Is There a Role for Antiinflammatory Treatment in COPD?

Abstract

By definition, chronic obstructive pulmonary disease (COPD) is associated with an abnormal inflammatory response of affected lungs. Therefore, the search for an effective anti-inflammatory therapy for this debilitating disease is intense. However, to date, there is no such anti-inflammatory treatment for COPD. While there are some modest effects of inhaled corticosteroids on selected clinical endpoints in COPD, it remains to be proven that the observed effects are due to changes in the underlying inflammation, in particular since relevant clinical endpoints of COPD can be significantly improved by treatments not targeting inflammation. Therefore, it appears justified to reconsider the present knowledge about any linkage of local and systemic inflammation and clinical features of COPD, including lung function, exacerbations, disease progression, and mortality. Any such link needs to be carefully established before future anti-inflammatory therapies for COPD are developed and investigated in clinical trials, in particular since proof-of-concept trials aiming merely at inflammatory markers in COPD may not be predictive of clinical success or failure. The present review summarizes current knowledge about the role of inflammation in COPD, and critically analyzes results from clinical trials with inhaled corticosteroids and phosphodiesterase-4 inhibitors in COPD, the two classes of putative antiinflammatory agents with the richest body of evidence from controlled studies.

Keywords:

• COPD
• Inflammation
• Treatment
• Inhaled corticosteroids
• Phosphodiesterase inhibitors

BACKGROUND

Inflammation in COPD

Chronic obstructive pulmonary disease (COPD) is a debilitating disease defined by chronic and poorly reversible airway obstruction in the presence of an abnormal or amplified inflammatory response of the lungs towards noxious gases, in particular cigarette smoke ([1]). Besides infiltration of bronchial mucosa and the alveolar space by CD8⁺-lymphocytes ([2]), increased presence of neutrophils and macrophages in the airways of affected patients is assumed to be a hallmark of COPD ([3],4,5,6,7). Further, there is evidence of vascular and, in particular, systemic inflammation in COPD ([8]). Despite a certain degree of overlap (e.g., presence of airway eosinophils in a subset of COPD patients, in particular during exacerbations), it is generally accepted that the pattern of inflammation in COPD is unique and distinct from that observed in patients with asthma ([1]).

However, while there is an undisputable association of COPD with airway inflammation, the extent to which inflammation contributes to the clinical features of COPD remains questionable. This is particularly due to a poor understanding of the underlying nature of inflammation in COPD which might differ from one person to the other. In that context, no consistent association has been shown between inflammatory markers and clinically important outcome parameters, such as lung function, exacerbation rates, disease progression, and mortality. It is interesting to note that while smoking cessation has been shown to reduce the rate of lung function decline in COPD ([9]), this effect was not linked to a reduction in inflammation, as shown by a biopsy study comparing histologic patterns of active smokers and quitters with COPD ([10]). This observation is also supported by the recent studies of Hogg et al., indicating that the presence of luminal mucus in peripheral airways of resected lung samples, but not the degree and severity of inflammation per se, was the only predictor of COPD progression in a cohort of COPD patients with emphysema ([11]).

Further, lung volume reduction surgery was found to cause a significant and meaningful reduction in the number of subsequent exacerbations ([12]), while the procedure lacks any direct antiinflammatory effect. Similar results have been obtained with long-term treatment with a long-acting anticholinergic bronchodilator, which significantly reduced COPD exacerbations, but has not been shown yet to alter the inflammatory pattern of the disease ([13]). While these findings do not necessarily implicate that antiinflammatory drugs in COPD will not produce positive outcomes, it underscores the difficulty to choose the adequate endpoints in a clinical trial. It might of course be argued that a reduction of inflammation per se is a valid and clinically significant outcome to achieve ([14], [15]), however this ultimately needs to be linked to a more general positive clinical health effect (physiologic markers, quality of life, morbidity or mortality, etc.).

