Abstract
Objective
To determine predictive factors for cochlear obliteration after translabyrinthine surgery for vestibular schwannoma.
Study Design
Case series with chart review.
Setting
Neurotology referral center.
Subjects and Methods
In total, 345 charts were reviewed, resulting in 103 patients who underwent translabyrinthine surgery between January 2010 and July 2015 and had postoperative magnetic resonance imaging (MRI) for review. Surveillance MRI performed after translabyrinthine resection of vestibular schwannomas was reviewed. Heavily T2-weighted MRI obtained an average of 21.8 months after surgery was reviewed to assess cochlear patency. Tumor size, preoperative audiograms, and MRI cochlear fluid-attenuated inversion recovery (FLAIR) intensity were compared between patients with retained cochlear patency and those without.
Results
Fifty-four percent of patients retained cochlear patency after translabyrinthine surgery. Tumor size did not differ statistically between the 2 groups. There was no statistically significant difference in speech reception thresholds, pure-tone average, or word recognition scores between patent and nonpatent groups. Preoperative MRI FLAIR intensity did not differ between groups.
Conclusions
More than half of patients retain cochlear patency after translabyrinthine vestibular schwannoma surgery. Cochlear patency is required for cochlear implant in patients with unilateral deafness. Preoperative tumor size, hearing performance, and intensity on MRI FLAIR do not predict cochlear patency. To prevent loss of opportunity for cochlear implantation, simultaneous implantation and cochlear lumen keeper placement are options.
Keywords
Since their approval in profoundly hearing-impaired adults by the US Food and Drug Administration in 1984, cochlear implants have seen progressive expansion of their indications, allowing implantation in patients with significant residual hearing. While not approved for the treatment of unilateral deafness, there are now multicenter studies of cochlear implantation in patients with single-sided deafness due to a variety of conditions, including the removal of vestibular schwannomas and other pathologies causing sensorineural hearing loss in a single ear. 1
One group of patients that may benefit from implantation are those with single-sided deafness after removal of vestibular schwannoma by the translabyrinthine approach. The many advantages of this approach include wide exposure of the facial nerve, minimal brain retraction, application to tumors of all sizes, completely extradural drilling, and avoidance of trauma to the suboccipital musculature. A disadvantage of the approach is the sacrifice of residual hearing as a result of labyrinthectomy.
Cochlear implantation requires preservation of the cochlear nerve and preservation of the cochlear patency. The cochlea may become obliterated by fibrous tissue or ossified by new bone formation after surgical manipulation. In 1 study of patients undergoing transmastoid labyrinthectomy for control of vertigo attacks associated with endolymphatic hydrops, 100% (n = 18) retained cochlear patency. 2 Cochlear implantation after labyrinthectomy has been described in patients with Ménière’s disease.3-6 Patients undergoing labyrinthectomy as part of translabyrinthine surgery for removal of vestibular schwannoma retain cochlear patency in as many as 67% of cases. 7 Factors that predict the rate of patency of the cochlea after surgical manipulation remain unknown.
Miller et al, 8 studying hearing preservation in vestibular schwannoma surgery via retrosigmoid craniotomy for vestibular schwannoma removal, found that the elevation of intracochlear proteins probably correlates with a lower chance of hearing preservation. The level of intracochlear proteins can be predicted by magnetic resonance imaging (MRI) of the cochlea using the fluid-attenuated inversion recovery (FLAIR) sequence.
Before cochlear implantation can be a reliable option for patients undergoing translabyrinthine surgery for vestibular schwannoma removal, more must be understood about the preservation of the cochlear nerve and duct with surgical manipulation. The present study reviewed our experience with translabyrinthine surgery to determine if preservation of the cochlea could be predicted by factors such as tumor size, preoperative hearing, and MRI FLAIR sequence of preoperative cochleae.
Methods
This study was approved by the Providence Hospital and Medical Centers Institutional Review Board.
We reviewed charts of 345 patients undergoing translabyrinthine surgery for vestibular schwannoma removal between January 2010 and July 2015. Patients were gathered by diagnosis codes for vestibular schwannoma and surgical billing codes for transtemporal surgery. Inclusion criteria were pathologic confirmation of vestibular schwannoma, confirmation of labyrinthectomy in the operative report, and postoperative MRI for review. Exclusion criteria included neurofibromatosis, pathology other than vestibular schwannoma, and approaches other than the translabyrinthine approach.
