Allergen challenge increases capsaicin-evoked cough responses in patients with allergic asthma

Allergen challenge increases capsaicin-evoked cough responses in patients with allergic asthma Imran Satia, MD, PhD,a,b Richard Watson, BSc,a Tara Scime, BSc,a Rachel J. Dockry, PhD,b,c Shilpi Sen, BSc,b,c James W. Ford, BSc,b,c Patrick D. Mitchell, MD,d Stephen J. Fowler, MD,b,c Gail M. Gauvreau, PhD,a Paul M. O’Byrne, MD,a,b and Jaclyn A. Smith, MD, PhDb,c Hamilton, Ontario, and Calgary, Alberta, Canada, and Manchester, United Kingdom GRAPHICAL ABSTRACT

From athe Department of Medicine, Division of Respirology, McMaster University, Hamilton; bthe Division of Infection, Immunity and Respiratory Medicine, University of Manchester, and the Manchester Academic Health Science Centre; cthe NHS Foundation Trust, University of Manchester; and dthe Department of Medicine, University of Calgary. This study was funded by the British Medical Association James Trust Award. I.S. is funded by a fellowship from the European Respiratory Society/European Union Marie Curie Award. Disclosure of potential conflict of interest: I. Satia reports grant from the BMA James Trust Award and the North West Lung Centre Grant; personal fees from Educational Talks for GPs; and sponsorship to attend conference meetings outside the submitted work. P. D. Mitchell reports grants from the Irish Asthma Society and Teva and personal fees from Educational Talks for GPs and specialists outside the submitted work. P. M. O’Byrne reports personal fees from the Oversight Committee for LABA Safety study; consulting fees from AstraZenca, GlaxoSmithKline, Merck, and Boehringer; and grants from AstraZeneca and Genentech outside the submitted work. J. A. Smith reports grants from the British Medical Association James Trust

during the conduct of the study and grants and personal fees from Ario Pharma, GlaxoSmithKline, NeRRe Pharmaceuticals, Menlo, and Bayer; personal fees from Boehringer Ingleheim, Genentech, and Neomed; grants and personal fees from Merck; nonfinancial support from Vitalograph; personal fees from Cheisi; and grants and personal fees from Afferent outside the submitted work. In addition, J. A. Smith has a patent A method for generating output data licensed. The rest of the authors declare that they have no relevant conflicts of interest. Clinical Trial Registration: www.controlled-trials.com ISRCTN79930571. Received for publication June 22, 2018; revised October 24, 2018; accepted for publication November 30, 2018. Corresponding author: Jaclyn A. Smith, MD, PhD, Division of Infection, Immunity and Respiratory Medicine, University of Manchester, University Hospital of South Manchester, Level 2, Education and Research Centre, Manchester M23 9LT, United Kingdom. E-mail: [email protected]. 0091-6749/$36.00 Ó 2019 American Academy of Allergy, Asthma & Immunology https://doi.org/10.1016/j.jaci.2018.11.050

1

2 SATIA ET AL

Background: Cough is a common and troublesome symptom in asthmatic patients, but little is known about the neuronal pathways that trigger cough. The mechanisms by which airway inflammation, airway hyperresponsiveness, and variable airflow obstruction cause cough are unclear. Objective: We sought to investigate the effects of allergen exposure on cough reflex sensitivity. Methods: We performed a 9-visit, randomized, single-blind, placebo-controlled, 2-way crossover study comparing cough responses to inhaled capsaicin in patients with mild atopic asthma after allergen challenge compared with diluent control. Full-dose capsaicin challenge was performed at screening to determine the capsaicin dose inducing a half-maximal response, which was subsequently administered at 30 minutes and 24 hours after inhaled allergen/diluent challenge. Spontaneous coughing was measured for 24 hours after allergen/diluent. Methacholine challenge and sputum induction were performed before and after allergen/diluent challenge. Results: Twelve steroid-naive subjects completed the study (6 female subjects; mean age, 34.8 years). Allergen inhalation caused both an early (mean 6 SD, 38.2% 6 13.0%) and late (mean 6 SD, 23.7% 6 13.2%) decrease in FEV1 and an increase in sputum eosinophil counts 24 hours later (after diluent: median, 1.9% [interquartile range, 0.8% to 5.8%]; after allergen: median, 14.9% [interquartile range, 8.9% to 37.3%]; P 5 .005). There was also an increase in capsaicin-evoked coughs after allergen exposure compared with diluent at both 30 minutes (geometric mean coughs, 21.9 [95% CI, 16.5-29.20] vs 12.1 [95% CI, 8.3-17.7]; P < .001) and 24 hours (geometric mean coughs, 16.1 [95% CI, 11.3-23.0] vs 9.8 [95% CI, 6.1-15.8]; P 5 .001). Allergen exposure was also associated with an increase in spontaneous coughs over 24 hours. Conclusion: Allergen-induced bronchoconstriction and airway eosinophilia result in increased cough reflex sensitivity to capsaicin associated with an increase in 24-hour spontaneous coughing. (J Allergy Clin Immunol 2019;nnn:nnn-nnn.) Key words: Asthma, allergen, capsaicin, cough, transient receptor potential vanilloid type 1

Asthma is a complex heterogeneous disease in which the pathophysiology is currently thought to consist of 2 predominant dimensions: airway inflammation and airway hyperresponsiveness (AHR). However, the relationship between the fundamental pathophysiology and development of symptoms, such as cough, wheeze, shortness of breath, and chest tightness, is unclear. Patients with severe asthma have demonstrated discordance between airway eosinophilia and symptoms,1-3 and although recent novel biologics have demonstrated improvements in airway eosinophilia, lung function, and risk of exacerbation, there has been a paucity of data showing clinically meaningful improvements in day-to-day symptoms and quality of life.4-7 Cough is a common1,2,8,9 and troublesome10 symptom in asthmatic patients and is triggered by airway nerves sending impulses to the central nervous system and, unlike other symptoms, can be quantified objectively and evoked experimentally. The relationship between airway inflammation, AHR, and cough requires further investigation to ascertain whether neuronal function is an important dimension in the pathophysiology of cough in asthmatic patients.

