gms | German Medical Science

GMS Current Topics in Otorhinolaryngology - Head and Neck Surgery

German Society of Oto-Rhino-Laryngology, Head and Neck Surgery (DGHNOKHC)

ISSN 1865-1011

Reconstructive methods in hearing disorders - surgical methods

Review Article

Search Medline for

GMS Curr Top Otorhinolaryngol Head Neck Surg 2005;4:Doc02

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/journals/cto/2005-4/cto000008.shtml

Published: September 28, 2005

© 2005 Zahnert.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Abstract

Restoration of hearing is associated in many cases with resocialisation of those affected and therefore occupies an important place in a society where communication is becoming ever faster. Not all problems can be solved surgically. Even 50 years after the introduction of tympanoplasty, the hearing results are unsatisfactory and often do not reach the threshold for social hearing. The cause of this can in most cases be regarded as incomplete restoration of the mucosal function of the middle ear and tube, which leads to ventilation disorders of the ear and does not allow real vibration of the reconstructed middle ear. However, a few are also caused by the biomechanics of the reconstructed ossicular chain. There has been progress in reconstructive middle ear surgery, which applies particularly to the development of implants. Implants made of titanium, which are distinguished by outstanding biocompatibility, delicate design and by biomechanical possibilities in the reconstruction of chain function, can be regarded as a new generation. Metal implants for the first time allow a controlled close fit with the remainder of the chain and integration of micromechanical functions in the implant. Moreover, there has also been progress in microsurgery itself. This applies particularly to the operative procedures for auditory canal atresia, the restoration of the tympanic membrane and the coupling of implants. This paper gives a summary of the current state of reconstructive microsurgery paying attention to the acousto-mechanical rules.

Keywords: middle ear surgery, middle ear reconstruction, tympanoplasty, stapessurgery, auditory canal reconstruction, middle ear implant, auditory canal atresia


1. Restoration of the acoustic function of the auditory canal

1.1. Acoustic function of the auditory canal

Before sound waves reach the tympanic membrane, they are modified through reflection and damping by the outer shape of the body, the auricle and the auditory canal. The auricle and the auditory canal form a funnel and act like a covered pipe stopped at one end by the tympanic membrane. Among the factors listed, the auditory canal and the auricle have the greatest influence on the resonance of the outer ear [1]. The amplification of acoustic pressure by the auricle depends on its shape and on the angle of incidence of the sound so that it possesses an important function in determining the sound's direction.

Helmholz investigated the resonance space of the outer ear using glass and metal tubes or spheres (Helmholz resonators), which could be applied to the external auditory meatus and he calculated a resonance frequency between 2-3 kHz [2]. As the speed of sound can be taken as constant, the geometry (radius, length, volume) of the auditory canal is a crucial quantity in the resonance frequency (see formula 1). Thus, the surgeon influences the resonance frequency when he changes the geometry of the auditory canal.

Equation 1 [3]

New measurements using probe microphones yielded resonance frequencies between 2800 Hz and 3100 Hz in healthy subjects [4], [5]. The acoustic pressure amplification is about 20 dB in the frequency range between 2 and 4 kHz.

1.2. The change in sound conduction with radical cavities

After classical radical cavity operations, the acoustic relationships of the auditory canal and middle ear change. Removal of the posterior wall of the auditory canal leads to a fall in resonance frequency through volume expansion. In principle:

- the bigger the entrance of the external auditory meatus, the higher the resonance frequency,

- the greater the volume of the cavity, the lower the resonance frequency [6].

Measurements in subjects with radical cavities showed a drop in the resonance frequency compared to healthy subjects from 2942 Hz to 1939 Hz, which can be attributed particularly to the expansion of the volume from 0.9 to 2.3 ml. In contrast, the width of the external auditory meatus and the nature of the wall lining the cavity (full of corners or smooth) had only slight influence. A loss of 10 dB in acoustic pressure amplification in the frequency range between 3 and 4 kHz is associated with the volume expansion. Hartwein attributes a high degree of importance to it for understanding speech and hearing music. The formant area from 2800-3200 Hz is an essential vehicle for the entire speech spectrum [6].

In radical cavities a further acoustic effect arises because of the reduced air volume behind the tympanic membrane. While the mastoid and middle ear have a volume of about 6 cm³ in the healthy ear, the volume behind the tympanic membrane is reduced to 1 cm³ in the case of radical cavities [7]. If the volume is reduced further with a flat tympanum (0.5 cm³), acoustic conduction disorders of 10 dB and more arise in the low frequencies according to calculations and experiments [8].

1.3. Surgical methods for restoring the auditory canal

1.3.1. Techniques of auditory canal reconstruction

Radical cavities can be avoided if the posterior auditory canal wall is removed only temporarily at operation and is finally reimplanted again (see Figure 1 [Fig. 1]) [9]. A precondition for the use of this technique is good pneumatisation of the mastoid in order to "drill free" the wall of the auditory canal before removing it along with an intact lateral attic wall as far as possible. Unfortunately these conditions are met only rarely in the case of large cholesteatomas so that the surgically simpler creation of a classical radical cavity has to be relied on. Today these are usually so reduced in size that their residual volume corresponds to a wide auditory canal. However, before the mastoid can be obliterated, all the mucosal cells must be removed by drilling [10]. Chronic infections and delayed wound healing can be avoided in this way. By widening the external auditory meatus, the ventilation of the operation cavity usually turns out so well that chronic infections of the cavity skin are an exception. The acoustic consequences of radical cavity reduction are an improvement in the acoustic pressure amplification by about 10 dB and an increase in the resonance frequency to the normal range [6], [11], [12].

The surgical techniques of radical cavity reduction are varied and are determined above all by the available material and the experience of the surgeon. Autologous materials should in principle be preferred to alloplastic materials in order to avoid foreign body reactions and the breakdown of the material by inflammation [13]. The autologous materials available include bone pate removed during drilling, cartilage from the concha or tragus and also muscle and perichondrium [14], [15], [16], [17], [18]. A combination of these materials must often be employed, particularly when large cavities are to be made smaller. If bone for the reconstruction is lacking, muscle flaps can nearly always be rotated into the cavity; these offer a well vascularised bed for the build-up of cartilage and enable a "dry ear" to be achieved 4-6 weeks postoperatively. The principles of radical cavity reduction are shown in Figure 2 [Fig. 2].

1.3.2. Creation of a neoauditory canal in auditory canal atresia

Atresia of the auditory canal is a congenital malformation of the ear, which is often associated with auricle deformities, deformities of the middle and occasionally of the inner ear. The incidence of auditory canal atresia is reported to be 1/10000-1/20000 [19].

In 1953 House described the construction of the auditory canal and middle ear in atresia as one of the most difficult operations in otology [20]. The causes of the usually high degree of difficulty are reported as [21]:

- the absence of landmarks during drilling through the compact bone,

- the altered anatomy of the middle ear and the facial nerve,

- the limited space for middle ear reconstruction,

- the problems in the healing process.

The best time for an operation in the case of bilateral atresia is said to be at the age of 5 -7 years [19], [22]. In the case of unilateral auditory canal atresia, in contrast, the indication for operation in childhood is controversial. While some surgeons justify an operation even with unilateral atresia in order to utilise the advantages of binaural hearing (directional hearing, understanding speech against background noise) for the development of hearing and speech in childhood [23], others argue in favour of waiting until adulthood. The affected patients and not the parents should weigh up the risks and benefits of the operation [24], [25], especially as the chances of success are rather moderate with regard to the hearing outcome. While postoperative acoustic conduction components of 25 dB can be expected in 70% of cases in the first hearing measurements [26], [27], only 50-60% of the patients achieve social hearing (30dB or better) after 2-5 years [24], [26], [28].

The causes of persistent sound conduction components are reported to be above all the narrower space in the tympanic cavity compared to normal patients and the area of the tympanic membrane which is often too small [28]. It is assumed that the mobilised ossicular chain is fixed again due to new bone formation in a confined space.

There is also a tendency to lateralisation of the tympanic membrane, which occurs with an incidence of 9-62% depending on the surgical technique [29], [30].

A further problem is restenosis of the auditory canal, which can be expected in about 30% of cases [24], [28], [31], [32]. Both the restenosis and the lateralisation tendency could be reduced significantly if split skin grafts were used instead of full-thickness skin grafts [19], [29], [33]. Critics of this technique are of the opinion that the relatively poorly perfused smooth bone is not an ideal basis for the split skin so rejection occurs in a few cases [34]. A few authors therefore try to use pedicled skin flaps, chondrodermal flaps or a flap technique using a combination of the two in the newly created auditory canal [35], [36], [37], [38].

A new technique for avoiding stenosis is reconstruction of the skin of the auditory canal on a bed of cartilage. To do this, a silicone cylinder is coated with cartilage and stable healing is achieved in a skin pocket. After 6 months, this prefabricated cartilage canal can be inserted into the newly created bony auditory canal and lined with split skin (see Figure 3 [Fig. 3]) [39], [40].

The intraoperative risks include noise trauma and facial nerve paresis. The risk of facial paresis has been reduced in recent years particularly by the use of nerve monitoring and improved preoperative diagnosis by means of high resolution computer tomography of the temporal bone (HR-CT) [21]. Transient paresis is reported in 1-1.5% of cases [28], [35], [41]. The risk of an artificial drop in bone conduction between 4 and 8 kHz is reported in about 15% of operations for atresia [32], [42] and the risk of deafness is similar to that in stapes operations, between 0.6 and 3% [28]. It can be reduced by drilling at a low speed or using a CO2 laser for resection of the atresia plate.

