Otology & Neurotology 25:833837 2004, Otology & Neurotology, Inc.

Hearing Loss in Wegeners Granulomatosis

*Sivasanker Bakthavachalam, *Mark S. Driver, *Clarke Cox, *Jeffrey H. Spiegel, *Kenneth M. Grundfast, and Peter A. Merkel
*Department of OtolaryngologyHead and Neck Surgery and the Sections of Rheumatology and Clinical Epidemiology, Department of Medicine, Boston University School of Medicine, Boston, Massachusetts, U.S.A.

Objective: To describe the frequency, type, and clinical course of hearing loss in Wegeners granulomatosis and assess hearing loss as an indicator of disease activity. Study Design, Setting, and Patients: Retrospective cohort review of all patients with Wegeners granulomatosis seen in 1 year at an academic medical center. Main Outcome Measures: Hearing loss documented by puretone audiogram. Results: Thirty-six patients were included in the analysis: 20 men and 16 women, with a mean age of 55.5 years (range, 2287 yr); 30 (83%) were antineutrophil cytoplasmic autoantibodiespositive, and the mean disease duration was 47 months (range, 2196 mo). Twenty patients (56%) had documented hearing loss: there were 17 (47%) cases of sensorineural hearing loss and 12 (33%) cases of conductive hearing loss. Seven of 12 cases of conductive hearing loss improved with immunosuppressive treatment of Wegeners granulomatosis, 2 worsened, and 3 remained stable. Of 17 cases of sensorineural hearing loss, 3 improved, 4 worsened, and 10 remained stable.

Seven patients had hearing loss requiring amplification. Five of 35 (14%) patients had established hearing loss months to years before diagnosis of Wegeners granulomatosis. Hearing loss occurred both on initial presentation and with disease relapse. The rates of conductive hearing loss (38%) and sensorineural hearing loss (31%) were also high in the subset of patients 65 years of age or younger and without history of noise exposure. Conclusions: Both sensorineural hearing loss and conductive hearing loss are common in Wegeners granulomatosis, may result in significant morbidity, and may precede the diagnosis of Wegeners granulomatosis by years. Both types of hearing loss in patients with Wegeners granulomatosis may be used as an indicator of disease. These data suggest that it may be appropriate to obtain screening audiograms in all patients with newly diagnosed or relapsing Wegeners granulomatosis. Key Words: AudiogramsHearing lossVasculitis Wegeners granulomatosis. Otol Neurotol 25:833837, 2004.

Head and neck complaints are extremely common among patients with Wegeners granulomatosis (WG), and are often present early in the disease course (13). This places the otolaryngologist in a pivotal role both in diagnosis of WG and in the detection of exacerbations. Although the nasal, tracheal, and middle ear findings of WG are well recognized by most otolaryngologists, sensorineural hearing loss (SNHL) is less appreciated. Approximately 50% of WG patients have otologic findings, including both SNHL and conductive hearing
Address correspondence and reprint requests to Peter A. Merkel, M.D., M.P.H., Vasculitis Center, E5, Boston University School of Medicine, 715 Albany Street, Boston, MA 02118, U.S.A.; Email: pmerkel@bu.edu Supported in part by research grants from The National Institute of Arthritis, Musculoskeletal and Skin Diseases. Dr. Merkel is supported in part by a Mid-Career Development Award in Clinical Investigation (National Institutes of HealthNational Institute of Arthritis, Musculoskeletal and Skin Diseases Grant K24 AR2224-01A1); the National Center for Research Resources; Boston University General Clinical Research Center Program Grant (National Institutes of HealthNational Center for Research Resources Grant MO 1RR00533); and the Boston University Wegeners Granulomatosis Research Fund.

loss (CHL) (4). The reported incidence of SNHL in WG varies, with recent reports citing rates of 2.8% (5) and 13% (6). SNHL is especially significant, because cranial nerve impairment suggests a neurodegenerative process that might warrant aggressive therapy (5). Furthermore, SNHL is believed to be largely irreversible, therefore potentially adding to the patients cumulative disability (7). Interpreting audiologic data in patients with WG can be problematic, however. For example, it is difficult to determine whether SNHL in a patient with WG is attributable to the autoimmune disorder itself or to other causes such as presbycusis or noise-induced hearing loss (NIHL). In addition, the usefulness of SNHL in guiding treatment decisions is unclear. This retrospective study of all patients with WG evaluated in 1 year at an academic medical center was undertaken to assess the significance and clinical spectrum of hearing loss in WG. PATIENTS AND METHODS Study patient eligibility criteria
Patients with WG were cared for by one rheumatologist (P. A. M.) through the Vasculitis Center at Boston University

833

834

S. BAKTHAVACHALAM ET AL.
bilized, or worsened), were also recorded. All patients underwent testing for ANCA by both immunofluorescence and ELISA techniques. Patients were considered ANCA-positive if at any time they had positive tests of cytoplasmic pattern (CANCA) coupled with antibodies to proteinase 3 (antiPR-3) by ELISA.

