Presbycusis: A human temporal bone study of individuals with flat audiometric patterns of hearing loss using a new method to quantify stria vascularis volume

Authors

  • Erik G. Nelson MD,

    Corresponding author
    1. Section of Otolaryngology–Head and Neck Surgery, University of Chicago Medical Center, Chicago, Illinois, U.S.A.
    • Correspondence to Erik G. Nelson, MD, Section of Otolaryngology–Head and Neck Surgery, University of Chicago Medical Center, 5841 S. Maryland Avenue, MC 1035, Chicago, IL 60637, U.S.A.
    Search for more papers by this author
  • Raul Hinojosa MD

    1. Section of Otolaryngology–Head and Neck Surgery, University of Chicago Medical Center, Chicago, Illinois, U.S.A.
    Search for more papers by this author

Abstract

Objective: The purpose of this study was to determine the prevalence of stria vascularis atrophy in individuals with presbycusis and flat audiometric patterns of hearing loss. Individuals with presbycusis have historically been categorized by the shape of their audiograms, and flat audiometric thresholds have been reported to be associated with atrophy of the stria vascularis. Stria vascularis volume was not measured in these studies. Study Design: Retrospective case review. Methods: Archival human temporal bones from individuals with presbycusis were selected on the basis of strict audiometric criteria for flat audiometric thresholds. Six temporal bones that met these criteria were identified and compared with 10 temporal bones in individuals with normal hearing. A unique quantitative method was developed to measure the stria vascularis volume in these temporal bones. The hair cell and spiral ganglion cell populations also were quantitatively evaluated. Results: Only one of the six individuals with presbycusis and flat audiometric thresholds had significant atrophy of the stria vascularis. This individual with stria vascularis atrophy also had reduced inner hair cell, outer hair cell, and ganglion cell populations. Three of the individuals with presbycusis had spiral ganglion cell loss, three individuals had inner hair cell loss, and all six individuals had outer hair cell loss. Conclusions: The results of this investigation suggest that individuals with presbycusis and flat audiometric patterns of hearing loss infrequently have stria vascularis atrophy. Outer hair cell loss alone or in combination with inner hair cell or ganglion cell loss may be the cause of flat audiometric thresholds in individuals with presbycusis.

The term presbycusis refers to hearing loss that is associated with the degenerative process of aging. By definition, presbycusis is bilateral, symmetrical, and slowly progressive.1 It is the most common cause of adult hearing deficiency. In the United States, presbycusis affects 40% of the population older than 75 years of age, and in our aging society, it is becoming more prevalent.2 Our lack of understanding of this disease process and our inability to remediate its progression will continue to have a great impact on the productivity and quality of life of these individuals.

The occurrence of presbycusis is thought to be determined predominantly by genetic factors, but it can be influenced by environmental factors.3,4 Preliminary investigations suggest that caloric diet restriction and antioxidants may reduce age-related hearing loss caused by mutations of the mitochondrial genome.4 Genetic testing, as it evolves, may eventually play a role in the diagnosis and management of presbycusis. At present, however, clinicians must use the family history, the history of onset and progression, and the results of audiometric testing to determine the degree of impairment, to estimate the potential for future hearing loss, and to make recommendations for amplification with hearing aids. With improvement of our understanding of presbycusis, treatment other than with amplification may be possible.

Until recently, the cochlea has been considered to lack the capability of regenerating new sensory and neural cells.1 Now experimental investigations suggest that hair cell regeneration in the inner ear is possible and may be stimulated by mitogenic agents.5,6 The results of audiometric studies and histopathologic studies of the temporal bone demonstrate great variability between individuals with presbycusis. Methods of identifying the specific abnormality of the cochlear elements in an individual with presbycusis may be necessary to recommend appropriate treatment to restore cochlear function. Evaluation with audiometric testing may serve such a role.

Presbycusis has been categorized into different types on the basis of audiometric patterns of hearing loss, and attempts have been made to correlate these patterns of hearing loss with atrophic changes of the individual cochlear elements observed histopathologically.1,7,8 The current method of classifying presbycusis is presented in Table I. Quantitative methods of evaluating the hair cells and spiral ganglion cells of the cochlea have been refined.9,10 Age-related hearing loss associated with atrophy of these cells also has been described extensively; however, whether a predictable pattern of hearing loss results from this atrophy remains unclear.3,11,12

Consensus has not been reached regarding the relationship between audiometric patterns of hearing loss and cochlear pathologic conditions in individuals with presbycusis, in part because of incomplete or nonquantitative methods of evaluating the cochlear elements and inconsistent application of uniform audiometric criteria. These approaches are prone to inaccuracy, and their use may lead to false conclusions. This project was established to further study the relationship between flat audiometric patterns of hearing loss in presbycusis and the associated abnormalities of the cochlear elements. Previously described methods of quantifying the stria vascularis have used visual estimates or have not evaluated the stria vascularis throughout the cochlea. Therefore, a method was developed to measure the total stria vascularis volume in archival temporal bones. A nonvolumetric method of quantifying the stria vascularis13 was used as a reference for comparison with this new volumetric method. It is hoped that the understanding of presbycusis will improve through the development of more accurate and comprehensive histopathologic assessments.