In addition, the phosphodiesterase-4 inhibitors cilomilast and roflumilast at the dose-range used, have demonstrated some antiinflammatory activity in vivo ([16], [17]), but their clinical effectiveness has been at best marginal ([18]).

This large body of data from well-controlled clinical studies has reinforced the debate about the clinical meaningfulness of anti-inflammation treatment in COPD ([19]). The present review summarizes the clinical effects of those antiinflammatory therapies for COPD, namely inhaled corticosteroids and PDE4-inhibitors.

Inhaled corticosteroids (ICS)

While ICS are the mainstay of antiinflammatory treatment in asthma, their role in COPD remains controversial. This is also reflected by the fact, that current guidelines recommend ICS only in a subset of COPD patients with severe disease (FEV₁ < 50% of predicted) and a history of recurrent exacerbations. Numerous studies with ICS in COPD have been performed, with diverging endpoints and outcomes. In general terms, the common ground of all these trials was the a priori assumption, that—given a somewhat “causal” link between inflammation and clinical features of COPD–ICS therapy would result in a favourable clinical outcome due to its effect on the underlying inflammation. However, the results of the majority of ICS trials in COPD has generated conflicting results, and it appeared particularily difficult to link any of the observed outcomes to the putative mechanism of action of ICS, namely reduction of inflammation ([20]). Finally, due to increased risk of pneumonia associated with ICS containing treatments ([21], [22], [23]), the individual benefit-risk ratio of ICS has to be taken into account.

Effect of ICS on pulmonary function

In some, albeit not all, studies a negative correlation of inflammation and pulmonary function, in particular FEV1, has been found. Histological examinations by Saetta and co-workers ([2], [5], [24], [25]) revealed a negative correlation of FEV₁ and the number of CD8+ T-cells in a small number of subjects. A negative correlation with FEV₁ has also been described for sputum neutrophils ([26], [27]). However, this observation is in contrast to a study by Lapperre et al., where only a small part of airway function loss was attributable to sputum neutrophils ([28]). In cross-sectional analyses by Hogg et al., inflammation increased with higher COPD stages (according to GOLD) ([29]), but an extended investigation on a similar population found no significant correlation between inflammatory markers (lymhoid follicles) and disease progression ([11]), thus a causal link between inflammation and airway obstruction remains questionable.

Hence, despite some evidence promoting a correlation of inflammation with airway obstruction, the overall picture remains inconclusive. Therefore, it cannot be necessarily concluded, that any therapeutic intervention targeting inflammation in COPD will have a beneficial impact on FEV₁.

The majority of short-term studies using ICS in COPD failed to show any beneficial effects on pulmonary function ([30],31,32,33,34,35). Only the studies using higher doses of ICS by Thompson et al. ([36]) (1000 μg Beclomethasone/day for 6 weeks), Thompson et al. ([37]) (1000 μg Fluticasone/day for 12 weeks), and Nishimura et al. ([38]) (3000 μg BDP/day for 4 weeks) showed a significant improvement of FEV₁.

In contrast, the majority of medium- and long-term trials with ICS in COPD found a significant improvement of FEV₁, although the magnitude of improvement was modest in most cases.

In the large-scale controlled parallel-group trials studying the effect of ICS/long-acting beta-agonist fixed combination products with their single components or placebo, effects of ICS monotherapy on lung function were comparable, with a difference to placebo of 105 ml ([39]), 112 ml ([40]), 95 ml ([41]), or 47 ml ([22]).

In contrast, Calverley et al. ([42]) failed to demonstrate any effect of budesonide 2x400 mcg per day versus placebo on FEV₁ after 1 year (2% difference, p = ns). Negative results with ICS in COPD were also observed in studies by Bourbeau et al. ([43]), where treatment with budesonide 1600 mcg per day over 6 months was not superior to placebo with regard to FEV₁. The influence of fluticasone propionate (FP) withdrawal versus continued treatment (1000 mcg per day) in 244 COPD patients was investigated in the COPE study ([44]). Here, the adjusted treatment difference after 6 months was 38 ml in favour of FP, but did not meet the level of significance (p = 0.056).