Preoperative data from the HR-CT are the most important parameters for estimating the acoustic success of operation. Apart from malformations of the cochlea, the prospects are also poor in the absence of the ossicles, a missing round or oval window or absent middle ear ventilation. In contrast, fixation of the foot plate can be expected only in 6% of cases [24].

Different scores have been developed with the aim of evaluating the chances of hearing success. Altmann [43] distinguishes mild, moderate and severe malformations of the ear. Gill [44] and De la Cruz [29] extended the classification developed by Altmann to include pneumatisation of the mastoid and middle ear and/or the course of the facial nerve. Jahrsdörfer developed a point system for assessing the chances of hearing success [27]. Using CT data, a total of 10 points were given, where sound conduction components < 25 dB can be expected in 80-90% of cases with over 8 points. With fewer than 5 points, the indication for operation should be considered critically. Siegert refined the system further HR-CT data [45].

There are fundamentally 2 different operation techniques available, the transmastoid and the anterior approach. In the transmastoid approach, the middle ear is sought through a mastoidectomy. The advantage is better orientation and identification of landmarks such as the lateral semicircular canal. With this approach, the atrresia plate is approached from behind during drilling and the ossicles can be identified in good time. Advocates of this technique emphasise the lower risk of inner ear damage through semicircular canal injury or drill noise after accidentally touching the ossicles [29].

Jahrsdoerfer, Mattox and Fisch described a so-called anterior approach in 1986, in which drilling is performed through the atresia plate, if possible without opening the mastoid cells [46], [47]. The only landmarks are the roof of the middle cranial fossa and the temporomandibular joint. Although this procedure must be regarded as clearly more hazardous with regard to noise injury of the inner ear, time-consuming obliteration of the often extensively pneumatised mastoid cavity is not required. The risk of facial paresis is low nevertheless with this approach as the nerve usually runs medial to the atresia plate [46]. Another advantage is the clearly shortened postoperative wound healing, because less granulation tissue forms than in a large operation cavity. In the opinion of Chadrasekhar the anterior approach should be favoured today because the restenosis rate is markedly lower [24]. However, if uncertainties arise during drilling with regard to the anatomy, one should not hesitate to open the mastoid until the anatomical orientation is restored in order to avoid inner ear trauma or facial nerve injuries.


2. Reconstruction of the tympanic membrane

2.1. Acoustic function of the healthy tympanic membrane

When there is an intact tympanic membrane and sound stimulus in the auditory canal, the airborne sound in the auditory canal is converted into mechanical vibrations of the tympanic membrane. The optimum sound absorption is between 2 and 5 kHz. Below 1 kHz only 1/100 of the energy is absorbed by the tympanic membrane and the rest is reflected. Between 1 and 10 kHz more than 1/10 of the energy is absorbed with an optimum at 3 and 7 kHz [48].

The intact tympanic membrane thus greatly damps the sound entering the tympanic cavity from the auditory canal depending on the frequency. Austin gives a value of - 17 dB (SPL) for the proportion of the sound then radiated by the tympanic membrane into the tympanic cavity compared to the sound level in the auditory canal [49]. Experimental investigations show frequency-dependent levels between -10 and -20 dB (SPL) for the secondary airborne sound remaining behind the tympanic membrane [48], [50], [51].

2.2. Techniques of tympanic membrane reconstruction

The tympanic membrane occupies a key position in reconstructive surgery as all ossicular reconstructions are unsuccessful without a membrane that can oscillate. While small defects can often be closed without loss of function, reconstruction of the entire tympanic membrane, particular in the presence of ventilation disorders, is a challenge from the acoustic aspect.

The purpose of tympanic membrane reconstruction is restoration of the anatomical and functional characteristics of this membrane. 3 aims are pursued:

- closing the defect securely

- creating adequate stability with regard to ventilation abnormalities

- producing acoustic characteristics that are similar to the healthy tympanic membrane

The acousto-mechanical characteristics of the reconstructed tympanic membrane are determined critically by the choice of graft material and the reconstruction technique. In principle, the more rigid the material, the more stable the reconstruction with regard to atmospheric pressure fluctuations and ventilation abnormalities (retractions) but the poorer the sound transmission (see Figure 4 [Fig. 4]). Depending on the individual pathology, the surgeon should be able to vary his reconstruction technique in order to find the optimum compromise between these two requirements.

2.2.1. Choice of graft material

Autologous temporalis fascia, perichondrium and cartilage from the concha or tragus are used most often today for tympanic membrane reconstruction [52], [53], [54], [55]. The materials can be obtained easily during the operation and do not represent any infection risk compared to preserved grafts. The choice of the material is guided by the pathological changes in the tympanic membrane and in the tympanic cavity and by the ventilation circumstances.

2.2.2. Reconstruction techniques of the tympanic membranes when mucosa and ventilation are healthy

When the mucosa and ventilation are normal and there are small to medium uninflamed tympanic membrane defects, temporalis fascia and perichondrium are the preferred materials. Better chances of healing and somewhat greater stability (lower tendency to shrinkage) are ascribed to perichondrium [56], [57].

The classical techniques of tympanic membrane reconstruction with membranous materials are the overlay and underlay techniques. The less risky and technically easier underlay technique is preferred today, where the fascial or perichondrial graft is attached to the medial surface of the tympanic membrane, that is, under the defect (see Figure 5 [Fig. 5]) [56]. Usually the graft lies on the bony limb of the posterior wall of the auditory canal and the floor of the auditory canal in the region of the tympanomeatal flap. Anteriorly the graft attaches to the tympanic membrane through adhesion forces if the remaining margin of the tympanic membrane is wide enough. Adequate moisture in the first days of wound healing and placing silicone sheets on the lateral side of the tympanic membrane are required for this adhesion. The moisture comes from the mucosa of the now closed tympanic cavity and by placing antibiotic-containing Gelfoam® sponges on the tympanic membrane. The operation can be performed through both an endaural or retroauricular approach. The retroauricular approach offers the advantage of better view, particularly when the defects are located far forward and leaves more space for working when the extent of the pathological changes cannot be estimated with certainty before the operation. For small defects in the posterior tympanic membrane region, the endaural approach offers the advantage of a comparatively small skin incision.

Particularly in the region of the anterior quadrants of the tympanic membrane, the underlay technique does not always lead to reliable adhesion when the residual tympanic membrane is narrow. Various suggestions have been made for solving this problem. These include supporting the graft with Gelfoam® sponges, quilting the graft to the residual tympanic membrane or the use of fibrin glue. When Gelfoam® sponges are used, adhesions between the tympanic membrane and the medial wall of the tympanic cavity are feared [58]. Use of fibrin glue is costly and might be employed only in selected cases. Our own experience is with the pointwise pull-through technique under the anterior tympanic membrane limbus (see Figure 6 [Fig. 6]). With only pointwise mobilisation of the limbus secure fixation of the anterior edge of the graft can be achieved without the risk of blunting. However, wide division of the limbus in the region of the anterior tympanic membrane should be avoided as far as possible.

For defects of the anterior tympanic membrane, the so-called over-underlay technique is recommended, in which the graft is laid circularly on all sides under the tympanic membrane in the case of large defects but is not pulled through under the handle of the malleus but is placed on it laterally (see Figure 6 [Fig. 6]) [59]. This fixation is said to offer greater stability of the reconstruction, improve the success rate with defects of the anterior tympanic membrane and has proven effective particularly when the position of the malleus handle is steep. Adhesions between the graft and the promontory with resulting sound conduction components can thus be avoided. However, the success of this technique is linked to careful removal of epithelial remnants on the handle of the malleus before covering it with the graft in order to prevent the development of a cholesteatoma in this region. Lateralisation of the tympanic membrane, that is, division of the graft from the handle of the malleus, can be avoided by splinting the tympanic membrane with a silicone sheet and Gelfoam® sponges [60].

Isolated alternatives were compared to the classical techniques, which may be appropriate particularly in the case of smaller tympanic membrane defects. These include gelfilm sandwich tympanoplasty [61], swinging door tympanoplasty [62], microclip technique [63], the fat plug technique [64] and even adhesive techniques [65], [66]. However, there have hitherto been only a few studies with low case numbers on the use of these techniques. Closure with a subcutaneous fat plug may, however, be a promising concept in the case of very small recurrent defects in order to avoid a further more major operation.

2.2.3. Reconstruction of the tympanic membrane in the case of ventilation abnormalities, pathological alterations of the mucosa and total defects of the tympanic membrane

The healing of cutaneous grafts is linked to good tympanic membrane vascularisation and almost normal ventilation conditions. With atelectasis or retraction pockets, the mechanical stability of these materials is inadequate. In this situation, cartilage has proved to be a reliable material. The success rate of ear drum repair is between 98 and 100% [67], [68], [69], [70], which is why this method is also preferred for recurrent defects.

The question of whether the greater rigidity of the cartilage material has a deleterious effect on acoustic transmission receives various responses. While no significant influence of the graft material on the hearing result can be found on pure tone audiometry after myringoplasty [71], experimental measurements show a change in tympanic membrane vibration behaviour because of the cartilage implantation. After tympanic membrane underlay with cartilage there was a rise in the auditory canal sound pressure level of up to 7 dB at 3 kHz as a sign of the increased tympanic membrane impedance [72]. Investigations in an auditory canal-tympanic membrane model yielded transmission characteristics for thin cartilage plates with a thickness of 200-300µm similar to those in the healthy tympanic membrane. With regard to atmospheric pressure fluctuations, however, the thinned cartilage was not equally stable. A thickness of 500 µm was found as a compromise in order to ensure good sound transmission characteristics with adequate mechanical stability [73].

