PURPOSE: Exacerbations of COPD result in worsening lung function and mortality. In the roflumilast pivotal trials, moderate or severe exacerbations were assessed by medical intervention (moderate=oral/parenteral corticosteroid use; severe=hospitalization/death). However, assessments of exacerbations across studies can be affected by the different definitions used. In this post hoc analysis, we examined primary outcome measure exacerbation events from the two roflumilast pivotal trials using two alternative exacerbation definitions based either on symptoms or rescue medication use.
METHODS: Data were pooled from two 52-week pivotal trials of roflumilast (500µg once-daily) in subjects with severe-to-very-severe COPD, chronic bronchitis, and a history of exacerbations. Each exacerbation event based on the primary outcome definition was analyzed using two alternative definitions. Exacerbations associated with symptoms were defined by either worsening of cough/sputum for two consecutive days within +/-5 days of moderate/severe exacerbation onset or worsening of total symptom score (≥0.5) within +/-3 days from average of 6 days prior to +/- 3 days window. Exacerbations associated with rescue medication were defined by an increase in rescue medication (≥3 puffs/day for ≥2 consecutive days) within +/-5 days of moderate/severe exacerbation onset.
RESULTS: A total of 1537 subjects received roflumilast and 1554 placebo. Based on primary outcome measure definitions there were 2978 moderate or severe exacerbations (roflumilast=1299; placebo=1679); of these the post hoc analysis found that 1833 were associated with symptoms (796 for roflumilast; 1037 for placebo) and 287 were associated with rescue medication (113 for roflumilast; 174 for placebo). The rate ratio of exacerbations associated with symptoms for roflumilast compared with placebo was 0.824 (95% CI 0.730, 0.931, p=0.0018). Similarly for exacerbations associated with rescue medication use, the rate ratio was 0.697 (95% CI 0.504, 0.964, p=0.0289). When exacerbations with two definitions were combined, the rate ratio was 0.802 (95% CI 0.712, 0.904, p=0.0003).
CONCLUSIONS: Medical intervention-based exacerbation definitions in the pivotal roflumilast trials were substantiated by this analysis of symptomatic and rescue medication exacerbation definitions.
CLINICAL IMPLICATIONS: This post hoc analysis proves that roflumilast’s efficacy for reducing the risk of exacerbations is robust.
DISCLOSURE: Nicola Hanania: Consultant fee, speaker bureau, advisory committee, etc.: Forest Laboratories, Inc., Grant monies (from industry related sources): Forest Laboratories, Inc.
Jill Karpel: Grant monies (from industry related sources): Forest, Boeringher-Ingelheim, Novartis, Genentech, Consultant fee, speaker bureau, advisory committee, etc.: Merck, Genentech, Astra-Zeneca, Boeringher-Ingelheim
Udo-Michael Goehring: Employee: Nycomed GmbH
Manja Brose: Employee: Nycomed GmbH
Dirk Bredenbroeker: Employee: Nycomed GmbH
Hassan Lakkis: Employee: Forest Research Institute
Paul Rowe: Employee: Forest Research Institute
No Product/Research Disclosure Information