METHODS

From our temporal bone collection, which contains 1700 specimens, only six cases with a clinical diagnosis of presbycusis that met our strict criteria for a flat audiometric hearing loss pattern were identified. These individuals had bilateral, symmetrical, progressive, age-related sensorineural hearing loss with a threshold greater than 25 dB and no greater than a 20 dB threshold difference between the frequencies of 250 Hz and 8000 Hz. Word discrimination test scores were not available. The individuals ranged in age from 52 to 87 years. The temporal bones of these six individuals with presbycusis were compared with 10 temporal bones from people with no hearing loss from 250 Hz through 8000 Hz.

The temporal bones of these individuals were removed within 48 hours after death, placed in cold 20% formalin for 24 hours, and then placed in cold 10% formalin for an additional 10 days. After decalcification with 1% nitric acid for 35 days and dehydration in graded concentrations of ethanol, the specimens were embedded in celloidin. Serial sections of the temporal bones were cut in the horizontal plane at a thickness of 20 μm. Every tenth section was mounted on a glass slide and stained with hematoxylin and eosin stain (H&E) for light microscopic examination.

The organ of Corti was graphically reconstructed according to the methods previously described by Guild14 and Schuknecht.9 The presence or absence of inner and outer hair cells was evaluated in each tenth section. The percentage of inner and outer hair cells present in each cochlea was calculated, which allowed the hair cell population in each cochlea to be represented as a single number for comparison.

The spiral ganglion of each cochlea was reconstructed according to the method of Hinojosa et al10 on tracing paper that had been superimposed on the reconstruction of the organ of Corti (Fig. 1). The borders of the spiral ganglion coils were measured by using the heads of the pillar cells in the first turn of the organ of Corti as the reference point. The nucleoli of the ganglion cells in each tenth section were counted with the use of an ocular grid, and the total number of nucleoli counted in each cochlea was multiplied by 10. The total number of nucleoli in each cochlea was multiplied by a correction factor of 0.95 to account for nucleoli at the point of sectioning that would otherwise be counted twice to determine the total number of ganglion cells in each cochlea. The organ of Corti was divided into consecutive 5 mm segments starting at the basal end of the cochlea. At each 5 mm division, a line was drawn radially from the organ of Corti to the spiral ganglion. Thus, each 5 mm segment of the organ of Corti demarcated a corresponding segment of the spiral ganglion.

The distance from the heads of the pillar cells of the organ of Corti to the basal cell layer of the stria vascularis was measured, and the stria vascularis was graphically reconstructed in two dimensions on tracing paper, as described for the spiral ganglion (Fig. 2). The total length of the reconstructed stria vascularis was measured. Using the Olympus C-2 Image Analyzer Program (Olympus Corp., Lake Success, NY) running on an IBM-compatible personal computer, the cross-sectional area of the stria vascularis was measured throughout the cochlea. Approximately 100 measurements were taken in each cochlea. The angle of sectioning of the stria vascularis in each slide was measured using the graphic reconstruction. The artifact introduced by measuring the cross-sectional area where the stria was sectioned at an angle was corrected by multiplying the measured cross-sectional area by the cosine of the angle of sectioning. The mean corrected cross-sectional area was determined for each millimeter segment of the stria vascularis, and the volume of the segment was calculated. In the tangential segments where the area could not be measured, the volume was assumed to be equal to the adjacent segment. The total stria vascularis volume of the cochlea was determined by summing the volume of each millimeter segment throughout the cochlea. At each 5 mm division of the organ of Corti described in the preceding paragraph, a line was drawn radially from the organ of Corti to the stria vascularis. Thus, each 5 mm segment of the organ of Corti demarcated a corresponding segment of the stria vascularis.

Table Table I.. Current Classification of Presbycusis
TypeAudiometric PatternHistopathology
  1. Data from references 1 and 7.

SensoryAbrupt high tone lossHair cell loss
NeuralDiminished word discriminationSpiral ganglion cell loss
StrialFlat lossStria vascularis atrophy
Cochlear conductiveGradual descending patternNo morphologic findings (presumed stiffening of basilar membrane)
MixedCombinations of flat, sloping, and abrupt high tone lossCombinations of hair cell, ganglion cell, and stria vascularis loss
IndeterminateFlat and/or abrupt high tone lossNo morphologic findings (presumed impaired cellular function)
Figure Fig. 1..

Case 6: Graphic reconstruction of the organ of Corti and spiral ganglion. The organ of Corti is divided into 5 mm segments beginning at the basal end. Radial lines drawn from the organ of Corti demarcate the corresponding segments of the spiral ganglion.

Figure Fig. 2..

Case 6: Graphic reconstruction of organ of Corti (broken line) and stria vascularis (solid line).

Our new method of measuring the total volume of the stria vascularis was compared with the method previously described by Pauler et al,13 which we used to measure the mean cross-sectional area of the stria vascularis. The stria vascularis in each of the temporal bones in our study was evaluated by using both methods. In the method of Pauler et al,13 the image analyzer was used to measure the cross-sectional area of the stria vascularis at 40 locations near the midmodiolar section, and the average cross-sectional area was calculated. The cross-sectional area was measured only in the apical half of the cochlea, and the measured cross-sectional area was not corrected for artifact introduced by sectioning at an angle.

