Abstract
Snoring and obstructive sleep apnoea (OSA) are increasingly common conditions, and confer a significant health and socioeconomic burden. Furthermore, untreated OSA represents a significant mortality risk. Patients require careful assessment, including detailed clinical history and examination, sleep study and drug-induced sleep endoscopy (DISE). Although nasal continuous positive airway pressure (nCPAP) is the gold standard treatment for moderate and severe OSA, multidisciplinary team assessment is often required to develop the best treatment plan for an individual, especially when nasal CPAP is poorly tolerated. There is a wide range of medical and surgical treatment options, and following appropriate patient selection and assessment, a focused site-specific, often multilevel, intervention is indicated. There is an increasing body of evidence in the literature supporting these multilevel interventions and with agreement on standardized outcome measures more trials are likely to improve the robustness of these data further.
Keywords
Introduction
Sleep-related breathing disorders encompass a broad spectrum of conditions, including obstructive sleep apnoea (OSA), where recurrent partial or complete obstruction of the upper airway occurs. The prevalence of OSA is approximately 2–4% in men and 1–2% in women.1–5 It has been established that OSA is an independent risk factor for cerebrovascular and cardiovascular disease and can consequently lead to significant morbidity and mortality.6,7 Alongside the health-related impacts, OSA can also confer a considerable social and economic burden for those affected, 8 and it has been suggested that the daytime sleepiness and resultant impaired cognitive function may contribute to job-related and motor vehicle accidents.9,10 Separately, moderate-to-severe OSA has been reported to independently significantly increase the risk of all-cause mortality. 11 The disease burden associated with OSA is only likely to rise as worldwide obesity levels within the population are increasing. 12
Paediatric OSA is mainly thought to be due to adenotonsillar hypertrophy and so the gold standard treatment for this condition is adenotonsillectomy.13,14 However, adult OSA has a multifactorial aetiology and often arises as a result of multilevel anatomical obstruction. Consequently, continuous positive airway pressure (CPAP) remains the gold standard treatment for moderate-to-severe OSA in the adult population in the UK. 15 Whilst CPAP is clearly an effective treatment option for some patients, compliance and adherence, particularly long term, can be poor and has been estimated to be as low as 40–85%.16–18 Clearly, in this context other nonsurgical treatments require consideration such as lifestyle modification, sleep position training and oral appliances. In some patients, however, these measures are either ineffective, insufficient or they cannot tolerate them long term. In this disparate group of patients, surgery may play a role; to either facilitate the use of CPAP, or as a treatment modality in itself. In this narrative review article, we will explore the range of conservative, medical and surgical treatments available for the treatment of adult OSA.
Clinical evaluation
Clinical evaluation of the patient is key to implementing appropriate investigations and an effective management plan. Patients will usually present to clinic with their partner, or at the behest of their partner. Patients with OSA will also usually mention episodes suggestive of apnoea, complain of daytime somnolence, morning headaches, poor concentration, restless sleep and feeling unrefreshed in the morning.
Risk factors should also be assessed in the clinical evaluation of a patient. Obesity is a major risk factor; not only is excess fat around the upper airway thought to increase the likelihood of airway collapse, but obesity has also been shown to affect lung volumes and therefore respiratory stability by causing a significant reduction in the residual functional capacity when asleep. 19 Male sex also predisposes to the development of OSA, with higher rates of central obesity and longer length airways thought to be possible contributing factors. 20 Age 21 and smoking 10 are also common risk factors. Skeletal framework abnormalities should be noted as retrognathia would compromise the hypopharynx and maxillary retrusion would similarly affect the retropalatal space. Obvious external deformity of the nose may have an impact on upper airway obstruction by increasing the flow limitation index.
OSA can lead to significant metabolic and cardiovascular disease. It is associated with systemic hypertension, 22 myocardial infarction and stroke.23,24 There is a complex relationship between obesity, diabetes and OSA, but despite this, OSA has been shown to be an independent risk factor for diabetes mellitus. 25
It is important to enquire about rhinological symptoms, as various nasal conditions can affect the patient’s ability to breathe through their nose at night and also potentially negatively impact on the use of nasal CPAP. Furthermore, patients with mouth-breathing have further retraction of the mandible which compromises the hypopharyngeal dimensions and can also exacerbate palatal and pharyngeal vibrations. Finally, an assessment of the patient’s comorbidities, and drug history should be taken, noting alcohol intake and any use of sleeping tablets such as benzodiazepines.
