Step-Down Therapy for Asthma Well Controlled on Inhaled Corticosteroid and Long-Acting Beta-Agonist: A Randomized Clinical Trial

Publication date: Available online 30 September 2017
Source:The Journal of Allergy and Clinical Immunology: In Practice
Author(s): Linda Rogers, Elizabeth A. Sugar, Kathryn Blake, Mario Castro, Emily Dimango, Nicola A. Hanania, Kyle I. Happel, Stephen P. Peters, Joan Reibman, Joy Saams, W. Gerald Teague, Robert A. Wise, Janet T. Holbrook
BackgroundStepping down therapy when asthma is well controlled on combination inhaled corticosteroids (ICSs) and long-acting beta-agonists (LABAs) is recommended, but it is not known whether lowering the ICS dose or stopping LABA is superior.ObjectiveTo evaluate whether step-down therapy with LABA is superior to one without; and, secondarily, to evaluate whether reducing the ICS dose while maintaining LABA is noninferior to remaining on stable-ICS/LABA.MethodsThe study was a randomized, double-masked 3-arm parallel group trial in participants aged 12 years or older. Following an 8-week run-in, 459 participants were randomly assigned to continue medium-dose ICS/LABA, reduced-dose ICS/LABA, or ICS alone (LABA-step-off) and followed for 48 weeks. The primary outcome was time to treatment failure, a composite of health care utilization, systemic corticosteroid use, increase in rescue therapy, decline in lung function, or participant or physician decision.ResultsTime to treatment failure did not differ significantly between reduced- ICS/LABA and LABA-step-off (hazard ratio, 1.07; 95.3% CI, 0.69-1.65, P = .76). Nor was there a difference between stable-ICS/LABA and reduced-ICS/LABA (hazard ratio, 1.11; 95% CI, 0.70-1.76; P = .67), but the 10% noninferiority margin was exceeded. Lung function declines and hospitalization rates were significantly greater in the LABA-step-off group.ConclusionsThe 2 step-down regimens did not differ in terms of treatment failure, although stopping LABA was associated with a decline in lung function and more hospitalizations. There was no evidence to support the noninferiority of reduced-ICS/LABA as compared with stable-ICS/LABA.