INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Obstructive sleep apnea hypopnea syndrome (SAHS) is characterized by recurrent episodes of complete or incomplete upper airway (UA) obstruction that are associated with repetitive transient hypoxia and arousals leading to sleep fragmentation. The prevalence of SAHS is 4% in men and 2% in women in the middle-aged active population (1). SAHS is associated with increased morbidity (2, 3) and mortality (4, 5) related to the cardiovascular and neurocognitive consequences of the disease.

The maintenance of UA patency throughout the respiratory cycle depends on the balance between dilating and collapsing forces and compliance, which is influenced by several factors such as UA shape and dimension (6), mucosal characteristics (7), upstream resistance, the characteristics of the force generated by the contraction of dilator muscles (8), and compliance (9). Therefore, the force generated by the contraction of dilator muscles plays an important role in the control of UA patency.

The musculus uvulae (MU) is located in the uvula and there is indirect evidence that it can be considered a UA dilator because (1) its contraction shortens the uvula, (2) this results in an anterior hooking of the uvula during the course of an obstructive apnea (10), (3) this hooking of the uvula precedes upper airway collapse (11), and (4) the contraction of this muscle shortens and stiffens the soft palate in humans (12). We have found that the force generated by the MU in vitro is greater in patients with SAHS than in snorers (13); we also reported that the proportion of fast twitch muscle fiber and the level of aerobic enzyme activity are higher in the former group. Similar results were obtained for the genioglossus muscle (14). These characteristics could be the result of an adaptive process in response to a long-term, high-intensity resistive loading of the UA because it is indeed admitted that the awake electromyographic (EMG) activity of UA dilator muscles is higher in patients with SAHS than in unaffected individuals (15). This could account for the physiological, biochemical, and histochemical differences in UA muscle characteristics observed between patients with SAHS and nonapneic snorers (14).

The influence of sleep on the force developed by UA dilator muscles is not clearly established. A detrimental effect of sleep on the reflex-mediated increase in UA EMG activity in response to negative pressure has been reported in different muscles (16, 17). This effect could interfere with the ability to generate an adaptive dilating force and then to maintain UA patency. This is partly supported by the results of Mezzanotte and coworkers (18), who found that the sleep onset-related decrease in tensor palatini (TP) activity is greater in patients with SAHS than in unaffected subjects. However, because the awake EMG activity was significantly higher in patients with SAHS, sleeping TP activity was still higher in patients with SAHS than in unaffected subjects. Furthermore, a significant increase in EMG activity was observed in 3 of 10 patients with SAHS during sleep. We have found that the force generated by MU contraction in vitro is positively correlated with UA collapsibility measured during sleep, suggesting that the force developed by this muscle is adapted to UA collapsibility. Therefore, sleeping UA muscle activity also contributes to determine the physiological adaptive response to resistance training that these muscles are facing.

On the basis of these results, it appears that UA dilator muscle activity is increased in patients with SAHS, both during wakefulness and sleep, compared with snorers. Even if dilator muscle activity is greater during sleep in patients with SAHS than in unaffected subjects, the sleep-induced decrease in this activity is sufficient to reduce significantly the dilating force that is developed and compromise UA patency. This could be explained by alterations in the mechanical efficiency of UA dilator contraction on tissue shortening and/or displacement in these patients compared with nonapneic subjects. Such impediment could be due to differences in structure of nonmuscular tissue surrounding UA muscles, which could interfere with the transmission of the dilating force to anatomical structures. The aim of the present study was to study the mechanical coupling between muscular and nonmuscular components of the uvula by comparing uvular tissue elastance and uvular shortening during in vitro MU stimulation in patients with SAHS and nonapneic snorers.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty-three men (age, 42 ± 10 yr; body mass index, 31.3 ± 4.1 kg/m²) who had undergone a uvulopalatopharyngoplasty (UPPP) for the treatment of SAHS or nonapneic snoring were included in the study. None of the patients with SAHS had been treated at the time of surgery. Subjects had no medication and their thyroid function was normal. All subjects had a conventional sleep recording within the month preceding UPPP.

