Abstract
Background High exhaled nitric oxide fraction (FENO) levels are associated with greater risk of asthma exacerbation. However, it is not clear how FENO can be used to guide safe reductions in inhaled corticosteroid (ICS) doses in asthma patients. This study assesses the ability of FENO to guide ICS reductions.
Methods Systematic searching of electronic databases identified prospective observational studies and randomised controlled trials which recruited participants with mild-to-moderate asthma aged ≥12 years and measured FENO before reducing ICS. We performed multilevel mixed-effects logistic regression in relation to acute exacerbations and estimated each participant's exacerbation risk using our logistic regression model.
Results We included data from seven out of eight eligible studies, representing 384 participants. ICS doses were halved in four studies and withdrawn in three studies. A baseline FENO measurement of ≥50 ppb was associated with increased risk of exacerbations (crude OR 3.14, 95% CI 1.41–7.00, p=0.005; adjusted OR 3.08, 95% CI 1.36–6.98, p=0.007) and corresponded to an estimated exacerbation risk cut-off of 15%. Reducing ICS when estimated exacerbation risk was <15% versus <10% would result in fewer patients remaining on the same ICS dose (40 (10.4%) out of 384 versus 141 (36.7%) out of 384), but similar proportions of patients avoiding exacerbations (222 (91.4%) out of 243, 95% CI 87.1–94.6% versus 311 (90.4%) out of 344, 95% CI 86.8–93.3%).
Conclusion In patients with mild-to-moderate asthma, gradual ICS reduction when FENO is <50 ppb may help decrease ICS use without increasing exacerbations. Future research should aim to validate these findings in larger populations.
Abstract
In nonsmoking patients with mild-to-moderate well-controlled asthma, stepping down treatment when FENO is <50 ppb reduces prescribing of inhaled corticosteroids without increasing exacerbations http://bit.ly/2SKaxSt
Introduction
Asthma is a complex condition characterised by marked variation in underlying pathophysiology and response to treatment. Inhaled corticosteroids (ICS) are considered to be the mainstay of treatment to prevent acute asthma exacerbations. However, only asthma driven by the type 2 pathway of airway inflammation is likely to respond to ICS [1].
Approximately 80% of UK asthma patients receive ICS [2]. However, despite an increase of almost one-third in ICS prescribing costs between 2008 and 2018 [3], asthma-related hospitalisation and exacerbation rates have not improved [2, 4]. Unnecessary ICS treatment can increase risk of local and systemic side-effects [5, 6]; therefore, clinical guidelines recommend that clinicians should consider gradually reducing treatment when patients have been stable for ≥3 months [7].
Systematic reviews of randomised controlled trials (RCTs) comparing ICS dose maintenance with ICS reduction or withdrawal have reported inconsistent findings. One review found that stopping ICS in stable asthma patients was associated with significantly increased risk of exacerbations [8]. However, another concluded that reducing ICS by ≥50% did not significantly increase exacerbation risk [9]. A Cochrane review [10] did not find differences in exacerbations between adults whose ICS dose was maintained versus reduced by 50–60%, but found that data were limited and of low quality.
The type 2 pathway of airway inflammation is driven by a range of cytokines, including interleukin (IL)-4, IL-13 and IL-5. IL-13 in particular has been described as having key roles in many aspects of asthma pathogenesis including mucus hypersecretion, goblet cell hyperplasia, subepithelial fibrosis and airway hyperresponsiveness [11, 12]. In addition, IL-13 activates inducible nitric oxide synthase, leading to increased production of nitric oxide in the airway [13]. Therefore, exhaled nitric oxide fraction (FENO) is an indicator of IL-13-driven corticosteroid-responsive airway inflammation. However, a systematic review and qualitative synthesis found insufficient evidence to determine the ability of FENO to guide ICS reductions [14].
A meta-analysis of data from RCTs which compared FENO-guided monitoring with guideline- or symptom-based monitoring found a significant reduction in exacerbations, but no difference in final daily ICS dose [15]. However, the clinical implications of this remain uncertain due to differences between studies in FENO cut-offs for adjusting treatment, characteristics of FENO-guided management algorithms and definition of exacerbations. An individual patient data meta-analysis can address these issues by using a consistent definition of exacerbations across all studies and analysing data from participants in whom treatment changes were not guided by prespecified cut-offs. Our study examines the ability of FENO to guide safe ICS reductions in asthma patients.
