Journal of Otology & RhinologyISSN: 2324-8785

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.
bahis siteleri bahis siteleri bahis siteleri casino siteleri

Research Article, J Otol Rhinol Vol: 4 Issue: 3

How Reliable is Preoperative Temporal Bone CT in Mastoid Surgery?

Neeraj Sethi*, Amy Rafferty and Ratnamala Kayarkar
Mid-Yorkshire NHS Trust, Pinderfields Hospital, Aberford Road, Wakefield, West Yorkshire, WF1 4DG, United Kingdom
Corresponding author : Neeraj Sethi
3 Victoria St, Leeds, West Yorkshire, LS7 4PA, UK
Tel: 07980-281 223
E-mail: [email protected]
Received: November 22, 2014 Accepted: April 17, 2015 Published: April 29, 2015
Citation: Sethi N, Rafferty A, Kayarkar R (2015) How Reliable is Preoperative Temporal Bone CT in Mastoid Surgery?. J Otol Rhinol 4:3. doi:10.4172/2324-8785.1000228


How Reliable is Preoperative Temporal Bone CT in Mastoid Surgery?

Objective: To evaluate the reliability of high resolution computerized tomography (HRCT) of the temporal bone in assessing local complications of active squamous chronic otitis media (cholesteatoma) preoperatively.

Methods: 156 patients having undergone mastoid surgery were identified. Prospectively collected intraoperative information was compared to preoperative findings for four specific parameters.

Results: Relatively low sensitivities for all four parameters were found ranging from 11% for facial nerve canal dehiscence to 35% for ossicular chain erosion.

Conclusions: This study emphasises the need for surgeons not to be misled by preoperative HRCT. The role of imaging in anticipating complications is by no means definitive.

Keywords: Otology; Computed tomography; Cholesteatoma; Chronic Otitis media; Radiology


Otology; Computed tomography; Cholesteatoma; Chronic Otitis media; Radiology


High resolution computed tomography (HRCT) imaging of the temporal bone is increasingly becoming routine for many otologists as part of their preoperative work-up for mastoid surgery. It remains controversial as to whether it should be performed in every case due to concerns over sensitivity and specificity [1-3]. The commonly accepted advantages include: a visual aid to pre-operative counseling of the patient and anticipation of complications of active squamous chronic otitis media (COM) also referred to as cholesteatoma [1].
Arguments that an over-reliance on the CT findings could lull surgeons into a false sense of security over the absence of a dehiscent facial nerve, tegmen or lateral semi-circular canal (LSCC) cause persistent consternation over the true value of preoperative HRCT.
Preoperative temporal bone HRCT is a routine part of the senior author’s practice. We aimed to evaluate the reliability of temporal bone HRCT in assessing anatomical variation and local sequelae of COM preoperatively.

Materials and Methods

This project was registered with the local audit department.
All adult patients having undergone mastoid surgery for acquired active squamous COM under the senior author over 60 months in the Mid-Yorkshire NHS Trust were identified.
As part of on-going audit the intraoperative findings regarding the presence of ossicular chain erosion, facial nerve canal dehiscence, lateral semi-circular canal dehiscence and the integrity of the tegmen were recorded at the time of surgery. This database was compared to the preoperative HRCT scan findings and the specificity and sensitivity calculated.
All HRCT scans were performed according to the same protocol with 1mm slice axial and coronal views. The HRCT scans were reported by 3 radiologists all of whom specifically commented on the presence, location and extent of cholesteatoma, the state of the ossicular chain, erosion of the tegmen and the integrity of the facial nerve canal and labyrinth.