Bar graphs representing the spiral ganglion cell loss and stria vascularis loss per 5 mm segment of the organ of Corti were constructed for the temporal bones from the individuals with presbycusis. A loss was depicted when the value in any 5 mm segment was less than the mean value from the group of 10 individuals with normal hearing. Bar graphs representing the inner and outer hair cell loss per 1 mm segment of the organ of Corti were also constructed for the temporal bones from individuals with presbycusis.

Statistical analysis was performed to determine normative values. Differences in sex, race, and the hearing loss group were evaluated with the use of Student's t-test. Fisher's Exact Test was used when normal distributions were not present.

RESULTS

The radius of curvature of the stria vascularis is greater than the radius of curvature of the organ of Corti (Fig. 2). Consequently, the length of the stria vascularis from the basal to the apical end of the cochlea is greater than the length of the organ of Corti (Table II). In the group of normal-hearing individuals, the mean length of the organ of Corti was 33.9 mm, and the mean length of the stria vascularis was 39.7 mm. The radius of curvature of the spiral ganglion is less than that of the organ of Corti, and consequently the length of the spiral ganglion is less than the length of the organ of Corti (Table II). The mean length of the spiral ganglion in the group of normal-hearing individuals was 14.2 mm.

The thickness of the stria vascularis—the distance from the marginal cell layer to the basal cell layer—is relatively constant throughout the cochlea, except where focal areas of degeneration have occurred. The width of the stria vascularis—the distance from the margin of the stria vascularis adjacent to the spiral prominence to the margin adjacent to the Reissner's membrane—is greater in the basal turn than in the apical turn (Fig. 3). Consequently, the cross-sectional area of the stria vascularis and the volume of the stria vascularis are greatest in the basal turn and become progressively smaller as the apical turn is approached (Table III). In most temporal bones, the basal turn of the cochlea measures approximately 15 mm in length. Approximately half of the stria vascularis volume is contained in the basal turn of the cochlea (Table III).

Table Table II.. Comparison of Lengths of the Organ of Corti, the Spiral Ganglion, and the Stria Vascularis in the Normal-Hearing Group
 Organ of CortiSpiral GanglionStria Vascularis
CaseLength, mmLength, mmLength, mm
133.712.841.1
237.516.142.8
331.915.236.4
436.915.442.8
533.315.138.5
632.913.137.4
732.313.638.9
831.713.438.1
934.212.941.2
1033.914.138.9
Mean33.814.239.7

When the directly measured cross-sectional area of the stria vascularis, without correcting for the angle of sectioning, is plotted along its length, a scalloped pattern is observed (Fig. 4). The peaked regions represent artifact due to the apparent increase in area where the stria vascularis is sectioned at an angle. In the midmodiolar section, the cross-sectional area of the stria vascularis is represented accurately; however, as the tangential areas of sectioning are approached, the angle of sectioning increases. This results in an apparent but not real increase in the thickness of the stria vascularis and a consequent increase in the cross-sectional area of the stria vascularis (Fig. 5). As mentioned earlier, by multiplying the measured cross-sectional area by the cosine of the angle of sectioning, the scalloped pattern becomes almost linear (Fig. 4), and an accurate representation of the cross-sectional area is observed.

The quantitative evaluation of the cochlear elements in our group of 10 normal-hearing individuals is presented in Table IV. In this group, the percentage of observed remaining inner hair cells ranged from 97 to 100%, and the percentage of observed remaining outer hair cells ranged from 93 to 98%. The method used for estimating the percentage of remaining hair cells involves the evaluation of every tenth section of the cochlea. In addition, in areas where tangential sectioning of the cochlea occurs, the hair cells cannot be evaluated completely. Therefore, a limited amount of undetected hair cell loss was possible in cases in which 100% of hair cells were observed to be present. The spiral ganglion cell population in this group of normal-hearing individuals ranged from 29,978 to 38,352. In our group of normal-hearing individuals, the stria vascularis volume ranged from 305.4 to 482.1 × 106 μm3, with a mean of 409.7 × 106 μm3.

Using the technique described by Pauler et al,13 the mean cross-sectional area of the stria vascularis found in our group of normal-hearing individuals was similar to that reported by Pauler et al. in their normal-hearing group. They reported a mean of 9716 μm2, with a range of 8405 to 11,124 μm2. In our group, we observed a mean of 9463 μm2, with a range of 7718 to 10,798 μm2 (Table IV).

Limited areas of stria vascularis degeneration were observed in the cochleae from individuals with normal hearing and from those with presbycusis. These areas of degeneration were usually limited to segments of the stria vascularis less than 1 mm in length. They ranged in severity from mild thinning of the stria vascularis to complete absence of the stria vascularis. These lesions often involved only a margin of the stria vascularis and not its entire width. Photomicrographs demonstrate these variations (Fig. 6).

The quantitative evaluation of the cochlear elements in our group of individuals with presbycusis and a flat audiometric pattern of hearing loss is presented in Table V. The organ of Corti reconstruction, the bar graphs of the cochlear elements, and the audiograms of the individuals with presbycusis are depicted in Figures 7 through 12.

Of the six temporal bones studied from individuals with presbycusis and flat audiometric patterns, only one (case 16) had a statistically significant (P < .05) reduction in stria vascularis volume and mean cross-sectional area when compared with the normal-hearing group (Table V). The other five bones had a stria vascularis volume and mean cross-sectional area within the normal range. Three (cases 11, 15, and 16) of six bones had a statistically significant (P < .05) reduction in the ganglion cell population. A mild but statistically significant (P < .05) reduction in inner hair cell population was observed in three (cases 12, 15, and 16) of six bones. The finding common to all six bones with presbycusis and flat audiometric patterns was a statistically significant (P < .05) reduction of the outer hair cell population.

