Bilateral Cochlear Implantation in Children With Anomalous Cochleovestibular Anatomy

Objective  To outline clinical experience with bilateral cochlear implantation in children with cochleovestibular anomalies.

Design  A prospective cohort study with a mean follow-up of 12 months.

Setting  An academic, pediatric, tertiary referral center.

Patients  All eligible children were prospectively recruited from January 1, 2007, through October 31, 2008. Ten children aged 9 to 33 months who had congenital inner ear malformations, including common cavity, incomplete partition, and cochleovestibular hypoplasia, participated.

Interventions  Bilateral cochlear implantation was performed sequentially with an interimplantation delay greater than 2 years in 7 children and less than 1 year in 1 child. Bilateral simultaneous implantation was performed in 2 children.

Main Outcome Measures  Complications, hearing outcomes, and balance outcomes.

Results  All children underwent successful implantation. Five children had a perilymph “gusher” (on 1 side only), and there were no other complications. All children had 22 active electrodes bilaterally and achieved speech reception. All 8 children who underwent closed-set speech perception testing scored above 75%, and 5 of the 7 children who underwent open-set testing achieved scores above 75%. Despite variable vestibular function before bilateral implantation, no prolonged imbalance occurred, although 3 children (30%) had transient unsteadiness for up to 2 weeks after the second implantation.

Conclusions  Bilateral cochlear implantation was performed safely and successfully in children with a spectrum of bilaterally anomalous cochleovestibular anatomy. Hearing outcomes suggest that these children should not be excluded from undergoing bilateral implantation. This study provides guidance on candidacy issues, surgical decision making, and surgical techniques in this group.

We have previously reported that abnormalities in cochleovestibular anatomy can provide greater-than-normal challenges for cochlear implant surgery and programming.¹ Because of the uncertainty that surrounds the potential for binaural processing in these children,² our academic, pediatric, tertiary referral center has cautiously progressed toward performing bilateral cochlear implantation in the presence of abnormal cochleovestibular anatomy. The primary objective of the present study is to determine whether bilateral cochlear implantation should be considered for children who have cochleovestibular abnormalities. Specifically, our aims were to (1) determine the safety of bilateral cochlear implantation, (2) describe a specific surgical technique and potential pitfalls, (3) explore the process of candidacy assessment, and (4) evaluate the potential for improved hearing relative to unilateral implantation.

Hearing-impaired children have been undergoing cochlear implantation to restore the sense of audition for more than 20 years,^3,4 and significant advances in surgical technique and device design have been made. Jackler et al⁵ found that 20% of children with sensorineural hearing loss have associated bony abnormalities of the labyrinth, and more recent studies^1,6 that used high-resolution computed tomography have reported an incidence as high as 30% to 35%. Malformation of the inner ear was once considered a contraindication to cochlear implant surgery.⁷ Numerous studies^1,8-14 from different centers have now described successful unilateral cochlear implantation in these children in terms of both surgical technique and audiologic outcomes.

Cochlear implantation in children with prelingual severe to profound sensorineural hearing loss can result in excellent speech and language outcomes, with the achievement of age-appropriate norms in some children.¹⁵ Despite this, children with unilateral cochlear implants typically have significant difficulty with the localization of sound and hearing in noise.¹⁶Focus on the provision of binaural hearing for children with implants has increased, and recent work has demonstrated better performance with bilateral implants than unilateral implants in speech perception in quiet (binaural summation effect),¹⁷ speech perception in noise (head shadow effect),^18,19 and sound localization.²⁰ Therefore, bilateral implantation may allow children to obtain binaural hearing and improved speech perception in situations such as the classroom and other noisy environments.

To date, little published information is available with regard to the role of bilateral implantation in children with anomalous inner ear morphology. The present study outlines our clinical experience with these children, including a discussion of preoperative imaging, candidacy, electrode array choice, complicating factors, and early outcomes.

