1.1 Introduction

This book has been written to describe which patients with obstructive sleep apnea (OSA) can be identified as appropriate candidates for upper airway surgery, to describe in detail the various procedures available, and to discuss how a particular surgery can be selected for a given patient. Upper airway surgery, in general, assumes that the primary cause of OSA is an anatomically small pharyngeal airway and that the surgical procedure can adequately correct this anatomical abnormality to allow for unobstructed breathing during sleep. It has been known for years that abnormal pharyngeal anatomy is an important part of OSA pathophysiology and this will be discussed further. However, the success or failure of upper airway surgery to reduce or eliminate disordered breathing is likely dependent on a number of variables only some of which relate to anatomy. These include:

• 
  

  The severity of the anatomical abnormality.

  
  
• 
  

  The site or sites, severity, and configuration of collapse and their proper identification.

  
  
• 
  

  The role and importance of nonanatomical traits in the pathogenesis of OSA in a particular patient.

The role of abnormal anatomy in the pathogenesis of OSA has been well described over the years. Haponik et al ¹ were the first to suggest this possibility when they demonstrated a smaller pharyngeal airway lumen in OSA patients compared with a control group using CT scanning in 1982. This led to numerous assessments of pharyngeal anatomy in OSA patients using a variety of imaging techniques including cephalometry, acoustic reflection, CT scanning, and MRI. The work of Richard Schwab ^{2 , 3} using MRI for the past 10 to 15 years has convincingly demonstrated that the upper airway lumen is smaller and that certain tissue structures are larger in OSA patients compared with various control groups. These enlarged structures include the tongue, the lateral pharyngeal walls, the parapharyngeal fat pads, and the soft palate/uvula among others. However, virtually all such imaging has been conducted during wakefulness when pharyngeal muscles are active, making pure assessment of anatomy difficult. Ultimately Isono et al ⁴ studied patients during neuromuscular blockade and found the pressure-area plots of patients with OSA to be quite different from controls, indicating a clear anatomic deficiency in the apnea patients.

In addition to studies directly assessing pharyngeal anatomy, metrics of pharyngeal collapsibility obtained during sleep have been used for the past 30 years to quantify the severity of the anatomic abnormality. Alan Schwartz and colleagues ^{5 , 6} at Johns Hopkins developed the methods to quantify the critical closing pressure (Pcrt) that can be measured with and without pharyngeal muscle activity (active versus passive Pcrt). This metric has proved quite valuable as a physiologic measure of the severity of the anatomic abnormality and has been used widely. However, no measure of anatomy whether quantified awake or asleep, with or without muscle activity, or using imaging or physiologic techniques has correlated well with the severity of OSA. ⁷ Indeed some patients without OSA have a similar or more collapsible airway than many OSA patients. ⁸ Patients with a very high passive Pcrt (above 2–3 cm H₂O) generally have more severe OSA than those with a lower Pcrt (−2 to 0 cm H₂O). However, there remains huge variability in the apnea–hypopnea index (AHI) at any measured level of pharyngeal collapsibility. ^{7 , 8} This failure to predict apnea severity can be interpreted in a number of ways. First, AHI may be a poor measure of OSA severity. Second, Pcrt may be a less quantitative assessment of the pure anatomy than originally thought. However, third, and most likely, is the fact that anatomy is not the entire explanation for the presence or severity of OSA. Other physiologic traits are likely important as well and these will be addressed later. That being said, anatomy is the most important trait in the majority of OSA patients, and OSA is likely to resolve if surgery can completely correct the anatomic deficiency yielding a minimally collapsible airway (Pcrt below −6 or −7 cm H₂O). ⁸ However, this is often difficult.

Recognizing that anatomy is not often the entire explanation for the presence of OSA, it is not surprising that the techniques described earlier for quantifying pharyngeal anatomy/collapsibility have not proved very useful in predicting who will and will not respond to upper airway surgery. Again, this may reflect our inability to adequately quantify collapsibility or the importance of nonanatomic traits. However, considerable effort has been put into at least identifying the site(s) of collapse as clearly as possible and directing the surgery at the appropriate site(s). Not surprisingly, assessing the site of collapse during sleep, generally using drug-induced sleep endoscopy (DISE), has proved the most fruitful. ⁹ However, in the author’s opinion, no study has combined DISE and measurement of Pcrt to predict surgical outcomes, and such a study is needed as both the site of collapse and the severity of collapsibility seem likely to dictate surgical success.

