Alternatives to CPAP in the treatment of the obstructive sleep apnoea syndrome

The obstructive sleep apnoea syndrome (OSAS) results in excessive daytime sleepiness, impaired quality of life, and is associated with an increased risk of traffic accidents and cardiovascular disease. Nasal continuous positive airway pressure (CPAP), the standard treatment for OSAS provides immediate relief of symptoms and has only minor side effects. Nevertheless, an alternative treatment is needed if CPAP is not feasible for medical or psychological reasons. Removable oral appliances that advance the mandible when fitted to the teeth during sleep also improve nocturnal breathing disturbances, symptoms, quality of life, vigilance and blood pressure in OSAS patients. Their long-term effectiveness and side effects require further study. In morbidly obese patients suffering from OSAS bariatric surgery should be considered as a treatment that reduces obesity and at the same time improves OSAS. In selected patients including those with adeno-tonsillar hypertrophy, and cranio-facial malformations various surgical techniques that enlarge the upper airway may be a treatment option for OSAS.

The obstructive sleep apnoea syndrome (OSAS) is characterised by intermittent collapse of the upper airway during sleep resulting in hypopnoea, apnoea, repetitive oxygen desaturation and sleep disruption [1].Daytime consequences include excessive sleepiness, impaired cognitive performance, disturbed mood, and reduced quality of life [2,3].The increased risk of traffic accidents due to sleepiness is an obvious individual and public health concern [4].Furthermore, OSAS is an independent risk factor for hypertension [5], myocardial infarction [6] and stroke [7].Epidemiological studies have suggested that 2 to 4% of the adult population suffer from OSAS [8] but the prevalence may have increased with the epidemic of obesity [9].The diagnosis of OSAS relies on typical symptoms including excessive daytime sleepiness and lack of concentration, habitual snoring, nocturnal choking, and witnessed apnoea [10].Male gender, obesity, a large neck size [11] and certain cranio-facial characteristics further enhance the suspicion of OSAS [12].The diagnosis is confirmed by a sleep study.
General treatment recommendations for patients with OSAS include sufficient and regular sleep hours (sleep hygiene), avoidance of smoking and alcohol consumption, and diet to reduce weight in obese patients.However, a persistent weight reduction is difficult to achieve and maintain, and behavioural modification is only minimally effective [13].The current standard treatment for OSAS consists in nocturnal application of continuous positive airway pressure (CPAP) via a nasal mask [14].Several randomised trials have established the effectiveness of CPAP in patients with various degrees of OSAS severity [15,16].Excessive sleepiness and other symptoms, quality of life, objective vigilance, driving simulator performance and other cognitive tasks are significantly improved by CPAP over baseline when compared to treatment with sham (placebo) CPAP or placebo medication [17,18].The effect size achieved with CPAP for several clinical outcomes is large or moderate, and is largest in the most severe cases of OSAS [17,19,20].In addition to providing symptomatic improvement, there is evidence that CPAP reduces fatal and nonfatal cardiovascular events in moderate to severe OSAS [6], and modifies cardiovascular risk since it reduces blood pressure [21] and circulating markers of cardiovascular risk such as cholesterol [22], C-reactive protein and interleukin-6 [23].Side effects of CPAP therapy are generally mild and reversible [16].
Current guidelines suggest treatment of symptomatic OSAS patients with more than 5 to 30 apnoea/hypopnoea per hour of sleep [24] depending Introduction on severity of symptoms and comorbidity.Generally, severely symptomatic patients with frequent apnoea/hypopnoea (>30 events/hour) readily improve with CPAP therapy.Conversely, in asymptomatic patients even with a markedly elevated apnoea/hypopnoea index (>30 events/h), there was no measurable benefit of CPAP therapy [25].In practice, a trial of CPAP therapy over a limited time period may help to identify patients who might benefit from long-term therapy [26].In a recent study we even found that a favourable response to a 2-week CPAP trial was more accurate in predicting OSAS patients successfully treated over >4 months than polysomnography, the conventional diagnostic "gold-standard" for OSAS [27].In summary, OSAS therapy is performed to improve symptoms and quality of life, and to prevent sleepiness-related accidents.Treatment of asymptomatic patients solely for correction of a laboratory abnormality, ie, for reducing an increased apnoea/hypopnoea index, or for reduction of the cardiovascular risk or prevention of potential disease progression is not established.
Despite its effectiveness CPAP therapy is used less then prescribed [28,29].This may relate to the inconvenience of the therapy, psychological factors, discomfort due to the pressure, skin irritation and other factors.In patients not able or not willing to perform CPAP therapy other treatment options have to be considered.The purpose of this article is to review the current alternatives to CPAP in the treatment of OSAS.
Removable oral appliances are attractive treatment options for patients with OSAS not able or not willing to tolerate the standard CPAP therapy [30][31][32][33].This is because oral appliances are easy to apply, handy, not dependent on electricity and thus particularly suitable for the use during travel.Furthermore, sleeping with an oral appliance is perceived as less socially disturbing than wearing a CPAP mask.The most effective oral appliances used today are designed to hold the mandible in an anterior position (protrusion).The so-called mandibular advancement devices (MAD) are fitted onto the lower and upper dental arches before going to sleep and removed in the morning (figure 1).They increase the upper airway lumen during sleep by protrusion of the mandible and tongue [34], increase the upper airway muscle tone [35], and reduce the passive pharyngeal wall compliance [36].
Several randomised studies, some of them using a sham appliance without mandibular advancement or medication as a placebo control, have demonstrated that MAD are effective in controlling symptoms of OSAS, nocturnal breathing disturbances, oxygenation and sleep disturbances, and even blood pressure [37][38][39].In one study, patients with mild sleep apnoea (less than 30 apnoea/hypopnoea per hour of sleep) were ran-domised to a sequence of 3 periods of 3 months with either a MAD, nasal CPAP or a placebo tablet [40].At the end of each 3-month period, the outcomes were compared.CPAP and MAD significantly improved symptoms to a similar degree.CPAP was more effective in reducing sleep disordered breathing, while a reduction in blood pressure was achieved with MAD only.Patients rated CPAP to be more effective than MAD but the latter to be more convenient.This may have contributed to a higher treatment adherence with MAD [40].In another study, MAD were shown to be effective in even severe cases of OSAS.The higher the number of apnoea/hypopnoea, the greater was their reduction by MAD [41].Side effects of MAD are relatively common but only rarely require discontinuation of therapy.They include tooth pain, hypersalivation, mucosal dryness, and occlusal changes [42].Potential limitations of MAD are the requirement of a minimal number of stable teeth (ie, at least 8 teeth in the upper and lower jaw), absence of gingival disease and temporo-mandibular joint pain.Individual adaptation by an experienced orthodontist in cooperation with a pulmonary physician is crucial for achieving an optimal therapeutic result.It may take several weeks and require repeated consulta- Monobloc mandibular advancement device as used at the University Hospital of Zurich.The custom-fitted appliance is snapped onto the dental arches for overnight treatment and held in place by metallic clasps.The protrusive forces acting onto the mandible are distributed over a large area of the alveolar process (modified from [33]).

