OBJECTIVES

: To assess the comparative efficacy of short-acting muscarinic antagonist (SAMA), long-acting muscarinic antagonists (LAMA), LAMA in combination with long-acting beta-agonists (LAMA/LABA) and inhaled corticosteroids in combination with LABA (ICS/LABA) for the maintenance treatment of chronic obstructive pulmonary disease (COPD).

METHODS

: Randomised controlled trials comparing any of the interventions (SAMA (ipratropium), LAMA (glycopyrronium, tiotropium, umeclidinium), LAMA/LABA (glycopyrronium/indacaterol, tiotropium/olodaterol, umeclidinium/vilanterol) and ICS/LABA (salmeterol/fluticasone, formoterol/budesonide) with each other or with placebo were identified in Pubmed and Embase by two independent reviewers. A network meta-analysis (NMA) combining direct and indirect evidence examined the change from baseline in trough forced expiratory volume in 1 second (FEV) at weeks 12 and 24 using a random effects model.

RESULTS

: Forty seven studies comprising 55,515 patients were included in the NMA, with 38 studies reporting results at week 12 and 29 studies at week 24. Inconsistency models were not statistically significant (week 12: p=0.0584, week 24: p=0.0728). All LAMAs, LAMA/LABAs and ICS/LABAs led to significantly greater improvement in trough FEV compared to SAMA at weeks 12 and 24. All LAMA/LABAs led to more statistically significant improvements than LAMAs and ICS/LABAs at weeks 12 and 24 for all comparisons except tiotropium/olodaterol versus umeclidinium. Among the LAMAs, umeclidinium showed statistically significant improvement in trough FEV at week 12 compared to tiotropium and glycopyrronium, but the results were not considered clinically significant. There were no significant differences in trough FEV for all LAMA-LAMA comparisons at week 24. Among the LAMA/LABAs, there were no significant differences at weeks 12 and 24.

CONCLUSIONS

: LAMA/LABA showed greatest improvement in lung function at weeks 12 and 24 compared with the other inhaled drug classes while SAMA showed least improvement.