Ear, Nose & Throat Journal2023, Vol. 102(11) 701–708© The Author(s) 2021Article reuse guidelines:sagepub.com/journals-permissionsDOI: 10.1177/01455613211022075journals.sagepub.com/home/ear
Purpose: To evaluate if a specific type of cochlear implant (CI) electrode array (EA) reveals higher rates/prevalence of vestibular symptoms and to characterize their respective relationship to intracochlear position and objective vestibular function. Methods: This retrospective study included 71 cochlear implantations in patients older than 18 years. The electrode position within the cochlea, electrode insertion angle, and cochlear coverage were determined from postoperative multiplanar reconstructed cone-beam computed tomography scans. All device manufacturers were represented. Data related to preoperative and postoperative PTA as well as vestibular symptoms in the preoperative and postoperative stages were collected from the patient’s records. Results: Twelve of the 71 (16.9%) CI patients experienced vertigo symptoms in the early postoperative period. In 5 (7.0%) patients, the vertigo complaints lasted until the time of the first activation (5-6 weeks postoperative). Postoperative onset of vestibular symptoms was more often seen in patients receiving lateral wall (LW)/straight EAs (19%) compared to perimodiolar/precurved EAs (7%), but this was only a trend and no statistical significance was observed. Moreover, preoperative pathologic caloric responses (CRs) better predicted the postoperative onset of vestibular symptoms. Conclusion: The preoperative consideration of a complicated CI-induced vertigo is important in the counseling particularly of elderly patients. We identified some risk factors for post-CI vertigo that should be considered in the patient’s counseling: preoperative pathologic CRs, the extent of surgical trauma, and possibly the use of an LW EA, regardless of the length.
Keywordscochlear implant, perimodiolar electrode array, lateral wall, postoperative vertigo, cochlear coverage, PTA shift
Postoperative complications in cochlear implant (CI) recipients requiring surgical revision or hospitalization for medical treatment are rare. The procedure has been standardized because of increasing experiences and controlled postoperative care.1-4 It is important to mark that vestibular impairments are one of the most common complications of cochlear implantation. During cochlear implantation, structural and functional changes occur in the labyrinth because of the anatomic proximity between the auditory and vestibular system.5,6 Data on the incidence of postoperative vestibular complaints after CI surgery show widely varying results between 0.33% and 75%.5,7-9 Among the factors discussed, the position of the electrode array (EA) within the cochlea appears to be crucial for many functions as well as postoperative complications.
With surgical and technical advances of CI, new designs of lateral wall (LW) and perimodiolar (PM) EAs have been devised to effectively and efficiently stimulate the auditory nerve.10 The LW EA is located along the lateral wall of the scala tympani (ST) and is believed to stimulate the neural fiber endings of the auditory neurons at the Organ of Corti.11 In contrast, by positioning the EA closer to the modiolus, the PM EA is believed to stimulate the spiral ganglion cells more directly and might lead to better localized neural stimulation.12,13 However, due to its greater stiffness and volume, the PM EA has a higher incidence of causing intracochlear trauma during electrode insertion,11 identified as scalar translocation into the scala vestibuli (SV).
Clinical trials studying the effect of different types of EA on the vestibular function and subjective vestibular complaints are not well reported.14 The purpose of our study is to analyze if the type of CI EA and its exact intracochlear position show a relationship of higher rates of postoperative vestibular complaints and peripheral vestibular function.
This retrospective study includes 71 cochlear implantations performed in patients older than 18 years at a tertiary referral university hospital between August 2018 and December 2018. We chose this period of time to maintain consistency of the providers directly involved with patient care as the same 2 otorhinolaryngologists looked after the patients on an outpatient basis. Inclusion criteria were first CI surgery ipsilateral and age more than 18 years. Exclusion criteria were removal of an ipsilateral vestibular schwannoma and previous temporal bone fracture.
The CI surgery was performed using the standardized surgical technique.4 Briefly, a mastoidectomy with opening of the facial recess and exposure of the round window niche was drilled. This was followed by gentle puncture and incision of the round window membrane, insertion of the EA into the ST, fixation of the EA in a bone slit inferior to the facial recess, and fixation of the internal part of the device. We used the round window approach as first choice when inserting EA. A postoperative cone-beam computed tomography (CBCT) was performed to verify the intracochlear position of the EA (Figures 1 and 2). In our tertiary care hospital, all patients undergoing a CI surgery are hospitalized for at least 4 days (including the surgery day).
