Aggressive treatment of asymptomatic GERD does not improve asthma

Daily POEMs

June 2009

Clinical question
Does treatment of asymptomatic gastroesophageal reflux improve outcomes in patients with poorly controlled asthma?

Bottom line
Although it was a compelling idea supported by some observational data, giving high doses of proton pump inhibitors (PPIs) to patients who have poor asthma control but no symptoms of gastroesophageal reflux (GERD) does not improve outcomes. PPIs may still have a role in symptomatic patients, but primarily for the control of reflux symptoms and not because they are likely to improve asthma control. Also, remember that most patients do quite well with 20 mg omeprazole and do not require high doses of esomeprazole. (LOE = 1b)

Reference
American Lung Association Asthma Clinical Research Centers, Mastronarde JG, Anthonisen NR, Castro M, et al. Efficacy of esomeprazole for treatment of poorly controlled asthma. N Engl J Med 2009;360(15):1487-1499.

Study design: Randomized controlled trial (double-blinded)

Allocation: Concealed

Setting: Outpatient (specialty)

Synopsis
GERD is common in patients with asthma, but often does not cause typical symptoms. This association has led to recommendations to test and treat patients for GERD who have poorly controlled asthma. In this study, 412 patients with poorly controlled asthma, despite moderate or high doses of inhaled corticosteroids, were randomized to receive esomeprazole 40 mg twice daily or matching placebo. Poor control was defined as a score of 1.5 or higher on a 6-point scale of control or at least one acute, unscheduled visit for asthma in the previous year. Patients were excluded if they had reflux symptoms, pregnancy, or other serious illness. The primary outcome was the number of episodes of poor control, defined as 30% worsening in peak flow rate on 2 consecutive days (a disease-oriented outcome) or either an unscheduled visit for asthma or the need for oral prednisone (patient-oriented outcomes). Adherence to medications was good, groups were balanced, and analysis was by intention to treat. Some patients were not included in the final analysis because they did not turn in any follow-up symptom diary cards (n = 9) or because their data were lost because of Hurricane Katrina (n = 10). Patients were evaluated monthly for 24 weeks. There was no difference between groups in the likelihood of an episode of poor control as defined above, the risk of any of its components, or the risk of any secondary outcomes (use of rescue medications, night awakening, symptom scores, or quality-of-life scores). There was no difference in serious adverse events. The authors estimate that they had a 77% to 97% chance of detecting a 33% decrease in exacerbations.

Copyright© 2009 John Wiley & Sons, Inc.