Quantitative studies of the vestibular system with serially sectioned human temporal bones have been limited because of difficulty in distinguishing hair cells from supporting cells and type I from type II hair cells. In addition, there is only a limited amount of normative data available regarding vestibular hair cell counts in humans. In this study, archival temporal bone sections were examined by Nomarski (differential interference contrast) microscopy, which permitted visualization of the cuticular plate and stereociliary bundle so as to allow unambiguous identification of hair cells. The density of type I, type II, and total numbers of vestibular hair cells in each of the 5 sense organs was determined in a set of 67 normal temporal bones that ranged from birth through 100 years of age. The mean total densities at birth were 76 to 79 cells per 0.01 mm2 in the cristae, 68 cells per 0.01 mm2 in the utricle, and 61 cells per 0.01 mm2 in the saccule. The ratio of type I to type II hair cells at birth was 2.4:1 in the cristae and 1.3:1 in the maculae. There was a highly significant age-related decline in all sense organs for total, type I, and type II hair cell densities that was best fit by a linear regression model. The cristae lost type I cells with advancing age at a significantly greater rate than the maculae, whereas age-related losses for type II cells occurred at the same rate for all 5 sense organs. Hair cell densities in the cristae were significantly higher at the periphery than at the center. There were no significant sex or interaural differences for any of the counts. Mathematical models were developed to calculate the mean and 95% prediction intervals for the total, type I, and type II hair cell densities in each sense organ on the basis of age. There was overall good agreement between the hair cell densities determined in this study and those reported by others using surface preparation techniques. Our data and related models will serve as a normative database that will be useful for comparison to counts made from subjects with known vestibular disorders.

Quantitative assessments of vestibular hair cells and Scarpa's ganglion cells were performed on 17 temporal bones from 10 individuals who had well-documented clinical evidence of aminoglycoside ototoxicity (streptomycin, kanamycin, and neomycin). Assessment of vestibular hair cells was performed by Nomarski (differential interference contrast) microscopy. Hair cell counts were expressed as densities (number of cells per 0.01 mm2 surface area of the sensory epithelium). The results were compared with age-matched normal data. Streptomycin caused a significant loss of both type I and type II hair cells in all 5 vestibular sense organs. In comparing the ototoxic effect on type I versus type II hair cells, there was greater type I hair cell loss for all 3 cristae, but not for the maculae. The vestibular ototoxic effects of kanamycin appeared to be similar to those of streptomycin, but the small sample size precluded definitive conclusions from being made. Neomycin did not cause loss of vestibular hair cells. Within the limits of this study (maximum postototoxicity survival time of 12 months), there was no significant loss of Scarpa's ganglion cells for any of the 3 drugs. The findings have implications in several clinical areas, including the correlation of vestibular test results to pathological findings, the rehabilitation of patients with vestibular ototoxicity, the use of aminoglycosides to treat Meniere's disease, and the development of a vestibular prosthesis.

Scarpa's ganglion cell counts were performed in 106 serially sectioned, normal human temporal bones from 75 individuals. Of these 106 bones, 15 were from neonates less than 30 days old, 14 were from infants between the ages of 1 and 12 months, and the remainder were distributed throughout each decade of life, with sample sizes ranging from 4 to 10 per decade. All temporal bones had to meet 2 criteria: no symptoms or signs of inner ear disease except for presbycusis in the medical case history and no abnormality in the inner ear on light microscopy. The total ganglion cell counts declined significantly with age at an average rate of 57 cells per year. The age-related decline was significantly greater in the superior division than in the inferior division. There was also a significant sex effect, independent of age: for any age, the count in men averaged 1,526 cells higher than in women. There was no significant interaural difference. Mathematical models were developed to compute the mean and 95% prediction intervals for Scarpa's ganglion cell counts in terms of age and sex parameters. The counts and models will serve as a normative database against which to compare counts made in temporal bones from subjects with known vestibular disorders.

Quantitative assessments of vestibular hair cells and Scarpa's ganglion cells were performed on temporal bones from 24 patients with well-documented Meniere's disease. Of these, 18 had unilateral disease and 6 had bilateral disease. Vestibular hair cell counts were made in each of the 5 sense organs by Nomarski (differential interference contrast) microscopy. Hair cell counts were expressed as densities: number of cells per 0.01 mm2 surface area of the sensory epithelium. The results were compared with age- and sex-matched normal data. The type I hair cell densities for all vestibular sense organs were within the range for normative data. On the other hand, there was a significant loss (p < .01) of type II hair cells for all 3 cristae and both maculae. There was also a significant loss of Scarpa's ganglion cells (p < .001) when compared with normative data. The findings indicate a selective loss of type II hair cells and Scarpa's ganglion cells in Meniere's disease. These new observations have implications regarding the pathophysiological mechanism and clinical manifestations of Meniere's disease.

This paper discusses the use of vestibular testing to discriminate between right (n = 29) and left (n = 27) Ménière's disease. We examined reduced vestibular response (RVR), directional preponderance, and spontaneous and positional nystagmus measurements from electronystagmography, as well as the asymmetry measurements from the sinusoidal harmonic acceleration test, to determine whether multivariate logistic regression could improve upon the discrimination performance of RVR alone. We found that patients with a spontaneous or positional nystagmus often had a “recovery nystagmus”, beating in the opposite direction of that predicted for an acute lesion. When present, the direction and magnitude of such nystagmus could be used in the classification algorithm to increase the discriminatory power over RVR alone, but in these patients the rotation test asymmetry measurements were rendered useless. In the absence of spontaneous or positional nystagmus, asymmetry measurements significantly enhanced right/left discrimination. Directional preponderance was insignificant in determining the side of lesion.

Although patient symptoms and an audiogram can typically identify the affected ear or ears in Ménière's disease, there are some cases where this differentiation is problematic. This paper concentrates on the sole use of vestibular test data to discriminate between unilateral and bilateral Ménière's disease. Patients that were known to have peripheral unilateral vestibular hypofunction (n = 104) were used as learning groups to define a region in multidimensional measurement space consisting of four vestibular test scores which summarized data from electronystagmography, sinusoidal harmonic acceleration, and computerized dynamic posturography tests. A multivariate boundary was created from the unilateral learning group that determined thresholds for identifying bilateral vestibular hypofunction. Patients with bilateral Ménière's disease (n = 23) and with bilateral ototoxicity (n = 19) were then used as test subjects to determine the sensitivity of the multivariate boundary. Results showed up to a increase in estimated test sensitivity (specificity = 95%) bilateral vestibular hypofunction.