Abstract
While asthma is known to be associated with an increased risk of progressive lung function impairments and fixed airflow obstruction, there is ongoing debate on whether inhaled corticosteroids (ICS) modify these long-term risks.
Searches were performed of the PubMed, Embase and CENTRAL databases up to 22 July 2019 for studies with follow-up ≥1 year that investigated the effects of maintenance ICS on changes in lung function in asthma.
Inclusion criteria were met by 13 randomised controlled trials (RCTs) (n=11 678) and 11 observational studies (n=3720). Median (interquartile range) follow-up was 1.0 (1–4) and 8.4 (3–28) years, respectively. In the RCTs, predominantly in individuals with mild asthma, ICS use was associated with improved pre-bronchodilator (BD) forced expiratory volume in 1 s (FEV1) across all age groups (2.22% predicted (95% CI 1.32–3.12), n=8332), with similar estimates of strength in association for children and adults. Improvements in post-BD FEV1 were observed in adults (1.54% (0.87–2.21), n=3970), but not in children (0.20% (−0.49–0.90), n=3924) (subgroup difference, p=0.006). Estimates were similar between smokers and nonsmokers. There were no RCT data on incidence of fixed airflow obstruction. In the observational studies, ICS use was associated with improved pre-BD FEV1 in children and adults. There were limited observational data for post-BD outcomes.
In patients with mild asthma, maintenance ICS are associated with modest, age-dependent improvements in long-term lung function, representing an added benefit to the broader clinical actions of ICS in asthma. There is currently insufficient evidence to determine whether treatment reduces incidence of fixed airflow obstruction in later life.
Abstract
In patients with mild asthma, maintenance inhaled corticosteroids are associated with modest age-dependent improvements in long-term lung function https://bit.ly/32K2ZTF
Introduction
Asthma is a common global disease, affecting over 350 million people, for which many patients are treated using inhaled corticosteroids (ICS). While the benefits of such treatment have been demonstrated in terms of asthma control, hospital admissions and mortality, [1–4], the long-term effects of ICS in asthma on changes in lung function remains unclear. This information is necessary to guide optimal asthma management, particularly for patients with early-onset persistent disease who are the greatest risk of progressive lung function impairments [5].
Recent studies have shown that several lung function trajectories lead to fixed airflow obstruction [6, 7]. These include impaired lung growth in the first two to three decades of life and accelerated lung function decline from the third decade onwards [6]. Depending on age of onset, asthma is now known to contribute to both pathways [8, 9]. Given the rising prevalence and global impact of obstructive airways diseases, strategies aimed at preserving lung function in asthma are urgently needed. There is also growing international interest in this area, including a European Respiratory Society clinical initiative, CADSET (Chronic Airway Diseases Early Stratification), which aims to explore the factors that influence lung function trajectories over the lifespan [10].
ICS are currently recommended as a first-line option in persistent asthma [11], and are associated with improved airway calibre through a reduction in airway inflammation, mucus hypersecretion and bronchial hyperresponsiveness [12]. To date, the available evidence on their long-term effects on lung function in asthma have not been adequately summarised. Updates to the Global Initiative for Asthma (GINA) recommendations in 2019 [13], which now strongly advocate for ICS-containing treatment in all patients with mild asthma, further increase the need to critically evaluate the evidence on this topic. Therefore, we aimed to systematically review all studies which have investigated the long-term effects of ICS on lung function in asthma and to undertake a meta-analysis where similar data were available.
Methods
Search strategy
We conducted systematic searches of the PubMed, EMBASE and CENTRAL databases from inception to 22 July 2019 for peer-reviewed studies. For full details refer to table e1. Reference lists of related review articles were manually screened for additional studies meeting our inclusion criteria.
Eligibility criteria
We included randomised controlled trials (RCTs) and observational studies with ≥1year follow-up in which the effects of maintenance ICS on change in lung function (growth or decline) and risk of fixed airflow obstruction was assessed. Studies involving either children or adults with current asthma were included. Studies were required to have an appropriate placebo or control comparison group. Inclusion was limited to English-language studies published in full text. Studies with concomitant use of maintenance systemic corticosteroids were excluded.
