Larynx vs. esophagus as reflexogenic sites for acid-induced bronchoconstriction in dogs

Teruhiko Ishikawa, Shin-Ichi Sekizawa, Franca B. Sant’Ambrogio, Giuseppe Sant’Ambrogio


Bronchoconstriction in asthmatic patients is frequently associated with gastroesophageal reflux. However, it is still unclear whether bronchoconstriction originates from the esophagus or from aspiration of the refluxate into the larynx and larger airway. We compared the effect of repeated esophageal and laryngeal instillations of HCl-pepsin (pH 1.0) on tracheal smooth muscle activity in eight anesthetized and artificially ventilated dogs. Saline was used as control. We used pressure in the cuff of an endotracheal tube (Pcuff) as a direct index of smooth muscle activity at the level of the larger airways controlled by vagal efferents. The Pcuff values of the first 60 s after instillations were averaged, and the difference from the baseline values was evaluated. Changes in Pcuff were significantly greater with laryngeal than with esophageal instillations (P = 0.0166). HCl-pepsin instillation into the larynx evoked greater responses than did saline (P = 0.00543), whereas no differences were detected with esophageal instillations. Repeated laryngeal exposure enhanced the responsiveness significantly (P < 0.001). Our data indicate that the larynx is more important than the esophagus as a reflexogenic site for the elicitation of reflex bronchoconstriction in response to acidic solutions.

  • gastroesophageal reflux
  • acid-induced laryngitis
  • laryngeal reflexes
  • esophageal reflexes

gastroesophageal reflux (GER) is often associated with bronchial asthma (2,19); however, the nature of the association is not fully understood. Although bronchoconstriction has been viewed as a result of gastric refluxate through a vagally mediated reflex (8, 18, 32), it is still controversial whether this reflex originates from the esophagus or the airway. Mendelson (20) was the first to describe “asthmatic-like” reactions induced by aspiration of gastric contents into the airway during induction of general anesthesia (Mendelson syndrome), and he considered aspiration into the airway to be the most important factor for the elicitation of bronchoconstriction. In fact, even if the refluxate does not reach the lower airways, small amounts of aspiration can cause several kinds of reflexes because of the high responsiveness of the larynx (4, 6, 22,25, 31). This contention is also supported by the observation that acid-induced laryngitis is often seen in GER patients (5, 10, 16, 27).

Conflicting results have been reported on the reflexogenic role of the esophagus in acid-induced bronchoconstriction. Several investigators (17, 18, 23, 26, 29) concluded that esophageal acidification does cause reflex bronchoconstriction, whereas others did not find any substantial evidence for the reactivity to acid solution. In fact, Ekstrom and Tibbling (7) and Wesseling et al. (30) did not find any association of bronchoconstriction with esophageal acidification. The difficulties with these reports are that most of the studies assessed bronchoconstriction by conventional, effort-dependent pulmonary function tests (e.g., peak expiratory flow; forced expiratory volume at 1 s). More direct evidence of reflex bronchoconstriction is desirable for disclosing reflexogenic sites.

Recently, we reported (14) that pressure in the cuff of an endotracheal tube (Pcuff) provides a more sensitive index of airway smooth muscle activity than does total lung resistance, especially when the response is mediated by the vagal efferents. In fact, this index reflects smooth muscle contraction or dilatation of the portion of tracheobronchial tree directly innervated by the vagus (14). We believe that this measurement could more clearly reveal whether the airway or the esophagus is more important for acid-induced reflex bronchoconstriction. The larynx was chosen from the rest of the airways, because it is first exposed to refluxate after the regurgitation has reached the pharynx.


General preparation.

