The electrically evoked auditory brain stem response in some cochlear
implant patients may be confounded by evoked potentials generated by v
estibular neurons. The magnitude of this contribution to the response
from the vestibular system is unknown, in part because the survival of
cells within Scarpa's ganglion in profoundly deaf humans is unknown.
Therefore, we undertook a quantitative study of Scarpa's ganglion in 4
8 deaf subjects who in life would have been candidates for cochlear im
plantation and in 5 subjects with normal hearing. The numbers of resid
ual cells in both Scarpa's ganglion and the spiral ganglion in deaf su
bjects were significantly less than in individuals with normal hearing
. Bivariate analysis demonstrated a highly significant positive correl
ation between cell counts of Scarpa's ganglion and the spiral ganglion
. The durations of hearing loss and of profound deafness were negative
ly correlated with Scarpa's ganglion cell counts. However, in contrast
to spiral ganglion cell survival, the cause of profound deafness did
not predict the number of Scarpa's ganglion cells. Multiple linear reg
ression analysis using a variety of clinical parameters demonstrated t
hat the best predictor of the number of Scarpa's ganglion cells in pro
foundly deaf humans was the number of remaining spiral ganglion cells.