ear autoantibodies in patients with rapidly progressive sensorineural hearing loss. Laryngoscope 1990;100:516-24.

(5.) Zeitoun H, Phil M, Gray J, et al. The prognostic significance of serum antibodies to inner-ear supporting cell antigens in progressive or sudden hearing loss. Laryngoscope. Submitted for publication.

(6.) Carey TE, Nair TS, Gray JP, et al. The search for the inner ear antigen of autoimmune sensorineural hearing loss. In: Veldman JE, Passati D, Lim DJ, eds. New Frontiers in Immunobiology. The Hague: Kugler, 2000:67-74.

(7.) Bowman CA, Nelson RA. Human leukocytic antigens in autoimmune sensorineural hearing loss. Laryngoscope 1987;97:7-9.

(8.) McCabe BF. Autoimmune sensorineural hearing loss. Ann Otol Rhinol Laryngol 1979;88:585-9.

(9.) Harris JP, Ryan AR Fundamental immune mechanisms of the brain and inner ear. Otolaryngol Head Neck Surg 1995;112:639-53.

(10.) Sismanis A, Wise CM, Johnson GD. Methotrexate management of immune-mediated cochleovestibular disorders. Otolaryngol Head Neck Surg 1997;116:146-52.

(11.) Harris JP. Experimental autoimmune sensorineural hearing toss. Laryngoscope 1987;97:63-76.

(12.) Trune DR, Kempton JB, Mitchell CR, Hefeneider SH. Failure of elevated heat shock protein 70 antibodies to alter cochlear function in mice. Hearing Res 1998;116:65-70.

(13.) Hughes GB, Moscicki R, Barna BP, San Martin JE. Laboratory diagnosis of immune inner ear disease. Am J Otol 1994;15:198-202.

(14.) Moscicki RA, San Martin JE, Quintero CH, et at. Serum antibody to inner ear proteins in patients with progressive hearing loss. Correlation with disease activity and response to corticosteroid treatment. JAMA 1994;272:611-6.

(15.) Nair TS, Raphael Y, Dolan DF, et al. Monoclonal antibody induced hearing loss. Hearing Res 1995;83:101-13.

(16.) Nair TS, Prieskom DM, Miller JM, et at. In vivo binding and hearing loss after intracochlear infusion of KHRI-3 antibody. Hearing Res l997;107:93-l01.

(17.) Disher MJ, Ramakrishnan A, Nair TS, et al. Human autoantibodies and monoclonat antibody KHRI-3 bind to a phylogenetically conserved inner-ear-supporting cell antigen. Ann N Y Acad Sci 1997;830:253-65.

[Figure 1 omitted]

Figure 2.

Mean PTA value before and after treatment in ears that responded
to various therapeutic regimens.


                       Pre-Tx PTA  Response PTA

All Patients              64.1         55.7
Steroids & cytotoxics     64.2         59.7
Steroids only             48.1         33.3

Note: Table made from bar graph
Figure 3.

Means SDS values before and after treatment in ears that responded
to various therapeutic regimens.

                             Percent
                       Pre-Tx    Response
                        SDS        SDS

All Patients             52         74
Steroids & cytotoxics    47.9       74.1
Steroids                 69.1       76

Note: Table made from bar graph
Table 1.

Response data in the different treatment groups

                  n (%) of ears meeting         Duration (mo)
Criterion         response criteria (*)       Mean        Median


Steroid only

I (+)(ss)               3/7 (42.9)             6.7         4.0
II (n)(#)               4/7 (57.1)             8.1         8.0
III (**)                1/7 (14.3)             3.5          --

Methotrexate

I                      6/17 (35.3)             6.3         3.0
II                     9/17 (52.9)             5.2         3.0
III                   12/17 (70.6)            10.5        12.0

Cyclophosphamide

I                       2/7 (28.6)            14.0        14.0
II                      3/7 (42.9)            10.0         5.0
III                     5/7 (71.4)            10.0         5.0

Azathioprine

I                       1/5 (20.0)             2.0          --
II                      3/5 (60.0)             5.5         3.0
III                     3/5 (60.0)            10.8         9.0


                      Duration (mo)
Criterion         Mode        Range


Steroid only

I (+)(ss)          --        2 to 14
II (N)(#)          --       2.5 to 14
III (**)           --          --

Methotrexate

I                  2.0       2 to 14
II                 2.0       2 to 14
III               14.0       2 to 22

Cyclophosphamide

I                  --        5 to 23
II                 --        2 to 23
III                5.0       3 to 23

Azathioprine

I                  --          --
II                 --       2.5 to 11
III                --       3.5 to 20


(*)Some ears have sustained their response past the time of data
acquisition.