The patients were scanned using a 3-D ACCUITOMO 170 Digital CBCT scanner (J. Morita Tokyo mfg Corp). The reconstruction of the Digital Imaging and Communications in Medicine data was performed using the Osirix MD software version 2.5.1.
The insertion angle (IA) in degrees, cochlear coverage (CC) in percent, insertion depth (ID) of the EA in the cochlea in mm, and scalar translocation were determined. An illustration of a postoperative segmentation and determination of the electrode IA and ID of the EA in the cochlea is given in Figure 3. By performing a cochlea and EA segmentation to measure cochlear duct length, CC and the ID methods described by Schurzig et al15 were used. Scalar translocation was assessed as an indirect way to determine intracochlear trauma. Translocation was defined when the EA had exceeded more than half of the section of the cochlea duct from ST into the SV (Figures 4 and 5).
Vestibular testing was conducted preoperatively and postoperatively in all patients. The postoperative stage was analyzed in the early postoperative period (ie, during the first 3 days after surgery) and at the time of the first fitting (initial activation) of the CI 5 to 6 weeks after implantation. Data related to vestibular symptoms (duration of vertigo, dizziness, imbalance, and concomitant vegetative symptoms) in the preoperative and postoperative stages were collected from the patient’s records. Measurements preoperative and in the early postoperative period included and recorded were spontaneous nystagmus (SN) and head-shaking nystagmus (HSN) and preoperative caloric test. Postoperative at the time of the first fitting SN and PN, Romberg and Unterberger test as well as clinical head impulse test (HIT) were performed and recorded in all patients. In cases of vestibular complaints at the time of the first CI fitting, the patients were interviewed about the time of onset, quality, frequency, duration, triggering factors (eg, special body position), and concomitant vegetative symptoms. Whenever nystagmus, a HIT impairment or a pathologic Romberg test and Unterberger test were identified, the caloric responses (CRs) were repeated according to the method of Hallpike.17 Whenever nystagmus, a HIT impairment or a pathologic Romberg test and Unterberger test were identified, the CRs were repeated according to the method of Hallpike.18
Multivariant logistic regression analysis and univariate analysis were conducted using χ2, Fisher exact tests, and Point-Biserial correlation for categorical variables using SPSS. A P value <.05 was set as a significant value.
Institutional review board exemption for this retrospective study was given (project identification code 1897-2013).
The adult population comprised 39 (54.9%) women and 32 (45.1%) men, with a median age of 58 years (range: 19-86 years) at the time of implantation. Of the 71 patients, 57 received their first CI and 14 (19.7%) had already been implanted on the contralateral side. Implanted cochlear devices included MED-EL Flex 28 (n = 17), Nucleus 532 (n = 7), Nucleus 522 (n = 14), Nucleus 512 (n = 6), Advanced Bionics (AB) HiFocus Slim J (n = 19), and Oticon straight soft EVO (n = 8). Table 1 lists the CI electrodes with their respective desired position in the cochlea and the mean values of ID, IA, and CC of each EA in the cochlea. In total, 14 (19.7%) patients received a PM EA, 57 patients an LW EA, of which 40 (56.3%) patients a short LW EA and 17 (23.9%) patients a long LW EA.
Preoperative vestibular symptoms. Of the 71 patients, 59 (83.1%) patients reported no vestibular symptoms prior to the surgery, whereas 6 (8.4%) patients described vestibular complaints and the remaining 6 (8.4%) had preoperative gait unsteadiness. The CR of the patients with vertigo complaints prior to the surgery revealed in 3 cases a canal paresis (CP) contralateral, 2 were implanted on the contralateral ear and 1 patient had MD on the contralateral ear. One patient with vestibular complaints showed a bilateral CP and 1 patient a normal CR. All preoperative caloric test results are listed in Table 2.
The patients were assigned into 1 of 3 different groups: The first group consisted of patients without vertigo (n = 59), the second group consisted of patients with vestibular symptoms in the early postoperative period (n = 12), and the third group comprehended patients with vestibular symptoms at the time of the first fitting of CI (n = 5).
Without vertigo group. Fifty-nine patients experienced no vestibular symptoms after the CI surgery. It must be emphasized that all of the 14 bilateral implanted patients were found within this group, that is, the patients with bilateral CI experienced no vestibular symptoms after the surgery.
