Published ahead of print on February 5, 2004, doi:10.1164/rccm.200309-1283OC
© 2004 American Thoracic Society Regional Effects of Selective Pharyngeal Muscle Activation on Airway ShapePulmonary, Allergy, and Critical Care Division, Department of Medicine and Center for Sleep and Respiratory Neurobiology, University of Pennsylvania; and Department of Medicine, Philadelphia Veterans Affairs Medical Center, Philadelphia, Pennsylvania Correspondence and requests for reprints should be addressed to Samuel T. Kuna, M.D., Philadelphia VAMC (111P), University and Woodland Avenue, Philadelphia, PA 19104. E-mail: skuna{at}mail.med.upenn.edu
Pharyngeal airway fiberoptic imaging was performed in 10 decerebrate cats to determine the effect of selective pharyngeal muscle activation on airway shape. At intraluminal pressures from 6 to –6 cm H2O, maximum anteroposterior and lateral diameters were measured in the rostral oropharynx, caudal oropharynx, and velopharynx with and without bilateral stimulation of the medial hypoglossus (HG), lateral HG, whole HG, glossopharyngeus, and pharyngeal branch of vagus nerves. At all three airway levels without nerve stimulation, the increase in diameter with increasing pressure was greater in the lateral than anteroposterior dimension. Stimulation of the hypoglossal and glossopharyngeus nerves caused greater increases in lateral than anteroposterior diameter in all three regions with different effects across nerves and regions. Stimulation of these four nerves frequently caused greater increases in both diameters, as the airway cross-sectional area was decreased by lowering airway pressure. Stimulation of the pharyngeal branch of vagus resulted in greater decreases in lateral than anteroposterior dimension in the caudal oropharynx and velopharynx, especially as airway cross-sectional area was increased by increasing intraluminal pressure. The results indicate that selective activation of pharyngeal muscles in cats frequently results in greater changes in lateral than anteroposterior airway diameter and that these effects are dependent on airway region and cross-sectional area.
Key Words: hypoglossus nerve • glossopharyngeus nerve • vagus nerve • velopharynx • oropharynx Pharyngeal muscles that dilate or constrict the airway are thought to play an important role in the pathogenesis of obstructive sleep apnea (1–12). Suppression of motor output to pharyngeal dilator muscles during non-REM and REM sleep promotes airway closure. Reopening of the airway at the termination of obstructive apneas and hypopneas is associated with a burst of pharyngeal muscle activity. Fiberoptic studies from this laboratory in decerebrate cats have examined the effects of selective pharyngeal muscle activation on the cross-sectional airway area at specific regions of the pharynx (13, 14). The results indicate that pharyngeal muscles have different regional effects on the pharyngeal airway. Some muscles are more effective in the rostral airway, whereas others exert their greatest effect in more caudal regions. These results also demonstrated that the effects of pharyngeal muscle activation in a given region of the pharynx are dependent on airway area. Generally, pharyngeal dilating muscles are more effective dilators when the airway area is reduced, and pharyngeal constricting muscles are more effective constrictors when the airway area is increased. Although many animal studies have examined the effects of pharyngeal muscle activation on airway volume, cross-sectional area, and stiffness, little is known about how pharyngeal muscle activation alters airway shape. A previous study from this laboratory found that regional changes in the airway cross-sectional area at atmospheric intraluminal pressure were primarily concentric (13). This finding was unexpected because studies from this laboratory and by other investigators in animals and humans indicate that interventions that increase or decrease airway size are generally associated with greater changes in the lateral than anteroposterior dimension. Because the regional dilating or constricting effects of pharyngeal muscle activation depend on airway cross-sectional area, we hypothesized that dilator muscle activation would produce greater increases in lateral than anteroposterior airway diameter when the airway area was reduced and that constrictor muscle activation would be associated with greater increases in lateral than anteroposterior diameter when airway area was increased. To test this hypothesis, fiberoptic images of the