Adverse events are defined as any unfavorable and unintended diagnosis, symptom, sign (including an abnormal lab finding), syndrome or disease which either occurs during study, having been absent at baseline, or, if present at baseline, appears to worsen. Serious adverse events are any untoward medical occurrences that result in death, are life threatening, require (or prolong) hospitalization, cause persistent or significant disability/incapacity, result in congenital anomalies or birth defects, or are other conditions which in judgments of the investigators represent significant hazards.

Safety population - all patients who received at least one dose of study drug whether or not they were randomized. Only patients with safety assessments were included in this analysis

Discontinuation rates are calculated using the Kaplan Meier method. The protocol allowed patients to discontinue outside the treatment window, hence we have a patient who discontinued at Day 388. Reasons for discontinuing treatment are Subject/guardian decision, Adverse event, Protocol deviation Lack of efficacy, Physician decision, Dosing error, Disease improvement under study, Pregnancy, Technical problems

The Safety set consisted of all patients that received at least one dose of study medication and had at least one post-baseline safety assessment. Patients were analyzed according to treatment received.

Change from baseline in pre-dose trough FEV1 was analyzed using a repeated measures analysis of covariance model which contained treatment, baseline FEV1, visit, baseline smoking status, baseline ICS use, COPD severity and treatment by visit, visit by baseline FEV1 interactions. An unstructured variance-covariance error matrix was used .Pulmonary function assessments were performed using centralized spirometry according to international standards. Pre-dose trough FEV1 was defined as the mean of FEV1 at -45 min and -15 min before the morning dose at Week 52. Baseline FEV1 was defined as the mean of the pre-dose FEV1 at -45 min and -15 min on Day 1.

The Full Analysis set (FAS) included patients who received at least one dose of study medication. Patients were analyzed according to the treatment to which they were randomized. Only participants from the full analysis set, who had outcome measure data with applicable fixed effects/covariates according to the analysis model, were analyzed.

Pulmonary function assessments were performed using centralized spirometry according to international standards. Baseline FEV1 was defined as the average of the pre-dose FEV1 measured at -45 minutes (min) and -15 min at day 1.

The Full Analysis set (FAS). At each day/time point, only subjects with a value at both baseline and the respective day/time point are included. Only participants with baseline and specific post baseline time points were included in the analysis for that time point.

Pulmonary function assessments were performed using centralized spirometry according to international standards. Baseline FVC was defined as the average of the pre-dose FVC measured at -45 minutes (min) and -15 min at day 1.

Full Analysis set (FAS) included all randomized patients who received at least one dose of study medication. Only participants with baseline and specific post baseline time points were included in the analysis for that time point.

The total symptom score was defined as the sum of individual cores for respiratory symptoms, cough, wheeze, amount of sputum, color of sputum, and reathlessness. Where a patient had a morning score and an evening score for an individual symptom on one particular day then the worst score was to be taken as the daily score for that symptom. Each symptom scale ranged from 0-3 where 0 was no symptoms and 3 was the worst. The total daily/daytime/nighttime symptom score consists of looking at the score for 6 symptoms and can therefore have a minimum score of 0 or a maximum of 18.

Full Analysis set (FAS) included all randomized patients who received at least one dose of study medication. Only participants from the full analysis set, who had outcome measure data with applicable fixed effects/covariates according to the analysis model, were analyzed."

Percentage of days with 'no daytime symptoms' A day with 'no daytime symptoms' was defined from the diary data as any day where the patient had recorded in the evening no cough, no wheeze, no production of sputum and no feeling of breathlessness (other than when running) and no puffs of rescue medication during the past 12 hours (approximately 8 am to 8pm). However, a patient was not considered symptom free if they had used rescue medication that day even if his/her total daytime symptoms score was zero. Percentage of nights with 'no nighttime awakenings' A night with 'no nighttime awakenings' was defined from diary data as any night where the patient did not wake up due to symptoms. The total number of nights with 'no nighttime awakenings' over the treatment period was divided by the total number of nights where diary recordings had been made in order to derive the percentage nights with 'no nighttime awakenings'.

Full Analysis set (FAS) included all randomized patients who received at least one dose of study medication. Only participants from the full analysis set, who had outcome measure data with applicable fixed effects/covariates according to the analysis model, were analyzed.

The number of puffs of rescue medication taken in the previous 12 hours was recorded by the patients in the eDiary in the morning and evening. The total number of puffs of rescue medication per day over the 52 week treatment period was calculated and divided by the total number of days with non-missing rescue data to derive the mean daily number of puffs of rescue medication taken for the patient. If the number of puffs was missing for part of the day (either morning or evening), then a half day was used in the denominator. Change from baseline in number of puffs were analyzed using a linear mixed model which contained treatment, baseline number of puffs, baseline smoking status, baseline ICS use and COPD disease severity as fixed effects with center as a random effect

Only participants from the full analysis set, who had outcome measure data with applicable fixed effects/covariates according to the analysis model, were analyzed.

COPD exacerbations are considered to be moderate if treatment with systemic corticosteroids and/or antibiotics was required. COPD exacerbations are considered to be severe if hospitalizations were required. Rates are calculated using the Kaplan Meier method.

The Full Analysis set (FAS) included all randomized patients who received at least one dose of study medication. Patients were analyzed according to the treatment to which they were randomized