#280 Effectiveness and safety of direct oral anticoagulants vs VKA in patients on chronic dialysis: a target trial emulation using nationwide registry data

Abstract

Abstract Background and Aims The mortality and morbidity rates among patients on dialysis remain high and cardiovascular disease is the leading cause of death in this population. In chronic kidney disease (CKD), a major thromboembolic risk is paradoxically associated with a major haemorrhagic risk. Hence, the prescription of oral anticoagulants in CKD patients is challenging. Clinical trials of direct oral anticoagulants (DOAC) are scarce and inconclusive in patients who are receiving dialysis, for whom DOAC are not labelled in Europe. In an exhaustive French nationwide registry of patients undergoing chronic dialysis, we compared the effectiveness and safety of off-label DOAC use versus approved vitamin K antagonist (VKA). Method Data on patients on dialysis were extracted from the French Renal Epidemiology and Information Network (REIN) registry and merged with data from the French national healthcare system database (Système National des Données de Santé, SNDS). Patients on dialysis who had initiated treatment with an oral anticoagulant between January 1st, 2012, and December 31st, 2020, were eligible for inclusion. The primary safety outcome was the occurrence of major bleeding events and the primary effectiveness outcome was the occurrence of thrombotic events. Using target trial emulation design, we compared the safety and effectiveness outcomes for DOAC and VKA using propensity-score-weighted cause-specific Cox regression. Results 8 954 patients received an oral anticoagulant (483 DOAC and 8 471 VKA) for the first time after the initiation of dialysis. Over a median [interquartile range] follow-up period of 1.7 [0.8-3.2] years, 2 567 patients presented a first thromboembolic event and 1 254 patients had a bleeding event. After propensity score adjustment, the risk of a thromboembolic event was significantly lower in patients treated with a DOAC than in patients treated with a VKA (weighted hazard ratio (HR) [95% confidence interval (CI)]: 0.66 [0.46; 0.94] (Figure). A non-significant trend toward lower risk of major bleeding events was found in patients treated with a DOAC compared to patients treated with a VKA (weighted HR [95% CI]: 0.68 [0.41; 1.12]) (Figure). Sensitivity analyses showed the same trends. Conclusion The fact that patients on dialysis are excluded from large, cardiovascular-disease-focused randomized controlled trials (such as those with DOACs) significantly limits the evidence needed to make clinical decisions. Most of the retrospective cohort studies comparing these two drug classes were performed in the USA, where, in contrast to Europe, apixaban has been approved for the treatment of non-valvular fibrillation in patients on haemodialysis. Our real-world study of patients on dialysis is one of the first of its type to have assessed the effectiveness and safety of off-label DOAC prescription (relative to VKA prescription) in Europe. In a large group of dialysis patients initiating oral anticoagulation, the off-label use of DOACs was associated with a statistically significant lower risk of thromboembolic events, and a non-significant trend toward lower risk of bleeding relative to VKA use. These results are reassuring with regard to the off-label use of DOACs and might facilitate the establishment of an expert consensus and guidelines on antithrombotic therapy in patients with ESKD.