Copyright ©ERS Journals Ltd 2008 The expiration reflex from the trachea and bronchi1 Dept of Pathophysiology, Faculty of Medicine, Comenius University, Martin, Slovakia, and 2 University of London, London, UK. CORRESPONDENCE: J. Widdicombe, University of London, 116 Pepys Road, London, SW20 8NY, UK. Fax: 44 208261815. E-mail: JohnWiddicombeJ{at}aol.com Keywords: Cough, expiration reflex, larynx, tracheobronchial tree, vocal folds
Received: May 26, 2007
The expiration reflex (ER) is a forced expiratory effort against a closed glottis that subsequently opens to eject laryngeal debris and prevent aspiration of material. It is distinct from the cough reflex. Its source is usually assumed to be restricted to the larynx and vocal folds, and its possible origin from the tracheobronchial (TB) tree has been suggested but never studied. The current authors re-analysed previous records with mechanical or chemical stimulation of the TB tree to see if an ER can consistently be elicited, and to see whether it has properties similar to that from the larynx and vocal folds. A random review was made of some of the extensive literature on TB "cough" to see if it confirmed the authors conclusions. The TBER was consistently seen in cats and rabbits, either alone or followed by cough. These results are consistent with many studies in other species, including humans. It was enhanced, relative to cough, by inflation of the lungs and by general anaesthesia. Tracheobronchial expiration reflex occurs frequently with mechanical stimulation of the tracheobronchial tree. It differs fundamentally from many of the properties of "true" cough. Its features similar to the laryngeal expiration reflex suggest that both should be labelled "expiration reflexes" and not cough. Its existence should be taken into account in experimental, and possibly clinical, studies on tracheobronchial cough. The expiration reflex (ER) is defined as the strong expiratory effort, not preceded by an inspiration or accompanied by closure and then opening of the glottis; it results from mechanical and chemical stimulation of the larynx, especially the vocal folds. It was first described, but not named, in 1841 by Williams 1, when he placed his finger inside the opened trachea of a dog and observed that when his finger reached the larynx the dog made powerful expiratory efforts. It was comprehensively analysed and named by Korpas and co-workers 2–6, Hanacek et al. 7 and Tatar et al. 8 in the 1970s. The ER and the cough reflex (CR) differ in many physiological and pharmacological ways and have different functions. These differences have been described elsewhere 2, 4, 9–11. In the present study it is only necessary to point out that, according to the classic definitions, ER starts with an expiration while CR starts with an inspiration; implying quite different sensory or afferent inputs and/or central nervous processing; that the central nervous pathways for the ER have many different features from those for the CR 12, 13; that drugs, anaesthesia, lung inflation and blood gas tensions affect the two reflexes in different ways 5, 9–11, 14, 15; and that the function of the ER is to prevent aspiration of material into the lower airways, while that of the CR is to draw air into the lungs to promote a more efficient subsequent expulsion of mucus and airway debris. In nearly all the studies of ER, especially those by Korpas and co-workers 2–6, Hanacek et al. 7 and Tatar et al. 8, it is implied that the ER can only be elicited from the larynx or vocal folds, an appropriate site in view of the function of the reflex. However, this does not seem to be explicitly stated. Nor is the possibility that the ER cannot be elicited from the tracheobronchial (TB) tree. A survey of some of the literature on "cough" from the TB tree (TBCR) shows many illustrations of mechanical or, occasionally, electrical or chemical irritation of this region causing isolated expiratory efforts (TBER; see Discussion) or expiratory efforts followed by cough. This is similar to the ER from the larynx (LER), where there may be expiratory efforts associated with, but distinct from, coughs, and sometimes followed by coughs. However, authors seldom comment on this phenomenon. The term "cough reflex" is nearly always applied to both responses resulting from stimulation of the TB tree. Two of the current authors (M. Tatar and J. Hanacek) have re-evaluated previous records on mechanical and chemical stimulation of the TB tree and larynx in cats and rabbits 2, 3, 7, 8, 14, 16. The aim of the present study is to present these results and discuss the significance of the TBER.