Cartilage can be used in the reconstruction of the tympanic membranes on its own or as a composite graft, still attached to the perichondrium. Different operation techniques have been published depending on the form of the cartilage graft. The following can be distinguished fundamentally (see Figure 7 [Fig. 7]):

- palisade technique,

- cartilage plate techniques,

- cartilage island techniques,

- cartilage tension ring (horseshoe).

The palisade technique can be regarded as the oldest cartilage reconstruction technique. According to Heermann tragus cartilage is cut into 6 strips and laid as a bridge over the tympanic cavity. The cartilage palisades are covered over with perichondrium or fascia. A disadvantage of the palisade technique is the large thickness of the cartilage strips, which can lead to a reduction in the volume of the tympanic cavity, to adhesions with the promontory or to blunting. The acoustic characteristics of palisade reconstruction are often better than the visual impression would suggest, as shown by clinical and experimental results. Vibration maxima occur between the cartilage strips, leading to reduced transmission characteristics which are still similar to those of the tympanic membrane [53]. How greatly the vibration is damped also depends on the number and thickness of the cartilage strips. To improve the acoustic characteristics, a reduction in the number of cartilage strips from 6 to 3 and thinning them to 0.5 mm was therefore suggested [74], [75].

The cartilage plate technique differs from the palisade technique in that the continuity of the cartilage is preserved. The cartilage plates are usually placed on the bone frame and so result in a stable reconstruction. In general, by placing the plates on bone, stiffening of the tympanic membrane reconstruction and thus greater acoustic impedance compared to cartilage palisades or the cartilage island technique have to be accepted. The acousto-mechanical characteristics of the cartilage plate are determined by its thickness, while the source of harvesting (tragus or concha) is not important [53]. In atelectasis of the tympanic membrane, cartilage thicknesses of 500 µm or more are recommended in order to achieve adequate stability. However, if ventilation is normal, thin film-like cartilage plates can be used, which lead to a natural tympanic membrane form and wide conditions in the tympanic cavity (see Figure 8 [Fig. 8]) [76].

In the cartilage island technique, the cartilage does not lie on the bony frame but lies over a cartilage-free margin. As a result, the sound transmission characteristics in the low frequencies improve in comparison with the plate technique. In the individual case, the relationship between the area and margin of the cartilage and the thickness of the cartilage in particular are important. In general it can be assumed that the smaller and thinner the cartilage island, the less the sound transmission of the reconstructed tympanic membrane is diminished.

A typical example is cartilage islands, which are pushed over the prosthetic plate to protect the tympanic membrane. Other cartilage islands are inserted as a composite graft together with the perichondrium and so can reconstruct the entire tympanic membrane [68], [77], [78], [79], [80].

Reconstruction of the entire tympanic membrane can be difficult particularly in the case of subtotal or total defects and a steep malleus handle. An special cartilage technique was proposed for these cases by Borkowski [81]. In order to achieve secure anchorage of the perichondrium, the membrane is set in a horseshoe-shaped cartilage tension ring, which presses against the bone beneath the annulus fibrosus. The cartilage recess serves to accept the handle of the malleus and enables the reconstruction to be tensed and relaxed during the implantation. However, the clinical success of this technique has so far been confirmed in only a few cases.


3. Reconstruction of the ossicular chain

3.1. Acoustic function of the chain

Sound is converted at the tympanic membrane into mechanical vibrations of the ossicular chain, then amplified and it produces a volume displacement of the cochlear fluid through movement of the footplate. The concept of sound pressure transformation used by Wullstein is known in English as ossicular coupling [82], [83].

Because of the different lever arms of the forces at the umbo and lenticular process, there is an amplification of force at the head of the stapes. The following assumptions are made when the lever action of the ossicles is considered:

1. The malleus and incus move as a rigid body (the incudo-malleolar joint is rigid.)

2. The rotation movement of the malleus and incus is around a fixed axis situated in the incudo-malleolar joint.

Leverage of 1.27 to 1.3 was calculated for the two boundary conditions [84], [85], [86], [87]. More recent experimental investigations on the vibration patterns of the ossicles showed that above 1 kHz the position and alignment of the rotation axis changes with the frequency and even within a phase cycle so that the result is not a hinged or piston-like movement of the ossicles but rather a complex movement pattern [88], [89], [90]. Tilting movements of the malleus-incus complex occur, which are transmitted as far as the footplate [91]. It can therefore no longer be said that there is a fixed lever ratio at the higher frequencies. The former assumption that the joints are functionally fixed during sound transmission and "give way" only at very high sound pressures must be corrected. Both in the incudo-malleolar joint and in the incudostapedial joint, sliding movements occur at frequencies of 1 kHz even at physiological sound levels [90], [91], [92], [93].

3.2. Acoustic function of the ossicular ligaments

Little is known so far about the acoustic function of the ligaments. Overall the impedance of the ligaments is regarded as low compared to annular ligament impedance [94], [95], [96]. Huber investigated the effect of division of the posterior incus ligament, as occurs in extended posterior tympanotomy, on the deflection of the footplate by means of laser Doppler interferometry intraoperatively and found no significant changes [97]. Other experimental investigations with separation and tensile strain of all the ligaments were able to show that the ligaments from the acoustic aspect exert less a storage function than a damping function on the transmission of the middle ear. This is expressed not only at the low frequencies but above all in the speech region. As in the case of a microphone membrane, resonances are unwanted in middle ear transmission [98].

3.3. Choice of transplant

From the biological aspect, autologous material cannot be surpassed, even if its long-term stability in the inflamed ear must today be called in question. For reconstruction of the ossicular chain autologous cartilage or bone transplants are suitable. Cartilage is a comparatively soft material which can help to damp vibrations in the chain reconstruction. In addition it has a tendency to shrinkage and resorption in long-term investigations, which is why cartilage can only be recommended for columella reconstructions in exceptional cases (e.g. in the case of a subluxed stapes).

Use of the autologous incus or head of malleus was hitherto accepted as the gold standard for reconstruction in the noninflamed middle ear [99]. The ossicular remnants can be carved to size as a columella and represent a light and hard natural material. In 75 to 100% of cases these interpositions are replaced by new bone after coupling to the ossicular remnants [100]. The revascularisation takes place by the penetration of connective tissue into the Haversian canals. If the tissue changes into inflammatory granulation tissue, the dividing bone is broken down [101], [102]. Depending on the frequency of inflammation there is thus a long-term danger of partial or complete resorption of the bone, which leads to instability. Since implants of cortical bone have much more open pores than ossicular bone, their risk of resorption is higher [56].

If no ossicular remnants are available, homologous preserved cadaver ossicles have been used since the 1970s. In recent years this has fallen almost completely into disuse because the risk of infection with HIV and Creutzfeld Jakob disease cannot be completely excluded [103]. This also applies to implants of dentine although in this case the risks of infection are extremely low because of the possibility of sterilisation.

3.3.1. Alloplastic implant materials

The implant materials available today differ in their mechanical characteristics and in their biocompatibility and biostability. Only the most commonly used materials will be listed here. For a detailed description, the review by Dost should be consulted [103].

Ceramics: Among the ceramics, the extremely hard aluminium oxides are the gold standard among the bioinert materials [103], [104]. The good tolerability proven in animal studies has also been confirmed in clinical long-term investigations, with a rejection rate of 0- 3% [105], [106].

Glass ceramics (e.g. Ceravital®) can stimulate bone growth as bioactive materials. The tendency to resorption in inflammation is a disadvantage, as long-term investigations demonstrate [107], [108]. The products Macor® and Bioverit® are said to have a lower tendency to resorption but tend to form adhesions with the bony walls of the tympanic cavity or other ossicles [103].

Hydroxyapatite resembles the ground substance of bone and the teeth, is biocompatible and is preferred as an implant material for prostheses especially in the USA [109]. It heals well into the adjacent bone but can be broken down in the inflamed ear in 4% of cases [110]. According to Costatino 1991 it has no osteogenetic characteristics but does have osteoconductive characteristics, i.e. it acts as a scaffolding for ingrowing bone [111].

Plastics: Among the plastics polyethylene and teflon in particular are widely used in middle ear surgery. The products Plastipore® and Polycel® are well-known. The extrusion rates are between 2 and 38% and thus higher than with the ceramics [103]. In more recent studies giant cells with signs of foreign body reactions were found in animal experiments and in explanted prostheses [112], [113].

Teflon is familiar especially from stapes surgery. After explantation of stapes prostheses 15 years postoperatively only a delicate wrapping layer of connective tissue was found but no foreign body reactions [114].

Metals: Among the metals the use of gold, titanium and platinum is most widespread. New technologies in material manufacture including laser welding methods have led to new possibilities in implant manufacture and the development of delicate light prostheses adapted to the shape of the natural middle ear.

Platinum is a precious metal with extreme resistance to oxidation and corrosion and thus good biocompatibility. It has proven suitable particularly in stapes surgery for fixing the piston to the long process of the incus. Necrosis of the process of the incus occurred in only 0.7% of cases and it is disputed whether this involved foreign body reactions, scar contraction or perfusion disorders due to the platinum loop [115].

