gms | German Medical Science

ACh Acetylcholine

AMPA α-amino-3-hydroxy-5-methyl-4-isoxazolproprionate

BBN Broadband noise

BDNF Brain-derived neurotrophic factor

CAP Compound action potential

CNS Central nervous system

DA Dopamine

DPOAE Distortion products of otoacoustic emissions

FGF Fibroblast growth factor

GABA γ-amino-butyric acid

GDNF Glial-cell-line derived neurotrophic factor

HCN Hydrogen cyanide

HSP Heat shock proteins

IHC Inner hair cells

kHz Kilohertz

MAK Maximale Arbeitsplatzkonzentration (maximum workplace concentration)

(M)OCB (Medial) olivocochlear bundle

MOCR Acoustic reflex of the medial olivocochlear bundle

NBN Narrow band noise

NMDA N-methyl-D-aspartate

OHC Outer hair cell

PTS Permanent threshold shift

RWM Round window membrane

ROS Reactive oxygen species

SPL Sound pressure level

TDT Tinnitus desensitization therapy

TEOAE Transient evoked otoacoustic emissions

TRT Tinnitus retraining therapy,

TTS Temporary threshold shift

(S)SNHL (Sudden) sensorineural hearing loss

Noise numbers amongst the most frequent causes of acquired hearing damage. Despite suitable measures for reducing noise pollution, e.g. at the workplace, by legally enforced noise level limits as well as numerous measures for preventing the origin of noise and its propagation, many adults, adolescents and children expose themselves both voluntarily and involuntarily to noise on a long-term basis. Approximately 25-30 million workers in Europe work in an environment with excessive noise levels [1], with members of some occupational groups are exposed to particularly extreme noise stress. The increasing amount of noise exposure outside the workplace caused by recreational/leisure activities is also added to that [2], [3], [4], [5], [6], [7]. The latter affects not only adults, but also babies, small children, schoolchildren and adolescents.

Noise can lead to inner-ear damage with temporary or permanent sensorineural hearing loss (SNHL) and tinnitus. Hearing loss and tinnitus can also occur separately. It is the only proven form of health damage caused by noise. Audiologically, a noise-induced hearing loss is expressed as a threshold shift in the pure tone audiogram, by a recruitment of above-threshold audiometry, as amplitude reductions or losses in otoacoustic emissions, and as a loss in speech intelligibility in the speech audiogram [8]. Tinnitus can appear in a compensated (subjective coping possible) or decompensated (subjective coping not possible) manner.

Exposure to continuous noise, impulse noise or blast injuries are the main culprits involved in bringing about SNHL. Impulse noise can lead to more severe damage to the inner ear than continuous noise can [9]. Important criteria for the development of noise-related injury are the sound pressure levels (SPL), the rapidity with which sound levels increase, the exposure time and individual sensitivity, i.e. the so-called vulnerability of the inner ear. Depending on these factors, a noise exposure may lead first to a temporary threshold shift (TTS) and/or tinnitus. If a complete restitution after a temporary threshold shift/ tinnitus does not occur in the recovery phase, a permanent inner ear injury results ("permanent threshold shift" = PTS). However, an immediate onset of PTS and/or permanent tinnitus is also possible. Smaller permanent damage to the inner ear can also already start without it being able to be recorded using conventional audiological diagnostics.

Temporary and permanent threshold shifts as well as tinnitus can be traced primarily to partly definable functional changes in the acoustic organ (organ of Corti) of the inner ear. Clinical and experimental studies have shown that noise can result above all in direct mechanical injury of the hair cells, whereby the outer hair cells are particularly vulnerable. The normally stiff and upright-standing stereocilia can soften and even collapse [8]. It is also well known that cochlear injury can arise from metabolic decompensation, the consequences of which may be cellular apoptosis in the inner ear. Upon noise of a very high intensity an immediate necrotic cell loss and mechanical avulsion can occur in the inner ear. Apart from the purely mechanical damage due to the acoustic hyperstimulation, the potassium release from the endolymph caused by the injury can lead to an intoxication and death of other sensory and nerve cells (review in [8]).

The conspicuous vulnerability of the outer hair cells is a reflection of the fact that the cochlear amplifier located in the outer hair cells and thus the "motor" for the active traveling wave movements is extremely vulnerable to noise [8]. The disrupted function of the outer hair cell cochlear amplifier after noise overexposure results in a less amplified traveling wave propagation. This leads to a decrease in sensitivity and frequency selectivity in the inner ear. Audiologically it can be recognized by a decrease or loss in otoacoustic emissions, a positive recruitment, and a deterioration in hearing thresholds (TTS, PTS) and speech comprehension.

Chronic hearing loss acquired in the workplace or during leisure time is important not just for one's development in private life; it can also have a negative impact on occupational life. Our society is characterized by many professions in which linguistic information needs to be conveyed directly or via the telephone or headphones; for such professions adequate hearing is an absolute necessity. Because of the large number of people affected, hearing loss from noise represents a serious national health problem despite the continuing pace of scientific progress.

Extensive literature already exists regarding the anatomical and physiological consequences of excessively loud noise exposure. See for example: Dieroff (1994) [10], Hawkins (1973) [11], Henderson and Hamernik (1995) [12], Lehnhardt and Koch (1994) [13], Pfander (1975) [14], Saunders et al. (1985) [15], Saunders et al. (1995) [16] and Zenner (1994) [8].

Over the last decades, numerous aspects of noise-induced hearing loss have been examined. These include the acoustic properties and time-course of the noise, audiometric findings and subclinical changes in hearing function, as can be shown for example by modifications in otoacoustic emission, histological alterations in the cochlea after noise-induced injury, and pathophysiological changes in the hair cells, auditory nerve cells and central sections of the auditory pathway with special consideration of the plasticity of the auditory system. Current research projects for studying acute acoustic trauma and chronic hearing loss unite modern molecular biology and genetics procedures with anatomical and system-physiological studies. This link-up should create the conditions to identify novel interventional procedures for preventing and treating acute acoustic traumas and chronic hearing loss. For the development of such preventive and therapeutic measures one requires a detailed understanding of the pathophysiological processes occurring with cell damage and death, as well as during recovery and repair in general and especially in the auditory system.

The goal of this review is to update and expand upon the already existent knowledge base regarding noise-induced hearing loss, so that novel aspects of hearing loss from noise can be dealt with in particular. However, gaps in our knowledge as well as failings in our health care system shall also be given adequate consideration.

Despite the pace of progress that has been enabled by extensive research, a range of questions remain unanswered. These concern especially (1) the great variety of cellular mechanisms leading to hearing loss (2) the possibilities for physically and pharmacologically preventive, protective and therapeutic intervention, (3) the early recognition of hearing loss, (4) the identification of individuals with a noise vulnerability of the inner ear using suitable investigative procedures and (5) the role of different co-factors. Numerous options are in principle conceivable for reducing noise-induced hearing loss. Both pre- and para-exposure preventive as well as post-exposure therapeutic measures can be considered. They are summarized in table 1 [Tab. 1].

It is well known that exposure to chemicals alone can lead to hearing disorders. However. chemicals in combination with simultaneous noise exposure can lead to injurious effects in the auditory system that are either additive or synergistic in nature. Since hearing loss can also arise at noise levels accepted as innocuous if there is a simultaneous or metachronous exposure to certain chemicals, existing control procedures and workplace analyses no longer suffice under certain circumstances, and these have to be reassessed. Furthermore, the growing sociomedical importance of hearing loss from leisure noise shall also be a topic of this review. It is primarily young people who suffer from this kind of permanent hearing loss, and health care systems have failed in the planning of plausible, purpose-built and adequate protective measures.

Findings from molecular-biological and molecular-genetic studies on the pathophysiology of noise-related injury open new perspectives to be derived for preventing and treating this disorder. Pharmacological studies have pointed towards potential substance classes that might intervene in cellular signal cascades leading to cell damage or cell death, or that might promote endogenous protective mechanisms and mechanisms for repair/recovery and thereby expose interesting aspects for future therapeutic intervention. Another approach worthy of discussion is the idea of reducing the sensitivity of the auditory organ towards harmful sound exposure by pre-exposing it to innocuous sound levels (sound conditioning or toughening).

The risk of suffering irreversible inner ear damage through noise increases with its amplitude and duration. Studies on large collectives of noise-exposed individuals in industry have shown that a dose-effect relationship is evident that allows a fairly reliable statistical prediction of the average impairment of hearing to be expected [17], [18]. The risk of injury varies with frequency. The typically damaged frequencies ranges are around 4 kHz followed by the neighboring frequencies of 3 and 6 kHz [19].

When measuring noise exposure levels several aspects like frequency and time weighting of the exposure need to be considered.

Sound level meters measure acoustic pressure and by international agreement they are calibrated in decibels (dB). The sound pressure level (Lp) is defined as Lp = 20log(P/P₀), with P is the measured root mean squared (rms) sound pressure and P₀ is the reference rms sound pressure (20µPa). Because the ear is relatively insensitive to very low and very high frequencies, sound level meters have weighted frequency responses. The frequency weighting 'A' best expresses the human ear's response to loudness. Practically all noise is measured using the 'A' filter. The sound pressure level in dB(A) gives a close indication of the subjective loudness of the noise. There are other frequency weighting curves like the 'C' and 'Z' for special applications. The fact that noise levels are usually fluctuating and not constant is considered in the "equivalent continuous sound pressure level" (Leq). Leq is the steady sound level that, over a specified period of time, would produce the same energy equivalence as the fluctuating sound level actually occurring. Impulse noise is noise having a high peak of short duration or a sequence of such peaks. If a noise contains a large amount of impulses, this needs to be especially considered, since impulse noise exerts an increased risk for damage of the cochlea.

Personal noise exposure (Lex, also known as Lepd), usually referrers to a daily 8 hour noise exposure rate.

If measured by conventional pure tone audiometry, sound pressure levels below 85dB(A) have - statistically - only a low influence on hearing over the long-term. Exceptions include individuals with a vulnerable inner ear. However, this level is still capable of inducing measurable hearing-losses in high frequency ranges. Impaired hearing can start to appear in the range between 85 and 89 dB(A), but only after long exposure periods [20], or perhaps earlier with vulnerable inner ears. From 90 dB on a clear risk to hearing must be reckoned with.

A new risk assessment by the (US) National Institute for Occupational Safety and Health

Cincinnati, Ohio (NIOSH, revised criteria 1998) incorporating the 4 kHz audiometric frequency in the definition of hearing impairment reaffirms support for the 85-dBA recommended exposure limit for occupational noise exposure (85 decibels, A-weighted, as an 8-hour time-weighted average). With a 40-year lifetime exposure at 85-dBA, the excess risk of developing occupational NIHL is 8%-considerably lower than the 25% excess risk at the 90-dBA level.

If one is inclined to carry out calculations with noise levels, it should not be forgotten that the decibel (dB) is a logarithmic ratio unit. Thus an increase in rating level of 3 dB (e.g. from 90 to 93 dB (A)) represents a doubled risk to hearing. In other words, a two-hour stress with 93 dB(A) engenders the same risk as a four hour stress at 90 dB(A). An exposure at 105 dB(A), as is frequently encountered in discos, entails the same risk already after 4.8 min as an eight-hour noise exposure at 85 dB(A).

When evaluating risks from person-related noise levels it should not be forgotten that the a limit exposure level can only provide a protective effect if hearing can recover for a sufficiently long period after the sound event (for instance after an 8-hour work day). This means that a certain recovery time (at least e.g. 10 h) with a sound level lower than 70 dB(A) is adhered to after work [21], [22]. Protective measures are only then effective if the recovery times are adhered to and noisy recreational activities are not indulged in.

A relatively precise recording of sound energy acting on a human being can be achieved in a fixed occupational setting. Extensive protective measures like ear protectors and recommended exposure limits are also adhered to as a rule. Despite this, 6673 cases of occupational hearing loss as an occupational disorder were recognized by industrial trade cooperatives in 2002 in Germany. Noise-related hearing loss is therefore the most frequent occupational disorder in Germany, followed by asbestosis and skin diseases (source: Hauptverband der gewerblichen Berufsgenossenschaften (German Union of Industrial Trade Cooperatives), 2003).

Isolated occupational hearing loss shall not be gone into further here, since numerous publications and reviews, e.g. that of Dieroff (1994), have been written on this subject [10]. Important deficits in occupational safety arising from two new aspects shall be dealt with instead, i.e. the facts that:

(1) combined exposure of noise and specific chemicals can lead to hearing losses even though isolated exposure to one factor is classified as innocuous.

(2) safety measures are only effective if in addition to the regular use of individual ear protectors the required hearing recovery is not interrupted by noise during one's leisure time.

5.2.1 Problem

Studies on various noxins as causes for occupational health afflictions refer frequently to the investigation of the effects of one isolated, defined exposure on health. Maximal workplace concentrations (MAK) or, as in the case of noise, recommended exposure limits, are based on such a focused approach. Since in the real work environment a combination of different noxins is usually encountered, one can only forecast the effects of a combination of several pollutants by summating the individual harmful effects. This causes problems when two or several harmful effects can not just be added, but instead the agents causing them potentiate each other's actions. In such a case permanent damage to the organism can already arise with an exposure level not classified as dangerous after a single exposure.

This situation is especially problematic if a risk is not recognized with combined exposure of different noxins because the exposure level of the individual noxins (pollutant concentrations, sound pressure levels) lie below the normatively set limits. Thus people working in such an occupational environment are not classified as being endangered. Necessary measures, such as regular health controls or restrictions of exposure, can not therefore be put into effect.

Although case reports on the otoxicity of chemicals were already published in the sixties, studies on the combined effect of several chemicals or chemicals in combination with noise found little favor initially, possibly because of their complexity. Discussion about biological interactions between noise and chemicals was driven by observations of Barregard and Axelsson (1984) who found that the incidence of inner ear hearing loss was higher amongst workers who were exposed both to solvents and noise than it was in those exposed to noise alone [23]. Also, a longitudinal workplace-related study over twenty years revealed that the incidence of inner ear hearing loss amongst workers exposed to chemicals and a noise level of 80-90 dB(A) was three times that seen amongst workers not exposed to chemicals, but exposed to a noisier environment of 95-100 dB(A) [24]. Results of research from the last two decades have emphasized the importance of this occupational health problem.

5.2.2 Relevant chemicals at the workplace

Up to now specific-ototoxic effects have been studied only for a relatively small number of industrial chemicals. These relate to substances whose general toxic or specific neurotoxic or nephrotoxic effects are known. There are also substances that induce the production of reactive oxygen species. Data on the ability of a substance to produce reactive oxygen species (see below) might indicate a potential ototoxicity of a substance. On the one hand, generation of free oxygen radicals is a basic mechanism for toxicity and on the other it is considered as a component of mechanisms that lead to noise-induced hearing loss [25]. Based on such criteria the highest priority has been awarded to the study of the following substances: solvents, asphyxiants, metals and pesticides/herbicides [Tab. 2] [26].

