Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Patterns of Cochlear Hair-Cell Loss in Guinea Pigs After Intense Stimulation by Sinusoidal Sound HARLOW W. ADES, CHARLES W. STOCKWELL, AND LYNN B. POCHE University of Illinais HANS ENGSTROM University of Uppsala SUMMARY Guinea pigs were individually exposed to intense sinusoidal sound stimulation of various frequen- cies and varying also in intensity and duration. After a suitable period to allow for degeneration of damaged hair cells, the animals were sacrificed and surface preparations of the organs of Corti were made according to the method of Engstrom. Cochleograms of each organ of Corti were constructed to map the position and condition of each cochlear hair cell. The cochleograms were coded for com- puter reduction of data. Intercomparisons of hair-cell damage were made in terms of variants of the three parameters of the exposure stimulus. Narrow regions of severe to total hair-cell destruction were seen in the ears exposed to higher frequency stimuli. In general, greater damage was seen in outer than in inner hair cells. This dif- ference was greatest in ears exposed to low frequencies, in which extensive outer-hair-cell damage was seen near the apex. Relations between damage and stimulation patterns are discussed in terms of the nonlinear response of the ear to high-intensity stimulation. INTRODUCTION There are two reasons for doing experiments in which damage patterns of cochlear hair cells are mapped after exposing animals to intense sound. The first is the obvious one of trying to determine the parameters of sound stimuli which will produce a given amount and pattern of hair-cell changes. The second rests on the hope of gaining inferential information on the normal pattern of stimulation by sounds of known characteristics, and assumes a direct corre- spondence between damage patterns and stimu- lation patterns. To complete the correlations thus implied, it would be necessary to add to the equation the exact pattern of hearing loss. Evidence on all of these points in the past has 1 This research was supported by NASA grant NGL 14-005-074. ranged from less than adequate to fragmentary. The greatest amount of functional data and the sketchiest amount of anatomical information have come from human experiments, whereas the fuller anatomical evidence, combined with minimal functional evidence, has been derived from animal experiments. The functional evi- dence from both human and animal subjects has been limited largely to the pure-tone audio- gram, a test which falls considerably short of the subtlety needed to discriminate crucially in terms of frequency analytical function. On the other hand, the damage to hair cells in both human and animal cochleas has been studied almost exclusively by Guild's method of graphic reconstruction from serial sections (ref. 1), a method which is severely limited in that it makes accurate orientation difficult, and leaves possibly crucial segments of the organ of Corti unex-
286 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION plored. The net result of many studies using these methods has been to leave in doubt whether intense stimulation by sinusoidal sound produces hair-cell damage which is truly sys- tematic or selective. This stands in marked con- trast to the apparently precise and systematic localization of tonal maxima along the length of the organ of Corti. Together with a growing list of students and associates, we have worked with a method developed in Professor Engstriim's laboratory, and it is called the surface-preparation tech- nique, which makes possible the examination of the organ of Corti in toto from its endolymphatic surface. It is quite feasible to chart all the hair cells, each in its proper position in relation to the rest, and to note their condition (i.e., intact, dam- aged, or replaced by a phalangeal scar). The method was exploited in the present study to explore systematically the relationship between hair-cell damage and frequency analytical func- tion of the cochlea. METHOD Forty-five female guinea pigs approximately 300 grams in body weight were used as experi- mental subjects. They were received from the supplier in groups of 10. In addition to the assurances of the breeder that these animals had been free of disease, had not been previously exposed to harmful acoustic stimulation, or to ototoxic drugs, two further precautions were taken: (1) The animals were isolated from the rest of the colony for at least a week, and (2) one or two animals from each group were used as con- trols, being treated in all respects like the experi- mental group except that they were not exposed to sound stimulation. Three variables of pure-tone exposure were sampled: stimulus frequency, stimulus intensity, and exposure duration. The frequencies in- cluded 4000, 2000, 1000, 500, and 125 Hz. In- tensities included 110, 120, 130,140, and 150 dB. Durations of exposures included 16, 8, 4, 2, and 1 hours; 30, 15, and 7.5 minutes. The original plan called for at least one animal at each com- bination of the variables; however, attrition dur- ing the postexposure period (from disease, acci- dent, and the like) left some of the design blocks unfilled, and generally unequal numbers in the others. Animals were exposed individually in the small plane-wave chamber illustrated in figure 1. This type of enclosure was selected because it provided the most uniform sound field available for the frequencies used. The maximum variation in exposure level, due to changes in head position, was Â±2 dB at 2000 Hz. (I will not go into the detail of the sound-gener- ating and monitoring equipment now. The information is available to anyone who is in- terested. The outline of the apparatus is seen GENERATOR fc, COUNTER VTVM V t ATTEN SPECTRUM ANALYZER T AMPLIFIER AMPLIFIER t t t 0 DRIVER(S) FROM MICRO PHONE GLASS FIBER TERMINATION Flr.URE 1.âSound-exposureapparatus. Top: Block diagram of sound-generating and measuring equipment. Bottom: Cross-sectional view of plane-wave chamber fitted with twin drivers and exponential horn. Microphone openings art not shown. They are located in the front and back walls of the chamber at the level of the animal's head. (VTVM: Vacuum tube voltmeter.)
