Rationale: Predicting corticosteroid response in COPD is important but difficult. Response is more likely to occur in association with eosinophilic airway inflammation, for which the fraction of exhaled nitric oxide (FeNO) is a good surrogate marker.
Objectives: We aimed to establish whether FeNO levels would predict the clinical response to oral corticosteroid in COPD.
Methods: We performed a double-blind, crossover trial of steroid in patients with COPD. After a 4-week washout of inhaled steroids, patients received prednisone 30 mg/d or matching placebo, in random order, with an intervening 4-week washout. The predictive values of FeNO for clinically significant changes in 6-minute-walk distance (6MWD), spirometry (FEV1), and St. George's Respiratory Questionnaire (SGRQ) were calculated.
Measurements and Main Results: A total of 65 patients (mean FEV1 = 57% predicted) were randomized. With prednisone, there was a net increase of 13 m in 6MWD (P = 0.02) and 0.06 L in postbronchodilator FEV1 (P = 0.02) compared with placebo. The change in SGRQ was not significant. Using receiver operator characteristic analysis, the area under the curve for an increase of 0.2 L in FEV1 was 0.69 (P = 0.04) with an optimum FeNO cut-point of 50 ppb. The positive and negative predictive values were 67 and 82%, respectively. FENO was not a significant predictor for changes in 6MWD or SGRQ.
Conclusions: FeNO is a weak predictor of short-term response to oral corticosteroid in COPD, its usefulness being limited to predicting increase in FEV1.
Clinical trial registered with www.anzctr.org.au (ACTRN12605000683639).
Only a minority of patients with chronic obstructive pulmonary disease (COPD) have steroid-responsive airway inflammation, and they are difficult to identify. Exhaled nitric oxide (FeNO) predicts steroid response in patients with nonspecific airway symptoms. Its usefulness in COPD has not been systematically assessed.
FeNO is a weak predictor of short-term steroid response in COPD, its usefulness being limited to predicting an increase in FEV1.
There is evidence that steroid responders are more likely to be characterized by the presence of eosinophilic airway inflammation. Studies have shown that sputum eosinophilia in patients with COPD is associated with a short-term response to corticosteroid, demonstrated by increased airway caliber and improved health-related quality of life (8–11). However, the clinical applicability of sputum induction is limited because it is technically demanding and results are not immediately available. In contrast, measurement of the fraction of nitric oxide in exhaled breath (FeNO) is simple and reliable (12). FeNO correlates with eosinophilic airway inflammation (13–15) and has usefulness as a predictor of corticosteroid response in patients with nonspecific chronic airways symptoms (16). To date, its application among patients with COPD has not been systematically assessed.
We hypothesized that FeNO levels could be used to predict short-term response to corticosteroid in COPD. We performed a double-blind, placebo-controlled, crossover trial to evaluate FeNO as a predictor of changes in functional exercise capacity, lung function, and health-related quality of life in response to a trial of treatment with oral prednisone. Oral rather than inhaled steroid was chosen so as to minimize drug response variability associated with inadequate inhalation technique and drug deposition.
We recruited patients with a diagnosis of COPD from our own research database and respiratory clinics between April 2003 and October 2008. Patients were 45 years of age or older, had a smoking history of more than 10 pack-years, persistent symptoms of chronic airflow obstruction, a postbronchodilator FEV1/FVC of less than 70%, and FEV1 of 30 to 80% predicted. Current smokers were excluded because of the effect of smoking on exhaled nitric oxide levels (17). Other exclusions were diagnoses of asthma, bronchiectasis, lung cancer, diabetes, or any other comorbidity likely to affect completion of the study. Patients taking regular oral corticosteroid or who had required oral corticosteroid for exacerbations more than twice during the previous 6 months were also excluded. Patients who developed an acute exacerbation (AECOPD) during the study were reviewed, treated, and once clinically stable were considered for reentry into the study. A second exacerbation resulted in withdrawal of the patient. The study was approved by the Canterbury and Otago Ethics Committees, and all patients gave written, informed consent.