All eligible children were prospectively recruited for our study of bilateral cochlear implantation as approved by the Research Ethics Board at The Hospital for Sick Children from January 1, 2007, through October 31, 2008. The children met appropriate suitability criteria for their first implantation through a multidisciplinary team assessment, as previously described.²¹ Before implantation in the first ear, each patient underwent high-resolution computed tomography with contiguous 0.625-mm-thick sections (GE Ultra-lightspeed; General Electric Healthcare, Milwaukee, Wisconsin). We performed 160-mA axial and 100-mA direct coronal planes through the petrous temporal bones. Nine of the included children also underwent magnetic resonance imaging of the brain and internal auditory canals by use of T2 fast spin echo and multishot echo-planar imaging with steady-state free precession axial views of the inner ears and direct sagittal oblique sequences of the internal auditory canals. Radiologic anomalies were classified by the use of previously described criteria.¹ All children with cochleovestibular anomalies who underwent bilateral cochlear implantation were included in this study. Children with complete cochleovestibular aplasia (Michel aplasia), cochlear nerve aplasia, or cochlear nerve canal aplasia were considered inappropriate candidates for cochlear implantation and therefore were excluded from this study.We consider children with isolated vestibular aqueduct enlargement indistinguishable from children with normal anatomy from the perspective of implant surgery and speech and language perception, and these children were also excluded from the study.

Ten infants and children (5 boys and 5 girls) were recruited for participation in this study. The age at the time of the first implantation ranged from 9 to 33 months (mean, 19 months; SD, 10 months). There was a delay between the first and second implantations of more than 2 years in 7 participants (mean, 5.1 years; SD, 1.3 years), a delay of 11 months in 1 participant, and simultaneous bilateral implantation in 2 participants. The mean follow-up period after bilateral implantation was 12 months (SD, 6 months; range, 4-26 months). For comparison, during the same period and outside this study, 5 unilateral implantations were performed for children with anomalous cochleovestibular anatomy, and 67 children with normal anatomy (including isolated vestibular aqueduct enlargement) received sequential or simultaneous bilateral implants.

Intraoperative surgical findings were reviewed to identify the presence of cerebrospinal fluid (CSF) or perilymph leakage, abnormalities of facial nerve anatomy, the number of implant electrodes inserted, and any complications or difficulties encountered. Postoperative outcomes were assessed by the review of behavioral measures of speech perception, which are routinely completed in all children in our program at follow-up visits. These outcomes include closed-set tests (Early Speech Perception Test, Word Identification by Picture Index, and Test of Auditory Comprehension), in each of which the child hears a word and chooses the appropriate object or picture from a set, and open-set tests (Phonetically Balanced Kindergarten Word test and Glendonald Auditory Screening Procedure test), in each of which the child must repeat the word without any additional cues. Scores are provided as the percentage of correct responses. We also assessed the number of electrodes used for the child, the most recent behavioral minimum threshold, and behavioral maximum comfort stimulation levels from his or her most recent cochlear implantation MAP. Vestibular-function testing (computerized rotation test, vestibular-evoked myogenic potentials, and electronystagmography, including caloric testing) was performed before and after the second cochlear implantation in children who were 4 years or older.

The 10 included children had a spectrum of cochleovestibular anomalies, with the inclusion of cochlear hypoplasia, incomplete partition, and common cavity deformities (Table 1 and Figure 1). Vestibular function tests were performed before the second cochlear implantation in the 6 children who were 4 years or older (Table 1). The 2 children with bilateral common cavity deformity demonstrated severe bilateral peripheral vestibular dysfunction. One child with bilateral cochlear hypoplasia had a weak left caloric response, having undergone implantation on the right side first, and another had normal function. Of the children with incomplete partition, one demonstrated some right impairment (the already-implanted side) and another had normal vestibular function. After the second cochlear implantation, 7 (70%) of the children demonstrated no evidence of imbalance and the other 3 had mild imbalance for up to 2 weeks. None of the children demonstrated prolonged imbalance.