PATTERNS OF COCHLEAR HAIR-CELL LOSS 287 in fig. 1. Fig. 2 shows the harmonic energy analysis for all the applied tones.) After exposure, the animals were maintained for 4 to 7 weeks to allow time for degeneration of any hair cells damaged by the exposure stimuli. At the end of the survival period, they were anesthetized with pentobarbital sodium and decapitated, and rapid dissection of the ears was carried out as previously described by Engstrrim, Ades, and Andersson (ref. 2). The specimens were fixed and stained in 1.5 percent Veronal-buffered osmic acid solution. Figure 3 shows how the organ of Corti was freed by seg- ments which were mounted in glycerin on a slide, covered, and were then ready for examination. Under phase-contrast illumination, the epi- thelium of the guinea pig's organ of Corti appears as in figure \A. It can be seen that the geo- metrical pattern of sensory cells is highly regular. Extending nearly from the apex to the base, there are three rows of outer hair cells (OHC), with an approximately equal number of cells in each row. The inner hair cells (IHC) form a single row. When a sensory cell has been dam- aged and allowed to degenerate, its position in the mosaic is preserved in the form of a "scar." In view of the high degree of spatial regularity found among these sensory cells, it was a simple, if tedious, matter to literally map the position and condition of each hair cell onto schematic forms that we called cochleograms. These forms underwent continuous evolution in the course of the study; figure 4fi shows the latest version. Each cell was represented by a space in the cochleogram and the number entered in the space specified its condition. The layout of the cochleogram and the numerical code were designed to permit direct punching of computer data cards. A few cells were usually lost through prepara- tion artifact. In some cases of severe damage, where the entire organ of Corti was obliterated, it was necessary to estimate the numbers of hair cells involved. While recourse to estimation was unfortunate, it was imperative to account for all the hair cells so that direct comparison between specimens could be made. The degree of uniformity in numbers of hair cells occupying a given segment of the organ of Corti allows I50 â 130 - ' 4000 Hz 110 : 90 f 1 . 1234 150 - 130 - n 2000 Hz *Â£ 110 â U n n O 90 - 234 to c 150 â â¢o 130 1000 Hz CM O O O 110 _ 6 90 â f1 234 CD TJ 150 â 130 NO - - 500 Hz Q. CO 90 U . i 234 150 - 130 - 125 Hz 110 â Dn 1 90 - 1234 HARMONIC NUMBER FIGURE 2.â A harmonic energy analysis of the applied tones 4000, 2000, 1000, 500, and 125 Hz at 130 dB (solid bars) and 150 dB (open bars). The 150-dB level was not used at 4000 Hz and 500 Hz.
288 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION STAPES BASAL TIP FIGURE 3.â Schematic diagram of the guinea pig's organ oj Corti. showing the extent of segments removed from the bon\ modiolus. Each segment corresponded approximately to one complete coil of the organ of Corti. Also shown are the positions of the stapes and round window. B 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 I I I I I \ I I I I I I I I I I I I I 1 1 I I J I 1 J I 1 1 2 2 I I I I 1 1 I 1 1 1 I 1 I I I I 1 1 1 I 4 i I I I 4 I I 1 1 4 2 FIGURE 4.â A: Phase-contrast micrograph of guinea pig's organ of Corti. B: Cochleogram of the segment shown above. 850 X confidence in the accuracy of estimating missing cells. Each OHC occupies a linear extent of approximately 8.35 microns. Inner hair cells are larger; they occupy a distance of 10.44 microns. When the cochlear partition is pre- pared using the surface-specimen technique, the fragile parts of the organ of Corti are supported by the bony spiral lamina. Thus, when the hair cells have been swept away either by the sound exposure or by clumsy dissection, the spiral lamina remains intact. Therefore, it was usually sufficient to scale the distance along the missing portion to make a fairly accurate estimate of the number of hair cells that were involved. RESULTS Hair-Cell Damage: Comparison Among Groups For purposes of analysis, the organ of Corti was visualized as consisting of a number of seg- ments lying end to end. The segments each contained 50 OHC of each row and 40 IHC. In each segment, the percentage of damaged hair cells was computed separately for each hair-cell row. A certain number of hair cells were un- observable in some segments due to preparation artifact. In these cases, the percentage of damaged cells was calculated on the basis of the number of cells that were observed. When the number of unobservable cells in a particular row exceeded 50 percent of the total number
