Search strategy

This study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement and supplemented by guidance from the Cochrane Collaboration Handbook. A protocol of this systematic review was designed a priori and was registered in the PROSPERO database (registration number: CRD42021237489) Searches were performed from January 1, 2016 until February 14, 2022 in the following databases: PubMed (MEDLINE), Science Direct, Web of Science, Scopus, and EMBASE databases using the following MeSH terms: ("Cochlear Implants"[Mesh] OR "Cochlear Implantation"[Mesh]) AND ("Ear, Inner/abnormalities"[Mesh] OR "Ear, Inner/growth and development"[Mesh]), where MeSH search was not available the following Boolean terms were used ("Cochlear Implant" OR "Cochlear Implantation") AND ("Inner ear abnormalities" OR "cochlear abnormalities" OR "Congenital malformation").

Study selection

Studies were excluded based on the following exclusion criteria: studies that did not fit the above characteristics, review articles, meta-analyses, abstracts only, conference proceedings, editorials/letters, case reports, or articles published before 2016. Articles prior to 2016 were excluded as they were previously evaluated in a systematic review [4]. Additionally, adult patients were excluded from analysis. The following inclusion criteria were used: data including closed-set word (CSW) tests, open-set word (OSW) tests, surgical data including intraoperative and postoperative complications, Categories of Auditory Performance (CAP), and/or Speech Intelligibility Rating (SIR) in patients undergoing unilateral or bilateral cochlear implantation for SNHL secondary to IEMs. All searched titles, abstracts, and full-text articles were independently reviewed by at least two reviewers (S.S., R.W., J.L., A.F., and C.D.). Disagreements over inclusion and exclusion criteria were resolved through a consensus between the reviewers or discussion with other investigators of this study. Articles were initially screened based on title and abstract screening before proceeding to full-text analysis.

Quality assessment

The “Tool to Assess Risk of Bias in Cohort Studies” published by the CLARITY group was used to perform quality assessment of the studies [12]. Eight areas were evaluated: population selection, exposure assessment, pre-existing exposure, control matching, confounding assessment, assessment of outcome, follow-up assessment, and co-intervention assessment. The risk of bias was rated as low, unclear, or high according to established Cochrane guidelines. At least two reviewers independently conducted this assessment (S.S., R.W., J.L., A.F., and C.D.), and any disagreements were resolved by consensus between the reviewers or discussion with other investigators of this study.

Data extraction

All data were extracted and separated by type of malformation [6, 7]. These malformations included: Michel deformity, common cavity (CC) deformity, cochlear aplasia, cochlear dysplasia, cochlear hypoplasia, enlarged vestibular aqueduct (EVA) syndrome, incomplete partition types I (IP-I), II (IP-II or Mondini deformity), and III (IP-III). Descriptions and audiological findings of these malformations are included in Table 1 [13]. Any discrepancies in classification or grouping were discussed with the senior author (A.E.) for a final decision.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Descriptions and audiological findings in individuals with IEMs [13].

https://doi.org/10.1371/journal.pone.0275543.t001

Intraoperative and perioperative complications were grouped into cerebrospinal fluid (CSF) leak or gusher, postoperative infections, electrode-induced facial nerve stimulation, and complete versus incomplete electrode insertion.

Speech perception tests were grouped into CSW or OSW tests depending on which category they fell into. Furthermore, data regarding speech perception abilities were extracted from multiple studies that included SIR scores. Finally, CAP results were also included to stratify auditory receptive skills apart from speech perception in patients following cochlear implantation.

Open-set testing

Overall, three studies included open-set results. Isaiah et al. included 78 patients with IEMs. One year after CI surgery, results showed that 33% of control patients failed open-set speech discrimination. Multiple groups with IEMs failed (defined by a total score of 0%) in both open- and closed-set speech testing. This included 77% with cochlear dysplasia and 89% with vestibular dysplasia [14]. Differences between the success rates of speech performance testing between these IEM groups were not statistically significant [14].

