General considerations
Historically, there have been more challenges in the evaluation and management of obstructive sleep apnea (OSA) in infants compared to older children, but more recently OSA has been increasingly recognized and studied in these youngest patients. OSA has been linked to age-specific comorbidities in this age group, including neurocognitive and developmental delays, failure to thrive, and even sudden infant death. Infants are particularly vulnerable to OSA due to anatomical features and REM-predominant sleep distribution and are at further increased risk when there are underlying neurologic or neuromuscular conditions, or issues with respiratory mechanics and ventilatory control. The management of OSA in infants is challenging due to limited age-specific outcome targets, a limited number of devices that would otherwise be appropriate for older children, and the higher proportion of comorbidities in this age group which requires more individualized treatment plans. The limited evidence for treatment guidelines in infants has resulted in variability in management between centers. Even when a treatment plan is put in place, its effectiveness can be affected by various extraneous factors, such as rapid changes in required support due to growth and development.
The decision to initiate treatment for OSA in infants and which modality to utilize may be dependent on a variety of factors, including the child’s age, presenting symptoms, etiology of upper airway obstruction, polysomnogram results, and the presence of additional comorbidities, either related to or in addition to OSA. In this chapter, OSA management options will be reviewed, including nonsurgical and surgical, for infants less than 2 years of age ( Table 7.1 ). Additionally, special considerations related to monitoring for treatment efficacy will be considered. Lastly, future directions for infant OSA management will be proposed.
| Treatment Modality | Mechanism of Action | Benefits | Challenges | Target Populations |
|---|---|---|---|---|
| Nonsurgical | ||||
| Positive airway pressure | Pressurized air acts as a pneumatic stent to maintain airway patency. | May be effective for even moderate-severe OSA. Use and efficacy can be monitored. Noninvasive. Strong evidence to support its efficacy in older patients. | Potential for increased risk of skin breakdown. Difficult to find well-fitting interfaces for infants. Desensitization may be challenging without adequate support. Increased use may predispose to midface hypoplasia. | Moderate-severe OSA when appropriate interface available. |
| High-flow nasal cannula | Heats and humidifies air, decreases dead space within the upper airway, and reduces upper airway collapse through a continuous stream of positive pressure. | Nasal prongs require less desensitization than PAP interface. | Potential risk for pneumothorax or pneumomediastinum due to air leak. May not be adequate for more severe OSA. Monitoring and titration limited. | Moderate OSA when PAP not available. |
| Supplemental oxygen | Reduce desaturation associated with obstructive events. | Widely available, nasal cannula more comfortable than PAP interface. | Insufficient for more severe OSA, possibility of blunting respiratory drive. | Mild-moderate OSA and/or hypoxemia without hypoventilation. |
| Oral/nasal appliances | Bypass oropharyngeal and nasopharyngeal obstruction, respectively. | Noninvasive. Less cumbersome compared to other nonsurgical methods. | Some not widely available. Mucosal irritation, ulcer formation, risk for infection. | Population depends on specific appliance; typically, infants with craniofacial conditions. |
| Prone positioning | Allows for anterior displacement of tongue base by gravity, relieving the obstruction. | Simple, requires no equipment or intervention. | Ineffective for more significant OSA. May increase risk for SIDS. | Adjunct in glossoptosis with mild OSA. |
| Surgical | ||||
| Adenotonsillectomy | Removal of obstructing adenoid and tonsillar tissue. | Highly effective first-line treatment for OSA related to adenotonsillar hypertrophy. | Infants may be at greater risk for significant postoperative complications. Most infants will not have adenotonsillar hypertrophy and will not be candidates. | Those with adenotonsillar hypertrophy as the primary cause of OSA. |
| Tongue-lip adhesion (TLA) | Surgical fixation between the lip and tongue, a form of glossopexy. Meant to prevent posterior displacement of the tongue resulting in obstruction. | May negate the need for tracheostomy in patients with RS and severe OSA. | Higher rates of complications and residual OSA as compared to MDO. Less likely to prevent the need for tracheostomy, compared to MDO, in patients with RS and severe OSA. | Those with micro- or retrognathia and moderate-severe OSA from glossoptosis. |
