For cochlear implant (CI) surgery, an increasing number of electrode array designs have been established over the last few decades. In addition, the indication for implantation has been expanded to include patients with residual hearing, especially in low frequencies. Therefore, in addition to years of deafness, age at implantation, scalar position, and neurocognition, other variables influencing the outcome need to be considered (1–4).

The focus of cochlear implant surgeons is on improving outcomes in terms of reducing surgical trauma with the aim of preserving residual hearing. It is well-known that cochlear anatomy varies between patients, requiring the use of different electrode arrays of varying length and design, distinguishing between straight and curved, resulting in individual electrode selection (5, 6). For this reason, the standard preoperative procedure for CI candidates includes either a cone beam computed tomography scan or a computed tomography scan. These imaging techniques enable the surgeon to identify individual variations in temporal bone anatomy and to measure the cochlea, as described in several studies (5, 7, 8). The customization of the electrode array, based on cochlear anatomy, is of increasing interest to manufacturers.

Primary insertion into the scala tympani has been shown to positively influence speech discrimination (1, 4, 9, 10). Perimodiolar electrode arrays, such as the Contour Advance (CA) electrode array developed by Cochlear™, demonstrated dislocation rates of 15.4% and scala vestibuli insertion of 16.1% in the study of Ketterer et al. (11). Previous studies have reported scala tympani dislocation rates ranging from 10 to 38.7%, with the higher rates including cases of scala vestibuli insertions at implantation (19 out of 49 ears implanted with the CA) (12, 13). On the contrary, in 2017, Aschendorff et al. described the slim modiolar electrode array (SMA) as being slim and atraumatic. Ketterer et al. also described very low dislocation rates and a small risk of intracochlear trauma (11, 14). In addition, Aschendorff et al. and Beck et al. described the SMA as a perimodiolar electrode array with a higher risk of tip fold overs, which can be reduced by surgical experience (14, 15). When comparing the electrode arrays of MED-EL, Ketterer et al. found that the risk of scalar dislocation is lower with the short electrode arrays Flex²⁴ and Flex²⁶ than with the longer electrodes Flex²⁸ and Flex^Soft (31.5 mm) (6). Furthermore, some studies show that especially short straight arrays have lower dislocation rates than stiffer perimodiolar arrays as the CA (12, 16).

This study aimed to compare audiological residual hearing preservation following CI depending on the implanted electrode array, in particular the difference between straight and perimodiolar electrode array design. We aimed to determine the electrode array design with the least cochlear trauma and thus the best audiological performance regarding residual hearing, and possibly to provide a recommendation for the selection of an electrode array for patients with existing residual hearing.

Due to further development of electrode arrays with different shapes and lengths as well as improved surgical techniques, CI surgery focuses on reducing intracochlear trauma and preserving residual hearing. As the indication criteria for CI have expanded over the past decade, more patients with existing residual hearing continue to be implanted. This increases the need for an atraumatic electrode array that enables the preservation of the patient’s residual hearing.

Our study shows that the electrode array design significantly influences the preservation of residual hearing, both in low-frequency PTA, i.e., PTAlow, and standard PTA4. The electrode arrays associated with the least cochlear trauma and the most favorable performance in terms of preserving residual hearing were the SSA of Cochlear™ and the Flex²⁴ of MED-EL. These straight lateral wall electrode arrays are characterized by a slim design and a short length. Consistent with our findings, previous studies have reported higher rates of hearing preservation with straight and mid-scala electrode arrays (e.g., precurved array from AB) compared to perimodiolar electrode arrays from Cochlear™ (18, 19). Nevertheless, Sweeney et al. included only 16 cases (18). Perkins et al. found a deterioration in low-frequency PTA when comparing patients implanted with precurved electrode arrays from the time of activation to the final follow-up (20). However, they suggested a selection bias, with surgeons favoring straight electrode arrays for patients with higher residual hearing or lower audiometric thresholds (20). This proposal can be validated within the scope of our study as well, as the manufacturer was chosen by the patients, while the specific electrode array was selected intraoperatively by the surgeon. These findings are consistent with another study that proposed that the amount of residual hearing is related to its preservation over the long term (21). The CA accounted for the largest number of included ears in our investigation; therefore, the results regarding residual hearing preservation are robust. The high level of residual hearing loss associated with the CA may be a consequence of traumatic surgical approach, typically cochleostomy, which can trigger fibrotic remodeling and autoimmune processes, leading to scarring. As described in previous studies, this fibro-osseous reaction can cause delayed loss of residual hearing (22, 23). However, it should be noted that fibrotic remodeling after surgery is primarily observed within the first year after implantation, whereas in our study, residual hearing loss was still observed after 1 year (20). Additionally, the composition of the study cohort may have influenced the poorer outcomes associated with the CA, as this CI was preferentially implanted in patients with minimal residual hearing, thus introducing preoperative bias due to the retrospective study design. Despite this, the early implantation of the CA allowed for a longer observation period over the years, suggesting that hearing loss over time may also be due to natural age-related hearing loss.