A range of definitions of asthma were accepted, including physician-diagnosed asthma, spirometrically defined asthma and survey-reported asthma. Current asthma was defined as: asthma symptoms or asthma-related healthcare utilisation within the last 12 months.
Outcomes of interest
There were two outcomes of interest. 1) Change in lung function from baseline defined as: change in forced expiratory volume in 1 s (FEV1), FEV1/forced vital capacity (FVC), forced expiratory flow at 25–75% of FVC (FEF25–75%) or bronchial hyperresponsiveness (BHR). 2) Fixed airflow obstruction defined as post-bronchodilator (BD) FEV1/FVC ratio <0.7 or less than the lower limit of normal.
Selection of studies
All studies identified in the search strategy were independently screened by two review authors (D.J. Tan and either D.S. Bui or X. Dai). Full texts of all studies considered eligible, potentially eligible or unclear were retrieved and assessed. Disagreements at this stage or further on were settled by consultation with a third author (D.S. Bui or X. Dai).
Data extraction
Using a standardised data extraction form, study characteristics and outcome data were extracted independently by two review authors (D.J. Tan and either D.S. Bui or X. Dai). Information extracted included the first author, date of publication, study design, study setting, date of study, number of participants, mean age, age range, sex, asthma definition, asthma severity, smoking history, ICS use (type, frequency, dose and duration), outcome definitions, confounders and interactions, effect estimates and 95% confidence intervals.
Quality assessment and risk of bias
Risk of bias was assessed independently by two review authors (D.J. Tan and either D.S. Bui or X. Dai) using the Cochrane Collaboration's Tool for individual RCTs, modified Newcastle–Ottowa scale (NOS) for individual observational studies [14], and GRADE (Grading of Recommendations Assessment, Development and Evaluation) guidelines for quality by outcome across a range of studies [15].
Statistical analysis
Where multiple studies reported data for a single outcome on different scales (e.g. % predicted or mL), standardised mean differences (SMDs) were calculated to enable pooling of the data. If studies presented data in both formats, data presented as % predicted were selected in preference to mL for conversion to SMD.
RCTs and observational studies were meta-analysed separately. Observational studies were required to have adjusted for asthma severity and smoking status in the statistical analysis or have accounted for these factors in their study design (e.g. inclusion criteria of only mild asthma) to be included in the meta-analysis.
Heterogeneity was assessed using the I2 statistic. Fixed-effects models were used if I2 was <25%. Random-effects models were used if I2 was between 25% and 75%. Pooled estimates were presented but considered to be unreliable if I2 was >75%. The selected model for each main outcome was then applied to any related subgroup meta-analyses. Data were pooled using the program Review Manager, version 5.3 [16].
Subgroup analyses
The following planned subgroup analyses were performed where possible: 1) age group: children (aged <18 years) versus adults (aged ≥18 years); 2) atopic status: atopic versus nonatopic; 3) blood or sputum eosinophil level: high versus low; threshold as specified by the individual studies; 4) smoking status: current versus never- or former smoker; 5) length of follow-up: RCTs: 1 year versus >1 year; observational studies: 10 years or <10 versus >10 years.
Sensitivity analyses
The follow sensitivity analyses were performed where possible: 1) risk of bias assessments: with versus without studies considered at high risk of bias; 2) meta-analysis methodology: fixed-effects models versus random-effects models.
Protocol registration
The protocol for this systematic review was registered prospectively in the PROSPERO database in September 2017 (registration number: CRD42017053543). FVC and BHR were added as outcomes of interest after protocol registration.
Results
Search results
6517 records were identified from the database searches and reference lists of related articles (figure 1). Of these, 365 records were selected for full-text screening and of these, 38 articles from 24 unique studies met the pre-specified inclusion criteria. Excluded records and their reasons for exclusion are listed in table e2. Characteristics of the included studies have been summarised in table 1 (RCTs) and table e3 (observational studies).
PRISMA flow diagram for the study selection process. ICS: inhaled corticosteroid.
