|
|
||||||||
Neuroradiology |
1 From the Department of Radiology, Hopital Huriez-Sud, Centre Hospitalier Universitaire de Lille, 1 Place de Verdun, 59037 Lille, France (F.D., O.E., L.L.); the Departments of Otoneurology (C.V., F.M.V.) and Neurosurgery (J.P.L.), Hopital Roger Salengro, Centre Hospitalier Universitaire de Lille, France. From the 1998 RSNA scientific assembly. Received November 2, 1998; revision requested December 21; final revision received September 17, 1999; accepted September 29. Address correspondence to F.D. (e-mail: fdubrulle@caramail.com).
| Abstract |
|---|
|
|
|---|
MATERIALS AND METHODS: Thirty-one patients with serviceable hearing underwent 1.5-T MR imaging before hearing-preservation surgery. The presence of cochlear fossa enhancement on T1-weighted spin-echo images obtained after the administration of contrast material was compared with the results of hearing-preservation surgery.
RESULTS: Cochlear fossa enhancement was present in 13 patients, and all of them had total hearing loss after surgery. There was no cochlear fossa enhancement in 18 patients; 15 maintained serviceable hearing after surgery, and three had postoperative hearing loss with no serviceable hearing (sensitivity, 81%; specificity, 100%).
CONCLUSION: Cochlear fossa enhancement on T1-weighted spin-echo images seems to be a reliable sign for analyzing the state of the cochlear nerve. The absence of cochlear fossa enhancement could become an additional criterion for selecting the surgical approach in vestibular schwannomas.
Index terms: Ear, MR, 213.121411, 213.12143 Ear, neoplasms, 213.364 Hearing loss, 213.1267 Magnetic resonance (MR), contrast enhancement, 213.12143 Magnetic resonance (MR), high-resolution, 213.121411, 213.12143 Nerves, cranial, 213.364 Nerves, MR, 213.121411, 213.12143 Schwannoma, 213.364
| Introduction |
|---|
|
|
|---|
This approach is attempted nowadays in patients who have small schwannomas and serviceably sufficient hearing. However, these criteria are insufficient to predict the state of the cochlear nerve, and there are failures in cochlear nerve preservation during surgery. The purpose of our study was to describe a sign in magnetic resonance (MR) imaging that could reflect the state of the cochlear nerve before hearing-preservation surgery in small vestibular schwannomas.
| MATERIALS AND METHODS |
|---|
|
|
|---|
The following three sequences were used with a head coil. A thin-section T1-weighted spin-echo sequence was used in the transverse plane, centered on the internal auditory canal, before the intravenous injection of 0.1 mmol of gadoterate meglumine (Dotarem; Guerbet, Roissy, France) per kilogram of body weight, and with the following imaging parameters: 550/20 (repetition time msec/echo time msec), 2.0-mm section thickness, 300 x 512 matrix, 280 x 280 field of view, three acquisitions, and an acquisition time of 5 minutes 10 seconds. A constructive interference in steady state (CISS) sequence centered on the internal auditory canal was then used with the following parameters: 12.25/5.90, 230 x 512 matrix, 165 x 220 field of view, and 0.7-mm section thickness. This study consisted of two additional sequences with a total acquisition time of 8 minutes 40 seconds (3). A T1-weighted spin-echo sequence was used in the transverse plane after the administration of contrast material. The parameters were the same as those for the T1-weighted spin-echo sequence before the administration of contrast material.
Sixteen grade I (intracanalicular) and fifteen grade II (extension in the cerebellopontine angle of less than 20 mm) schwannomas were analyzed (4). Tumor size, which was measured as the greater axis including the intracanalicular extension multiplied by the transverse axis, ranged from 3 x 2 mm to 22 x 10 mm, with a mean size of 12.2 x 7.3 mm. All schwannomas in this study had an extension to the fundus of the internal auditory canal.