Mandibular advancement devices
tions to complete the conctruction of MAD and one or several follow-up sleep studies to establish the desired effect.Whether MAD remain effective over several years, and whether there are relevant and potentially irreversible side effects such as orthodontic changes and damage to the mandibulomaxillary joints requires further study.It is important to realise that the favourable results reported in the cited studies [37][38][39] conducted with specific devices may not be extrapolated to the large number of other commercially available appliances which have not been subjected to a rigorous scientific evaluation.In conclusion, individually fitted and scientifically validated MAD are a valuable alternative for treatment of patients with mild and even severe OSAS if CPAP therapy is not feasible and if initial adaptation and long-term follow-up is performed by experienced professionals.
Obesity is a major risk factor for OSAS as well as for other diseases including hypertension and cardiovascular disease [9].Therefore, reduction of excessive weight is an important therapeutic goal independent of its beneficial effects on OSAS.While diet supported by pharmacological treatment has not resulted in major and persistent weight loss, favourable results have been achieved by bariatric surgery including adjustable gastric banding and gastric bypass surgery [43,44].A meta-analysis has evaluated the effects of bariatric surgery on OSAS [45].Among 12266 obese patients undergoing bariatric surgery and evaluated by sleep studies, OSAS was diagnosed in 2399 (20%).In 1636 of 1921 patients (85%) in whom the information was available, OSAS improved or was resolved.In addition to the favourable effects on sleep related breathing disturbances, bariatric surgery improved diabetes mellitus, hypertension and hyperlipidaemia in a substantial majority of the patients.Therefore, bariatric surgery has an important role in the treatment of obese OSAS patients.