The study was conducted, after approval of the Institutional Animal Care and Use Committee, on eight healthy mongrel dogs of either sex (weight 12.3–18.2 kg). The animals were sedated with ketamine (10 mg/kg im) and were anesthetized with a mixture of urethan and α-chloralose (500 and 50 mg/kg iv, respectively). The dog was placed in a supine position on an operating table. The femoral artery and vein were catheterized to monitor blood pressure and to administer drugs. Tracheostomy was established, and an endotracheal tube with a high-volume, low-pressure cuff (7.5-mm ID; Rüchlit, Waiblingen, Germany) was inserted in the extrathoracic portion of the trachea. The cuff was filled with enough distilled water (2–5 ml) to provide a complete seal with the tracheal wall. A pressure transducer (P10EZ; Gould, Valley View, OH) was connected to the liquid-filled cuff to measure Pcuff; the transducer was placed at the same height as the tracheal cuff. The animal was then artificially ventilated with a constant-volume-type ventilator (model 607; Harvard Apparatus, Millis, MA) and paralyzed with doxacurium chloride (2.5 mg iv). This muscle relaxant has virtually no autonomic effects at the dose used (3, 13). Tidal volume (12–16 ml/kg) and breathing frequency (10–18 breaths/min) were adjusted to maintain end-tidal CO2 at a concentration of 5.8–6.2% as measured by a precalibrated infrared gas analyzer (Medical Gas Analyzer LB-2; Beckman, Schiller Park, IL).

Supplemental doses of the anesthetics and neuromuscular blockade were administered regularly (urethan, 150 mg ⋅ kg−1 ⋅ h−1; α-chloralose, 15 mg ⋅ kg−1 ⋅ h−1; doxacurium, 0.5 mg/h). In addition to Pcuff and blood pressure, tracheal airflow and pressure were also obtained by using a no. 1 Fleisch-type pneumotachograph connected to a differential transducer (model 15196; Gould, Hato Rey, Puerto Rico) and a differential transducer (model 16720, Gould) with one port open to atmosphere. All signals were digitized at 100 Hz per channel with an analog-to-digital converter (AT-CODAS Data-Acquisition Card; Dataq Instruments, Akron, OH) and stored in an IBM-compatible computer.

Experimental protocol.

The larynx and the esophagus were challenged with a solution of hydrochloric acid containing 0.2 g/l pepsin (HCl-P; pH 1.0); saline (0.9% NaCl unbuffered, pH 6.4) was used as control. The order of the challenges was randomized among dogs, between larynx and esophagus, but saline was always given first. A volume of 4 ml was instilled into the larynx with a polyethylene catheter inserted through the tracheal stoma and advanced cranially until the tip reached the larynx. This catheter has multiple side holes over the distal 3 cm to allow the instilled solution to reach a wide area. To minimize the possible mechanical stimulation by the catheter, 10–15 s were allowed between insertion and instillation; in any case, we made sure that Pcuff was in the baseline condition before starting the instillation. For the esophageal instillations (8 ml), the same type of catheter was placed into either the proximal or the distal portion of the esophagus (4 dogs each) and left in situ. The determination of proximal and distal portion of the esophagus was established as follows: a catheter was inserted through the mouth into the esophagus and gently pushed forward until the tip reached the lower esophageal sphincter (the catheter could not be pushed any further). Then the catheter was withdrawn 5 cm for the distal esophageal instillations; proper placement was confirmed postmortem. For the proximal esophageal instillations, the catheter tip was placed midway in the extrathoracic portion of the esophagus; this position was confirmed by direct palpation of the catheter tip across the esophageal wall. All instillations were performed at end-expiratory volume. At 5 min after instillation, the solution was removed by suctioning, and an interval of 10 min was allowed before the next instillation. One experimental session consisted of one saline and one HCl-P instillation into the laryngeal and the esophageal lumina. A total of six sessions was completed in each dog.

Data and statistical analysis.

The response was quantified by calculating the difference between the average of the Pcuff value in the first 60 s after an instillation and the corresponding baseline value (ΔPcuff). The Pcuff values corresponding to the three consecutive breaths immediately preceding the instillation were averaged and was used as the baseline value.

