Restorative procedures in cases of impaired voice function following complete laryngectomy.

Surgical voice restoration with a tracheo-oesophageal fistula using an alloplastic voice prosthesis is the current standard in Germany for patients with laryngectomy. With the increasing widespread use of this type of rehabilitation new duties emerge, not only for ones colleagues in the hospital but also for those in general practice. Care of these patients close to their home must be the aim of us all. With the use of voice prostheses on the increase any problems arising for the clinician are few and easily treatable; for these a therapy concept should be borne in mind. Surgical voice restoration is indicated only in individual cases due, among other reasons, to the high operational costs involved. If surgical voice restoration is impossible or unsuccessful, oesophageal voice replacement and electronic voice support are realistic alternatives. Improvements can be expected as regards the durability of voice prostheses, methods of replacing them, and speech procedure with finger-free tracheostomy closure. Greater use of the artificial larynx without tracheostomy is the objective for the distant future.

1. Introduction

Despite all progress made in organ-preservation therapy in larygeal and hypopharyngeal carcinomas, laryngectomy is still seen as an important pillar in the treatment of extensive tumours; it is also seen in the background of increasing incidence of these tumours [1]. Using modern surgical voice restoration possibilities, laryngectomy has for patients largely solved the problem of loss of voice [2], [3], [4], [5]. In 1990 Hagen stated [6] that in 5.6% of laryngectomies performed in Germany a voice prosthesis was implanted, whilst in 60% oesophageal speech was used, and in 28% voice restoration was accomplished using an electronic synthesizer. In 1995 Maier et al [7] reported on a figure of 16.5% of voice prostheses. Seinsch [8] observed collectively in a German rehabilitation clinic a rise in patients with a voice prosthesis from 1% in 1992 to 60% in 1999. These references illustrate the development in the last 10 years in the field of voice restoration following total laryngectomy.

2. Surgical restoration of voice

Surgical voice restoration following laryngectomy has developed dramatically in the 1990s. The causes lie on the one hand in the production and wider availability and use of functionally-reliable voice prostheses which, due to efficient handling, guarantee high safety when aspirating, and on the other hand in the development of plastic surgical methods in the construction of a voice fistula with safe aspiration.

For all surgical voice restoration procedures the construction of a permanent tracheo-oesophageal fistula is seen as a basic principle whereby, in the case of a classical total laryngectomy, airway and digestive tract are no longer completely separated. This artificial fistula produces a solution to two problems:

1. An almost resistance-free opening for the passage of expired air from the trachea in the direction of the oesophagus/pharynx when speaking

2. When no speaking is advised, a guarantee of a safe, water-tight closure of the trachea for nutrition and salivation

These pre-conditions are fulfilled by either a relatively costly plastic surgical procedure or the simple implantation of a voice prosthesis.

The advantages of a surgical voice restoration lie in the availability of the total volume of expired air of the lungs for speaking, which lead to longer speech cycles, a more powerful voice and a larger voice volume in comparison with oesophageal speaking [9]. Speaking via a voice fistula - compared with an oesophageal voice or electronic aid - is clearly easier to learn, the voice sounds more natural and the possibility of successful voice formation in the overwhelming majority of patients is high [10], [11], [12].

For patients with a voice fistula the following usage problems arise:

- Speaking requires closure of the tracheostomy. As a rule this is done using the finger so that one hand for normal activity is not available which, in public, could lead to certain stigmatisation.

- When using a voice prosthesis it should be remembered that this has a limited usage period. First symptom of this wearing out is the looseness of the valve in consecutive aspirations. Replacing the voice prosthesis is, as a rule, required 2 to 3 times per year, which thus necessitates closer liaison with the specialist doctor or treatment clinic during the usual oncological after-care intervals [13].