Methods
Our study protocol is registered on the PROSPERO International Prospective Register of Systematic Reviews (www.crd.york.ac.uk/PROSPERO/; identifier number CRD42017071826).
Data sources and study selection
We performed systematic electronic database searches of MEDLINE and MEDLINE In-Process (OvidSP) [1946–], Embase (OvidSP) [1974–], Cochrane Central Register of Controlled Trials (Cochrane Library, John Wiley & Sons) and Web of Science Core Databases (Web of Science, Thomson Reuters) until September 13, 2019 with no language restrictions. Appendix 1 summarises our MEDLINE search strategy. In addition, we searched clinicaltrials.gov and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (http://apps.who.int/trialsearch/Default.aspx) to identify trial protocols. To avoid omission of relevant studies, we screened the reference lists of included studies and relevant systematic and narrative reviews and consulted experts in the field.
One reviewer (KW) screened article titles to exclude studies that were obviously irrelevant. Two reviewers (KW, JV) independently assessed abstracts and full-text articles for eligibility and resolved any disagreements by discussion.
We included prospective observational studies or RCTs conducted in community healthcare settings which recruited participants aged ≥12 years with clinician-diagnosed asthma treated with low- or moderate-dose ICS (i.e. <1000 μg fluticasone propionate equivalents per day [7]), and measured FENO before reducing ICS. We excluded studies that recruited highly selected populations (e.g. occupational asthma), stepped down ICS only in individuals whose FENO measurements were below predefined decision thresholds, introduced alternative treatments after stepping down ICS, or did not collect data on acute exacerbations.
We defined acute exacerbations as acute asthma-related episodes requiring systemic corticosteroids or antibiotics, hospital admission or unscheduled healthcare visits due to asthma during the 12-week period after stepping down ICS. Study authors were asked to provide data on these outcomes as separate variables where possible. However, composite outcome data on acute exacerbations were accepted if definitions of these were consistent with our study definition.
Data extraction and quality assessment
Authors whose studies met eligibility criteria were approached for provision of individual patient data including baseline characteristics (age, sex, smoking status, body mass index, atopy), baseline ICS dose, FENO before ICS dose was stepped down, and acute exacerbations. Original individual patient data were kept on a secure server and prepared in a consistent format for all studies.
Study quality was independently assessed by two reviewers (JB and AF-N) using a modified version of the Quality in Prognosis Studies tool [16] (appendix 2). Discrepancies were resolved by discussion involving a third reviewer (KW). Study-level data were summarised and compared with published findings. Study authors were contacted for assistance with clarifying any discrepancies identified.
Data analysis
We conducted a one-stage individual patient data meta-analysis using mixed effects multilevel logistic regression using the melogit command in STATA (version 14; StataCorp, College Station, TX, USA). Our primary outcome was presence of one or more acute exacerbations during the 12-week period after stepping down ICS. Our secondary outcome was acute exacerbations requiring systemic corticosteroids. Our regression models accounted for within-study clustering and included age, sex and FENO as baseline covariates.
FENO was classified as low (≤20 ppb), intermediate (>20–<50 ppb) or high (≥50 ppb), thus incorporating cut-offs recommended for children and adults by the American Thoracic Society guidelines [17]. We assessed model performance by calculating C-statistics with 95% confidence intervals and used our regression equation to estimate acute exacerbation risk for each participant. We set hypothetical decision thresholds based on the distribution of estimated exacerbation risks and summarised percentages of clinical decisions which were appropriate according to each threshold with 95% confidence intervals.
We conducted prespecified subgroup analyses according to age (12–59 years inclusive versus ≥60 years), smoking status and baseline ICS dose before stepping down treatment. Baseline ICS dose was classified as below, within or above the range associated with greatest therapeutic benefit (100–250 μg fluticasone propionate equivalents per day) [18].
Results
Study selection
Figure 1 summarises the results of our electronic database search, which retrieved 13 757 articles excluding duplicates and animal studies. We assessed 29 full-text articles, of which eight were suitable for inclusion [19–26]. The authors of four studies provided individual patient data for all participants who met our eligibility criteria [20, 23, 24, 26]. The authors of one study [25] were unable to provide individual patient data from participants recruited at three recruitment centres due to internal reasons, but provided individual patient data for all other participants. Data from two studies [21, 22] were obtained from the National Heart, Lung, and Blood Institute Biologic Specimen and Data Repository Information Coordinating Center repository. The authors of one study [19] declined to provide individual patient data.
