Airway inflammation assessed by invasive and noninvasive means in severe asthma: Eosinophilic and noneosinophilic phenotypes

Airway inflammation assessed by invasive and noninvasive means in severe asthma: Eosinophilic and noneosinophilic phenotypes Catherine Lemière, MD, MSc, a Pierre Ernst, MD, b Ron Olivenstein, MD, b Yasuhiro Yamauchi, PhD, b Karuthapillai Govindaraju, PhD, b Mara S. Ludwig, MD, b James G. Martin, MD, b and Qutayba Hamid, MD, PhD b Montreal, Quebec, Canada Background: Airway inflammation assessed by bronchial biopsies demonstrates distinct eosinophilic and noneosinophilic phenotypes in severe asthma, but their relationship to other biomarkers of disease (induced sputum and nitric oxide [NO]) is not clear. Objectives: We sought to compare airway inflammation using noninvasive (induced sputum, exhaled NO), and invasive (bronchial biopsies) methods in moderate and severe asthma and to assess whether induced sputum and exhaled NO would allow the identification of eosinophilic and noneosinophilic phenotypes in severe asthma. Methods: We performed a cross-sectional study of 32 subjects with severe asthma and 35 subjects with moderate asthma, from whom we obtained bronchial biopsies, induced sputum, and exhaled NO measurements. Results: Among subjects with severe asthma, we identified eosinophilic and noneosinophilic phenotypes using both bronchial biopsies and sputum cell counts. However, the vast majority of subjects with high sputum eosinophil counts did not have high mucosal eosinophil counts. Exhaled NO was increased in the eosinophilic phenotype as judged from bronchial biopsy findings, but not on the basis of induced sputum. Subjects with high sputum eosinophil counts experienced more asthma exacerbations than the subjects with low sputum eosinophil counts. In contrast, we did not find any differences in the clinical characteristics between eosinophilic and noneosinophilic phenotypes that were identified by bronchial biopsies. Conclusion: The use of sputum cell counts allowed the identification of a subgroup of subjects with severe asthma who were at risk of more frequent asthma exacerbations. From a the Hôpital du Sacré-Coeur de Montréal, University of Montreal; and b Meakins Christie Laboratories, McGill University, and the Montreal Chest Institute of the McGill University Health Centre. Supported by the Richard and Edith Strauss Canada Foundation and GlaxoSmithKline Canada. Dr Lemière is a scholar of the Canadian Institutes of Health Research. Disclosure of potential conflict of interest: P. Ernst has been a consultant to Altana, AstraZeneca, GlaxoSmithKline, Merck, and Novartis, has received a research grant from GlaxoSmithKline, and has been a speaker for AstraZeneca, GlaxoSmithKline, Merck, and Novartis. The rest of the authors have declared that they have no conflict of interest. Received for publication May 9, 2006; revised August 2, 2006; accepted for publication August 3, 2006. Available online September 26, 2006. Reprint requests: Catherine Lemière, MD, MSc, Department of Chest Medicine, Sacré-Coeur Hospital, 5400 West Gouin, Montreal, Quebec, Canada, H4J 1C5. E-mail: catherine.lemiere@umontreal.ca. 0091-6749/$32.00 Ó 2006 American Academy of Allergy, Asthma and Immunology doi:10.1016/j.jaci.2006.08.003 Clinical implications: Monitoring sputum eosinophil counts in subjects with severe asthma may allow identifying the subjects with the greatest disease activity. (J Allergy Clin Immunol 2006;118:1033-9.) Key words: Severe asthma, bronchial biopsies, exhaled nitric oxide, induced sputum Although the currently available asthma medications are very effective in controlling asthma in the majority of patients, 10% to 15% of patients with asthma still have disease that is difficult to control despite maximal asthma therapy. The causes of an inadequate response are likely multiple. 1 Despite the heterogeneity of the group, there are distinguishable phenotypes. For example, noneosinophilic and eosinophilic phenotypes have been described on the basis of the presence or absence of eosinophils in bronchial biopsies. 2 Identifying these phenotypes is relevant in clinical practice because it is likely that eosinophilic phenotypes will benefit from higher doses of systemic corticosteroids. 3 Airway inflammation has been assessed in patients with asthma using relatively invasive means for example, bronchial biopsies or bronchoalveolar lavage. However, these methods are too invasive to be performed repetitively. In the last 15 years, noninvasive methods for assessing airway inflammation have been developed. It has been shown that exhaled nitric oxide (NO) is increased in uncontrolled asthma, 4 decreases with anti-inflammatory therapy, 5 and correlates with sputum eosinophil counts 6 and airway responsiveness. 