The study was a randomized, double-blind, placebo-controlled, crossover trial of oral prednisone (30 mg/d) for 3 weeks per treatment period (Figure 1). Any ICS treatment was withdrawn 4 weeks before the first treatment period. The two treatment periods were separated by a 4-week washout period during which the patients did not receive any inhaled or oral corticosteroid. Patients attended the research clinic before and after each treatment period and performed a fixed sequence of assessments at each visit: St. George's Respiratory Questionnaire (SGRQ) (18), FeNO measured according to current recommendations (19), spirometry before and 15 minutes after 400 μg of inhaled albuterol, 6-minute-walk test according to current guidelines (20), and then sputum induction. Further details of procedures are available in the online data supplement.
The primary outcome was change in 6-minute-walk distance (6MWD) after prednisone, with postbronchodilator FEV1 and SGRQ total score as coprimary outcomes. It was determined that a minimum of 48 completed patients would be required to determine a significant treatment-related difference in 6MWD of 35 m (α = 0.05, β = 0.2). Secondary outcomes were changes in individual SGRQ domains, FeNO, and sputum eosinophil counts.
Randomized patients were excluded from analysis for nonadherence or because of adverse events other than deteriorating respiratory function. Patients who withdrew during the first treatment period were excluded from analysis. Patients who withdrew during the second treatment period or the between-treatment washout were assigned a net change of zero for outcome variables for the second treatment period.
Results for FeNO and sputum eosinophil counts were logarithmically transformed before analysis. Comparisons of treatment-related outcomes were performed using paired t tests and analysis of variance. The significance of the change in outcomes across tertiles was analyzed by linear regression (Table 3). Correlations were determined using Spearman rank correlation. Receiver operator characteristic (ROC) analyses were performed to determine the predictive usefulness of FeNO for improvement in primary outcomes with prednisone. The clinically relevant cut-points for FeNO and sputum eosinophil percentage were greater than 25 ppb and greater than 2%, respectively. Minimal clinically important differences for each of the primary endpoints were: 35 m for the 6MWD (21), 0.2 L for FEV1 (22, 23), and −4 units for the SGRQ (22, 24). Analyses were performed using SPSS 16.
A total of 82 patients were recruited, of whom 65 proceeded to randomization. Of the 17 patients not randomized, 13 were symptomatic after withdrawal of ICS, 2 were too busy to continue, 1 had inadequate FeNO technique, and 1 had an unrelated illness. Data from 62 patients were included in the analysis: 2 patients were excluded because of nonadherence and 1 was excluded because of a new diagnosis of angina. A total of 55 patients completed all parts of the study. Of the seven who completed only the first treatment arm, four withdrew during the washout after receiving prednisone (three with AECOPD, one restarted smoking), two withdrew while on placebo in the second treatment arm (both with AECOPD), and one withdrew in the washout after placebo (AECOPD). Adherence to treatment, assessed by pill count from retrieved medication containers, was 96%. Treatment order had no significant effect on the primary and coprimary outcomes.
The baseline data for each of the study endpoints, stratified by baseline FeNO, are shown in Table 1. With prednisone, the 6MWD increased by 13 m (95% confidence interval [CI]: 3–22 m, P = 0.02) compared with placebo, and FEV1 increased by 0.06 liters (95% CI: 0.02–0.11 L, P = 0.02) (Table 2). There was a nonsignificant decrease in SGRQ score of −2.4 (−5.3 to 0.6, P = 0.16). The number of responders who demonstrated changes greater than or equal to the minimal clinically important difference was 8 for 6MWD (12.9%), 14 for FEV1 (22.6%), and 21 for SGRQ (33.9%).