All patients were selected for hearing- preservation surgery on the basis of the criteria developed by Wade and House (5). Since the publication of their article, other criteria have been suggested, but they were based on more or less the same proposals (1,2,6,7).
The selection criterion for attempting hearing preservation was strict. Our inclusion criterion was small schwannoma, defined as grade I or grade II and less than 10 mm in the cerebellopontine angle, with serviceable hearing, defined as pure tone average decrease lower than 50 dB, speech reception threshold lower than 50 dB, speech discrimination score greater than 70%, and good general condition. All patients had normal facial function before surgery.
Two independent observers (F.D., L.L.) separately analyzed the appearance of the schwannomas in the fundus of the internal auditory canal, and in particular they analyzed the appearance of the cochlear fossa. The cochlear fossa is the canal containing the cochlear nerve, which arises from the fundus and goes to the cochlea. The cochlear fossa must be well differentiated from the pregeniculate facial nerve, which is situated above the cochlear fossa. In continuity with the pregeniculate facial nerve is the geniculate ganglion; in continuity with the cochlear fossa is the cochlea. A thin-section transverse T1-weighted spin-echo sequence (2.0-mm section thickness) was useful for separating these two elements. The thin sections of the CISS sequence (0.7-mm section thickness) were also useful for differentiating the pregeniculate region from the cochlear fossa. The observers were blinded to patient name and clinical information. These observations were made before surgery in all patients.
The extension of the schwannomas in the cochlear fossa was considered positive when there was hook-shaped enhancement in the area of the cochlear fossa on transverse T1-weighted spin-echo images obtained after the administration of contrast material (Figs 1a, 2). This observation was compared with observations made on nonenhanced T1-weighted spin-echo images. The extension in the cochlear fossa was considered negative when there was no enhancement in the cochlear fossa on postcontrast transverse T1-weighted spin-echo images (Fig 3a).
|
|
|
|
|
Surgery was performed within 3 months after the MR imaging (mean delay, 10 days). At our institution (Centre Hospitalier Universitaire de Lille, France), hearing preservation consisted of a combined approach to the internal auditory canal (8): a middle fossa approach and a retrosigmoid approach. This combined method was used by a combined oto- and neurosurgical team in 29 patients in this study. In the smallest schwannomas in two patients in this study, an exclusively middle fossa approach was used.
The aim of these methods was to remove the vestibular schwannoma while preserving the other nerves in the internal auditory canal (ie, the facial nerve and the cochlear branch of the vestibulocochlear nerve). Preservation of the cochlear branch made hearing preservation possible. Surgery was performed with monitoring of the brainstem-evoked auditory responses.
Sensitivity, specificity, and predictive values of the positive or negative extension in the cochlear fossa visible on postcontrast T1-weighted spin-echo MR images were calculated in comparison with surgical results on the basis of the criteria of serviceable hearing after 3 months (9,10). Hearing preservation was considered a success when the pure tone average decrease was lower than 30 dB with a speech discrimination score greater than 50 dB (8,10). Hearing preservation was also compared with the size and grade of the schwannoma. A
2 test was used for statistical analysis.
| RESULTS |
|---|
|
|
|---|
Thirteen of 31 patients had cochlear fossa enhancement on postcontrast images. In five patients, surgery showed lesions of the cochlear nerve. These lesions were either an extension of the schwannoma in the cochlear fossa along the cochlear branch or a complete tumoral inclusion of the nerve at the level of the cochlear fossa. In these five patients, the cochlear nerve could not be preserved during surgery. In eight patients, dissection of the cochlear nerve was difficult because of the close proximity of the tumor to the cochlear nerve in the cochlear fossa. In these eight patients, the surgeon (F.M.V.) tried to preserve the nerve during dissection, but the brainstem-evoked auditory responses disappeared during the procedure, and total hearing loss was observed after surgery. Thus, hearing could not be preserved in the 13 patients with cochlear fossa enhancement at MR imaging.
