Triple Therapy in COPD: Time for Adaptive Selection Trials

Recent trials reported significant reductions in all-cause mortality with single-inhaler triple therapy for chronic obstructive pulmonary disease (COPD). However, reviews of these trials identified inconsistencies in the findings and methodological issues with the design and analysis, including the "adverse impact of inhaled corticosteroid (ICS) withdrawal rather than the addition" of the triple therapy. Indeed, ICS were discontinued in over 70% of the patients in these trials and 40% already using triple therapy, muddying the interpretation of the data. The "adaptive" clinical trial design is an efficient approach that allows continual modification of the study treatment allocation during follow-up. In this article, we propose the "adaptive selection" trial design, which applies the adaptive concept to the selection of patients into the trial by adapting the randomization choices to the treatment already used by the patients. With such a design, patients already on triple therapy would be excluded outright from trials of triple therapy effectiveness, while the others are randomly allocated to specific treatment arms according to their current treatment, avoiding issues of treatment withdrawal effects. Adaptive selection trials should be the norm for studies of COPD therapies. This approach would avoid the vexing effects of treatment withdrawal that have afflicted the recent triple therapy trials. This concept of adaptive selection has been applied in COPD to the question of whether patients can be safely de-escalated from ICS. It is time to also apply it to studies of the effectiveness of treatment escalation.