pharyngeal airway in decerebrate cats from a recent study from this laboratory were reanalyzed to determine the regional effects of selective pharyngeal muscle activation on airway shape across a range of intraluminal pressures above and below atmospheric pressure. Differential regional effects of pharyngeal muscle activation on the pharyngeal airway are not unexpected. The pharyngeal airway muscles have complex anatomic relationships and receive their motor output from several different nerves (15). The hypoglossal and glossopharyngeal nerves innervate pharyngeal dilator muscles. The whole hypoglossus (HG) nerve provides motor output to the intrinsic and extrinsic tongue muscles (16). The medial branch of the HG innervates the genioglossus and geniohyoid muscles, both tongue protrudors. The lateral branch of the HG innervates the tongue retractors: styloglossus and hyoglossus. Both HG branches carry motor output to the intrinsic tongue muscle fibers that help shape the tongue. The glossopharyngeus nerve provides motor output to the stylopharyngeus muscle (15). In addition, recent studies indicate that the glossopharyngeus also supplies the levator veli palatini, pharyngeal constrictors, and cricopharyngeus muscles via inputs to the pharyngeal plexus (17, 18). In contrast to these nerves, the pharyngeal branch of the vagus provides the primary motor output to the superior, medial, and inferior pharyngeal constrictor muscles. For this analysis, each of the previously mentioned nerves was stimulated bilaterally to determine the effect of selective pharyngeal muscle activation on airway shape.
Animal Experiments The protocol was performed in 10 decerebrate, tracheotomized, spontaneously breathing adult cats as approved by the Animal Care Committee. Anesthesia and general surgical preparation were described previously (13, 14). The nose and mouth were sealed with a fast-setting dental impression material (Frantz Design, Inc., Austin, TX). Cuff electrodes stimulated the bilateral distal cut ends of the following nerves: medial HG, lateral HG, glossopharyngeus, and pharyngeal branch of vagus. The specific stimulation methods and fiberoptic imaging techniques are detailed in previous studies (13, 14). Video recordings of the fiberoptic images at a given airway level were obtained with and without bilateral stimulation of individual nerves and simultaneous bilateral stimulation of both medial and lateral HG branches (whole HG) over a pressure range of 6 to –6 cm H2O in 2-cm H2O increments (13). The specified pressures were produced by connecting the rostral trachea to adjustable positive (Respironics, Murrysville, PA) or negative pressure (Ametek, Kent, OH) sources.
Data Analysis In further analyses, the predicted mean changes in anteroposterior and lateral measures from the analysis of variance model were used to compare the effect of stimulation of the different nerves on anteroposterior and lateral diameter separately in a given region. These analyses modeled change scores (i.e., change from the no stimulation condition) and assessed the effect of measure, nerve, and pressure on the change scores. Post hoc pairwise comparisons were made using Tukey-adjusted contrasts. After adjusting the p values over the five comparisons (nerves), the significance level was assumed at p < 0.05.
The overall statistical results indicated a significant four-way interaction among the factors of stimulation, region, and nerve (F[8, 1994] = 2.17; p = 0.027), such that the effect of stimulation on anteroposterior versus lateral measures was different for different nerves depending on airway region.
Pressure–Diameter Curves in Different Regions without Stimulation
Effects of Nerve Stimulation at a Given Region on Anteroposterior and Lateral Diameter Stimulation of the medial HG increased anteroposterior and lateral diameters at all three airway levels (p values  0.007; Figure 2) . In the rostral oropharynx, the effects of stimulation on anteroposterior and lateral diameter were greater at lower pressure levels (p < 0.0001). There was a trend for greater increases in lateral than anteroposterior diameter with stimulation at lower pressures (p = 0.092). In the caudal oropharynx, medial HG stimulation had similar effects on anteroposterior and lateral diameters (p = 0.478), with relatively greater increases at lower pressure levels (p < 0.0003). In the velopharynx, the effect of stimulation was fairly consistent at all pressures (p = 0.531) and was similar for both anteroposterior and lateral diameters (p = 0.212).
