Centralized spirometry according to internationally accepted standards was used. The model contained treatment, baseline smoking status and baseline inhaled corticosteroid (ICS) use as fixed effects with the baseline measurement of FEV1, FEV1 prior to inhalation of short acting bronchodilators and FEV1 post inhalation of short acting bronchodilator as covariates and center nested in region as a random effect. If trough FEV1 was missing at week 12, the latest non-missing pre-dose trough FEV1 (the mean of 45 and 15 min pre-dose measurements) from day 29, 57 or 84) was carried forward. These measurements had to have been taken before the next dose of study medication. FEV1 measurements within 6 hours of rescue medication use or within 7 days of systemic corticosteroid use were not included in the analysis.

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

Centralized spirometry was used according to internationally accepted standards was used. The trapezoidal rule was applied to calculate FEV1 Area Under the Curve (AUC) and then normalized to the length of time. Whether the participants had complete or incomplete FEV1 assessments in respective time ranges, their AUCs were calculated based on the existing FEV1 measurements (i.e., the missing FEV1 measurements were not interpolated). Specifically, for those participants who had a FEV1 assessment at only one time-point, their AUC was approximated by the observed FEV1. FEV1 measurements within 6 hours of rescue medication use or within 7 days of systemic corticosteroid use were not included in the analysis.

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

Centralized spirometry was used according to internationally accepted standards was used. Peak FEV1 was defined as the maximum FEV1 during the first 4 hours post morning dosing. The model contained treatment, baseline smoking status and baseline ICS use as fixed effects with the baseline measurement of FEV1, FEV1 prior to inhalation of short acting bronchodilators and FEV1 post inhalation of short acting bronchodilator as covariates and center nested in a region as a random effect. If all FEV1 measurements were missing from 30 minutes onward, the peak FEV1 was not calculated. FEV1 measurements within 6 hours of rescue medication use or within 7 days of systemic corticosteroid use were not included in the analysis.

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

Centralized spirometry according to internationally accepted standards was used. FEV1 was measured at all post-dose time points up to 4 hours, and at 23 hours 15 minutes and 23 hours 45 minutes, by visit. The model contained treatment, baseline smoking status and baseline ICS use as fixed effects with the baseline measurement of FEV1, FEV1 prior to inhaltion of short acting bronchodilators and FEV1 post inhaltion of short acting bronchodilators as covariates and center nested in region as a random effect. FEV1 measurements within 6 hours of rescue medication use or within 7 days of systemic corticosteroid use were not included in the analysis.

Full Analysis Set

FVC was calculated at each time point up to 4 hours post-dose and at 23 hours 15 minutes and 23 hours 45 minutes, by visit. The model contained treatment, baseline smoking status and baseline ICS use as fixed effects with the baseline measurement of FVC, FEV1 prior to inhaltion of short acting bronchodilators and FEV1 post inhaltion of short acting bronchodilators as covariates and center nested in region as a random effect. FVC measurements within 6 hours of rescue medication use or within 7 days of systemic corticosteroid use were set to missing.

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

Inspiratory Capacity (IC) was measured at 20 min pre-dose and at post-dose at 25 minutes, 1 hour 55 minutes, 3 hours 55 minutes and 23 hours 40 minutes, by visit. The model contained treatment, baseline smoking status and baseline ICS use as fixed effects with the baseline measurement of IC, FEV1 prior to inhalation of short acting bronchodilators and FEV1 post inhalation of short acting bronchodilators as covariates and center nested in region as a random effect. IC measurements within 6 hours of rescue medication use or within 7 days of systemic corticosteroid use were set to missing.

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 in the patient diary in the morning and evening. The total number of puffs of rescue medication per day over the whole active 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.

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

Dyspnea was measured at baseline using the Baseline Dyspnea Index (BDI) and during treatment using the Transitional Dyspnea Index (TDI). Analysis was done via mixed model. The BDI and TDI each have three domains: functional impairment, magnitude of task and magnitude of effort. BDI domains were rated from 0 (severe) to 4 (unimpaired) and rates summed for baseline focal score ranged from 0 to 12; lower scores mean worse severity. TDI domains were rated from -3 (major deterioration) to 3 (major improvement) and rates summed for transition focal score ranged from -9 to 9; negative scores indicate deterioration. A TDI focal score of 1 is considered a minimal clinically important difference.

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 symptoms (respiratory, cough, wheeze, sputum color, sputum production, breathlessness, sore throat, nasal discharge or congestion, and fever) for the whole active treatment period was analyzed using a mixed model, which contained treatment, baseline smoking status and baseline ICS use as fixed effects with the baseline symptom score, FEV1 prior to inhalation of short acting bronchodilators and FEV1 post inhalation of short acting bronchodilators as covariates and center nested in region as a random effect. Each symptom was scored as 0, 1, 2 or 3 where the description for each score varied. For each of the symptoms, the range of scores from 0 to 3 represented an increase in symptoms where 0 represented little to no symptom and 3 represented severe or worst symptom. The total symptom score, which is the sum of the individual scores, ranged from 0 (best possible outcome) to 27 (worst possible outcome).

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

All study emergent adverse events including Chronic Obstructive Pulmonary Disease exacerbations were monitored from screening through the end of study.

Safety Set: The safety set included all patients who received at least one dose of study drug whether or not they were randomized.