In summary, the data on the effect of ICS monotherapy on lung function in COPD suggest a small, but demonstrable effect, when the duration of treatment was at least 6 months. The effect size can be roughly estimated to be as large as 50–100 ml over placebo ([45]). Redelmeier et al. concluded that in patients with severe COPD (FEV₁ 35% of predicted), an FEV₁ improvement of 114 ml was associated with an improvement in dyspnea ([46]). However, in the majority of COPD studies with ICS assessing dyspnea as an outcome parameter, no beneficial effects were observed ([40],41,42. [44], [47],48,49,50).

In any case, the lung function improvements seen with ICS are much lower than those observed with long acting bronchodilators ([51], [52], [41], [49]), and there is as yet no consistent evidence that the improvements of FEV1 seen with ICS are related to a suppression of inflammation. On the contrary, chronic ICS use treatment does not reduce sputum neutrophilia, and may actually increase tissue airway neutrophilia in biopsies from COPD patients ([53]). Therefore, possible alternative mechanisms underlying the effect of ICS on lung function need to be considered, such as the reduction of epithelial plasma leakage or mucus production. An often proposed mechanism for the enhancement of clinical effects of long-acting beta-agonists by concomitant use of ICS is the up-regulation of beta-receptors by glucocorticoids ([54]). While there is ample evidence that such an effect is demonstrable in vitro ([55]), and clinical studies at least indicate that adding an ICS to a LABA may improve some outcomes in COPD ([49],41,22), one has to consider that many studies in asthma have failed to demonstrate any relevant effect of ICS usage on beta-receptor downregulation in vivo ([56],57,58,59). Thus, it remains unproven that upregulation of beta-receptors by ICS in COPD patients leads to improved outcomes.

Effects of ICS on disease progression

It is still unclear whether disease progression of COPD, defined as accelerated loss of lung function (FEV₁) over time is—at least in part—mediated by inflammation. To date, studies supporting a major role of inflammation in disease progression are scant. Stanescu et al. reported a post hoc-analysis showing a correlation of sputum neutrophilia and FEV₁ decline in a small group of COPD patients ([60]). An excess of mucus production, was associated with a more rapid FEV₁ decline in the study by Vestbo et al. ([61]). Further, the cross-sectional pathology studies by Hogg et al. suggest an increase of inflammatory markers with increasing COPD stages ([29]). Until recently, it was assumed that the beneficial effect of smoking cessation on the decline of FEV₁ over time was due to a reduction in airway inflammation. However, Gamble et al. comparing airway inflammation in active and former smokers with COPD found no difference between the two groups ([10]). This finding also concurs with data by Rutgers et al., who found that airway inflammation may persist in COPD even after smoking cessation ([62]). Despite the fact, that smoking cessation is undoubtedly of benefit in COPD patients, it remains to be proven whether this effect is linked to an amelioration of the underlying airway inflammation ([63]).

Accordingly it is not surprising that numerous large-scale trials investigating the effect of long-term ICS on FEV₁ decline in COPD have produced negative results with regard to this endpoint ([43], [47], [65],66,67,68,69,70,71,72). It was argued, that the lack of any statistically significant effect of ICS on FEV₁ decline could be due to the small “true” effect size and therefore underpowering of the single studies. To address this issue, two metaanalyses generating confliciting results were done. The analysis of Highland et al. ([73]), demonstrated a non-significant (p = 0.11) difference of 5 ± 3.2 ml/FEV₁ decline per year. This outcome was hampered by a calculation error, falsely citing the effect of the original study by Vestbo ([69]) to be as large as −3.1 ml/year, while the true effect was in fact +3.1 ml/year. The analysis by Sutherland et al. ([74]), reported a statistically significant effect of ICS of 7.7 ml/year (p = 0.02) included a study by Derenne et al. ([66]), accounting for a large proportion of the calculated ICS effect. The validity of this study was severly questioned, since it had never been published in a peer-reviewed journal ([73]).

