Letter to editor – a response to: ‘efficacy and safety of triple combination therapy for treating chronic obstructive pulmonary disease: an expert review’

Dear Editors,

We read with interest the review article by Ritondo et al. on triple therapy in chronic obstructive pulmonary disease (COPD) [1], but disagree with some of the points raised by the authors in this article.

The authors note in their discussion that ‘In the IMPACT study, many patients assigned to the LABA/LAMA group were abruptly stepped down in their treatment and ICS withdrawn, which could have triggered COPD exacerbation’[1]. In this regard, we would like to highlight that IMPACT was not designed as an ICS withdrawal study, and that owing to the 2:2:1 ICS/LAMA/LABA:ICS/LABA:LAMA/LABA randomization scheme the vast majority of patients (~85%) did not experience ICS withdrawal at randomization since they either had not been on an ICS prior to the study or remained on an ICS-containing medication. Recently published post hoc analyses of IMPACT show that the benefits of triple therapy versus LAMA/LABA in terms of exacerbation reduction or improvements in lung function and health-related quality of life are not related to abrupt ICS withdrawal [2]. The authors also question the results of IMPACT and TRIBUTE, arguing that ‘both TRIBUTE and IMPACT studies permitted the inclusion of patients with a prior history of asthma, an important risk factor for severe COPD exacerbation that could have influenced the results and perhaps led to an exaggeration of the benefits of ICS’. It is curious that these two trials are highlighted in this way, since a history of asthma was not an exclusion criterion in any of the other studies included in the article by Ritondo et al. In fact, reversibility to salbutamol was lower in IMPACT (18% of patients were reversible to salbutamol) [3], than in KRONOS (43%) [4], or ETHOS (30.7%) [5] (percentage of patients reversible to salbutamol were not reported in studies of BDP/FOR/GLY).

The authors also mention that ‘there are doubts concerning the potential role of blood eosinophils in COPD patients since available data mainly derive from post-hoc and retrospective analyses of several large trials’ [1]. In this regard, we just wish to highlight that the analyses by eosinophil subgroup in both IMPACT and ETHOS, the two largest trials to date, were pre-specified [3,5]. In both studies, the exacerbation reduction benefits of triple therapy versus either ICS/LABA or LAMA/LABA dual therapy was seen irrespective of blood eosinophil count, with greater benefits in patients with higher blood eosinophil counts for triple therapy versus LAMA/LABA [3,5].

Finally, they conclude that current evidence indicates that combinations including budesonide (BUD) or beclomethasone dipropionate (BDP) are characterized by a superior efficacy/safety profile to fluticasone furoate (FF). A critical examination of the data fails to support this conclusion.

The efficacy of all inhaled corticosteroids (ICS) in COPD is very similar. Regardless of background bronchodilator therapy, the addition of an ICS reduces exacerbations by about 25%. This was first demonstrated in the ISOLDE study with a background of short-acting bronchodilators [6], and was confirmed by the Cochrane review of ICS/long-acting β₂-agonist (LABA) compared with LABA alone showing almost identical results for BUD and fluticasone propionate (FP) [7]. It is further supported by recent studies of single-inhaler triple therapy, which have shown that the addition of either BUD or FF to a dual bronchodilator produces a reduction in exacerbations of approximately 25% [3,5]. Other efficacy outcomes such as forced expiratory volume in 1 second (FEV₁) and St George’s Respiratory Questionnaire (SGRQ) are similar [3,8,9]. Therefore, one must conclude that in terms of efficacy all ICSs produce very similar results.

To state there is a difference in the therapeutic ratio therefore requires the authors to demonstrate that there is a difference in the safety profile, of which the most important aspect is pneumonia. There are two factors that need to be taken into account in interpreting the data. Firstly, the risk of pneumonia is significantly influenced by several risk factors, including age, body mass index, airflow obstruction, as well as exacerbation and pneumonia history, with patients with more severe COPD having increased pneumonia risk [10,11]. The second important factor is the method by which the pneumonia is determined. In some studies, such as IMPACT, pneumonia events were captured from the adverse event report, according to the clinical judgment of the investigator; chest x-rays were required by protocol for all suspected pneumonias or exacerbations and were independently reviewed [3]. In contrast in the ETHOS study, the pneumonia rate was from pneumonias adjudicated as such by an expert panel [5]. These two different methods of capturing pneumonia will produce different rates, with adjudicated pneumonia rates expected to be lower, and it is therefore important to compare the rate ratio between the ICS-containing group and the LABA/LAMA group within each study.