Figure Fig. 3..

Case 6: Photomicrographs of stria vascularis at the midmodiolar section in the basal (A), second (B), and apical (C) turns demonstrating the decreasing size of the stria vascularis as the apical turn is approached (original magnification, × 144). Graphic reconstruction demonstrates location of each photomicrograph.

Statistical comparison of the normal-hearing group and the presbycusis group revealed no significant difference (P > .05) in stria vascularis volume, stria vascularis mean cross-sectional area, and ganglion cell population. There was a significant difference in inner hair cell population (P = .046) and outer hair cell population (P = .009) between the two groups. There was no significant difference (P > .05) in inner ear morphologic findings between men and women in the normal-hearing and presbycusis groups. No significant difference (P > .05) was observed with regard to inner ear morphologic findings between black and white racial groups in the normal-hearing group. Only one white subject and five black subjects are present in the hearing loss group, making statistical comparison based on race invalid.

Table Table III.. Stria Vascularis Volume* per 5 mm Segment of the Organ of Corti, Mean of Control Group of 10 Normal-Hearing Individuals
Parameter0–55–1010–1515–2020–2525–3030–end
  1. * × 106 μm3

Volume per 5 mm segment of the organ of Corti75.471.457.660.257.952.737.4
Volume per mm organ of Corti15.014.311.512.011.610.59.73

DISCUSSION

Animal studies are useful in gaining insight into human disease processes. They allow the study of populations under controlled conditions without the numerous uncontrolled factors present in human studies. In addition, differentiating the contribution of individual cochlear elements in presbycusis requires postmortem histopathologic evaluation, which also limits human studies. Although age-related hearing loss has been observed in animals,15 the causes and mechanisms may be different from those in humans. For example, in several rodent species, the age-related pattern of hair cell loss is greatest in the apical end of the cochlea,16–18 unlike in humans, in whom hair cell loss is greatest in the basal turn.8,19–22 Therefore, animal studies cannot replace the investigation of age-related hearing loss in humans.

Figure Fig. 4..

Case 6: Cross-sectional areas of stria vascularis plotted along its length. Circles, directly measured cross-sectional area; x, cross-sectional area corrected for the artifact introduced by sectioning at an angle—the directly measured cross-sectional area multiplied by the cosine of the angle of sectioning (area × cosine Ø). The arrows indicate points of tangential sectioning.

Figure Fig. 5..

Case 6: Photomicrographs of stria vascularis at midmodiolar section (A) and at a near tangential section (B) in the basal turn demonstrating the apparent increase in size of the stria vascularis when sectioned at an angle (original magnification, × 144). Graphic reconstruction demonstrates location of each photomicrograph and angle of sectioning. A, Ø = 0 degrees; B, Ø = 53 degrees.

Table Table IV.. Quantitative Evaluation of the Cochlear Elements in the Normal-Hearing Group
CaseSexRaceAgeSVVSVMCAGCIHCOHC
  1. SVV, stria vascularis volume (x106 μm3); SVMCA, stria vascularis mean cross-sectional area (μm2); GC, ganglion cells; IHC, percentage of inner hair cells remaining; OHC, percentage of outer hair cells remaining; M, male; F, female; B, black; W, white.

1MB20482.110,79831,96810097
2MW27378.0725029,97810094
3MW26372.3944831,7689994
4MW9456.010,12538,35210096
5FB76305.4771834,5239797
6MB68434.5969035,61510093
7MB51453.310,53534,74110098
8FB50397.810,39832,73710098
9FW39459.910,42032,93510093
10FW63358.0824531,20810095
Mean   409.7946333,38310096

Light microscopic evaluation of the temporal bone effectively demonstrates pathologic changes that result from cell death and tissue loss. Studies using immunohistochemical techniques suggest that the functional integrity of the stria vascularis is maintained when its structural integrity appears normal under light microscopy.23 Schuknecht et al,24 Nadol,25 and Kimura and Schuknech26 used electron microscopy to demonstrate degenerative changes in the cochlear elements that were not apparent by light microscopy. Electron microscopy requires immediate postmortem acquisition of tissue, however, which frequently is not possible in human studies and is limited to the evaluation of small samples of tissue. Therefore, light microscopy remains the most practical method of comprehensively evaluating the cochlear elements in larger groups of individuals.

Investigators have categorized presbycusis on the basis of the shape of the audiometric curves and have attempted to correlate these patterns of hearing loss with pathologic changes of individual cochlear elements. Crowe et al8 identified one group of individuals with presbycusis with abrupt high-tone loss who had atrophy of the organ of Corti in the basal turn of the cochlea and a second group with gradual high-tone loss who had partial atrophy of the neural fibers supplying the basal turn of the cochlea. A third group did not have significant histopathologic abnormalities of the temporal bone.