Tertiles by FeNO (ppb)
Subjects Completing 1st Treatment Only
|All Subjects||< 19.0||19.0–30.2||> 30.2|
|No. of patients||62||21||21||20||7|
|Age, yr (range)||72 (59–86)||70 (59–81)||73 (63–79)||73 (61–86)||75 (62–81)|
|Sex, female||18 (29%)||7 (33%)||8 (38%)||3 (15%)||1 (14%)|
|BMI, kg/m2||27.1 (26.0–28.2)||26.7 (24.9–28.5)||28.6 (26.6–30.5)||26.0 (24.1–27.9)||25.4 (23.2–27.6)|
|No. of pack-years||47 (41–53)||46 (36.5–56.2)||48 (37–58)||47 (36–58)||50 (35–65)|
|ICS dose, μg (range)||475 (0–2000)||498 (0–2000)||229 (0–1000)||710 (0–2000)||1257 (0–2000)|
|Patients taking long-acting bronchodilator||11 (18%)||5 (24%)||1 (5%)||5 (25%)||3 (43%)|
|FEV1/FVC, % (postbronchodilator)||49 (47–51)||50 (46–54)||50 (46–54)||47 (44–49)||45 (41–50)|
|FEV1, L (postbronchodilator)||1.58 (1.46–1.71)||1.68 (1.48–1.88)||1.50 (1.28–1.72)||1.57 (1.34–1.80)||1.45 (1.11–1.78)|
|GOLD classification (stage 2/stage 3)||42/20||14/7||15/6||13/7||3/4|
|Patients with bronchodilator reversibility*||29 (47%)||9 (43%)||9 (43%)||11 (55%)||7 (100%)|
|6MWD, m||482 (460–503)||492 (450–534)||471 (440–502)||482 (445–520)||432 (359–506)|
|SGRQ symptoms score||53.3 (48.1–58.5)||47.7 (37.8–57.6)||49.7 (41.8–57.5)||62.8 (55.2–70.4)||60.8 (48.8–72.9)|
|SGRQ activity score||53.6 (47.9–59.3)||52.8 (42.9–62.8)||48.8 (38.5–59.1)||59.2 (50.0–68.5)||60.7 (45.4–76.0)|
|SGRQ impact score||24.9 (20.7–29.1)||25.2 (17.2–33.3)||20.8 (14.3–27.4)||28.5 (21.6–35.4)||30.4 (12.8–48.1)|
|SGRQ total score||38.1 (33.9–42.3)||37.2 (29.1–45.2)||33.9 (27.3–40.5)||43.3 (36.7–49.8)||44.5 (29.3–59.7)|
|FeNO, ppb†||24.9 (21.8–28.5)||14.0 (12.7–15.5)||24.7 (23.3–26.2)||46.0 (39.8–53.2)||38.5 (21.3–69.7)|
|Eosinophils, %†||2.05 (1.24–3.37)||0.56 (0.31–0.99)||2.78 (1.30–5.91)||4.78 (1.89–12.13)||7.29 (1.69–31.50)|
|Neutrophils, %||69.4 (64.4–74.5)||79.7 (73.8–85.7)||68.7 (61.6–75.9)||60.9 (50.3–71.6)||55.4 (33.4–77.4)|
|Macrophages, %||15.1 (12.7–17.6)||13.8 (9.6–18.1)||16.6 (12.9–20.3)||14.7 (9.8–19.6)||16.4 (7.4–25.3)|
|Lymphocytes, %||0.8 (0.6–1.0)||0.7 (0.3–1.2)||0.9 (0.6–1.2)||0.9 (0.5–1.2)||0.6 (0.2–1.0)|
|Epithelial cells, %||5.2 (3.3–7.2)||4.4 (1.6–7.2)||4.3 (2.1–6.5)||7.0 (2.2–11.7)||3.0 (1.0–4.9)|
Change after Placebo
Change after Prednisone
|6MWD, m||483 (11)||481 (12)||−2 (4)||481 (11)||491 (11)||11 (3)||0.02|
|FEV1, L (postbronchodilator)||1.57 (0.06)||1.57 (0.06)||−0.01 (0.01)||1.56 (0.06)||1.62 (0.06)||0.06 (0.02)||0.02|
|FVC, L (postbronchodilator)||3.22 (0.11)||3.24 (0.11)||0.02 (0.03)||3.24 (0.11)||3.29 (0.10)||0.05 (0.04)||0.57|
|SGRQ total||38.5 (2.1)||39.4 (2.3)||0.9 (0.9)||38.7 (2.1)||37.2 (2.0)||−1.5 (1.2)||0.16|
|SGRQ symptoms||54.6 (2.6)||56.8 (2.7)||2.1 (1.6)||54.2 (2.7)||52.0 (2.6)||−2.2 (1.7)||0.08|
|SGRQ activity||55.9 (2.8)||55.3 (3.1)||−0.7 (1.6)||54.6 (3.0)||54.3 (2.8)||−0.2 (1.8)||0.88|
|SGRQ impacts||23.9 (2.0)||25.4 (2.3)||1.4 (1.1)||25.2 (2.1)||23.3 (2.0)||−1.9 (1.3)||0.06|
|FeNO, ppb*||23.6 (0.03)||24.4 (0.03)||0.97 (0.02)||26.1 (0.03)||19.8 (0.03)||1.32 (0.03)||< 0.001|
