Coronal computed tomography prevalence of superior semicircular canal dehiscence

Coronal computed tomography prevalence of superior semicircular canal dehiscence ROBERT A. WILLIAMSON,

MD,

JEFFREY T. VRABEC,

MD,

NEWTON J. COKER,

OBJECTIVES: The relatively new clinical entity superior canal dehiscence syndrome (SCDS) is diagnosed by clinical symptoms and signs. Coronal computed tomography (CT) has been used to confirm the diagnosis. A consecutive series of temporal bone CT scans was reviewed to define the prevalence of a dehiscent-appearing superior semicircular canal. STUDY DESIGN AND SETTING: Temporal bone CT scans performed over a 2-year period at a university-based tertiary referral center were reviewed independently by 3 individuals. Scans were excluded if coronal images were not obtained or reconstructed from axial images. Prevalence figures for dehiscent-appearing superior semicircular canal were determined by consensus. Medical records of selected individuals with a dehiscentappearing canal were reviewed for study indications and otologic symptoms. RESULTS: A dehiscent-appearing superior semicircular canal was seen in 9% of studies. Correlation among examiners was greater than 94%. Medical records indicated symptoms suggestive of or compatible with the diagnosis of SCDS in rare cases. CONCLUSION: The prevalence of a dehiscent-appearing superior semicircular canal on coronal CT of the temporal bones performed with 1.0-mm collimation is substantially greater than that predicted by temporal bone histologic study. Clinical symptoms compatible with the diagnosis were seldom recorded, suggesting low specificity. The high sensitivity and low specificity of CT scan create a risk for overdiagnosis of SCDS if the coronal CT scans are not correlated with clinical symptoms. (Otolaryngol Head Neck Surg 2003;129:481-9.)

S uperior canal dehiscence syndrome (SCDS) is a newly described inner ear disorder characterized From the Bobby R. Alford Department of Otorhinolaryngology and Communicative Sciences (Drs Williamson, Vrabec, and Coker) and the Department of Radiology (Neuroradiology), The Methodist Hospital, Baylor College of Medicine (Dr Sandlin). Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, San Diego, CA, September 22-25, 2002.

MD,

and MARLIN SANDLIN,

MD,

Houston, Texas

by specific neurotologic signs and symptoms. As described by Minor1 and Cremer et al,2 patients with SCDS develop sound- and/or pressure-induced vertigo associated with predictable verticaltorsional eye movements aligned with the plane of the affected superior semicircular canal (SSCC). They may also experience chronic dysequilibrium and unsteadiness. Performance of various maneuvers that alter middle ear or intracranial pressure typically evokes symptoms. Confirmation of the diagnosis is often made with high-resolution temporal bone computed tomography (CT) imaging through the labyrinth, which may depict absence or abnormal thinning of bone overlying the SSCC.1,3-5 At times, individuals without a classical clinical history of SCDS will have CT findings compatible with the diagnosis. In these cases, the clinician and neuroradiologist must interpret imaging data and determine if the CT scan is overly sensitive and also correlate CT findings with the clinical signs that are present. However, neither the clinical nor radiographic prevalences of a dehiscent SSCC are known, and the risk of erroneous or overinterpretation of imaging data exists.6 Additional insight may be gained by defining the radiographic prevalence of a dehiscent SSCC in routine temporal bone CT imaging and in comparing these data with previously-published temporal bone histopathologic studies on SSCC dehiscence.7,8 STUDY DESIGN AND METHODS Study design and methodology were reviewed and approved by the Institutional Review Board of Baylor College of Medicine. The Department of Reprint requests: Jeffrey T. Vrabec, MD, Bobby R. Alford Department of Otorhinolaryngology and Communicative Sciences, Baylor College of Medicine, 6550 Fannin, Suite SM1727, Houston, TX 77030; e-mail, jvrabec@bcm. tmc.edu. Copyright © 2003 by the American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. 0194-5998/2003/$30.00 ⫹ 0 doi:10.1016/S0194-5998(03)01391-3 481