Schuknecht19 subsequently classified the histopathology of presbycusis into four categories. The first type, sensory presbycusis, was characterized by atrophy of the organ of Corti in the basal end of the cochlea and was associated with an abrupt high-tone hearing loss. The second type, neural presbycusis, demonstrated a loss of cochlear neurons and poor word discrimination in the presence of stable pure-tone thresholds. The third type, metabolic presbycusis, was associated with atrophy of the stria vascularis and hearing loss with a flat pure-tone audiometric threshold and good word discrimination scores. The fourth type, mechanical presbycusis, was characterized by normal morphologic findings and a linear descending pure-tone audiogram, which was considered to be related to abnormal motion mechanics of the cochlear partition such as stiffening of the basilar membrane.

Figure Fig. 6..

Photomicrographs demonstrating atrophy of the stria vascularis (original magnification, x144). A, Case 2: segmental degeneration of the stria vascularis with absence at the margins adjacent to the spiral prominence and the Reissner membrane and normal appearance of remaining stria vascularis. B, Case 3: absence of stria vascularis adjacent to the spiral prominence and degeneration of remaining stria vascularis. C, Case 8: absence of stria vascularis adjacent to the Reissner membrane and thinning adjacent to the spiral prominence. The central portion is thicker due to the presence of vessels adjacent to the basal layer. D, Case 9: diffuse thinning of the stria vascularis along the entire width. E, Case 14: irregular thinning of the stria vascularis. F, Case 16: encapsulated basophilic material in an area of complete stria vascularis degeneration. G, Case 16: area of complete stria vascularis atrophy.

Table Table V.. Quantitative Evaluation of the Cochlear Elements in the Presbycusis with Flat Audiogram Group
CaseSexRaceAgeSVVSVMCAGCIHCOHCHL
  1. SVV, stria vascularis volume (x106 μm3); SVMCA, stria vascularis mean cross-sectional area (μm2); GC, ganglion cells; IHC, percentage of inner hair cells remaining; OHC, percentage of outer hair cells remaining; HL, pure tone average 0.5, 1, 2 kHz; F, female; M, male; B, black; W, white.

11FB72409.110,09321,5181008240
12FB52360.4852531,046928648
13FB78450.6954032,604998542
14MB63488.3935736,086977635
15MB87335.210,11320,347957557
16FW62290.4432825,815955540

Schuknecht and Gacek7 further categorized cases of presbycusis without abnormalities of the cochlea detectable by light microscopy into two types: the previously described cochlear conductive type of presbycusis, defined by a gradually sloping pattern of hearing loss and attributed to stiffening of the basilar membrane, and an indeterminate type of presbycusis, defined by a flat and/or abrupt high-tone hearing loss. They suggested that the indeterminate type of presbycusis was due to impaired cellular function rather than to cellular attrition and was therefore undetectable by light microscopy. In their report, Schuknecht and Gacek7 also described a mixed type of presbycusis that was characterized by combinations of flat, gradually sloping, and abrupt high-tone hearing losses with observable light microscopic abnormalities of multiple cochlear elements. Other investigators, such as Suga and Lindsay11 and Engstrom et al,12 observed no correlation between the audiometric pattern of hearing loss and the histologic findings in presbycusis. Similarly, Jennings and Jones27 commented that the categorization of presbycusis on the basis of audiometric patterns of hearing loss suggested by Schuknecht19 and Schuknecht and Gacek7 was poorly supported by the findings of other investigators. Studies that have evaluated the relationship between histopathology and hearing loss are discussed in the following paragraphs.

Review of the literature reveals consensus regarding the pattern and progression of hearing loss in aging populations.28–33 Although most studies describe a predominant sloping audiometric pattern, flat audiometric patterns have also been described.11,19,24,34–37 Some investigators have reported that the individuals with flat audiometric patterns of hearing loss differ from those with sloping patterns in that their hearing loss results from atrophic changes in the stria vascularis13,19,24 or degeneration in the central auditory pathways.36 Other investigators have observed spiral ganglion cell loss and stria vascularis loss or spiral ganglion cell loss alone in cases of presbycusis with flat audiometric patterns of hearing loss; however, no correlation between the type of audiometric curve and the abnormalities present in the cochlear elements were evident in the Suga and Lindsay study.11

Although various patterns of hearing loss progression have been described in individuals with sloping patterns of hearing loss,32 in one study, only 8% of individuals with a flat audiometric pattern of hearing loss changed to a sloping pattern over time.34 In general, histopathologic studies of presbycusis have involved fewer individuals than audiometric studies, and histopathologic studies may be prone to selection bias. Therefore, audiometric studies may reflect the prevalence of different patterns of presbycusis more accurately than histopathologic studies.

Figure Fig. 7..

Case 11: Presbycusis with flat audiogram, reconstruction of cochlea.

Few studies have examined the relative contribution of cochlear versus central auditory pathway pathologic conditions in cases of presbycusis. Tests for recruitment suggest that most individuals with presbycusis have a cochlear abnormality.30,37 Additional auditory brainstem response studies,38 electrocochleography studies,38 and otoacoustic emission studies39,40 suggest that hair cell loss may be the predominant abnormality in individuals with presbycusis. Numerous histopathologic studies in individuals with presbycusis have confirmed that hair cell loss is a common finding.7,8,11,19,20,22,24,41–47 Two of these studies have reported spiral ganglion cell loss as the most frequently observed feature in the temporal bones of individuals with presbycusis.11,44 In these reports, however, the central nervous systems of the individuals studied were not evaluated.

Figure Fig. 8..