Eighteen of 31 patients had no cochlear fossa enhancement on postcontrast images. In these 18 patients, the surgeon identified a normal aspect of the cochlear fossa and preserved the cochlear nerve during surgery. Hearing was evaluated 3 months later. Hearing preservation was considered a success in 15 patients, with the same hearing as before surgery in nine patients and with only a slight hearing loss in six patients in whom pure tone average decreased lower than 30 dB and the speech discrimination score was greater than 50 dB. Hearing preservation was considered a failure in three patients who had an important hearing decrease or complete hearing loss after surgery.
The facial nerve was uninvolved by tumor extension in all patients, and all patients had normal facial function after surgery.
The sensitivity of positive cochlear fossa enhancement compared with hearing preservation was 81%, with a specificity of 100%. The positive predictive value was 100%, with a negative predictive value of 83%. The results of surgery were significantly different between the patients with cochlear fossa enhancement and those without cochlear fossa enhancement (P < .001). Six patients with grade I schwannomas and seven patients with grade II schwannomas had cochlear fossa enhancement. There was no statistical correlation between grade I or grade II and hearing preservation (P > .05), nor was there any correlation between the size of the schwannoma and hearing preservation (P > .05).
In the 13 patients with cochlear fossa enhancement on postcontrast T1-weighted images, the CISS images were considered abnormal: The cochlear branch was not identified, and there was no high signal intensity at this level. In the 18 patients without cochlear fossa enhancement, the observers described a normal cochlear fossa on the CISS images in 14 patients; in the remaining four, the cochlear fossa could not be analyzed on the CISS images because of the absence of fluid in the internal auditory canal owing to its very small size. In these four patients, the cochlear fossa on the asymptomatic side was not well analyzed: The distinction among the four nerves in the fundus was difficult due to the small size of the internal auditory canal.
| DISCUSSION |
|---|
|
|
|---|
In this study, we tried to evaluate the state of the cochlear nerve before surgery and to suggest MR criteria of extension in the cochlear fossa. We also attempted to correlate these criteria with the probability of achieving hearing preservation. In our study, the analysis of the cochlear fossa on MR images enabled us to describe a sign concerning the cochlear fossa when schwannomas reached the fundus of the internal auditory canal: cochlear fossa enhancement on thin-section T1-weighted spin-echo images.
To allow detection of slight cochlear fossa enhancement, we used thin sections with T1-weighted spin-echo sequences (2.0-mm section thickness, 300 x 512 matrix). These technical parameters produced a relatively long acquisition time (5 minutes 10 seconds), but they were justified to obtain excellent spatial resolution and permit meticulous analysis of the fundus of the internal auditory canal, in particular the cochlear fossa.
The T1-weighted images obtained before the administration of contrast material were used only to depict cochlear fossa enhancement after the injection of contrast material. The cochlear fossa enhancement was well detected on transverse postcontrast T1-weighted spin-echo images but not in the coronal plane, in which this small sign is not identified. This could be explained by the transverse orientation of the cochlear fossa.
The CISS images showed the same findings as the postcontrast images didthat is, a normal cochlear fossa in the absence of cochlear fossa enhancement and an abnormal cochlear fossa in the presence of cochlear fossa enhancement. However, in four patients (13%), the analysis of the cochlear fossa was not possible because of the small size of the internal auditory canal. In these patients, the four nerves were difficult to identify in the internal auditory canal, and the observers could not analyze the cochlear fossa. In our study, vestibular schwannomas were always detected on the CISS images; the screening of vestibular schwannomas seems possible with the CISS sequence alone (16,17). However, the CISS sequence can be useful for differentiating the pregeniculate facial nerve from the cochlear fossa.