Stimulation of the lateral HG resulted in significant increases in both anteroposterior and lateral diameters in the rostral oropharynx (p < 0.0001) that were relatively similar across pressure levels (p = 0.113; Figure 2). In the caudal oropharynx, lateral HG stimulation had no practical effect on anteroposterior diameter, whereas there was a significant increase in lateral diameter with stimulation at lower pressure levels (p < 0.0001). In the velopharynx, the effect of stimulation was greater in the lateral than anteroposterior dimension (p = 0.003), and the effect was more pronounced for both measures at lower pressure levels (p = 0.002). Stimulation of the whole HG resulted in a significant increase in anteroposterior and lateral diameter at all three airway levels (all p values < 0.0001; Figure 2). In the rostral oropharynx, stimulation had more of an effect on lateral than anteroposterior diameter (p = 0.0006), especially at lower pressure levels (p = 0.013). Similar results were obtained in the caudal oropharynx. In the velopharynx, no differences were apparent in the increase in anteroposterior and lateral measures with stimulation (p = 0.177). Greater effects occurred in both dimensions at lower pressure levels (p = 0.0002).
Glossopharyngeal nerve stimulation caused a significant increase in anteroposterior and lateral diameters at all three airway levels (p values < 0.0001; Figure 3)
. In the rostral oropharynx, the increases in anteroposterior and lateral diameters were similar (p = 0.908) and relatively constant across pressure levels (p = 0.793). Stimulation caused a greater increase in lateral than anteroposterior dimension in both the caudal oropharynx and velopharynx (p
Changes in Anteroposterior and Lateral Diameter with Stimulation of the Different Nerves in a Given Airway Region Tukey adjusted post hoc tests were performed looking at the differences among the five nerves within a particular region. This was done separately at each of the three regions for the anteroposterior measure and then again for the lateral measure. Because the interaction terms containing these effects were statistically significant, the differences among nerves (and in some cases the interaction with pressure) were examined. Across the pressure range in the rostral oropharynx, stimulation of the medial HG and whole HG produced similar increases in anteroposterior diameter (p = 0.464) that were greater than those resulting from stimulation of either the lateral HG or glossopharyngeus (p < 0.001). The greatest increase in lateral diameter in the rostral oropharynx occurred with stimulation of the whole HG. The next largest increase in lateral diameter occurred with medial HG stimulation (p = 0.004). In the caudal oropharynx, glossopharyngeal stimulation produced the greatest increase in both anteroposterior and lateral diameters (p < 0.001). The increases in anteroposterior diameter in the caudal oropharynx with whole HG and medial HG stimulation were similar (p = 0.912), but greater increases in lateral diameter were present with whole HG compared with medial HG stimulation (p = 0.019). In the velopharynx, no statistically significant differences in change in anteroposterior diameter were present between stimulation of the glossopharyngeus versus stimulation of the whole HG or either of its branches (p > 0.110). In contrast, a greater increase in velopharyngeal lateral diameter resulted from glossopharyngeal stimulation than from any of the other nerves (p < 0.0001). Stimulation of the medial HG and whole HG had similar effects on the increase in anteroposterior (p = 1.000) and lateral diameters (p = 0.153) in the velopharynx.
