IntroductionThe auditory system seems better equipped to deal with injuriesin lower species than in mammals. In fish and amphibians, theinner ear will produce new sensory cells (hair cells) throughouttheir life and, consequently, injured cells can be replaced contin-uously. Birds lose this ability during embryonic development, butpossess the capacity to replace the injured sensory cells by regen-eration and thus maintain hearing function. In contrast, mam-malian hair cell loss has always been considered irreversible.The mechanism of cell death in the cochlea is produced intwo ways: through “necrotic cell death” mediated by very loudnoise, or “apoptosis,” mediated by the activation of cysteine protease family within the cells, the caspases [very loud noise canalso induce immediate apoptosis (1)]. Originally these mecha-nisms, necrosis versus apoptosis, were thought to operate with different initiators (as an extrinsic cellular pathway and an intrin-sic cellular pathway, respectively), but it may be assumed thatthese mechanisms are more or less under statistical control in thatdependent on the characteristics of the stimulus the extent of celldeath and damages is brought about by one of these two majormechanisms. Each of these mechanisms provides the possibility toreduce and, in some cases, to prevent cochlear cell death throughactive intervention with pharmacotherapy.Recently, many researchers have investigated the role ofantioxidant agents in different models of peripheral hearingdisorders. It has been found that antioxidants protect thecochlea from noise-induced trauma, as well as cisplatin andaminoglycoside exposure (2–4). Van De Water et al. recentlysuggested that protection of auditory sensory cells from cisplatinis carried out at the molecular level by three mechanisms: pre-vention of reactive oxygen species (ROS) formation; neutralisa-tion of toxic products, and blockage of apoptotic pathways (5).Several genes regulate the differentiation of cochlear haircells and supporting cells from their common precursor cellsduring mammalian embryogenesis. Recent experiments haveprovided new and exciting information about the processesrelated to inner ear damage. For example, in the mammalianvestibular system, hair cell regeneration has been shown to occurunder certain circumstances (6). The situation in the auditorysystem is less clear. There is evidence of hair cell regeneration innewborn mice given explants of cochlear duct (7) and in replac-ing the damaged hair cells by converting the supporting cells (8).A key gene is Atoh1 (also known as Math1). This is the mousehomologue of the drosophila gene atonal that encodes a basichelix-loop-helix transcription factor (9). Overexpression ofAtoh1 in nonsensory cells of the normal cochlea generates newhair cells, both in vitro and in vivo. Atoh1 has been shown to actas a “prohair cell gene” and is required for the differentiation ofhair cells from multipotent progenitors. Recently, Izumikawaet al. (2005) demonstrated that in mammals by using gene ther-apy, the lost hair cells will regenerate and that hearing may bereturned to the profoundly deaf mammalian ear (10). This find-ing opens new perspectives for the treatment of hearing loss andjustifies the efforts to encapsulate nucleotides encoding the Math1gene within the nanostructures for the treatment of deafness.In addition, a moderate degree of spontaneous recovery ofhearing after noise trauma has been observed in humans, implyingthat humans may also have the capacity to regain hearing function(11). However, the mechanisms behind the recovery have not yetbeen fully delineated. There is, however, substantial evidence thatcochlear damage induced by noise can be prevented by the appli-cation of different pharmacologically active substances (12). Thus,there are grounds to expect that hearing disorders in mammalsmay, under certain circumstances, be successfully treated.Drugs can reach the inner ear by systemic application(orally, intravenously, or via the cerebrospinal fluid) or locally[from the middle ear over the round window membrane(RWM) through permeation, direct injection through theRWM or the oval window, and also with an osmotic pump bypassing through the lateral wall of the cochlea]. However, notall of these approaches are clinically possible.Pharmacotherapyof the inner earIlmari Pyykkö, Esko Toppila, Jing Zou,Erna Kentala161181 Chap16 3/29/07 6:36 PM Page 219
Mechanisms of noise-inducedhearing lossNormal auditory stimulation elicits pressure differences acrossthe cochlear partition causing a number of mechanical eventswithin the organ of Corti: vibrations, shearing motion, anddeflection of the stereocilia (13). The end result is excitation ofthe outer and inner hair cells and, following release of trans-mitter substances, increased activity in the cochlear nerve. Theouter hair cells are activated and react in a linear manner tosinusoidal sound stimulation with one impulse to one sinusoidup to 1000 Hz. At higher frequencies, other mechanisms areinvolved in coding the amplification of the signal. These arenot known in detail.Assisting the tight coupling between the tectorial mem-brane and the basilar membrane, the tips of the stereocilia ofthe outer hair cells are buried within the tegmentum. The con-tractions of the outer hair cell bodies amplify the basilar mem-brane vibration and transduce the vibration to shear forces thatwill activate the inner hair cells. The perceived and activelyenhanced basilar membrane vibration is transmitted into thecentral auditory system and is perceived as sound. The role ofthe supporting cells is not clear yet but they may serve as a sup-porting organ to provide stability and damping of excessivevibration. Damage to the cochlea may also lead to hyperacusisand we hypothesise that this symptom may be linked to sup-porting cell damage (Fig. 16.1).Obviously, noise or excessive auditory stimulation willelicit shear forces in the cochlea but at much larger amplitudes.There are two fundamentally different ways by which overstim-ulation may lead to cochlear injury: mechanical or metabolic(14). Intense noise exceeding 125 dB sound pressure level(SPL) in animal experiments leads to large amplitude vibrationthat may mechanically alter or disrupt cochlear structures caus-ing mechanical damage to cell membranes and nerve endingsand disturb the blood circulation. Cellular distortion, disorgan-isation of the stereocilia, and possible rupture of cell membranesdisable the cochlear fluid barriers and will cause immediatereduction of auditory sensitivity (15).At