This retrospective study was approved by the local Institutional Ethics Committee (decision number AKBE/175/2022). The project conforms to The Code of Ethics of the World Medical Association (Declaration of Helsinki).

This study performed a retrospective analysis of data obtained from the medical records of patients diagnosed with CHARGE syndrome. Twenty temporal bones of ten consecutive patients diagnosed with the syndrome were subjected to detailed CT analysis focusing on the inner ear and internal acoustic canal. One patient did not have a complete set of audiological tests; however, he was included in the study group due to the CT findings worth detailed radiological assessment. Two independent, experienced investigators (also authors of this manuscript) evaluated the patient's CT scans using the same study protocol and questionnaire.

The patients were diagnosed in two large co-working tertiary centers due to a typical CHARGE combination of clinical features. One center provides genetic and audiological testing and is a full children's tertiary hospital (covering all specialties needed for CHARGE syndrome patients and more). The second one specializes in audiological testing and hearing prostheses, including cochlear implants in children. According to the literature, about 60% of patients with CHARGE present autosomal dominant mutations in the CHD7 gene. More than 1000 mutation variants that can cause CHARGE syndrome have been identified.6,8–12 In this study, six analyzed patients underwent genetic testing confirming the gene mutations. Four were not analyzed genetically due to parents' no agreement with these tests but presented typical specific and recognizable patterns of CHARGE anomalies.

All ten patients underwent CT scanning of the temporal bone. This study included only the medical records of patients with CHARGE syndrome and CT scans. Patients with CHARGE syndrome diagnosis but no CT were not included. Performing full audiological diagnostics and CT scanning of the temporal bone, which in most cases in small children requires sedation, is not always a priority due to the other more serious components of the syndrome.

A scan analysis protocol scheme was used for CT evaluation, which had proven satisfactory in previous studies. In addition, a questionnaire with key features was created to assist the investigators with CT evaluation and to standardize their assessment and thus enable the comparison of results. Table 1 presents the mentioned questionare, pointing out the key features of computed tomography analysis. An attempt was made to classify the identified malformations using the newest Sennaroglu classification.13

All scans were analyzed using RadiAnt DICOM Viewer 2022.1 64-bit (Medixant, Poznan, Poland) with the MultiPlanar Reconstruction (MPR) option.

Initially, in each case, the researcher set the window level to 500 Hounsfield Units (HU) and the window width to 3000 HU. Then, the protocol developed in earlier studies was used as follows. The MPR tool assessed the semicircular canals in three reference planes. Initially, the axial plane in which the Lateral Semicircular Canal (LSC) should show a typical “signet ring appearance”. Then, investigators searched for the plane of the Anterior Semicircular Canal (ASC, as described by Pöschl) and the Posterior Semicircular Canal (PSC) with common crus.

The next challenge was to identify and describe the cochlea and vestibule. At first, the longitudinal section of the cochlea was set to assess the basal, middle, and apical turns around the modiolus. Afterward, the adjusted axial plane was set at an angle that allowed for the evaluation of the modiolus and basal turn of the cochlea. Subsequently, a perpendicular plane was defined concerning the modified axial plane, enabling the visualization of the modiolus and all three cochlea turns. However, due to the severity of malformations in analyzed patients, the reference points for determining such planes were disturbed, making the analysis much more difficult.

Internal Acoustic Canal (IAC) diameter was assessed in an axial section of CT by performing measurements at the level of the internal acoustic pore. Firstly, a longitudinal axis line was drawn, and then the IAC diameter was measured as a line drawn perpendicularly to the longitudinal axis of the IAC.

For better visualization and understanding of found malformations in the inner ear in analyzed patients with CHARGE syndrome, examples are shown in Figs. 1, 2 and 3.

Figure 1.

Cochlear malformations identified in computed tomography evaluation and presented in the planes selected for the most precise visualization. Each panel represents a particular patient: (a) Case #2 right ear, cochlear hypoplasia type II; (b) Case #1 right ear, cochlear hypoplasia type III; (c) Case #10 right ear, (c1) Cochlear Aplasia with a Dilated Vestibule (CADV); (c2) Rudimentary otocyst. White arrows indicate anomaly location. Individual cases are supplemented with ASSR reconstruction of audiograms (hearing thresholds in dBnHL), respectively.

(0.58MB).

Figure 2.

Semicircular canals and vestibule malformations identified in computed tomography evaluation and presented in dedicated planes. Panel a and b) represent the same patient (case #2) with a hypoplastic vestibule and the absence of all three semicircular canals. Panel c (case #10) shows a dysmorphic superior semicircular canal with an accompanying dilated vestibule.

(0.17MB).

Figure 3.

Exemplary measurements of the Internal Auditory Canal (IAC) diameter in an axial section of a computed tomography made at the internal acoustic opening (porus acusticus internus), the IAC diameter (solid line) was determined perpendicular to the longitudinal axis (dashed line) of the IAC; Panel a – narrowed IAC with diameter 2.70 mm (case #2); Panel b ‒ normal sized IAC with diameter 3.76 mm (case #1). Cochlea (C) and Vestibule (V) are marked with a line art over for easier identification.

(0.22MB).

The assessment of the patient's hearing was carried out in a way suitable for their developmental stage and age. Audiological test results of hearing thresholds (ASSR) were available for analysis in 9 patients (18 ears) and were performed at the age between 3 months to 2.5 years. The results confirmed hearing loss. ASSR results for 500, 1000, 2000, and 4000 Hz were analyzed for the purposes of this study. In ASSR, if the threshold for a particular frequency was absent for the highest intensity level tested (120 dBnHL result as no response), its value was set as 130 dBnHL to facilitate the analyses.

