Abstract
Objective
Superior canal dehiscence is defined by missing bony coverage of the superior canal against the middle cranial fossa. The gold standard in diagnosis is high-resolution computed tomography (CT). A false-positive CT scan, identifying a dehiscence when one is not present, could lead to unnecessary surgical therapy. This study aims to compare postmortem CT scans with autopsy findings with regard to superior canal dehiscence.
Study Design
Postmortem study.
Setting
Tertiary referral center.
Subjects and Methods
Twenty-two nontraumatic death cases within a 3-month period (January to March 2017) were included with 44 temporal bones. Each body underwent postmortem head CT prior to medicolegal autopsy. The middle fossa floor was exposed, and if present, the superior semicircular canal dehiscence was identified and measured. In each case, 3 comparable photographs were taken during the autopsy (left temporal bone, right temporal bone, overview).
Results
Autopsy findings revealed bony dehiscences in 11% of the temporal bones, whereas CT scan revealed bony dehiscences in 16%. The length of the dehiscences were longer when measured by CT imaging.
Conclusion
The diagnosis of superior canal dehiscence syndrome requires high-resolution CT with clinical symptoms and physiologic evidence of a third mobile window. Our study underlines a mismatch between multislice CT imaging in the coronal plane and the presence of a dehiscence on autopsy.
Superior canal dehiscence (SCD) was first described by Minor et al in 1998, and it is characterized by missing bony coverage of the superior canal against the middle cranial fossa. 1 Patients with SCD syndrome may present with different vestibular and audiologic symptoms. In addition to oscillopsia and rotational vertigo, Tullio phenomenon, Hennebert sign, and Valsalva-related vertigo can be present with vestibular symptoms.1,2 Hyperacusis to bone-conducted sound, autophony, low-frequency air-bone gap, and negative bone conduction thresholds are characteristic.1,3 Pathophysiologically, the symptoms are assumed to be related to a third window phenomenon, which should lead to a modified labyrinthine perception of noise and pressure-related changes inside the labyrinthine fluid system.1,4-6
In a study of 1000 temporal bones among 596 people, incidence of bony dehiscence was 0.5%. 7 With clinical symptoms, the radiologic evaluation (high-resolution, flat panel, or cone beam computed tomography) with a maximum slice thickness of 0.5 mm is important to detect a superior semicircular canal dehiscence. Besides radiologic findings, further diagnostic points (eg, pure tone audiogram, cervical vestibular evoked myogenic potential) have to be observed to determine that pressure is being transmitted via the dehiscence, causing symptoms.8,9
The association of the SCD length and the signs and symptoms of SCD syndrome is currently under discussion. A larger defect was shown to correlate with vestibulocochlear manifestations, 10 while other studies did not find an association between the SCD length and clinical presentation.11-13 Audiologically, the air-bone gap increased with a larger SCD,13-15 while other studies found no correlation between the SCD length and the degree of hearing loss.12,16 Some of these studies comparing length with the air-bone gap used CT data,12,14 and others used direct intraoperative observation via the middle fossa. 13 The location of the dehiscence closer to the ampulla seems to have a greater impact on auditory and vestibular symptoms. 15 Relatively small sample sizes and varied methods of measuring SCD length might explain these varying results.
The surgical therapeutic options include transmastoid canal occlusion or occlusion via the middle fossa. Both surgical strategies showed good results with a control of vertigo and hearing preservation.17,18
In the transmastoid approach, the dehiscence itself cannot be seen, whereas via the middle fossa, the dehiscence can be identified and plugged or sealed. However, the dehiscence might be challenging to identify due to missing anatomic landmarks. 19 In addition to contributing to the diagnosis, radiologic scans can affect preoperative planning.
The present study aims to evaluate the accordance of the radiologically determined size of the superior semicircular canal dehiscence and the real anatomic dehiscence.
Material and Methods
The study was approved by the institutional review board at the Unfallkrankenhaus Berlin (IRB-ukb-HNO-2017/01) and was conducted according to the principles expressed in the Declaration of Helsinki.
