Abstract
Objectives
To evaluate the role of electrically evoked auditory brainstem responses (EABRs) in assessing auditory pathway integrity and predicting auditory and speech outcomes in patients undergoing cochlear reimplantation (CRI), especially those with cochlear ossification or complex anatomy.
Methods
A retrospective analysis was conducted of 40 CRI patients (12 ossification, 28 non-ossification) from 2019 to 2024. Intraoperative and postoperative EABR assessments were performed to evaluate neural responses and guide cochleostomy. Parameters included the elicitation rate, V-wave amplitude, threshold, and Gibson score. Auditory–speech performance was measured by the Categories of Auditory Performance and Speech Intelligibility Rating scales, and word-recognition scores. Regression analyses identified predictors of postoperative outcomes.
Results
The EABR elicitation rate was 100% in non-ossification cases and 80% in ossification cases. Gibson scores were higher, and V-wave thresholds lower, in non-ossification ears (p < 0.05). Latencies and amplitudes did not differ significantly. Auditory–speech outcomes were superior in non-ossification patients. Implantation age, cochlear ossification, and middle-electrode Gibson score significantly predicted postoperative performance; middle-electrode Gibson score correlated most strongly with the disyllabic word recognition rate (R2 = 0.406, p = 0.0007).
Conclusions
EABRs reliably assess auditory pathway integrity, inform surgical guidance, and predict postoperative auditory–speech outcomes in CRI, especially in ossified or anatomically complex cochleae.
Significance
The findings support the use of EABRs as a clinical adjunct to improve surgical precision and functional outcomes in challenging CRI cases.
Keywords: cochlear implantation, EABR, ossification, performance, reimplantation
1. Introduction
Cochlear implants (CIs) are the standard treatment for individuals with bilateral severe-to-profound sensorineural hearing loss (House, 1987; Mudry and Mills, 2013). Since the first implantation in 1961, CIs have been widely adopted worldwide. However, the limited lifespan of electronic devices makes them susceptible to various malfunctions that may necessitate reimplantation in both the short and long term. It has been estimated that children with hearing impairment may require two to three CI devices over their lifetime.
The first reported cochlear reimplantation (CRI) procedure was performed by Hochmair-Desoyer and Burian in 1985 in two patients using Vienna CIs (Hochmair-Desoyer and Burian, 1985). Reported reimplantation rates across studies are currently 0.5%–30%, depending on the patient population and the duration of follow-up (Asghari et al., 2024; Smetak et al., 2023). As the number of CI recipients continues to increase, the incidence of CRI—primarily driven by device failure, medical complications, and, to a lesser extent, technological upgrades—is also expected to rise steadily. The indications for reimplantation can be categorized as device-related (hard failure or soft failure) or medical, with device-related complications remaining the most common cause (Chen et al., 2024; Kim et al., 2023).
Overall, CRI is considered a safe procedure, and most studies have demonstrated that it can preserve or even improve auditory and speech performance. Nevertheless, approximately 15%–16% of children experience deterioration in outcomes after reimplantation (Blanc et al., 2022). Factors, such as cochleostomy-related trauma, foreign body reactions to the electrode array, and intracochlear inflammatory responses are associated with cochlear ossification and fibrosis following initial implantation (Danielian et al., 2021), potentially complicating electrode reinsertion. In addition, these newly formed tissues may disrupt the electrode–neural interface, thereby negatively affecting auditory and speech perception and potentially contributing to device malfunction and soft failure (Donati et al., 2023; Kamakura and Nadol, 2016).
In addition to intracochlear tissue formation, other proposed causes of postoperative performance deterioration include scala vestibuli insertion of the electrode array, initial diagnostic inaccuracies, soft tissue collapse within the scala tympani that impedes proper reinsertion, prolonged periods of device non-use before reimplantation, and poor adherence to postoperative speech rehabilitation programs (Blanc et al., 2022).
The electrically evoked auditory brainstem response (EABR) is a diagnostic modality analogous to acoustic auditory brainstem response testing, but it uses direct electrical stimulation of the cochlea to assess auditory brainstem function. EABRs can be elicited either extracochlearly, such as at the round window niche (Causon et al., 2019), or intracochlearly through a CI (Firszt et al., 2002). Because post-implantation outcomes may correlate with the number and functional integrity of surviving auditory nerve fibers, EABR thresholds have shown prognostic associations with performance in CI users, particularly within the low-to-mid frequency range (Chen et al., 2025). Therefore, EABRs may also have value in predicting postoperative outcomes following CRI.
Furthermore, the surgical technique used during the initial implantation and the intraoperative conditions encountered may introduce uncertainty during reimplantation, particularly in complex cases such as severe cochlear ossification. We, therefore, hypothesized that preoperative or intraoperative assessment of neural pathway integrity using EABRs could facilitate successful reimplantation and provide predictive information regarding postoperative outcomes.
In this study, we retrospectively analyzed the surgical, electrophysiological, and postoperative outcomes of 40 CRI cases treated at the China–Japan Friendship Hospital. We compared EABR findings with postoperative speech rehabilitation outcomes in patients with normal cochlear morphology and those with cochlear ossification or inner-ear deformities. The study had two primary objectives: to characterize EABR waveform patterns in patients with cochlear ossification or abnormal cochlear anatomy, and to identify electrophysiological parameters that may predict auditory and speech rehabilitation outcomes following CRI.
2. Materials and methods
2.1. Subjects
A retrospective study was conducted of 40 CRI cases with an EABR electrophysiological assessment that was performed at our center between January 2019 and September 2024. Among these, one case was identified from a cohort of 1,123 patients who underwent initial cochlear implantation at our center between March 2007 and September 2024. The remaining 39 cases underwent initial implantation at other domestic cochlear implant centers and were subsequently referred to our institution for reimplantation surgery. The experimental protocol and study design were reviewed and approved by the Ethics Committee of the China–Japan Friendship Hospital. Written informed consent was obtained from all participants or their legal guardians before study initiation.
2.2. Clinical data collection
The clinical data collected from the patients included the timing of the initial cochlear implantation, the timing of CRI, and the indications for reimplantation. Surgical records from the initial implantation, preoperative temporal bone computed tomography (CT) scans, and audiological evaluations performed at 3 months after CRI were reviewed. Manufacturer-specific device integrity test reports, as well as device programming and rehabilitation records, were also collected.
The indications for CRI were categorized as device failure or medical complications. Hard failure was defined as the complete absence of auditory stimulation from the CI. Cases with normal device integrity testing but persistent complaints, such as dizziness or declining auditory performance, were classified as soft failure. Medical complications included wound or flap-related problems, meningitis, implant extrusion, magnet migration, improper initial implantation, and facial-nerve stimulation.
2.3. Evaluation of cochlear ossification and fibrosis
Cochlear ossification can be confirmed by CT and magnetic resonance imaging (MRI). Because MRI was contraindicated or infeasible for patients with CRI, the diagnosis of cochlear ossification and fibrosis relied primarily on preoperative temporal bone CT and intraoperative exploration.
(1). Temporal bone CT findings: narrowing of the bony cochlear duct, irregularities of the inner wall of the cochlear duct, and bony hyperplasia around the cochleostomy site.
