INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

During wakefulness, individuals with obstructive sleep apnea (OSA) have augmented activity of the pharyngeal dilator muscles (including the genioglossus). This is thought to represent a neuromuscular compensatory mechanism for an anatomically compromised pharyngeal airway (1). Augmented upper airway dilator muscle activity is lost at sleep onset in the apnea patient, and its loss is associated with pharyngeal collapse (2, 3).

The activity of the pharyngeal dilator muscles is influenced by numerous variables, including Pa_O2, Pa_CO2, sleep-wake state, estrogen/progesterone, inspired air temperature, lung volume, and intrapharyngeal pressure (4). The relative role that each of these variables plays in the augmented level of muscle activity seen in patients with OSA has not been established. Evidence for the importance of intrapharyngeal pressure derives from the well-described negative-pressure reflex (NPR). In both animals and humans, the NPR is characterized by a short-latency increase in upper airway dilator muscle (including genioglossus) activity in response to a negative-pressure stimulus (9). However, in the majority of human studies in which the NPR has been assessed, supraphysiological negative pressures (10 to 25 cm H₂O) have been used to assess the reflex. Therefore, the clinical importance of negative pressure in driving pharyngeal muscle activity during basal breathing is unclear, since pressure swings in awake humans are generally small (i.e., < 4 cm H₂O). However, it seems likely that topical receptor mechanisms, responsive to local conditions (e.g., negative pressure, flow, resistance, deformation, or muscle stretch) are present and protect airway patency on a breath-by-breath basis. Such local reflex mechanisms could play a role in the neuromuscular response to individual variability in airway characteristics (9).

In addition to being physiologically appealing, this negative-pressure hypothesis is supported by the observation that positive airway pressure (e.g., continuous positive airway pressure; CPAP) leads to a substantial decline in pharyngeal dilator muscle activation in apnea patients (10). In addition, pharyngeal anesthesia leads to a diminished reflex and a decrement in basal muscle activity, further suggesting an association between airway pressure and pharyngeal dilator muscle activity (11, 12). However, definitive demonstration that physiological levels of negative pressure drive genioglossus activation in awake patients with OSA is currently lacking.

Individuals with OSA who have previously undergone tracheostomy provide an opportunity to measure pharyngeal dilator muscle activation in the presence and absence of upper airway respiratory stimuli (13). If upper airway receptors are important in driving activation of the genioglossus, one would predict that during nasal breathing (NB), the upper airway would be exposed to respiratory stimuli (such as negative pressure), and that a high level of genioglossus activation should be present. On the other hand, with breathing through a tracheal stoma (TSB), the upper airway would not be exposed to these stimuli, and one would predict a lower level of genioglossus activation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We studied five patients (four men and one woman) who had been previously tracheostomized for treatment of OSA. Their mean age was 51.8 ± 3.4 yr (mean ± SEM), and each had an apnea-hypopnea index > 50 events/h prior to tracheostomy. Informed consent was obtained from each patient, with the study protocol having received prior approval from the Human Subjects Committee of the Brigham and Women's Hospital.

Instrumentation and techniques for the study were as follows: Patients wore a nasal mask (Healthdyne Technologies, Marietta, GA) connected to a two-way valve partitioning inspiration and expiration. Inspiratory flow during NB was determined with a pneumotachometer (Fleisch, Lausanne, Switzerland) and a differential pressure transducer (Validyne Corp., Northridge, CA) calibrated with a rotameter. During NB, patients were instructed to breathe exclusively through the nose, and were carefully monitored by video camera to ensure that the mouth was completely closed. Ventilation was not quantified during TSB, but the general absence of or a marked reduction in flow at the nose was documented with the equipment described earlier. Wakefulness was also documented with the video camera, to ensure that the subject's eyes were completely open throughout the study.

The genioglossal electromyogram (GG EMG) was recorded with two stainless steel, Teflon-coated, 30-gauge wire electrodes. Each was inserted 15 to 20 mm into the body of the genioglossus with a 25-gauge needle, which was quickly removed, leaving the wire in place. The two electrodes were inserted on opposite sides of the muscle, with each located 3 mm lateral to the frenulum and close to the insertion of the genioglossus onto the mandible, as described previously (1). Electrodes were referenced to a common ground lead (placed on the forehead) to yield a bipolar recording. The raw EMG was amplified, bandpass filtered (between 30 Hz and 1,000 Hz), rectified, and electronically integrated on a moving-time-average (MTA) basis with a time constant of 100 ms (CWE, Inc., Ardmore, PA). The GG EMG was quantified in arbitrary units and normalized in each patient so that the peak phasic EMG during NB was defined as 100 units.

