Background: Atrial fibrillation (AF) is a major risk factor for stroke and death conferring a 5-fold risk of stroke compared to patients without AF. The appropriate and timely anticoagulant therapy of patients at risk with vitamin K antagonist (VKA) or non-vitamin K-oral anticoagulants (NOACs) is one of the core principles of AF management. As factor Xa inhibitor marketed for stroke prevention in patients with non-valvular AF, edoxaban has shown to be at least as effective and safe as the VKA warfarin in pivotal trials. However, phenprocoumon is the predominant VKA in clinical practice in Germany and comparative data in terms of effectiveness and safety for edoxaban versus phenprocoumon are scarce.

Objectives: The purpose of this study was to evaluate the suspected clinical advantages of edoxaban versus phenprocoumon in non-valvular AF patients in a German real-world setting.

Methods: The study was conducted as a non-interventional retrospective new-user cohort study using longitudinal German Statuary Health Insurance claims data of the Institute for Applied Health Research Berlin (InGef) from January 1^st, 2014 through June 30^th, 2019. The research database comprises anonymized healthcare claims of about four million insured persons. Continuous enrolment for 12 months before anticoagulant initiation was required to assess baseline characteristics. New users of edoxaban and phenprocoumon were compared to assess the risk for stroke/systemic embolism (SSE), all strokes (AS), ischemic stroke (IS), hemorrhagic stroke (HS) and all-cause mortality (effectiveness events), as well as major bleedings (MB), intracranial bleedings (ICB), gastrointestinal bleedings (GIB) and any bleedings (AB) (safety events) in non-valvular AF patients. Hazard ratios (HR) were estimated through multiple outcome-specific cox proportional hazard regression models adjusting for differences in baseline characteristics. As sensitivity analysis, a propensity score matching was performed.

Results: Between January 1^st, 2015 and December 31^st, 2018, a total of 7,975 vs 13,319 patients newly initiated treatment with edoxaban vs phenprocoumon. After adjusting for baseline confounders, edoxaban revealed a numerically lower risk of SSE (HR: 0.85, 95% CI: 0.70–1.02, p = 0.087) and AS (HR: 0.87, 95% CI: 0.71–1.06, p = 0.177) compared with phenprocoumon without reaching statistical significance. Edoxaban was associated with significantly lower risk for HS (HR: 0.52, 95% CI: 0.33–0.83, p = 0.006), MB (HR: 0.69, 95% CI: 0.58–0.81, p < 0.001), ICB (HR: 0.48, 95% CI: 0.35–0.67, p < 0.001) and AB (HR: 0.79, 95% CI: 0.73–0.86, p < 0.001) than phenprocoumon. The risk for GIB was numerically but not statistically higher for edoxaban (HR: 1.11, 95% CI: 0.95–1.29, p = 0.176) while the risks of IS (HR: 0.99, 95% CI: 0.80–1.24, p = 0.964) and all-cause mortality (HR: 1.00, 95% CI: 0.91–1.11, p = 0.924) were similar between the treatment groups. Results were consistent in analysis using propensity score matching.

Conclusion: The results of this real-world analysis indicated better effectiveness and safety outcomes in patients with non-valvular AF initiating edoxaban treatment compared to phenprocoumon. The findings are largely consistent with the results of comparable real-world studies and confirm the findings of the pivotal trial (ENGAGE AF-TIMI 48) in a German real-world population.