Gold was first used as a stapes piston in middle ear surgery. It is similar to aluminium oxide ceramic in biocompatibility [103] and is also believed to inhibit bacterial growth [116]. Steinbach used PORP and TORP implants for reconstruction of the ossicular chain from gold. It has also been used for the fabrication of tympanic tubes and tube wires. When used as a piston, isolated falls in bone conduction were observed [117]. Dost suspects a toxic effect of the material on the opened inner ear. In other studies these reactions were not observed but an even greater increase in the bone conduction threshold was recorded compared to the lighter teflon pistons [103], [118]. Gold implants have recently been increasingly superseded by titanium implants.

Titanium: Titanium is a much lighter and more rigid material compared to gold (4 times lower mass). Compared to gold it has elastic spring characteristics so that it is suitable e.g. for fabrication of clip mechanisms. It is valued as one of the most biocompatible materials as it forms an oxide layer in the middle ear which in turn acts as a base for connective tissue cells [103]. Bony adhesions with the surroundings or foreign body reactions were not found in either animal experiments or in explanted prostheses [119], [120]. There has so far been no evidence of osseointegration of the titanium in the middle ear.

Bone cements: Glass ionomer cement

This cement, which was first employed in dentistry, is composed of a mixture of calcium aluminium fluorosilicate glass and polyalkane acid, which hardens intraoperatively to a solid material in a chemical reaction [103]. Aluminium intoxication with fatal outcome was observed after contact of the unhardened cement with the cerebrospinal fluid (CSF) in the reconstruction of large bone defects at the skull base [121]. After that the material was withdrawn from the market although when used sparingly for reconstruction of the ossicular chain (defects of the process of the incus) it is an ideal material both acoustically and mechanically [122]. However, the idea of bridging small bony ossicle defects with modellable cement has not been abandoned. In a more recent study, glass ionomer cement was again used for reconstruction of the long process of the incus with complication-free healing after a year [123].

3.3.2. Which material is preferred?

The historical development of implants and the preference for a material differs regionally and between continents. In the USA hydroxyapatite is used predominantly for TORP or PORP implants, 82%, autologous incus or malleus 72% followed by autologous cartilage 62% [109]. It is noteworthy that in recent years biomaterials have been used more frequently instead of autologous bone or cartilage. Among the biomaterials hydroxyapatite leads with 82% followed by Plastipore with 59%. In recent years metal implants (titanium) have been added and while they represent only 12% they achieve the highest percentage satisfaction levels.

In Germany titanium implants have been used in the middle ear since the 1990s. The number of publications suggests that this material is being used increasingly, where the delicate design, the low weight and the high rigiditiy in particular are considered responsible for the good hearing results [70], [99], [124], [125], [126], [127].

3.3.3. Hearing results with different implants

With a postoperative sound conduction component of up to 20 dB in 40-70% of cases with an intact stapes and in 20-55% when the upper part of the stapes is missing, the hearing results after tympanoplasty must be regarded as rather unsatisfactory [128], [129], [130], [131], [132]. Although these results cannot be blamed only on the reconstruction technique or the implant, there has been a gradual move in recent years from the purely empirical development of the implants to targeted optimising of the mechanical characteristics in model calculations and experiments [8], [133], [134], [135]. Publications of clinical results permit only limited assessment of the influence of the material on the hearing result because the compared implants usually differ considerably not only in material characteristics but also in design. In addition there are different coupling situations which make it difficult to compare different studies. Experimental investigations and calculations should therefore be evaluated more highly in general. Critical material characteristics for sound transmission are the mass and the rigidity. It can generally be concluded from previous impedance calculations on the middle ear model that the rigidity of a middle ear prosthesis should be much greater than the impedance of stapes and cochlea together [52]. Experimental investigations on models suggest that the rigidity has an influence on sound transmission particularly at the high frequencies [136], [137]. According to experimental investigations on prosthesis mass, this should not be more than 15 mg in order to avoid losses of transmission at the high frequencies [138]. A hitherto little heeded factor in ossicular chain reconstruction is the quality of the coupling of the implant to the ossicle remnants. New possibilities result from the design and the elastic spring characteristics of the titanium.

3.4. Reconstruction of the ossicular chain in defects of the long incus process

Retraction pockets or small cholesteatomas often destroy the long process of the incus. In these cases there are 3 possibilities for reconstruction (see Figure 9 [Fig. 9]):

- bridging the defect with autologous bone or cartilage

- bridging the defect with adhesives or cements

- bridging the defect with an implant

From the biomechanical aspect, closure of the defect by adhesives or cements is an elegant solution. Acoustically stiff connections are produced through the connection with the bone that can be achieved only with difficulty through the interposition of bone or cartilage. Moreover, this reconstruction allows compensation of atmospheric pressure fluctuations as the normal anatomy of the chain is preserved. The problems of the long-term biocompatiblity and availability of these materials have already been mentioned.

Cartilage should not be used for reconstruction of large defects of the long process of the incus because the very soft material cannot restore the lever situation and losses of transmission by damping are to be feared. Small defects of the lenticular process, which can be bridged by pushing in a disc of cartilage, are an exception. However, a bone chip is more suitable in these cases from the acoustic aspect.

If thick discs of cartilage inserted onto the head of the stapes contact the tympanic membrane, the reconstruction can be compared to a type 3 tympanoplasty. Although good hearing results are described here initially, the long-term results depend on the state of inflammation and the ventilation. If the cartilage is resorbed with recurrent inflammation or if its consistency alters, abnormalities in the coupling must be anticipated.

The Plester angle prosthesis is available as an implant for reconstruction of the long process of the incus [122]. It is almost the only implant that allows a largely physiological reconstruction of the chain. However, coupling to the long process of the incus is not always easy to bring off. The relatively rigid titanium clips can often not be fixed securely when the process of the incus is short. In contrast, coupling of gold plates to the head of the stapes succeeds much better. When clipping, strain of the annular ligament by tilting of the stapes should be watched for.

3.5. Reconstruction with incus interpositions

In the non-inflamed middle ear, interposition of autologous incus is favoured if it has not been affected by cholesteatoma. When used correctly, the acoustic results of incus interposition can be compared with those of modern implants. Geyer investigated 354 patients in a retrospective study over an observation period of 1.5 years on average and compared incus, ionomer cement and titanium implants using the residual sound conduction components. After incus interposition the average residual sound conduction fell below 10 dB, after titanium PORP below 10-15 dB and after ionomer cement PORP below 15 dB [99].

Nevertheless, the long-term tendency to bone resorption and the risk of adhesions with the surroundings must be assessed critically. The interposed body of the incus is usually very voluminous and can touch the promontory or the bony frame of the tympanic membrane. This can lead to long-term losses in sound transmission by friction or fixation (see Figure 10 [Fig. 10]). Another problem is coupling to the head of the stapes, which cannot be performed under vision, compared to titanium implants. For secure anchorage with regard to atmospheric pressure fluctuations, deep drilling in the incus remnant is necessary (see Figure 11 [Fig. 11]). If this succeeds, strong adhesions must be feared so that dislocation of the stapes threatens in revision operations. In this point, the incus is inferior to modern implants.

To simplify the complicated coupling of the incus, teflon implants were developed which are drilled into the incus remnant and sit on the head of the stapes with a cap [139]. However, the advantage of the autologous material is reduced so that in these cases a complete switch to an implant can be made.

From today's aspect an incus interposition should be inserted only when the likelihood of a revision is regarded as low and the middle ear is completely free from inflammation. In the absence of the upper part of the stapes and when the oval niche is narrow the chance of success with modern delicate TORP implants is greater than when the incus is used.

3.6. Reconstruction with implants when the stapes superstructure is intact

3.6.1. Coupling to the head of the stapes

The design of most implants is adapted to the shape of the head of the stapes either preoperatively or intraoperatively. Either tube, cup or bell shapes are produced. It is unclear how firmly an implant should be bound to the head of the stapes. The contact region between prosthesis and bone has so far hardly been investigated systematically. Few experimental studies show the effect of a more or less rigid prosthetic coupling. Eiber was able to measure the rigidity of the connection between bell and head of stapes and found that the prosthesis should grip the head with 50 mN in order to avoid resonances and overtones [140]. According to subjective observations it is suspected that after healing of the implant even in a natural manner a firm connection between bell and head of stapes might occur because of connective tissue tethering of the metal to the head [55]. So far there have been no known animal studies of this connective tissue tethering with measurement of the transmission function. According to histological investigations in the animal model (rabbit) titanium in the middle ear is covered only with normal mucosa [119].

With titanium implants, clipping them to the head of the stapes can be risky because higher forces are necessary for bending the limbs of the bell than with the comparatively soft gold. In this situation titanium implants with elastic spring limbs offer the possibility of a close fit because of the material and design [126]. The danger of osseointegration, which is familiar with titanium from implantology, does not result from this contact with the bone, which is only superficial, so that the possibility of a low-risk implant exchange is maintained [120].

3.6.2. Coupling to the tympanic membrane

Besides the contact zone between the prosthetic bell and the stapes there is a further contact area between the tympanic membrane and prosthetic plate, which has an influence on the acoustic transmission characteristics. Nishihara studied the influence of the prosthetic angle when coupling a PORP to the tympanic membrane. A prosthesis placed on the stapes with a flat plate was compared with a prosthesis with a plate angled 30° [138]. After surface contact with the tympanic membrane, transmission losses at 1 kHz and gains at between 1 and 3 kHz of 10 dB in each case were found compared to the angled prosthesis. A flat contact between the prosthesis plate and the tympanic membrane is desirable.

