Mouse, Man, and Machine: Combining Model Systems to Develop a Biomarker for Cochlear Deafferentation in Humans

Project Summary
Clinical testing for peripheral auditory dysfunction focuses on the audiogram. However, many auditory perceptual
deficits, such as tinnitus, hyperacusis, and difficulty with speech perception, cannot be fully explained by the
audiogram. Cochlear deafferentation (i.e., loss of inner hair cells, spiral ganglion cells, or cochlear synapses),
may contribute to these perceptual problems. However, there is currently no method for diagnosing
deafferentation in living humans. This prevents us from determining the prevalence of deafferentation in humans,
identifying deafferentation risk factors and perceptual consequences, or testing potential drug treatments.
Several non-invasive physiological measures are sensitive to loss of cochlear synapses (a form of
deafferentation) in animal models, including the auditory brainstem response (ABR), the envelope following
response (EFR), and the middle ear muscle reflex (MEMR). However, it is unclear how cochlear gain loss (e.g.,
due to outer hair cell damage) impacts the relationship between deafferentation and these physiological
measures, hindering translation to a diagnostic test for deafferentation. The overall objective of this proposal is
to develop a computational model that can estimate deafferentation from non-invasive physiological
measurements in humans with varying degrees of cochlear gain loss. The central hypothesis is that cochlear
gain loss can be predicted from distortion product otoacoustic emissions (DPOAEs) and deafferentation can be
predicted from a combination of ABR, EFR, and MEMR measurements. This hypothesis will be tested by
pursuing four specific aims: 1) Expand a computational model of the auditory periphery (CMAP) to predict ABR,
EFR, MEMR, and DPOAE responses in mice and humans based on both cochlear gain and afferent function, 2)
Validate and refine the CMAP by collecting physiological and histological data from mouse, 3) Predict
deafferentation in individual human subjects from physiological measurements by fitting the CMAP using
Bayesian regression, and 4) Evaluate deafferentation predictions for their relationship with risk factors and
predicted perceptual consequences of deafferentation. This approach is innovative because it extends prior work
to animal and human models with both cochlear gain loss and deafferentation, uses computational modeling to
bridge the gap between model systems, and combines multiple physiological measurements to predict
deafferentation in individual human subjects. The proposed research is significant because we currently have
no means of diagnosing deafferentation. Thus, the prevalence, associated risk factors, and perceptual impacts
of this condition are unclear. This project is expected to result in a biomarker of deafferentation for individual
patients that is based on their physiological measurements. This will enable us to identify peripheral auditory
damage that is independent of cochlear gain loss. If the biomarker is correlated with risk factors such as noise
exposure and auditory perceptual deficits such as speech perception difficulty, it will allow for the development
of targeted treatments for auditory perceptual deficits and strategies for damage prevention.

Grant Number: 5R01DC020423-04
NIH Institute/Center: NIH
Principal Investigator: Naomi Bramhall