Introduction
Autoimmune inner-ear disease is characterised by hearing loss and/or dizziness caused by antibodies or immune cells that attack the inner ear. The incidence of immune-mediated inner-ear disease is difficult to determine, although it probably accounts for less than 1 per cent of the causes of hearing loss or vertigo.Reference Bovo, Ciorba and Martini1 However, it may be underdiagnosed, as there is no specific diagnostic test. Hearing loss can appear suddenly, and follow a fluctuating or rapidly progressive course, and up to 50 per cent of cases are accompanied by instability, ataxia or vertigo. True vertigo is infrequent.
As these manifestations are sufficiently non-specific to cover most inner-ear disorders, there is a need to establish clinical diagnostic criteria that guide the application of a treatment. The diagnosis is based primarily on clinical criteria (a high-risk clinical profile that includes, but is not restricted to: female sex, age of 25–50 years at onset, unexplained progressive bilateral sensorineural hearing loss (SNHL), rapid progression of symptoms over weeks, normal otoscopic examination findings, injury to an ear from different causes and the onset of progressive hearing loss in the contralateral ear, and coexistence of systemic autoimmune disease) and a positive response to corticosteroids.Reference García-Berrocal, Trinidad, Ramírez-Camacho, Lobo, Verdaguer and Ibañez2
Advances in magnetic resonance imaging (MRI) have made it possible to differentiate the compartments of the inner ear and enable in vivo visualisation of the endolymphatic space enlargement in patients diagnosed with Ménière's disease.Reference Naganawa, Satake, Kawamura, Fukatsu, Sone and Nakashima3 Endolymphatic hydrops has also been observed in Cogan's syndrome (prototype of autoimmune disease leading to inner-ear injury).Reference Jung, Nadol, Folkerth and Merola4 The incidence of endolymphatic hydrops in patients with immune-mediated inner-ear disease is not well-known, although some anecdotal evidence has been reported.Reference Tuñón, ómez, Lobo Duro, Brea Álvarez and García-Berrocal5
Animal studies and very few human histopathological studies of autoimmune inner-ear disease have shown degeneration of inner-ear structures, atrophic changes in the auditory nerve and development of endolymphatic hydrops.Reference Lobo6 Hence, the main objective of this study was to demonstrate the presence of endolymphatic hydrops in patients with a clinical diagnosis of immune-mediated inner-ear disease. In addition, the clinical variables associated with endolymphatic hydrops are described so as to generate hypotheses and guide future research.
Materials and methods
Participants
Participants were recruited prospectively from patients presenting to the out-patient clinic at the otolaryngology department with a clinical diagnosis of immune-mediated inner-ear disease, from October 2015 to December 2016. These patients entered a clinical trial approved by the hospital's ethics committee (approval details: proyecto de investigaciòn (research project) number PI 14814, Acta 306, 12-01-2015).
Inclusion criteria for the study were any patient with a clinical diagnosis of immune-mediated inner-ear disease, as determined by: major and minor criteria;Reference García-Berrocal, Trinidad, Ramírez-Camacho, Lobo, Verdaguer and Ibañez2 no other diagnosis for the hearing loss after laboratory and imaging tests; and a response to corticoids or immunosuppressive therapy, defined as a complete response (full recovery at week eight of therapy, after onset or full recovery of the previous threshold after relapses) or a partial response (improvement of 15 dB at one frequency or 10 dB at two frequencies, or 15 per cent improvement in word recognition).
Patients were excluded if they were pregnant, aged under 14 years, or if they had chronic middle-ear disease, a tympanic membrane perforation, renal failure, previous allergic reactions to gadolinium contrast agents, metal foreign bodies (including cerebrovascular aneurysm clips), a pacemaker, or claustrophobia.
A structured patient history was taken and clinical records were examined to assess: the duration of symptoms (hearing loss), the presence of other symptoms (dizziness, tinnitus, ear fullness), other associated autoimmune diseases, and other findings revealed by laboratory and imaging studies such as immunophenotypic analysis of peripheral blood lymphocytes, positron emission tomography (PET)-computed tomography or MRI. These data were collected by one of the ENT specialists prior to review of the MRI results and subsequently analysed without knowledge of the MRI results.
Audiometry
Standard pure tone audiometry and extended high-frequency audiometry was performed before and after immunosuppressive therapy to ascertain positive responses to therapy.
