| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
Dr. Brown is Medical Director, New Mexico Center for Sleep Medicine, Lovelace Health Systems, and The Division of Pulmonary, Allergy, and Critical Care, Department of Internal Medicine, University of New Mexico School of Medicine.
Correspondence to: Lee K. Brown MD, FCCP, Medical Director, New Mexico Center for Sleep Medicine, Lovelace Health Systems, 4700 Jefferson Blvd NE, Suite 800, Albuquerque, NM 87109; e-mail: kbrown{at}alum.mit.edu
There is measure in all things.
"Satires" Book 1, Horace
Horace was certainly not referring to sleep apnea hypopnea syndrome (SAHS) when he composed this well-known aphorism, even though there are descriptions of this disorder dating back to antiquity. Nevertheless, it is an apt introduction to a discussion of the article by Dempsey et al appearing in this issue of CHEST (see page 840), which represents the latest contribution to an already extensive literature describing measurements of upper airway anatomy in patients with SAHS.
The story of anatomy and SAHS begins with the observation that periodic failure of upper airway dilator muscle activity is a key to pathogenesis.1 2 From the very beginning, however, abnormal upper airway anatomy was also suspected as contributory to the syndrome. In fact, the earliest polysomnographic reports concentrated on sleep apnea in the pickwickian syndrome,3 and a common etiologic speculation centered on reduced upper airway caliber caused by adjacent accumulations of adipose tissue. Not long thereafter, reports of SAHS in patients with specific facial dysmorphologies appeared, such as the recessed mandible one author referred to as the "Andy Gump appearance,"4 and another referred to as "bird-like face syndrome."5 Differences in airway caliber between individuals with and without SAHS were also reported by, among others, Haponik and colleagues6 and Bradley et al.7 More or less simultaneously, nasal continuous positive airway pressure was introduced as a treatment modality,8 followed a few years later by the clever application of continuous negative airway pressure as an experimental technique.9 The result, as elegantly summarized by Gold and Schwartz,10 was the critical pressure theory.
Critical pressure (Pcrit) is an attribute of a fluid mechanical system consisting of a collapsible segment of tubing interposed between two rigid segments, all of which are enclosed in a sealed container. Intensivists will recognize this as the three zones of the pulmonary vasculature à la John West, while pulmonary physiologists should see visions of flow limitation in the small airways; it is a ubiquitous concept with wide application in biology. When applied to obstructive sleep apnea, the oropharynx and hypopharynx together represent the collapsible segment, and the Pcrit is an inherent property of the collapsible segment analogous to the transluminal pressure required to open the collapsible segment. Schwartz and colleagues9 demonstrated that normal individuals without sleep-disordered breathing require continuous negative airway pressures of approximately -9 cm H2O in order to induce obstructive apnea during sleep. Thus, the Pcrit of a normal sleeping subject is - 9 cm H2O (analogous to West’s zone 3); that is, the pharynx of such an individual has an inherent tendency to resist collapse. A patient who exhibits obstructive apneas during sleep, however, requires a positive degree of continuous airway pressure in order to prevent pharyngeal collapse; this individual has a positive Pcrit (West’s zone 1), and a pharynx that requires active intervention (dilator muscle tone) in order to stay patent. Patients with an intermediate value of Pcrit11 exhibit inspiratory flow limitation that may manifest as simple snoring, the respiratory effort-related arousals of the upper airway resistance syndrome or as hypopneas (West’s zone 2).
What then determines Pcrit? The inherent tendency of the pharynx to collapse presumably depends on body position and pharyngeal compliance and anatomy. Compliance may be largely controlled by upper airway dilator muscle tone,12 but there may also be contributions from local pathology, eg, trauma from snoring.13 14 Gross descriptions of facial morphology or aggregate measurements of upper airway size may serve as a first approximation of the anatomic contribution to Pcrit, but have proven insufficient to characterize the specific anatomic abnormalities that lead to clinically significant SAHS. Consequently, an extensive literature correlating measurements obtained by radiographic imaging of the upper airway with severity of SAHS has emerged. The article currently under discussion represents the latest, but surely not the last, of these studies.