Reduction of excess weight
Various techniques of soft tissue surgery for treatment of OSAS have been proposed but their role still remains controversial [46][47][48].An exception is adeno-tonsillectomy which is successfully performed in children and in adult OSAS patients with adeno-tonsillar hypertrophy [49].Uvulopalato-pharyngoplasty (UPPP) performed by conventional techniques or with the use of laser [50] has provided inconsistent and unpredictable results in regard to improving OSAS and snoring.In one study, temperature-controlled radiofrequency tissue ablation applied in local anaesthesia to the tongue base and palate over the course of several weeks was compared to the effects of a sham procedure in patients with mild OSAS (mean apnoea/hypopnoea index 20/h).The verum surgery improved subjective sleepiness, quality of life and objectively measured reaction time more than the sham procedure but the apnoea/hypopnoea index remained unchanged.Side effects included haematomas, ulceration, pain and difficulty swallowing for several weeks in some patients.The follow-up time was not mentioned [51].Tongue base procedures including suspension or resection have been performed in small patient groups and the results require further confirmation [52].Tracheotomy and maxillo-facial surgery are considered too aggressive to be recommended routinely as first-line therapy.The Stanford step-by-step approach for surgery in OSAS is considered in patients not successfully treated with CPAP [46,53].The first stage comprises limited mandibular osteotomy (with or without UPPP, genioglossus advancement, hyoid myotomy, and hyothyroidopexy).Maxillo-mandibular advancement osteotomy, stage II surgery, is considered if stage I is not successful, or in the first place if cranio-facial dysmorphia is present [54].In a single centre report on 51 OSAS patients treated surgically, stage I surgery was performed in 44 patients [54].In only 10 of them satisfactory improvement was achieved while 34 were treatment failures.In 13 of these patients maxillomandibular advancement osteotomy was subsequently performed and additional 7 patients with cranio-facial dysmorphias underwent maxillo-facial surgery in the first place.Maxillomandibular advancement osteotomy was successful in 15 of the 20 operated patients, 5 patients were treatment failures.Over the same time period in which the 51 OSAS patients underwent upper airway surgery, 939 patients were started with CPAP at this center.Thus, the staged surgical concept seems to be successfully applicable in a very minor fraction of OSAS patients, namely those not successfully treated with CPAP or a MAD, and those with significant adeno-tonsillar hypertrophy, or cranio-facial dysmorphias and other anatomical obstacles amenable to surgery.

Upper airway surgery
It has been a long-standing clinical observation that snoring is particularly common in patients with nasal obstruction, and early reports have shown that experimental nasal occlusion promotes obstructive sleep apnoea [55].Whether nasal obstruction due to chronic rhinitis, nasal polyposis or nasal septal deviation is a predisposing factor for OSAS is not clear.Epidemiological studies have shown that chronic rhinitis symptoms, and increased nasal resistance measured by rhinomanometry are associated with habitual snoring but a similar association was not demonstrated for OSAS [56,57].Nevertheless, treating OSAS patients with chronic rhinitis with fluticasone administered intra-nasally for one month im-proved sleepiness, and reduced the apnoea/hypopnoea index with statistical significance, though only to a minimal degree compared to placebo [58].In another non-randomised study, 19 OSAS patients and 7 snorers with impaired nasal breathing underwent nasal surgery [59].Sleep related breathing disturbances were not significantly changed but patients reported being less sleepy after the intervention.Nasal obstruction appears therefore to have a minor role in the pathophysiology of OSAS.Treating impaired nasal breathing may still be beneficial in selected patients since it improves subjective symptoms and sleep quality [60] and may contribute to successful nasal CPAP therapy in OSAS patients.

