Abstract
This paper summarizes the value of diffusion-weighted magnetic resonance imaging in the evaluation of temporal bone pathology. It highlights the use of different types of diffusion-weighted magnetic resonance imaging in the different types of cholesteatoma, prior to first stage surgery and prior to second look surgery. The value of diffusion-weighted magnetic resonance imaging in the evaluation of pathology of the apex of the petrous bone and the cerebellopontine angle is also discussed.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
Introduction
Image evaluation of diseases of the temporal bone has till now mainly been guided by clinical and audiological findings. One can state that conductive hearing loss is mainly evaluated using computed tomography (CT) and that sensorineural hearing loss is mainly examined using magnetic resonance imaging (MRI).
Most pathologies causing conductive hearing loss are situated in the middle ear whereas most causes of sensorineural hearing loss are found in the inner ear or in the central auditory pathways.
However, this somewhat artificial subdivision in image evaluation of pathology of the temporal bone has become more vague for the past few years due to newer technical developments in MRI, such as diffusion-weighted (DW) MRI.
MRI and more specifically DW MRI have gained increasing importance in the evaluation of pathology of the entire temporal bone region.
This review aims to provide an overview of the current status of the published data on DW MRI in the temporal bone region.
A short introduction of this technique is provided followed by the current status in the evaluation of middle ear lesions and cholesteatoma more specifically. The appearance of the most frequent lesions of the petrous bone apex and cerebellopontine angle (CPA) on DW MRI will also be discussed.
Diffusion-weighted magnetic resonance imaging
The mechanism of DW MRI is based on the Brownian motion of water molecules in tissue and, more importantly, on the hindrances/facilitations of water molecule movements in various types of tissue. In order to make an MRI sequence sensitive to the diffusion of water molecules, the sequence is expanded with a diffusion-sensitizing gradient scheme, usually a very fast, single-shot gradient-echo data collecting sequence (echoplanar—EP). The amount of diffusion sensitizing applied is usually indicated by the b-value. In clinical practice, images are generally acquired with a b-value of 0 and 1,000 s/mm2 [1, 2].
However, numerous artefacts can be generated during acquisition of DW MRI, such as eddy current artefacts, susceptibility artefacts, ghosting artefacts, chemical shift, and motion artefacts [3]. With the use of higher magnetic fields, these artefacts and image distortions on EP DW imaging are even more pronounced.
Due to the low incidence of movement artefact, the high brain homogeneity and high signal-to-noise ratio research at the onset was mainly focused on the brain.
This resulted in the fact that clinically, DW MR imaging is an established method used routinely for the diagnosis of acute stroke [1, 2].
However, over the last few years, several extracranial applications of DW MRI have been developed. Major applications have been described in the diagnosis and treatment follow-up of tumoral lesions in the upper abdomen, pelvis, and head and neck region [1].
In the temporal bone region, the interface between air, bone, and the temporal lobe is particularly prone to susceptibility artefacts [3]. For the evaluation of the temporal bone region, several types of non-echoplanar-based diffusion-weighted sequences have been described [3–5]. These turbo spin echo (TSE) or fast spin echo (FSE)-based diffusion-weighted have a higher spatial resolution, generate thinner slices (down to 2 mm), and do not suffer at all from susceptibility artefacts [3]. The single-shot turbo spin echo diffusion-weighted sequence uses a 180° radiofrequency refocusing pulse for each measured echo, which explains the reduction of the susceptibility artefact. It permits fast multiplanar imaging in artefact-prone regions, such as the posterior fossa and the inferior frontal and temporal lobes [3].
Cholesteatoma
Congenital versus acquired cholesteatoma
Congenital cholesteatoma
Congenital cholesteatoma is a rare entity originating at the time of closure of the neural tube in which the ectoderm—the later skin—gets entrapped in the skull, extradurally, in the temporal bone [6, 7].
When the ectoderm gets entrapped in the skull intradurally, the term epidermoid cyst is used. The term epidermoid tumor is considered a misnomer as this entity is regarded as a congenital anomaly rather than a tumoral entity.
Both congenital cholesteatoma and epidermoid cyst are the same histological entity, comprised of entrapped ectoderm or skin. The lesion starts to expand by the continuous exfoliation and desquamation of the epithelium, giving rise to a slowly expansile lesion with pressure changes on the surrounding tissues. Both lesions contain exfoliated skin or keratin inside.
Congenital cholesteatoma can be found anywhere in the temporal bone pyramid. In the middle ear, it is most frequently found in the anterior superior part of the middle ear. It can extend posterior towards the ossicular chain with possible associated erosion of the ossicular chain [6, 7] (Fig. 1). By definition, a middle ear congenital cholesteatoma is found behind an intact tympanic membrane without any signs of associated infection [6, 7].