2.1 Tracheo-oesophageal fistulae without valve

Following the Hagen laryngoplasty [14] and the further development of the Asai methods in Germany [15] begun by Maier and Weidauer, plastic surgery procedures for voice restoration have progressed in parallel to modern voice prostheses. The advantage is in the independent voice restoration by means of foreign body implants and the associated after-care. Therefore all plastic surgery procedures are accompanied by a different high aspiration rate, the remedy for which - alongside voice restoration - is an important aim of the operation. The implementation of these procedures requires to some extent special surgical knowledge and competence, as the cost of an operation is considerable when compared to that of voice restoration using a prosthesis.

During discussions about cost in health-care, those concerning rehabilitation must also be considered. Certainly in current literature no data exist concerning the difference in cost between voice restoration by plastic surgery and by using long-term voice prostheses.

2.1.1 Procedures without tissue displacement

The Staffieri and Asai voice fistulas marked the start of modern surgical voice restoration [16]. These procedures have lost their significance. Minnigerode et al [17] reported a high aspiration rate following these operations and in 1988 recommended the oesophageal voice-replacement as standard for voice restoration, as did other authors [18], [19].

Vieira et al [20] and Kinishi et al [21] wrote on the creation of the Amatsu voice-shunt [22] from parts of the posterior tracheal wall and oesophageal muscles. Strome et al [23] in 1986 described the formation of a mucous membrane tube as a fistula between trachea and oesophagus. Brasnu et al [24] reported satisfactory results but also on the relatively high aspiration rate of the method. Altogether these methods no longer appear to be of great significance in a hospital's working day.

2.1.2 Procedure with tissue displacement

2.1.2.1 Pedicle grafts

Maier et al. in 1994 introduced a method by which a voice fistula was constructed by using a pedicle graft from the musculus pectoralis [7], [15]. From the residual pharyngeal mucous membrane alongside the closure of the hypopharyngeal cone the authors created a cranial mucus membrane strip stem, which they sutured to the upper edge of the tracheal root.

This mucous membrane illustrated the posterior wall of the voice fistula which was to be formed. The frontal and lateral walls were created using a myofascial musculus pectoralis major graft. The fistula ran from the upper edge of the tracheal root to the lower tonsillar pole. In 90% of patients an acceptable vocal result was achieved [7]. One problem with this procedure lay in the relatively high aspiration rate of 24%. After post-operative radiotherapy a similarly high proportion resulted in scarred shrinkage of the fistula. The time taken for creating the pedicle and the development of the graft was increased when compared to a „simple" laryngectomy. Generally speaking this procedure could not become widespread.

2.1.2.2 Free microvascular tissue transfer

2.1.2.2.1 Hagen laryngoplasty

In 1990 [14] Hagen publicised the construction of a voice fistula from a fascial-cutaneous radial-free forearm flap. The basic surgical principle was the production of an epithelialised tube from the radial free forearm flap, whose caudal opening at the tracheal root and the cranial opening was stitched into the pharyngeal seam beneath the base of the tongue. Additionally, a replacement epiglottis was created from a part of the radial lobe with a piece of ear cartilage [7]. Hagen reported [7] on a temporary aspiration in 28% and a permanent aspiration in 5% of patients studied. In 6% of patients a fistular stenosis developed, which required further therapeutic measures.

This very costly surgical procedure also could not be carried out up to the present time. The causes lie in the significantly high operative costs, the necessary knowledge of microvascular surgery and attendant risks. Due to accompanying illnesses and the accumulated noxiousness, numerous patients with laryngeal or hypo-pharyngeal carcinoma are at risk whereby microvascular surgery is not indicated. The great advantage of this costly method is certainly the relatively safe and stable functional outcome, which renders the patient independent of implanted foreign bodies. Despite the high costs, the procedure therefore seems to appear as a voice restoration possibility especially to younger patients, whose long-term cancer prognosis is favourable.

2.1.2.2.2 Interposition of a jejunum section

The first description of a free transplant of a jejunum section as a pharyngeal replacement and a vocal fistula goes back to the Ehrenburger work-team in Vienna in 1985 [25]. A jejunal section and vessel stem were removed by an abdominal surgeon which the E.N.T. surgeon then transplanted by microvascular re-anastomosis to be used jointly as pharyngeal replacement and vocal fistula [26]. The placement of a membrane-covered tube as a vocal fistula has the advantage of the absence of scar formation with stenosis of the fistula.

