Journal Pre-proof Single-inhaler triple therapy fluticasone furoate/umeclidinium/vilanterol versus dual therapy in current and former smokers with COPD: IMPACT trial post hoc analysis

Abstract


Introduction
Chronic obstructive pulmonary disease (COPD) 1 is a leading cause of morbidity and mortality worldwide, characterized by chronic airflow limitation that causes persistent respiratory symptoms.
Smoking is the main risk factor for developing COPD and other comorbidities such as cardiovascular disease. Inhalation of cigarette smoke causes lung inflammation and results in tissue damage and destruction. [1,2] Continued smoking results in worse respiratory symptoms and more severe COPD [3] and is a major contributor to the risk of developing COPD exacerbations. [4] Smoking cessation is recommended for all patients with COPD and has been identified as the most influential factor in altering the disease course.  [2] Studies suggest that continued cigarette smoking impairs responses to ICS. [5][6][7][8][9][10][11] The mechanism underlying ICS response impairment may include pro-inflammatory effects of smoking and/or reduction in histone deacetylase 2 activity (HDAC2) by oxidative stress generated in the lungs from cigarette smoking. [12] The IMPACT trial demonstrated that once-daily single-inhaler triple therapy with ICS/LAMA/LABA (fluticasone furoate, umeclidinium and vilanterol [FF/UMEC/VI]) significantly reduced the annual rate of moderate/severe exacerbations and improved lung function and health status compared with FF/VI or UMEC/VI in patients ≥40 years of age with symptomatic COPD and a history of exacerbations. [13] A previous post hoc analysis of IMPACT has shown smoking status may modify the relationship between blood eosinophil count and the efficacy of therapy on moderate or severe exacerbations, Transition Dyspnea Index (TDI), and forced expiratory volume in 1 second (FEV 1 ), with current smokers having a reduced response to ICS at any eosinophil count compared with former smokers. [7] However, the effect of smoking status on the efficacy of triple versus dual therapy on clinical outcomes at multiple time points and safety has yet to be fully reported. Here, we aim to ascertain if 1 Abbreviations: AEs, adverse events; AESIs, adverse events of special interest; BDI, baseline dyspnea index; COPD, chronic obstructive pulmonary disease; FEV 1 , forced expiratory volume in 1 second; FF, fluticasone furoate; GOLD, Global Initiative for Chronic Obstructive Lung Disease; HDAC2, histone deacetylase 2 activity; ICS, inhaled corticosteroid; ITT, intent-to-treat; LABA, long-acting β 2 -agonist; LAMA, long-acting muscarinic antagonist; SAE, serious adverse event; SD, standard deviation; SGRQ, St George's Respiratory Questionnaire; TDI, transition dyspnea index; UMEC, umeclidinium; VI, vilanterol smoking status affects treatment outcomes by evaluating the efficacy and safety of FF/UMEC/VI versus FF/VI and UMEC/VI in current and former smokers in a post hoc analysis of the IMPACT trial.

Study design
IMPACT (GSK Study CTT116855/NCT02164513; N=10,355) was a Phase III, 52-week randomized, double-blind, multicenter study, which compared the efficacy and safety of once-daily single-inhaler triple therapy with FF/UMEC/VI 100/62.5/25mcg with once-daily dual therapy with FF/VI 100/25μg or UMEC/VI 62.5/25μg, all administered via the Ellipta dry-powder inhaler. [13,14] In order to reflect routine clinical practice, patients continued their existing COPD maintenance therapy during a 2- week run-in period prior to being randomized (2:2:1). This post hoc analysis evaluated the efficacy and safety of triple therapy versus dual therapy by smoking status at screening. Patients were classified as former smokers if they had not smoked for >6 months prior to the screening visit in order to minimize recall bias, or as current smokers if they had stopped within the last 6 months or were active smokers at the time of the screening visit. Smoking status was consistent across the 52 weeks during the trial in all treatment arms.

Study population and endpoints
The inclusion and exclusion criteria have been described previously. [13,14] Briefly, eligible patients were ≥40 years of age with symptomatic COPD (COPD Assessment Test score ≥10) and with either a FEV 1 <50% of predicted normal values and ≥1 moderate or severe exacerbation in the previous year or FEV 1 50-80% of predicted normal values and ≥2 moderate or ≥1 severe exacerbations in the previous year. [13] All patients were required to have a ≥10 pack-year smoking history. Exclusion criteria included a current diagnosis of asthma. However, patients with a history of asthma were eligible to increase the generalizability of the trial. [13] All patients provided written informed consent. The study was conducted in accordance with Good Clinical Practice guidelines and the provisions of the Declaration of Helsinki and received approval by local ethics review boards of the participating sites.
The efficacy endpoints assessed in this analysis included rate of on-treatment moderate/severe exacerbations; risk (time-to-first) of on-treatment moderate/severe exacerbation; change from baseline in trough FEV 1 at Weeks 4, 16, 28, 40, and 52; change from baseline in St George's Respiratory Questionnaire (SGRQ) total score and percentage of SGRQ responders (patients with a decrease in SGRQ total score from baseline of ≥4 points) at Weeks 4, 28 and 52; and TDI focal score at Weeks 4, 28, and 52. The safety endpoints included incidences of adverse events (AEs), AEs of J o u r n a l P r e -p r o o f special interest (AESIs) and serious AEs (SAEs) from the start of randomization until safety follow-up at Week 52. Moderate exacerbations were defined as exacerbations that required treatment with oral/systemic corticosteroids and/or antibiotics (not involving hospitalization or resulting in death).
Severe exacerbations were defined as exacerbations that required hospitalization or resulted in death.