Congenital cholesteatoma can also be found in the region of the inner ear and membranous labyrinth in which it has a predilection for the region around the geniculate ganglion [6, 7] (Fig. 2). Very often, it has a component protruding in the middle ear and a component invading the membranous labyrinth. The component protruding in the middle ear is often causing a conductive hearing loss due to impingement and/or erosion of the ossicular chain. Clinically, the inner ear component is very often causing signs of facial nerve palsy and/or sensorineural hearing loss. Debate still exists whether an isolated petrous bone apex congenital cholesteatoma exists, whether a petrous bone apex cholesteatoma is not always a middle ear cholesteatoma, extending into the petrous bone apex (Fig. 3).
Due to their location, surgical treatment of these congenital cholesteatoma is often very difficult. Care should be taken that in case of a component protruding in the middle ear, the inner ear component is resected in toto. Otherwise, the inner ear part of the congenital cholesteatoma will continue to grow further.
Acquired cholesteatoma
The acquired cholesteatoma is originating at a so-called retraction pocket most frequently situated at the upper and posterior part of the tympanic membrane or the pars flaccida.
This retraction pocket gradually starts to fill up Prussak’s space, and when it gets sealed off, this retraction pocket is gradually starting to enlarge by the continuous desquamation of the epithelium, the so-called cholesteatoma matrix [8, 9].
Growing slowly, this pars flaccida cholesteatoma will start eroding surrounding structures. It will start to erode the scutum—the bony spur of Prussak’s space—and the ossicular chain at the level of the malleus head and the body and short process of the incus. The lateral epitympanic wall can also be eroded. By further expanding, the ossicular chain is displaced medially in case of a pars flaccida cholesteatoma [8, 9].
When the cholesteatoma expands further, it gradually fills up the attic towards the mastoid and it can erode several structures with possible complications. Extension towards the roof of the middle ear cavity can erode the tegmen with possible invasion into the middle cranial fossa and subsequent complications of meningitis or a temporal lobe abscess [8].
Extension towards the medial wall of the middle ear will erode most frequently the bony layer over the lateral semicircular canal with—in case of invasion—a possible secondary labyrinthitis in a limited number of cases (Fig. 4). The tympanic segment of the facial nerve canal can also be eroded with subsequent facial nerve palsy.
A less frequent type of cholesteatoma that can be found is the pars tensa cholesteatoma. These cholesteatomas originate from the mesotympanic part of the tympanic membrane and grow medially and upwards medial to the ossicular chain, displacing the ossicular chain laterally [8, 9].
Surgery of a pars flaccida cholesteatoma is most frequently performed using a canal wall up tympanoplasty (CWU), in which a mastoidectomy is performed, the wall of the external auditory canal is preserved, and disease is eradicated from the middle ear and mastoid. This technique, however, carries the risk of leaving cholesteatomatous tissue behind and makes clinical follow-up difficult as the intact external auditory canal does not allow inspection of the antrum and mastoid [10].
For this reason, about 1 year after first-stage surgery, a second look surgery is performed to evaluate the presence of a “residual” cholesteatoma. This second look surgery is performed in about 60% to 65% of patients with a higher percentage of second look surgery in children, reaching 80% [11]. The percentage of residual cholesteatoma found at second look surgery in adults is about 10% to 15%. In children, the percentage of residual cholesteatoma is higher varying between 23% and 44%, with recurrent cholesteatoma around 20% [11].
In order to lower the residual and recurrent rates of cholesteatoma in CWU techniques, primary bony obliteration techniques (PBOT) have been developed. This technique can be used during primary cholesteatoma surgery in order to treat middle ear and mastoid cholesteatoma but can also be assessed in revision cases of recurrent cholesteatoma. When using this technique, the canal wall up tympanoplasty cavity is subsequently filled up with a mixture of bone and bone pâté in order to diminish the number of recurrence through new retractions of the tympanic membrane [12–14]. A functional ossicular chain reconstruction is performed either in the same stage or in a second stage.
The number of residual cholesteatomas (15%) is lower than in the CWU technique with a recurrence rate of about 2% [14, 15].
Cholesteatoma: diffusion-weighted magnetic resonance imaging
Introduction
Evaluation of an acquired middle ear cholesteatoma was mainly done using CT scan. CT nicely demonstrates the erosion of the lateral epitympanic wall and ossicular chain. Erosion of the lateral semicircular canal can also be evaluated using CT scan [8, 9].
To see the direct effect of an associated invasion in the membranous labyrinth, MRI including T2-weighted images and post-gadolinium (Gd) T1-weighted sequences are required. On 3D heavily T2-weighted sequences, the fluid content and signal intensity of the membranous labyrinth can be evaluated. On post-Gd T1-weighted sequences, the enhancement of the membranous labyrinth should be looked after. Most frequently, it is the associated inflammation invading the membranous labyrinth and not the cholesteatoma itself that is causing enhancement (Fig. 4).
