We congratulate Wesseler et al. (1) for their informative and important report of endolymphatic hydrops magnetic resonance (MR) imaging with intratympanic contrast agent application. They examined 31 patients (52 ears) and show images of the inner ear fluid spaces in a quality that is rarely encountered in today's publications on this topic. Hydrops MRI, in their study, performed better than functional audiovestibular tests in identifying patients with Menière's Disease (MD). Interestingly, two of their patients did not fulfill the diagnostic criteria proposed by the Barany Society in 2015, but they had endolymphatic hydrops (ELH) on MR imaging and fulfilled the possible MD criteria of the AAO-HNS 1995. This corresponds well to our clinical experience where the 2015 criteria, which suffer from significant shortcomings (2), are adding no benefit over the 1995 criteria. In a recent descriptive study of patients with MRI-confirmed endolymphatic hydrops (3), the discrepancy between the traditional clinical diagnostic criteria and the confirmed underlying pathology (endolymphatic hydrops) hydrops has been demonstrated, indicating the necessity of a new formulation of clinical diagnostic criteria which take into account the important knowledge gathered after more than a decade of clinical hydrops MR imaging. Wesseler et al.'s findings confirm the strong and significant correlation between cochlear hydrops and hearing loss. A specific strength of their study is the bilateral examination with IT-Gd MRI, which allowed them to clearly demonstrate that the hearing loss is more pronounced in the ear with the greater degree of cochlear hydrops. This result could be achieved by a high image quality with a strong perilymph signal due to the IT-Gd application route, whereas in our experience the intravenous Gd application route does not achieve the same image quality, especially in the region of the thin cochlear duct, where a high signal-to-noise ratio is of utmost importance for reliable visualization. We would like to comment on two parts of the article by Wesseler et al. to offer the reader a broader view upon them: the correlation between vestibular evoked myogenic potentials (VEMP) and hydrops and the safety of intratympanic Gd application. Concerning the topic of the correlation between VEMP and vestibular hydrops: Wesseler et al. did not find a correlation, but their VEMP study population was relatively small (n = 11). Katayama et al. (4) examined 40 patients (among them 19 patients with clinically definite MD and 17 patients with other subtypes of hydropic ear disease). They classified cVEMP as “present” or “absent” and the hydrops as “none/mild/significant” and reported a significant correlation between VEMP and vestibular hydrops. Seo et al. (5) examined 26 definite MD patients, the vestibular hydrops was rated as present or absent, and they found no correlation between hydrops and VEMP. Okumura et al. (6) examined cVEMP and oVEMP in 21 patients with definite MD, vestibular hydrops was rated “present” or “absent,” they reported a positive correlation between VEMP pathology and vestibular hydrops: Among 14 patients who showed vestibular hydrops on Gd-MRI, 10 (71.4%) had abnormal cVEMP asymmetry ratios. All patients who had normal cVEMP and oVEMP results showed no vestibular hydrops on MRI. Further, all MD patients without vestibular hydrops on Gd-MRI had normal cVEMPs and oVEMPs. Therefore, their data also revealed a positive correlation between VEMP pathology and vestibular hydrops. Jerin et al. (7) examined oVEMP frequency tuning in certain MD patients versus healthy controls and could for the first time provide direct evidence that the oVEMP frequency tuning is altered in the presence of ELH. They did not analyze the VEMP results according to the ELH severity, but merely compared the three groups (affected ears of MD patients, unaffected ears of MD patients, healthy controls) and did not find group differences for oVEMP amplitudes. This study, in line with other reports about VEMP frequency tuning in MD patients, indicates that VEMP frequency tuning parameters are diagnostically more useful for the “functional” detection of ELH than VEMP amplitudes (or amplitude ratios), which are known present as enlarged, reduced, or absent in patients with MD. Wesseler et al. did not mention two other studies on the topic of VEMP/ELH correlation: Gürkov et al. (8) examined 27 patients with definite MD (AAO-HNS 1995), the hydrops was graded as “none/mild/marked/extreme,” and the cVEMP were described as interaural amplitude ratio. We found a significant positive correlation between hydrops severity and cVEMP pathology. Another study (9) examined 23 patients with definite MD (AAO-HNS 1995) whose diagnosis was confirmed with electrocochleography, using quantitative vestibular endolymph area ratios (percentage) for the description of vestibular hydrops severity and the interaural amplitude ratio for description of the cVEMP. There, we also found a significant positive correlation between cVEMP pathology and hydrops severity. For the interpretation of these studies it is important to note the inhomogeneity of methods concerning several important parameters, e.g., diagnostic criteria, hydrops quantification, and VEMP quantification. It is our impression from the published studies that if the diagnostic certainty and the image quality are high and if the