Vestibular symptoms in the early postoperative period group. Twelve (16.9%) patients experienced vertigo symptoms in the early postoperative period after surgery. All 12 patients who had vertigo manifested within the first postoperative days were unilaterally implanted and were treated using intravenous steroids once daily for 3 days (250 or 500 mg Soludecortin H—depending on secondary diagnosis) prior to discharge. Table 3 summarizes the implanted devices, intraoperative specialties, vestibular complaints, and vestibular test results within this group. The Fisher exact test found a significant association between vertigo in the early postoperative period and preoperative CR pathology, χ2 (1) = 4.261, P = .050, n = 70. The association is weak (CC = 0.247, P = 0.05, Cramer V = 0.247, P = .05). These finding was regardless of the electrode type as provided in Table 4.
Vestibular symptoms at first fitting. In 3 of the 12 patients with vestibular symptoms in the early postoperative period, the subjective vertigo complaints lasted until the time of the first fitting. No patient of the study population developed new vestibular complaints after discharge. In 2 of these patients with vestibular symptoms in the early postoperative period, a pathologic result in at least 1 of the abovementioned testing results was noticed. These patients with remaining vestibular complaints occurred in the early postoperative period and persisting until the time of the first fitting are printed in bold type in Table 3.
In the group of patients without vertigo, 1 patient showed an obliteration of the oval window on the implanted ear. Two patients in the group of vertigo in the early postoperative period showed intraoperative abnormalities. In one case, a bilateral pericochlear otosclerosis was found, with an early obliteration of the basal cochlea turn. In this case, an insertion test device ‘‘Insertion probe L’’ by MED-EL was inserted prior to the electrode insertion of a MED-EL Flex 28 electrode and an extension of the round window anteroinferiorly was drilled. In another case, the CI EA was dislocated in the vestibulum after the first electrode insertion and was then removed and inserted correctly using an AB HiFocus Slim J electrode. These 2 patients with intraoperative difficulties did not have any vertigo complaints preoperatively and normal caloric results preoperatively.
Out of the group of patients without vertigo 47 received an LW EA (31 of them a short LW and 16 a long LW EA) and 12 a PM EA. Within the group of patients with vestibular symptoms in the early postoperative period, 11 of the 12 patients received an LW EA. Within the group of patients with vestibular symptoms until the first fitting, 4 patients were implanted with a short LW EA and 1 with a long LW EA.
In total, 19.3% of the patients with LW EA developed vertigo symptoms and 7.1% of the patients with PM EA. Although there was a tendency of LW EA and development of vertigo symptoms, the Fisher exact test found no statistically significant association neither between vestibular symptoms in the early postoperative stage and electrode type, χ2 (1) = 4.287, P = .146, n = 71, nor vestibular symptoms at the time of first fitting and electrode type, χ2 (1) = 1.630, P = .613, n = 71.
The evaluation of the postoperative CBCT scans showed that the mean IA of the PM EA was 342 degrees (SD ±47), the mean CC was 59.25% (SD ±0.05), and the mean ID was 16.37 mm (SD ±1.18). The mean IA of the short LW EA was 375 degrees (SD ±40), the CC was 62.35% (SD ±0.04), and the ID was 21.0 mm (SD ±2.07). The IA of the long LW EA was 509 degrees (SD ±72) with a mean CC of 77.46% (SD ±0.04) and a mean ID of 25.57 mm (SD ±1.60). The analysis of IA as well as CC and ID between the group of patients without vertigo and with vestibular symptoms in the early postoperative period and at the time of first fitting showed no statistical significant difference.
Twelve (16.9%) implantations showed an EA dislocation into the SV in the basal turn (5 PM EA, 2 short LW, and 5 a long LW EA). There was no statistically significant difference regarding scalar translocation and either EA type or vestibular symptoms postoperatively.
Over the past decades, many studies have investigated the occurrence of vertigo after cochlear implantation. Overall, when patients with CI present a complication postoperative, vertigo is a frequent complaint. In the majority of the postoperatively dizzy patients in our study, vestibular symptoms were temporary, a reassuring finding. Approximately 17% of the patients who reported vestibular problems did it during the first week after surgery and only in 4.2% the subjective complaints lasted until the time of the first fitting. Dizziness and vertigo complaints were of new onset in all 12 (16.9% of 71) patients and vestibular function deteriorated in 3 cases (4.2% of 71; Table 3). In summary, the follow-up of in the early postoperative period occurred vestibular complaints showed that a relatively high number of patients (7/12, 58.3%) recovered after treatment with intravenous steroids, requiring no other treatment.