It was already noted in early reports that differentiation between inflammation and cholesteatoma is possible using Gd-enhanced T1-weighted images as cholesteatoma is by definition “avascular tissue” and does not enhance in contrast to inflammation which clearly enhances [16].
Fitzek was one of the first to demonstrate the hyperintense aspect of cholesteatoma on b1000 EP DW images. This hyperintense aspect was completely different from the complete lack of hyperintensity of middle ear inflammatory changes on b1000 EP DW images [17]. Major drawbacks of the EP DW sequences are, however, the low resolution, the rather thick slices, and the susceptibility artefacts at the interface temporal lobe and temporal bone limiting clearly the capability of this sequence to detect smaller cholesteatomas [18].
Recent papers have highlighted the advantages of non-echoplanar-based diffusion-weighted sequences. These sequences are most frequently single shot or multishot-based turbo spin echo diffusion-weighted sequences. They have a thinner slice thickness, a slightly higher resolution, and a complete lack of artefacts compared to echo planar diffusion-weighted sequences [3].
Congenital cholesteatoma
The congenital cholesteatoma situated in the middle ear is very difficult to evaluate clinically.
The presence of a congenital middle ear cholesteatoma is often suspected in the presence of a whitish lesion behind an intact tympanic membrane on otoscopic evaluation. Most oftenly, a conductive hearing loss is discovered. Congenital cholesteatomas are most frequently detected during childhood.
CT, however, is able to nicely demonstrate the nodular soft tissue lesion located very often in the middle ear around the ossicular chain. The intact tympanic membrane and the nodular soft tissue mass lesion in a young patient on an atypical location must attract the attention of the radiologist towards the diagnosis of a congenital cholesteatoma.
Due to the fact that congenital middle ear cholesteatomas are very often small lesions, EP DW MRI is unable to detect and characterize congenital middle ear cholesteatoma [18]. Non-EP DW MRI, however, is able to detect and characterize the congenital middle ear cholesteatoma due to its thinner slice thickness, higher resolution, and lack of artefacts. It nicely shows the very small nodular hyperintensity on b1000 images.
Even a congenital cholesteatoma as small as 2 mm can be detected and characterized on non-EP DW MRI [19] (Fig. 1). Moreover, correlation with apparent diffusion coefficient (ADC) maps nicely shows the hypointensity of the congenital cholesteatoma caused by the diffusion restriction in the cholesteatoma.
In case of congenital cholesteatoma situated in the petrous bone near or in the membranous labyrinth, the use of CT and MRI is mandatory. CT is needed to demonstrate the sharply delineated punched-out soft tissue lesion and to show the relation to the geniculate ganglion and the structures of the membranous labyrinth. On DW MRI, the diagnosis is straightforward showing the clear hyperintensity of the lesion on b1000 images. On EP DWI, the lesions look rather distorted, while congenital cholesteatoma shows a sharply delineated, nodular hyperintense aspect on non-EP DWI. It is even helpful in the description of the extension by clearly showing the hyperintensity of the congenital cholesteatoma (Fig. 2).
Even though the diagnosis can be made on DW MRI images, correlation with standard TSE T2-weighted images, 3D thin slice heavily T2-weighted images, and thin slice post-Gd T1-weighted is complementary in the description of the exact extension and invasion in the different structures of the membranous labyrinth.
Differential diagnosis with other lytic lesions in or around the membranous labyrinth can be made based upon the location and aspect of the lesion, its delineation, and the signal intensity on b1000 diffusion-weighted images. Congenital cholesteatoma is the only lesion showing a frank hyperintensity on b1000 diffusion-weighted images. Endolymphatic sac tumor and glomus jugulo-tympanicum have a rather specific predilection site. Endolymphatic sac tumor displays mixed signal intensities on standard MRI sequences with enhancement. Glomus jugulo-tympanicum has, apart from its specific predilection site and strong enhancement with salt and pepper appearance on standard sequences, a characteristic MR angiographic appearance. On unenhanced 3D TOF MR angio, the high signal intensity of the small intralesional vessels can be seen. On CT, it shows a very irregular permeative lytic bone pattern. Metastatic lesions to the temporal bone also have a very irregular lytic aspect compared to the sharply delineated and regular lytic aspect of congenital cholesteatoma. Metastatic lesions also enhance.
The facial nerve schwannoma also has a predilection site for the geniculate ganglion, but it follows the course of the facial nerve whereas the congenital schwannoma extends beyond the limits of the facial nerve course. Moreover, the facial nerve schwannoma always enhances. Hemangiomas located in the geniculate ganglion region limit themselves to the region of the geniculate ganglion but also display characteristic calcifications on CT. Moreover, both facial nerve schwannomas and hemangiomas do not display any hyperintense signal on b1000 DW MRI.
A non-EP DW MRI sequence should definitely be included in the MR evaluation of a middle ear soft tissue lesion and in the evaluation of a lytic lesion in the temporal bone pyramid.