J ALLERGY CLIN IMMUNOL nnn 2019

Abbreviations used AHR: Airway hyperresponsiveness EAR: Early asthmatic response ED50: Capsaicin dose inducing half-maximal response IQR: Interquartile range LAR: Late asthmatic response TRPA1: Transient receptor potential ankyrin-1 TRPV1: Transient receptor potential vanilloid type 1

We have provided evidence for neuronal dysfunction in patients with mild-to-moderate stable asthma, as demonstrated by an exaggerated response to inhaled capsaicin compared with that seen in healthy volunteers.11 Cough responses were independent of methacholine AHR, baseline lung function, or fraction of exhaled nitric oxide, although this was in a stable group of patients with asthma with minimal airflow obstruction. We subsequently investigated the interaction between bronchoconstriction and capsaicin-evoked cough and showed that bronchoconstriction induced by inhaling methacholine resulted in an increase in cough responses to inhaled capsaicin, which gradually diminished as airway caliber improved.12 This suggested that smooth muscle contraction increases neuronal function, as measured based on the cough reflex. An important missing component in these studies was directly assessing the influence of airway inflammation on airway nerves that trigger the cough reflex. Previous observational cross-sectional studies have demonstrated no relationships between airway inflammation, asthma control, and spontaneous coughs.13 However, the effect of directly increasing airway inflammation in individual patients on spontaneous or experimentally evoked coughs using capsaicin was not assessed. Therefore the aim of this study was to compare cough responses to inhaled capsaicin in patients with mild steroidnaive atopic asthma during and 24 hours after inhaled allergen challenge compared with diluent (saline) control. The primary end point was the number of evoked coughs after inhaling the capsaicin dose inducing half-maximal response (ED50) after a full-dose capsaicin challenge determined at the baseline visit. We also explored differences in spontaneous hourly cough rates for 24 hours after the allergen and diluent challenges.

METHODS Subjects Participants with mild steroid-naive atopic asthma who demonstrated an early and late bronchoconstriction response to an inhaled allergen were recruited. All participants had a baseline FEV1 of 70% of predicted value or greater and evidence of methacholine AHR (PC20 < 16 mg/mL) with wellcontrolled asthma according to Global Initiative for Asthma classification and were receiving stable medication for at least 4 weeks. We excluded current smokers, those with a recent exacerbation or uncontrolled symptoms, and use of medication that might have altered cough responses (eg, opiates, gabapentin, anticholinergics, and theophylline). The protocol was approved by the local research ethics committees (15/NW/0787 and HiREB:1042) and Health Canada (Control no. 191467 and protocol ISRCTN7993057). All subjects provided written informed consent.

Study protocol and procedures This was a 9-visit, 2-center, randomized, single-blind, placebo-controlled, 2-way crossover study comparing cough responses to inhaled capsaicin in

SATIA ET AL 3

J ALLERGY CLIN IMMUNOL VOLUME nnn, NUMBER nn

FIG 1. Summary of study design and visits. *Subjects in the Manchester center did not attend for methacholine challenge after the screening allergen challenge.

patients with allergic asthma 30 minutes and 24 hours after exposure to allergen compared with diluent (saline) control (Fig 1). Full details of all study procedures are found in the Methods section in this article’s Online Repository at www.jacionline.org. Participants were invited for 3 initial screening visits. On day 1, history and examination were performed, followed by spirometry; a full-dose capsaicin cough challenge, as described previously11; skin prick testing and allergen dose titration14,15; and a methacholine challenge (PC20) using the 2-minute tidal breathing method.16 The ED50 during the full-dose capsaicin challenge was used as a single-dose challenge at visits 5, 6, 8, and 9 (Fig 1). On day 2, participants returned for an incremental inhaled allergen challenge to measure the early asthmatic response (EAR; ie, greatest decrease in FEV1 between 0 and 3 hours) and late asthmatic response (LAR; greatest decrease in FEV1 between 3 and 7 hours), as previously described.17 Subjects at the McMaster site attended on visit 3 for a methacholine challenge after completing the screening allergen challenge, whereas in Manchester safety and adverse events were assessed over the telephone. Only patients with both an EAR _20% decrease in FEV1) and a LAR (> _15% decrease in FEV1) were included (> in subsequent visits. Patients who met the inclusion criteria were randomized to 2 study periods of 3 consecutive days at least 2 to 4 weeks apart and 2 to 4 weeks after the screening allergen challenge. At visit 4, methacholine challenge was performed, followed by sputum induction. At visit 5, the day after, participants were fitted with an ambulatory cough monitor (VitaloJAK; Vitalograph, Buckinghamshire, United Kingdom) for the next 24 hours and asked to inhale either allergen/diluent in a randomized order. Thirty minutes after the end of the allergen/diluent challenge, participants were administered 4 inhalations of capsaicin ED50 30 seconds apart, and the number of evoked coughs was measured. Patients were monitored for 7 hours after the allergen/diluent challenge with regular measurements of FEV1. Salbutamol was administered at the end of 7 hours, and patients were told to take additional inhalations at home, if required. At visit 6, 24 hours after the start of the challenge, patients performed baseline spirometry, followed by repeat capsaicin ED50 cough challenge. This was followed by a methacholine challenge and sputum induction. At least 2 weeks later, participants returned to repeat the above triad at visits 7, 8, and 9.