In the study of Melo et al., patients were assessed once between 8–180 months post-implantation. Although not statistically significant, the average test scores on monosyllabic word sets in patients with IEMs was 78.1% compared to 69.6% in the controls [15].

In Ozkan et al., open-set sentence recognition testing was performed [16]. At the 1–3-year mark, IEM patients without common cavity deformity had a maximum score of 80%, while IEM patients with common cavity did not score high enough on the pattern perception test for sentence recognition testing.

Closed-set testing

Five studies with closed-set speech perception tests were included in the final dataset. In the study of Celik et al., control patients did not have congenital IEMs. IEMs in the study included cochlear hypoplasia, IP-II, IP-I, and CC. Both groups scored similarly on the 6-month postoperative Monosyllabic-Trochee-Polysyllabic Test (MTP) in the 3-word groups (Control = 7.5, IEMs = 7.7), 6-word groups (Control = 10.4, IEMs = 7.7), and 12-word groups categories (Control = 14.3, IEMs = 14.0) [17]. The results of this study were not statistically significant.

Isaiah et al. evaluated the CSW ability of their participants using the Early Speech Perception (ESP) test. In patients with IEMs, 9% of those with cochlear dysplasia achieved a non-zero score on the closed-set evaluation alone, along with 6% of patients with vestibular dysplasia and 10% with EVA syndrome [14].

Ronner et al. discussed the preoperative and postoperative early speech perception word intelligibility scores for cochlear implantation patients. Patients were stratified into normal anatomy, low-risk anatomy, and high-risk anatomy (Table 4) [18].

In the study of Ozkan et al., closed-set pattern perception testing was done on both control and experimental patients. Patients with IEMs were compared against one another; those with dilatation of the vestibule, IP-I, cochlear hypoplasia, and CC deformity performed statistically significantly worse when compared to patients with EVA, IP-II, and IP-III. At the 1-3-year postoperative time point, EVA patients scored 90% on pattern perception test, while patients with CC deformity scored approximately 35%.

Patel et al. found that when comparing patients with IP-II to patients without IP-II undergoing cochlear implantation, patients with IP-II had a statistically significant decrease of 30.2% in word recognition score (WRS) following cochlear implantation to patients with normal anatomy [19].

CAP score

Three studies included a preoperative and postoperative CAP score. Zhang et al. evaluated children with CC deformity against a control group matched in sex, age, and time of implantation (n = 6) [20]. Both the control and IEM groups showed improvement, though patients with CC deformity improved significantly less than those with normal inner ear anatomy. The mean CAP improvement in patients with IEMs was 3.50 compared to 9.50 in the control group [20].

Melo et al. assessed CAP-II scores in patients with and without IEMs that received CIs. The authors defined major IEMs as cochlear hypoplasia and IP-II, while all other anomalies were considered minor IEMs [15]. These values were not statistically significant, and there was no statistically significant difference between the patients with IEMs and those in the control group at 12, 24, and 36 months post-cochlear implantation [15].

In Choi et al., CAP scores were measured in patients with IP-III deformity and in control patients with normal inner ear structure before and after surgery. Audiologic assessments occurred pre-cochlear implantation and at 3, 6, 12, 18, and 24 months after implantation, and the mean CAP scores in the IP-III patients were 0.8, 2.6, 3.0, 3.6, and 3.9, respectively [21]. When compared to the control groups, the scores were not statistically significant at any time point before 24 months postoperatively. At 24 months, the patients with IP-III deformity performed significantly poorer than the controls with a mean CAP of 3.9 versus 5.0.

SIR score

Three studies included SIR scores as an outcome measure. Zhang et al. evaluated SIR scores in patients with CC deformity compared with a sex, age, and time of implantation matched control group. In congruity with their CAP score outcomes, both the CC and normal anatomy groups demonstrated improvement following cochlear implantation; however, those with CC deformity showed significantly less improvement in SIR 1-year postoperatively. The mean SIR score in the common cavity group was 3.75, while the control group scored 9.25 [20].