Acquired middle ear cholesteatoma
Introduction
In the past few years, the role of diffusion-weighted MRI in the evaluation of patients with acquired middle ear cholesteatoma has increasingly grown.
In the evaluation of middle ear cholesteatoma, non-EP DW imaging has the advantage over EP DW imaging regarding thinner slices, a higher resolution, and a complete lack of artefacts [3]. Non-EP DW imaging is able to detect cholesteatoma as small as 2 mm in size whereas this is limited to 5 mm for EP DW imaging [18, 19]. In the evaluation of acquired middle ear cholesteatoma, preference should definitely be given to non-EP DW sequences over EP DW sequences.
Acquired middle ear cholesteatoma
In case of a clinical and otoscopical straightforward cholesteatoma, imaging evaluation prior to first stage surgery is done by CT, showing the erosive changes to the lateral epitympanic wall and the ossicular chain [8, 20, 21]. In case of a clinical and/or otoscopical unequivocal diagnosis of a cholesteatoma, detection and diagnosis of a middle ear cholesteatoma can be performed using non-EP DW sequences as a screening tool. In order to localize the cholesteatoma, axial and coronal TSE T2-weighted sequences are added. Gadolinium administration is—in those cases—not required [22].
However, in case of clinical suspicion of associated infection and/or inflammation, the combination of non-EP DW sequences and delayed post-Gd T1-weighted and TSE T2-weighted sequences will be able to differentiate the cholesteatoma from the surrounding infection or inflammation. In case of an associated complication, the combination of standard sequences after gadolinium administration and non-EP DW sequences is recommended (Fig. 4).
There are two reasons for false-negative examinations on DW sequences. Evacuation of the keratin from the cholesteatoma sac—by auto-evacuation or suction cleaning—will cause a false-negative examination as DW sequences detect the keratin inside the cholesteatoma (Fig. 5) [17–19]. Those cholesteatomas are also called “mural” cholesteatomas [9]. They form the major subgroup of false-negative DW examinations. Delineation of such an evacuated cholesteatoma is difficult even on standard MR sequences after gadolinium. Theoretically, one should be able to delineate the cholesteatoma matrix or epithelium on post-Gd T1-weighted sequences, but this remains very difficult due to the often surrounding inflammation (Fig. 5).
Motion artefacts degraded examinations will also result in a false-negative examination as in these cases the signal intensities of small cholesteatomas on b1000 non-EP DW sequences get smeared out over multiple pixels [19]. This should be taken into consideration especially in the pediatric population.
False-positive results on DW sequences are rare. There have been reports of false-positive results in cases with acute middle ear infection [17], bone powder [4], silastic sheets [23], granulation tissue [23], and scar tissue [24].
From our own experience, we know that accumulation of cerum in the external auditory canal can give a hyperintense signal on DW sequences. Correlation with standard TSE T2-weighted sequences nicely depicts the cerum in the external auditory canal (Fig. 6). Accumulated sebum in a sebum cyst also gives rise to a hyperintense signal. Careful interpretation of the images will show that the lesion is situated outside the temporal bone (Fig. 7).
Pre-second look evaluation
One of the greatest challenges in the past decade has been the question whether MRI could replace second look surgery and if patient selection for second look surgery is possible using MRI.
First and foremost, differentiation should be made between residual and recurrent cholesteatoma [25, 26]. Residual cholesteatomas are cholesteatomatous pearls left behind at first-stage surgery. It is known that the use of CWU tympanoplasty implies a high number of residual cholesteatomas. They can be found anywhere in the middle ear or mastoid cavity and are usually rather small at the time of detection. Due to this small size and the fact that they can be found anywhere in the middle ear and mastoid cavity, they are difficult to detect (Fig. 8).
Recurrent cholesteatoma is cholesteatoma originating again in a retraction pocket at the tympanic membrane or tympanic membrane graft (Fig. 9). Recurrent cholesteatoma takes more time to develop and is usually larger at the time of detection (Fig. 10).
Early reports on the value of standard MRI sequences replacing second look surgery were very disappointing [27]. However, in the last decade, two major types of MR protocols have emerged in the evaluation of cholesteatoma patients prior to second look.
Delayed post-Gd T1-weighted sequences have been reported to be able to detect residual cholesteatoma prior to second look as small as 3 mm. The rationale is based upon the fact that postoperative changes take time to enhance. Therefore, immediate post-Gd T1-weighted imaging can result in a false-positive finding of cholesteatoma. In literature, using this protocol, patients usually get preselected based upon CT findings [28, 29].
Echoplanar diffusion-weighted imaging has various reported sensitivities ranging from 12.5% to 86% [18, 23, 24, 30, 31]. Specificity varies from 73% to 100% [18, 23, 24, 30, 31].