(2). Intraoperative observations: bony hyperplasia around the cochleostomy site, narrowing of the cochlear duct, resistance encountered during electrode extraction and reinsertion, and partial electrode extrusion surrounded by dense fibrous tissue causing resistance during extraction. In such cases, enlargement of the cochleostomy and removal of fibrous tissue within the basal turn of the cochlea were required to facilitate insertion of the new electrode array.
2.4. Assessment of speech and listening ability
Auditory and speech performance were evaluated at 3 months after CRI using Mandarin Speech Perception word lists, with outcomes quantified by monosyllabic and disyllabic word-recognition scores. Stimuli were presented in the sound field at 65 dBA through a single loudspeaker, with the subject seated 1 m from the loudspeaker and directly facing it.
In addition, standardized Chinese versions of the Categories of Auditory Performance (CAP) and Speech Intelligibility Rating (SIR) scales were administered to evaluate postoperative auditory behavioral responses and speech intelligibility. The CAP scale consists of nine hierarchical levels (0–9), whereas the SIR scale comprises five levels (1–5).
2.5. EABR procedures
2.5.1. EABR recording
Evoked auditory brainstem response evoked via the round window membrane: EABR recordings were obtained using Neuro-Audio NET1.0.103.3 (Neurosoft, Ivanovo, Russia). Needle electrodes were used as the recording, ground, and reference electrodes and were placed at the midline forehead, between the eyebrows, and the contralateral shoulder, respectively. Electrical stimulation was generated using an electromyography external electrical stimulator (Neurosoft, Ivanovo, Russia). A facial-nerve stimulation probe (Medtronic, Minneapolis, MN, USA) was used as the stimulating positive electrode. A stainless-steel needle electrode served as the reference electrode and was positioned over the ipsilateral external occipital protuberance. The initial stimulation intensity for EABR recording was set at 2.0 mA. Each stimulation intensity was tested twice, and the two waveforms were superimposed before proceeding to the next intensity level. The maximum stimulation intensity was 3.0 mA.
Postoperative EABRs: The recording electrodes and recording parameters used for postoperative EABRs were identical to those described above. Alternating biphasic pulses in monopolar mode (MP 1 + 2), with a pulse width of 50 μs (MED-EL devices [MED-EL GmbH, Innsbruck, Austria] set at 37 CU]; Nurotron devices [Nurotron Biotechnology Co., Ltd., Hangzhou, China] set at 50 CU) and an interphase gap of 5 μs, were used as stimuli. Stimuli were delivered at a rate of 30 Hz through the Nurotron Processor Interface and MED-EL, controlled by dedicated software (NuroSound 1.1.0; Nurotron Biotechnology). To explore the EABR characteristics from high- to low-frequency stimulation sites, EABRs were obtained by applying electric stimulation at one apical, one middle, and one basal electrode (i.e., electrodes 1, 11, and 22) representing low, middle, and high frequencies, respectively.
The EABR output signal was filtered online using a band-pass filter of 0.1–3 kHz and averaged over 512 sweeps at each stimulation level within a 15-ms recording window. Electrical stimulation consisted of alternating wave pulses with a duration of 100 μs delivered at a rate of 21 Hz. Electrode impedance was maintained at below 1 kΩ.
2.5.2. Surgical procedures
Following mastoidectomy and removal of neo-osteogenic and hyperplastic fibrotic tissue within the facial recess, posterior tympanotomy was performed to expose the original electrode array. To minimize interference from muscle activity during EABR recording, patients received cisatracurium at a dose of 0.5 mg/kg according to body weight. After removal of the original electrode array from the cochlea, the stimulation probe was positioned at the original cochlear fenestration site for EABR recording.
2.5.3. EABR data processing and analysis
All EABR analyses were independently performed by two experienced professionals, including an otolaryngologist and an audiologist.
EABR evoked via the round window membrane: Repeatable and stable EABR waveforms were evoked at the maximum stimulation level.
Postoperative EABRs: Routine parameters were evaluated according to the Gibson scoring system, in which a maximum score of 3 indicates the presence of clearly identifiable waves II, III, and IV–V, with a wave V amplitude exceeding 0.5 μV (Chen et al., 2024). The latencies of eIII and eV at 20 CL above threshold were recorded. Two threshold definitions were adopted. The EABR threshold was defined as the stimulation intensity immediately before the waveform amplitude decreased below 0.1 μV or before stable, reproducible, high-quality wave V responses could no longer be identified. The 0.5-μV threshold was defined as the stimulation intensity immediately before the waveform amplitude decreased below 0.5 μV. The response rate was calculated by dividing the number of patients with positive EABRs by the total number of subjects included in the study.
2.6. Statistical analysis
Statistical analyses were performed to evaluate differences in study variables between groups with and without cochlear ossification. Normally distributed continuous variables were analyzed using independent-samples t-tests, whereas non-normally distributed continuous variables were analyzed using the Wilcoxon signed-rank test. Categorical variables were compared using the chi-square test or Fisher’s exact test, as appropriate. Data normality was assessed using the Shapiro-Wilk test.
To investigate the relationship between EABR parameters and postoperative outcomes, univariate analyses were performed using CAP, SIR, and monosyllabic and disyllabic word-recognition scores as dependent variables, while EABR parameters and age at implantation were treated as independent variables. Age at implantation was identified as a significant factor in the univariate analysis. Therefore, multivariate analyses were subsequently conducted with age at implantation included as a covariate to assess the influence of EABR parameters on CAP, SIR, and monosyllabic and disyllabic word-recognition scores. Adjusted odds ratios and 95% confidence intervals were calculated to quantify associations.
All statistical analyses were performed using R software version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria). A p-value of <0.05 was considered statistically significant.
3. Results
In total, 12 patients were identified as having cochlear ossification based on preoperative CT imaging and intraoperative findings. Among these, one case (No. 37) was associated with a medical complication. This patient had developed localized skin infection and necrosis 3 years after the initial cochlear implantation, which eventually resulted in flap breakdown and device exposure. Following debridement and removal of the receiver–stimulator, the patient underwent CRI 3 months later. During surgery, ossification of the basal turn of the cochlea was observed, and complete electrode insertion was achieved after partial removal of hyperplastic bone and fibrous tissue. The remaining 11 cases underwent reimplantation because of device failure, including nine cases of soft failure and two cases of hard failure. Detailed clinical characteristics of these 12 patients are summarized in Figure 1 and Table 1.
FIGURE 1.
(a) Temporal computed tomography (CT) revealed cochlear ossification in the basal turn, consistent with the intraoperative finding of substantial ossification obstructing the original cochleotomy site during cochlear reimplantation (CRI) surgery (b). Evoked auditory brainstem responses (EABR) recordings obtained via the round window membrane demonstrated repeatable and distinct wave V responses (c), further confirming that the pore served as the appropriate insertion pathway. Postoperative EABR showed favorable waveforms in the middle electrodes (d). (e–k) Correspond to Patients 8 and 10, respectively. Both patients exhibited cochlear ossification on CT scans and flat EABR responses, with Waves III and V not being identified. (j,k) Correspond to Patient 7, in whom cochlear ossification was identified in the basal turn. EABR recorded via the round window membrane demonstrated unfavorable waveforms, and only partial implantation was achieved, accompanied by poor postoperative EABR responses.
TABLE 1.