482 WILLIAMSON et al

Radiology database at the Methodist Hospital (the primary tertiary-referral hospital of Baylor College of Medicine) was searched to identify all patients who had undergone temporal bone CT imaging during the period between January 2000 and December 2001. High-resolution temporal bone CT imaging protocols were performed with the following parameters on GE Hi-Speed CTI scanners (GE Medical Systems, Milwaukee, WI): 9.6 cm FOV, 1 mm slice thickness, zero interslice gap, 120 kV, 500 mA, and 512 ⫻ 512 matrix, with edge algorithm. Digital files were retrieved and transferred to a GE Pathspeed Workstation (GE Medical Systems, Version 8.1, 2001) for review. All available studies were reviewed to detect dehiscence of the SSCC. Studies were excluded if no coronal images were obtained, if coronal images were reconstructed from axial data, if significant motion artifact was present, or if prior surgical procedures or disease process (eg. labyrinthectomy, middle fossa craniotomy for vestibular schwannoma, or labyrinthine ossification) significantly altered or distorted relevant labyrinthine anatomy. The resolution limit of the image acquisition parameters described above is 0.324 mm.9 Images were reviewed independently by 3 different examiners (2 neurotologists and 1 neuroradiologist), all blinded to the clinical status of the patient and indication for CT imaging. Each superior canal was classified as dehiscent or normal. A study was deemed positive for dehiscence by consensus of 2 of the 3 examiners. Correlation between examiners’ individual interpretations of scans was also determined. Medical records were then reviewed in consensus cases to determine the indication for CT examination and whether any symptoms or signs of SCDS were present. RESULTS The total number of temporal bone CT scans identified over the 2-year period was 272. From this group, 20 were excluded due to lack of data acquisition in the coronal plane, 7 were excluded for motion artifact, and 22 were not available. An additional 4 temporal bones were individually excluded for anatomic reasons: 3 with labyrinthine ossification and 1 for labyrinthectomy. Thus, a total of 442 temporal bones were examined from 223 different individuals. The mean age of the

Otolaryngology– Head and Neck Surgery November 2003

study group was 45 years, with a range of 7 to 87 years. There was an equal distribution according to gender. CT scans were obtained for various indications; the most common diagnoses were cholesteatoma, hearing loss (including sensorineural hearing loss), and chronic otitis media. Table 1 lists the relevant clinical data in those cases with radiographic dehiscence. A total of 39 (9%) temporal bones were determined to have a radiographically dehiscent SSCC. This prevalence is markedly greater than a previously published temporal bone histologic survey of SSCC dehiscence (P ⬍ 0.001).7 There are no other published reports on radiographic prevalence available for comparison. An additional 15 temporal bones (3%) were judged to be dehiscent by a single examiner. Correlation among examiners was greater than 94%. Of the total 39 temporal bones determined to be dehiscent, 14 (36%) were found to be dehiscent at the apex or superiormost aspect of the SSCC (Fig 1). Fourteen others (36%) were dehiscent along the posterior aspect of the SSCC, often in the region of the groove formed by the superior petrosal sinus (Fig 2). The remaining 11 (28%) were found to be dehiscent along the anterior aspect of the SSCC (Fig 3). Nine individuals had bilateral dehiscence, accounting for over 46% of the cases. Medical records were available in 26 of the 30 individuals with radiographic dehiscence. None of the CT scans in those cases were ordered for suspicion of SCDS. One patient with nonspecific dizziness as a presenting complaint underwent CT scan during workup; however, a history of Tullio phenomenon, fistula testing, and provocative maneuvers were all negative. A second patient with fluctuating sensorineural hearing loss and aural fullness underwent contrast-enhanced CT (in place of magnetic resonance imaging) and subsequently responded well to standard therapy for Me´ nie˘ re’s disease. Of the remainder, there were no cases in which symptoms of low-frequency hearing loss, pressure-induced vertigo, Tullio phenomenon, or motion intolerance/dysequilibrium was recorded. This retrospective review cannot assess whether direct questioning for these symptoms was performed.

Otolaryngology– Head and Neck Surgery Volume 129 Number 5

WILLIAMSON et al 483

Table 1. Clinical data from patients with evidence of SSCC dehiscence on temporal bone CT images Patient

Age (yr)

Gender

Indication for CT scan

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

67 16 57 35 54 54 30 57 57 20 65 47 53 72 76 57 64 61 40 55 49 65 51 66 35 47 53 65 35 45

F M F F M F F F F F M M F M M M F F F F F M F M F F F M M M

Pulsatile tinnitus Lympanosclerosis SNHL (Meniere’s) SNHL Trauma Petrous apex lesion Trauma Facial paralysis COM EAC stenosis Otorrhea Otorrhea Cholesteatoma Cholesteatoma Petrous apex lesion Pulsatile tinnitus Cholesteatoma COM COM Cholesteatoma Skull base osteomyelitis COM SNHL Pulsatile tinnitus Facial paralysis Dizziness Tinnitus with hearing loss Acoustic neuroma CHL Cholesteatoma

Right SSCC dehiscence

Left SSCC dehiscence

Anterior Anterior Apex Anterior

Anterior Anterior

Apex Apex Posterior Posterior

Anterior Posterior Apex Anterior Anterior

Posterior Posterior Posterior Anterior Posterior Apex Anterior Posterior Posterior

Posterior Apex Apex Apex Posterior Apex Apex Posterior

Apex Posterior Apex Apex Posterior Anterior Apex

SSCC, Superior semicircular canal dehiscence; CT, computed tomography; SNHL, sensorineural hearing loss; COM, chronic otitis media; EAC, external auditory canal; CHL, conductive hearing loss.