A disc of cartilage is usually interposed between the tympanic membrane and the prosthesis plate. So far there have only been examples of calculations in a model which described smoothing of resonances by the cartilage [141]. Initial experimental investigations by means of laser Doppler vibrometry show that the interposed cartilage has only a slight influence on the vibration behaviour [142]. However, it can be assumed that the effect is dependent on the ratio of the cartilage mass to the prosthesis mass. Concha or tragus cartilage should therefore be thinned. Interposed cartilage with an original thickness of 1 mm and a diameter of 4 mm weighs 8 mg. In contrast, cartilage with a thickness of 0.5 mm and the same diameter has a mass of only 4 mg. Thick cartilage should therefore be combined with a light transplant in order to avoid losses of transmission at the high frequencies.

The cartilage interposed between prosthesis and tympanic membrane occupies a key position in the reconstruction when it forms a bridge between the prosthesis plate and the bone margin. When thick plates of cartilage are used, indirect fixation of the prosthesis and losses of transmission can occur. A particularly critical situation is tympanoplasty after creation of a radical cavity. Here the distance between the prosthesis and the lateral semicircular canal is very small. In these cases thin plates of cartilage or cartilage islands should be used if the ventilation circumstances permit (Figure 12 [Fig. 12]).

3.6.3. Coupling to the handle of the malleus

Vlaming and Nishihara determined the centre of the handle of the malleus as the preferred position [138], [143]. Calculations with the finite element model confirmed these experimental results [144], [145]. However, there are situations where a prosthesis should not be coupled to the handle of the malleus but rather to the tympanic membrane. With a poorly ventilated tympanic cavity, steep malleus handle and high head of stapes, the prosthesis shaft is tilted with regard to the vertical axis of the stapes as a result of which the annular ligament which is more yielding in the lateral direction is thought to be tensed. Nishihara obtained an angle of 45° which should not be exceeded in order to avoid losses of transmission [138]. In this situation a tympanic membrane position near the handle of the malleus is to be preferred and the plate of the prosthesis should contact the handle of the malleus laterally. In that way the advantages of malleus handle coupling can be better used at the high frequencies (see Figure 13 [Fig. 13]) [98].

3.6.4. Coupling to the footplate

The position of the prosthesis on the footplate and the stability of the bone contact are further factors in sound transmission. Previous investigations show that a position in the centre of the footplate is optimal [146]. With coupling in front there are losses at the high frequencies and at the low frequencies with coupling behind. Apparently an optimum is found for most frequencies at the centre of the footplate, onto which the prosthesis shaft should be placed at an angle of about 5-15°.

It is difficult to keep the foot of the prosthesis in position on the smooth bone of the centre of the footplate. Various solutions have been proposed. The use of spike or spandrel techniques can ensure good mechanical anchorage [147], but involves a risk of CSF fistula or footplate fracture, so use in the contaminated ear cannot be recommended. In contrast, stabilisation of the prosthesis foot on the footplate by cartilage can be regarded as nonhazardous (see Figure 14 [Fig. 14]). Micropunches are already available for preparing a cartilage shoe to accept the prosthesis [148]. According to our own experimental investigations, such a cartilage shoe damps the footplate vibration by a maximum of 6 dB even with firm adhesions to the surroundings. However, the cartilage should be thinned to 0.5 mm here also.

3.7. Restoration of hearing function with ossicular chain fixation and ligament fixation

3.7.1. Fixation by ligament sclerosis

When examining the mobility of the ossicular chain, the ligaments should also be palpated with the dissecting needle. Sclerotic fixation of the anterior mallear ligament can affect sound transmission considerably. In experimental investigations using laser Doppler vibrometry a reduction in sound transmission through the middle ear of 0-12 dB was found after artificial fixation of the anterior mallear ligament [149], [150]. Calculations with the Finite Elemente (FE) model show values of 6 dB after 1000-fold stiffening of the anterior mallear ligament [151]. There is little information so far on the frequency of this fixation. It is suspected that the anterior mallear ligament is to blame in 38% of cases with residual sound conduction components after stapedoplasty [152]. To divide the ligament fixation, the ligament should be divided with the crescent knife or laser and then the bony origin of the ligament should be removed with the House curette so that a wide gap is produced between the anterior process of the malleus and the bony margin. Placing a silicone sheet between them can prevent further adhesions in the individual case.

The procedure with a sclerosed tensor tendon is less problematic. It can be divided with fine scissors without producing deleterious forms of vibration on the tympanic membrane. This can also often eliminate a steep position of the malleus handle with contact to the promontory.

3.7.2. Fixation of the head of the malleus

The diagnosis of idiopathic fixation of the head of the malleus is usually made by examination of movements under the Siegle's pneumatic ear funnel. There are essentially 2 different ways of solving the problem. The more complex but more physiological way consists of removal of the bony bridge between the epitympanum and the head of the malleus via mastoidectomy. However drilling directly on the head of the malleus must be avoided. The bone can first be thinned with the drill at a low speed in the epitympanum, but removal of the bony bridge to the head of the malleus should rather be done with the House curette or laser in order to prevent noise injury to the inner ear. It should be recalled that sound levels of 120dB and more occur when the drill touches the malleus [153].

A second way consists of removal of the incus, punching out the head of the malleus and grinding the incus to fit as an interposition. Even though mastoidectomy is avoided with this procedure and the risk of inner ear trauma is low, the operation represents the destruction of an intact ossicular chain.

3.7.3. Fixation of the annular ligament

Fixation of the annular ligament is a surgical challenge because of the risk of deafness and the often narrow conditions in the oval niche although stapedoplasty has become established as a largely standardised operation technique. While stapes mobilisation or fenestration of the semicircular canals was done in the past in otosclerosis, insertion of a piston has become popular since the 1950s [154]. Through this technique the risks of the operation could be reduced and the long-term hearing results improved. Performance of a stapedotomy, in which a small hole is made in the footplate to accept the piston, can essentially be recommended today [155], [156], [157], [158]. Compared to partial or complete removal of the footplate, this technique offers the advantage of a largely atraumatic opening of the inner ear, a smaller perilymph fistula compared to stapedectomy and thus a reduction in the risk of infection. As the perforation of the footplate has to be smaller than the diameter of the footplate 1.33 ±0.11 mm [159] to avoid a transverse fracture of the footplate, piston diameters of 0.4 mm up to a maximum of 0.8 mm have become accepted. As reported in the review articles of Häusler and Hüttenbrink, there is no agreement in clinical studies on which piston diameter promises the best hearing success [157], [160]. Calculations show that better sound transmission can be expected at the low frequencies (up to about 2 kHz) with increasing piston diameter, which was also confirmed in our own clinical studies [8], [161]. On the other hand, whether losses in sound transmission actually occur above this frequency with increasing piston diameter, as would be suspected theoretically, cannot be decided with certainty from the conflicting clinical investigations and the model calculations which at present are reliable only up to 2 kHz.

The material most often employed in stapedoplasty is the platinum-teflon piston. However, titanium pistons have also been used recently. A distinction is made in the surgical procedure between the conventional technique of stapes dissection and the laser technique. Supporters of the laser technique praise the contact-free and largely blood-free dissection, while opponents point to the danger of heat or noise trauma and the time and cost involved. The results of conservative instrumental methods and laser surgery are similar [162]. Regarding modern instrumental techniques and the use of lasers, the articles of Häusler and Jovanovic can be consulted [157], [162].

There are interesting innovations in the coupling of stapes prostheses to the long process of the incus. Micromechanical clip mechanisms not only avoid the manual gripping of the prosthesis which is often laborious with titanium but also enable it to be fitted tightly to the incus. The acoustic circumstances might thus be stabilised on the long process of the incus. As investigations using laser Doppler vibrometry have shown, the close fit of the piston on the process of the incus has a significant influence on vibration behaviour. With only loose contact, transmission losses of 10 dB on average and in individual cases up to 28 dB can be expected (see Figure 15 [Fig. 15]) [163].


4. Prospects

If the implant developments of recent years are considered, an improvement in mechanical characteristics can be seen because of new technologies in the manufacture of titanium implants. The implants have not only become lighter and more delicate but also allow better coupling to the ossicles through clip mechanisms. A hitherto unsolved problem with conventional implants is the rigidity of the columella. Along with the acoustic transmission function, the healthy ossicular chain has the task of compensating atmospheric pressure fluctuations and thus avoiding damage to the inner ear. This function has not been looked after by previous implants. Although the tympanic membrane and the annular ligament are strong enough in most cases to compensate atmospheric pressure fluctuations, this function is desirable for future implant developments in order to avoid dislocations of the prosthesis, protrusions or penetration into the vestibule especially when the annular ligament has been damaged previously (see Figure 16 [Fig. 16]). There are attempts at a solution by the integration of micromechanical joints in the prosthesis. For example, a joint between the plate and the shaft can help the prosthesis plate placed below the handle of the malleus to move with the movements of the malleus with atmospheric pressure changes. Other proposals envisage integration of the joint in the shaft of the prosthesis (see Figure 17 [Fig. 17]). That way, the prosthesis can buckle when the handle of the malleus moves in. The joint must be so designed that no losses of sound transmission occur in the joint. By means of laser Doppler vibrometry the characteristics of implants with integrated joints can be optimised. For middle ear surgery these implants would signify more safety for the protection of the inner ear and help to reduce the risk of prosthesis dislocation.