Detailed clinical characteristics of 12 patients with cochlea ossification.
| Subject no. | Sex | Side | Mode | Etiology | CI type | Sound processor | CRI reason | Electrode insertion | Pathology | Radiology | ||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | M | R | B | NA | CS-10A | NSP-60B | 4.8 | 2.2 | Soft | Complete | – | – |
| 2 | M | R | B | NA | CS-10A (TM) | Voyager | 13.8 | 6.6 | Soft | Complete | Acute and chronic inflammatory cell infiltration within the cochlear fibrous tissue, accompanied by a multinucleated giant cell reaction | Basal cochlear ossification |
| 3 | F | R | M | NA | CS-10A (TM) | Enduro | 10.3 | 5.8 | Soft | Complete | NA | Basal cochlear ossification |
| 4 | M | R | M | NA | CS-20A (TM) | Voyager | 7.3 | 2.4 | Soft | Complete | NA | Basal cochlear ossification |
| 5 | F | L | M | NA | CS-10A (TM) | Enduro | – | – | Soft | Complete | Intracochlear tissue revealed scant fibrous connective tissue with focal calcifications, and foreign-body giant cell reactions | Basal cochlear ossification; large bony fragment |
| 6 | M | R | B | EYA1 chr8:72127661; CNC narrowing | CS-10A | Enduro | 9.3 | 1.3 | Soft | Partial/18 | Chronic inflammation of the fibrous connective tissue, accompanied by deposition of cholesterol crystals | Pericochlear bone hyperplasia |
| 7 | F | R | M | NA | CS-10A | Enduro | 8 | 4.2 | Soft | Partial/8 | – | – |
| 8 | M | R | M | NA | CS-20A | Voyager | 16.75 | 9.9 | Soft | Complete | – | – |
| 9 | M | L | B | Auditory neuropathy spectrum disorder | Helix adv.1j/Q30 | Harmony | 4 | 1 | Medical | Complete | Granulation tissue with multinucleated giant cells, fibrous encapsulation of the electrode, and chronic inflammatory infiltrates (lymphocytes, plasma cells, and eosinophils) | – |
| 10 | M | R | B | IP-III; POU3F4, CHM | CS-10A | NSP-60C | 8 | 6 | Soft | Partial/19 | – | – |
| 11 | M | L | M | NA | CS-10A | NSP-60C | 19 | 8.5 | Hard | Complete | – | – |
| 12 | M | R | B | LVAS, Mondini | CS-10A (TM) | Voyager | 14 | 5 | Hard | Complete | – | – |
The non-ossification group included 28 patients. Among them, seven cases were attributed to hard failure, 20 to soft failure, and 1 to a medical complication. This patient (No. 26) developed recurrent meningitis due to a meningeal defect 11 years after the initial implantation. The defect was surgically repaired, and CRI was successfully performed 3 months later without evidence of cochlear ossification. The EABR waveforms recorded from three electrode locations in this patient demonstrated good waveform differentiation.
Table 2 summarizes the key baseline characteristics of both groups, including age at reimplantation, duration of deafness, etiology, and electrode type. As shown in Table 2, there were no significant differences in age at reimplantation, duration of deafness, etiology, or electrode type between the ossification and non-ossification groups, which further supports the reliability and validity of our findings.
TABLE 2.
Baseline comparison between the non-ossification and ossification groups.
| Region/variable | Non-ossification group (n = 28) |
Ossification group> (n = 12) |
Total (N = 40) |
t/χ2/Z | P-value |
|---|---|---|---|---|---|
| Testing age, y | 185 | 0.342 | |||
| Median [Q1, Q3] | 13 [7, 16.60] | 9 [7, 14.00] | 11 [7, 16.20] | – | – |
| Implantation age/y | 167 | 0.706 | |||
| Median [Q1, Q3] | 5 [2, 9.15] | 5 [2, 6.60] | 5 [2, 8.30] | – | – |
| Sex | – | 0.725 | |||
| F | 9 (32.14%) | 3 (25.00%) | 12 (30.00%) | – | – |
| M | 19 (67.86%) | 9 (75.00%) | 28 (70.00%) | – | – |
| Implanted ear | 0.725 | ||||
| L | 9 (32.14%) | 3 (25.00%) | 12 (30.00%) | – | – |
| R | 19 (67.86%) | 9 (75.00%) | 28 (70.00%) | – | – |
| Mode | – | 0.728 | |||
| B | 11 (39.29%) | 6 (50.00%) | 17 (42.50%) | – | – |
| M | 17 (60.71%) | 6(50.00%) | 23(57.50%) | – | – |
| Etiology | – | 0.489 | |||
| Aminoglycoside antibiotics | 1 (3.57%) | 0 (0.00%) | 1 (2.50%) | – | – |
| EYA1 | 0 (0.00%) | 1 (8.33%) | 1 (2.50%) | – | – |
| GJB2 | 6 (21.43%) | 0 (0.00%) | 6 (15.00%) | – | – |
| IP-III, POU3F4 | 0 (0.00%) | 1 (8.33%) | 1 (2.50%) | – | – |
| LVAS | 3 (10.71%) | 1 (3.57%) | 4 (14.29%) | – | – |
| PTPN11 | 1 (3.57%) | 0 (0.00%) | 1 (2.50%) | – | – |
| Hypoplastic cochlear nerve | 0 (0.00%) | 1 (8.33%) | 1 (2.50%) | – | – |
| NA | 17 (60.71%) | 8 (66.67%) | 25 (62.50%) | – | – |
| Type of RCI | – | 0.454 | |||
| CS-10A | 11 (39.29%) | 5(41.67%) | 16 (40.00%) | – | – |
| CS-10A (TM) | 12 (42.86%) | 4 (33.33%) | 16 (40.00%) | – | – |
| CS-20A | 0 (0.00%) | 1 (8.33%) | 1 (2.50%) | – | – |
| CS-20A (TM) | 4 (14.29%) | 1 (8.33%) | 5 (12.50%) | – | – |
| Helix adv.1j/ Q30 | 0 (0.00%) | 1 (8.33%) | 1 (2.50%) | – | – |
| SONATA TI100+F28 | 1 (3.57%) | 0 (0.00%) | 1 (2.50%) | – | – |
| Type of sound processor | – | 0.805 | |||
| Enduro | 8 (28.57%) | 4 (33.33%) | 12 (30.00%) | – | – |
| NSP-60B | 4 (14.29%) | 1 (8.33%) | 5 (12.50%) | – | – |
| NSP-60C | 4 (14.29%) | 2 (16.67%) | 6 (15.00%) | – | – |
| Sonata/Opus 2 | 1 (3.57%) | 0 (0.00%) | 1 (2.50%) | – | – |
| Voyager | 11 (39.29%) | 4 (33.33%) | 15 (37.50%) | – | – |
| Advanced Bionics | 0 (0.00%) | 1 (8.33%) | 1 (2.50%) | – | – |
RCI auditory performance was available for 35 of 40 patients (87.5%). In the non-ossification group, the mean CRI CAP and SIR scores were 6.19 ± 1.5 and 3.77 ± 0.95, respectively, and the mean monosyllabic and disyllabic word-recognition rates were 71.2 ± 25.5% and 73.6 ± 24.8%, respectively. In the ossification group, the mean pre-failure CAP and SIR scores were 4.17 ± 1.34 and 2.25 ± 0.87, respectively, and the mean monosyllabic and disyllabic word-recognition rates were 38.0 ± 23.5% and 41.5 ± 24.7%, respectively. All patients with available data obtained a measurable auditory benefit from their initial implant prior to failure, confirming that the auditory pathway was functional before reimplantation in these cases.