Study characteristics of the included randomised controlled trials
RCTs
Included studies
13 RCTs with a total of 11 678 participants were included [12, 17–27]. Five RCTs were performed in children (n=1250) [18, 21, 23, 26, 27], seven in adults (n=3263) [12, 17, 19, 20, 22, 24, 25] and one in both children and adults (n=7165) [3]. Three studies contributed 90% of the participants (START: n=7165; SYGMA1: n=2559; CAMP: n=729) [3, 24, 27]. Asthma was physician-diagnosed in all studies. BHR or bronchodilator reversibility was an additional inclusion criterion in nine studies (n=10 889) [3, 19, 20, 22–27]. Seven studies were in individuals with mild asthma (n=10 708) [3, 17–19, 21, 24, 26], two were in individuals with mild-to-moderate asthma (n=764) [12, 27], one was in individuals with moderate-to-severe asthma (n=54) [23], and three did not limit inclusion by asthma severity (n=152) [20, 22, 25]. Eight studies used budesonide (n=11 128) [3, 17, 21–25, 27], three used fluticasone (n=149) [12, 19, 20] and two used beclomethasone (n=401) [18, 26]. Based on GINA criteria [11], ICS were categorised as low dose in eight studies (n=10 459) [3, 17, 19–22, 24, 25], medium dose in four studies (n=1184) [18, 23, 26, 27] and high dose in one study (n=35) [12]. Median (interquartile range) follow-up was 1.0 (1–4) years.
Risk of bias assessments
Methodological quality in the RCTs was good overall (figure 2). All RCTs were assessed to be at low risk of selection, performance and detection bias. Five studies were assessed to be at high risk of attrition bias, related to high dropout rates [19–21, 23, 25]. Reporting bias was unclear in eight mainly older studies for which prospective protocols were not available [18–23, 25, 26]. GRADE certainty of evidence across outcomes ranged from low to high as outlined in tables 2 and 3.
Cochrane Collaboration Tool (randomised controlled trials): risk of bias assessments for each included study.
Meta-analysis of randomised controlled trials for pre-bronchodilator (BD) outcomes stratified by age group
Meta-analysis of randomised controlled trials for post-bronchodilator (BD) outcomes stratified by age group
Studies excluded from the quantitative analysis
Two small RCTs were excluded from the quantitative analysis due to methodological factors or nonrepresentative study samples [20, 23]. One study compared budesonide to placebo in 47 children with moderate-to-severe asthma, but had significant and selective dropout in the placebo arm, which resulted in early discontinuation [23]. The budesonide arm was continued to study completion at 3 years and the authors did not account for differences in treatment duration in the comparisons. The second study was published as a brief report and compared fluticasone to placebo in 45 adults with asthma and accelerated lung function decline (>80 mL·year−1) [20]. The study sample was considered to be highly selected due the declining lung function inclusion criteria and not representative of the general asthma population.
Pre-BD lung function
Pre-BD FEV1
Eight studies measured changes in pre-BD FEV1 (% pred) (table 2, figure 3) [3, 12, 19, 21, 22, 25–27]. Improvements were observed with treatment across all age groups (2.22% (95% CI 1.32–3.12), n=8332), with similar estimates of strength in association in children (2.08% (95% CI 0.71–3.44), n=4151) and adults (2.47% (95% CI 1.64–3.29), n=4181). Findings were similar in four studies which measured pre-BD FEV1 in mL (74 mL (95% CI 54–94), n=3603) [18, 24, 25, 27] and when SMDs were calculated and pooled (0.21 SMD (95% CI 0.12–0.30), n=11 131).
Forest plot comparison (randomised controlled trials): change in pre-bronchodilator forced expiratory volume in 1 s (% predicted) stratified by age. ICS: inhaled corticosteroid.
Pre-BD FVC
Two studies measured changes in pre-BD FVC (% pred) and found no evidence of a treatment benefit (0.10% (95% CI −1.23–1.43), n=804) [25, 27]. Findings were similar in two studies which reported pre-BD FVC in mL (−13 mL (95% CI −149–124), n=804) [25, 27] and when SMD were calculated and pooled (0.01 SMD (95% CI −0.13–0.15), n=804)
Pre-BD FEV1/FVC
No studies measured changes in pre-BD FEV1/FVC (% pred). One study performed in children measured change in pre-BD FEV1/FVC (ratio multiplied by 100) and found significant improvements with treatment (1.60 (95% CI 0.64–2.56), n=729) [27].
