Comparative effectiveness and safety of single inhaler triple therapies for chronic obstructive pulmonary disease: new user cohort study
BMJ 2024; 387 doi: https://doi.org/10.1136/bmj-2024-080409 (Published 30 December 2024) Cite this as: BMJ 2024;387:e080409Linked Editorial
Environmentally friendly inhaler regimens for COPD
- William B Feldman
, assistant professor1 2 3, - Samy Suissa, professor4,
- Aaron S Kesselheim, professor1 3,
- Jerry Avorn, professor1 3,
- Massimiliano Russo, assistant professor5,
- Sebastian Schneeweiss, professor1 3,
- Shirley V Wang, associate professor1 3
- 1Division of Pharmacoepidemiology and Pharmacoeconomics, Department of Medicine, Brigham and Women’s Hospital, Boston, MA 02120, USA
- 2Division of Pulmonary and Critical Care Medicine, Department of Medicine, Brigham and Women’s Hospital, Boston, MA, USA
- 3Harvard Medical School, Boston, MA, USA
- 4Centre for Clinical Epidemiology, Lady Davis Institute for Medical Research, Jewish General Hospital, Montréal, McGill University, Montreal, QC, Canada
- 5Department of Statistics, Ohio State University, Columbus, OH, USA
- Correspondence to: W B Feldman wbfeldman{at}bwh.harvard.edu (or @wbfeldman on X)
- Accepted 6 November 2024
Abstract
Objective To compare the effectiveness and safety of budesonide-glycopyrrolate-formoterol, a twice daily metered dose inhaler, and fluticasone-umeclidinium-vilanterol, a once daily dry powder inhaler, in patients with chronic obstructive pulmonary disease (COPD) treated in routine clinical practice.
Design New user cohort study.
Setting Longitudinal commercial US claims data.
Participants New initiators of budesonide-glycopyrrolate-formoterol or fluticasone-umeclidinium-vilanterol between 1 January 2021 and 30 September 2023 who had a diagnosis of COPD and were aged 40 years or older.
Main outcome measures In this 1:1 propensity score matched study, the main outcome measures were first moderate or severe COPD exacerbation (effectiveness) and first admission to hospital with pneumonia (safety) while on treatment. Potential confounders were measured in the 365 days before cohort entry and included in propensity scores. Hazard ratios and 95% confidence intervals (CIs) were estimated using a Cox proportional hazards regression model.
Results The study cohort included 20 388 propensity score matched pairs of new users initiating single inhaler triple therapy. Patients who received budesonide-glycopyrrolate-formoterol had a 9% higher incidence of first moderate or severe COPD exacerbation (hazard ratio 1.09 (95% CI 1.04 to 1.14); number needed to harm 38) compared with patients receiving fluticasone-umeclidinium-vilanterol and an identical incidence of first admission to hospital with pneumonia (1.00 (0.91 to 1.10)). The hazard of first moderate COPD exacerbation was 7% higher (1.07 (1.02 to 1.12); number needed to harm 54) and the hazard of first severe COPD exacerbation 29% higher (1.29 (1.12 to 1.48); number needed to harm 97) among those receiving budesonide-glycopyrrolate-formoterol compared to fluticasone-umeclidinium-vilanterol. Prespecified sensitivity analyses yielded similar findings to the primary analysis.
Conclusions Budesonide-glycopyrrolate-formoterol was not associated with improved clinical outcomes compared with fluticasone-umeclidinium-vilanterol. Given the added climate impact of metered dose inhalers, health systems seeking to decrease use of these products may consider steps to promote further prescribing of fluticasone-umeclidinium-vilanterol compared with budesonide-glycopyrrolate-formoterol in people with COPD.
Study registration Center for Open Science Real World Evidence Registry (https://osf.io/6gdyp/).
