Serious Asthma Events with Fluticasone plus Salmeterol versus Fluticasone Alone

BACKGROUND The safe and appropriate use of long-acting beta-agonists (LABAs) for the treatment of asthma has been widely debated. In two large clinical trials, investigators found a potential risk of serious asthma-related events associated with LABAs. This study was designed to evaluate the risk of administering the LABA salmeterol in combination with an inhaled glucocorticoid, fluticasone propionate. METHODS In this multicenter, randomized, double-blind trial, adolescent and adult patients (age, ≥12 years) with persistent asthma were assigned to receive either fluticasone with salmeterol or fluticasone alone for 26 weeks. All the patients had a history of a severe asthma exacerbation in the year before randomization but not during the previous month. Patients were excluded from the trial if they had a history of lifethreatening or unstable asthma. The primary safety end point was the first serious asthma-related event (death, endotracheal intubation, or hospitalization). Noninferiority of fluticasone–salmeterol to fluticasone alone was defined as an upper boundary of the 95% confidence interval for the risk of the primary safety end point of less than 2.0. The efficacy end point was the first severe asthma exacerbation. RESULTS Of 11,679 patients who were enrolled, 67 had 74 serious asthma-related events, with 36 events in 34 patients in the fluticasone–salmeterol group and 38 events in 33 patients in the fluticasone-only group. The hazard ratio for a serious asthmarelated event in the fluticasone–salmeterol group was 1.03 (95% confidence interval [CI], 0.64 to 1.66), and noninferiority was achieved (P = 0.003). There were no asthma-related deaths; 2 patients in the fluticasone-only group underwent asthmarelated intubation. The risk of a severe asthma exacerbation was 21% lower in the fluticasone–salmeterol group than in the fluticasone-only group (hazard ratio, 0.79; 95% CI, 0.70 to 0.89), with at least one severe asthma exacerbation occurring in 480 of 5834 patients (8%) in the fluticasone–salmeterol group, as compared with 597 of 5845 patients (10%) in the fluticasone-only group (P