Background
Sudden sensorineural hearing loss is an unexplained acute sensorineural hearing loss of greater than 30 decibels (dB) over a minimum of three pure tone audiometric averaged threshold test frequencies, occurring within 72 hours of onset.Reference Chandrasekhar, Tsai Do, Schwartz, Bontempo, Faucett and Finestone1 Proposed risk factors for sudden sensorineural hearing loss include advanced age, low folate levels, a diet low in vegetables, and metabolic syndrome.Reference Cadoni, Agostino, Scipione and Galli2–Reference Lam, Bao, Hua and Sommer4 Spontaneous hearing recovery occurs in 32–65 per cent of sudden sensorineural hearing loss patients,Reference Byl5–Reference Nakashima, Sato, Gyo, Hato, Yoshida and Shimono9 and is affected by a variety of factors including hearing loss severity, patient age and concomitant vestibular symptoms.Reference Chandrasekhar, Tsai Do, Schwartz, Bontempo, Faucett and Finestone1,Reference Byl5 Most cases are idiopathic, with only 10–15 per cent of cases having an identifiable cause.Reference Chandrasekhar, Tsai Do, Schwartz, Bontempo, Faucett and Finestone1 Cochlea ischemia or viral infection are the two predominant proposed pathogenetic mechanisms in idiopathic sudden sensorineural hearing loss.Reference Yamada, Kita, Shinmura, Nakamura, Sahara and Misawa10 Studies of plasma and serum microRNAs in sensorineural hearing loss patients have identified known oxidative stress related microRNAs and others that target oxidative stress related biological pathways.Reference Safabakhsh, Wijesinghe, Nunez and Nunez11–Reference Nunez, Wijesinghe, Nabi, Yeh and Garnis13 This pathogenic mechanism is supported by the efficacy of hyperbaric oxygen therapy in sudden sensorineural hearing loss patients.Reference Joshua, Ayub, Wijesinghe and Nunez12 Interestingly, many of these microRNAs are predominant in the nervous system, suggesting that pathogenetic lesions in sudden sensorineural hearing loss may be localised to the auditory nerve rather than the cochlea in some patients.Reference Nunez, Wijesinghe, Nabi, Yeh and Garnis13 Therefore, we adapted an established hearing-loss categorisation system to distinguish neural hearing loss patients from non-neural hearing loss patients, and to determine if the two groups differ in functional recovery.
Schucknecht and Gacek described five types of age-related hearing loss that correlate with audiometric patterns of hearing loss: sensory, neural, strial, cochlear conductive and other.Reference Schuknecht and Gacek14 The neural type may present with any kind of pure tone average (PTA) slope; however, it is characterised by a speech reception threshold elevation that is greater than the PTA elevation by at least 15 dB.Reference Schuknecht and Gacek14 On the other hand, the Sagers et al.Reference Sagers, Landegger, Worthington, Nadol and Stankovic15 study of primary neuronal degeneration concluded that PTA and word recognition scores decreased to the same degree with progressive neuronal loss.
We elected to adopt an audiometric characterisation of neural type hearing loss more aligned with Schucknecht and Gacek'sReference Schuknecht and Gacek14 description, with the aim of only classifying patients with the most severe audiometric features of neuronal degeneration as having a neural pattern of hearing loss. We used this model to investigate differences in hearing recovery, as determined by word recognition score and PTA4, between neural-type sudden sensorineural hearing loss and non-neural sudden sensorineural hearing loss patients. The 2019 American Academy of Otolaryngology – Head and Neck Surgery (AAO–HNS) hearing recovery criteria were used.
Methods
Study design and participants
This study was approved by the University of British Columbia's Clinical Research Ethics Board (H18-00736). A retrospective chart review of patients who presented with sudden sensorineural hearing loss to clinics at Vancouver General Hospital, a tertiary care urban hospital, from 1 November 2013 to 30 June 2019 was performed. The records of Vancouver General Hospital, Division of Otolaryngology–Head and Neck Surgery Urgent Access clinics were searched electronically using International Classification of Diseases 9 code 388.2 (sudden hearing loss, unspecified). Additionally, a hand search of hardcopy diagnostic record sheets of patients who attended the Vancouver General Hospital Division of Otolaryngology–Head and Neck Surgery Urgent Access clinic, and Vancouver General Hospital Audiology Department using key terms “sudden sensorineural hearing loss” or “sudden deafness” or “decrease in hearing” was performed. Prospectively collected data of sudden sensorineural hearing loss patients who attended the Vancouver General Hospital subspecialist Otology and Neurotology clinic and who were willing to participate in sudden sensorineural hearing loss studies were also studied.