|Eosinophils, %*||2.04 (0.10)||2.17 (0.11)||0.94 (0.06)||1.82 (0.11)||0.37 (0.09)||4.94 (0.11)||< 0.001|
|Neutrophils, %||72.0 (2.4)||67.1 (2.7)||−4.9 (2.2)||71.4 (2.6)||78.7 (2.0)||7.3 (2.5)||< 0.01|
|Macrophages, %||15.9 (1.7)||17.5 (1.5)||1.6 (1.6)||14.5 (1.2)||14.8 (1.3)||0.3 (1.4)||0.47|
|Lymphocytes, %||0.92 (0.11)||1.03 (0.14)||0.1 (0.1)||0.78 (0.12)||0.46 (0.07)||−0.3 (0.1)||0.03|
|Epithelials, %||4.19 (0.67)||6.58 (1.24)||2.4 (1.3)||5.81 (1.28)||5.51 (1.36)||−0.3 (1.2)||0.12|
The geometric mean FeNO decreased from 26.1 ppb to 19.8 ppb after prednisone compared with an increase from 23.6 ppb to 24.4 ppb after placebo (P < 0.001). There was a significant decrease in geometric mean sputum eosinophil count after prednisone from 1.8 to 0.4%, compared with an increase from 2.0 to 2.2% after placebo (P < 0.001). There was a significant correlation between off-steroid FeNO and sputum eosinophil percentage (r = 0.46, P < 0.01).
The correlation coefficients for the relationship between baseline FeNO and the primary endpoints were: 6MWD, r = 0.10, P = 0.45; FEV1, r = 0.32, P = 0.01; SGRQ, r = 0.12, P = 0.36). A significant improvement from the lowest to the highest FeNO tertile was observed for FEV1 (P = 0.03) but not for 6MWD or SGRQ (Figure 2 and Table 3). Results for other outcomes are shown in Table 3. The relationships between baseline sputum eosinophils and the primary endpoints were all nonsignificant.
Tertiles by FeNO (ppb)
|< 19.0||19.0–30.2||> 30.2||P Value|
|6MWD, m||+15 (−11 to +41)||+7 (−7 to +22)||+15 (6 to +25)||0.96|
|FEV1, L (postbronchodilator)||−0.01 (−0.09 to +0.07)||+0.07 (−0.01 to +0.15)||+0.12 (+0.04 to +0.21)||0.03|
|FVC, L (postbronchodilator)||−0.07 (−0.22 to +0.09)||−0.01 (−0.19 to +0.16)||+0.18 (−0.04 to +0.40)||0.06|
|SGRQ total||−1.3 (−6.0 to +3.3)||−1.7 (−8.0 to +4.5)||−4.1 (−10.3 to +2.1)||0.50|
|SGRQ symptoms||+0.8 (−8.8 to +10.5)||−7.4 (−14.8 to 0.0)||−6.8 (−14.1 to +0.5)||0.20|
|SGRQ activity||−3.0 (−10.6 to +4.5)||+2.2 (−9.2 to +13.6)||+2.3 (−8.8 to +13.5)||0.46|
|SGRQ impacts||−1.1 (−6.4 to +4.2)||−2.1 (−7.7 to +3.5)||−6.9 (−13.4 to -0.4)||0.17|
|FeNO, ppb*||+1.1 (+0.9 to +1.3)||+1.4 (+1.1 to +1.7)||+1.8 (+1.4 to +2.2)||< 0.01|
|Eosinophils, %*||+1.5 (+0.7 to +3.6)||+6.6 (+2.9 to +15.1)||+12.0 (+3.2 to +45.0)||0.01|
|Neutrophils, %||+6.5 (−3.2 to +16.3)||+4.6 (−5.0 to +14.2)||+19.2 (+10.2 to +28.3)||0.07|
Using ROC analyses, the predictive value of baseline FeNO for an increase of 0.2 L in FEV1 with prednisone was significant (area under the curve [AUC] 0.69, P = 0.04) with an optimum FeNO cut-point of 50 ppb (sensitivity 29%, specificity 96%, positive predictive value 67%, negative predictive value 82%; Table 4). The predictive values of baseline FeNO for an increase in FEV1 of 20% were of even greater significance, with an AUC of 0.80, P less than 0.01 (Table 4). Six patients had a baseline FeNO greater than 50 ppb, and the