Pressures were monitored in the nasal mask (Validyne) and in the airway at the level of the choanae and the epiglottis. One nostril was decongested (oxymetazalone HCl) and anesthetized (lidocaine HCl), and two pressure-tipped catheters (MPC-500; Millar, Houston, TX) were inserted through this nostril and localized to determine choanal and epiglottic pressures. Before insertion of the catheters, the signals for nasal mask, choanal and epiglottic pressure were calibrated simultaneously in a rigid cylinder, using a standard water manometer. These three signals, as well as the flow signal, were shown to be without lags in amplitude or phase at up to 2 Hz.

Negative airway pressure stimuli were generated through use of a partially evacuated 50-L canister and a solenoid valve, as described previously (11). Each negative-pressure application (NPA) had a rapid onset and offset and a total duration of < 0.5 s. NPAs were made through the nasal mask during both NB and TSB conditions. Each NPA generated from 8 to 14 cm H₂O pressure at the choanae, with a goal of 10 to 12 cm H₂O. The mean choanal negative pressures were matched for two conditions of NB and TSB. Airway collapsibility was assessed according to an index during NPA, as previously described (11). The measure of collapsibility used was the pressure difference between the choanae and the epiglottis during NPA.

Each patient reported to the laboratory during the day, having been without food intake for at least 4 h. After informed consent was obtained, from the patient, the pressure catheters, EMG wires, and nasal mask were attached and the patient assumed a supine posture. Each patient was studied under the two conditions of NB and TSB. During NB the tracheal stoma was occluded and all airflow was directed through the nose. During TSB the stoma was opened and the vast majority of flow occurred through the stoma. However, because the patients did not have cuffed endotracheal tubes, a small quantity of flow continued to occur through the upper airway.

Each patient was initially monitored under both NB and TSB conditions during tidal breathing for approximately 15 min, after which the responses to approximately 40 NPAs were recorded. The order of the two breathing conditions was assigned arbitrarily.

All signals (GG EMG [raw and MTA], airway pressure [mask, choanal, epiglottic], inspiratory flow, and VT were recorded on a 16-channel Grass Model 78 polygraph (Grass Instruments, Quincy, MA). Certain signals (GG EMG MTA, airway pressures, and inspiratory flow) were also recorded by a computer, using signal-averaging software (SIG-AVG; Cambridge Electronic Design, Ltd., Cambridge, UK). During basal breathing the computer sampled every third or fourth breath, with sampling being continued until at least 50 breaths had been recorded. All breaths were subsequently signal-averaged, yielding a single calibrated waveform of each signal for subsequent data analysis. During tidal breathing, peak phasic and tonic GG EMG data were calculated for each patient under both NB and TSB conditions.

All NPAs occurred during early inspiration with signal averaging beginning with the onset of negative pressure at the choanae as the zero time point during NB. During TSB, a pressure catheter was inserted a short distance into the stoma and the NPAs triggered the early negative pressure of inspiration. The latency of the GG EMG response was measured as the interval from time zero until a rapid rise in the GG EMG signal was detected. The GG EMG response to negative pressure was quantified in three ways: (1) as the increase in the GG EMG signal from time zero to the maximum signal during the negative-pressure stimulus (GG EMG); (2) as the peak GG EMG value obtained; and (3) as the percent increase in the GG EMG signal [GG EMG/time zero EMG]*100).

All statistical analyses were performed with commercially available software (Excel 97; Microsoft, Inc., Seattle, WA). Comparisons of data for NB and TSB were made with two-tailed paired t tests. For all analyses, was set at 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Figure 1 shows the raw data for one patient, demonstrating a striking decrement in GG EMG activity when ventilation was switched from NB to TSB. This decrement in GG EMG activity was consistent among patients, as is shown in Table 1 and Figure 2.

View larger version (92K): [in this window] [in a new window]   Figure 1.   Both the raw and MTA GG EMG from one patient fell during tracheal stomal (trach) breathing. Note that both epiglottic pressure and inspiratory flow were substantially attenuated following the transition from nasal to tracheal breathing.

View larger version (50K): [in this window] [in a new window]   Figure 2.   The peak phasic genioglossus GG EMG fell in all five patients after the transition from NB to TSB (trach). The nasal peak phasic GG EMG values are normalized to 100 arbitrary units in each patient (see text).

Although it was not statistically significant, all five study patients showed an attenuated negative-pressure reflex during TSB as compared with NB (Table 1). Collapsibility (measured by the pressure decrement between the choanae and epiglottis during negative-pressure pulses) increased substantially during TSB as compared to with NB (Table 1).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results indicate that the withdrawal of upper airway respiratory stimuli, such as airflow and negative pressure, causes a large and immediate decrement in pharyngeal muscle dilator activation in awake individuals with OSA. This decrement was observed for both the tonic and phasic GG EMG in our patients. The muscle also tended to be less responsive to negative pressure (albeit not significantly in this small group of patients), whereas the upper airway was substantially more collapsible during TSB than during NB. These results strongly support the importance of upper airway receptors (possibly responsive to negative pressure) in driving the augmented genioglossus activation seen in awake apnea patients.

