Coughing During Mannitol Challenge Is Associated With Asthma*

Transcription

1 clinical investigations Coughing During Mannitol Challenge Is Associated With Asthma* Heikki O. Koskela, MD; Liisa Hyvärinen, MD; John D. Brannan, PhD; Hak-Kim Chan, PhD; and Sandra D. Anderson, PhD, DSc Study objectives: To define whether coughing during mannitol challenge is a nonspecific side effect of this challenge or is associated with asthma. Design: A prospective study. Setting: University hospital. Participants: Thirty-seven steroid-naive, asthmatic subjects and 10 healthy subjects. Measurements: The participants completed a symptom questionnaire, recorded peak expiratory flows (PEFs), and underwent spirometry, skin tests, and bronchial provocations with mannitol, histamine, and cold air. Seventeen of the asthmatic subjects were treated with budesonide, 800 g per day, and the measurements were repeated after 3 and 6 months of treatment. Coughs were recorded during the mannitol challenges, and the cough sensitivity was expressed as the cumulative number of coughs divided by the cumulative dose of mannitol. Results: The asthmatic subjects coughed more during the mannitol challenge than the healthy subjects (8.3 coughs per 100 mg [95% confidence interval (CI), 6.2 to 11.0] vs 1.1 coughs per 100 mg [95% CI, 0.4 to 3.0]; p < ). Even those asthmatic subjects who did not develop bronchoconstriction after the maximal cumulative dose of mannitol (635 mg) coughed significantly more than the healthy subjects (53 coughs [95% CI, 34 to 72] vs 12 coughs [95% CI, 4 to 21]; p 0.003). Budesonide treatment decreased the cough sensitivity (p 0.023), which was significantly associated with improvements in overall symptom frequency, cough frequency, diurnal PEF variation, FEV 1, and bronchial hyperresponsiveness. Conclusions: Coughing during mannitol challenge is associated with asthma and occurs independently of bronchoconstriction. It can be used to study the mechanisms of asthmatic cough. Furthermore, the measurement of the mannitol-provoked coughing may be useful both in the diagnosis of asthma as well as in the assessment of the effects of an anti-inflammatory therapy on this common disorder. (CHEST 2004; 125: ) Key words: asthma; bronchial hyperresponsiveness; bronchial provocation; cough; cough reflex; mannitol challenge Abbreviations: CDR coughs-to-dose ratio; CDRp partial coughs-to-dose-ratio; CI confidence interval; HETE hydroxyeicosatetraenoic acid; PEF peak expiratory flow; RAR rapidly adapting receptor; RDR response-to-dose ratio Bronchial provocation using the dry powder of mannitol is a novel test for demonstrating airway hyperresponsiveness. 1 Mannitol, 2 in keeping with *From the Department of Respiratory Medicine (Drs. Koskela and Hyvärinen), Kuopio University Hospital, Kuopio, Finland; the Department of Respiratory Medicine (Drs. Brannan and Anderson), Royal Prince Alfred Hospital, Camperdown, NSW, Australia; and the Department of Pharmacy (Dr. Chan), University of Sydney, NSW, Australia. This study was supported by Kuopio University Hospital. Manuscript received August 14, 2003; revision accepted November 25, other non iso-osmolar substances, 3 7 can provoke cough when inhaled. The mechanisms of the mannitol-provoked cough are not known. It might simply be due to the nonrespirable, large particles of mannitol powder, which impact in the upper airways during inhalation, causing nonspecific mechanical Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( permissions@chestnet.org). Correspondence to: Heikki Koskela, MD, Department of Respiratory Medicine, Kuopio University Hospital, PL 1777, Kuopio, Finland; heikki.koskela@kuh.fi CHEST / 125 / 6/ JUNE,