Patients aged 19 years or older with hearing loss greater than 30 dB over three contiguous frequencies within 72 hours of onset were included in the study. Attending staff otolaryngologists or residents under their supervision made all sudden sensorineural hearing loss diagnoses. Patients with insufficient presenting or follow-up audiometric data or with hearing loss and identifiable causes such as MRI-confirmed acoustic neuroma, Meniere's disease, familial hearing loss, autoimmune hearing loss, ototoxic medication, infectious and/or inflammatory middle- and/or inner-ear disease were excluded.
Parameters recorded included patient age, sex, history of smoking and alcohol usage, coexisting audiovestibular complaints (e.g. vertigo), medical comorbidities (e.g. diabetes, hypertension, and dyslipidemia), and treatment type (oral steroid, intratympanic steroid, hyperbaric oxygen therapy, and combinations of these treatments). The audiometric data recorded consisted of initial and final follow-up bone- and air-conduction pure tone audiometric thresholds at 0.5 kHz, 1, 2, 3, 4, 6 and 8.0 kHz and word recognition score. A pure tone average (PTA4) score was calculated from the mean threshold across four frequencies: 0.5, 1, 2, and (3 or 4) kHz. A case was categorised as a neural type sudden sensorineural hearing loss when: (1) the initial word recognition score was less than 60 per cent and (2) the magnitude of the word recognition score reduction was 20 per cent greater than expected based on the PTA4.
Evaluation of hearing outcomes
Initial and final post-treatment PTA4 and word recognition scores were used to assess hearing outcome. Patients’ hearing outcomes were categorised based on the AAO–HNS 2019 Sudden Hearing Loss Clinical Practice Guideline criteria: “completely recovered” (return to within 10 dB of the unaffected ear and recovery of word recognition score to within 5 per cent to 10 per cent of the unaffected ear); “partially recovered” (an improvement of ≥ 10 dB in pure tone thresholds but not complete recovery); or “not recovered” (less than 10 dB improvement).Reference Chandrasekhar, Tsai Do, Schwartz, Bontempo, Faucett and Finestone1,Reference Alexander and Harris6
Statistical analysis
Neural and non-neural inter-group mean age, initial and final PTA4, initial and final word recognition score, PTA4 change, and word recognition score changes were statistically compared by independent samples t-test. The inter-group proportions of patients who recovered hearing, received combination treatment and the sex distribution of patients were compared with Pearson's chi square test or Fisher's exact test. Statistical analyses were conducted with the Statistical Package for the Social Sciences (SPSS) version 25.0 (IBM, Armonk, NY, USA). A p value of less than 0.05 was considered as significant. A Bonferroni corrected p value of less than 0.0071 was considered significant for multiple t tests. Box and whisker plots were generated using GraphPad Prism version 9.0 (GraphPad Software, La Jolla, California, United States).
Results
Eighty patients of the 132 cases that were identified and reviewed for eligibility, were excluded due to incomplete initial and/or final audiogram data. The remaining 52 patients were reviewed for exclusion criteria: three patients were excluded for concomitant use of ototoxic medications, and one patient was excluded due to a history of inflammatory middle-ear disease. Forty-eight patients who satisfied the inclusion criteria were studied (Figure 1), 12 of whom were classified as neural and 36 as non-neural sudden sensorineural hearing loss cases. The mean ages standard deviation (± SD) of the neural and non-neural type patients 57.7 ± 14.9 and 55.3 ± 15.2 years respectively, were statistically similar. The male:female ratios in neural and non-neural type patients (7:5 and 4:5, respectively) were not statistically different (Table 1). The initial PTA4 in the affected ears in neural and non-neural type patients (69.6 ± 15.8 dB and 54.2 ± 29.4 dB, respectively) were statistically similar (Table 2). The initial word recognition scores (per cent) (± SD: 17.1 ± 17.6 and 71.5 ± 35.5 in neural and non-neural type patients, respectively) were significantly different (p < 0.0001, t-test) (Table 2).