Medical Center during 2002. The Boston University Vasculitis Center is a tertiary referral center for inflammatory vasculitis and serves patients from the entire United Sates. The study was approved by the Boston University Institutional Review Board. Thirty-seven patients were initially considered eligible for the study. However, one patient died during the study period, and data from that patient were not included in this analysis. Two patients with mixed losses had a history of prolonged exposure to loud noise, and they were counted as having conductive loss, but were excluded from analysis of sensorineural deafness in an attempt to exclude non-WG causes of hearing loss. Diagnosis of WG was based on a modification of the 1990 American College of Rheumatology classification criteria for WG with a fifth criterion of antineutrophil cytoplasmic autoantibodies (ANCA) positivity for by measurement of antibodies to proteinase-3 (PR-3) using enzyme-linked immunosorbent assay (ELISA) (8).

Statistics
Group comparisons were made with Students t test for continuous variables and Fishers exact test for categorical variables. All tests were two-tailed with a significance level ( ) of 0.05.

Data presentation
In reporting the patient baseline data, disease duration was defined as the time from diagnosis of WG to the end of the study period (December 31, 2002). Patients with MHL were counted as having one case each of SNHL and CHL, leading to the number of cases of hearing loss exceeding the number of patients with hearing loss. Hearing loss was classified as a head and neck manifestation. Severe disease manifestations are defined as those that impart risk of permanent damage to an organ, necessitating cyclophosphamide therapy (5,8). Examples of severe manifestations include alveolar hemorrhage, renal disease, scleritis, gangrene, nervous system involvement, or SNHL. Statistical analysis was performed on the entire cohort as well as a subset of patients aged 65 years or younger without a history of noise exposure. This subset was identified as that most likely to have hearing loss unrelated to NIHL or presbycusis, because the prevalence of age-induced hearing loss dramatically increases above the age of 65 (11). Audiogram series were analyzed to define progression as improvement, no change, or worsening.

Hearing loss
Hearing loss was defined using the criteria set forth by the American National Standard Specification for Audiometers (9) as follows: CHL was defined as the presence of an air-bone gap of greater than 10 dB. SNHL was defined as both air and bone conduction below 15 dB with no significant air-bone gap (>10 dB). Mixed hearing loss (MHL) was defined as both SNHL and CHL components on the audiogram. To ensure that CHL and SNHL were treated as separate entities, patients with MHL were included by assigning then to both sensorineural and conductive hearing loss categories. All audiograms were reviewed and interpreted by three authors of this article who were trained in interpreting audiograms: a medical student (S. B.), an otolaryngology resident (M. S. D.), and the Director of the Division of Audiology (C. C.) in the Department of OtolaryngologyHead and Neck Surgery at Boston Medical Center.

Data collection
Data obtained from patients medical records included demographic information, clinical detail, and laboratory studies. Additional audiograms were obtained from outside institutions when available. Data were collected beginning with patients initial presentations of disease at Boston Medical Center or an outside institution and for each documented WG flare. Manifestations of disease other than hearing loss were tallied according to the Birmingham Vasculitis Activity Score for Wegeners Granulomatosis (BVAS/WG) (10). The major categories in the BVAS/WG evaluation form are organized by major organ systems. A flare was defined as any recurrence of signs or symptoms of disease after a period of remission (8,10). The first flare represented initial presentation with WG. Treatment dates and responses to treatment (improved, sta-

RESULTS Table 1 displays baseline data of all patients. For analysis of air conduction, 36 patients, 20 men and 16 women, diagnosed with WG were included. Thirty-four were included in the group assessed for sensorineural loss (18 men and 16 women). Ages ranged from 22 to 87 years, with a mean of 55.5 years. The mean disease duration was 35.7 months. Thirty patients (83%) tested positive for antibodies to the PR-3 antigen. Twenty-three patients underwent at least one audiogram and 20 patients (56%) had a documented hearing loss on at least one audiogram. Allowing for 9 patients who had mixed losses, the incidence of SNHL and CHL was 17 (47%) and 12