Study selection. FENO: exhaled nitric oxide fraction; ICS: inhaled corticosteroids.
Another study [27], which involved ICS reductions at 8-week intervals was considered, but not included, because its method of measuring FENO resulted in values which were not comparable to those obtained by other included studies, and we were unable to identify a method for converting these values to measurements which would have been comparable.
Overview of studies that provided individual patient data
Studies whose authors agreed to provide individual patient data were published between 2001 and 2016. Table 1 summarises the characteristics of our included studies.
Characteristics of included studies
Three studies were conducted in outpatient clinics [20, 24, 25] and one in primary care [26]. The other three studies were conducted in academic centres and the surrounding community [21–23]. Four studies included patients with asthma diagnosed according to Global Initiative for Asthma (GINA) [20, 24, 25] or American Thoracic Society [21] guidelines. One study included primary care patients with a recorded diagnosis of asthma [26]. Other studies diagnosed asthma based on symptoms [23] or spirometry and airway hyperresponsiveness [22].
We included data from two prospective observational studies which halved ICS doses in all participants [25, 26] and two open-label RCTs [20, 24] which included groups in which ICS dose was halved. In two studies, ICS and long-acting β2-agonist (LABA) doses were halved [24, 25]. Additionally, we included data from the placebo groups of three RCTs which involved an initial ICS treatment period followed by use of placebo inhalers [21–23]. Furthermore, we included data from the “rescue beclometasone” group of one RCT [23], whereby ICS dose was reduced from regular use to rescue use only.
Four studies included participants whose asthma had been well controlled for ≥3 months [20, 24–26]. Three RCTs included participants whose asthma had been well controlled for 6 weeks [21, 22] or 4 weeks [23] and included some participants who had not previously used controller treatment. Two studies defined well-controlled asthma based on GINA criteria [24, 25], one based on US National Asthma Education and Prevention Program asthma care guidelines [23], and three according to clinical criteria [20, 21, 26]. One study did not define well-controlled asthma [22], but withdrew participants who experienced a significant exacerbation during the initial ICS treatment phase.
Two studies provided individual patient data on exacerbations as a single variable [24, 25]. Other included studies provided exacerbation data as separate variables for exacerbations requiring systemic corticosteroids [20–23, 26], hospital admissions [20, 21], unscheduled healthcare visits [21] or antibiotics [26].
Risk of bias
Risk of bias was generally low for study participation, study attrition and measurement of acute exacerbations (table 2). Risk of bias for study participation was only considered high in one study [21], which only collected baseline FENO measurements in 26 (46%) out of 56 participants in the placebo group.
Risk of bias
The findings of one study [25] were felt to be at high risk of confounding, as the analysis did not stratify or adjust for differences in baseline characteristics between participants with high (≥37 ppb) or low (<37 ppb) FENO. Another study stated that differences in FENO and clinical measurements between participants who remained stable versus those who had exacerbations after reducing ICS were not statistically significant, but did not present findings to substantiate this [26]. Therefore, risk of bias was considered to be unclear in relation to confounding, statistical analysis and reporting. No important issues were identified in checking individual patient data.
Characteristics of included participants
Individual patient data were provided for 417 participants, out of which 33 (7.9%) were excluded from our dataset due to missing data on baseline FENO measurements before ICS treatment was stepped down (n=30 [21], n=2 [26], n=1 [24]). Among the 384 participants included in our dataset, 43 (11.2%) had an acute exacerbation within 12 weeks of stepping down their ICS dose (n=0 [20], n=3 [21], n=2 [22], n=6 [23], n=7 [24], n=1 [25], n=24 [26]).
Table 3 summarises baseline participant characteristics. Approximately three-quarters of participants had never smoked and nearly two-thirds had a history of atopy. FENO was ≤20 ppb in approximately half of participants.
Participant characteristics
Multilevel logistic regression analysis
Risk of exacerbation after stepping down ICS was significantly greater in participants with high versus low FENO after adjustment for within-study clustering (crude OR 2.70, 95% CI 1.16–6.26; p=0.021). This finding was still statistically significant after further adjustment for age and sex (adjusted OR 2.56, 95% CI 1.08–6.09; p=0.033).
