Sleep Endoscopy

Background

Sleep endoscopy, also known as sleep nasoendoscopy (SNE) or drug-induced sleep endoscopy (DISE), is a powerful tool for studying the dynamic airway in a sleeping patient with obstructive sleep apnea (OSA). Using the knowledge gained from sleep endoscopy, the surgeon can tailor the operative procedure to the patient's specific condition. ^{[1, 2, 3, 4]}

Due to the difficulty in establishing the site of obstruction in the conscious patient who carries a diagnosis of OSA, the diagnosis and treatment of OSA is a complex and multidimensional issue. Croft and Pringle first proposed sleep endoscopy in 1991. Using midazolam as a sedating agent, they demonstrated the utility of passing a fiberoptic endoscope (see the image below) through a sleeping patient’s nasal cavity to assess pharyngeal structures for evidence of obstruction. They were able to induce the preexisting snoring in 95% of their patients. ^[5]

In 1993, Croft and Pringle developed a grading scale that utilized sleep endoscopy to categorize snoring and obstruction. Grading was based on whether the obstruction was palatal, multilevel, or tongue-based. ^[6] Sleep endoscopy, in combination with the grading scale, allows the physician to directly observe pharyngeal structures in the sedated patient with OSA and categorize the obstruction.

Another grading system that uses sleep endoscopy to assess airway obstruction utilizes 3 separate evaluations of the pharynx. The first analysis uses a dichotomous assessment to identify individual areas of obstruction in the palate and hypopharynx regions. The second analysis assesses the percentage of obstruction in each area: less than 50%, 50-75%, and more than 75%, representing mild, moderate, and severe obstruction, respectively.

Based on the level and pattern of airway obstruction in a patient with OSA, sleep endoscopy allows the physician to tailor the treatment plan to each patient. This can improve the results of surgical intervention and/or minimize the scope of intervention. Sleep endoscopy may also provide information that erases the need for surgery altogether.

For example, nearly 70% of patients surveyed in an outpatient setting by Hewitt et al were determined to have a palatal cause of obstruction and were prescribed surgical intervention. However, after the patients underwent sleep endoscopy, that figure dropped to 54%, reducing the number of procedures performed. ^[7]

The high success rate of customized treatment for OSA has been attributed to the targeted selection offered by sleep endoscopy. ^[8] The multimedia clip below shows typical findings seen on routine DISE.

Other techniques for OSA assessment

Otolaryngologists use numerous modalities to locate the site of snore-sound generation and obstructive locations. These tools include a complete head-and-neck exam, the Müller maneuver, lateral radiographic cephalometry, computed tomography (CT) scanning, and magnetic resonance imaging (MRI). The static nature of these exams, coupled with increased resting muscle tone while the patient is awake, makes pinpointing the source of true obstruction difficult. While each of these methods offers reasonable diagnostic efficacy, none offers direct and dynamic visualization of pharyngeal structures during sleep.

Indications and Contraindications

Any patient who has documented OSA and is a potential candidate for OSA surgery should be considered for DISE via propofol. ^[9] This dynamic exam will study the patient’s real-time OSA under light sedation, complementing the complete head-and-neck exam and (utilizing the Müller maneuver) awake endoscopy.

The ideal patient should have:

• Documented OSA

• An airway deemed manageable by the anesthesiologist and the operating surgeon

• A patent nasal airway and nasopharynx, to allow the fiberoptic scope to pass unimpeded

Absolute contraindications include patients who are pregnant or who have a known history of propofol allergy. Other contraindications are significant nasal obstruction that impedes passage of the flexible fiberoptic laryngoscope (FFL), an “unsafe” airway, a frank aspiration history, and allergies to propofol components such as egg lecithin or soybean oil.

Outcomes

Due to the subjective nature of evaluating airway collapse during sedation, the question of sleep endoscopy’s reliability is a concern.

When comparing assessments by 2 independent reviewers of prerecorded sleep endoscopy procedures, Kezirian et al demonstrated moderate to substantial interrater reliability. This was significant in the identification of primary structures involved in obstruction versus individual structures. ^[10]

This same study demonstrated a higher interrater reliability for assessment of the palatal region for obstruction in general versus assessment of individual structures that cause obstruction in the palatal region. The authors stated that the lower reliability in assessing individual structures is less important in palatal obstruction, because traditional uvulopalatopharyngoplasty (UPPP) treatment is the same regardless of the structure involved, be it the soft palate or velopharynx lateral pharyngeal wall.