This is explained by the fact that in some studies residual cholesteatomas [18, 23] were evaluated, and in other studies, a mixture of residual and recurrent cholesteatomas was used [24, 30, 31]. EP DW studies have a reported size limit of 5 mm. This is the main reason why EP DW sequences are unable to detect residual cholesteatomas and cannot replace second look surgery [18].
Using EP DW sequences, very high positive predictive values (PPV) were reported (between 80% and 100%), meaning that a positive diagnosis of a cholesteatoma can be made if a hyperintense lesion is detected, provided that artefacts are not misinterpreted as a cholesteatoma [18, 23, 24, 30, 31]. Negative predictive values (NPV) are reported to be variable (between 40% and 75%). This is explained by the fact that—due to size limits—a lot of false-negative cases are reported.
However, non-EP DW sequences have been reported to be able to replace second look surgery. Using the combination of non-EP DW sequences and delayed post-Gd sequences, a sensitivity of 90% and a specificity of 100% can be achieved [32]. One Australian group even screens patients prior to second look using non-EP DW sequences alone [33, 34].
Recent literature has indeed demonstrated that non-EP DW sequences alone have the same sensitivity, specificity, PPV, and NPV than the combination of non-EP DW sequences and delayed post-Gd T1-weighted sequences together. Non-EP DW sequences have also a significantly higher sensitivity, specificity, PPV, and NPV than delayed post-Gd T1-weighted sequences [22].
Therefore, evaluation of cholesteatoma patients can be performed prior to second look surgery using non-EP DW sequences alone [22, 33, 34]. In order to be able to localize suspected lesions on the non-EP DW sequences, an axial and coronal TSE T2-weighted sequence is added [22].
Cholesteatoma patients should definitely be selected for second look surgery using non-EP DW sequences (Fig. 8) [22, 33, 34]. CT scanning should no longer be used as the first imaging tool prior to second look surgery. CT scanning prior to second look should only be used in case of a positive MR examination using non-EP DW sequences in an immediate presurgical setting. This will inevitably reduce the number of useless irradiated patients prior to second look surgery.
MRI, including DW sequences, is extremely useful in the evaluation of possible postoperative complications. Particularly in case of an associated tegmen defect, one cannot differentiate soft tissues in the middle ear. MRI and DW sequences in particular have the possibility to differentiate associated soft tissue lesions. Meningoceles and meningoencefaloceles can easily be detected on MRI (Fig. 11).
Following a PBOT, imaging becomes even more important as there is still the risk of burying a residual cholesteatoma in the bony obliterated cavity. The rate of residual and recurrent cholesteatoma has been reported to be much lower than in the CWU technique.
EP DWI seems to be useless in the evaluation of these patients as these lesions are very small [35]. However, again, non-EP DW sequences are reported to have the highest sensitivity and specificity [36]. Recurrences are situated in most cases superficial to the obliterated mastoid or at the interface between the obliterated mastoid and the middle ear cavity (Fig. 12). In these cases, CT is complementary to MRI—using non-EP DW sequences—in showing the corresponding CT soft tissues densities to the MR reported hyperintensities (Fig. 12).
Petrous apex lesions
DW MRI is also very helpful in the differentiation of various lesions in the petrous bone apex [37, 38] (Table 1). Asymmetry in aeration of the petrous bone apex is one of the most common pseudo-tumorous lesions of the petrous bone apex (Fig. 13). If one apex is aerated, it displays a signal void on standard MR sequences as well as on diffusion-weighted sequences. The contralateral non-aerated apex contains fat with a high intensity on T1-weighted sequences and a moderate intensity on T2-weighted sequences. It lacks, however, a clear high intensity on b1000 DW MRI. Some hyperintensity can, however, be noted due to T2 shine through. The intensity on T2-weighted sequences is much lower that the intensity of cholesterol granuloma (see below) [37, 38].
An aerated petrous bone apex can get fluid filled and infected, displaying a low intensity on T1-weighted sequences with possible peripheral enhancement. On T2-weighted sequences, signal intensity is high. On b1000 DW MRI, there is no high signal intensity making the differential diagnosis with a petrous bone apex cholesteatoma possible (Fig. 14).
In the chronic phase, a mucocele of the petrous apex can develop areas of higher signal intensity on T1-weighted images and low signal intensity on T2-weighed images with an expansile appearance of the lesion on MRI and CT. Again, this lesion displays no hyperintensity on the b1000 diffusion-weighted sequence [37, 38].
A petrous bone apex cholesterol granuloma is the most common primary lesion of the petrous apex accounting for about 60% of all lesions in that region. They can occur bilateral. A history of serous or chronic otitis media has been identified as a major risk factor for the development of a cholesterol granuloma. They contain a brownish liquid with cholesterol crystals. Repetitive cycles of hemorrhage and granulomatous reaction initiated by an obstruction of the ventilation outlet of the apex are believed to be the cause [37, 38].