| Mandibular distraction osteogenesis (MDO) | Involves bilateral osteotomies of the mandibular rami and distractor placement bilaterally, allowing for sequential lengthening of the hypoplastic mandible. | Highly effective in many patients, even with very severe OSA. May be associated with improved ability to feed. | Potential complications may include malunion or misalignment of the involved bones, and infection. | Those with micro- or retrognathia and moderate-severe OSA from glossoptosis. |
| Tongue reduction | Debulking of an enlarged tongue to reduce oropharyngeal obstruction. | May negate the need for invasive respiratory support. | No consensus on appropriate timing. Potential complications include wound dehiscence and surgical revision. May impact dental outcomes and dentoalveolar development. | Those with moderate-severe OSA related to macroglossia (i.e., Beckwith-Wiedemann syndrome). |
| Supraglottoplasty | Includes making an incision in the aryepiglottic folds releasing the epiglottis and removing redundant tissues which result in collapse of the larynx on inspiration. | May obviate the need for tracheostomy, previously the primary treatment for those with OSA and laryngomalacia. | Potential complications include cough and dysphagia. Potential need for surgical revision. Efficacy of procedure versus natural history of condition unclear. | Those with laryngomalacia. |
| Tracheostomy | Use of an adjunctive and artificial airway to bypass upper airway obstruction. | Definitive treatment to bypass upper airway obstruction. Improved symptoms in adult populations. | Invasive procedure with potential complications (i.e., tracheostomy tube occlusion, accidental decannulation). | Those with limited treatment options or OSA refractory to surgical intervention. |
The etiology of sleep-disordered breathing in infants is multifactorial and pathophysiology is unique from older children and adults. Factors such as compromised upper airway anatomy, disadvantaged pulmonary mechanics, ventilatory control immaturity, lower arousal threshold, laryngeal chemoreflex, and a REM-predominant sleep state distribution make infants particularly vulnerable to OSA. Though not currently well described in the literature, OSA risk factors in infants can be conceptualized as discreet but sometimes overlapping endotypes ( Fig. 7.1 ). Broadly speaking, these endotypic categories can be divided into (1) bone abnormalities, (2) soft tissue abnormalities, and (3) other abnormalities that can be targets of management strategies described below.
Medical management of infant OSA
Given the potential repercussions of untreated OSA in infants, clinicians must be aware of management options, especially for more severe OSA. Further challenging management, the natural history of OSA in some conditions associated with OSA in infants and young children is poorly understood, and the need for and type of intervention may change over time. For example, patients with nonsyndromic micrognathia may have severe OSA during the neonatal period, but mandibular growth is expected to occur. In these instances, temporizing therapies with a nasopharyngeal airway, positive airway pressure, or other nonsurgical therapies may be sufficient. In contrast, infants with syndromic micrognathia such as Treacher Collins syndrome may not experience sufficient mandibular growth and may be more appropriately referred for surgical intervention, discussed below.
Ultimately, the decision about the most appropriate therapy should incorporate a variety of factors, including the severity of the OSA from objective testing, as well as accompanying comorbidities, the projected prognosis for the patient’s underlying medical condition, and available treatment options. For mild OSA without significant clinical concerns, the decision to manage with watchful waiting may be appropriate. There are a variety of nonsurgical treatment options for infants with OSA, reviewed below.
Positive airway pressure
While adenotonsillectomy is considered first-line treatment for pediatric OSA, positive airway pressure (PAP) is an acceptable option for those without adenotonsillar hypertrophy or where adenotonsillectomy is contraindicated. , PAP has previously been shown to improve symptoms in children with OSA and has led to improvements on polysomnogram including both decreased apnea-hypopnea indices and increased arterial oxygen saturation, as well as improved quality of life, and positive impacts on cardiovascular risk. Adherence to PAP can be challenging and thus a multimodal approach can help ease this burden, particularly in infants.