SPLs of less than 125 dB, sound-induced overstimula-tion and overactivity of the cochlea can result in disturbedcochlear homeostasis and subsequent functional impairment inthe absence of direct and immediate mechanical damage.Experimental evidence suggests a critical level about 125 dBSPL, at which the cause of damage changes from predominantlymetabolic to mechanical (16). Thus, at moderate SPLs, damagewould mainly be caused by metabolic mechanisms while athigher levels, mechanical mechanisms would predominate. Aschanges in homeostasis may also occur in mechanical traumaand the effects of metabolic stress are also likely to be expressedas mechanical damage, it is not meaningful to make a strict sep-aration between metabolic and mechanical causes of noise-induced hearing loss.When the metabolic and/or mechanical stress is too large,the cells will die and a permanent hearing loss results. Celldeath is a result of either apoptosis or necrosis. Apoptosis is astrictly controlled process to eliminate dysfunctional cells with-out affecting the surrounding tissue. It can be viewed as a coun-terbalance to cell division, and a disturbance may, for example,result in degenerative disorders or tumour growth. Necrosis onthe other hand is a more passive type of cell death, involving arapid and disorganised breakdown of a cell, often as a conse-quence of acute trauma (toxic substances, ischaemia, etc.). Asthe cell contents are released directly into the surrounding tis-sue and an inflammatory reaction usually follows. Thus, for theorganism, apoptosis is the preferred method when it is necessaryto eliminate cells. In the auditory system, there is no conclusiveevidence that apoptosis does play a significant role. Structuralobservations of DNA fragmentation may suggest the involve-ment of either apoptotic or necrotic mechanisms during peri-and postnatal development of the inner ear (17). A recentstudy on autopsy materials from subjects with no history ofacoustic trauma suggests that apoptosis does not contribute sig-nificantly to the regulation of the cell population in the normaladult inner ear (18). Nevertheless, apoptosis may be involvedduring noise-induced trauma, although there is to date no directevidence in humans.Changes in cochlear blood flow have generally been sug-gested as contributing to noise-induced hearing loss (19).Recent findings have clearly demonstrated noise-induced alter-ations in the cochlear microcirculation causing local ischaemia(20). The effect varies with the intensity and duration of theexposure, but when vascular insufficiency is manifest, thereduced oxygen and energy supply to the cochlea and theaccumulation of metabolites will be accompanied by severefunctional alterations. It has been shown experimentally thatapplying drugs blocking vasoconstriction prevents a noise-induced microcirculatory disorder and maintains normal hear-ing (21). However, the exact role of local blood flow alterationsis unclear and it should be noted that it has been observed thathearing loss and cochlear hypoxia may actually precede changesin cochlear blood flow (22).220 Current managementFigure 16.1 Schematic drawing of organ of Corti. Abbreviations: BM, basilarmembrane, DC, dieters cells; IHC, inner hair cells; OHC, outer hair cells; PC, pillarcell; ST, stereocilia; TM, tectorial membrane.1181 Chap16 3/29/07 6:36 PM Page 220
There are several mechanisms leading to cellular damageafter acoustical overstimulation (Fig. 16.2). The damage can berepaired or can be irreversible leading to cell death. Some of themechanisms are mainly related to metabolic changes, e.g.,oxidative stress, synaptic hyperactivity, and altered cochlearblood flow, while others are predominantly mechanical. It islikely, however, that the resulting damage to the auditory sys-tem is partly mediated by similar mechanisms irrespective of thecause. Although definite evidence of a common final pathwayis missing, experimental data suggest that free radicals and otherhighly reactive endogenous substances play a significant role innoise-induced hearing loss.The mechanisms causing cell death through necrosis arefundamentally different from those in apoptosis. Table 16.1summarises the differences. The apoptotic mechanism is, in thedevelopmental stage and in some disease stages, such as in can-cer or granulomatous infection (for example in tuberculosis), anormal and vital part of life. With these mechanisms, the bodyshelters from infection, eliminates small tumours, and controlsthe growth of larger tumours.Apoptotic mechanisms andfree radicalsIt is well known, from other biological systems, that reactiveoxygen metabolites (ROMs) are important mediators of cellinjury. ROMs are free radicals or other molecules, which have achemical structure, making them extremely reactive. As theyreact very easily with cellular components such as lipids,proteins, and DNA, they are potentially cytotoxic. ROMs areproduced continuously as part of normally occurring reactions,e.g., in the mitochondria. However, protection is offered byseveral endogenous antioxidants. These are either enzymescatalysing reactions to neutralise the ROM, or scavengers bind-ing them. When there is an imbalance between the productionof ROMs and the endogenous protective mechanisms, the tissueis under oxidative stress. Increased ROM production can causecell death, whereas overactive protective mechanisms may leadto tumour growth. In the auditory system, there are severalreports demonstrating both elevated levels of either ROMs orantioxidants following noise exposure (23), and reduced hearingloss by treatments increasing the antioxidant level (24).The key element in apoptosis is the caspase-induced celldeath pathway. Caspases consist of a family of cysteine proteasesthat are present in the cells in an inactive form. In short, whenthe cell is damaged, a lethal chain reaction occurs that is trig-gered by activation of Bax gene. In the reaction, apaf-1 inter-acts with cytochrome C that is located on the mitochondrialsurface, the complex interacts with procaspase-9 (a complexcalled to apoptosome) that cleaves and results in the caspase-9that finally activates the caspase-3 through cleavage of someother procaspases. The “killer” caspase-3 reacts with the mito-chondrial membrane and causes membrane lysis by liberatingthe lysosome enzymes from the cell leading to degradation ofDNA and the proteins and disintegration of the cell (Fig. 16.3).There are today 14 members of caspase family, but not allmembers of caspase family participate in the apoptosis, ascaspase-1 and -11 function