Statistical analysis was conducted in the STATISTICA program (TIBCO Software Inc. 2017, version 13.3). The data were tested for normality, parametric, and non-parametric criteria. Detailed statistical analysis was performed using Spearman's correlation test. The level of statistical significance was set at p = 0.05.

The structure of the cochlea was considered in terms of the basal turn and then jointly in terms of the middle and apical turns. Moreover, the absence or presence of modiolus was assessed. Then, the findings were summarily classified to a specific malformation type according to the newest Senneroglu classification.13 The basal turn was diagnosed as normal in 90% of cases. However, one patient with cochlear aplasia and rudimentary otocysts was also identified. More variety in structure was found in the middle and apical turns. Besides cochlear aplasia and rudimentary otocyst, the following abnormalities were discovered, which are essential concerning the entire bony labyrinth. Respectively: 2 cases (10%) of dysmorphic (cystic) middle and apical turns, 12 cases (60%) of shortened (less than two turns), and 4 cases (20%) of normal, fully developed cochlea. Modiolus was absent in 20% of cases. In the rest of the cases, the modiolus was considered present. Cochlear anomalies identified in analyzed patients are presented in Table 1, and examples are shown in Fig. 1.

Lateral semicircular canal defects were the most common anomaly in analyzed patients with CHARGE syndrome, affecting 100% of cases. The second most common anomaly concerned the absence of the posterior semicircular canal in all temporal bones except one, where the canal was preserved in its dysmorphic form (5%). The superior semicircular canal was affected by a deep anomaly to a slightly lesser extent. Its absence was observed in 13 temporal bones (65%). In 6, it was hypoplastic (30%), and in 1 case – dysmorphic (5%). However, the vestibule itself was normal in 6 cases (30%), hypoplastic in 12 cases (60%) and absent in one case (5%). Also, the vestibule was dilated in one case, which coexisted with cochlear aplasia. Semicircular canals defects and vestibular anomalies identified in analyzed patients are presented in Table 1, and examples are shown in Fig. 2.

Cochlear Aplasia with a Dilated Vestibule (CADV) and rudimentary otocyst was accompanied by the absence of IAC, which for statistical analysis purposes in this study was considered 0.00 mm in diameter. In the remaining cases, the IAC diameter ranged from a minimum dimension of 2.27 mm–6.08 mm. The median IAC diameter of all ears was 3.41 mm, and the Standard Deviation (SD) was 1.37 mm. IAC measurement examples are shown in Fig. 3.

In the final stage of the CT investigation, the analysis of individual elements with a summary aimed at assigning specific malformations to each case according to the Senneroglu classification13 was carried out. Cochlear Aplasia with a Dilated Vestibule (CADV) and rudimentary otocyst (Fig. 1c) was seen in both ears in one patient (10% of the analyzed ears, patient #10). It was also related to the almost complete aplasia of the semicircular canals. Only in the right ear was a structure that could be regarded as a remnant of the dysplastic superior semicircular canal (Fig. 2c). Cochlear hypoplasia type II (Fig. 1a) was also seen in both ears in one patient (10% of the analyzed ears, patient #2). Cochlear hypoplasia type III (Fig. 1b) was the most common malformation found in 12 ears (60%). In 4 cases, no cochlear anomaly was identified. Still, in half of them, normal cochlea was accompanied by isolated semicircular canals anomaly and, in two other cases, by a hypoplastic vestibule. In two patients, the defects were not symmetrical in both ears. In both of them, the right ear was more severely affected with noticeable cochlear hypoplasia type III, while the left ear showed only isolated vestibular labyrinth anomaly.

In addition, some correlations were observed between the severity of malformations of individual structures. Firstly, the correlation between the IAC diameter and the severity of vestibular malformations was statistically significant. The more severe the vestibule malformation, the smaller the IAC diameter (p = 0.013). In addition, the severity of vestibular defect significantly correlated with the severity of hypoplasia of the basal turn of the cochlea (p = 0.0057). Moreover, the correlation occurred in the combination of vestibular malformation with the presence or absence of modiolus. Its absence correlated with the vestibular malformation (p = 0.0125).

Audiological results were unavailable for one patient (patient #6). The remaining nine patients underwent audiological evaluation. The results are shown in Table 2. The ASSR results have been converted to reconstructed audiograms for better understanding and visualization. In the ASSR, there was a statistically significant correlation between the severity of the inner ear malformation and the profoundness of the hearing loss for all of the frequencies, with p-values ranging from 0.0001 to 0.0432. The most advanced malformations, such as Cochlear Aplasia with a Dilated Vestibule (CADV) and rudimentary otocyst, showed non-response, which is understandable considering not only the condition of the cochlea (absence) but the absence of IAC on both sides as well. Subsequently, type II hypoplasia correlated with worse hearing level outcomes than type III hypoplasia. Ultimately, patients with the best-developed cochlea had the best audiological results. When inspecting the structure of the cochlea more deeply, taking into consideration the state of the basal, middle, and apical turns and the presence or absence of the modiolus, it was also observed that the better development of these structures, the hearing thresholds become lower (p-value ranged from 0.0000 to 0.0432). A significant correlation was also observed for IAC morphology, where absence, narrowing, or normal size correlated statistically significantly with audiological results (p-value ranged from 0.0001 to 0.0002). Moreover, a significant correlation with the ASSR scores was observed regarding the exact dimension of the IAC (p-value ranged from 0.0076 to 0.0288). The results of the statistical analysis are presented in Table 3.