The greatly increased risk of abdominal intervention certainly counters this advantage. Such an intervention cannot be carried out in parallel to the E.N.T. procedure, which would extend the operation time and the combined intervention would require suitably qualified surgical teams. To minimise the trauma of an operation there were attempts to obtain a jejunum section by laparoscopy [27]. The functional results of the reconstruction are good but with joint usage as both pharyngeal replacement and vocal fistula, there is certainly an increased danger in aspirating [28], [29].

Remmert [30] publicised the use of a loop from the anterior musculus digastricus to restrain the functional narrowing of the section between the two arms of the jejunum. By means of this plasty a minimal aspiration rate was able to be achieved [31].

This very costly process of complex reconstruction of the oesophagus and vocal function could not be carried out in the last 10 years. The operation costs are many times higher than a Hagen laryngoplasty; therefore this operation is, as a rule, performed only in large hospitals with suitably-qualified staff [32]. In individual cases this reconstruction can be taken into consideration in the operative therapy of extensive hypopharyngeal or upper oesophageal carcinoma, but always emphatically bearing in mind the unfavourable oncological prognosis in contrast to the cost and trauma of an operation.

2.1.2.2.3 Transplantation of an ileo-coccal patch

Kobayshi et al [33] described in 2003 the patch transplant with the ileo-coccal valve acting as tracheo-oesophageal fistula with autologous valve in 7 patients. In 6 cases they reported good voice restoration without aspiration. At the moment there are no further publications on this procedure.

2.1.2.2.4 Laryngeal transplant

Transplantation of a complete larynx is at present not regarded as a meaningful alternative to voice restoration due to the requisite immuno-suppression in tumour patients [34], [35]. In the literature there is a procedure report on this operation about traumatic damage to the larynx [36], [37].

2.1.2.2.5 Artificial larynx

As of today no artificial larynx model is ready to go into mass-production. The complex demands of such an organ, (opening during respiration, achieved narrowing for voice formation, closure when sipping and their complex control mechanisms) have technically still not been met [38].

2.2 Tracheo-oesophageal fistula with voice prosthesis

Lascaratos et al [39] reported that Nicolas Tapas in Constantinople at the end of the 19th century had developed the first functional voice prosthesis, but this was not publicised. The modern development of voice prostheses began over 30 years ago. The first publication of a working voice prosthesis was written in 1972 by Mozolewski et al [40] but as this was in Polish, this development was not generally recognised. Then in 1979 Blom and Singer [41] reported on their voice prosthesis. In Germany, Herrmann [42] and the working-groups from the Netherlands with their various types of voice prostheses - Panje [43], Nijdam [44], [45], Groningen [46], Provox® [47]- were further pioneers in the development of functional voice prostheses.

The creation of an artifical tracheo-oesophageal fistula and the implantation of a voice prosthesis for opening the oesophagus characterises the basic surgical principle. There is differentiation with the Liegedauer replacement and long-term prostheses [48]:

1. Replacement prostheses are as a rule changed by the patients or their relatives daily or every few days [49], [50].

2. Long-term prostheses remain in the fistula for longer (around 6 months, as a rule) and are changed by medical specialist staff (in Germany by an E.N.T. specialist doctor) [51], [52], [53], [54], [55], [56]. Today these prostheses are the most used.

The older voice prosthesis models (Blom-Singer, Eska-Herrmann, Panje) necessitate a relatively high pressure at the opening due to their low inner diameter and clumsy valve mechanism, whilst today's models (Groningen, Provox®) open with low pressure [55], [56], [57], [58]. Similarly, Blom later developed a long-term prosthesis with low opening pressure [59].

Concerning the moment of implantation of the voice prosthesis, differentiation is made between the primary implantation (during the laryngectomy) and the secondary implantation (moved within a given time for larygectomy). During the early 1990s secondary implantation was still being discussed but today's standard is primary implantation [8], [60], [61].