Five children demonstrated a CSF or perilymph leak on cochleostomy, and in all patients this occurred only in the left ear (25% of study ears). This was the first ear operated on in 2 patients and the second ear operated on in 3 patients. Of these 5 leaks, 1 was of a low-volume oozing type consistent with perilymph, and 4 were high-volume “gushers” consistent with CSF leakage. Three of the leaks were managed by electrode array insertion and the packing of the cochleostomy with several small pieces of temporalis fascia. The fourth and fifth gushers required additional application of Tisseel fibrin glue. In the fifth only, a continuous lumbar drain was inserted during the time on the operating table because of significant concern about the risk of late dislodgement of the seal with physiologic elevation in CSF pressure.

Two patients had evidence of unilateral abnormal facial nerve anatomy (10% of study ears), with the facial nerve displaced anteriorly. This finding did not require any change in surgical technique. One child had an asymptomatic, incidental, unilateral, congenital cholesteatoma identified on preoperative magnetic resonance imaging and computed tomography, which was confirmed at the time of surgery. This child was treated in 2 stages. First, the cholesteatoma was surgically treated and the anomalous cochlea implanted in the contralateral healthy ear. Second, after 12 months, a second-look operation was performed on the ear with previous congenital cholesteatoma that had not undergone implantation. When absence of residual disease was confirmed, we proceeded to perform cochlear implantation on this side, without complication. No other complications, such as wound infection, facial nerve weakness, or nonstimulation, were encountered in this series. Details of the implant electrode arrays used and the number of electrodes inserted are outlined in Table 2.

All 10 children were able to perceive the full complement of 22 active electrodes bilaterally, confirmed by pure tone audiometry. The mean MAP dynamic ranges (maximum stimulation level minus minimum stimulation level) were calculated for electrodes 20, 9, and 3, thereby representing the distal, middle, and proximal portions of the electrode array (Table 3). These were compared for the first and second implantations (Figure 2) and found to be statistically significantly lower in the ears that underwent implantation second (Wilcoxon signed rank test, P < .001).

Limited speech outcome data are available for the children in the present study because of difficulty with speech perception testing in young children and the short follow-up period. In Table 4, hearing outcome data are summarized. In general, all 10 children were detecting speech and most were developing good speech perception scores on closed-set and/or open-set tests.

The potential advantages of binaural input are particularly relevant for children with anomalous cochleovestibular anatomy, who may have increased challenges and in some cases underperform with a unilateral cochlear implant.¹ To establish the safety of the undertaking of bilateral implantation in these children, we have taken a step-by-step approach. The initial step has been to establish an early age for the first implantation in children with normal anatomy who undergo sequential implantation. Through development of our surgical and anesthetic techniques, it became routine to perform early simultaneous bilateral implantation in children with normal anatomy. Once we had developed extensive experience with unilateral implantation in children with anomalous cochleovestibular anatomy,¹ we were able to combine this with our experience of sequential bilateral implantation and most recently simultaneous implantation (patients 8 and 9).

Two patients in the present series (1 with bilateral cochlear hypoplasia and 1 with bilateral incomplete partition) were noted to have abnormal facial nerve anatomy with the facial nerve on 1 side displaced anteriorly. This proportion is in keeping with the prevalence rates of aberrant facial nerves associated with cochleovestibular anatomy in other published studies (16%-17%)^22,23 and in our own series of children who underwent unilateral implantation (14%).¹

Although the unpredictability of perilymph or CSF gushers after cochleostomy is recognized in this population,^12,24 our policy has been to implant the side with the less anomalous anatomy for the first procedure (Figure 1), which reduces the risk of gushers. Hence, 3 of the 5 perilymph or CSF gushers occurred during the second side procedure (in the more anatomically anomalous ear). Despite this policy, 2 of the patients had a significant leak only during the procedure on their first (less anatomically anomalous) ear, which emphasizes the difficulty in the prediction of a perilymph or CSF gusher from the radiologic appearances. In bilateral simultaneous implantation, we would only proceed to the second side in the absence of uncontrolled perilymph or CSF leakage or concern about the facial nerve function after the first-side procedure.