7 Furthermore, exhaled NO provides helpful insight for the titration of inhaled steroids in patients with asthma. 8 Exhaled NO has recently been shown to have the capability of distinguishing eosinophilic and noneosinophilic phenotypes in severe asthma. 9 Analysis of induced sputum cellularity is a reliable method for monitoring the of asthma. Indeed, sputum eosinophils are increased by exposure to common 10-13 allergens and by reduction in steroid. 14 In contrast, they are decreased by with corticosteroids. 15,16 There is evidence that the presence of sputum eosinophilia may be a good predictor of response to inhaled steroids. 17 Furthermore, adapting the asthma according to the counts of sputum eosinophils allows for a decrease in asthma exacerbations. 18,19 However, sputum eosinophil counts reportedly do not correlate closely with mucosal eosinophilia. 20 Therefore, it is 1033

1034 Lemière et al J ALLERGY CLIN IMMUNOL NOVEMBER 2006 Abbreviations used ATS: American Thoracic Society BrEos: Bronchial eosinophil GR: Glucocorticoid receptor MBP: Major basic protein NO: Nitric oxide speosh: High sputum eosinophil count speosl: Low sputum eosinophil count unclear whether induced sputum has the ability to distinguish the eosinophilic and noneosinophilic phenotypes identified by bronchial biopsies in severe asthma. The aim of this study was, first, to compare airway inflammation using noninvasive (induced sputum, exhaled NO) and invasive (bronchial biopsies) methods in subjects with moderate or severe asthma, and second, to assess whether induced sputum and exhaled NO will allow the distinction between eosinophilic and noneosinophilic phenotypes in severe asthma. METHODS Study design This is a cross-sectional study comparing airway inflammation assessed by induced sputum, exhaled NO, and bronchial biopsies between subjects with moderate asthma and subjects with severe asthma. Induced sputum and exhaled NO were collected on the same day within no more than 3 weeks of the bronchoscopy. Subjects Subjects 18 years of age and older with moderate and severe asthma were enrolled over a 2-year period in 2 specialized respiratory clinics: the Montreal Chest Institute of McGill University Health Center and Sacré-Coeur Hospital. Asthma was defined according to the American Thoracic Society (ATS) criteria. 21 Subjects were considered as having severe asthma if, on enrollment, they fulfilled 1 major criterion and 2 minor criteria as previously defined by the ATS workshop on difficult-to-treat asthma. 1 Major criteria were (1) with daily oral steroids for >50% of the previous 12 months, and (2) high dose of inhaled steroids equivalent to more than fluticasone 1000 mg/d with at least 1 other add-on therapy (long-acting b-agonist, leukotriene receptor antagonist, or theophylline) continuously over the period of the previous 12 months. Minor criteria were need for daily short-acting b-agonist, persistent airflow obstruction (prebronchodilator FEV 1 < 70% and FEV 1 /forced vital capacity <80% predicted), 1 or more emergency care visits in the last 12 months, 3 or more steroid bursts in the last 12 months, prompt deterioration with 25% or lower dose reduction of oral corticosteroids, and near-fatal asthma event in the last 3 years. Subjects were considered to have moderate asthma if they had asthma controlled on a minimum of 200 mg/d fluticasone (or equivalent) and not more than 1000 mg/d fluticasone (or equivalent) with or without concomitant therapy with a long-acting b-agonist, leukotriene receptor antagonists, or theophylline; they had no more than 2 steroid bursts in the past year, and none in the past 3 months, with total days on oral steroids less than 30 days in the previous 12 months; and they had a FEV 1 > 70% predicted and >90% of personal best from the past 2 years. They had a maximum of 1 unscheduled visit for asthma in the previous 12 months. Asthma exacerbations that occurred in the year preceding the study were defined by the use of oral steroids, visit to the emergency department, or hospitalization for asthma during this period. Subjects who were currently smoking or who were known to have other pulmonary diseases, including a previous diagnosis of chronic obstructive pulmonary disease (COPD), were excluded. The study was approved by the ethics committees of both participating institutions. All subjects provided signed informed consent. Procedures Symptom severity was graded according to the Juniper asthma control questionnaire. 22 Spirometry was performed according to the ATS standards. 23 Allergy skin prick tests with commercial extracts from common allergens were performed using the modified prick method; results were regarded as positive if the wheal was 3 mm. Sputum was induced using inhalations of increasing concentrations (3%, 4%, and 5%) of hypertonic saline 24 and processed as previously described. 