Intratympanic gadolinium
The contrast agent used was an 8-fold dilution of 0.5 mmol/ml solution of gadoteric acid (Dotarem® = 6d-DOTA; Guerbet, Aulnay-sous-Bois, France) with normal saline. The detailed methods for intratympanic gadolinium injection have been reported previously.Reference Tuñón, ómez, Lobo Duro, Brea Álvarez and García-Berrocal5, Reference Nakashima, Naganawa, Sugiura, Teranishi, Sone and Hayashi7, Reference Naganawa, Sugiura, Kawamura, Fukatsu, Sone and Nakashima8
Magnetic resonance imaging
All patients underwent 3 T MRI of the temporal bone with an eight-channel phased array coil to rule out schwannoma or other causes for the symptoms. From 24 to 28 hours after bilateral intratympanic contrast administration in the otolaryngology department, a three-dimensional (3D) real inversion recovery sequence was performed in an Achieva 3 T scanner (Philips, Best, The Netherlands) with an eight-channel sensitivity encoding (‘SENSE’) antenna, with the following parameters: repetition time = 6000 ms, echo time = 107 ms, inversion time = 1650 ms, flip angle = 180°, number of signals averaged = 1, sensitivity encoding factor = 1.5 × 1, field of view = 160 × 160 × 18 mm, and voxel size = 0.55 × 0.63 × 1 mm/pixel. The sequence was performed in the axial plane and the scan time was 12.42 minutes.
Two experienced neuroradiologists, who were unaware of the audiology results, the side, or the unilaterality or bilaterality of symptoms, and who had the minimal amount of clinical information required by the radiology request documentation, qualitatively analysed the images separately. The degree of endolymphatic hydrops was established visually in both the cochlea and the vestibule, and was graded according to the Baráth criteria.Reference Baráth, Schuknecht, Naldi, Schrepfer, Bockisch and Hegemann9
The axial plane of the vestibule was chosen as the anatomical reference for measuring the degree of vestibular hydrops, which includes practically all the lateral semicircular canal. In cases where the utricle and saccule are delineated separately, but are enlarged, volumes were measured by freehand area delimitation. The axial projection centred in a cut plane through the middle plane of the modiolus was used for the evaluation of cochlear hydrops, according to a published scheme.Reference Tuñón, ómez, Lobo Duro, Brea Álvarez and García-Berrocal5
The study protocol was reviewed and approved by the hospital's ethics committee. Written informed consent was obtained from all patients prior to their participation in the study.
Statistical analysis
Categorical variables are presented as means of absolute and relative frequencies; numerical variables are presented as means (standard deviations) or medians (percentiles 25 and 75). Categorical variables were analysed with the chi-square test or Fisher's exact test, and the numerical variables were analysed using the Mann–Whitney U test. A p value of less than 0.05 was considered statistically significant. The statistical analysis was conducted using Stata software, version 14 (StataCorp, College Station, Texas, USA).
Results
Seventeen patients with a clinical suspicion of immune-mediated inner-ear disease were enrolled in the study. The demographic data are shown in Table I. All patients underwent pure tone audiometry and extended high-frequency audiometry (bilateral), and MRI of the temporal bone with prior gadolinium instillation. Mean age at the time of entering the study was 47.9 years, and mean age at onset of the first symptoms was 41.3 years (range, 14–72 years), with a mean evolution time of immune-mediated inner-ear disease of 6.6 years (range of 1–43 years, mode of 1 year, median of 1.5 years). There were 7 males and 10 females.
Demographics, clinical symptoms and results of hydrops on MRI
* Unilateral; otherwise bilateral. †Indicates patients with secondary immune-mediated inner-ear disease. MRI = magnetic resonance imaging; M = male; F = female; N/A = not available
Twelve patients presented with primary and five presented with secondary immune-mediated inner-ear disease. Among these patients, four had autoimmune thyroiditis. The presenting symptoms were: sudden SNHL, in 35 per cent (6 out of 17), or fluctuating SNHL, in 65 per cent (11 out of 17). Four patients had unilateral (23.5 per cent) and 13 had bilateral SNHL (76.5 per cent). Only three patients (18 per cent) had experienced vestibular symptoms at some point since the onset of symptoms, and only three patients (18 per cent) complained of aural fullness. In contrast, tinnitus was a much more frequent complaint, with 13 patients (76 per cent) having a stable or irregular ringing in one or both ears. Hearing loss after an injury to the contralateral ear occurred in two patients (one had an acoustic neurinoma, and the other suffered an acoustic trauma and later developed a rapidly progressive SNHL in the contralateral ear that stabilised after immunosuppressive treatment).