changes in outcomes for these patients were mean (range): FEV1 0.21 L (range −0.19 to 0.39 L); 6MWD 13.5 m (range −8 to 32 m); SGRQ −7.4 (range −28.7 to 9.7). The predictive values of baseline FeNO for either a 35-m increase in 6MWD or a four-point reduction in SGRQ total score were not significant (AUC for 6MWD: 0.467, P = 0.97; AUC for SGRQ: 0.569, P = 0.38). Somewhat surprisingly, the baseline sputum eosinophil count (%) was not a significant predictor for any of the clinical endpoints (AUC for 35-m increase in 6MWD: 0.49, P = 0.921; AUC for 0.2 L increase in FEV1: 0.63, P = 0.15; AUC for four-point reduction in SGRQ total score: 0.65, P = 0.06).
FeNO Cut-point (ppb)
|Increase in FEV1 of ≥ 0.2 L||25||71||56||32||87||60|
|Increase in FEV1 of ≥20%||25||86||55||19||97||58|
The results of the present study demonstrate that FeNO is a weak predictor of short-term response to oral corticosteroid in patients with stable, moderately severe COPD. Using ROC analyses, the AUC was significant for FEV1 (0.69, P = 0.04) but not for 6MWD or SGRQ. The positive predictive value for the FeNO value with greatest accuracy for predicting an increase in FEV1 (50 ppb) was 67%. Interestingly, this cut-point corresponds very closely with results from other studies in which the optimum cut-point for FeNO to predict steroid response was found to be almost identical (16, 25). More importantly, the negative predictive value of a low (normal) FeNO (< 25 ppb) was 87%, providing evidence that a normal FeNO result helpfully predicts the absence of a response. In the context of treating COPD, in which at best only 20% of patients will demonstrate steroid responsiveness (5, 6), this information would help the clinician to avoid prescribing unnecessary ICS treatment.
Both FeNO and sputum eosinophil percentage were reduced by prednisone treatment (Tables 2 and 3), with the greatest reductions occurring in patients with the highest baseline FENO levels. For eosinophils, there was a 12-fold reduction in patients in the highest FeNO tertile. Similarly, overall there was a statistically significant change in 6MWD and FEV1, but not SGRQ, with prednisone. These outcomes and their relationship to baseline levels of airway inflammation are consistent with those of a previous study (9), and indicate that in COPD, just as in asthma, the eosinophilic component of a disease with mixed airway inflammation is steroid responsive. FeNO measurements, which correlated with percent sputum eosinophils (r = 0.46, P < 0.01) provided a similar but easier-to-obtain perspective. Despite this, the predictive values for FeNO (AUCs) were significant only for change in FEV1 and not for change in 6MWD or SGRQ, reflecting the fact that significant correlations do not necessarily equate with high predictive values. One explanation may be that there is a somewhat poorer relationship between changes in airway inflammation and changes in 6MWD or SGRQ over the short term.