in the regulation of cytokines. Theinitiator includes caspase-9 and -8 and the effector includesPharmacotherapy of the inner ear 221Figure 16.2 Schematic drawing representing the necrotic and apoptotic celldeath mechanisms, as excitotoxicity caused by glutamate.Table 16.1 Comparison between apoptosis and necrosisApoptosis NecrosisMay happen under both Only happens underphysiological and pathological conditionspathological conditionsA gene-directed process Not a gene-directed processAn energy-dependent process Not an energy-dependentprocessProtein synthesis is increased Protein synthesis isdecreasedATP content is normal ATP content is decreasedSingle cell involved Several cells involvedA delayed degeneration An immediate degenerationprocess processCellular shrinkage Cellular oedemaOrganelles are intact Organelles are destroyedChromatin condensation Chromatin destructionLate membrane damage Early membrane damageDoes not cause inflammation Causes inflammationAbbreviation: ATP, adenosine triphosphate.1181 Chap16 3/29/07 6:36 PM Page 221
caspase-3, -6, and -7. The initiation of the caspase reaction canbe regulated by the external cell death receptor pathwaythrough the Fas ligand—receptor activation or through theintrinsic cell death pathway [the mitogen-activated proteinkinase (MAPK)/Jun N-terminal kinases (JNK) pathway]. Boththese pathways trigger responses that lead to final stage activa-tion of caspase-3 that acts as the executioner molecule for thecell. Nevertheless, caspase-3 appears to participate in the nor-mal development and maturation of the membranous labyrinthand its cochleovestibular ganglion, so that a loss of functionmutation of the gene for caspase-3 could result in maldevelop-ment of the inner ear and a hearing deficit.ROS and the caspase-induced celldeath pathwayOxidative stress is a key factor in apoptosis with the creation ofROS and other free radicals (e.g., hydroxyl radical), which acti-vate the apoptotic pathway through cellular mechanisms thatare linked to caspase activation (5). These ROS and other rad-icals damage the affected cell’s organelles and internal mem-branes resulting in mitochondrial membrane damage and a lossof the membrane potential. This loss of mitochondrial mem-brane integrity results in a release of cytochrome C from thedamaged mitochondria into the cytoplasm. Once cytochromeC enters the cytoplasm, it combines with a facilitating moleculetermed apoptotic protease-activating factor-1, dATP (anenergy-supplying molecule), and procaspase-9 to form theapoptosome (also known as the aposome), which cleaves pro-caspase-9 and generates activated caspase-9 (25). A small mito-chondrial “proapoptosis molecule” facilitates apoptosis of anaffected cell by inhibiting some of the damaged cell’s naturallyoccurring caspase-inhibitory molecules (e.g., NIAP-neuronalinhibitor of apoptosis protein). Once procaspase-9 has beenactivated, its downstream targets are effector caspases, e.g., caspase-3, -6, and -7 (26). The naturally occurring cellular apop-tosis-inhibitory proteins are thought to target activated effectorcaspases such as caspases-3 and -7 for deactivation (27).The activated effector caspases can interact with a largenumber of targets within an affected cell to bring about itsdestruction by apoptosis. Some of the cellular molecules tar-geted by the caspases are summarised by van de Water et al. (5)as: poly (ADP-ribose) polymerase (PARP-1); DNA within thenucleus and a DNA repair enzyme; nuclear lamin molecule A,B, and C; DNA fragmentation factor 45; inhibitor of caspase-activated DNase; receptor-interacting protein; DNA topoiso-merase; signal transducer and activator of transcription-1; Rb;X-linked inhibitor of apoptosis; U1 small nucleoprotein; fodrin;vimentin; and procaspase-2, -6, and -10. Caspase-3 has beensuggested as being the primary executioner in most cellularapoptosis during both normal developmental cell death and theremoval of damaged cells after injury (i.e., apoptosis) (25).Hu et al. examined noise trauma–initiated apoptosis ofcochlear outer hair cells in the chinchilla (28). In a double-labelled study, the authors localised the activated caspase-3 tothe cell bodies of damaged hair cells undergoing apoptosis. Theresults show a relationship between post–noise exposure pro-gression of hair cell loss, apoptosis of damaged hair cells, andactivation of caspase-3. The study also demonstrated thatactivation of caspase-3 persists for at least two days after theinitial noise trauma exposure. There was a correlation betweenpost–exposure loss of noise-damaged outer hair cells, apoptoticchanges in the outer hair cell nuclei, and the presence ofactivated caspase-3, -8, and -9 in the cell bodies of damagedsensory cells. The finding also indicates that the treatmentwindow for noise-induced apoptosis of cochlea lasts at leasttwo days.The MAPK/JNK-induced cell deathpathwayThe extrinsic cell pathway involves the binding of cell deathreceptors that are members of tumour necrosis factors (TNF)-␣pathway. In this, there are two receptors in the cell that are bothactivated in the shear stress injury by TNF-␣, the type 1 (p55receptor pathway) and type 2 (p75 receptor pathway). As inmost instances in upregulation of cellular function, one leads tocell death and the other tries to rescue the cell. The TNF-␣type1 receptor pathway is the apoptotic pathway. TNF, through itstwo types of receptors, activates two signalling pathways withincells (29). One, linked to receptor type 1 leads to programmedcell death (apoptosis), whereas the other, linked to receptor 2,counters the death signal and leads to survival. When bothreceptors are expressed, the type 2 receptor of the TNF-␣mayenhance the receptor 1–mediated death pathway. The final con-sequence may depend on the level of type 1 expression. The sur-vival pathway activates a transcription factor, nuclear factor(NF)-B, which works by turning on a set of antiapoptotic genes.222 Current managementFigure 16.3 A diagram showing activation of caspase pathway in the finalapoptotic pathway leading to disintegration of the cell.1181 Chap16 3/29/07 6:36 PM Page 222