Twenty-two death cases within a 3-month period (January to March 2017) were included in this prospective autopsy study, including both temporal bones from each body (n = 44). The inclusion criteria were absence of putrefaction and any known head trauma and/or surgery prior to death, completely removed dura from the middle fossa, and completed CT scan. Prior to medicolegal autopsy, each body was scanned with a 16-slice CT scanner (Toshiba Activion; Toshiba Medical Systems GmbH, Neuss, Germany) with a slice thickness of 0.5 mm and an overlap of 0.3 mm. Reconstructions of 2- and 3-dimensional volume-rendered images were performed with the accessory software as described previously.20,21 The acquired CT images were analyzed and interpreted by an experienced neuroradiologist. The CT scans were reformatted coronal to the longitudinal axis of the temporal bone in the Poschl view. The images were obtained in a separate bone window in addition to a soft tissue window.
After the CT scan, all bodies underwent a forensic autopsy according to current German guidelines. The skull cap was opened circular with an oscillating bone saw, and the brain and the dura were removed carefully with an elevator. The middle fossa was exposed, and the superior semicircular canal dehiscence was identified and measured if present. Despite the putridity of the specimens, soft tissue remained in contact with bone over the superior semicircular canal. Three photographs were taken in each case (right temporal bone, left temporal bone, overview of the skull base) with a Sony α6000 camera (24.3 megapixels; Sony Corp, Thailand) and a Sony objective (SELP 1650). Each picture was approximately 5 MB in size with a resolution of 350 dpi. The medical information about audiologic, vestibulocochlear, or neurologic symptoms was not available.
Results
Eighteen bodies were male and 4 bodies were female, with a mean age of 43 years (range, 23-75 years). The causes of death included intoxication (n = 1), hanging (n = 16), drowning (n = 1), smoke gas inhalation (n = 2), and hemorrhagic shock (n = 2). Of the 44 temporal bones, 2 showed radiologic dehiscences of the superior semicircular canal on the right side and 5 on the left side ( Figure 1 ). One case was identified with a bilateral dehiscence. In total, 16% (7 of 44 temporal bones) showed bony dehiscences on CT scans.
Reformatted 2-dimensional computed tomography scan of specimen 14 of the left side: Poschl view. (a, b) The white arrows point to the dehiscence; (b) the black arrow points to the superior semicircular canal.
Autopsy findings revealed bony dehiscences in 11% (5 of 44 temporal bones; Figure 2 ). Three dehiscences of the superior semicircular canal were found on the left side and 2 on the right, whereas 1 specimen showed a bilateral dehiscence. Only in the case with the bilateral dehiscences did radiology and autopsy findings correlate ( Figures 3 and 4 ). In the other cases, autopsy and radiology findings did not match ( Figures 5 and 6 ). Fisher’s exact test was conducted between the radiologic and autopsy findings of the bony dehiscences. There was not a statistically significant association between the radiologic and autopsy findings (P = .182). The sensitivity for CT scan identifying a dehiscence was 40%, and the specificity was 87%.
Autopsy findings of the left middle fossa of specimen 14. A bony dehiscence of the superior semicircular canal was not seen, despite the radiologic findings in Figure 1 . A, anterior; IAC, internal auditory canal; L, lateral; M, medial; MF, middle fossa; P, posterior; PF, posterior fossa.
Autopsy findings of the right middle fossa of specimen 19. A bony dehiscence of the superior semicircular matched with the radiologic findings in Figure 4 . The arrow points to the dehiscence.
Reformatted 2-dimensional computed tomography scan of specimen 19 of the right side, coronal to the longitudinal axis of the temporal bone. The white arrow points to the dehiscence, the black arrow to the labyrinth.
Autopsy findings of the left middle fossa of specimen 4. A bony dehiscence of the superior semicircular was seen during autopsy (circled) but not on the computed tomography scan. A, anterior; IAC, internal auditory canal; L, lateral; P, posterior.
Reformatted 2-dimensional computed tomography scan of specimen 4 of the left side, coronal to the longitudinal axis of the temporal bone. The black arrow points to the intact semicircular canal.