Amongst the substances mentioned, solvents occur most often after asphyxiants. Solvents are used in detergents, adhesives, paints, varnishes and diluents in many industries as well as to a limited extent in households. They occur as mixed products in solutions within the chemical and petroleum/gas processing industries and are easily inhaled because of their volatility.

A substance that has been studied particularly intensively in recent years regarding interaction with noise is styrene, an agent that has found diverse uses amongst other places in the plastics industry. In such an environment, workers are frequently exposed simultaneously or consecutively to styrene vapors and loud machine noises [27]. Asphyxiants such as carbon monoxide and hydrogen cyanide are encountered in the work environment at least as frequently as solvents. Hydrogen cyanide is employed in the extraction of iron ores, in galvanization, and as a reactant in the chemicals industry. Carbon monoxide is always encountered naturally as a component of exhaust fumes where combustion engines are used and is therefore a pollutant occurring almost ubiquitously. It is also found in considerable quantities in association with the coal and steel industry, in cellulose and paper processing industries as well as at the sites of fires where firefighters carry out their work. Herbicides and pesticides (paraquat, organophosphate) are of course used in agriculture. Hydrogen cyanide is also used for pest control.

5.2.3 Studies on co-exposure

Several animal-experimental and occupational health studies have shown that permanent hearing loss can even arise when the maximum workplace concentrations (MAK values) and noise thresholds are adhered to [26]. Such work environments can not be formally recognized as dangerous in everyday life and employees can not to be provided appropriate preventive measures. Results from a range of animal-experimental and occupational exposure studies with noise combined with chemicals (co-exposure) are summarized in tables 3 [Tab. 3] and 4 [Tab. 4]. When interpreting these studies the following aspects must be considered in particular:

1) While noise is harmful primarily to the cochlea, industrial chemicals exert harmful effects both on the cochlea and central structures of the auditory system due to their general neurotoxicity [28], [29], [30], [31].

2) As already mentioned, chemicals can act ototoxically on their own (for a review see: Johnson and Nylen, 1995) and/or synergistically (additively or potentiating) with noise [26], [32], [33].

3) Different species respond differently to chemicals. The auditory system of the guinea pig for example seems to be less sensitive towards styrene exposure than that of the rat. Since the metabolism of the rat (e.g. with solvents) is similar to that of humans, this animal has found favor for studying the ototoxic effects of solvents [34], [35].

4) Damage to peripheral and central parts of the auditory system depends on the exposure parameters: dose (concentration, noise level), exposure duration, pauses (recovery times) between the exposures and the sequence of exposure (parallel, sequential).

5) Individual, endogenous (e.g. genetic) factors must be considered.

6) Frequently workers are exposed not just to one specific chemical, but to a mixture of different solvents.

7) The use of large audiological test batteries for the detailed characterization of hearing loss is restricted in most epidemiological studies because of a lack of time and money. Because of the general neurotoxicity of many chemicals, pure-tone and speech audiograms are not sufficient for recording damage to the central auditory system. The inclusion of tests for the central auditory system would be desirable [31].

8) Evidence that ototoxic effects can also continue after chemical exposure, and that alcohol can potentiate these effects requires special consideration [35], [36]. In this way interactions between different noxins can also occur outside of the workplace and thereby escape any form of preventive control.

5.2.4 Conclusions

The study of the effects of combined exposure with several pollutants such as chemicals and noise represents an important challenge for future occupational safety research. Until recently this was given only relatively scant attention, partly because of the complexity of the problem. Future studies on biological associations and epidemiological studies must lay the foundation for rational monitoring and intervention in the working environment.

A current program of the European Union "NoiseChem" shall provide better consideration of mixed exposure to noise and chemicals [37]. This shall occur by:

1) developing tests for evaluating noise- and solvent-caused damage to the hearing and balance systems;

2) determining dose/effect relationships among 2,000 workers exposed to different solvent-noise combinations;

3) using tests on humans and animal models to see where and how solvents and noise exert their effects; and

4) developing hearing conservation schemes taking both factors into account.

5.3.1 Lacking hearing recovery phase

Many adolescents and young adults work in a noisy environment and expose themselves during the period defined as the hearing recovery phase to noises described under leisure noise (see below). In addition to risk from occupational and leisure noise, the shortening of the hearing recovery phase is another risk factor. As already stated above, adequate occupational safety at a noise-intensive workplace requires a 10 hour hearing recovery period per working day during which exposure lies below 70 dB(A). However, tables 5 [Tab. 5] and 6 [Tab. 6] illustrate that moderate stress with leisure noise already exceeds the recovery level so that even when hearing moderate music levels the recovery period is shortened. Even without an increase in mean sound level, an increased risk to the inner ear arises from the combination of the recreational and workplace noise sources.

5.3.2 10% subpopulation

This affects particularly the 10% subpopulation of an age-group mentioned below (section "leisure noise"), since there is a higher incidence of poor schooling and low social status, unskilled occupations combined with at noisy workplaces and excessive exposure to leisure noise [38].

Regarding the subjective stressing of the population and the individual perception of a nuisance, environmental noise plays a more important role than occupational noise. The most significant sources of noise in our environment are:

• Traffic noises of all kinds (e.g. from automobiles, aircraft, railways and ships),

• Industrial and construction work noise, e.g. noise radiating from industrial plants and building sites in the neighborhood,

• Noise from the residential and leisure activities, i.e. from neighbors (e.g. noise from radios and TVs, lawnmowers, do-it-yourself work, steps) or sound emanating from leisure activities of any type (e.g. from sports, music and dancing events, shooting practice, toy pistols, fireworks, and model airplanes).

With the exception of leisure noise, sound intensity in the environment does not normally reach damaging levels. However, cardiovascular (cardiac ischemia) and hormonal effects (cortisone, adrenaline) have been attributed to it, although this remains to be further investigated [39].

An important exception is recreational or leisure noise. In children, adolescents and young adults in particular, activities are indulged in more often that are associated with high and even exceptionally high sound pressure levels that may be even more dangerous than those encountered in the workplace. Even babies and small children are affected. Loud toys, fireworks and electrically amplified loud music of all genres are just as damaging to the inner ear as workplace noise from loud machines. A techno fan who exposes himself to loud music from a portable music player is just as exposed as a worker in a steel mill without ear protection [38].

5.5.1 Hearing loss from leisure noise

In recent years, a rise in the number of individuals with impaired hearing has been observed amongst adolescents who have never worked in a noisy workplace [40], [41], [42], [43]. Leisure noise is presumed to underlie such findings that normally only come to light in occupational medical check-ups. Sound pressure level measurements in discos average between 92 and 111 dB(A), and with headphones or earphones inserted inside the external auditory canal (e.g. with portable music players) the maximum level can reach 120 dB(A) with an average of 100 dB(A) [44], [45], [46], [47], [48]. Loudspeakers are another health-endangering electroacoustic noise source at large music events (e.g. open air concerts). In a survey in Germany, two thirds of 1814 young men had already experienced "buzzing in the ears", "ear whistling" or "deaf ears" (as an indication of a temporary shift in hearing thresholds) after loud music events [49]. Symptoms had already been encountered repeatedly in most cases. In 24% of the individuals examined, irreversible hearing-loss typical for noise-induced hearing impairment was found. Risk analysis shows a clearly increased risk of hearing impairment for firework use, frequent clubbing and the use of loud walkman devices [49], [50], [51].

5.5.2 Children's toys

With toy guns, a peak level of more than 135 dB(A) at 1 m from the exposed ear can be achieved, while at 2.5 cm it can even reach 163-173 dB(A) [Tab. 7]. Like pistols, jumping crackers ("Knackfrösche") are also held to the ear for fun, and reach levels of 135 dB(A). Small toy trumpets can reach an intensity of 125 dB(A) [7]. With the equally loud toy mobile phones (measurements: A. Limberger, personal communication) the devices are actually intended for use next to the ear.

5.5.3 Electroacoustically amplified music

In discos, continuous sound levels of more than 100 dB are frequently measured. According to measurements of the "Physikalisch-Technische Bundesanstalt, Braunschweig, Germany" (German Federal Institute for Physical and Technical Research) continuous sound pressure levels of up to 110 dB (corrected for free field), and maximum levels up to 120 dB(A) (external auditory canal measurements) are reached for portable mini cassette players [42], [44], [47], [48], [52].

Headphones: The average levels determined in tests are not harmful to hearing [38]. However, these average levels do not say anything about the extreme listening habits of some adolescents. From figure 1 A [Fig. 1] it can be seen that 10% of 11- 17 year olds examined (the so-called 10% subpopulation) set their levels to 100 dB(A) and more. With dynamic averaging one comes to a mean listening level of 98 dB(A) in the 10% group [53]. Figure 1B [Fig. 1] lists the figures regarding the duration of music listening [46]: 10 to 18 year old adolescents listen on average for 1 h at most per day, but 10% of 13-19 year olds listen to at least 3 h of music per day [46], [49], [54].

With lower school status the proportion of pupils exposed to an equivalent continuous sound load (related to 40 week hours) with music levels of 90 dB(A) and above increases markedly. There was no age effect here, since the population sample for these assessments was restricted to a group of 17 year olds.

In conclusion the following noise values were found for adolescents and teenagers in Germany:

• An initially unproblematic looking median of Lm=78 dB(A) for portable music devices, but an alarming 98 dB(A) for the 10% subpopulation [Fig. 1].

As such the extreme noise heard by 10% of an age-group of children, adolescents and young adults is alarmingly high in Germany.

Discos: The average noise level over time [Fig. 2] in the discos surveyed lay between 89 dB(A) and 110 dB(A), where the distribution maximum was 100 dB(A) [42], [53]. The dynamically averaged Lm for the discos surveyed amounted to:

• 1988: Lm=102.3 dB(A) (29 discos),

• 1994/1997: Lm=102.1 dB(A) (14 discos),

and as such remained unchanged over these years. It was also found that the levels increase during the course of the night by about 2 dB(A)/hour [55].

The exposure times increase with increasing number of disco visits. In 1980 Bickerdike and Gregory determined an average individual stay of 2.5 h [56]. On the basis of approx. 10,000 surveyed adolescents the picture provided in Figure 3 [Fig. 3] appeared up until 1993. Individuals between 12 and 22 years of age attended a disco on average 1 to 2 times a month, while older people attended a little more often. For public health reasons an important fact is that until 1993 10% of the surveyed individuals attended discos regularly 1-2 times per week. In more recent investigations higher figures were reported, showing that young men aged 16-24 years now attend discos on average 5 h weekly while 80% of 18-19 year olds attend for 6 h per week [49].

In conclusion, the following noise values can be derived for adolescents in Germany:

• In discos an averaged estimate of Lm=90 dB(A, harmful to the ears) and a much higher value of 95 dB(A) for a 10% subpopulation of an age-group („10% value, [Fig. 2], [Fig. 3]).

Also here, the extreme noise exposure for 10% of an age-group of children, adolescents and young adults is alarmingly high in Germany.

Large music events: Sound pressure level information relating to open-air concerts and large events in closed arenas is listed in table 8 [Tab. 8]. In the direct vicinity of the loudspeakers the danger of suffering ear damage is at its greatest. If the distance from the loudspeaker is doubled, the sound level falls by 3-6 dB depending on the size of the loudspeakers.

5.5.4 Firearms sports

Representative studies have revealed that firearms sport plays a role amongst 2-12% of the surveyed individuals. Shot noise can reach maximum sound pressure levels of 173 dB. Hunters can not wear any ear protection because hearing is an indispensable tool for detecting game while hunting.

5.5.5 Fireworks

Annually in Germany about 8000 patients suffer acute acoustic trauma as a result of New Year's Eve celebrations [2]. There is hardly an ENT practice or clinic that does not treat patients with such inner ear injuries regularly at the start of the New Year.

Acute acoustic trauma by fireworks or flare pistols is caused by impulse noise. Since the subjective sensation with impulse noise is markedly reduced due to the short duration of the sound, high sound pressure levels that damage the ear can be experienced without them being perceived as loud. Sound impulses from fireworks can reach maximum levels of 145 to more than 160 dB SPL at the ear if they explode at a distance of approximately 2 m [57]. With public fireworks displays, peak levels of up to 190 dB can be measured in their immediate vicinity that fall to approx. 150 dB SPL [58] at normal observation distances. Flare and scare pistols are also frequently used around New Year's Eve. Depending on the distance and the angle to the ear, peak levels of around 160 dB to a maximum of 181 dB SPL can be measured [59], [60]. Sound reflections by nearby standing buildings or improper use in closed rooms can also lead to an increase in sound energy reaching the ear.

In an own prospective epidemiological study we stressed this exceptional risk amongst male children, adolescents and young adults in particular regarding this kind of leisure noise [2]. Within the scope of this prospective study epidemiological and audiological data from patients with hearing loss due to fireworks or flare pistols during the changeover from 1999 to 2000 was obtained from 562 representative study centers in Germany. After extrapolation of the data from the patients studied, a total number of 8160 patients for the whole of Germany was projected (95% confidence range: 7580 - 8740). The absolute incidence was 9.9 per 10⁵ inhabitants and is as such comparable to the incidence of idiopathic sudden sensorineural hearing loss occurring over an entire year in countries of the western world [61]. Male patients were affected three times as frequently as female patients and were on average younger (medians; m: 22 years, f: 25 years). 59% of the patients were less than 25 years old. For the age-group from 6 to 25 there was a much greater incidence (28 / 10⁵) with a maximum of 107 per 10⁵ inhabitants amongst 19 year old men ([Fig. 4] A). About 69% of all patients complained of at least one-sided subjective hearing loss. Tinnitus was either an accompanying or the main symptom in 84% of all cases. 5% of all cases were associated with a unilateral and 1% with a bilateral tympanic membrane perforation. Hearing losses were also evaluated using audiograms. In 79% of all patients for whom an audiogram was available, and who showed no preexisting hearing loss in the medical history, a unilateral hearing loss at least could be confirmed (median: 30 dB at 4 and 6 kHz). In 52% of the patients this hearing-loss was even more marked (median: 40 dB). Interestingly, the absolute number of traumas was not unduly raised for the turn of the millennium, as a comparison with results of a nationwide pilot study for 1998/1999 showed [62]. Also, assessment of patients with hearing loss due to New Year's Eve fireworks in Giessen from the years 1998 to 2002 showed a similar number of cases for each turn of year [63].

Approximately four of five (81%) of the traumas recorded in the nationwide epidemiological study were not caused by the patient him/herself, as opposed to the 19% who admitted to have been at least partially responsible. Amongst the first mentioned group, 15% were designated as being deliberately construed while the rest were accidental [2]. It is therefore clear that hearing damage caused by fireworks can justify considerable claims for compensation from those agents responsible for causing it. A recent review on the civil legal and criminal aspects of hearing loss due to New Year's Eve fireworks was published by Beyer et al. 2003 [64].