25 All slides were read at the same center. Exhaled NO was collected by using a validated offline technique with sample bags in accordance with the ATS recommendations. 26 Exhaled NO was measured from these sample bags by the chemiluminescence technique using a Nitric Oxide Analyzer (model Sievers 280i; Sievers Instruments Inc, Boulder, Colo) interfaced through an analog-to-digital converter board to a personal computer. Bronchoscopy was performed under mild sedation. Six endobronchial biopsies were obtained from the right lung at various segmental and subsegmental carinae. Two blocks were frozen, embedded in optimal cutting temperature medium, sectioned, and fixed in acetone/methanol. Tissues were fixed with 10% formalin and embedded in paraffin wax before cutting. Tissue sections (5 mm) were immunostained for specific eosinophil (eosinophil major basic protein [MBP]) and neutrophil (neutrophil elastase) markers using the alkaline phosphatase technique. Immunoreactivity was detected by using Fast Red, visualized under light microscopy, and analyzed by using image analyses. As described in a previous article, 27 glucocorticoid receptor (GR) b and " antibodies were obtained from ABR Affinity BioReagents (Golden, Colo). Immunoreactivity was evaluated in the epithelium with a scoring technique based on the percentage of positive cells. Although 6 biopsies were taken, only 2 were available for immunocytochemistry. Results from the 2 biopsies were consistent. The cell counts were normalized per tissue area and expressed as cells per square millimeter of airway submucosa. Tissue analysis was performed blind to the clinical data by 2 observers. The interobserver variation was less than 5%. Analysis Results were expressed as means and SDs, except data with nonnormal distribution, which were expressed as median and interquartile ranges. Sputum and bronchial cell counts as well as exhaled NO data were analyzed after logarithmic transformation. Subjects were divided according to the level of eosinophilic inflammation found in induced sputum and bronchial biopsies. According to previous studies, 9,17 they were divided into low sputum eosinophil count (speosl) if they had sputum eosinophils lower than 3%, and high sputum eosinophil count (speosh) if they had sputum eosinophils equal to or greater than 3%. 17 They were divided into BrEos if they had bronchial eosinophils lower than 22 cells/mm 2 and BrEos 1 if they had mucosal eosinophils equal to or greater than 22. 9 Other normally distributed data were compared by using a 2-sided Student t test. Correlations between the different variables of interest were examined using a Pearson correlation test. Significance was accepted at the level of 95%. The analysis was performed using the SPSS 12.0 statistical package (Chicago, Ill).

J ALLERGY CLIN IMMUNOL VOLUME 118, NUMBER 5 Lemière et al 1035 TABLE I. Clinical characteristics of subjects with moderate and severe asthma RESULTS Thirty-five subjects with severe asthma and 32 subjects with mild to moderate asthma were enrolled. Their clinical characteristics are summarized in Table I. As expected, subjects with severe asthma had a lower FEV 1, a higher number of asthma exacerbations in the year preceding the study, and more asthma symptoms in spite of a higher dose of inhaled steroids or systemic steroids than the subjects with mild to moderate asthma (Table I). Bronchial biopsies Moderate asthma Severe asthma n 35 32 Sex (male/female) 20/15 13/19 Atopy 22 24 Age (y) 45.2 6 10.8 47.7 6 11.4 Duration of asthma (y) 20.9 6 14.0 19.81 6 16.5 Dose of ICS (mg) 517.1 6 266.2 1390.6 6 385.9* Subjects on LABA (n) 26 32 Subjects on oral 0 10 (11.5 6 7.2) steroids (n [mg]) Asthma exacerbations in 0.4 6 0.7 3.1 6 0.55* the previous year (n) Exsmokers (n [PY]) 17 (9.2 6 7.4) 13 (13.5 6 8.3) Symptom score 0.9 6 0.7 1.9 6 1.3* FEV 1 (% predicted) 87.2 6 11.7 69.4 6 24.6* FEV 1 /FVC (%) 73.6 6 7.0 65.6 6 12.7* FVC, Forced vital capacity; ICS, inhaled corticosteroid; LABA, long-acting b 2 -agonist; PY, pack-year. *P.05. Bronchial biopsies were obtained in 46 subjects 22 moderate and 24 severe. Nine subjects with severe asthma and 6 subjects with moderate asthma refused to undergo bronchoscopy because of the relatively invasive nature of the test. GR b and GR " expression was measured in 12 subjects with severe asthma and 16 subjects with moderate asthma. We did not find any differences in the intensity of the granulocytic inflammation, as determined by area corrected numbers of MBP or elastase-positive cells in biopsies, between the subjects with moderate and severe asthma (Fig 1; Table II). Although there was no difference in the GR " expression between subjects with severe (2.5 6 2.2) and moderate asthma (3.6 6 2.5; P 5.3), there was an increased expression of GR b in subjects with severe asthma (4.7 6 1.6) compared with subjects with moderate asthma (1.8 6 0.9; P<.001). GR "/GR b ratio was higher in subjects with moderate (2.1 6 0.6) compared with severe (0.6 6 0.6; P 5.03) asthma, as previously reported. 