Introduction
International guidelines recommend triple inhaler treatment for some patients with chronic obstructive pulmonary disease (COPD), consisting of an inhaled corticosteroid, a long acting muscarinic antagonist, and a long acting β-agonist.1 Two single inhalers with these triple therapies are marketed in the US: budesonide-glycopyrrolate-formoterol (Breztri Aerosphere), a twice daily metered dose inhaler, and fluticasone-umeclidinium-vilanterol (Trelegy Ellipta), a once daily dry powder inhaler. Both combinations have shown superiority over long acting muscarinic antagonist-long acting β-agonists and inhaled corticosteroid-long acting β-agonists in pivotal randomized controlled trials in select patients with COPD,23 and may be prescribed according to clinical guidelines when inhaled corticosteroid-long acting muscarinic antagonist-long acting β-agonists are indicated.1
Many health systems worldwide have sought to reduce reliance on metered dose inhalers like budesonide-glycopyrrolate-formoterol since they contain propellants (hydrofluoroalkanes) that contribute to greenhouse gas emissions that are not found in dry powder inhalers; for example, emissions associated with metered dose inhalers represent 3% of the entire carbon footprint of the National Health Service in the UK.45 Yet, metered dose and dry powder formulations of single inhaler triple therapy have not been compared head-to-head in randomized controlled trials.
Research into the comparative effectiveness and safety of triple therapy in people with COPD has generated mixed findings. The adverse effect of pneumonia has been consistently shown in clinical trials and observational research of inhaled corticosteroids.23678 But one recent study using data from the UK found that budesonide-based triple therapy was associated with fewer admissions to hospital with pneumonia and reduced all cause mortality compared with fluticasone-based triple therapy.9 This finding was consistent with earlier observational studies and a systematic review showing decreased pneumonia risk for patients with COPD who received budesonide-based rather than fluticasone-based regimens.10111213 Some researchers have hypothesized that fluticasone could increase pneumonia risk due to slower airway absorption, higher lipophilicity, and more potent immunosuppressive effects.13
These studies, however, have not analyzed potential intraclass differences among patients receiving single inhaler triple therapy. The recent UK study included patients receiving triple therapy via separate inhalers and using a variety of inhaler types (single inhaler triple therapy was not yet available during the study period); other studies focused on dual combination rather than triple combination regimens, which may limit generalizability to patients with more severe disease. In addition, prior observational studies comparing budesonide and fluticasone analyzed short acting fluticasone propionate rather than the longer acting fluticasone furoate, which is found in fluticasone-umeclidinium-vilanterol.
In contrast to studies suggesting a potential safety advantage for budesonide-based triple therapy, other recent research has suggested reduced effectiveness. A network meta-analysis of randomized controlled trials found that budesonide-glycopyrrolate-formoterol was associated with more annual moderate or severe COPD exacerbations than fluticasone-umeclidinium-vilanterol.14 However, this study aggregated data from trials with different inclusion criteria, with some systematically enrolling patients with more severe COPD, and thus comparisons across trials may be subject to bias. In addition, the authors did not analyze pneumonia risk in the two treatment groups because of inconsistencies in how different trials defined the outcome.
Given ongoing clinical uncertainty, we sought to compare the effectiveness and safety of budesonide-glycopyrrolate-formoterol and fluticasone-umeclidinium-vilanterol in patients with COPD who are treated in routine clinical practice. Rigorous studies of longitudinal health care data offer an important tool to help inform treatment guidelines and shape prescribing practices when randomized controlled trials have not been performed.15 Such research may be especially valuable for developing a generalizable evidence-base in COPD because patients treated in routine clinical practice tend to be older, have more comorbidities, and include more women than patients who are generally enrolled in randomized controlled trials.1617181920212223
Methods
Study cohort
The study was completed using Optum’s de-identified Clinformatics Data Mart, a database of administrative health claims for members of large commercial and Medicare Advantage health plans (appendix methods). We included patients who initiated budesonide-glycopyrrolate-formoterol (exposure) or fluticasone-umeclidinium-vilanterol (referent) between 1 January 2021 and 30 September 2023 in the analysis. The study began in 2021, the first full year after budesonide-glycopyrrolate-formoterol was approved in the US; both the exposure and referent were marketed in the US throughout the study period.