3.1. Intraoperative EABR results
Intraoperative EABR recordings via the round window niche were successfully obtained in 11 patients during CRI surgery.
3.2. Comparison of postoperative EABR parameters and auditory–speech performance between the two groups
In the non-ossification group, 27 of 28 patients demonstrated well-differentiated EABR waveforms, yielding an elicitation rate of 100%; one patient (No. 28) did not undergo EABR testing. In the ossification group, two patients did not undergo EABR testing, and one patient had EABR recordings obtained only from the apical electrode. Among the remaining patients, two failed to demonstrate reproducible EABR waveforms across all three electrode positions, resulting in an elicitation rate of 80% (n = 8/10).
Gibson scores recorded from the apical, middle, and basal electrodes were significantly higher in the non-ossification group than in the ossification group (apical: p = 0.002; middle: p = 0.002; basal: p = 0.0001) (Figure 2). Analysis of Wave III and Wave V latencies revealed no significant differences between the two groups across all three electrode positions. However, V-wave thresholds, including both the 0.5-μV threshold (Figures 3a, b) and the disappearance threshold (Figures 3c, d), were significantly lower at the middle and basal electrodes in the non-ossification group compared with the ossification group (p-values are presented in Table 3). No significant differences in V-wave amplitudes were observed between the two groups.
FIGURE 2.
Gibson scores recorded from the apical, middle, and basal electrodes were significantly higher in the non-ossification group than in the ossification group.
FIGURE 3.
V-wave thresholds, including both the 0.5-μV threshold (a,b) and the disappearance threshold (c,d), were significantly lower at the middle and basal electrodes in the non-ossification group compared with the ossification group. *p < 0.05, **p < 0.01.
TABLE 3.
Comparison of evoked auditory brainstem response (EABR) measures and speech and auditory performance between the non-ossification and ossification groups.
| EABR measures | P-value | Statistical value |
|---|---|---|
| Gibson–apical | 0.002298 | N/A |
| Gibson–middle | 0.001532 | N/A |
| Gibson–basal | 0.0001436 | N/A |
| Apical-V latency/ms | 0.8269 | t = 0.2253 |
| Middle-V latency/ms | 0.6251 | t = 0.5085 |
| Basal-V latency/ms | 0.2255 | t = −1.3912 |
| Apical-III latency/ms | 0.1736 | Z = 1.382, U = 57.5 |
| Middle-III latency/ms | 0.8316 | Z = 0.234, U = 84.5 |
| Basal-III latency/ms | 0.3769 | Z = 0.91, U = 43 |
| V threshold 0.5–apical/nC | 0.07189 | Z = −1.816, U = 207.5 |
| V threshold-0.5–middle/nC | 0.02899 | Z = −2.2, U = 200.5 |
| V threshold-0.5–basal/nC | 0.03364 | Z = −2.141, U = 198 |
| V threshold-apical/nC | 0.07779 | Z = −1.779, U = 206.5 |
| V threshold-middle/nC | 0.01236 | Z = −2.518, U = 210 |
| V threshold-basal/nC | 0.00836 | Z = −2.654, U = 214 |
| V amplitude–apical | 0.1977 | Z = 1.237, U = 68 |
| V amplitude–middle | 0.2237 | Z = 1.084, U = 60 |
| V amplitude–basal | 0.2883 | Z = 3.77, U = 45.5 |
| CAP | 0.0001735 | Z = 3.921, U = 42 |
| SIR | 9.393e-05 | Z = 3.603, U = 46 |
| MONO | 0.000333 | Z = 3.638, U = 45 |
| DSP | 0.0002907 | Z = −1.816, U = 207.5 |
Regarding auditory–speech performance, the ossification group demonstrated significantly poorer outcomes in CAP and SIR scores (Figures 4a, b), as well as lower monosyllabic and disyllabic word-recognition rates (Figures 4c, d), compared with the non-ossification group.
FIGURE 4.
The ossification group demonstrated significantly poorer Categories of Auditory Performance (CAP) and Speech Intelligibility Rating (SIR) scores (a,b), as well as lower monosyllabic and disyllabic word-recognition rates (c,d), compared with the non-ossification group. ***p < 0.001.
3.3. Factors influencing postoperative auditory–speech performance following CRI
Univariate regression analysis demonstrated that implantation age; cochlear ossification; Gibson scores recorded at the apical, middle, and basal electrodes; and V-wave thresholds were significantly associated with auditory–speech performance outcomes, including CAP, SIR, and monosyllabic and disyllabic word-recognition scores. Detailed R2 and p-values are presented in Table 4. Among these variables, the Gibson score recorded at the middle electrode showed the strongest correlation with the disyllabic word-recognition rate (R2 = 0.4057, p = 0.0007).
TABLE 4.
Univariate analysis of the correlation between evoked auditory brainstem response (EABR) parameters and auditory–speech performance.