All 458 study centers participating in the prospective epidemiological study [2] on patients with acute hearing loss from fireworks or flare pistols were asked to carry out follow-ups (up to 12 months after the event, 2000). From the 455 study centers, follow-up data was provided by 265 (58%) of the centers from 401 (20%) of the initial 1,999 participating patients. No follow-up was carried out in 720 patients according to information provided by the study centers [65].

More than half (54%) of the patients undertaking a follow-up revealed ongoing symptoms such as tinnitus (44%), subjective hearing-loss (34%) or hyperacusis, or the investigator diagnosed them as having permanent sensorineural hearing loss attributable to the bang or blast trauma. Assuming that these distributions of symptoms and findings also apply amongst patients for whom no data from long-term follow-ups was available, the absolute extrapolated number of patients with permanent hearing loss would be 4436 for the year of the study, corresponding to an incidence of 5.4 per 100,000 inhabitants. As with acute hearing loss, ongoing hearing loss affected male patients much more frequently than it did female patients. More than half of the patients were younger than 25 years, although only 27% of the population belong to this age-group ([Fig. 4] B). However, such results have to be weighed against the validity of the small sample of patients who underwent follow-ups (401 of 1,999 patients) and the different time-points at which they were undertaken. It is safe to assume for example that patients with ongoing complaints are more likely to return to ENT care, while recovery processes can be observed in some cases up to nine months after the initial acute event [66].

Other longitudinal studies produced similar, albeit somewhat smaller figures for permanent hearing loss from acute acoustic trauma:

(1) in one long-term surveillance study the recovery of hearing function was registered in patients of the University ENT clinic Tübingen over a period of up to two years after a acute acoustic trauma arising from New Year's Eve fireworks or flare pistols [66]. Six months to two years after the trauma, and despite early therapy, at least a third of all patients in the study suffered hearing loss of more than 20 dB HL in the frequency range typical for acute or chronic acoustic trauma-induced tinnitus.

(2) In a longitudinal study, Mrena et al. examined troops and veterans in Finland who had received treatment for an acute acoustic trauma during military service [67]. At the examination time-points some weeks and ten to fifteen years after the bang trauma, a chronic tinnitus was found by the investigators in 29% and 19% of the patients, respectively. The recovery of the hearing threshold was not examined in this Finnish study.

(3) Amongst participants of the Indian "Deepawali" festival where fireworks are used extensively, every fourth patient with firework-induced hearing-loss revealed a permanent shift in hearing threshold of ≥ 30 dB at 4 and 6kHz [68].

The epidemiological data currently available stresses the not-to-be underestimated sociomedical importance of hearing loss due to New Year's Eve fireworks and assist in the identification of target groups for directed preventive measures.

Hearing damage can in principle arise from acoustic overstimulation, an increased vulnerability of the ear, or a combination of both factors.

A hypothesis to explain the consequences of sound overexposure assumes two general and fundamental mechanisms of hearing damage which was already put forward as long ago as 1969 [69]: firstly, mechanical trauma may lead directly to damage of the architecture of the inner ear structures, and secondly the increase in pathological activity caused by excessive sound stimulation may result in a metabolic stress response. Two important general starting points for therapeutic intervention, i.e. physical reduction of sound energy reaching the ear and the pharmacological alteration of cellular response to stress stimuli, are based on this hypothesis.

Hearing damage may also result from statistically innocuous sound exposure. This situation, as well as the fact that excessive stimulation does not always induce ear damage, is discussed in the section "vulnerability".

In order to be able to fully explain noise-related injury to the inner ear at a systemic, cellular and molecular level, there remains an abundance of unresolved problems. This is because a broad range of different conditions have been tested under which hearing damage has been studied until now. Under controlled experimental conditions, however, a few alterations have been described until now that can be reliably reproduced [53], [70], [71], [72], [73].

The pathological mechanisms after an acute acoustic trauma can be described according to physical parameters such as the intensity of sound and the length of exposure. In addition, various pathological states can be observed depending on the time elapsed after the traumatic event. A progression from reversible to irreversible alterations can also be observed. Such qualitative and quantitative changes concern different tissue structures and cellular elements in the cochlea. The affected tissue structures in the cochlea are

• the organ of Corti with

• Hair cells and

• Supporting cells

as well as the tissues located medially to the organ of Corti

• Limbus

and the ones lateral

• Spiral ligament and the

• Vascular stria of the cochlear duct.

In addition

• Afferent neurons of the auditory nerve and spiral ganglion cells can be damaged.

Usually, not all the abovementioned cellular elements are affected in a single individual. However, the outer hair cells are almost always involved. For this reason the audiological consequences of their functional loss shall be gone into in particular:

6.1.1 Damage to the outer hair cells

A specific property of the outer hair cells - motility [74], [75], [76], [77] - forms the basis for the local, active amplification mechanism in the cochlea. The motor protein prestin has now been identified as the molecular basis for this motility [78]. A disease-induced, isolated damage or loss of outer hair cells with amplification of the inner hair cells results in a considerable degradation in the propagation of the traveling wave with shifts in hearing thresholds and recruitment (review in Preyer and Gummer, 1996 [79]). At low sound levels the traveling wave is no longer actively amplified, so that the sound is only heard above the comparatively high physiological threshold of the inner hair cells, i.e. at approximately 50-70 dB HL [80]. Clinically a threshold loss up to 50-70 dB HL may be seen in the pure-tone audiogram. In addition the sharp peak of the traveling wave which is important for frequency selectivity is [81]. This can explain the loss in speech discrimination in the speech audiogram. Furthermore, recruitment occurs and the amplitude of acoustically evoked otoacoustic emissions decreases or they disappear. Binaural hearing properties such as lateral spatial hearing and recognition of signals amongst background noise can also be expected to be limited quite markedly [82].

6.1.2 Recruitment

Upon damage to the outer hair cells the decreasing (i.e. non-linear) amplification of the traveling wave with increasing sound pressure is altered (see Preyer and Gummer for a review 1996 [79]). For this reason in a diseased individual a loud tone that follows a softer tone (e.g. in the Békésy audiogram) leads to a relatively stronger ("linear") increase in the traveling wave than occurs under physiological conditions. In figure 5 [Fig. 5] the increase in amplitude of the traveling wave upon increase in sound pressure in the damaged ear follows a linear relationship while in the healthy ear it follows a non-linear, sigmoidal relationship (the process in the healthy ear results from the amplification by the outer hair cells). For this reason, wherever the pathologically shifted hearing threshold is exceeded a sound pressure level difference is perceived earlier in patients than it is in healthy individuals. With high sound pressure levels individuals with impaired and normal hearing behave similarly again, and this is perceived as a volume equalization. Recruitment can therefore be understood as a sign of damage to the cochlea, and particularly the outer hair cells.

Noise that endangers hearing seldom leads to displacements in the inner ear exceeding 800 nm (at 150 dB SPL) [83]. For comparison, outer hair cells reach a length of up to 80,000 nm (80 µm) and a diameter of 5,000 nm (5 µm).

6.2.1 Rough mechanical injury

Because of the abovementioned disparities in dimensions, rough mechanical injuries are in fact rare. Only exorbitant effects, as occur with some blast traumas or in some animal experiments (210 dB), lead to histologically visible, gross mechanical disruptions in the inner ear [84]. Destructive mechanical damage in finer detail takes the form of structural alterations of the organ of Corti organ in particular. Rupture of cell associations at the apical surface, the reticular layer, or damage of supporting cells (especially the pillar cells) number amongst these [12], [16]. The fact that extremely high potassium concentrations can contribute to cell death following injury-induced release from the endolymph space should not be forgotten. The destructive tissue alterations result above all in irreversible, necrotic loss of hair cells [85]. Cell death through necrosis is explained below.

6.2.2 Micromechanical injuries

As a rule, because of the dimensional aspects mentioned above, usually micromechanical or even no mechanical damage to the architecture of the inner ear is expected with most acoustic traumas. As explained further below, metabolic stress as a consequence of functional overstimulation may play the most significant part in noise-induced damage.

Micromechanically one observes a decoupling of the stereocilia of the hair cells in the acoustic organ from the tectorial membrane [86]. As a result of decoupling, hearing loss occurs that then may lead to a recovery upon recoupling, a phenomenon that may contribute to a temporary threshold shift (TTS).

An early micromechanical reaction is floppiness of the stereocilia due to a loss in rigidity of the stereocilia. The normally stiff and upright-standing stereocilia of the damaged hair cells appear as floppy and bent, but not broken. As shown below, this is actually a response to metabolic injury, and is due to a degradation of cytoskeletal proteins with micromechanical consequences [87]. Mechano-electrical transduction is then no longer possible.

With a longer duration and higher intensity of the harmful stimulus, mechanical damage can lead to ultimate loss of the stereocilia. Here as well, mechano-electrical transduction is no longer possible. At the inner hair cells, the loss of the longest stereocilia already leads to a reduction of the resting current through the stereocilia membrane [88], and a disruption of the transduction process is expected.

For cells in general, and with that also for the main functional cells of the cochlea, even very diverse agents (mechanical stress, hypoxia, poisons, chemical substances, pathogens etc.) can ultimately lead to similar cellular metabolic response, and therefore to the activation of the same or at least similar cellular cascades (common final pathway). According to our current state of knowledge it is the massive increase in extracellular glutamate concentration, the influx of calcium into the cell and the formation of oxygen free radicals that play major roles in the central nervous system and the auditory periphery [89], [90], [91], [92]. However, it is not so much an activation of a specific or parallel running cascade, but more a mesh-like networking of biochemical processes that occurs.

In response to acoustical overexposure, cells subject themselves to a so-called metabolic stress. As a result of this metabolic stress, the affected cell might degrade parts of its cytoskeleton or even die (decompensated metabolic stress), or counter-regulatory metabolic protective and/or repair mechanisms might manage to prevent this (compensated metabolic stress). An important form of metabolic stress is oxidative stress (see below). This is related to the production of oxygen free radicals that occurs especially within the hair cells. Excitoxicity is related to an increased extracellular glutamate concentration and increased calcium influx, which in the cochlea involves afferent neurons of the auditory nerve in particular. A molecular cytoskeletal degradation is typical for stereocilia.

6.3.1 Degradation of cytoskeletal proteins

The phenomenon of floppy stereocilia with its micromechanically effective weakening of the stereocilia is a very specific and well documented early metabolic reaction of hair cells [93]. The physiologically stiff and upright standing stereocilia appear bent or even fused with neighboring stereocilia under the scanning electron microscope when they are damaged [87]. Biochemically and biophysically, a molecular change in the stereocilia actin filaments underlies this effect. They form the hair cell skeleton responsible for stereocilia rigidity [94]. In the pathological state a degradation of part of the molecules that cross-links neighboring actin filaments occurs in the stereocilia [87], [95]. Acoustic trauma may also lead to F-actin degradation. This disassembly can be prevented by a caspase-3 inhibitor [96]. A direct relationship therefore exists with the below-mentioned metabolic apoptosis pathway. The extent to which continuous re-modelling of actin filament structures in stereocilia shown in recent in vitro experiments plays a role in the repair of floppy stereocilia shall be the subject of future investigations [97].

6.3.2 Oxidative stress

Metabolic activity is typically coupled to the formation of reactive oxygen species. Important reactive oxygen species include the superoxide anion (O₂ ^-), the hydroxyl radical (OH•) and hydrogen peroxide (H₂O₂). The superoxide-anion can react with nitric oxide (NO) to form peroxinitrite (ONOO^-). Under normal conditions the (potentially toxic) oxygen radicals formed in the mitochondria are rendered innocuous by intracellular endogenous anti-oxidative systems [90]. If this does not succeed (see below decompensating oxidative stress), the cell progresses to an apoptotic death.

Oxidative stress cascades in the cochlea: Animal-experimental studies have shown that formation of reactive oxygen and nitrogen compounds in the cochlea represents a primary mechanism of damage after trauma. Here it is also important that a range of differing conditions, e.g. acoustic trauma, hypoxia/ischemia, addition of ototoxic aminoglycosides, chemotherapeutics or chemicals, and withdrawal of growth factors, are able to lead to oxidative stress [25], [92], [98], [99].

Induction of oxidative stress reactions by noise: The role of reactive oxygen compounds in noise-related injury to the ear is emphasized by the following results of animal-experimental studies: (1) Excessive acoustic stimulation leads to an increase in harmful, reactive oxygen species (ROS) in the cochlea, as has been shown from measurements of hydroxyl radicals [100] and superoxide anions [101]. (2) After excessive acoustic exposure the activity of endogenous anti-oxidative mechanisms increases in the inner ear, e.g. glutathione [102]. (3) The activation of anti-oxidative systems attenuates acoustic trauma-induced hair cell loss and threshold shifts [Tab. 9], [103], [104], [105], [106], [107], [108], [109]. (4) The weakening of endogenous anti-oxidative systems by pharmacological inhibition of the glutathione system (e.g, with buthionine sulfoximine, BSO) or genetic manipulation (deletion of the superoxide dismutase gene or mutation of the glutathione peroxidase gene) promotes morphological and functional damage from noise exposure [110], [111].

Decompensated oxidative stress: If protection can not sufficiently be provided by endogenous systems due to a disorder in the balance between reactive oxygen/nitrogen species and endogenous anti-oxidative systems (decompensated oxidative stress), this leads to the destruction of cell and nuclear membranes by lipid peroxidation, DNA fragmentation, the activation of proteases, to an increase in intracellular calcium and ultimately to apoptotic cell death (see below, for a review see: [90], [112]).

This transition between metabolic damage and cell death can progress slowly in the post-traumatic phase. It lasts up to two weeks and with typical acoustic trauma spreads from the basal half of the cochlea. In functional terms the death of cells, especially the sensory cells, entails a permanent sensorineural hearing loss (permanent threshold shift, PTS).

6.3.3 Nitric oxide with noise

Apart from the formation of reactive oxygen compounds, tissue damage induced also by nitric oxide (NO) is thought of as a cause for hearing loss after acoustic trauma. Enzymes synthesizing nitric oxide, NO synthetases, have been demonstrated in the cochlea [113], [114]. A rise in NO in the perilymph of the guinea pig cochlea as well as in the inner and outer hair cells after acoustic trauma was shown in an in vivo study by Shi et al. (2002) [115]. It is presumed that the cytotoxic effect develops as a result of nitric oxide reacting with the superoxide anion to produce peroxynitrite (ONOO^-). Peroxynitrite possesses strong oxidizing properties through which amino acids and nucleotides can be modified and intracellular signal cascades can be influenced [116]. The application of peroxynitrite scavenger molecules, such as the seleno-organic compound ebselen, can reduce cochlear damage and permanent hearing-loss in animal experiments [103].