Thus, the current knowledge based on orginal studies, does not indicate a relevant effect of ICS on COPD progression.

Effect of ICS on exacerbations

Exacerbations of COPD are assumed to represent a clinical outcome parameter that is to some extent related to the underlying inflammatory process ([14], [75]). Hence, an effective anti-inflammatory therapy is expected to have a beneficial impact on the frequency and/or severity of exacerbations. Qiu et al. found an increase of neutrophils in bronchial biopsies of COPD patients during an acute exacerbation ([76]). Other authors have also confirmed the presence of increased inflammatory markers during exacerbations including eosinophils ([77]) when compared to stable disease ([78]). While it appears reasonable that bronchial and also systemic inflammation increase during an exacerbation – even more so, since the majority of exacerbations is triggered by infections ([79]) – there is inconclusive data about the predictive value of the degree of inflammation on prospective exacerbation rates. In a study by Bhowmik et al., increased sputum neutrophils did not predict exacerbations in a cohort of COPD patients ([80]). A similar observation was made by Gompertz et al., who failed to detect any significant difference in the inflammatory pattern of frequent and infrequent exacerbators ([81]).

It is therefore important to view any therapeutic drug effects on exacerbations with caution, because exacerbations can also be effectively prevented by long-acting bronchodilators without altering the underlying and yet poorly understood inflammatory process.

The effect of ICS treatment on exacerbation rates in COPD has been investigated by numerous well-controlled, often large-scale clinical trials. Publications from recent years have generated a relatively consistent body of evidence supporting the benefit of ICS in reducing exacerbations in certain COPD subgroups. However, it should be noted that by far not all studies have produced a positive outcome in this regard, due to differences in study duration, dosage, size effect, exacerbation definition, and study population.

In a study by Dompeling et al. ([68]), the exacerbation rate under beclomethasone (2 × 400 mcg) was equal to the rate during a control period. Further, Renkema et al. ([67]) reported a similar exacerbation rate over two years for budesonide, budesonide plus prednisolone, and placebo. A non-significant overall reduction of exacerbations was found in a study by Weir et al. (BDP 2 × 750 mcg/2 × 1000 μg over 2 years) ([47]). A similar observation was made by Paggiaro et al., where FP 2 × 500 mcg numerically, but not significantly reduced exacerbations. However, moderate to severe exacerbations were significantly lower with FP in this trial ([48]).

In the ISOLDE study, FP 2 × 500 mcg reduced the exacerbation rate by 25% (0.99 versus 1.32 during placebo, p = 0.026). However, using a post-hoc analysis, this effect was restricted to a subgroup of COPD patients with severe COPD (post-bronchodilator FEV₁ < 50% of predicted), whereas no overall effect was found for less severe COPD patients ([82]).

The TRISTAN study ([41]) found a significant reduction of exacerbations with high-dose fluticasone propionate (FP) monotherapy (500 mcg bid) over placebo (exacerbation rate 1.05/year versus 1.30/year, p = 0.003). Similar results were reported by Szafranski et al. using budesonide (BUD) 400 mcg bid in subjects with moderate to severe COPD. However, BUD reduced only mild exacerbations by 41%, while there was no significant effect on more severe (requiring systemic steroids and/or antibiotics and/or hospitalisation) exacerbations.