Repeated-measures multivariate analysis of variance was performed for the comparison between instillations into the esophagus and larynx. For the planned comparison between specific sets of means of interest, such as acid instillation into larynx vs. esophagus, appropriate contrasts were constructed to reveal the difference of the effect (contrast analysis). P < 0.05 was considered significant.


Figure 1 shows experimental records from one dog; HCl-P instillation into the larynx clearly increased Pcuff, whereas instillation into the esophagus did not. Biphasic responses in Pcuff were often seen with HCl-P instillations into the larynx. The second phases of the biphasic responses with laryngeal HCl-P instillations were sustained for ∼1–2 min. Figure2 represents the mean values of ΔPcuff obtained from all the sessions and shows the sustained effects of HCl-P instillation into the larynx compared with the other types of instillation. This sustained effect of HCl-P instillations on Pcuff is more evident after the third experimental sessions. Figure3 represents the overall results with statistical comparisons. The responses with the proximal and distal esophageal instillation were nested, because no visible difference was observed.

Fig. 1.

Experimental records of the 5th esophageal and laryngeal instillation session in 1 dog. HCl-P, HCl-pepsin; ▾, instillations; BP, arterial blood pressure; V˙, tracheal flow; Ptr, tracheal pressure; Pcuff, pressure in the cuff of the endotracheal tube. Virtually no response was observed in Pcuff after esophageal instillations. Pronounced responses were seen in Pcuff and arterial blood pressure both with saline and HCl-P instillation. Note biphasic response in Pcuff with HCl-P laryngeal instillation; second phase was sustained for nearly 2 min. Bar, 20 s.

Fig. 2.

Average responses of change in Pcuff (ΔPcuff) for the 6 sessions (n = 8 dogs). Data points are presented every 2 s; error bars are omitted for clarity. Instillations were performed at time 0. Note sustained effect of HCl-P instillation into the larynx insessions 4–6.

Fig. 3.

Summary of changes in Pcuff. Values are means ± SE. For clarity, significance (P < 0.05) within session is indicated only for laryngeal vs. esophageal HCl-P instillations (*) and for laryngeal saline vs. HCl-P instillations (#).

Overall, stronger reflexogenic effects were detected from the larynx than from the esophagus (P = 0.0166). HCl-P elicited stronger responses when instilled into the larynx than into the esophagus (P = 0.00635).

HCl-P laryngeal instillations caused significantly larger airway smooth muscle contraction than did saline (P= 0.00543), whereas no statistical difference was detected for the esophageal instillations. Repeated exposure of the larynx, but not of the esophagus, to HCl-P enhanced the response of airway smooth muscle significantly (P < 0.001).


Our present results indicate a much greater importance of the larynx, compared with the esophagus, as a reflexogenic site in response to HCl-P. Our findings confirm the lack of effect of esophageal acidification on respiratory impedance in asthmatic subjects, as reported by Wesseling et al. (30), and on cough elicitation in cats, as reported by Tatar and Pecova (28). The present results are also in partial agreement with studies by other authors who attributed a significant role of the airway as reflexogenic site in humans (15), cats (29), and rabbits (31). These results are also well supported by our companion study on esophageal receptors (26a), in which 28 esophageal receptors from the cervical vagal nerve were tested, and only three were specifically but poorly activated by HCl-P instillation into the esophagus.

On the other hand, our present findings in anesthetized dogs do not agree with other studies conducted on conscious human subjects (17, 18,23, 26). Species differences and anesthesia could account for the discrepancy. In particular, depth of anesthesia has been shown to have profound effects on respiratory responses to irritant stimuli (21). In addition, we should also realize that, in most of the studies on human subjects, bronchoconstriction was assessed by conventional pulmonary function tests, such as forced expiratory flow in 1 s, which is generally less specific and sensitive than are direct measurements of smooth muscle activity. In fact, these tests cannot eliminate factors related to the particular efforts of the subjects (1, 14). Our present study instead relies on a direct measurement of smooth muscle contraction in an innervated portion of the airway, which is thus directly involved in reflex responses. Moreover, the large airways give a larger contribution to total airway resistance than do the small ones.