Statistical analysis Data analysis was done with SPSS software (version 22.0; IBM, Armonk, NY). Individual data are shown for baseline demographics. This study generated data for numbers of capsaicin-evoked coughs, changes in spontaneous 24-hour cough rates, FEV1, methacholine PC20, and sputum eosinophilia. Summary data are presented as means and SDs or medians and interquartile ranges (IQRs). The primary outcome was the number of capsaicin-evoked coughs at 30 minutes and 24 hours after allergen challenge

compared with diluent challenge. Prior data indicated that the difference in the number of coughs at the ED50 dose of capsaicin of matched pairs is normally distributed with an SD of 2.78. Twelve subjects were required for 90% power to detect a difference in the mean response of matched pairs of 61.28 by using a 2-sided a value of .05. The effect of bronchoconstriction and sputum eosinophilia on capsaicin-evoked and spontaneous coughs after allergen/diluent challenge was analyzed by using generalized estimating equations (exchangeable correlation structure). ED50 capsaicin-evoked coughs and spontaneous 24-hour coughs were log transformed. These generalized estimating equation models were used to estimate means and 95% CIs.

RESULTS Subjects We recruited 15 subjects with mild steroid-naive asthma, 12 of whom completed the study (Table I). Of the 2 who did not complete the study, 1 patient had an upper respiratory tract infection and changed residence, and 2 did not have an LAR. All participants had seasonal allergic rhinitis, no clinical history of gastroesophageal reflux disease, good baseline lung function (mean 6 SD percent predicted FEV1, 92.4% 6 10.2%), and evidence of methacholine AHR (mean 6 SD PC20, 2.1 6 1.6 mg/ mL) and were sensitive to house dust mite allergen extract, cat, or ragweed. During the screening allergen challenge, the mean percentage decrease in FEV1 in the EAR was 38.0 6 10.3 and that in the LAR was 27.0 6 14.1 relative to baseline. All patients coughed during the baseline full-dose capsaicin cough challenge; median maximum cough response evoked by any concentration of capsaicin was 19.5 coughs (15.5-35.8 coughs), and ED50 was 31.3 mmol/L (15.6-62.5 mmol/L). Lung function after allergen and diluent challenge There were significant differences in the decrease in percentage of FEV1 from baseline after inhaled allergen compared with diluent challenge in both the EAR (mean 6 SD decrease in FEV1 percentage, 38.2% 6 13.0% after allergen vs 4.5% 6 2.8% after diluent; P < .001) and LAR (mean 6 SD decrease in FEV1 percentage, 23.7% 6 13.2% after allergen vs 2.2% 6 2.1% after diluent; P < .001, Fig 2). There was a recovery in FEV1 24 hours after allergen and diluent challenge compared with baseline FEV1 on the day of the challenge (mean 6 SD percentage decrease in FEV1, 8.6% 6 9.4% after allergen vs

4 SATIA ET AL

J ALLERGY CLIN IMMUNOL nnn 2019

TABLE I. Baseline demographics and screening data of individual participants Patient ID 1 Age (y) Sex BMI (kg/m2) FEV1 (L [% predicted]) FVC (L [%predicted]) Methacholine PC20 (mg/mL) Allergen EAR screen (% decrease in FEV1) LAR screen (% decrease in FEV1)

Emax (coughs [no.]) ED50 (mmol/L capsaicin)

2

3

4

5

6

7

8

9

10

11

12

Mean

SD

19 19 25 24 30 68 27 51 23 46 53 32 34.8 15.8 M F M F M M F F F F M M 23.1 21.3 25.9 19.1 29.9 24.6 24.5 31.9 24.7 25.3 23.5 30.4 25.3 3.8 3.51 (91) 3.70 (97) 3.84 (110) 2.60 (90) 3.77 (99) 3.33 (100) 2.96 (85) 2.59 (89) 2.71 (84) 1.95 (71) 3.80 (103) 3.36 (90) 3.18 (92.4) 0.61 (10.2) 4.76 (105) 4.54 (103) 4.40 (108) 3.03 (89) 4.51 (100) 4.27 (95) 3.71 (91) 3.27 (87) 3.13 (84) 2.79 (81) 4.87 (102) 4.52 (98) 3.98 (95.3) 0.75 (8.6) 0.1

2.7

0.7

4.0

0.3

4.3

2.4

1.3

HDM 31.8

HDM 28.1

HDM 31.1

HDM 41.5

HDM 42.1

Cat 63.6

Cat 43.0

Cat 31.0

65.2

16.4

27.8

19.8

16.9

18.2

32.2

15.7

18 7.8

17 15.6

14 62.5

13 3.9

15 31.3

19 125

58 15.6

36 125

3.6

0.2

Ragweed Ragweed 47.1 29.7

1.6

3.8

2.1

1.6

Cat 38.0

HDM 29.0

38.0

10.3

27.0

14.1

21.0

22.5

28.1

40.0

35 62.5

21 31.3

20 62.5

53 15.6

Median

IQR

19.5 31.3

15.5-35.8 15.6-62.5

BMI, Body mass index; Emax, maximum cough response evoked by any concentration of capsaicin; F, female; HDM, house dust mite; M, male.

