Relationship between Metabolomics Profile of Perilymph in Cochlear-Implanted Patients and Duration of Hearing Loss

Abstract

Perilymph metabolomic analysis is an emerging innovative strategy to improve our knowledge of physiopathology in sensorineural hearing loss. This study aims to develop a metabolomic profile of human perilymph with which to evaluate the relationship between metabolome and the duration of hearing loss. Inclusion criteria were eligibility for cochlear implantation and easy access to the round window during surgery; patients with residual acoustic hearing in the ear to be implanted were excluded. Human perilymph was sampled from 19 subjects during cochlear implantation surgery. The perilymph analysis was performed by Liquid Chromatography-High-Resolution Mass and data were analyzed by supervised multivariate analysis based on Partial Least-Squares Discriminant Analysis and univariate analysis. Samples were grouped according to their median duration of hearing loss. We included the age of patients as a covariate in our models. Statistical analysis and pathways evaluation were performed using Metaboanalyst. Nineteen samples of human perilymph were analyzed, and a total of 106 different metabolites were identified. Metabolomic profiles were significantly different for subjects with ≤ 12 or > 12 years of hearing loss, highlighting the following discriminant compounds: N-acetylneuraminate, glutaric acid, cystine, 2-methylpropanoate, butanoate and xanthine. As expected, the age of patients was also one of the main discriminant parameters. Metabolic signatures were observed for duration of hearing loss. These findings are promising steps towards illuminating the pathophysiological pathways associated with etiologies of sensorineural hearing loss, and hold open the possibilities of further explorations into the mechanisms of sensorineural hearing loss using metabolomic analysis.

Keywords: cochlear implant; duration of hearing loss; metabolomics; perilymph.

Figure 1

Statistical analysis to compare the metabolome profile from perilymph fluid of cochlear implants (CIs) patients according to duration of hearing loss. (A) Multivariate analysis by partial least squares discriminant analysis (PLS-DA), discriminating between patients according to the median duration of hearing loss. The red circles show data for patients with duration of hearing loss ≤ 12 years, and the green circles show data for patients with duration of hearing loss > 12 years. Components 1 and 2 represent a linear combination of relevant metabolites expressing the maximum variance. After mean-centering and scaling to unit variance, the data are used for the computation of the first principal component, that is the line in the K-dimensional space that best approximates the data in the least squares sense. Importantly one principal component is insufficient to model the systematic variation of a data set, and a second principal component is calculated. The second PC is also represented by a line in the K-dimensional variable space, which is orthogonal to the first PC. (B) The rank of the different metabolites (the top 15) identified by the PLS-DA according to the VIP score on the left and according to the coefficient score on the right. (C) Univariate analysis via volcano plot based on fold change and p-value, highlighting 1 metabolite. Legend: VIP: Variable Influence of Projection; OPLS-DA: Orthogonal Partial Least Squares Discriminant Analysis; p: p-value; FC: fold change.