PATTERNS OF COCHLEAR HAIR-CELL LOSS 289 in that segment, that row of cells was dropped from the analysis of the segment. Hair-cell surfaces are uniform in length and all segments were made to contain an equal number of hair cells, so all segments were nearly equal in length. Therefore, a particular segment of one organ of Corti occupied approxi- mately the same position (with respect to the base) as the corresponding segment of any other organ of Corti. Mean percent damage values were calculated for each hair-cell row in corresponding segments of ears receiving identical exposures. A mean damage curve was constructed for each group of ears by plotting mean damage values of the segments with the segments arranged along the abscissa in the positions they occupy on the organ of Corti. Figure 5 presents the mean damage curves for groups that were exposed to 130-dB sound pressure level (SPL). The mean curves for groups exposed to 150-dB SPL are shown in figure 6, along with the mean damage curve for a group of normal ears. The curves presented in figures 5 and 6 should be interpreted with caution because the vari- ability within groups was large and sample size differed widely, but several trends are noted: A comparison between the amount of damage produced by 130- and 150-dB sound exposures at 125, 1000, and 2000 Hz reveals that, as ex- pected, exposure to 150 dB produced considerably more hair-cell damage than did the comparable 130-dB exposure. A more interesting result is the almost complete loss of damage localiza- tion caused by the more intense exposure at 1000 and 2000 Hz (fig. 6fi through E). The increased sound intensity caused damage to spread primarily toward the base. Hair-cell loss in all rows was nearly 100 percent over the basal 80 percent of the organ of Corti. Hair-cell damage in ears exposed to 125 Hz at 150 dB appears to be restricted largely to outer hair cells. The damage is shown in only one ear exposed for 4 hours (fig. 66'), so conclu- sions here must remain tentative, but this ear showed the most extensive damage to OHC of any ear in the sample, while IHC damage averaged only 11 percent. In ears exposed to the same sound for 1 hour (fig. 6F), a similar effect was seen, but only in the apical two- thirds of the organ of Corti. The tendency for OHC damage to be proportionately greater than IHC damage near the apex appears also in the groups of ears exposed to 130 dB. In figure 5, it can be seen that, when damage is present near the apex, it is confined to OHC. This effect was present in all groups regardless of the exposure frequency. The relationship between stimulation frequency and position of maximum damage is a matter of paramount interest. All place theories depend on a functional relationship between tonal fre- quency and position on the cochlear partition at 1000 Hl- I Hour - i i I i i i 50OH2- I Hour - i i i t I i i i i I i i i i I i i i H 123 Hl - I Hour - n. 20 iIl. 13O dB 40 to 0 100 â¢0 â¢o 40 CO 0 100 â¢0 60 40 to 1000 Hl- 4 Hours - I I I I I I I I 1 I I I I I I I I I ! [ â¢ /v ' â¢' 50O Hi 4 Hours I i i i I I i i i i I I I I I I i i 123 Hz 4 Hours i I i i i i I i i 15 10 5 2O 15 MILLIMETERS FROM BASAL 10 END FIGURE 5. â Mean damage curves for groups exposed to 130- dB SPL. The small arrow above each curve indicates the position of maximum stimulation for the exposure frequency. Blackened areas indicate damage to all four hair cell rows; . IHC; , OHC 1; OHC 2; ...OHC 3.
290 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION â¢troi" A : 80 - - 60 - NORMAL - 40 - - 20 "MiII i i i i 1 i i i i 1 > i i i~ O 20 13 10 ISO dB 15 10 5 20 15 10 MILLIMETERS FROM BASAL END FIGURE 6. âMean damage curves for groups exposed to 150- dB SPL. The small arrow above each curve indicates the position of maximum stimulation for the exposure frequency. Blackened areas indicate damage to all four hair cell rows; . IHC; OHC 1; , OHC 2; OHC .'). The mean curve for 12 normal (unexposed) ears is shown at the top of the figure. which it vibrates with maximum amplitude. Evi- dence of such a relationship has been provided by Steinberg (ref. 3), Culler et al. (ref. 4), and von Bekesy (ref. 5). The frequency "maps" of the cochlear partition developed by these in- vestigators differ somewhat in detail, but agree in showing an approximate inverse relation be- tween the logarithm of frequency and distance from the base. Greenwood (ref. 6) constructed an empirical function, based on von Bekesy's observations (ref. 5), which relates frequency to the position of maximum vibration, and is prob- ably the best available expression of the relation- ship between tonal frequency and position of maximum vibration on the cochlear partition. It is, nevertheless, only an approximation and subject to some errors of assumption. One way of examining the relation between hair-cell stimulation and hair-cell damage is to compare points of maximum stimulation and maximum damage in relation to stimulus fre- quency. Figure 5 shows that the position of maximum damage bears some relation to ex- posure frequency. As frequency is increased, the position of maximum damage moves toward the base. The small arrow above each curve in the figure indicates the position of maximum stimulation (calculated with Greenwood's for- mula) for the exposure frequency of that group. Exposure to 2000 and 4000 Hz (fig. 5,4 through C) produced damage that was most sharply localized. The agreement between stimulation and damage maxima at these frequencies was reasonably close, but not impressive. Exposure at 1000 Hz (fig. 5D and E) produced less localized damage than did exposure to higher frequencies, and maximum damage was closer to the position stimulated by 2000 Hz than to the position stimu- lated by 1000 Hz. Exposures at 500 and 125 Hz (fig. 5f through /) produced IHC damage that was slight and whose maxima consistently fell at points basalward from the places supposedly receiving maximum stimulation from the exposure tones. Damage confined to OHC appeared near the apex. It appears that correspondence between maxi- mum stimulation and maximum damage positions on the organ of Corti is best for higher frequen- cies (where maxima are sharper). If a frequency "map" of damage maxima were made, the separation between frequencies in general would be less than it is in the frequency map of stimu- lation maxima, with the lower frequencies, in particular, shifting toward the basal end. How- ever, since the damage maxima are not very sharp in these mean curves, support for this statement cannot be very strong. To determine whether 4-hour exposures cause more hair-cell damage than 1-hour exposures at 130 dB, the groups exposed to 2000, 1000. 500, and 125 Hz were combined and a Mann- Whitney Latest (see Siegel, ref. 7) was applied to compare effects of the two exposure dura- tions. Each hair-cell row was compared separately. The large sample size (ni = 16, n2 = 24) justified referring the obtained V-
PATTERNS OF COCHLEAR HAIR-CELL LOSS 291 value to the normal distribution. The 0.01 significance level was chosen and a one-tailed test was used. Damage to IHC in ears exposed for 4 hours was not significantly greater than damage sustained after 1-hour exposure; how- ever, damage in all three rows of OHC con- tinued to increase after 1 hour. Differences in the amount of outer-hair-cell damage between 1 and 4 hours were significant beyond P= 0.001. The small number of ears in some groups exposed at 150 dB do not permit any statistical treatment, but the curves in figure 6 suggest that, for 125-Hz exposures, the picture is similar to that seen at 130 dB. IHC damage appears essentially complete after 1 hour, but OHC damage continues to grow between 1 and 4 hours. At 1000 and 2000 Hz, no appreciable additional damage in any hair-cell row was sustained after an exposure of 1 hour. It ap- pears that hair cells which are susceptible to damage at these frequencies have already sustained damage after 1 hour when the intensity islSOdB. Variability was noted in amount or pattern of damage among animals with similar exposures, and even between the two ears of individual subjects. I should like to call attention to one type of variation; namely, that found in the dis- tribution of damage along the organ of Corti. The mean curves presented in figure 5 generally show poorly localized hair-cell loss. Consider, for example, figure 5E which represents the mean curve for ears exposed to 1000 Hz at 130 dB for 4 hours. IHC loss was most severe in the middle regions, and OHC loss was spread rather evenly in the apical half of the organ of Corti. In contrast with their mean curve, the individual damage curves for these ears, shown in figure 7, display much more localized hair-cell loss occurring in a series of sharp peaks. The occurrence of these peaks in different places in the individual ears is expressed by rather broad maxima in the mean curve. In five of the six ears, there was a single sharp peak of damage involving both IHC and OHC. In four of these (83R, 83L, 88R, and 132L), the peak was located 10 to 11 mm from the base. The fifth (132R) displayed a peak at 13 mm from the base. In the remaining ear (88L), there d I I z â¢ o r i i 80 60 40 n ^f z 1- z u : 83L : 20 83R I at LLl 1 1 1 1 1 1 1 1 1 1 1 1 1 1 IJ L n I 1 1 1 1 1 1 1 I 1 1 1 III 1 I 1 1 MM FROM BASAL END MM FROM BASAL END (T U e 20 19 10 5 MM FROM BASAL END 20 15 10 9 MM FROM BASAL END I32L I I l I I I I I 20 19 10 9 MM FROM BASAL END 100 â¢0 60 40 to 0 " i n i i u.ujj.i i .i- LI..I i j i 20 19 10 9 MM FROM BASAL END FIGURE 7.âIndividual damage curves for ears exposed to 1000 Hz at 130 dB for 4 hours. The small arrow above each curve indicates the position of maximum stimulation for 1000 Hz. Blackened areas indicate damage to all four hair cell rows; , IHC; , OHC I; , OHC 2; OHC 3. were three such peaks, located at 9, 10.5, and 12 mm fr.om the base. These peaks represent areas of total damage on the organ of Corti. The existence of sharply defined areas of total damage was a common finding in the 56 ears exposed to 130 dB. A photomicrograph of such an area appears in figure 8. In 29 of these ears, a single totally damaged area was found, while in an additional 7 ears, there were 2 or more such areas spaced at intervals along the organ of Corti. These areas varied in extent from 0.1 to 1.5 mm. They could readily be identified as soon as the cochlea was opened during dissection. Even under low-power magnification, an inter- ruption was observed in the organ of Corti where hair cells, pillar cells, and supporting elements had been swept away. In addition, degenera- tion of the radial nerve supply was plainly visible.
