To the best of our knowledge, the present study is the first to report the presence of different types of JB, according to the classification by Manjila and Semaan (13), and the anatomical variables detected using HRCT and 3DRC in hydropic and non-hydropic ears were confirmed by Gd-MRI, suggesting a potential relationship between the height of JBs and the non-visualization of the VA and the development of EH. There was a likelihood of presence of EH related to HJB in type 4 JB, but not in types 2 and 3. An obliteration of the VA, as determined by 3DRC, could be deemed a specific radiological sign for screening causative HJB in MD.
Meniere disease or endolymphatic hydrops has been associated with HJB in several studies (4-6). As there is no consensus on the exact definition of HJB, the results regarding its role in the etiology of MD are controversial. Redfern et al and Park et al have described a higher prevalence of JB abnormalities in patients with MD, as compared with the general population (11,18). However, numerous studies have reported no difference in the prevalence of HJB between affected and unaffected ears (19, 20). In the present study, the Manjila and Semaan classification was used to divide JBs into types1 (Fig. 1B), 2 (Fig. 1C), 3 (Fig. 3A, B and C) and 4 (Fig. 4A and B). A significantly higher prevalence of type 4 JB was found in hydropic ears, as compared with non-hydropic ears, but there were no differences in the presence of type 2 and type 3 JB and overall incidence of HJB between hydropic and non-hydropic ears. It was also identified that the height of JB, and not its size, contributed to the likelihood of the HJB-related endolymphatic hydrops. Therefore, type 4 JB with an upward extension of the bulb to a sufficient height has shown a great potential for the development of EH in some MD patients. As type 4 JB accounted for 8.7% of all JB cases in hydropic ears, the majority of JB cases in hydropic ears didn’t show a cause-consequence relation between HJB and EH.
In addition, the JB abnormalities, such as JB diverticulum (JBD) and JBVAD have been linked to a variety of cochleovestibular symptoms, depending on their impact on surrounding structures (8,21). Those without a HJB were more likely to have dehiscence or diverticulum (6). In the present study, no difference in the frequency of JB diverticulum or dehiscence was observed between hydropic and non-hydropic ears, suggesting there is no association between the presence of JBD and JBVAD, and the development of endolymphatic hydrops.
In addition, as the link between HJB and MD is based on the hypothesis that HJB disturbs the VA to the point of producing hydrops (6,9), research has focused on how the VA is affected by JB abnormalities. Several studies have radiologically demonstrated hypoplastic VAs and narrowed or obliterated endolymphatic ducts in MD (22-24). However, although the majority of the literature has shown a narrow VA, the results of certain studies could be interpreted in different ways. According to Ikeda and Sando, 21% of the normal population has a hypoplastic VA (25). Sando and Ikeda (26) also reported that 40% of patients with MD have a normoplastic or hyperplastic VA. Therefore, there are conflicting points of view regarding the role of the VA in MD. Considering the complex VA anatomy, the inconsistent results could be attributed to different imaging criteria used for the measurement of the VA. Prior studies have shown that a Pöschl plane image could reveal nearly the entire external aperture of the VA with maximum accuracy, as compared with the conventional axial plane (27, 28). In addition, as the VA is located between the common crus and PSCC, the current proposal of the Manjila and Semaan classification for JB accounts for the relationship of PSCC; therefore, it is reasonable to postulate that an upward extension of JB above the inferior margin of the PSCC is likely to erode the region of the VA, which could be shown as VA absence (non-visible type) in a Pöschl plane image (Fig. 4A). The present results demonstrated a significantly higher prevalence of non-visualization of the VA in hydropic ears, as compared with that in non-hydropic ears, suggesting that the presence of hydrops may be linked to HJB-caused VA obliteration.
However, non-visualization of the VA was not bound to VA obliteration, a non-visible type aqueduct can occur in normal ears (25, 26). As VA dimensions and JB abnormalities can vary, only a few of the structures of the VA and HJB could be viewed, and their spatial relationships could not be accurately observed on conventional CT imaging. 3DRC imaging has been used to evaluate VA and membrane labyrinth in MD patients (29, 30) and has yielded more precise images than those generated by conventional CT. In the present study, 3DRC imaging systematically displayed the detailed structures of the temporal bone which enabled us to illustrate the distinct spatial relationship of VA and JB. There was a significant difference in type III relationship of VA and JB between hydropic and non-hydropic ears, suggesting a significant association between VA obstruction and the presence of EH. No significant differences were observed in the prevalence of types I and II between hydropic and non-hydropic ears, suggesting there was no relationship between the JB neighboring VA or JB touching VA and the development of EH. In addition, in type III, 3DRC displayed the upward JB penetration of the VA or encroachment upon the VA, which caused the obliteration of its distal part (Fig. 4C and D) with the presence of EH (Fig.4E). By contrast, in types I and II, stereoscopic images showed the VA intact by means of varying the angles of view on the computer display (Fig.3D and E) without the presence of EH (Fig. 3F); this provided insights into the pathogenetic mechanisms underlying the HJB-related development of EH.
As All type IIIs were found in hydropic ears with type 4 JB in our cohort, It can be postulated that the disruption of VA may occur only when the HJB reaches to a sufficient height, such as above the inferior margin of the IAC, and such pathological feature of VA related HJB was likely to be screened with the specific radiological sign revealing by 3DRC. Furthermore, the non-visualization of the VA shown on Pöschl plane occurred in 9 (25%) hydropic and 2 (6.5%) non-hydropic ears, whereas the obliteration of the VA on 3DRC was found in 5 (13.9%) hydropic and no non-hydropic ears. This finding showed a higher accuracy and optimal specificity for evaluating VA obliteration using stereoscopic images.
Additionally, figure 4 showed the results of CT and 3DRC, and the dynamic change of EH with Gd-MRI prior to and following surgery in one of five MD patients with visualization of the obstruction of VA on 3DRC images, who was performed endolymphatic sac surgery on the right affected ear. Not surprisingly, the proximal portion of endolymphatic sac was found absent as the encroachment of HJB. Then, the blockage of endolymphatic duct was opened to drain excess endolymphatic fluids. Two weeks following surgery, Gd-MRI showed a complete reversal of hydrops in both cochlea and vestibule (Fig 4E and F), suggesting that drainage of endolymphatic duct could be an effective means of relieving hydrops. This result further confirmed an obstruction of the VA by HJB was indeed the cause of endolymphatic hydrops in some MD patients. Although the prevalence is low, the identification of this causative factor is very important for both the diagnosis of MD and surgical planning in MD treatment. As discontinuity of the VA, the ES shunting/decompression procedures, which targeted the ES to improve the fluid resorptive functions of the ES, most likely couldn’t work in this patient due to the obliterated VA that separates the ES from the other labyrinthine fluid spaces.
The present study had several limitations. First, the disruption of VA was confirmed surgically in only one patient; in other 4 patients, the lack of surgical or pathological confirmation of VA obliteration shown in Pöschl plane or 3DRC images decreasing the accuracy of the results. Secondly, the radiological finding of an intact VA on 3DRC reflected the results of the morphological evaluation of the VA; whether existence of dysfunction in these intact VA is unknown. A hypothesis has been proposed that the presence of any JB abnormality may contribute to the development of Ménière’s symptoms, presumably by interfering with neighboring inner ear structures to alter the direction of endolymphatic flow or ES venous drainage (6, 31). The present results still need to be confirmed by further studies.