A useful adjunct to this history is the use of screening tools, of which a variety exist. One such tool is the Epworth sleepiness score (Table 1), and although this is a tool designed to explore daytime sleepiness, it has been shown that a patient score of greater than 10 correlates with a higher likelihood of significant OSA. 26 An alternative is The STOP-Bang questionnaire, which is a disease-specific validated screening tool for OSA. A recent meta-analysis confirmed sensitivities ranging from 90–96% for mild-to-severe OSA, respectively. 27
Epworth Sleepiness Scale 26 : How likely are you to doze off or fall asleep in the following situations, in contrast to just feeling tired? Please score as follows: 0 = would never doze; 1 = slight chance of dozing; 2 = moderate chance of dozing; 3 = high chance of dozing.
Clinical examination should involve a detailed examination specifically looking for evidence of retrognathia, craniofacial abnormalities and dental malocclusion. Assessment of a patient’s body mass index (BMI) is particularly important, as this has been shown to correlate with surgical outcomes. Flexible nasopharyngolaryngoscopy is particularly helpful to fully view the pharynx, tongue base and larynx, and also allows for dynamic assessment of the airway to determine levels of airway collapse, particularly when undertaken with Muller’s manoeuvre or simulated snore, although these evaluations have proven difficult to standardize. 28 As part of the oral cavity examination, the Friedman or Mallampati grading for tongue position is noted, as this can help to predict the likelihood of OSA.29–31
Investigation of obstructive sleep apnoea
Hospital-based polysomnography is the optimal investigation for patients with suspected OSA, however, this is both expensive and inconvenient for the majority of patients. Most patients are assessed using an ambulatory sleep study that is performed in the comfort of their own home. This will record oxygen saturation, nasal and oral airflow, as well as respiratory effort. From these parameters, an apnoea/hypopnoea index (AHI), mean oxygenation, oxygen desaturation index, and flow limitation index can be calculated. The American Academy of Sleep Medicine Task Force defined ‘apnoea’ as a cessation of airflow for at least 10 seconds and a ‘hypopnoea’ episode as a reduction in tidal volume or vital capacity by at least 30%. 32 The AHI score has been used to classify the severity of the OSA, with mild OSA as 5 < AHI < 15, moderate as 15 < AHI < 30, and severe OSA as an AHI > 30. 32 Respiratory disturbance index (RDI) is synonymous with AHI but also consists of respiratory-effort-related arousals (RERAs). RERAs are arousals from sleep but technically do not link with definitions of apnoea and hypopnea but do disrupt sleep. 33
Assessment of the upper airway in awake patients has limited usefulness, as the level and patterns of airway obstruction can be different in the sleeping patient as a result of variation in the muscle tone. The concept of drug-induced sleep endoscopy (DISE) was first introduced by Croft and Pringle in 1991, 34 to allow the dynamic assessment of upper airway obstruction in a ‘drug-induced sleep’. This allows for a targeted and tailored approach to treatment, and may help us to understand why CPAP fails in some patients. Critics of the technique, however, have claimed that natural physiological sleep differs from drug-induced sedation, and argue that assessment is subjective, with a lack of standardization. Conversely, advocates of DISE argue that sedation affects all the anatomical areas evenly allowing meaningful evaluation of the airway. Additionally, there have been efforts, at a recent European Congress, to standardize this assessment. 35 In most centres, DISE remains the evaluation technique of choice as it allows visualization of three-dimensional anatomy of the obstructive upper airway during sleep, albeit drug induced, and is preferred to other methods such as imaging or using pressure transducers.
Treatment options for obstructive sleep apnoea
Limitations
There is a lack of high-quality evidence in this field, particularly in relation to the surgical interventions but this is true of surgery in general. Furthermore, there are no agreed standard outcome measures by which to judge surgical outcome. This is further compounded by the fact that surgery can be undertaken as a primary treatment or as an adjunctive procedure to improve CPAP compliance. Certainly, given the complexity of the decision making in these patients, the management strategies for individual patients should be planned in multidisciplinary teams with representation from sleep medicine (respiratory physicians), orthodontics, maxillofacial and otolaryngology surgeons. Finally, management should be in an incremental stepwise fashion (detailed below), with regular reassessment after each intervention by sleep study and DISE.