No significant differences in exacerbation risk were demonstrated between participants with intermediate versus low FENO (crude OR 0.67, 95% CI 0.31–1.43, p=0.299; adjusted OR 0.64, 95% CI 0.29–1.37, p=0.248). However, risk of exacerbation was significantly greater in participants with high versus intermediate FENO (crude OR 4.03, 95% CI 1.57–10.4, p=0.004; adjusted OR 4.03, 95% CI 1.55–10.5, p=0.004) and in participants with high versus low or intermediate FENO (crude OR 3.14, 95% CI 1.41–7.00, p=0.005; adjusted OR 3.08, 95% CI 1.36–6.98, p=0.007).
Our crude model (adjusted for within-study clustering and FENO) performed modestly in relation to predicting acute exacerbations (C-statistic 0.61, 95% CI 0.53–0.70). Additional adjustment for age and sex as baseline covariates did not significantly improve model performance (C-statistic 0.65, 95% CI 0.56–0.74).
The distribution of estimated exacerbation risks calculated for each participant using our regression model (linear predictor = −2.417 + (0.0008 × age (years) + (0.623 if male) + (−0.453 if FENO >20 ppb and <50 ppb) + (0.941 if FENO ≥50 ppb)) was positively skewed (median 8.7%, range 5.4–31.1%). Based on this distribution, we set hypothetical decision thresholds of 10%, 15% and 20% in relation to estimated risk of exacerbations and summarised the numbers and percentages of clinical decisions which were appropriate at each of these thresholds with 95% confidence intervals. Based on our model, similar percentages of participants would not have an exacerbation after stepping down ICS, or potentially avoid an exacerbation as a result of their ICS dose being maintained, irrespective of which risk of exacerbation threshold was used to guide these decisions (figure 2, appendix 3). However, the same ICS dose would be continued in fewer participants if ICS was stepped down in patients whose estimated risk was <15% (40 (10.4%) out of 384, 95% CI 7.5–13.9%) versus <10% (141 (36.7%) out of 384, 95% CI 31.9–41.8%). A FENO cut-off of 50 ppb corresponded to an estimated exacerbation risk cut-off of 15% (table 4).
Appropriateness of clinical decisions guided by different risk of exacerbation thresholds. Appropriate step down of inhaled corticosteroids (ICS) was defined as reductions in ICS dose in patients whose estimated risk of exacerbation was below a given threshold and in whom no acute exacerbation was observed during the following 12-week period. Appropriate continuation of ICS was defined as maintaining the same ICS dose in a patient whose estimated risk of exacerbation was at or above a given threshold, thereby potentially preventing an acute exacerbation during the following 12-week period.
Distribution of exhaled nitric oxide fraction (FENO) measurements according to estimated risk of exacerbation
We were unable to perform our planned secondary outcome analysis because we were only able to identify 14 acute exacerbations requiring systemic corticosteroids (n=0 [20], n=3 [21], n=2 [22], n=6 [23], n=3 [26]).
Subgroup analyses
Table 5 summarises the results of our prespecified subgroup analyses. Due to limited numbers of participants with exacerbations in each subgroup, only crude odds ratios with 95% confidence intervals were calculated. Exacerbation risk was significantly higher in individuals with high versus intermediate FENO in participants who had never smoked (crude OR 3.71, 95% CI 1.15–11.9; p=0.028) and participants aged ≥60 years (crude OR 10.2, 95% CI 1.76–59.2; p=0.010). In addition, exacerbation risk was significantly higher in individuals with high versus low FENO (crude OR 3.93, 95% CI 1.15–13.5; p=0.029) and high versus intermediate FENO (crude OR 5.85, 95% CI 1.38–24.8; p=0.016) in the subgroup whose ICS dose was above the optimal therapeutic range.
Prespecified subgroup analyses for exhaled nitric oxide fraction (FENO) in relation to acute exacerbations of asthma after stepping down inhaled corticosteroids
Post hoc analyses
Appendix 4 summarises the results of two post hoc analyses. Our first post hoc analysis excluded data from participants in the study by Wilson et al. [26], which contributed 189 (49.2%) out of 384 participants in our dataset including 24 (55.8%) out of 43 participants who had an acute exacerbation. After excluding these data, exacerbation risk was still significantly higher in individuals with high versus intermediate FENO (crude OR 3.75, 95% CI 1.06–13.2; p=0.040). However, the same comparison was no longer statistically significant when only data from participants from Wilson et al. [26] were included (crude OR 4.19, 95% CI 0.99–17.8; p=0.051). Our second post hoc analysis was stratified according to whether ICS treatment was halved versus withdrawn or reduced to as-needed use. In the withdrawn subgroup, exacerbation risk was significantly higher in individuals with high versus low or intermediate FENO, but crude OR estimates were associated with very wide 95% confidence intervals since exacerbations were observed in only 11 out of 108 participants.