The entity of a petrous bone apex cholesterol granuloma presents as an expansile lesion with a clear high signal intensity on T1-weighted as well as on T2-weighted images (Fig. 15). This lesion does not display a high signal intensity on diffusion-weighted sequences. A slight hyperintensity can, however, be seen due to a T2 shine through effect. However, signal intensity on an ADC map remains high in contrast to the drop of signal intensity in case of a cholesteatoma. The contralateral petrous apex is usually pneumatized [37, 38].
Petrous bone apex cholesterol granuloma should be differentiated from the petrous bone apex cholesteatoma. This is again an expansile lesion with signal intensities comparable to the signal intensities of middle ear cholesteatoma: low signal intensity on T1-weighted sequences and a high signal intensity on T2-weighted sequences. To differentiate such a lesion from an opacified petrous bone apex or a petrous bone apex mucocele, diffusion-weighted MRI is very helpful and will display a clear hyperintensity on b1000 images with a drop in signal intensity on ADC map (Fig. 3).
Discussion still exists if a petrous bone apex cholesteatoma is a congenital cholesteatoma or an acquired cholesteatoma extending from the middle ear to the petrous bone apex [39]. There is a tendency towards the hypothesis that a petrous bone apex cholesteatoma is always an extension of a middle ear cholesteatoma into the petrous bone apex.
Petrous apex cephaloceles may be confused with “cystic”-like appearing lesions of the petrous apex such as the opacified petrous bone apex or the cholesteatoma. They result from herniation of the posterolateral dural wall of Meckel’s cave into the anterolateral aspect of the petrous bone. It is frequently associated to a hydrops of Meckel’s cave and to an empty sella. Signal intensities on standard sequences are those of cerebrospinal fluid (CSF) and the lesion shows a sharply delineated scalloping of the anterior and lateral aspect of the temporal bone [37, 38, 40]. Apart from the fact that the lesion is eccentric to the petrous bone apex, signal intensities on diffusion-weighted imaging are also low [40]. An overview of intensities on various sequences of the most frequent lesions and pseudolesions of the petrous apex is given in Table 1.
CPA lesions
The scala of CPA lesions is large and diverse and is beyond the scope of this paper. The top three of space occupying lesions in the cerebello-pontine angle are vestibule-cochlear schwannoma, meningeoma, and epidermoid cyst [41, 42].
The biggest group consists of vestibulo-cochlear schwannomas and meningeomas, accounting for about 90% of all tumorous lesions of the CPA. Both lesions display an isointensity to brain on unenhanced T1-weighted images, have a slight hyperintensity on T2-weighted images, and demonstrate a clear enhancement on post-gadolinium T1-weighted images. Vestibulo-cochlear schwannomas are always situated in the course of the vestibulo-cochlear nerve with often a component inside the internal auditory canal (IAC) and a component protruding in the CPA. The CPA component has sharp angles with the posteromedial surface of the temporal bone and can rarely display a dural tail [41, 42].
The meningeoma is usually eccentric to the IAC, but it may cover the IAC and even extend into the IAC. It has obtuse angles with the posteromedial surface of the temporal bone and displays a dural tail caused by either tumorous infiltration either caused by inflammatory changes [41, 42].
The epidermoid inclusion cyst is the third most common lesion of the CPA, accounting for about 5% of all mass lesions in the CPA. It can have a supratentorial extension with components into the middle cranial fossa, Meckel’s cave, the suprasellar region, and the quadrigeminal plate cistern. It is considered a congenital lesion rather than a tumoral lesion originating at the time of closure of the neural tube. It is a slowly growing mass due to continuous desquamation of the lining epithelium. Symptoms are usually caused due to compression of cranial nerves, the brain stem, and/or the cerebellum [42].
The epidermoid inclusion cyst has as a CSF-like appearance on standard MRI sequences, looking isointense on T1-weighted images and hyperintense on T2-weighted images. On these sequences, differentiation with an arachnoid cyst is very difficult. Arachnoid cysts are frequent in the CPA and can have a compressive effect on the 7th and 8th cranial nerve or on the 9th, 10th, and 11th cranial nerve depending on its location. Arachnoid cysts also have a CSF-like appearance.
On heavily T2-weighted sequences and fluid-attenuated inversion recovery (FLAIR) sequences, arachnoid cysts still have CSF appearance whereas epidermoid inclusion cysts have a mixed hyper–hypointense appearance on these heavily T2-weighted sequences and an inhomogeneous hyperintense appearance on FLAIR sequences. The appearance on heavily T2-weighted sequences and FLAIR sequences is caused by the fact that an epidermoid cyst is constituted of lamellated desquamated keratin. Diffusion-weighted imaging clearly shows a hyperintense signal intensity in case of an epidermoid inclusion cyst as epidermoid inclusion cysts are histologically exactly the same as acquired or congenital cholesteatoma (Fig. 16). Arachnoid cysts do not display hyperintensity on diffusion-weighted images [42, 43] (Fig. 17).