Whilst Hilgers et al [47] and Kerr et al [62] reported on worse functional results following secondary implantation, some examinations [63] and the experiences of Lavertu et al [64] showed equally good results following primary and secondary implantation [65]. The surgical risks in primary implantation can be practically neglected; with secondary implantation a higher risk of oesophageal perforation exists due to the necessarily rigid nature of an oesophagoscopy and the post-therapeutic changes in the underside of the throat [66].

Various modifications to simplify the procedure of secondary implantation of a voice prosthesis have been publicised (flexible endoscopy [67], [68], puncture with retrograde placement [69], puncture with needle [70], gastroscopy [71], modified trocar [72], out-patient [73]). We still give priority to primary implantation [74]. We carry out secondary implantation only in patients with laryngectomy carried out in another place and failed development of vocal eructation.

We see contra-indications to primary restoration using voice prostheses only in cases of sub-glottal tumours penetrating the trachea, non-co-operation of the patient, severe cardio-pulmonary and neurological complaints (e.g. hemiplegia, Parkinson's disease).

Implantation of a voice prosthesis following pharyngeal reconstruction is also possible using a stomach pull-up [75], a radialis or pectoralis patch [76], [77], a colon transposition [78] and a freely transplantable jejunum segment [79].

2.2.1 Types of voice prostheses

Technical developments last year have resulted in functional and stable voice prostheses. Whilst in the 1980s and early 1990s the fewest implantations of Blom-Singer [41], [80] and Eska-Herrmann [42] were carried out in Germany, the Provox® voice prosthesis established itself during the 1990s [81]; this was developed by the Hilgers working-group [47].

There are Provox® 1 and Provox® 2 prostheses. By comparison with the Provox® 1 the Provox® 2 [63] is equipped with a softer oesophageal retaining disc and can therefore be exchanged antegrade using the replacing system through the tracheostomy. This exchange is simpler and less of a strain for both patient and doctor than the retrograde change of the Provox® 1 system, which is effected by using a probe over the hypopharyngeal cone and the oral cavity. Despite good local anaesthesia, a somewhat severe choking sensation was often observed in Provox® 1-changing [63].

More recently, a further model was developed with the Provox-Active valve® [82]. With this prosthesis the valve flap is closed by a magnetic mechanism. The durability of the prosthesis is thereby extended and the problem of the valve opening due to low pressure when sipping is overcome. This realisation is based on the observation that with some patients, when sipping and breathing at the same time, the Provox® 1 and 2 prostheses can open due to low pressure in the oesophagus.

Fluids will overflow into the trachea. Aspiration results as a rule in replacement of the prosthesis, although the prosthesis itself functions correctly. This phenomenon is possibly the cause of decidedly frequent change of prosthesis in individual patients. For the new Provox model there are at present no long-term results. It is therefore uncertain to what extent the promised longer durability time compared to the other models is realised. It should be similarly observed that this prosthesis at the moment in Germany costs 8 times as much as the other Provox models (information personally given by Atos Medical); this means that obtaining approval via the responsible cost-centre is not made any easier.

In 1998 Schouwenburg et al introduced the Voice-Masters® prosthesis [83]. This prosthesis distinguishes itself using a valve mechanism with a titanium ball, which is retained by movable silicone arms. However, this prosthesis cannot generally be used so far. The Staffieri working-group also developed in the meanwhile its own prosthesis [84]. Voice prostheses are only valves which themselves produce no sounds. In recent times there have been attempts to introduce sound-producing elements into the prostheses, especially to better overcome the ‚masculine' sound level, too deep for laryngectomy in women [85], [86].

At the time of writing, voice restoration following laryngectomy using the Provox® voice prosthesis is shown to be the most widely-used method in Germany.

2.2.2 Management of problems with voice prostheses

Possible problems with voice restoration using a voice prosthesis are divided into those seen immediately post-operatively (i.e. no voice restoration) and those later surfacing in connection with the prosthesis, the fistula or the tracheostomy.