Statistical analyses
Study population characteristics, efficacy and safety endpoints were assessed in the intent-to-treat (ITT) population. The rate of on-treatment moderate/severe COPD exacerbations in current smokers and former smokers was analyzed using a generalized linear model assuming a negative binomial distribution, with covariates of treatment group, sex, exacerbation history (≤1, ≥2 moderate/severe), geographical region, post-bronchodilator percent predicted FEV 1 (screening), and treatment group by smoking status (screening) interaction. Interaction testing was performed to determine whether the treatment effect on the primary endpoint was modified by the following factors: gender, exacerbation history (≤1 or ≥2 moderate/severe exacerbations), smoking status at screening, geographical region or post-bronchodilator % predicted FEV 1 at screening. Interactions were deemed significant if P<0.1.
Time-to-first event (risk) analyses for on-treatment moderate/severe COPD exacerbations were conducted using a Cox proportional hazards model with covariates of treatment group, sex, exacerbation history (≤1, ≥2 moderate/severe), geographical region, post-bronchodilator percent predicted FEV 1 (screening), smoking status (screening), and treatment group by smoking status (screening) interaction. Change from baseline in trough FEV 1 , change from baseline in SGRQ total score and TDI focal score were analyzed using a repeated measures model with covariates of treatment group, smoking status (screening), geographical region, visit, treatment group by visit, treatment group by smoking status (screening), visit by smoking status (screening), and treatment group by visit by smoking status (screening) interactions. The analyses of trough FEV 1 and SGRQ total score also included baseline as a covariate and baseline by visit interaction term, and TDI focal score also included Baseline Dyspnea Index (BDI) focal score as a covariate and BDI focal score by visit interaction term. The proportion of responders according to SGRQ total score was assessed for current and former smokers; the analyses did not include an interaction term, instead each subgroup was analyzed separately. Safety endpoints were assessed in the ITT population using descriptive statistics. AESIs were defined as AEs within specified areas of interest for FF, UMEC, and/or VI, or the overall COPD population. All programming for statistical analyses was performed using SAS Version 9.4.

Study population
Baseline characteristics by smoking status at screening and treatment group are displayed in Table 1

GOLD grades b , n (%)
J o u r n a l P r e -p r o o f

Interaction testing
Of the prespecified factors tested in IMPACT, only exacerbation history (P=0.010) and smoking status at screening (P=0.023) demonstrated a significant interaction of treatment on the primary endpoint (  (Figure 1).
J o u r n a l P r e -p r o o f

Efficacy endpoints Exacerbations and lung function
In both current and former smokers, the annual rate of on-treatment moderate/severe COPD exacerbation was significantly lower with FF/UMEC/VI compared with FF/VI and UMEV/VI (Figure 1).
For both current and former smokers, the mean improvement from baseline in trough FEV 1 was significantly greater with FF/UMEC/VI versus FF/VI and UMEC/VI at all time points (Figure 3; Figure   S1). However, in former smokers, the difference in improvement across time points with FF/UMEC/VI versus UMEC/VI was between 58 and 61mL, whereas in current smokers it was 40-50mL (Figure 3; Figure S1).
J o u r n a l P r e -p r o o f    (Figure 5).
There was no difference in the mean TDI focal score with FF/UMEC/VI versus FF/VI and UMEC/VI at Week 52 in either current or former smokers ( Figure S3)     J o u r n a l P r e -p r o o f 20 Note: n=Number of subjects, #=Number of events. Note: Rate is event rate per 1000 subject-years, calculated as the number of events x 1000, divided by the total duration at risk AEs, adverse events; AESIs, adverse events of special interest; FF, fluticasone furoate; ITT, intent-to-treat; LRTI, lower respiratory tract infections; SAEs, serious adverse events; SMQ, standard Medical Dictionary for Regulatory Activities (MedDRA) queries; UMEC, umeclidinium; VI, vilanterol The lowest rate of pneumonia was observed in current smokers treated with UMEC/VI (58.0 per 1,000 patient-years). Differences in the occurrence of local corticosteroid effects were also observed between current smokers (6-10%) and former smokers (5-7%) across treatment arms. Current smokers had higher exposure-adjusted rates of local corticosteroid effects compared with former smokers across all treatment arms ( Table 2).
The most common on-treatment AESIs were cardiovascular effects, occurring with a similar incidence and rate across all treatment groups among both current smokers and former smokers (10-11% and 155.0-169.1 per 1,000 patient-years) ( Table 2).