27 The group of subjects with severe asthma consisted of a BrEos 1 group of 9 subjects who had a median (interquartile range) of 75.7 (65.0) eosinophils/mm 2 and a BrEos group of 15 subjects who had 7.5 (12.9) mm 2 mucosal FIG 1. Sputum neutrophils and elastase-positive cells in bronchial biopsies in severe and moderate asthma. There was no difference in sputum neutrophils and mucosal neutrophils between subjects with severe and moderate asthma. TABLE II. Characteristics of airway inflammation assessed by bronchial biopsies, induced sputum, and exhaled NO in subjects with moderate and severe asthma Moderate asthma Severe asthma Sputum eosinophils (%) 0.2 (3.6) 1.3 (17.9)* Sputum neutrophils (%) 24.7 (32.7) 25.5 (33.7) Exhaled NO (ppb) 12.2 (9.1) 13.2 (17.6) MBP-positive cells/mm 2 10.57 (22.74) 14.2 (62.37) Elastase-positive cells/mm 2 59.2 (59.0) 78.8 (68.6) *P.05. eosinophils. Fig 2 illustrates the presence of MBP-positive cells in the airway wall of a subject with severe asthma. We did not find any difference in the clinical parameters symptoms (1.5 6 1.2 vs 2.3 6 1.2; P 5.08), FEV 1 (72.2 6 24.2% predicted vs 69.8 6 25.7% predicted; P 5.8), number of exacerbations during the year preceding the study (3.3 6 3.3 vs 3.1 6 3.3; P 5.9), or duration of asthma (15.1 y 6 14.7 y vs 20.8 y 6 16.5 y; P 5.3) between the BrEos 1 and BrEos groups. There was a higher level of exhaled NO in the BrEos 1 group (26.8 [47.0] ppb) compared with the BrEos group (11.3 [14.9] ppb; P 5.05). The BrEos 1 group also had a higher number of neutrophils (89.5 [80.0]/mm 2 ) than the BrEos group (63.8 [65.0]/mm 2 ; P 5.02). There was no difference in the GR b expression (3.7 6 0.6 vs 5.0 6 1.7; P 5.2), the GR " expression (1.5 6 0.7 vs 2.8 6 2.5; P 5.5), and the GR b/gr " ratio (0.5 6 0.3 vs 0.7 6 0.7; P 5.7) between the BrEos 1 and the BrEos group. The moderate group was composed of a BrEos 1 group of 6 subjects who had 66.3 (60.13) eosinophils/mm 2 and a BrEos group of 16 subjects who had 8.5 (7.9) mm 2 mucosal eosinophils. We did not find any difference in the clinical parameters symptoms (0.4 6 0.3 vs 0.9 6 0.7; P 5.08), FEV 1 (92.5 6 11.4 % predicted vs 85.3 6 11.8 % predicted; P 5.3), level of exhaled NO (17.6 [55.1] ppb vs 12.0 [7.65] ppb; P5.2), number of exacerbations during the year preceding the study (0.7 6 1.0 vs 0.4 6 0.6; P 5.4), or duration of asthma (16.4 y 6 14.3 y

1036 Lemière et al J ALLERGY CLIN IMMUNOL NOVEMBER 2006 FIG 2. Immunostaining for MBP in the airway wall of a subject with severe asthma. vs 24.2 y 6 16.3 y; P 5.3) between BrEos 1 and BrEos moderate groups. However, the BrEos 1 group had a higher number of neutrophils (91.6 [73.7]/mm 2 ) than the BrEos group (46.9 [47.5] neutrophils/mm 2 ; P 5.02). There was no difference in the GR b expression, the GR " expression, and the GR b/gr " ratio between the BrEos 1 compared with the BrEos group (data not shown). We did not find any significant correlation between the number of MBP or elastase-positive cells and any of the clinical parameters, symptoms, FEV 1, number of exacerbations, and duration of asthma in subjects with moderate or severe asthma. There was no correlation either with the GR b expression, the GR " expression, and the GR b/ GR " ratio (data not shown). Exhaled NO Exhaled NO was obtained in 61 subjects 30 moderate and 31 severe. We did not find any differences in the level of exhaled NO between subjects with moderate and severe asthma (Table II). Furthermore, we did not find any significant correlation between exhaled NO and any of the collected clinical parameters, symptoms (r 5 0.3; P 5.1), FEV 1 (r 5 0.2; P 5.2), number of exacerbations (r 5 0.1; P 5.5), and duration of asthma (r 5 0.05; P 5.9). Sputum cell counts Sputum was successfully obtained in 61 subjects 30 moderate and 31 severe. Overall, the subjects with severe asthma had a higher percentage of sputum eosinophils (1.3 [17.9] %) than subjects with mild/moderate asthma (0.2 [3.6] %). The group of subjects with severe asthma consisted of a group of 13 subjects with speosh (median eosinophil count of 22.5 [38.8] %), with 6 on oral steroids; and a group of 18 subjects with speosl (0.5 [1.05] %), with 7 on oral steroids. The speosh group had more asthma exacerbations (4.7 6 3.8 vs 1.9 6 2.1; P 5.04) during the year before their