Will SUMMIT reach the peak in COPD?

INTRODUCTION The Study to Understand Mortality and Morbidity in COPD (SUMMIT) is a randomised controlled trial conducted in patients with COPD and comorbid cardiovascular disease, designed to assess the ‘impact of Fluticasone Furoate/Vilanterol combination (FF/VI) and its individual components on the survival of patients with moderate COPD and either a history of CVD or at increased risk for CVD’. Fluticasone furoate is an inhaled corticosteroid (ICS) and vilanterol is a longacting β-agonist (LABA). The trial is enrolling 16 000 such patients with moderate COPD randomly assigned to once daily treatment with this ICS–LABA combination, the ICS only, the LABA only or placebo. Patients in this event-driven trial will be followed up until 1000 deaths occur, which is expected to take up to 44 months. Mortality is the primary endpoint and the primary comparison is between the ICS–LABA combination and placebo. The study is designed as a superiority trial with 90% power to detect a 30% reduction in all-cause mortality when comparing the ICS–LABA with placebo at the two-sided 1% significance level. Secondary endpoints will also be studied, including lung function decline and composite cardiovascular events. The design of SUMMIT is similar to that of the Towards a Revolution in COPD Health (TORCH) trial, with the exception that SUMMIT focuses on a study population of patients with moderate COPD and at cardiovascular risk. As well, while all patients in TORCH were followed for an exact 3-year period, SUMMIT being event-driven is expected to follow-up patients between 15 and 44 months. At first glance, the SUMMIT trial appears to have set unattainable goals. Indeed, the magnitude of the targeted effect in terms of power, a 30% reduction in all-cause mortality (HR=0.70), seems overly ambitious. In the TORCH trial, the target for this same outcome was a 27% reduction (HR=0.73) with an even smaller sample size. Moreover the primary outcome is all-cause mortality, rather than cause-specific mortality. Because of the nature of the study population, one would expect that respiratory or cardiovascular mortality be the focus of the outcome, particularly in view of a rationale for this trial, namely, that ‘combination treatment seemed to have a similar effect on respiratory and cardiovascular mortality in the primary TORCH report’. 3 Including non-respiratory and non-cardiovascular deaths in the outcome can be thus perceived as potentially diluting any beneficial effects of the treatment. In all, it can be construed that SUMMIT set the bar much too high to find significant effectiveness for the drug and, if a benefit is indeed reached with this trial, it will be against all odds. Not necessarily so. In fact, several methodological aspects of the study design could affect the results in the other direction. The design of the TORCH trial revealed several concerns that could have led to biased estimates of the effects of the ICS–LABA combination inhaler on mortality and on the secondary outcomes. Many of these methodological limitations are still present in the design of the SUMMIT trial and could also affect the estimates of this trial. In this paper, we review these methodological issues in the context of the SUMMIT trial and discuss how they could potentially affect the results of this trial and the interpretation of the data.