All patients were required to have a diagnosis of COPD based on International Classification of Diseases, tenth revision, clinical modification (ICD-10-CM) codes (J41.x, J42.x, J43.x, J44.x). We included patients with three outpatient claims or one inpatient claim in the prior three years (positive predictive value 0.82 (95% confidence interval (CI) 0.72 to 0.89)) to capture patients with active COPD.24 We excluded people younger than 40 years to increase the specificity of the COPD diagnosis, and we required that all patients have at least 365 days of continuous enrollment in the dataset before cohort entry. Patients with COPD who also had prior asthma diagnoses were included in the analysis. Patients were excluded if they had received either budesonide-glycopyrrolate-formoterol, fluticasone-umeclidinium-vilanterol, or an inhaled corticosteroid-long acting muscarinic antagonist-long acting β-agonist combination via separate inhalers (defined as dispensing of inhalers with ingredients from all three classes via any combination within 30 days of each other) during the 365 days before cohort entry. We also excluded patients who received both budesonide-glycopyrrolate-formoterol and fluticasone-umeclidinium-vilanterol or who received triple therapy plus another maintenance inhaler on the cohort entry date. By contrast, patients who received monotherapy (long acting muscarinic antagonist, long acting β-agonist, or inhaled corticosteroid) or dual therapy (long acting muscarinic antagonist-long acting β-agonist or inhaled corticosteroid-long acting β-agonist) during the baseline assessment period were included in the analysis.
Assessment of covariates
Potential confounders, including baseline lung disease, comorbidities, healthcare use, and medication use, were measured leading up to and including the 365 days before cohort entry and were included in propensity scores (see appendix methods for a complete list of covariates included in the propensity score model). We assessed socioeconomic covariates (mean copayments, total copayments, and unique brand-to-generic prescription fills) until the day before cohort entry. Any prior history of asthma was assessed using all available data. Eosinophil levels can fluctuate over time, therefore, we used the most recent measured value in the 180 days before cohort entry. We measured demographic covariates (age, gender, race, and region), calendar year and season of index prescription, and whether the index prescription was written by a pulmonologist on the day of cohort entry (see appendix figure 1 for a timeline of covariate assessment).
Primary outcomes and follow-up
The primary outcomes were the first moderate or severe COPD exacerbation (effectiveness) and the first admission to hospital with pneumonia (safety). Moderate exacerbations were defined by fills of prednisone prescribed for 5-14 days (positive predictive value 0.73),25 and severe COPD exacerbations by admission to hospital with a COPD diagnosis code (specified above) in the primary position (positive predictive value 0.86 based on ICD-9-CM codes; conversion to ICD-10-CM codes was performed based on clinical review).26 If a patient met criteria for both a moderate and severe COPD exacerbation within 14 days of each other, the exacerbation was considered to be a severe exacerbation starting on the day when criteria for the exacerbation were first met. This categorization of COPD exacerbations requiring systemic steroids (moderate) versus those requiring admission to hospital (severe) is used in the Global Initiative for Chronic Obstructive Lung Disease guidelines and has been widely used in randomized controlled trials and other observational studies.123678272829 Admission to hospital with pneumonia was defined based on ICD-10-CM codes (J.09.X1, J10.xx-J18.x, A01.03, A02.22, A37.01, A37.11, A37.81, A37.91, A54.84, B01.2, B05.2, B06.81, B77.81, J85.1, J22) in any position(positive predictive value 0.88 based on ICD-9-CM codes; conversion to ICD-10-CM codes was performed based on clinical review).30 Patients were followed up for up to one year, the end of data, or until they had an outcome of interest, discontinued treatment (with a grace period of 60 days permitted between inhaler fills), added or switched maintenance inhalers, died, or were disenrolled from insurance.
Prespecified secondary, subgroup, and sensitivity analysis
Secondary outcomes included all cause mortality, first moderate exacerbation, first severe exacerbation, annual rate of moderate or severe exacerbations (analyzed separately and as a composite), and annual rate of admission to hospital with pneumonia. Unadjusted and adjusted results are reported for primary and secondary outcomes.
