Commentary

The review by Singh et al and the trial by Tashkin et al were done to address uncertainty about the risks and benefits of inhaled anticholinergic therapy in COPD. However, understanding the discrepant findings is a challenge. Why did the meta-analysis show a 53% increase in myocardial infarction and a 26% increase in all-cause mortality with inhaled anticholinergics, whereas the subsequent large UPLIFT trial found relative reductions of 27% and 11% with tiotropium for these same outcomes?

Contradictory findings may reflect differences in individual trial characteristics. No trial was designed with mortality or cardiovascular events as a primary outcome. Most trials had high dropout rates that differed between groups, creating substantial opportunity for biased estimation of treatment effects, especially for secondary outcomes. However, mortality data for the UPLIFT trial were 98% complete, whereas no other long-term tiotropium trial collected mortality data after patients withdrew. Also, the UPLIFT trial followed patients for 4 years, allowing capture of more events than previous shorter trials. As a striking illustration, the UPLIFT trial reported deaths in 15.7% of participants compared with 1.8% in the meta-analysis.

Singh et al’s decision to pool placebo-controlled trials with trials using active comparators is questionable. Absence of statistical heterogeneity among trials provides insufficient reassurance if there is a high index of suspicion that other drugs have effects on the end points of interest. Consequently, the meta-analysis findings could reflect differing efficacy of these active agents for reducing cardiovascular events. However, the UPLIFT trial uniquely permitted other COPD therapies as cointerventions, many of which were excluded in other trials. Thus, it remains possible that inhaled anticholinergics increase risk of cardiovascular events when given alone but not when combined with an inhaled corticosteroid (ICS) and long-acting β-agonist (LABA). In addition, the extent to which patients in the placebo group of the UPLIFT trial received short-acting anticholinergic medications is not completely clear.

Nevertheless, the UPLIFT trial provides important reassurance that tiotropium, added to usual COPD therapy, does not increase cardiovascular events or mortality. Although an updated meta-analysis incorporating the UPLIFT trial would be helpful, it would be dominated by UPLIFT’s findings because its total events vastly outnumber those of all other trials combined. However, such a meta-analysis would be useful in helping us understand the extent to which chance can explain differences in study findings.

The finding of the UPLIFT trial that adding tiotropium in patients already receiving an ICS–LABA combination further reduces exacerbations is a new and important observation. It extends our knowledge from a previous trial of adding an ICS–LABA combination in patients already receiving tiotropium, which failed to show a significant difference in exacerbations but found benefits for secondary outcomes.1 This new evidence supports current guideline recommendations that all 3 medications should be used together in patients with COPD who have frequent exacerbations.2 3

At the same time, the failure of all of these trials to find significant effects on mortality or decline in FEV₁ is disheartening and suggests we may have reached the limit of what can be achieved with current inhaled medications. Further progress in combating the morbidity and mortality associated with severe COPD may now require shifting our attention to new therapeutic pathways.