INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

INTRINSIC LARYNGEAL MUSCLES contract to bring about changes in glottic size, which can affect resistance to airflow into and out of the lungs. Phasic laryngeal abduction during inspiration widens the glottic opening and decreases resistance to airflow (4, 5). Phasic expiratory activity of muscles that adduct or narrow the vocal folds results in increased resistance to airflow out of the lungs (5, 12).

Pressure and/or airflow has been shown to modulate the phasic respiratory activity of intrinsic laryngeal muscles. Phasic activity of the posterior cricoarytenoid muscle (PCA), a laryngeal abductor, increases when airflow is diverted from tracheotomy to upper airway (UA) breathing (25) and in response to negative pressure pulses (38). Insalaco et al. (10) showed increased activity of the thyroarytenoid muscle (TA), a laryngeal adductor, in response to increased resistive loading.

In anesthetized neonatal animal models, constant flow and/or pressure in the isolated UA has been shown to result in a ventilatory inhibition characterized by a reduction in breathing frequency and tidal volume (VT) (1, 6, 14, 17, 20). This ventilatory inhibition is abolished after bilateral section of the superior laryngeal nerve (SLN) or topical anesthesia of the laryngeal region (1, 6, 20). However, chronic section of the SLN does not affect the breathing pattern in conscious kittens (20), suggesting that the inhibitory effects of SLN afferents are not a major factor during eupnea in the conscious animal but can be magnified under certain circumstances or that adaptation occurs after ablation of these reflex pathways.

Sant'Ambrogio and co-workers (27) identified and characterized pressure- and flow-sensitive endings in the internal branch of the SLN. "Flow" receptors actually function as thermoreceptors, stimulated by cooling of the UA mucosa from body temperature to room temperature. Pressure-sensitive endings respond to positive or negative transmural pressure applications; they seldom respond to both (27).

To evaluate the contribution of UA receptors sensing pressure and flow to the controlled respiratory phase-specific activation of the PCA, TA, and respiratory pump muscles, we studied the steady-state effects of presumably stimulating specific UA receptors (flow and/or pressure) in anesthetized, tracheotomized, spontaneously breathing piglets with an isolated UA with oscillatory airflows in phase with the animal's tracheal breathing pattern. The magnitudes of pressure and flow applied to the isolated UA were independent of those in the lower airway and were within the physiological range for the respiratory drive used in our studies. By altering the temperature of the air applied to the isolated UA, we presumably stimulated pressure receptors only (with air at body temperature) or both pressure and flow receptors (with air at room temperature). We hypothesized that UA pressure and flow would preferentially modulate the phasic activity of laryngeal muscles compared with respiratory pump muscles in a respiratory phase-specific fashion. We also assessed the relative contributions of UA flow and/or pressure to this modulation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The protocol was approved by the Institutional Animal Care and Use Committee at the Robert Wood Johnson Medical School.

Surgical preparation. Twenty-seven piglets of either gender (15.4 ± 1.9 days of age, range 3-33 days; 3.92 ± 0.28 kg body wt) were premedicated for surgical preparation with acepromazine (1 mg/kg im) and ketamine (45 mg/kg im). The animals were secured, in the supine position with the neck extended, on a surgical table. Rectal temperature was monitored (Yellow Springs Instruments telethermometer probe) and maintained at 37-39°C with a heating pad. A few drops of Xylocaine (2% lidocaine HCl, 20 mg/ml sc) or a topical anesthetic skin refrigerant (ethyl chloride) were used to premedicate areas before incisions. Cannulas were placed in the femoral artery for blood pressure monitoring and blood-gas sampling and in the femoral vein for maintenance of anesthesia by constant infusion of ketamine (0.3 mg · min¹ · kg¹ as required to maintain anesthesia). A longitudinal neck incision was made to expose the larynx and trachea. The trachea was sectioned between the third and fourth ring, and an endotracheal tube was inserted in the caudal segment. A differential pressure transducer (Validyne) and a pneumotachograph (Hans Rudolph) connected to a transducer (Validyne) were attached to monitor tracheal flow and pressure, respectively. The rostral segment of the trachea was intubated below the larynx and attached to a pneumotachograph (Fleisch) connected to a differential pressure transducer (Validyne) for the measurement of UA flow and to a differential pressure transducer (Validyne) for the measurement of pressure.

Isolated UA breathing circuit. The UA was isolated, and a breathing circuit was constructed for the generation of oscillatory pressure and flow with air at room temperature or warmed to 40°C by passage through coiled tubing submerged in a temperature-controlled bath to achieve body temperature when the air passed through the upper airway.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Isolated upper airway (UA) breathing circuit for delivery of oscillatory or constant pressure and flow. A: tracheal inspiration triggered opening of the servo-controlled valve connected below the larynx and exposed the isolated UA to a negative pressure source, pulling bias air available at the snout in the inspiratory direction to produce oscillatory negative pressure and inspiratory flow, during inspiration only. B: tracheal expiration triggered opening of the servo-controlled valve, which exposed the isolated UA to a positive pressure source and delivered air in the expiratory direction, during expiration only.