It has become evident over the past 10 to 15 years that nonanatomical traits are important in OSA pathogenesis. ^{10 , 11 , 12} Most evidence suggests that there are four phenotypic traits that dictate who will and will not develop OSA. As outlined earlier, anatomy is one such trait. The others include:

• 
  

  The upper airway response: This is the ability of the pharyngeal dilator muscles to respond to standard stimuli (airway negative pressure and rising PCO₂) during sleep. Conceptually, virtually all patients with OSA breathe normally while awake with little difficulty maintaining a patent upper airway. Thus pharyngeal dilator muscles can compensate for even the most anatomically deficient airway during wakefulness. ¹³ It is the failure of this compensatory muscle activation during sleep that leads to airway collapse in most patients. However, there is considerable variability in the responsiveness of these muscles during sleep. In some patients these muscles can activate quickly during sleep yielding stable respiration despite considerable airway collapsibility. In other patients with a minimally abnormal airway, the muscles cannot compensate during sleep and airway collapse occurs rapidly. Thus, variability in upper airway muscle responsiveness importantly dictates who does and who does not develop sleep apnea.

  
  
• 
  

  Arousal threshold to respiratory stimulation: Based on what has been said earlier, when there is deficient pharyngeal anatomy, upper airway muscles must respond during sleep to open the airway if apneas or hypopneas are to be prevented. However, the stimuli to the upper airway muscles (increasingly negative airway pressure and rising PCO₂) develop slowly after the onset of airway obstruction. ¹⁴ Thus the individual must stay asleep long enough for these stimuli to reach adequate levels to activate the dilator muscles and restore airway patency. If the individual awakens from sleep before the required level of muscle activation has been reached, stable breathing during sustained periods of sleep will become difficult to achieve. Thus a high respiratory arousal threshold may help prevent disordered breathing by allowing pharyngeal dilator muscles adequate time to activate. On the other hand, individuals with a low respiratory arousal threshold will readily awaken shortly after each episode of airflow obstruction and therefore cycle between sleep and wake.

  
  
• 
  

  Loop gain (ventilatory control instability): Loop gain is an engineering term used to describe the gain of any system controlled by feedback loops. Respiratory control is very much a feedback-controlled system designed primarily to control arterial PCO₂. Loop gain is most easily understood as the respiratory response to a disturbance divided by that disturbance, i.e., response/disturbance. A high loop gain is characterized by a large respiratory response to a small disturbance. In such a case, a small increase in PCO₂ leads to a large increase in ventilation to correct it. When this occurs, ventilation can become unstable with waxing and waning between hyperpnea and hypopnea/apnea. Thus individuals with a high loop gain have a tendency toward unstable respiratory control. ¹⁵ If that same individual has a collapsible upper airway, then small airway obstructions can lead to large ventilatory overshoots, which leads to the next obstructive event.

Thus, although upper airway anatomy is fundamentally important in the pathogenesis of OSA, it is far from the entire cause of this disorder.

The fact that sleep apnea has a multifactorial cause should not deter efforts focused primarily on improving pharyngeal anatomy such as upper airway surgery. As stated previously, if the anatomy can be adequately addressed, the other traits generally become unimportant. It is when surgery is unsuccessful or only partially successful that the other contributors to OSA need to be considered. Currently if surgery does not lead to an adequate reduction in AHI, doctors revert to standard approaches to apnea management, i.e., continuous positive airway pressure (CPAP), mandibular advancing devices, or further surgery; all approaches again focus on correcting the anatomy. An alternative approach would be to recognize that improving anatomy increases the potential role of the nonanatomic traits described before and makes therapies focused on these other traits more likely to be successful. As an example, two recently completed but unpublished studies (oral communication from Dr. Scott Sands in September 2017) indicate that the single most important trait distinguishing OSA patients who failed pharyngeal surgical procedures versus being cured was a particularly high loop gain, not badly abnormal anatomy. This would suggest, although the studies have not been performed, that treating this high loop gain in the surgical failures with acetazolamide or nocturnal oxygen might well turn failures into successes.