2 irritation of the laryngeal rapidly adapting receptors (RAR). 2,8 If this were the case, the mannitolprovoked cough might be minimized by changing the preparation of dry powder and the device used for delivery, thus minimizing the impaction of the particles in the upper airways. Another possibility is that the mannitol-provoked cough is related to the increased sensitivity of the asthmatic airways to various osmotic stimuli. 9 If so, it could be due to at least two mechanisms. First, mannitol-provoked cough might be secondary to mannitol-induced bronchoconstriction, since tracheobronchial RARs can be stimulated by bronchoconstriction. 8 Second, mannitol could stimulate RARs or other airway receptors in the lower airways directly, and independently of bronchoconstriction. If the mannitol-provoked cough is associated with asthma, it would provide a new, noninvasive index to study the mechanisms of asthma. The present study was conducted to find out whether the mannitol-provoked cough is associated with asthma and, if so, whether it is secondary to mannitol-induced bronchoconstriction or whether it occurs independently. Materials and Methods The material for the present study is obtained from two prospective clinical studies that have been carried out in our institute. The first study 10 was carried out to compare mannitol challenge with cold air and histamine challenges to demonstrate airway hyperresponsiveness in subjects with steroidnaive, difficult-to-diagnose asthma. The second study 11 was carried out to investigate the effect of inhaled budesonide on the responsiveness to those three challenges in a subgroup of the asthmatic subjects who participated in the first study. To explore the significance of mannitol-provoked cough, all coughs were manually recorded during every mannitol challenge in the studies. Subjects We recruited consecutive subjects with a new diagnosis of asthma in our outpatient clinic over a period of 18 months. The diagnosis of asthma was based on patient history and clinical examination findings, in addition to objective evidence of reversible airway obstruction, according to the Finnish Social Insurance Institute criteria. 12 At least one of the following criteria had to be fulfilled for the diagnosis of asthma: (1) at least a 15% fall in FEV 1 after exercise challenge; (2) at least a 15% improvement in FEV 1 after inhaling a bronchodilating drug in spirometry; (3) at least a 20% spontaneous diurnal peak expiratory flow (PEF) variation in home peak flow monitoring on at least 2 days; and (4) at least a 15% improvement in PEF after inhaling a bronchodilating drug during home peak flow monitoring on at least 2 days. The exclusion criteria were previous use of inhaled or oral corticosteroids, febrile respiratory tract infection within 4 weeks, and FEV 1 of 50% predicted. 13 Subjects were also excluded if the staff physician considered COPD to be the most probable diagnosis, even if the reversibility criteria were fulfilled. Forty-two asthmatic subjects were included. Five subjects discontinued the study, three due to personal reasons not related to the study, one due to unstable asthma, and one due to acute sinusitis. Of the 37 asthmatic subjects who completed the study, criterion 1 was fulfilled in none of the subjects (exercise challenge is not included in the routine clinical evaluation in the authors hospital), criterion 2 was fulfilled in four subjects, criterion 3 was fulfilled in 12 subjects, and criterion 4 was fulfilled in 28 subjects. In many subjects, more than one criterion was fulfilled. The subjects had refrained from taking short-acting 2 agonists for 6 h before the challenges. In addition, 10 healthy volunteers were recruited. They had no chronic respiratory diseases or symptoms and were lifelong nonsmokers. These 47 subjects formed the group that completed the first study, and their basic characteristics are given in Table 1. Nineteen asthmatic subjects who were found to be hyperresponsive to both mannitol and histamine in the first study were recruited to the second study assessing the effects of budesonide treatment. One patient discontinued the treatment within 1 month due to hoarseness, and another patient discontinued treatment without giving a reason. The remaining 17 asthmatic subjects formed the group that participated in the second study, and their basic characteristics are given in Table 2. The Finnish National Agency of Medicines and the institutional ethics committee approved the studies, and all subjects gave their informed written consent for participation. Protocol In the first study, the subjects completed a symptom questionnaire, and the information of the diagnostic home PEF monitoring was recorded. Spirometry (model M9449; Medikro Ltd; Kuopio, Finland) was carried out before and 15 min after inhalation of 0.4 mg salbutamol (Ventoline Evohaler; GlaxoSmithKline Ltd; Uxbridge, UK) according to the American Thoracic Society guidelines. 