Nonsurgical treatment of obstructive sleep apnoea
CPAP
In the UK, the National Institute of Clinical Excellence (NICE) guideline recommends nasal CPAP for patients with moderate or severe OSA. The American College of Physicians also recommends CPAP as a first initial therapy for OSA, 36 and hence this is considered the gold standard treatment worldwide. For a more in-depth understanding of CPAP for the treatment of OSA, please see the review by Virk and Kotecha. 37
Lifestyle modification
Many patients with OSA have an increased body mass index (BMI). It has been shown that appropriate weight loss can reduce the incidence of significant OSA.36,38 Consequently, the first intervention in these patients is to suggest controlled, maintained weight loss. 36 The American College of Physician guidelines looked at various studies examining the effect of weight loss strategies in OSA patients with a BMI > 30kg/m2,39–41 and concluded that intensive weight loss programmes in this subset of patients may be beneficial in reducing the severity of OSA. An increased BMI does correlate with surgical outcomes, and whilst there are no nationally or internationally agreed guidelines regarding a BMI cut off for surgery, local guidelines within our own institution state that patients will not be considered for surgery in our institution unless their BMI is <35 kg/m2 but preferably <32 kg/m2. Additionally, alcohol consumption has been linked to worsening existing OSA and so, recommending that the patient reduces (or abstains) their alcohol intake can be a useful intervention.42–44
Sleep position can have a significant effect on the severity of the OSA, with up to 55.9% of patients with OSA having positional OSA syndrome. 45 Sleep positioning therapy can also be highly effective in those patients who have a significant supine element to their OSA.46–49 This can be as simple as stitching a tennis ball into the back of one’s pajamas, and although effective, was very poorly tolerated long term. 50 More recently, several devices have been designed to limit the supine position whilst asleep. The vibro-tactic neck-based device (‘Night Shift’) has been shown to be effective to reduce supine sleeping and leads to a reduction in AHI. 51 The effect of the prone sleep position has also been explored recently with the development of a customized mattress and pillow, which reduced the median AHI from 23 to 7 events per hour in those patients who exhibited a positional element to their OSA. 52
Medical treatment
A recent Cochrane review concluded that there was insufficient evidence for recommending drug therapy for the treatment of OSA, 53 largely due to small study numbers and hence no meaningful conclusions could be drawn. Although no drugs have yet been convincingly developed to treat OSA, several efforts have been made to address the breathing disorder itself. Increasing the arousal threshold whilst not affecting muscular tone in the upper airway has been posited as one potential pharmacological strategy for the treatment of OSA. In this context, Zopiclone has been evaluated with one study reporting possible benefit, although acknowledging that patient selection is crucial as incorrectly selected patients can suffer hypoxia. 54 Conversely, another study did not show any benefit. 55 Similarly, trazodone has been assessed with promising results, initially, in two studies56,57 and then in a clinical trial, 58 but further work is needed before this agent could be used more widely. Stabilization of ventilation has been seen as another potential strategy for improving OSA, and acetazolamide has been shown to be useful in this context,59,60 however, again, there are no trials with this drug outside of the high-altitude setting. Another approach has been to improve upper airway muscle stability with desipramine which has been reported by one group with promising results. 61 More speculatively, others have reported on L-tryptophan as a potential agent for improving sleep.62,63 Finally, several investigators have attempted to delineate drug discovery strategies or identify therapeutic targets for future medical treatments of OSA.64–66
Nasal conditions such as allergic and nonallergic rhinitis, nasal polyposis, and CPAP rhinitis, make it difficult for patients to breathe through their nose at night, or indeed, tolerate a nasal CPAP mask. Interestingly, patients with OSA have significantly more rhinitis than those people who do not suffer with OSA. 67 Medical interventions to improve nasal patency at night can improve sleep quality, however, improvements in specific objective sleep parameters have not yet been identified. 68 In those patients with an obstructive nasal passage, treatment of this condition has also been shown to improve the efficacy of CPAP and patients’ compliance with it. 68 Thus allergic or nonallergic rhinitis (including CPAP-induced rhinitis) should be treated, with a combination of allergen avoidance, saline nasal douches, intranasal steroids and antihistamines. Other coexisting conditions should be excluded (or treated, if present) such as hypothyroidism, and medical treatment of obesity. 69