FIG 2. Changes in lung function after allergen/diluent challenges. Data shown are means and 95% CIs.

21.8% 6 5.4% after diluent; ie, there was a 1.8% improvement in FEV1 percentage at 24 hours after diluent challenge).

Capsaicin cough responses after allergen challenge There was an increase in capsaicin ED50 coughs 30 minutes after an allergen challenge compared with diluent (geometric mean coughs, 21.9 [95% CI, 16.5-29.2] vs 12.1 [95% CI, 8.3-17.7]; P < .001; Fig 3, A). There was a reduction in capsaicin ED50 coughs after 24 hours, but participants still coughed more 24 hours after allergen than after diluent (geometric mean coughs, 16.1 [95% CI, 11.3-23.0] vs 9.8 [95% CI, 6.1-15.8]; P 5 .001; Fig 3, B). There was no period or sequence effect. There was a significant interaction between the decrease in FEV1 and capsaicin-evoked coughs both at 30 minutes and

24 hours (P < .001), and there was an independent effect of the allergen at both time points (P 5 .028), implying allergeninduced airway inflammation also affected capsaicin-evoked coughs independent of a decrease in FEV1.

Spontaneous 24-hour cough rates after allergen challenge There was an increase in spontaneous cough rate over 24 hours after allergen compared with diluent challenge (geometric mean 24-hour cough rate, 0.60 [95% CI, 0.32-1.11] vs 0.25 [95% CI, 0.16-0.41]; P < .001). The decrease in FEV1 after allergen/diluent was related to log spontaneous cough frequency (b 5 20.72, P < .001; Fig 4).

SATIA ET AL 5

J ALLERGY CLIN IMMUNOL VOLUME nnn, NUMBER nn

FIG 3. Decrease in FEV1 from baseline (left y-axis) with capsaicin ED50 cough values (right y-axis) after 30 minutes (A) and 24 hours (B). Bars represent geometric means calculated by using generalized estimating equations modeling.

FIG 4. Spontaneous cough over 24 hours after allergen compared with diluent challenge. Bars represent geometric mean and 95% CI estimates using generalized estimating equations modeling.

Changes in methacholine PC20 before and after allergen/diluent challenge There was a decrease in methacholine PC20 24 hours after inhaled allergen challenge compared with diluent challenge (median PC20, 1.3 mg/mL [IQR, 0.4-2.3 mg/mL] after diluent vs 0.5 mg/mL [IQR, 0.1-0.8 mg/mL] after allergen; P 5 .05). There was no difference in methacholine PC20 before the allergen/ diluent challenge (median PC20, 2.1 mg/mL [IQR, 0.6-2.8 mg/ mL] before diluent vs 1.1 [IQR, 0.4-2.3 mg/mL] before allergen; P 5 .27; Fig 5). Changes in sputum eosinophil counts before and after allergen/diluent challenge Allergen challenge increased the percentage of sputum eosinophils at 24 hours after inhaled allergen compared with that after diluent challenge (median eosinophilia, 1.9% [IQR, 0.8% to 5.8%] after diluent vs 14.9% [IQR, 8.9% to 37.3%] after allergen;

P 5 .007; Fig 6, A). Capsaicin-evoked coughs 24 hours after allergen/diluent were independently related to both sputum eosinophilia (b 5 0.007, P < .001) and the decrease in FEV1 (b 5 20.27, P 5 .004).

Safety of capsaicin and allergen challenge There were no serious adverse events throughout the entire study period. Capsaicin-evoked coughing was not associated with any additional decrease in FEV1 at 30 minutes or 24 hours. After allergen challenge, 1 subject demonstrated a decrease in FEV1 of 57% during the LAR. Salbutamol was administered before discharge from the clinical research facility, and the subject took additional salbutamol overnight but did not require prednisone or hospitalization. The subject presented the morning after with an FEV1 34% below baseline, and hence methacholine challenge was not performed, but capsaicin cough challenge was performed, and there was no decrease in FEV1.

6 SATIA ET AL

FIG 5. Changes in methacholine PC20 before and after allergen and diluent challenge. Horizontal bars represents median values. *One patient whose FEV1 was greater than 20% below baseline value the day after allergen and in whom methacholine challenge was not performed.

DISCUSSION To our knowledge, this is the first study to investigate both experimentally evoked cough responses and objectively quantified 24-hour spontaneous cough frequency both during and after an allergen challenge. Our data provide evidence to suggest that inhaled allergen is associated with an increase in capsaicinevoked cough response at 30 minutes that persists at 24 hours. The change in cough response at 30 minutes was predicted by a decrease in FEV1, although at 24 hours, it was predicted by both FEV1 and percentage of sputum eosinophils. Allergen inhalation was also associated with an increase in overall spontaneous coughs similarly related to changes in FEV1. These results suggest that inhaled allergen challenge in patients with atopic asthma (indirect bronchial provocation challenge) can sensitize the sensory nerve afferents responsible for causing cough through bronchoconstriction and inflammation. The results of our study are in contrast to those of Minogushi et al,18 who, despite an increase in sputum eosinophil counts and AHR, showed no change in the dose of concentration of capsaicin inducing at least 5 coughs after allergen challenge. However, that small study was a parallel-group design, and 3 of the 9 patients did not have an LAR. Moreover, the end point of concentration of capsaicin inducing at least 5 coughs might not be the optimal end point12,19 and was only measured 24 hours after allergen. There have been no other studies that have directly investigated the effects of inducing airway inflammation and AHR on cough, and hence our understanding of asthma pathophysiology and cough has been based on observational studies in human subjects, which previously showed no relationship.12 Evidence that airway inflammation and cough are related comes from indirect circumstantial evidence demonstrating an increase in the concentration of citric acid causing 2 and 4 coughs 1 month after treatment with an inhaled steroid and salbutamol.20 Careful consideration must be given to the relationships between airflow obstruction and capsaicin-evoked coughs to fully understand the possible mechanisms involved in this study. In the EAR the mean decrease in percent predicted FEV1 after inhaled allergen was 38.2% compared with 4.5% after diluent, with an associated 10 additional capsaicin-evoked coughs (mean total coughs, 12.1 to 21.9 coughs). In a previous study we