References

1.
Shaw EAG. Eardrum representation in middle-ear acoustical networks. Meet Acoust Soc Am 1977:94.
2.
Helmholtz H. Die Lehre von der Tonempfindung; 1862.
3.
Trendelenburg F. Einführung in die Akustik, 3.Aufl. ed. Berlin/Göttingen/Heidelberg: Springer; 1961.
4.
Ackley RS, Traynor RM. Measurement of ear canal resonance. Hearing Instruments 1989;40:11.
5.
Rodriguez Jorge J, Zenner HP, Hemmert W, Burkhardt C, Gummer AW. Laservibrometrie. HNO 1997;45:997-1007.
6.
Hartwein J. Untersuchungen zur Akustik der offenen Mastoidhöhle (sog. "Radikalhöhle") und deren Beeinflußbarkeit durch chirurgische Maßnahmen. Teil II: Klinische Untersuchungen. Laryng Rhinol Otol 1992;71:453-461.
7.
Molvaer OI, Vallersnes FM, Kringlebotn M. The size of the middle ear and the mastoid air cell. Acta Otolaryngol 1978;85:24-32.
8.
Rosowski JJ, Merchant SN. Mechanical and acoustic analysis of middle ear reconstruction. Am J Otol 1995;16:486-497.
9.
Mercke U. The cholesteatomatous ear one year after surgery with obliteration technique. Am J Otol 1987;8:534-536.
10.
Palva P. Surgical Treatment of Chronic Middle Ear Disease. Acta Otolaryngol (Stockholm) 1987;104:279-284.
11.
Hartwein J. Untersuchungen zur Akustik der offenen Mastoidhöhle (sog. "Radikalhöhle") und deren Beeinflußbarkeit durch chirurgische Maßnahmen. Teil I: Physikalische Grundlagen, experimentelle Untersuchungen. Laryng Rhinol Otol 1992;71:401-406.
12.
Satar B, Yetiser S, Ozkaptan Y. Evolving acoustic characteristics of the canal wall down cavities due to neo-osteogenesis by periosteal flap. Otol Neurotol 2002;23:845-849.
13.
Mahendran S, Yung MW. Mastoid obliteration with hydroxyapatite cement: the Ipswich experience. Otol Neurotol 2004;25:19-21.
14.
Roberson JB, Jr., Mason TP, Stidham KR. Mastoid obliteration: autogenous cranial bone pAte reconstruction. Otol Neurotol 2003;24:132-140.
15.
Tsai TL, Lien CF, Guo YC. Mastoid-obliteration surgery with cartilage for suppurative cholesteatomatous ears. Zhonghua Yi Xue Za Zhi (Taipei) 2002;65:523-528.
16.
Soo G, Tong MC, van Hasselt CA. Mastoid obliteration and lining using the temporoparietal fascial flap. Laryngoscope 1997;107:1674.
17.
Moffat DA, Gray RF, Irving RM. Mastoid obliteration using bone pate. Clin Otolaryngol 1994;19:149-157.
18.
East CA, Brough MD, Grant HR. Mastoid obliteration with the temporoparietal fascia flap. J Laryngol Otol 1991;105:417-420.
19.
Chang SO, Min YG, Kim CS, Koh TY. Surgical management of congenital aural atresia. Laryngoscope 1994;104:606-611.
20.
House HP. Management of congenital ear canal atresia. Laryngoscope 1953;63:916-946.
21.
Bauer GP, Wiet RJ, Zappia JJ. Congenital aural atresia. Laryngoscope 1994;104:1219-1224.
22.
McKinnon BJ, Jahrsdoerfer RA. Congenital auricular atresia: update on options for intervention and timing of repair. Otolaryngol Clin North Am 2002;35:877-890.
23.
Cole RR, Jahrsdoerfer RA. Congenital aural atresia. Clin Plast Surg 1990;17:367-371.
24.
Chandrasekhar SS, De la Cruz A, Garrido E. Surgery of congenital aural atresia. Am J Otol 1995;16:713-717.
25.
Crabtree JA. Congenital atresia: case selection, complications, and prevention. Otolaryngol Clin North Am 1982;15:755-762.
26.
Bellucci RJ. Congenital Aural Malformations: Diagnosis and Treatment. Otolaryngol Clin North Am 1981;14:95-124.
27.
Jahrsdoerfer RA, Yeakley JW, Aguilar EA, Cole RR, Gray LC. Grading system for the selection of patients with congenital aural atresia. Am J Otol 1992;13:6-12.
28.
Lambert PR. Congenital aural atresia: stability of surgical results. Laryngoscope 1998;108:1801-1805.
29.
De la Cruz A, Linthicum FH, Jr., Luxford WM. Congenital atresia of the external auditory canal. Laryngoscope 1985;95:421-427.
30.
Minatogawa T, Nishimura Y, Inamori T, Kumoi T. Results of tympanoplasty for congenital aural atresia and stenosis, with special reference to fascia and homograft as the graft material of the tympanic membrane. Laryngoscope 1989;99:632-638.
31.
Schuknecht HF. Congenital aural atresia. Laryngoscope 1989;99:908-917.
32.
Jahrsdoerfer RA. Congenital atresia of the ear. Laryngoscope 1978;88:1-48.
33.
Sperling NM, Kay D. Diagnosis and management of the lateralized tympanic membrane. Laryngoscope 2000;110:1987-1993.
34.
Yotsuyanagi T, Urushidate S, Nihei Y, Yokoi K, Sawada Y. Reconstruction of congenital stenosis of external auditory canal with a postauricular chondrocutaneous flap. Plast Reconstr Surg 1998;102:2320-2324.
35.
Chang SO, Jeon SJ, Jeong HS, Kim CS. Prevention of postoperative meatal stenosis with anteriorly and inferiorly based periosteal flaps in congenital aural atresia surgery. Otol Neurotol 2002;23:25-28.
36.
Bell DR. External auditory canal stenosis and atresia: dual flap surgery. J Otolaryngol 1988;17:19-21.
37.
Dhooge IJ, Vermeersch HF. The use of two pedicled skin flaps in the surgical treatment of acquired atresia of the outer ear canal. Clin Otolaryngol 1999;24:58-60.
38.
Furuta S, Noguchi M, Takagi N. Reconstruction of stenotic external auditory canal with a postauricular chondrocutaneous flap. Plast Reconstr Surg 1994;94:700-704.
39.
Siegert R, Weerda H. Two-step external ear canal construction in atresia as part of auricular reconstruction. Laryngoscope 2001;111:708-714.
40.
Siegert R. Combined reconstruction of congenital auricular atresia and severe microtia. Laryngoscope 2003;113:2021-2027; discussion 2028-2029.
41.
Jahrsdoerfer RA. The facial nerve in congenital middle ear malformations. Laryngoscope 1981;91:1217-1225.
42.
Lambert PR, Dodson EE. Congenital malformations of the external auditory canal. Otolaryngol Clin North Am 1996;29:741-760.
43.
Altmann F. Congenital atresia of the ear in man and animals. Ann Otol Rhinol Laryngol 1955;64:824-858.
44.
Gill NW. Congenital atresia of the ear. A review of the surgical findings in 83 cases. J Laryngol Otol 1969;83:551-587.
45.
Siegert R, Weerda H, Mayer T, Bruckmann H. [High resolution computerized tomography of middle ear abnormalities]. Laryngorhinootologie 1996;75:187-194.
46.
Mattox DE, Fisch U. Surgical correction of congenital atresia of the ear. Otolaryngol Head Neck Surg 1986;94:574-577.
47.
Jahrsdoerfer RA, Hall JW, 3rd. Congenital malformations of the ear. Am J Otol 1986;7:267-269.
48.
Rosowski JJ. The effects of external and middle ear filtering on auditory threshold and noise induced hearing loss. J Acoust Soc Am 1991;90:124-135.
49.
Austin DF. Acoustic mechanisms in middle ear sound transfer. Otolaryngologic Clinics of North America 1994;27:641-654.
50.
Peake WT, Rosowski JJ, Lynch TJ. Middle-ear transmission: acoustic versus ossicular coupling in cat and human. Hear Res 1992;57:245-268.
51.
Wever EG, Lawrence M. Physiological Acoustics. In, Princeton; 1954.
52.
Merchant SN, McKenna MJ, Rosowski JJ. Current status and future challenges of tympanoplasty. Eur Arch Otolaryngol 1998;255:221-228.
53.
Murbe D, Zahnert T, Bornitz M, Huttenbrink KB. Acoustic properties of different cartilage reconstruction techniques of the tympanic membrane. Laryngoscope 2002;112:1769-1776.
54.
Dornhoffer JL. Hearing Results With Cartilage Tympanoplasty. Laryngoscope 1997;107:1094-1099.
55.
Hüttenbrink K-B. Zur Rekonstruktion des Schallleitungsapparates unter biomechanischen Gesichtspunkten. Laryng Rhinol Otol 2000;79:23-51.
56.
Hüttenbrink K-B. Die operative Behandlung der chronischen Otitis media. I-III. HNO 1994;42:582-593, 648-657, 701-718.
57.
Hildmann H, Sudhoff H, Jahnke K. Grundzüge einer differenzierten Cholesteatomchirurgie. Laryng Rhinol Otol 2000;Suppl. 2:73-94.
58.
Jahnke K. Fortschritte in der Mikrochirurgie des Mittelohres. HNO 1987;35:1-13.
59.
Stage J, Bak-Pedersen K. Underlay tympanoplasty with the graft lateral to the malleus handle. Clin Otolaryngol 1992;17:6-9.