enrollment in the current study and had asthma for a shorter time (10.1 y 6 8.1 y vs 26.2 y 6 17.6 y; P 5.01) compared with the speosl group. Sputum neutrophils FIG 3. Correlation between sputum eosinophils and asthma exacerbations during the year preceding the study. There was a positive significant correlation between the number of asthma exacerbations in the year preceding the study and the percentage of sputum eosinophils in subjects with severe asthma. were similar in subjects with moderate and severe asthma (Fig 1; Table II). There was a significant correlation between the number of asthma exacerbations reported by the subjects with severe asthma during the year preceding the study and the percentage of sputum eosinophils (r 5 0.38; P 5.04; Fig 3). This correlation was not found in the group of subjects with moderate asthma (r 5 0.2; P5.4). There was also a significant inverse correlation between the duration of asthma and the percentage of sputum eosinophils in subjects with severe asthma (r 5 0.5; P 5.006; Fig 4). This correlation was not found in subjects with moderate asthma (r 5 0.1; P 5.6). There was no correlation between the GR b expression, the GR " expression, and the GR b/gr " ratio and the sputum eosinophil counts in subjects with severe or moderate asthma (data not shown). Correlation between sputum cell counts, exhaled NO, and MBP and elastase-positive cells The speosh group did not show a high number of eosinophils in their bronchial mucosa. Indeed, there was no correlation between sputum eosinophil count and eosinophils found in the bronchial mucosa in subjects

J ALLERGY CLIN IMMUNOL VOLUME 118, NUMBER 5 Lemière et al 1037 FIG 4. Correlation between sputum eosinophils and duration of asthma. There was an inverse correlation between the duration of asthma and the number of sputum eosinophil in subjects with severe asthma, indicating that the subjects with the highest sputum eosinophilia had asthma for the shortest time. FIG 5. Correlation between exhaled NO and MBP-positive cells in bronchial biopsies. Subjects with severe asthma with high exhaled NO also showed high mucosal eosinophilia. with moderate (r 5 0.1; P 5.5) or severe asthma (r 5 0.09; P 5.7). There was a significant correlation between exhaled NO and mucosal eosinophils (r 5 0.4; P 5.05) in subjects with severe asthma but not moderate asthma (r 5 0.2; P 5.4; Fig 5). Overall, there was a significant correlation between exhaled NO and mucosal eosinophils (r 5 0.3; P 5.05). We did not find any correlation between exhaled NO and sputum eosinophils in subjects with moderate (r 520.1; P 5.5) or severe (r 5 0.1; P 5.5) asthma or in the whole study population (r 5 0.08; P 5.5). We found a significant correlation between sputum neutrophils and elastase-positive cells in the airway wall in subjects with moderate (r 5 0.6; P 5.004) but not severe asthma (r 5 0.06; P 5.7). Exhaled NO was not significantly correlated with elastase-positive cells in subjects with moderate (r 5 0.2; P 5.3) or severe (r 5 0.2; P 5.4) asthma. There was a correlation between the number of MBP-positive cells and elastase-positive cells in moderate asthma (r 5 0.5; P 5.004) that did not reach statistical significance in severe asthma (r 5 0.06; P 5.7). DISCUSSION As previously described, 12 we found that subjects with severe asthma had eosinophilic and noneosinophilic phenotypes. These phenotypes were identified using both bronchial biopsies and sputum cell counts. However, most (9 out of 13) subjects with high sputum eosinophil counts did not have high mucosal eosinophil counts. Exhaled NO was increased in the eosinophilic phenotype identified from bronchial biopsy findings, but not on the basis of induced sputum. Subjects with high sputum eosinophil counts had more asthma exacerbations in the year preceding enrollment than the subjects with low sputum eosinophil counts. In contrast, we did not find any differences in clinical characteristics between subjects with eosinophilic and noneosinophilic phenotypes according to bronchial biopsies or exhaled NO. These results are consistent with previous studies showing a good correlation between sputum eosinophil counts and eosinophils in bronchoalveolar lavage, but a poor correlation between sputum eosinophils and eosinophils from bronchial biopsies. 20,28 However, those studies were performed only in subjects with mild to moderate asthma. It is the first time to our knowledge that induced sputum, bronchial biopsies, and exhaled NO were compared in subjects with severe asthma. We enrolled subjects with moderate asthma as control subjects rather than enrolling subjects with mild asthma or healthy subjects to reduce the potential confounding effect of with inhaled steroids on airway inflammation.