We examined LABA use as an effect modifier by performing an additional post hoc sensitivity analysis excluding data from studies whose participants were on ICS/LABA preparations [24, 25]. The findings of this analysis were consistent with those of our main analysis. Risk of exacerbation after stepping down ICS was significantly greater in participants with high versus low FENO (adjusted OR 2.94, 95% CI 1.10–7.82; p=0.031), intermediate FENO (adjusted OR 4.00, 95% CI 1.37–11.7; p=0.011) and intermediate or low FENO (adjusted OR 3.32, 95% CI 1.31–8.39; p=0.011). However, no statistically significant difference in risk was observed between participants with intermediate versus low FENO (adjusted OR 0.73, 95% CI 0.32–1.71; p=0.474).
Discussion
Summary of main findings
In nonsmoking patients with mild-to-moderate well-controlled asthma, stepping down ICS when FENO is <50 ppb on existing treatment is associated with reduced ICS prescribing without increasing exacerbations in the following 12 weeks. A FENO cut-off of 50 ppb corresponds to an estimated exacerbation risk of 15%.
Comparison with existing literature
Previous study findings suggest lower FENO cut-offs to guide ICS dose reductions. In children with stable asthma, a FENO of ≥22 ppb is a significant predictor of future exacerbations [28]. Additionally, a study in adults with well-controlled moderate asthma concluded that halving ICS/LABA doses was safe in patients with FENO of ≤28 ppb since no significant differences in numbers of exacerbations were observed before and after reducing treatment [29].
Conversely, a longitudinal analysis found that a higher baseline FENO was associated with lower risk of severe exacerbations in adults with mild-to-moderate asthma [30]. However, it was not possible to determine whether exacerbations were due to ICS dose reductions or other causes because the follow-up period was long (≥12 months) and only approximately three-quarters of participants were using ICS at baseline.
Strengths and limitations
We managed to obtain individual patient data from seven out of eight relevant studies. Our findings are unlikely to have been unduly influenced by omission of the study whose authors declined to provide individual patient data [19], which only included 32 children aged 7–14 years.
Data on baseline FENO measurements were missing from <8% of participants, justifying a complete case analysis, and our regression models accounted for potential within-study clustering of exacerbation outcomes, which may have occurred due to differences between study populations in baseline FENO measurements and other potential prognostic factors. We acknowledge that we were only able to include data from a relatively small number of patients. However, the generalisability of our findings to patients with clinician-diagnosed asthma is strengthened by our inclusion of studies which used pragmatic approaches to diagnosing asthma, thereby reflecting current clinical practice for establishing asthma diagnoses in community and outpatient healthcare settings.
For pragmatic reasons, we examined the predictive value of clinic FENO measurements but were unable to account for factors which may have affected these, including eosinophilic chronic rhinosinusitis [31], ICS adherence [32] and recent allergen exposure [33]. In addition, did not have sufficient data to examine FENO measurement process or device as a potential confounder, as only two studies used Sievers rather than Aerocrine analysers [20, 21] and only three participants across these two studies had one or more exacerbations. All except one study [24] excluded patients with recent exacerbations. Data on other risk factors, such as seasonal exacerbation patterns [34, 35], raised blood eosinophil counts [36] and duration of symptom stability before reducing medication [37] were not available.
We chose a 12-week follow-up period based on guidelines for identifying patients in whom ICS reductions should be considered [7]. While extending this period may have provided a more complete reflection of asthma stability, this may also have resulted in inclusion of later exacerbations due to reasons unrelated to reducing ICS, for example acute respiratory tract infections [38] or poor medication adherence [39]. Additionally, most exacerbations in asthma symptoms occur within 28 days of reducing ICS [40, 41]. We did not examine the value of FENO in predicting single versus multiple exacerbations, as UK primary care data suggest that only 3–5% of asthma patients experience multiple exacerbations over a 1-year period [36, 42].
