Rationale This analysis aimed to obtain a comprehensive and objective safety assessment of the combination of tiotropium (T), a long-acting muscarinic antagonist, with olodaterol (O), a long-acting β₂-agonist, (T+O) in patients with moderate to very severe chronic obstructive pulmonary disease (COPD).

Methods Data from two, 52-week, pivotal Phase III trials investigating T+O 5/5 µg and T+O 2.5/5 µg versus T 2.5 µg, 5 µg and O 5 µg were pooled, and patient narratives and profiles of serious adverse-event (SAE) reports were reviewed by an independent Adjudication Committee. The committee members independently assessed all SAEs to determine if any deaths, hospitalisations or intubations were respiratory-related, cardiovascular-related, cerebrovascular-related or other event-related. For an SAE adjudicated as respiratory-related, determination was made if it was related to COPD or pneumonia. For an SAE adjudicated as cerebrovascular-related, determination was made if it was related to stroke or other cerebrovascular events. Incidences of the composite end point (death, hospitalisation and intubation for respiratory-, cardiovascular-, cerebrovascular- or other-related events) and the individual components of this end point were evaluated.

Results The safety population for the primary analysis included patients from two trials (NCT01431274 and NCT01431287) in which 799/5162 (15.5%; range across treatments: 14.3–16.5%) had any adjudicated event of interest. As expected in a moderate to very severe COPD population, most SAEs were respiratory-related (8.1%; 420 patients). Eighty-three (1.6%) patients had cardiovascular-related SAEs and 27 (0.5%) had cerebrovascular-related SAEs; 363 (7.0%) had SAEs that were adjudicated as non-respiratory-, non-cardiovascular- or non-cerebrovascular-related. Most adjudicated SAEs (763 patients; 14.8%) were hospitalisations, while there were 26 (0.5%) patients with intubation and 75 (1.5%) with fatal SAEs (86 [1.7%] had fatal SAEs when including vital status follow-up).

Conclusions The adjudicated analysis of SAEs demonstrated that the risk of having an event (composite end point of hospitalisations, intubations and death whether related to respiratory, cardiovascular, cerebrovascular or other cause) was similar for T+O 5/5 μg compared to T+O 2.5/5 μg or any of the monotherapy components. Similar conclusions can be drawn for the individual events of hospitalisations, intubations and death.