DISCUSSION
This is the first human study to determine whether both peripherally acting and centrally acting GABAB agonists modulate cough responses to inhaled capsaicin compared with placebo. The data suggests that the actions of GABAB agonists, in healthy volunteers occur in the central rather than the peripheral nervous system. Lesogaberan (peripherally acting GABAB agonist) had no significant inhibitory effect on the cough reflex whereas baclofen (centrally acting GABAB agonist) increased the ED50. Interestingly, lesogaberan had a small but significant increase on Emax i.e. the subjects coughed, slightly, more on this drug.
The dose response curve from our data (Figure 3) demonstrates a subtle but significant right shift of the baclofen curve, suggesting a stronger stimulus is required to evoke coughing when on baclofen treatment, consistent with the existing literature. Dicpinigaitis et al, studied a similar group pf subjects to those in this study (20 healthy volunteers, mean age 30) who were dosed chronically with baclofen 10mg three times a day (6) and following that study he went on to study 20mg dosing in a similar group (7). In both of these studies the baclofen group demonstrated a significant increase in the log C5 capsaicin concentration. This is the concentration of capsaicin that induces 5 coughs which is broadly comparable to the ED50 in our study. There are two small case series in the literature which have suggested baclofen might also treat chronic cough. One was an open-label study evaluating a 4-week course of low-dose, oral baclofen against ACE inhibitor-induced cough (8) suggesting that baclofen treatment caused a prolonged antitussive effect. The other study reported suppression of cough and capsaicin cough responses in two chronic cough patients after a 14-day course of baclofen, 10mg; three times a day (5). Participants in both of these studies did not report significant adverse effects, however, in clinical practice patients report significant dizziness (up to 28% of patients) and drowsiness (up to 10% of patients) which limits its use (10).
It is worth noting that changes in experimentally evoked cough do not necessarily equate to clinical anti-tussive effects on spontaneous coughing. This was demonstrated by Belvisi et al., 2017 who studied the effect of a potent TRPV1 antagonist (XEN-D0501) vs placebo on capsaicin evoked cough and spontaneous 24-hour cough frequency (11). Despite XEN-D0501 leading to significant reduction in the capsaicin Emax, there was no improvement in objective cough frequency. In addition to this some agents (such as morphine and gabapentin) that have been shown to improve cough specific quality of life, and are currently used off-license in a clinical setting, do not appear to modulate the C5 endpoint in cough challenge (12). However, it remains unclear which challenge agents, methodology and endpoints are optimal at present. Thus the main role of cough challenge seems to be in confirming target engagement and understanding whether mechanisms translate from animal studies to humans.
Interestingly, the data from this study suggests a difference between the effects of lesogaberan in humans compared with a guinea pig model. Canning et al., found that both lesogaberan and baclofen significantly reduced citric acid evoked coughs in guinea-pigs (4). Possible explanations for this include differences in the blood brain barrier between species and also relatively higher levels of dosing in the guinea pig, which may have saturated the GATs responsible for inhibiting the CNS effects of lesogaberan.
Despite the relatively small number of subjects in this study, statistically significant changes between the treatments were teased out by the dose response inhalational capsaicin challenge, suggesting the utility and power of this particular method and these endpoints when assessing cough reflex responses. The small but statistically significant increase in Emax with lesogaberan therapy is interesting, paradoxical in the context of the inhibitory nature of GABAB and difficult to explain. Our prediction was that Emax would have reduced with lesogaberan, mirroring the changes in animal studies. One potential explanation for the lack of reduction in Emax with lesogaberan in healthy volunteers is that GABAB is involved in endogenous inhibitory mechanisms controlling cough. In the healthy state, these control mechanisms are intact and thus additional GABAB agonism may have no effect on cough responses, whereas in patients with chronic cough there is increasing evidence that these descending inhibitory mechanisms may be deficient (13-15) and in such circumstances GABABagonism could have an effect in chronic cough patients not seen in healthy controls.
This study has some limitations. Firstly we used single doses of lesogaberan and baclofen, as opposed to multiple dosing. However, we were still able to replicate previous findings with baclofen after two weeks of dosing. Another possible limitation of this study is that capsaicin was used for the cough challenge, whereas the animal studies used citric acid. This choice was made in light of the fact that all previous human studies suggesting an effect of baclofen on cough responses had all employed capsaicin challenge, furthermore this particular challenge methodology has not been performed using citric acid. An effect of lesogaberan on citric acid but not capsaicin evoked cough cannot be excluded.
In conclusion, this data suggests GABAB receptor agonists are inhibitory of the human cough reflex in health, but only through their activity in the central nervous system. It remains to be seen whether peripheral GABAB agonists have any effect on the sensitized cough reflex in chronic cough patients.