Figure 1. Flow diagram of patient charts that were included and excluded in our analysis. Forty-eight of 132 patients were included after being reviewed for eligibility. Eighty patients were excluded for missing audiogram data, and upon review of exclusion criteria, three additional patients were excluded for ototoxic medications and one for inflammatory middle-ear disease. Twelve patients were classified into the neural SSNHL (sudden sensorineural hearing loss) category, and 36 patients were classified into the non-neural SSNHL category.
Table 1. Summary of neural and non-neural type SSNHL (sudden sensorineural hearing loss) patient's demographic, medical history, and treatment details. IT = intra-tympanic; HBO = hyperbaric oxygen; NA = not statistically analysed
Table 2. Summary of audiometric data analyses by SSNHL (sudden sensorineural hearing loss) type (neural and non-neural types compared). Significant p values after Bonferroni correction are indicated by *. PTA = pure tone average; SD = standard deviation; WRS = word recognition score
There was no statistically significant inter-group difference in the proportions of patients who received combination oral prednisone and intratympanic dexamethasone: 16.7 per cent (2/12) and 11.1 per cent (4/36) of the neural and non-neural sudden sensorineural hearing loss patients, respectively. Likewise, there was no statistically significant difference in the 50.0 per cent (6/12) proportion of neural sudden sensorineural hearing loss patients and 27.8 per cent (10/36) proportion of non-neural sudden sensorineural hearing loss patients who received combination steroid and hyperbaric oxygen therapy (Table 1). Additional patient treatment details medical comorbidities and lifestyle factors are summarised in Table 1.
The hearing recovery rates 66.7 per cent (8/12) and 66.7 per cent (24/36) (p = 1.0, Fisher's exact test, Table 1), final PTA4 44.3 ± 22.5 versus 44.3 ± 27.9 (p = 0.998, t-test, Table 2) and final word recognition score scores 60.3 ± 30.6 versus 68.1 ± 37.7 (p = 0.538, t-test, Table 2) in neural and non-neural types of sudden sensorineural hearing loss respectively, were similar. However, the affected ear's word recognition score change (per cent) ± SD with treatment: 46.9 ± 29.8 and 3.2 ± 25.8 (p <0.0001, t-test), in the neural and non-neural sudden sensorineural hearing loss groups, respectively, were significantly different (Table 2, Figure 2).
Figure 2. Box and whisker plots illustrate the neural and non-neural type SSNHL (sudden sensorineural hearing loss) patients’ audiometric data before and after treatment. The lower quartile (Q1) or 25th percentile of the dataset forms the lower margin of the box; the median (Q2) or 50th percentile of the dataset is illustrated as a line within the box; the upper quartile (Q3) or 75th percentile of the dataset is represented by the upper margin of the box. The inter-quartile range is the distance from Q1 to Q3. Minimum and maximum dataset values were used to create the whiskers, and the individual patient scores shown as points are superimposed in each plot. Actual p values are illustrated in each panel; a Bonferroni corrected p value < 0.0071 was considered significant. PTA = pure tone average; WRS = word recognition score.
Discussion
The proportion of patients experiencing hearing recovery did not vary significantly between the neural and non-neural sudden sensorineural hearing loss groups. No significant difference was seen in the initial PTA4, PTA4 gain, final word recognition score and final PTA4 between the neural and non-neural sudden sensorineural hearing loss patients. However, neural sudden sensorineural hearing loss patients had significantly poorer initial word recognition scores and demonstrated greater word recognition score gains in the affected ear with treatment compared to non-neural sudden sensorineural hearing loss patients.
The cochlea is generally thought to be the site of the pathological changes resulting in hearing loss in sudden sensorineural hearing loss patients, although different mechanisms are proposed.Reference Yamada, Kita, Shinmura, Nakamura, Sahara and Misawa10 This contrasts with sudden onset vertigo, where different sites of pathology are accepted as possible causes for the vertigo. Specifically, the vestibular nerve is accepted as a possible site of disease, as reflected by the diagnosis of vestibular neuritis, while the labyrinth is the presumed site when vertigo is accompanied by sensorineural hearing loss, as in Meniere's disease. In sudden sensorineural hearing loss there are no widely accepted guidelines for distinguishing neural (auditory nerve) sudden sensorineural hearing loss from sensory (cochlea specific) sudden sensorineural hearing loss. We used Schucknecht and Gacek'sReference Schuknecht and Gacek14 criteria for defining neural and non-neural types of sudden sensorineural hearing loss and observed a difference in mean word recognition score at presentation consistent with the application of our diagnostic criteria.