VEMP quantification and the hydrops quantification are of a high resolution (fine-scaled), then the VEMP pathology is more likely to be correlated to the hydrops severity. Concerning the potential toxicity of intratympanically applied Gd contrast: On this topic, Wesseler et al. only refer to one study by Liu et al. (10) and state that there were no long-term studies on the matter. Liu et al. performed audiometry 24 hours before and 24 hours and 1 month after i.t. administration (via Eustachian tube perfusion) of Gd, and the report of their safety outcome is confined to the verbal statement that no significant changes in audiometry and tympanometry were noted and that no adverse effects of the intratympanic injection of Gd were observed. Our group, however, specifically studied the potential ototoxic effects of intratympanic Gd administration in the short term (11) (1 d), middle term (12) (1 wk), and long term (13) (20 mo, median; 7–34 mo). These studies reported audiometric data before and after the i.t. Gd application. In none of these studies could an ototoxic effect be detected. For hydrops MRI, intravenous and intratympanic contrast applications each has their own advantages and disadvantages (e.g., the extremely low total dose to which the patient is exposed by the intratympanic delivery route), but we believe this information is relevant for the otologist to choose wisely the best approach for his patients.
In Reply: Dr. Attyé has written a comment in response to the review article “Menière and Friends: Imaging and Classification of Hydropic Ear Disease,” (1) which here is referred to as “original article.” In the original article, 1) evidence emanating from histopathologic studies of temporal bones is described, 2) evidence emanating from inner ear magnetic resonance imaging (MRI) studies is described (including work from Dr. Attyé's group), 3) previously used classifications of Menière's disease (MD) are described. These different perspectives are combined with the author's experience in neurotology and diagnostic and therapeutic management of patients with (suspected) hydropic ear disease including inner ear magnetic resonance (MR) image analysis to formulate a new classification of MD and its variants. The concept of Hydropic Ear Disease, which was already published previously (2–6), represents the first logical and comprehensive framework encompassing all variants of MD and is applicable in clinical practice as well as research, even without the availability of a 3T MR scanner. The comment by Dr. Attyé, however, is concerned with details of inner ear imaging techniques and with the presentation of inner ear MR images in publications. It therefore does not seem well addressed and would have been better addressed at one of the many publications from different groups which present novel findings of inner ear MR imaging. Nevertheless, in the following, his comments shall be discussed. Since the landmark discovery of separate visualization of endolymphatic and perilymphatic inner ear fluid spaces (7) in living patients using clinical MR equipment and the first demonstration of endolymphatic hydrops in patients with hydropic ear disease (MD) (8), a vast body of literature on MR imaging of endolymphatic hydrops has been produced by many different working groups around the world (9), most recently documenting the highly significant correlation between cochlear endolymphatic hydrops and hearing loss in a cohort of almost 200 patients (10). Clearly, the leading role in the refinement of MR imaging techniques of endolymphatic hydrops is well documented by numerous publications from the Nagyoa group, displaying excellent image quality and an unsurpassed number of technical innovations in the field, e.g., (11–13). The techniques applied worldwide for MR imaging of endolymphatic hydrops include intratympanic as well as intravenous contrast application, at least five different contrast agents, different models of MR scanners produced by Siemens, General Electrics and Philips, different types of receiver coils, and many different types of sequences as well as image analysis procedures. It is self-evident that not every group has already published their images. It is self-evident that not every group produces inner ear MR images of perfect quality. It is self-evident that every institution should aim to achieve an image quality which allows for a distinction between normal endolymphatic space and endolymphatic hydrops. It should not be forgotten that the inner ear fluid spaces have a complex geometry and are extremely small, they measure about 200 μl in total! As it is true for any kind of measurement procedure that involves 1) variable technical equipment and procedures and 2) human—and therefore subjective—analysis and decision making, there must be differences between the images produced by different working groups. In a recent article (14), Dr. Attyé et al. report that by using a three-dimensional (3D) fluid attenuated inversion recovery (FLAIR) sequence, their group could not reproduce the findings reported previously by other groups. They examined 30 healthy controls and when using the conventional axial image analysis (15), reported to find moderate cochlear endolymphatic hydrops (EH) in 16 healthy subjects, severe cochlear EH in 4 healthy subjects, moderate vestibular EH in 18 healthy subjects, and severe vestibular EH in 9 