In contrast, Calverley et al. using a different design (pretreatment with formoterol plus oral steroid prior to randomization to either budesonid 400 mcg bid, formoterol 9 mcg bid, budesonide/formoterol combination product, or placebo) ([42]), reported opposite results: the number of severe, corticosteroid-dependent exacerbations was significantly reduced by BUD monotherapy versus placebo (0.87/year versus 1.14/year, p = 0.044), while there was no effect on the time to first exacerbation (p = 0.512) or overall exacerbation rate (p = 0.308). However, the COPE ([44]) study, using a similar “steroid withdrawal” design, showed an increased exacerbation risk after withdrawal of FP compared with continued FP treatment (RR 1.5, 95% CI 1.05–21.). Finally, in the TORCH trial ([22]), FP monotherapy also reduced the overall moderate and severe exacerbation rate over 3 years by 18% versus placebo (p < 0.001).

In summary, the majority of controlled clinical trials found a beneficial effect of chronic ICS treatment on moderate and severe exacerbation frequency in COPD patients. However, the results were variable, and most of the studies reporting a positive outcome used high doses of ICS. Further, the effects of ICS monotherapy on exacerbations are not superior to those observed with long-acting bronchodilator monotherapy, when directly compared ([22],41). Thus, it remains questionable, whether the observed effects are truly representative of a direct anti-inflammatory effect of ICS in COPD. Further, the effects of ICS on exacerbations appear to be restricted to severe COPD patients with a history of repeated exacerbations. This is also in contrast to studies with long acting anticholingergic bronchodilators, showing a reduction in exacerbations also in milder COPD patients ([83]).

Effects of ICS on mortality

The effect of ICS on COPD mortality was at centre stage of a long-prevailing dispute in the respiratory community. While potential positve effects were reported from some retrospective database studies, the discussion very much focussed on methodological issues in statistical trials involving clinical databases. On the one hand, a beneficial effect of ICS on overall mortality in COPD was postulated, while on the other hand, these results were claimed to be invalid due to an immortal time bias ([84],85,86,87,88,89,90). Until prospective studies investigating the effect of ICS on mortality were undertaken, this controversy remained unresolved.

The question whether ICS may reduce mortality in COPD has finally been addressed in a large-scale prospective clinical trial ([22]), the TORCH study, where ICS monotherapy – in sharp contrast to earlier hypotheses - actually did not signifcantly differ from death rates in the placebo group (hazard ratio for FP 1.06, p = 0.53), LABA monotherapy (hazard ratio 0.879 versus placebo, p = 0.18), and LABA/ICS fixed combination (hazard ratio versus placebo 0.825, p = 0.052, hazard ratio versus FP monotherapy 0.774, p = 0.007). In this regard, a solid linkage of bronchial or systemic inflammation with mortality in COPD still remains to be established.

Effects of ICS on surrogate markers of inflammation

Studying the effect of ICS on markers of inflammation in COPD is complicated by the fact, that there are few accepted inflammatory “biomarkers”. Bronchial biopsies are regarded as gold standard to evaluate inflammation in COPD, however, a concise correlation of biopsy markers with clinical endpoints has not been established. Besides biopsies, the most frequently used method to study bronchial inflammation in COPD is induced sputum cell counts ([14]).

Therapeutic agents with a putative anti-inflammatory mode of action should have a beneficial effect on the typical pattern of inflammation in COPD. In turn, any potential clinical effects, e.g., on lung function, should be confirmed by consistent effects on inflammation to make a true “anti-inflammatory” effect plausible on a biological level. What appears logical at first glance, does not necessarily need to be linked with COPD, as described for the PDE4- inhibitors below, which have shown anti-inflammatory effects in controlled trials, whereas their impact on major clinical endpoints has not been proven. On the other hand, whether the clinical benefits seen with ICS, i.e., the modest reduction of exacerbations and improvements in lung function, are due to the anti-inflammatory mode of action or to other, as yet undiscovered mechanisms, needs to be critically scrutinized.