In a normal cochlea, the ideal electrode position is with medially directed stimulating electrodes in close proximity to the modiolus to maximize stimulation of auditory neurons. The situation is complicated in the presence of anomalous cochlear anatomy, and consideration must be given to the placement and the type of electrode array to be used. In children with incomplete partition without shortening (incomplete partition type 2), it should be possible to fully insert a precurved modiolus-hugging electrode array, as in the normal cochlea. In incomplete partition with a shortened and enlarged basal coil (incomplete partition type 1), full insertion of all stimulating electrodes is still usually possible. In the common cavity deformity, the interscalar septum and modiolus are absent, so the precise location of the stimulatable neural tissue is not definitively known. In these cases we advocate the use of a straight electrode array with circumferential stimulating electrodes. This array is inserted into the cavity with the stimulating elements directed laterally toward the outside of the cavity, where contact with neural elements should theoretically be maximized. A similar principle applies to some cases of incomplete partition type 1, which have significant absence of the modiolus.

In any anomalous cochlea, aggressive attempts at full insertion may result in misplacement of the electrode array through the deficient modiolus into the internal auditory canal.¹¹ This happens because a lack of space does not allow the electrode array to advance distally within the cochlea; secondarily, the array becomes directed toward the potentially deficient modiolus. In common cavity and severe incomplete partition type 1 deformities, we overcome the risk of misdirected insertion into the internal auditory canal by the use of a straight electrode array: we gently push it against the promontory before insertion into the cochleostomy to create a slight curvature over the first 3 to 5 electrodes. We believe this allows us to steer the electrode array toward the modiolus and avoid misplacement into the internal auditory canal.

In common cavities, alternative insertion techniques in which the array is advanced through the anlage of the vestibule or the lateral semicircular canal would avoid the risk of the approach toward the middle ear through a facial recess and give good access to the common cavity.²⁵ This technique was used in 1 child in our series on the second, more anomalous cochlea, but we have generally preferred to use the standard approach if possible. The decision to use a precurved or straight electrode array therefore depends largely on the extent of the modiolus and the degree of partitioning. The preoperative radiologic imaging allows careful assessment of the cochlear partitioning and in most cases provides the information required to make the correct electrode array choice before surgery.

In all children in this series, all 22 electrodes were active on both sides and provided children with a range of stimulation levels as indicated in Table 3. This confirmed the potential for improved hearing through cochlear implants in both ears. The dynamic ranges with the second implant were found to be significantly lower than with the first implant. This was possibly because the ear that underwent implantation first in each patient was considered to have the least dysplastic cochlea, which gives it the potential for better stimulation. Previous work by 1 of the co-authors¹ demonstrated that children with more severe cochlear anomalies indeed had reduced dynamic ranges compared with those with less severe anomalies or normal anatomy. Alternatively, the difference might reflect evolution in the cochlear implant array design because some of the children received an updated generation of implant electrode array for their second procedure (Table 2).

The children in this series all demonstrated the ability to hear with implants in each ear independently by the use of pure-tone audiometry testing. As indicated in Table 4, all have achieved speech reception, and all 8 tested children had closed-set speech perception scores above 75% (as determined through the use of the Word Identification by Picture Index, Test of Auditory Comprehension, and Early Speech Perception Test). In all 8 patients, this result was achieved before the second implantation was performed, except in patient 7, who was too young to be tested until after the second implantation. Seven tested children achieved open-set speech perception scores greater than 75% (as determined through the use of the Phonetically Balanced Kindergarten Word, Phonetically Balanced Kindergarten Phonemes, and Glendonald Auditory Screening Procedure tests), whereas the 3 children who had a common cavity deformity did not collectively score greater than 75%. Despite the poorer performance in the children with common cavity anomaly, cochlear implantation can be justified because these children perform significantly better than the worst-performing children with normal cochleae.¹ It will be some years before late-outcome data will become available for report, but the early outcomes reported in this study confirm that children with cochleovestibular anomalies perform similarly to their peers with normal cochleas. We therefore suggest that these children should not be excluded from consideration for bilateral implantation on the grounds of the anomalous anatomy.