PAP is prescribed as either continuous positive airway pressure (CPAP) or bilevel positive airway pressure. Auto-titrating PAP is available, but not currently appropriate for infants. PAP works by delivering pressurized air, acting as a pneumatic stent to maintain airway patency, preventing obstruction. Most PAP devices for the treatment of OSA are designed for adults and larger children. In many cases, infants will not meet the minimum weight requirement for PAP machines and will require a portable home ventilator. When fitting a particular infant for an interface with which to supply the prescribed PAP, there are fewer options available and additional challenges for patients with craniofacial differences. Traditionally, nasal masks ( Fig. 7.2 ) are the preferred option for infants due to availability, fit, and safety. While there are some naso-oral or full-face interfaces, there could be risk of harm by aspiration with these in the home environment, particularly if the child is not being directly observed. Another option is nonoccluding nasal prongs, although these come with their own challenges, including a lack of occlusion with air leak, resulting in lower pressures, and making this device less likely to be effective for more severe OSA. After initiating PAP, polysomnography should be used to titrate to an effective pressure, as with older children and adults.
While there are a limited number of studies to objectively assess the utility of PAP in infants with OSA, there is some data to support its effectiveness. In an early report of 74 infants less than 12 months, followed closely for a mean duration of nearly 3 years, all but two were successfully able to use PAP. Another study from 1999 evaluated a cohort of infants who underwent overnight polysomnography to assess the severity of their OSA and followed them longitudinally. The authors found that CPAP was useful in preventing obstruction and improving sleep disturbance associated with OSA. Most of the cohort (18/24, 75%) used CPAP at home, in some cases for greater than 4 years. In a more recent retrospective study from our group, the efficacy of CPAP in 41 infants less than 6 months old was compared to a cohort of school-age children (5–10 years of age) with OSA underwent both baseline and titration polysomnography after having started PAP. Despite having more severe baseline OSA, a similar proportion of infants were effectively titrated with PAP compared to their older counterparts, and machine downloads showed that infants used PAP more. Surprisingly, the barriers to use were similar between the two groups.
While PAP can be an effective treatment modality for infants with OSA, adherence can be a challenge, as with older patients. Identifying potential barriers to PAP use is important for the medical team to address, and caregiver efficacy is an important concept. Though studied in older children and adolescents, it has been shown that caregiver-reported self-efficacy plays an important role in predicting PAP adherence. Additional challenges in infants include frequently changing infant sleeping patterns and rapid growth, which may necessitate changes in the interface as well as required pressures. For infants, our practice is to frequently reevaluate PAP use and barriers with families in the clinic and reevaluate the need for PAP and required pressures with titration studies in the sleep laboratory, particularly during the first year of life, when there can be frequent changes. Potential side effects of PAP in infants may include nasal congestion, sensitive skin, and an increased risk for skin breakdown. With prolonged use, there is also a potential risk for midface hypoplasia, particularly in infants with craniofacial clefts. The long-term success of CPAP for infants lies within the abilities and capabilities of the medical team to not just prepare the child for home but the family as well.
High-flow nasal cannula
High-flow nasal cannula (HFNC) is a noninvasive treatment option that has been studied in both adults and children with respiratory insufficiency. HFNC both heats and humidifies air, and its mechanism lies in both decreasing dead space within the upper airway and preventing upper airway collapse through substantial airflow that can provide positive pressure. HFNC is delivered through a nasal cannula interface, sometimes occluding a greater portion of the nares than a traditional low-flow cannula.
HFNC could be an alternative to PAP in young children and infants, particularly when there is not a PAP interface with a good fit, such as very small infants or those with craniofacial differences. A 2013 randomized controlled trial compared the efficacy of HFNC to nasal CPAP among 432 infants in a neonatal intensive care unit who were candidates for PAP. They found that infants randomized to the HFNC arm of the trial were no more likely to require intubation or supplemental oxygen compared to those randomized to CPAP. Specific to OSA, it has been previously demonstrated that HFNC can result in a significant reduction in obstructive events and can help alleviate symptoms. A recent 2020 retrospective study of 22 infants with OSA found that HFNC led to a significant reduction in obstructive events (28.9 events/hour versus 2.6 events/hour, P <0.001). Overall, it appears that HFNC might be an alternative treatment modality for infants who would otherwise be candidates for PAP.