However, caution should be exercised when interpreting the outcomes regarding residual hearing preservation for the SMA. The SMA ears included in this study were primarily earlier cases in the study cohort, during which surgical procedures were subject to a learning curve, as described by Aschendorff et al. (24). In addition, newly developed electrode arrays are not primarily used in patients for whom the preservation of residual hearing is a secondary goal of CI surgery. Moreover, our dataset encompasses only 23 ears that underwent SMA insertion, with the first implantations following the introduction of this electrode array were not specifically aimed at preserving residual hearing.

Nevertheless, perimodiolar and straight electrode arrays are described to demonstrate different dislocation behaviors and intracochlear trauma risk in multiple studies (1, 5, 11, 13). Variations in the angle between the first and second cochlear turns, as well as diminished lengths typical of underdeveloped cochleae, can exacerbate insertion challenges and elevate the risk of scalar dislocation (25, 26). Another factor influencing the angular and linear insertion depth is the cochlear size, as already described by Ketterer et al. (5). The length of the electrode influences the insertion depth and thus the probability of cochlear trauma. According to our data, short electrode arrays such as the Flex²⁴ and the SSA show better outcomes regarding residual hearing. The Flex²⁶ did not demonstrate scalar dislocation or radiographically intracochlear trauma, as described by Ketterer et al. (6). However, residual hearing outcomes, as investigated in this study, should be examined with a larger study cohort (n = 15). O’Connell reported poorer results in terms of preserving residual hearing with deeper insertion of straight electrode arrays, although this did not reaching statistical significance (27). Other investigations found no difference in angular insertion depth in terms of postoperative shifts in audiometric threshold (20, 28). Erixon et al. included 21 straight electrode arrays and did not find a significant correlation between insertion depth and residual hearing preservation (29). Causon et al. described in their review a significant influence of angular insertion depth on residual hearing (30). Even though most studies described electrode array length as a risk factor for damaging residual hearing (4, 31, 32), some studies did not find such an association (33, 34).

Nevertheless, the preservation of residual hearing depends on many other factors, such as insertion speed, complete opening of the round window membrane, and the use of corticosteroids, as described in multiple previous studies (35–38). Perkins et al. indicated that hearing preservation may reach stability 1 year postoperatively, which could be potentially attributable to insertion trauma and immediate inflammatory responses following surgery (20). However, in our extensive study cohort, this was not confirmed. While Figures 5, 6 demonstrate that the initial decline in residual hearing is most pronounced in the initial days, it is not necessarily stable even 1-year post-implantation, exhibiting a decreasing trend across all electrode arrays examined. Nonetheless, the SSA exhibits the most stable curves over time. Due to the retrospective nature of our study design, we must acknowledge potential bias stemming from incomplete data and the lack of a consistent follow-up interval for determining audiometric thresholds. Additionally, variability between surgeons may serve as a confounding factor. Furthermore, data on the final scalar position as well as the cochlear approach were not collected and analyzed; for example, for round-window insertion, no differences in terms of residual hearing preservation were found when comparing straight and perimodiolar electrodes (39).

The results presented in this study provide sufficient data to suggest that flexible and slim electrodes are more suitable for residual hearing preservation. In our study, the SSA and Flex²⁴, in particular, showed the best results in terms of long-term and stable postoperative residual hearing preservation. However, no statistical difference was observed when compared with other flexible and slim modiolar electrodes, likely due to the number of factors to consider.

In the future, robotic implantation is likely to improve residual hearing preservation outcomes by providing a consistent, atraumatic insertion speed, thereby reducing insertion force as a factor in improving scope results.