Other publications have yielded a varying incidence of postoperative vertigo after CI surgery of 0.33% to 75%.1,6,7,9,19-21 However, a direct comparison with other studies on postoperative subjective or objective vertigo complaints is difficult because several different methods have been used to evaluate the vestibular impairment. The meta-analysis by Abouzayd et al10 regarding the best objective measurements for analyzing vertigo complaints in patients with CI revealed a sensitivity of 21% for caloric tests, 32% for the cervical vestibular evoked myogenic potentials (c-VEMP), and 50% for HIT.10 Given the low sensitivity of the caloric responses in this meta-analysis, postoperative caloric testing was rather indicated in case of clinical signs of lateral ampulla impairment such as the presence of corrective saccades in the HIT, what we also investigated in the patients before performing a caloric test. Previous studies reported about a lack of correlation between vertigo symptoms and functional impairment on objective test results using VEMPs, CR, and HIT.22 These results suggest that the clinical assessment and subjective patient reports are equally important to the vestibular tests.
In the present study, we investigated the vestibular function with a special attention on the subjective complaints described by the patients. Therefore, we subjectively screened the vestibular function by asking the patients about their complaints and later obtained objective information regarding these symptoms by examining the abovementioned test battery on an outpatient basis. The impaired vestibular function did not correlate with vertigo symptoms in every case. Two of the five patients who showed spontaneous or head-shaking nystagmus or corrective saccades in the HIT did not have subjective vertigo complaints at the time of investigation (Table 3).
A study by Fischer et al16 evaluated the EA placement of 63 CIs in 2015 and found an EA dislocation in 5 cases (all long LW EA). In 3 of the 5 cases, the EA was inserted via a cochleostomy, and in all cases, the EA penetrated the basal membrane in the first 45 degree segment of the basal turn of the cochlea. And according to Ketterer et al,17 PM EA (21.6% dislocation rates of the Cochlear Contour Advance EA) and long LW EA (dislocation rate of 26.3% of the MedEl Flex soft) have the highest rate of dislocations nowadays. Insertion of the EA choosing the wrong angle and wider basal diameters of EA could be the underlying cause for dislocations in the basal turn. A histopathological study of 21 temporal bones conducted by Adunka and Kiefer23 in 2006 has provided evidence that intracochlear trauma increases with deep insertions and an IA greater than 360 degrees. In our investigation, neither the CC nor the IA nor the scalar translocation did correlate with vertigo complaints or deterioration of the vestibular function. Regarding the type of EA used, the majority (n = 58; 81.7%) of our 71 implanted patients received an LW EA and in most cases of postoperative vertigo complaints an LW EA was inserted. The question remains whether or not this type of EA prompts the vestibular complaints. The findings of our study showed that 11 of the 12 patients with postoperative vertigo symptoms were implanted with an LW EA. A limiting factor must be taken into account, as a statistical error could occur due to our small sample of patients who received a PM EA and the predominance of patients implanted with LW EA.
A possible hypothesis regarding the onset of vestibular symptoms could be a pathophysiological mechanism where the LW EA is applying minimal but persistent pressure to the inner ear veins in the wall and floor of the ST. Inner ear veins drain mainly through separate accessory channels along the cochlear and vestibular aqueducts. Interference of the cochlear blood flow might occur as a result of compression since these vessels are not protected by bone.24 The disruption leads to instantaneous pathological changes in the inner ear, and a disturbed microcirculation may be etiologically linked to different inner ear disorders, such as sudden deafness, noise-induced hearing loss, vestibular neuritis, and MD.25-28
In contrast, a study by Nordfalk et al29 showed that only 3 (7.7%) of 39 patients experienced new symptoms of vertigo after the implantation of LW EA. In other previous studies comparing precurved and flexible EA, no significant change was found regarding postoperative vertigo.6,30,31 However, in these studies, postoperative vestibular function and symptoms were not the primary outcome and type of device not the main variable. In contrast, the study by Frodlund et al14 found a significant difference in decline of CR in patients implanted with a straight EA compared with implantation with precurved or flexible EA. To clarify whether the type of device and especially the intracochlear position of the EA are influencing long=term vestibular function, larger prospective observational studies with pairing groups by EA designs are recommended. In order to avoid confounding variables accounting for the observations of postoperative vestibular symptoms, a prospective study with pairing groups by EA designs should be planned with unilateral implanted patients without history of vestibular symptoms preoperative and no CR pathology preoperative.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Farnaz Matin https://orcid.org/0000-0002-3953-112X
1 Otorhinolaryngology Department, Head and Neck Surgery, Hanover Medical University, Hannover, GermanyReceived: January 27, 2021; revised: April 28, 2021; accepted: May 15, 2021
Corresponding Author:Farnaz Matin, Otorhinolaryngology Department, Head and Neck Surgery, Hanover Medical University, Hannover 30625, Germany.Email: matin.farnaz@mh-hannover.de