Subgroups of interest included patients who: (1) had Global Initiative for Chronic Obstructive Lung Diseasegroup E disease; (2) had at least one moderate or severe COPD exacerbation during the baseline assessment period; (3) had at least one severe COPD exacerbation during the baseline assessment period; (4) received any baseline maintenance inhaler before cohort entry; (5) had a prior diagnosis code for asthma; (6) had a diagnosis of asthma in the prior three years; (7) had eosinophil levels above 300 µL; (8) underwent spirometry during the baseline assessment period; and (9) received the initial budesonide-glycopyrrolate-formoterol or fluticasone-umeclidinium-vilanterol prescription from a pulmonologist (appendix table 1). All subgroup analyses were done for patients meeting the relevant subgroup criterion and for the remainder of patients not meeting the subgroup criterion.
To check the robustness of our findings, we conducted several prespecified sensitivity analyses. We varied the grace period permitted between prescription fills (from 60 days to 30 days and 90 days), performed an as-started analysis (with no censoring for treatment discontinuation or switching in the matched population), and excluded early events (in the first 30 days and first 60 days after cohort entry) (appendix table 2). We also varied the definitions of outcomes in our analysis (appendix methods), and we did a post-hoc sensitivity analysis limiting follow-up to 180 days.
Statistical analysis
Our primary analysis was a 1:1 propensity score matched analysis. We used nearest neighbor matching and a caliper of 0.01 on the propensity score scale. We estimated propensity scores using logistic regression, including all covariates previously mentioned without further variable selection. The targeted estimand was the average treatment effect among patients with COPD receiving budesonide-glycoypyrrolate-formoterol. In the case of missing data for covariates in our model, we used a missing indicator. For our time-to-event analyses, we estimated hazard ratios (HRs) and 95% CIs using a Cox proportional hazards regression model. We estimated absolute risk differences at 365 days, and we calculated the number needed to harm as 1/absolute risk difference when the 95% confidence interval for the risk difference excluded the null.31 When analyzing annual rates of COPD exacerbations and admission to hospital with pneumonia in secondary analyses, we performed negative binomial regression. In prespecified sensitivity analysis, we conducted 1:1 high dimensional propensity score matching to adjust for hundreds of covariates generated through automated selection based on outpatient and inpatient diagnosis codes, procedures, and pharmacy claims.323334
All analyses were completed using the Aetion Evidence Platform v4.73 (Aetion, Inc), which carries out some statistical computations using R v3.4.2, and Stata, v16.1 (StataCorp). The Aetion Evidence Platform has been used extensively in the academic literature, including in studies showing the reproducibility of real world evidence and successfully emulating randomized controlled trials.3536 The platform has been through comprehensive internal quality checks and validation processes and has been directly tested against both de novo programming and other quality checked analytical workflows used by the US Food and Drug Administration Sentinel Initiative.37 The study was approved by the Mass General Brigham Institutional Review Board (2023P000164) and the protocol was preregistered at the Center for Open Science Real World Evidence Registry (https://osf.io/6gdyp/) before the study began.
Patient and public involvement
Our dataset included only de-identified patients, and data use agreements precluded us from contacting these patients. In addition, our funding did not support patient or public involvement.
Results
The cohort included 87 751 patients (67 356 new users of fluticasone-umeclidinium-vilanterol and 20 395 new users of budesonide-glycopyrrolate-formoterol) (fig 1), from which 20 388 matched pairs were identified for the primary analysis (see table 1 for select baseline characteristics of the two treatment groups and appendix table 3 for the full set of baseline characteristics). The populations were highly overlapping in the unadjusted cohort and further aligned after matching (appendix figure 2), with mean absolute standardized differences of 0.028 before matching and 0.004 after matching.
Study cohort, including the number of patients excluded and the rationales for exclusion. COPD=chronic obstructive pulmonary disease
Select baseline characteristics of patients initiating single-inhaler triple therapy in the matched cohort
Primary effectiveness and safety outcomes
Patients in the matched cohort had 7729 moderate or severe exacerbations during 15 229 person years of follow-up, giving a crude incidence of 507.5 events per thousand person years. Those receiving budesonide-glycopyrrolate-formoterol had a 9% increased hazard of first moderate or severe COPD exacerbation compared with patients receiving fluticasone-umeclidinium-vilanterol (HR 1.09 (95% CI 1.04 to 1.14)) (table 2). The absolute risk difference at 365 days was 2.6% (95% CI 0.8% to 4.4%) and the number needed to harm 38. Median follow-up time was 88 days (interquartile range 65 to 164 days) among patients receiving budesonide-glycopyrrolate-formoterol and 113 days (74 to 206 days) among patients receiving fluticasone-umeclidinium-vilanterol (reasons for censoring are given in appendix table 4).
