With the current international pocket guide for diagnosing and managing COPD running to 40 pages,1 a summary of the occasionally confusing current treatment strategies might be useful. The international GOLD guidance suggests that symptomatic patients with COPD may be commenced on single long-acting bronchodilators, in the form of long-acting beta2 agonists (LABAs) or long-acting muscarinic antagonists (LAMAs), and if they remain symptomatic then dual bronchodilator therapy (LABA/LAMA) should be initiated. The addition of inhaled corticosteroid (ICS) to the long-acting bronchodilator(s) may be considered in those who continue to experience exacerbations.1

There is no specific guidance about when to escalate from dual to triple therapy and in light of studies showing an increased risk of pneumonia in patients using ICS, questions remain over when the benefits of ICS outweigh the risks. It is clear that better evidence is needed.

KRONOS study

The KRONOS study2 investigated a new triple inhaler in an attempt to provide such evidence. Using a double-blind, parallel-group, phase 3 randomised controlled design, 1,902 patients with COPD were recruited from hospital clinics and care centres in Canada, the US, Japan and China. Eligible patients were 40 to 80 years old, ex- or current smokers and on dual inhaler therapy. They were randomised to 24 weeks of budesonide/glycopyrrolate/formoterol fumarate (triple therapy), glycopyrrolate/formoterol fumarate (LABA/LAMA) or one of two formulations of budesonide/formoterol fumarate (LABA/ICS).* This approach allowed the authors to compare LABA/LAMA directly with LABA/ICS, and then to see what happens when ICS is added to LABA/LAMA, and when LAMA is added to LABA/ICS. Finally, the researchers analysed what is the best biomarker to predict treatment response.

The results suggest that, when comparing the two, LABA/ICS was better at exacerbation reduction, and LABA/LAMA at improving bronchodilation, measured in this study using the forced expiratory volume in one second (FEV1). And thus, as might be expected, adding a LAMA to LABA/ICS improved bronchodilation, while adding ICS to LABA/LAMA reduced exacerbation frequency. Interestingly, in this study there was no increase in pneumonia risk with inhaled corticosteroid use. It is worth noting that previous work suggests that budesonide, as used in this study, appears safer as regards pneumonia risk than fluticasone propionate.3

Which patients benefit most from triple therapy?

So this all suggests that triple therapy is effective at exacerbation reduction and improving spirometry, and demonstrates a good safety profile. However, it would be helpful if we could target more precisely those who might benefit from the addition of ICS.

Current recommendations rely on exacerbation frequency, based at least partly on the fact that many such studies only recruit ‘frequent exacerbators’. In this study the majority (74%) had not experienced exacerbations in the previous year, yet ICS therapy did still appear to halve their future exacerbation risk, with a predicted fall from around one exacerbation per year to one every two years. So using past exacerbation frequency may not be the best predictor of ICS response.

Is there a biomarker that might do the job better? Based on evidence from previous retrospective studies, blood eosinophils were prospectively trialled in this study. The results fairly convincingly show that the higher the blood eosinophil count the more beneficial the addition of ICS, both in terms of exacerbation frequency (evident from blood eosinophil levels of 0.1 x 109/litre and above) and FEV1 (evident from blood eosinophils 0.25 x 109/litre and above).4


It would seem that patients with symptomatic COPD have around 100ml improvement in FEV1 when LAMA is added to a LABA/ICS, though notably this did not translate to improvement in quality of life scores. Patients appeared to benefit from the addition of ICS irrespective of previous exacerbation frequency, suggesting that past exacerbations are not a good predictor of response to ICS. Blood eosinophils, however, appear much more promising: in those whose eosinophils exceed 0.1 x 109/litre (around 75% of this COPD population) ICS addition convincingly reduced exacerbation frequency. Reassuringly, adding this particular ICS (budesonide) did not increase the frequency of pneumonia. This is a relatively short study at 24 weeks, and a larger and longer study prospectively stratifying patients by blood eosinophil count is currently recruiting. It remains to be seen, however, whether we will have to wait for the results of this trial for our treatment guidelines to change.

Dr Gareth Hynes is a clinical research fellow in severe asthma at the University of Oxford.

*The reason for the two formulations of LABA/ICS is that, for the purposes of a fair comparison, all formulations used in this trial were metered-dose inhalers (MDIs), but the licensed formulation of this particular LABA/ICS combination is a dry powder inhaler (DPI). Therefore both the MDI and DPI versions of LABA/ICS were included to ensure there was no difference between the two formulations (and essentially there were no differences).

This project was initiated and funded by Teva Respiratory. Teva have had no influence over content. Topics and content have been selected and written by independent experts.