The MRI scans obtained were valid and of sufficient quality to assess the endolymphatic space in all patients. The validity of the examination was established by comparing the images of asymptomatic and non-endolymphatic hydrops ears of the study patients.
There were no severe complications associated with the intratympanic injection. Five patients suffered mild otalgia and four patients experienced mild dizziness that resolved the same day of the procedure (Table II).
Degree of hydrops on MRI* and adverse effects
* Graded according to the Baráth criteria: 0 = none; 1 = mild; 2 = significant.Reference Baráth, Schuknecht, Naldi, Schrepfer, Bockisch and Hegemann9 †Indicates patients with secondary immune-mediated inner-ear disease. MRI = magnetic resonance imaging
Hydrops was observed in 12 patients (Tables I and II): 5 patients had cochlear hydrops, 3 had vestibular hydrops, and 4 had both cochlear and vestibular hydrops. An unpredicted case of endolymphatic hydrops was observed in one patient (8.3 per cent), wherein endolymphatic hydrops was revealed in the cochlea of the contralateral ear. In two patients with bilateral hearing loss, endolymphatic hydrops was observed in one ear. One of these patients had a profound hearing loss in the right ear and several years later had developed a fluctuating hearing loss in the left ear (delayed endolymphatic hydrops).
Gadolinium 3D MRI diagnosed hydrops in 11 of 12 patients with primary immune-mediated inner-ear disease (92 per cent). Of these, seven patients (64 per cent) had only cochlear (five cases) or predominantly cochlear hydrops (two cases). A positive MRI result was observed in only one of five patients (20 per cent) with secondary immune-mediated inner-ear disease. This patient had congenital cytomegalovirus and autoimmune thyroiditis, with profound SNHL in the left ear, and fluctuating SNHL in the right ear, which fully responded to oral, intravenous corticosteroids and intratympanic therapy on the five occasions he was treated. A grade 2 cochlear and vestibular endolymphatic hydrops was observed in this patient. The MRI scan also showed hydrocephalus in this patient.
In three patients with a positive gadolinium MRI result, only vestibular hydrops was observed. Two of these patients presented with vertigo and a clinical picture comparable to that of Ménière's disease, and had a complete response to corticosteroid therapy whenever treatment was given.
A PET scan was performed on 14 patients and the results were positive in only 4. No uptake in the inner ear was observed in any case. Uptake was observed in the thyroid gland in two patients: one had autoimmune thyroiditis, and the other had no history of autoimmune thyroiditis or any other autoimmune diseases apart from immune-mediated inner-ear disease. Uptake was observed in the aortic arch and supra-aortic trunks in one patient, who also had autoimmune thyroiditis. In the other patient, uptake was observed in the tonsils and cervical lymph nodes.
Immunophenotypic analysis of peripheral blood lymphocytes was performed on nine patients and the results were positive in five (55 per cent).
Patients received oral, intratympanic or intravenous corticosteroids, both at the onset of symptoms and on relapse. After therapy at onset, nine patients achieved a complete response and six patients had a partial response. Of the remaining two patients, one received no treatment and in the other case the information was not available. Fourteen patients received therapy after relapses: six patients received oral corticoids, nine received intratympanic corticoids (four of these received both oral and intratympanic therapy), two patients received intravenous corticoids, and one patient received another immunosuppressant or immunomodulatory therapy (etanercept, methotrexate). All patients achieved a complete or partial response to treatment.
The clinical data and the MRI findings are shown in Table I. The degree of endolymphatic hydrops and the adverse effects are shown in Table II. Figure 1 shows the various extents of vestibular and cochlear endolymphatic hydrops in the affected ears of several patients. Although extended high-frequency audiometry was performed in all patients, these data are not shown here.