healthy subjects. In their hands, axial image analysis of inner ear MR imaging using 3D FLAIR was not useful for differentiating between healthy controls and MD patients. They claim that this method was clinically not useful and propose an alternative method of image analysis, based on multiplanar reconstruction and on the sagittal visualization of the sacculus and utriculus. This method is referred to as sacculus-utriculus-ratio-inversion (SURI) grading here. They define the SURI grading as follows: If the saccule can be visualized separately from the utricle and is smaller than the utricle, this would be regarded as normal. If the saccule is larger than the utricle, this would be regarded as EH. The logical error in this analysis is the following: When the saccule is enlarged, it seems as confluent with the utricle on 3D FLAIR imaging and cannot be differentiated from the utricle (moreover, it cannot even be differentiated from the surrounding bone). Hence, it is impossible to measure the size of the utricle or of the saccule individually, and therefore it is impossible to calculate an area ratio of saccule vs utricle as they propose. Accordingly, Figure 4 in their article clearly shows a confluent vestibular endolymph compartment with no distinction between the saccule and the utricle. This method proposed by Dr. Attyé is therefore not useful for the quantification of saccule or utricle size in the dilated state of endolymphatic hydrops. A prior publication from a Chinese group (16), which was published in 2011, used a very similar MR sequence (3D FLAIR) and image analysis procedure (sagittal section analysis of saccule and utricle), and found an endolymph area ratio of 9 to 28% for the cochlea and 14 to 40% for the vestibule in 20 healthy controls, which corresponds well to the axial grading system proposed by the Nagoya group. The same Chinese group extended their report to 60 healthy controls with very similar results (17) and published those in 2015. Furthermore, they were able to distinguish between healthy controls and MD patients (18). Therefore, in summary, the inability of Dr. Attyé's group to reproduce the findings of previously published reports may have many possible reasons, including image quality, but it does not mean that the image analysis methods used by several groups (e.g., Nagoya, Munich, Zurich, Beijing) were not valid. In the first paragraph of Dr. Attyé's comment, he is wrongly referring to “the diagnosis of endolymphatic hydrops by MRI at least 4 hours after intravenous contrast media injection” by citing an article (reference No. 2 in Dr. Attyé's comment) which does not deal with intravenous contrast injection. Concerning the specific criticisms of the original article (1), the comment on Figure 2 is correct. The double arrow in this figure is indeed pointed toward the fundus and not the basal cochlear duct. This is a simple mistake with no bearing on the conclusions of this case report (which was published initially in 2014 (19)) or on the clinical management of Hydropic Ear Disease at all. It seems very unlikely that any researcher with experience in hydrops imaging would change his diagnosis because of such a mistake. Nevertheless, it is important to point out such mistakes and to carefully evaluate inner ear anatomy when analyzing MR images. For example, in the clinical study by Dr. Attyé's group (14) that deals with inner ear anatomy and different methods of image analysis, it seems that the authors did not examine the inner ear anatomy using a heavily T2-weighted sequence (MR cisternography), as it would be expected from any state-of-the-art inner ear imaging examination. This represents a serious flaw of their study, since knowledge of inner ear total fluid space anatomy (e.g., ruling out tumors, malformations, fibrosis) is essential to the interpretation of endolymph/perilymph imaging such as in Hydropic Ear Disease. In the same study, the authors report that the ampulla was visible in 32 MD patients, but their study supposedly did include only 30 MD patients, so one of the two numbers must be false. Also, the age of the included patients is described as greater than 40 years in the publication, while the study protocol published at www.clinicaltrials.gov and cited in their publication states that only patients above 50 years were included, so one of the two numbers must be false. Further, in Figure 2A published by Dr. Attyé's group in another article (20), the arrow marking the saccule seems to be placed exactly above the vestibule and therefore hides the vestibule (which is the region of interest in this figure) from the reader's eye. It would be recommended to place this arrow lateral of the saccule, where it does not overlie the region of interest (and the potentially enlarged endolympatic space). Further, in Figure 1 of another article (21) published by Dr. Attyé's group, the labels A, B and C, D are incorrectly used in the figure legend. In the same Figure 1A, a hypointense area in the center of the cochlea is described as endolymphatic duct. Due to the spiral geometry of the cochlea, this area most likely is in fact located in the modiolus and not in the endolymphatic duct. To complicate matters further, in the neighboring figure, the same hypointense area in the center of the cochlea is not marked as endolymphatic duct. Additionally, in Figure 6 of the same article, the term “endolymphatic