INTRODUCTION
Cochlear implantation is an effective treatment for the habilitation of children with congenital severe to profound sensorineural hearing loss. Most children with prelingual deafness who undergo cochlear implantation at an appropriate age can achieve age-appropriate language and speech development. However, small number of children with prelingual deafness benefit little after cochlear implantation. These include children with cochlear nerve deficiency (CND) or severe cochlear malformations. CND refers to a small (hypoplastic) or absent (aplastic) cochlear nerve displayed on magnetic resonance imaging (MRI) scans (Adunka et al. 2006; Sennaroglu 2010). Previous studies have demonstrated that the responsiveness of cochlear nerves to electrical stimulation in children with CND was worse than that in children with normal-sized cochlear nerves (He et al. 2018). As the ability of the cochlear nerves to respond well to electrical stimulation is essential for good auditory sensitivity, cochlear implant (CI) benefits are reduced in children with CND.
Previous evidence indicates that the outcomes of cochlear implantation in children with CND are poor and greatly varied among individuals (Young et al. 2012; Vincenti et al. 2014; Wu et al. 2015; Chao et al. 2016). A recent meta-analysis by Vesseur et al. (2018) showed the auditory and speech outcomes with CI use of 108 patients with CND from 15 published studies. They concluded that only approximately 25% attained open-set speech perception, 34% attained closed-set speech perception, and the remaining 41% showed improvements in only sound awareness or worsened. Furthermore, compared with patients with hypoplastic cochlear nerves, patients with aplastic cochlear nerves tend to benefit less from CIs (Kutz et al. 2011; Peng et al. 2017). However, most previous studies have been small case series with outcomes evaluated at a single time point post-implantation. To date, few studies have revealed the long-term development of the auditory and speech abilities of children with CND who underwent cochlear implantation. Moreover, although some patients with CND can achieve simple speech after implantation, few studies have reported open-set speech recognition results over time (Young et al. 2012). Consequently, little is known about the auditory and speech trajectories of children who have deficient cochlear nerves.
The present study was undertaken to review the long-term outcomes of cochlear implantation in children with deficient cochlear nerves and to compare their auditory skills and speech ability development to those of children with normal-sized cochlear nerves who received implantation.
MATERIALS AND METHODS
Participants
This retrospective case-control study was approved by the Ethical Committee of Shandong Provincial ENT Hospital affiliated with Shandong University (No. XYK20170906). Data regarding children diagnosed with CND who underwent cochlear implantation at Shandong provincial ENT hospital, a tertiary referral center, from September 2012 to December 2018 were comprehensively reviewed. All participants were required to meet the following inclusion criteria: (1) had bilateral cochlear nerve aplasia diagnosed using imaging results, (2) had congenital bilateral severe to profound sensorineural hearing loss before implantation, (3) underwent unilateral cochlear implantation, (4) underwent postoperative rehabilitation for at least 1 year, (5) had been using the device for at least 2 years, and (6) had a normal cochlear structure, or had only Mondini malformation. The exclusion criteria were as follows: patients (1) with severe comorbidities, (2) with other severe cochlear malformations (such as common cavity, cochlear hypoplasia, or cochlear incomplete partitions type I and III), (3) who underwent bilateral cochlear implantation, and (4) who did not wish to participate in this study. The frequency matching in terms of implantation age, sex, preoperative auditory and speech abilities, and the type of electrode array was used to select the control group participants. Children with normal-sized cochlear nerves who underwent unilateral cochlear implantation at the same center were chosen in terms of implantation age, preoperative auditory and speech abilities, and the type of electrode array. The required sample size in the control group was calculated using PASS 11.0 software. The native language of all participants was Mandarin Chinese. At pre-implantation, the severity of hearing loss was assessed using the distortion product otoacoustic emission, auditory brainstem evoked response test, behavioral audiometry with a wobble tone, or auditory steady state response.
Radiological Assessment
All participants underwent MRI and high-resolution temporal bone computed tomography (HRCT) to evaluate cochlear nerve status and other inner ear malformations before the operation, following previously described protocols (Purcell et al. 2015; Chao et al. 2016). MRI was performed using a clinical 3.0 T MRI system (MAGNETOM Verio; Siemens, Erlangen, Germany). MRI sequences included axial T1-weighted and T2-weighted imaging and three-dimensional fast spin echo T2-weighted sequences. Direct oblique sagittal images perpendicular to the long axis of the internal auditory canal (IAC) were obtained to show the cochlear, facial, superior vestibular, and inferior vestibular nerves in the IAC. The cochlear nerve was evaluated based on the results of the MRI scan following a previously described protocol (Casselman et al. 1997). Cochlear nerve aplasia was diagnosed if the cochlear nerve could not be identified on any plane of the MRI scans. The diameter of the auditory nerves was measured on the axial images at the cerebellopontine angle. HRCT was performed using a 64-slice multidetector CT scanner (Somatom Sensation Cardiac 64, Siemens Medical Solutions, Forchheim, Germany). Both axial and coronal images were obtained. Cochlear structure was evaluated to determine whether it had a malformation. In the CND group, the diameters of the IAC and bony cochlear nerve canal (BCNC) were measured on the axial plane for each participant. IAC stenosis was defined as an IAC diameter of less than 3 mm, and BCNC stenosis was diagnosed if the BCNC diameter was less than 1.5 mm, while BCNC atresia was diagnosed if BCNC was absent on any plane on the HRCT scans (Purcell et al. 2015; Lim et al. 2018).