Melo et al. also included SIR scores as post-cochlear implantation audiometric outcomes. Mean SIR score 24 months following surgery in those with major IEMs was 3.71 (+/-1.43), while those with minor IEMs was 4.41 (+/-0.90) [15]. Both groups had an average SIR score of 4.3 with a variability of +/-1.0 in patients without IEMs versus a +/-1.1 in patients with IEMs [15]. Average SIR score in patients with IEMs were lower when compared to the controls at 12-, 24-, and 36-months following surgery; however, these differences were not statistically significant [15].

Qi et al. conducted a study evaluating SIR scores in patients for 5 years with rating assessments occurring before surgery and at 1, 3, 6, 12, 24, 36, 48, and 60 months postoperatively. Both groups demonstrated significant improvements between each of the time points after 3 months except for the 6–12-month period and the 36–48-month period; no statistically significant difference in SIR was found at these time points [22].

Operative findings

Operative findings were included in a total of five studies within this review. In all patients in Kamogashira et al., the standard facial recess approach was adopted, and a CI electrode was inserted through a cochleostomy. Intraoperative CSF gushers were encountered in 38% of children with cochlear hypoplasia and 29% with IP. No patients experienced facial nerve palsy; however, postoperative facial nerve stimulation by CI was frequently observed in children with CC (50%) and cochlear hypoplasia (38%) [23].

In Vashist et al., 29% of patients with IEMs experienced CSF gushers (2 EVA and 2 IP-II) or CSF oozing (1 IP-II) while only 3% with normal inner ear anatomy suffered from CSF gushers [24]. Only patients with IP-I, IP-II, IP-III, EVA syndrome, dysplastic semicircular canal, and narrow internal auditory canal (IAC) were included. CSF oozing was defined as spontaneous resolution of CSF leakage within 5–10 minutes, while CSF gusher was defined as resolution of leakage within 15–20 minutes only after packing of the round window site [24]. All 9 patients with CSF leakage had complete insertion of the electrode, and there were no cases of CSF otorhinorrhea or meningitis in the follow-up period.

In Isaiah et al., a total of 5 patients (6.4%) had an intraoperative complication of CSF gusher; 4 of these patients had associated cochlear dysplasia alone or in conjunction with another IEM. In addition, 6.4% of patients had incomplete insertion of the electrode (3 cochlear dysplasia), while 3.8% had full insertion with difficulty [14].

In Melo et al. 9 patients had cochlear hypoplasia, 5 had IP-II, 3 had EVA syndrome, 3 had EVA syndrome with semicircular canal (SCC) aplasia, 5 had partial SCC aplasia, and 1 patient had cochlear nerve hypoplasia [15]. Only 3.8% of patients with cochlear hypoplasia had surgery complicated by incomplete electrode insertion; all other patients had successful complete insertion of the electrodes. CSF leakage occurred in 7.6% of patients, all of which had IP-II. The study did not assess postoperative facial nerve stimulation.

In Choi et al., all 8 patients with IP-III experienced a CSF gusher at the cochleostomy site during the procedure [21].

In Parent et al., patients with cochlear malformations suffered from an increased rate of statistically significant complications. Multivariate analysis found that adult patients with cochlear malformations had a higher rate of complications than pediatric patients [25]. The combined minor and major complication rate was 10.12% (6.4%-15.6%) in patients suffering from cochlear malformations; however, the type of complication was not stratified by etiology of deafness [25].