60.
Kartush JM, Michaelides EM, Becvarovski Z, LaRouere MJ. Over-under tympanoplasty. Laryngoscope 2002;112:802-807.
61.
Karlan MS. Gelatin film sandwich in tympanoplasty. Otolaryngol Head Neck Surg 1979;87:84-86.
62.
Schwaber MK. Postauricular undersurface tympanic membrane grafting: some modifications of the "swinging door" technique. Otolaryngol Head Neck Surg 1986;95:182-187.
63.
Williams JD. Microclip application in tympanoplasty. Ann Otol Rhinol Laryngol 1977;86:223-226.
64.
Liew L, Daudia A, Narula AA. Synchronous fat plug myringoplasty and tympanostomy tube removal in the management of refractory otorrhoea in younger patients. Int J Pediatr Otorhinolaryngol 2002;66:291-296.
65.
Kartush JM. Tympanic membrane Patcher: a new device to close tympanic membrane perforations in an office setting. Am J Otol 2000;21:615-620.
66.
McFeely WJ, Jr., Bojrab DI, Kartush JM. Tympanic membrane perforation repair using AlloDerm. Otolaryngol Head Neck Surg 2000;123:17-21.
67.
Wiegand H. [Tympanic membrane repair with cartilage and double tissue-layered grafts (author's transl)]. Hno 1978;26:233-236.
68.
Milewski C. Composite graft tympanoplasty in the treatment of ears with advanced middle ear pathology. Laryngoscope 1993;103:1352-1356.
69.
Amedee RG, Mann WJ, Riechelmann H. Cartilage palisade tympanoplasty. Am J Otol 1989;10:447-450.
70.
Neumann A, Schultz-Coulon HJ, Jahnke K. Type III tympanoplasty applying the palisade cartilage technique: a study of 61 cases. Otol Neurotol 2003;24:33-37.
71.
Müller J, Schön F, Helms J. Realities in tympanoplasty. In: Middle ear mechanics in research and otosurgery, Dresden; 1997. 151-157.
72.
Schöttke H, Hartwein J, Pau HW. Einfluß unterschiedlicher Transplantatmaterialien bei der Tympanoplastik Typ 1 auf den Schalldruckpegel im Gehörgang. OtoRhinoLaryngol Nova 1992;2:318-320.
73.
Zahnert T, Bornitz M, Hüttenbrink K-B. Akustische und mechanische Eigenschaften von Trommelfelltransplantaten. Laryng Rhinol Otol 1997;76:717-723.
74.
Bernal Sprekelsen M, Barberan MT, Lliso MDR. Preliminary functional results of tympanoplasty with palisade cartilage. Acta Otorrinolaringol Esp (Spain) 1997;48:341-346.
75.
Bernal-Sprekelsen M, Romaguera Lliso MD, Sanz Gonzalo JJ. Cartilage palisades in type III tympanoplasty: anatomic and functional long-term results. Otol Neurotol 2003;24:38-42.
76.
Zahnert T, Hüttenbrink K-B, Mürbe D, Bornitz M. Experimental Investigations of the use of Cartilage in Tympanic Membrane Reconstruction. Am J Otol 2000;21:322-328.
77.
Levinson RM. Cartilage-perichondrial composite graft tympanoplasty in the treatment of posterior marginal and attic retraction pockets. Laryngoscope 1987;97:1069-1074.
78.
Duckert LG, Muller J, Makielski KH, Helms J. Composite autograft "shield" reconstruction of remnant tympanic membranes. Am J Otol 1995;16:21-26.
79.
Hartwein J, Leuwer RM, Kehrl W. The total reconstruction of the tympanic membrane by "crowncork" technique. Otolaryngol 1992;13:172-175.
80.
Tolsdorff P. Tympanoplastik mit Tragusknorpeltranplantat "Knorpeldeckel-Plastik". Laryngol Rhinol Otol 1983;62:97-102.
81.
Borkowski G, Sudhoff H, Luckhaupt H. [Autologous perichondrium-cartilage graft in the treatment of total or subtotal perforations of the tympanic membrane]. Laryngorhinootologie 1999;78:68-72.
82.
Rosowski J, Merchant S (eds). The function and mechanics of normal, diseased and reconstructed middle ears.), The Hague, The Netherlands: Kugler Publications; 2000.
83.
Wullstein HL. Operationen zur Verbesserung des Gehörs. Stuttgart: Thieme; 1968
84.
Bekesy GV. Experiments in hearing. (Reprint). Huntington/NY: Robert E. Krieger Publ. Comp; 1960/1980
85.
Stuhlmann O. The nonlinear transmission characteristics of the Auditory Ossicles. J Acoust Soc Amer 1937;9:119-128.
86.
Fischler H, Frei EH, Spira D. Dynamic response of Middle-Ear Structures. J Acoust Soc Am 1967;41:1220-1231.
87.
Gundersen T. Prostheses in the ossicular chain. 1. Mechanics of movement. Arch Otolaryngol 1972;96:416-425.
88.
Decraemer WF, Khanna SM, Funnell WR. Malleus vibration mode changes with frequency. Hear Res 1991;54:305-318.
89.
Decraemer WF, Khanna SM. Vibrations on the Malleus Measured Through the Ear Canal. In: Middle Ear Mechanics in Research and Otosurgery, Dresden; 1997. 32-39.
90.
Huber A et al. Analysis of ossicular vibration in three dimensions. In Middle ear mechanics in research and otosurgery, Dresden; 1997. 82-87.
91.
Zahnert T. Laser in der Ohrforschung. Laryngorhinootologie 2003;82 Suppl 1:157-180.
92.
Willi UB, Ferrazzini MA, Huber AM. The incudo-malleolar joint and sound transmission losses. Hear Res 2002;174:32-44.
93.
Decraemer WF, Khanna SM. New insights into vibration of the middle ear. In Second International Symposium on Middle-Ear Mechanics in Research and Otosurgery, Boston, MA, USA, 1999; 2000. 23-38.
94.
Hüttenbrink KB. Zur funktionellen Anatomie des Mittelohrs: Die Haltebänder der Ossikelkette. Arch f Ohren-, Nasen- und Kehlkopfheilkunde 1988;Suppl.:102-103.
95.
Wada H, Koike T, Kobayashi T. Three-dimensional Finite-Element Method (FEM) analysis of the human middle ear. In: Middle ear mechanics in research and otosurgery, Dresden; 1997. 76-81.
96.
Hudde H, Engel A. Measuring and modeling basic properties of the human middle ear and ear canal. Part I: Model structure and measuring techniques. Acustica & acta acustica 1998;84:720-738.
97.
Huber A, Linder T, Ferrazzini M, Schmid S, Dillier N, Stoeckli S, Fisch U. Intraoperative assessment of stapes movement. Ann Otol Rhinol Laryngol 2001;110:31-35.
98.
Zahnert T. Laserinterferometrische Untersuchungen zur Dynamik des gesunden, pathologisch veränderten und rekonstruierten Mittelohres. Habilitationsschrift 2003.
99.
Geyer G, Rocker J. [Results after rebuilding the ossicular chain using the autogenous incus, ionomer-cement-and titanium implants (tympanoplasty type III)]. Laryngorhinootologie 2002;81:164-170.
100.
Lang J, Kerr AG, Smyth GD. Transplanted ossicles after two decades. J Laryngol Otol 1989;103:471-472.
101.
Steinbach E. Zur Bedeutung der unveränderten Gehörknöchelchenform bei Transplantationen im menschlichen Mittelohr. Arch f Ohren-, Nasen- und Kehlkopfheilkunde 1973;205:146-149.
102.
Eitschberger E. Gefäßentwicklung in Ossikulartransplantaten und -implantaten. Laryng Rhinol Otol 1980;59:238-243.
103.
Dost P. Biomaterials in reconstructive middle ear surgery. Laryngo-Rhino-Otol 2000;79:53-72.
104.
Jahnke K, Plester D. Aluminium oxide ceramic implants in middle ear surgery. Clin Otolaryngol 1981;6:193-195.
105.
Yamamoto E. Aluminum oxide ceramic ossicular replacement prosthesis. Ann Otol Rhinol Laryngol 1985;94:149-152.
106.
Jahnke K, Plester D, Heimke G. Experiences with Al2O3--ceramic middle ear implants. Biomaterials 1983;4:137-138.
107.
Reck R. 5 Jahre klinische Erfahrung mit Ceravitalprothesen im Mittelohr. HNO 1985;33:166-170.
108.
Brewis C, Orrell J, Yung MW. Ceravital revisited: lessons to be learned. Otol Neurotol 2003;24:20-23.
109.
Goldenberg RA, Emmet JR. Current use of implants in middle ear surgery. Otol Neurotol 2001;22:145-152.
110.
Grote JJ, van Blitterswijk CA, Kuijpers W. Hydroxyapatite ceramic as middle ear implant material: animal experimental results. Ann Otol Rhinol Laryngol Suppl 1986;123:1-5.
111.
Costantino PD, Friedman CD, Jones K, Chow LC, Pelzer HJ, Sisson GA, Sr. Hydroxyapatite cement. I. Basic chemistry and histologic properties. Arch Otolaryngol Head Neck Surg 1991;117:379-384.
112.
Gjuric M, Mladina R, Koscak J. [Plastipore prosthesis in the animal experiment]. Laryngol Rhinol Otol (Stuttg) 1987;66:522-525.
113.
Jahnke K, Dost P, Schrader M. Biocompatibility studies of implants for reconstructive middle ear surgery. In: Transplants and Implants in Otology III, Amsterdam; 1996. 41-46.