In conclusion, upper airway surgery in appropriately selected OSA patients is a viable therapy focused on correcting deficient pharyngeal anatomy, the primary cause of sleep apnea. Thus, if the anatomic defect is not too great, the site and type of surgery are correct, and the other nonanatomic traits are not terribly abnormal, surgery should be successful. However, when surgery is not successful, continuing to address only anatomy may not be in the best interest of the patient and efforts to improve the other traits may yield surprisingly good results. Such an approach, however, will require further research before it is broadly accepted.

1.2 A Short History of Surgery for Sleep-Disordered Breathing: From the Uvula Crusher to the Stimulation of the Hypoglossus Nerve

1.2.1 Antiquity

Observations about symptoms and surgical procedures to influence snoring and OSA are transmitted from antiquity. Let us start with the so-called “Sleeping Lady of Malta,” a 5,000-year-old terracotta sculpture from the Stone Age in the Hypogeum of Malta. She looks like a snoring woman suffering from obesity and hypersomnia.

In Greek mythology, the family of the Gods of the Underworld is associated with sleep in a fascinating interrelationship: Nyx (night) is the mother of the twins Hypnos (sleep) and Thanatos (death), and Morpheus (dreams) is the son of Hypnos. Detailed observations on the multiple facets of sleep-disordered breathing (SDB) were documented by the ancient Greeks and Romans, not only as to medical aspects but also in the Greco–Roman literature. Albert Esser (1885–1972) gave us an inspiring insight into these ancient sources about SDB. ¹⁶ Snoring may be caused by exogenous factors such as excessive eating and drinking, supine position, dropping of the lower jaw, or an epidemic disease of the nose. Esser also cites endogenous or constitutional factors that induce snoring such as age, small children and old people, or the pyknic or plethoric sleeper with a well-fed neck. In the ancient literature there are some detailed observations of the characteristics of the snoring sounds, including the extremely loud and interrupted snoring and the apneic snoring combined with a stop of breath. The main inspiratory type of snoring was recognized, as were the respiratory effort and body movements associated with snoring.

Comparable body dimensions as “The Sleeping Lady of Malta” can be supposed for Dionysius (360–305 BC), tyrant of Heraclea on the Euxine (Black Sea). He loved eating and Athenaeus (“Deipnosophistae” ca. 200 AD) transmitted Dionysius’ wish on how to die: “One thing for my own self I desire – and this seems to me the only death that is a happy dying – to lie on my back with its many rolls of fat, scarce uttering a word, gasping for breath, while I eat and say ‘I am rotting away in pleasure.’” ¹⁷ Athenaeus also reported that the physicians prescribed that he should get some fine needles, exceedingly long, which they thrust through his ribs and belly whenever he happened to fall into a very deep sleep. Now up to a certain point under the flesh, completely calloused as it was by fat, the needle caused no sensation; but if the needle went through so far as to touch the region which was free of fat, then he would be thoroughly aroused. At last Dionysius was choked by his own fat. ¹⁸

In the 1980s, 2,300 years later, a modern form of bariatric surgery was applied in the United States to successfully treat obesity-related OSA. ¹⁹ A study about 330 patients who were successfully treated by bariatric surgery was published by Kleinhans and Verse. ²⁰

Hippocrates (460–359 BC), clearly described how a nasal polyp causes snoring 2,400 years ago: “When the polyp comes from the nose, hanging down the middle cartilages like a uvula, softly expanding with expiration outside the nose, retracting with inspiration, it causes a croaky voice and snoring during sleep.” ²¹ Hippocrates surgically treated the polyps with loops or using the sponge method which was nicely depicted by Baldewein ²² (▶Fig. 1.1).

In the past, uvula pathology appears to have been a common cause of different symptoms among them those associated with SDB. As the access to the uvula is easy, no wonder it became the scapegoat being sacrificed for all types of diseases for thousands of years as first documented in the Indian Sushruta 3,000 years ago. Hippocrates clearly recommended when to crush the uvula (Prognostic XXII): “It is dangerous to cut away or lance the uvula while it is red and enlarged, … Where, however, … forming what is called ‘the grape,’ that is when the front of the uvula is enlarged and vivid, while the upper part is thinner, it is safe to operate.”

Early instruments to amputate the uvula were excavated in Roman tombs. Usually they were manufactured from bronze. The forceps for crushing the uvula, termed staphylagra, is shown in ▶Fig. 1.2. It was buried in about 275 AD and found in a physician’s tomb in Paris. ²³

Stay updated, free articles. Join our Telegram channel

Join