NF-B is normally composed of two subunits, p50 and p65.It is usually held captive in the cytoplasm of a cell because itassociates with an inhibitor protein called inhibitory protein-B(IB), which stops NF-B from entering the nucleus. After cellsare treated with TNF-␣, the IB protein becomes labelled withphosphate groups, which marks it out for degradation. With itsjailer destroyed, NF-B is free to move into the nucleus, where itbinds to relevant sites in its target genes and activates a newprogramme of gene expression, ensuring that the cell survives.The IB kinase complex mediates the key phosphorylation of IBin this chain of events.In Mongolian gerbils, it was found that changes in the lev-els of apoptosis-related proteins correlated with decreases incochlear function as measured by distortion product otoacousticemissions (DPOAEs) (30). The apoptosis-related proteins thatcorrelated with a decrease in DPOAEs were (i) an antiapoptoticfast response gene Bcl, bcl-2, which was decreased in the tissuesof the aged cochlea and (ii) activated caspase-3 molecules,which increased in the tissues of the aged cochlea when thesetissues were compared with the same tissue types obtained fromthe cochleae of the young animals. The level of Bax (a proapop-tosis protein caused by cell death gene Bax) did not show anyageing-related increase or decrease. Both bcl-2 and activatedcaspase-3 are involved in the control and execution of theMAPK/JKN-mediated cell death pathway, which is thought tobe the primary mediator of oxidative stress-induced apoptosis ofinner ear sensory cells. Thus the reported higher shear stressvulnerability of older animals may be linked to differences inregulation of the components in the apoptotic pathway (31).During intense sound exposure, the inner hair cells are over-stimulated resulting in synaptic hyperactivity and an excessiverelease of transmitter substance. The afferent neurotransmitter ismost likely to be glutamate, which, like other excitatory aminoacids, has toxic effects when released in large amounts. Theresulting overstimulation of the glutamate receptors elicits aninflow of calcium ions, which, in combination with other ions,brings about the entry of water and subsequent swelling of thenerve endings. The result may be a total disruption of thesynapses between the inner hair cells and the afferent nervefibres in the cochlear nerve (32). A dorsal root acid sensing ionchannel has been detected in the spiral ganglion cells (SGCs)and the organ of Corti including the nerve fibres innervating theorgan of Corti (33). It is known that opening of the acid sensingion channel may flux Ca2⫹and induce cell death (34). Thismechanism may also be involved in noise-induced hearing lossand ischaemia-induced hearing loss because both shear stressand ischaemia can result in a low pH extracellular homeostasis.In addition to the accumulation of ROMs seen followingmetabolic and/or mechanical stress, it has been demonstratedthat acoustical overstimulation leads to a significant rise inintracellular calcium levels in the outer hair cells (35). A sus-tained increase in the intracellular calcium concentration isknown to result in severe cell injuries such us cytoskeletalbreakdown, membrane defects, and DNA damage (36). Oneprobable consequence of the increased calcium concentrationin the outer hair cells is the loss of cell body stiffness observedafter intense acoustical stimulation (37). Moreover, a structuralreorganisation of the organ of Corti has recently been demon-strated following acoustical overstimulation (15). The noise-induced changes in cellular stiffness and structure of thehearing organ seem to be, at least partly, reversible and theresults may thus contribute to knowledge of the mechanismsinvolved (Fig. 16.4).Pharmacotherapy of the inner ear 223Figure 16.4 Life-and-death decision in the cells. Cellular stimulation with TNF-␣(top) simultaneously activates survival (left) and death (right) signallingpathways. The survival pathway leads to the activation of NF-B, which inducesthe expression of antiapoptotic genes in the nucleus. NF-B (subunits p50 andp65) is normally held captive in the cytoplasm by the IkB protein. Cellularstimulation with TNF-␣leads to activation of the IKK complex, whichphosphorylates IkB. The phosphate tag (circled “p”) singles out IkB fordestruction. NF-B is then free to move into the nucleus and activate its targetgenes. Hoeflich et al. have revealed an unexpected requirement for GSK-3b inthe NF-B–mediated activation of genes needed for survival (39). It is not yetclear how GSK-3b works in this pathway, but it probably involves a critical stepfollowing the movement of NF-B to the nucleus. Targeted disruption in miceof any of the molecules coloured red leads to death of the embryo,accompanied by TNF-␣–induced apoptosis of hepatocytes. Abbreviations: GSK,glycogen synthase kinase; IKB, inhibitory kinase-B; IKK, IkB kinase; NF, nuclearfactor; TNF, tumour necrosis factor. Source: From Ref. 38.1181 Chap16 3/29/07 6:36 PM Page 223
Necrosis induced by soundstimulationVery loud sound leads to mechanical damage of the organ ofCorti with fractures of the cellular membrane, liberation oflysosome content, and, exposure of the cell content to extra-cellular fluids. In necrosis, there is abundance of proinflamma-tory cytokines such as interleukin (IL)-1, IL-6, and TNF-␣,and migration of inflammatory cells (40).TNF-␣is involved in cellular survival/damage mechanismespecially through the TNF receptor 1 (p55). The survival path-way can induce activation of c-Jun NH2-terminal kinases andNF-B (41). There is also another TNF-␣–mediated pathwaythat acts through receptor 2 (p75), the necrosis stimulating path-way. This pathway is also self-feeding; as the receptor 2 pathwayis activated, it enhances production of TNF-␣. The upregulationof receptor type 2 leads to inflammation and cytotoxic effects.Upregulation of growth factors innoise traumaThe role of upregulation of growth factors is not yet well known.One of the nerve growth factors (NFG), the NFG1, is expressedin the cochlea during traumatizing noise leading to permanentthreshold shift (PTS) but not during nontraumatic noise leadingto temporary threshold shift (TTS) (41). One of the immediateresponses by NFG1 gene production is activation of c-fos. Loudnoise at a damaging level also upregulates genes producing glialcell line–derived neurotrophic factor (GDNF) in rats. Theupregulation starts after four hours, peaks after 12 hours, and lev-els out after 12 hours from cessation of the noise exposure (42).GDNF has a protective effect on noise-induced cellular damagebut the exact mechanism has not yet been delineated. It hasbeen hypothesised that GDNF is involved in the consolidationof recovery function from noise damage. It is also possible thatGDNF has a function that is related to protection from addi-tional noise-induced stress, rather than recovery from the firststress. GDNF upregulation may be related to the training effect,toughening, or conditioning of the organ of