Antiinflammatory effects of long-acting beta2-agonists in patients with asthma: a systematic review and metaanalysis.

Background: Long-acting beta(2)-agonists (LABAs) are recommended as add-on therapy to antiinflammatory treatment in patients with chronic persistent asthma. Results from individual studies evaluating the in vivo antiinflammatory effect of LABAs are conflicting. The purpose of this metaanalysis was to determine whether LABAs have an in vivo antiinflammatory effect compared to placebo and whether the addition of a LABA to therapy with inhaled corticosteroids (ICSs) has a synergistic or additive antiinflammatory effect.

Methods: A systematic search was performed of online databases for randomized controlled trials evaluating the antiinflammatory effects of the following: (1) LABAs compared to placebo; and (2) a LABA plus ICS vs ICS alone in adults and children with asthma. Inflammatory outcome measures included cell counts and markers of cell activation in sputum, BAL fluid, bronchial biopsy specimens, serum, and exhaled nitric oxide (ENO). Data were independently extracted by two study investigators and analyzed to generate standardized mean differences using either a fixed or random-effects metaanalysis depending on the degree of heterogeneity.

Results: Thirty-two studies (n = 1,105 participants) met the inclusion criteria. The mean number of participants and mean number of studies for each inflammatory outcome in the metaanalysis was 113.1 (SD, 46.2) and 3.5 (SD, 1.3), respectively. There was no effect of LABA therapy on sputum, BAL fluid, or mucosal inflammatory cells in adults or in children. LABA therapy decreased ENO levels and BAL fluid albumin levels in adults. In children, LABA therapy was associated with a small decrease in serum eosinophils and interleukin-4.

Conclusion: LABA therapy does not appear to have any clinically important antiinflammatory or proinflammatory effect. LABA therapy decreases BAL fluid albumin levels, suggesting a possible modulating effect on microvascular leakage.