For many years, identifying markers of steroid responsiveness in COPD has been a “holy grail.” Earlier studies explored the value of both bronchodilator reversibility and airway hyperresponsiveness as predictors of steroid responsiveness. Unfortunately, the relationships between these measurements and treatment outcomes are weak, and these tests are unreliable in this setting (26). Clinicians often resort to a trial of steroid in COPD, but here again evidence to justify this strategy is poor (26). In practice, n-of-1 trials are difficult, involving repeat consultations and spirometric measurements, and they are now no longer recommended (2). Empiric therapy is often undertaken and patients may remain on long-term treatments in the absence of confirmatory evidence regarding their efficacy. To date, using a biomarker to identify potential therapeutic responses has not been carefully investigated in COPD. It is theoretically desirable to use a biomarker in any disease state if the biomarker in question (in this case a marker of airway inflammation) reflects the underlying pathology and is responsive to a disease-modifying treatment intervention. Given that this is the case for the relationship between FeNO and eosinophilic inflammation, this is a potentially important advance in the management of airways disease (12). In COPD, only one previous study with FeNO has shown a correlation between baseline FeNO and change in FEV1 after ICS (27), although there is also evidence that an alternative biomarker of airway inflammation, sputum eosinophils, for which FeNO is a surrogate, is associated with steroid responsiveness (8–11). In these studies, however, predictive values, ideally the issue of interest, were not calculated for FeNO or sputum eosinophils.
A potential criticism of our study design is that oral rather than inhaled corticosteroid was used. The relationship between outcomes after a short-term trial of oral steroid and outcomes with long-term inhaled steroid in patients with COPD is not a consistent one (28). Thus, arguably, it would have been more clinically relevant to determine whether FeNO measurements predict the response to long-term ICS treatment. This issue was keenly debated when our study was designed. We came to the conclusion that it was important at this stage to answer the question: “Does FeNO predict response to ICS in COPD if this is at all possible?” By choosing to administer oral prednisone, the confounding effects of variable inhaler technique and airway drug deposition on treatment responses were minimized. Ideally, our study would have involved a sequential trial of oral followed by inhaled steroid in each patient. However, it has been highlighted that despite the overall lack of effect of ICS on COPD outcomes, the results of a short-term trial of oral steroid do indeed have predictive significance (29). In the ISOLDE study, subjects with the greatest increase in FEV1 after prednisolone had the largest reduction in exacerbations during subsequent treatment with inhaled fluticasone (28). This observation provides indirect support for using FeNO as a predictive biomarker, given its usefulness as a predictor of change in FEV1 with prednisone. Furthermore, it has been shown that long-term ICS treatment is more likely to be of benefit in patients whose pretreatment airway inflammation includes a significant eosinophilic component (30). Given that FeNO measurements are a surrogate marker for sputum eosinophil counts, one would therefore speculate that the predictive values for FeNO in relation to long-term outcomes might be equally or even more significant than in the present report. A further study to test this hypothesis would be justified.
The results of our study might have been more definitive had it been possible for all enrolled subjects to enter the randomized phase of the study. Unfortunately, 13 patients were unable to tolerate cessation of ICS during the run-in, resulting in a potential selection bias and underestimation of the beneficial effects of prednisone. We can only speculate that these 13 patients might have had eosinophilic airway inflammation with correspondingly elevated FeNO levels. In other studies of steroid withdrawal in COPD, a similar proportion of patients who are clearly “steroid-requiring” has been identified when treatment is discontinued (31, 32). It is worth noting that the seven patients who withdrew after randomization (of whom six suffered an AECOPD) had marked sputum eosinophilia (7.3%) and elevated FeNO (38.5 ppb) at randomization (Table 1).
FeNO measurements are affected by a number of factors, including current cigarette smoking and, obviously, inhaled corticosteroid use (33). FeNO levels are approximately 30 to 40% lower in current smokers although adjustments may be applied (17). The effect of exposure to corticosteroid may be 4 to 6 weeks in duration. Thus our results are only applicable in patients who are ex-smokers and are currently steroid-free.
In conclusion, we have demonstrated that FeNO measurements in patients with COPD are a predictor for changes in airflow obstruction, but not improvements in functional exercise capacity or health-related quality of life, with corticosteroid therapy. Low FeNO values are highly predictive that improvements in FEV1 are unlikely. Despite the fact that the indications for using ICS in COPD are limited and there are risks of adverse effects (2), they are widely and empirically prescribed, largely because objective data on which rational therapeutic choices may be based are not easily obtainable. Using an appropriate biomarker, such as FeNO, has the potential to improve this situation.