14 Skin prick tests were carried out against 16 common aeroallergens (Soluprick SQ; Table 1 Summary Characteristics of the 47 Subjects Who Participated in the First Study* Characteristics Asthmatic Patients Healthy Subjects (n 37) (n 10) Age, yr 49 (44 54) 41 (34 48) Gender Female 16 5 Male 21 5 Current smokers, No. 6 0 Atopic subjects, No Duration of asthmatic 24 (2 396) symptoms, mo Distribution of asthma severity according to Global Initiative for Asthma classification Intermittent 1 Mild persistent 7 Moderate persistent 29 FEV 1, % predicted 81 (76 86) 99 (95 104) Improvement in FEV 1 after 0.4 mg salbutamol, % 6.6 ( ) 2.3 ( ) *Values given as mean (95% CI), unless otherwise indicated. Values given as median (minimum maximum). Predicted values are those of Viljanen et al clinical investigations

3 Table 2 Basic Characteristics of the 17 Asthmatic Patients Who Participated the Second Study* ALK-Abelló; Hörsholm, Denmark). Atopy was defined as the appearance of a wheal reaction of at least 3 mm to any of the allergens. After these assessments, mannitol, cold air, and histamine challenges were carried out in random order, on separate days. The cold air challenge consisted of a single 4-min period of frigid air hyperpnea. 15 The histamine aerosol was delivered by a dosimetric nebulizer in progressively increasing doses. 16 In the second study, the measurements were repeated after 3 and 6 months of treatment with inhaled budesonide (Pulmicort Turbuhaler, 400 g per dose; Astra- Zeneca Ltd; Lund, Sweden), one dose twice a day. The investigators asked about compliance with treatment, and the inhalation technique was checked during each visit. Questionnaire In a self-administered questionnaire, questions were asked about the frequency of dyspnea, cough, wheezing, and sputum production during the last month. The possible answers for each symptom were as follows: less than once per month (1 point on the sum symptom score); one to five times per month (2 points); about once per week (3 points); several times per week (4 points); and daily (5 points). The sum symptom score was calculated as the sum of points for the four symptoms asked about. The daily use of bronchodilating drugs during the last month was defined, as well as the smoking habits. On the basis of symptom frequency and lung function information before treatment, the subject s asthma severity was determined using the Global Initiative for Asthma classification. 17 Home PEF Monitoring Characteristics Values Age, yr 51 (43 58) Gender distribution Male 12 Female 5 Current smokers, No. 3 Atopic subjects, No. 7 Duration of asthmatic symptoms before start of 24 ( ) treatment, mo Distribution of asthma severity according to Global Initiative for Asthma classification Mild persistent asthma 2 patients Moderate persistent asthma 15 patients FEV 1, % predicted 77 (68 85) Improvement in FEV 1 after 0.4 mg salbutamol, % 6.8 ( ) *Values given as mean and (95% CI), unless otherwise indicated. Values given as median (minimum maximum). Predicted values are those of Viljanen et al. 13 The asthmatic subjects documented three PEF values in the morning and in the evening for 2 weeks (Mini Wright; Clement Clarke International; Harlow, UK), and the best of three values was used in the analysis. The variation for each day was calculated by dividing the difference between the evening and the morning PEF values by the mean of these values, expressed in percentages. 18 The mean of these values for 14 days was calculated. In four subjects, the baseline monitoring was not technically satisfactory, and these recordings were omitted. Mannitol Challenge The mannitol powder was prepared by spray drying, and 57% of the emitted particles were 7 m in size when measured by dispersion (Multi-Stage Liquid Impinger; Astra- Zeneca Ltd) at 60 L/min using the same device (Inhalator; Boehringer Ingelheim Pty LTD; Ingelheim, Germany) that was used to deliver the mannitol to the subjects. The mannitol powder was weighed (5, 10, 20, and 40 mg) into gelatin capsules and was inhaled in doubling doses (0, 5, 10, 20, 40, 80, and 160 mg). The 80-mg and 160-mg doses were given in multiples of 40-mg capsules, and the 160-mg dose was delivered three times. 