The methods have been fully described elsewhere 2–4, 8, 16, 17. However, the current authors have re-analysed the experimental results of these previous studies. Briefly, cats and rabbits, of either sex, weighing 2.4–3.8 kg were anaesthetised with pentobarbitone sodium (30 mg·kg–1, i.p. and i.v., respectively). Supplementary doses were added to maintain appropriate levels of anaesthesia. The depth of anaesthesia was assessed from pattern of breathing, level and stability of end-tidal CO2 tension and arousal responses. Tracheal tubes were inserted. Unanaesthetised cats with healed chronic tracheal cannulations were also used 14. In cats, a silon fibre (0.4 mm diameter) was inserted through the tracheal cannula either to the level of the tracheal bifurcation or to the laryngeal region, and gently pushed up and down for 7 s. In rabbits, a soft venous catheter was used to stimulate the TB tree and the larynx. The TB irritation would stimulate sensors both in the trachea and in the larger bronchi. For the larynx both the vocal folds and the adjacent areas would be stimulated. To stimulate only the vocal folds in rabbits, a silon fibre loop (3–5 mm diameter) was moved across the folds from the tracheal side. The current authors define the CR as an inspiration followed immediately by a strong expiration, and the ER as a strong expiration not immediately preceded by an inspiration (see Discussion). In anaesthetised animals, cough and the ER were recorded as the changes in intrapleural pressure, via an intrapleural cannula connected to a pressure transducer and electromanometer. To evaluate the effects of lung inflation in cats an oesophageal balloon catheter was connected to a pressure transducer. The lungs of cats and rabbits were inflated by 0.5, 1.0 and 1.5 kPa maintained positive pressures and the tracheobronchial region or the vocal folds were stimulated. Aerosols of capsaicin solution (100 µg·mL–1, median particle diameter 6 µm) were administered to cats via the tracheal cannula for 3–7 breaths; tracheobronchial capsaicin is ineffective as a cough stimulant in rabbits 18. In conscious cats, cough and the ER were recorded as the changes of lateral intratracheal pressure via a tracheal cannula connected to a pressure transducer 8, 16. The effects of general anaesthesia and of inflation of the lungs were statistically analysed nonparametrically, using the Chi-squared test. A p-value of <0.05 was taken as significant.
Pattern of response to TB stimulation Table 1
Stimulation of the larynx and vocal folds Table 2
Stimulation restricted to the vocal folds in rabbits always caused ERs (83%) or an ER followed by cough (17%).
Anaesthesia and TB stimulation
Effects of lung inflation
With stimulation of the vocal folds in rabbits, lung inflation caused little change in the response, which was predominantly ERs. Isolated ERs changed from 83 to 81% (nonsignificant) and ERs followed by CR changed from 17 to 19% (nonsignificant).
Capsaicin stimulation
The current authors analysed all appropriate and available records of previous research on cats and rabbits (conducted by M. Tatar and J. Hanacek). Analysis of the results shows the following. 1) Mechanical stimulation of the TB tree of cats and rabbits frequently causes an ER. This has been illustrated before but seldom identified as such. 2) In general, the reflexes from the TB tree of cats are similar in pattern to those from the larynx, but in rabbits the ER is more frequent from the larynx than from the TB tree. 3) In cats, anaesthesia significantly increases the incidence of the TBERs compared with the TBCRs, but makes little difference in the response to laryngeal stimulation. 4) Inflation of the lungs in cats and rabbits makes the TBER compared with coughs significantly more frequent, but makes little difference to the response to stimulation of the vocal folds, in which case the ER dominates. 5) In cats, capsaicin causes TBCRs but not TBERs. The definitions of the TBCR and the TBER are stated previously (see Methods). While it is customary in clinical studies to define cough as all expiratory efforts, with no distinction between the TBCR and the TBER, the present authors believe that in analytical and laboratory studies the distinction and definitions are important, since they point to different mechanisms, as established for the larynx 2, 3, 5, 9–11 and, as indicated in the present study, for the TB tree. Indeed a recent Task Force Report 19 supported this use of two definitions, although it did not name them. The current authors suggest that TBER is an appropriate name in view of its similarity with the LER, a term well established in the literature. It is appreciated that in a bout of coughing (a cough "attack" or "epoch") distinction of the ER from the expiratory effort that is part of the CR is often difficult; but since the TBER can occur in isolation or before a cough it must be a genuine phenomenon distinct from cough. Statistical analysis of most of the current results has not been attempted since they were derived from several experimental runs, which were not always carried out under identical conditions. In addition, the conclusion that the ER can frequently be elicited from the TB tree is essentially a qualitative one. In quantitative terms the sizes and frequencies of the responses will depend on many factors, such as presence, type and depth of anaesthesia, species, and nature and duration of airway stimuli. Statistics have only been applied to assess the changes due to lung inflation and anaesthesia.