6.3.4 Excitotoxicity

Excitotoxicity is a mechanism harmful to the afferent auditory nerve fibers that can be induced by an acoustic trauma. The fundamentals of excitotoxicity and neural transmission at the afferent-efferent synaptic complex of the IHC were described in full by Pujol and Puel et al. and by Ehrenberger and Felix et al. [117], [118].

The primary neurotransmitter between the inner hair cells and the afferent auditory nerve fibers is glutamate [119], [120]. During the normal hearing process glutamate binds to the fast AMPA receptors of the auditory nerve fibers, while the slow NMDA receptors that are also present apparently play no physiological role. More recent physiological-pharmacological studies at the synapses of the IHC with the afferent nerve fibers, comparing the effects of an NMDA receptor antagonist (D-2-amino-5-phosphonopentanoate) with that of a selective AMPA receptor antagonist (GYKI53784), did indeed show that the rapid, excitatory transmission between the IHC and the primary auditory nerve fiber is mediated preferably via AMPA receptors, while NMDA receptors appear to play a lesser role [121], [122].

With excessive stimulation, however, the NMDA receptors are activated greatly because of the large glutamate excess, a fact that has fatal implications for the nerve cell [118].

The synapses between the inner hair cells and the afferent nerve fibers: The vast majority (90-95%) of cochlear afferent nerve fibers innervate the inner hair cells as myelinated type-I afferents directly without branching out [123], [124]. The main transmitter at the synapse of the IHC with the afferent nerve fibers is glutamate [120], [125], [126]. Glutamate receptors are subdivided according to their agonists into N-methyl-D-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolproprionate (AMPA) and kainate receptors. Indications that all three receptor types probably exist at the synapse to the inner hair cell were obtained by immunohistochemistry, in situ hybridization and gene expression studies which all showed that primarily auditory neurons express subunits of NMDA (NR1 and NR2A-D), AMPA (GluR2-4), kainate (GluR5-7) receptors and the high affinity kainate binding protein (KA1, KA2) [127], [128], [129], [130], [131].

Mechanisms of cochlear excitotoxicity: If the neurotransmitter glutamate is released excessively into the synaptic cleft and/or it is not removed, neurotoxic effects ensue at neural synapses, a phenomenon termed excitotoxicity [132]. The fact that the application of kynurenate, a glutamate antagonist with a broad spectrum of activity, clearly acts protectively regarding morphological damage and hearing-function, indicates the special role of synaptic glutamate release from IHC with excessive noise exposure [133].

The NMDA receptor, which contains ion channels (i.e. an ionotrophic receptor), is of particular importance for understanding excitotoxicity. The channels allow not only monovalent cations, but also calcium ions into the cell interior. The channels of the NMDA receptors are blocked by extracellular Magnesium (Mg²⁺) ions (an interesting, therapeutically usable functional property). This blockade is voltage dependent, so that calcium permeable NMDA receptor-associated channels only open if two conditions are fulfilled: 1) glutamate must be available at a high concentration and, 2) the nerve cell must be depolarized [134]. Both apply if excessive noise stimulation is present. With a noise-induced glutamate surplus (the inner hair cells release a disproportionate amount of glutamate in the presence of noise) it is assumed that the Mg²⁺ block of the NMDA receptors is lifted and in this way the normally inactive NMDA receptor is activated. The increased calcium influx that is then induced causes a long-term potentiation of synaptic transmission which can lead to calcium intoxication of the afferent auditory nerve fiber.

The resulting excitotoxicity is characterized by a two-step process. Firstly, swelling of the postsynaptic structures occurs due to ion (particularly calcium) and the associated water influx until there is a functional loss of the afferent dendrites contacting the nerve cells. Secondly, a calcium influx-triggered cascade is set in motion with release of reactive oxygen radicals and activation of proteases and endonucleases, a process that ultimately leads to a metabolic decompensation with apoptotic death of the primary auditory neuron [135].

6.3.5 Cell death

In many cases cell death results either from mechanical trauma or, more frequently, metabolic decompensation as a stress response particularly in the auditory organ and in the afferents as well. The outer hair cells are affected particularly frequently and seem to be especially vulnerable. As a result of cell death a histological scar may be seen in place of functional inner ear cells. Two general forms of cell death should be distinguished: necrosis and apoptosis.

Necrosis: Animal-experimental studies have shown that necrosis plays an important role particularly where the cochlea is exposed to extreme noise [136], [137]. The first cells generally perish directly from the mechanical action of the sound. There can either be a direct mechanical tearing of the cell, or a potassium intoxication after injury to the endolymph space. The resulting necrosis is a passive and disordered form of cell death where the cells release their contents in toxic form, and other areas of the cochlea are put at risk. In the acoustically traumatized cochlea, necrotic cells are therefore found over wide areas of the acoustic organ [136], while apoptotic hair cells (see below) are mainly restricted to the tonotopic area of the harmful sound frequency, i.e. usually in the basocochlear region.

In general, necrosis is characterized by a classical morphological appearance. One observes a disintegration of the cell membrane, a condensation of nuclear material and a swelling of the cellular organelles. Upon destruction of the cell membrane, proteolytic enzymes are released from the cytoplasm that can destroy neighboring cells. Ultimately, an inflammatory reaction is caused by the attraction of immune cells. A cellular necrosis therefore tends to spread out over time (for a review see Majno and Joris, 1995 [138]).

Apoptosis: This form of cell death is distinguished from the above-mentioned necrosis. Animal-experimental studies revealed that in the cochlea it is mainly cell death through apoptosis that plays an important role for the occurrence of permanent threshold shift [136], [137]. Of course cell death from apoptosis, particularly that of the sensory cells, entails a permanent and not a reversible loss of hearing.

Unlike the passive, disordered necrosis, a highly ordered, genetically controlled program is activated in the cell upon apoptosis. Apoptosis occurs in regenerative cells, i.e. not in hair cells, even as a normal physiological process and serves to regulate cell number by removing superfluous cells. However, if apoptosis occurs in the adult cells of the inner ear, this represents an undesirable pathological process that is a consequence of noise-induced metabolic stress.

With apoptosis observed upon acoustic trauma, cells react to the intracellular signals of the acoustically-induced oxidative stress. In the initial phase the cell nucleus, cytoplasm and mitochondria shrink without there being a loss of cellular integrity. As a result, no proteolytic enzymes are released, and an inflammatory reaction does not usually occur. At this stage chromatin (DNA and associated structure proteins) compacts and becomes destroyed. The cell volume decreases, intense membrane blebbing appears at the cell membrane and the cell loses its contact with neighboring cells. Finally the cell breaks down into membrane enclosed apoptotic bodies that are eliminated by phagocytosis without any local inflammatory reaction (for reviews see Majno and Joris, 1995 [138] and Reed, 2000 [139]).

Very generally, apoptosis, depending on cell type and trigger, can be initiated by external signals (extrinsic pathway) or internal signals (intrinsic pathway) within the cells. The common final pathway for these different apoptosis-inducing routes is the initiation of a signal-cascade at whose center proteases stand (particularly caspases). These enzymes irreversibly initiate programmed cell death [140]. Caspases form a family of aspartate-specific cysteine proteases. Because of their sequence and presumed functions they are classified either as initiators (e.g. caspase-1,-8,-9 and -10) or effectors (e.g. caspase-3,-6 and -7) of apoptosis. Caspase-3 is regarded as an important effector protease [140], [141]).

With an intrinsically activated apoptosis pathway (type II), caused e.g. by noise-induced formation of reactive oxygen radicals, the protein Apaf-1 (apoptotic protease activating factor) is released from the mitochondrial membrane protein Bcl-2. As a result of this release membrane pores open in the external mitochondrial membrane and cytochrome C amongst other constituents is released into the cytoplasm. Within the cytoplasm, cytochrome C, Apaf-1 and pro-caspase-9 form the so-called apoptosome with which caspase-9 is activated in combination with dATP. Both Caspase-8 and -9 belong to the regulators of the activation of the effector caspase-3 (see for a review Ashe and Berry 2003 [140], [141]). Recently more and more publications have also stressed the role of caspase-independent signal cascades that also lead to cell death alongside the caspase dependent apoptosis pathways [142].

With extrinsically initiated apoptosis pathways (type I), intracellular caspase-8 und-10 and through that caspase 3 are activated by binding of external ligands to cell membrane receptors ("death receptors").

In a recent publication Nicotera et al. examined the time-dependent activity of caspase-3,-8 and -9, cytochrome C and morphological markers of apoptosis and necrosis after an acoustic trauma to the chinchilla. The authors found an increase in caspases of the intrinsic (caspase-9) and extrinsic (caspase-8) activation pathways and their common final pathway (caspase-3) in apoptotic cells. It is presumed that apoptotic mechanisms play an important role in the progression of hair cell damage over time after the acoustic trauma is discontinued [143]. As mentioned above, the transition from metabolic damage and cell death into the post-traumatic phase can occur very slowly. It lasts for up to two weeks and with a typical noise trauma spreads out over the basal half of the cochlea.

Local progression on the other hand is restricted. Apoptosis does not typically spread over the entire cochlea. Instead one finds apoptotic hair cells (characteristic for a typical acoustic trauma) mainly in the tonotopic zone responsible for the appropriate frequency of the acoustic trauma, while necrotic cells (characteristic for extreme acoustic trauma) can spread out over much wider areas of the cochlea [136].

A further signaling cascade of apoptosis, activated in a caspase-dependent or -independent manner, is the JNK signal pathway [144]. Pirvola et al. showed that JNK, also known as stress activated protein kinase, is associated with apoptosis induced by acoustic trauma [145].

6.3.6 Increased vulnerability towards noise

Different sensitivity to noise: Not all people exposed to noise will suffer a clinically relevant acoustic trauma. A phenomenon observed both clinically and in animal experiments on noise induced hearing loss is highly variable sensitivity of the auditory organ to noise exposure within a population [146], [147], [148]. Apart from excessive stimulation, an individually increased sensitivity of the inner ear plays a major role in noise-induced hearing loss. There are also individuals who experience acoustic traumas below noise exposures classified as being dangerous to hearing (the so-called vulnerable inner ear).

Despite numerous advances in the study of the pathophysiology of noise-related injury, the biological mechanisms underlying the extremely varied sensitivity towards noise are for the most part unclear. Of special interest are studies involving audiological, electrophysiological and genetic procedures to evaluate differences in inner ear vulnerability. However, a reliable clinical prognostic indicator for individual acoustic susceptibility has not yet been identified.

This situation complicates the regulatory setting of sound pressure level limits and exposure times to prevent noise-induced hearing loss. For a prevention of hearing loss from noise it would be desirable to identify those individuals who have an enhanced inner ear vulnerability. Also, in order to understand the pathophysiology of noise-related injury and the therapeutic interventional strategies to be derived from it, more needs to be known about the causes of variable acoustic vulnerability. This broad topic can not be given an exhaustive discussion here. However, because of its importance for preventive medicine and the therapy of noise-conditioned inner ear damage it shall be mentioned and some aspects shall be gone into.

Audiological aspects: Controlled studies on human individuals involving generation of a PTS are unethical and longitudinal studies on larger groups of individuals are very arduous and technically barely feasible. Therefore psychoacoustic procedures have gained favor for seeking ways to predict individual acoustic vulnerability; these have measured the extent of a temporary threshold shift (TTS) as a predictor to assess vulnerability towards a permanent threshold shift (PTS) (for a review see Henderson et al., 1993 [147]). However, it is not so much the extent of a TTS, but far more the course of recovery of the hearing threshold after a TTS where some correlation is described with a PTS [149], [150]. A detailed study by Plinkert et al. examined for example soldiers with a TTS after shooting practice under standardized laboratory conditions and identified transientlz evoked otoacoustic emissions (TEOAE) as the most sensitive objective measurement procedure at the time in order to record and possibly predict a subtle hearing impairment caused by noise [151].

Over recent years, study of the function of the medial olivocochlear system regarding protection against hearing impairment has received constant attention (see below). The medial olivocochlear system expresses a bilateral acoustically evoked reflex that can be measured from its influence on otoacoustic evoked emissions [152], [153]. An interesting approach to study individual noise sensitivity was described by Maison and Libermann (2000) [154]. In an animal-experimental study on guinea pigs the authors were able to show that (1) an activation of the olivocochlear system leads to protection of the inner ear from PTS, and (2) the intensity of the acoustic reflex of the medial olivocochlear bundle (MOCR) inversely correlated with the observed, individual vulnerability towards noise exposure (PTS). This test in modified form might find use as a non-invasive screening procedure for identifying people with a increased acoustic vulnerability [154].

Genetic aspects: Current progress in decoding the human genome [155] and the characterization of individual genes involved in heriditary hearing impairment [156] have stressed the relevancy of genetic aspects also with noise induced hearing loss. An important aspect in the identification of patients at risk from acoustically-induced hearing loss is the search for a genetic basis for acoustic vulnerability. Some animal-experimental studies have shown that mice homozygous for the Ahl gene show a increased susceptibility towards noise [157], [158], [159]. Other studies found mouse strains with exceptional resistance towards noise exposure [160]. An up-to-date review was published by Davis et al. [161].

The ear clearly possesses mechanisms for noise protection and hearing recovery. TTS and PTS are not distinguished without reason. Also, the risk of a PTS is higher with impulse noise when the peak level is reached within a few milliseconds (e.g. with a bang) than it is when the same peak level is reached over a longer period of time.

However, audiological recovery does not necessarily entail a complete biological recovery. An increasing risk of a PTS exists where repeated exposures are endured even if there is intermittent audiological recovery [14]. Obviously the ear has a acoustic trauma "memory", so that over time TTS can eventually lead to PTS in a manner dependent on total exposure time, recovery times, frequency and sound intensity.

7.1.1 Middle ear muscles

The middle ear reflex leads to a reduction in the intensity of incoming acoustic signals by bringing about an impedance change [162]. Signs of a potential clinical effectiveness arose from observations on patients with idiopathic facial palsy [163], who during paralysis showed TTS more easily than was the case after paralysis. However, because of the long latency time the reflex is unsuitable for protecting against impulse noise.

7.1.2 Efferents of the auditory nerve

Efferents of the auditory nerve control the inner ear. A protective role against noise has also been ascribed to them. One can distinguish the lateral olivo-cochlear system that controls the afferents of the auditory nerve emanating from the IHC from the medial olivo-cochlear bundle which acts mainly as an inhibitor of outer hair cell motor activity.

Lateral olivo-cochlear system: The efferent nerve fibers to the IHC originate from the lateral superior olive nucleus and do not terminate directly at the sensory cells themselves, but at the dendrites of the type-I afferents below the IHC [164], [165]. Different substances have been identified as neurotransmitters or neuromodulators at this synapse: acetylcholine, γ-amino-butyric acid (GABA), dopamine (DA), enkephalins, dynorphins and "calcitonin gene-related peptide" [120]. The lateral efferents ensure a modulation of input signals from the IHC into the central nervous system as an efferent feedback system [120], [166].

