Tympanic membrane findings of otitis media with anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (OMAAV)

Objective Otitis media with anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (OMAAV) is characterized by adult otitis media refractory to conventional treatments. OMAAV is either an aural manifestation of existing ANCA-associated vasculitis (AAV) or an initial aural manifestation of AAV. OMAAV occasionally causes an irreversible profound sensorineural hearing loss that may require a cochlear implant even in the latter case. In such a case, prompt diagnosis of OMAAV is important but sometimes difficult. When diagnosing OMAAV, repetitive otitis media with effusion (OME) in adults is the most difficult differential diagnosis. Precise evaluation of tympanic membrane (TM) findings would help to achieve a prompt diagnosis. The objective of this study was to discriminate OMAAV from adult OME based on tympanic TM findings. Methods 10 with OMAAV and 10 with adult OME were included. We established a sc oring system of O MAAV t ympanic membrane (SCOT) to evaluate TM findings of OMAAV consisted of following three characteristic findings: thickening of pars tensa, vasodilation of pars tensa, and posterior wall swelling. Each TM finding in OMAAV and OME was scored from 0 to 3 by 20 otolaryngologists who never knew the diagnosis. Reliability of the scoring system in terms of consistency between examiners was evaluated by intraclass correlation coefficients (ICC). Validity was tested by comparing the TM scores between OMAAV and OME and by the area under the curve (AUC) of receiver operating characteristic (ROC) curve to discriminate OMAAV from OME. Correlations between the TM scores and various systemic markers of OMAAV including white blood cell count, C-reactive protein, myeloperoxidase-anti-neutrophil cytoplasmic antibody, and Birmingham Vasculitis Activity Score were examined. Results The ICC of each score was over 0.95. Each of and the total TM scores were significantly higher in OMAAV than in OME. AUC of ROC curve was 0.9134. The cut-off value set at 2 points had the best combination of sensitivity (93.0%) and specificity (74.0%) to distinguish OMAAV from OME. No significant correlations were found between the total score of SCOT and systemic markers. However, the total score of SCOT significantly correlated with the average hearing level of both air (p = 0.021) and bone conductions (p = 0.032). Conclusion Reliability and validity of SCOT in discriminating OMAAV from adult OME, the most difficult differential diagnosis, were demonstrated, suggesting that SCOT would be useful to make an early diagnosis of OMAAV. Correlation of SCOT with hearing level suggests that SCOT is also useful to evaluate disease status of OMAAV.