Unventilated Airway Is Time-dependently Constricted in Paralyzed Dogs

Our study protocol was approved by Animal Care and Use Committee of the University of Hirosaki School of Medicine (Hirosaki, Japan). Twenty-one adult mongrel dogs (8–12 kg) were anesthetized with intravenous pentobarbital (30 mg/kg + 2.0 mg · kg⁻¹· h⁻¹) and were assigned to three groups: an apnea–100% O²group (n = 7), an apnea plus vagotomy group (n = 7), and a one-lung ventilation group (n = 7). The lungs of these dogs were ventilated with 100% O². The tracheas were intubated with single- or double-lumen endotracheal tubes in the apnea and apnea plus vagotomy or one-lung ventilation groups, respectively. The distal cuff of the double lumen endotracheal tube was inflated in the left lung. Neuromuscular blockade was obtained with pancuronium at 0.2 mg · kg⁻¹· h⁻¹for mechanical ventilation with oxygen using a volume-controlled respirator (Servoventilator 900C; Siemens-Elema AB, Solna, Sweden), and end-tidal carbon dioxide was maintained at 4.0–4.5%. The femoral artery was cannulated for monitoring of arterial blood pressure and blood sampling. The femoral vein was also cannulated for insertion of a double-lumen catheter through which fluid and drugs were administered. In the apnea plus vagotomy group, the vagus nerves were isolated and then cut bilaterally.

In addition, to determine whether inspiratory O²concentration affected the results, seven dogs were studied with the same protocol as the apnea–100% O²group, except that apnea was produced after ventilation with 50% O²plus 50% N²(apnea–50% O²group). In this group, we measured airway tone not only during apnea but also after ventilation was restarted.

In all groups, the bronchial cross-sectional area (BCA) of the third bronchial bifurcation in the right lung was monitored via  a superfine fiberoptic bronchoscope (2.2 mm OD, AF type 22A; Olympus, Tokyo, Japan) to assess bronchial tone as reported previously. 12–15Briefly, the image of the third bifurcation was printed out with a video printer (VY-170; Hitachi, Tokyo, Japan) during the end-expiratory pause and then was taken into a Macintosh computer (Power Macintosh 7100/80 AV; Apple Computer Inc, Cupertino, CA) by the scanner (Scan-Jet 4c; Hewlett Packard Co., Singapore) to measure the BCA using image-analyzing software (MacSCOPE 2.56; Mitani Co., Fukui, Japan). This image processing was performed by an investigator who was blinded to the study protocol.

In the apnea–100% O²and apnea plus vagotomy groups, a respirator was turned off at the end-expiratory pause to produce apnea for 5 min. However, in the apnea–50% O²group, the respirator was turned off only for 3 min because we preliminarily found that apnea for more than 3 min causes hypoxemia (< 60 mmHg). In the one-lung ventilation group, the right-sided lumen was blocked for 5 min to cause an absence of ventilation of the right lung, and 15 min later, the left-sided lumen was blocked for 5 min to cause only right-lung ventilation. The BCA was assessed before and 1, 2, and 3 min after the ventilator was turned off (at the end-expiratory pause) and on again in the apnea–50% O²group, and it was assessed before and 1, 2, 3, 4, and 5 min after absence of ventilation began in the other groups. When the ventilator was turned off or one-lung ventilation was started, the lumen of the endotracheal tube was occluded by the clamp. In addition, arterial blood (1 ml) was simultaneously sampled to assess arterial oxygen tension (Pao²), arterial carbon dioxide tension (Paco²), and p  H.

All data are expressed as mean ± SD. The BCA is presented as percent of basal BCA. Data were appropriately analyzed with one-way or two-way repeated-measures analysis of variance followed by Student-Newman-Keuls test using Sigma Stat for Windows (Jandel Scientific Software, Chicago, IL). P < 0.05 was considered significant.

In the apnea plus vagotomy group, vagotomy increased the basal BCA by 10.3 ± 9.7%. After the start of apnea, BCA decreased time-dependently in the apnea–100% O²group, whereas it did not in the apnea plus vagotomy group (fig. 1). In the apnea–50% O²group, BCA was reduced time-dependently, similar to the apnea–100% O²group. In addition, the reduction in BCA rapidly returned to baseline after ventilation was restarted (fig. 2).