When that is done, in studies that enrolled patients with a low exacerbation rate, such as KRONOS, TRIBUTE or SUMMIT, there is little evidence that ICS increase the rate of pneumonia [4,12,13]. That is to be expected because such patients are at low risk of ICS-induced pneumonia [10]. In contrast, in studies that enrolled patients with a history of frequent exacerbations, such as IMPACT and ETHOS, there is evidence of an increased pneumonia rate with triple therapy compared with the LAMA/LABA arm, with very similar pneumonia rate ratios between these treatment arms in both studies (1.6 in IMPACT and 1.7 in ETHOS) [3,5]. In terms of head-to-head comparisons, in the FULFIL study there was a low rate of pneumonias and while there was an increased incidence of pneumonia with FF/umeclidinium/vilanterol (FF/UMEC/VI) versus BUD/FOR at 24 weeks, this difference disappeared by 52 weeks [14].

This is recognized in a statement of the European Medicines Agency’s Pharmacovigilance Risk Assessment Committee, which states that there is no evidence for intraclass difference in pneumonia rates and that it is a class effect [15]. In addition, as noted above, cross-study comparisons should be avoided as trial populations have different underlying pneumonia risk and definition and capture of pneumonia events vary between studies. It is therefore our view that the conclusion regarding the differential benefit:risk profile of triple therapies in the review by Ritondo et al. is incorrect, and if the data are analyzed correctly the benefit:risk ratio for all ICSs is very similar.

Declaration of interest

NC Barnes, P Jones, and DA Lipson are employees of GSK and hold stocks and shares in GSK. D. Singh has received personal fees from GSK, Cipla, Genentech and Peptinnovate, and personal fees and grant support from AstraZeneca, Boehringer Ingelheim, Chiesi, Glenmark, Menarini, Mundipharma, Novartis, Pfizer, Pulmatrix, Theravance and Verona. FJ Martinez has received personal fees and non-financial support from AstraZeneca, Boehringer Ingelheim, Genentech, GSK, Inova Fairfax Health System, Methodist Hospital, National Society for Continuing Education/Haymarket, Novartis, Pearl Pharmaceuticals, PeerView Communications, Physicians Education Resource, Chiesi, CSL Behring, Sunovion, University of Alabama Birmingham, Physicians Education Resource, Canadian Respiratory Network, CME Outfitters, Teva, Vindico and Dartmouth, non-financial support from Bioscale/ProterrixBio, Nitto and Zambon, personal fees from MD Magazine, New York University, UpToDate, WebMD/MedScape, Patara/Respivant, Rockpointe, Rare Disease Healthcare Communications, France Foundation and Prime Education and other support from AstraZeneca (non-personal travel support; academic co-authorship), Boehringer Ingelheim (steering committee; non-personal travel support), Bioscale/ProterrixBio (advisory board attendance with no direct financial compensation), Afferent/Merck (IPF Steering Committee), Gilead Sciences (co-authorship on IPF Study Steering Committee), Patara/Respivant (IPF Steering Committee), Stromedix/Biogen, Veracyte (co-authorship on IPF Study Steering Committee), Prometic (IPF Steering Committee), Bayer (ILD Steering Committee), Bridge Biotherapeutics (IPF consultation), Bristol-Myers Squibb (IPF consultation), Physicians Education Resource (IPF Advisory Board), ProMedior/Roche (IPF Steering Committee), twoXR (IPF Teleconsulation), and Gala (Advisory Board attendance). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Acknowledgments

D Singh is supported by the National Institute for Health Research (NIHR) Manchester Biomedical Research Centre (BRC). Editorial support (in the form of collating author comments, grammatical editing and referencing) was provided by Chrystelle Rasamison, at Fishawack Indicia Ltd, part of Fishawack Health, and was funded by GlaxoSmithKline (GSK).

Additional information

Funding

The IMPACT trial was funded by GSK (study CTT116855; NCT02164513).