(However, there has been developing interest in UPPP modifications. These include the expansion pharyngoplasty, ^[11] uvulopalatal flap, ^[12] anterior palatoplasty, ^[13] and Z-palatopharyngoplasty. ^[14] Each of these modifies the palate in various ways, creating either a superior and lateral pull, an anterior pull, or a combination pull on the palate and pharynx. Application of these specific palate procedures based on DISE could change how the palate is addressed.)

Kezirian et al's study also mentions that the tongue, epiglottis and lateral pharyngeal walls are the 3 structures most commonly involved in obstruction in the hypopharynx. At this site, there is a moderate to substantial interrater reliability in assessing individual hypopharyngeal structures that cause obstruction. Because there are varying treatment options for the different structures that are involved, sleep endoscopy can help to determine which hypopharyngeal and oropharyngeal procedure will be the most efficacious.

A study by Rodriguez-Bruno et al concluded that sleep endoscopy has good reliability, particularly in the evaluation of hypopharyngeal structures. The investigators looked at test-retest reliability, comparing the results from 2 distinct exams analyzed by 1 person. ^[15]

When retrospectively reviewing more than 2,400 procedures involving patients with symptoms of sleep-disordered breathing, Kotecha et al demonstrated greater than 98 percent effectiveness of sleep endoscopy in producing snoring in patients. This conclusion was important, because in order for sleep endoscopy to be a valid tool for evaluating obstruction, it has to be proficient in recreating sleeplike conditions. ^[16]

Concerns regarding the potential for false-positives with sedation revolve around the premise that sedation-induced sleep can cause a greater degree of muscle relaxation than physiologically natural sleep does. ^[17] Critics argue that snoring may be induced in the patient who otherwise would not exhibit symptoms during normal sleep. ^[18]

However, when nonsnorers who underwent similar sedation techniques were compared with individuals with self-described snoring problems, the nonsnorers were not induced to snore with sedation. ^{[19, 20]}

Another concern with sleep endoscopy is whether or not the sedation-induced sleep alters the sleep profile. Rabelo et al showed that patients induced with propofol did not enter rapid eye movement (REM) sleep during sedation and that these patients tended to remain in slow-wave sleep. When the apnea-hypopnea indexes (AHIs) between propofol-induced patients and those whom slept without sedation were compared, there was little difference between the groups. Although the fundamental sleep architecture is changed in a patient with OSA, propofol has been shown to not change the respiratory pattern in patients with apnea; ^[20]

Another study demonstrated a reduction in the duration of REM sleep in patients undergoing DISE; however non-REM sleep patterns were unchanged. ^[21] It is important to note that, although it is believed that the majority of apneic events occur during REM sleep, research has shown that AHIs measured during REM and non-REM sleep in patients with OSA do not differ significantly. ^[22]

Intraprocedural grading using any of the methods described above typically correlates well with results of AHI, and it has been shown that AHIs measured after targeted therapy directed by sleep endoscopy are significantly lower.

In a study comparing 207 primary snorers without OSA with 117 subjects with OSA after receiving sedation, a higher degree of collapsibility was seen in the OSA group, with a correlation observed between the AHI during natural sleep and the degree of hypopharyngeal obstruction during sleep endoscopy. ^[23]

It is becoming more and more recognized that the utility of DISE easily surpasses the information gained from awake endoscopy in a clinic. A recent study concluded that DISE yielded better results as to specific sites, degree, and patterns of obstruction compared with the awake Muller maneuver. ^[24]

Complications in sleep endoscopy

Complications associated with sleep endoscopy include the following:

• Epistaxis from the flexible laryngoscope

• Laryngospasm

• Aspiration

• Loss of the airway

• Need for a surgical airway

Applications of Sleep Endoscopy

The following are examples of various uvulopalatopharyngoplasty and hypopharyngeal modifications for select dynamic airway findings seen on DISE:

• Lateral collapse of the oropharyngeal airway - Expansion sphincter palatopharyngoplasty (Tucker/Pang source)

• Oropharyngeal anterior-to-posterior collapse - Uvulopalatal flap, anterior palatoplasty

• Oropharyngeal concentric collapse - Combination of lateral and anterior-to-posterior collapse techniques, Z-palatopharyngoplasty

• Hypopharyngeal base of tongue collapse - Submucosal minimally invasive lingual excision (SMILE), radiofrequency ablation of the base of tongue

• Epiglottic collapse - Hyoid suspension (this also helps with lateral hypopharyngeal collapse)

Sleep endoscopy is a great tool for teaching others about airway management, and it is helpful for anesthesiology and otolaryngology residents who are learning about airway anatomy and physiology.