Corti (42).Vascular endothelial growth factor (VEGF) is also upregu-lated during shear stress that leads to traumatic changes (43).The expression of VEGF is limited to the hair bundles andSGCs after traumatic vibration. VEGF receptor 1 (VEGF R1)is not detected in the vibrated cochlea, whereas VEGF R2expression is present in the lower part of the outer hair cells,Dieters’ cells, Hensen’s cells, Claudius cells, the basal mem-brane of the organ of Corti, the internal sulcus cells, nucleus ofthe SGCs, the lateral wall of scala tympani, and the spiral liga-ment (Fig. 16.6). No expression of VEGF R2 was observed inthe stria vascularis.It is well accepted that shear movement exists in the organof Corti, but there is no documentation of shear stress withinthe SGCs. Shear forces within the bone matrix stimulate bonecells and mechanically transform them causing upregulation ofgenes in the cells (44). The SGCs are surrounded by perilymphand bone matrix. The shear force produced by the transcranialvibration is conducted to the SGCs and is able to cause shearstress. In our study, VEGF and VEGF R2 gene expression in theSGCs supports this hypothesis (Fig. 16.5).Vibration induces VEGF and VEGF R2 expression in thecochlea, but not VEGF R1. Our results confirm the biologicalsignificance of a previous in vitro study, which indicated that invascular endothelial cells, high shear stress induced an increasein VEGF R2 expression. This upregulation reached its maxi-mum and was in a linear relationship to the stress strengthwithin a range of 2 to 40 dyne/cm2(45). The authors inter-preted this as showing that an increase in the shear stress in thevasculature by postischaemic reperfusion stimulates VEGF R2expression, resulting in an increase in vascular permeability andleading to neovascularisation. After myocardial infarction, thenewly formed myofibroblasts express VEGF and VEGF R2 thatseem to play a significant role in tissue repair/remodelling (46).When the cochlea is exposed to mechanical vibration, a shearstress presented in the various cell types of the cochlea withconcomitant increase of expression of VEGF and VEGF R2.Thus, VEGF may contribute to tissue remodelling and angio-genesis at the site of damage in an autocrine manner and maybe important in preventing further damage to the cochlea. Themost enhanced expression was located in the SGCs, stereocilia,supporting cells, the internal sulcus cells, and epithelial cells ofthe lateral wall of the scala tympani. No obvious expression wasfound in the hair cells. This means that the hair cells are ratherstable and not affected by the VEGF-induced reaction and seemnot to be able to be remodelled/repaired by VEGF when theyare damaged. Conversely, the spiral ganglion may be repairedwith the assistance of VEGF because both VEGF and VEGF R2are expressed there.Shear stress–induced VEGF expression seems to be time-dependent. After acute shear stress within six hours of the224 Current managementFigure 16.5 Expression of vascular endothelial growth factor in the hairbundles of the outer and inner hair cells after shear stress–induced cochleartrauma in a guinea pig. It is quite possible that this staining covers the tip links.1181 Chap16 3/29/07 6:36 PM Page 224
exposure, it is not expressed (47); whereas after longer shearstress, its expression is increased up to 14 days (48). In thecochlea, we observed expression of VEGF and VEGF R2 one tothree days after vibration. This is in accordance with time limitof previous reports. In pathological states, especially at theacute phase of brain ischaemia and myocardial infarction,VEGF was expressed and could induce oedema, which is dele-terious. This happened within six hours (49). The naturallyoccurring upregulation of VEGF at a later phase (six hourslater) means that VEGF and VEGF R2 responses are protectiveresponses in the individual. There is no evidence to showwhether VEGF and VEGF R2 are expressed in the spiralganglion within six hours of cochlea shear stress. It is worthinvestigating this response to provide reference data for clinicaltreatment.Pharmacotherapy of the inner earFree radical scavengersAt least three important ROS are generated in the reduction ofO2to H2O: superoxide anion (O2–), hydrogen peroxide (H2O2),and hydroxyl radical (OH–). It has been demonstrated that ROSsare involved in noise trauma (50–52), cisplatin ototoxicity(53–55), and aminoglycoside ototoxicity (56). Direct evidence ofROS ototoxicity has been demonstrated using isolated outer haircells and by intraperilymphatic infusion (57). ROS ototoxicity isbelieved to be mediated by deleterious effects at multiple sitesincluding lipid peroxidation, DNA strand breaks, and alterationsin carbohydrate and protein structures.Increased knowledge of the processes leading to cellularinjuries is of fundamental importance in order to develop clin-ical means for protection and repair. Many recent reports on theprotection against noise-induced hearing loss offered by drugssuch as antioxidants and neurotrophins (NTs) are promising.Table 16.2 shows the antioxidants for which this has beenexperimentally demonstrated, and some are currently in use.In addition to these agents, there are several other com-pounds that have been tried and some may be useful, but thereare insufficient data on their efficacy in preventing or healingcochlear injury.There are several different pharmacologically active agentsthat have been tried or are in use to treat sudden acoustictrauma. In general, few experiments have been prospective withrelevant control material. The experiments carried out in mili-tary camps with the use of Mg2⫹are effective and usable, but thelimitation in their use is that Mg2⫹should be administeredbefore exposure to inner ear trauma. The efficacy seems to belimited to preventative action by alleviating the accumulationof excessive Ca2⫹in the cochlea.An iron chelator and free radical scavengers have beenshown to attenuate cochlear damage caused by noise (58). Alsothe antioxidant D-methionine has proved to be useful in pre-venting gentamicin-induced ototoxicity (4). N-acetylcysteine(NAC) is metabolised to cysteine (among other molecules) andmay provide cellular needs for glutathione (GSH) in the pres-ence of ROS. Several recent articles demonstrate that L-NACor related drugs could reduce noise-induced hearing loss (59).ROS-induced damage may occur in vibration-induced hearingloss, and ROS scavengers such as NAC may prevent vibration-induced hearing loss. In an animal model, NAC could notprevent vibration-induced hearing loss, although differentPharmacotherapy