TABLE 1. Data on all study patients

All (n Mean age (yr) (range) Men (%) Women (%) Disease duration (mo) (range) ANCA/anti-PR-3 (%) Medication usage (ever) Glucocorticoids (%) Methotrexate (%) Azathioprine (%) Cyclophosphamide (%) HL, hearing loss. Otology & Neurotology, Vol. 25, No. 5, 2004 36) All SNHL (n 62 (2487) 19 (59) 8 (47) 35 (12108) 14 (77) 18 (100) 13 (72) 5 (27) 16 (88) 17) All CHL (n 58 (3981) 5 (41.7) 7 (58.3) 31 (684) 9 (75) 12 (100) 9 (75) 4 (33) 11 (91) 12) Any HL (n 60 (2487) 10 (50) 10 (50) 33 (6108) 16 (80) 20 (100) 15 (75) 5 (23) 17 (85) 20) No HL (n 16)

55.5 (2287) 20 (56) 16 (44) 36 (3120) 30 (83) 36 (100) 26 (72) 8 (22) 32 (89)

48 (2265) 10 (63) 6 (38) 39 (3120) 14 (88) 16 (100) 11 (69) 3 (19) 15 (94)

HEARING LOSS IN WEGENERS GRANULOMATOSIS (33%), respectively. The patients with hearing loss were significantly older than those without hearing loss (60 versus 48 years old, p 0.014). Two patients had audiometric patterns consistent with NIHL (characterized by a deficit at 4,000 Hz), and two of these had a history of noise exposure. Seven had audiograms suggestive of presbycusis (characterized by high-frequency loss). Seventeen of 20 patients (85%) with hearing loss tested positive for C-ANCA/antiPR-3 antibodies. Sixteen of 20 patients (80%) had bilateral hearing loss. Seven patients used hearing amplification. Five of 36 (14%) patients had established hearing loss months to years before diagnosis of WG. Hearing loss occurred both on initial presentation and with disease relapse. Comparative statistics between the subgroup of patients with SNHL and those with CHL are not possible because patients with MHL, of which there were many, contribute data to both groups. Table 2 displays data for patients aged 65 years or younger with no known history of noise exposure. In this subgroup, cases of SNHL are highly likely to be secondary to WG alone. This restriction yielded 26 patients, with a mean age of 45 years (range, 2263 yr). Eleven of 26 patients (42%) in the younger group had hearing loss, compared with 9 of 10 patients (90%) in the group older than 65 (p 0.022). SNHL and CHL prevalences were similar (31% versus 38%) in the younger group. Of the 10 patients who were older than 65 years old, 9 had SNHL (90%) and 2 had CHL (20%). Table 3 shows the nonotologic manifestations of WG in the entire cohort. These are organized by major systems according to the BVAS/WG categorization (10). Thirty-three patients (92%) had a nonotologic head and neck manifestation of the disease, and many had pulmonary (69%) and renal manifestations (67%). SNHL was associated with renal disease in 67% of cases and with CHL in 42% of cases. Of 29 patients classified as having severe disease, 17 (59%) had SNHL and 16 (55%) of the patients had bilateral hearing loss. Seven patients had SNHL as their only severe manifestation and five had bilateral hearing loss. Four of the seven patients were put on cyclophosphamide for their severe WG. Two of these had worsened hearing, one remained stable, and one had hearing that improved. Subsequently, of the remaining three patients who were not given cyclophosphamide,
TABLE 2. Data on study patients aged
All (n Mean age (yr) (range) Males (%) Females (%) Disease duration (mo) (range) ANCA/anti-PR-3 (%) Medication usage (ever) Glucocorticoids (%) Methotrexate (%) Azathioprine (%) Cyclophosphamide (%) HL, hearing loss. 26) All SNHL (n 53 (2463) 4 (57) 3 (38) 43.0 (24186) 4 (50) 8 (100) 6 (75) 2 (25) 5 (63)