Changes in Anteroposterior and Lateral Diameter with Stimulation of the a Given Nerve across Airway Regions
This is the first report to examine systematically the effects of pharyngeal muscle activation on the shape of the airway. The findings extend a previous report, using the same set of animals, by demonstrating that dilator muscles have differential effects on the pharyngeal airway, not only in terms of changes in airway area but also with regard changes in airway shape (14). Under passive conditions (i.e., no nerve stimulation), airway area in the rostral oropharynx, caudal oropharynx, and velopharynx increased with increasing pressure. This analysis shows that the increase in area under passive conditions was due to a greater increase in lateral than anteroposterior dimension at all three airway levels. In the previous report, the effect of muscle activation on airway area depended on nerve, region, and airway pressure (14). Stimulation of the medial, lateral, and whole HG nerves increased the airway area in all three regions examined with the greatest effect in the rostral oropharynx. Stimulation of the glossopharyngeal nerve also increased airway area in all three regions, with the greatest effect in the caudal oropharynx. The current analysis indicates that in the majority of cases the increases in airway area were associated with greater increases in lateral than anteroposterior diameter. Greater increases in lateral than anteroposterior dimension occurred in the rostral oropharynx with whole HG stimulation and in the caudal oropharynx and velopharynx with glossopharyngeus, whole HG, and lateral HG stimulation. In contrast to the other nerves, stimulation of the pharyngeal branch of the vagus decreased airway area in the caudal oropharynx and velopharynx (14). The results of this analysis reveal that the decreases in airway area were due to greater decreases in lateral than anteroposterior dimension, especially at higher pressures. The results indicate that across nerves and regions. selective stimulation of pharyngeal muscles frequently results in greater changes in the lateral than anteroposterior airway diameter. Figure 4 is a schematic that illustrates the changes in airway lateral and anteroposterior dimension that accompanied the changes in airway area. The increase in airway area with dilator muscle stimulation was either due to a concentric increase in airway dimensions or a greater increase in the lateral compared with the anteroposterior dimension. Stimulation of a given nerve had different effects depending on airway region, and within a given region, different effects were observed with stimulation of the individual nerves. For the nerves innervating dilator muscles (medial HG, lateral HG, whole HG, glossopharyngeus), nerve stimulation caused greater increases in lateral compared with anteroposterior dimension in 7 of the 12 nerve/region combinations (three regions, four nerves to dilating muscles). As reported previously, the change in airway area with selective pharyngeal muscle activation depends on airway size. Activation of the whole HG and its individual branches generally caused greater increases in airway area when intraluminal pressure was subatmospheric, that is, when airway size was reduced (14). The increase in airway area in all three regions with glossopharyngeal stimulation was uniform across the pressure range. These results show similar interactions between pressure and the changes in lateral and anteroposterior dimension. In 9 of 12 nerve/region combinations (three regions, four nerves to dilating muscles), stimulation caused greater increases in lateral diameter when airway size was reduced. In seven of these nine combinations, a similar interaction was observed between airway size and anteroposterior diameter. In contrast, stimulation of the pharyngeal branch of the vagus was associated with greater decreases in airway area and lateral diameter at all three levels when airway size was increased by intraluminal pressures above atmospheric.
The results also demonstrate that the change in airway shape with selective muscle stimulation depends on airway intraluminal pressure, that is, the size the airway prior to stimulation. Figure 5 illustrates the two most common responses observed with stimulation of nerves providing motor output to pharyngeal dilators. In one case, the increase in diameter with muscle stimulation is independent of pressure (i.e., airway size). The pressure–diameter relationships with and without muscle stimulation are parallel, and stimulation is associated with a uniform increase of the particular airway dimension that is independent of airway size. In the other case, the change in diameter is dependent on airway pressure, with greater increases in diameter when airway size is reduced by lowering intraluminal pressure. These two responses were observed across nerves and regions in both the anteroposterior and lateral dimension. The increased mechanical effectiveness under these conditions would be advantageous in helping to restore airway patency when airway size is narrowed or closed. Such a condition would occur at the termination of an obstructive apnea when a large burst of motor output to pharyngeal airway muscles on arousal reopens the pathologically narrowed or closed pharyngeal airway.