Cochlear Implant Programming
For all participants, CI programming and assessment were performed regularly postoperatively (usually at the first, third, sixth, ninth, 12th, 18th, and 24th months post-activation and every 6 months to 1 year thereafter). Participants’ MAPs were created using Custom Sound (version 4.3) software (Cochlear Ltd., Macquarie, NSW, Australia). The speech processor used was Freedom or Nucleus 5 (Cochlear Ltd.). The strategy used was the advanced combination encoder, the sensitivity was set at 12, and the stimulation mode used was monopolar 1 plus 2 (MP1 + 2) in all participants’ speech processors. All electrodes were activated unless the child definitively showed no response on the electrode. Default pulse widths (25 μs/phase) and stimulation rates (900 pps) were used in the control group. Our previous studies showed that the responses of the cochlear nerve to electrical stimulation were reduced and the recovery period was prolonged in children with CND, which indicates that they may require higher stimulation levels for their MAPs, and a lower pulse rate might be more beneficial (He et al. 2018). In this study, the pulse widths used were 37 to 88 μs/phase and the stimulation rates used were 500 or 720 pps in the CND group. The comfort levels (C-levels) and threshold levels (T-levels) were set using approaches based on age and the capacity to provide behavioral responses. T-levels were set to the lowest levels the children could respond to, and C-levels were set to the maximum comfortable levels. For children who did not understand the concept of loudness of sound owing to poor auditory-verbal skills, C-levels were set at the highest levels to which the patient showed no signs of discomfort. For children whose C-levels were higher than the compliance value, the C-levels were set below the compliance value limit. For children with facial nerve stimulation, the C-levels were set at the highest levels that would not cause facial nerve stimulation, or alternatively, the responsible electrodes were deactivated. After programming, loud sounds were used to confirm that the children had a comfortable listening experience. The overall C-levels would be slightly decreased if the children felt the sounds were too loud. It was challenging to program for young children (especially children with CND) who could not cooperate with the testing of their behavioral levels. For these children, electrically evoked compound action potential thresholds, aided hearing thresholds with CIs in the sound room, and feedback from parents were used as references. For most children with CND, an appropriate MAP could be achieved 6 to 12 months after the first activation.
Audiological Assessments
Post-implantation, the auditory perceptual capabilities and speech ability development were assessed using an array of appropriate tests, including the Categories of Auditory Performance (CAP) scales, the Speech Intelligibility Rating (SIR) scales, behavioral audiometry with wobble tone, the Mandarin early speech perception (MESP) test, and the Mandarin Bamford–Kowal–Bench test. The CAP and SIR are parent questionnaires assessing the child’s auditory perceptual capabilities with CIs and speech intelligibility (Albalawi et al. 2019). These two questionnaires were administered at each visit. The aided hearing threshold (including 500 Hz, 1 kHz, 2 kHz, and 4 kHz) was collected in a free-field sound room using visual reinforcement audiometry or play audiometry. This test was also performed at each visit. The MESP test is a closed-set word recognition test, in which a limited number of potential choices are available to the participant (Zheng et al. 2010). The MESP test was attempted when the child was able to participate and when the CAP scale was higher than level 4. The Mandarin Bamford–Kowal–Bench test is an open-set word- (including monosyllabic and bisyllabic words) and sentence recognition test (Xi et al. 2012). This was attempted when the child passed the closed-set testing and could express simple words or sentences. All speech perception tests were performed in a sound room using live voice or recorded stimuli, presented at 30 dB HL higher than the average hearing thresholds or at the maximum output level (75 dB HL).