Of the inhibitory neurotransmitters in the lateral olivocochlear system, the role of dopamine has been examined best up until now. Intracochlear application of dopamine reduces the compound action potential (CAP) in the auditory nerve and leads to a threshold shift and drop in the spontaneous discharge rate without affecting the tuning properties of the auditory nerve fiber [167]. Upon stimulation at the characteristic frequency, dopamine lowers the stimulus-induced activity and reduces the dynamic range of the discharge rate from the auditory nerve fiber. Specific dopamine agonists lower the CAP-amplitude as does dopamine itself. The influencing of the spontaneous discharge rate of the type I afferents by dopamine antagonists depends on the spontaneous discharge rate pertaining to the respective nerve fiber subtype. In the fiber subtypes with lower intrinsic spontaneous activity, the application of dopamine antagonists results in an increase in spontaneous activity, an improvement in the threshold and a decrease in the maximal discharge rate of the nerve fiber. In fiber subtypes with high intrinsic spontaneous activity, a decrease in spontaneous activity below the initial level occurs after the initial rise in spontaneous activity following application of DA antagonists [167]. Morphological study of the synapses after application of DA antagonists reveals swellings of dendrites, as is also seen with excitotoxic damage [167]. These findings indicate that dopamine exerts a tonic-inhibitory effect on the activity of the primary auditory neuron and loss of this inhibition leads to early signs of excitotoxicity.

Medial olivo-cochlear system: The medial olivocochlear bundle (MOCB) includes those efferent fibers of which almost all terminate at the OHC. Efferent signals between the nerve fibers of the MOCB and the outer hair cells are conveyed mainly via the release of acetylcholine (ACh) and GABA (gamma amino butyric acid) [168], [169]. The ACh receptors of the outer hair cells share properties with nicotinic ACh receptors [168], [170]. The ACh receptors are ligand-activated cation channels of the α9 subtype (α9-nAChR) that are functionally co-expressed as hetero-oligomeric complexes with α10-nAChR [171]. Acetylcholine induces a calcium influx into the OHC, followed by a calcium-activated potassium efflux that hyperpolarizes the OHC and inhibits the rapid inhibitory efferent effects [172], [173]. Under physiological conditions the medial olivocochlear bundle is activated by contralateral sound signals [174], [175] so that the rapid motor activity of the outer hair cells is inhibited. Cody and Johnstone (1982) as well as Maffi and Aitkin (1987) showed that contralateral auditory stimulation can reduce an ipsilateral acoustic trauma [176], [177]. In the presence of strychnine (blocker of the olivocochlear bundle) protection is no longer found [178], [179], [180]. The observations are consistent with the proposition that inhibition of rapid hair cell motility combined with possible tonic outer hair cell movements (displacement of the operation point) may provide acoustic protection and adaptation of the cochlea at high oscillatory amplitudes [8]. Ryan et al. (1990) activated efferents by electric stimulation of the inferior colliculus in the guinea pig and found a reduction of an acoustically-induced TTS when they measured the N₁ threshold [181]. In contrast, Libermann (1991) found no signs of protection against excessive acoustic stimulation when the olivocochlear bundle was electrically stimulated in the cat [182]. The striking difference in results with guinea pigs may reflect real differences in the functions of the efferents in the two species, or the existence of different cochlear feedback mechanisms.

7.2.1 Endogenous anti-oxidative systems

As already discussed, excessively loud acoustic stimulation does not necessarily lead to an increase in fatally harmful reactive oxygen species in the cochlea [100], [101], but can lead to the activation of endogenous anti-oxidative mechanisms, such as an increase in glutathione in the cochlea [102].

For this purpose cells of the cochlea are equipped with a complex array of anti-oxidative defense mechanisms. Vitamins (ascorbic acid, alpha-tocopherol), glutathione and enzymes are included amongst others. With the enzymes one can differentiate between those involved in glutathione metabolism (glutathione-S-transferase, γ-glutamyl-cysteine-synthase, glutathione reductase, glutathione-peroxidase) and those involved in reactions with superoxide anions and hydrogen peroxide (superoxide-dismutase, catalase [92], [98]). The enzymes of glutathione metabolism show a higher activity in cochlear tissue than in other neurosensory tissues [183]. These endogenous, cochlear, anti-oxidative defense systems can inactivate reactive oxygen species and in so doing exert a protective effect against these noxins [25].

7.2.2 Neurotrophic factors

Amongst the counter-regulatory processes activated by acoustic trauma is the induction of neurotrophic factors (neurotrophins) such as glial cell-line derived neurotrophic factor (GDNF) [184]. Neurotrophins are small peptides that after binding to specific receptors of the cell membrane activate intracellular signal cascades necessary for development, cell differentiation, cell survival or (axonal) regeneration (see for a review Bibel and Barde, 2000 [185]). Keithley et al. showed that after local application of GDNF a protection of the inner ear is achieved towards noise exposure [186]. For this reason, and because expression of GDNF in the cochlea is increased after excessive acoustic stimulation, this protein has been assigned a role as an endogenous protective factor [184].

"Brain derived neurotrophic factor" (BDNF) and neurotrophin 3 (NT3) are other neurotrophins of the auditory system that enable and maintain the survival and neuritogenesis of neonatal auditory projections [187], [188]. Recent studies have also revealed important functions of these neurotrophins (BDNF and NT3) in the adult auditory organ [189], [190]. A protective effect on neurons and sensory cells in the inner ear has also been discussed for fibroblast growth factor (FGF) [191]. The expression of receptors for FGF and epidermal growth factor (EGF) in the hair cells and vascular endothelial growth factor (VEGF) receptors in the cochlea are also interesting, and knowledge here might lead to the use of these factors as hair cell protectors [191], [192], [193].

7.2.3 Heat shock proteins

Heat shock proteins (HSP) help protect organisms against various metabolic stress situations (for a review see Parsell and Lindquist 1993 [194]). They have also been verified in the cochlea [195] and their upregulation can be observed with heat stress [196], [197], transient ischemia and excessive acoustic stimulation [198]. Noise exposure to the rat (TTS) leads 6-8 hours after exposure to an upregulation of HSP72 [199]. The results of various studies have indicated that the activation of heat shock proteins induces a protective counter-regulation of the inner ear with threatened hearing loss. Yoshida et al. [197] exposed CBA/CaJ mice to a whole body heat stress (41.5 ºC for 15 min) before exposing them to noise (8-16 kHz, 100 dB SPL, 2h). Measurement of HSP70-mRNA by quantitative PCR revealed a 100-200 fold increase in HSP70 values in heat-exposed animals compared to controls. The animals with heat stress before noise exposure showed a much lower acoustic trauma-induced permanent threshold shift compared to control animals without heat stress.

7.2.4 Bcl-2 proteins

Apoptosis is tightly regulated by genes and proteins to which the family of intracellular Bcl-2 proteins belong ("B-cell follicular lymphoma"). It plays a decisive role in cellular survival and apoptotic death. Both pro-apoptotic and anti-apoptotic proteins belong to this group. Anti-apoptotic proteins of the Bcl-2 family play a special role in the inhibition of cytochrome-C release from mitochondria and are therefore inhibitory towards the activation of caspases (see above for the intrinsic pathway of apoptosis activation). Current studies on apoptotic processes in the cochlea after traumatic acoustic stimulation with or without sound conditioning revealed (1) a release of cytochrome C into the cytosol of outer hair cells after acoustic trauma alone, (2) an increased gene expression of the apoptosis suppression gene bcl-2 and (3) a reduced release of cytochrome C by "sound conditioning" [200]. These data allow an important role of bcl-2 to be presumed as an acoustically inducible, protective factor in the cochlea. If sound conditioning induces apoptosis inhibition via this pathway, this might represent an interesting approach for future medical intervention following acoustic trauma (see below).

7.2.5 Further counter-regulatory mechanisms

During the recovery phase after a noise-induced temporary threshold shift (TTS), further intracellular signal cascades are activated in mammals that serve cellular protection. Modern molecular-genetic screening procedures make it possible to search for genes whose protein products might participate in counter-regulatory cascades upon noise after gene activation. Lomax et al. examined differential gene expression in the cochlea of the rat after noise exposure using a microarray technique [201]. The TTS of 30-50 dB SPL (complete recovery after three hours) as well as the PTS of >70 dB was confirmed using brain stem potentials and cytocochleograms. Immediately after the 90-minute sound exposure the cochlea was removed and the RNA was isolated. Then the RNA was radioactively labeled by RT-PCR and a hybridization was carried out on cDNA arrays that contained 1176 genes. In the comparison between sound-exposed groups with non-sound-exposed control groups the ratios and the differences between the radioactive signals provided indications on differential gene expression in the two hearing loss models (TTS or PTS). RT-PCR analysis confirmed that various genes encoding for transcription factors involved in growth regulation cascades and which are activated at a very early stage after cell damage revealed a higher expression after PTS exposure compared to TTS exposure [Tab. 10]. These gene groups are of special interest for future experimental studies since they are looked upon as an interlink between studies at the molecular and the systematic level [202]. In addition to these genes, small secreted signal molecules were upregulated upon PTS exposure, but not after TTS exposure [Tab. 10].

Other research groups also used the gene microarray technique in a similar way, in order to study altered gene expression after TTS noise exposure. Studies on the chinchilla revealed classes of genes that were found to be upregulated in a time-dependent manner. These overexpressed genes have a direct influence on protein synthesis (increase in ribosomal and ribonuclear proteins), cell metabolism (rise in metabolic enzymes), proteins of the cytoskeleton (polypeptide neurofilament, tubulin-beta-5, "actin-related protein", myosin, alpha-actinin), calcium related proteins (calcium binding protein S100A1, calmodulin 3, calpactin, annexin VI) and heat shock proteins (HSP 74kD) [203]. In all, the abovementioned authors identified over 150 gene transcripts in their experiments whose expression is at least doubled upon TTS exposure. Interestingly, the authors could not verify any upregulation of gene transcripts associated with enzymes of the endogenous anti-oxidative system although an increase e.g. in the enzymes glutathione-reductase, gamma-glutamyl-cysteine-synthetase and catalase could be shown under similar stimulus conditions [204].

Although it is possible using gene array techniques to identify a large number of genes that play a role in inner ear reactions after sound exposure, and good candidates have indeed been described, the reliability and data evaluation of these experiments has not yet been adequately optimized. The generation of extremely large amounts of data and their subsequent filtering and evaluation present particular challenges. Furthermore, the data that can be produced with the aid of this technology can not often be verified by other means. For this reason a range of additional studies are required to identify the actual genes involved in cochlear cell responses upon noise exposure. The gene array technique here therefore represents only a first step.

The fact that noise-related injury is often mixed in nature creates problems for the interpretation of studies on the induction of cochlear signal cascades after noise trauma. In addition to a PTS component, a TTS component with partial recovery of the hearing threshold exists. Thus after "PTS damage" of the mammalian cochlea, mechanisms of protection, recovery and repair are activated in addition to signal cascades responsible for programmed cell death (apoptosis). Just as little as the upregulated genes found in response to noise exposure in chickens can be assigned to the stress response or recovery (see below), it is only possible within limits to assign differentially expressed genes after "PTS damage" in mammals to a specific signal cascade. These genes can be components of protection mechanisms, apoptosis or repair processes, or may even have multiple functions.

The stimulus parameters and observation periods have also varied in earlier investigations, as have the mammalian species studied. A systematic investigation of differential gene expression with various stimulus parameters, observation periods and with different families and species shall allow more information to be derived regarding the stress response, apoptosis, recovery, repair and the regeneration of molecular processes.

(Footnote: We thank Dr H. Löwenheim, Tübingen, for his critical reviewing of this section.)

When considering recovery processes in the inner ear, the concepts of repair and recovery must be clearly distinguished. With repair the process is a recovery of reversibly damaged, but surviving cells, while regeneration is the new formation of irreversibly damaged cells and cells lost through cell death. While in the mammalian cochlea regeneration of sensory cells is not usually observed, there are signs of repair processes after acoustic trauma. Recovery with temporary hearing-loss can be based on such repair processes in surviving cellular elements in the cochlea.

Since the transition between reversible and irreversible damage is smooth, and the advancement of cellular loss can be observed up to two weeks after a damaging event, a post-traumatic time window exists within which both processes are observed in parallel [85].

More recent experimental investigations using in vitro models suggest a regrowth of the stereocilia [97], [205], so that with appropriate damage a repair of the hair cells by regrowth of stereocilia seems feasible. According to observations on postnatal in vitro models the actin structures of stereocilia are renewed every 48 hours or so [97].

The hair cells are surrounded by supporting cells at which reversible damage is also observed [85]. Such reversible damage in the acoustic organ and the associated TTS is for the most part reversed within 2 weeks of the acoustic trauma.

After excitotoxic damage, a complete or partial morphological and functional recovery can occur within a few days through the new formation of synaptic complexes at the inner hair cells [206]. Synaptic repair processes could be shown after acoustic traumatic damage as well, and these processes occurring in the first days after an acoustic trauma played a role in the recovery of temporary threshold shifts (TTS) [133]. The synaptic repair processes are of particular interest in a pathophysiological and therapeutic respect. After excitotoxic damage in the cochlea, an overexpression of NMDA receptor mRNA (subunit NR1) is observed. This overexpression of NMDA receptors can cause an increased discharge rate of the primary auditory nerve fiber and in this way form the starting point for a peripheral post-traumatic tinnitus [206]. The experimental blockade of NMDA receptors leads to a delay in repair processes at the synapse, and a delay in functional recovery. This indicates a neurotrophic effect of glutamate during the post-traumatic synaptoneogenesis, i.e. NMDA receptors and glutamate might be important for retraining the correct synaptic "architecture" after an acoustic trauma [207].

Temporary contacts of the lateral efferent nerve fiber with the cell body of the IHC [206] are a further aspect of the synaptic repair process at the IHC after excitotoxic damage, a state that is also observed during the development of the cochlea. For this reason a hypothesis was formulated that a temporary, direct activation of the IHC promotes synaptic repair processes through lateral efferents via stimulation of glutamate release at the IHC [208].

Since the synaptic repair processes after excitotoxic damage progress in the same way as they do in normal development, pharmacological approaches for promoting the correct re-innervation of the dendrites with the IHC might be derived from knowledge of the developmental-biological processes involved in the formation of synaptic complexes at the IHC.

Regarding excitotoxicity due to excessive sound exposure, other disorders, such as mechanical damage to the stereocilia or the cell bodies of the outer hair cells have to be taken into account with the recovery processes that overlie the kinetics of the recovery/repair process at the synapses of the inner hair cells.

(Footnote: We thank Dr H. Löwenheim, Tübingen, for his critical reviewing of this section.)