| CAP | SIR | MONO | DSP | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Electrode/parameter | R2 | F statistic | P-value | R2 | F statistic | P-value | R2 | F statistic | P-value | R2 | F statistic | P-value |
| Implantation age | 0.0994 | 4.1942 | 0.04752 | 0.113 | 4.8426 | 0.03392 | 0.1213 | 5.1082 | 0.02979 | 0.1449 | 6.2692 | 0.01682 |
| Age difference | 0.01377 | 0.5304 | 0.4709 | 0.006048 | 0.2312 | 0.6334 | 0.01992 | 0.7521 | 0.3914 | 0.02893 | 1.1022 | 0.3006 |
| Apical III latency/ms | 0.04277 | 1.4296 | 0.2406 | 0.03244 | 1.0728 | 0.3081 | 0.03007 | 0.9609 | 0.3345 | 0.02962 | 0.9461 | 0.3382 |
| Middle III latency/ms | 0.09113 | 3.2087 | 0.08271 | 0.08183 | 2.8521 | 0.101 | 0.06647 | 2.2074 | 0.1475 | 0.07206 | 2.4075 | 0.1309 |
| Basal III latency/ms | 0.01691 | 0.5161 | 0.4781 | 0.02289 | 0.7028 | 0.4085 | 0.02087 | 0.618 | 0.4382 | 0.02629 | 0.783 | 0.3835 |
| Apical V latency/ms | 0.09738 | 3.5603 | 0.068 | 0.06582 | 2.325 | 0.1368 | 0.09901 | 3.5165 | 0.06991 | 0.0934 | 3.2969 | 0.07879 |
| Middle V latency/ms | 0.1096 | 4.0635 | 0.05202 | 0.05753 | 2.0144 | 0.1652 | 0.08674 | 3.0393 | 0.09087 | 0.08525 | 2.9823 | 0.09382 |
| Basal V latency/ms | 0.0465 | 1.463 | 0.2359 | 0.01226 | 0.3723 | 0.5463 | 0.01922 | 0.5684 | 0.457 | 0.01762 | 0.5201 | 0.4766 |
| V threshold 0.5-apical/nC | 0.2073 | 9.4161 | 0.004072 | 0.1807 | 7.9418 | 0.007799 | 0.2501 | 11.6746 | 0.001621 | 0.2327 | 10.6147 | 0.002498 |
| V threshold 0.5–middle/nC | 0.208 | 9.1934 | 0.004552 | 0.162 | 6.7662 | 0.01352 | 0.2548 | 11.6265 | 0.001691 | 0.2344 | 10.4069 | 0.002775 |
| V threshold 0.5–basal/nC | 0.1543 | 6.3869 | 0.01617 | 0.1103 | 4.3399 | 0.0446 | 0.08691 | 3.2364 | 0.0809 | 0.09626 | 3.6215 | 0.06553 |
| V threshold–apical/nC | 0.2228 | 10.3209 | 0.002771 | 0.2034 | 9.1948 | 0.004481 | 0.2694 | 12.9036 | 0.0009972 | 0.2547 | 11.9619 | 0.001445 |
| V Threshold–Middle/nC | 0.2243 | 10.122 | 0.003067 | 0.2207 | 9.9096 | 0.003353 | 0.2702 | 12.5883 | 0.001157 | 0.2532 | 11.5268 | 0.00176 |
| V threshold–basal/nC | 0.2932 | 14.5215 | 0.0005378 | 0.2126 | 9.4512 | 0.004075 | 0.3497 | 18.2837 | 0.0001458 | 0.3041 | 14.8611 | 0.0004897 |
| V amplitude–apical | 0.00004989 | 0.0016 | 0.9684 | 0.001806 | 0.0579 | 0.8114 | 0.001455 | 0.0452 | 0.8331 | 0.002955 | 0.0919 | 0.7638 |
| V amplitude–middle | 0.0004441 | 0.0138 | 0.9073 | 0.009331 | 0.292 | 0.5928 | 0.01078 | 0.3268 | 0.5718 | 0.01142 | 0.3467 | 0.5604 |
| V amplitude–basal | 0.002991 | 0.09 | 0.7662 | 0.004422 | 0.1332 | 0.7177 | 0.01259 | 0.3698 | 0.5479 | 0.01001 | 0.2933 | 0.5923 |
| Gibson-apical | 0.2805 | 6.8234 | 0.003146 | 0.2359 | 5.4041 | 0.009009 | 0.3205 | 8.0177 | 0.001404 | 0.3283 | 8.3108 | 0.001152 |
| Gibson–middle | 0.2989 | 4.69 | 0.007772 | 0.2609 | 3.8832 | 0.01755 | 0.3956 | 6.9829 | 0.0009599 | 0.4057 | 7.2805 | 0.0007426 |
| Gibson–basal | 0.2902 | 6.9518 | 0.002944 | 0.2448 | 5.5112 | 0.00845 | 0.3896 | 10.5333 | 0.0002898 | 0.3712 | 9.7422 | 0.0004731 |
| Cochlear ossification | 0.3174 | 18.1365 | 0.0001254 | 0.3805 | 23.9574 | 0.00001745 | 0.3049 | 16.671 | 0.0002206 | 0.3032 | 16.5345 | 0.0002319 |
CAP, Categories of Auditory Performance; SIR, Speech Intelligibility Rating; MONO, monosyllabic word recognition; DSP, disyllabic word recognition.
In the regression model adjusted for implantation age and cochlear ossification, several EABR parameters remained significantly associated with auditory–speech outcomes. Specifically, Wave III latency recorded at the apical and middle electrodes, as well as Wave V latency at the apical electrode, demonstrated significant negative correlations with CAP scores. In addition, Wave III latency at the apical and middle electrodes was significantly negatively correlated with SIR scores, whereas the Gibson score at the basal electrode showed a significant positive correlation with monosyllabic and disyllabic word-recognition rates. These findings are summarized in Table 5.
TABLE 5.
Multivariate regression analysis of the correlation between EABR parameters and auditory–speech performance (covariates: implantation age + presence of ossification).
| CAP | SIR | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Outcome measure | R2 | F statistic | P- value | β | T-value | β_P-value | R2 | F statistic | P-value | β | T-value | β_P-value | |
| Apical III latency/ms | 0.4995 | 9.9795 | 0.0001011 | −1.18432407 | −2.1583 | 0.039035 | 0.5977 | 14.8576 | 0.000004134 | −0.78731745 | −2.3529 | 0.0253764 | |
| Middle III latency/ms | 0.5001 | 10.0051 | 0.00009921 | −1.48421797 | −2.2965 | 0.028804 | 0.5913 | 14.4674 | 0.000005216 | −0.93814894 | −2.4361 | 0.0209961 | |
| Basal III latency/ms | 0.4115 | 6.5258 | 0.001738 | −0.73377312 | −1.2052 | 0.238208 | 0.5415 | 11.0229 | 0.00005946 | −0.61125262 | −1.6859 | 0.102935 | |
| Apical–V latency/ms | 0.5312 | 11.7068 | 0.00002708 | −0.88517617 | −2.0998 | 0.0439789 | 0.5938 | 15.105 | 0.000003084 | −0.45689601 | −1.7485 | 0.0902767 | |
| Middle–V latency/ms | 0.5099 | 10.7498 | 0.00005291 | −0.77146085 | −1.7727 | 0.0861087 | 0.5573 | 13.0074 | 0.00001138 | −0.27262809 | −1.0173 | 0.316863 | |
| Basal–V latency/ms | 0.4148 | 6.6153 | 0.001612 | −0.584631 | −1.2723 | 0.213746 | 0.5044 | 9.4998 | 0.0001712 | −0.20856143 | −0.731 | 0.470851 | |
| V threshold 0.5–apical/nC | 0.4828 | 10.5808 | 0.00004604 | −0.00864735 | −0.7105 | 0.482216 | 0.5394 | 13.2724 | 0.000006753 | −0.00064656 | −0.0843 | 0.933332 | |
| V threshold 0.5–middle/nC | 0.4716 | 9.8157 | 0.00008922 | −0.00837469 | −0.6729 | 0.505715 | 0.522 | 12.0146 | 0.00001781 | 0.00123529 | 0.1573 | 0.87598 | |
| V threshold 0.5–basal/nC | 0.4767 | 10.0201 | 0.0000763 | −0.00801833 | −0.8839 | 0.383151 | 0.5221 | 12.0186 | 0.00001776 | −0.0010035 | −0.1745 | 0.86257 | |
| V threshold–apical/nC | 0.5021 | 11.427 | 0.00002461 | −0.01761493 | −1.3554 | 0.184212 | 0.5509 | 13.903 | 0.000004434 | −0.00772767 | −0.9374 | 0.355186 | |