The authors thank Grant Hastie and Sue Weaver for the manufacture and blinding of the trial medications.
|1.||Suissa S, McGhan R, Niewoehner D, Make B. Inhaled corticosteroids in chronic obstructive pulmonary disease. Proc Am Thorac Soc 2007;4:535–542.|
|2.||Rodriguez-Roisin R, Rabe KF, Anzueto A, Bourbeau J, Calverley P, Casas A, deGuia TS, Fukuchi Y, Hui DSC, Jenkins J, et al.; Global Initiative for Chronic Obstructive Lung Disease (GOLD). Global strategy for diagnosis, management, and prevention of chronic obstructive pulmonary disease [Internet]. Bethesda, MD: NHLBI; 2008 (Accessed 2009 May 8). Available from:|
|3.||Drummond MB, Dasenbrook EC, Pitz MW, Murphy DJ, Fan E. Inhaled corticosteroids in patients with stable chronic obstructive pulmonary disease: a systematic review and meta-analysis. JAMA 2008;300:2407–2416.|
|4.||Marsh SE, Travers J, Weatherall M, Williams MV, Aldington S, Shirtcliffe PM, Hansell AL, Nowitz MR, McNaughton AA, Soriano JB, et al. Proportional classifications of COPD phenotypes. Thorax 2008;63:761–767.|
|5.||Weir DC, Burge PS. Effects of high dose inhaled beclomethasone dipropionate, 750 micrograms and 1500 micrograms twice daily, and 40 mg per day oral prednisolone on lung function, symptoms, and bronchial hyperresponsiveness in patients with non-asthmatic chronic airflow obstruction. Thorax 1993;48:309–316.|
|6.||Weir DC, Gove RI, Robertson AS, Burge PS. Corticosteroid trials in non-asthmatic chronic airflow obstruction: a comparison of oral prednisolone and inhaled beclomethasone dipropionate. Thorax 1990;45:112–117.|
|7.||Singh S, Amin AV, Loke YK. Long-term use of inhaled corticosteroids and the risk of pneumonia in chronic obstructive pulmonary disease: a meta-analysis. Arch Intern Med 2009;169:219–229.|
|8.||Brightling CE, McKenna S, Hargadon B, Birring S, Green R, Siva R, Berry M, Parker D, Monteiro W, Pavord ID, et al. Sputum eosinophilia and the short term response to inhaled mometasone in chronic obstructive pulmonary disease. Thorax 2005;60:193–198.|
|9.||Brightling CE, Monteiro W, Ward R, Parker D, Morgan MD, Wardlaw AJ, Pavord ID. Sputum eosinophilia and short-term response to prednisolone in chronic obstructive pulmonary disease: a randomised controlled trial. Lancet 2000;356:1480–1485.|
|10.||Leigh R, Pizzichini MM, Morris MM, Maltais F, Hargreave FE, Pizzichini E. Stable COPD: predicting benefit from high-dose inhaled corticosteroid treatment. Eur Respir J 2006;27:964–971.|
|11.||Pizzichini E, Pizzichini MM, Gibson P, Parameswaran K, Gleich GJ, Berman L, Dolovich J, Hargreave FE. Sputum eosinophilia predicts benefit from prednisone in smokers with chronic obstructive bronchitis. Am J Respir Crit Care Med 1998;158:1511–1517.|
|12.||Pavord ID, Shaw DE, Gibson PG, Taylor DR. Inflammometry to assess airway diseases. Lancet 2008;372:1017–1019.|
|13.||Jatakanon A, Lim S, Kharitonov SA, Chung KF, Barnes PJ. Correlation between exhaled nitric oxide, sputum eosinophils, and methacholine responsiveness in patients with mild asthma. Thorax 1998;53:91–95.|
|14.||Payne DN, Adcock IM, Wilson NM, Oates T, Scallan M, Bush A. Relationship between exhaled nitric oxide and mucosal eosinophilic inflammation in children with difficult asthma, after treatment with oral prednisolone. Am J Respir Crit Care Med 2001;164:1376–1381.|
|15.||Berlyne GS, Parameswaran K, Kamada D, Efthimiadis A, Hargreave FE. A comparison of exhaled nitric oxide and induced sputum as markers of airway inflammation. J Allergy Clin Immunol 2000;106:638–644.|