Axial three-dimensional real inversion recovery sequence magnetic resonance imaging scans, obtained 24 hours after intratympanic injection of Dotarem® = 6d-DOTA (gadoterate meglumine). (a) Patient 7: shows severely enlarged endolymphatic space in the cochlea (short arrow) and mildly enlarged endolymphatic space in the vestibule (arrow) as negative signal intensity values in the left ear and mild cochlear hydrops (short arrow) in the right ear. (b) Patient 14: demonstrates mildly enlarged endolymphatic space in the right cochlea (short arrow) and mildly enlarged endolymphatic space in the left cochlea (short arrow) as negative signal intensity values. (c) Patient 15: shows severely enlarged endolymphatic space in the cochlea (short arrows) and severely enlarged endolymphatic space in the vestibule (arrows) in both ears. (d) Patient 17: demonstrates mildly enlarged endolymphatic space in the left cochlea (short arrow).
The statistical analysis revealed no significant correlations between: the duration of symptoms and degree of endolymphatic hydrops; the extent of cochlear endolymphatic hydrops and pure tone audiometry thresholds on the affected side at low, middle or high frequencies; or the degree of vestibular or cochlear endolymphatic hydrops and aural fullness or tinnitus when considered with or without vertigo.
Discussion
Gadolinium instilled in the middle ear 24 hours before MRI gave adequate imaging contrast and was relatively well-tolerated by most patients, with very few complications. This result accords with several reports, as no deterioration of the hearing function of healthy volunteers or patients has been described after intratympanic Dotarem® (gadoterate meglumine) injection.Reference Liu, Huang, Meng, Wang, Liu and Chen10, Reference Louza, Krause and Gürkov11
In Ménière's disease, the distribution of endolymphatic hydrops has often been shown to include the cochlea and vestibule, regardless of whether patients presented with only cochlear symptoms or only vestibular symptoms.Reference Hornibrook, Flook, Greig, Babbage, Goh and Coates12 The results of our study show a similar pattern in patients with immune-mediated inner-ear disease, but cochlear hydrops alone or predominantly cochlear hydrops were mostly observed in patients with primary disease.
Endolymphatic hydrops is observed most frequently in the saccule, followed by the cochlea, utricle and the semicircular canals.Reference Okuno and Sando13 Recent MRI studies have also confirmed these histopathological findings.Reference Fiorino, Pizzini, Beltramello and Barbieri14 In a recent meta-analysis, which included 184 temporal bones from different collections of temporal bones in the USA, endolymphatic hydrops always arose in the cochlear apex, even in non-symptomatic cases, and then involved the saccule, utricle, ampullae and canal system in that precise sequence.Reference Pender15 In our study, endolymphatic hydrops was most commonly observed in the cochlea.
In two of three patients with vestibular endolymphatic hydrops only, the clinical picture was similar to that of Ménière's disease. There is some overlap between these two diseases, as an immune-mediated mechanism is assumed in a small proportion of patients with Ménière's disease. Tinnitus is a common symptom in Ménière's disease and was present in 13 (76 per cent) of our immune-mediated inner-ear disease patients. In contrast, aural fullness, another frequent complaint in definite Ménière's disease patients, was reported by only three (18 per cent) of our patients. Recent studies suggest that vestibular hypofunction occurs primarily in the first 5–10 years following disease onset.Reference Huppert, Strupp and Brandt16 This was not investigated in our study, but only three (18 per cent) of the patients had vestibular symptoms. Moreover, in Ménière's disease patients, there is a characteristic upward-sloping curve after subtracting the normal ear hearing threshold;Reference Wu, Dai, Zhao and Sha17 however, in our patients, the audiogram more frequently showed a downward-sloping pattern, with more severe hearing loss in the high frequencies.
The contralateral ears of patients with unilateral Ménière's disease show asymptomatic saccular hydrops in around 30 per cent of patients.Reference Morita, Cureoglu, Nomiya, Nomiya, Joglekar and Harada18 In our study, endolymphatic hydrops in the asymptomatic ear was observed in 8 per cent of patients (1 out of 12), which is in line with recent studies.Reference Ito, Kitahara, Inui, Miyasaka, Kichikawa and Ota19
Despite the differences found in the study, endolymphatic hydrops may not be useful to differentiate between immune-mediated inner-ear disease and Ménière's disease; however, it does raise questions regarding the pathophysiology of both conditions.
We found no correlation between the duration of immune-mediated inner-ear disease and the expansion of endolymphatic fluid in the vestibule and cochlea, or between specific symptoms and the degree of endolymphatic hydrops. However, as the evolution of most patients’ disease was less than two years, it is not possible to reach any firm conclusions. Nor was there an association between the degree of hydrops and hearing loss on conventional audiometry or high-frequency audiometry.