835

one had fluctuating hearing loss throughout his disease course that never fully resolved, and the other two had stable hearing loss. Table 4 depicts the changes in hearing loss in all cases. Although there appears to be a trend indicating cases of CHL showing improvement more often than cases of SNHL (58% versus 18%), this difference was not statistically significant. All of these patients except one with CHL had been treated with cyclophosphamide. Fourteen of 17 (82%) cases of SNHL were unremitting despite therapy. In analyzing the subgroup of patients younger than 65 years old with no noise exposure (Table 5), similar trends in hearing outcome exist compared with the entire group. There are fewer cases of SNHL, as expected, but still a low incidence of improvement in SNHL (13%). Although more cases of CHL showed improvement (40%) than SNHL (13%), this difference was also not statistically significant. All eight of the patients who presented with hearing loss as their initial complaint had SNHL at some point in their course. Five of the eight had stable deficits, whereas two worsened and one improved. Five patients also had CHL and, of these, two improved, two worsened, and one remained stable. Of these eight patients who initially complained of hearing loss, four were younger than 60 years of age, with no known history of noise exposure. Three of these four patients had hearing loss that progressed and one had stable hearing loss. None of the four patients hearing improved. DISCUSSION Although the nasal, upper airway, pulmonary, and renal manifestations of WG are well known, the otologic ones are less appreciated (1215). In the current study, 20 (56%) patients with WG had some form of hearing loss, which compares to the upper limit of the previously reported prevalences [1961% (16), 47% (13)]. The prevalence of SNHL in our patients with WG (47%) is higher than previously reported (2.843%) (46). This is possibly explained by the comprehensive review of medical records for evidence of hearing loss used in the current study. Hearing loss can be an early indicator of WG. Sixteen
65 years and without a history of noise exposure
8) All CHL (n 10) Any HL (n 47 (2463) 5 (45) 6 (55) 39 (684) 7 (64) 8 (100) 9 (82) 3 (27) 9 (82) 11) No HL (n 15)

45.0 (2263) 14 (54) 12 (46) 40 (12119) 20 (77) 26 (100) 16 (62) 5 (26) 21 (81)

45.0 (2455) 5 (50) 5 (50) 41 (6186) 6 (60) 10 (100) 8 (80) 2 (20) 6 (60)

44 (2263) 9 (60) 6 (40) 42 (4120) 14 (93) 15 (100) 8 (53) 1 (0.07) 12 (80)

Otology & Neurotology, Vol. 25, No. 5, 2004

836

S. BAKTHAVACHALAM ET AL.
TABLE 3. Manifestations of Wegeners granulomatosis in study group
All (n 36) SNHL (n 17) CHL (n 12) Any HL (n 16 (80.0) 2 (10.0) 10 (50.0) 20 (100.0) 1 (5.00) 0 (0.00) 15 (75.0) 12 (60.0) 5 (25.0) 20) No HL (n 11 (68.8) 8 (50.0) 9 (56.2) 11 (68.8) 0 (100.0) 0 (100.0) 10 (62.5) 12 (75.0) 1 (6.3) 16) Bilateral HL (n 12 (75.0) 1 (6.25) 7 (43.7) 16 (100.0) 1 (6.3) 0 (0.00) 10 (62.5) 11 (68.8) 4 (25.0) 16)

General Cutaneous (%) Mucous membrane/eye (%) Head and neck (%) Cardiovascular (%) Gastrointestinal (%) Pulmonary (%) Renal (%) Neurologic (%)

27 (75.0) 10 (27.8) 19 (52.8) 33 (91.7) 1 (2.8) 0 (0.00) 25 (69.4) 24 (66.7) 6 (16.7)

15 (83.3) 2 (11.1) 8 (42.1) 18 (100.0) 1 (5.6) 0 (0.00) 12 (66.7) 12 (66.7) 5 (27.8)

10 (83.3) 1 (8.33) 6 (50.0) 12 (100.0) 1 (8.3) 0 (0.00) 10 (83.3) 5 (41.7) 4 (33.3)