Surprisingly, the neural sudden sensorineural hearing loss patients demonstrated greater improvement in word recognition scores with treatment than the non-neural patients. The groups did not vary by age, sex or severity of PTA4 on presentation. There were too few cases with vertigo to allow robust inter-group statistical comparison (2 of 12 (16.7 per cent) and 9 of 36 (25 per cent) of patients in the neural and non-neural groups, respectively). Analysis by treatment received likewise did not suggest a significant difference in the neural and non-neural sudden sensorineural hearing loss patients. Importantly, the proportion of neural sudden sensorineural hearing loss (50 per cent) who received triple therapy was not significantly different from the proportion of non-neural patients (27.8 per cent) undergoing the same (Table 1). Therefore, it is not expected that an inter-group difference in the treatment effect of hyperbaric oxygen will be discerned in this study, although the benefit of hyperbaric oxygen on PTA threshold improvement is documented.Reference Joshua, Ayub, Wijesinghe and Nunez12
This study found a difference in word recognition score recovery based on the pathological site of lesion in sudden sensorineural hearing loss patients which has not been described before. Chang et al.,Reference Chang, Ho and Kuo16 support the importance of classifying audiological patterns of presenting hearing loss in sudden sensorineural hearing loss as these appear to determine prognosis. However, their classification system is based only on the PTA and does not include speech discrimination findings.Reference Chang, Ho and Kuo16 The current study emphasises the need to classify initial patterns of hearing loss comprehensively using both PTA and word recognition score.
• The literature suggests that the pathological lesion in sudden sensorineural hearing loss is located in the cochlea
• This research aimed to determine if there is a subset of patients where the pathological focus is on the auditory nerve, and if the two groups differ in recovery
• Whereas pure tone audiometry is a commonly used audiometric finding to measure hearing loss outcomes, a novel approach was used to classifying sudden sensorineural hearing loss by incorporating word recognition scores
• Patients with a neural type of hearing loss demonstrated greater word recognition score recovery after treatment than those in the sensory group
In common with other retrospective studies, missing and incomplete data documentation is a limitation. Forty-eight patients were included in the study after reviewing 132 case records for eligibility. Eighty cases were excluded due to missing data, which reduced the number of patients available for analysis, an issue that is best rectified with a prospective study design. The small sample sizes of the neural and non-neural subgroups precluded robust comparison of sub-group patient characteristics (e.g. vertigo) known to influence hearing prognosis.
Conclusion
This study utilised audiometric findings at presentation in sudden sensorineural hearing loss patients to classify hearing losses as neural or non-neural. Our hypothesis that patients with a neural pattern of hearing loss are as likely to recover as those with a non-neural pattern was supported. However, sudden sensorineural hearing loss patients with a neural type hearing loss demonstrated greater word recognition score gains in response to treatment than those with a sensory type of loss. Future studies should prospectively investigate the differences in hearing outcomes in the neural type of sudden sensorineural hearing loss, including how the presence of features such as severity of hearing loss, age at presentation, comorbidities, dose and timing of treatment, and associated vertigo affect prognosis.
Ethics approval
Ethics approval was obtained from the University of British Columbia, Clinical Research Ethics Board.
Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Funding and competing interests
The authors declare that they received no funding and have no competing interests.
Contribution of authors
conception and design: RX, TJ, DN, PW; acquisition of data: TJ, RX, PW, AA, ML, RG, DN; statistical analysis and interpretation of data: PW, RG, DN; drafting of manuscript: RX, PW, TJ, AA, ML, RG; revision of manuscript for intellectual content: DN, PW, RG, RX, TJ; final revision and manuscript approval: RG, RX, PW, TJ, AA, ML, DN; supervision: DN.
Acknowledgements
We thank Dr. Li Qi, Dr. Joshua Gurberg, Dr. Lisa Ying, Dr. John Lee, Dr. Ameen Amanian, and Dr. Jennifer Ham for helpful discussions and assistance with data collection.
Open access