The slope of the pressure–diameter relationship could be interpreted as an "effective compliance" of the pharyngeal airway. However, the greater changes in lateral compared with anteroposterior airway dimension do not necessarily mean that the lateral airway walls are more "compliant." Caution is advised in making any inferences about the mechanical properties of the pharyngeal walls based on fiberoptic or radiographic imaging of the airway. Greater expansion of the airway in the lateral direction could conceivably result from an unfolding of the lateral walls of the airway. Instead of acting directly on the lateral walls, the pharyngeal dilating muscles may be exerting their primary effect on the hyoid apparatus, displacing the anterior pharyngeal wall ventrally and causing a secondary expansion of the lateral airway. This possible explanation is similar in concept to inflating a globe-shaped Chinese lantern from a closed position—its initial opening being greatest in the dimension perpendicular to the applied force. Techniques that assess tissue motion in the pharyngeal wall are needed to determine the mechanical properties of the pharyngeal wall and the response of the lateral and anteroposterior pharyngeal walls during interventions that expand and collapse the pharyngeal airway. This study supports several human studies reporting that interventions that change airway size are associated with greater movement in the lateral than anteroposterior direction. Kuna and colleagues (21) studied the effect of nasal continuous positive airway pressure on upper airway size and configuration by computerized tomography in obese adults with obstructive sleep apnea and normal adults during wakefulness. Linear increases in airway area that occurred across a pressure range of 0 to 15 cm H2O were primarily due to lateral expansion of the airway. Schwab and colleagues reported similar findings in the retropalatal and retroglossal regions over a similar pressure range (22). Neither of these studies tested the response to subatmospheric pressure. Other studies in human subjects emphasize the importance of the lateral airway dimension and lateral pharyngeal walls in determining airway size. Schwab and colleagues performed a magnetic resonance imaging of the pharyngeal airway in normal adult humans and patients with obstructive sleep apnea and found that the airway narrowing in patients with obstructive sleep apnea was predominantly in the lateral dimension and was associated with lateral pharyngeal wall thickening (23). These investigators also studied changes in airway dimension on magnetic resonance imaging during spontaneous breathing in awake normal subjects, snorers, and patients with obstructive sleep apnea (24). Lateral dimensional changes during respiration were larger than the anteroposterior dimensional changes in all subject groups at most of the pharyngeal regions examined. Trudo and colleagues (25) performed magnetic resonance images of the pharynx in normal adults during wakefulness and sleep and found a concentric decrease in pharyngeal airway size and a significant increase in lateral pharyngeal wall thickness in the retropalatal region during sleep. Wheatley and colleagues (26) measured dimensional changes in the oropharyngeal airway by fluoroscopy in awake supine normal adults during exposure to negative intraluminal pressures generated by external suction at the mouth during voluntary glottic closure without inspiratory effort. There was a substantial narrowing in both anteroposterior and lateral diameters of the pharyngeal airway at the C4 vertebral level during the maneuver. It is difficult to compare the previously discussed findings in humans with the current findings in cats. The fiberoptic technique used in this study only provided information about airway size and did not allow assessment of lateral wall thickness. Changes in pharyngeal muscle activity were not measured in the human studies. The results may be of potential clinical importance because they relate to the treatment of obstructive sleep apnea. Previous studies have attempted to prevent pharyngeal closure during sleep in humans by stimulating the medial HG and genioglossus muscle (27, 28). Despite this focus on tongue protrudor muscle(s), clinical studies indicate that the pharyngeal airway closes at different locations in patients with obstructive sleep apnea, and complex patterns of closure involving multiple pharyngeal segments are know to occur (29, 30). Of the nerves tested in this study, stimulation of the glossopharyngeal nerve and HG nerves exhibited dilating effects not only in the rostral but also in the caudal pharyngeal airway. Although the effects of medial HG stimulation on lateral and anteroposterior diameter in the caudal oropharynx and velopharynx were relatively small, even small changes in diameter can have important effects on airway area and therefore airflow resistance. However, the regional differences seen in this study suggest that stimulation of one nerve or muscle may not be uniformly effective in preventing upper airway closure in patients with obstructive sleep apnea. Development of a therapeutic approach that activates overall motor output to pharyngeal dilator muscles might be preferable to the selective activation of a particular muscle or muscle group. In summary, changes in pharyngeal airway shape with selective activation of pharyngeal muscles are dependent on region of the pharyngeal airway, specific pharyngeal muscles activated, and airway cross-sectional area. The increase in airway area with selective activation of pharyngeal muscles is generally associated with greater changes in lateral than anteroposterior diameter. In many cases, the increase in lateral airway dimension with selective activation of dilator muscles became greater at airway pressures below atmospheric pressure, that is, when the pharyngeal airway size is decreased. Techniques need to be developed that will assess how changes in pharyngeal wall tissue motion effect the observed changes in airway size and shape.
Allison McCormick and Sharif Branham provided technical assistance. Jacqueline Cater provided statistical assistance.
Conflict of Interest Statement: S.T.K. has no declared conflict of interest. Received in original form September 15, 2003; accepted in final form February 4, 2004
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