| V threshold–middle/nC | 0.4724 | 9.8472 | 0.00008709 | -0.01169985 | −0.7095 | 0.483 | 0.5242 | 12.1212 | 0.00001653 | −0.00437916 | −0.4215 | 0.676131 | |
| V Threshold-Basal/nC | 0.4836 | 10.3009 | 0.00006169 | −0.012794 | −1.1102 | 0.274945 | 0.5217 | 11.9989 | 0.000018 | 0.00033711 | 0.0458 | 0.963738 | |
| V amplitude-apical | 0.477 | 9.1197 | 0.0001916 | −0.35143753 | −1.2359 | 0.226084 | 0.5766 | 13.6164 | 0.00000877 | −0.31775199 | −1.8554 | 0.0733836 | |
| V amplitude-middle | 0.4293 | 7.2705 | 0.0008839 | −0.13234147 | −0.478 | 0.636208 | 0.5196 | 10.4563 | 0.00007895 | −0.20219433 | −1.2021 | 0.239055 | |
| V amplitude-basal | 0.3954 | 6.1044 | 0.002492 | −0.35193748 | −0.8102 | 0.424686 | 0.4791 | 8.5854 | 0.000336 | −0.26975604 | −0.9866 | 0.332298 | |
| Gibson-apical | 0.5183 | 8.8752 | 0.00005579 | 0.12083285 | 0.0792 | 0.937327 | 0.5558 | 10.3213 | 0.00001559 | 0.41617914 | 0.4254 | 0.673286 | |
| Gibson–middle | 0.5093 | 6.4339 | 0.0003299 | −0.15248494 | −0.0917 | 0.927523 | 0.5378 | 7.2142 | 0.0001401 | 0.39929605 | 0.3729 | 0.711745 | |
| Gibson–basal | 0.4826 | 7.4621 | 0.00023 | 0.88835299 | 0.9236 | 0.362631 | 0.5228 | 8.7638 | 0.00006776 | 0.16610318 | 0.271 | 0.78814 | |
| MONO | DSP | ||||||||||||
| R2 | F statistic | P-value | β | T-value | β_P-value | R2 | F statistic | P-value | β | T-value | β_P-value | ||
| 0.4551 | 8.0744 | 0.0004628 | −0.15061684 | −1.6442 | 0.110944 | 0.5339 | 11.074 | 0.00005155 | −0.14513666 | −1.7372 | 0.0929644 | ||
| 0.3946 | 6.3014 | 0.002003 | −0.19009314 | −1.7113 | 0.0976944 | 0.4607 | 8.2577 | 0.0004008 | −0.19000723 | −1.8263 | 0.0781182 | ||
| 0.3166 | 4.1686 | 0.01505 | −0.09793996 | −1.0449 | 0.305353 | 0.4302 | 6.7957 | 0.001469 | −0.10908196 | −1.2822 | 0.210676 | ||
| 0.4942 | 9.772 | 0.0001177 | −0.12416034 | −1.6442 | 0.110565 | 0.5601 | 12.7307 | 0.00001536 | −0.11049361 | −1.5868 | 0.123036 | ||
| 0.4185 | 7.1956 | 0.0008881 | −0.09424085 | −1.228 | 0.22899 | 0.4745 | 9.0279 | 0.0002054 | −0.08444607 | −1.1644 | 0.25344 | ||
| 0.2921 | 3.7139 | 0.02341 | −0.02599792 | −0.3489 | 0.729873 | 0.3965 | 5.9141 | 0.00308 | −0.0145403 | −0.2126 | 0.833225 | ||
| 0.4923 | 10.6652 | 0.00004701 | −0.0022469 | −1.1627 | 0.253294 | 0.5223 | 12.028 | 0.00001764 | −0.00179457 | −0.956 | 0.34601 | ||
| 0.4525 | 8.8144 | 0.0002097 | −0.00250596 | −1.2878 | 0.207064 | 0.4815 | 9.9073 | 0.00009002 | −0.00197356 | −1.0403 | 0.306001 | ||
| 0.4247 | 7.8735 | 0.0004503 | −0.00026652 | −0.181 | 0.857483 | 0.4652 | 9.2766 | 0.0001459 | −0.0003712 | −0.261 | 0.795747 | ||
| 0.5208 | 11.9568 | 0.00001854 | −0.00377314 | −1.8439 | 0.0741952 | 0.5493 | 13.407 | 0.000006912 | −0.00341054 | −1.7161 | 0.0955225 | ||
| 0.453 | 8.8349 | 0.0002063 | −0.00335091 | −1.3015 | 0.202385 | 0.4822 | 9.9325 | 0.00008833 | −0.00265945 | −1.0596 | 0.297254 | ||
| 0.4773 | 9.7391 | 0.0001023 | −0.00320923 | −1.8044 | 0.0805828 | 0.4896 | 10.2306 | 0.00007066 | −0.00222871 | −1.2658 | 0.214736 | ||
| 0.4503 | 7.9177 | 0.0005239 | −0.03838697 | −0.8036 | 0.428192 | 0.5136 | 10.2057 | 0.0000942 | −0.03287671 | −0.7456 | 0.461887 | ||
| 0.3943 | 6.0766 | 0.002552 | −0.03933264 | −0.8992 | 0.376191 | 0.4677 | 8.1999 | 0.0004508 | −0.03766666 | −0.9397 | 0.355398 | ||
| 0.3552 | 4.9578 | 0.007193 | −0.06706934 | −0.9693 | 0.341007 | 0.4368 | 6.9815 | 0.001262 | −0.06211509 | −0.988 | 0.331912 | ||
| 0.5342 | 9.1758 | 0.00004684 | 0.04487951 | 0.1867 | 0.853049 | 0.5726 | 10.719 | 0.00001259 | 0.05099904 | 0.2212 | 0.826336 | ||
| 0.5473 | 7.2526 | 0.0001482 | −0.06196578 | −0.2528 | 0.802138 | 0.5915 | 8.6878 | 0.00003517 | −0.06673182 | −0.2861 | 0.776779 | ||
| 0.5084 | 8.0152 | 0.0001473 | 0.33190436 | 2.306 | 0.0279646 | 0.5336 | 8.8668 | 0.00006805 | 0.30016007 | 2.137 | 0.040594 | ||
CAP, Categories of Auditory Performance; SIR, Speech Intelligibility Rating; MONO, monosyllabic word recognition; DSP, disyllabic word recognition.
4. Discussion
In this study, we used EABRs to evaluate auditory brainstem pathway function in patients undergoing CRI. In patients with a normal cochlear anatomy, EABRs typically demonstrated a well-defined Wave V with normal latency, reflecting synchronous neural conduction through the auditory brainstem pathways, particularly the lateral lemniscus and inferior colliculus. In contrast, patients with significant cochlear ossification exhibited lower EABR elicitation rates, poorer waveform differentiation reflected by lower Gibson scores, and higher V-wave thresholds than patients without ossification. These findings may be explained by the fact that cochlear ossification impedes smooth insertion of the electrode array, reduces the number of effectively stimulated neural populations, and disrupts neural synchrony. Consequently, higher stimulation levels are required, resulting in broader or delayed neural activation caused by weak and desynchronized input to the brainstem. Furthermore, our findings demonstrate that pre-implantation EABRs recorded via the round window niche serve as a valuable localization tool during CRI surgery, particularly in cases with abnormal cochlear anatomy or ossification. However, no significant differences were observed in EABR latencies or V-wave amplitudes between the two groups. Cochlear ossification had a significant negative impact on postoperative auditory–speech performance. EABR thresholds and Gibson scores were significantly associated with postoperative outcomes, while latency and waveform characteristics also provided prognostic information. Collectively, our electrophysiological findings imply that EABRs, as an objective assessment of auditory brainstem pathway integrity, can facilitate successful execution of complex CRI procedures, support minimally invasive surgical approaches, and provide valuable prognostic information. These findings demonstrate the potential role of EABRs as an important adjunct for optimizing surgical planning and improving outcomes in challenging CRI cases.