In this investigation, Dempsey and coauthors analyzed upright and supine lateral cephalometric radiographs, body mass index (BMI), and polysomnographic variables in patients with obstructive sleep apnea and working adults with and without known habitual snoring. Using the desaturation index (DI) as the dependent variable, they report that six of the cephalometric measurements together predicted about one fourth of the variability in DI, and BMI predicted another one fourth. Of the six cephalometric measurements, the horizontal distance from the porion vertical to the supradentale (PV-A) was the most important factor; this is a measure of maxillary (mid-facial) sufficiency. Another cephalometric index reflective of midfacial dimensions (distance from posterior nasal spine to spheno-occipital point or PNS-SO), and a measurement of posterior airway space at the level of the midface (PAS1) together with PV-A, accounted for about 18% of the variability in DI. The two measurements associated with mandibular sufficiency (the angle defined by the sella, nasion, and deepest anterior point in the concavity of the anterior mandible or S-N-B and the angle defined by the gonion, spheno-ethmoidal point and posterior nasal spine or Go-SE-PNS) and a measurement of hyoid bone position (the distance from the anterior-superior hyoid point to the line joining the retrognathion and the anterior inferior point on the third vertebra or H-[Rgn-C3]) accounted for a small percentage of the variability. Interactive terms combining BMI with five of the six cephalometrics added approximately 15% to the prediction of variance. Age and the cephalometrics were the only important predictors of variance in nonobese individuals, while BMI and cephalometrics continued to account for half of the variance of DI in obese subjects. Finally, receiver operating curve analysis identified a cutoff value for PV-A that provided the best ability to predict normal vs abnormal breathing during sleep: a PV-A value < 97 mm identified a threefold and fivefold increase in risk of mild sleep-disordered breathing in obese and nonobese individuals, respectively, and a threefold and 7.5-fold increase in moderate-to-severe sleep- disordered breathing.
Cephalometric radiographs in SAHS have been reported in numerous publications. Abnormalities that have been associated with increased SAHS severity were summarized a few years ago by Schwab,15 and consist of mandibular and maxillary deficiency, reduced dimension of the posterior airway space (equivalent to PAS2 in the current article, measured at the base of the tongue), enlarged tongue, enlarged soft palate, and caudally displaced hyoid.15 These studies varied widely in quality, however. As pointed out by Miles and colleagues,16 control groups may be matched, unmatched but from a similar population as the patients, a convenience sample from some other population, or only compared to normative data from the literature; and sample size may be quite small. However, a virtual blizzard of more recent studies has largely borne out these earlier results.17 18 19 20 21 22 23 24 25 26 A meta-analysis of earlier studies16 found that only the distance from the gonion to the gnathion (Go-Gn), a measure of the lateral dimension of the mandible, consistently predicted SAHS. However, even a cursory review of the numerous studies available to date finds that distance of the hyoid from the mandibular plane, increased soft palate and tongue, and mandibular and maxillary deficiency are mentioned so many times as correlating with degree of sleep-disordered breathing that a new meta-analysis would likely prove their significance as well.
Statistical association does not, however, prove causality. While it is extremely likely that changes in upper airway anatomy lead to SAHS, it is also at least plausible that preexisting upper airway obstruction, particularly during childhood, can cause developmental changes in upper airway morphology.27 28 Fortunately, several studies reinforce the likelihood that the anatomic changes identified by cephalometry do promote SAHS. Sforza and colleagues29 studied 57 male patients with SAHS and were able to demonstrate significant associations between Pcrit and both soft palate length and hyoid bone position. Shortly thereafter, Kato et al30 reported on six patients with SAHS who were subjected to paralysis and general anesthesia for measurement of Pcrit; determinations were made with three progressive degrees of mandibular advancement produced by an appliance. They were able to demonstrate a dose-dependent reduction in Pcrit with increasing amounts of mandibular advancement, thus establishing a direct causal relationship between mandibular anatomy and Pcrit.30
Is there "measure in all things," or at least in the anatomic basis of SAHS? The progressive contributions to our knowledge of anatomy and SAHS support the notion that function follows form, but gaps remain in our knowledge. First, the data so far published on cephalometric measurements and their correlation with SAHS severity are voluminous, use diverse measurements and techniques, and at times disagree with respect to the specific anatomic markers that most impact on SAHS. Consequently, a meta-analysis combining the primary data of all of these investigators could bring needed clarity to the field. Second, the important findings of Kato et al30 require replication, both using external manipulations of anatomy and by correlating cephalometric findings with Pcrit. Finally, pertinent cephalometric measurements should be incorporated into prediction models for SAHS, since existing models using clinical variables analyzed by logistic regression, neural networks, and so forth31 32 33 have largely failed to provide the necessary degree of accuracy. Until we address these issues, we may remain confident that there is "measure in all things"—including SAHS—but we must more accurately identify what to measure.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