J ALLERGY CLIN IMMUNOL nnn 2019

demonstrated that methacholine-induced bronchoconstriction (without inducing inflammation)12 showed that a 19% decrease in FEV1 percentage after methacholine resulted in an increase in coughs by approximately 5 (mean, 8.4 increasing to 13.9). When compared with the current study, a remarkably similar relationship between FEV1 and evoked cough was demonstrated; that is, a near doubling decrease in FEV1 was associated with an almost doubling of capsaicin-evoked coughs. As such, we have now shown that acute bronchoconstriction after both a direct and indirect challenge consistently increases the cough response to capsaicin, implying that cough reflex sensitivity and airway caliber are implicitly linked. It should be noted that the mechanisms underlying this linkage remain unclear; numerous processes have the potential to heighten capsaicin cough responses during bronchoconstriction. In addition to the direct effects of mechanical stimulation and mediator release on nerve function, it is also possible that slowly and rapidly adapting receptor firing during bronchoconstriction indirectly modifies C-fiber function. Bronchoconstriction can also alter exposure of airway nerve terminals to capsaicin as a result of folding and reorientation of the epithelium or even change airway deposition of capsaicin. Unfortunately, the influences of such effects are difficult to predict because little is known about the patterns of innervation of the human airways or indeed the deposition of inhaled capsaicin. Unlike the shorter-acting methacholine challenge, allergen challenge resulted in an LAR in our subjects. This allowed us to investigate the effects of airway inflammation in addition to bronchoconstriction. Although allergen-induced bronchoconstriction improved 24 hours after allergen challenge, the mean FEV1 percentage was still lower than baseline by 8.6%, and hence the excessive coughs demonstrated the day after allergen challenge could be attributed to a combination of both bronchoconstriction and an increase in eosinophilic airway inflammation. This was confirmed in our statistical models when the effects of a decrease in FEV1 and allergen challenge were both independently predictive of capsaicin-evoked cough responses at 24 hours and then also with a decrease in FEV1 and sputum eosinophil percentages. Therefore we could speculate that cough responses at 24 hours are due to the potential sensitizing effects of airway inflammation on airway nerves in addition to a modest degree of bronchoconstriction; the exact mechanism of this effect is unclear and will require further exploration. In particular, more direct evidence could be provided, with bronchial biopsy specimens showing colocation of eosinophils and airway sensory nerves after allergen challenge. There are several possible neuronal mechanisms through which the exaggerated cough responses after inhaled allergen can be explained. Acute bronchoconstriction in the EAR is thought to be elicited by IgE-mediated release of histamine, tryptase, cysteinyl leukotrienes, and prostaglandins.21 The LAR is less well understood, but current evidence suggests that allergen inhalation activates dendritic cells in the airway, which orchestrate a complex type 2 inflammatory response, resulting in stimulation, maturation, trafficking, and degranulation of eosinophils, basophils, and neutrophils to release cysteinyl leukotrienes and histamine.22 These mediators are thought to directly cause smooth muscle contraction, and hence capsaicin-sensitive airway nerves could be sensitized directly because of the mechanical effects of bronchoconstriction or through release of inflammatory mediators in the airways. An alternative explanation would be that airway nerves are not sensitized at all but that the presence of inflammatory mediators as a result of allergen challenge along with

SATIA ET AL 7

J ALLERGY CLIN IMMUNOL VOLUME nnn, NUMBER nn

FIG 6. Changes in percentage sputum eosinophil counts (A) and total cell counts (106 cells per gram; B) before and after allergen and diluent challenge. Bars represent medians.

capsaicin inhalation has a synergistic effect in stimulating airway nerves. Evidence for either mechanism from human studies is currently lacking; however, studies in guinea pigs have implicated a central role for ATP-activating purinergic receptors on nerves,23 mechanical stress stimulating ATP release through TGF-b,3,24 and histamine indirectly stimulating nerves through ATP release from smooth muscle contraction.23 Prostaglandin D2 and prostaglandin E2 have also been shown to cause depolarization in isolated guinea pig and human vagus nerve preparations.25,26 The relative contribution of bronchoconstriction and allergeninduced inflammatory mediators to neuronal hypersensitivity could be addressed in a future study in which subjects were given a bronchodilator at 30 minutes and 24 hours after allergen challenge to negate the effects of bronchoconstriction. Neuronal phenotypic switching is an alternative mechanism potentially underlying our results. Ovalbumin-sensitized and exposed guinea pigs express novel transient receptor potential vanilloid type 1 (TRPV1) receptors on A-d fibers after allergen challenge, making them capsaicin sensitive.27 This switching of nerve fibers to express de novo functional TRPV1 ion channels was mimicked by instilling either brain-derived or glial-derived neurotrophic factor to the trachea. Increased levels of neurotrophins have been detected in serum and airway samples in patients with allergic asthma28-30 and after segmental allergen challenge.31 Hence it is plausible that the increase in coughs seen at 24 hours after allergen was due to addition of a new set of A-d fibers now becoming capsaicin responsive. Animal studies have also implicated an alternative ion channel from the TRP family: transient receptor potential ankyrin-1 (TRPA1). In a series of experiments in rats and mice, Raemdonck et al32 showed the LAR can be attenuated by anesthesia, ruthenium red (a non-selective TRP blocker), a TRPA1 antagonist and anti-cholinergic, but not TRPV1. This suggests that the LAR is driven by TRPA1 found on sensory nerve endings, which synapse in the brainstem, ultimately causing bronchoconstriction through the parasympathetic efferent nerves. However, it is still unclear how allergen-induced inflammation activates the TRPA1 channel, and this is yet to be translated to human subjects. There are limitations to this study. First, our subjects had mild steroid-naive atopic asthma. It is unclear how applicable these results are to a much broader asthmatic population, but this model has commonly been used to study mechanisms underlying asthma.