Appliances
Simple appliances such as nasal dilators can help to increase nasal airflow at night. These come in various forms and all function by holding open or preventing collapse of the internal nasal valve region and thus improving nasal airflow. These can be effective in patients who are ‘simple snorers’ with nasal obstruction only and if they are successful then surgery to address the internal nasal valve should be considered. 70
In appropriately selected patients, mandibular advancement splints (MASs) are an effective adjunct to CPAP, or as a treatment alone for mild-to-moderate OSA. 71 MASs are specifically indicated in those patients who are retrognathic or who on DISE, are seen to have significant tongue base collapse. 72 This functions by protruding the hyoid and mandible anteriorly, thereby stabilizing the tongue base and preventing it from occluding the airway when the patient sleeps. MASs have been shown to be effective at improving OSA, 72 but similar to CPAP, compliance can be poor due to discomfort and are only tolerated by 60–70% of people. 73 Furthermore, they are contraindicated in patients who have poor dentition or temporomandibular joint problems. 74
Surgical treatment of obstructive sleep apnoea
Surgical treatment for OSA has traditionally been considered of limited benefit. The ‘one size fits all’ approach of a radical uvulopalatopharyngoplasty for all sleep-disordered breathing has quite rightly come under scrutiny in the past. However, with the advent of DISE assessment, surgical intervention has become much more focused and targeted, resulting in improved outcomes for patients.75–82 It is hugely important, therefore, to identify and accurately select which patients might be suitable for surgical intervention and address the need for weight loss prior to embarking on surgical management. 36 In our institution, surgery is undertaken in three groups of patients. First, in those who have mild OSA (AHI < 15) but who snore and have raised flow limitation indices (FLI) of greater than 15%. Second, in those who have moderate OSA (15 < AHI < 30) or severe OSA (AHI > 30) who cannot tolerate CPAP or who have failed a trial of MAS. In these patients, surgery is performed as an adjunct to CPAP with the aim of improving the airway, and thus reducing the CPAP pressures required, facilitating improved compliance. Thirdly, surgery can be undertaken in patients with moderate or severe OSA, who cannot tolerate CPAP, with curative intent aiming to improve their AHI to less than 15. In all these groups of patients, DISE is used to assess the airway and helps in deciding the appropriate surgical intervention,34,83–85 which can be a single-site procedure but more commonly, it is multilevel.
Nasal surgery
The aim of nasal surgery, in its various forms in the context of OSA, is to help the simple snorer (AHI < 5, FLI < 15%) or to improve CPAP compliance. 86 There is no evidence that nasal surgery alone will significantly improve AHI score but a meta-analysis has shown that nasal surgery can reduce CPAP pressure requirements, thereby improving discomfort levels and hence, compliance. 86 The nasal surgery performed can vary greatly and depends on the anatomical issues present. Commonly performed procedures include; septoplasty, reduction of inferior turbinates, endoscopic sinus surgery, septorhinoplasty or nasal valve surgery.
Palate surgery
Palatal surgery (Figure 1) in its many forms is the most commonly performed surgery in patients with OSA. Generally, the vogue is to approach the surgery in a stepwise fashion, starting with less aggressive interventions and moving as required to more radical solutions. Obviously, assessment of the palate both during wakefulness and whilst performing DISE is essential. Simple observations such as Friedman’s tongue position can be very informative. For example, patients with position 1 will often benefit from palatal surgery whereas positions 3 or 4 are less likely to benefit. 87 More generally, the surgery can be divided up into procedures that will stiffen the soft palate such as chemical injections (e.g. sodium tetradecyl), radiofrequency treatments or palatal implants; or into procedures that alter the shape of the soft palate. Although in the past, the evidence for these less aggressive procedures has been poor,88–90 there is now more compelling evidence supporting the use of radiofrequency palatal surgery (Figure 2) in the treatment of both snoring and OSA.79,80,91 More recently, radiofrequency applications to the soft palate have indeed been approved by NICE for use in the UK. 90

Pre- and postoperative views of the oropharynx following palatal surgery.

Diagram showing radiofrequency application to the palate. Ten applications are shown: four median and six lateral, each 10 watts.