The findings of our subgroup and post hoc analyses should be interpreted with caution. Due to low numbers of acute exacerbations observed in the groups analysed, only crude odds ratios could be calculated; the 95% confidence intervals associated with these were too wide to be able to draw robust conclusions about their clinical importance. Furthermore, we were unable to estimate pooled sensitivities and specificities in relation to specific FENO thresholds because one study had no exacerbation events [20].
Implications of findings for clinical practice and future research
Our findings provide evidence to support safe gradual ICS reductions in nonsmoking patients with well-controlled mild-to-moderate asthma if FENO is <50 ppb. However, successful maintenance of these reductions should also address broader patient-level factors, including chronic upper airway complications and medication adherence [43], and involve ongoing monitoring of FENO levels before further reductions are considered.
Future research should aim to validate our findings using longer follow-up periods and larger populations, which will allow more robust evaluation of the predictive value of FENO in relevant subgroups. In particular, RCT evidence is needed to compare the safety and effectiveness of different FENO decision thresholds and ICS dose reduction strategies. More research is also needed to determine the value of FENO in predicting single versus multiple exacerbations, as well as the predictive value of follow-up FENO measurements. Variability in FENO is reported around the time of exacerbations [44]. FENO at 4 weeks after withdrawing ICS is also reported to be a strong predictor of deterioration in exacerbation frequency [45].
The predictive value of FENO alongside other biomarkers of airway inflammation should also be investigated. Concurrently raised FENO and blood eosinophil counts are associated with increased exacerbation risk in adults with asthma [46]. Conversely, low peak expiratory flow rate variability, high serum IL-10 and low serum IL-33 are reported as significant predictors of good symptomatic control in adults with asthma following a 50% reduction in ICS [47].
Conclusions
Clinicians may consider gradual reduction of ICS in nonsmoking, well-controlled, mild-to-moderate asthma patients whose FENO is <50 ppb on existing treatment. Future research should aim to prospectively validate these findings in larger study populations and explore ways of enhancing the value of FENO in predicting future exacerbations.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Appendix_1 ERJ-02150-2019_Appendix_1
Appendix_2 ERJ-02150-2019_Appendix_2
Appendix_3 ERJ-02150-2019_Appendix_3
Appendix_4 ERJ-02150-2019_Appendix_4
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-02150-2019.Shareable
Acknowledgements
The views expressed are those of the authors and not necessarily those of the National Health Service, the National Institute for Health Research or the Department of Health and Social Care. This manuscript was prepared using PRICE, SOCS/SLIC and TREXA Research Materials obtained from the National Heart, Lung, and Blood Institute Asthma Clinical Research Network and Childhood Asthma Research and Education Network. The authors would like to thank Dave Mauger (Penn State University, Hershey, PA, USA) and Vernon Chinchilli (Penn State College of Medicine, Hershey, PA) for their helpful comments on the manuscript and assistance with identifying and obtaining individual participant data for our meta-analysis.
Footnotes
This article has an editorial commentary: https://doi.org/10.1183/13993003.01319-2020
This article has supplementary material available from erj.ersjournals.com
Ethical approval: Overall ethical approval was not required as this study did not require use of patient identifiers. Collaborating groups obtained individual approvals for data sharing from Research Ethics Committees at Juntendo University, Kinki University Hospital, Keio University Hospital, Kagoshima University Medical and Dental Hospital and Shizuoka General Hospital.
Conflict of interest: K. Wang reports grants from National Institute for Health Research, during the conduct of the study.
Conflict of interest: J.Y. Verbakel has nothing to disclose.
Conflict of interest: J. Oke has nothing to disclose.
Conflict of interest: A. Fleming-Nouri has nothing to disclose.
Conflict of interest: J. Brewin has nothing to disclose.
Conflict of interest: N. Roberts has nothing to disclose.
Conflict of interest: N. Harada reports personal fees from AstraZeneca, outside the submitted work. In addition, N. Harada has a patent Japanese Patent Application 2018-097070 pending.
Conflict of interest: R. Atsuta has nothing to disclose.