Case 12: Presbycusis with flat audiogram, reconstruction of cochlea. The organ of Corti is absent owing to preparation artifact.

Figure Fig. 9..

Case 13: Presbycusis with flat audiogram, reconstruction of cochlea.

Figure Fig. 10..

Case 14: Presbycusis with flat audiogram, reconstruction of cochlea.

Figure Fig. 11..

Case 15: Presbycusis with flat audiogram, reconstruction of cochlea.

Several audiometric studies have suggested that progressive central nervous system abnormalities may result in diminished speech reception ability that is due in part to decreased cognitive functioning.31,48,49 Other investigators have reported atrophic changes of the neuron cell bodies, axons, and myelin in the central auditory pathways of elderly individuals, which they interpreted as the histopathologic correlates of presbycusis.36,50–52 In one of these studies,51 the hearing level of the individuals studied was not known. Three of these studies36,50,51 did not compare their findings with a group of young normal-hearing individuals. In addition, three of these studies50–52 did not evaluate nonauditory parts of the brain, and two of them51,52 did not evaluate the temporal bones of the individuals studied.

Figure Fig. 12..

Case 16: Presbycusis with flat audiogram, reconstruction of cochlea. The organ of Corti is absent owing to preparation artifact.

In contrast, a quantitative study of the ventral cochlear nucleus revealed no reduction of the neuron population in older individuals when compared with younger control subjects.53 In this study, however, the audiograms of the individuals were not reported, and the individuals studied were not known to have presbycusis. Although some of these reports suggest that abnormalities of the central auditory pathways may be a cause of presbycusis, comprehensive, quantitative evaluation of the brain and temporal bones of individuals with presbycusis confirmed by audiometric testing is needed to better understand the prevalence and extent of central auditory pathway abnormalities in the process of presbycusis.

Loss of word discrimination has been reported to be associated with elevated pure-tone audiometric thresholds30 and particularly with elevated low-frequency thresholds.37 Another report described a relationship between loss of speech discrimination and elevated thresholds in the midfrequency range.54 Some investigators1,3,7,19,24,54,55 have associated a decline in speech discrimination with a loss of spiral ganglion cells and have categorized this presentation as a specific type of presbycusis referred to as neural presbycusis. Neural presbycusis was characterized by a progressive loss of word discrimination in the presence of stable pure-tone thresholds that were not affected until neuronal loss exceeded 90%. In the cases depicted in these reports of neural presbycusis with a range of 64 to 89% loss of spiral ganglion neurons, however, all had moderate to severe sloping pure-tone thresholds,1,7 and no cases were presented with normal pure-tone thresholds and a loss of word discrimination. In one study,55 it was observed that a minimum of 10,000 spiral ganglion cells was necessary to maintain some capacity for speech discrimination; however, normal word discrimination and pure-tone thresholds were observed even when the spiral ganglion cell population was as low as 20,000. Other investigators8,11,44 have reported that spiral ganglion cell loss is a common finding in presbycusis and is often associated with sloping pure-tone thresholds.

Investigators have suggested that abnormalities of the basilar membrane, which are difficult to detect by light microscopy, may be a cause of hearing loss in cases of presbycusis.3,7,19,25,45,56,57 Schuknecht3,19 reported that cases of presbycusis with gradually sloping audiometric patterns usually lacked sufficient light microscopic histopathologic abnormalities of the cochlea to explain the observed hearing loss and suggested that abnormal motion mechanics of the basilar membrane were the cause of hearing loss in these cases. Ramadan and Schuknecht45 also reported six such cases with gradually sloping audiograms, which they attributed to an abnormality of the basilar membrane.

In contrast, although Crowe et al8 reported that 25% of the temporal bones in their study of presbycusis had no observable histopathologic abnormalities to explain the individuals' hearing loss, they observed that gradual sloping audiometric patterns of hearing loss were usually associated with atrophy of the neural elements of the cochlea, and Wright and Schuknecht58 observed an association of spiral ligament atrophy with a descending audiometric pattern of hearing loss. The concept of a basilar membrane abnormality as a cause of presbycusis was originally reported by Mayer,56 who attributed the hearing loss of presbycusis to calcification and stiffening of the basilar membrane, although other investigators8 observed no correlation between this histologic finding and hearing loss. Nomura57 reported the presence of lipid deposits in the basilar membrane, which he suggested may be the cause of hearing loss in some individuals with presbycusis. Nadol25 described the electron microscopic findings of one case of presbycusis with marked thickening of the basilar membrane that consisted of an increased number of fibrils and an accumulation of amorphous material. Bhatt et al59 observed an age-related thickening of the basilar membrane in the basal end of the cochlea; however, no difference was observed between individuals with hearing loss and age-matched normal-hearing control subjects.

The stria vascularis is active metabolically,60–62 secretes potassium-rich fluid into the endolymphatic space,63–65 and produces the endocochlear potential.66,67 Although acoustic trauma,68 ototoxic drugs,68 and venous obstruction69 have been shown to result in stria vascularis degeneration, the factors that cause age-related degeneration remain undefined. Saxen41 associated atrophy of the stria vascularis with sclerotic changes in its vasculature, and Johnsson and Hawkins70 observed atrophy of the stria vascularis in association with a loss of capillaries in the spiral ligament. Schuknecht et al24 suggested that genetic factors are the most common cause of stria vascularis degeneration. They also stated that stria vascularis atrophy is probably not a secondary effect of vascular lesions, because of a lack of association with generalized vascular disease, its frequent onset at an earlier age, and the absence of histologic evidence for antecedent vascular disease in the capillary system of the stria vascularis.