Non-EP diffusion-weighted sequences are to be preferred over EP diffusion-weighted sequences because of their lack of susceptibility artefacts.
Diffusion-weighted sequences are also crucial in the postoperative follow-up as it allows confirmation of the presence of a possible residual tumor [42] (Fig. 18).
In the evaluation of a CPA lesion, a diffusion-weighted sequence should always be included. It narrows differential diagnosis in lesions in which standard MR sequences are not equivocal (Fig. 19).
References
Thoeny HC, Keyzer D (2007) Extracranial applications of diffusion-weighted magnetic resonance imaging. Eur Radiol 17:1385–1393
Bammer R, Holdsworth SJ, Veldhuis WB, Skare ST (2009) New methods in diffusion-weighted and diffusion tensor imaging. Magn Reson Imaging Clin N Am 17:175–204
De Foer B, Vercruysse JP, Pilet B et al (2006) Single-shot, turbo spin-echo, diffusion-weighted imaging versus spin-echo planar, diffusion-weighted imaging in the detection of acquired middle ear cholesteatoma. ANJR Am J Neuroradiol 27:1480–1482
Dubrulle F, Souillard R, Chechin D et al (2006) Diffusion-weighted MR imaging sequence in the detection of postoperative recurrent cholesteatoma. Radiology 238:604–610
Lehmann P, Saliou G, Brochart C et al (2009) 3 T MR imaging of postoperative recurrent middle ear cholesteatomas: value of periodically rotated overlapping parallel lines with enhanced reconstruction diffusion-weighted MR imaging. AJNR Am J Neuroradiol 30:423–427
Nelson M, Roger G, Koltai PJ et al (2002) Congenital cholesteatoma: classification, management and outcome. Arch Otolaryngol Head Neck Surg 128:810–814
Kutz JW Jr, Friedman RA (2007) Congenital middle ear cholesteatoma. Ear Nose Throat J 86:654
Lemmerling M, De Foer B (2004) Imaging of cholesteatomatous and non-cholesteatomatouvarga, cholesteatomas middle ear disease. In: Lemmerling M, Kollias SS (eds) Radiology of the petrous bone. Springer, New York, pp 31–47
Barath K, Huber AM, Stämpfli P, Varga P, Kollias S (2010) Neuroradiology of cholesteatomas. AJNR Am J Neuroradiol Apr 1 (Epub ahead of print).
Brown JS (1982) A ten year statistical follow-up of 1142 consecutive cases of cholesteatoma: the closed versus the open technique. Laryngoscope 92:390–396
Schilder AG, Govaerts PJ, Somers T, Offeciers FE (1997) Tympano-ossicular allografts for cholesteatoma in children. Int J Pediatr Otorhinolaryngol 42:31–40
Mercke U (1987) The cholesteatomatous ear one year after surgery with obliteration technique. Am J Otol 8:534–536
Gantz BJ, Wilkinson EP, Hansen MR (2005) Canal wall reconstruction tympanomastoidectomy with mastoid obliteration. Laryngoscope 115:1734–1740
Offeciers E, Vercruysse JP, De Foer B, Casselman J, Somers T (2008) Mastoid and epitympanic obliteration. The obliteration technique. In: Ars B (ed) Chronic otitis media. Pathogenesis oriented therapeutic treatment. Kugler, Amsterdam, pp 299–327
Vercruysse JP, De Foer B, Somers T, Casselman JW, Offeciers E (2008) Mastoid and epitympanic bony obliteration in pediatric cholesteatoma. Otol Neurotol 29:953–960
Martin N, Sterkers O, Nahum (1990) Chronic inflammatory disease of the middle ear cavities: Gd-DTPA-enhanced MR imaging. Radiology 176:39–405
Fitzek C, Mewes T, Fitzek S et al (2002) Diffusion-weighted MRI of cholesteatomas of the petrous bone. J Magn Reson Imaging 15:636–641
Vercruysse JP, De Foer B, Pouillon M et al (2006) The value of diffusion-weighted MR imaging in the diagnosis of primary acquired and residual cholesteatoma: a surgical verified study of 100 patients. Eur Radiol 16:1461–1467
De Foer B, Vercruysse JP, Bernaerts A et al (2007) The value of single-shot turbo spin-echo diffusion-weighted MR imaging in the detection of middle ear cholesteatoma. Neuroradiology 49:841–848
7Lemmerling MM, De Foer B, VandeVyver V, Vercruysse JP, Kl V (2008) Imaging of the opacified middle ear. Eur J Radiol 66:363–371
Lemmerling MM, De Foer B, Verbist BM, VandeVyver V (2009) Imaging of inflammatory and infectious diseases in the temporal bone. Neuroimaging Clin N Am 19:321–337
De Foer B, Vercruysse JP, Bernaerts A et al (2010) Middle ear cholesteatoma: non-echo-planar diffusion-weighted MR imaging versus delayed gadolinium-enhanced T1-weighted MR—value in detection. Radiology 255:866–872