patients had hearing loss at presentation, and in eight it was the primary complaint. The most common otologic manifestations of WG are secondary to granuloma formation within the middle ear, eustachian tube, or nasopharynx. These pathologic changes can result in CHL, serous otitis, chronic otorrhea, and otomastoiditis, a painful, draining ear, sometimes associated with facial nerve paralysis. The evaluation of new hearing loss, either CHL or SNHL, in an adult should, therefore, include audiometry, followed by upper airway endoscopy in the presence of serous otitis to search for inflammatory lesions. Hearing loss in WG can result in permanent disability as evidenced by the use of amplification in seven patients (19%) in our cohort. Several other patients would have undoubtedly benefited from hearing aids, and many patients reported deficits affecting their daily activities. SNHL is a significant finding in WG, and its detection is therefore important for appropriate patient management. The presence of SNHL can be a warning of severe WG and is thought to necessitate initial treatment with cyclophosphamide rather than either methotrexate or azathioprine. Furthermore, SNHL is significant in that it can be associated with other severe manifestations. Although 7 patients had SNHL as their only severe manifestation, 11 of 22 (50%) of the patients with severe disease had SNHL associated with other severe disease manifestations. For example, 12 patients (67%) with SNHL also had renal disease. Of 29 patients with other manifestations of severe disease, 17 (59%) had SNHL. This emphasizes the importance of SNHL as a potential indicator of severe disease. Finally, fewer patients with SNHL than with CHL had improvement, but these findings were not statistically significant in this small subsample and can only be considered a possible trend. Together, these findings provide arguments for performing

a screening audiogram in all patients newly diagnosed with WG and subsequent serial audiograms for those with hearing loss and/or hearing complaints. Conductive hearing loss has long been known to occur when granulomatous disease involves the middle ear, but the extent to which hearing can be improved for patients with conductive hearing loss from WG is unclear. Certainly, if a patient with WG has only a middle ear effusion as the reason for conductive hearing loss in an affected ear, then removing the fluid from the middle ear and inserting a ventilating tube might result in sustained hearing improvement. However, when the patient with WG has middle ear or mastoid involvement, usually associated with a disease flare, then surgery without concomitant remission of the extratemporal bone manifestations tends to be unsuccessful in achieving long-term improvement in hearing. WG-related audiometric patterns have been described as typically flat, sometimes with additional high tone losses (4), findings that are sometimes difficult to distinguish from those caused by noise exposure or age. In an attempt to minimize the influence of these confounders, we list the audiometric findings for those 26 patients who were younger than 65 years old and with no history of noise exposure. This group had a SNHL incidence of 31%, an exceptional rate when compared with the 8% reported incidence of SNHL in laborers who comply fully with workplace recommendations (17). Among the study patients older than 65 years, the prevalence of SNHL (90%) was higher than would be expected to be caused by presbycusis alone. The accepted prevalence of presbycusis is 25% in those aged 65 to 75 years and 40 to 50% in those older than 75 years (17). Therefore, these data further support the findings that SNHL in patients with WG is likely attributable to the vasculitis and should not readily be dismissed as age- or noise-related.

TABLE 4. Hearing outcome for all study patients with hearing loss
SNHL (n 17) Improved (%) Stable (%) Worse (%) 3 (18) 10 (59) 4 (24) (n CHL 12)

TABLE 5. Hearing outcome in study patients with hearing loss aged 65 years and without a history of noise exposure
SNHL (n 8) Improved (%) Stable (%) Worse (%) 1 (13) 4 (50) 3 (38) (n CHL 10)

7 (58) 3 (25) 2 (17)

4 (40) 4 (40) 2 (20)

Otology & Neurotology, Vol. 25, No. 5, 2004

HEARING LOSS IN WEGENERS GRANULOMATOSIS This study has several strengths. The cohort of 36 patients is the largest reported with detailed audiometric data. The patients were comprehensively followed up with clinical data collected with the most current methodology. Furthermore, multiple audiograms were obtained for most patients, and these were reviewed by multiple assessors using a uniform system of interpretation. This study also has certain limitations. Although we found a high incidence of both CHL and SNHL among patients with WG, the true overall incidence of hearing loss may be even higher. This is because only 22 of 36 patients had objective audiometric evaluations, resulting in a remarkable 91% of audiograms showing a hearing deficit. It is likely that some of the remaining 14 patients, who did not have audiograms, had hearing loss that remained undetected. Furthermore, the audiograms obtained were not timed regularly, and five patients had only one audiogram. Because of this, transient episodes of hearing loss would have been easily missed. In addition, because nine patients had mixed hearing loss and, therefore, contributed data to both the CHL and SNHL categories, statistical comparisons between these groups were of limited utility. For example, despite the relatively large cohort in this study, an even larger one would likely be needed to demonstrate a statistically significant association between SNHL and renal disease. These results can only be directly applied to patients with WG and not other ANCA-associated vasculitides such as microscopic polyangiitis or Churg-Strauss syndrome. The cohort is drawn from a referral-based rheumatology practice, but the spectrum of disease is generally consistent with other larger published cohorts (3,5,18). The high incidence of Wegeners-related SNHL in this study is partly confounded by SNHL of other causes, most notably age- and noise-induced hearing loss. However, the analysis plan made adjustments to correct for this possibility, and we still found an extremely high rate of SNHL. Thus, our 48% incidence of SNHL is significant even when considering these factors and is a conservative estimate. CONCLUSION The findings of this study suggest a higher incidence of SNHL in WG than has previously been reported. Furthermore, the findings suggest that SNHL can possibly be viewed as one of several indicators of increasing severity of disease. For these reasons, we propose the following uses of audiometric data in patients with WG: (1)