1038 Lemière et al J ALLERGY CLIN IMMUNOL NOVEMBER 2006 Indeed, subjects with severe asthma are treated with high doses of inhaled corticosteroids, whereas subjects with mild asthma or healthy subjects take low doses or no inhaled corticosteroids. This is likely to have an important effect on airway inflammation as measured especially by induced sputum 29 and exhaled NO. 30 The phenotypes defined according to sputum eosinophils seem to correlate better with asthma severity than the phenotypes defined by bronchial biopsies. Although the levels of exhaled NO and mucosal eosinophils were similar in subjects with severe and moderate asthma, subjects with severe asthma had more sputum eosinophils than subjects with moderate asthma, which is consistent with previously published studies. 31 Subjects with severe asthma with high sputum eosinophil counts had more asthma exacerbations than subjects with severe asthma with low sputum eosinophil counts. It is likely that sputum eosinophilia is a good reflect of the disease activity. Indeed, Romagnoli et al 32 showed that sputum cell counts were more closely correlated with asthma control than with asthma severity. Subjects with severe asthma have been shown to have a higher proportion of activated sputum eosinophils than subjects with mild asthma. 33 Eosinophils in the airway lumen may be in a different state of activation than eosinophils found in the bronchial mucosa or may be indicative of greater concentrations of intraluminal chemoattractants such as eotaxin and RANTES. If the presence of intraluminal eosinophils reflects the degree of epithelial activation, then it is not surprising that this asthma phenotype is more severe. An alternative explanation is that induced sputum represents the convergence of cells entering the airways in different sites within the bronchial tree. Large airways are sampled at the beginning of sputum induction, whereas peripheral airways and alveoli are sampled at later times. 34 Bronchial biopsies were restricted to the large airways. Therefore, induced sputum is likely also to reflect inflammation of the small airways, which may be prominent in severe asthma. Indeed, Wenzel et al 35 reported significant distal neutrophilic inflammation in subjects with severe corticosteroid-dependent asthma. Suppressing sputum eosinophils by using increasing doses of inhaled steroids has been shown to reduce the number of subsequent exacerbations. 18,19 In our study, subjects with high sputum eosinophil counts were more prone to have had an asthma exacerbation in the past and thus were more likely to be at risk of experiencing subsequent asthma exacerbations. Treating these subjects with higher doses of inhaled steroids or increasing the dose of oral steroids may have improved the management of their asthma. 3 The persistence of airway eosinophilia in these subjects reflects a failure of usually adequate doses of corticosteroid to suppress inflammation. 36 Indeed, the b isoform of the glucocorticoid receptor, frequently associated with steroid resistance, was overexpressed in the airways of the severe asthma group compared with the moderate group, as previously reported. 27 We could not find any correlation between the number of sputum or mucosal eosinophils and the GR b expression. However, it is likely that the current study was underpowered to demonstrate such an association. Ten Brinke et al 37 had previously shown that high sputum eosinophil counts were associated with persistent airflow limitation in subjects with severe asthma. Although we could not find any differences in airflow obstruction between subjects with high and low eosinophil counts according to sputum or bronchial biopsies, we found that subjects with severe asthma with high sputum eosinophil counts had asthma for a shorter duration than subjects with low sputum eosinophilic counts. Those subjects may be at greater risk to develop airway obstruction than the subjects with low eosinophils as their asthma progresses. We did not find any correlation between sputum eosinophils and exhaled NO. Although several studies found a correlation between these 2 parameters, the majority of these studies were performed in subjects with mild to moderate asthma. In a large study of 566 subjects with asthma, Berry et al 38 failed to show a significant correlation between sputum eosinophils and exhaled NO in exsmokers, but found a positive correlation between NO and sputum eosinophils in smokers and nonsmokers. The subjects studied appeared to be a group of subjects with milder asthma with lower inhaled steroid doses than the group of subjects we studied. Most of the study subjects were treated with high doses of inhaled steroids and some of them with systemic steroids. Because exhaled NO is suppressed by corticosteroids, it is likely that the high doses of inhaled steroids taken by the study subjects prevented us from finding differences in exhaled NO between the 2 groups. The subjects with eosinophilia of the bronchial mucosa showed higher NO levels. Whether these observations are causally related is not clear. Perhaps both measurements are a reflection of steroid resistance or persistent activation of epithelial cells, which are the major site of synthesis of NO by inducible nitric oxide synthase (inos) and a source of eosinophil chemoattractant factors. In conclusion, the use of sputum cell counts allowed the identification of a subgroup of subjects with severe asthma experiencing more frequent asthma exacerbations than subjects with low sputum cell counts. Bronchial biopsies also identify a subgroup with high mucosal eosinophils, but this characteristic did not seem to be associated with clinical markers of asthma severity. Induced sputum may be a better reflection of current disease activity than the analysis of bronchial biopsies or exhaled NO. REFERENCES 1. Proceedings of the ATS workshop on refractory asthma: current understanding, recommendations, and unanswered questions. American Thoracic Society. Am J Respir Crit Care Med 2000;162:2341-51. 2. Wenzel SE, Schwartz LB, Langmack EL, Halliday JL, Trudeau JB, Gibbs RL, et al. Evidence that severe asthma can be divided pathologically into two inflammatory subtypes with distinct physiologic and clinical characteristics. Am J Respir Crit Care Med 1999;160:1001-8.