COPD exacerbations, admission to hospital with pneumonia, and all cause mortality in patients receiving single inhaler triple therapy
A total of 1812 people had their first admission to hospital with pneumonia in the matched cohort during 17 281 person years of follow-up (crude incidence of 104.9 per 1000 person years). The hazard of first admission to hospital with pneumonia was identical between the two treatment groups (HR 1.00 (95% CI 0.91 to 1.10); absolute risk difference 0.4% (95% CI −0.6% to 1.3%)). Median follow-up time was 105 days (interquartile range 88 to 197 days) among people receiving budesonide-glycopyrrolate-formoterol and 135 days (88 to 243 days) among those receiving fluticasone-umeclidinium-vilanterol (reasons for censoring are given in appendix table 5).
Kaplan-Meier curves for the primary effectiveness and safety analyses showed consistent effects over the 365 days of follow-up (appendix figures 3 and 4). Sensitivity analyses yielded similar findings to the primary analysis when examining the outcomes of first moderate or severe COPD exacerbation (fig 2) and first admission to hospital with pneumonia (appendix figure 5).
Hazard ratios and 95% confidence intervals of first moderate or severe chronic obstructive pulmonary disease (COPD) exacerbation in new users of budesonide-glycopyrrolate-formoterol versus fluticasone-umeclidinium-vilanterol across a range of prespecified sensitivity analyses. Patients receiving budesonide-glycopyrrolate-formoterol had a slightly higher hazard of first moderate or severe COPD exacerbation in most sensitivity analyses
Moderate versus severe COPD exacerbations and all cause mortality
When separately analyzing moderate and severe COPD exacerbations, patients in the matched cohort receiving budesonide-glycopyrrolate-formoterol had a 7% increased relative hazard (HR 1.07 (95% CI 1.02 to 1.12)) and a 1.9% increase in absolute risk ((95% CI 0.1% to 3.6%); number needed to harm 54) of first moderate COPD exacerbation compared with patients receiving fluticasone-umeclidinium-vilanterol. People receiving budesonide-glycopyrrolate-formoterol had a 29% increased relative hazard (1.29 (1.12 to 1.48)) and a 1.0% increased absolute risk ((0.1% to 1.9%); number needed to harm 97) of first severe COPD exacerbation. All cause mortality was similar between the two groups (HR 1.04 (95% CI 0.93 to 1.16); absolute risk difference 0.4% (95% CI −0.6% to 1.3%)). Sensitivity analyses of these secondary endpoints are given in appendix figure 6 (moderate exacerbation), appendix figure 7 (severe exacerbation), and appendix figure 8 (all cause mortality).
Cumulative events
Patients in the matched cohort who received budesonide-glycopyrrolate-formoterol had 8% more cumulative moderate or severe COPD exacerbations (incident rate ratio 1.08 (95% CI 1.04 to 1.13)), 7% more cumulative moderate exacerbations (1.07 (1.02 to 1.12)), and 26% more cumulative severe exacerbations (1.26 (1.09 to 1.45)) during a maximum of one year of follow-up. Rates of cumulative admission to hospital with pneumonia were similar between the two treatment groups (1.03 (0.93 to 1.14)).
Subgroup analysis
Higher hazards of first moderate or severe COPD exacerbation in the matched cohort were observed in patients more prone to exacerbations, including people with at least one baseline moderate or severe exacerbation (HR 1.12 (95% CI 1.07 to 1.17)), at least one baseline severe exacerbation (1.15 (1.04 to 1.28)), prior baseline maintenance inhaler therapy (1.14 (1.08 to 1.20)), and eosinophil concentrations of more than 300 cells/µL (1.20 (1.02 to 1.41)) (fig 3). Such differences were not observed in patients who did not have some of the key indications for initiating triple therapy, including people with no baseline exacerbations (1.02 (0.92 to 1.14)), no baseline maintenance therapy (1.01 (0.94 to 1.10)), eosinophil counts of 100 cells/µL or lower (1.07 (0.94 to 1.22)); or eosinophil counts of between 101-300 cells/µL 1.03 (0.93 to 1.14). The risks of first admission to hospital with pneumonia among patients receiving budesonide-glycopyrrolate-formoterol versus fluticasone-umeclidinium-vilanterol were similar for all subgroups analyzed (appendix figure 9).