Apnea also significantly decreased Pao²and p  H and increased Paco²in the apnea–50% O², apnea–100% O², and apnea plus vagotomy groups (tables 1 and 2). However, hypoxemia (< 60 mmHg) was not seen in all dogs of the apnea–50% O²group. In the apnea–100% O²group, no dogs showed hypoxemia until 4 min after apnea started, but 6 of 14 dogs did 5 min after apnea started. In addition, Pao²rapidly returned to baseline after ventilation was restarted in the apnea–50% O²group (table 2).

In the one-lung ventilation group, the BCA of the unventilated lung decreased time dependently, whereas the area of the ventilated lung did not change (fig. 3). Paco²slightly increased from 44 ± 7 to 46 ± 5 mmHg 5 min after one-lung ventilation was started. In addition, although Pao²significantly decreased from 468 ± 59 to 131 ± 158 mmHg, hypoxemia was seen in only 1 of 12 dogs (Pao²= 59 mmHg;table 3).

In the current study, during apnea, BCA decreased in a time-dependent manner with a decrease in Pao²and an increase in Paco². Several factors should be considered to understand the mechanism of apnea-induced bronchoconstriction.

Apnea causes hypoxia and hypercapnia. Previous in vitro  studies 4–6show that these attenuate contraction of airway smooth muscles. In contrast, in vivo  studies 1–3,7suggest that these may induce airway constriction. Iscoe and Fisher 7showed that pulmonary resistance in response to hypoxia (fraction of inspired oxygen [Fio²]= 0.1) and to hypercapnia (inhalation of 5% CO²) increased by 49% and 59%, respectively, in decerebrate cats. In the current study, although hypercapnia and respiratory acidosis occurred in all dogs, bronchoconstriction was observed even without hypoxia. Therefore, hypoxia may not be a main contributor to apnea-induced airway constriction.

The report of Iscoe and Fisher 7suggests that the airway constriction may occur by means of hypercapnia and that it could be abolished by atropine or vagotomy, similar to the current study showing that vagotomy abolished apnea-induced bronchoconstriction. However, in the one-lung ventilation group, despite the fact that Paco²increased by only 2 mmHg 5 min after one-lung ventilation was started, the BCA of unventilated lung time-dependently decreased. Therefore, it is unlikely that hypercapnia is a main factor for apnea-induced airway constriction.

It is known that absorption atelectasis easily occurs by means of a high inspired O²concentration. 16In addition, as apnea substantially decreases the ventilation–perfusion ratio, 16lung collapse is likely to occur and may reflect reduction in BCA. However, it is also known that a lower inspired O²concentration causes lung collapse more slowly. 16However, in the current study, reduction in BCA in the apnea–50% O²group was not less than that in the apnea–100% O²group. In addition, the reduction was not observed even after ventilation with 100% O²in vagotomized dogs. Therefore, the passive collapse is not involved in bronchoconstriction.

An in vitro  study reported by Fredberg et al.  17clearly showed that tidal stretch of airway smooth muscles attenuates active force or muscle stiffness. Such stretching of airway smooth muscle during tidal breathing may contribute to attenuation of bronchoconstriction. In fact, the airways in normal humans and dogs do not constrict to closure even when maximally stimulated. 18In contrast, maximal stimulation can produce airway closure in isolated lobes and bronchial segments. 18Thus, in the absence of tidal stretch, such as apnea, airway narrowing may easily occur. However, Fredberg et al.  17reported that the attenuation of active force or muscle stiffness by tidal stretch may not be neurally mediated because tetrodotoxin did not affect the results. The current study shows that because vagotomy abolished apnea-induced airway narrowing, tidal stretch may not be involved in the mechanism.

The fact that vagotomy completely abolished apnea-induced bronchoconstriction suggests involvement of the neural reflex. Similarly, Brown et al.  19also reported that breath holding after ventilation with 100% O²produces airway narrowing, which can be antagonized by atropine pretreatment. It is known that activation of airway afferent nerve fibers results in reflexes that serve to protect the airways. 11Because vagotomy or atropine abolished the bronchoconstriction during apnea, the afferent neurons of the vagus nerve should be involved in the mechanism of apnea-induced bronchoconstriction. The afferent neurons have been classified into four subclasses: slowly adapting stretch receptors, rapidly adapting stretch receptors (irritant receptors), pulmonary C fibers, and bronchial C fibers. 11Therefore, some of these vagal afferent nerves may be involved in apnea-induced bronchoconstriction.