Children in our program who are considered for bilateral implantation have routinely undergone vestibular function testing before the second implantation, with the inclusion of the computerized rotation test, vestibular-evoked myogenic potentials, and electronystagmography with caloric testing. These tests were performed in children older than 4 years because of compliance difficulties in younger children, and this factor limited the availability of data in our series to 6 children (60%). The residual function before the second implantation varied among these children. There was severe dysfunction bilaterally in the 2 children with common cavity deformity, but this is likely to have been a congenital deficiency, and these children did not demonstrate imbalance either before or after their cochlear implantations. Imbalance after the second implantation (but interestingly not the first) was only experienced by 3 children (2 with incomplete partition and 1 with cochlear hypoplasia), and all these children had early resolution of symptoms.

Other work^26,27 has failed to demonstrate significant clinical vestibular effects after unilateral and bilateral implantation in children with normal cochleas. One study²⁸ used preoperative testing of implantation candidates: it demonstrated 68% to have vestibular deficits, with only 21% who showed any deterioration in the ear that was to undergo implantation. We have recently found that children with acquired absence of vestibular function secondary to meningitis perform worse than their peers in a standardized balance test but can balance well in everyday situations.²⁹ These studies, combined with the lack of any prolonged imbalance sequelae in the present series, suggest that bilateral implantation in these children does not appear to be contraindicated or influenced by the status of the vestibular system after the first implantation. Currently, we continue to perform vestibular testing before a second implantation and accumulate data in patients with both normal and anomalous anatomy, but we may eventually have enough evidence to suggest that vestibular testing is not routinely required.

The present study provides evidence that it is safe to perform bilateral implantation in children with cochleovestibular anomalies, with our series showing no surgical complications or persistent vestibular compromise. Early hearing outcomes in these children have been in keeping with their peers with normal anatomy. We will have a clearer understanding of the role of bilateral implantation when long-term outcome data are available, which will allow comparison of children with anomalous anatomy who have undergone bilateral implantation with children with anomalous anatomy who have undergone unilateral implantation and children with normal anatomy who have undergone bilateral implantation. Because the present study has shown no undue risk in undergoing a second implantation (delayed or simultaneously) and a potential for hearing benefit, we will continue to offer this option to children in our implantation program as part of our research protocol of the exploration of the potential benefits of bilateral implantation.

Correspondence: Neil K. Chadha, MPH, FRCS(ORL-HNS), MBChB(Hons), Department of Otolaryngology, The Hospital for Sick Children, 555 University Ave, Toronto, ON M5G 1X8, Canada (neil.chadha@utoronto.ca).

Submitted for Publication: November 16, 2008; final revision received April 1, 2009; accepted April 2, 2009.

Author Contributions: Dr Chadha had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Chadha, James, Gordon, and Papsin. Acquisition of data: Chadha, Blaser, and Papsin. Analysis and interpretation of data: Chadha, James, Gordon, Blaser, and Papsin. Drafting of the manuscript: Chadha. Critical revision of the manuscript for important intellectual content: James, Gordon, Blaser, and Papsin. Statistical analysis: Chadha. Administrative, technical, and material support: Chadha and Blaser. Study supervision: James, Gordon, and Papsin.

Financial Disclosure: Dr Papsin has acted as a consultant for Cochlear Corporation, a manufacturer of cochlear implants.

Additional Contributions: We gratefully acknowledge the contribution of all members of the team in the Cochlear Implant Program at The Hospital for Sick Children.