The total length of the bony dehiscences in the autopsy findings was 1 mm in 1 left temporal bone and 2 mm in the other temporal bones. In the radiologic evaluation, 3 dehiscences measured 1 mm; 1 measured 1.5 mm; and 1 measured 2 mm. Both temporal bones where autopsy and radiologic findings matched up had 2 mm in autopsy and 2.5 mm in radiologic evaluation ( Table 1 ).
Overview of All Subjects. a
Abbreviations: CT, computed tomography; F, female; M, male; PM, postmortem.
Values are presented as semicircular canal dehiscence in millimeters. Blank cells indicate no findings.
Discussion
SCD syndrome is characterized by missing bony coverage to the middle fossa. Symptoms may vary, but they include audiologic and vestibular symptoms.
The gold standard to detect the SCD is the high-resolution CT of the temporal bone. If the clinical presentation is in accordance with the radiologic findings, surgical therapy can be an option. 22 The surgical planning will be based on the CT scan. In our study, we found differences between the radiologic and autopsy findings ( Figures 1 and 3 ). Only 2 dehiscences detected by radiology were discovered on autopsy. Radiologically, 4 temporal bones showed small bony coverage over the superior semicircular canal. The bony patency of the middle fossa was confirmed in these cases. Different authors stated a relation between the size of the dehiscence and the clinical appearance. An extended defect can correlate with vestibulocochlear manifestations,9,13 while other studies did not find an association between the SCD length and the clinical presentation or degree of hearing loss.8,11 In our study, clinical symptoms were not part of the investigation, as medically relevant information was not available. The bony dehiscences were between 1 and 2 mm in total length. In comparison with the radiologic evaluation, our autopsy findings showed a smaller dehiscence. This is in line with the study by Tavassolie et al, who demonstrated that the real size of a bony dehiscence is usually smaller than that predicted by the CT scans. 8 Furthermore, they found 2 patients with radiologic dehiscences that were not seen intraoperatively but did show typical symptoms of the SCD sydrome. 8
The transmastoid occlusion of the superior canal is a surgical option in cases in which patients are suffering from symptoms of SCDS. Besides the transmastoid occlusion, an access above the middle fossa is described, including a resurfacing or occlusion of the superior semicircular canal. The middle fossa approach 23 gives the surgeon a direct view of the bony dehiscence but may pose greater risk to the patient. The transmastoid approach is safe, highly effective, and very familiar for otologists. 24 This approach could pose less risk for the patient but does not permit a direct view of the dehiscence. Nevertheless, resolution of symptoms is more important than the approach.25-27
At autopsy, we were able to identify 5 dehiscences, while CT scans identified 7 dehiscences. This makes up 11% of autopsy findings and 16% of radiologic findings. Compared with histologic or radiologic studies, which state the prevalence of SCD to occur in 0.5% to 3%, our study population shows a clearly higher prevalence of SCD.7,28 The small number of individuals included in this study might bias these findings.
Postmortem CT is a standard instrument in autopsy practice. The advantage of this virtual autopsy is the generation of a 3-dimensional preautopsy snapshot of the body with excellent visualization of skeletal pathologies. 29 Despite these advantages, the virtual autopsy cannot replace the traditional autopsy. Our findings underline the importance of the autopsy as it outmatches the virtual autopsy.
Our study has some limitations. Medical records were not available for all enrolled specimens. Neither was it known if the specimen with SCD had any auditory or vestibular symptoms. Our study population was limited to specimens with nontraumatic causes of death and made up only 22 specimens with 44 temporal bones. Nevertheless, in our study population, 11% of the temporal bones had bony dehiscences. This contrasts with the literature and may be explained by the small cohort. Furthermore, the quality of our CT scans is not typical for the radiologic diagnosis of SCD in clinical practice.
Conclusion
Superior semicircular dehiscence syndrome should be diagnosed by clinical and radiologic findings. We found a divergence between radiologic and autopsy—that is, anatomic findings in terms of occurrence and length.
Author Contributions
Disclosures
Footnotes
No sponsorships or competing interests have been disclosed for this article.