1 After every dose, the coughs were manually recorded. The test continued until the FEV 1 had fallen 15% or until the maximal cumulative dose of 635 mg had been administered. The response-to-dose ratio (RDR) was calculated as the percentage fall in FEV 1 after the last dose, divided by the cumulative dose of mannitol. The coughs-to-dose ratio (CDR) was calculated as the cumulative number of coughs from the inhalation of the first capsule to the last measurement of FEV 1, divided by the cumulative dose of mannitol, in milligrams. To get more easily understandable values, this value then was multiplied by 100 and, thus, represents the number of coughs per 100 mg mannitol administration. The partial CDR (CDRp) was similarly calculated but using the cumulative number of coughs from the inhalation of the first capsule up to the inhalation of the dose causing the FEV 1 to fall permanently 5% from baseline (the coughs induced by that dose were not included). Statistical Analysis The difference in the distribution of the cough indexes from normal distribution was studied using one-sample Kolmogorov- Smirnov test. The distribution of the absolute number of coughs per challenge was sufficiently close to normal in the asthmatic subjects (p 0.34). On the contrary, the CDR and CDRp values were not normally distributed in them (p and 0.004, respectively). Therefore, log-transformed CDR and CDRp values (p 0.75 and p 0.69, respectively) were used in the statistical analysis. If no coughs had occurred, one cough was added to be able to perform the log transformation. The RDR values were also log-transformed. The results are given as the mean and 95% confidence interval (CI), except for CDR, CDRp, and RDR values, which are geometric means and 95% CIs. The difference of the responses between the groups was analyzed using the Student unpaired t test. When small groups of subjects were compared, the nonparametric Mann-Whitney test also was used. The Pearson correlation coefficient was used in the correlation analysis. The association of various subject characteristics with the CDR and CDRp was studied by linear regression analysis. Repeated-measures analysis of variance was used to analyze the effects of budesonide on the various indexes. A p value of 0.05 was considered to be statistically significant. All analyses were carried out using a statistical software package (SPSS for Windows, version ; SPSS; Chicago, IL). Results In the asthmatic subjects, the number of coughs provoked by mannitol was proportional to the actual dose of mannitol inhaled, and the higher the dose, the greater the number of coughs, without signs of refractoriness (Fig 1). By contrast to the CHEST / 125 / 6/ JUNE,

4 Figure 1. Relation of the noncumulative doses of mannitol to the mean noncumulative number of coughs provoked by the doses in 10 healthy subjects and 37 asthmatic patients. All healthy subjects tolerated the maximal dose of mannitol, but the number of asthmatic patients diminished with the higher mannitol doses. Error bars are omitted for clarity. asthmatic subjects, mannitol provoked only infrequent coughing in the healthy subjects and the frequency was unrelated to the dose administered. Thus, the values for CDR and CDRp were significantly greater in the 37 steroid-naive asthmatic subjects compared with those for the 10 healthy subjects (CDR: 8.3 coughs per 100 mg [95% CI, 6.2 to 11.0] vs 1.1 coughs per 100 mg [95% CI, 0.4 to 3.0]; p ; CDRp: 11.2 coughs per 100 mg [95% CI, 7.3 to 17.1] vs 1.1 coughs per 100 mg [95% CI, 0.4 to 3.0]; p ) [Fig 2]. Of note in the asthmatic subjects, 60% of the total number of coughs occurred before the dose of mannitol that induced a 5% fall in FEV 1 was administered. At that stage, 57% (mean, 272 of 474 mg) of the total dose of mannitol had been inhaled. There were 18 steroid-naive asthmatic subjects and all 