Relevant Anatomy

The pharynx is bounded by the base of the skull superiorly; the cricoid cartilage inferiorly; and the nasal cavities, the oropharyngeal inlet, and the base of the tongue anteriorly.

The boundaries of the oropharynx are the lower edge of the soft palate superiorly and the hyoid bone inferiorly. The anterior border is formed by the oropharyngeal inlet and the base of the tongue, and the posterior border is formed by the superior and middle pharyngeal constrictor muscles and their overlying mucosa.

Inferiorly, the posterior one third of the tongue, or the base of the tongue, continues the anterior border of the oropharynx. The vallecula, which is the space between the base of the tongue and the epiglottis, forms the inferior border of the oropharynx. This is typically at the level of the hyoid bone.

The borders of the hypopharynx are the hyoid bone superiorly and the upper esophageal sphincter (UES), or cricopharyngeus muscle, inferiorly.

The anterior boundary of the hypopharynx consists largely of the laryngeal inlet, which includes the epiglottis and the paired aryepiglottic folds and arytenoid cartilages. The posterior surface of the arytenoid cartilages and the posterior plate of the cricoid cartilage complete the anteroinferior border of the hypopharynx. Lateral to the arytenoid cartilages, the hypopharynx consists of the paired piriform sinuses, which are bounded laterally by the thyroid cartilage.

For more information about the relevant anatomy, see Throat Anatomy. Also see Mouth Anatomy, Nasal Anatomy, and Pharynx Anatomy.

The most important task for the surgeon is communication. First, the surgeon needs to discuss the goals of the obstructive sleep apnea surgery with the patient. Will the proposed OSA surgery be for cure or for palliation? Second, the surgeon needs to communicate the idea of sleep endoscopy and how it will be applied.

Third, the surgeon needs to communicate with the anesthesia team. It is ideal to have a preferred anesthesiologist who understands the goal of DISE. Snoring that leads to witnessed obstructive sleep apnea is the goal of DISE, and the anesthesiologist must not be afraid of obstructive events on the table. Desaturations are expected and should be tolerated within reason. It is helpful to point out that patients with OSA experience obstructive episodes and desaturations continually in the privacy of their own bedroom.

Blow-by oxygen is permitted. No paralytics should be administered until the time of endotracheal intubation.

Equipment

Equipment used in the procedure includes the following:

• Flexible laryngoscope with defogging solution of choice

• Monitors for pulse oximetry, routine vital signs

• Oxygen and mask and/or nasal cannula; Ambu bag

• Flexible suction and Yankauer suction

• Tracheotomy tray

• Audiovisual (AV) tower - An AV system with recording capability is a great way to review cases and to improve one’s decision algorithm and technique; the recordings also make great teaching tools for staff, patients, residents, and students

• A microphone is optional to record airway sounds and snoring

• Adjustable operating table

Equipment for intubation for emergent airways includes the following:

• Laryngoscope with various Miller and Macintosh blades

• Magill forceps

• Laryngeal mask airways (LMAs) and intubating LMAs

• Eschmann stylet

• Combitube

• Endotracheal tubes of various sizes

• Jet ventilation and 14-gauge needle

• Nasopharyngeal airway

• Oropharyngeal airway

By definition, patients with OSA have potentially difficult airways. While it is rare to lose an airway by this method, one must always be prepared for the worst-case scenario.

Patient Preparation

Medications administered to the patient include the following:

• Glycopyrrolate 0.2 mg IV x 1 - This must be given in the preoperative suite at least 15 minutes prior to DISE; it will decrease the salivary secretions, allowing optimal viewing during DISE

• Oxymetazoline nasal spray - Two sprays into both nostrils 15 minutes prior to the procedure; it is best to avoid any topical lidocaine, as this could potentially remove and blunt any natural airway reflexes

• Propofol infusion - Start 100 μ g/kg/min and titrate to patient’s snoring and OSA.

Try to avoid propofol boluses. Too much propofol will lead to a central apneic episode, making it difficult to distinguish between OSA and central apnea in the patient. This concern must be communicated with the anesthesiologist.