In total, 18 papers describing mechanical (and occasionally electrical or chemical) stimulation of the TB mucosa to cause coughs have been identified at random. Of these, 11 studies showed initial expiratory efforts, sometimes isolated and sometimes followed by cough 20–30, and the total list is almost certainly much greater. Few of the authors referred to the ER, probably because most of the studies were published before the naming of the laryngeal ER by Korpas and co-workers 2–8 in the 1970s. The species studied include anaesthetised and unanaesthetised cats, dogs and rabbits, and humans under light anaesthesia 27, 28. Most of these studies involved a brief punctate stimulus, mechanical or electrical, and the response was often a single ER without subsequent cough. Figure 2
For the results assessed in the present analysis the mechanical stimulus was maintained for 7 s, so it bore little relationship to the timing of the respiratory cycle, and it is probably not surprising that the reactions were usually lengthy and generally included both ERs and cough. An ER has eventually to be followed by an inspiration (since lung volume has been greatly reduced), and this could be the first phase of a cough. In lightly anaesthetised humans, the same stimulus to larynx and trachea produced almost identical responses, an ER followed by coughs 27, 28. Although in most, but not all, of the studies cited, general anaesthesia was used (which would depress cough relative to the ER) the TBER was also clearly seen in unanaesthetised animals 20, 21, 25. Anaesthesia might preferentially inhibit inspiration and prolong expiration; such a change might affect the inspiratory efforts of the CR, leaving the expirations of the ER more prominent. However, in cats pentobarbitone anaesthesia decreased the expiratory efforts of TBCR without significantly changing inspiratory values 14. The present authors have not analysed the results of the study with this question in mind.
The distinction between the TBER and cough is reinforced by the following observations, most of which are confirmed in the present analysis. 1) Inflation of the lungs enhances the ER relative to cough 4. The inflation would stimulate slowly adapting pulmonary stretch receptors, and this might cause a preferential inhibition of inspiratory, compared with expiratory, efforts. 2) Capsaicin, unlike mechanical stimulation, induces cough but not the ER 31. 3) Opioid drugs are more effective in blocking the TBCR than the TBER (fig. 2 While the teleological purpose of the LER is clear, i.e. to prevent aspiration of material into the lower airways, the purpose of the TBER is less obvious. Presumably, if material is aspirated the TBER is a second line of defence. One can speculate that the ER is a more primitive reflex in evolution than cough. Some fish can cough either forwards or backwards 32–34. If a chemical irritant enters the mouth, the fish makes an expulsive movement to expel the irritant forwards out of the mouth, which would protect the delicate gills behind the pharynx. If the irritant enters the gill cavity from behind, it is propelled backwards to remove it from the gills. There seems to be no equivalent in fish of the initial inspiration that is part of the definition of cough in mammals. One might suggest, perhaps with a flight of fancy, that the forward expulsion from the oropharynx is the equivalent of the ER from the larynx in mammals, and prevents entry of the irritant; while the backwards expulsion from the gills is like the ER from the TB tree in mammals (apart from the direction of flow), and expels the irritant once it has gained entry. Of course this is not to suggest that the neurological mechanisms of "cough" and the ER are identical in fish and mammals. Another possibility is that the TBER might be due to "vagal overflow". Cough can be induced from the oesophagus 35 and from the external ear 36, both innervated by the vagus nerves; these actions seem to be examples of "unintelligent design", since they appear to provide no apparent useful function. However the TBER is not in this category; its useful function is as a reserve defence process if aspirated material overcomes the powerful LER (the laryngeal gatekeeper).
What is not clear is which neural sensors (afferent receptors) are responsible for the two reflex patterns. There is currently intensive research on the sensors for cough 37–39. These must differ from those for the TBER, since the CR starts with an inspiration and the ER an expiration. One can conceive that the same particular afferent pathway could cause either inspiration or expiration and that the direction of the response is determined by a second input to the brainstem, which acts as a gate or switch in the brainstem; however, this is still saying that the neural inputs for the CR and the ER must be different. Forced expirations, such as those seen with the ER and the CR, will stimulate airway receptors and set up feedback loops that will secondarily influence the defensive reflexes. The reflex role of these feedbacks, or even their existence, has not been established, but the forced expiration of the ER could promote inspiration as the first component of a cough. The cough response to capsaicin, i.e. coughs without ERs, may be related to the fact that capsaicin is a rather selective stimulant of nociceptors, such as C-fibre sensors 37. There is no known stimulus that always causes ERs but never coughs, although mechanical stimulation of the rabbit vocal folds comes close. The ER from the larynx in humans has latency (from laryngeal stimulation to abdominal muscular response) of In conclusion, the current authors believe that there is a true expiration reflex from the tracheobronchial tree, with many physiological and pharmacological properties similar to the well-established expiration reflex from the vocal folds and larynx. Although there have been clues to this observation in the literature, the present paper seems to be the first attempt to emphasise the existence of the tracheobronchial expiration reflex. This conclusion, if accepted, is important because the tracheobronchial expiration reflex must be taken into account during laboratory and clinical research on defensive reflexes from the lower airways.
None declared.
The authors are grateful to J. Korpas (Comenius University, Martin, Solvakia), whose inspiring research and leadership was the foundation for this study; and to G. Fontana (Universita di Firenze, Firenze, Italy) for helpful suggestions.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||