Postoperative MRI was then reviewed. The latest available MRI was used. Particular attention was given to the fast imaging employing steady-state acquisition (FIESTA) sequence. Cochleae were graded as patent or obstructed based on the presence of loss of fluid signal in the residual cochlea (see Figure 1 ).
Fast imaging employing steady-state acquisition sequence magnetic resonance imaging showing postoperative preservation (green) and obliteration (red) of the cochlea of 2 different patients.
Preoperative audiograms were reviewed and pure-tone average (PTA), speech reception thresholds (SRTs), and word recognition scores (WRSs) were recorded. Patients were grouped according to the following: PTA <50 dB or ≥51 dB, SRTs <50 dB or ≥51 dB, and WRSs >50% or ≤49%. These groups were created to reflect patients with serviceable and unserviceable hearing.
Preoperative MRI was reviewed and tumor size in centimeters was recorded. Patients were grouped according to the following: 0 to 1 cm, 1.1 to 2 cm, 2.1 to 3 cm, and >3 cm.
Finally, FLAIR sequence of cochlea was examined and graded as done by previous examiners and designated as either limited hyperintensity (LH) or extensive hyperintensity (EH). 8
Statistical analysis was performed between the 2 groups (patent or obstructed) for each of the factors (hearing, tumor size, FLAIR intensity). Chi-square analysis was used for discreet variables and Student t test for continuous variables.
Results
Of 345 patients undergoing translabyrinthine surgery during the study period, 103 patients fitting the inclusion/exclusion criteria were included in this study. Nineteen patients were excluded due to pathology other than schwannoma. Seventeen patients were excluded due to retrosigmoid approach. Three patients were excluded due to middle fossa approach. Two patients were excluded due to gamma knife treatment. One patient was excluded due to diagnosis of neurofibromatosis type 2. Of the remaining patients, all had MRI studies with reports available, but only 103 had digital images available for review.
All patients included had at least 1 postoperative MRI. The average time between surgery and final MRI was 21.8 months. The graded MRI was the final MRI available. Fifty-seven patients (54%) retained patency of their cochlea as indicated by maintenance of fluid signal on FIESTA sequence. Forty-six patients were determined to have obliteration of the cochlea. There were 12 patients who had preoperative FLAIR images.
In the 46 patients with cochlear obliteration, all had this finding at the first postoperative MRI. The average time to this first MRI was 16.9 months. The shortest time to first MRI was 1 month. The longest time to first MRI was 39 months.
Ninety-seven patients had a preoperative MRI report available for review of maximum tumor dimension. In patients with tumors 0 to 1 cm, 1.1 to 2 cm, 2.1 to 3 cm, and >3 cm, patency rate was 53%, 51%, 68%, and 39% respectively (see Table 1 ). There was no significant difference between these groups as determined by χ2 analysis (P > .05 between all groups). Using >3 cm tumor size as a cutoff point, there was still no statistically significant difference in patency between tumors 0-3 cm and >3 cm (P = .17).
Relationship between Cochlear Patency and Preoperative Tumor Size.
P value for χ2 test comparing each group to patency in total patient population.
Preoperative hearing data were examined for a relationship with cochlear patency. PTA data were available for 95 patients. For patients with PTA <50 dB, patency rate was 57% compared with patients with PTA >50 dB and a patency rate of 48%. This was not a statistically significant difference (P = .50). SRT data were available for 97 patients. Patients with SRTs <50 dB had a patency rate of 56% compared with 48% in patients with SRTs >50 dB. This was not a statistically significant difference (P = .52). WRS data were available for 98 patients. Patients with WRSs >50% had a patency rate of 53% compared with 56% in patients with WRSs <50%. This was not significant (P = .83). For summary of these findings, see Table 2 .
Relationship between Cochlear Patency and Preoperative Hearing. a
Abbreviations: PTA, pure-tone average; SRT, speech reception threshold; WRS, word recognition score.
PTA was not a predictor of patency (P = .50). SRT was not a predictor of patency (P = .52). WRS was not a predictor of patency (P = .83).
Preoperative MRI FLAIR cochlear images were available for 12 patients. Review revealed patency rates of 67% and 83% for LH and EH, respectively. Comparing the rate of patency in LH cochleae to the overall population rate of 54%, there was no statistically significant difference (P = .69). For the EH rate of 83% patency compared with the overall rate of 54%, there was no statistically significant difference (P = .24). Comparing LH to EH patency rates, there was no statistically significant difference (P = 1). For cochleae deemed LH, PTA, SRTs, and WRSs were 29 dB, 24 dB, and 86%, respectively. For those deemed EH, PTA, SRTs, and WRSs were 49 dB, 51 dB, and 66%, respectively. See Table 3 for summary of these findings.