In this respect, it becomes clear that there is no consistent effect of ICS on markers of inflammation in the literature. Gizycki et al. ([91]) and Hattotuwa et al. ([92]) studied the effect of high-dose FP (1000 mcg per day) on the cellular pattern of mucosal biopsies in COPD patients. Neither analysis by light microscopy ([92]) nor by transmission electron microspopy ([91]) revealed an effect of FP on the number of CD3+−, CD4+− or CD8+− lymphocytes, EG2+-cells (eosinophils), CD45+ (Memory-T)-cells, neutrophil elastase + cells (neutrophils), or CD68+−cells (macrophages). There was a small effect on mast cell numbers in both studies (p = 0.04,⁹² and p < 0.05,⁹¹ respectively).

Verhoeven et al. studied the effect of 6 months treatment with high-dose FP (100 mcg per day) in a subgroup of COPD patients with demonstrable airway hyperreactivity ⁹³. Despite improvements in FEV₁, there was no effect on hyperreactivity, and inflammatory cells compared to placebo. Similar results have been observed in a recent biopsy study comparing the effects of a LABA/ICS combination versus ICS monotherapy and placebo on inflammatory cells in COPD patients ([53]). While the LABA/ICS fixed combination significantly reduced some types of inflammatory cells, ICS monotherapy did not show any beneficial effect versus placebo. In contrast, it was shown that airway neutrophilia was significantly increased by FP versus placebo, a finding compatible with the in vivo and in vitro effects of corticosteroids on neutrophil biology ([94]). Thus, these and other ([95]) recent findings raise the question, whether the “anti-inflammatory” component of LABA/ICS combinations contributes to any anti-inflammatory effects.

Studies using induced sputum to investigate anti-inflammatory effects of ICS in COPD are often limited by their small size or short-term duration. While studies by Sugiura ([96]) (4 weeks, 800 mcg BDP), Keatings ([97]) (2 weeks, 1600 mcg budesonide), Culpitt ([98]) (4 weeks 1000 mcg FP), Loppow ([99]) (4 weeks, chronic bronchitis, 1000 mcg FP) showed no effect of the ICS on sputum total cells and differential, Confalonieri ([100]) (8 weeks, 1500 mcg BDP), Mirici ([84]) (12 weeks, 800 mcg budesonide), and Yildiz ([101]) (8 weeks, 1500 mcg FP) demonstrated some positive effects of ICS treatment on neutrophils or lymphocytes ([84]) in induced sputum samples.

Two further studies investigated the effect of ICS treatment on bronchoalveolar lavage cells in COPD. Ozol et al. ([102]) (6 months, 800 mcg budesonide) and Balbi et al. ([103])(6 weeks, 1500 mcg BDP) found a significant reduction of BAL neutrophils and interleukin-8.

Gan et al. conducted a meta-analysis including all above cited studies on the effect of ICS on sputum cells in COPD. In their pooled analysis, a total effect of a approximate 2% reduction of sputum neutrophils by ICS was calculated. Given the high number of sputum neutrophils in COPD usually approaching 70% of all sputum cells ([104]), the clinical relevance of their calculated effect remains unclear.

Selective PDE-4 inhibitors

The failure of anti-inflammatory drugs like glucocorticoids to have a major impact on the progression of COPD has stimulated the development of novel anti-inflammatory therapies such as selective phosphodiesterase-4 (PDE4) inhibitors. PDE4 inhibitors increase the intracellular concentration of cyclic adenosine monophosphate (cAMP), which has a broad spectrum of anti-inflammtory effects on various target cells commonly linked to airway inflammation in COPD such as neutrophils, macrophages and T cells ([105]). The second generation, orally active PDE4-inhibitors in the late phase III of clinical development, are roflumilast and cilomilast, the latter being less receptor selective ([106]).