The slowly adapting stretch receptors are located in the tracheobronchial tree but are concentrated in the large conducting airway and have an important role in the Hering-Breuer reflex. 11In addition, the activation of these units decreases parasympathetic drive to the airway and leads to airway smooth muscle relaxation during inflation. 11Therefore, it is possible that as a consequence of reduction of the slowly adapting stretch receptors by apnea, an increase in the parasympathetic drive may lead to bronchoconstriction. The current data showing that vagotomy abolished bronchoconstriction support this hypothesis.

The rapidly adapting receptors are located in circumference of the entire tracheobronchial tree but are concentrated in the large airway. 11These receptors respond to mechanical stimulation with a burst of action potential during the dynamic phase and then rapidly adapt during the static phase of stimuli, although the activity of these receptors is not coordinated with inspiration or expiration of normal tidal breathing. 11Relatively large changes in transmural pressure, evoked by either distention or collapse of the airway, are required to activate the rapidly adapting receptors. 11In addition, activation of some of these fibers is known to produce cough, bronchoconstriction, and secretion, and these fibers are found in the large airway. 11The current study shows that volumes of the large airway may not change for 5 min of apnea because BCA did not change in the vagotomized dogs. Therefore, unchanged volume in the large airway may not lead to bronchoconstriction, i.e. , the rapidly adapting stretch receptors may not be involved.

The classification of pulmonary and bronchial C fibers are based on their location in the peripheral airways (J receptors) and the large airways, respectively. 11Both C fibers are mechanically sensitive, and stimulation of these fibers results in defensive pulmonary reflexes, such as cough, apnea, bronchoconstriction, and mucous secretion. 11Both C fibers are stimulated by hyperinflation of the lung but not by deflation. 11In addition, a considerable increase in Paco²stimulates bronchial C fibers. 11However, bronchoconstriction was observed without a considerable increase in Paco²in the one-lung ventilation group. Therefore, both C fibers may be uninvolved in apnea-induced bronchoconstriction.

Pentobarbital per se  has been reported to attenuate the airway reflex. 7However, because vagotomy increased BCA by 10% during pentobarbital anesthesia in the current study, the dose of pentobarbital used could not inhibit the reflex completely. Similarly, using high-resolution computed tomography, Brown et al.  19reported that 0.2 mg/kg atropine caused an increase in baseline airway area to 115 ± 5% during thiopental anesthesia in dogs. The bronchodilation could be due to blockade of parasympathetic drive to the airway by vagotomy or atropine.

Pancuronium was used to prevent spontaneous breathing during apnea in the current study. Cembela et al.  20showed that a clinical concentration of pancuronium inhibits muscarinic receptor subtype M2 rather than M1 or M3. Okanlami et al.  21also reported that a small dose of pancuronium produces M2 antagonistic effects and M3 effects predominantly at the larger dose. Because M2 activation inhibits the release of acetylcholine, 22M2 antagonism by pancuronium is likely to potentiate vagally mediated bronchoconstriction. 21,23Therefore, apnea-induced bronchoconstriction may also be potentiated by pancuronium.

In the current study, we used a fiberoptic bronchoscopy method to assess airway caliber. Although airway caliber has classically been assessed with indirect measurement methods, such as airway resistance and compliance, direct methods are more specific than the indirect ones that Brown et al.  24reported. Therefore, we have developed a portable direct method using a fiberoptic bronchoscope. 12To validate our method, we previously compared changes in BCA with percent of dynamic pulmonary compliance (%C^dyn) and percent of airway resistance (%R^aw) in a histamine bronchoconstriction model. We observed that the %BCA completely returned to the prehistamine value by means of intravenous epinephrine, yet %C^dynand %R^awdid not. 12In addition, there was a significant correlation between %BCA and %C^dynor %R^aw. 12Therefore, the bronchoscopic method can assess airway caliber more specifically than indirect ones can. 12 

In conclusion, the current study shows that the unventilated airway constricts spontaneously. This constriction could be vagally mediated (maybe slowly adapting stretch receptors) but is not due to hypoxia and hypercapnia. Further studies are required to elucidate the mechanism.