of the inner ear 225Table 16.2 Possible reactive oxygen species scavengers1. Glutathione is a nucleophilic scavenger and an electron donor via the SH group of it business residue, cysteine2. N-acetylcysteine, the acetylated variant of the amino acid L-cysteine, is an excellent source of SH groups and is converted in the bodyinto metabolites capable of stimulating glutathione synthesis, promoting detoxification and acting directly as free radical scavengers3. Ascorbic acid and its sodium, potassium, and calcium salts are commonly used as antioxidant food additives. These compounds arewater soluble and thus cannot protect fat from oxidants: For this purpose, the fat soluble esters of ascorbic acid with long-chain fattyacids (ascorbyl palmitate or ascorbyl stearate) can be used as food antioxidants4. Salicylic acid is able to absorb hydroxyl ions and thus impede a main step in the process of membrane lipid peroxidation5. Melatonin, once oxidised, cannot be reduced to its former state because it forms several stable end products upon reacting with freeradicals. Therefore, it has been referred to as a terminal (or suicidal) antioxidant6. Tocopherols are the most abundant and efficient scavengers of hydroperoxyl radicals in biological membranes7. The iron chelator (desferrioxamine) forms a stable complex with ferric iron, decreasing its availability for the production of reactiveoxygen species. Desferrioxamine is a powerful inhibitor of iron-dependent lipid peroxidation and hydroxyl radical formation8. Mannitol is free radical scavenger of the hydroxyl radical to which the aldehyde moiety of mannitol reacts and binds. This forms amannitol radical that undergoes disproportionation or dimerises, and thus becomes less cytotoxic than the former hydroxyl radical,causing less damage to the cellular ultrastructureAbbreviation: SH, sulfhydryl.1181 Chap16 3/29/07 6:36 PM Page 225
226 Current managementTable 16.3 Clinical trial of drug treatment of acute acoustic traumaDrugs Sample number Method Efficacy AuthorsVitamin A Re No Ward and Glorig, (64)Dextran 72 Re,Co Yes Martin and Jakobs, (65)209 Re,Co No Eibach and Borger, (66)Dextran;⫹ pentoxifylline 147 Pr,Ra No Probst et al., (67)50 Pr,Co No Eibach and Borger, (68)Bencyclan 85 Re,Co No Eibach and Borger, (68)Xantinol nicotinate 85 Re,Co No Eibach and Borger, (68)ATP 136 Re,Co No Eibach and Borger, (68)Vitamin A, B, E 85 Re,Co No Eibach and Borger, (68)Methionine 85 Re,Co No Eibach and Borger, (68)Cinnarizine 57 Re,Co No Eibach and Borger, (68)Betahistine 122 Pr,Ra No Pilgramm and Schumann, (69)Magnesium 320 Pr,Ra Yes Joachim et al., (70)4 g, drink 80 No Pilgram et al., (71)10mg/kg, infusion 300 Pr,Co Yes Attias et al., (63)167 mg, drink 100 Pr,Ra No Maurer et al., (72)DiltiazemAbbreviations: Co, control group; Pr, prospective study; Ra, randomised study; Re, retrospective study.Source: From Refs. 63–72.administration approaches have been tested (60). Figure 16.6shows the results of NAC on vibration-induced hearing loss. Infact, NAC appears to have synergistic neurotoxic effects incombination with glutamate, which may be the primary affer-ent neurotransmitter of the cochlea (61,62).Several compounds have been tried in the prevention ortreatment of noise-induced hearing loss in humans (Table 16.3).Few of these experiments have control group or arerandomised and prospective. So far, based on evidence inhumans, only Mg2⫹seems to be effective in prevention of noise-induced hearing loss. The study of Attias was carried out inIsraeli army forces and included controls, indicating thatreplacement of Ca2⫹ions in body led to protection from noisedamage (63).In animal studies, the control of noise dose and environ-mental factors can be minimised. There has been muchresearch conducted in animals with several pharmacologicalcompounds. In general, all seem to work in animal experimentsthat have been tried for prevention, but their clinical valueneeds to be documented. For treatment of sudden deafness,there are several substances suggested for use. These are listedin Table 16.4.Figure 16.6 Mean hearing threshold change after exposure of the guinea pigson traumatizing vibration causing shear stress, which was administered withNAC before vibration. Note: 0—round window delivery; 1—80mg/m2; 2—800 mg/m2; 3—2000 mg/m2; 4—8000 mg/m2; 5—control. Abbreviation: NAC, N-acetylcysteine.1181 Chap16 3/29/07 6:36 PM Page 226
TNF-␣and its antagonistsAfter shear stress of the cochlea, Zou et al. demonstrated a weakTNF receptor 1 staining mainly in Hensen’s cells, Claudiuscells, the internal sulcus cells, and the capillaries of the spiralganglion (Fig. 16.7) (43). Much stronger expression of TNFreceptor 2 was found mainly in the SGCs, Henson’s cells,Claudius cells, the internal sulcus cells, Dieters’ cells, the basalmembrane of the organ of Corti, the spiral ligament, the spiralvascular prominence, with weaker staining in the lower part ofthe out hair cells (43). No TNF receptor expression wasdetected in the normal cochlea.Although TNF-␣, TNF receptor 1, and receptor 2 wereobserved in the vibrated cochlea, the expression of TNF receptor2 was more prominent in the cochlea. The combination of TNF-␣with TNF receptor 2 is capable of activating JNK and NF-B(41). The activation of JNK and NF-B has the function of anantiapoptotic agent (41). On the other hand, the activation ofTNF receptor 1 induces apoptosis (77). When both receptors areexpressed, the activation of TNF receptor 2 enhances the effectsPharmacotherapy of the inner ear 227Table 16.4 Animal studies of drug treatment of acute acoustic traumaDrugs Animal species Efficacy Parameters AuthorsNAC Guinea pig ⫹⫹ ROS, ABR, hair cell loss Ohinata et al., (73)NAC; ⫹ acetyl-salicylic acid Chinchilla ⫹⫹⫹ ABR, cytocochleogram Kopke et al., (59)HES70, ES200, pentoxifylline, Guinea pigs HES ⫹⫹⫹; pO2, CAP, ABR, CM Lamm and Arnold, (21)ginkgo biloba, betahistidine betahistidine ⫹ ;othersAllopurinol Guinea pigs ⫹ ROS, ABR Attanasio et al., (74)Allopurinol, SOD-PEG Rat ⫹ ABR, CAP Seidman et al., (75)Dipyridamol, allopurinol Guinea pigs ⫹ CAP Bergmann, (76)DFO, mannitol, GDNF Guinea pigs ⫹⫹ Cytocochleogram, ABR Yamasoba et al., (58)Abbreviations: ABR, brainstem evoked response; CAP, compound action potential; CM, cochelar microphonics; DFO, desferroxamine; GDNF, glial cell line–derivedneurotrophic factor; HES, hydroxyethyl starch; NAC, N-acetylcysteine; PEG, polyethylene glycol; ROS, reactive oxygen species; SOD, superoxide dismutase. Source: From Refs. 21, 58, 59, 73–76.Figure 16.7 Expression of TNF-␣and receptor upregulation in guinea pig cochlea after induction of powerful shear stress. TNF-␣(A) and receptor 2 © are markedly expressed in the vibrated cochlea. Receptor 1 (B) only showed slight expression in the vibrated cochlea. Abbreviation: TNF, tumour necrosis factor.1181 Chap16 3/29/07 6:36 PM Page 227
