This was a multinational, randomized, double-blind, double-dummy, placebo-controlled, multi-dose, Phase III, crossover study in patients with moderate-to-severe COPD. The study was conducted in accordance with the Declaration of Helsinki (1989) and local applicable laws and regulations. Institutional Review Board or Independent Ethics Committee approval was obtained for each participating study center. All participants provided informed written consent prior to taking part in the study.
Male and female patients aged ≥40 years with moderate-to-severe COPD (GOLD 2006), a smoking history of at least 10 pack-years (current or previous smokers), post-bronchodilator FEV1 ≥30% but <80% of the predicted normal value, and post-bronchodilator FEV1/forced vital capacity (FVC) <70% were included in the study.
Patients were excluded from the study if they had been hospitalized for a COPD exacerbation in the 6 weeks prior to screening or during the run-in period; had a history of asthma; or had concomitant pulmonary disease or a significant unstable cardiovascular or metabolic comorbidity.
Study design and treatments
The study comprised a pre-screening visit, a 14-day screening period, followed by three 14-day treatment periods, each of which was separated by a 14-day washout. At the pre-screening visit, patients' ongoing COPD medications were reviewed and, if necessary, adjusted to exclude prohibited COPD therapies. On completion of the screening period, eligible patients were randomized using a validated automated system to receive three of the four treatments (with a different treatment in each treatment period), each once-daily. An incomplete-block design was selected (with three treatment periods rather than four) to reduce the overall burden on patients, and to increase the likelihood of patients completing the study (given the complexity in delivering the third-party blinded study medication - see below). Due to the use of different inhalers to deliver the study drugs, patients were randomized not only to a treatment sequence, but also to a sequence of inhalers. Indacaterol 150 or 300 μg was delivered via single-dose dry powder inhaler (SDDPI); tiotropium 18 μg was delivered via the manufacturer's proprietary single-dose dry powder inhalation device (HandiHaler®). Study drugs were inhaled between 06:00 h and 10:00 h on each day throughout the treatment periods.
Indacaterol and its matching placebo were made identical in appearance and were dispensed in such a manner so as to make them indistinguishable to patients and all blinded study personnel. As an exact physical match to tiotropium was not available, full blinding was achieved by third-party blinding procedures. These procedures were as follows: study drug was prepared and provided to the patient each morning, either at home or in the clinic, by persons who were independent of the other clinical trial processes (referred to as 'independent study blinding coordinators' or 'ISBCs') to preserve the integrity of the blind. Two ISBCs were required for each daily study drug administration to each individual patient. The first (unblinded) ISBC (who had no contact whatsoever with the patient) prepared the study drugs and devices, maintained patient, investigator and the second ISBC blinding and ensured that the patients strictly adhered to their allocated drug sequence. The second (blinded) ISBC provided the patient with the prepared study drug and devices and monitored administration of the drug by patients and also ensured that the blinding was maintained throughout. Both ISBCs completed the Third Party Blinding Log for every drug administration, in order to evidence that the blinding procedure was strictly followed.
Allowable concurrent COPD therapies included the use of inhaled corticosteroids (ICS), provided the regimen had been stabilized for at least 1 month prior to the screening visit. Patients taking fixed-dose combinations of ICS and LABAs were switched to equivalent ICS monotherapy at a dose and dosage regimen maintained for the duration of the study. The following medications could not be used after the screening visit (except as study medication): long- or short-acting anticholinergic agents, long- or short-acting β2-agonists, xanthine derivatives, and parenteral or oral corticosteroids. Salbutamol was the only rescue medication permitted throughout the study, although not within 6 h prior to the start of each visit.
Assessments and outcomes
All clinic visits started in the morning. In addition to the assessments during the screening visits, serial spirometry was performed on Day 1 and Day 14 of each treatment period, at 50 and 15 min pre-dose and at 5, 15 and 30 min and 1, 2, 4, 8, 10, 12 and 14 h post-dose. Spirometry was also assessed on Day 2 and Day 15 of each treatment period at 23 h 10 min and 23 h 45 min post-dose (based on the time of study drug administration on the previous day) to enable trough values of FEV1 to be determined. All spirometry evaluations were performed according to American Thoracic Society/European Respiratory Society standards .
Adverse events (AEs) and serious AEs were recorded, along with their severity, duration and relationship to study drug. Other safety assessments included: urinalysis; regular monitoring of hematology, blood chemistry (including serum potassium and blood glucose) and vital signs; and assessment of corrected QT interval (QTc).
The primary objective of the study was to determine whether indacaterol was superior to placebo as assessed by trough FEV1 after 14 days of treatment, with trough FEV1 defined as the mean of FEV1 measurements at 23 h 10 min and 23 h 45 min post-dose. The key secondary objective was a non-inferiority comparison between indacaterol and tiotropium for this endpoint (and if achieved, to then test for superiority). Other efficacy variables included trough FEV1 after the first dose, and FEV1 measurements at individual timepoints after the first dose and on Day 14 in each treatment period.
Sample size calculation and statistical analysis
The study was powered for the key secondary objective, the non-inferiority comparison of indacaterol versus tiotropium for trough FEV1 after 14 days, where a non-inferiority margin of 55 mL based on a Cochrane review  was adopted. An advantage of 30 mL for indacaterol over tiotropium was assumed and a standard deviation of 220 mL for the difference between repeated measures on the same patient (based on information from previous studies). Taking account of the incomplete block nature of the design, 126 evaluable patients would provide a power of 90% for a one-sided test at the 1.25% significance level (half the usual alpha level to adjust for multiplicity). Allowing for a dropout rate of 15%, a total of 148 patients were planned to be randomized into this study. This number of patients would give 99% power for the primary endpoint, assuming a minimum clinically important difference (MCID) of 120 mL.
All efficacy variables, including the primary efficacy variable, were analyzed for the modified intent-to-treat (mITT) population comprising all randomized patients who received at least one dose of study drug. The non-inferiority comparison between indacaterol and tiotropium for trough FEV1 after 14 days was analyzed for the per-protocol population, which included all patients in the mITT population who had no major protocol deviations. The safety population included all patients who received at least one dose of study drug. Patients were analyzed according to treatment received.
An analysis of covariance model was used to analyze the primary endpoint and included terms for treatment, period, patient and period baseline value (pre-dose FEV1 on Day 1 of each treatment period), with results presented as least squares means, i.e., means adjusted for the covariates in the model. To allow for multiplicity, a Bonferroni adjustment was applied to maintain the overall Type I error rate at 5%. A similar model was used to analyze the secondary endpoints (with the non-inferiority and superiority comparisons between indacaterol and tiotropium also controlled for multiplicity).