10 healthy subjects in whom the change in FEV 1 was 15% from baseline after the maximal cumulative dose of mannitol. Within this group of 28, there were significantly more coughs in the asthmatic subjects compared with the healthy subjects (53 coughs [95% CI, 34 to 72] vs 12 coughs [95% CI, 4 to 21], respectively; p 0.003) [Fig 3]. In 8 of the 18 asthmatic subjects, the final fall in FEV 1 was 4.5%, a value similar to the greatest fall documented in a healthy individual. Even in these eight asthmatic subjects, mannitol provoked significantly more coughs than in the healthy subjects (52 coughs [95% CI, 20 to 83] vs 12 coughs [95% CI, 4 to 21], respectively; p [Student unpaired t test] or p [Mann-Whitney test]). The cough indexes were not associated with age, sex, atopy, dermatographism, current smoking status, or smoking history. Table 3 shows the correlation of the CDR and CDRp values with each other, and with airway responsiveness to mannitol, cold air, and histamine. The treatment of the asthmatic subjects with inhaled budesonide decreased their CDR and CDRp values, which approached those in the healthy subjects after 6 months of therapy (CDR, 3.0 coughs per 100 mg [95% CI, 1.4 to 6.5]; CDRp, 4.6 coughs per 100 mg [95% CI, 1.7 to 12.4]) [Fig 4]. Budesonide also significantly decreased the sum symptom score, cough frequency, daily use of bronchodilating drugs, diurnal PEF variation, and mannitol, cold air, and histamine airway responsiveness, as previously reported. 11 The magnitude of the budesonideinduced change in the cough indexes was highly significantly associated with improvements in sum symptom score, cough frequency, diurnal PEF variation, FEV 1, and bronchial hyperresponsiveness (Table 4). Figure 2. CDR values (top) and CRPp values (bottom) in 10 healthy subjects and 37 asthmatic patients (p between the groups) clinical investigations

5 Figure 3. Number of coughs during the entire mannitol challenge in the subjects with a 15% fall in FEV 1 after the maximal dose of mannitol was administered (p between the groups). Discussion This study shows that coughing during mannitol challenge is not just a nonspecific side effect of the challenge but is related to the increased sensitivity of the asthmatic airways to various osmotic stimuli. The following findings of the present study support this view: first, there were statistically highly significant differences in cough indexes between asthmatic and healthy subjects; second, there were significant associations between the cough indexes and airway responsiveness to mannitol and cold air; third, the cough indexes decreased during treatment with the inhaled budesonide; and fourth, there were highly significant associations between the budesonideinduced decrease in cough indexes and the budesonide-induced changes in several indexes of asthma severity. This study also shows that mannitol-provoked cough in asthmatic subjects is not simply a manifestation of mannitol-induced bronchoconstriction, as documented by changes in FEV 1. This was demonstrated by the following findings. First, in asthmatic subjects, 60% of all coughs had occurred before inhalation of the dose causing a 5% fall in FEV 1. Second, asthmatic subjects coughed more than the healthy subjects before any significant bronchoconstriction was documented. Third, even those asthmatic subjects who had 4.5% fall in FEV 1 in response to the maximal dose of mannitol coughed significantly more than the healthy subjects. The difference in cough sensitivity between healthy and asthmatic subjects may be more than quantitative. The number of coughs increased with each doubling dose of mannitol only in the asthmatic subjects, whereas the infrequent coughing in the healthy subjects was unrelated to the dose. This finding suggests that coughing is not a normal response to a rise in airway lining fluid osmolarity but is a pathologic response that is associated with asthma. This view is supported by the observations that spontaneous cough is usually not provoked by the inhalation of hypertonic saline solution in patients with chronic bronchitis or cystic fibrosis, although it increases the mucociliary clearance in those patients. 