Relationship between FLAIR Sequence, Hearing, and Cochlear Patency. a
Abbreviations: EH, extensive hyperintensity; FLAIR, fluid-attenuated inversion recovery; LH, limited hyperintensity; MRI, magnetic resonance imaging; PTA, pure-tone average; SRT, speech reception threshold; WRS, word recognition score.
Student t test comparing means for LH vs EH groups.
Discussion
In considering cochlear implantation after translabyrinthine surgery, cochlear patency is a requirement. Previous investigations suggest that the rate of obstruction is as high as 67%. 7 However, here we demonstrate that over half of cochleae are preserved approximately 2 years after surgery.
Unfortunately, it is difficult to predict which cochleae will become obstructed. Preoperative tumor size cannot be relied upon to predict patency. Even in tumors >3 cm, the rate of cochlear patency is not statistically different from smaller tumors. Hearing performance lends little information about cochlear patency as well. Patients with lower pure-tone thresholds preoperatively did not have statistically significant higher preservation of cochlear patency. Average WRSs were actually better in patients who went on to develop obstruction postoperatively, although again without statistical significance. Characterizing preoperative hearing as either serviceable or unserviceable does not reliably predict postoperative cochlear patency. With regard to preoperative MRI FLAIR intensity of the cochlea, hyperintensity is associated with poorer hearing by other investigators. 8 In our data, FLAIR intensity does not predict cochlear patency or hearing performance with any statistical significance. This finding was significantly underpowered as there were only 12 sets of preoperative MRI images recovered for review.
We acknowledge a limitation in this study with regard to the availability of imaging for review. As cochlear fluid is not routinely discussed in the radiologist’s reports, we required digital images be available for inclusion. The availability of digital images was limited as many patients travel a great distance to our facility and undergo imaging at outside facilities. The availability of preoperative imaging is even further limited by this effect because it is typically obtained by outside physicians. This made preoperative MRI FLAIR data rare. In addition, tumor measurements can vary between various clinicians. Reported measurements are more variable than standardized measurement obtained by a single clinician. Therefore, we used the largest reported dimension in an effort to decrease variability. We believe that the patients included in the study are a random cross section of all patients undergoing translabyrinthine removal of vestibular schwannoma at our institution, and as such, their rate of cochlear patency is representative of the rate of cochlear patency in patients with vestibular schwannoma treated by this modality.
Given the substantial rate of cochlear obliteration, the timing of obliteration is also important. Prior investigators have shown that obliteration can occur as early as 3 months postoperatively. 7 In the current study, all patients who showed obliteration on examination of their final MRI had developed this change by the first MRI. In these patients, time to first MRI was 16.9 months, but the shortest MRI was done at 1 month postoperatively. This similar finding to prior studies suggests that if cochlear obliteration is to occur, it will occur quickly after surgery.
Although cochlear obliteration is difficult to predict, it need not prevent consideration of cochlear implantation as a possibility for these patients. Two strategies exist for the circumvention of this problem. Simultaneous cochlear implantation has been done safely and effectively. 9 However, in addition to the technical demands of cochlear nerve preservation and complete schwannoma removal, financial obstacles in the form of limited insurance approval in the setting of contralateral normal hearing remain. For situations in which access to cochlear implant is precluded by normal contralateral hearing, placement of silastic lumen keeper may be an option to prevent the substantial possibility of obliteration and allow for future cochlear implantation.
Conclusions
At 54%, the rate of preservation of the cochlear patency after translabyrinthine surgery is higher than reported in other studies. However, preoperative tumor size, hearing status, and cochlear FLAIR intensity are not reliable predictors of postoperative cochlear patency. If cochlear nerve preservation is achieved during translabyrinthine surgery, cochlear lumen keeper and simultaneous cochlear implantation are options to prevent loss of opportunity for cochlear implantation due to cochlear obliteration.
Author Contributions
Disclosures
Footnotes
Acknowledgements
We thank Nancy Jackson, PhD, for her contributions in statistical analysis. We also thank Seilesh Babu, MD, Dennis Bojrab, MD, Eleanor Chan, MD, Robert Hong, MD, Michael LaRouere, MD, Ilka Naumann, MD, and John Zappia, MD, of the Michigan Ear Institute for their contributions in providing patient data and suggestions to this study.
No sponsorships or competing interests have been disclosed for this article.