Data Analyses
In this study, the average hearing thresholds were calculated using thresholds of 500 Hz, 1 kHz, 2 kHz, and 4 kHz. When there was no measurable response, the threshold was considered to be 10 dB HL higher than the maximum output intensity. For some children who could not fully participate in the behavioral threshold test, the mean aided hearing thresholds were calculated with two or three frequencies. Dependent variables evaluated in this study included CAP scores, SIR scores, hearing thresholds, and word recognition rates. Generalized linear mixed-effect models (GLMMs) were selected to analyze the data in this study, as they can robustly handle missing data. The difference between the study group and the control group was assessed using the GLMMs with the participant group and test time as the fixed effects and the participant as the random effect. For post-hoc tests, Bonferroni correction was used to adjust the p values of multiple comparisons. In addition, participants in the CND group were divided into two subgroups depending on the implantation age. Children who underwent the implantation at less than 3 years of age were classified under the CND group 1, and children who underwent the implantation at more than 3 years of age were classified under the CND group 2. The difference between these two subgroups was assessed using the GLMMs with the group as the fixed effect and the participant as the random effect. Statistical significance was set at p < 0.05. All statistical analyses were performed using the Statistical Package for the Social Sciences version 26.0 (IBM SPSS Statistics for Windows, Version 26.0; IBM Corp, Armonk, NY).
RESULTS
Basic Characteristics of Participants
A total of 55 children with cochlear nerve aplasia met the inclusion criteria. The mean age at the time of implantation in the study group was 2.9 years (SD = 1.6, range = 0.9 to 6.9), and 60% of the children received their implants before the age of 3 years. All participants were implanted with a Cochlear Nucleus device with a contour electrode array, either a Freedom (24RE[CA]) or a CI512 CI. The electrode array was fully inserted in all participants. The control group included 35 implanted children with normal-sized cochlear nerves. The control group did not differ from the study group in terms of the age at implantation (p > 0.05), pre-implantation CAP and SIR scores (p > 0.05), or pre-implantation average aided hearing thresholds in the implanted ear (p > 0.05).
One participant in the study group had given up wearing her device owing to poor outcomes for more than 2 years, but all remaining participants continuously used their devices. At the time of the last follow-up, the mean age of the participants in the study group was 7.5 years (SD = 2.3, range = 3.9 to 13.4), and the mean duration of CI use was 4.5 years (SD = 1.5, range = 2.0 to 9.5). Preoperatively, all participants failed to pass the distortion product otoacoustic emission, and auditory brainstem evoked response responses were not seen at the limits of the equipment. The average hearing thresholds with hearing aids were 80.3 dB HL (SD = 18.2) in the implanted ears and 78.6 dB HL (SD =16.4) in the contralateral ears. The paired t test revealed no significant difference in the aided hearing threshold between the implanted ear and the unimplanted ear at pre-implantation (t = 0.56, p > 0.05). In this study, 22 participants continuously or intermittently used hearing aids in the contralateral ear at post-implantation and the average hearing threshold with hearing aids was 66.5 dB HL (SD = 20.8; range: 30.0 to 96.3). Participants' demographic information and baseline characteristics are presented in Table 1 (see individual participant information in Supplemental Digital Content 1, https://links.lww.com/EANDH/B71).
TABLE 1. -
Demographic information of subjects who participated in this study
| Variable |
Cochlear Nerve Aplasia Group |
Control Group |
| Sex, No. (%) |
|  Female |
29 (53) |
20 (57) |
|  Male |
26 (47) |
15 (35) |
| Implanted ear, No. (%) |
|  Right |
25 (45) |
26 (74) |
|  Left |
30 (55) |
9 (26) |
| Implanted age |
2.9 (1.6) |
2.4 (1.5) |
| Mean (SD) (range), yrs |
0.9–6.9 |
1.1–6.5 |
| Duration of implant use |
4.5 (1.5) |
5.0 (0.8) |
| Mean (SD) (range), yrs |
2.0–9.5 |
2.8–6.2 |
|  2–3 yrs |
55 |
35 |
|  3–4 yrs |
50 |
35 |
|  4–5 yrs |
36 |
28 |
|  >5 yrs |
23 |
17 |
| Electrode array |
|  24 RECA |
53 |
30 |
|  CI512 |
2 |
5 |
Imaging Characteristics of Children With Cochlear Nerve Aplasia
In the CND group, the cochlear nerves were absent in all participants on any planes of the MRI scans bilaterally. Except for one participant, all had two nerve bundles (vestibulocochlear nerve and facial nerve) that presented in the IAC of the tested ears. For participant CND54, only one nerve bundle was present in the IAC (which was considered to be the facial nerve, as she had normal facial nerve function). The mean diameter of the auditory nerve in the tested ear was 1.43 mm (SD = 0.33; range = 0.70 to 2.17). HRCT scans indicated that, except for four children with bilateral Mondini malformation, all children had bilateral normal cochlea. Of the 55 tested ears in the CND group, 49 (89%) had BCNC stenosis, five (10%) had BCNC atresia, and only one (1%) had a normal BCNC. The mean width of the BCNC was 0.85 mm (SD = 0.36; range = 0 to 1.60). In addition, 16 (29%) children had IAC stenosis, whereas the other 39 (71%) children had normal IACs. The mean width of the IAC was 3.53 mm (SD = 1.06; range = 1.48 to 6.24). Detailed imaging results (including the width of the IAC, width of the BCNC, and diameter of the auditory nerve) of individual participants in the CND group are presented in Supplemental Digital Content 1, https://links.lww.com/EANDH/B71.