of receptor 1 activation. The final fate of the cells should berelated to the expression ratio of both receptors. Shear stressinhibits TNF-␣–induced apoptosis by activating phosphatidyli-nositol 3-kinase and inhibiting caspase-3 (Fig. 16.7) (78).Recently anticytokine therapies have become a commontreatment in diseases of autoimmune origin such as rheumatoidarthritis and Crohn’s disease (79). Treatment with monoclonalantibodies against TNF-␣suppresses inflammation andimproves patient well being (79). TNF-␣is a proinflammatorycytokine released during infection or inflammation, which callsthe immune system to action (38). Anti–TNF-␣antibodyadministration in vivo results in the rapid downregulation of aspectrum of cytokines, cytokine inhibitors, and acute-phaseproteins (79).Etanecerp and infliximab are drugs that potently and selec-tively bind TNF-␣in the cellular microenvironment, therebypreventing TNF-␣from interacting with membrane-boundTNF receptors on target cells. Etanecerp is a recombinantfusion protein of the soluble type 2 TNF receptor on a humanIgG1 backbone, whereas infliximab is a chimeric anti–TNF-␣monoclonal antibody containing a murine TNF-␣–bindingregion and human IgG1 backbone.Both etanecerp and infliximab are reported to have a positiveeffect on hearing loss or hearing fluctuation in Menière’s diseaseand idiopathic sensorineural hearing loss (80). In the animalmodel in which keyhole limpet hemocyanin (KLH) was used toinduce autoimmune hearing loss in guinea pigs, etanecerp couldeffectively alleviate the hearing loss and cochlear damage in theanimal model (81). The findings were confirmed in later study inthe same animal model (82). However, a multicentre study onimmunomediated cochleovestibular disorders by Matteson et al.could not demonstrate that etanecerp was effective in alleviatingvertigo and tinnitus or improving hearing in these patients (83).Zou et al. studied the effect of infliximab on the prevention of hear-ing loss after shear stress–induced cochlear trauma (unpublisheddata). In this trauma, TNF-␣and its receptor 1 and 2 are upregu-lated in the cochlea. Infliximab was administered through differentapproaches in the experiments, intravenously, intraperitoneally,and transtympanically. None of the administration methods couldprevent the animals from developing hearing loss. In a subsequenttrial, four patients with vertigo and bilateral severe sensorineuralhearing loss were followed up for three months and infliximab withazathioprine were administered intravenously according to proto-col used for treatment of severe rheumatoid arthritis. In none ofthe patients was hearing improved or preserved. It is noteworthywas that one of the subjects responded to corticosteroids with animprovement of hearing of 50 dB, but did not show a similarresponsiveness to infliximab.Infliximab may cause severe adverse effects, the main beinghypersensitivity reactions, development of antinuclear antibod-ies, possibly lymphoproliferative disorders, and reactivation oflatent tuberculosis. Also a case has been reported with severeneutropaenia and thrombocytopaenia associated with inflix-imab (84). Infliximab infusions are accompanied by acute reac-tions in approximately 5% of infusions (85).To summarise the findings of TNF-␣, it seems well docu-mented that in a damaged cochlea, there is upregulation ofTNF type 1 and 2 receptors but the efficacy of the blockingagents have not yet been demonstrated, so that neitheretanecerp nor infliximab can be recommended for treatment ofhearing loss in humans.Neuroprotection: calpain, nitric oxide,and glutamate receptorsThe accumulation of free radicals severely damages the innerear and other tissues. Through a complex chain of events, thisdamage can then cause a release and accumulation of glutamateand calpains. Nitric oxide (NO) plays a role in a great range ofimportant functions in the organism, such as vasodilatation,relaxation of muscles, neurotransmission, and neuromediation.NO has been found to cause ototoxicity. Ruan et al. demon-strated that sodium nitroprusside, a NO donor, produced bothouter hair cell and inner hair cell damage when it was appliedat the cochlear round window (86). NO synthase (NOS) hasbeen shown to play an active role in the initiation of degener-ation of the SGCs of the rat cochlea (87). It has been suggestedthat noise-induced hearing loss is partly due to excessive releaseof the excitatory amino acids such as glutamate and conse-quently exciting the postsynaptic receptors leading to swellingof the nerve endings (88). It has also been suggested that theototoxicity of noise trauma and aminoglycosides may resultfrom the same excitatory process at the glutamate receptor (89).NO mediates the effects of excitatory amino acids in the cen-tral nervous system and may play a similar role in the peripheralauditory system, since glutamate is considered to be the afferentneurotransmitter at the inner hair cell synapses. NO plays animportant role in kainic acid–induced ototoxicity (90). A studydemonstrated that 7-nitroindazole, a competitive inhibitor ofneuronal NOS, could attenuate the compound action potential(CAP) threshold shift caused by kainic acid, suggesting thatNO is coupled to a glutamate receptor (91). Indeed, Amaeeet al. suggested that NO might be involved in sensorineuralhearing loss–induced by bacterial meningitis (92).Recently Barkdull et al. used cochlear microperfusion totreat sensorineural hearing loss caused by inflammation in aguinea pig model (93). The microperfusion was effective inthe acute phase that is associated with elevations in cytokines,NO, and cellular infiltrates and the breakdown of theblood–labyrinthine barrier. The chronic phase leads to irre-versible ossification of the labyrinth demanding other kinds oftreatment to facilitate removal of inflammatory cells and theirbyproducts. The benefit of microperfusion may be sustainedwhen combined with local delivery of immunosuppressiveagents to the inner ear.Studies have shown that excessive glutamate may play a rolein the production of tinnitus. They also show that glutamateantagonists can have a protective effect on the inner ear and pos-sibly be a treatment for peripheral tinnitus, which is generated bythe inner ear. Several such drugs are currently under investigation228 Current management1181 Chap16 3/29/07 6:36 PM Page 228
