Introduction
Obstructive sleep apnoea (OSA) is a disorder of breathing during sleep, characterised by prolonged partial upper airway obstruction, and/or intermittent complete obstruction (obstructive apnoea), which disrupts ventilation during sleep and normal sleep patterns.Reference Marcus, Chapman, Ward, McColley, Herrerias and Stillwell1 Between 6 and 12 per cent of children are quoted to show significant snoring, and 1–3 per cent have OSA.Reference Sharma, Kanona, Kumar and Kotecha2 Adenotonsillectomy is the most common treatment for children with OSA and generally the first line of therapy.
A recent UK survey investigating the management of paediatric snoring and OSA revealed great heterogeneity in the assessment and management of patients, with only 25 per cent of respondents having a dedicated management protocol for paediatric OSA.Reference Sharma, Kanona, Kumar and Kotecha2 In addition, it revealed low follow-up rates, with a mean duration of 6.8 ± 1.2 weeks. This creates uncertainty as to whether the treatment of adenotonsillectomy we are currently performing for OSA actually works and is beneficial in the long term.
Controversy also remains regarding the post-operative management of these patients. For instance, do they require an overnight stay on the ward or paediatric high dependency unit, and can we identify which patients are at increased risk of post-operative cardiorespiratory complications?
Recent surges in research activity have helped us to build on and improve our knowledge base to answer these pertinent questions. This systematic review aimed to assess: the evidence base for tonsillectomy versus tonsillotomy in the management of paediatric OSA, anaesthetic considerations, the evidence for the benefits of treating OSA surgically, and the medical treatments available for residual OSA after adenotonsillectomy.
Materials and methods
In preparation of this systematic review, we followed the guidelines of the Preferred Reporting Items for Systematic reviews and Meta-Analyses for Protocols (‘PRISMA-P’) 2015. We assessed the quality of papers using the National Institute for Health and Care Excellence quality assessment tool, as it allows some analysis of case reports and series (lower level evidence).
We sought high-quality prospective clinical studies, reviews or laboratory work relevant to the definition, diagnosis, classification and non-surgical management of paediatric OSA. We searched Medline, Embase and Cochrane Library databases from inception up to March 2017, using the following key words and combinations thereof: ‘obstructive sleep apnoea’, ‘obstructive sleep apnea’, ‘paediatric’, ‘pediatric’, and ‘controlled trials/systematic review/evidence base’.
Abstracts, identified from a review of article titles, were evaluated for inclusion by two authors (LP and KB) working independently, with consensus if opinions differed. Papers were chosen if the abstracts suggested systematic reviews or meta-analyses, prospective controlled studies, or original basic science findings from laboratory studies. Audits, and larger case series and cohort studies, especially those offering any comparative study between groups, or before and after intervention, provided a lower level of evidence. Papers suggesting algorithms and consensus views for treatment, together with case reports suggesting complications and the earliest reports of the condition were also included.
Abstracts were excluded if they suggested isolated case reports or small and uncontrolled series that presented no new insight. No language restrictions were applied.
Results
We selected 35 papers that met the criteria for this review (and which are referred to in the text): 18 systematic reviews;Reference Marcus, Chapman, Ward, McColley, Herrerias and Stillwell1, Reference Leong and Davis3–Reference Manickam, Shott, Boss, Cohen, Meinzen-Derr and Amin19 5 randomised controlled trials;Reference Marcus, Rosen, Davidson Ward, Halbower, Sterni and Lutz20–Reference Quante, Wang, Weng, Rosen, Amin and Garetz24 3 non-randomised controlled or comparative studies;Reference Bar, Tarasiuk, Segev, Phillip and Tal25–Reference Tarasiuk, Simon, Tal and Reuveni27 6 case series or cohort studies;Reference Goldstein, Post, Rosenfeld and Campbell28–Reference Hoey, Foden, Hadjisymeou Andreou, Noonan, Chowdhury and Greig33 3 surveys, consensus documents or questionnaire studies;Reference Sharma, Kanona, Kumar and Kotecha2, Reference Nardone, McKee-Cole and Friedman34, Reference Chng, Goh, Wang, Tan and Ong35 0 historical articles; and 0 case reports.