DISCUSSION Radiographic evidence of SSCC dehiscence in this study of temporal bone CT scans obtained over a 2-year period occurred in 39 of 442 temporal bones, for a radiographic prevalence of 9%. An additional 15 (3%) bones were judged to be dehiscent by 1 examiner. Nine individuals who underwent CT scanning were found to have bilateral dehiscence. These figures are markedly greater than the prevalence of a dehiscent SSCC noted in temporal bone histopathologic specimens (P ⬍ 0.001) reviewed by Carey et al.7 In that study, histopathologically confirmed dehiscence occurred in 0.5% of temporal bone specimens, with an additional 1.4% of specimens noted to have markedly thinned bone (⬍0.1 mm) overlying the SSCC either at the superiormost aspect of the

superior semicircular canal or posteriorly in the region of the superior petrosal sinus. Although those investigators thought their estimates of SSCC dehiscence to be conservative, the radiographic prevalence of dehiscence appears to be significantly higher. The inability to accurately resolve bone with a thickness of less than 0.324 mm is responsible for this disparity. This would suggest a tendency for overdiagnosis of SCDS if radiographic criteria are too heavily relied on. It is also useful to compare bilateral involvement and location of SSCC dehiscence noted in this study with that of Carey et al.7 Of 596 individuals studied histologically, marked thinning or dehiscence was noted in 12 (2%). Of those individuals, 50% were noted to have bilateral involvement. Radiographically, we noted bilateral in-

484 WILLIAMSON et al

Otolaryngology– Head and Neck Surgery November 2003

Fig 1. Representative coronal CT images from patients with dehiscence at the apex or superiormost aspect of the SSCC. (A) Left temporal bone (arrow indicates dehiscent SSCC). (B) Left temporal bone from a different patient (arrow indicates dehiscent SSCC).

Otolaryngology– Head and Neck Surgery Volume 129 Number 5

WILLIAMSON et al 485

Fig 2. Representative CT images from patients with dehiscence at the posterior aspect of the SSCC. (A) Left temporal bone, coronal view (dehiscence noted by arrow). (B) Right temporal bone, coronal view, from the same patient (bilateral dehiscence on coronal images, noted by arrow). (C) Axial view from a different patient, with right SSCC dehiscence (arrow) along posterior groove formed by superior petrosal sinus.

Otolaryngology– Head and Neck Surgery November 2003

486 WILLIAMSON et al

Fig 2. (Continued).

volvement in 9 of 30 individuals with SSCC dehiscence, which compares favorably with histologic data. Regarding location of dehiscence, 14 (36%) of 39 were localized to the superiormost aspect of the SSCC. In addition, 14 (36%) of 39 of dehiscences occurred in the region of the superior petrosal sinus. Histologically, 4 (80%) of 5 confirmed dehiscences occurred in the region of the groove formed by the superior petrosal sinus. The finding of dehiscence along the anterior aspect of the SSCC in this study, a site not described in the anatomic study, immediately raises the issue of imaging artifact. The anterior SSCC dehiscence may be due to partial volume-averaging artifact in the imaging process (discussed later). There are several sources of potential error in both CT imaging technique and also in image interpretation that must be considered. Ideally, a temporal bone imaging protocol should capture coronal data along a reference axis perpendicular to the patient’s hard palate. The patient is typically positioned prone with the neck extended in the gantry, and the imaging sequence is then devel-

oped based on scout film information. In practice however, the positioning of the patient’s head varies with each CT scan, and consequently, there is a variable and nonstandardized amount of angulation present in coronal images. The CT images may be oriented tangential to the curvature of the SSCC such that overlying bone signal may not be captured in the slice data (Fig 4). This likely increases partial volume-averaging effects and results in a higher rate of error when interpreting scans for dehiscence, especially along the anterior curvature of the SSCC. To overcome this problem, we recommend modifying CT imaging protocols in suspicious cases to use 0.5-mm slice thickness, with zero interslice gap, as a means of reducing volume-averaging effects and improving accuracy. This modification was used by Minor,1 where ultra-high-resolution temporal bone CT scans were obtained using 0.5-mm thin collimation with 0.8-mm incremental reconstruction performed in the coronal plane. Belden et al10 evaluated standard temporal bone CT scans that were suspicious for SSCC dehiscence and repeated im-