Second, we have used capsaicin to evoke coughs, which is a specific TRPV1 agonist. It is unclear whether other tussive challenges, such as citric acid, or hypo-osmolar/hyperosmolar challenge agents would produce similar results. Nonetheless, we chose capsaicin as a challenge agent because it is the agent with which we have the most clinical experience, and we have previously demonstrated safety after bronchoconstriction with methacholine. Third, we did not determine capsaicin ED50 at more time points throughout the 24-hour challenge period because in this first study we were uncertain of the safety of multiple capsaicin challenges during allergen exposure. Fourth, our primary end point was to study experimentally evoked cough responses, and hence we did not plan for a sample size powered for changes in spontaneous 24-hour coughing or airway inflammation. In conclusion, we have safely performed a capsaicin cough challenge with allergen exposure in this proof-of-concept mechanistic study in patients with mild atopic asthma. The results of this study suggest that airflow obstruction and airway inflammation both contribute to a heightened cough reflex sensitivity, which is associated with increased 24-hour spontaneous coughing. This implies that cough reflex sensitization is an important dimension of asthma pathophysiology. We thank all the subjects who participated in the study, the National Institute for Health Research (NIHR) Manchester Clinical Research Facility (CRF), and the Cardio-Respiratory Allergen Research Unit at McMaster University Medical Centre.

Key messages d

We safely performed a capsaicin cough challenge during and after allergen challenge and demonstrated heighted responses during the EAR and 24 hours later.

d

Heightened cough response to capsaicin was related to changes in airway caliber and airway eosinophilia and associated with increased spontaneous 24-hour coughing.

REFERENCES 1. Haldar P, Pavord ID, Shaw DE, Berry MA, Thomas M, Brightling CE, et al. Cluster analysis and clinical asthma phenotypes. Am J Respir Crit Care Med 2008;178: 218-24.

8 SATIA ET AL

2. Siroux V, Basagana X, Boudier A, Pin I, Garcia-Aymerich J, Vesin A, et al. Identifying adult asthma phenotypes using a clustering approach. Eur Respir J 2011;38: 310-7. 3. Moore WC, Meyers DA, Wenzel SE, Teague WG, Li H, Li X, et al. Identification of asthma phenotypes using cluster analysis in the severe asthma research program. Am J Respir Crit Care Med 2010;181:315-23. 4. Bleecker ER, FitzGerald JM, Chanez P, Papi A, Weinstein SF, Barker P, et al. Efficacy and safety of benralizumab for patients with severe asthma uncontrolled with highdosage inhaled corticosteroids and long-acting b2-agonists (SIROCCO): a randomised, multicentre, placebo-controlled phase 3 trial. Lancet 2016;388:2115-27. 5. Wenzel S, Castro M, Corren J, Maspero J, Wang L, Zhang B, et al. Dupilumab efficacy and safety in adults with uncontrolled persistent asthma despite use of medium-to-high-dose inhaled corticosteroids plus a long-acting b2 agonist: a randomised double-blind placebo-controlled pivotal phase 2b dose-ranging trial. Lancet 2016;388:31-44. 6. Ortega HG, Liu MC, Pavord ID, Brusselle GG, FitzGerald JM, Chetta A, et al. Mepolizumab treatment in patients with severe eosinophilic asthma. N Engl J Med 2014;371:1198-207. 7. Corren J, Parnes JR, Wang L, Mo M, Roseti SL, Griffiths JM, et al. Tezepelumab in adults with uncontrolled asthma. N Engl J Med 2017;377:936-46. 8. Manfreda J, Becklake MR, Sears MR, Chan-Yeung M, Dimich-Ward H, Siersted HC, et al. Prevalence of asthma symptoms among adults aged 20-44 years in Canada. CMAJ 2001;164:995-1001. 9. Mincheva R, Ekerljung L, Bjerg A, Axelsson M, Popov TA, Lundb€ack B, et al. Frequent cough in unsatisfactory controlled asthma—results from the population-based West Sweden Asthma Study. Respir Res 2014;15. 10. Osman LM, McKenzie L, Cairns J, Friend JAR, Godden DJ, Legge JS, et al. Patient weighting of importance of asthma symptoms. Thorax 2001;56:138-42. 11. Satia I, Tsamandouras N, Holt K, Badri H, Woodhead M, Ogungbenro K, et al. Capsaicin-evoked cough responses in asthmatic patients: evidence for airway neuronal dysfunction. J Allergy Clin Immunol 2017;139:771-9.e10. 12. Satia I, Badri H, Woodhead M, O’Byrne PM, Fowler SJ, Smith JA. The interaction between bronchoconstriction and cough in asthma. Thorax 2017;72:1144-6. 13. Marsden PA, Satia I, Ibrahim B, Woodcock A, Yates L, Donnelly I, et al. Objective cough frequency, airway inflammation, and disease control in asthma. Chest 2016; 149:1460-6. 14. Cockcroft DW, Davis BE, Boulet LP, Deschesnes F, Gauvreau GM, O’Byrne PM, et al. The links between allergen skin test sensitivity, airway responsiveness and airway response to allergen. Allergy 2005;60:56-9. 15. Killian D, Cockcroft DW, Hargreave FE, Dolovich J. Factors in allergen-induced asthma: relevance of the intensity of the airways allergic reaction and non-specific bronchial reactivity. Clin Exp Allergy 1976;6:219-25. 16. Crapo RO, Casaburi R, Coates AL, Enright PL, Hankinson JL, Irvin CG, et al. Guidelines for methacholine and exercise challenge testing-1999. This official statement of the American Thoracic Society was adopted by the ATS Board of Directors, July 1999. Am J Respir Crit Care Med 2000;161:309-29.