Supplemental oxygen
Low-flow supplemental oxygen has previously been used to treat OSA in adult populations, and its use in children has also been documented. In a limited number of pediatric studies, low-flow supplemental oxygen has been shown to improve oxyhemoglobin saturation and decrease obstructive events primarily by decreasing desaturation needed to score hypopnea, without necessarily a concomitant decrease in ventilatory drive. , While the literature is scarce, there may be utility in using low-flow supplemental oxygen in the treatment of OSA in infants. One recent study evaluated polysomnography data in a cohort of infants who underwent two sequential sleep studies, initially without and then with supplemental oxygen. This study found that the use of low-flow supplemental oxygen was associated with improved oxygenation and fewer obstructive events. Importantly, there was no significant difference in alveolar ventilation as measured by both end-tidal CO 2 or transcutaneous CO 2 monitoring. While additional studies may be needed to evaluate long-term outcomes, there appears to be some preliminary data to support the use of low-flow supplemental oxygen in this age group.
Oral and nasal appliances
Primarily studied in patients with Robin sequence (RS), who classically present with micrognathia and/or retrognathia with glossoptosis, oral appliances may result in some improvement to OSA due to relief of tongue-based obstruction. OSA in patients with RS results from posterior movement of the tongue base toward the posterior pharyngeal wall, which can be accompanied by inward collapse of the lateral pharyngeal walls. Treatment should focus on both shifting the tongue forward and stabilizing the pharynx. Unlike strategies which are more invasive, including tongue-lip adhesion, mandibular distraction osteogenesis, and tracheostomy, noninvasive alternatives that allow correction of tongue displacement and pharyngeal stability are needed. Oral appliances, including the pre-epiglottic baton plate (PEBP), consist of a hard acrylic plate made to fit the patient based on a silicone impression of the oropharynx, attached to a hard acrylic velar extension. One case series with long-term follow-up showed a significant decrease in the mixed-obstructive-apnea index at time of discharge and near resolution 3 months later in a small cohort of infants with RS and OSA. The PEBP, and a similar apparatus named the Tuebingen palatal place, have been used in other populations as well, including infants with femoral facial syndrome and Down syndrome. In addition to OSA, these devices may also allow for both swallowing and phonation development. These devices may not be widely available, and thus providers may need to consider alternative options.
Like an oral appliance, an oropharyngeal airway or nasopharyngeal airway (NPA) also has been used in children to bypass nasopharyngeal obstruction. Predominantly used in adults, it has also been studied in infants with syndromic craniosynostosis or glossoptosis with concurrent OSA. With midface hypoplasia known to be a contributing factor to OSA in patients with syndromic craniosynostosis, a study out of the United Kingdom looked at the efficacy of using an NPA to bypass midface obstruction in both infants and children with moderate-severe OSA. In this study, 27 children aged 5 months through 48 months with syndromic craniosynostosis and moderate to severe OSA were treated with an NPA. Results showed an improvement in AHI in 96% of the cohort, in addition to oxyhemoglobin saturation, although most still had mild residual OSA. Risk of the NPA include mucosal irritation resulting in pain or bleeding and nasal ulcers/infections, some of which can be mitigated by adjustments to NPA tube size, shape, and material. Most studies assessing its efficacy have done so in the postoperative period or for short-term use, and there is little data to support its long-term use, which may impact both its availability and its use in general.
Positioning
In some patients, prone positioning during sleep can reduce the effect of gravity on upper airway collapsibility, particularly related to tongue-based obstruction. Therefore, prone positioning may be used in infants, including those with glossoptosis, although its efficacy remains unclear. One case series of three infants with micrognathia and severe OSA found that there was a significant improvement in obstructive polysomnographic indices with prone positioning compared to supine positioning, with one having complete resolution and one other with mild residual OSA. However, a recent study compared polysomnographic data in 11 infants with RS in the supine and prone positions who were turned halfway through a diagnostic sleep study. In this small cohort, results showed that there was a significant decrease in the OAHI in some infants when turned prone versus supine, but significant OSA persisted in most cases. In a retrospective chart review of 23 infants with cleft palate who slept in both the supine and prone positions during a polysomnogram, there were no significant improvements in polysomnographic metrics, including apnea-hypopnea index and gas exchange data, between the two sleep positions. Even in the setting of improved OSA, one must also consider the concurrent risk of sudden infant death syndrome in young infants who are turned prone while asleep. Overall, the data to support prone positioning is limited, and additional evidence is necessary to support its use in infancy. Polysomnography should be performed in infants in the prone position to evaluate for efficacy.