292 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION FIGURE 8. â Phase-contrast micrograph of a restricted region of total hair-cell damage located about 9 mm from the base. This was one of several such areas seen in an ear exposed to 2000 Hz at 130 dB for 4 hours. Displacement of cells in the region to the right of the totally damaged area is the result of dissection artifact. 350 X The positions on the organ of Corti of total destruction define, as closely as possible, points of maximum damage with which to compare maxi- mum stimulation points predicted by Green- wood's function. The comparisons are made in figure 9 for ears that were exposed at 130 dB. Sixteen ears were exposed at 4000 Hz; of these, 12 ears contained one or more totally damaged areas. Figure 9 shows that areas of total damage in these 12 ears are clustered about a point ap- proximately 9 mm from the base and slightly apical to the point of maximum stimulation for 4000 Hz. Three more such areas lie nearer the base at intervals of about 1 mm; these three areas were contributed by a single ear that also con- tained an area of total damage between 8.1 and 9.6 mm from the base. Each of the 10 ears ex- posed at 2000 Hz contained one or more areas of total damage. (This group, as well as the ones exposed at 1000 and 500 Hz, includes the ears exposed both for 1 hour and for 4 hours.) Total damage areas for the ears exposed at 2000 Hz were clustered at approximately the same place as total damage in ears exposed at 4000 Hz and predominantly below the maximum stimulation position predicted by Greenwood's function for 2000 Hz. Nine of ten ears exposed at 1000 Hz showed areas of maximum damage that occupied positions between 8.75 and 12.75 mm from the base. Among 12 ears exposed at 500 Hz, only five showed an area of total damage; all of these areas were well below the point of maximum stimulation for 500 Hz. None of the ears ex- posed at 125 Hz showed areas of total damage. Place theories assume that ears exposed to the same frequency will receive maximum stimula- tion at the same point on the basilar membrane. Greenwood's function should probably be re- garded as a rule of thumb at best, and so some deviation between predicted stimulation maxima â¢BtEMWOOC'S FUNCTION MILLIMETERS FROM BASAL END FIGURE 9. â Histograms of totally damaged areas for groups exposed at a 130-dB SPL. Ears exposed both for 1 and for 4 hours are included. Open circles indicate the position of maximum stimulation for the exposure frequency. See text for information on sample size of the various groups.
PATTERNS OF COCHLEAR HAIR-CELL LOSS 293 and observed damage maxima would not be in- consistent with a statement that hair-cell damage mirrors hair-cell stimulation; however, maximum damage points for ears exposed to the same fre- quency are spread over 3 to 5 mm, which, if Greenwood's function is correct, corresponds to a frequency range of almost two octaves. It seems improbable that the variation in a fre- quency localization between ears is that large; moreover, the distribution of maximum damage with respect to exposure frequency, as seen in figure 9, clearly departs from the distribution pre- dicted from Greenwood's function. As exposure frequency decreases, the damage maximum falls increasingly basalward in relation to the stimula- tion maximum. Less than half the ears exposed at 500 Hz and no ears exposed at 125 Hz showed any areas of total damage. In these ears, damage was confined largely to OHC. In general, damage was not so sharply localized as it was in the ears exposed to higher frequencies, but rather tended to be distributed in multiple peaks. The ears exposed to 1000 Hz (fig. 7) show best the char- acteristics of damage both from higher and lower frequency exposure. Besides the localized areas of total damage characteristic of high frequencies, these ears display damage limited largely to OHC near the apex, which is charac- teristic of low frequency exposures. The prominence of multiple damage peaks seen in these ears exposed to a pure tone stimulus pro- vides further argument against a simple corre- spondence between stimulation and damage patterns. Hair-Cell Damage: Comparisons Among Hair-Cell Rows Three differences between damage to IHC and OHC have already been mentioned: (1) Damage to OHC continued to increase after exposure for 1 hour, whereas IHC damage did not. (2) OHC sustained more damage than IHC. This was true at all points along the organ of Corti in ears exposed to 130 dB; however, excep- tions appear in two groups exposed to 150 dB. Two ears exposed at 1000 Hz for 4 hours (seen in fig. 6E) showed more damage to IHC than to OHC on either side of the totally damaged region. In the ears exposed to 125 Hz for 1 hour (fig. 6Â£). IHC damage exceeded damage to OHC between 4 mm from the base and the basal tip. In all other groups exposed to 150 dB, OHC damage was either equal to or exceeded IHC damage in all regions of the organ of Corti. (3) The difference between the amount of of damage to IHC and OHC became progressively larger toward the apex. This effect appeared in groups of ears exposed at 130 dB (shown in fig. 5). It is part of a more general relationship between the radial distribution of damage and distance along the organ of Corti. Damage nearer the base is more likely to involve hair cells lying closer to the modiolus than is damage located farther along the organ of Corti. This relationship can be seen more clearly