Introduction

Cholesteatoma imaging has gone through many stages of development, not only in the type of imaging used but also in the use of the scan by the otologist. Computed tomography (CT) scans can be used to identify bony structures such as ossicles, lateral semi-circular canals and the facial nerve canal, and some otologists use these images before operating to familiarise themselves with any anatomical variation. However, CT differentiates poorly between different soft tissues and as such is less useful for cholesteatoma diagnosis. Granulation tissue, scar tissue, cholesterol granulomas and oedematous mucosa can all be mistaken for cholesteatoma on CT scanning, and vice versa. In CT images showing bony erosion apparently caused by a mass of soft tissue, the ability to predict cholesteatoma is of course improved. However, in cases with previous mastoid surgery the anatomy will have been altered at the initial operation, and the diagnosis is more difficult.

Non-enhanced magnetic resonance imaging (MRI) shows less detail of the temporal bone than CT. As regards temporal bone imaging, the first useful MRI technique was enhanced scanning using gadolinium contrast, which is slowly taken up by living tissue in the middle ear and mastoid.Reference Martin, Sterkers and Hahum¹ Following a 45-minute delay between contrast administration and scanning, vascular middle-ear and mastoid structures give a hyperintense signal, whereas cholesteatoma, being avascular, does not enhance at all and can therefore be distinguished from other, vascular tissue. However, this is very time-consuming, and gadolinium is also known to occasionally cause side effects.

Echo-planar diffusion-weighted MRI scans are created with a 1.5 T superconductive unit scanner. Although not always available, such scanners are commonplace in neuroradiological departments and are used for brain scanning to detect small ischaemic cerebrovascular accidents.Reference Huisman² Cholesteatoma is seen on echo-planar diffusion-weighted MRI as a hyperintense signal, distinguishing it from other middle-ear and mastoid structures. (See Figure 1) This imaging modality is therefore useful for the pre-operative diagnosis of primary or recurrent cholesteatoma when the clinical diagnosis is uncertain, and also for the post-operative detection of residual cholesteatoma following canal wall up mastoidectomy.

Fig. 1

Coronal modified echo-planar diffusion-weighted magnetic resonance imaging scan, showing primary right-sided cholesteatoma.

Unfortunately, when bone and air occur next to each other, a higher intensity MRI signal is produced, termed a susceptibility artefact. Where the temporal lobe lies next to the temporal bone and the mastoid airspace, a characteristic high intensity curvilinear artefact is often seen. As this is an area in which cholesteatomas can form, differentiation between susceptibility artefact and cholesteatoma can be difficult, making accurate diagnosis of small cholesteatomas problematic. There is also the risk of false positives, when the artefact is mistaken for cholesteatoma, and false negatives, when the cholesteatoma is hidden in the hyperintense artefact signal.

Non-echo-planar diffusion-weighted MRI uses a single shot fast (or turbo) spin echo diffusion-weighted image, which is a quicker sequence using a different magnet arrangement. Scans can take as little as 35 seconds, and images do not suffer from susceptibility artefacts. This means that small cholesteatomas of 3 or even 2 mm can be detected within the mastoid as a bright, hyperintense signal diagnostic of cholesteatoma.Reference De Foer, Vercruysse, Bernaerts, Deckers, Pouillon and Somers^3, Reference De Foer, Vercruysse, Bernaerts, Maes, Deckers and Michiels⁴ Non-echo-planar diffusion-weighted MRI has been shown to be more sensitive and specific for small cholesteatomas than echo-planar diffusion-weighted MRI.Reference De Foer, Vercruysse, Bernaerts, Maes, Deckers and Michiels^4, Reference Kodama, Yano, Tamura, Tono and Machida⁵ Non-echo-planar diffusion-weighted MRI is especially useful in detecting small residual pieces of cholesteatoma retained after canal wall up mastoidectomy, thereby avoiding the need for a ‘second look’ operation (often performed to check for residual cholesteatoma that would not be seen on otoscopy after a canal wall up mastoidectomy).

The equipment and systems needed to perform non-echo-planar diffusion-weighted MRI are in addition to those needed for standard echo-planar imaging. The main use for such equipment is for cholesteatoma detection; other clinical uses are limited. Therefore, the financial argument for purchasing such equipment may not be sufficiently strong in the current, stringent economic climate, especially in smaller departments where its use may be limited. As many departments already have echo-planar diffusion-weighted MRI units, it is logical to make the best use of these existing facilities in order to improve the quality and reliability of MR imaging.

In the out-patient clinic, the clinician often cannot be completely confident in diagnosing cholesteatoma, due to difficulty in visualising the deep canal in a discharging, infected and painful ear. It is our regular practice to undertake temporal bone imaging as a diagnostic investigation, and also to aid pre-operative planning by determining the extent of cholesteatoma.