Association Between Use of Inhaled Corticosteroids and the Risk of All-Cause

[gmohan]

August 24, 2022

Association Between Use of Inhaled Corticosteroids and the Risk of All-Cause Mortality in Patients With COPD
Chest


TAKE-HOME MESSAGE
To assess whether maintenance therapy with inhaled corticosteroids (ICS) reduces the risk of mortality in patients with chronic obstructive pulmonary disease (COPD), this meta-analysis evaluated 60 randomized controlled trials involving 60,552 patients receiving ICS and 42,482 patients not receiving ICS.
Regimens containing ICS—particularly when used for more than 6 months and as a part of triple inhaler therapy—conferred a statistically significant reduction in the risk of all-cause mortality. Notable predictors of the mortality benefit included peripheral eosinophil count ≥200/μL and a low dose or medium dose of ICS.

Prior studies have demonstrated contrasting results regarding the outcomes of patients with COPD receiving ICS. The present meta-analysis indicates a mortality benefit, highlights subgroups most likely to benefit, and may have implications for ICS dosing in patients with COPD.

The meta-analysis of 60 randomized trials by Chen et al reports that inhaled corticosteroids (ICS) in the treatment of COPD reduce all-cause mortality by at least 10%, and up to 40% in some subgroups.1 Does this analysis imply that all patients with COPD should be on ICS-based therapy?

Taking a closer look at these trials, we note that the results of the meta-analysis are largely driven by three large trials, SUMMIT, IMPACT and ETHOS, with the latter two contributing 41% and 32% reductions in mortality with ICS use, respectively. However, these triple therapy trials are affected by the abrupt discontinuation of the patients’ treatment before randomization.2-4 Indeed, regular ICS use was withdrawn in 70% and 80% of the patients who were randomized to a dual LABA/LAMA bronchodilator in IMPACT and ETHOS, respectively.5,6 How does this affect the results of these trials?

First, mortality was reduced with triple therapy mainly in the first 3 months after randomization (by 76% in IMPACT and by 63% in ETHOS), compared with dual bronchodilators, with no reduction in mortality during the subsequent 9 months under triple therapy. Could ICS really act so quickly and remarkably on preventing death, with no sustained benefit? Unlikely; it is more compatible with an effect of withdrawal of prior ICS use.

Second, the analysis restricted to the non-users of ICS before randomization found that mortality was not reduced with triple therapy (non-significant 25% and 49% increases in IMPACT and ETHOS, respectively), compared with dual bronchodilators.5,6 This does not support a beneficial effect of ICS on mortality, on the contrary.

Historically, the suggestion that inhaled corticosteroids could reduce mortality in COPD was first reported in a 2001 observational study, followed by several others, shown to be affected by the notorious immortal time bias. While randomized trials are the “gold standard” of evidence, it is noteworthy that some trials can be affected by bias, and their meta-analyses will inherently produce biased results. The meta-analysis of ICS-containing therapy in COPD associated with lower mortality is one such example.

As we await more reliable trials, unaffected by this bias, ICS-containing inhaled therapy for patients with COPD should be reserved for those with a history of asthma, multiple recent exacerbations, significant eosinophilia, and poor lung function, in line with treatment guidelines.

G Mohan