Subgroup Analyses of Lung-Function Change From the WISDOM Study

Abstract

PURPOSE: WISDOM showed that in patients with severe to very severe chronic obstructive pulmonary disease (COPD) and exacerbation history receiving long-acting muscarinic antagonist with long-acting beta2-agonist, stepwise withdrawal of inhaled corticosteroids (ICS) was not inferior to continued ICS for risk of first moderate or severe COPD exacerbation, but there was a greater decrease in lung function following complete ICS withdrawal. To evaluate factors associated with this, further subgroup analyses were performed. METHOD(S): Patients were randomized to triple therapy or ICS withdrawal over 12 weeks (wks). A post hoc subgroup analysis of change from baseline in in-clinic trough forced expiratory volume in 1 s (FEV ) at wks 6, 12, 18, and 52 was performed using restricted maximum likelihood repeated measures models. Also, multiple linear regression using backward elimination was performed for change from baseline in in-clinic trough FEV1 at wk 52 to identify potential subgroups associated with a reduction in lung function over 1 year. RESULT(S): Consistent with the overall population, post hoc subgroup analysis of ontreatment trough FEV did not indicate notable differences between treatments (continued ICS vs ICS withdrawal) during ICS withdrawal at wks 6 and 12 for any subgroups. At wk 18 (6 wks after complete withdrawal) and wk 52, the overall adjusted mean difference between treatment groups was significant, favoring ICS for in-clinic trough FEV1. This was also generally reflected in the subgroups. Multivariate analyses at wk 52 identified 3 subgroups as significantly associated with trough FEV1 changes independent of treatment group. These were geographic region (greater decreases in trough FEV1 for Asian region vs non-Asian [p=0.003]), Global initiative for chronic Obstructive Lung Disease (GOLD) at screening (greater decreases for GOLD 4 [p=0.003] vs GOLD 3), and sex (p=0.004). Only sex was associated with treatment effect, with men showing a greater decrease in trough FEV1 after ICS withdrawal vs continued ICS (p=0.001). No differences were observed for women between treatment groups (p=0.876). CONCLUSION(S): Men demonstrated a greater decrease in lung function following ICS withdrawal. Female gender was associated with lung-function changes irrespective of treatment. CLINICAL IMPLICATIONS: Adequately powered prospective clinical studies will need to confirm these findings. Funding(s): Boehringer Ingelheim. Editorial assistance: Complete HealthVizion.