Static respiratory mechanics measurements. Pressure-volume relationships were constructed for the total respiratory system and lung only, postmortem, before and after thoracotomy in 10 piglets. Animals were euthanized, and the tracheotomy tube was fitted with a three-way connector attached to a 60-ml syringe and a pressure transducer. To determine the pressure-volume relationship for the total respiratory system, lung inflations (5, 10, 15, 20, 25, 30, 35, and 40 ml) and deflations (5, 10, 15, and 20 ml), relative to resting volume, were applied at a constant rate. Once each target inflation or deflation volume was reached, it was maintained to obtain a plateau on the pressure recording, and then the volume was returned to resting levels. Subsequently, once the pressure returned to resting levels, the procedure was repeated until all pressure measurements were obtained for all target volumes. A thoracotomy was performed via a longitudinal incision alongside the sternum to expose the lungs, the chest wall was widely retracted, and the pressure-volume relationship for the lung only was determined by repeating the protocol.

Data collection and analysis. Data were recorded on chart paper (model TA 4000, Gould) and digital tape (model 4000A, Vetter). Pressure, flow, and EMG MTA data were digitized at a sampling rate of 25 Hz for computer-assisted analysis on-line on a desktop microcomputer fitted with data-acquisition hardware and storage (CODAS, DATAQ). EMG activities were quantitated as the peak height of the phasic MTA activity above tonic levels during inspiration (PCA and Dia) or expiration (TA and Abd). Ventilatory parameters [inspiratory (TI), expiratory (TE), and total respiratory time (TT) and VT] were measured or calculated breath by breath using data-acquisition software (Advanced CODAS, DATAQ). TI was defined as the time from the onset to maximum phasic activity of the MTA Dia EMG above tonic levels for each breath. TE was defined as the time from the maximum level of phasic MTA Dia EMG above tonic values to the onset of the subsequent breath. TT was defined as the sum of TI and TE. VT was calculated from digital integration of tracheal flow signals. Additional calculations were performed with spreadsheet software (Quattro Pro for Windows, Borland).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of UA negative pressure and inspiratory flow applied during inspiration only. Experiments to assess the effect of UA negative pressure and inspiratory flow, applied during inspiration only, on respiratory activity were conducted under exposure to hypercapnia or, when phasic PCA activity was present, under hyperoxia. Figure 2 illustrates the typical pattern of responses observed, which consists of UA airflows at body and room temperature, resulting in enhancement of phasic PCA EMG while Dia and Abd remained unaffected compared with baseline values (absence of pressure and flow in the isolated UA).

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Representative example of the respiratory response to negative pressure and inspiratory flow applied to the isolated UA during inspiration only. PCA, Dia, and Abd, moving-time-averaged posterior cricoarytenoid, diaphragm, and abdominal electromyograms, respectively. This record was obtained under exposure of the lower airways to hypercapnia, which explains the larger tracheal pressure swings than those typically observed for tracheotomized, spontaneously breathing piglets.

Effect of UA positive pressure and expiratory flow applied during expiration only. Experiments to assess the effect of UA positive pressure and expiratory flow, applied during expiration only, on respiratory activity were conducted under exposure to hyperoxia when phasic TA activity was present. Animals were exposed to hypoxia if phasic TA activity was absent under hyperoxic inspired gas mixtures. Phasic PCA and Abd activity was rarely present. Figure 3 illustrates the typical pattern of responses, which consists of UA airflows at body and room temperature, resulting in enhancement of phasic TA EMG compared with baseline values. Figure 3 also shows a reduction of respiratory frequency with room or body temperature oscillatory pressure and flow.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Representative example of the respiratory response to positive pressure and expiratory flow applied to the isolated UA during expiration only. TA, moving-time-averaged thyroarytenoid electromyogram.

Static respiratory compliance measurements. Maturational changes in CT, CL, Cw, and Cw/CL were assessed in 10 piglets (16.5 ± 3.1 days of age, range 5-33 days; 4.17 ± 0.49 kg body wt; Fig. 4). CT and Cw diminish with age, with the majority of the changes occurring within the 1st wk and stabilizing by the 2nd and 3rd wk of life. The correlation between Cw/CL and age (Pearson coefficient = 0.605, r² = 0.366, P = 0.064), although not statistically significant, suggests an inversely proportional trend between the two variables. Cw/CL remains above unity during the 1st mo of life of the piglet.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Maturational changes in total respiratory system (CT), lung (CL), and chest wall (Cw) compliance (A) and in ratio Cw/CL (B) in 10 piglets. CT and Cw diminish with age, with the majority of the changes occurring within the 1st wk of life and stabilizing by the 2nd and 3rd wk. Correlation between Cw/CL and age (Pearson coefficient = 0.605, r2 = 0.366, P = 0.064), although not statistically significant, suggests an inversely proportional trend between the 2 variables. Cw /CL remains above unity during the 1st mo of life of the piglet.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

