Background: Global initiative for chronic Obstructive Lung Disease (GOLD) strategy recommends treating patients with chronic obstructive pulmonary disease (COPD) who have recurrent exacerbations with inhaled corticosteroids (ICS). The Withdrawal of Inhaled Steroids During Optimized bronchodilator Management (WISDOM) study was designed to evaluate the effects of stepwise withdrawal of ICS in patients with severe COPD (GOLD 3–4) and a history of exacerbation who were treated with a combination of a long-acting muscarinic antagonist plus a long-acting β2-agonist (LAMA+LABA).

Methods: All patients in this 12-month, double-blind, parallel-group, active-controlled study (NCT00975195) received triple therapy (tiotropium18 µg QD, salmeterol 50 µg BID, and fluticasone 500 µg BID) for a 6-week run-in period. Following this, baseline assessments were made and patients were randomized 1:1 to continue triple therapy or undergo stepwise withdrawal of ICS over 12 weeks, with dose reduction every 6 weeks. We report time to first moderate to severe on-treatment exacerbation (primary end point; non-inferiority margin pre-defined as a hazard ratio [HR] <1.2), along with results for the following subgroups: geographic region, age, sex, smoking status, baseline body mass index, and previous COPD therapy.

Results: 2485 patients were randomized and treated. Most patients were male (2049), with a mean age of 63.8 years and mean forced expiratory volume in 1 second (FEV1) of 0.98 L (34.2%) at baseline (Week 0). The risk of experiencing any on-treatment COPD exacerbation with ICS withdrawal was non-inferior to continued ICS therapy (HR 1.058; 95% confidence interval [CI] 0.941, 1.189). Similar results were seen for moderate or severe exacerbations (HR 1.05; 95% CI 0.93, 1.18). Consistent findings were also seen in a sensitivity analysis that included post-treatment exacerbations (HR 1.061; 95% CI 0.945, 1.192) and in all pre-defined sub-groups analyzed. At Week 18, the adjusted mean decrease from baseline in trough FEV1 was 38 mL greater on complete ICS withdrawal than on ICS (p<0.0001), with similar results at Week 52. No significant safety signals were identified.

Conclusions: The risk of moderate to severe exacerbation after stepwise withdrawal of ICS was comparable to continued ICS treatment in patients with GOLD 3–4 COPD receiving LAMA+LABA. The clinical significance of the small but significant decrease in FEV1 after ICS withdrawal is not known. These data suggest that ICS may be successfully withdrawn in COPD patients receiving LAMA+LABA without significantly increasing the risk of exacerbation.