Both cilomilast and roflumilast have been studied in trials using bronchial biopsies and/or induced sputum to evaluate their effect on mucosal and airway inflammation in COPD. Both drugs reduced inflammation in biopsies and sputum in well-controlled clinical trials ([16],17). However, these anti-inflammatory effects were not accompanied by meaningful clinical benefits, underlining the pitfalls and imponderabilities of developing anti-inflammatory drugs for COPD, when there is no clear link between inflammation and accepted clinical endpoints. To date, four placebo-controlled, double-blind randomised clincial trials (RCTs) of 647 up to 825 patients with moderate to severe COPD have evaluated the efficacy of cilomilast 15 mg twice daily over 24 weeks and shown variable effects on exacerbations and quality of life with small improvements in measures of lung function (FEV₁), which was also not consistently demonstrable throughout all trials. Adverse events including nausea, headache and GI events were generally mild and self-limiting ([107],106).

The clinical experience with roflumilast is similar to that of cilomilast except for a better safety and tolerability profile with the main adverse events being diarrhea and nausea ([108]). In a 6-month phase III double-blind, placebo-controlled RCT of 1411 patients with moderate to severe COPD, Roflumilast 500 μg once daily produced modest but significant improvements in lung function without changing the frequency of moderate to severe exacerbations or health-related quality of life in a clinically relevant manner (109_. These findings were basically confirmed in another double-blind, placebo-controlled RCT over 1 year in 1513 patients with severe to very severe COPD ([18]). A post-hoc analysis indicated, however, that roflumilast 500 μg once daily may reduce the rate of moderate to severe exacerbations in a subgroup of very severe COPD patients.

Based on the available data, the superior anti-inflammatory activity and additional therapeutic value of selective PDE4 inhibitors versus inhaled glucocorticoids as potential combination partners to long-acting bronchodilators in patients with severe forms of COPD still remains to be shown. So far, clinical trials with selective PDE4 inhibitors have not met the expectations of a novel disease-modifying drug and may indicate that dose-limiting side effects may prevent the full clinical potency of current selective PDE4 inhibitors from being realized.

SUMMARY AND CONCLUSION

To date, there is still no effective anti-inflammatory treatment for COPD. While there are some modest effects of ICS and to a lesser extent of PDE4 inhibitors on clinical endpoints such as exacerbations, it remains to be proven whether the observed effects are due to changes of the underlying inflammation. This is even more important, since exacerbations can also be significantly reduced by non anti-inflammatory treatments, namely long-acting bronchodilators and lung volume reduction surgery. Furthermore, inhaled corticosteroids have not been shown yet to alter the progression of COPD including mortality. Thus, it may be necessary to fundamentally re-evaluate the scientific foundations of a connection between inflammation and well-accepted clinical endpoints in COPD. This would require a fundamental better understanding of the varied underlying and stage-dependent inflammatory mechanisms involved in this disease and will also form the basis for targeted treatments in the future. Clearly, one has to acknowledge the difficulty in non-invasively determining inflammation in COPD patients and to identify potentially important unique aspects of individual patients rather than the whole group, given the heterogenous nature of inflammation. However, until then, current evidence supports updated COPD guidelines that recommend the additional use of anti-inflammatory treatment with ICS only in those severe COPD patients with a history of repeated exacerbations.

Declaration of interest

KMB received compensation for serving on advisory boards for Novartis, Germany, and Boehringer Ingelheim, Germany. He has participated as a speaker in scientific meetings or courses organized and financed by various pharmaceutical companies (AstraZeneca, GSK, Boehringer, Novartis, Pfizer) in 2003–2008. The institution where KMB currently is employed has received compensations for designing and participating in single or multi-centre clinical trials in 2004–2008 from several companies (Altana, AstraZeneca, Boehringer, Cytos, Novartis, GSK, Revotar Biopharmaceuticals, EpiGenesis, Corus Pharma, Almirall Prodesfarma, Merck Sharp & Dohme, Fujisawa, Pfizer, Medapharma). KMB has been reimbursed for travel expenses by Boehringer, GSK, Novartis, and Pfizer for attending and presenting at conferences. TG is currently an employee of Boehringer Ingelheim, Germany.