114.
Himi T, Igarashi M, Kataura A. Temporal bone histopathology over 15 years post-stapedectomy. Acta Otolaryngol Suppl 1988;447:126-134.
115.
Schimanski G. [Erosion and necrosis of the long process of the incus after otosclerosis operation]. Hno 1997;45:682-689.
116.
Heumann H, Steinbach E, Seuffer R. [A clinical and experimental study on precious metal ventilation tubes (author's transl)]. Laryngol Rhinol Otol (Stuttg) 1982;61:17-19.
117.
Jahnke K, Dost P, Mißfeldt N. Revisionen nach Stapeschirurgie. HNO 1999;47:427.
118.
Tange RA, Grolman W, Dreschler WA. Gold and titanium in the oval window: a comparison of two metal stapes prostheses. Otol Neurotol 2004;25:102-105.
119.
Schwager K. Epithelisierung von Titanprothesen im Mittelohr des Kaninchens. Modellvorstellungen zur Mukosaentwicklung. Laryngo-Rhino-Otol 1998;77:38-42.
120.
Schwager K, Müller J, Schön F. Verträglichkeit von Titan-Prothesen im Mittelohr - Ergebnisse einer klinischen Kontrolle. HNO 1999;47:430.
121.
Hantson P, Mathieu P, Gersdorff M, Sindic C, Lauwerys R. Encephalopathy with seizures after use of aluminium-containing bone cement. Lancet 1994;344:1647.
122.
Maassen MM, Zenner HP. Tympanoplasty Type 2 With Ionomeric Cement and Titanium-Gold- Angle Prostheses. The American Journal of Otology 1998;19:693-699.
123.
Ozer E, Bayazit YA, Kanlikama M, Mumbuc S, Ozen Z. Incudostapedial rebridging ossiculoplasty with bone cement. Otol Neurotol 2002;23:643-646.
124.
Zenner HP, Freitag HG, Linti C, Steinhardt U, Rodriguez JJ, Preyer S, Mauz PS, Surth M, Planck H, Baumann I, Lehner R, Eiber A. Acoustomechanical properties of open TTP titanium middle ear prostheses. Hear Res 2004;192:36-46.
125.
Stupp GH, Dalchow C, Grün D, Stupp HF, Wustrow J. Titan-Prothesen im Mittelohr. 3-Jahres-Erfahrungsbericht. Laryng Rhinol Otol 1999;78:299-303.
126.
Huttenbrink KB, Zahnert T, Wustenberg EG, Hofmann G. Titanium clip prosthesis. Otol Neurotol 2004;25:436-442.
127.
Dalchow CV, Grun D, Stupp HF. Reconstruction of the ossicular chain with titanium implants. Otolaryngol Head Neck Surg 2001;125:628-630.
128.
Goldenberg RA. Ossiculoplasty with composite prostheses. PORP and TORP. Otolaryngologic Clinics of North America 1994;27:727-745.
129.
Slater PW, Rizer FM, Schuring AG, W.H. L. Practical Use of Total and Partial Ossicular Replacement Prosthesis in Ossiculoplasty. Laryngoscope 1997;107:1193-1198.
130.
Vartiainen E, Nuutinen J. Long-term hearing results of one-stage tympanoplasty for chronic otitis media. Eur Arch Otolaryngol 1992;249:329-331.
131.
Fisch U. Total reconstruction of the ossicular chain. Otolaryngologic Clinics of North America 1994;27:785-797.
132.
Merchant SN, Ravicz ME, Voss SE, Peake WT, Rosowski JJ. Middle ear mechanics in normal, diseased and reconstructed ears. J Laryng Otol 1998;112:715-731.
133.
Goode RL, Nishihara S. Experimental models of ossiculoplasty. Otolaryngol Clin North Am 1994;27:663-675.
134.
Hüttenbrink K-B. Akustisch optimierte Mittelohrprothesen. Neue Techniken zur zukünftigen Erforschung und Entwicklung verbesserter Implantate. HNO 1997;7:509-511.
135.
Zahnert T, Schmidt R, Hüttenbrink K-B, Hardtke H-J. FE-Simulation of vibrations of the Dresden middle ear prosthesis. In Middle ear mechanics in research and otosurgery, Dresden; 1997. 200-206.
136.
Meister H, Walger M, Mickenhagen A, Stennert E. Messung der Schwingungseigenschaften von Mittelohrimplantaten mit einem mechanischen Mittelohrmodell. HNO 1998;46:241-245.
137.
Meister H, Stennert E, Wolger M, Klinter HD, Mickenhagen A. Ein Meßsystem zur Überprüfung des akustomechanischen Übertragungsverhaltens von Mittelohrimplantaten. HNO 1997;45:81-85.
138.
Nishihara S, Goode R. Experimental study of the acoustic properties of incus replacement prostheses in a human temporal bone model. Am J Otol 1994;15:485-494.
139.
Schuring AG, Lippy WH. Semibiologic middle ear prostheses: ossicle cup and ossicle columella. Otolaryngol Head Neck Surg 1982;90:629-624.
140.
Eiber A, Freitag H-G, Schimanski G, Zenner HP. On the coupling of prostheses to the middle ear structure and its influence on sound transfer. In: Second International Symposium on Middle-Ear Mechanics in Research and Otosurgery, Boston, MA, USA, 1999; 2000. 297-308.
141.
Eiber A, Freitag H-G, Burkhardt C, Hemmert W, Maassen MM, Jorge JR, Zenner H-P. Dynamics of middle ear prostheses - simulations and measurements. Audiology & Neuro-Otology 1999;4:178-184.
142.
Morris DP, Bance M, van Wijhe RG, Kiefte M, Smith R. Optimum tension for partial ossicular replacement prosthesis reconstruction in the human middle ear. Laryngoscope 2004;114:305-308.
143.
Vlaming MS, Feenstra L. Studies on the mechanics of the reconstructed human middle ear. Clin Otolaryngol 1986;11:411-422.
144.
Koike T, Wada H, Kobayashi T, Takasaka T. Finite-element method analysis of a reconstructed middle ear. Proceedings ARO-Meeting (Association of Research in Otology) in Florida 1997.
145.
Koike T, Wada H, Kobayashi T. Modeling of the human middle ear using the finite-element method. The Journal of the Acoustical Society of America 2002;111:1306-1317.
146.
Asai M, Huber A, Goode RL. Analysis of the best site on the stapes footplate for ossicular chain reconstruction. Acta Otolaryngol 1999;119:356-361.
147.
Fisch U, May J. Tympanoplasty, mastoidectomy and stapes surgery. Stuttgart, New York: Thieme 1994, p. 39-40
148.
Huttenbrink KB, Zahnert T, Beutner D, Hofmann G. [The cartilage guide: a solution for anchoring a columella-prosthesis on footplate]. Laryngorhinootologie 2004;83:450-456.
149.
Huber A, Koike T, Wada H, Nandapalan V, Fisch U. Fixation of the anterior mallear ligament: Diagnosis and consequences for hearing results in stapes surgery. Ann OtoI Rhinol Laryngol 2003; 112: 348-355
150.
Nakajima HH, Peake WT, Rosowski JJ, Merchant SM. The Effects of Stiffening the Anterior Mallear Ligament. Singapore: World Scientific Publishing Co. Pte.Ltd.; 2004. p. 197-203.
151.
Koike T, Wada H, Kobayashi H. Effects of Individual Differences in Size and Stiffness of the Middle Ear on its Sound Transmission. Singapore: World Scientific Publishing Co.; 2004. p. 68-75.
152.
Huber A, Koike T, Wada H, Nandapalan V, Fisch U. Fixation of the anterior mallear ligament: diagnosis and consequences for hearing results in stapes surgery. Ann Otol Rhinol Laryngol 2003;112:348-355.
153.
Zou J, Bretlau P, Pyykko I, Starck J, Toppila E. Sensorineural hearing loss after vibration: an animal model for evaluating prevention and treatment of inner ear hearing loss. Acta Otolaryngol 2001;121:143-148.
154.
Shea JJ. A 20-year report on fenestration of the oval window. Trans Sect Otolaryngol Am Acad Opthamol Otolaryngol 1976;82:ORL21-29.
155.
Sedwick JD, Louden CL, Shelton C. Stapedectomy vs stapedotomy. Do you really need a laser? Arch Otolaryngol Head Neck Surg 1997;123:177-180.
156.
Fisch U, Dillier N. Technik und Spätresultate der Stapedotomie. HNO 1987;35:252-254.
157.
Häusler R. Fortschritte in der Stapeschirurgie. Laryng Rhinol Otol Suppl 2 2000;79:95-139.
158.
Kursten R, Schneider B, Zrunek M. Long-term results after stapedectomy versus stapedotomy. Am J Otol 1994;15:804-806.
159.
Kirikae J. The middle ear. The University of Tokyo Press 1960.
160.
Huttenbrink KB. Biomechanics of stapesplasty: a review. Otol Neurotol 2003;24:548-557; discussion 557-549.
161.
Teig E, Lindemann H. Stapedotomy piston diameter - is bigger better? Oto Rhinolaryngologia Nova 1999;9:252-256.
162.
Jovanovic S. Laser in der Otologie. Laryngo-Rhino-0tol 2003;82:21-53.
163.
Huber AM, Ma F, Felix H, Linder T. Stapes prosthesis attachment: the effect of crimping on sound transfer in otosclerosis surgery. Laryngoscope 2003;113:853-858.