Speech perception results

Studies suggest that patients with IEMs take longer to develop speech perception skills than those without IEMs [14, 15, 17–19]. However, as time of post-implantation progresses, patients with IEMs show no statistically significant difference in speech perception on open-set tests compared to children with normal cochlear anatomy. Although most IEM patients fare well in speech perception 2–3 years after implantation, IEMs such as common cavity deformity or severe cochlear hypoplasia that disrupt cochlear architecture or cochlear nerve formation have limited improvement and tend to perform much worse in all speech perception tests [26, 27]. Patients with milder forms of cochlear hypoplasia tend to perform better. Many studies suggest that earlier age of implantation resulted in better performance on open-set word speech perception tests [28, 29]. Similar to their performance on the open speech perception test, IEM patients performed comparably to controls in the closed speech perception test at multiple time points post-implantation, including at 6, 12, and 24 months [21].

Auditory outcomes

The three studies that included a CAP score indicated its usefulness in determining auditory perception beyond speech perception. This is important given that children with IEMs may have more difficulty understanding speech than environmental sounds. In all studies analyzed, patients with CC deformity performed the worst in CAP scores, likely due to concurrent malformation of the cochlear nerve. Although CC deformity patients did not perform well, they still experienced improvement compared to preoperative CAP scores, as did all patients with other IEMs. Most studies showed that, in patients with IP-II and IP-III malformations, there were no statistically significant differences in CAP score when compared to normal anatomy controls. Choi et al. showed that improvements in CAP score in patients with IP-III performance halted at 24-months following cochlear implantation and that they scored worse than control patients. However, Daneshi et al. emphasizes that CAP score results are subjective and commonly rely on parental statements and observations rather than objective measures [30].

The speech perception was analyzed in three studies using the SIR score. In conjunction with CAP scores, patients with CC deformity performed poorer on SIR. Patients with CC deformity frequently have decreased spiral ganglion cells of the cochlear nerve, resulting in reduced performance on speech perception [20, 31]. It could be hypothesized that this decreased auditory performance may then translate to a decreased overall capability of speech perception. However, because only three studies were included that reported SIR results, it is difficult to conclude whether IEM significantly impacts SIR.

Operative outcomes—CSF gusher

In conjunction with speech perception and production, operative outcomes are crucial to prepare for and expect during surgery. CSF oozing and gushers occurred most in patients with IP-II and IP-III malformation, EVA syndrome, or a combination of the two [2, 21, 32]. Benchetrit et al. deduced that the increased incidence of CSF gusher in patients with IP-II and EVA syndrome is likely due to direct communication between CSF and the inner ear [33]. However, all cases of gushers were conservatively managed and controlled by either soft tissue packing, reverse Trendelenburg positioning during surgery, or lumbar drainage in more challenging cases [34–36]. Bayazit found that revision due to recurrence of CSF leakage was rare, only occurring among patients with EVA and IP-1 [2]. Choi et al. found that all patients with IP-III had CSF gusher, highlighting the increased risk due to the absence of the modiolus resulting in direct communication between the cochlear base and the internal auditory canal [21]. These results emphasize that for patients with IEMs undergoing cochlear implantation, even the most severe intraoperative risks can usually be managed conservatively, reserving invasive surgical intervention for rare, severe cases. Preoperative imaging and meticulous surgical planning are crucial to reducing the risk of gusher occurrence intraoperatively.

Operative outcomes—facial nerve stimulation

Postoperative facial nerve stimulation was more prominent in patients with CC deformity with an incidence of greater than 50% [23]. The aberrant positioning of the facial nerve and combined cochleovestibular structure in patients with CC deformities increases the risk of facial nerve stimulation following cochlear implantation. As described by Sennaroğlu, the facial nerve migration during embryogenesis may be abnormal if the basal turn of the cochlea is not properly formed, as is the case with CC deformity [6]. Patients with IP (I, II, or III) had the lowest rate of aberrant postoperative facial nerve stimulation [37, 38]. Facial nerve aberrations require detailed surgical planning as well as anatomical mapping with surgical modifications to ensure a successful cochlear implantation in patients with IEMs [39]. Surgical intervention in these patients may further require alternative approaches such as retrofacial approach, transaditus approach, transcanal approach, or a combined transmeatal approach with posterior tympanotomy [35, 39–41]. Most cases of aberrant facial nerve stimulation were controlled by turning off select electrodes, decreasing the C-level, or readjusting the electrode map, as discussed in the study of Naito et al. [42].