2.2.2.1 Failure of voice formation using a prosthesis

If speech using a prosthesis is not possible post-operatively, the reliability of the prosthesis and the resistance of the tracheo-oesophageal section can be relatively easily checked by means of insufflation of air using a flexible suction catheter (Ch 14) and a Politzer cuff over the prosthesis. If resistance by the hypopharyngeal cone seems too great - which becomes recognisable in the majority of the air-insufflation of the stomach - suspicion falls upon a stenosis of the pharyngo-oesophageal section. As the next step, an oesophageal barium meal examination during sipping (to eliminate organic stenosis) - and during attempts at speech -should follow [87], [88], [89].

If a broad hypopharyngeal cone is evident, which narrows during an attempt at speech above the voice prosthesis, a functional stenosis of the pharngo-oesophageal elements comes into question [90]. To prevent such muscular functional disturbances, neurectomy of the plexus myentericus or primary myotomy of the pharyngeal muscular system during laryngectomy is recommended [80], [90], [91], [92], [93]. In the hospital in Jena we have wide experience in primary neurectomy [94], where in the area of the upper oesophageal sphyncter of too narrow proportions as well, a localised myotomy can be required.

Should, however, a post-operative rise in muscle tone in pharyngo-oesophageal section occur, an injection of Botox® 100 IE into the pharyngeal muscular system has shown itself to be today's therapy of choice [95], [96], [97], [98]. As a rule, a single injection is sufficient to prepare the way for a good voice. Repeat injections known in other applications are seldom necessary. In individual cases where Botox injection therapy fails, a secondary myotomy of the pharyngeal muscular system can be indicated, despite the increased risks (fistula formation) [99], [100]. From own experiences in such cases, frequent pronounced scarring around the pharynx is found between operations which, moreover, is clearly prevented by the motility of the pharyngo-oesophageal section. As an alternative to the classic open procedure, trans-oral myotomies using laser are now being suggested [101].

A stage-by-stage logopaedic practice programme, which should also include the proper tracheostomy closure and the paving of the way for ructus phonation, is part of the restoration programme. With this, around 90% of laryngectomy patients with a voice prosthesis can today expect good voice restoration [57], [65].

2.2.2.2 Patients with voice prosthesis: problems with the follow-up

2.2.2.2.1 Replacing the prosthesis

Voice prostheses are susceptible to material ageing, which results in a loosening of the valve and the passage of fluids from the oesophagus into the trachea. The first symptom for the patients is often the observation in the secretions of coffee particles, noticeable by their colour. On average a valve lasts around half a year [102]. Durability of voice prostheses can certainly vary from a few days to many years [63], [103], [104], [105].

Valve insufficiency with aspiration is the most frequent reason for a replacement [63]. For the widely-distributed Provox prostheses, replacement with the Provox® 2 system is the method of choice [50]. This is simply a question of the procedure to be carried out which is also designed for E.N.T. specialists active on the move. Certainly from our experience there are at present only a few General Practitioners or family physicians who replace voice prostheses themselves. The result nowadays is many patients being unnecessarily tied too much to the operative hospital.

Complications with replacing a prosthesis are the exception. There are cases on which the emergence of a mechanical ileus due to a voice prosthesis having being swallowed [106] and a tracheal dislocation with aspiration in the right bronchus [63] were reported. The Provox® voice prostheses are available in different lengths, with the differences being in the various distances between the tracheal and oesophageal retaining disc; the diameter of each prosthesis is the same [47]. For the others less widely distributed models in Germany there are similar simple replacement procedures [41], [42], [83], [84].

Ageing of the voice prosthesis has its causes chiefly in the colonisation of candida albicans and the destruction of silicone due to this fungal colonisation [107]. There are various attempts at stabilising the material surface of the prosthesis with special ‚coatings' [108] or using a laser [109] against fungal colonisation. Nutrition also has an influence beside radiotherapy as regards the bacterial and fungal flora of the pharynx, seen always in the same place [110]. Some examinations [65] have shown that too frequent mechanical cleaning of the voice prosthesis with a brush leads to premature valve insufficiency.