Post-implantation Auditory and Speech Performance in Children With Cochlear Nerve Aplasia
Two children demonstrated sound detection with no further improvement on the test battery compared with the baseline condition. Compared with pre-implantation, the mean aided hearing threshold significantly improved by the first year post-implantation (p < 0.01), and further improved by the second year post-implantation (p < 0.01). The aided hearing threshold stabilized at 3 years post-implantation, and there were no significant changes subsequently (p > 0.05). The preoperative and postoperative CAP scores in the study group are shown in Figure 1A. Pre-implantation CAP scores were less than 3 for all children, with a mean score of 0.40 (SD = 0.56). The GLMM test showed that the duration of implant use had a significant influence on the CAP scores (F(5, 207.08) = 382.99, p < 0.001). Compared with pre-implantation, CAP scores significantly improved by 1-year post-implantation (p < 0.001) and continued to improve over time. The pairwise comparisons with Bonferroni correction showed significant differences in the CAP scores between the first and second years (p < 0.001) and between the second and fourth years (p < 0.05). However, there were no significant differences in the CAP scores between the second and third years (p = 1.00), between the third and fourth years (p = 0.39), between the fourth and fifth years (p = 1.00), or between the third and fifth years (p = 0.08). The preoperative and postoperative SIR scores in the study group are shown in Figure 1B. At pre-implantation, the mean SIR score was 1.18 (SD = 0.43), with one participant scoring 3, two participants scoring 2, and all other participants scoring 1. Similarly, the GLMM test results showed that the duration of implant use significantly influenced the SIR scores (F(5, 207.68) = 97.51, p < 0.001). The SIR scores also gradually improved over time, post-implantation. The pairwise comparisons with Bonferroni correction showed significant differences in the SIR scores between the first and second years (p < 0.001) and between the third and fourth years (p < 0.01). However, there were no significant differences in the SIR scores between the second and third years (p = 0.33) or between the fourth and fifth years (p = 1.00). The CAP scores were higher in CND group 1 than in CND group 2 for all tests. The SIR scores were lower in CND group 1 at pre-implantation, but were higher in the fourth and fifth years post-implantation than in CND group 2. Subgroup comparisons of GLMMs revealed no significant group effect on CAP scores (F(1, 67.43) = 0.17, p = 0.67) or SIR scores (F(1, 57.56) = 1.07, p = 0.30). In addition, linear regression analysis showed that there were no significant correlations between the width of the BCNC and CAP scores or SIR scores (p = 0.07 to 0.63).
;)
Fig. 1.: Pre- and post-implantation CAP scores (A) and SIR scores (B) in children with cochlear nerve aplasia. Round and square symbols represent CAP and SIR scores for individual children with cochlear nerve aplasia. The boxes represent mean scores. The error bars represent ±1 SD. **Means p < 0.01; *means p < 0.05. CAP indicates Categories of Auditory Performance; SIR, Speech Intelligibility Rating.
The numbers of participants in the study group who underwent the closed-set or open-set speech perception tests at different time points post-implantation are displayed in Table 2. During the initial 2 years post-implantation, 35 (64%) children could not participate in the closed-set test, 18 (32%) children with CND participated in the closed-set testing, and only nine (16%) of them passed the test. Moreover, only two children (4%) could participate in the open-set speech perception test, with monosyllabic and bisyllabic word recognition rates of 20% and 22% for one child and 31% and 34% for the other, respectively. With the extension of CI use duration, more children could pass the closed-set testing, and participation in the open-set testing increased. Overall, 27 participants (49%) had some degree of open-set speech perception skills (monosyllabic words score > 0%), 10 participants (18%) passed the closed-set speech perception test, and the other 18 (33%) children could not pass or cooperate with the closed-set speech perception test at the last follow up.