If at evaluation, the patient has excessive redundant tissue or they have failed a simpler intervention previously, they are likely to require more aggressive surgery. Fujita first characterized the uvulopalatopharyngoplasty 92 (UPPP), although there have been several variations on this operation since. This procedure has been criticized in the past for high rates of postoperative nasopharyngeal stenosis and velopharyngeal insufficiency. As a result, at our institution, we have moved to using the Kotecha technique laser-assisted uvulopalatoplasty78,93,94 (LAUP). When there is less redundant tissue in the patient’s soft palate, the procut palatoplasty technique is often performed in combination with interstitial radiofrequency to the palate. Coblation of the palate can also serve as a good adjunct to procut palatoplasty. 95
It is particularly important to reassess patients in the form of a sleep study and repeat DISE after each intervention, as previous surgical interventions may alter the mechanics and dynamics of the upper airway. In those patients where the AHI has failed to sufficiently improve following a LAUP, and in who the palate is still contributing to airway obstruction, the modified Ellis LAUP 96 may be of benefit. Other surgical options for a primary palatal issue includes expansion sphincter palatoplasty, palatal advancement flaps, and pillar implant surgery have been shown to help improve AHI scores.97,98
Tongue base surgery
Tongue base collapse is a common finding on DISE assessment, and there may also be an associated epiglottic contribution. 99 Various surgical strategies exist, but suboptimal transoral access to the tongue base (and epiglottis) remains a significant issue and is therefore challenging to address. Minimally invasive techniques such as radiofrequency thermotherapy or coblation channelling of the tongue base act to reduce the bulk of the tongue and to stiffen it.76,79,91,100,101 If these procedures fail to help, midline glossectomy, genial tubercle advancement and hyoid suspension are alternative operations that have been reported in the literature with mixed results.75,102,103 Transoral robotic surgery (TORS) allows for significantly improved visualization of the anatomical areas in question, combined with multiplanar tissue manipulation at any angle. Studies looking at robotic assisted-laser reduction of the tongue base and laser resection of the epiglottitis (Figure 3) have shown promising results,104–106 with significant improvements in mean AHI. Arora and colleagues reported an overall 51% reduction, with normal sleep study results in 36% postoperatively 99 and more recently, a meta-analysis has also shown a significant reduction in AHI, along with improvements in Epworth scores and snoring visual analogue scale scores. 107

Postoperative view of the larynx following transoral laser resection of the epiglottis.
Hypoglossal nerve stimulation has been attempted to improve upper airway muscle tone during inspiration and although this is undoubtedly an interesting approach, more work is needed before it could be a mainstream treatment option. Certainly, some worrying complications have been reported, such as infections associated with the device requiring explantation, as well as device malfunction, necessitating replacement.108,109
Other surgery
Tracheostomy remains the gold standard cure for OSA with the highest success rates but clearly this is a major intervention. Indications include those patients who refuse CPAP and have failed other surgical treatment modalities, or can be used as an airway adjunct in order to safely perform other surgical procedures. 110 Maxillomandibular advancement surgery has been shown nearly as good as tracheostomy at improving snoring and OSA.111–113 However, the patient must adhere to a soft diet for 2 months postoperatively and serious complications have been reported. 114 Obesity is a major factor in the development of OSA and as a result, bariatric surgery is gaining popularity as a mode of treatment. However, there are no compelling data in the literature showing that it can significantly improve AHI scores but it certainly has been reported to improve quality of sleep. 115
Although all the above medical and surgical options have been discussed in turn, in actuality, these treatment options are often deployed together or sequentially after repeated DISE assessment. Multilevel surgery is now commonly performed, particularly minimally invasive surgery such as radiofrequency treatment to palate, tongue base and inferior turbinates, which can be combined with tonsillectomy, if necessary. Similarly, whilst performing TORS, one may address the base of the tongue and the epiglottis simultaneously. Table 2 demonstrates the surgical techniques for OSA and Table 3 demonstrates our local policy in terms of decision making for certain anatomical and dynamic airway findings.
Anatomical site and associated surgical technique.
Surgical decision making for obstructive sleep apnoea.
Performed more commonly in women.
Performed more commonly in men.
Conclusions
Snoring and OSA are common conditions with significant morbidity. Careful assessment of these patients is required, including clinical examination, sleep study and DISE. After multidisciplinary team discussion, medical and surgical interventions are frequently used in combination and progressive stepwise multilevel treatments may be adopted after repeat DISE assessment. CPAP remains the gold standard treatment for moderate/severe OSA but other interventions can help improve compliance of nasal CPAP or abrogate the need for it. Although there is increasing evidence for surgical procedures in the context of snoring and OSA, more studies are required, particularly if standardized outcome measures can be agreed.
Footnotes
Acknowledgements
MS Ferguson and JC Magill made equal contribution to this work.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement
The authors declare that there is no conflict of interest.