We recommend to our patients cleaning of the voice prosthesis with a brush only if resistance becomes greater when speaking, or if it leads to aspiration via the prosthesis. We do not advise regular cleaning of this kind. Less common reasons for prosthesis replacement are greater resistance when speaking, leaking around the fistula (see 2.2.2.2.5) and loss of the prosthesis (see 2.2.2.2.3).

2.2.2.2.2 Granulation around the voice prosthesis

Granulation is less of a problem around the voice prosthesis, so long as the prosthesis does not move [65]. This applies particularly to granulatory covering at the sides of the voice prosthesis, often observed. If this leads to a movement of the prosthesis, removal by cauterisation, surgical resection or CO2 or diode laser is indicated. If the tendency towards granulation continues, replacement with a longer prosthesis should follow in order to minimise possible tropical causes due to pressure of the flange of the prosthesis. In cases of distinct increase in granulation a recurrence of tumour must be discounted.

2.2.2.2.3 Loss of voice prosthesis

In some patients the voice prosthesis can be lost, i.e. the patient comes to the hospital with an empty fistula. Speech is still possible but with some patients aspiration is not noticed at all. Often patients are unable to relate the circumstances of the loss of the prosthesis and the defective valve. In the case of a radio-opaque Provox® prosthesis lying in the respiratory or gastro-intestinal tract, thoracic and abdominal X-rays [47] are recommended to locate this.

If the prosthesis shows up in the intestine the patient should be monitored until it is passed naturally, in order to recognise complications in good time, as - in special cases - an ileus [106]. If the prosthesis is located in the thoracic region, removal by immediate tracheo-bronchoscopy is indicated. As in most cases the fistula can be probed, a new prosthesis can be inserted without any problem. In such cases we would give preference to the Provox® 1 prosthesis due to its stable position in the oesophagus.

2.2.2.2.4 Spontaneous closure of fistula

A more recent problem is the spontaneous closure of the oesophageal opening by the oesophageal mucous membrane. Typically the patient cannot speak any more, and probing of the voice prosthesis lying normally in the anterior tracheal wall is impossible. Patients report a gradual deterioration of voice quality until speech is no longer possible. Blood can be found at the tip of the brush after attempts at cleaning. In these cases, use of the flexible oesophagoscope under local anaesthetic proved its worth to us in securing a result.

If mucous membrane is seen to grow over the oesophageal arm of the voice prosthesis, perforation of the mucous membrane wall by careful sharp probing of the prosthesis under endoscopic monitoring can be attempted. If this does not work, a new fistula construction under anaesthetic is necessary, using the lumen of the old voice prosthesis as a guide for the probe. We have observed this phenomenon so far only with the Provox® 2 prosthesis [65].

2.2.2.2.5 Enlargement of a tracheo-oesophageal fistula

When widening the tracheo-oesophageal fistula saliva and other fluids run over into the trachea between prosthesis and fistula edge. We have seen this so far only with Provox® 2 prostheses [90], [107], [111]. During attempts at sipping coloured fluids (e.g. milk) prior to prosthesis replacement, the result can be secured [112]. We have developed the following staged therapy concept:

1. Removal of the voice prosthesis, spontaneous shrinking of the prosthesis (approx. 3 days), insertion of a new prosthesis,

2. Purse-string suture of the fistula where the voice prosthesis is situated,

3. Intra-mural injections of human fibrin adhesive (no testing necessary) around the fistula,

4. Operative, three-layered fistula closure, followed later by implantation of a new voice prosthesis [113].

In the references, injections of own fat [114], hyaluronic acid [115], bioplastique [116], [117] and collagen [118] are recommended.

In patients in whom post-operative radiation dates back a long time, an indication of operative fistula closure is very critical to check. Due to radiation fibrosis, major problems with wound-healing are to be expected.