TABLE 2. -
Number of subjects in the study group participated in the closed-set or open-set speech perception tests at different times post-implantation
|
Closed-set Testing (%)/Passed (%) |
Open-set Testing (%) |
Unable to Participate (or Not Available) (%) |
Totals (%) |
| 2 yrs |
18 (32)/9 (16) |
2 (4) |
35 (64) |
55 (100) |
| 3 yrs |
20 (36)/9 (18) |
16 (29) |
19 (35) |
50 (100) |
| 4 yrs |
9 (25)/2 (5) |
21 (58) |
6 (17) |
36 (100) |
| ≥5 yrs |
7 (30)/1 (13) |
11 (48) |
5 (22) |
23 (100) |
The results of the word recognition tests performed at 3 years post-implantation in the study group are displayed in Figure 2. All participants had limited open-set speech performance, even after long-term CI usage. Overall, the word recognition rates increased gradually from the third to the fifth year post-implantation. At the third, fourth, and fifth years post-implantation, the mean monosyllabic word recognition rates were 33.7% (SD = 11.6%), 34.3% (SD = 11.2%), and 38.2% (SD = 14.5%), respectively, and the mean bisyllabic word recognition rates were 29.8% (SD = 14.6%), 33.5% (SD = 13.6%), and 39.4% (SD = 11.8%), respectively. The GLMM showed that the duration of implant use had significant influence on bisyllabic word recognition rates (F(2, 25.13) = 5.38, p < 0.05) but no significant influence on monosyllabic recognition rates (F(2, 26.39) = 2.55, p = 0.09). The pairwise comparisons with Bonferroni correction showed significant differences in the bisyllabic word recognition rates between the third and fifth years post-implantation (p < 0.05). No significant differences in the monosyllabic word recognition rates were found between any two test times (p > 0.05). In addition, word recognition rates varied greatly among individuals, with the monosyllabic word recognition rates ranging from 12% to 68% and the bisyllabic word recognition rates ranging from 10% to 66%. In addition, no significant correlations were found between the width of the BCNC and monosyllabic (p = 0.43) or bisyllabic word recognition rates (p = 0.31).
Fig. 2.: Means and SDs of post-implantation monosyllabic and bisyllabic word recognition rates in the study group, tested 3 years post-implantation. Word recognition rates for individual participants are displayed as round symbols or square symbols. Testing time points are indicated on the X axis.
Comparison With Children With Normal-Sized Cochlear Nerves
The performance of children with CIs and cochlear nerve aplasia was further compared with that of the children in the control group. Figure 3 displays the average hearing thresholds of the two groups tested at different times. Aided hearing thresholds also significantly improved in the control group after activation (p < 0.01), and thresholds further significantly decreased in the second year post-implantation (p < 0.05). Children in the CND group had a significantly higher aided hearing threshold than those in the control group at all test time points (p < 0.01) (Fig. 3). Figure 4 displays the development of the CAP and SIR scores between the CND and control groups. In the control group, CAP and SIR scores significantly improved during the initial 2 years post-activation, which were 6.0 (SD = 0.4) and 3.8 (SD = 0.6), respectively, in the second year post-implantation. For group comparison, the results of the GLMMs showed significant effects of the study group (F(1, 88.05) = 73.87, p < 0.001), testing time (F(5, 361.65) = 1133.78, p < 0.001), and interaction between group and testing time (F(5, 361.65) = 21.63, p < 0.001) on CAP scores. The results of the pairwise comparisons with Bonferroni correction showed that the CND group had significantly lower CAP scores than the control group at all test times (p < 0.01). Similarly, the results of the GLMMs showed significant effects of the study group (F(1, 87.26) = 84.42, p < 0.001), testing time (F(5, 362.29) = 381.64, p < 0.001), and an interaction between group and testing time (F(5, 362.29) = 36.86, p < 0.001) on SIR scores. The results of the pairwise comparisons with Bonferroni correction showed that the CND group had significantly lower SIR scores than the control group at all test times (p < 0.01). Figure 5 shows the comparison of the open-set speech recognition rates between the CND and control groups. The results of the GLMMs showed significant effects of the study group (F(1, 41.90) = 424.25, p < 0.001) and testing time (F(2, 56.92) = 9.26, p < 0.001) for monosyllabic word recognition rates but no significant effect of the interaction between group and testing time (F(2, 56.96) = 0.22, p = 0.80) for monosyllabic word recognition rates. Similarly, the results of the GLMMs showed significant effects of the study group (F(1, 46.90) = 399.92, p < 0.001) and testing time (F(2, 49.12) = 10.62, p < 0.001) for bisyllabic word recognition rates but no significant effect of the interaction between group and testing time (F(2, 49.12) = 1.08, p = 0.34) for bisyllabic word recognition rates. The results of the pairwise comparisons with Bonferroni correction showed that the CND group had significantly lower monosyllabic and bisyllabic word recognition scores than the control group (p < 0.01).
Fig. 3.: Means and SDs of post-implantation aided hearing thresholds in the study and control groups. **Indicates a significant difference (p < 0.01) between the study and control groups. ##Indicates a significant difference (p < 0.01) found between the first year and second year post-implantation in the study group. #Indicates a significant difference (p < 0.05) found between the first year and second year post-implantation in the control group.
Fig. 4.: Means and SDs of CAP and SIR scores in the study and control groups at different time points. CAP and SIR scores gradually improved in both groups at post-implantation. However, the improvement progresses were significantly slower in the CND group (p < 0.01). CAP indicates Categories of Auditory Performance; CND, cochlear nerve deficiency; SIR, Speech Intelligibility Rating.