when, in addition, one considers differences among the three individual rows of OHC. Hair- cell damage in ears exposed to 4000 Hz (fig. 5/4) was confined to the middle region of the organ of Corti. In this region, the most severely damaged OHC row was the innermost row, OHC 1. On the other hand, ears exposed to lower frequencies showed damage near the apex that was increasingly confined to OHC. In these groups, the outermost row, OHC 3, was the one showing the most severe damage. For example, in ears exposed to 500 and 125 Hz (fig. 5F through /), damage was least severe in OHC 1, greater in OHC 2, and still greater in OHC 3. Distance along the organ of Corti is not, however, the only variable related to these differences in the radial distribution of damage. The above comparisons were made among groups of ears that were exposed to tones of different frequencies. Perhaps radial effects are related to distance along the organ of Corti because both are a function of the same variable; that is, frequency of exposure. The damage curves in figure 6F and G suggest that this is the case. The ears shown here were exposed to 125 Hz at 150 dB. Differences among OHC rows are not seen, but these ears show the large differences between OHC and IHC damage that are characteristic of ears exposed to low frequencies. They failed to show appreciable
294 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION IHC damage even in the middle region of the organ of Corti, where IHC damage had become prominent in other ears exposed to higher frequencies. The low-frequency tone was capa- ble of producing damage, largely confined to OHC, in all regions of the organ of Corti. This result suggests that exposure frequency rather than position along the organ of Corti is the relevant variable determining the radial dis- tribution of damage. DISCUSSION In the introduction, it was suggested that previous studies failed to reveal selective cochlear lesions because the method of graphical reconstruction from serial sections precluded precise assessment of hair-cell damage, particularly when damage was relatively light. Use of the surface-preparation technique per- mitted a more complete assessment of hair- cell damage than was possible before. More selective and discontinuous hair-cell loss was seen than has been reported previously, par- ticularly in ears in which damage was relatively slight. Presumably this feature of hair-cell damage went unnoticed in ears prepared by the traditional method. In general, the gross pattern of hair-cell damage seen in the present material affirms the conclu- sions based oh earlier observations (refs. 8 to 10). Lesions first appeared near the stimulation maximum for the exposure frequency. More intense exposure caused lesions to spread from that point toward the base. It should be noted that Elliott and McGee (ref. 11) described a different pattern of spread in cats. They stated, "The spread of damaging effects is at least as great toward the apical end as toward the basal end of the cochlea, and possibly greater." A species difference cannot be completely ruled out, but our results support the conclusion that the spread of hair-cell damage is primarily toward the base. The greater susceptibility of OHC than IHC to damage has been reported consistently, but the observation that the radial distribution of damage depends upon exposure frequency has not been mentioned before. Beagley (ref. 12) has stated that OHC 1 was the most damaged OHC row in guinea pig ears exposed to a 500-Hz tone. It was shown in the present study that OHC 1 was most severely damaged in ears ex- posed to 4000 Hz, but, contrary to Beagley's results, OHC 1 showed the least damage of any OHC row after exposure to 500 Hz. Other in- vestigators using the surface-preparation tech- nique (refs. 2 and 13) have demonstrated dif- ferential radial damage, but they noted no systematic trends. Other results reported here are relevant to two of the more general problems that plague in- vestigators of auditory damage: (1) Individual ears vary greatly in their susceptibility to damage, and (2) hair-cell damage and hearing losses do not correspond to exposure parameters in a simple way. The first problem is of considerable clinical importance. Large individual differences re- tard progress in establishing workable damage- risk criteria. Extensive research has been di- rected toward predicting individual susceptibility to noise damage using temporary threshold shift as a predictor, but only limited success has been achieved (ref. 14). Development of more suc- cessful predictors has been limited largely by the lack of knowledge concerning the relevant variables. In this regard, it is worthwhile to point out that, in the present study, ears exposed to highly controlled acoustic stimulation never- theless exhibited large differences in the amount of hair-cell damage. A considerable portion of the variability occurred between the two ears of the same animal. The magnitude of the in- dividual differences seen in the present data indicates that a significant source of variation lies between the entrance of the ear canal and the receptors. A search for factors producing this variation could contribute to successful clinical prediction of susceptibility to hearing loss. The other major problem that concerns investi- gators of hearing loss arises as a result of the expectations of place theory; each frequency is presumed to cause maximum stimulation at a specific site on the organ of Corti. Hair cells that receive maximum stimulation by a particular frequency should be the first ones to be damaged when the intensity is raised: however, the greatest hearing loss usually occurs one-half