Venail F, Bonafe A, Poirrier V, Mondain M, Uziel A (2008) Comparison of echo-planar diffusion-weighted imaging and delayed postcontrast T1-weighted MR imaging for the detection of residual cholesteatoma. AJNR Am J Neuroradiol 29:1363–1368
Jeunen G, Desloovere C, Hermans R, Vandecavye V (2008) The value of magnetic resonance imaging in the diagnosis of residual or recurrent acquired cholesteatoma after canal wall-up tympanoplasty. Otol Neurotol 29:16–18
Sheehy JL, Brackman DE, Graham MD (1977) Cholesteatoma surgery: residual and recurrent disease. A review of 1024 cases. Ann Otol Rhinol Laryngol 86:451–462
Brackman DE (1993) Tympanoplasty with mastoidectomy: canal wall up procedures. Am J Otol 14:380–382
Vanden Abeele D, Coen E, Parizel PM, Van de Heyning P (1999) Can MRI replace a second look operation in cholesteatoma surgery? Acta Otolaryngol 119:555–561
Williams MT, Ayache D, Albert C et al (2003) Detection of postoperative residual cholesteatoma with delayed contrast-enhanced MR imaging: initial findings. Eur Radiol 13:169–174
Ayache D, Williams MT, Lejeune D, Corré A (2005) Usefulness of delayed postcontrast magnetic resonance imaging in the detection of residual cholesteatoma after canal wall-up tympanoplasty. Laryngoscope 115:607–610
Aikele P, Kittner T, Offergeld C et al (2003) Diffusion-weighted MR imaging of cholesteatoma tin pediatric and adult patients who have undergone middle ear surgery. AJR Am J Roentgenol 181:261–265
Stasolla A, Magliulo G, Parrotto D, Luppi G, Marini M (2004) Detection of postoperative relapsing/residual cholesteatoma with diffusion-weighted echo-planar magnetic resonance imaging. Otol Neurotol 25:879–884
De Foer B, Vercruysse JP, Bernaert A et al (2008) Detection of postoperative residual cholesteatoma with non-echo-planar diffusion-weighted magnetic resonance imaging. Otol Neurotol 29:513–517
Dhenpnorrarat RC, Wood B, Rajan GP (2009) Postoperative non-echo-planar diffusion-weighted magnetic resonance imaging changes after cholesteatoma surgery: implications for cholesteatoma screening. Otol Neurotol 30:54–58
Rajan GP, Ambett R, Wun L et al (2010) Preliminary outcomes of cholesteatoma screening in children using non-echo-planar diffusion-weighted magnetic resonance imaging. Int J Pediatr Otorhinolaryngol 74:297–301
De Foer B, Vercruysse JP, Pouillon M et al (2007) Value of high-resolution computed tomography and magnetic resonance imaging in the detection of residual cholesteatoma in primary bony obliterated mastoids. Am J Otolaryngol 28:230–234
Vercruysse JP, De Foer B, Somers T, Casselman J, Offeciers E (2010) Lont-term follow up after bony mastoid and epitympanic obliteration: radiological findings. J Laryngol Otol 124:37–43
Lemmerling M (2004) Petrous apex lesions. In: Lemmerling M, Kollias SS (eds) Radiology of the petrous bone. Springer, New York, pp 171–179
Schmalfuss IM (2009) Petrous apex. Neuroimaging Clin N Am 19:367–391
Silveira Filho LG, Ayache D, Sterkers O, Williams MT (2010) Middle ear cholesteatoma extending into the petrous apex. Otol Neurotol 31:544–545
Alkilic-Genauzeau I, Boukobza M, Lot G, George B, Merland JJ (2007) CT and MRI features of arachnoid cyst of the petrous apex: report of 3 cases. J Radiol 88:1179–1183
Bonneville SJ, Chiras J (2007) Imaging of cerebellopontine angle lesions: an update. Part 1: enhancing extra-axial lesions. Eur Radiol 17:2472–2482
Bonneville F, Savatovsky J, Chiras J (2007) Imaging of cerebellopontine angle lesions: an update. Part 2: intra-axial lesions, skull base lesions that may invade the CPA region, and non-enhancing extra-axial lesions. Eur Radiol 17:2908–2920
Liu P, Saida Y, Yoshioka H, Itai Y (2003) MR imaging of epidermoids at the cerebellopontine angle. Magn Reson Med Sci 2:109–115
Conflict of interest
We declare that we have no conflict of interest.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
De Foer, B., Vercruysse, JP., Spaepen, M. et al. Diffusion-weighted magnetic resonance imaging of the temporal bone. Neuroradiology 52, 785–807 (2010). https://doi.org/10.1007/s00234-010-0742-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00234-010-0742-1