Background
Exacerbations of chronic obstructive pulmonary disease (COPD) are associated with both short-term loss and long-term decline of lung function [
1‐
3]. Previous studies have shown that lung function drops at the time of an exacerbation and does not always fully recover to pre-exacerbation levels [
4,
5]. Much of what is known about the impact of exacerbations on lung function is from analysing long-term lung function changes in patients who experienced exacerbations over the course of a clinical trial, with lung function measured at scheduled in-clinic visits rather than at the time of the event [
2,
3].
Some studies have measured lung function at the time of and immediately after an event [
6,
7]. However, there are limited data available on the time course of lung function changes preceding exacerbations and immediately following exacerbations. Seemungal et al. [
4] showed that peak expiratory flow rate (PEFR) remains relatively stable before the onset of an exacerbation, while symptoms already deteriorate. By contrast, Calverley et al. [
8] observed a decrease of around 8% in peak expiratory flow (PEF) in the 2–3 weeks before an exacerbation.
There remains a need to better understand the spirometric changes around the time of an exacerbation, as it is currently difficult to identify exacerbations early. This is of clinical relevance as treating exacerbations earlier is associated with faster recovery and reduced risk of hospitalization [
9].
The WISDOM study was a 12-month inhaled corticosteroid (ICS) withdrawal study in patients with severe-to-very severe COPD. It showed that there was no increased risk of exacerbations following stepwise withdrawal of fluticasone propionate in patients receiving tiotropium and salmeterol [
10]. In the WISDOM study, following suitable training, patients performed daily spirometry at home, providing an opportunity to examine the changes in forced expiratory volume in 1 s (FEV
1) that occur prior to and immediately after an exacerbation.
The aim of the present post hoc analysis was to characterize the lung function profile of patients prior to, during and following the first moderate/severe and severe exacerbation during the WISDOM trial.
Methods
Study design
The WISDOM study methodology has been published previously [
10,
11]. In brief, this was a multinational, randomized, double-blind, parallel-group, active-control study. Patients entered a 6-week triple-therapy run-in with long-acting muscarinic antagonist (LAMA) and long-acting β
2-agonist (LABA)/ICS (tiotropium and salmeterol/fluticasone propionate). Patients were then randomized (study Week 0) to either continue triple therapy for 52 weeks or to continue receiving salmeterol and tiotropium (dual bronchodilator therapy) whilst discontinuing ICS in a stepwise manner over 12 weeks. ICS were completely discontinued in this group at Week 12. The primary endpoint was the time to the first moderate or severe COPD exacerbation during the 12-month study period. These results and further post hoc analyses have been published previously [
10,
12,
13]. The study was performed in accordance with the Declaration of Helsinki, the International Conference on Harmonisation’s Harmonised Tripartite Guideline for Good Clinical Practice and local regulations. The protocol was approved by the ethics research board of the respective institutions, and all patients provided written informed consent.
Patients and treatments
Patients were ≥ 40 years of age, were either current or former smokers, had been diagnosed with severe or very severe COPD (defined as an FEV
1 < 50% of the predicted volume and < 70% of the forced vital capacity after bronchodilation), and a history of at least one documented exacerbation in the 12 months prior to screening. Further inclusion and exclusion criteria have been reported elsewhere [
11].
Patients were randomized to continue receiving tiotropium, salmeterol and fluticasone propionate, or to have fluticasone propionate withdrawn over a 12-week period (further details are given in the Supplementary Methods in the Additional file
1).
Exacerbations
Patients completed a simple daily paper diary, recording changes in respiratory symptoms and the use of medications between visits. Diaries were used to help patients report symptoms and medication changes to investigators. Moderate exacerbations were defined as an increase of at least two lower respiratory tract symptoms related to COPD (shortness of breath, sputum production [volume], sputum purulence, cough, wheezing or chest tightness), or the new onset of two or more such symptoms, with at least one symptom lasting 3 or more days and for which antibiotics, systemic glucocorticoids or both were prescribed. A severe exacerbation was defined as an exacerbation requiring hospitalization. The start date of an exacerbation was defined as the date of onset of the first COPD symptom that was part of the exacerbation and was determined by the investigator reviewing the diary cards. The end of the exacerbation was determined by the investigator reviewing the diary cards and stopping exacerbation treatment.
In the case of an exacerbation, patients could be treated with oral glucocorticoids, and/or antibiotics and further treatments as deemed medically necessary.
Lung function
Following suitable training, patients performed daily spirometry throughout the 52-week study. Home-based measurements were performed by patients each morning before administration of the study drug using a portable, battery-operated, ultrasound, transit-time-based electronic spirometer (EasyOne® NDD Medical Technologies, Chelmsford, MA, USA, and Zurich, Switzerland). Patients performed at least three efforts each day, and the highest FEV
1 was selected for analysis [
14]. Automatic feedback was provided to the patient if effort was not in line with American Thoracic Society (ATS)/European Respiratory Society (ERS) criteria [
15]. Data were retrieved at clinic visits, and a pulmonologist performed a central over-read of all lung function data and graded it for acceptability according to ATS/ERS criteria. All spirometric techniques and equipment used were in accordance with ATS/ERS recommendations [
15].