If the intensity and duration of an acoustic trauma lead to the cell losses mentioned elsewhere, particularly regarding sensory cells, a restoration of function can only be conceived if the lost cellular elements are reformed. In mammalian cochlea the regeneration of sensory cells is not usually observed. A future approach, however, is to gain an understanding of the cell-biological processes necessary for recovery in the auditory organ. As a matter of principle, cells in the acoustic organ are actually capable of regeneration, but the recovery process is genetically suppressed. The first indications are provided by studies of the cell cycle inhibiting factor p27, the inhibition of which is associated with mitosis in the organ of Corti [209]. Löwenheim in his review in 2003 provided an extensive description on this topic [210].

Another option is to learn from species in which recovery occurs. Birds are examples of these.

After the loss of hair cells from excessive noise exposure, mechanisms of cellular regeneration are activated in the inner ears of birds. The lost hair cells can be reformed from cell divisions of neighboring support cells. As a result of the hair cell recovery a functional recovery of hearing thresholds to values near those observed before sound exposure occurs [211], [212], [213], [214], [215], [216]. A still existing residual loss in the hearing threshold of 10-20 dB may be due to an incomplete re-innervation of the afferent synapses at the regenerated hair cells [217]. Even if auditory function recovers significantly often subtle functional deficits remain, for instance the changes in rate-intensity coding in the chick auditory nerve after complete recovery of auditory threshold after noise trauma [198].

From the cloning and characterization of bird genes that are less or more active after an acoustic trauma, i.e. differentially expressed, molecular processes and intracellular biochemical processes should be identified that play a decisive role e.g. in sensory cell degeneration with subsequent recovery. A global approach for identifying genes that become differentially expressed after an excessive acoustic stimulation in the regenerating inner ear of the chicken, is pursued using a number of approaches including "differential display" and "subtractive hybridization" [218], [219], [220]. Earlier studies on differential gene expression after acoustic trauma in the chicken revealed numerous genes with an altered expression pattern upon noise exposure. These might play an important role in recovery processes. On the other hand the expression of these genes may also be a sign of the general cellular stress response, as also occurs in mammals.

The genes that are differentially expressed in the inner ear of the chicken after acoustic trauma can be divided into various classes: proteins of intracellular signal cascades, cytoskeletal proteins and regulators of the actin cytoskeleton [Tab. 11] [201]. While some of the genes that are upregulated after noise exposure code for known signal pathways (CaM kinase II, PTHrP), other proteins are presumed to be involved in other intracellular signal cascades (UBE3B, CRISP). Proteins of the cytoskeleton are required for the production of new hair cells with their actin-rich stereocilia bundles. Regulators of the actin cytoskeleton participate in mediating actin polymerization/depolymerization and are necessary for normal hearing function. Myosin and kinesin, also proteins associated with the cytoskeleton, also play an important role in hair cells. Of special importance is the differential expression of the COCH gene after acoustic trauma in the chicken. The COCH gene codes for an extracellular matrix protein. Mutations in the COCH gene amongst affected members of DFNA9-families lead to senescent hearing loss and equilibrium disorders [221].

The avian cochlea is a valuable cell-biological model, but it remains an open question whether it can provide conclusive answers about the human cochlea at the molecular level. The human acoustic (Corti's) organ and the avian basal papilla differ considerably in their construction and cellular structure.

Prevention must be a core aspect in the reduction of noise-induced hearing loss. Prevention means that the tympanic membrane receives no or a reduced sound level from the noise source. The decrease in sound exposure at the ear can be brought about by

• elimination of the sound source or

• noise reduction/sound-absorbing measures at the sound source,

• avoidance of the sound source

• personal ear protectors worn by the individual.

In order to act against recreationally-induced risks to hearing amongst young adults, adolescents and children, prevention by education of those most exposed to music should be provided from physicians, schools, youth and social workers, and also the media. A noise-related health education program within the public curriculum of all schools in Germany was recently prepared by the German Federal Center for Public Health Education. Further school educational measures were provided by the initiative "Take care of your ears" of the German Society of ENT Physicians as well as the German Society of Oto-Rhino-Laryngology, Head and Neck Surgery. Since age-appropriate preventive measures can not be delivered to adolescents and young adults using a "wagging finger" approach, the instructional materials made available to teachers represents the more helpful way.

Together with advice provided by physicians and parents, such measures fail to influence all adolescents and young adults. A recent study on the influence of information about the risks from hearing excessively loud music on hearing protection behavior amongst adolescents in Austria showed that although the information sensitized individuals to the issue of subjective volume assessment, such individuals wore hearing protection and attended discos to the same extent as uninformed individuals of the same age [222]. It can also be assumed that the 10% group with the extremely excessive leisure noise exposure are the most hostile to any form of protection counseling.

The lack of regulatory measures regarding noise exposure in discos and large music events, as well as with New Years Eve celebrations is deplorable. Proposals by the German Federal Medical Council have been made [223].

The Federal Immission Control Law (BImSchG) in particular lays the basis for general measures on noise control in its amended version from July 1995, according to which humans must be protected from noise immissions that according to their nature, intensity and duration are sufficient to entail risks, considerable disadvantage or considerable nuisance for the community or neighborhood. In addition the 18^th Decree for Implementing the Federal Immission Control Law exists for sports/recreational institutions (18. BImSchV) with references to "Recreational facility noise". They serve to protect residents in the neighborhood. Regarding the type and intensity of outgoing sound emissions coming from a facility - even a facility that is not required to certify itself - as well as the immissions within the facility itself, the regulations in the "technical instructions on noise", the 6th general control regulation of the BImSchG must be followed. Accordingly, facilities must be run in such a way that damaging environmental effects must be prevented as best as possible according to the current state of the art, and that unavoidable harmful noise effects are restricted to a minimum according to the current state of the art for noise reduction. The "technical instructions on noise" include measurement and evaluation procedures that consider the special characteristics of noises and quote immission standards that should not be exceeded if considerable nuisances and risks are not to be experienced.

Further regulations are contained in state immission control laws as well as noise or police regulations in the state (Länder) and local statutes. They set out at which times an increased need for quiet exists, and limit the use of loudspeakers and music systems to an extent that third parties are no longer nuisanced as far as is unavoidable according to the circumstances. These regulations, however, do not suffice to protect attendees of music events etc.

Fortunately, the norms for sound immissions from toys as well as music reproduction devices are currently being developed at a European level so that future reductions in risks to hearing can be counted upon in CE certified devices.

The already mentioned control measures and protection rules for the workplace are laid out in the "accident prevention regulation noise" (UVV Lärm) [21]: "Hearing damage can already occur at a noise level whose rating level exceeds 85 dB(A). Noise areas are areas in which noise occurs at which the rating level reaches or exceeds 90 dB(A). They must be categorized as such". From 85 dB(A) personal noise protection must be made available by the company according to VDI guideline 2560, while from 90 dB(A) that protection must be used. Regarding the early recognition of pathological changes due to the actions of damaging occupational noise, the fundamentals for professional/trade association for occupational medical check-ups (BG G20 "Noise") provide advice regarding the realization of audiometric tests and their potential diagnostic scope.

As much as the "accident prevention regulation noise" is an important step for protecting the employee at the workplace, simultaneous chemical exposure at the workplace must also be taken into account in the preventive assessment in the future [26].

The considerable noise exposure from recreational activities in approximately 10% of an age-group amongst adolescents and young adults must also be considered since these exceptionally disrupt the assumed recovery times for the "accident prevention regulation noise".

The most expedient measures for preventing noise-related injury to the ear are of course the above-mentioned ones that reduce the level of sound energy reaching the tympanic membrane. However, such measures can not prevent all forms of noise-related injury to the ear. The inherent limits of such measures include:

1) Very high sound levels for which the attenuation effect of the ear protectors is not sufficient

2) Sound propagation via the skull bones

3) Suboptimal fitting of ear protectors

4) Situations in which no protection of the ears can be worn (e.g. military engagements, orchestra musicians)

5) Compliance problems

Protective therapeutic procedures that are applied pre-exposure and/or during noise exposure (para-exposure) therefore represent a desirable element of a comprehensive strategy with the goal of preventing noise-induced hearing loss as the most frequent cause of acquired inner ear hearing loss. The protective procedures when applied rationally should either block an unwanted pathophysiological pathway or activate endogenous protective mechanisms.

11.1.1 Inhibition of excitotoxicity

One pre-/para-exposure therapeutic procedure of evidence level Ib exists. In the section on excitotoxicity it was pointed out that blockade by extracellular magnesium ions represents an interesting, possibly therapeutically exploitable functional property of the NMDA receptor. Blockade should act against the unwanted influx of excessive calcium ions through the NMDA receptor channel into the afferent auditory nerve fiber and in this way prevent calcium ion-mediated excitotoxicity.

In animal experiments a positive correlation between serum magnesium levels and a decrease in chronic noise-induced hearing loss was observed [224], [225]. The pre-exposure application of magnesium reduces morphological and functional damage from an acoustic trauma in guinea pigs [226], [227]. An explanation for this effect might lie in magnesium's role in controlling the NMDA receptors of the afferent auditory nerve fiber that are involved in excitotoxicity, or an effect influencing the function of the OHC [224]. The first double-blind, randomized studies (evidence level Ib, see description in [Tab. 9]) with pre- and para-exposure application of magnesium aspartate are consistent with these ideas [228]. Three hundred healthy young soldiers were given magnesium aspartate during a 2-month military firing practice in a randomized, placebo controlled, prospective study. The verum group received a Mg-aspartate drink containing 167 mg of the agent daily, while the control group drank a placebo. In the placebo group hearing loss, and particularly bilateral hearing loss, was found significantly more frequently. Side effects were not observed [228].

Pharmacological mimicry of the protective dopamine effect appears to represent another therapeutic possibility. As mentioned above, dopamine presumably participates in protecting the auditory nerve from noise-induced injury by reducing excitotoxic effects [229], [230]. The first animal-experimental approaches with the dopamine-2 agonist piribedil (an antiparkinson- drug) during excessive acoustic stimulation revealed a clear protective effect regarding the hearing threshold as well as the morphology of the afferent-efferent synaptic complexes of the afferent auditory nerve fibers in the guinea pig [230].

In addition, NMDA receptor blockers are also available for clinical use (for the role of the NMDA receptor, see the above section on excitotoxicity). Examples include caroverine as well as memantine. They have been discussed as potential treatments for tinnitus in particular (see section below on tinnitus [231], [232]). Using iontophoretic techniques, Oesterreicher et al. were able to show in animal experiments a marked inhibitory effect of both substances on a glutamate-induced, presumably NMDA associated activity in afferent auditory nerve fibers [233].

The regulation and control of the extracellular glutamate concentration is another potential future approach for preventing excitotoxic damage in the cochlea.

11.1.2 Antioxidants

As already discussed, acoustic stimulation leads to an increase in reactive oxygen species in the cochlea that in turn can initiate apoptosis [100], [101] (see also section 5.2). If endogenous anti-oxidative defense systems such as glutathione are not able to inactivate the reactive oxygen species [25], [108], this paves the way for a supplementary pharmacotherapy. Vitamin C and salycilate might be used for this purpose since they act as radical scavengers. Another radical scavenger is alpha-liponic acid, although this is hardly ever applied for protective purposes. Furthermore, an anti-oxidative hydroxyl radical scavenger property for caroverine has also been reported [234].

A general review of anti-oxidative therapy strategies with acute damage to the central nervous system was published by Gilgun-Sherki et al in 2002 [235]. Table 12 [Tab. 12] shows a list of antioxidants that have revealed a protective effect against acoustic trauma in animal experiments.

11.1.3 Inhibition of apoptosis

Even if up until now there has been too little research on the role of signal cascades that lead to apoptosis in cochlea cells after acoustic trauma, the first results with inhibitors of these signal cascades point towards therapeutic possibilities for the future. As described above, apoptosis is tightly regulated by genes and proteins to which caspases also belong. Currently, extensive basic research is being conducted on various neuron related illnesses (Parkinson's, apoplexy) to see whether inhibition of caspase pathways can be clinically exploited [236].

For acoustic trauma, caspases represent an interesting potential target for pharmacological intervention [137]. A further possibility for preventing or restricting apoptosis in general and specifically after acoustic trauma is the JNK signal pathway [237]. Inhibitors of this signal cascade can reduce the morphological and physiological consequences of an acoustic trauma in animal experiments [145], [238].

Recently, the idea of reducing the sensitivity of the auditory organ to harmful sound exposure by a preceding auditory stimulus, also known as sound conditioning, toughening, "priming" or here also as "noise stress", has regained a certain popularity. This idea was already developed in 1963 by Miller et al. on the basis of animal-experimental studies [239].

Auditory protection by a preceding exposure to a therapeutically effective sound signal does not appear to be impossible. However, any clinical application at the workplace, or against recreational sources, fails because of our lacking knowledge on the properties of the sound signal required. Previous findings allow us to assume that if such protection is possible, it will have to be induced by a specific acoustic signal that is not necessarily encountered in everyday life or at the workplace. This specific sound signal is unknown.

11.2.1 Studies

Miller et al. [239] exposed cats on 16 consecutive days to sound stimuli interrupted regularly with pauses, and observed a reduction in threshold shifts following noise exposure on the 16th day compared to the first day.

There are also signs of possible protective effects from a biochemical perspective. After stress-induced upregulation of heat shock proteins (HSP, details see above) by acoustic stimulation, a subsequent protection was found towards excessive acoustic stimulation. A precondition was that the traumatic noise injury occurred during a period after the conditioning in which the HSP were still raised [197], [198]. These observations suggest that heat shock proteins might play a role in reducing the sensitivity of the auditory organ towards injurious sound stimuli with preceding application of acoustic stimuli.

Several recent studies indicate that continuous or regularly interrupted sound exposure with low, harmless levels, can induce metachronous protective effects with subsequent harmful sound levels in some of the experiments. This phenomenon was examined in different mammalian species such as mice, desert gerbils, chinchillas, guinea pigs, rats, rabbit as well as in man (review in Niu and Canlon, 2002 [240]). A number of sound exposure strategies have been found to be successful in animal experiments under laboratory conditions, while others have not.

11.2.2 Acoustic stimulation paradigms

Two different acoustic stimulation paradigms designed to reduce the sensitivity of the auditory organ to traumatizing sound stimulation can be distinguished. They are designated as either "conditioning" or "toughening" of the auditory organ.

Sound conditioning: This is generally understood as the continuous or intermittent application of an innocuous acoustic stimulus, i.e. with a low sound level, before the traumatizing noise exposure. This phenomenon has been examined in different species such as the desert gerbil [241], [242], [243], the rat [244], the guinea pig [245], [246] and in man [247].

Sound toughening: This is generally understood as the intermittent stimulation with sound levels that are loud enough to cause a temporary threshold shift (TTS). During daily repeated stimulation the TTS observed may become smaller and smaller. In some experiments TTS no longer occurred despite the ongoing presence of daily stimulation. This phenomenon was also examined in different species, e.g. in desert gerbils [248] and particularly in chinchillas [249], [250], [251], [252], [253]. However, the semantic use of the terms sound conditioning, toughening and "resistance" is not always applied religiously in the literature, and the terms are sometimes even arbitrarily exchanged.