Operative outcomes incomplete insertion

Adunka et al. concluded that incomplete insertion of the electrode array was most common in patients with IEM [43]. Of the 121 patients whose insertion rate was discussed within the included studies, 6 patients (4.9%) had incomplete insertion, which is larger than reported by the study of Farhood et al. Anatomy that shortens the length of the cochlea (such as cochlear hypoplasia) or makes the cochlea difficult to navigate should be considered, depending on the severity of the hypoplasia, when deciding the size of the electrode to use for cochlear implantation. Given the increased risk of insertion of the electrode array into the vestibule or internal auditory canal in IEM patients, modified surgical approaches may be needed, especially in cases of difficult cases such as IP-III in which risk of erroneous insertion into the internal auditory canal is high [44]. Electrodes must be chosen based on preoperative imaging to decrease the risk of electrode failure or operative complications [45]. In IP-III patients, specifically, there is a higher risk of insertion into the internal auditory canal. A recently published study by Minami et al. categorized IEMs based on the severity of modiolus deficiency and internal auditory canal/cochlear nerve deficiency to help address the types of electrodes that could be used for particular IEMs [46]. This study suggests that a perimodiolar electrode should not be utilized in patients with IP-I, IP-III, or common cavity, due to cochlear dysplasia or modiolus defects. However, in patients where the modiolus is present, but the cochlear nerve is deficient, such as IAC stenosis, cochlear aperture stenosis, or IP-II, a perimodiolar electrode can provide increased intensity to perpetuate signals through a deficient cochlear nerve [46]. Historically, a one-size-fits-all method has been used, however, electrodes have undergone multiple technological advancements over the past 30 years. Newer electrodes utilize various lengths and orientations as well as advanced processing systems and software, permitting the clinicians to tailor the electrode towards specific patient anatomy [47].

Limitations

There is very little standardization for the timing and administration technique of speech perception tests in CI [48]. In general, closed-set tests are easier for patients to take as their answer choices are limited to the choices provided to them, while open-set tests are more difficult [49]. Few studies included in this review did not separate open-set tests from closed-set tests, making deduction of speech perception issues difficult to pinpoint.

Interpretation and comparison of results was also complicated by varying length of data collection and follow-up. For example, several studies collected data over 15–20 years, during which time surgical techniques and technology have improved, making it difficult to directly compare patients at the beginning of a study to those at the end. Appropriate comparison of cochlear implantation outcomes in patients with IEMs against patients with normal anatomy may also provide better insight into the expectations that need to be discussed with parents of patients with IEMs. However, given the rarity of IEMs, it is difficult to fault studies for gathering data in this method. In addition, to ensure inclusion of studies with comparable controls, many chart reviews and case series studies with strong data were not included. Also, given the commonality of IEM with concurrent neurological disorders, it would be interesting in future studies to evaluate the effect of additional handicap (autism spectrum disorders and pervasive developmental disorders) on audiometric outcomes [2].

Although rare, post-operative infection, including meningitis, can have significant and devastating long-term effects on cochlear implant outcomes. As meningitis has been shown to be more common among patients with IEM, there is a need to identify infection sources as well as tailor the post-operative treatment course in patients with IEM [50]. To limit cases of meningitis, it has been suggested to encourage vaccination of patients prior to surgery and have shorter follow-up intervals in patients who experienced perilymph gushers, perilymph oozers, otorrhea, or rhinorrhea [50].

Furthermore, this review was limited to studies searchable through PubMed (MEDLINE), Science Direct, Web of Science, Scopus, and EMBASE databases and mapped to the MeSH search terms commonly used in systematic reviews, which may have impacted the results reported.