Clinical performance, safety, and patient-reported outcomes of an active osseointegrated bone-conduction hearing implant system at 24-month follow-up

This study collected 12- and 24-month follow-up data from the enrolled group of Osia System recipients, who had existing data from a previous study, at pre-implantation and 6 months post-implantation [1]. The current study was approved by respective local ethics committees as per local regulations, conducted in accordance with the Declaration of Helsinki [6] and ISO14155:2011[7] and was registered on ClinicalTrials.gov with identifier NCT04754477.

The investigational device was the Cochlear™ Osia 2 System (Cochlear Ltd., Sydney, Australia) consisting of an external sound processor (SP) (Osia 2 SP) magnetically retained on the skin over the site of an internal implant (OSI200 Implant), which is fixed to the temporal bone with an osseointegrating implant (BI300 screw fixture, 3 or 4 mm). Surgical technique has been published previously [1] and the SP was individually fitted to each subject's hearing loss using Osia Fitting Software 2.0.

The objective of the clinical study was to compare safety parameters, patient-reported outcomes (PRO), and change in free-field hearing thresholds and speech reception threshold (SRT) in noise (dB SNR) with the Osia 2 System at 6 months and 12 and 24 months post-implantation. As part of the current study, visits were carried out at 12 and 24 months after implantation with the Osia 2 System for efficacy and safety evaluation. Data collected between baseline (pre-implantation) and 6-month follow-up post-implantation with the Osia 2 System were retrospectively collected from the previous clinical study sponsored by Cochlear (ClinicalTrials.gov NCT04041700). Safety parameters were recorded throughout the current clinical study and retrospectively collected for enrolled patients in the previous clinical study between implantation with the Osia 2 System and 6-month follow-up.

The tests were performed in a sound-insulated audiometric booth using calibrated equipment with the non-test ear blocked using both a foam plug and earmuffs in case of normal or near-normal hearing or a large asymmetry between ears. During testing, SPs were set to fixed directionality mode. Threshold audiometry was performed using warble tones presented via a loudspeaker located 1 m in front of the subjects at 0-degree azimuth, with 1 m of free space surrounding the test subject. PTA4 hearing thresholds were calculated by taking an average of hearing thresholds measured at 0.5, 1, 2, and 4 kHz. Changes in audiometric thresholds of > 10 dB at a single frequency are considered as clinically relevant as outside of the test–retest variability [8].

For speech in noise testing, both speech and noise were presented in free field at 0-degree azimuth (front). In Melbourne and Sydney, the AuSTIN test was administered and sentences were presented at a constant level of 65 dB SPL throughout the test, and babble noise was adapted stepwise according to the software used to establish the signal-to-noise ratio (SNR) providing a 50% level of correctly repeated morphemes. In Hong Kong, the CHINT methodology was used, where noise was maintained at a constant 65 dB SPL, and sentence speech stimuli were presented adaptatively, with stepwise adjustments to the speech level made by the software to establish the SNR where the test subject repeated 50% of the material correctly. Improvements of greater than 3 dB SNR are considered to be clinically relevant [9].

Patient-reported outcome measures were collected using validated questionnaires: the Health Utilities Index (HUI-3) [10, 11], the Abbreviated Profile of Hearing Aid Benefit (APHAB) [12, 13], and the Speech, Spatial and Qualities of Hearing Scale 12 (SSQ12) [14, 15]. HUI-3 evaluates eight health-related quality of life (QoL) dimensions (vision, hearing, speech, walking/mobility, dexterity, self-care, emotion, and cognition) and a comprehensive health state attribute. A change in global HUI-3 score of 0.03 or higher is considered clinically relevant [16]. APHAB is a hearing-related PRO instrument, which includes four subscales (ease of communication, reverberation, and background noise, aversiveness) and a global score. A change in score of higher than 10 for global score is generally regarded as clinically relevant [17]. SSQ12 is a short (12-item) version of the original 49-item SSQ questionnaire [18] that measures the self-reported auditory disability in everyday life across three subdomains (speech, spatial, and qualities of hearing). Changes of 1.0 unit or greater on SSQ subscales indicate a clinically relevant change [18]. Data between baseline and 6 months post-implantation were retrospectively collected from the previous clinical study. PRO at 12 and 24 months post-implantation were prospectively collected as part of the current study.

Patient-reported daily usage, wearing comfort, and retention were collected at all study visits after activation. Daily use communicated by the patient was reported as the average hours of daily SP use during the period preceding the visit. Comfort and retention were assessed subjectively by indicating on a visual analog scale consisting of a straight line running from 0 to 100 mm (0 mm indicated no comfort at all/insufficient retention, and 100 mm indicated the most comfortable situation imaginable/excellent retention).

Analyses were performed on all subjects enrolled in the current study. All statistical analyses were paired and nonparametric and the Friedman’s test was used to test for paired observations. All significance tests were two-tailed and performed at the 0.05 significance level. All significance values were adjusted by the Bonferroni correction for multiple tests. Statistical analysis was performed in SAS version 9.4 (SAS Institute Inc.) and IBM SPSS Statistics, version 28 (Armonk, NY: IBM Corp.). Demographics, baseline characteristics, daily use, comfort, adverse events, and device deficiencies are presented descriptively.