Studies on this subject differ with regard to species, the parameters of the preceding stimulation and the parameters for the subsequent acoustic trauma (frequency spectrum, intensity, duration, length of pause between the conditioning/toughening and traumatic sound exposures, impulse trauma or the continuous traumatizing stimulus). As such the results of these studies must be assessed in a differentiated manner and interpreted cautiously. Within the same species, small modifications in acoustic parameters can result in either no or a clear protective effect of the sound conditioning. While Fowler et al. showed initially no positive effect of sound conditioning in mice [254], later experiments with other stimulus parameters did show a protective effect of a preceding sound stimulation [255]. No change in permanent hearing loss induced by an acoustic trauma (narrow-band noise at 4,5 kHz, 107, 110, or 117 dB SPL, 24 h) was shown after continuous or interrupted sound conditioning with narrow band noise of 4.5 kHz and various combinations of intensity and duration of the "training" (80-96 dB SPL, 7 or 10 days) compared to control animals without previous "conditioning" [254]. However, Yoshida and Libermann, also using mice, showed both with shorter-term (8-16 kHz, 89 dB SPL, 15 min) and longer conditioning (8-16 kHz, 81 dB SPL, 1 week) a decrease in hearing-loss induced by a subsequent acoustic trauma (8-16 kHz, 100 dB SPL, 2h) [255]. Since both studies employed different auditory parameters, they are not contradictory. Instead, this comparison illustrates how important the systematic investigation of this phenomenon is before any extrapolations can be made to the human case.

Under no circumstances do observations suggest that noises from everyday life or the workplace can prevent acoustically-conditioned hearing loss. On the contrary, the extraordinary variability in the requirements for any possible protective sound signal renders their existence in everyday life or the normal workplace as unlikely.

11.2.3 Mechanisms of action of pre-exposure acoustic stimulation

Various hypotheses concerning the mechanisms that might underlie the observed protective effects of sound conditioning have been investigated over the last years. These have dealt with the role of the middle ear muscles and the medial olivocochlear system as well as the stabilization of various endogenous cochlear protective systems, such as antioxidants or radical scavengers, heat shock proteins (HSP), changes in the expression of glutamate receptors, neurotrophic factors and calcium buffer systems; some shall now be described in more detail.

Middle ear muscles: It is well known that the middle ear reflex reduces the strength of incoming auditory signals by altering impedance [162]. In this respect it is clear that the middle ear reflex might play a role in sound conditioning. Three independent animal-experimental studies examined the protective effect of sound conditioning after severing of the middle ear muscles in the desert gerbil [242] and the chinchilla [256] as well as after paralysis in the guinea pig [257]. All three studies showed a protective effect of the sound conditioning despite these operations, and they all therefore came to the conclusion that the middle ear muscles exert no appreciable influence on this phenomenon.

Medial olivo-cochlear system: The role of the olivo-cochlear system (OCB, details see below) in sound conditioning and toughening is not yet clarified and the experimental results are inconsistent. Severance of the olivocochlear bundle in chinchillas led to larger TTS and PTS and more extensive hair cell loss than was seen in animals with an intact efferent system [258]. Kujawa and Liberman on the other hand observed less distinct phenomena in the guinea pig [246]. After complete severance of all olivocochlear fibers there was a decrease in the protective effects of sound conditioning, but not when only the crossed parts of the OCB were severed. However, a stronger PTS was seen in animals with complete severance of the OCB and sound conditioning than was seen in animals operated on without sound conditioning. Since independently of sound conditioning a protection appears also in cases with surgical lesions in the brain stem without complete severance of the olivocochleaer bundle ("sham surgery"), the authors attributed the protection more to the general stress reaction than the sound conditioning [246]. Experiments in which occlusion of an ear during the sound conditioning led to a loss in the protective effect indicate more a local process in the cochlea than a systemic process [259]. From a combination of sound conditioning experiments and local application of strychnine to deactivate the OCB, the same authors revealed that although the medial efferent system is involved in the protection of the inner ear against noise exposure, it appears to play no role in sound conditioning [260].

Endogenous antioxidants: Sound conditioning leads to an increase in catalase and glutathione reductase in the chinchilla cochlea and accordingly augments potential anti-oxidative defense systems in that tissue [261].

Heat shock proteins: After stress-induced upregulation of heat shock proteins a protection was found against excessive acoustic stimulation. Here it did not matter whether the upregulation of HSP was induced by heat stress or acoustic stress, provided that subsequent traumatic noise-induced injury occurred during a time window after conditioning within which the HSP level were still increased [197], [198]. Such observations suggest that heat shock proteins might play a role in reducing the sensitivity of the auditory organ towards harmful sound exposure when a preceding acoustic stimulation is provided.

Neurotrophic factors: Nam et al showed an increased expression of GDNF in the cochlea in response to excessive acoustic stimulation [184]. To what extent this and other neurotrophic factors might play a role with acoustic trauma and protection by sound conditioning is currently the subject of investigation.

Bcl-2 proteins: It is presumed that anti-apoptotic proteins of the bcl-2 family prevent the release of cytochrome C and thus the activation of caspases [140]. Current investigations on apoptotic processes in the cochlea after traumatic sound stimulation with and without sound conditioning have shown an amplified gene expression of the bcl-2 gene and a decreased release of cytochrome C upon sound conditioning [200]. These data allow us to presume a possible role of bcl-2 as an acoustically inducible, protective, anti-apoptotic factor in the cochlea. If sound conditioning acts via this pathway to inhibit apoptosis, this might represent an interesting approach for possible future therapeutic interventions.

Temporary inactivation of mechano-electric transduction channels: A further, interesting approach to explain the protective effect of pre-traumatic sound exposure is provided by the (1998) "inactivation model" described by Patuzzi. By experimenting on himself the author brought about temporary threshold shifts (TTS) and described the multi-exponential course of the threshold recoveries observed using a mathematical model of the various kinetic states of the outer hair cell mechano-electric transduction channels [262], [263]. If conformational changes in the mechano-electric transduction channels occur during a TTS due to excessive acoustic stimulation, this might contribute towards the reduced vulnerability to the immediately following acoustic trauma.

11.2.4 Conclusions

Sound conditioning i.e. the targeted, continuous or intermittent acoustic stimulation with "innocuous" sound levels before an acoustic trauma might represent an opportunity to decrease the extent of a noise-induced hearing loss when applied under precisely defined and very differentiated conditions. The conditions, however, required for human application are unknown.

A reliable post-exposure pharmacotherapy does not exist. The goal is to provide and/or develop pharmacological strategies, based on known mechanisms of cellular damage, death and recovery, that can beneficially influence subcellular and molecular processes after an acoustic trauma by stimulating recovery and repair. For this purpose the natural biological protection mechanisms should be augmented or the harmful processes should be suppressed. This also applies to any future drugs. There are also rational justifications from the pathophysiological circumstances outlined in this review for certain pharmaceuticals that are already available today.

Figure 6 [Fig. 6] provides an overview of the acute metabolic consequences of an excessive acoustic stimulation and the possibilities for intervention derived from them. A selection of results from recent animal-experimental studies on pharmacological-preventive or protective-therapeutic intervention is provided in table 12 [Tab. 12]. From figure 6 [Fig. 6] and table 12 [Tab. 12] it can be seen that the same pharmacological intervention can be provided as pre-exposure, para-exposure or post-exposure protection.

The following therapeutic approaches shall be discussed:

1. Activation of protein biosynthesis for cell repair

2. Anti-oxidative strategies

3. Anti-apoptotic strategies

4. Anti-excitotoxic therapy

5. Anti-inflammatory therapy

12.1.1 Cell repair for activation of protein biosynthesis

A protein-specific activation, e.g. in the form of a specific increase in gene expression, would of course be desirable. This includes above all the gene classes for which a time-dependent upregulation has been measured [203]. Their pharmacological or gene-therapeutic activation could be exploited to bring about a deliberate increase in ribosomal and ribonuclear proteins, the metabolism of cytoskeletal proteins (polypeptide neurofilaments, tubulin-beta-5, "actin-related protein", myosin, alpha-actinin), proteins related to calcium (calcium binding protein S100A1, calmodulin 3, calpactin, annexin VI) and also heat shock proteins (HSP 74kD) [203]. An increase in actin proteins together with link proteins of the stereocilia might assist in the reconstruction of stereocilia and postsynaptic nerve endings of the auditory nerve upon acute noise-induced hearing loss.

A general, non-specific activation of cellular protein biosynthesis can in principle be induced by cortisone and its derivatives [264]. Given at sufficiently high doses (at least 250 mg of prednisolone in man [265]) it reaches the intracochlear fluid spaces even after systemic application [266], [267]. According to the guidelines for sudden sensorinerual hearing loss (SSNHL) of the German Society of Oto-Rhino-Laryngology: Head and Neck-Surgery, a high dose prednisolone application has now become part of a modern therapeutic concept. However, an acute acoustic trauma is not the same as idiopathic SSNHL. Nevertheless one can still assume that in some sudden deafness patients there are some common cytopathological characteristics with acute acoustic trauma due to the fact that both processes share common final pathways for injury. Common therapeutic approaches can therefore be derived from this.

12.1.2 Anti-oxidative therapy

The highly reactive oxygen species that lead especially to apoptosis can be challenged therapeutically using currently available oxygen radical scavengers. According to the above guidelines, alpha-liponic acid can be included as a radical scavenger in a current therapeutic concept. Additional oxygen radical scavengers include vitamin C and Salicylate. Caroverin functions as an hydroxyl radical scavenger [234] and showed otoprotective effects in animals after acoustic trauma [268].

The following approaches, still in the phase of research, are also being pursued (see also table 12 [Tab. 12]):

• Exogenous application of anti-oxidative enzymes and their precursors

• Stabilization of endogenous anti-oxidative systems

• Exogenous application of metal ion chelators

• Exogenous application of lipid peroxidation inhibitors

12.1.3 Anti-apoptotic strategies by specific protease inhibition

Such a therapy is still not available today. Various proteases play an important role in apoptosis. Amongst these include the caspase family and calpain. Inhibition of these proteases therefore represents 1) a fundamental target for intervention in order to prevent the apoptotic death of cells in the inner ear after a traumatic insult, and therefore also 2) a starting point for the therapy of acoustic trauma [269].

The following therapeutic strategies are being investigated:

• Inhibition of caspases

• Inhibition of calpain

• Upregulation of anti-apoptotic genes (Bcl-2)

12.1.4 Anti-excitotoxic therapy

Clinically, NMDA receptor blockers are available for use (role of NMDA receptors: see above section on excitotoxicity). Examples include caroverine as well as memantine. They have been suggested for the therapy of tinnitus in particular (see section below on tinnitus [117]). These drugs might also be suitable for the treatment of acute noise-induced hearing loss. However, the fact that NMDA receptors might be needed for the repair of afferent nerve endings appears to contradict this [207].

In the section on protection it is pointed out that the blockade of NMDA receptor channels by extracellular magnesium ions is an interesting, apparently protective functional property. A blockade should positively influence the unwanted entry of excessive calcium ions into the afferent auditory nerve fiber. Clinically, positive results are now available from the first double-blind, randomized controlled studies (evidence level Ib) involving pre- and para-exposure application of magnesium aspartate [228], [270].

In the mean time, recent animal-experimental studies have shown that an early post-exposure magnesium application after impulse noise-induced acoustic trauma leads to a time-dependent decrease in morphological and functional damage of the inner ear [271], [272], [273].

Further research perspectives include the pharmacological mimicry of the above-mentioned protective dopamine effect using dopamine agonists [230]. The regulation and control of extracellular glutamate concentration also seems to represent a future possibility for the prevention of excitotoxic damage in the cochlea. As such, the following targets appear to exist for pharmacological intervention to prevent excitotoxic damage after excessive noise exposure:

1) the pre- and postsynaptic repair processes at the synapse between the IHC and the afferent dendrites of the primary auditory nerve,

2) the synaptic connection of the lateral efferent system with these dendrites and

3) the IHC themselves. Regarding the IHC, current studies on glutamate reuptake from the synaptic cleft point towards the excitotoxic role of potential defects in membrane proteins important for the reuptake of excitatory amino acids [274].

4) Furthermore, between mechanical overstimulation and excessive glutamate release with subsequent excitotoxicity there is a step which in principle is also capable of being pharmacologically modulated. This is the calcium influx through voltage-controlled calcium-channels that induces the exocytosis of glutamate [275]. More than 90% of these Ca channels belong to the class D L-type Ca channels [276], [277]. Since the sensitivity of the Ca channels on the IHC is only very low towards the known L-type Ca channel antagonists nimodipine and nifedipine, and selective antagonists for class D Ca channels are not yet known, a pharmacological intervention at this level can not yet be carried out.

12.1.5 Anti-inflammatory therapy

An anti-inflammatory therapy appears to be indicated particularly when a necrosis is presumed from the case history and the audiogram. Apart from conventional anti-inflammatory substances, one can also consider the anti-inflammatory effect of cortisone derivates. Together with the above-described activation of protein biosynthesis, two desirable, albeit non-specific effects of one substance can be achieved.

12.1.6 Neurotrophic factors

Generally, neural growth factors supplied by (1) exogenous application, (2) genetic over-expression or (3) endogenous stimulation of their production, promote the neural cell functions of survival and growth (see for a review [278]). Given before a traumatic sound exposure, local application of GDNF results in a clear morphological and functional protection in animal experiments [279]. These results provide hope that the role of neurotrophins in the CNS [278] can also be transferred to the auditory sensory cells and neurons.

The currently available conventional drugs can be applied systemically. However, future specific therapeutic forms may require a topical application of cytobiological or gene therapeutic agents. As such the following section is dedicated to local therapy.

12.3.1 Application strategies

Now that the elementary understanding of the pathophysiology and pathobiochemistry of sensorineural hearing loss has grown considerably and cell-biological as well as molecular-biological methods have opened up previously locked doors, current research activities are now directed towards the use of cell biological and gene therapeutic agents for treating sensorineural hearing loss including noise-induced hearing loss. Their experimental application has already allowed us to recognize that their local application shall be of more benefit since systemic side effects can be minimized and therapeutic effects can be optimized. The necessity of local application becomes particularly apparent with gene transfer into the cochlea for the therapy of inner ear disorders [280], [281], [282], [283].

As such another area to be researched can be recognized, i.e. the topical application of medicines to the middle ear for the treatment of the inner ear. For this, the fundamental mechanisms of substance transport from the middle ear to the inner ear as well as in the inner ear itself must be understood so that effective therapies can be developed in the future. Amongst the application strategies include intratympanic injection through the tympanic membrane or, as part of a tympanoscopy, the application via a partially or fully implantable catheter and pump system. The intratympanic injection can occur singly or repeatedly as has already been done for years with the gentamycin therapy in Meniere's disease [284].