Fig. 5.: Means and SDs of post-implantation monosyllabic and bisyllabic word recognition rates in the study and control groups. Both monosyllabic and bisyllabic word recognition rates were significantly lower in the CND group than in the control group at all time points (p < 0.01). **p < 0.01; *p < 0.05. CND indicates cochlear nerve deficiency.
DISCUSSION
Children with cochlear nerve aplasia are a cohort of high clinical concern owing to poor outcomes and great variability in outcomes after cochlear implantation. Overall, most previous studies on CI use in children with CND have been limited by small participant cohorts with limited follow-up data. Our study was the first to longitudinally investigate the development of auditory and speech abilities in a large cohort with a long-term follow-up.
One meaningful finding in this longitudinal study was that the progress of auditory and speech development in children with cochlear nerve aplasia was slower than and inferior to that of children with normal-sized cochlear nerves. For children with normal-sized cochlear nerves, the CAP and SIR scores showed a dramatic improvement during the first 2 years post-implantation. Most children with normal-sized cochlear nerves had CAP scores of 6 or 7 and SIR scores of 4 or 5 at 2 years post-implantation, which indicated that they obtained both the ability to understand conversation without lipreading and intelligible spoken language. For children with cochlear nerve aplasia, both auditory perception and speech intelligibility also significantly improved during the first 2 years post-implantation. However, their CAP and SIR scores were significantly lower than those of children with normal-sized cochlear nerves. At 2 years post-implantation, approximately 40% of children with cochlear nerve aplasia still had CAP scores of less than 5, meaning that they could not discriminate common phrases or receive effective speech information input with their CIs. In addition, it was not until 3 years post-implantation that only half of the children with cochlear nerve aplasia developed some ability to understand simple intelligible spoken language. This poor and slow development of auditory and speech abilities in children with cochlear nerve aplasia with CIs mainly points to the poor ability of the cochlear nerves to translate speech sounds into the auditory cortex. Furthermore, children with CND required a significantly longer time to obtain appropriate programming stimulation parameters than those with normal-sized cochlear nerves, which also impaired their perception ability in the post-implantation period. Furthermore, a previous study showed that approximately half of the children with CND have concurrent neurological deficits, which would also affect their cognitive function and learning ability (Birman et al. 2016).
Another important finding in this longitudinal study was that both auditory perception and speech intelligibility tended to improve over a long period post-implantation in children with cochlear nerve aplasia. Although the CAP and SIR scores significantly improved in the first 2 years post-implantation, auditory ability and speech intelligibility did not achieve meaningful development in children with cochlear nerve aplasia. Both CAP and SIR scores significantly improved from the second to the fourth year post-implantation in children with cochlear nerve aplasia. In addition, the likelihood of measuring closed-set and open-set speech perception in children with cochlear nerve aplasia increased over time. These results indicate that auditory and speech abilities tended to improve in the long term. In this study, the developmental trajectories of the auditory and speech abilities were different in children with cochlear nerve aplasia. Although there was no significant improvement in CAP scores, the SIR scores greatly improved from the third to the fifth year post-implantation. The development of speech ability may be due to the accumulation of auditory input and speech rehabilitation training. A previous study also revealed that the development of speech ability may be slower than the development of auditory ability (Guo et al. 2020). Overall, approximately half of the children attained open-set speech perception in this present study. The percentage of children with open-set speech perception ability in this study was higher than those previously reported. Buchman reported that only four (19%) children with CND achieved open-set speech perception ability (Buchman et al. 2011). Young et al. reported that three out of 10 children with CND with CI use achieved open-set speech perception ability (Young et al. 2012). This difference might be due to the large cohort size and long-term use of CIs. Furthermore, appropriate CI mapping is also attributed to these better outcomes.
Furthermore, auditory performance of CIs for children with cochlear nerve aplasia was limited. In this study, children with cochlear nerve aplasia seldom used an exclusively oral communication mode, and most required visual supplementation and lipreading as aids, even after a long period of CI use. This is similar to the results of previous studies. In an analysis of 22 participants with CND treated with CIs, only one child used oral communication (Buchman et al. 2011). In the study by Birman et al., 15 out of 51 children with CND used speech language alone, and half of them used sign language (Birman et al. 2016). In addition, approximately half of the children with cochlear nerve aplasia had SIR scores of 3 or less after more than 5 years of CI use, indicating that it was difficult for them to express intelligible connected speech even after an extended CI use. The limited auditory perception ability further restricts the development of spoken language. In clinics, we usually recommend the families of patients use some form of visual augmentation as a supplement when performing auditory and speech rehabilitation. Consistent rehabilitation and training may help improve the intelligibility of spoken language.