J ALLERGY CLIN IMMUNOL VOLUME 118, NUMBER 5 Lemière et al 1039 3. ten Brinke A, Zwinderman AH, Sterk PJ, Rabe KF, Bel EH. Refractory eosinophilic airway inflammation in severe asthma: effect of parenteral corticosteroids. Am J Respir Crit Care Med 2004;170:601-5. 4. Crater SE, Peters EJ, Martin ML, Murphy AW, Platts-Mills TA. Expired nitric oxide and airway obstruction in asthma patients with an acute exacerbation. Am J Respir Crit Care Med 1999;159:806-11. 5. Massaro AF, Gaston B, Kita D, Fanta C, Stamler JS, Drazen JM. Expired nitric oxide levels during of acute asthma. Am J Respir Crit Care Med 1995;152:800-3. 6. Leuppi JD, Downs SH, Downie SR, Marks GB, Salome CM. Exhaled nitric oxide levels in atopic children: relation to specific allergic sensitisation, AHR, and respiratory symptoms. Thorax 2002;57:518-23. 7. Langley SJ, Goldthorpe S, Custovic A, Woodcock A. Relationship among pulmonary function, bronchial reactivity, and exhaled nitric oxide in a large group of asthmatic patients. Ann Allergy Asthma Immunol 2003;91:398-404. 8. Smith AD, Cowan JO, Brassett KP, Herbison GP, Taylor DR. Use of exhaled nitric oxide measurements to guide in chronic asthma. N Engl J Med 2005;352:2163-73. 9. Silkoff PE, Lent AM, Busacker AA, Katial RK, Balzar S, Strand M, et al. Exhaled nitric oxide identifies the persistent eosinophilic phenotype in severe refractory asthma. J Allergy Clin Immunol 2005;116: 1249-55. 10. Pin I, Freitag A, O Byrne P, Girgis-Gavardo A, Watson R, Dolovich J, et al. Changes in the cellular profile of induced sputum after allergeninduced asthmatic responses. Am Rev Respir Dis 1992;145:1265-9. 11. Fahy J, Liu J, Wong H, Boushey H. Analysis of cellular and biochemical constituents of induced sputum after allergen challenge: a method for studying allergic airway inflammation. J Allergy Clin Immunol 1994; 93:1031-9. 12. Wong B, Dolovich J, Ramsdale E, O Byrne P, Gontovnick L, Denberg J, et al. Formoterol compared with beclomethasone and placebo on allergen-induced asthmatic responses. Am Rev Respir Dis 1992;146:1156-60. 13. Pizzichini M, Kidney J, Wong B, Morris M, Efthimiadis A, Dolovich J, et al. Effect of salmeterol compared with beclomethasone on allergen-induced asthmatic and inflammatory responses. Eur Respir J 1996;9:449-55. 14. Gibson P, Wong B, Hepperle M, Kline P, Girgis-Gavardo A, Guyatt G, et al. A research method to induce and examine a mild exacerbation of asthma by withdrawal of inhaled corticosteroid. Clin Exp Allergy 1992; 22:525-32. 15. Pizzichini MM, Pizzichini E, Clelland L, Efthimiadis A, Mahony J, Dolovich J, et al. Sputum in severe exacerbations of asthma: kinetics of inflammatory indices after prednisone. Am J Respir Crit Care Med 1997;155:1501-8. 16. Hargreave FE. Induced sputum and response to glucocorticoids. J Allergy Clin Immunol 1998;102:S102-5. 17. Pavord ID, Brightling CE, Woltmann G, Wardlaw AJ. Non-eosinophilic corticosteroid unresponsive asthma [letter]. Lancet 1999;353:2213-4. 18. Green RH, Brightling CE, McKenna S, Hargadon B, Parker D, Bradding P, et al. Asthma exacerbations and sputum eosinophil counts: a randomised controlled trial. Lancet 2002;360:1715-21. 19. Jayaram L, Pizzichini MM, Cook RJ, Boulet LP, Lemiere C, Pizzichini E, et al. Determining asthma by monitoring sputum cell counts: effect on exacerbations. Eur Respir J 2006;27:483-94. 20. Grootendorst DC, Sont JK, Willems LN, Kluin-Nelemans JC, Van Krieken JH, Veselic-Charvat M, et al. Comparison of inflammatory cell counts in asthma: induced sputum vs bronchoalveolar lavage and bronchial biopsies. Clin Exp Allergy 1997;27:769-79. 