Discussion
Question one
Tonsillectomy or tonsillotomy? Does adenoid removal confer much added benefit?
Adenotonsillectomy for OSA is the most commonly performed procedure and treatment of choice for paediatric OSA, to reduce the obstruction caused by large tonsils and adenoids. There is good evidence that adenotonsillectomy is efficacious in children with OSA and adenotonsillar hypertrophy. The majority of children show marked improvement in their polysomnography parameters, with a success rate, in otherwise healthy and non-obese children, of 75 per cent.Reference Dehlink and Tan13 In addition, this treatment is associated with a significant reduction in healthcare utilisation.Reference Tarasiuk, Simon, Tal and Reuveni27 Recent research activity and debate has focused on whether tonsillotomy is a safer alternative to the traditional approach of extracapsular tonsillectomy and how much added benefit is offered by adenoidectomy.
Tonsillectomy versus tonsillotomy
Tonsillotomy, also known as subtotal tonsillectomy, intracapsular tonsillectomy or partial tonsillectomy, aims to remove approximately 90 per cent of tonsillar tissue, leaving a thin rim laterally to avoid disruption of the tonsillar capsule and prevent damage to the surrounding pharyngeal muscles. This proposes to reduce the severity and duration of post-operative pain, with less frequent post-operative bleeding and an efficacy equal to that of total tonsillectomy for the treatment of sleep-disordered breathing.Reference Wood, Cho and Carney8, Reference Wilson, Merer and Moscatello21
A meta-analysis comparing 10 prospective studies (1029 participants) directly compared tonsillotomy and tonsillectomy in the management of paediatric sleep-disordered breathing.Reference Wang, Fu, Feng, Guan and Yin10 It found that tonsillotomy may be advantageous over tonsillectomy in short-term measures, with lower haemorrhage rate, shorter operation time and faster recovery to a pain-free state. Long-term follow up showed no clinically significant differences between the tonsillotomy and tonsillectomy groups in terms of: resolution of upper airway obstructive symptoms (including polysomnography outcomes), quality of life or post-operative immune function. However, a subgroup analysis showed a difference with regard to the rate of sleep-disordered breathing recurrence between the tonsillotomy and tonsillectomy groups, where the risk ratio of sleep-disordered breathing recurrence was found to be 3.33 (95 per cent confidence interval = 1.62, p = 0.001), favouring tonsillectomy at an average follow-up time of 31 months.
A retrospective cohort study of 52 children comparing extracapsular and intracapsular tonsillectomy with adenoidectomy found comparable short-term outcomes (improved post-operative polysomnography findings) with both treatment modalities.Reference Mukhatiyar, Nandalike, Cohen, Sin, Gangar and Bent32 However, in a subset of patients with asthma and obesity, intracapsular tonsillectomy was associated with residual OSA. Obesity and other co-morbidities such as asthma are independent risk factors for the development of OSA, and, when they co-exist, the additive effects can contribute to treatment failure. There is thus a need for more extensive studies, especially into the long-term follow up, to allow better assessment of the maintenance of treatment effects.Reference Anderson, Holm and Homoe14, Reference Ross, Storfer-Isser, Hart, Kibler, Rueschman and Rosen31, Reference Chng, Goh, Wang, Tan and Ong35
There have been recent advances in Coblation® technology. In addition, Hoey et al. reported on its positive application in tonsillotomy for OSA.Reference Hoey, Foden, Hadjisymeou Andreou, Noonan, Chowdhury and Greig33 These authors described a case series of 500 children (mean age of 5.1 years) who underwent Coblation intracapsular tonsillectomy for obstructive and/or infective symptoms. Long-term follow-up data (pre- and post-operative parent-reported outcome questionnaire findings) revealed highly significant improvements in the scores post-operatively in patients with obstructive symptoms, with minimal pain control requirements and low haemorrhage rates (there were no primary bleeds, and two patients had a secondary bleed, managed conservatively). Of the patients, 2.4 per cent required revision tonsil surgery (10 for obstructive symptoms, 2 for infective symptoms), with the majority being very young children. Further studies with greater sample sizes would help to establish the role and advantages of Coblation intracapsular tonsillectomy for OSA treatment.