UA flow-sensitive endings can be inactivated by perfusion of the isolated UA with body temperature air, whereas room temperature air can be utilized to stimulate pressure and flow receptors (29). Cooling of the laryngeal mucosa, while stimulating flow-sensitive endings, diminishes but does not inactivate pressure receptor discharge compared with air at body temperature (26). When the response of the PCA to oscillatory negative pressure and flow is compared, the degree of enhancement was not statistically significantly different between air temperature treatments. The response of the PCA suggests that, despite pressure receptor inhibition due to cooling, feedback from flow receptors is sufficient to result in the same degree of enhancement under both temperature conditions.

Studies of the TA response to receptor stimulation showed that activation of UA positive pressure receptors only or both pressure and flow receptors favors adduction. The pressure receptor population stimulated in these experiments differs from that stimulated with negative pressure, inasmuch as laryngeal pressure receptors have been described as being sensitive to positive or negative transmural pressure applications, but seldom to both (27). Because there were no statistically significant differences between afferent stimulation due to oscillatory room temperature and body temperature air in the isolated UA, feedback from flow receptors is sufficient to counteract the inhibitory effect of cooling on pressure receptors and result in the same degree of enhancement under both temperature conditions.

Although our results showed statistically significant changes in Dia EMG activity and respiratory timing with UA flow and/or pressure, these small changes are unlikely to be physiologically significant. However, our results suggest that laryngeal endings do modulate the phasic respiratory activity of laryngeal muscles. Does removal of this sensory feedback modify eupneic breathing pattern? Stockwell et al. (33) compared ventilatory parameters in humans before and after lidocaine-induced anesthesia of the SLN. They did not observe any differences in parameters attributable to SLN conduction blockade and concluded that SLN afferents do not play a significant role in resting breathing pattern in humans. Mortola and Rezzonico (20) addressed the same question by studying anesthetized and unanesthetized kittens. Although acute section of the SLN abolished inhibitory ventilatory effects of steady airflows applied to isolated UA, ventilatory inhibition is still present with chronic SLN section. The authors allude to other non-SLN UA receptors developing compensatory mechanisms. One must consider the possible role of sensory innervation of the nasopharynx supplied by the glossopharyngeal nerves investigated by Widdicombe et al. (37) and Mathew and Sant'Ambrogio (15). Another strong candidate is innervation of the nasal mucosa supplied by the ophthalmic and maxillary divisions of the trigeminal nerve (29). Tsubone (34) found cold- and pressure-sensitive endings in the nasal cavity innervated by the ethmoidal nerve, a branch of the ophthalmic division of the trigeminal nerve. Unlike studies from the laboratories of Sant'Ambrogio and Mortola, which largely bypassed nonlaryngeal UA receptors by placing a cannula immediately above the larynx, the isolated UA route of breathing in our experimental design spanned nasal-to-subglottic areas, thus stimulating nasal, pharyngeal, and laryngeal pressure and flow receptors.

Mathew et al. (13), as well as Al-Shway and Mortola (1), were able to eliminate reflex responses to change in UA pressure and/or flow by topical anesthesia of the laryngeal region, suggesting a superficial location for these receptors. Although the majority of the UA pressure- and flow-sensitive endings alluded to in this study do travel via the internal branch of the SLN, because additional receptors are located in the nasal and pharyngeal UA, bilateral sectioning of the SLN in our preparation would not necessarily eliminate the reflexes described here.

A recent review by Sant'Ambrogio et al. (29) comments that, under eupneic breathing conditions, laryngeal afferents have not been shown to substantially modify breathing patterns. However, the authors suggest a possible role in maintenance of respiratory homeostasis during hyperventilation or "special conditions" that challenge the respiratory system. In our experimental design, animals were challenged with hypercapnia or hypoxia. Although these conditions were present throughout control and UA receptor-stimulated trials, they may have been sufficient to provide a different set point from eupnea for the respiratory system to bring about the modulating effects of UA pressure and flow receptor stimulation. This may play a role in the fact that UA positive pressure and expiratory flow were shown to affect TA activity more than negative pressure and inspiratory flow affected PCA activity (10-15 vs. 45-50%). Furthermore, a stronger reflex activation of TA than of PCA activity from UA receptor stimulation can be speculated when the maturational age of the animal model used in our studies is considered.