It should be noted that in the five patients with secondary immune-mediated inner-ear disease, only one (20 per cent) had hydrops on MRI. The only patient in the primary immune-mediated inner-ear disease group with a negative MRI result showed PET scan uptake in the thyroid gland. The only patient with secondary immune-mediated inner-ear disease and documented endolymphatic hydrops on MRI had congenital cytomegalovirus, which is a confounding factor and could explain endolymphatic hydrops by itself.Reference Rarey and Davis20–Reference Fukuda, Keithley and Harris22 The absence of a correlation between auditory symptoms and the presence of hydrops in patients with secondary immune-mediated inner-ear disease has previously been documented in human temporal bone studies (cochlear hydrops was found in only 1 of 14 ears in 7 patients).Reference Sone, Schachern, Paparella and Morizono23 Kariya et al. examined 15 temporal bone samples from 8 systemic lupus erythematosus patients, along with 17 samples from 10 age-matched healthy control patients.Reference Kariya, Kaya, Hizli, Hizli, Nishizaki and Paparella24 They found a positive correlation between the loss of cochlear hair cells and the duration of systemic lupus erythematosus, and concluded that it represented the histopathological basis for the cochlear dysfunction, including SNHL, experienced by systemic lupus erythematosus patients.
Endolymphatic hydrops is considered to result from the disruption of inner-ear water homeostasis, which involves excessive endolymph production and/or reduced endolymph absorption. The experimental model that has contributed most to the pathophysiology of immune-mediated inner-ear disease is the model of experimental labyrinthitis by keyhole limpet haemocyanin, developed by Harris.Reference Harris25 The inoculation of keyhole limpet haemocyanin in rats and guinea pigs can trigger sterile labyrinthitis, leading to a functional alteration with the loss of sensory cells, and, ultimately, to endolymphatic hydrops, cochlear fibrosis and osteoneogenesis. The loss of or damage to spiral ligament fibrocytes that express sodium (Na+)/potassium (K+) adenosine triphosphatase (ATPase) contributes to cochlear dysfunction by altering K+ homeostasis. Extravasation of leukocytes to the tympanic ramp leads to the formation of fibrous matrix and granuloma.Reference Lobo6 In the present study, hydrops was present in 71 per cent of the patients, a lower rate than in those with definite Ménière's disease, which could be explained by their different pathophysiology.
The main limitation of the present investigation is the small number of patients. However, immune-mediated inner-ear disease is a rare disease, and these preliminary results may guide the planning of future research that will shed light on this issue. One question that remains uncertain is whether primary and secondary immune-mediated inner-ear disease behave in different ways regarding the presence of endolymphatic hydrops, which could be due to different pathophysiological mechanisms. In our study, autoimmune thyroiditis was the most frequent autoimmune disease associated with immune-mediated inner-ear disease, but a larger sample should include other illnesses such as systemic lupus erythematosus or granulomatosis with polyangiitis.
Another limitation of the study is the lack of universal agreement on the diagnostic criteria for immune-mediated inner-ear disease. Serological test findings were not positive in all patients, and we relied on our proposed major and minor criteria.
A further study with larger samples should be conducted in the near future, and should include patients with other autoimmune diseases associated with immune-mediated inner-ear disease. We are currently recruiting more patients with presumed secondary immune-mediated inner-ear disease to further study the presence of endolymphatic hydrops among other study variables.
• Animal histopathological studies of autoimmune inner-ear disease have shown inner-ear structure degeneration, atrophic auditory nerve changes and endolymphatic hydrops development
• Magnetic resonance imaging (MRI) has become useful for assessing endolymphatic hydrops in Ménière's disease
• This study evaluated and confirmed the presence of endolymphatic hydrops in immune-mediated inner-ear disease patients
• The findings support the role of endolymphatic hydrops in the pathogenesis of immune-mediated inner-ear disease
• Demonstration of endolymphatic hydrops on MRI can be useful in diagnosing inner-ear diseases that present with vertigo or hearing loss
Conclusion
Endolymphatic hydrops has been observed in different inner-ear disorders. This study demonstrates the presence of endolymphatic hydrops in patients with immune-mediated inner-ear disease. However, the results suggest that within immune-mediated inner-ear disease, there are groups with different pathophysiological mechanisms.
Acknowledgement
The authors are indebted to Ana Rayuela for her assistance with the statistical analysis.
Competing interests
None declared.