Hazard ratios and 95% confidence intervals of first moderate or severe chronic obstructive pulmonary disease (COPD) exacerbation in new users of budesonide-glycopyrrolate-formoterol versus fluticasone-umeclidinium-vilanterol across a range of prespecified subgroup analyses. Patients receiving budesonide-glycopyrrolate-formoterol had a slightly higher hazard of first moderate or severe COPD exacerbation across most subgroup analyses; differences were not observed for those with milder forms of disease, including people who, during the baseline assessment period, had no COPD exacerbations, filled no scripts for maintenance inhalers, received no spirometry, and had a low eosinophil count. GOLD=Global Initiative for Chronic Obstructive Lung Disease
Discussion
Budesonide-glycopyrrolate-formoterol was associated with a slightly higher incidence of first moderate or severe COPD exacerbation and an identical incidence of first admission to hospital with pneumonia, compared with fluticasone-umeclidinium-vilanterol in patients with COPD treated in routine clinical practice. Similar findings were observed when analyzing cumulative annual exacerbations, which more fully reflect the burden of disease experienced by patients. The risks associated with use of budesonide-glycopyrrolate-formoterol were more pronounced when separately analyzing severe COPD exacerbations in a subgroup analysis—with patients receiving budesonide-glycopyrrolate-formoterol having a 29% higher hazard of first severe COPD exacerbation compared with people receiving fluticasone-umeclidinium-vilanterol. Given these outcomes and the differential climate impact of single-inhaler triple therapies, health systems designing formularies and setting treatment guidelines may consider steps to increase use of fluticasone-umeclidinium-vilanterol relative to budesonide-glycopyrrolate-formoterol for patients with COPD.
Mechanistic hypotheses
Several potential reasons could explain why patients receiving fluticasone-umeclidinium-vilanterol seemed to have slightly fewer COPD exacerbations in our study. Fluticasone-umeclidinium-vilanterol is a once daily medication while budesonide-glycopyrrolate-formoterol requires twice daily dosing, and prior studies have found better adherence to inhalers with less frequent dosing.38394041 Although we censored patients at treatment discontinuation, people receiving budesonide-glycopyrrolate-formoterol could have skipped doses at higher rates while on treatment, meaning less consistent therapeutic drug levels before censoring. Another potential explanation concerns the different techniques needed to operate the two inhalers; whereas metered dose inhalers require patients to time their breaths with actuation, dry powder inhalers only require deep inspiration. Research has suggested lower error rates with dry powder inhalers containing fluticasone-umeclidinium-vilanterol than metered dose inhalers.42 Finally, apart from any differences in dosing or delivery devices of the two treatments, the active moieties in fluticasone-umeclidinium-vilanterol could be more effective in preventing COPD exacerbations than the active moieties in budesonide-glycopyrrolate-formoterol. Further research is needed to understand what may be driving the observed differences in our study.
Comparison with previous studies
The similar rates of admission to hospital with pneumonia that we observed between people receiving budesonide-glycopyrrolate-formoterol and fluticasone-umeclidinium-vilanterol stand in contrast to earlier observational studies analyzing budesonide- versus fluticasone-based therapies.9101112 One possible explanation for our findings is that the longer acting version of fluticasone (fluticasone furoate) in single inhaler triple therapy may lead to a lower pneumonia risk than the shorter acting version (fluticasone propionate) analyzed in prior studies. Further research is needed to directly compare inhalers containing fluticasone furoate versus fluticasone propionate in patients with COPD. However, because these prior studies did not control for inhaler device type, another possibility is that the devices themselves mediated the risk of pneumonia. Although budesonide-containing dry powder inhalers for COPD are not available in the US, they are available elsewhere, which would enable comparisons of dry powder inhalers containing budesonide versus fluticasone furoate.