Positioning

The anesthesiologist should be at the head of the bed in preparation for planned or urgent endotracheal intubation. The bed should be level; a pillow can be provided for patient comfort. Bed belts should be used to secure the patient to the bed and prevent any misadventures.

Assuming that the surgeon is right handed, place the AV tower across from him or her. The anesthesiologist should have a view of the monitor too.

Approach Considerations

In my practice, I usually perform sleep endoscopy immediately before the proposed OSA surgery. Any planned nasal surgery is usually clear in the clinic, but the sleep endoscopy findings can definitely influence which oropharyngeal and/or hypopharyngeal procedure I perform on the patient. I discuss every possible scenario with my patient, and these visits can easily extend to 30 minutes for counseling purposes.

Another alternative is to perform sleep endoscopy in the operating room and then perform the definitive surgery after the sleep endoscopy discussion in clinic. This is helpful when I do staged, multilevel surgery. I will perform DISE and nasal surgery and then bring the patient back for oropharyngeal/hypopharyngeal OSA surgery. In the interim, I will have discussed the specifics of the DISE with the patient and how I will apply this knowledge in the operating suite.

Consider avoiding muscle relaxant medications such as benzodiazepines, as these can relax the airway too much and possibly give false positives. Croft and Pringle ^[5] used only midazolam for their DISE, and it was reliable for them. While midazolam does make the patient more at ease in the preoperative area, the propofol dose may be less, compared with a scenario in which the patient does not require midazolam preoperatively.

The safest way to perform sleep endoscopy is with a team approach with your anesthesiologist, with DISE performed in a monitored setting. As described above, communication with your patient and the operative team is important.

In order to get a reliable and valid exam, patience is needed. Allow the propofol infusion to work, and allow the patient to settle down into snoring and then an eventual obstruction.

Sleep Endoscopy Procedure

Once the AV tower and FFL are ready, signal the anesthesiologist to begin the propofol infusion. The anesthesiologist must carefully titrate the propofol infusion in order to cause obstructive apnea, but no central apnea. This can easily be observed.

When the patient can no longer be aroused by voice, defog the FFL and introduce it into the nasal cavity. Fully examine the nasal cavity for any airway obstructions and then advance into the nasopharynx. Wait in the nasopharynx until the patient begins snoring.

Examine the nasopharynx, velopharynx, and hypopharynx. Pay attention to the degree of collapse; the following grading scale can be used:

• 0-25% collapse

• 26-50% collapse

• 51-75% collapse

• 76-100% collapse

Pay attention to the following structures:

• Obstruction at the level of the palate - Palate, tonsils, and lateral pharyngeal wall

• Obstruction at the level of the hypopharynx - Base of tongue, epiglottis, and lateral pharyngeal wall (one study mentions that 23% of patients had epiglottic collapse that was not seen on awake endoscopy ^[24] )

Advance the mandible 5-10 mm forward to reproduce the action of a genioglossal advancement or mandibular advancement device.

Perform a jaw thrust maneuver or insert a nasal trumpet or oral airway to convince oneself or the anesthesiologist of the effectiveness of these simple procedures, which can establish an airway in an obstructing patient.

If there is suspicion preoperatively of hypopharyngeal collapse, one can elect to dissect down to the hyoid bone. During direct visualization of the hypopharynx in DISE, the hyoid bone can be manipulated to simulate hyoid suspension or advancement. If there is improvement in the hypopharyngeal airway during manipulation, the appropriate procedure can be performed (hyoid suspension, myotomy, hyoid advancement, etc).

Objective grading schemes

VOTE is a proposed system of classification that may be used in order to help establish a universal way for otolaryngologists to classify and communicate objective findings of airway obstruction in OSA patients. Based on multiple DISE interventions, four common locations of obstruction were established. These locations can either contribute to OSA individually or in combination. The levels are velum, oropharynx, tongue base, and epiglottis (VOTE). A grade of 0-2 is assigned based on degree of obstruction: 0 = no obstruction; 1 = partial obstruction; 2 = total obstruction. ^{[24, 25]} It is also helpful to describe the type of collapse: lateral, anterior-posterior, or concentric.

In a recent study looking at DISE in pediatric patients, authors developed a quantitative way of distinguishing airway obstruction in pediatric DISE. The evaluated locations were adenoids, velum, lateral pharyngeal wall, tongue base, and supraglottis. Obstructions were scored on a specific 4-point grading system. ^[26]

Both grading schemes provide a reliable and reproducible method of evaluating the upper airway during sleep endoscopy.