Sleep Studies

Sleep studies were done according to conventional polysomnographic techniques (19, 20). Nasal and mouth flows were measured with thermocouples (Grass Instruments Co., Quincy, MA); electrocardiograms were performed; thoracoabdominal movements were monitored by respiratory inductive plethysmography (Respitrace; Ambulatory Monitoring, Ardsley, NY) (21); arterial oxyhemoglobin saturation was measured with an ear oximeter (Criticare 504; CSI, Waukesha, WI); and breathing sounds were recorded with two microphones placed at the head of the bed (22). All variables were collected on computer.

Muscle Sampling

Uvular tissue was resected during conventional UPPP procedures as previously described (14). Caution was taken not to modify the local vascular tone, which could influence muscular characteristics as well as soft tissue properties. For these reasons, no vasoactive agents were injected in the soft palate, or in the uvula before it had been resected. For the same reason, no electrocautery was used during that part of surgery. Resected tissue was placed in ice-cold Krebs-Ringer buffer and immediately brought to the laboratory of one of the investigators (C.C.). Measurements were made within 20 min of the uvular resection.

Measurement of Uvular and Musculus Uvulae Properties

Measurements were made in Krebs-Ringer solution maintained at 35° C and supplemented with glucose and curare (20 µg/ml). The solution was stabilized at pH 7.4 by equilibration with a 95% O₂-5% CO₂ gas mixture. One end of the uvula was sutured to a rigid support and the other end to a dual-mode servo system (force and displacement transducer, model 300 H; Cambridge Technology, Cambridge, MA). Supramaximal stimuli were delivered through platinum field electrodes. The output of the stimulator was adjusted to obtain an ~ 25-V stimulation in the bathing solution.

The optimal length (Lo), which corresponds to the length at which maximal isometric twitch tension was reached under supramaximal stimulation, was determined first. The maximum tetanic tension (Po) was then determined. Passive specific uvular elastance was quantified by measuring the passive tension/length relationship of uvular tissue; tissue lengthening was measured during a progressive stretching of the uvular piece up to 140% of Lo by 5% steps and the slope of the tension/length relationship (g/% Lo) was calculated (23). Active uvular elastance was estimated using the rate of force production (dP/dt) during a maximum tetanic contraction and expressed as percent Po/ms. MU contraction-related uvular shortening was measured during isotonic stimulation, using the stimulation parameters required to develop Po and a fixed load equal to 50% Po; it was expressed as a percentage of Lo. The uvular tissue was then dissected to isolate MU muscle. MU Lo and Po values were determined as previously described and MU elastance was determined in seven subjects.

The density of uvular tissue was calculated as the mass/volume ratio. Tissue volume was obtained by measuring the volume of Krebs buffer displaced by the uvular tissue when immersed in a graduated cylinder allowing readings with a precision of 0.125 ml. MU lactate was measured in two patients with SAHS and two nonapneic snorers. Lactate content was determined on neutralized PCA extract using a spectrophotometric method (24).

Statistical Analysis

Passive uvular stiffness, uvular and isolated MU tetanic tensions, active uvular stiffness, and uvular shortening during isotonic stimulation characteristics were compared in patients with SAHS and nonapneic snorers with an unpaired t test and a Bonferroni correction (significant difference inferred for p values < 0.01). Correlation between uvular tissue characteristics (uvular shortening and passive uvular stiffness) and nocturnal breathing disorders was studied by least-squares regression analysis.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The diagnosis of SAHS was made for 15 patients whose apnea indexes were > 5/h and whose apnea + hypopnea indexes were > 15/h; the other 8 subjects were classified as nonapneic snorers. Table 1 presents the characteristics of both groups of subjects. Patients with SAHS all complained of diurnal hypersomnolence as illustrated by the high Epworth sleepiness score (normal, < 10) (25) (Table 1); this score was normal in the nonapneic snorer group. No difference was found in age or body mass index between the two groups. There was no difference in uvular weight (1.1 ± 0.2 g in patients with SAHS and 1.3 ± 0.4 g in snorers) or in uvular density (0.99 ± 0.13 and 0.95 ± 0.20 g · cm³, respectively) between the two groups.