19 The coughing of the healthy subjects may be due to a nonspecific irritation of the upper airways by the impaction of the powder or to some other phenomena unrelated to airway lining fluid osmolarity. It is generally accepted that laryngeal and tracheobronchial RARs are the afferent nerves that are most probably associated with the cough reflex in healthy subjects. 8,20 There is evidence showing that airway lining fluid hyperosmolarity can fire RARs. 21 Hyperosmolarity also could induce airway mucosal mast cells to release histamine, a stimulant of lung RARs. 8,22 Hyperosmolarity also can fire the unmyelinated, capsaicin-sensitive bronchial C-fibers. 21 Table 3 Pearson Correlation Coefficients Between CDR and CDRp Values, and Between Them and Indices of Airway Hyperresponsiveness in 37 Steroid- Naive Subjects With Asthma CDR CDRp Variables r Value p Value r Value p Value CDRp Mannitol RDR Cold air-induced change in FEV 1 Histamine RDR Figure 4. The change in CDR and CRPp during treatment with inhaled budesonide in 17 subjects with asthma (p for the change in CDR; p 0.05 for the change in CDRp). CHEST / 125 / 6/ JUNE,

6 Table 4 Pearson Correlation Coefficients Between the Budesonide-Induced Changes in CDR and CDRp Values and Budesonide-Induced Changes in Several Indices of Asthma Severity* CDR CDRp Variables r Value p Value r Value p Value Sum symptom score 0.64 ( ) 0.47 (0.005) Cough frequency 0.48 ( 0.004) 0.41 (0.015) Daily doses of bronchodilating drugs 0.27 (0.12) 0.15 (0.40) Mean diurnal PEF variation 0.78 ( ) 0.44 (0.025) FEV 1 % of predicted 0.47 (0.005) 0.25 (0.15) Mannitol RDR 0.58 ( ) 0.59 ( ) Cold air-induced change in FEV (0.002) 0.37 (0.033) Histamine RDR 0.41 (0.015) 0.30 (0.082) *In each correlation analysis, 34 pairs of values were used, representing 17 asthmatic subjects studied at two time points after the start of treatment with inhaled budesonide. However, in the analysis of mean diurnal PEF variation, only 26 pairs of values were used. Information is obtained from the symptom questionnaire. However, the role of airway C-fibers in initiating cough is uncertain. 23 C-fibers may participate in the cough reflex by releasing sensory neuropeptides, which in turn stimulate RARs. 8 We have previously shown that nasal challenge with mannitol is associated with a rapid release of 15-hydroxyeicosatetraenoic acid (HETE) to nasal lavage fluid in subjects with allergic rhinitis, but not in healthy subjects HETE is the major arachidonic acid metabolite in human bronchi, and its source is probably the airway epithelial cell. 25 Interestingly, 15-HETE is one of the most potent activators of C-fiber capsaicin (vanilloid) receptors. 26 Thus, mannitol could stimulate the airway epithelial cells of asthmatic subjects to release 15-HETE, which, in turn, could stimulate airway C-fibers, resulting in coughing. This theory could explain the great difference in the severity of mannitol-provoked coughing between healthy and asthmatic subjects. Previous studies using capsaicin, citric acid, or tartaric acid to document the cough threshold have consistently failed to show statistically significant differences between asthmatic and healthy subjects, and this has limited the use of these tests in asthma research. On the contrary, mannitol provoked significantly more coughs in the asthmatic patients than in the healthy subjects. Therefore, asthma-associated cough may be stimulus-specific and should probably be studied by using mannitol rather than by the substances used previously. The significance of coughing during histamine and cold air challenges remains obscure because we did not record the coughs during these challenges. The present study can give some clinically useful information. First, if a subject with a suspicion of asthma does not develop an abnormal fall in FEV 1 during the mannitol challenge, the challenge result is usually considered to be negative. 