Atrial fibrillation (AF) is a common disease causing the heart to beat rapidly and irregularly. As a result, AF increases the risks of clots and strokes. As individuals age, they face higher risk to develop AF and stroke, particularly those aged 80 years and older. To prevent strokes, patients with AF are often prescribed oral anticoagulation (OAC) drugs such as vitamin K antagonists (VKAs) like warfarin or direct oral anticoagulants (DOACs). The more recently approved DOACs, which include dabigatran, rivaroxaban, apixaban and edoxaban, are currently recommended by most international guidelines over warfarin for most patients with AF to prevent strokes. Despite their similar or superior effectiveness at preventing strokes compared to VKAs, the lack of clear guidelines regarding individuals with AF over 80 and under-prescription of DOACs in this population complicates AF management. Only two studies directly compared DOACs to evaluate their effectiveness and safety in patients with AF over 80 years. Their findings suggested the superiority of apixaban over dabigatran and rivaroxaban, However, no study compared edoxaban with other DOACs, given its more recent approval. Thus, we will investigate the safety and effectiveness of edoxaban compared with apixaban. We will form two distinct groups of patients with AF treated with either edoxaban or apixaban. We will then compare the number of patients who experience a stroke and a major bleeding between the groups to inform clinicians about the most effective choice of DOAC in this population.
With ageing population, AF and strokes are expected to prevail in the elderly, since AF and advancing age independently increase thromboembolic risk. Although most guidelines recommend DOACs over warfarin for most patients with nonvalvular AF (NVAF), further real-world evidence studies on patients over 80 are needed to inform clinicians on DOAC prescription choice for an optimal AF management. Previous studies suggested the superiority of apixaban over rivaroxaban and dabigatran, but comparative studies with edoxaban are scarce. Consistent with their increasing prescription rates, we will conduct a UK population-based cohort study to assess the effectiveness and safety of edoxaban compared with apixaban in patients with NVAF aged ≥80. We will use the CPRD, linked to the Hospital Episode Statistics inpatient database and Office for National Statistics mortality database. We will assemble a cohort of all patients with NVAF aged ≥80 newly treated with edoxaban or apixaban between 2015 and 2021. The primary effectiveness endpoint of interest will be a composite of hospitalisation with an incident ischaemic stroke, transient ischaemic attack, or systemic embolism. The primary safety endpoint of interest will be major bleeding defined as any bleeding requiring hospitalisation. Patients will be followed until the occurrence of the outcome (depending on the studied outcome), all-cause mortality, OAC discontinuation or switching, end of registration with the general practice, or end of the study period, whichever occurs first. Weighted Cox proportional hazards models will be used to separately compute the hazard ratios (HR) and 95% confidence intervals (95%CI) of the studied outcomes associated with edoxaban compared with apixaban. In secondary analyses, we will evaluate if these risks vary with patients’ characteristics, previous OACs use and DOAC doses. We will conduct sensitivity analyses to assess the robustness of the results.
Primary outcomes: composite outcome of ischaemic stroke, transient ischaemic attack, and systemic embolism; major bleeding.
Secondary outcomes: all-cause mortality; composite outcome of stroke/TIA, systemic embolism, gastrointestinal bleeding, and intracranial haemorrhage.
Samy Suissa - Chief Investigator - Sir Mortimer B Davis Jewish General Hospital
Christel Renoux - Corresponding Applicant - McGill University
Richard Chiv - Collaborator - McGill University
Sarah Beradid - Collaborator - Sir Mortimer B Davis Jewish General Hospital
HES Admitted Patient Care;ONS Death Registration Data