Fibrous tissue formation and neo-osteogenesis have been documented in histopathological studies of temporal bone specimens (Kamakura and Nadol, 2016). Kamakura and Nadol (2016) reported that postoperative CNC word scores were negatively correlated with the volume of newly formed bone within the scala tympani, scala media/vestibuli, and the cochlea. In our ossification group, patients demonstrated substantial ossification obstructing the basal turn insertion pathway, which likely altered electrical stimulation patterns around the electrode array. This observation is consistent with our findings of elevated EABR thresholds and poorer Gibson scores in the ossification group. Ossified tissue and fibrous bands may reduce the efficiency of the electrode–neural interface, leading to increased auditory thresholds and reduced dynamic ranges. Among the EABR parameters evaluated, the Gibson score demonstrated the strongest predictive value, particularly for mid- to high-frequency auditory regions. Although ossification significantly affected waveform elicitation, EABR latencies—which may reflect neural plasticity and maturation—remained relatively stable in CRI patients. Consequently, the Gibson score, which reflects the overall differentiation of Waves III and V, appeared to be the most suitable parameter for predicting postoperative outcomes following CRI.
In recent years, clinical attention to EABRs has declined, largely because of the widespread adoption of electrically evoked compound action potentials (ECAPs), which provide rapid, convenient, and universally applicable testing. ECAPs can provide comprehensive electrode-specific data for predicting postoperative outcomes and determining T- and C-levels during dynamic range mapping (Antony et al., 2025), gradually overshadowing the advantages of EABRs in primary cochlear implantation. However, in CRI surgery, both our findings and reports from other centers indicate that complex conditions such as cochlear fibrosis, ossification, and hyperplastic tissue formation are frequently encountered (Donati et al., 2023, Kamakura and Nadol, 2016). CRI is often performed after medical complications, including meningitis, flap necrosis, infection, or severe pre-existing cochlear malformations. Bacterial infection, foreign-body reactions to electrode materials, bone dust, or blood contamination may promote neo-ossification, resulting in insertion difficulties and uncertain electrode positioning (Kamakura and Nadol, 2016). Under these conditions, conventional postoperative ECAPs cannot provide pre-implantation localization or prediction of neural responsiveness because of the disruption of the normal cochlear architecture. In our study, Patient No. 6 in the ossification group presented with severe ossification of the basal turn, making identification of the cochlear insertion pathway difficult after bone removal. EABR testing was used to confirm the appropriate insertion site, allowing partial electrode insertion (18/24 electrodes). This case illustrates that pre-implantation EABRs enable assessment of neural responsiveness directly at the drilling site and may provide important guidance for electrode reimplantation.
Previous studies have demonstrated that most patients undergoing CRI achieve postoperative outcomes comparable to or better than those obtained after the initial implantation (Gawecki et al., 2016). In most cases, CRI permits successful electrode reinsertion, and residual hearing may even be preserved without additional cochlear trauma (Ikeya et al., 2013). However, perielectrode ossification can complicate CRI surgery by impairing complete electrode insertion and optimal positioning, thereby adversely affecting postoperative outcomes (Terry et al., 2015). In the present study, the ossification group demonstrated significantly poorer auditory–speech performance than the non-ossification group (CAP: p = 0.00017; SIR: p < 0.0001; monosyllabic recognition: p = 0.00033; disyllabic recognition: p = 0.00029), indicating that cochlear ossification/fibrosis and the ability to achieve complete electrode insertion are critical determinants of CI outcomes. Importantly, Gibson scores and EABR V-wave thresholds demonstrated the strongest correlations with postoperative outcomes. The Gibson score is used to evaluate the quality and configuration of EABR waveforms, particularly Waves III and V, with higher scores indicating more robust and synchronized neural responses along the auditory pathway. In addition, the input/output slope reflects neural recruitment and synchrony (Chen et al., 2024). The Gibson score recorded at the middle electrode demonstrated the strongest correlation with the disyllabic word-recognition rate (R2 = 0.4057, p = 0.0007). Furthermore, in multivariate models adjusted for implantation age and cochlear ossification, Wave III latencies at the apical and middle electrodes showed significant negative correlations with CAP and SIR scores, whereas the Gibson score at the basal electrode demonstrated significant positive correlations with monosyllabic and disyllabic word-recognition rates (Table 5).
Because the human cochlea is tonotopically organized, the middle electrode region generally corresponds to frequencies between 1,000 and 2,000 Hz, which are critically important for encoding disyllabic speech sounds. In contrast, the basal turn encodes higher frequencies (approximately 2,000–8,000 Hz), which contribute disproportionately to several aspects of speech perception, including consonant identification, discrimination of voiceless fricatives and affricates, and temporal envelope processing. Therefore, favorable basal-turn electrode electrophysiological characteristics may indicate superior monosyllabic and disyllabic word-recognition performance, particularly under noisy listening conditions. Meanwhile, the middle electrode region may represent a “transition zone” between relatively preserved apical neural structures and the more vulnerable basal turn. A high Gibson score at the middle electrode may therefore reflect better preservation of neural synchrony and intact III–V interpeak intervals, indicating preserved temporal processing integrity. These mechanisms may explain the strong correlation between middle-array electrode parameters and postoperative disyllabic word-recognition performance. Our findings are consistent with those of Gallego et al., who reported that EABR Wave V latency and the III–V interpeak interval were associated with speech-recognition outcomes. Collectively, these findings imply that intraoperative and postoperative electrophysiological assessments are valuable for early prognostic evaluation and optimization of postoperative patient management.
Histological studies have demonstrated that round window insertion causes less inner-ear trauma and results in lower degrees of fibrosis and ossification than does cochleostomy (Bi et al., 2021). Minimally invasive surgical techniques, including low-speed cochleostomy, prevention of bone dust and blood contamination within the cochlea, and slow electrode insertion to avoid injury to the osseous spiral lamina and basilar membrane, may reduce cochlear ossification and fibrosis while preserving residual hearing. Somdas et al. (2007) suggested that cochlear ossification and fibrosis may extend from the apex toward the base, with the most severe ossification typically occurring in the basal turn. Similar findings were observed in our study, in which two cases of severe ossification demonstrated marked basal turn involvement. In Patient No. 6, partial electrode insertion was achieved by opening the basal turn and removing hyperplastic bone, thereby preserving a degree of auditory and speech function. Previous studies have also indicated that cochlear ossification and fibrosis may contribute to spiral ganglion cell loss, potentially affecting long-term CI outcomes.
5. Limitations
Several limitations of this study should be acknowledged. First, the relatively small sample size, particularly in the ossification group (n = 12), limits statistical power and precludes more robust subgroup analyses. Second, the 3-month follow-up period is relatively short; longer-term studies (6–12 months) are needed to confirm the stability of these outcomes. Third, missing EABR data for several patients, particularly in the ossification group, may have introduced selection bias. Fourth, the prognostic associations we observed are correlational; external validation in an independent cohort is necessary before these findings can be considered definitive. Future multicenter, prospective studies with larger sample sizes are needed to confirm and extend our results.