Pre-BD FEF25–75
One study measured changes in pre-BD FEF25–75%, but was excluded from the quantitative analysis due to concerns around methodology [23]. The study reported benefits in pre-BD FEV25–75% with ICS use over the treatment period.
Post-BD lung function
Post-BD FEV1
One study in adults [3] and two in children [3, 27] measured changes in post-BD FEV1 (% pred) (table 3, figure 4). Significant improvements were observed with treatment in adults (1.54% (95% CI 0.87–2.21), n=3970), but not in children (0.20% (95% CI −0.49–0.90), n=3924) (p-subgroup differences0.006). One study reported change in post-BD FEV1 (mL) in children and also did not demonstrate a treatment benefit (−40 mL (95% CI −115–35, n=729) [27]. Findings were similar when SMD were calculated and pooled, with benefits found in adults (0.14 SMD (95% CI 0.08–0.21), n=3970), but not in children (0.02 SMD (95% CI −0.05–0.09), n=4924).
Forest plot comparison (randomised controlled trials): change in post-bronchodilator forced expiratory volume in 1 s (% predicted) stratified by age. ICS: inhaled corticosteroid.
Post-BD FVC
One study performed in children measured changes in post-BD FVC (% pred) and found no evidence of treatment benefit (−0.20% (95% CI −1.40–1.00), n=729) [27]. The same study also measured changes in post-BD FVC (mL) and a borderline adverse effect of treatment would have been reported if this measure was used (−60 mL (95% CI −120– −0), n=729) [27].
Post-BD FEV1/FVC
No studies measured changes in post-BD FEV1/FVC (% pred). One study in children measured change in post-BD FEV1/FVC (ratio) and found an association in the direction of a benefit (0.70 (95% CI −0.08–1.48), n=729), although this did not reach statistical significance [27].
Post-BD FEF25–75%
One study measured changes in post-BD FEF25–75%, but was excluded from the quantitative analysis due to methodological factors [23]. The study reported benefits in post-BD FEF25–75% with ICS use over the treatment period.
Fixed airflow obstruction
There were no RCTs that reported data on incidence of fixed airflow obstruction.
BHR
Three studies measured BHR as change in doubling concentrations of methacholine provocative concentration causing a 20% fall in FEV1 (PC20) [19, 21, 22]. Treatment increased methacholine PC20 compared to placebo (1.07 doubling concentrations (95% CI 0.65–1.49), n=223). Findings were similar when BHR was expressed as change in methacholine PC20 mg·mL−1 (0.99 (95% CI 0.32–1.66), n=161) [26], change in methacholine PC20 factor increase from baseline (1.10× (95% CI 0.61–1.59), n=729) [27] and change in histamine PC20 factor increase from baseline (2.59× (95% CI 1.18–4.00), n=75) [25].
Subgroup analyses
Stratification by smoking status
The largest study (START) reported changes in pre-BD and post-BD lung function stratified by smoking status. No significant differences were found between smokers and nonsmokers for change in either pre-BD FEV1 (p-subgroup difference0.22) or post-BD FEV1 (p-subgroup difference0.37) [3].
Stratification by duration of follow-up
When stratified by duration of follow-up, the greatest benefits for pre-BD FEV1 were found in the first year of follow-up [18, 19, 22, 24, 25] when compared to studies with follow-up >1 year [3, 27] (p-subgroup differences0.004) (figure e1). Stratification for post-BD lung function by follow-up could not be performed as all studies had follow-up durations >1 year.
Other pre-specified subgroup analyses
No studies provided outcome data for the other pre-specified subgroup analyses by atopic status, blood or sputum eosinophil counts.
Sensitivity analyses
Sensitivity analyses performed by meta-analysis model and risk of bias assessment did not significantly affect the results for pre-BD or post-BD lung function (table e4).