Surgical management of infant OSA
General indications for surgical management of OSA include moderate-to-severe OSA on polysomnography as well as overnight and daytime symptoms and failure or lack more conservative medical management. Surgical options are targeted to the site of obstruction, which may include the nasal cavity, nasopharynx, oropharynx, and hypopharynx. In children, adenotonsillectomy is considered first-line treatment for children with obstructive sleep apnea who have adenotonsillar hypertrophy. While some patients less than 2 years old with OSA do have adenotonsillar hypertrophy, there are a variety of additional anatomic factors specific to this age group which are important to address and will dictate surgical management. In this section, we will review the most common surgical treatments used in the treatment of infant OSA.
Adenotonsillar hypertrophy
Adenotonsillar hypertrophy is a major contributor to OSA in older children, and the definitive treatment for many children is adenotonsillectomy. In contrast, in neonates and infants, a smaller portion will have enlarged adenoids and tonsils, particularly as the peak age for adenoid and tonsillar hypertrophy with associated OSA may be between 3 and 6 years of age. , Though adenotonsillectomy can be performed in infants, particularly 12 to 24 months old, there is the potential for increased complications, so the procedure should be done with additional monitoring in place for complications such as oxyhemoglobin desaturation, laryngospasm, airway obstruction due to edema, and postobstructive pulmonary edema. ,
The American Academy of Otolaryngology—Head and Neck Surgery clinical practice guidelines, most recently updated in 2019, provides evidence-based recommendations for pre-, intra-, and postoperative care for children 1 to 18 years of age undergoing tonsillectomy. For children with sleep-disordered breathing, polysomnography is recommended for children less than 2 years old in addition to those who are obese, have Down syndrome, craniofacial abnormalities, and mucopolysaccharidoses, among others. With regards to postoperative monitoring, the 2019 guidelines recommend overnight, inpatient monitoring for children less than 3 years of age, or who have severe OSA (defined as an apnea-hypopnea index (AHI) ≥10 events per hour and/or an SpO 2 nadir <80%).
Several studies have evaluated the safety and efficacy of adenotonsillectomy in infants. One recent study that compared adenoidectomy complications, including postoperative bleeding and hypoxemia, in infants under 1 year of age to those between 3 and 5 years old found that the rates of complication were low and did not vary significantly across age groups. A separate case series looked at the efficacy of adenotonsillectomy in 25 infants with OSA younger than 1 year of age, of which 12 were otherwise healthy and 13 had significant comorbid conditions including Down syndrome, prematurity, hypoxic-ischemic encephalopathy, and heart disease. They found that adenotonsillectomy was successful in treating OSA through a combination of subjective improvement in symptomatology and postoperative polysomnography; however, the success rate was higher in those who were otherwise healthy compared to infants with comorbid conditions. Postoperative complications, including dehydration, persistent hypoxemia, and bleeding, occurred in 25% and 33% of healthy patients and those with comorbid conditions, respectively.
While the use of adenotonsillectomy in infants remains more complex than in older children, studies have demonstrated usefulness in treating OSA. As explored below, other etiologies for OSA, including some comorbid conditions, can drastically have an impact on management options.
Robin sequence/micrognathia
RS results from maldevelopment of the mandible resulting in micrognathia, subsequent glossoptosis, and ultimately upper airway obstruction. While some infants with RS have an identified syndrome such as Stickler syndrome or Treacher Collins syndrome, others have micrognathia as an isolated finding. The severity of OSA in infants with RS is high variable, and management includes both nonsurgical and surgical modalities. While in more mild cases, conservative methods may be used (see above), those with severe obstruction often require surgical therapy during the neonatal period.