Implications for clinical practice
Our study may provide reassurance to health systems seeking to decrease greenhouse gas emissions by reducing use of metered dose inhalers, because the single inhaler triple therapy with the lower carbon footprint (the dry powder inhaler, fluticasone-umeclidinium-vilanterol) was also associated with slightly improved clinical outcomes. Hydrofluroalkane-134a, the propellant in budesonide-glycopyrrolate-formoterol, has 1430 times the global warming potential as carbon dioxide and contributes to greenhouse gas emissions from metered dose inhalers that are 20 times greater than the emissions from dry powder inhalers during the lifecycles of these products.443 Pharmacies dispense more than 100 million metered dose inhalers in the US each year, representing nearly 90% of all inhalers prescribed, with emissions equivalent to approximately 550 000 gas fueled cars driven annually.4434445 Although dry powder inhalers are associated with other environmental impacts (eg, fossil depletion and marine ecotoxic effects),46 the global warming potential of metered dose inhalers has prompted efforts by many health systems worldwide to increase use of dry powder inhalers, with some countries, including Sweden, Denmark, and Japan, reaching rates of dry powder inhaler prescribing exceeding 50%.45 As health care systems move toward lower carbon inhalers, data for the comparative effectiveness and safety of these products are important to help ensure that patients with chronic lung disease receive optimal, evidence based care.
Limitations
This study has several important limitations. Firstly, while the distributions of characteristics among people receiving the two single inhaler triple therapies were highly overlapping even before matching, the possibility of some residual confounding cannot be ruled out. Metered dose inhalers may be preferred in patients with frailty and poor inspiratory force, and such patients could have been more likely to have exacerbations in follow-up. However, under reasonable assumptions for the prevalence of suboptimal peak inspiratory force, patients receiving budesonide-glycopyrrolate-formoterol would need to be approximately 1.8 times as likely to have suboptimal peak inspiratory force. In addition, patients with suboptimal inspiratory force would need to have approximately 1.8 times the risk of experiencing a moderate or severe COPD exacerbation (appendix figure 10).47 Yet, studies suggest that patients receiving metered dose inhalers in clinical practice are only slightly more likely to have suboptimal peak inspiratory force and only slightly more likely to experience exacerbations.48 We adjusted for numerous covariates associated with frailty and predisposition for exacerbations, including a validated frailty index.49 We also observed similar results when using high dimensional propensity scoring matching, which adjusts for hundreds of empirically selected covariates that may serve as proxies for confounders that are not directly measured. Because some insurance formularies in the US cover only one triple inhaler, the therapy prescribed may be dictated more by formulary design than clinical preference. Still, we cannot exclude the possibility of residual confounding as an explanation for our observation that patients receiving budesonide-glycopyrrolate-formoterol had a 9% higher hazard of moderate or severe COPD exacerbation compared with patients receiving fluticasone-umeclidinium-vilanterol.
Secondly, although we analyzed perhaps the most important safety signal related to inhaled corticosteroids (admission to hospital with pneumonia), we did not analyze other known risks such as oral thrush, osteoporosis, and adrenal insufficiency. Thirdly, rates of non-persistence were high and follow-up time was short, reflecting the reality of routine clinical practice in which patients with COPD often choose to stop taking recommended treatments.50515253 Fourthly, because the study was completed using healthcare claims, we did not have data for daily inhaler use and technique and thus we were unable to draw further conclusions about the source of observed differences in outcomes between patients receiving budesonide-glycopyrrolate-formoterol and fluticasone-umeclidinium-vilanterol. Fifthly, data for eosinophil concentrations were only available for a subset of patients. Sixthly, the study included a broad range of patients with commercial insurance and Medicare Advantage but may not generalize to other groups in the US, particularly the uninsured. Seventhly, we included patients with COPD in our study who also had prior asthma diagnosis codes. Subgroup analyses showed similar findings to the primary analysis when analyzing patients with COPD with no asthma diagnoses in the three years before cohort entry. However, inclusion of these patients, while perhaps increasing the generalizability of our study, could have led to possible treatment heterogeneity among patients with COPD. Eighthly, while we observed that the effect size for severe exacerbations in the matched cohort (HR 1.29 (95% CI 1.12 to 1.48)) became larger after confounding adjustment, this finding occurred in the context of our exploration of multiple subgroups. Therefore, further research is needed to clarify whether this is a chance finding or whether there may be a stronger effect for more severe outcomes. Finally, our study analyzed the only two single agent triple inhalers on the US market; additional studies are needed of other triple therapies available in different parts of the world and of metered dose versus dry powder inhalers across other therapeutic classes, both for the treatment of asthma and COPD.