Substance control in the middle ear is important here. Access to the round window membrane can be hindered because of the frequently encountered occlusion of the round window niche by collections of connective tissue or false membranes [285], [286]. Furthermore, it has to be ensured that the medicine remains for a defined period of time at a defined concentration in the middle ear [287], [288]. As such, volume stabilization has been proposed in the middle ear, as provided e.g. by hyaluronic acid and fibrin glue, that conveys a potential benefit over direct injection of liquids that can easily flow away via the eustachian tube [289], [290]. It can be expected that systems for controlled medication release, e.g. medication pumps and pharmaceutical-technical modulators of release, shall play an increasing role in this area of research. Partially implantable catheter systems with an external medication pump [291], [292] or fully implantable medication delivery systems need to be distinguished [293], [294]. Medication application in this way can proceed either in a stepwise manner [294] or continuously [292].

12.3.2 Pharmacokinetics in the inner ear

Regarding pharmacokinetic principles in the inner ear after local application of medications to the round window membrane (RWM), only a few findings have been published up until now. Some studies have qualitatively described the permeability of the round window membrane for various substances depending on molecular size and charge (see for a review Goycoolea, 2001 [295]). Quantitative figures on round window membrane permeability and time-dependent concentration changes at different target structures in the inner ear are rare. Some authors have reported concentration-time curves after topical application to the round window membrane [266], [267], [296]. However, sampling from the inner ear is technically difficult and often afflicted with artifacts [297]. Computer simulations can assist the establishment of a theoretical basis for substance distribution in the inner ear fluids after local application as well as the interpretation of experiments for quantifying concentration time courses [287], [288]. On this basis experiments can be planned and interpreted while different application strategies for local pharmaceutical application to the round window membrane can be compared with one another and improved if necessary. This represents an absolute necessity for the increased use of this application form in clinical practice.

Current treatment procedures for chronic hearing loss consist of the fitting of hearing aids or the implantation of electronic hearing-implants wherever these are indicated [298], [299].

A regenerative therapy with de novo formation of outer hair cells is not currently possible. Incited by the partial recovery observed in lower species (see above), however, the first basic scientific results have been published that show that a recovery of outer hair cells in the adult cochlea is not fundamentally impossible. Thus, an increase in cell proliferation is observed in the cochlea upon the loss of the cell cycle factor p27 [209]. Details of recent regenerative therapeutic approaches can be found in reviews published for instance by Loewenheim in 2003 [210] or by Minoda et al. in 2004 [300].

In a certain proportion of chronically afflicted individuals comorbidities are observed that can be attributed to the permanent hearing loss. Although rare, they are in principle the same as those observed in tinnitus, and for this reason they are addressed in the following section on tinnitus.

This review up until now has focused almost exclusively on hearing loss. For a tinnitus associated with a noise-induced hearing loss one does of course assume that similar peripheral, pathophysiological and biological mechanisms are involved as those seen in hearing loss. This applies also for isolated tinnitus that can be a precursor stage for a later occurring hearing loss. A narrow frequency band hearing loss (e.g. between the measurement points of a pure tone threshold audiogram) can also escape detection in an audiological examination.

For acoustically-induced acute tinnitus, the same etiopathological, preventive and therapeutic considerations apply as is the case for acute hearing loss.

A similar situation applies for chronic tinnitus, as long as the cochlea is only affected. Unlike the case with isolated hearing loss, however, an additional secondary centralization occurs in many individuals with chronic tinnitus [301], [302].

The pathophysiological basis for the centralization is presumably either a

• central, mainly cognitive sensitization (Zenner model) or a

• central, tinnitus signal amplification (Jastreboff model).

Furthermore , as is also the case with isolated hearing loss, a

• psychological coping disorder is possible.

Furthermore considerable

• comorbidities can occur.

14.2.1 Central tinnitus amplification

The model of central tinnitus hyperactivation assumes a massive pathological increase in the activity of the tinnitus signal in the auditory pathway. The tinnitus-specific activity increase leads to perception of a pathological loudness (a so-called phantom perception) for the tinnitus. Subconscious, central associations of the tinnitus signals especially with emotions appear to be involved in the increase in activity. Jastreboff, and later also his colleague Hazel, can be thanked for the pioneering achievement of this model [303], [304]. From their model of a neurophysiological tinnitus amplification at the perceptive level they suggested an acoustic re-training therapy (see below) for influencing perception.

14.2.2 Central sensitization

A more recent model of central sensitization does not assume a tinnitus hyperactivation. The central aspect is rather a cognitive hypersensitivity (sensitization) towards the tinnitus caused by a reduction in the cognitive threshold for the tinnitus [305], [306]. Like the activation model, subconscious, emotional associations play an important role. However, five important associations and links at the conscious level i.e. (i) arousal experiencing the tinnitus as (ii) noxious and (iii) unpredictable and inducing (iv) loss of control and (v) fear [Tab. 13], occurring at a cognitive level, are also involved. Such associations supposedly cause the reduction in threshold [307], [308].

Sensitization is a fundamental neurophysiologic mechanism of the final cognitive pathway of the auditory system which leads in this case to a pathological hypersensitivity in tinnitus perception with chronic tinnitus. It represents the outcome of specific, central nervous, neurophysiologic learning processes that is based on the plasticity of the central auditory system [307], [308].

14.2.3 Comorbidities

In chronic, acoustically-induced hearing loss, but even more so with chronic tinnitus, comorbidities are widespread. Here the review shall be restricted to a shortlist of conditions [Tab. 14].

For the treatment of tinnitus centralization, two central learning procedures are currently employed that are each based on the two above-mentioned neurophysiologic models: the passive acoustic tinnitus retraining therapy and the active tinnitus desensitization therapy (TDT), which is as neurootological, cognitive behavioral therapy. Both therapies aim at tinnitus habituation. Neurophysiologically, habituation is a specific learning process [307]. People learn more effectively when they participate in a learning process actively and at their own initiative (so-called specific cognitive "behavior") instead of passively receiving information and a sound signal [308].

14.3.1 Passive learning therapy: acoustic retraining

According to Jastreboff and Hazell, the goal of an acoustic retraining therapy (tinnitus retraining therapy, TRT) is to reduce the raised neural activity caused by the tinnitus by habituating to an external sound signal [304], [309]. A passive exposure to sound for several hours, every day over months is used for that and this usually takes the form of a white-noise signal. In addition, TRT also involves face-to-face counseling where appropriate information is provided to allow the patient to understand the clinical condition and therapy (so-called "directive counseling" [Tab. 15], [310]). Additional therapeutic procedures, especially group therapy, behavioral therapy or psychotherapy, are not indicated by Jastreboff-Hazell, and should not to be considered as part of a TRT. In Germany, TRT is not rarely combined with the cognitive behavioral therapy mentioned below. In this case this is no longer simply a TRT anymore [311].

A large number of aids are available for instrumental treatment with a tinnitus retraining therapy: these include specially constructed maskers (e.g. Noiser®, Tinnitus Control Instrument®), hearing aids, tinnitus instruments (combined masker and hearing aid). The tinnitus instruments can be switched over to a hearing aid function, or both functions can be employed simultaneously [312].

Prospective, randomized controlled studies of sufficient statistical power are lacking [313]. According to non-experimental empirical observations, the psychosomatic tinnitus stress and the daily sensation are reduced while the quality of life and the noise discomfort threshold are improved in approx. 40-80% of cases within 12 months of wearing the device for approx. 6 hours per day [309], [312], [314].

It is remarkable, however, that the authors got the same results when patients only underwent a tinnitus counseling without undergoing the acoustic part of the re-training therapy. When counseling was stopped, a drastic deterioration of the results occurred [315], [316]. This indicates a cognitive part of the "direct counseling" of the TRT, that obviously represents the main basis of the therapeutic effect [310].

The results suggest that the indication of white-noise procedures in TRT should be looked at critically. If the white-noise device is renounced, a cognitive therapeutic element clearly remains that by definition is no longer a TRT (a therapy without sound application is not a TRT). Modern cognitive therapy is now discussed in the following section on desensitization.

14.3.2 Active therapy: cognitive desensitization

As its name suggests, this collective term (abbr.: TDT, tinnitus desensitization therapy) includes intervention procedures that lead to a cognitive desensitization. As a result of active, behavioral-medical processes that specifically influence tinnitus cognition (with active, specific cognitive behavior of the patients), it is attempted to raise the sensitivity threshold of cognition to the extent that a habituation may be achieved. As already stated, this represents a learning process. For this learning process one typically uses established, tinnitus-specific cognitive behavioral procedures [310], [313], [315], [317]: this is therefore a therapy involving active behavioral co-operation of the patient. The active cooperation leads to a considerably more effective learning process than a passive TRT procedure [308]. Because of the tinnitus specificity, a large part of the intervention procedures show a typical neurootological content [305].

The goal of carrying out cognitive behavioral procedures is to upregulate the cognitive sensitivity threshold for tinnitus so that the so-called LCCS (limited capacity control system) associated with the auditory system is no longer occupied by the complex tinnitus cognition. Modern sensory-physiological brain research, e.g. by Birbaumer [307], has shown that because of the LCCS the auditory system can only recognize one complex stimulus pattern at a time. This important discovery of the corticosubcortical limited capacity control system (LCCS) is being therapeutically exploited. If the negative, complex tinnitus cognition is replaced by another complex stimulus cognition, the negative tinnitus cognition is displaced from the auditory LCCS (disengagement). In analogy with medication, this can be described as a competitive inhibition of tinnitus cognition [305]. A typical cognitive antagonist is positive imagination [313]. This is a brain-produced tinnitus substituting alternative that is actively processed and trained during the 5-15 hour course of treatment [313], [315].

A second approach is cognitive tinnitus modification. After a cognitive modification the tinnitus is positively occupied and substitutes the negative tinnitus cognition. The therapy requires much time so that the tinnitus displacement process induced by cognitive modification and/or a cognitive antagonist is finally stored in the long-term memory of the auditory system. After having stored positive imagination or tinnitus modification the patient learns to react to his tinnitus by eliciting positive imagination or tinnitus modification (disengagement process). The therapy at onset is active and indeed strenuous for the patient. If the reaction is actively repeated for a longer period of time, the tinnitus displacement is learned to follow automatically, i.e. without any conscious effort (habituation). The incoming tinnitus stimulus is held for some milliseconds in a sensory memory. It then elicits the therapeutically produced stimulus reaction with the result that the reaction, i.e. the learned tinnitus disengagement induced by cognitive tinnitus modification and/or a cognitive antagonist, functions automatically. The tinnitus is therefore ignored, and as a matter of definition a tinnitus habituation, i.e. the opposite of a sensitization, is achieved [305] (sensitization and habituation are recognized as antagonistic neurophysiologic processes [307]). The path that therapy follows from sensitization to habituation is referred to as desensitization. Habituation together with a successful combating of the tinnitus is the therapeutic outcome.

According to the validity criteria for evidence based medicine (EbM, see Table 9 [Tab. 9]), the therapeutic procedures for a neurootologically cognitive TDT are of evidence level IIb: two high-quality, prospective controlled studies have been published [313], [315]. Both Kroener-Herwig et al. as well as Delb et al. applied the therapeutic approaches used in TDT in controlled group studies and undertook a step-by-step modification process to optimize the procedure. The psychosomatic tinnitus stress and the daily perception time were decreased lastingly and quality of life was improved. The Goebel-Hiller/Hallam tinnitus score decreased from 49.5 to 34.6 [315].

For the ENT specialist it is important to become familiar with current fundamental knowledge about the pathophysiological course of an acoustic trauma. Of great importance is the realization that as a result of sound overexposure, mid-term, functionally damaging or cell destroying metabolic cascades in the inner ear are activated that may involve oxidative stress, excitotoxicity and apoptosis. More obtuse forms of damage are rare.

On the basis of this pathophysiological knowledge, the healing doctor can then provide the patient with a correct model for his/her illness. He will also be able to engage himself preventively in advising children, adolescents and adults.

Nevertheless, despite important regulatory measures for noise protection at the workplace it has to be stated that with chemical co-exposure additional risks arise for the inner ear that are not adequately taken into account by current occupational safety measures. In addition, leisure noise stress means that an important precondition for the efficacy of noise protection at the workplace is no longer fulfilled, i.e. a recovery phase in one's leisure time. Here physicians must implore legislators to pass suitable regulatory measures for protecting children and adolescents.

Leisure noise has therefore presented itself overall as an important aspect of public health: 10% of an age-group of adolescents and young adults are already threatened with hearing loss and a loss of speech comprehension in their mid 20s due to the excessively high leisure noise stress they expose themselves to.

Amongst the protective drugs against unavoidable noise magnesium aspartate has proven itself as an effective substance of evidence level Ib for pre- and para-exposure prophylaxis.

For a post-exposure acute therapy, cortisone is available to increase protein biosynthesis as well as for its anti-inflammatory properties. In addition, alpha-liponic acid is available for disrupting the pathological pathway leading to apoptosis by acting as an oxygen radical scavenger. There is also a rational justification for an acute therapy with NMDA receptor blockers, since they can suppress excitotoxicity of the afferent auditory nerve fiber. Here, a therapy is begun at the latest 24 hours after the acoustic trauma and is terminated after a week. A longer therapy might suppress the desired post-exposure recovery in an unwanted way since the function of NMDA receptors is required then.

For chronic hearing loss only hearing aids and electronic implants are available today for treatment. Substances specifically designed for regenerative treatment are not currently available in the clinics. Current knowledge, however, does not rule out that these might become available in the future.

Chronic tinnitus presents a special problem on its own. Here, a secondary centralization can develop as part of a chronic complex tinnitus. Based on state-of-the-art neurophysiological concepts, a neurootologically cognitive desensitization therapy is currently finding favor, since this pursues the goal of renormalizing the reduced threshold of cognition.

For helpful discussions on the individual sections we would like to thank Larry Fechter, Ph.D, Thais Morata, Ph.D., Elmar Oesterreicher, M.D. and the following colleagues of the Department of Otorhinolaryngology: Head and Neck Surgery at the University of Tübingen and the Tübingen Hearing Research Centre (THRC): Jutta Engel, Ph.D., Hartmut Hahn, Ph.D., Marlies Knipper, Ph.D., Susan Kupka, Ph.D., Hubert Loewenheim M.D. and Wolfgang Wagner, M.D.. The epidemiological data on the incidence of hearing loss from fireworks represents the results of data collection from approximately 800 physicians in 31 ENT university clinics, 87 ENT departments of city hospitals, and 444 ENT practices in Germany. A list of the participating centers and physicians can be found under: http://www.medizin.uni-tuebingen.de/hno/gehoerschaeden_feuerwerkskoerper.htm.