PATTERNS OF COCHLEAR HAIR-CELL LOSS 295 to one octave above the exposure frequency (ref. 15). Even more disturbing is the fact that a variety of nonsinusoidal stimuli, e.g., impulses and wideband noise, produce their greatest losses in the region of 4000 Hz. Various hypotheses have been offered to account for this. In general, they postulate that, for various anatomical reasons, the region of the organ of Corti focally receptive to 4000 Hz is either more vulnerable to damage or is the site at which particularly destructive forces develop. The results of the present study showed that tonal stimuli caused damage that was generally related to the site of maximum stimulation for the exposure frequency. Damage was not confined to the region of the organ of Corti stimulated by 4000 Hz. If the hypothesis were valid, an indication of special susceptibility in that region should have appeared in ears exposed at low frequencies. Von Bekesy (ref. 5, p. 504) found that the basilar membrane of the guinea pig vibrates in phase when stimulated by a pure tone below the frequency of 200 Hz. Thus, at 125 Hz, stimula- tion along the entire organ of Corti should be fairly uniform, and. if the region receiving 4000 Hz is especially susceptible to damage, then damage should be most severe in that region. According to the present results, there was, in fact, no evidence of especially heavy damage in the 4000-Hz region in such ears. Therefore, those hypotheses designed to account specifically for the effects of non- sinusoidal stimulation do not appear relevant to a discussion of damage patterns caused by pure tones. Two general notions exist as to the mechanism of hair-cell damage. The first of these attributes hair-cell damage to stresses within the cochlea sufficient to produce mechanical injury. Schu- knecht and Tonndorf (ref. 16) have made the fol- lowing analysis of the distribution of mechanical stress on the cochlear partition. Stress is de- fined as mass times acceleration per unit area. Schuknecht and Tonndorf assumed that the mass of the cochlear partition is approximately the same over its whole extent; therefore, the stress upon any segment of the partition during its displacement is directly proportional to its acceleration. Acceleration depends on the ratio of the displacement amplitude and the square of the period of vibration. By analyzing time-displacement patterns in cochlear models, Schuknecht and Tonndorf demonstrated that the period of vibration is least at the basal end of the cochlear partition and increases expo- nentially thereafter. Since high frequencies produce maximum displacement amplitudes nearer the base, it follows that high frequencies produce greater acceleration, and therefore greater stress, in the region of maximum ampli- tude than do lower frequencies. High frequen- cies should cause more hair-cell damage than low frequencies by the mechanism of mechanical stress. The other notion is the so-called physico- chemical theory which attributes hair-cell damage to exhaustion of cytochemical or enzy- matic materials as a result of prolonged exposure to acoustic stimuli (ref. 17). This idea is sup- ported by evidence of ultrastructural changes following fairly intense acoustic stimulation (refs. 18 to 21). The earliest change was an increase in the number of osmiophilic inclusions within apices of OHC that were interpreted as products of intensified oxidative metabolism. More prolonged stimulation produced additional morphological alterations thought to be associated with a lack of oxygen in the inner ear. These changes were most prominent in OHC. Koide et al. (ref. 20) demonstrated a reduction in oxygen tension in the perilymph during exposure to loud acoustic stimulation. Thus, two mechanisms of hair-cell damage have been postulated: (1) Mechanical stress, which is thought to be more important at higher fre- quencies; and (2) exhaustion of metabolic mate- rials, which is thought to depend more on ex- posure duration and to involve mainly OHC. The present results suggest that both of these mechanisms must be adduced to account for the hair-cell damage that was observed. Two find- ings are relevant: (1) IHC loss was essentially complete after 1 hour of exposure, but OHC loss continued to increase between 1 and 4 hours, indicating that OHC are more sensitive than IHC to gradually accumulating effects of exposure; (2) OHC loss was proportionately greater than