We have previously shown that home-based FEV
1 is a reliable measure, and there was a good agreement between the home-based spirometry results and the in-clinic spirometry in the WISDOM trial. However, the home-based measures of FEV
1 were consistently approximately 50 mL lower than in-clinic values [
14].
Statistical analysis
For the present post hoc analysis, we calculated absolute and percentage change from pre-exacerbation period baseline (mean of the period 6–8 weeks before the exacerbation) in on-treatment daily FEV1 for 8 weeks before and after the first day of the first moderate/severe on-treatment COPD exacerbation, as well as change in on-treatment daily FEV1 in the subsets of moderate and severe on-treatment COPD exacerbations. We also calculated weekly means for FEV1 and the change from Week − 8.
No inferential statistical analyses were conducted. Descriptive statistics were calculated for demographic variables. Data for change in on-treatment daily and weekly FEV1 up to and following the first day of the first exacerbation (moderate, moderate/severe and severe) were plotted graphically for visual interpretation. For this analysis, patients who had at least one measurement per week in a 56-day period before and after the start of a moderate or severe exacerbation were included. Missing values were imputed by linear interpolation (intermittent), backfilling (beginning) or carry forward (end).
The analysis includes patients’ first exacerbation after the ICS withdrawal visit.
Discussion
This is the first study to show a decrease in daily-measured lung function before an exacerbation in a large group of patients with severe COPD and a low FEV
1 at baseline (33% predicted). This decrease in FEV
1 was observed from approximately 2 weeks before the start of an exacerbation, and although improvements were seen post-exacerbation, FEV
1 did not recover to pre-exacerbation levels up to 8 weeks after the start of the exacerbation. The post-exacerbation FEV
1 levels could either reflect the fact that the drop in FEV
1 never recovered, did not recover within 8 weeks after the start of the exacerbation, or that there was progressive lung function decline over time. Interestingly, another analysis of a large clinical trial has shown that the rate of lung function decline increases following a single exacerbation [
16].
There are limited data in which the daily lung function around the time of an exacerbation has been studied, though our results are consistent with previous studies that have shown that exacerbations are associated with loss in lung function [
1‐
3]. One other study that did report daily lung function was published by Seemungal et al. [
4] – a study in which a cohort of 101 patients recorded daily PEFR measurements over 2.5 years. In their cohort, there was no change in PEFR in the days prior to an exacerbation, unlike the results presented here. The PEFR results did, however, show incomplete recovery of lung function [
4], as was observed in this WISDOM analysis. Another study that measured PEF before and after exacerbations showed a small drop in PEF prior to the exacerbations that then returned to pre-exacerbation levels within 2 weeks of the exacerbation; however, increased symptoms were still seen 4–8 weeks after the exacerbation [
17].
Interestingly, another small study measured lung function at baseline and during an exacerbation in a severe COPD population similar to ours, and found a 40 mL drop in FEV
1 at the time of the exacerbation [
18], an observation consistent with our results.
The daily lung function data also show that the time course of the changes in lung function around exacerbations was similar in patients who withdrew from ICS to those in patients who continued ICS. This suggests that the objective change associated with an exacerbation occurring during ICS treatment was similar to that experienced when this therapy was not used.
The main limitation of the analysis is that it only included patients who could continue to perform home-based spirometry manoeuvres during and after their exacerbation, and patients with more severe symptomatic exacerbations may be more likely to stop performing the home-based measurements. It is possible that prior to more severe exacerbations the drop in FEV
1 may be more pronounced. However, the demographics of the patients included in this analysis are similar to the demographics of the wider WISDOM population. Another limitation of this study is the inherent variability in FEV
1, as the repeatability of the measure in individuals is greater than the differences we observed [
19]. This raises questions about the possibility of using the drop in lung function as a predictive measure, but does not contradict our findings about the drop in group mean FEV
1.
One strength of the analysis is that, unlike in observational data, maintenance therapy was standardized. The WISDOM study offers a unique opportunity to closely study the daily lung function of a relatively large number of patients before, during and after exacerbations of COPD occurring during triple therapy compared with dual bronchodilation.
Identifying exacerbations before they fully develop could improve their treatment. Although blood eosinophil count and exacerbation history can both identify patients at increased risk of exacerbation [
13,
20‐
22], it remains difficult to predict exacerbations of COPD. It is unlikely that all COPD exacerbations could be predicted by a drop in FEV
1 before the event, but it is possible that there is a subset of exacerbations that could be identified by daily home-based spirometry [
14] in patients at high risk of exacerbations. Future work is needed to identify cut-offs and to develop algorithms for prediction of exacerbations.
Acknowledgements
The authors acknowledge the significant contribution of Helen Finnigan, previously an employee of Boehringer Ingelheim, who performed most of the analyses presented in this manuscript and reviewed an early outline.