1 However, if such a subject coughs frequently (ie, 35 coughs approximately) during the challenge, a clinician is given a hint that the subject might still be experiencing an asthma-like airway pathology. It is clear from the literature that there are patients with eosinophilic bronchial inflammation and respiratory symptoms who have normal lung function and no significant airway responsiveness Mannitol challenge combined with cough recordings would appear to be an attractive alternative to investigate such patients. Second, the cough indexes proved to be valid in monitoring the effect of treatment with inhaled corticosteroids. Validity in this study represents convergence validity (ie, the relationship of the instrument to other instruments that measure the same thing). 37 During the budesonide treatment, the CDR values decreased most in subjects who also showed the greatest improvements in sum symptom score, cough frequency, diurnal PEF variation, FEV 1, and bronchial hyperresponsiveness. In other words, a decrease in CDR in an individual subject during treatment with inhaled corticosteroid probably indicates a healing of asthma and not just the presence of therapy. Therefore, cough recordings may increase the usefulness of mannitol challenge in monitoring the effect of antiasthma therapy. 38 CDR was superior to CDRp in this respect and may thus be the index of choice to be used in the future. Although the magnitude of the budesonideinduced change in CDR was highly significantly associated with the improvement in the sum symptom score, it should be remembered that a selfadministered symptom questionnaire has limitations. Significant proportions of asthmatic subjects tend to underestimate or overestimate their symptoms compared with objective measurements of lung func clinical investigations

7 tion. 39 Therefore, the use of a self-administered questionnaire may be a potential weakness in the present study. In conclusion, this study shows that coughing during mannitol challenge occurs independently of bronchoconstriction and is associated with asthma. This phenomenon may add to our understanding about the mechanisms of asthmatic cough. Furthermore, the measurement of the mannitol-provoked coughing may be useful both in the diagnosis of asthma as well as in the assessment of the effects of an anti-inflammatory therapy on patients with this common disorder. ACKNOWLEDGMENT: The authors thank Pirjo Vänttinen, RN, for her assistance. References 1 Anderson SD, Brannan J, Spring J, et al. A new method for bronchial-provocation testing in asthmatic subjects using a dry powder of mannitol. Am J Respir Crit Care Med 1997; 156: Brannan JD, Koskela H, Anderson SD, et al. Responsiveness to mannitol in asthmatic subjects with exercise- and hyperventilation-induced asthma. Am J Respir Crit Care Med 1998; 158: Sheppard D, Rizk NW, Boushey HA, et al. Mechanism of cough and bronchoconstriction induced by distilled water aerosol. Am Rev Respir Dis 1983; 127: Boulet LP, Legris C, Thibault L, et al. Comparative bronchial responses to hyperosmolar saline and methacholine in asthma. Thorax 1987; 42: Lowry RH, Wood AM, Higenbottam TW. Effects of ph and osmolarity on aerosol-induced cough in normal volunteers. Clin Sci (Lond) 1988; 74: Fahy JV, Wong HH, Geppetti P, et al. Effect of an NK1 receptor antagonist (CP-99, 994) on hypertonic salineinduced bronchoconstriction and cough in male asthmatic subjects. Am J Respir Crit Care Med 1995; 152: Rodwell LT, Anderson SD. Airway responsiveness to hyperosmolar saline challenge in cystic fibrosis: a pilot study. Pediatr Pulmonol 1996; 21: Sant Ambrogio G, Widdicombe J. Reflexes from airway rapidly adapting receptors. Respir Physiol 2001; 125: Anderson SD, Smith CM. Osmotic challenges in the assessment of bronchial hyperresponsiveness. Am Rev Respir Dis 1991; 143:S43 S46 10 Koskela HO, Hyvärinen L, Brannan JD, et al. Responsiveness to three bronchial provocation tests in patients with asthma. Chest 2003; 124: Koskela HO, Hyvärinen L, Brannan JD, et al. Sensitivity and validity of three bronchial provocation tests to demonstrate the effect of inhaled corticosteroids in asthma. Chest 2003; 124: Peittola S, Ruponen M. Pharmaca fennica. Rauma, Finland: Lääketietokeskus, 2002; Viljanen AA, Halttunen PK, Kreus K-E, et al. Spirometric studies in non-smoking, healthy adults. Scand J Clin Lab Invest 1982; 42: American Thoracic Society. Standardization of spirometry: 1994 update. Am