Uvular tissue optimal length was similar in the two groups (22.4 ± 3.9 mm in patients with SAHS and 24.5 ± 3.0 mm in snorers, p = 0.2). No difference was found in maximal tetanic tension produced by uvular tissue between the two groups (47.2 ± 14.8 g in patients with SAHS and 39.1 ± 16.5 g in snorers, p = 0.2). Individual values of passive specific uvular elastance in patients with SAHS and nonapneic snorers are presented in Figure 1. Specific tissue elastance was significantly less in nonapneic snorers (0.36 ± 0.27 g/% Lo) compared with patients with SAHS (0.84 ± 0.39 g/% Lo) (p = 0.006). The difference between the two groups persists when the apneic outlier is removed from the analysis (0.76 ± 0.26, p = 0.004). For technical reasons, measurements of uvular shortening during supramaximal stimulation with 50% Po loading could be measured in only five patients with SAHS and three nonapneic snorers. In this subset, there was no difference in uvular shortening between the two groups, whether it was expressed in millimeters (3.0 ± 1.3 mm in snorers and 2.1 ± 1.0 mm in patients with SAHS) or in percentage of Lo (12.1 ± 6.5 and 10.6 ± 4.0% Lo, respectively). Interestingly, there was a significant negative correlation between uvular shortening and passive uvular specific elastance (r = 0.69, p = 0.05). Active uvular elastance was 0.86 ± 0.17% Po/ms in patients with SAHS and 0.84 ± 0.13% Po/ms in nonapneic snorers (p = 0.5).

View larger version (10K): [in this window] [in a new window]   Figure 1.   Individual values of passive specific uvular elastance measured in patients with SAHS and nonapneic snorers. Elastance was significantly lower in snorers than in patients with SAHS. *p < 0.05.

MU lactate was 2.76 ± 0.91 µmol · g¹. Maximal tetanic tension developed by isolated MU was higher in patients with SAHS than in nonapneic snorers (45.8 ± 23.1 and 30.0 ± 8.3 g, respectively), this difference being borderline significant using a Bonferroni correction (p = 0.04). Maximal tetanic tension normalized for the muscle area was similar in the two groups (6.4 ± 4.0 N/cm² in patients with SAHS and 5.6 ± 2.4 N/cm² in snorers, p = 0.7). No difference was found in passive MU specific elastance between the two groups (0.43 ± 0.14 and 0.60 ± 0.03 g/% Lo, respectively).

To evaluate if uvular tissue mechanical properties could correlate with anthropomorphic and sleep-related breathing characteristics, regression analysis was done between these different variables. A strong positive relationship was found between the apnea index and specific uvular elastance (r = 0.57, p = 0.005) (Figure 2). This parameter was also positively correlated with neck circumference (r = 0.48, p = 0.02). No correlation was found between uvular characteristics and body mass index.

View larger version (11K): [in this window] [in a new window]   Figure 2.   Relationship between the apnea index and passive specific uvular elastance. Open circles represent nonapneic snorers and closed circles represent patients with SAHS. There is a strong positive correlation between these parameters (r = 0.57).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Important studies on the pathophysiological mechanisms of SAHS have previously focused on various factors that modify UA collapsibility, on the influence of UA anatomical characteristics, and on the UA dilator EMG activity during wakefulness and sleep. To our knowledge, the present work is the first to evaluate the importance of UA tissue mechanical properties in the occurrence of sleep-related obstructive breathing disorders.

The study of uvular tissue is biologically significant because it is known that (1) UA closure occurs at the velopharyngeal level in the majority of patients with SAHS (26), (2) MU muscle, which is present in the uvula, has the same EMG pattern as that of other UA dilator muscles (27), and (3) because these tissues can be dissected intact and used for in vitro testing after UPPP resection. We have used these tissues to characterize MU muscle physiological, histochemical, and biochemical features in patients with SAHS and snorers (14), to compare MU and genioglossus characteristics in these subjects (15), and to determine the influence of these characteristics on UA collapsibility (28).

The results obtained for MU in the present study are consistent with those previously reported, with a higher maximum tetanic tension in patients with SAHS than in nonapneic snorers (14). However, our present results demonstrate that mechanical properties of UA tissues dramatically interfere with the net effects of dilator muscle contraction on tissue displacement. This is clearly demonstrated by the significant difference found in uvular elastance between patients with SAHS and nonapneic snorers, by the negative correlation that exists between passive uvular elastance and uvular shortening, and by the absence of any difference in uvular maximum tetanic tension between the two groups even if the tension developed by MU alone is higher in patients with SAHS. These sets of data are consistent, one with the other, demonstrating that mechanical properties of UA tissues dramatically interfere with the net effects of dilator muscle contraction on the tension developed by UA tissue. Because specific elastance evaluates tissue stiffness, these two important results imply that UA tissues are stiffer in patients with SAHS than in snorers, and that the higher the tissue stiffness the smaller the tissue displacement during muscle contraction for a given tissue load. This can explain the apparent paradox that UA closure occurs in these patients even if their UA dilator muscles are able to develop greater tension.

The increase in tissue elastance observed in patients with SAHS could be related to histological differences in UA tissues, compared with nonapneic snorers. In an animal model of SAHS, Petrof and coworkers have reported a significant increase in connective tissue content in sternohyoid muscle of apneic bulldogs compared with control dogs (29). Data obtained with magnetic resonance imaging in this animal model demonstrated the presence of edema and fibrosis in UA muscles compared with nonrespiratory muscles (30). Data from analysis of nasal lavage in humans suggest the presence of mucosal inflammation in patients with SAHS compared with unaffected individuals, with an increased proportion of polymorphonuclear leukocytes and proinflammatory mediators in the former group (31). Histopathological evidence of interstitial fibrosis and intrafascicular interstitial infiltration has also been observed in the soft palate and uvula of patients with SAHS and snorers (32). These processes possibly represent nonspecific consequences of UA tissue injuries, leading to inflammatory cell infiltration and the development of muscle and soft tissue fibrosis (33). Such an increase in tissular edema, inflammation, and fibrosis could account for our results by increasing the stiffness of muscular and perimuscular structures, thus impeding the transmission of the produced tension to nonmuscular tissues during muscle contraction and altering the effects of UA dilator activation on tissue displacement and/or shortening. The occurrence of obstructive breathing disorders could therefore be accounted for by an abnormal linkage between UA muscle and perimuscular tissues. The importance of this phenomenon is illustrated by the positive relationship that we found between tissular elastance and the apnea index. However, it is not possible to determine which is the primary dependent variable in this relationship: increasing the frequency of obstructive breathing disorders should promote tissular injury, inflammation, edema, and fibrosis; this latter factor should alter the mechanical effect of muscle contraction, which in turn should worsen sleep-related UA instability and the occurrence of UA closure in a self-worsening phenomenon.

It could be argued that the histological abnormalities may not play such an important role in SAHS pathophysiology because patients with SAHS and nonapneic snorers do not qualitatively differ in their histopathologic findings (36). A similar comment could be adressed to adipose tissue content because uvular density was similar in the two groups. However, we believe that the mechanical consequences of these abnormalities may not be related to the extent of these histologic disorders but to their distribution within soft tissues, particularly around and within UA dilator muscles. This hypothesis is supported by the similar values for weight and density of uvular tissues obtained in patients with SAHS and snorers in the present study.

It should be asked how our results can be interpreted, considering the increase in UA compliance that characterizes patients with SAHS. UA occlusion occurs at the velopharyngeal level in the majority of these patients (26); the increase in UA compliance observed in OSA is due to the propensity of the velar and uvular structures to collapse with the posterior UA wall in an anteroposterior movement. In the present study we measured the longitudinal uvular elastance because this axis corresponds to the orientation of MU. In this regard, the increase in longitudinal uvular elastance may not change the tendency of the velopharyngeal structure to be passively sucked in by the transpharyngeal negative pressure gradient and its posterior movement leading to UA closure. This posterior movement will be further enhanced if velopharyngeal anatomical abnormalities decrease the UA caliber (Bernoulli principle). When the activity of UA dilator muscles progressively increases, the increase in longitudinal elastance impedes the mechanical effects of muscle contraction (tissue shortening and or stiffening), therefore decreasing the ability of UA dilators to maintain UA patency. These results suggest that the elastance of UA tissues measured along the axis where the dilating forces are applied may actually increase the UA anteroposterior compliance.

We conclude that tissue elastance significantly differs between patients with SAHS and nonapneic snorers, and that it influences the mechanical effects of MU contraction. These UA tissue characteristics may be involved in the pathophysiology of UA by increasing UA compliance. These data could help explain the poor diagnostic value of UA anatomical findings in the diagnosis of SAHS (37), because their repercussions on the tissular effects of UA dilator muscle contraction may not be related to anatomical characteristics but to differences in the histological architecture of UA tissue.