J ALLERGY CLIN IMMUNOL nnn 2019

17. Diamant Z, Gauvreau GM, Cockcroft DW, Boulet LP, Sterk PJ, De Jongh FHC, et al. Inhaled allergen bronchoprovocation tests. J Allergy Clin Immunol 2013; 132:1045-55.e6. 18. Minogushi H, Minogushi K, Tanaka A, Matsuo H, Kihara N, Adachi M. Cough receptor sensitivity to capsaicin does not change after allergen bronchoprovocation in allergic asthma. Thorax 2003;58:19-22. 19. Hilton ECY, Baverel PG, Woodcock A, Van Der Graaf PH, Smith JA. Pharmacodynamic modeling of cough responses to capsaicin inhalation calls into question the utility of the C5 end point. J Allergy Clin Immunol 2013;132:847-55, e1-5. 20. Di Franco A, Dente FL, Giannini D, Vagaggini B, Conti I, Macchioni P, et al. Effects of inhaled corticosteroids on cough threshold in patients with bronchial asthma. Pulm Pharmacol Ther 2001;14:35-40. 21. Cockcroft DW, Hargreave FE, O’Byrne PM, Boulet L-P. Understanding allergic asthma from allergen inhalation tests. Can Respir J 2007;14:414-8. 22. Gauvreau GM, El-Gammal AI, O’Byrne PM. Allergen-induced airway responses. Eur Respir J 2015;46:819-31. 23. Weigand LA, Ford AP, Undem BJ. A role for ATP in bronchoconstriction-induced activation of guinea pig vagal intrapulmonary C-fibres. J Physiol 2012;590: 4109-20. 24. Gonzalez EJ, Heppner TJ, Nelson MT, Vizzard MA. Purinergic signalling underlies transforming growth factor-b-mediated bladder afferent nerve hyperexcitability. J Physiol 2016;594:3575-88. 25. Maher SA, Birrell MA, Adcock JJ, Wortley MA, Dubuis ED, Bonvini SJ, et al. Prostaglandin D2 and the role of the DP1, DP2 and TP receptors in the control of airway reflex events. Eur Respir J 2015;45:1108-18. 26. Maher SA, Birrell MA, Belvisi MG. Prostaglandin E2 mediates cough via the EP3 receptor: implications for future disease therapy. Am J Respir Crit Care Med 2009; 180:923-8. 27. Lieu TM, Myers AC, Meeker S, Undem BJ. TRPV1 induction in airway vagal lowthreshold mechanosensory neurons by allergen challenge and neurotrophic factors. Am J Physiol Lung Cell Mol Physiol 2012;302:L941-8. 28. Nassenstein C, Braun A, Erpenbeck VJ, Lommatzsch M, Schmidt S, Krug N, et al. The neurotrophins nerve growth factor, brain-derived neurotrophic factor, neurotrophin-3, and neurotrophin-4 are survival and activation factors for eosinophils in patients with allergic bronchial asthma. J Exp Med 2003;198:455-67. 29. Bonini S, Lambiase A, Bonini S, Angelucci F, Magrini L, Manni L, et al. Circulating nerve growth factor levels are increased in humans with allergic diseases and asthma. Proc Natl Acad Sci U S A 1996;93:10955-60. 30. Watanabe T, Fajt ML, Trudeau JB, Voraphani N, Hu H, Zhou X, et al. Brainderived neurotrophic factor expression in asthma. Association with severity and type 2 inflammatory processes. Am J Respir Cell Mol Biol 2015;53:844-52. 31. Virchow JC, Julius P, Lommatzsch M, Luttmann W, Renz H, Braun A. Neurotrophins are increased in bronchoalveolar lavage fluid after segmental allergen provocation. Am J Respir Crit Care Med 1998;158:2002-5. 32. Raemdonck K, De Alba J, Birrell MA, Grace M, Maher SA, Irvin CG, et al. A role for sensory nerves in the late asthmatic response. Thorax 2012;67:19-25.