Through the use of light and electron microscopy, Takahashi71 categorized stria vascularis atrophy into two types: localized and diffuse. In the localized type, severe degenerative changes result in a thin cell layer on the spiral ligament, whereas the adjacent stria vascularis retains its normal appearance. This type of degeneration is often present in the margin adjacent to the spiral prominence and Reissner membrane. In the second type, diffuse atrophic changes are present throughout the entire width of the involved segment of the stria, which may result in irregular areas of decreased strial thickness. Both of these patterns of strial degeneration can be present in normal-hearing ears.

The frequency and importance of stria vascularis atrophy in presbycusis is a contentious issue. Some investigators have attributed less importance to the role of stria vascularis atrophy in presbycusis,8,11,43,44 whereas others have reported that it has great significance.3,7,13,19,24,41,70 Schuknecht3,19 described strial presbycusis as being distinguished by flat or slightly descending pure-tone audiometric thresholds with excellent word discrimination scores; when the total loss of stria vascularis tissue exceeded 30%, a flat audiometric hearing loss was expected.1,7 Although stria vascularis atrophy has been reported by Johnsson and Hawkins70 and Schuknecht et al24 to occur more frequently in the middle and apical turns of the cochlea, it also occurs in the basal turn8,71 (Fig. 12). The endolymphatic space is continuous along the cochlear duct, and therefore activity of the stria vascularis in the basal turn may affect function throughout the cochlea. Schuknecht et al24 agreed with this concept and stated that the flat audiometric pattern of hearing loss that they observed in individuals with a loss of stria vascularis tissue was caused by the chemical changes in the endolymph that impaired the function of the organ of Corti equally throughout the cochlea.

Our study demonstrates that approximately one half of the total stria vascularis volume is contained in the basal turn (approximately 0–15 mm) of the cochlea (Table III). For these reasons, evaluation of the stria vascularis throughout the cochlea may be more appropriate than evaluation of the apical portion alone, as Pauler et al13 reported. Schuknecht1 stated that a 30% loss of stria vascularis tissue was necessary to observe a flat audiometric pattern of hearing loss in individuals with presbycusis. Pauler et al13 reported a statistically significant relationship between the amount of stria vascularis atrophy and the severity of hearing loss in 17 cases of strial presbycusis. However, they did not evaluate the amount of stria vascularis tissue present in the basal turn of the cochlea in their study, which may have compensated for the losses that they reported in the middle and apical turns. In their study, only the apical half of the stria vascularis tissue was evaluated, which suggests that a 15% loss of stria vascularis tissue in the entire cochlea may result in hearing loss, given that they considered the loss of stria vascularis tissue in the basal turn to be negligible. This concept is not supported by a study in aging gerbils in which a loss of up to 70% of the stria vascularis resulted in only a 25% decline in the magnitude of the endocochlear potential.23 In addition, the amount of hair cell loss was reported in only one case by Pauler et al,13 and the ganglion cell population was reduced below the range of their seven normal control subjects in 4 of 17 cases; however, 13 of 17 cases had ganglion cell populations below the range of normal observed in another study10 and in our study.

The criteria defining a flat audiometric pattern of hearing loss were not established in these reports describing strial presbycusis.1,3,7,13,19,24 In the series of 17 cases reported by Pauler et al,13 the pure-tone threshold difference between 500 and 2000 Hz ranged from 0 to 30 dB. In addition, the audiometric thresholds at 4000 and 8000 Hz were not reported. In the report by Schuknecht et al24 describing atrophy of the stria vascularis as a cause of flat audiometric pattern of hearing loss, 10 of the 12 audiograms depicted had a threshold difference between 250 and 8000 Hz greater than 20 dB, and four of these had a 35 dB or greater difference. Also, in all four cases of strial presbycusis reported by Schuknecht,1 a 25 to 35 dB difference in the pure-tone thresholds between 250 and 8000 Hz was depicted. Although these studies have generated a wealth of information concerning presbycusis, the lack of strict audiometric criteria in categorizing presbycusis types and the use of estimates instead of complete quantitative histologic evaluation may lead to false conclusions about the relationship between stria vascularis atrophy and flat audiometric patterns of hearing loss.

In our study, we used strict audiometric criteria in selecting cases for study with flat audiometric hearing losses to minimize the potential for inclusion of cases with mixed features. The six temporal bones selected for our study of presbycusis were from individuals with bilateral, symmetrical, progressive age-related sensorineural hearing loss, a threshold greater than 25 dB, and no greater than a 20 dB threshold difference between the frequencies of 250 Hz and 8000 Hz. These cases of presbycusis were selected solely on the basis of audiometric criteria, and no cases that met these criteria were excluded. Our analysis identified only one case (case 16) with significant atrophy of the stria vascularis; this case also had ganglion cell, inner hair cell, and outer hair cell losses below the range of the normal group. Four cases (cases 11, 12, 15, and 16) in our study had statistically significant mixed losses of cochlear structures, and two cases (cases 13 and 14) had statistically significant outer hair cell loss only when compared with the group of normal-hearing individuals. Three cases (cases 11, 15, and 16) had statistically significant reduced ganglion cell populations, and three cases (cases 12, 15, and 16) had statistically significant inner hair cell losses when compared with the normal group.