In conclusion, this study highlights the utility of EABR as an objective method for assessing auditory brainstem pathway integrity during CRI, particularly in patients with cochlear ossification or anatomical abnormalities. EABR parameters, including V-wave thresholds, Gibson scores, and waveform latencies, provide valuable prognostic information regarding postoperative auditory–speech outcomes. These findings support the integration of EABRs into surgical planning and intraoperative assessment in complex CRI cases to optimize electrode placement and improve patient prognosis.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the National High Level Hospital Clinical Research Funding (grant no. 2023-NHLHCRF-PY-02).
Footnotes
Edited by: Paul Yoo, University of Toronto, Canada
Reviewed by: Anandhan E. Dhanasingh, MED-EL, Austria
Lana Shahabaddin, Hawler Medical University, Iraq
Gislaine Wiemes, Self-Employed, Brazil
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.
Ethics statement
The studies involving humans were approved by the China-Japan Friendship Hospital Ethical Committee. The Ethics Committee of the China-Japan Friendship Hospital approved the study protocol (approval no. 2023-KY-217-1). Informed consent was obtained from all individual participants involved in the study. The studies were conducted in accordance with local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants’ legal guardians/next of kin.
Author contributions
YJ: Writing – original draft. QB: Writing – review & editing, Conceptualization. YLv: Methodology, Writing – review & editing. ZC: Data curation, Writing – review & editing, Conceptualization. YLiu: Validation, Writing – review & editing. JL: Writing – review & editing, Methodology. YLi: Writing – review & editing, Supervision. WY: Visualization, Writing – review & editing, Validation. XX: Resources, Writing – review & editing, Project administration. YL: Writing – review & editing, Writing – original draft. ZW: Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
- Antony P., Saravanan P., Puttapasappa M., Chinnaraj G., Megha. (2025). Predicting most comfortable listening levels of ESRT from ECAP thresholds in paediatric cochlear implant users. Int. J. Pediatr. Otorhinolaryngol. 188:112194. 10.1016/j.ijporl.2024.112194 [DOI] [PubMed] [Google Scholar]
- Asghari A., Farhadi M., Daneshi A., Hashemi S., Mirsalehi M., Saki N., et al. (2024). Cochlear reimplantation rate, causes, and outcomes: A multicenter study. Cochlear Implants Int. 25 477–486. 10.1080/14670100.2024.2448905 [DOI] [PubMed] [Google Scholar]
- Bi Q., Chen Z., Lv Y., Luo J., Wang N., Li Y. (2021). Management of delayed-onset skin flap complications after pediatric cochlear implantation. Eur. Arch. Otorhinolaryngol. 278 2753–2761. 10.1007/s00405-020-06348-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Blanc F., Blanchet C., Sicard M., Merklen F., Venail F., Mondain M. (2022). Audiological outcomes and associated factors after pediatric cochlear reimplantation. J. Clin. Med. 11:3148. 10.3390/jcm11113148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Causon A., O’Driscoll M., Stapleton E., Lloyd S., Freeman S., Munro K. (2019). Extracochlear stimulation of electrically evoked auditory brainstem responses (eABRs) remains the preferred pre-implant auditory nerve function test in an assessor-blinded comparison. Otol. Neurotol. 40 47–55. 10.1097/MAO.0000000000002055 [DOI] [PubMed] [Google Scholar]
- Chen Z., Bi Q., Lv Y., Li Y., Yang W., Xu X., et al. (2024). Cochlear reimplantation outcomes over 20 years: Expertise in reimplantation surgery and auditory-speech rehabilitation. Am. J. Otolaryngol. 45:104400. 10.1016/j.amjoto.2024.104400 [DOI] [PubMed] [Google Scholar]
- Chen Z., Bi Q., Lv Y., Liu Y., Tian Y., Liu J., et al. (2025). Exploring the relationship between electrophysiological measures of the electrically evoked auditory brainstem response and speech perception outcomes post-cochlear implantation. Eur. Arch. Otorhinolaryngol. 282 3005–3017. 10.1007/s00405-024-09185-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Danielian A., Ishiyama G., Lopez I., Ishiyama A. (2021). Predictors of fibrotic and bone tissue formation with 3-D reconstructions of post-implantation human temporal bones. Otol. Neurotol. 42 e942–e948. 10.1097/MAO.0000000000003106 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Donati G., Nassif N., Redaelli de Zinis L. O. (2023). Osteoneogenesis at the round window: A possible cause of cochlear implant failure? Audiol. Res. 14 1–8. 10.3390/audiolres14010001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Firszt J., Chambers R., Kraus, Reeder R. M. (2002). Neurophysiology of cochlear implant users I: Effects of stimulus current level and electrode site on the electrical ABR, MLR, and N1-P2 response. Ear Hear. 23 502–515. 10.1097/00003446-200212000-00002 [DOI] [PubMed] [Google Scholar]
- Gawecki W., Karlik M., Borucki L., Szyfter-Harris J., Wrobel M. (2016). Skin flap complications after cochlear implantations. Eur. Arch. Otorhinolaryngol. 273 4175–4183. 10.1007/s00405-016-4107-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hochmair-Desoyer I., Burian K. (1985). Reimplantation of a molded scala tympani electrode: Impact on psychophysical and speech discrimination abilities. Ann. Otol. Rhinol. Laryngol. 94 65–70. 10.1177/000348948509400114 [DOI] [PubMed] [Google Scholar]
- House L. (1987). Cochlear implant: The beginning. Laryngoscope 97 996–997. [PubMed] [Google Scholar]
- Ikeya J., Kawano A., Nishiyama N., Kawaguchi S., Hagiwara A., Suzuki M. (2013). Long-term complications after cochlear implantation. Auris Nasus Larynx 40 525–529. 10.1016/j.anl.2013.04.012 [DOI] [PubMed] [Google Scholar]
- Kamakura T., Nadol J., Jr. (2016). Correlation between word recognition score and intracochlear new bone and fibrous tissue after cochlear implantation in the human. Hear. Res. 339 132–141. 10.1016/j.heares.2016.06.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kim J., Choi Y., Kang W., Park H., Ahn J., Chung J. (2023). The experience of device failure after cochlear implantation. J. Otolaryngol. Head Neck Surg. 52 45. 10.1186/s40463-023-00652-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mudry A., Mills M. (2013). The early history of the cochlear implant: A retrospective. JAMA Otolaryngol. Head Neck Surg. 139 446–453. 10.1001/jamaoto.2013.293 [DOI] [PubMed] [Google Scholar]
- Smetak M., Fernando S., O’Malley M., Bennett M., Haynes D., Wootten C., et al. (2023). Electrode array positioning after cochlear reimplantation from single manufacturer. Cochlear Implants Int. 24 273–281. 10.1080/14670100.2023.2179756 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Somdas M., Li P., Whiten D., Eddington D., Nadol J. (2007). Quantitative evaluation of new bone and fibrous tissue in the cochlea following cochlear implantation in the human. Audiol. Neurootol. 12 277–284. 10.1159/000103208 [DOI] [PubMed] [Google Scholar]
- Terry B., Kelt R., Jeyakumar A. (2015). Delayed Complications after cochlear implantation. JAMA Otolaryngol. Head Neck Surg. 141 1012–1017. 10.1001/jamaoto.2015.2154 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.