Level 1: Likely reliable evidence
A systematic Cochrane review evaluated 47 randomized placebo-controlled trials of inhaled corticosteroids (ICS) involving 13,139 patients with stable chronic obstructive pulmonary disease (COPD) (Cochrane Database Syst Rev. 2007:CD002991). Most trials were of good quality.
Comparing ICS vs. placebo, there were no significant differences in mortality: 8% vs. 8% in nine trials with 8,390 patients; 1.4% vs. 2.2% in three trials lasting one year with 1,907 patients; 9.4% vs. 9.3% in five trials lasting two years or more with 6,483 patients.
In an analysis of five trials with 2,586 patients, ICS were associated with 0.26 fewer exacerbations per patient-year. The mean number of exacerbations per year was 0.93 with ICS vs. 1.13 with placebo (P <.05) in the largest unpublished trial with 3,056 patients. ICS were also associated with a lower rate of decline (5.8 mL/year) in forced expiratory volume in one second (FEV1) (P=.06) in meta-analysis of six trials with 3,610 patients and a slower rate of decline in quality of life in five trials with 2,507 patients.
Patients using ICS had an increased risk of oropharyngeal candidiasis as well as hoarseness. There were no significant differences in fractures and bone mineral density. The response to ICS was not predicted by oral steroid response, bronchodilator reversibility, or bronchial hyperresponsiveness in COPD patients.