Overall 157 patients were identified. One scan was deemed inadequate due to movement artefact and therefore this patient was excluded. All operations were performed by or supervised by the senior author and the intraoperative findings recorded by her personally contemporaneously.
The mean age of patients was 38 yrs (range: 18-74 yrs). Primary surgery was performed in 121 patients with 35 being revision cases. The median time lag between the HRCT scan being performed and the time of surgery was 91 days (range: 8-370 days).
Integrity of the ossicular chain
Seventy-four preoperative scans showed erosion of the ossicular chain but intraoperatively only 23/74 were found to be eroded. Whilst 82 scans showed an intact ossicular chain intraoperatively only 39/82 were found to be intact (Figure 1). This produced a sensitivity of 35% and specificity of 43%.
Figure 1: To show comparison of CT findings to intraoperative findings regarding ossicular chain erosion.
Facial nerve canal dehiscence
Dehiscence was identified on 30 HRCT scans, of these only 3/30 were found to be dehiscent at the time of surgery. One hundred and twenty-six were interpreted as intact on HRCT but 4/126 were found to be dehiscent at the time of operation (Figure 2). This gave a sensitivity of 11% and specificity of 79%.
Figure 2: To show comparison of CT findings to intraoperative findings regarding facial nerve canal dehiscence.
Lateral semi-circular canal dehiscence
One hundred and forty-nine lateral semi-circular canals were found to be intact on HRCT but 6/149 intraoperatively were found to have fistulae. Whilst 7 preoperative HRCT scans were shown to have LSCC fistulae only 1/7 was present at the time of surgery (Figure 3). This produced a low sensitivity of 14% but high specificity of 96%.
Figure 3: To show comparison of CT findings to intraoperative findings regarding lateral semicircular canal dehiscence.
Tegmen dehiscence
The tegmen was found to be intact on 139 preoperative HRCT scans but 11/139 were found to be dehiscent intraoperatively. Seventeen HRCT scans were reported as showing a dehiscent tegmen but only 2/17 were found to be dehiscent at the time of surgery (Figure 4). This produced a sensitivity of 15% and specificity of 90%.
Figure 4: To show comparison of CT findings to intraoperative findings regarding tegmen dehiscence.
Positive and negative predictive values were also calculated for these 4 parameters (Table 1).
Table 1: Positive and negative predictive values.


In the literature there is a wide variation in the reported sensitivity and specificity of HRCT of the temporal bone. In general our findings were similar to those reported (Table 2) [4-10].
Table 2: Comparison of sensitivity and specificity to those reported in the literature.
One area we did not examine was the ability of HRCT to diagnose cholesteatoma. This was based on the known limitations of CT to differentiate soft tissue disease. Previously studies have shown it to be a poor diagnostic tool in this area [11-13]. The authors would also like to stress that acquired active squamous COM is a condition that is diagnosed clinically.
Throughout this study the decision to operate was made clinically and not based on findings of the HRCT scan. The scans were used as adjuncts to provide additional information for surgical planning and to aid informed consent. This is supported by Banerjee et al’s paper in 2003 [2].
The higher specificity results compared to sensitivity suggest that whilst HRCT can be reasonably good at excluding LSCC fistula and dehiscent tegmen it is not good enough at detecting them for the surgeon to allow any complacency in surgical technique.
However, temporal bone HRCTs still does play an important role in counseling patients preoperatively and in training. They are a valuable aid in understanding the challenging three-dimensional anatomy of the temporal bone. Real benefit can be gained by using HRCT scans to plan and aid anticipation of potential complications and contemporaneously correlating clinical findings to radiological images.
The limitations of this study are recognized however. Firstly there is the impact of the time lag between the scan being performed and surgery. There was a wide variation in this ranging from 8 days to just over a year. This is likely reflected in many surgeons’ practice due to constraints of the service, waiting lists and patient procrastination. Though no definitive data is available on the growth rates of cholesteatoma, in vivo, this lag could certainly have an effect on the accuracy of HRCT findings [14].
We also raise the question over whether the length of time (and therefore the extent of disease) that the patient has had COM could affect scan accuracy. The specific inflammatory status of the patient at the time of the scan could also have influenced the scan interpretation i.e. was the ear actively discharging or was it quiescent? Trojanowska et al suggest that delayed post-contrast images provide a sensitive imaging modality for delineating cholesteatoma from granulation tissue but this was not used in our study [13].
Interobserver variation between radiologists was not comparable reliably in this study as a single radiologist reviewed each patient’s images. This may have introduced a bias into the resulting sensitivity and specificity.
Of interest the authors found that there is no clear radiological definition of what is a bony defect/dehiscence (i.e. what is the minimum size detectable by HRCT). This lack of clarity may contribute to the low sensitivity and variable specificity.


In conclusion this study highlights the potential limitations of preoperative HRCT of the temporal bone and the authors strongly advocate the view that it is a relatively rough guide to the surgical anatomy rather than a definitive study. We would like to stress that this study is not intended to cast aspersions as to the skills of the interpreting clinician but rather demonstrates the limitations of the imaging modality in this context. Anecdotally the senior author has noticed a considerable increase over the last decade in the proportion of otologists obtaining HRCT prior to mastoid surgery. Perhaps reflecting the ease of obtaining this investigation combined with increasing confidence in interpretation and awareness of litigation.
It is particularly important for trainees to recognise that the scan images do not represent a safety net and meticulous, anticipatory surgical technique is the best approach. We do however support the use of HRCT as a counselling and training tool prior to mastoid surgery. It would be mutually beneficial for both surgeon and radiologist to interpret HRCT results, correlating them with surgical findings.


Track Your Manuscript

Media Partners