P6. Glucocorticoids

The effect of inhaled glucocorticoids on the decline in lung function seen in COPD has been the subject of a series of controlled trials and systematic reviews over recent years, and the effect remains unclear. A Cochrane systematic review (Yang et al., 2007) that pooled results from 9 studies found no significant effect of inhaled glucocorticoids on the decline in lung function for studies of 2 years or longer duration, weighted mean difference = 5.8 mls/yr (95% CI) -0.28 to 11.88, 2,333 participants). However this analysis did not include a recently published large randomised controlled trial, the TORCH study (Calverley et al., 2007). Recent further analyses of TORCH reported by Celli (Celli et al., 2008b) suggested that all treatment arms in this study (fluticasone alone, salmeterol alone and the combination) reduced FEV₁ decline compared with placebo. The reported adjusted rates of decline in mls/year for active treatment compared to placebo were 13.0, 95% CI 4.3-21.7, p=0.003 for fluticasone and 16.3, 95% CI 7.7-24.9 p<0.001 for fluticasone plus salmeterol. Salmeterol results were similar to those of fluticasone. However, a critique of both the methods of the original study and this later analysis has been published by Suissa (Suissa, 2008) (see also Suissa and Barnes, 2009) who cautions that the original FEV₁ results were incomplete, that a true intention to treat analysis was not performed, and that some of the results may represent "regression to the mean."

The Cochrane systematic review also found a significant effect of inhaled glucocorticoids on the decline in quality of life. For studies over 6 months duration this effect was a decrease in the mean fall in the St Georges Respiratory Questionnaire score of -1.22 (95% CI -1.83 to -0.60, 2,507 participants) units compared to placebo, although the studies were of varying duration from 12 to 36 months. This review did not include participants from the TORCH study, which also found a significant reduction in the decline in QOL for participants on fluticasone compared to placebo, mean difference -2.0 (95% CI -2.9 to -1.0, 2,479 participants, p<0.001).

While these data do not support the use of inhaled glucocorticoids in all people with COPD, they are indicated for those with more severe disease (FEV₁ <50% predicted), frequent exacerbations or a documented response to inhaled glucocorticoids. While the long-term adverse effects of inhaled glucocorticoids are unknown, caution is needed if ceasing inhaled glucocorticoid treatment given the observation that abrupt withdrawal maybe associated with increased symptoms.(Wouters et al., 2005)