Between 6- and 24-month follow-up, there were nine adverse events in six subjects that were either possibly, probably or definitely related to the device or procedure. Three of these events (pain behind implant, skin irritation, and prominence of the posterior inferior edge of system) were related to both device and procedure and six were related to the device only (discomfort from SP heating up, increased tinnitus, two reports of frustration, and two reports of non-use). The two reports of non-use were classified as moderate in severity and the remaining events were all classified as mild in severity. Figure 1 presents the cumulative adverse event rate for all related adverse events from implantation to 24-month follow-up. The majority of related adverse events occurred within the first 3 months post-implantation and the rate decreased between 3 and 24 months. All but three adverse events (two reports of non-use and one report of frustration) related to device or procedure reported in two subjects were resolved by study end.

Mean change in pure-tone audiometry four frequency average (PTA4) from 6-months to 24-months follow-up was − 3.7 dB HL (SD: 5.2 dB HL, range − 5.0 to 13.8. dB HL), while the mean change in PTA4 from 6-months to the 12-month follow-up period was − 2.2 dB HL (SD: 5.1 dB HL, range − 6.3 to 13.8. dB HL). There were no statistically significant differences in paired comparisons between the 6-month follow-up and 24-month follow-up (all p values > 0.05), indicating stable improvements up to 24-month follow-up with the investigational device for subjects with MHL/CHL and those with SSD (Fig. 2). There were also statistically significant (all p-values < 0.001) improvements in PTA4 hearing thresholds between pre-implantation and all subsequent aided follow-up measurements. At the 12-month follow-up, the mean improvement in PTA4 from the unaided situation was 26.7 dB HL (SD: 7.0 dB HL, range − 43.8 to − 13.8 dB HL). At the 24-month follow-up, the mean improvement in PTA4 from the unaided situation was 25.5 dB HL (SD: 6.5 dB HL, range − 37.5 to − 12.5 dB HL).

For the adaptive speech recognition in noise test, there was no statistically significant difference when comparing the investigational device at the 12- and 24-month follow-up to the aided situation at the last performed assessment 6 months after surgery in the previous clinical study (all p values > 0.05). Mean change in SRT in noise from 6 months to the 24-month follow-up visit was -0.244 dB SNR (SD: 0.913 dB SNR, range − 1.7 to 1.4 dB SNR). While the mean change in SNR from 6-months to the 12-month follow-up period was − 0.106 dB SNR (SD: 2.726 dB SNR, range − 3.8 to 9.2 dB SNR). These results indicate stable improvements for speech recognition up to 24-month follow-up for subjects with MHL/CHL and SSD (Fig. 3). Compared to the unaided situation, the SRT in noise improved by 8.7 dB SNR (SD 7.2 dB, range − 22.5 to 0.0 dB) when aided with the Osia System at the 12-month follow-up. At the 24-month follow-up, the average improvement with the Osia System was 8.7 dB (SD: 7.2 dB, range − 21.3 to − 1.7 dB) compared to the unaided situation at baseline.

The acceptable low rate of adverse events reported in this study is also in line with a recent evidence synthesis of adverse event rates for active bone-conduction hearing implant systems, including the Osia System [19]. Events reported during the 24-month follow-up (N = 19) were classified as mild or moderate. It was apparent that the rate of adverse events decreased markedly after the initial 3–6-month follow-up period post-implantation and that this continued to decrease with time. Rates of adverse events in studies with limited follow-up (i.e., ≤ 6 months) may therefore not be useful to model complication rates over the longer term and may lead to an overestimation of complication rates. These data can be used to inform counseling and future cost-effectiveness modeling with active bone-conduction hearing implants. Regarding the two subjects who had stopped using their device regularly, one retired subject with SSD did not feel the need for the Osia device due to social isolation during COVID-19 restrictions. Non-compliance rates with bone-conduction hearing implants are typically higher in subjects with SSD compared to those with MHL/CHL [20]. The other non-user with CHL/MHL was unhappy with the sound quality of the device after fitting and remained a non-user, highlighting the importance of patient-centered counseling, rehabilitation, and managing post-operative expectations.

The extended follow-up data confirmed that most patients continue using their Osia System and remain satisfied with the device. Importantly, the clinically relevant improvements in audiometric thresholds (including at higher frequencies), speech reception thresholds in noise, and health-related QoL reported in the previous study [1] remained stable during this 24-month follow-up study. These findings are echoed in an additional study in which Osia System recipients were followed for at least 12 months [21]. Results of the SSQ and HUI questionnaires in this study improved with longer Osia System experience and aided hearing thresholds and speech recognition in quiet were also significantly improved in comparison to the unaided situation [21]. The clinical benefits experienced with the Osia System soon after fitting are therefore expected to remain over the longer term.

The strength of this study is that it overcomes a previous limitation regarding the paucity of longer term follow-up data with the Osia System. The study was also focused on patient safety and had rigorous safety follow-up, ensuring that all adverse events were captured. Regarding limitations, the study cohort is relatively small, and the number of subjects with SSD prevented a sub-analysis being performed for this subgroup. However, there are published data available demonstrating the safety and efficacy of the Osia System in this population [2]. As seen for subjects with MHL/CHL, we would also expect the benefits observed after short-term observation to remain stable for those with SSD, as observed for a small cohort of subjects with SSD and extended follow-up [21]. Studies are now emerging that compare outcomes for hearing-impaired subjects implanted with active and passive bone-conduction systems [22]. Additional studies with larger patient numbers are required to truly capture the benefits of the system in comparison with other bone-conduction hearing implant systems.