Discussion
This post hoc analysis of data from the IMPACT trial demonstrates that single-inhaler triple therapy with FF/UMEC/VI significantly reduces the rate of on-treatment moderate/severe exacerbations and improves trough FEV 1 and SGRQ total score compared with FF/VI and UMEC/VI in both current and former smokers. A significant interaction of smoking status at screening with treatment was observed for the primary analysis; in line with this finding, subgroup analysis showed that the benefit of FF/UMEC/VI versus UMEC/VI on rate of on-treatment moderate/severe exacerbations was seen among both current and former smokers, however the benefit was greater in former versus current smokers, potentially due to a relative corticosteroid resistance in current smokers. [5] Similar trends were seen for both lung function and health status. These findings emphasize the importance of smoking cessation in the moderate-to-severe COPD population, to enable optimal response to treatment.
These differences in treatment effects by smoking status may be related to patients being more likely to stop smoking if they have more severe disease and worse health status at the time of COPD diagnosis. [15] Although in our study, current and former smokers had similar symptoms and exacerbation rates at baseline, a higher proportion of former smokers were receiving triple therapy prior to the study in line with previous studies which have shown former smokers are more symptomatic [15] and have a greater number of comorbid conditions. [16] In former smokers, triple therapy demonstrated greater effects compared with UMEC/VI for the rate and risk of exacerbations and health status (as measured by SGRQ); however, for lung function, the added benefit of triple therapy was more pronounced versus FF/VI. In current smokers, the added benefit of triple therapy versus FF/VI was greater than versus UMEC/VI for all endpoints. These findings illustrate the utility of the ICS in the treatment regimen for former smokers. In current smokers, however, concomitant use appears to blunt the clinical response to ICS. Impairment of the response to ICS in smokers has been reported for both asthma and COPD. [5, 7-10, 17, 18] Whilst this effect could be a result of J o u r n a l P r e -p r o o f increased mucus production impairing the absorption of ICS, [19] several possible mechanisms have been proposed for this finding, including a decrease in the enzyme HDAC2 or of the antioxidative transcription factor nuclear factor erythroid 2-related factor 2. Additionally, reduced glucocorticoid receptor nuclear translocation due to receptor phosphorylation at Ser226 by p38 mitogen-activated protein kinase can confer corticosteroid insensitivity. [20,21] Alternatively, a less likely explanation may be that acute anti-inflammatory effects associated with substances present in cigarette smoke such as carbon monoxide may result in ICS being superfluous. [22] Our findings complement previous analyses and published data. In an earlier analysis of the IMPACT The incidence of AESIs were similar between current and former smokers, except for pneumonia and local corticosteroid effects. Pneumonia incidence was higher in former smokers compared with current smokers, particularly in the ICS treatment arms (6.3% and 5.8% among current smokers vs 8.4% and 7.7% among former smokers in the FF/UMEC/VI and FF/VI treatment arms). Former smokers with COPD may have ceased smoking due to necessity from ill health, symptoms or frequent COPD exacerbations. [15] Local corticosteroid effects were higher in current smokers; however, local corticosteroid effects are also common side effects of smoking, notably increasing dysphonia, and oral candidiasis in those with immune dysfunction such as HIV.
Some limitations should be considered in the interpretation of the findings of this study. Analyses were conducted post hoc and the IMPACT trial, whilst extensive with a large sample size, was not powered to analyze endpoints by smoking status. Smoking status was measured at screening via self-reporting, which may have led to an underestimation of current smoking. Without cotinine levels being checked, it is possible some patients self-reported as former smokers when they were still actively smoking. As stated earlier, the baseline characteristics between the current and former smoker populations were not balanced, with former smokers more likely to be older, male and receiving triple therapy at baseline. As such, formal statistical comparisons between current and former smokers were not conducted as these differences may confound the results of any formal comparison between the groups. Strengths of the study include that IMPACT was a large, prospective COPD clinical trial and was designed to be generalizable to clinical practice. [13] Conclusion FF/UMEC/VI improved the exacerbation rate, lung function, and health status compared with FF/VI and UMEC/VI in patients with symptomatic COPD and a history of exacerbations, regardless of smoking status. The optimal benefit for triple versus LAMA/LABA dual therapy was observed in former smokers, potentially due to relative corticosteroid resistance in current smokers. This emphasizes the importance of smoking cessation in this symptomatic COPD population at risk of exacerbations, as amongst its many advantages, it leads to greater benefits from corticosteroid-J o u r n a l P r e -p r o o f GSK, MSD, Pfizer, Sanofi, Novartis and Roche. NB was an investigator on the IMPACT study and reports receiving consulting fees and serving as a participant in advisory boards for Kamada, AstraZeneca, Boehringer Ingelheim, GSK, Sanofi Genzyme and Novartis, as well as serving as a participant in advisory boards for Teva. He has also received lecture fees for Kamada, AstraZeneca, Boehringer Ingelheim, GSK and Sanofi Genzyme.

Funding
This work was funded by GSK (study number CTT116855; NCT02164513). The funders of the study had a role in the study design, data analysis, data interpretation and writing of the report.