Conflict of interest: K. Takahashi reports grants and personal fees from Chugai Pharmaceutical Co., Ltd, grants and personal fees from Nippon Boehringer Ingelheim Co., Ltd, grants and personal fees from MSD K.K., grants from GlaxoSmithKline Consumer Healthcare Japan K.K., grants from Nippon Shinyaku Co., Ltd, grants from Tsumura & Co., grants and personal fees from Pfizer Inc., grants and personal fees from AstraZeneca K.K., grants and personal fees from Taiho Pharmaceutical Co., Ltd, grants from Astellas Pharma Inc., grants and personal fees from Kyorin Pharmaceutical Co., Ltd, grants from Kyowa Hakko Kirin Co., Ltd, grants and personal fees from Teijin Pharma Limited, grants from Mochida Pharmaceutical Co., Ltd, grants from Toyama Chemical Co., Ltd, grants from Sanofi K.K., grants and personal fees from Ono Pharmaceutical Co., Ltd, grants and personal fees from Nobelpharma Co., Ltd, grants and personal fees from Novartis Pharma K.K., grants from Shionogi & Co., Ltd, grants and personal fees from Eli Lilly Japan K.K., grants from Nipro Corporation, grants from Torii Pharmaceutical Co., Ltd, grants from MiZ Company Limited, personal fees from Sumitomo Dainippon Pharma Co., Ltd, personal fees from Bristol-Myers K.K., personal fees from Meiji Seika Pharma Co., Ltd, personal fees from Otsuka Pharmaceutical Co., Ltd, personal fees from Parexel International Corporation, personal fees from Eisai Co., Ltd, personal fees from Mitsubishi Tanabe Pharma, outside the submitted work.
Conflict of interest: K. Mori has nothing to disclose.
Conflict of interest: T. Fujisawa has nothing to disclose.
Conflict of interest: T. Shirai has nothing to disclose.
Conflict of interest: T. Kawayama has nothing to disclose.
Conflict of interest: H. Inoue reports grants from Astellas, AstraZeneca, Boehringer-Ingelheim, Chugai Pharm, GlaxoSmithKline, Pfizer, Merck Sharp & Dohme, Novartis, Teijin-Pharma, personal fees from Astellas, AstraZeneca, Boehringer-Ingelheim, Chugai Pharm, GlaxoSmithKline, Kyorin, Merck Sharp & Dohme, Meiji Seika Pharma, Novartis, Otsuka, Pfizer, Taiho, outside the submitted work.
Conflict of interest: S. Lazarus reports grants from NIH/NHLBI, during the conduct of the study; grants from NIH/NHLBI, grants from American Lung Association – Airway Clinical Research Centers Network (ALA-ACRC), outside the submitted work.
Conflict of interest: S. Szefler reports other from Boehringer-Ingelheim, other from Genentech, other from GlaxoSmithKline, other from AstraZeneca, other from Daiichi Sankyo, grants from GlaxoSmithKline, other from Propeller Health, other from Sanofi, other from Regeneron, outside the submitted work.
Conflict of interest: F. Martinez reports grants from NIH/NHLBI, grants from NIH/NIEHS, grants from NIH/NIAID, grants from NIH/Office of Director, grants from Johnson & Johnson, personal fees from Copeval, personal fees from Commense Inc, outside the submitted work.
Conflict of interest: D. Shaw reports personal fees from AstraZeneca, GSK, TEVA and Novartis, outside the submitted work.
Conflict of interest: I.D. Pavord reports personal fees from AstraZeneca, personal fees from Boehringer Ingelheim, personal fees from Aerocrine, personal fees from Almirall, personal fees from Novartis, personal fees from GlaxoSmithKline, personal fees from Genentech, personal fees from Regeneron, speakers’ fees from Teva, speakers’ fees from Chiesi, advisory board fees from Sanofi, advisory board fees from Circassia, advisory board fees from Knopp, grants from NIHR, outside the submitted work.
Conflict of interest: M. Thomas reports personal fees from GSK, personal fees from Novartis, personal fees from Boehringer Ingelheim, outside the submitted work; and recent membership of the BTS SIGN Asthma guideline steering group and the NICE Asthma Diagnosis and Monitoring guideline development group.
Support statement: This paper presents independent research funded by the National Institute for Health Research under its Postdoctoral Fellowship Programme (grant reference number PDF-2016-09-007). Funding information for this article has been deposited with the Crossref Funder Registry.
- Received November 5, 2019.
- Accepted February 13, 2020.
- Copyright ©ERS 2020
References
