Level 2: Mid-level evidence
The Towards a Revolution in COPD Health (TORCH) trial evaluated the effect of treatments for chronic obstructive pulmonary disease (COPD) on morbidity and mortality outcomes. The study randomized 6,184 patients with COPD aged 40-80 years to placebo vs. salmeterol (Serevent) 500 µg vs. fluticasone (Flovent Diskus) 50 µg vs. salmeterol/fluticasone combination in a single inhaler (Advair Diskus) twice daily for three years (N Engl J Med. 2007;356:775-789). Only 3,769 patients (61%) completed the study, but vital status at three years was known for 6,111 patients (99%).
Comparing combination therapy vs. placebo, the annual rate of moderate-to-severe exacerbations was 0.85 vs. 1.13 (P <.01) and the annual rate of severe exacerbations (requiring hospitalization) was 0.16 vs. 0.19 (P = .03). While combination therapy clearly reduced COPD exacerbations, the statistical significance of the effect on mortality varied with adjustments for other factors. Rates of all-cause mortality were 12.6% vs. 15.2% (P=.052, NNT 39) and rates of adjusted mortality were 10.3% vs. 12.6% (P=.03, NNT 44); there was no significant difference in COPD mortality (4.7% vs. 6%, P=.11).
In addition, there were no significant differences comparing salmeterol monotherapy with placebo or fluticasone monotherapy with placebo. Comparing placebo vs. salmeterol vs. fluticasone, the rates of all-cause mortality were 15.2% vs. 13.5% vs. 16%. The rates of adjusted mortality were 12.6% vs. 10.9% vs. 13.3% and rates of COPD mortality were 6% vs. 6.1% vs. 6.9%. The annual rate of moderate-to-severe exacerbations was 1.13 vs. 0.97 vs. 0.93 and the annual rate of severe exacerbations was 0.19 vs. 0.16 vs. 0.17.