liquid” is used in the context of a hyperintense signal on 3D FLAIR. This seems to be a confusion with perilymphatic fluid which in fact usually seems hyperintense on 3D FLAIR. Further, in another study published by Dr. Attyé's group (22), the 3D FLAIR sequence is improved by changing the inversion time (TI), and two methods of image analysis—the Nagoya grading versus the SURI grading—are compared regarding their diagnostic accuracy. For the Nagoya grading method, they found a sensitivitiy/specificity of 1/0.24, which was markedly improved to 1/0.71 after changing the TI from 2,300 to 2,400 milliseconds. The SURI method, in contrast, was negatively affected by this change in TI time, with the sensitivity/specificity dropping from 1/1 to 0.86/0.88. Nevertheless, Dr. Attyé in his comment criticizes the Nagoya grading method as “dangerous, potentially leading to abnormal false positives in healthy volunteers” while referring to a comparative study (14) which did not use this improved 3D FLAIR sequence with a quite high diagnostic accuracy as reported by himself. One should not forget that the acquisition of high-quality inner ear MR images and the image analysis procedure are two separate things, and it seems that in his argumentation, Dr. Attyé tends to confuse the two with each other. Concerning the comment “the absence of a visible saccule in Figure 4 A” which Dr. Attye claims to be “problematic”: Dr. Attyé himself has defined the absence of a visible saccule as the absence of saccular EH (22), so it makes no sense to call this “problematic.” Furthermore, in this figure (initially published in 2012 (23)), the vestibular endolymph space is not described at all in the legend; all the text in the legend is regarding the cochlea. The words “vestibule,” “sacculus,” “utriculus” are not mentioned in the figure legend. Most importantly, there is no statement made concerning the absence or presence of the saccule; the axial section was chosen to clearly display the cochlea and not the saccule. The small nondilated saccule would be visible in a neighboring section and is typical for a healthy control subject such as the one displayed in the figure. So in summary, there is nothing “problematic” in this figure at all. The comment by Dr. Attyé “(…) especially when displaying in the same illustration (Fig. 4B) smaller cochlear endolymphatic duct dilatation” does not seem to make sense. Figure 4B in fact displays a subject with mild cochlear duct dilation and has nothing to do with Figure 4A, as it is clearly described in the figure legend. The speculation by Dr. Attyé regarding Figure 2 in the original article “(…) probably due to problems of registration between the two MR sequences” makes no sense. As the figure legend clearly states, the image was obtained using a 3D Real-Inversion Recovery sequence. This sequence, first discovered by the Nagoya group (24) in 2008, has the great advantage of separate visualization of perilymph, endolymph and bone on a single image (single image acquisition). Using this sequence, the discovery of hydropic herniation into the semicircular canal was made (25), a finding that has also been reported by Dr. Attyé's group (21) (but without a reference to the original discovery). Obviously, there is no transformation or coregistration involved. This sequence has also been used previously, but in combination with MR cisternography, for MR volumetry of endolymphatic space (26). Again, in this context, it should not be forgotten that differences between the image qualities of different groups may be due to a very large number of variables including MR scanner hardware, contrast agent, route of application of the contrast agent, MR sequence, image acquisition, and processing software. A deeper discussion of the details of “pseudo inversion recovery contrast” inner ear MR imaging, which is based on more than one image acquisition and subsequent coregistration and mathematical transformation is far beyond the scope of this reply to a comment, but the interested reader is recommended to study the extensive literature published by the Nagoya group on this topic. In summary, the following key points should be noted: 1) The concept of Hydropic Ear Disease is inspired by many sources of evidence, including inner ear MRI, but stands alone and is independent of technical differences between different working groups performing inner ear MRI. 2) When comparing the image quality of different working groups, a great number of parameters have to be taken into account, not only the MR sequence and not only the MR image analysis procedure. 3) Obviously, it is important to examine the complete inner ear anatomy, including the saccule and the utricle, when evaluating images for the presence of endolymphatic hydrops. 4) Using the 3D FLAIR sequence, it is impossible to separately measure the utricle and the saccule in the presence of (significant) endolymphatic hydrops, because both the structures have the same signal intensity and they cannot be differentiated from the surrounding bone. Finally, I would like to encourage fellow inner ear researchers to promote and exercise critical and constructive scientific debate and to seek interdisciplinary and international collaboration or at least exchange of experiences, preferably through direct peer-to-peer communication, which in my experience is the most efficient way of scientific progress.