The finding common to all cases in the presbycusis group with flat audiometric thresholds was a loss of outer hair cells below the range of the normal-hearing group. When comparing the normal-hearing group with the presbycusis group with flat audiometric thresholds, a statistical difference was observed in the inner hair cell and outer hair cell populations but not in the stria vascularis volume and the ganglion cell population.

These results do not support a correlation between flat audiometric patterns of hearing loss and atrophy of the stria vascularis in individuals with presbycusis and brings into question the prevalence of stria vascularis atrophy as the primary pathologic condition when flat audiometric hearing loss is observed clinically. These results do not exclude atrophy of the stria vascularis as a potential cause of hearing loss. The functional integrity of the stria vascularis is uncertain in archival human histopathologic studies because the ionic composition of the endolymph and the endocochlear potential cannot be measured. However, a study of age-related hearing loss in gerbils, which demonstrated a correlation between reduced immunostaining for Na-K-adenosine triphosphate in the stria vascularis and diminished endocochlear potentials, reported that areas of reduced immunostaining for Na-K-adenosine triphosphatase occurred rarely without observable atrophy of the stria vascularis.23 These findings suggest that quantitative measurements of stria vascularis volume in human temporal bones may accurately reflect stria vascularis function.

The interdependence of one type of cochlear element on another has been discussed in many reports; however, it remains unclear to what extent a primary degeneration of one cochlear element will lead to a secondary degeneration of another cochlear element. Some investigators20,25,42 have concluded that hair cell loss may lead to the degeneration of nerve fibers. Other investigators1,7,19 have suggested that nerve fiber integrity is not dependent on hair cells, but that a loss of supporting cells in the organ of Corti may lead to a loss of nerve fibers in the osseous spiral lamina and, to a lesser extent, a loss of spiral ganglion cells. Developmental abnormalities in humans demonstrate that normal-appearing hair cell populations can be present in the absence of spiral ganglion cells and nerve fibers, although the functional capacity of the hair cells is unknown.72 Saxen41 reported that degeneration of the stria vascularis may result in atrophy of the organ of Corti; however, experimental studies in animals demonstrate that the integrity of the hair cells is not dependent on the presence of the stria vascularis.73 These observations suggest that the integrity of the cochlear elements may be independent of one another.

It remains unclear whether a single factor is common to all individuals with presbycusis, which affects the cochlear elements in variable degrees depending on individual susceptibility, or whether presbycusis is a group of different disease processes that are specific to each cochlear element. Audiometric studies have supported genetic effect on the inheritance of presbycusis.74 Both nuclear75 and mitochondrial76,77 genetic factors have been demonstrated. The hair cells, spiral ganglion cells, and stria vascularis cells have an abundance of mitochondria that may predispose them to this age-related process of degeneration. Mitochondrial genes have a higher spontaneous mutation rate than nuclear genes, and mitochondrial DNA mutations have been proposed as a major factor in the aging of muscle and nervous tissues.78 Although the common mitochondrial deletion has been observed in the inner ear tissues of some individuals with presbycusis,77 a combination of other mutations of the mitochondrial genome that have not yet been evaluated may be even more prevalent.76 These mutations result in a loss of mitochondria function with diminished cellular efficiency and possible cell death.

With these considerations in mind, further studies are needed that involve randomly selected cases of presbycusis with comprehensive light and electron microscopic evaluation of the temporal bones and brains to resolve questions related to hearing loss patterns and the pathology of the auditory system. Certainly, as genetic testing evolves, it will provide a method of identifying individuals who are at risk for developing presbycusis. It is hoped that a better understanding of presbycusis will result in practical methods for its prevention and treatment.

CONCLUSIONS

Investigators have categorized presbycusis on the basis of the shape of audiometric curves. In addition, they have tried to predict the pathologic state of the cochlear elements from the audiometric pattern of hearing loss. It has been reported that in individuals with presbycusis, flat audiometric patterns of hearing loss are related to stria vascularis atrophy. This investigation was undertaken to determine the prevalence of stria vascularis atrophy in individuals with presbycusis and flat audiometric patterns of hearing loss. A method of measuring the total volume of stria vascularis tissue in archival temporal bones was developed to achieve this goal. To our knowledge, a method of measuring the stria vascularis volume has not been reported previously. In this investigation, the cochlear elements in six individuals with presbycusis and flat audiometric patterns of hearing loss were quantitatively evaluated and compared with those of 10 normal-hearing individuals. A summary of our findings follows:

1. Individuals with presbycusis and flat audiometric patterns of hearing loss infrequently have significant atrophy of the stria vascularis.

2. Outer hair cell loss alone or in combination with inner hair cell loss or ganglion cell loss may be the cause of flat audiometric thresholds in individuals with presbycusis.

Acknowledgements

The authors thank Xiling Liu and Professor John Bailor of the Department of Health Studies at the University of Chicago for their assistance with the statistical analysis of the experimental data. Appreciation is expressed for the insight and support of Drs. Harold F. Schuknecht, Cesar Fernandez, Robert I. Kohut, Vijay S. Dayal, and Gregory J. Matz during the course of this investigation and manuscript preparation.

Ancillary