A substantial body of animal and human data indicate that local airway stimuli (such as negative pressure) can activate pharyngeal dilator muscles (14, 15). However, large (nonphysiologic) negative pressures were generally used in the studies that produced these data. In addition, apnea patients, with substantially augmented muscle activity during wakefulness, have only marginally greater negative pressure in the airway than do normal controls. As a result, the role of negative pressure (or other local stimuli) in driving this compensatory neuromuscular activity has been questioned. However, our study strongly supports the concept that some localized upper airway receptor mechanism drives the pharyngeal dilator muscle activation observed in these patients. A number of studies using upper airway anesthesia also support this concept (11, 12).

Explanations other than negative pressure for the augmented pharyngeal dilator muscle activation seen in apnea patients are certainly possible. First, inspiratory flow through the upper airway changed considerably with the alternate breathing routes used in our study. Younes and colleagues and others have previously suggested that flow receptors exist in the upper airway and that they may affect respiratory pattern (16, 17). However, definitive studies for distinguishing the relative influences of pharyngeal pressure and airflow on upper airway dilator muscle activation have not so far been done.

Second, upper airway resistance could be the primary variable sensed in the mechanism leading to dilator musle activation. This would imply the ability to detect pressure in two regions (e.g., nasal and laryngeal) and to react to increasing differences between the two with increasing dilator muscle activation. Studies that involve inspiratory resistive loading to induce flow limitation, however, have not consistently shown strong genioglossus activation, despite high levels of velopharyngeal resistance. Therefore, although its role is possible, resistance is not likely to be the primary drive to genioglossus activation.

Third, exhaled carbon dioxide could influence dilator muscle activity. Laryngeal CO₂ receptors are clearly present and can increase genioglossus activation. In theory, if exhaled CO₂ is important in driving genioglossus activation, muscle dilator activation would fall during TSB, when the larynx would not be exposed to CO₂. However, apnea patients do not generally have higher exhaled CO₂ levels than do controls, and rebreathing of CO₂ by normal subjects does not lead to the huge increases in GG EMG activity seen in apnea patients. Moreover, the increased GG EMG activity of the apnea patient could be a learned behavior, protecting airway patency with little or no ongoing receptor stimulation required for muscle activation. However, the results of the present study argue strongly against this last explanation.

Our failure to find a statistically significant decrease in the negative-pressure reflex among patients with OSA is not a complete surprise. By design, we applied similar negative pressure at the choanae under both NB and TSB conditions, and one might therefore expect similar muscle activation under both conditions. However, we did not match negative pressure at the epiglottis. Because the pharyngeal airway of the apnea patient is considerably more collapsible during TSB than during NB, the pressure reaching the epiglottis was lower (6.99 cm H₂O nasal versus 3.07 cm H₂O tracheal, p = 0.046) during TSB. Because laryngeal receptors are believed to play an important role in this reflex (9), this could explain the trend toward a reduced response of the genioglossus during TSB.

A number of potential limitations must be considered in interpreting our study. First, because the patients were stable outpatients, they did not have cuffed tracheostomy tubes, and some negative pressure was therefore still measured at the epiglottis during TSB. In all patients, however, there was substantially less negative pressure during TSB than during NB. The presence of this residual epiglottic negative pressure during TSB should, if anything, have reduced the decrease in genioglossus activation that we observed. Therefore, a cuffed tube would probably have made our findings more robust. The absence of a cuffed tube did, however, limit our ability to measure flow and VT during TSB. We do not believe that any of our patients had important changes in respiratory rate or VT with the change in breathing routes used in the study, although we cannot completely exclude this possibility.

Second, one could argue that the upper airway of a tracheostomized apnea patient is not identical to that of a nontracheostomized apnea patient. The lack of repetitive collapse during sleep, and the "sensory deprivation" of the upper airway mucosa, could influence upper airway reflexes (18). We believe, however, that this is unlikely, since our tracheostomized apnea patients breathe through their upper airways during wakefulness and through their tracheostomy only during sleep. Because these patients experience pharyngeal airflow throughout the day during wakefulness, we feel that "sensory deprivation" is an unlikely explanation for the diminished negative-pressure reflex seen in apnea patients. Third, the resistance to breathing during TSB was almost certainly less than occurred during NB, which could influence muscle activation. However, in our hands, modest changes in airflow resistance (2 to 3 cm H₂O/L/s) do not lead to major changes in muscle activation, As a result, we doubt that resistance to breathing played a major role in our observations.

In conclusion, we believe that our results strongly support both the importance of local receptor mechanisms in the upper airway and the influence of these mechanisms on pharyngeal dilator muscle activity. Although other explanations are possible, we favor negative pressure as the critical stimulus in such muscle activity. The postulated local receptor mechanisms are likely to allow precise neuromuscular adaptation to the local environment, and can probably respond to changes in multiple variables, including airway size, airway shape, and tissue characteristics. How this information is processed and integrated by the central nervous system remains poorly understood, as does the effect of sleep upon these mechanisms.