It has been accepted for more than 20 years that asthma is an inflammatory disease of the airways, but it is more recent that airways inflammation has been accepted to be important in chronic obstructive pulmonary disease (COPD).1

Airways inflammation is found in heavy smokers who have not developed airflow obstruction and is qualitatively similar to that seen in smoking-related COPD.2 3 Some, but not all, studies have shown a relationship between the severity of airflow obstruction and the degree of inflammation, lending support to the hypothesis that airway inflammation is important in COPD.4 5 The airway inflammation is characterised by an increase in CD8+ T cells and macrophages in the airway wall. Increased numbers of neutrophils are seen in the airway wall, particularly in more severe disease, but neutrophils are particularly prominent in sputum. 2 6 7 B cells are increased in more severe disease and there are increased numbers of mast cells.2 The main site of airflow obstruction in COPD is considered to be in the small airways which are inaccessible to normal bronchoscopic biopsy techniques; however, the pattern of airway inflammation seen in the large airways accessible to a bronchoscope and in the smaller airways is essentially similar.2 Although airway inflammation is found in COPD, knowing which are the most important cells, cytokines and mediators is fraught with difficulty. Ultimately, understanding which of the pathways are important will depend on intervention studies which attempt to associate changes in inflammation with clinical benefit.

There are many intervention studies in asthma, investigating the effect of anti-inflammatory treatment on airway inflammation in chronic asthma or following allergen challenge. However, there are very few studies in COPD and the study by Bourbeau et al published in this issue in Thorax is therefore of value (see page 938).8^–12 Previous studies by our group and others have shown in double blind, placebo controlled studies that inhaled corticosteroids given alone for 3 months have only a minor effect on airway inflammation with a reduction only in the number of mast cells.8^–10 Well constructed studies of the effect of inhaled or oral steroids on measures of inflammation in sputum have generally shown no effect on neutrophils.6 7 A large study reported by our group comparing combination therapy with fluticasone/salmeterol over a 3-month period with placebo showed reductions in a wide range of inflammatory cells, particularly CD8+ T cells and CD4+ T cells in biopsy specimens and a reduction in the percentage of neutrophils in sputum.11 Although this study had the strength of being large with a statistical power to show changes in the three primary outcome variables, it had the disadvantage of only comparing combination therapy with placebo. Bourbeau et al have compared placebo, fluticasone 500 μg twice daily and a combination of fluticasone 500 μg twice daily and salmeterol 50 μg twice daily in well characterised patients with COPD. The inclusion of the inhaled steroid alone adds considerable support to the previous observations. They found a significant reduction in CD8+ T cells in the combination treated group with no change in the group treated with fluticasone alone. They also found a significant reduction in macrophages (CD68+) but no change in neutrophils or eosinophils in biopsy specimens. The reduction in macrophages is different from the finding in our study where, for reasons that are unclear, the number of macrophages in both the placebo and active treatment groups rose over the 3 months of the study with no difference between the groups. In our study the percentage of neutrophils was reduced in sputum, but we have not yet reported data on neutrophil numbers in biopsy specimens. Previous studies, both in asthma and COPD, have shown that measurements of inflammation in the airway wall made from biopsy specimens and in the airway lumen either by sputum or bronchoalveolar lavage do not necessarily give the same picture of anti-inflammatory changes. Bourbeau et al do not report measuring increased sputum inflammation.12 The percentage reductions in cell counts in their study and in our study are of the order of 40%, which are likely to be of biological significance.

If airways inflammation is important in COPD, what might be the clinical effects produced by a reduction in inflammation? Given the severity of the structural changes that occur in the airways of patients with COPD, it is a little unlikely that any marked improvement in airflow obstruction would occur, at least in the short term. It is more likely that a reduction in inflammation might decrease the chance of having an exacerbation of COPD, reduce mortality or reduce the rate of decline in forced expiratory volume in 1 s (FEV₁). It is tempting to speculate that the reduction in exacerbations of COPD, which have been shown with the combination of inhaled steroids and long-acting β₂ agonists, may be related to reductions in inflammation.13^–16 The reduction in rate of decline of FEV₁ and the possible reduction in mortality seen in the TORCH study may also be related to reduced inflammation.16

The results of studies of the effect of pharmacotherapy on inflammation in COPD have a broader significance. They show that, in relatively low numbers of subjects in studies over a 3-month period, treatments can have a significant anti-inflammatory effect. As new anti-inflammatory drugs become available to be investigated in COPD, these types of studies may be important at an early stage to demonstrate anti-inflammatory action, as studies in which the end point is the rate of decline in FEV₁ or a reduction in exacerbations would require much longer studies in much larger numbers of subjects. The fact that a combination of an inhaled steroid and a long-acting β₂ agonist can significantly reduce inflammation in COPD does indicate that this inflammation is tractable to pharmacological intervention which, again, should encourage those who are developing novel anti-inflammatory treatments for COPD to continue to pursue this aim.17 A criticism which is frequently levelled at these studies is that they are only looking at the large airways when the important site of inflammation may be the small airway but, until techniques which can safely and reliably sample small airways are developed, studies looking at airway inflammation by bronchial biopsies combined with induced sputum or bronchoalveolar lavage to sample the lumen of the airways will remain an invaluable way of understanding the information of airway inflammation in COPD.