Increased phasic TA activity or adduction of the vocal folds during expiration lengthens the expiratory time constant for airflow, resulting in "braking" of expiratory airflow. Expiratory airflow braking is a mechanism by which end-expiratory lung volumes can be increased above passive levels (24). Breathing from elevated end-expiratory lung volume is a characteristic of neonates (11, 19, 21), who must do so to compensate for highly compliant chest walls (8, 16, 23). Our static respiratory mechanics measurements in piglets showed that, although most of the changes in Cw occur by the 1st wk of life (Fig. 4A), Cw/CL is still above unity by the 3rd wk of life (Fig. 4B). Thus critical control of laryngeal adduction can be instrumental in these animals to compensate for their mechanical disadvantage. Although our results did not reveal significant age differences in the respiratory response to UA pressure and flow for piglets within the 1st mo of life, we speculate that restricting the study to piglets within the 1st wk of life might reveal such differences. Our static respiratory mechanics measurements are in agreement with those reported by Fisher and Mortola (7) for piglets at 3 ± 2 days of age. We are unaware of any other reports of static compliance measurements in piglets within the 1st mo of life.

The study of single-unit SLN afferent fiber recordings by Sant'Ambrogio et al. (27) showed that UA pressure and flow receptors can be stimulated for short or long periods of time, suggesting that the reflex response to laryngeal receptor stimulation could be affected by stimulus modality. To ensure activation of UA receptors similar to that stimulated during eupneic breathing, we applied pressure and/or flow waveforms to the isolated UA in phase with those generated by the animal. Although transmural pressure or flow applications to the UA have been previously shown to affect laryngeal muscle activity, the studies presented are unique, in that the modality of pressure and/or flow application is a more physiological-like stimulus than constant, square-wave, or intermittent pressure applications used by other investigators. Furthermore, the UA pressure levels utilized in the studies presented here are closer to physiological values than the levels used by other investigators. For example, Fisher et al. (6) used positive and negative pressures from 2 to 20 cmH₂O in puppies; in rabbit studies, Woodall and Mathew (39) used up to 15 cmH₂O positive and negative pressure pulses of 200-ms duration. Although high pressures can result in considerable laryngeal distortion and thus activate other endings such as joint receptors, low pressures can lead to the lack of reflex responses (15). However, the range of receptor discharge responses to pressure suggests that most of the response would be activated in the low pressure range. The chosen ranges of UA pressures used in our studies allowed for minimal effects on respiratory timing (10% reduction in TT or TE) while still enhancing laryngeal muscle activity. This can account for differences between studies in which a more drastic reduction in breathing frequency was observed in response to constant negative pressures, such as the rabbit study by Mathew et al. (13). Finally, the focus of our study was on a steady-state response, present after the first 10 breaths following stimulus application, which differs from previous studies that measured the immediate effects on stimulus application (1, 6).

Our choice of animal model may influence our results, inasmuch as species differences have been found in the relative distribution of pressure and flow receptors in the UA, in their location within the UA, and in the response of these receptors to respiratory stimuli. For example, although flow receptors have been identified in the SLN of dogs (28) and rabbits (18), they were not found in guinea pigs (35). However, cooling of the guinea pig larynx with l-menthol, a selective stimulant of laryngeal cold receptors, results in apnea (22), suggesting the possibility that these endings exist but are difficult to isolate. In the nose, cooling receptors have been identified in cats (36), rats (34), and guinea pigs (32).

Mechanoreceptors with a respiratory modulation have been found in the SLN of dogs (27), rats (31), guinea pigs (35), and rabbits (18). However, unlike laryngeal mechanoreceptors in dogs, the majority of pressure receptors in these species are stimulated by positive pressure and inhibited by collapsing pressure, possibly because of their location on the epiglottis and its movement during respiration (29).

Although the prevalence of laryngeal negative pressure receptors may differ between species, this does not necessarily imply a species difference in the reflex response to UA pressure and flow. UA pressure and flow receptors, although in different numbers and locations, exist in all of these species. For example, the low percentage of laryngeal negative pressure receptors in some species may be compensated by receptors found elsewhere in the UA, inasmuch as nasal pressure receptors have been identified in cats, rats, and guinea pigs (29, 30, 36). We are unaware of any studies that identified pressure and flow receptors in the UA of piglets. However, our study of the respiratory modulation of laryngeal muscle activity with UA pressure and flow suggests the potential existence of these receptors in piglets.

Because UA pressure and flow during inspiration did not affect Dia or Abd activity while favoring enhanced laryngeal abduction, we conclude that the resultant decrease in inspiratory UA resistance should lessen the load imposed on respiratory pump muscles and promote UA patency. Similarly, inasmuch as laryngeal adduction is favored during expiration, increased expiratory UA resistance assists in braking of expiratory airflow, lessening the load imposed on chest wall muscles to achieve the same level of braking. We conclude that UA afferent information plays an important role in modulating phasic respiratory activity of intrinsic laryngeal muscles on a breath-by-breath basis.