Conclusions
In a cohort of patients with COPD treated in routine practice, people receiving fluticasone-umeclidinium-vilanterol did not have improved clinical outcomes compared with people receiving budesonide-glycopyrrolate-formoterol. Dry powder inhalers may not be suitable for all patients with COPD. However, fluticasone-umeclidinium-vilanterol represents a safe and effective alternative to budesonide-glycopyrrolate-formoterol for health systems seeking to decrease use of metered dose inhalers.
What is already known on this topic
Triple inhaler therapy is recommended for some patients with chronic obstructive pulmonary disease (COPD) and is available in the United States as budesonide-glycopyrrolate-formoterol and fluticasone-umeclidinium-vilanterol
However, the comparative effectiveness and safety of budesonide-glycopyrrolate-formoterol, a twice daily metered dose inhaler, and fluticasone-umeclidinium-vilanterol, a once daily dry powder inhaler, is unknown
What this study adds
In a cohort of patients with COPD treated in clinical practice, budesonide-glycopyrrolate-formoterol was not associated with improved clinical outcomes compared to fluticasone-umeclidinium-vilanterol
People receiving budesonide-glycopyrrolate-formoterol had a 9% higher hazard of moderate or severe COPD exacerbation and an identical hazard of first admission to hospital with pneumonia compared with people receiving fluticasone-umeclidinium-vilanterol
Ethics statements
Ethical approval
The study was approved by the Mass General Brigham Institutional Review Board (2023P000164).
Data availability statement
No additional data available. Data use agreements do not permit sharing of source data or derivative analytic cohorts.
Footnotes
Contributors: All authors contributed to the conception and design of the manuscript; WBF, MR, and SVW contributed to the statistical analysis; all authors contributed to the interpretation of the data; WBF drafted the manuscript; all authors contributed to critical revision of the manuscript; SSu, ASK, JA, SSc, and SVW supervised the research; WBF obtained the funding. WBF had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.
Funding: This work was funded by a grant (National Heart, Lung, and Blood Institute K08HL163246) to WBF. The funders had no role in the design and conduct of the study; collection, management, analysis, interpretation, of the data; review, or approval of the manuscript; and decision to submit the manuscript for publication.
Competing interests:All authors have completed the ICMJE uniform disclosure form at www.icmje.org/disclosure-of-interest/ and declare: funding from the National Heart, Lung, and Blood Institute; outside the submitted work, WBF serves as a consultant for Alosa Health and as an expert witness in litigation against inhaler manufacturers. He previously served as a consultant for Aetion. SS attended advisory committee meetings for Atara, Novartis and Panalgo, and received speaking fees from Covis Pharma and Novartis. SS is participating in investigator-initiated grants to the Brigham and Women’s Hospital from Boehringer Ingelheim, Takeda, and UCB unrelated to the topic of this study. He owns equity in Aetion Inc, a software manufacturer. He is an advisor to Temedica GmbH, a patient-oriented data generation company. His interests were declared, reviewed, and approved by the Brigham and Women’s Hospital in accordance with their institutional compliance policies. SVW has been a consultant for Veracity Healthcare Analytics, Exponent Inc, and MITRE a federally funded research and development center for the Centers for Medicare and Medicaid for unrelated work.
Transparency: The lead author (the manuscript’s guarantor) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.
Dissemination to participants and related patient and public communities: The work will be disseminated to policymakers, professional societies, and lay audiences via press releases, social media engagement, and a companion blog post.
Provenance and peer review: Not commissioned; externally peer reviewed.
This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.