296 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION IHC loss at low stimulating frequencies (see figs. 5 and 6), indicating that frequencies some- how differ in their effect. Both of these observa- tions together suggest the following interpre- tation: At higher frequencies, mechanical stress was the most important mechanism of hair-cell damage. Injuries were seen both to OHC and to IHC, although OHC were affected more. On the other hand, lower frequencies at the same SPL developed less mechanical stress, so this factor was relatively less important as a cause of hair-cell damage: hence, most IHC survived low- frequency exposure, while OHC were damaged in greater numbers. This fact and the fact that OHC damage increased between 1 and 4 hours suggest that OHC are more sensitive to gradually accumulating effects of stimulation, and that this factor was a more important mechanism of dam- age at lower frequencies. In any case, formal analysis of hair-cell-damage patterns depends on the ability to assign fre- quencies to specific locations on the organ of Corti in the present material. Accordingly, the interpretations offered here must remain qualita- tive and tentative. They are based, for the most part, on indirect evidence. It is apparent that complex relations exist between the patterns of hair-cell damage and patterns of stimulation. In the fuller elaboration of these relationships, particular attention must be paid to the ear's non- linear response to high-intensity stimulation and to differential susceptibility between ears, and perhaps between different regions of the cochlea, to various stimulation parameters. REFERENCES 1. GUILD, S. R.: A Graphic Reconstruction Method for the Study of the Organ of Corti. Anat. Record, vol. 22, 1921, pp. 141-157. 2. ENGSTROM, H.; ADES, H. W.; AND ANDERSSON, A.: Structural Pattern of the Organ of Corti. Almqvist & Wiksell, Stockholm, 1966. 3. STEINBERG, J. C.: Positions of Stimulation in the Cochlea by Pure Tones. J. Acoust. Soc. Am., vol. 8, 1937, pp. 176-181. 4. CULLER, E. A.; COAKLEY, J. D.; LOWY, K.; AND GROSS, N.: A Revised Frequency Map of the Guinea-Pig Cochlea. Am. J. Psychol., vol. 56, 1943, pp. 475-500. 5. VON BEKESY, G.: Experiments in Hearing. McGraw- Hill, I960. 6. GREENWOOD, D. D.: Critical Bandwidth and the Fre- quency Coordinates of the Basilar Membrane. J. Acoust. Soc. Am., vol. 33, 1961, pp. 1344-1362. 7. SIEGEL, S.: Nonparametric Statistics. McGraw-Hill, 1956. 8. COVELL, W. P.: Histologic Changes in the Organ of Corti With Intense Sound. J. Comp. Neurol., vol. 99, 1953, pp. 43-49. 9. LURIE, M. H.; DAVIS, H.; AND HAWKINS, J. E., JR.: Acoustic Trauma of Organ of Corti in Guinea Pig. Laryngoscope, vol. 54. 1944, pp. 375-386. 10. WEVER, E. G.: AND SMITH. K. R.: The Problem of Stimulation Deafness: 1. Cochlear Impairment as a Function of Tonal Frequency. J. Exptl. Psychol., vol. 34, 1944, pp. 239-245. 11. ELLIOTT, D. N.; AND McGEE.T. M.: Effects of Cochlear Lesions Upon Audiograms and Intensity Discrimina- tions in Cats. Ann. Otol., vol. 74. 1965, pp. 386-408. 12. BEAGLEY, H. A.: Acoustic Trauma in the Guinea Pig. I. Electrophysiology and Histology. Acta Oto- Laryngol., suppl. 60, 1965, pp. 437-451. 13. BREDBERG. G.: Cellular Pattern and Nerve Supply of the Human Organ of Corti. Acta Oto-LaryngoL suppl. 236, 1968. 134 pp. 14. WARD. D. W.: Auditory Fatigue and Masking. Modern Developments in Audiology, J. Jerger, ed.. Academic Press, 1963, pp. 240-286. 15. KRYTER, K. D.: The Effects of Noise on Man. J. Speech Dis.. suppl. 1, 1950. 16. SCHUKNECHT. H. F.; AND ToNNDORF, J.: Acoustic Trauma of the Cochlea From Ear Surgery. Laryngo- scope, vol. 70, 1960, pp. 479-505. 17. SALMIVALLI. A.: Acoustic Trauma in Regular Army Personnel. Acta Oto-Laryngol., suppl. 222. 1967. 85pp. 18. BEAGLEY. H. A.: Acoustic Trauma in the Guinea Pig. II. Electron Microscopy Including the Morphology of Cell Junctions in the Organ of Corti. Acta Oto- Laryngol., suppl. 60, 1965, pp. 479-495. 19. ENGSTROM, H.; AND ADES, H. W.: Effect of High Intensity Noise on Inner Ear Sensory Epithelia. Acta Oto- Laryngol.. suppl. 158. 1960, pp. 219-229. 20. KOIDE, Y.: YOSHIDA. M.; KONNO. M.; NAKONA. Y.: YOSHIKAWA, YL: NAGADA. M.: AND MORIMOTO. M.: Some Aspects of the Biochemistry of Acoustic Trauma. Ann. Otol.. vol. 69. 1960, pp. 661-697. 21. SPOENDLIN. H.: Ultrastructural Features of the Organ of Corti in Normal and Acoustically Stimulated Animals. Ann. Otol., vol. 71, 1962. pp. 657-677.
PATTERNS OF COCHLEAR HAIR-CELL LOSS 297 DISCUSSION Bredberg: Previous studies on the pathology of the organ of Corti have almost completely failed to show multiple areas of damage. I should like to show a few slides from a study on the human cochlea (G. Bredberg: Cellular Pattern and Nerve Supply of the Human Organ of Corti. Acta Oto- Laryngol., suppl. 236, 1968, 134 pp.) illustrating some cases with multiple areas of degeneration. The nerve supply in the normal human cochlea appears macroscopically to be evenly distributed throughout all coils except in the most basal few millimeters of the basal coil. Figure Dl shows the nerve supply in the basal coil in a case of noise exposure, and figure D2 shows the cochlea from a woman 86 years of age. From my material, it appears that neural damage was most commonly distributed in an uneven, "patchy" way. FIGURE Dl. â Basal coil of the right cochlea from a man 71 years of age, exposed during life to high-intensity noise in a sawmill for many years. The cochlea shows an almost complete degeneration of radial nerves in the osseous spiral lamina as well as a corresponding degeneration of the organ of Corti in the region between 10.5 mm and 14.0 mm from the base. A few thin radial-nerve bundles are found in this area of otherwise complete degeneration (arrows 2 and 3). In addition there are a few narrow areas showing nerve loss (arrows 1, 4, and 5).
298 THE ROLE OF THE VESTIBULAR ORGANS IN SPACE EXPLORATION FIGURE D2. â Basal coil of the left cochlea from a woman aged 86 years, with no known excessive noise exposure. The cochlea shows an overall degeneration of nerve fibers, most marked in the base. Moreover, there is a patchy degeneration of nerves and the corresponding areas of the organ ofCorti [arrows). The etiology of this patchv loss is unknown.