In this study, although most children achieved improvement in auditory discrimination abilities, there were still two children for whom only sound detection ability improved. Many previous studies also indicated that children with aplastic cochlear nerves are less likely to benefit from CIs than children with hypoplastic cochlear nerves (Kang et al. 2010; Kutz et al. 2011; Peng et al. 2017). For children who receive no benefits from CIs, auditory brainstem implantation (ABI) might be an alternative treatment. Some children with ABI can identify speech sounds and obtain open-set speech perception (Sennaroglu et al. 2016; Asfour et al. 2018; van der Straaten et al. 2019). However, the outcomes of ABI also vary greatly among patients (Asfour et al. 2018). Currently, there is no effective assessment to preoperatively predict the benefits of CIs or ABI for an individual child with CND. A previous study showed that the outcomes of CI use in patients with CND could not be predicted using preoperative MRI or CT scans (Yamazaki et al. 2015). Our study further revealed no significant relationships between BCNC width and the outcomes of CI use. Thus, it is challenging to provide appropriate treatment (cochlear implantation or ABI) for an individual child with cochlear nerve aplasia. Previous reports recommend cochlear implantation before ABI, as some children attained closed- or open-set speech perception after cochlear implantation (Vesseur et al. 2018). Based on the results of this present study, we also suggest that for children with cochlear nerve aplasia, clinicians should choose cochlear implantation as the priority treatment.
Limitations
This study has several limitations. One limitation is that some of the patients with cochlear nerve aplasia had contralateral hearing aids. However, the benefits of hearing aids were poor for most children. Thus, we did not analyze the impact of hearing aids on the progress of auditory and speech abilities in children with cochlear nerve aplasia. Whether using hearing aids or CIs on the contralateral ear is more helpful to the development of speech will be investigated in our future study. In addition, CAP and SIR scores can only be used to partially assess the auditory and speech abilities of children with cochlear nerve aplasia; therefore, several important aspects of the development of spoken language remain unanswered, such as the learning of pronunciation. Because our previous studies have shown that the deterioration of the cochlear nerve increased from the basal to the apical parts of the cochlea in children with CND (He et al. 2018), the ability of the cochlear nerve to encode and transmit auditory information containing different frequencies might be influenced. Whether this would affect the learning of different phonemes for children with CND requires further investigation.
CONCLUSION
This study provides a long-term longitudinal investigation of auditory and speech performance post-implantation in a large cohort of children with cochlear nerve aplasia. The results demonstrated that most children with cochlear nerve aplasia could benefit from cochlear implantation, but the development process of auditory and speech abilities was significantly slower and worse than that in children with normal-sized cochlear nerves. Consistent CI usage and auditory rehabilitation may be helpful to achieve meaningful improvements in auditory and speech ability development. With long-term rehabilitation, most children with cochlear nerve aplasia could acquire the ability to distinguish common phrases and develop simple spoken language.
ACKNOWLEDGMENTS
This work was supported by grants from the National Natural Science Foundation of China (No. 81800905, 82071053), and the Natural Science Foundation of Shandong Province Grant (No. ZR2016QZ007).
REFERENCES
Adunka O. F., Roush P. A., Teagle H. F., Brown C. J., Zdanski C. J., Jewells V., Buchman C. A. (2006). Internal auditory canal morphology in children with
cochlear nerve deficiency. Otol Neurotol, 27, 793–801.
Albalawi Y., Nidami M., Almohawas F., Hagr A., Garadat S. N. (2019). Categories of Auditory Performance and speech intelligibility ratings in prelingually deaf children with bilateral implantation. Am J Audiol, 28, 62–68.
Asfour L., Friedmann D. R., Shapiro W. H., Roland J. T. Jr, Waltzman S. B. (2018). Early experience and health related quality of life
outcomes following auditory brainstem implantation in children. Int J Pediatr Otorhinolaryngol, 113, 140–149.
Birman C. S., Powell H. R., Gibson W. P., Elliott E. J. (2016).
Cochlear implant outcomes in cochlea nerve aplasia and hypoplasia. Otol Neurotol, 37, 438–445.
Buchman C. A., Teagle H. F., Roush P. A., Park L. R., Hatch D., Woodard J., Zdanski C., Adunka O. F. (2011). Cochlear implantation in children with labyrinthine anomalies and
cochlear nerve deficiency: implications for auditory brainstem implantation. Laryngoscope, 121, 1979–1988.
Casselman J. W., Offeciers F. E., Govaerts P. J., Kuhweide R., Geldof H., Somers T., D’Hont G. (1997). Aplasia and hypoplasia of the vestibulocochlear nerve: Diagnosis with MR imaging. Radiology, 202, 773–781.
Chao X., Luo J., Fan Z., Shi H., Han Y., Wang R., Song Y., Wang G., Wang H., Xu L. (2016). Usefulness of radiological findings for predicting cochlear implantation
outcomes in children with
cochlear nerve deficiency: A pilot study. Acta Otolaryngol, 136, 1051–1057.