Otolaryngology– Head and Neck Surgery Volume 129 Number 5

WILLIAMSON et al 487

Fig 3. Representative coronal CT images of dehiscence along the anterior aspect of the SSCC. Dehiscence in this location has not been described in a prior temporal bone histologic survey.7 Left temporal bone (dehiscence noted by arrow).

aging using 0.5-mm collimation. The data were also reconstructed parallel to the plane of the SSCC, allowing visualization of the entire ring of the SSCC and its surrounding bone. This modified protocol was found to improve the specificity and positive predictive value of CT imaging in the diagnosis of SCDS. The observations in our study would appear to agree with those findings. Clinicians and neuroradiologists should therefore interpret CT images cautiously, especially when a dehiscence is radiographically apparent at a location (especially anterior to the apex region) along the SSCC that has not been noted to occur histologically. It may also be helpful to review scout film data in each suspicious case to more fully appreciate the patient’s head position (Fig 4B) before interpreting CT images. Finally, imaging findings should be carefully correlated with clinical symptoms and signs before any treatment is recommended. The radiographic prevalence of SSCC dehiscence in this study was calculated based on temporal bone CT scans acquired at a university-

based tertiary referral center over a 2-year period. It is possible that our data vary from the actual radiographic prevalence by nature of the tertiary referral process and also by the availability of experienced clinicians and neuroradiologists to review imaging data. There may also be a learning curve that has not been accounted for in the process of interpreting CT scans for the presence of SSCC dehiscence. Regardless, the difference between radiographic and histologic prevalence appears to be very large, and the risk for overinterpretation still exists and should be considered carefully. CONCLUSIONS Radiographic dehiscence of the SSCC is present in at least 9% of coronal CT scans of the temporal bone, significantly exceeding the prevalence of 0.5% noted in a histologic survey. In many scans, this likely represents an imaging artifact as patient records rarely indicated any symptoms or signs suggestive of SCDS. Caution is advised to avoid overdiagnosis of SCDS based on imaging data alone.

488 WILLIAMSON et al

Otolaryngology– Head and Neck Surgery November 2003

Fig 4. Effect of gantry angle on imaging of the SSCC. (A) True coronal plane (black line) is perpendicular to the hard palate and likely perpendicular to the central axis of the SSCC. A typical gantry angle is illustrated by the gray line. This can alter interpretation of SSCC dehiscence, especially anteriorly, due to partial volume-averaging effects. (B) Typical scout view used for determining coronal CT imaging sequence. The black line represents the true coronal plane; the white line is the coronal plane that is actually used within the gantry to capture image data. There is a difference of approximately 47 degrees between the 2 planes. This patient’s scan was interpreted to have bilateral anterior SSCC dehiscence.

Otolaryngology– Head and Neck Surgery Volume 129 Number 5

REFERENCES

1. Minor LB. Superior canal dehiscence syndrome. Am J Otol 2000;21:9-19. 2. Cremer PD, Minor LB, Carey JP, et al. Eye movements in patients with superior canal dehiscence syndrome align with the abnormal canal. Neurology 2000;55:1833-41. 3. Brantberg K, Bergenius J, Mendel L, et al. Symptoms, findings and treatment in patients with dehiscence of the superior semicircular canal. Acta Otolaryngol 2001;121: 68-75. 4. Mong A, Loevner LA, Solomon D, et al. Sound- and pressure-induced vertigo associated with dehiscence of the roof of the superior semicircular canal. AJNR Am J Neuroradiol 1999;20:1973-5. 5. Ostrowski VB, Byskosh A, Hain TC. Tullio phenomenon with dehiscence of the superior semicircular canal. Otol Neurotol 2001;22:61-5.

WILLIAMSON et al 489

6. Minor LB, Solomon D, Zinreich JS, et al. Sound- and/or pressure-induced vertigo due to bone dehiscence of the superior semicircular canal. Arch Otolaryngol Head Neck Surg 1998;124:249-58. 7. Carey JP, Minor LB, Nager GT. Dehiscence or thinning of bone overlying the superior semicircular canal in a temporal bone survey. Arch Otolaryngol Head Neck Surg 2000;126:137-47. 8. Tsunoda A. Arcuate eminence in Caucasian populations. J Laryng Otol 2001;115:9-13. 9. Product Data Sheet, LightSpeed 16 CT Scanner System, GE Medical Systems: Milwaukee, WI. Available from: http://www.ge-medicalsystems.com. 10. Belden CJ, Weg N, Minor LB, et al. CT evaluation of bone dehiscence of the superior semicircular canal as a cause of sound- and/or pressure-induced vertigo. Radiology 2003;226:337-43.