837

screening audiograms should be performed in all patients with WG; (2) all patients with documented hearing loss should have follow-up audiometric assessment; and (3) audiometric data should be considered when making treatment decisions. Further research focusing on discerning the cause and developing more effective treatment of hearing loss in WG is needed. REFERENCES
1. Rasmussen N. Management of the ear, nose, and throat manifestations of Wegener granulomatosis: an otorhinolaryngologists perspective. Curr Opin Rheumatol 2001;13:311. 2. ODevaney K, Ferlito A, Hunter BC, Devaney SL, Rinaldo A. Wegeners granulomatosis of the head and neck. Ann Otol Rhinol Laryngol 1998;107:43945. 3. Hoffman GS, Kerr GS, Leavitt RY, et al. Wegener granulomatosis: an analysis of 158 patients. Ann Intern Med 1992;116:48898. 4. Kornblut AD, Wolff SM, Fauci AS. Ear disease in patients with Wegeners granulomatosis. Laryngoscope 1982;92:7137. 5. WGET Research Group. Limited versus severe Wegeners granulomatosis: baseline data on patients in the Wegeners granulomatosis etanercept trial. Arthritis Rheum 2003;48:2299309. 6. Takagi D, Nakamaru Y, Maguchi S, Furuta Y, Fukuda S. Otologic manifestations of Wegeners granulomatosis. Laryngoscope 2002; 112:168490. 7. Dekker PJ. Wegeners granulomatosis: otological aspects. J Otolaryngol 1993;22:3647. 8. WGET Research Group. Design of the Wegeners Granulomatosis Etanercept Trial (WGET). Control Clin Trials 2002;23:450468. 9. Canalis RF, Lambert PR. The Ear: Comprehensive Otology. Philadelphia: Lippincott Williams & Wilkins, 2000. 10. Stone JH, Hoffman GS, Merkel PA, et al. A disease-specific activity index for Wegeners granulomatosis: modification of the Birmingham Vasculitis Activity Score. International Network for the Study of the Systemic Vasculitides (INSSYS). Arthritis Rheum 2001;44:91220. 11. Collins JG. Prevalence of selected chronic conditions, United States, 19831985. NCHS Advance Data, No. 155, 114. Philadelphia: Lippincott Williams & Wilkins, 1988. 12. Calonius IH, Christensen CK. Hearing impairment and facial palsy as initial signs of Wegeners granulomatosis. J Laryngol Otol 1980;94:64957. 13. Illum P, Thorling K. Otological manifestations of Wegeners granulomatosis. Laryngoscope 1982;92:8014. 14. McDonald TJ, DeRemee RA. Wegeners granulomatosis. Laryngoscope 1983;93:22031. 15. Banerjee A, Armas JM, Dempster JH. Wegeners granulomatosis: diagnostic dilemma. J Laryngol Otol 2001;115:467. 16. McCaffrey TV, McDonald TJ, Facer GW, DeRemee RA. Otologic manifestations of Wegeners granulomatosis. Otolaryngol Head Neck Surg 1980;88:58693. 17. Adams PF, Hendershot GE, Marano MA. Current estimates from the National Health Interview Survey: National Center for Health Statistics. Vital Health Stat 1999;10(200). 18. Reinhold-Keller E, Beuge N, Latza U, et al. An interdisciplinary approach to the care of patients with Wegeners granulomatosis: long-term outcome in 155 patients. Arthritis Rheum 2000;43: 102132.

Otology & Neurotology, Vol. 25, No. 5, 2004