for hearing loss and tinnitus as, for example, memantine, carover-ine and magnesium. Caroverine has been shown to restrict theactivity of glutamate receptors and protect the hearing of guineapigs. Its safety and tolerance have been demonstrated in someclinical studies. In one study, 63% of patients treated with intra-venous caroverine reported a significant improvement in theirtinnitus immediately after intravenous infusion (94). Over 48% ofpatients remained stable after one week. No severe adverse effectswere identified for the majority of patient. However, a few patientsexperienced mild transient side effects. There is, however, con-flicting data that suggest that the placebo effect may have beenresponsible for the reduction in tinnitus. More clinical studiesneed to be conducted to resolve the controversy.Glutamate receptor antagonists have been found to protectthe cochlea from noise trauma and aminoglycoside ototoxicity.Excitotoxicity can be prevented by a non-NMDA (NMDA, N-methyl-D-aspartic acid) receptor antagonist (95). Swelling ofthe dendrites under the inner hair cells induced by the glutamateagonist alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionicacid (AMPA) can be partly prevented by the non-NMDA recep-tor antagonist dinitroquinoxaline-2,3-dione (DNQX). Noise-induced swelling of the dendrites under the inner hair cells hasbeen found to be prevented by either dizocilpine (MK) 801, anNMDA receptor antagonist, or kynurenic acid, a wide glutamatereceptor antagonist for both NMDA and non-NMDA receptors(96). Aminoglycoside-induced hearing disorders could be pre-vented by the NMDA receptor antagonist MK 801 (97). Thissuggests that the glutamate receptor plays an important role innoise and drug-induced hearing loss.Puel et al. observed total disruption of all synapses betweenthe inner hair cells and spiral ganglion neurone dendrites,together with the disappearance of cochlear potentials afterapplying AMPA, a glutamate agonist, to the cochlea (98). Inaddition, recovery of both the normal pattern of inner hair cellinnervation and the physiological responses has been observedwithin five days.Treatment of cochlea trauma with nervegrowth factorsNTs, including nerve growth factor (NGF), brain-derived NGF(BDNF), NT-3, and GDNF, are known to play a role in the sur-vival of injured cochlear neurones both in vitro and in vivo.Schindler et al. found that NGF significantly prevented damage tospiral ganglion neurones from neomycin in vivo (99). BDNF andNT-3 have been shown to protect spiral ganglion neurones fromototoxicity of cisplatin and aminoglycoside both in vitro and invivo (100). BDNF and GDNF have also been found to protect thecochlea from noise-induced damage (101). In addition, Pirvolaet al. found that fibroblast growth factor (FGF) receptor (FGFR)-3 mRNA was present in the organ of Corti following acousticoverstimulation and suggested that FGFR-3 could be involved inprotecting the cochlea from noise-induced damage (102).Recent findings that GDNF, BDNF, NT-3, and transform-ing growth factor-␣can protect the auditory hair cells fromacoustic trauma or aminoglycoside ototoxicity in vivo raise thequestion of whether other neurotrophic factors can also protectthe hair cells in vivo (103–106). FGF-2 can protect hair cellsfrom neomycin ototoxicity in vitro, and an in vivo study hasshown upregulation of FGFR-3 in the cochlea following noiseexposure, suggesting that some FGF family members might playa role in protection or repair of the cochlea from damage (107).However, no significant difference in threshold shifts wasobserved between the treated and untreated ears in any of thegroups (108). The extent of hair cell damage was also compara-ble among the different treatment groups. These findings indi-cate that exogenous FGF-1 or FGF-2 does not influencenoise-induced hair cell damage under the experimental condi-tions used in this study, suggesting that these FGFs are not goodcandidates as auditory hair cell protectors in vivo.Zou et al. demonstrated, in the guinea pig, that afterdamage under shear stress in cochlea, the hearing loss could bealleviated by combining BDNF and Connective Tissue NitrientFormula (CTNF) (60). Because BDNF ⫹ Ciliary Neuro TrophicFactor (CNTF) can improve the survival of SGCs while afford-ing no protection to hair cells from noise, protection from hear-ing loss with BDNF ⫹ CNTF suggests that ganglion cell damagemay be important in vibration-induced hearing loss (109).Apparently we need to protect the hair cells, ganglion cells, andpossibly other structures such as supporting cells and strial cellsfrom vibration-induced hearing loss. Because of severe sideeffects from BDNF ⫹ CNTF, it is still too risky to give NTs sys-temically but, in the future, local application may be useful inpreventing inner ear trauma.The development, within the mammalian cochlea, of neu-rite sprouting and integrity of SGCs is influenced by membersof several growth factor families. Among these NGF, BDNF,NT-3, and NT-⅘ are important (110). NGF, BDNF, NT-3, andNT-⅘ can promote the survival of postnatal mammalianSGCs in culture (111). Delivery of exogenous BDNF, NT-3,and NGF to the mammalian inner ear can prevent loss of SGCsfollowing administration of ototoxic drugs (100). NTs havebeen associated with regenerating neurones in avian cochleae(112). CNTF and leukaemia-inhibitory factor are members ofthe neuropoietic cytokine family and can also promote the sur-vival of SGCs (113). These cytokines and NTs act in concertupon mammalian SGCs. For example, the combination ofCNTF and NT-3 is more effective in promoting the survival ofdissociated SGCs in vitro than either factor alone (113).In recent years, studies on antioxidants and/or NTs showpromise in protecting the inner ear (hair cells and SGCs) fromtrauma such as noise (58). The next question, of clinical rele-vance, is to assess such pharmacological treatment in the pre-vention of surgically induced trauma to the human inner ear. Ina previous experimental study, we showed that vibration resultedin significant hearing loss (31). Therefore, experimental studiesmimicking the clinical situation are required to provide infor-mation on the mechanism underlying this kind of damage.Based on the evidence of animal experiments on the neu-rotrophic factors, it is possible in the future, when targeted drugPharmacotherapy of the inner ear 2291181 Chap16 3/29/07 6:36 PM Page 229
therapy will become feasible, that the NTs may be the key mol-ecules used in hearing preservation.