Blame for widening of the voice fistula can be placed upon changes caused by post-operative radiotherapy. Furthermore, we suspect a possible explanation lies in the softer retaining-disc of the Provox® 2 voice prosthesis as against the Provox® 1. This therefore results in insignificant movements of the prosthesis in the fistula during sipping. The outcome is a widening of tissue at the edge of the fistula.

2.2.3 Tracheostomy valves

Surgical voice restoration necessitates a new thinking with regard to the size of the tracheostomy. A stable tracheostomy should have the necessary width for breathing but at the same time not too large that a digital occlusion for speech is possible [119], [120]. One hand is therefore not available for other actions during speech [121]. A certain stigmatisation in public is unavoidable. Contact with tracheal secretions illustrates a hygiene problem for some patients and a handshake is often regarded as unpleasant [122]. A marked improvement is possible by using a tracheostomy filter [123].

Consistent use of heat and moisture exchangers also results in a better vocal quality by improved lung function [124]. Digital occlusion is more hygienic because of the absence of direct contact with tracheal secretions. An optimal closure of the tracheostomy must meet 3 pre-conditions:

1. A wide opening for breathing

2. Easier closure when speaking

3. Quick opening when coughing

Additionally, the mechanism must be easy to operate and fastening in the tracheostomy must be painless, air-tight and secure.

The Eska-Herrmann tracheostomy valve [125] requires special surgical shaping of the tracheostomy to hold the valve and, due to its magnetic valve, causes clicking noises when speaking so that some patients only modestly accept this tracheostomy valve.

The Blom-Singer tracheostomy valve [126], [127] can be attached around the tracheostomy with adhesive or carried on a cannula to work almost silently. A mechanism to enable quick opening during coughing is missing in both systems, the result being a dislocation of the valve during coughing. This problem was better solved using two newer models („Provox Free-Hands®" and „Window®") [128], [129], [130].

As before, a further basic problem with tracheostomy valves lies in the fixation in the tracheostomy. Adhesive fixation frequently results in skin irritation. Carrying the valve on a cannula can lead to mechanical irritation of the tracheal mucous membrane. With all the various fixation methods a problem with looseness exists, so that during speaking, air escapes. This does not happen when speaking using a voice prosthesis.

According to our experiences the solution to this problem lies not just in a more stable fixation of the tracheostomy valve but in less resistance, which the pharyngo-oesophageal section of the expiring air flowing in during the speech procedure opposes (see 2.2.2.1). The best voice restoration in a laryngectomy patient with a well-functioning voice prosthesis is at present, according to our experience, to be achieved using a Blom-Singer tracheostomy valve, carried on a short silicone cannula [65]. Another good possibility is the use of the Hilgers „Free-Hands" HME tracheostomy valve, carried on a cannula [130].

3. Non-surgical voice restoration

3.1 Esophageal voice

In recent years there have been no significant developments in the area of oesophageal speech. For patients for whom a contra-indication to surgical voice restoration exists or who have not learned to speak using these restorative measures, oesophageal speech presents itself as a good alternative [131]. The advantages of ructus speech lie in their being independent of implanted foreign bodies and that both hands are available for actions whilst speaking. The disadvantage is the high cost until the patient has learned to speak, with lower prospects of success in good voice formation.

In our own collective we observed two patients who, following successful voice restoration using a prosthesis, also learned to speak via the oesophagus. They then wanted their voice prosthesis removed.

Despite the simplicity of surgical voice restoration, all patients with a voice prosthesis should at least learn the rudiments of speaking via the oesophagus in order to have an alternative to a voice prosthesis.

3.2 Electronic speaking aids

Electronic speech aids today also constitute a good possibility in voice restoration for patients who cannot speak in other ways. With developments in electronics in recent years trials have been undertaken to improve electronic speech aids in relation to sound quality and to programme individual voice patterns [132], [133], [134]. Initial trials of the direction of an implantable electronic voice-producer via the EMG of the muscular system of the throat have been published [133].

Acknowledgements

My heartfelt thanks to Prof. Frans M. Hilgers, Amsterdam, for his critical examination of this paper and for his friendly support.