Adenoidectomy alone?
Although adenotonsillectomy is successful in treating OSA, there are concerns regarding the risk of post-operative bleeding in children aged less than three years. In addition, the role of adenoidectomy alone, in the presence of small tonsils, has been investigated. A systematic review and meta-analysis yielded a total of three studies of adenoidectomy alone with pre- and post-operative sleep study data (47 patients, level 4 evidence).Reference Reckley, Song, Chang, Cable, Certal and Camacho15 It concluded that adenoidectomy alone can significantly improve obstructed breathing during sleep, as determined by the improved apnoea/hypopnoea index and lowest oxygen saturation level. However, given the small number and low quality of the studies, adenoidectomy alone is an area for additional research rather than being the current standard practice in children.
Summary points
Current screening tests show a relatively high sensitivity but a low specificity. Most are better at detecting moderate-to-severe rather than mild OSA. Emerging evidence suggests that even mild forms of OSA (and possibly even primary snoring) carry associated morbidity. There are limited studies investigating the threshold at which the apnoea/hypopnoea index predicts morbidity, focusing on cardiac and neurocognitive disorders. The ability of screening tests other than polysomnography to predict post-operative complications after adenotonsillectomy is unclear.
Question two
Anaesthetic management. What level of post-operative care is required? What are the indications for paediatric intensive care unit admission and ventilation? Can it be predicted?
Children with OSA are at risk of respiratory complications following adenotonsillectomy, so it is important that they are appropriately identified and pre-assessed.Reference Wood and Besecker11 Children with an apnoea/hypopnoea index of 5 or more events per hour are seven times more likely to have a post-operative respiratory complication. Moreover, an obese child with OSA has been shown to experience significantly more respiratory complications after adenotonsillectomy compared with children of normal weight.Reference Nardone, McKee-Cole and Friedman34
Adenotonsillectomy for OSA is one of the most commonly performed paediatric surgical procedures. In order to reduce practice variability and increase post-operative safety, the American Academy of Otolaryngology – Head and Neck Surgery (AAO-HNS) published clinical practice guidelines for the assessment and management of paediatric adenotonsillectomy.Reference Baugh, Archer, Mitchell, Rosenfeld, Amin and Burns6 However, there is limited guidance, and little research on the duration and level of monitoring in such paediatric OSA patients after adenotonsillectomy. Also, in children without co-morbidities, there is still a risk of significant residual obstruction on the first post-operative night. Therefore, it is important to establish the duration and level of monitoring required, even in this subset of OSA patients, because of significant variability in practice.Reference Nardone, McKee-Cole and Friedman34
Pre-operative management
The AAO-HNS guidelines advocate polysomnography prior to adenotonsillectomy. This is particularly important in children with the following co-morbidities: obesity, Down syndrome, craniofacial abnormalities, neuromuscular disorders, sickle cell disease or mucopolysaccharidosis.Reference Baugh, Archer, Mitchell, Rosenfeld, Amin and Burns6 Given the co-morbid profile of these patients, they can often undergo anaesthesia for several simultaneous procedures, and therefore may have additional post-operative needs.Reference Schwengel, Sterni, Tunkel and Heitmiller5
There is much evidence to show that OSA has an impact on various physiological processes in the body, and the anaesthetist must be vigilant in assessing for these disturbances. Though cardiovascular evaluation is not routine, children who show signs of cardiac involvement, such as right ventricular dysfunction or systemic hypertension, should have a cardiac evaluation, with echocardiography recommended. However, research is lacking in substantiating this guidance.Reference Schwengel, Sterni, Tunkel and Heitmiller5, Reference Harper, Rawling and Brown12
Children with OSA are at high risk of airway obstruction, desaturation and laryngospasm during anaesthetic induction. Those children identified as having severe OSA or cardiovascular complications associated with OSA should be considered for pre-operative continuous positive airway pressure (CPAP) or bilevel positive airway pressure therapy, which potentially improves airway patency and pulmonary hypertension.Reference Leong and Davis3, Reference Schwengel, Sterni, Tunkel and Heitmiller5
Intra-operative management
Pre-operative sedation should be used cautiously, as medications such as benzodiazepines, barbiturates, propofol and opioids can persist into the post-operative period, and cause respiratory complications that include airway collapse and respiratory depression, a significant risk in children with OSA.Reference Wood and Besecker11 Agents such as ketamine and dexmedetomidine are less likely to cause post-operative complications as they help to preserve pharyngeal muscle tone and airway reflexes.Reference Harper, Rawling and Brown12 If obstruction occurs, the use of manual manoeuvres including the jaw thrust and an oral airway, with positive pressure delivered through a facemask, may relieve the obstruction.