(+)Criterion I: 10-dB improvement in at least three frequencies between
250 and 8,000 Hz.

(ss)Includes one patient who responded to criterion I for 2 months and
was then lost to followup.

(n)Criterion II: 15-dB improvement in at least two frequencies between
250 and 8,000 Hz.

(#)Includes one patient who responded to criterion II for 2.5 months at
the time of data acquisition and who has continued to sustain
improvement during the preparation of this report.

(**)Criterion III: 20% improvement in SDS.
Table 2.

Mean pretreatment and response PTA and SDS values for ears treated
with cytotoxic agents following steroid therapy

                                    PTA (dB)                SDS (%)
                    n (%)      Pre-Tx    Post-Tx    Pre-Tx    Post-Tx

Methotrexate

Responders        14 (60.9)     70.6      63.1       45.9      73.0
Failures           9 (39.1)     71.4      74.9       39.7      37.1

Cyclophosphamide

Responders         5 (50.0)     65.1      62.6       37.7      68.4
Failures           5 (50.0)     70.1      73.7       42.3      38.9

Azathioprine

Responders         5 (71.4)     56.8      53.4       68.8      85.6
Failures           2 (28.6)     85.0      93.0        4.0       9.3
Table 3.

Incidence of side effects

Side effect (total)          Steroids  Methotrexate  Cyclophosphamide

Hemorrhagic cystitis (*)(1)     0           0               1
Lymphoblastic
 vasculitis (*)(1)              0           0               0
Pancytopenia/sepsis(*)(1)       0           0               0

Elevated liver enzymes (7)      0           5               1
Nausea (3)                      0           3               0
Anorexia (2)                    1           0               1
Diarrhea (2)                    0           1               1
Insomnia (2)                    2           0               0

Fatigue (1)                     0           0               0
Hair Loss (1)                   0           1               0
Headache (1)                    0           0               0
Leukopenia (1)                  0           0               0
Menorrhagia (1)                 0           0               0
Weight gain (1)                 1           0               0

Side effect (total)          Azathioprine

Hemorrhagic cystitis (*)(1)       0
Lymphoblastic
 vasculitis (*)(1)                1
Pancytopenia/sepsis(*)(1)         1

Elevated liver enzymes (7)        1
Nausea (3)                        0
Anorexia (2)                      0
Diarrhea (2)                      0
Insomnia (2)                      0

Fatigue (1)                       1
Hair Loss (1)                     0
Headache (1)                      1
Leukopenia (1)                    1
Menorrhagia (1)                   1
Weight gain (1)                   0

(*)Major side effects.
Table 4.

Comparison of early and pretreatment mean PTA and SDS values among
treated controls

         PTA (db)                    SDS (%)        n (%) ears meeting
Earliest (*)  Pre-Tx (+)  Earliest (*)  Pre-Tx (+)   response criteria

    47.8         52.6         73.5         67.7         1/22 (4.5)

(*)Data obtained during the patient's initial visit.

(+)Data obtained just prior to the start of immunosuppressive therapy.
Table 5.

Comparison of early and late mean PTA and SDS values among untreated
controls

        PTA (db)                    SDS (%)        n (%) ears meeting
Earliest (*)  Final (+)  Earliest (*)  Final (+)  response criteria (ss)

    41.0        35.6         88.7        87.3         2/6 (33.3)

(*)Data obtained during the patient's initial visit.

(+)Data obtained at the final audiometric evaluation.

(ss)Adequate audiometric data were available on four of five patients in
the no-treatment group. The two ears that met our response criteria were
in a single patient with bilateral SSNHL. This patient satisfied
criteria I and II for 2 months and was then lost to followup.