The Failure Index as Marker of the Neural Health Profile in MED-EL CI Users: A Retrospective Analysis

Cochlear implants (CIs) enable hearing via direct electrical stimulation of the spiral ganglion neurons (SGN). Outcomes with a CI depend, in part, on the number and excitability of the SGNs. In an animal model, we introduced the Failure Index (FI) as an electrically-evoked compound action potential (eCAP)-derived marker of cochlear health. The FI informs about the presence, site, and size of SGN lesions. Here, we translated the FI to clinical recordings from human MED-EL CI users. In this retrospective study, we included patient data recorded between 2016 and 2024 from 199 ears with postlingual hearing loss and 79 ears with prelingual hearing loss. The averaged FI values over all contacts of the array were stable within the analysis period (3^rd month to 1^st year postoperatively). The FI increased with age and was elevated in etiologies associated with higher SGN loss. Using 3D reconstruction from cone-beam computed tomography scans, we confirmed that the FI was independent of electrode-to-modiolus distance (0.1-2.5 mm). The FI showed individual patterns along the array, with maxima usually found at basal contacts, corresponding to reduced neural health at high-frequency cochlear regions. In a selected group of postlingual-deaf ears, we found higher correlation coefficients between speech-perception scores and the FI compared to other eCAP-derived markers (i.e., threshold, slope, and amplitude). These results were in line with the hypothesis that the FI may serve as a clinical tool to identify implanted ears with reduced neural health and to identify contacts stimulating areas of reduced SGN survival and integrity.