21. Standards for the diagnosis and care of patients with chronic obstructive pulmonary disease (COPD) and asthma. This official statement of the American Thoracic Society was adopted by the ATS Board of Directors, November 1986. Am Rev Respir Dis 1987;136:225-44. 22. Juniper EF, O Byrne PM, Ferrie PJ, King DR, Roberts JN. Measuring asthma control: clinic questionnaire or daily diary? Am J Respir Crit Care Med 2000;162:1330-4. 23. American Thoracic Society. Standardization of spirometry: 1994 update. Am J Respir Crit Care Med 1995;152:1107-36. 24. Pin I, Gibson P, Kolendowicz F, Girgis-Gabardo A, Denburg J, Hargreave F, et al. Use of induced sputum cell counts to investigate airway inflammation in asthma. Thorax 1992;47:25-9. 25. Pizzichini E, Pizzichini M, Efthimiadis A, Evans S, Morris M, Squillace D, et al. Indices of airway inflammation in induced sputum: reproducibility and validity of cell and fluid-phase measurements. Am J Respir Crit Care Med 1996;154:308-17. 26. Recommendations for standardized procedures for the on-line and offline measurement of exhaled lower respiratory nitric oxide and nasal nitric oxide in adults and children 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 1999;160:2104-17. 27. Bergeron C, Fukakusa M, Olivenstein R, Lemiere C, Shannon J, Ernst P, et al. Increased glucocorticoid receptor-beta expression, but not decreased histone deacetylase 2, in severe asthma. J Allergy Clin Immunol 2006;117:703-5. 28. Maestrelli P, Saetta M, Stefano AD, Calcagni P, Turarto G, Ruggieri M, et al. Comparison of leukocyte counts in sputum, bronchial biopsies, and bronchoalveolar lavage. Am J Respir Crit Care Med 1995;152: 1926-31. 29. Jayaram L, Parameswaran K, Sears MR, Hargreave FE. Induced sputum cell counts: their usefulness in clinical practice. Eur Respir J 2000;16: 150-8. 30. Kharitonov SA, Barnes PJ. Clinical aspects of exhaled nitric oxide. Eur Respir J 2000;16:781-92. 31. Bartoli ML, Bacci E, Carnevali S, Cianchetti S, Dente FL, Di Franco A, et al. Clinical assessment of asthma severity partially corresponds to sputum eosinophilic airway inflammation. Respir Med 2004;98:184-93. 32. Romagnoli M, Vachier I, Tarodo de la Fuente P, Meziane H, Chavis C, Bousquet J, et al. Eosinophilic inflammation in sputum of poorly controlled asthmatics. Eur Respir J 2002;20:1370-7. 33. Jang AS, Choi IS, Park CS. Immunohistochemically stained activated eosinophils in sputum in patients with asthma. Respiration 2000;67: 183-8. 34. Gershman NH, Liu H, Wong HH, Liu JT, Fahy JV. Fractional analysis of sequential induced sputum samples during sputum induction: evidence that different lung compartments are sampled at different time points. J Allergy Clin Immunol 1999;104:322-8. 35. Wenzel SE, Szefler SJ, Leung DY, Sloan SI, Rex MD, Martin RJ. Bronchoscopic evaluation of severe asthma: persistent inflammation associated with high dose glucocorticoids. Am J Respir Crit Care Med 1997; 156:737-43. 36. Chakir J, Hamid Q, Bosse M, Boulet LP, Laviolette M. Bronchial inflammation in corticosteroid-sensitive and corticosteroid-resistant asthma at baseline and on oral corticosteroid. Clin Exp Allergy 2002; 32:578-82. 37. ten Brinke A, Zwinderman AH, Sterk PJ, Rabe KF, Bel EH. Factors associated with persistent airflow limitation in severe asthma. Am J Respir Crit Care Med 2001;164:744-8. 38. Berry MA, Shaw DE, Green RH, Brightling CE, Wardlaw AJ, Pavord ID. The use of exhaled nitric oxide concentration to identify eosinophilic airway inflammation: an observational study in adults with asthma. Clin Exp Allergy 2005;35:1175-9.