We also found that repeated exposures of the larynx to HCl-P solution significantly enhanced the contractile response of airway smooth muscle. This observation may have clinical significance, because chronic laryngitis is frequently associated with GER (5, 16, 27). Thus, in patients with chronic acid-induced laryngitis, repeated aspirations of gastric content into the larynx might cause increasingly severe bronchoconstriction. Although the present study was not aimed at disclosing the precise mechanisms of this enhancement by repeated HCl-P exposures, we may speculate that inflammatory processes or acid-induced injury of the laryngeal epithelium might be involved (24). However, for two main reasons, we do not think that the epithelial damage directly influenced the response of the tracheal smooth muscle:1) the increases in Pcuff were reversible and 2) the responses were recorded from a site (trachea) different from that where the instillations were performed (larynx). We might speculate that repeated instillations remove the protective layer of mucus and injure the epithelium, exposing the afferent endings more directly to the irritant stimuli and/or decreasing their response threshold. In any case, the contribution of chronic laryngitis to bronchial asthma has not been fully explored.

Laryngeal HCl-P instillations frequently caused biphasic responses in Pcuff. This type of response was very weak or absent with laryngeal saline instillations. We consider the second phase as a very important factor to explain the difference between HCl-P and saline instillations. In fact, the slopes of the increase in Pcuff immediately after laryngeal saline or HCl-P instillations are similar (Fig. 3). This might suggest that the response was rapidly adapting, possibly mostly due to the mechanical effect of the instillations. The second phase, instead, with its prolonged pattern, could be caused by the chemical stimulation of HCl-P. Similar responses were observed by Bradley et al. (4) with single fiber recordings from the superior laryngeal nerve of the sheep. They reported that the typical response to HCl (pH 2.0) applied to the epiglottis started as a high-frequency discharge, followed by low-frequency discharge and again a high-frequency discharge. The biphasic response could thus originate at the receptor level. In any case, even if their results are not directly applicable to our study, they are consistent with our observations.

Hong and Lee (11) and Hong et al. (12) also have reported a biphasic bronchoconstriction caused by cigarette smoke inhalation in guinea pigs. According to their study, the first phase was induced by a combination of cholinergic reflex and tachykinins, whereas the second phase involved cyclooxygenase metabolites. The involvement of noncholinergic system in the pathogenesis of GER was suggested by Hamamoto et al. (9). They proved that esophageal acid instillation caused neurogenic inflammation of the airways, and a neurokinin-1 receptor antagonist suppressed the response. Therefore, we speculate that the bronchoconstriction induced by HCl-P instillation into the larynx could also involve several mechanisms, including a cholinergic reflex as well as a nonadrenergic, noncholinergic system.

Generally, it is thought that the distal portion of the esophagus is more reactive than is the proximal segment. However, we did not find any difference between the two portions.

Concomitantly with smooth muscle contraction, cardiovascular responses were also noted with laryngeal instillations. Sustained hypertension was commonly seen, and, with HCl-P instillation, asystoles were also occasionally observed. These mechanisms can only be speculated on and go beyond the scope of this study.

In summary, our present results indicate a relatively greater importance of the larynx, compared with the esophagus, as a reflexogenic organ in GER-induced reflex bronchoconstriction. The response caused by laryngeal HCl-P instillation featured a biphasic response of airway smooth muscle contraction as well as an enhancement in magnitude after repeated exposures.


This study was supported by National Heart, Lung, and Blood Institute Grant HL-20122. T. Ishikawa was supported in part by Chiba University Hospital, Chiba, Japan.


  • Address for reprint requests and other correspondence: G. Sant’Ambrogio, Dept. of Physiology and Biophysics, The Univ. of Texas Medical Branch, 301 Univ. Blvd., Galveston, TX 77555-0641 (E-mail:gsantamb{at}

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 26a.
  28. 27.
  29. 28.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
View Abstract