The objectives of this study were 1) to assess the influence of a cochlear implantation on peripheral vestibular receptor function in the inner ear in the implant and in the nonimplant side, and 2) to analyze a possible correlation with resulting vertigo symptoms.Prospective clinical study.Cochlear implant center at tertiary referral hospital.A total of 32 patients, aged 15 to 83 years, undergoing cochlear implantation were assessed pre- and postoperatively for caloric horizontal semicircular canal response and vestibular-evoked myogenic potentials of the sacculus, and postoperatively for subjective vertigo symptoms. Patients with vertigo were compared with patients without symptoms with regard to the findings of the vestibular function tests.Cochlear implantation represents a significant risk factor for horizontal semicircular canal impairment (P < 0.001) and sacculus impairment (P = 0.047) in the implanted ear. In eight of 16 patients with preoperatively preserved caloric response, this response was decreased postoperatively. Before surgery, 14 of 30 patients had regular vestibular-evoked myogenic responses. Two months after implantation, six patients had a new loss and another six showed depression of sacculus function on vestibular-evoked myogenic potentials testing. The impaired vestibular function did not correlate with vertigo symptoms. Function on the contralateral side remains unaffected (P > 0.05).Cochlear implantation is a relevant risk factor for damage of peripheral vestibular receptor function. Therefore, preservation not only of residual hearing function but also of vestibular function should be aimed for, by using minimally invasive surgical techniques.
Bilateral vestibulopathy (BVP) is a debilitating disorder characterized by the hypofunction of both vestibular end organs or nerves. The most frequent identifiable causes of BVP are ototoxic drug effects, infectious and autoimmune disorders. However, the majority of cases remain idiopathic. Very recently, the first discovery of a clinical case of Amiodarone-associated BVP has been reported.An overview of the literature concerning the relation between amiodarone toxicity and BVP is presented and discussed.Older reports on amiodarone-induced symptoms of vertigo and gait instability lack a description of vestibular function test results. Recent evidence from retrospective studies including vestibular function testing in patients taking amiodarone have identified the drug as the hitherto unsuspected potential cause of a relatively large proportion of cases with "idiopathic" BVP.Patients who receive amiodarone should be monitored with vestibular function testing in order to recognize potential adverse effects on the vestibular system and allow for an informed decision on possible drug reduction or withdrawal.
Objective To study the correlation between the degree of endolymphatic hydrops as detected in vivo by magnetic resonance (MR) imaging and the auditory and vestibular function in patients with definite Ménière’s disease. Study Design Prospective observational study. Setting Tertiary referral center for balance disorders. Subjects and Methods In this prospective study, 41 patients who fulfilled the criteria for definite unilateral Ménière’s disease according to the American Association of Otolaryngology–Head and Neck Surgery and who showed a summating potential–to–action potential ratio of greater than 0.4 on electrocochleography were included. Intratympanic contrast-enhanced 3 Tesla MR imaging of the inner ear was performed using a 3D Inversion Recovery Turbo Spin Echo sequence. The degree of endolymphatic hydrops was graded on a Likert scale (0–3) in the cochlea and by vestibular endolymph space ratio in the vestibulum. The degree of hydrops was then analyzed with respect to its correlation with audiometric hearing levels, interaural amplitude ratios of vestibular evoked myogenic potentials, degree of horizontal semicircular canal paresis on caloric irrigation, and disease duration. Results The degree of hearing loss and the disease duration correlated significantly with cochlear hydrops (r = 0.85; r = 0.34). Sacculus dysfunction was significantly correlated with vestibular hydrops (r = −0.47). There was no significant correlation between horizontal semicircular canal paresis and vestibular hydrops. Conclusion In patients with clinically and electrocochleographically confirmed definite Ménière’s disease, the degree of MR morphological hydrops severity correlates significantly with impairment of hearing function and sacculus function.