Observational studies
Included studies
11 observational studies with a total of 3720 participants were included [28–37]. Four studies were in children (n=787) [28, 30, 34, 36] and seven in adults (n=2933) [29, 31–33, 35, 37, 38]. Asthma was physician-diagnosed in eight studies (n=1044) [28, 30–34, 36, 37] and self-reported in three studies (n=2676) [29, 35, 38]. ICS use was measured at baseline in five studies (n=750) [28, 31, 32, 34, 36], at follow-up in four studies (n=2663) [29, 30, 37, 38], and at both baseline and follow-up in two studies (n=307) [33, 35]. BHR or bronchodilator reversibility was required as an additional inclusion criterion in five studies (n=2108) [29, 31, 32, 34, 37]. Asthma was of mild-to-moderate severity in two studies (n=301) [28, 37] and the remaining nine did not restrict inclusion by asthma severity (n=3419). Median (interquartile range) follow-up was 8.4 (3–28) years.
Quality assessment
Methodological quality of the cohort studies varied (table e5), with a median (interquartile range) NOS score of 7 (4–9). Three studies were assessed as being of “very good” quality (NOS ≥9) [35, 36, 38], three of “good” quality (NOS 7–8) [28, 31, 33], four of “satisfactory” quality (NOS 5–6) [29, 32, 34, 37], and one of “unsatisfactory” quality (NOS 0–4) [30]. Quality was most often negatively impacted by inadequate adjustment for critical confounders (i.e. asthma severity and smoking status). Six studies also had high rates of loss to follow-up or selective patterns of dropout in the comparison groups [29, 30, 33, 35, 37, 38].
Studies excluded from the quantitative analysis
Seven observational studies were excluded from the quantitative analysis as critical confounders (asthma severity and smoking status) were not adjusted for in the statistical analysis or through study design [28–32, 34, 37]. Sensitivity analyses with these studies included are provided in table e6. Of the remaining four studies, NOS assessments ranged from 8 to 10 [33, 35, 36, 38].
Pre-BD lung function
Sufficient data to perform a meta-analysis were available for only one outcome, pre-BD FEV1 (mL), with three studies, all performed in adults, finding significant per-year treatment benefits over their follow-up periods (14 mL·year−1 (95% CI 2–26), n=939) [33, 35, 38]. One study, performed in children also found treatment benefits in pre-BD FEV1 (% pred) (0.81% per year (95% CI 0.01–1.61), n=190) [36].
Post-BD lung function
None of the four studies measured post-BD outcomes or incidence of fixed airflow obstruction.
Subgroup analyses
Stratification by smoking status
One observational study reported data on pre-BD FEV1 stratified by smoking status [35]. Similar to the START trial, this study found no significant differences between smokers and nonsmokers.
Other pre-specific subgroup analyses
No observational studies provided data for other pre-specified subgroup analyses by atopic status, blood or sputum eosinophil counts, or duration of follow-up.
Sensitivity analyses
Sensitivity analyses performed by meta-analysis model and risk of bias assessment did not significantly affect the results for pre-BD outcomes (table e6). There were insufficient studies to perform sensitivity analyses for post-BD outcomes.
Discussion
This is the first systematic review to comprehensively evaluate the long-term effects of ICS on lung function in patients with asthma. Within the RCTs, maintenance ICS was associated with modest improvements in pre-BD FEV1 across all age groups, whereas improvements in post-BD FEV1 were observed only in adults. The greatest benefits were observed in the first year of treatment, and estimates were similar between smokers and nonsmokers. Due to the characteristics of the included studies, these findings are most applicable to children and adults with mild asthma treated with low-dose ICS. However, we found no RCT data on the effect of ICS on incidence of fixed airflow obstruction in asthma. Moreover, while the observational studies were generally of longer duration, many were assessed as being at high risk of bias due to inadequate adjustment for confounding factors.
Current asthma guidelines recommend ICS as a first-line option for persistent asthma. ICS therapy improves asthma control, reduces exacerbations and provides a modest increase in short-term measures of lung function over weeks to months [2, 3, 39, 40]. While we found these benefits on lung function persisted beyond the first year of treatment, the magnitude of benefit was relatively modest and below conventional minimally clinically important difference cut-offs for FEV1 (230 mL) [41]. Additionally, in the subgroup analysis by duration of follow-up, the greatest benefits were obtained within the first year and then appeared to decrease with time. This trend was similarly observed in START (year 1 versus year 3) and CAMP (year 1 versus year 4) [3, 27]. However, it should be noted that this observation was confounded by several time-dependent factors, including decreasing compliance to ICS with time in the intervention arms and increasing use of nonintervention asthma medications in the placebo arms [3, 24, 27]. Therefore, rather than a cumulative benefit on lung function over time in these mild asthma trials, we found that maintenance ICS use was associated with a modest absolute benefit which was maintained to an extent with continued treatment. These effects could be more pronounced in patients with moderate-to-severe disease and in those receiving higher-dose ICS. However, there were limited data to evaluate such effects in these subgroups.