The enhanced transverse T1-weighted spin-echo and the CISS sequences were useful for analyzing the fundus of the internal auditory canal, but a meticulous analysis of the cochlear fossa in all patients required thin-section postcontrast T1-weighted spin-echo images that were more sensitive for detection of the state of the cochlear fossa. Postcontrast T1-weighted spin-echo images demonstrated tumor enhancement, whereas T2-weighted images only implied tumor presence when a normal structure (ie, the cochlear nerve) was not identified. This created a dilemma when the internal auditory canal was small and its normal structures were difficult to identify.
There was a substantial difference concerning hearing preservation between the groups with positive or negative cochlear fossa enhancement, with specificity and positive predictive values of 100%. None of the patients with cochlear fossa enhancement had cochlear nerve preservation after surgery, and all had subsequent total hearing loss (13 of 13 patients). Conversely, 83% (15 of 18) of the patients without cochlear fossa enhancement had sufficient hearing preservation, and only 17% (three of 18) had a mean decrease in their hearing.
This cochlear fossa enhancement on T1-weighted images seemed to be a reliable sign for analyzing the state of the cochlear nerve and foreseeing that the preservation of the cochlear nerve would probably be impossible during surgery. This sign never gave any false-positive results: When it was present, hearing was never preserved, hence the importance of its detection before adopting a surgical strategy and our assertion that it should become a new criterion for hearing-preservation surgery.
If MR imaging can be used to foresee that the cochlear nerve cannot be preserved when there is cochlear fossa enhancement, the surgeon may inform the patient and, perhaps, choose a different attitude: refraining from hearing-preservation surgery (translabyrinthine approach), radiosurgery, or observation. On the other hand, hearing preservation can be attempted with a good chance of success (15 [83%] of 18 patients in our study) when there is no cochlear fossa enhancement. The surgeon must advise the patient that there is a small failure rate (three [17%] of 18 patients in our study). The causes of these failures are numerous, such as vascular complications after surgery (interruption of the cochlear blood supply) (15,18) or minimal injury of the cochlear nerve during dissection (6).
In conclusion, the results of this study show that cochlear fossa enhancement on postcontrast T1-weighted spin-echo images is a specific sign for predicting the inability to preserve the cochlear nerve if surgery is to be performed for vestibular schwannoma. Hence, the new considerations for hearing-preservation surgery should be small vestibular schwannomas less than 10 mm in diameter in the cerebellopontine angle without cochlear fossa enhancement in patients with criteria of serviceable hearing.
| Acknowledgments |
|---|
| Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, L.L., F.D.; study concepts, L.L., F.D.; study design, F.D.; definition of intellectual content, L.L., F.D., O.E.; literature research, F.D., C.V.; clinical studies, C.V., F.M.V., J.P.L.; data acquisition, F.D.; data analysis, F.D., O.E.; statistical analysis, F.D., O.E., C.V.; manuscript preparation and editing, F.D.; manuscript review, L.L., O.E., C.V.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. Nakamura, S. Naganawa, T. Koshikawa, H. Fukatsu, Y. Sakurai, I. Aoki, A. Ninomiya, and T. Ishigaki High-Spatial-Resolution MR Cisternography of the Cerebellopontine Angle in 90 Seconds with a Zero-Fill Interpolated Fast Recovery 3D Fast Asymmetric Spin-Echo Sequence AJNR Am. J. Neuroradiol., September 1, 2002; 23(8): 1407 - 1412. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. Dubrulle, J. Delomez, A. Kiaei, P. Berger, C. Vincent, F.-M. M. Vaneecloo, and L. Lemaitre Mass Screening for Retrocochlear Disorders: Low-Field-Strength (0.2-T) versus High-Field-Strength (1.5-T) MR Imaging AJNR Am. J. Neuroradiol., June 1, 2002; 23(6): 918 - 923. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. L. Salzman, H. C. Davidson, H. R. Harnsberger, C. M. Glastonbury, R. H. Wiggins, S. Ellul, and C. Shelton Dumbbell Schwannomas of the Internal Auditory Canal AJNR Am. J. Neuroradiol., August 1, 2001; 22(7): 1368 - 1376. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |