INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The obstructive sleep apnea syndrome (OSAS) is a major public health problem (1). It is a frequent disease that affects about 2% of women and 4% of men, and it associates significant morbidity and mortality (2). OSAS is characterized by the repetition of episodes of pharyngeal closure during sleep. The pathogenesis of these episodes, which may occur hundreds of times each night, is not fully understood (5). However, it is accepted that the maintenance of pharyngeal patency depends on the equilibrium between several occluding and dilating forces (5). Among the latter, the activity of the pharyngeal dilator muscles is of fundamental importance (5). The genioglossus (GG) is the most important pharyngeal dilator muscle (6, 7). As such, therefore, it is conceivable that the GG might play a role in the pathogenesis of OSAS.

Hypothetically, the GG can be abnormal in OSAS by two different mechanisms. On the one hand, a primary myopathy of the GG may enhance pharyngeal collapsibility in these patients. Alternatively, because each episode of pharyngeal obstruction finishes with a vigorous contraction of the GG (5), a secondary process of muscle injury and repair may ensue. In turn, this may jeopardize the mechanical properties of the GG and render the pharynx more collapsible (8).

Treatment with continuous positive airway pressure (CPAP) is highly effective in OSAS (9). In these patients CPAP acts as an effective pneumatic splint that prevents pharyngeal collapse (9, 10). As such, CPAP effectively rests the GG (9, 11, 12). It is well established that a transition between different muscle fiber types occur in response to changes in the pattern and/or level of muscle activation (8). Therefore, CPAP has the potential to revert a hypothetical GG dysfunction, particularly if this was the consequence (rather than the cause) of the episodes of pharyngeal closure that characterize OSAS.

In this study, we hypothesized that the GG may be abnormal (both functionally and structurally) in patients with OSAS. To test this hypothesis, we compared the in vitro contractile properties and the fiber type distribution of GG samples obtained from subjects with and without OSAS. To distinguish a potential primary GG myopathy (where no reversibility after CPAP should theoretically be expected) from a secondary remodeling process of the muscle (where, at least some, reversibility should be expected after CPAP), we also investigated GG samples obtained from patients with OSAS who had been under treatment with CPAP for at least 1 yr.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We studied 16 patients with OSAS and 11 control subjects. We included only male subjects to avoid any influence of sex upon muscle contractile properties and/or structure (13). The diagnosis of OSAS was established by polysomnography (Ultrasom Nicolett, Madison, WI), which included recording of oronasal flow by thermocouples mounted on nasal prongs, thoracoabdominal movements by strain gauges, electrocardiogram, submental electromyogram, electro-occulogram, electroencephalogram (C₄ A₁, C₃ A₂), and oxygen saturation (Criticare Systems Inc., Milwaukee, WI). Nine of the 16 patients with OSAS were studied at the time of diagnosis, before any treatment had been started. The remaining seven patients with OSAS were studied after having been treated with CPAP (Respironics; REM Star, Madison, WI) for at least 1 yr. Compliance with treatment was confirmed by the timer built into the CPAP device. Control subjects were volunteers that required subglottic surgery because of malignant (n = 10) or benign lesions (n = 1). In all of them, OSAS was excluded clinically, according to the criteria of Kapuniai and colleagues (14). None of them was a habitual snorer or reported excessive daytime sleepiness. This investigation was approved by the Ethics Committee of our institution. All subjects (patients and controls) gave their signed consent after being fully informed of the nature and characteristics of the study.

GG Biopsy Technique

In 14 of the 16 patients with OSAS, the GG biopsy was obtained under local anesthesia (1% lidocaine with epinephrine). Extreme care was taken to minimize the amount of anesthesia used and to infiltrate only the skin and subcutaneous tissue. In no patient was anesthesia injected in or near the site of the biopsy. Through a 3-cm-wide horizontal incision, 2 cm below the inferior limit of the chin, the mylohyoid muscle was exposed and separated vertically in the midline. Then, the genioglossal muscle was identified (close to the inner aspect of the mandible) and a piece 4 by 4 mm (weight, 100 to 300 mg) was resected and sent to the laboratory, covered by saline-soaked gauze. After adequate hemostasis, the mylohyoid muscle was sutured vertically, and the skin was closed with subdermal stitches. In all the control subjects, and in the remaining two patients with OSAS (who refused treatment with CPAP but accepted treatment with uvulopalatopharyngoplasty [UPPP]), the GG biopsy was obtained under general anesthesia (propofol, midazolam, and fentanyl) following the same surgical technique mentioned above. In all participants, the GG biopsy was excised without the use of electrocautery.

In vitro Functional Studies

Immediately after biopsy, the muscle strip was mounted vertically in a tissue bath (Radnoti Glass Technology, Monrovia, CA) containing (in mM) 135 NaCl, 5 KCl, 2.5 CaCl₂, 1 MgSO₄, 1 NaH₂PO₄, 15 NaHCO₃, and 11 glucose (15). One end of the strip was tied to an immobile hook and the other to a calibrated high-sensitivity isometric force transducer (Radnoti). The salt solution was continuously gassed with 95% O₂ and 5% CO₂, its temperature was kept constant at 37° C by circulating water through the outer jacket of the tissue chamber (Ex series bath circulator; Neslab Instruments Inc., Newington, NH), and its pH was adjusted to 7.40.

Once in the tissue bath, muscle fiber length was adjusted by the micropositioner until maximal isometric twitch force responses were obtained (optimal fiber length [Lo]). Muscle strips were always stimulated (Model S-48; Grass Instruments Co., Quincy, MA) with supramaximal pulses of 1.0 ms delivered through platinum electrodes (Radnoti). All studies were performed at Lo. Muscles were allowed to equilibrate with the incubation medium for 30 min (18). At the end of the equilibration period, a twitch contraction was elicited, and the following measurements were obtained: the maximal twitch tension, the time to twitch tension (contraction time), the time from peak tension to half-maximal tension (half-relaxation time), and the force-frequency relationship in response to stimulation frequencies of 10, 20, 35, 50, 80, and 100 Hz for 700 ms (18). During determination of the force-frequency relationships, muscles were given a 60-s rest between each tetanic contraction, and maximal tetanic tension was then recorded (18). After characterization of the contractile properties, a 5-min rest period was allowed and the endurance of each muscle was tested. The fatigue protocol consisted in eliciting one 330-ms tetanic contraction at 40 Hz/s for 3 min (18). The loss of force over time was quantified every 30 s for the 3 min of the repetitive stimulation protocol. Muscle length was measured in the chamber, and muscle biopsy was weighed at the end of the experiment.

Calibration of the force transducer was performed by hanging from it known weights, allowing quantification of muscle force in grams. Muscle force production was digitized (Atlantis for Windows; Lakeshore Tech, Inc., Chicago, IL) and stored on a computer. Measures of force (and time) were made manually off-line (Pegasus for Windows; Lakeshore Tech). The absolute maximal forces were normalized to muscle cross-sectional area, as previously described (17). During the fatigue protocol, force was normalized to that produced during the first stimulation train (19); force-frequency relationship were normalized in each muscle strip for the maximal tetanic force (100 Hz) (15, 20).

Fiber Type Distribution

After the fatigue protocol (18), GG samples were frozen in isopentane, cooled with liquid nitrogen, and stored at 70° C until analysis. Then, 6-µm-thick sections of the biopsies were cut in a cryostat (20° C) (Cryocut 1800; Leica Instruments, Bensheim, Germany). Samples were also stained with hematoxylin-eosin and oil-red-o (21, 22), respectively, to investigate potential tissue damage and/or the presence of muscle fat infiltration, central nuclei, nuclei number, and inflammation. Type I ("slow twitch") and type II ("fast twitch") fibers were identified by the standard ATPase (at 9.4 and 4.6 pH preincubation) and NADH-TR stains following standard methodology (21), as previously reported in our laboratory (23). Fiber type percentage were calculated using a microscope (Carl Zeiss, Essen, Germany) by two independent observers. Because the reproducibility of these measurements was good (intraclass correlation coefficient of 0.92) (24), results of the two observers were averaged.

Statistical Analysis

Data are presented as mean ± SEM. The statistical significance of differences was assessed by analysis of variance (two-way ANOVA), followed by post-hoc contrast (Scheffe test) when appropriate. The Kruskall-Wallis test was used to compare the percentage of fiber types between groups. A p value lower than 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Clinical Data

None of the participants experienced any significant complication from the GG biopsy other than mild local pain and a small hematoma at the site of incision. The main clinical characteristics of the subjects included are shown in Table 1. Age was similar in the three groups of subjects. As expected, body mass index (BMI) was higher in the patients with OSAS than in the control subjects. All participants had a normal spirometry. Both groups of patients with OSAS had severe disease as indicated by the apnea-hypopnea index (Table 1). Patients used CPAP for an average of 7 ± 0.5 h/night (range, 5 to 8 h/ night).

In vitro Contractile Properties of the GG

Representative twitch contractions obtained in one control subject, one patient with OSAS (at diagnosis), and one patient with OSAS (after CPAP), are shown in Figure 1. As shown in Table 2, the average values of maximal twitch tension, contraction time, and half-relaxation time were not significantly different between groups. The mean force-frequency relationships in each group are shown in Figure 2. These relationships were not significantly different. By contrast, as shown in Figure 3, GG strips from patients with OSAS depicted a significantly higher fatigability than did those obtained from control subjects (two-way ANOVA, p < 0.001). This was also the case (data not shown) in those two patients with OSAS in whom the GG was obtained under general anesthesia (as it was the case for control subjects). Interestingly, after 1 yr under CPAP, the endurance of the GG was normalized (Figure 3). Accordingly, it was no longer significantly different from that in the control subjects, but it was now significantly different from patients with untreated OSAS (Figure 3).

View larger version (14K): [in this window] [in a new window]   Figure 1.   Representative twitch contraction tracing obtained (from left to right) in one control subject, one patient with OSAS (before any treatment had been started), and one patient with OSAS after being treated with CPAP.

View larger version (15K): [in this window] [in a new window]   Figure 2.   Force-frequency relationship obtained in control subjects, patients with OSAS at diagnosis, and patients with OSAS treated with CPAP. Values represent mean ± SEM in each group. Force is normalized to that at 100 Hz. No significant differences between groups were observed.

View larger version (15K): [in this window] [in a new window]   Figure 3.   Time course of decrement in force production of the GG (mean ± SEM) in response to repetitive stimulation at 40 Hz in all subjects studied. Force was normalized to that produced during the first stimulus train; *p < 0.05 between untreated patients with OSAS and control subjects; +p < 0.05 between untreated and treated patients with OSAS.

Fiber Type Distribution

At diagnosis, patients with OSAS showed a significantly higher percentage of type II fibers than did control subjects (p < 0.001) (Table 2). This difference disappeared in patients with OSAS who had been under CPAP treatment for at least 1 yr (Table 2). Representative tissue sections of the GG in one control subject and in two patients with OSAS (at diagnosis and after CPAP) are shown in Figure 4. There is no evidence of necrosis, increased connective tissue, and/or pathologic signs of myopathy (increased number of nuclei or central nuclei) or inflammatory infiltrates in these biopsies (Figure 5). Also of interest was the absence of fat.

View larger version (95K): [in this window] [in a new window]   Figure 4.   Representative tissue sections of the GG muscle (from left to right) in one control subject, one patient with OSAS (before any treatment had been started), and one patient with OSAS after being treated with CPAP (NADH-TR staining). Original magnification: ×100.

View larger version (115K): [in this window] [in a new window]   Figure 5.   Representative tissue sections of the GG in one patient with OSAS at the time of diagnosis. On the left panel, hematoxylin-eosin staining demonstrating normal skeletal muscle architecture. On the right panel, oil-red-o staining demonstrating no fat infiltration. Original magnification: ×400.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study shows that the in vitro fatigability of the GG is enhanced in patients with OSAS, that this is due to a higher proportion of type II fibers, and that both the structure and the function of the GG is normalized after treatment with CPAP.

Previous Studies

No previous study has investigated the in vitro mechanical properties of the GG in patients with OSAS. Sériès and colleagues (18) reported the isometric contractile properties of musculus uvulae in snorers and in patients with mild OSAS (mean apnea hypopnea index, 34/h). Although differences did not reach statistical significance, patients with OSAS showed a trend towards increased fatigability (18). Probably, the use of snorers as control subjects may have limited the possibility of finding a significant difference (8). More importantly, the physiologic importance of the musculus uvulae in keeping the upper airway open is lower than that of the GG (5, 25, 26). Some previous studies have investigated the fiber type distribution of several upper airway muscles, including the GG (18, 27, 28). All of them showed, as in our study, that the proportion of fast-twitch type II fibers was increased in patients with OSAS. Interestingly, this same observation has been reported in the sternohyoid muscle of the English bulldog, a type of dog that has an anatomically compromised upper airway and sleep-disordered breathing (29). Finally, the effects of CPAP upon the structure and contractile properties of the GG in OSAS have not been investigated before.

Interpretation of Findings

Petrof and colleagues (8) proposed the hypothesis that activity-induced upper airway muscle remodeling and injury may well play a role in the pathogenesis of OSAS. Our study provides experimental evidence to support this hypothesis. First, we observed that the GG in patients with OSAS was more susceptible to in vitro fatigue than that in control subjects (Figure 3). Therefore, it is likely that in vivo the GG will be less effective in maintaining pharyngeal patency with time in OSAS. Interestingly, the episodes of upper airway obstruction that characterizes OSAS increase in frequency and length through the night (30, 31). Second, this enhanced GG fatigability (Figure 3) had a structural counterpart in the higher percentage of type II fibers ("fast twitch") (Figure 4), which are known to be less resistant to fatigue than type I fibers (32). This fiber change, however, may be advantageous in these patients because it may increase muscle efficiency and force-generating capacity during the sudden bursts of GG recruitment necessary to open the collapsed airway (8). Third, and finally, we found that these functional and structural abnormalities of the GG were entirely normalized by CPAP (Figures 3 and 4, and Table 2). CPAP is known to prevent airway collapse (10) and to rest the GG (9, 11). By doing so, a change between type I and type II fiber, as the one we observed, can be expected (8). This reversibility, and the absence of any sign of muscle damage in GG samples (Figure 5) excludes a primary myopathy of the GG as a potential initiating event in OSAS, suggests that the observed GG changes are likely to be the consequence of OSAS (rather than its cause), and supports experimentally the hypothesis of activity-induced muscle remodeling (8).

Methodologic Limitations

Three potential methodologic limitations of our study deserve comment. First, in the control subjects we did not exclude OSAS by polysomnography (14). However, it is unlikely that control subjects may have had clinically significant OSAS because none of them snored habitually or complained of daytime hypersomnolence. Further, we think that the observation of complete reversal of the structural and functional changes of the GG by CPAP validates our conclusions, independently of the control group. Second, we used local anesthesia to obtain GG biopsies in most patients but general anesthesia in control subjects. Ethical reasons precluded a different approach. Yet, we are confident that this has not influenced our results significantly because: (1) the two patients with OSAS in whom the GG biopsy was obtained under general anesthesia (UPPP) also showed enhanced fatigability (compared with the control subjects); (2) a potential depressant effect of anesthesia upon muscle function should be evident, if anything, in those subjects under general anesthesia (control subjects) (33, 34). We found exactly the opposite phenomenon, that is, that control subjects showed more GG endurance than did patients; (3) extreme care was taken to ensure that local anesthesia was limited only to the skin and subcutaneous tissue, and in no case was the GG itself infiltrated directly. Experiments performed in our laboratory in skeletal muscle (anterior and posterior tibialis) in rats showed that this approach did not have any effect upon muscle function studied in vitro (data not shown); and (4) the functional and structural GG alterations were fully reverted by CPAP. This minimizes the potential influence of different types of anesthesia because GG biopsy in patients with OSAS (both at diagnosis and after CPAP) was obtained using local anesthesia in both cases and, nonetheless, the functional and structural results were clearly different. Third, and finally, the absolute twitch tension values recorded in this study (Table 2) are lower than those previously determined in experimental animals (15). This is probably due to the fact that we used cut fibers. It is known that muscle fibers whose ends are cut depolarize, resulting in impairment of contraction. This approach, though, is probably unavoidable in a study like the one herein reported. In any case, we think that this limitation does not affect comparison between groups (because the same methodology was used in each of them) or the assessment of force-frequency relationship or endurance properties (because they were normalized relative to the maximal and initial force, respectively).

Summary

This study has shown that the GG is abnormal both functionally and structurally in patients with OSAS. Because the enhanced fatigability of the GG and the abnormal fiber type distribution observed in these patients is normalized after treatment with CPAP, we suggest that these abnormalities are the consequence not the cause of the episodes of pharyngeal closure that characterize OSAS.