The age-based differences for post-BD FEV1 were unexpected. In meta-analyses with small numbers of contributing trials, subgroup differences can arise from subtle differences in methodology and this was important to exclude. However, the majority of the data for this outcome, for both children and adults, came from one large study (START), which recruited patients using standardised methodology across age groups [3]. CAMP also contributed data on post-BD FEV1 and did not identify a treatment benefit in children [27], despite their participants having a slightly higher disease severity at baseline than participants in START (mild-to-moderate versus mild only) [27]. While there appeared to be a trend toward a benefit with treatment on post-BD FEV1/FVC in CAMP, this did not reach statistical significance.
Several reasons may account for these age-based differences. First, the diagnosis of asthma in children is often less clear than in adults and physician-diagnosed asthma in the absence of objective tests can be uncertain [42, 43]. Inclusion of children who may not have benefited from ICS in the first place, such as those with viral-induced or episodic wheeze, would have reduced any observed treatment benefit. Secondly, childhood asthma often improves over adolescence and treatment differences could also have been reduced by improved asthma control or asthma remission as part of the natural course [44]. This theory is supported by data from START, in which adolescents (aged 11–17 years) were less likely to benefit from maintenance ICS compared to both younger children (aged 5–10 years) and adults (aged 18–66 years). Lung function trajectory studies have also shown that small deficits established early in life can diminish with time, with some children with below-normal lung function entering a “catch-up” phase during puberty [7, 45]. Consequently, rapid growth in lung function during childhood might also explain the observed treatment differences between children and adults.
While the observational studies had longer durations of follow-up compared to the RCTs, few met pre-specified criteria to be included in the quantitative analysis. Of those included, ICS use was variably assessed at baseline and/or follow-up, and change in medication use between these time-points were not clearly assessed. Acknowledging these important limitations, ICS use as defined in these observational studies was associated with improved trends in pre-BD FEV1 across the follow-up periods, which ranged from 9 to 13 years. The observational studies were also more likely to have observed any effects of ICSs on fixed airflow obstruction in asthma, which may develop over years to decades [7]. Unfortunately, we did not identify any studies reporting data for this outcome.
The effects of ICSs on lung function in asthma appear to be mediated, at least in part, through an effect on asthma exacerbations. Exacerbations represent intermittent periods of intense airways inflammation, and have been associated with structural airways remodelling and rapid lung function decline [46, 47]. In a post hoc analysis of START, O’Byrne et al. [47] showed that low-dose budesonide attenuated these adverse effects, with treatment associated with a reduction in both the rate and impact of severe exacerbations on post-BD FEV1. While budesonide was not found to improve post-BD FEV1 in children in the overall analysis, treatment was associated with a significant reduction in post-BD FEV1 impairment in children who experienced one or more severe exacerbations during the 3-year follow-up (−2.38% predicted on budesonide versus −6.29% predicted on placebo). This interaction, also observed in adults, supports the hypothesis that treatment effects are likely to be greater in patients with more severe disease.
Following completion of the recent SYGMA1 and Novel START trials, there is now strong evidence on the benefits of maintenance ICSs compared to short-acting β-agonist (SABA) reliever monotherapy for exacerbation prevention in mild asthma [17, 24]. Accordingly, the 2019 GINA update has now recommended against SABA reliever monotherapy in mild asthma, and instead recommends ICS containing treatment for all mild asthma patients. With new evidence supporting the use of as-needed budesonide–formoterol as an alternative to maintenance ICS in step 2 treatment [13, 17, 24, 48], the relative effects of these regimens on long-term lung function should also be examined.