|16.||Smith AD, Cowan JO, Brassett KP, Filsell S, McLachlan C, Monti-Sheehan G, Peter Herbison G, Robin Taylor D. Exhaled nitric oxide: a predictor of steroid response. Am J Respir Crit Care Med 2005;172:453–459.|
|17.||McSharry CP, McKay IC, Chaudhuri R, Livingston E, Fraser I, Thomson NC. Short and long-term effects of cigarette smoking independently influence exhaled nitric oxide concentration in asthma. J Allergy Clin Immunol 2005;116:88–93.|
|18.||Jones PW, Quirk FH, Baveystock CM, Littlejohns P. A self-complete measure of health status for chronic airflow limitation. The St. George's Respiratory Questionnaire. Am Rev Respir Dis 1992;145:1321–1327.|
|19.||American Thoracic Society. Recommendations for standardized procedures for the on-line and off-line 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;160:2104–2117.|
|20.||American Thoracic Society. ATS statement: guidelines for the six-minute walk test. Am J Respir Crit Care Med 2002;166:111–117.|
|21.||Puhan MA, Mador MJ, Held U, Goldstein R, Guyatt GH, Schunemann HJ. Interpretation of treatment changes in 6-minute walk distance in patients with COPD. Eur Respir J 2008;32:637–643.|
|22.||Cazzola M, MacNee W, Martinez FJ, Rabe KF, Franciosi LG, Barnes PJ, Brusasco V, Burge PS, Calverley PM, Celli BR, et al. Outcomes for COPD pharmacological trials: from lung function to biomarkers. Eur Respir J 2008;31:416–469.|
|23.||Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J 2004;23:932–946.|
|24.||Jones PW. Interpreting thresholds for a clinically significant change in health status in asthma and COPD. Eur Respir J 2002;19:398–404.|
|25.||Pijnenburg MW, Hofhuis W, Hop WC, De Jongste JC. Exhaled nitric oxide predicts asthma relapse in children with clinical asthma remission. Thorax 2005;60:215–218.|
|26.||Yang IA, Fong KM, Sim EHA, Black PN, Lasserson TJ. Inhaled corticosteroids for stable chronic obstructive pulmonary disease (review). Cochrane Db Syst Rev 2007;18:CD002991.|
|27.||Zietkowski Z, Kucharewicz I, Bodzenta-Lukaszyk A. The influence of inhaled corticosteroids on exhaled nitric oxide in stable chronic obstructive pulmonary disease. Respir Med 2005;99:816–824.|
|28.||Burge PS, Calverley PM, Jones PW, Spencer S, Anderson JA. Prednisolone response in patients with chronic obstructive pulmonary disease: results from the ISOLDE study. Thorax 2003;58:654–658.|
|29.||Pavord ID, Siva R, Brightling CE. Prednisolone response in patients with COPD. Thorax 2004;59:179. [Author reply.]|
|30.||Siva R, Green RH, Brightling CE, Shelley M, Hargadon B, McKenna S, Monteiro W, Berry M, Parker D, Wardlaw AJ, et al. Eosinophilic airway inflammation and exacerbations of COPD: a randomised controlled trial. Eur Respir J 2007;29:906–913.|
|31.||O'Brien A, Russo-Magno P, Karki A, Hiranniramol S, Hardin M, Kaszuba M, Sherman C, Rounds S. Effects of withdrawal of inhaled steroids in men with severe irreversible airflow obstruction. Am J Respir Crit Care Med 2001;164:365–371.|
|32.||van der Valk P, Monninkhof E, van der Palen J, Zielhuis G, van Herwaarden C. Effect of discontinuation of inhaled corticosteroids in patients with chronic obstructive pulmonary disease: the COPE study. Am J Respir Crit Care Med 2002;166:1358–1363.|
|33.||Taylor DR, Pijnenburg MW, Smith AD, De Jongste JC. Exhaled nitric oxide measurements: clinical application and interpretation. Thorax 2006;61:817–827.|