J Respir Crit Care Med 1995; 152: Koskela HO, Räsänen SH, Tukiainen HO. The diagnostic value of cold air hyperventilation in adults with suspected asthma. Respir Med 1997; 91: Sovijarvi AR, Malmberg LP, Reinikainen K, et al. A rapid dosimetric method with controlled tidal breathing for histamine challenge: repeatability and distribution of bronchial reactivity in a clinical material. Chest 1993; 104: National Heart Lung and Blood Institute. Global initiative for asthma: global strategy for asthma management and prevention: NHLBI/WHO workshop report. Bethesda, MD: National Heart Lung and Blood Institute, 1995; Publication No Higgins BG, Britton JR, Chinn S, et al. The distribution of peak expiratory flow variability in a population sample. Am Rev Respir Dis 1989; 140: Popov TA, Pizzichini MM, Pizzichini E, et al. Some technical factors influencing the induction of sputum for cell analysis. Eur Respir J 1995; 8: Canning BJ. Interactions between vagal afferent nerve subtypes mediating cough. Pulm Pharmacol Ther 2002; 15: Pisarri TE, Jonzon A, Coleridge HM, et al. Vagal afferent and reflex responses to changes in surface osmolarity in lower airways of dogs. J Appl Physiol 1992; 73: Eggleston PA, Kagey-Sobotka A, Lichtenstein LM. A comparison of the osmotic activation of basophils and human lung mast cells. Am Rev Respir Dis 1987; 135: Lee LY, Pisarri TE. Afferent properties and reflex functions of bronchopulmonary C-fibers. Respir Physiol 2001; 125: Koskela H, Di Sciascio MB, Anderson SD, et al. Nasal hyperosmolar challenge with a dry powder of mannitol in patients with allergic rhinitis: evidence for epithelial cell involvement. Clin Exp Allergy 2000; 30: Kumlin M, Ohlson E, Björck T, et al. 15(S)-hydroxyeicosatetraenoic acid (15-HETE) is the major arachidonic acid metabolite in human bronchi. In: Samuelson B, ed. Advances in prostaglandin, thromboxane, and leukotriene research (vol 21). New York, NY: Raven Press; 1990; Hwang SW, Cho H, Kwak J, et al. Direct activation of capsaicin receptors by products of lipoxygenases: endogenous capsaicin-like substances. Proc Natl Acad Sci U S A 2000; 97: Bickerman HA, Barach AL. The experimental production of cough in human subjects induced by citric acid aerosols: preliminary studies on the evaluation of antitussive agents. Am J Med Sci 1954; 228: Pounsford JC, Birch MJ, Saunders KB. Effect of bronchodilators on the cough response to inhaled citric acid in normal and asthmatic subjects. Thorax 1985; 40: Fujimura M, Sakamoto S, Kamio Y, et al. Cough receptor sensitivity and bronchial responsiveness in normal and asthmatic subjects. Eur Respir J 1992; 5: Schmidt D, Jorres RA, Magnussen H. Citric acid-induced cough thresholds in normal subjects, patients with bronchial asthma, and smokers. Eur J Med Res 1997; 2: Fujimura M, Kamio Y, Hashimoto T, et al. Airway cough sensitivity to inhaled capsaicin and bronchial responsiveness to methacholine in asthmatic and bronchitic subjects. Respirology 1998; 3: Millqvist E, Bende M, Lowhagen O. Sensory hyperreactivity: a possible mechanism underlying cough and asthma-like symptoms. Allergy 1998; 53: Di Franco A, Dente FL, Giannini D, et al. Effects of inhaled corticosteroids on cough threshold in patients with bronchial asthma. Pulm Pharmacol Ther 2001; 14: CHEST / 125 / 6/ JUNE,

8 34 Gibson PG, Dolovich J, Denburg J, et al. Chronic cough: eosinophilic bronchitis without asthma. Lancet 1989; 1: Rytila P, Metso T, Heikkinen K, et al. Airway inflammation in patients with symptoms suggesting asthma but with normal lung function. Eur Respir J 2000; 16: Brightling CE, Bradding P, Symon FA, et al. Mast-cell infiltration of airway smooth muscle in asthma. N Engl J Med 2002; 346: Petitti DB. Epidemiologic issues in outcomes research. In: Brownson RC, Petitti DB, eds. Applied epidemiology. 1st ed. New York, NY: Oxford University Press, 1998; Brannan JD, Koskela H, Anderson SD, et al. Budesonide reduces sensitivity and reactivity to inhaled mannitol in asthmatic subjects. Respirology 2002; 7: Nguyen BP, Wilson SR, German DF. Patients perceptions compared with objective ratings of asthma severity. Ann Allergy Asthma Immunol 1996; 77: clinical investigations