Review: Effect of drugs on human cough reflex sensitivity to inhaled capsaicin

Capsaicin, the pungent extract of red pepper (capsicum), has gained widespread use as a research tool among clinical investigators, as it induces cough in humans in a safe [1], dose-dependent, and reproducible manner [2, 3]. Capsaicin cough challenge in humans was first described in 1984 [4], and has since been used to evaluate the effect of numerous pharmacological agents on cough reflex sensitivity. Although many drugs have been shown to inhibit induced cough in the laboratory, others have failed to do so, including agents widely regarded as clinically effective antitussives.

A United States National Library of Medicine (PubMed) search was performed in September, 2012 using the search terms “cough” and “capsaicin” limited to human studies published in English. The abstracts of the 328 articles meeting those search criteria were reviewed and 56 studies were identified in which capsaicin cough challenge was employed to assess the effect of a pharmacological intervention on cough reflex sensitivity. Studies in which a positive drug effect was demonstrated (n = 33) are listed in Table 1[5‐37]; trials in which no effect was noted (n = 30) are summarized in Table 2[4, 10, 12, 17‐19, 35, 37‐59]. In seven of these studies, multiple drugs and/or multiple subject groups were evaluated, resulting in both positive and negative results in terms of assessment of drug activity. As the purpose of this review was to assess drug trials in which a potential therapeutic (antitussive) effect of a drug was being evaluated, studies demonstrating enhancement of cough reflex sensitivity by angiotensin-converting enzyme (ACE) inhibitors or other agents were excluded.

This review has identified 33 studies in which a pharmacological intervention was demonstrated to inhibit cough reflex sensitivity to inhaled capsaicin in a variety of subject populations, thus supporting the role of cough challenge as a useful clinical tool in the evaluation of potential antitussives [3]. A striking difference between the studies showing a positive drug effect (Table 1), and those failing to demonstrate a change in cough reflex sensitivity (Table 2) is the predominant subject populations studied. Of the negative studies, 70% involved evaluation of healthy volunteers. Among the trials displaying a positive drug effect, only 27% evaluated healthy volunteers, while the majority (73%) investigated various forms of pathological cough. Of note, multiple agents were shown to inhibit cough reflex sensitivity in pathological cough, while having no effect in healthy volunteers, including guaifenesin [10, 17] and tiotropium [12] in cough due to acute viral upper respiratory tract infection (URI; common cold). The leukotriene receptor antagonist zafirlukast inhibited capsaicin-induced cough in subjects with cough-variant asthma [21], but not in stable asthmatics without cough and healthy volunteers [44]. Interestingly, gabapentin has recently been shown to improve cough-specific quality of life in patients with refractory chronic cough, without affecting cough reflex sensitivity [38]. This particular study highlights the concept that the optimal approach to the evaluation of a potential antitussive agent should be multifaceted, with cough reflex sensitivity measurement complementing other measures, such as objective cough counting and subjective symptom-based questionnaires.

Conspicuous in their absence from the list of agents having demonstrated the ability to inhibit cough reflex sensitivity to capsaicin during URI are codeine and dextromethorphan, two of the most commonly used agents worldwide for the treatment of cough due to the common cold [60]. The only agents demonstrating the ability to inhibit cough reflex sensitivity to capsaicin in healthy volunteers were theobromine [15], baclofen [25, 27], inhaled lignocaine [31, 36], sulindac [34], systemic opiates [37], menthol [5] and, in healthy smokers, nicotine [11]. Interestingly, this list includes drugs thought to be centrally acting antitussives, as well as agents whose cough-inhibiting properties are presumed to occur through a peripheral mechanism.

Limiting the evaluation of a potential modulator of cough reflex sensitivity to a study group of healthy volunteers, whose cough reflex is not hyperresponsive, may not allow the drug to demonstrate its inhibitory effect. Thus, subjects with pathological cough appear to comprise the optimal study population when evaluating the effects of a potential antitussive agent on cough reflex sensitivity. The particular type of pathological cough best suited for evaluation of a novel antitussive may depend on the specific pharmacological action of the drug, and currently remains a question under vigorous debate.