J ALLERGY CLIN IMMUNOL VOLUME nnn, NUMBER nn

METHODS Study procedures in detail Allergen challenge. The concentration of allergen causing a 20% decrease in FEV1 is predicted by using the methacholine PC20 value, and the titration of allergen was determined based on results of skin prick tests. The starting dose of allergen was determined by using the Cockcroft formula, and allergen challenge was performed, as previously described.E1 Briefly, subjects were asked to inhale a low concentration of an allergen for 2 minutes through a nebulizer. Lung function was measured 10 minutes after the end of the 2 minutes of inhalation. If FEV1 decreased by less than 10% from the highest baseline FEV1, the next concentration was given. If the decrease was between 10% to 20%, the FEV1 was measured again after a further 10 minutes. If it was the same or had started to increase, the next concentration or half concentration was given. Inhalations were stopped when FEV1 had decreased by 20% or more or at the discretion of the investigator. After the last inhalation, FEV1 was measured at 20, 30, 45, 60, 90, and 120 minutes and then at hourly intervals until 7 hours after the last allergen inhalation. The EAR was defined as a 20% or greater decrease in FEV1 between 0 and 3 hours after allergen, and the LAR was defined as a 15% or greater decrease in FEV1 between 3 and 7 hours after allergen inhalation. Full dose-response capsaicin challenge testing. A full dose-response capsaicin challenge was performed at visit 1.E2 Patients inhaled increasing doubling concentrations (0.49-1000 mmol/L) of capsaicin through a nebulizer adapted to control dosage and inspiratory flow rate. Each dose of capsaicin was inhaled in 4 single breaths separated by 30 seconds. After each inhalation, the total number of coughs evoked (defined as explosive cough sounds) at each concentration within the first 15 seconds was recorded. The total maximum cough response evoked by any concentration of capsaicin is documented along with the ED50. The ED50 was used throughout visits 4 to 9. The lapel microphone of the VitaloJAK cough recorder was attached to the patient throughout the duration of the challenge, and an independent observer verified the number of coughs induced at each dose of capsaicin. Single-dose capsaicin challenge. The ED50 during the full challenge at visit 1 was used as a single-dose challenge in visits 5, 6, 8, and 9. Four inhalations were administered 30 seconds apart, and the number of coughs evoked in the first 15 seconds was counted and verified manually. Cough monitoring. Participants were asked to wear a cough monitor (VitaloJAK) for 24 hours from the start of the allergen/diluent challenge (visits 5 and 8) but also for a short duration while 4 inhalations of capsaicin were being performed (visits 6 and 9). This involved attaching an air microphone to the patient’s lapel, which recorded all the evoked coughs. The verification process was done after the end of the visits. Twenty-four-hour files were compressed into files of shorter duration by using custom-built software, and individual coughs were tagged and counted by using an audio editing package (Audition, version 3; Adobe Systems, San Jose, Calif). The number of coughs were expressed as coughs per hour.

SATIA ET AL 8.e1

Methacholine challenge test. Nebulized methacholine was administered in doubling concentrations from 0.0625 up to 16 mg/mL through the 2-minute tidal breathing method.E3 At each concentration, FEV1 was measured 30 and 90 seconds after the end of each inhalation. The percentage decrease in FEV1 from baseline was calculated after each dose by using the best FEV1, and the test was terminated once a 20% decrease in FEV1 was documented or the maximum concentration of 16 mg/mL was reached. Salbutamol was withheld for at least 4 hours. Sputum induction and processing. Before performing sputum induction, spirometry was repeated to ensure FEV1 had returned to within 10% of the baseline value before the methacholine challenge test. This occasionally required 400 mg of salbutamol in addition to the post– methacholine challenge test salbutamol. Nebulized hypertonic saline (5%) was administered at 5-minute intervals provided the FEV1 did not decrease by more than 20% from the presputum baseline value. If there was a decrease in FEV1 of between 10% and 20%, then a lower saline concentration was administered. Sputum was processed, as described previously.E4 Total cell counts were determined by using a Neubauer hemocytometer chamber (Hausser Scientific, Blue Bell, Pa) and expressed as the number of cells per gram of sputum. Two slides were stained with Diff Quik (American Scientific Products, McGaw Park, Ill), and the differential cell counts (400 cells per slide) of the 2 slides were averaged. Differential cell counts were expressed as percentages of total cell counts (400 nonsquamous cells). Skin prick testing and dose titration. A panel of allergen extracts was applied with a single-head metal lancet on the patient’s forearm. A negative control and a positive histamine control were also assessed. A positive result was demonstrated by a wheal of 2 mm or larger. The allergen with the largest wheal was used to titrate down to much lower concentrations, and the concentration of allergen that caused a skin wheal of at least 2 3 2 mm in size was noted. This was used to calculate the starting dose of the allergen challenge described previously.E1 REFERENCES E1. Diamant Z, Gauvreau GM, Cockcroft DW, Boulet LP, Sterk PJ, de Jongh FHC, et al. Inhaled allergen bronchoprovocation tests. J Allergy Clin Immunol 2013; 132:1045-55.e6. E2. Satia I, Tsamandouras N, Holt K, Badri H, Woodhead M, Ogungbenro K, et al. Capsaicin cough responses in asthma: evidence for airway neuronal dysfunction. J Allergy Clin Immunol 2017;139:771-9. E3. Crapo RO, Casaburi R, Coates AL, Enright PL, Hankinson JL, Irvin CG, et al. Guidelines for methacholine and exercise challenge testing-1999. This official statement of the American Thoracic Society was adopted by the ATS Board of Directors, July 1999. Am J Respir Crit Care Med 2000;161:309-29. E4. Pizzichini E, Pizzichini MMM, Efthimiadis A, Hargreave FE, Dolovich J. Measurement of inflammatory indices in induced sputum: effects of selection of sputum to minimize salivary contamination. Eur Respir J 1996;9: 1174-80.