This systematic review has a number of key strengths. We performed comprehensive searches of three electronic databases and two authors independently screened references, extracted data and performed quality assessments. We also evaluated both experimental and observational evidence. The RCTs provided higher quality evidence but recruited highly selected samples of asthma patients suited for drug efficacy trials. In contrast, the observational studies evaluated a broader range of asthma patients over longer durations of follow-up. Given the strong existing evidence base and ethical implications, we believe there are unlikely to be further placebo-controlled trials conducted. However, high-quality observational data on the longer term effects of ICS treatment in moderate-to-severe asthma, particularly for post-BD lung function, are still required
This review also had several limitations. First, one of our objectives of examining the effect of ICS on incidence of fixed airflow obstruction in asthma could not be achieved due to the lack of available evidence. Secondly, our review was limited to English language publications and relevant non-English studies may have been excluded. Thirdly, the Cochrane Collaboration now recommends the ROBINS-I tool for assessing risk of bias in nonrandomised studies [49] and while limited observational data were included, risk of bias assessments for these studies were performed using the formerly recommended NOS tool [14]. Fourthly, many of the included studies compared low-dose budesonide to placebo and recruited only patients with mild asthma. Therefore, our results may not be representative of more severe asthma, higher-dose or other ICS agents. Fifthly, most data for several outcomes came from one large RCT (START), which included patients with recent-onset asthma. It is unclear whether outcomes in this sample may have been more pronounced compared to patients with longer term established disease. Finally, most studies focused on changes in FEV1. Other routinely collected spirometry indices (e.g. FVC and FEV1/FVC) were infrequently reported, which raised concerns of selective reporting for these outcomes.
In conclusion, maintenance ICS are associated with modest, age-dependent improvements in long-term lung function in patients with mild asthma. These improvements represent an added functional benefit to the more accepted clinical actions of ICS and reinforce the importance of ICS as a recommended initial treatment option even for mild asthma. Further research is still needed to determine whether ICS use modifies the risk of developing fixed airflow obstruction in asthma, especially in smokers and high-quality observational data from existing prospective cohorts could be used to address these important knowledge gaps. Treatment adherence, correct inhaler technique and correct diagnosis of asthma are factors likely to influence these outcomes in real-world settings.
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.
Supplementary_tables ERR-0185-2020_Supplementary_tables
Footnotes
This article has supplementary material available from err.ersjournals.com
This study is registered at www.crd.york.ac.uk/prospero with identifier CRD42017053543
Provenance: Submitted article, peer reviewed.
Author contributions: D.J. Tan, S.C. Dharmage, J.L. Perret, C.J. Lodge, A.J. Lowe, P.S. Thomas, D. Jarvis, M.J. Abramson and E.H. Walters contributed to the study concept and drafted the review protocol. D.J. Tan, D.S. Bui and X. Dai performed the screening, quality assessment and data extraction.
Conflict of interest: D.J. Tan has nothing to disclose.
Conflict of interest: D.S. Bui has nothing to disclose.
Conflict of interest: X. Dai has nothing to disclose.
Conflict of interest: C.J. Lodge has nothing to disclose.
Conflict of interest: A.J. Lowe reports grants from National Health and Medical Research Council, during the conduct of the study.
Conflict of interest: P.S. Thomas reports having participated in advisory board meetings for GSK and Astra Zeneca, unrelated to this work.
Conflict of interest: D. Jarvis has nothing to disclose.
Conflict of interest: M.J. Abramson reports grants from Pfizer and Boehringer Ingelheim, personal fees and non-financial support from Sanofi, and personal fees from GSK, outside the submitted work.
Conflict of interest: E.H. Walters has nothing to disclose.
Conflict of interest: J.L. Perret reports a travel grant from Boehringer-Ingelheim, outside the submitted work.
Conflict of interest: S.C. Dharmage has nothing to disclose.
Support statement: This study was supported by funds from the National Health and Medical Research Council (NHMRC) of Australia European collaborative grant scheme (1101313) as part of ALEC (Ageing Lungs in European Cohorts funded by the European Union's Horizon 2020 Research and Innovation Programme under grant agreement 633212). The funding bodies were not involved in the included work nor the decision to publish. DJT was supported by a NHMRC Postgraduate Scholarship and Royal Australian College of Physicians (RACP) Woolcock Scholarship. Funding information for this article has been deposited with the Crossref Funder Registry.
- Received June 9, 2020.
- Accepted August 18, 2020.
- Copyright ©ERS 2021.
This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0.
References