Post-operative management
Other complications include laryngospasm, pulmonary oedema, pulmonary hypertensive crisis and pneumonia, which raises the question as to the duration of post-operative monitoring.Reference Schwengel, Sterni, Tunkel and Heitmiller5 Most studies have found that children younger than three years are at greatest risk, though some studies have quoted under two years.Reference Leong and Davis3 A retrospective review of 2315 children demonstrated a 2-fold increase in respiratory complications after adenotonsillectomy for sleep-disordered breathing children younger than 3 years, and a national survey found that 95 of 234 of respondents were using 3 years as the cut-off criteria for admission.Reference Nardone, McKee-Cole and Friedman34 After adenotonsillectomy, even when children have been fully extubated and are awake, they carry a 16–27 per cent risk of respiratory complications. There is also a small risk of post-obstructive pulmonary oedema developing in those with severe obstruction, which may require reintubation and ventilation.Reference Leong and Davis3 Patients with disorders of pharyngeal tone or craniofacial anatomy may have residual airway obstruction and will require close observation.
It is recommended that overnight oximetry is monitored on the first night post-operatively in those patients considered high risk (young or with co-morbidities), in a setting which can respond effectively to drops in oxygen saturation levels and with non-invasive ventilation facilities.Reference Nardone, McKee-Cole and Friedman34 There is no definitive guidance on whether children should be monitored on the ward, a step-down unit or a paediatric intensive care unit. However, a small retrospective chart review found that children without a post-anaesthesia care unit event after prolonged post-anaesthesia care unit monitoring were unlikely to later develop an adverse event.Reference Wood and Besecker11
A cross-sectional survey conducted in the USA and Canada investigated the peri-operative practice for children undergoing adenotonsillectomy in tertiary care units.Reference Nardone, McKee-Cole and Friedman34 Of the 48 respondents, 44 per cent reported that they had no official admission policy for children with sleep-disordered breathing. The AAO-HNS polysomnography criteria for severe OSA were used by 40 per cent of the respondents and 15 per cent used the American Academy of Pediatrics criteria (apnoea/hypopnoea index of more than 24 events per hour or oxygen saturation nadir of less than 80 per cent). A minimum time for observation was reported by 90 per cent of the centres, with a median of 2 hours and a range of 1–6 hours. Though this study noted the wide variation in the length of the post-operative period, it failed to identify the most common practice in this regard and rightly calls for the need for more research in this area.
Summary points
Children with OSA are an anaesthetic risk, and they require careful assessment and management in the peri-operative period. Despite guidelines, there is still varying practice regarding peri-operative anaesthetic management. There is a consensus as to pre-operative risk assessment, but more research into post-operative management is needed.
Question three
What evidence is there for the benefits of surgical intervention? Does early intervention give better results than delayed intervention?