Guo Q., Lyu J., Kong Y., Xu T., Dong R., Qi B., Wang S., Chen X. (2020). The development of auditory performance and speech perception in CI children after long-period follow up. Am J Otolaryngol, 41, 102466.
He S., Shahsavarani B. S., McFayden T. C., Wang H., Gill K. E., Xu L., Chao X., Luo J., Wang R., He N. (2018). Responsiveness of the Electrically stimulated cochlear nerve in children with
cochlear nerve deficiency. Ear Hear, 39, 238–250.
Kang W. S., Lee J. H., Lee H. N., Lee K. S. (2010). Cochlear implantations in young children with
cochlear nerve deficiency diagnosed by MRI. Otolaryngol Head Neck Surg, 143, 101–108.
Kutz J. W. Jr, Lee K. H., Isaacson B., Booth T. N., Sweeney M. H., Roland P. S. (2011). Cochlear implantation in children with cochlear nerve absence or deficiency. Otol Neurotol, 32, 956–961.
Lim C. H., Lim J. H., Kim D., Choi H. S., Lee D. H., Kim D. K. (2018). Bony cochlear nerve canal stenosis in pediatric unilateral sensorineural hearing loss. Int J Pediatr Otorhinolaryngol, 106, 72–74.
Peng K. A., Kuan E. C., Hagan S., Wilkinson E. P., Miller M. E. (2017).
Cochlear nerve aplasia and hypoplasia: Predictors of
cochlear implant success. Otolaryngol Head Neck Surg, 157, 392–400.
Purcell P. L., Iwata A. J., Phillips G. S., Paladin A. M., Sie K. C., Horn D. L. (2015). Bony cochlear nerve canal stenosis and speech discrimination in pediatric unilateral hearing loss. Laryngoscope, 125, 1691–1696.
Sennaroglu L. (2010). Cochlear implantation in inner ear malformations--A review article. Cochlear Implants Int, 11, 4–41.
Sennaroglu L., Sennaroglu G., Yucel E., Bilginer B., Atay G., Bajin M. D., Mocan B. O., Yaral M., Aslan F., Cnar B. C., Ă–zkan B., Batuk M. O., Kirazl C. E., Karakaya J., AtaÅŸ A., Saraç S., Ziyal I. (2016). Long-term results of ABI in children with severe inner ear malformations. Otol Neurotol, 37, 865–872.
van der Straaten T. F. K., Netten A. P., Boermans P., Briaire J. J., Scholing E., Koot R. W., Malessy M. J. A., van der Mey A. G. L., Verbist B. M., Frijns J. H. M. (2019). Pediatric auditory brainstem implant users compared with
cochlear implant users with additional disabilities. Otol Neurotol, 40, 936–945.
Vesseur A., Free R., Snels C., Dekker F., Mylanus E., Verbist B., Frijns J. (2018). Hearing restoration in
cochlear nerve deficiency: The choice between
cochlear implant or auditory brainstem implant, a meta-analysis. Otol Neurotol, 39, 428–437.
Vincenti V., Ormitti F., Ventura E., Guida M., Piccinini A., Pasanisi E. (2014). Cochlear implantation in children with
cochlear nerve deficiency. Int J Pediatr Otorhinolaryngol, 78, 912–917.
Wu C. M., Lee L. A., Chen C. K., Chan K. C., Tsou Y. T., Ng S. H. (2015). Impact of
cochlear nerve deficiency determined using 3-dimensional magnetic resonance imaging on hearing outcome in children with cochlear implants. Otol Neurotol, 36, 14–21.
Xi X., Ching T. Y., Ji F., Zhao Y., Li J. N., Seymour J., Hong M. D., Chen A. T., Dillon H. (2012). Development of a corpus of Mandarin sentences in babble with homogeneity optimized via psychometric evaluation. Int J Audiol, 51, 399–404.
Yamazaki H., Leigh J., Briggs R., Naito Y. (2015). Usefulness of MRI and EABR Testing for Predicting CI
Outcomes Immediately After Cochlear Implantation in Cases With
Cochlear Nerve Deficiency. Otol Neurotol, 36, 977–984.
Young N. M., Kim F. M., Ryan M. E., Tournis E., Yaras S. (2012). Pediatric cochlear implantation of children with eighth nerve deficiency. Int J Pediatr Otorhinolaryngol, 76, 1442–1448.
Zheng Y., Soli S. D., Meng Z., Tao Y., Wang K., Xu K., Zheng H. (2010). Assessment of Mandarin-speaking pediatric
cochlear implant recipients with the mandarin early speech perception (MESP) test. Int J Pediatr Otorhinolaryngol, 74, 920–925.