Tongue-lip adhesion (TLA) is a procedure that dates to the 1940s when it was popularized as a treatment for airway obstruction in RS and has since gone through various iterations. The procedure itself involves surgical fixation between the lip and tongue, a form of glossopexy, to prevent posterior displacement of the tongue resulting in obstruction. While reports of this procedure include improvements in OSA among infants, the cohorts of patients being studied are small, and to our knowledge, there is little longitudinal follow-up to support its efficacy. Furthermore, some patients have been found to require either repeat TLA or secondary surgery such as mandibular distraction or tracheostomy. In one 2009 study of 22 patients who underwent TLA for RS, 55% of the cohort developed a complication, with the 2 most common being wound dehiscence and chin abscess formation. While TLA might be an effective treatment modality for some, there are few studies that have included polysomnography and other objective outcomes of OSA improvement.
Mandibular distraction osteogenesis (MDO) is another option for surgical management of micrognathia in infants with severe OSA. MDO involves bilateral osteotomies of the mandibular rami and distractor placement bilaterally, allowing for sequential lengthening of the hypoplastic mandible. There have been several studies assessing the efficacy of MDO on OSA and feeding in the neonatal period. A longitudinal analysis of 31 neonates with RS and severe OSA who underwent MDO (median age at time of surgery of 23 days) found significant improvement in polysomnographic parameters, including a reduction in OAHI and improvements in both sleep efficiency and SpO 2 nadir. Of note, though there was substantial improvement in the severity of OSA postoperatively, all patients had some degree of residual OSA. Some patients with RS and severe airway obstruction may need tracheostomy, particularly those with underlying neurologic disease, but MDO may be an alternative. One recent study showed successful treatment of severe OSA using MDO in a series of five infants with tongue-based obstruction without micrognathia, with all avoiding tracheostomy and two requiring respiratory support by either nasal cannula or CPAP postoperatively. Complications of MDO may include tooth injury, nerve injury, infection, scarring, injury to the temporomandibular joint, and device failure. Careful planning and proper surgical technique can help mitigate some of these occurrences, but these complications are important to keep in mind.
A recent 2018 review of 67 studies compared the effectiveness between TLA and MDO in children with RS. They found that higher percentages of patients with MDO, as compared to TLA, avoided tracheostomy (95% versus 89%) and achieved full oral feeds (87% versus 70%). Additionally, rate of recurrent intervention due to residual obstruction was lower in the MDO group (4%–6%) as compared to the TLA group (22%–45%). There was too much variability between the study methodology to allow for metaanalysis. In a study of 1289 infants with micrognathia hospitalized in tertiary-care neonatal intensive care units who required surgical intervention, those undergoing MDO had significantly lower rates of secondary airway surgery and greater rates of exclusive oral feeding at hospital discharge compared to TLA. Ultimately, both TLA and MDO appear to be viable surgical options for those with RS and OSA. Patient selection, like with other surgical procedures, appears to be crucial to ensure successful outcomes in this cohort. Additional studies are necessary to monitor the long-term outcomes of this procedure, as are studies looking at individual syndromes and clinical trials comparing different surgeries (or surgery versus nonsurgical management) to compare outcomes.
Macroglossia
Like mandibular hypoplasia, tongue-based airway obstruction in infants may occur because of macroglossia ( Fig. 7.3 ), which may occur as part of an underlying syndrome, from a tumor, or in isolation. The degree of OSA in infants with macroglossia is related to the relative size of the tongue to the rest of the oral cavity as well as muscle tone and the presence of other airway abnormalities. The most common congenital syndrome resulting in macroglossia is Beckwith-Wiedemann syndrome (BWS), a pediatric overgrowth disorder characterized by hyperinsulinism, omphalocele, hemihypertrophy, distinct facial features, and increased risk for embryonal tumors. , As many as 50% of infants and children with BWS have OSA, and infants may be at the greatest risk due to more severe relative macroglossia. , In addition, macroglossia may be seen in infants with mucopolysaccharidoses, Down syndrome, and with hemangiomas and lymphangiomas.