Neurocognitive effects of adenotonsillectomy
Neurobehavioural deficits related to childhood OSA are well established, with more recent studies suggesting some impairment of cognition and behaviour across the whole spectrum of sleep-disordered breathing.Reference Bourke, Anderson, Yang, Jackman, Killedar and Nixon30 These deficits can be wide-ranging, and include attention, executive functions, language, memory, and behavioural components such as hyperactivity, anxiety, depression and aggression.Reference Row, Kheirandish, Neville and Gozal26
Multiple studies have shown that adenotonsillectomy can improve neurocognitive function in children with OSA, but many are limited by small sample sizes, lack of randomisation, and reliance on parent questionnaires rather than formal neuropsychological testing.Reference Konstantinopoulou and Tapia16, Reference Goldstein, Post, Rosenfeld and Campbell28
In order to address these mixed results of the effect of adenotonsillectomy on restoring neurocognition, a recent systematic review and meta-analysis of neurocognitive outcomes after paediatric adenotonsillectomy for OSA was conducted. It found an improvement in neurocognitive function and intelligence quotient after paediatric adenotonsillectomy, especially in pre-school aged children.Reference Song, Tolisano, Cable and Camacho17 However, of the 19 prospective studies included, the findings of neurocognitive improvement after adenotonsillectomy were not generalisable to children of all age groups, and few studies assessed pre-school-aged children or included follow-up data beyond one year. Only three studies evaluated pre-school children, all of which found improved neurobehavioural parameters, suggesting that earlier intervention strongly influences the efficacy of surgery. The only randomised, multicentre, controlled study was the Childhood Adenotonsillectomy Trial (‘CHAT’), but this included older children (aged five to nine years).Reference Marcus, Moore, Rosen, Giordani, Garetz and Taylor22 This study showed that adenotonsillectomy did not significantly improve attention or executive function, as measured by a change in Developmental Neuropsychological Assessment (‘NEPSY’) scores – a domain shown to be sensitive to intermittent hypoxaemia related to OSA. This may suggest that there is a possible fixed neurocognitive deficit after a threshold age, which is yet to be defined and requires further investigation.
Furthermore, there is no consistent use of a standardised assessment tool to facilitate the interpretation of results across studies.Reference Song, Tolisano, Cable and Camacho17 The Neuropsychological Developmental Assessment and the Stanford Binet Intelligence Scales are recommended for future studies, as they are both normed and well-validated instruments, with robust validity and reliability. However, a suggested disadvantage of using the Developmental Neuropsychological Assessment is that it is a more clinically orientated objective test which does not reflect ‘real-world settings’, whereas the Behaviour Rating Inventory of Executive Function (‘BRIEF’) scoring system may be better suited as an assessment tool.Reference Manickam, Shott, Boss, Cohen, Meinzen-Derr and Amin19, Reference Marcus, Moore, Rosen, Giordani, Garetz and Taylor22 Therefore, further consensus is required on the most appropriate assessment tool for evaluating the neurobehavioural and cognitive effects, in order to standardise assessment and drive research quality.
Post-adenotonsillectomy effects on growth
Growth failure has been frequently reported in children with OSA due to dysregulation in insulin-like growth factor 1, which reduces growth potential. These effects are reversed with adenotonsillectomy.Reference Bar, Tarasiuk, Segev, Phillip and Tal25 A systematic review addressing weight gain after adenotonsillectomy found that the only randomised controlled trial investigated short-term weight gain after adenotonsillectomy, with a number of possible reasons for post-operative weight gain.Reference Van, Khan and Hussain18 If a child is underweight prior to adenotonsillectomy, attaining their normal weight potential could be considered beneficial. However, as shown in a subset of patients in the Childhood Adenotonsillectomy Study, adenotonsillectomy in children with OSA who are also overweight pre-operatively leads to further weight gain post-operatively, with further risks of obesity.Reference Katz, Moore, Rosen, Mitchell, Amin and Arens23
Cardiovascular effects of adenotonsillectomy
Although it is well documented that treatment of OSA in adults leads to significant improvements in cardiovascular function, there is a lack of research and understanding of long-term outcomes in children. In the short-term, there is evidence for the improvement of cardiovascular parameters such as blood pressure, heart rate, cardiac morphology and cardiac function in children undergoing adenotonsillectomy for OSA.Reference Teo and Mitchell7 Furthermore, a systematic review and meta-analysis of echocardiographic findings in OSA children after adenotonsillectomy showed improvements in right-sided heart structure and function around six months post-treatment.Reference Weber, Pierri Carvalho, Ridley, Williams and El Dib9 Despite these findings, there is still a paucity of well-designed and powered studies.