Treatment of nasal obstruction
Unfortunately there are currently no drugs that allow effective pharmacological therapy of OSAS.Research in this field however is ongoing.Recently, interesting observations have been made in children with minimal adeno-tonsillar enlargement and very mild sleep disordered breathing [66].They were treated for 16 weeks with the leukotriene modifier montelukast achieving reductions in adenoid size and in sleep related breathing disturbances.Although these effects were modest, the results are promising.In compliant OSAS patients with residual hypersomonolence despite exclusion of other causes and effective CPAP treatment, modafinil, a drug prescribed to treat hypersomnolence in narcoleptics, has been used as an adjunct to improve alertness.In a randomised placebo controlled cross-over study with a 2 week period on modafinil no improvement in subjective sleepiness assessed by the Epworth score, no change in the multiple sleep latency test (MSLT) and only minor improvement in the ability to stay awake in a sleep-seductive environment (maintenance of wakefulness test) were found [67].In another placebo-controlled parallel trial extending over 4 weeks with a final daily dose of 400 mg of modafinil, a statistically significant but clinically minor improvement in objective vigilance measured by the MSLT was demonstrated [68].During extended open label use of modafinil over 4 months, patients continued to perceive an overall benefit in regard to improved sleepiness and quality of life [69].Side effects of modafinil were generally mild and most commonly consisted of headache (28%), anxiety (16%), nervousness (14%), insomnia (11%), and nausea (11%) [69].Despite the relatively favourable results regarding residual sleepiness the use of modafinil is still controversial.Arguments in favour of modafinil in this setting are some subjective and objective improvements of alertness and of quality of life.Con-versely, the use of a stimulant may reduce compliance with CPAP treatment as shown in the cited studies [67,69] and the stimulant drug may divert from inappropriate functioning of CPAP therapy and expose the OSAS patients to increased cardiovascular risk.
In mild positional OSAS, sleep in lateral position is often recommended.One randomised trial has evaluated the effect of 2 weeks of positional treatment consisting of a backpack with a soft ball inside.Thirteen patients with positional OSAS defined by more than twice the number of apnoea/hypopnoea in the supine as compared to the lateral position were included [70].Positional training improved subjective sleepiness, maintenance of wakefulness time measured objectively, and psychometric test performance to a similar degree as CPAP but the latter was more effective in reducing apnoea/hypopnoea and oxygen desaturations.Therefore, positional treatment is a reasonable therapy for selected patients with positional OSAS who are not tolerating CPAP.A nasal dilator [71] and a number of other appliances have also been promoted for treatment of snoring and OSAS but have not revealed consistent effects and can therefore not be recommended.
Nocturnal electrical stimulation of the hypoglossal nerve by an implanted pace-maker has been thought to prevent sleep related upper airway collapse in OSAS patients by activating submandibular muscles [61].However, this treatment is still experimental and there is insufficient evidence to support its clinical use.In one randomised, placebo-controlled trial, tongue muscle training by electrical neurostimulation applied twice during daytime for 8 weeks has been found to reduce snoring but not sleep apnoea in 33 patients [62].Whether this treatment modality is acceptable and effective in the long-term treatment of snoring remains open.The initial success

Drug therapy, adjunctive and experimental measures
achieved with atrial overdrive pacing as a therapy for OSAS patients treated with a cardiac pacemaker for other reasons has not been confirmed in subsequent trials, and this treatment is therefore not recommended [63,64].Whether biventricular pacing in heart failure patients with ventricular asynchrony improves co-existing sleep apnoea requires further studies [65].
Figure 2 summarises the current options for treatment of OSAS and the suggested sequence of evaluations.CPAP remains the standard treatment for the vast majority of OSAS patients due to its immediate and persistent effectiveness and the lack of major side effects.For patients in whom CPAP therapy is not feasible because of mask intolerance or other reasons, a custom-fitted MAD may be a valuable and effective alternative treatment but long-term effectiveness and side effects need to be monitored.In morbidly obese OSAS patients, bariatric surgery should be considered as an option to treat both excess weight, and OSAS and to prevent cardiovascular consequences.Upper airway surgery has a role in children and adults with enlarged adenoids and tonsils, or with cranio-facial dysmorphia and other selected patients in whom CPAP is not an option.

I
thank Karsten Fritsch, MD, for careful review of the manuscript.Correspondence: Konrad E.Bloch, MD Pulmonary Division, Dept. of Internal Medicine University Hospital of Zürich Rämistrasse 100, CH-8091 Zürich Switzerland E-Mail: pneubloc@usz.unizh.chConclusions : treat with topical steroids, consider nasal surgery Cranio-facial malformation: consider maxillo-facial surgery Adeno-tonsillar hypertrophy: consider adeno-tonsillectomy CPAP unacceptable or not feasible >8 upper and lower teeth, no temporo-mandibular joint pain: consider mandibular advancement device Evaluate: CPAP use, mask fit/leak nasal obstruction sleep hygiene alternative sleep disorder Consider modafinil as adjunct for residual sleepiness

Figure 2
Figure 2Treatment options for the obstructive sleep apnoea syndrome (OSAS).The standard therapy consisting of nasal continuous positive airway pressure (CPAP) and alternative treatment modalities along with suggested evaluations are outlined.