The Childhood Adenotonsillectomy Trial secondary outcomes also addressed cardiometabolic health, evaluating the effect of adenotonsillectomy for OSA on blood pressure, heart rate, lipids, glucose, insulin and C-reactive protein.Reference Quante, Wang, Weng, Rosen, Amin and Garetz24 They found no significant change in these parameters over the seven-month interval in the adenotonsillectomy group, compared to the watchful waiting group. Overnight heart rate appeared to be a sensitive marker of OSA severity and a measure of responsiveness to OSA treatment. This suggests a need for further research to clarify the role of adenotonsillectomy and its cardiovascular effects.
Summary points
A consensus on a standard method for evaluating the neurobehavioural and cognitive effects of OSA would help drive research standards and data validity. The effects of adenotonsillectomy on growth and its relation to a child's pre-operative weight need further investigation. There is some good evidence to show short-term improvement in cardiovascular parameters following adenotonsillectomy in children with OSA, but long-term data are lacking.
Question four
What medical treatment options are there for residual OSA after adenotonsillectomy?
Although adenotonsillectomy is the primary treatment of choice for paediatric OSA, complete resolution of OSA reportedly occurs in only 75 per cent of affected children.Reference Dehlink and Tan13 A suggested possible reason for residual disease is the presence of a multilevel obstruction, particularly common in children with co-morbidities such as Down's syndrome, obesity and craniofacial abnormalities.Reference Friedman, Wilson, Lin and Chang4 Depending on the severity of the residual OSA, treatment options vary from conservative management to CPAP, anti-inflammatory medication, oral appliances, orthodontic treatment and surgery.Reference Manickam, Shott, Boss, Cohen, Meinzen-Derr and Amin19
If adenotonsillectomy fails, second-line treatment would include investigations to identify another site of obstruction (e.g. laryngomalacia) and potentially treat surgically. Continuous positive airway pressure is another option for persistent symptoms following surgery, but adherence rates are low in children.Reference Marcus, Rosen, Davidson Ward, Halbower, Sterni and Lutz20 Nasal CPAP is another alternative increasingly used in children with residual OSA, with success rates of up to 50–100 per cent, coupled with a focus on parent education and motivation for use.Reference Massa, Gonsalez, Laverty, Wallis and Lane29
There is currently no consensus regarding the optimal methods for identifying sites of obstruction or treating children with persistent disease after surgery. Drug-induced sleep endoscopy and cine magnetic resonance imaging are the most commonly used tools reported in the literature to identify the site of obstruction in persistent OSA in children, but with no clear linkage to outcomes.Reference Manickam, Shott, Boss, Cohen, Meinzen-Derr and Amin19 Furthermore, evidence for further surgical intervention is very limited and involves mainly lingual tonsillectomy and supraglottoplasty.Reference Manickam, Shott, Boss, Cohen, Meinzen-Derr and Amin19
Summary points
Nasal CPAP may be a potential alternative to traditional CPAP for children with residual OSA after adenotonsillectomy, as it may be better tolerated. However, significant parent education and motivation is required.
Conclusion
The expansion in literature addressing the many facets of paediatric OSA has presented an opportunity for better evidence-based management of this condition for clinicians. One such example is the emerging evidence base for intracapsular tonsillectomy, in particular Coblation tonsillectomy, with its benefits of reduced haemorrhage rates and lower post-operative pain requirements. However, there still remain important questions, such as the effects of adenotonsillar regrowth and recurrence rates of OSA in the long term.
Children with OSA may pose a significant anaesthetic risk. Pre-operatively, there is a clear consensus on how best to stratify children according to their risk and we can use this to help plan post-operative care. However, there is little evidence to support best practice post-operatively, with wide variations in practice. It is clear that there is a need for more research in this area.
In addition, although there is good evidence for the benefits of surgical intervention, such as improvement in neurocognitive and behavioural outcomes, a standardised method of evaluating these will help improve the validity of the data and research standards.
