Efficacy and Safety of Non-Vitamin K Anticoagulants for Atrial Fibrillation in Relation to Different Renal Function Levels: A Network Meta-Analysis

BACKGROUND: We performed a network meta-analysis (NMA) comparing the efficacy (stroke or systemic embolism) and safety (major bleeding) among different non-vitamin K antagonist oral anticoagulants (NOACs) in patients with atrial fibrillation (AF) and renal impairment, with the aim of recommending the proper drug and the dose based on renal function. METHODS: We searched PubMed, EMBASE, Web of Science, and Cochrane Library with the items “dabigatran, edoxaban, apixaban, rivaroxaban, warfarin, and atrial fibrillation” through August 2019. NMA was analyzed with R (version 3.5.1, R Foundation for Statistical Computing) with the packages gemtc recalling JAGS (version 4.3.0) for the efficacy and safety of each drug with regard to different levels of renal function. NetMetaXL (version 1.6.1) and winBUGS (version 1.4.3) were used to obtain the cumulative ranking curve (SUCRA) of each drug. RESULT: In patients with normal renal function, dabigatran150 was ranked as the most effective drug (SUCRA 0.90), followed by dabigatran110 (SUCRA 0.68), apixaban (SUCRA 0.66), and rivaroxaban (SUCRA 0.59). With regard to the safety for preventing major bleeding, there was high probability that edoxaban30 (SUCRA 0.99) ranked first, compared to dabigatran110 (SUCRA 0.78) and edoxaban60 (SUCRA 0.66). For patients with mild renal impairment, with respect to the most effective drug for preventing stroke or systemic embolism, edoxaban60 ranked first (SUCRA 0.98), in comparison with dabigatran150 (SUCRA 0.74) and apixaban (SUCRA 0.64). Possibility of ranking first for the safest drug was edoxaban30 (SUCRA 0.99), followed by dabigatran110 (SUCRA 0.70) and apixaban (SUCRA 0.69). In patients with moderate renal function, dabigatran150 (SUCRA 0.95) ranked as the most effective drug in comparison with apixaban (SUCRA 0.66). Dabigatran110 (SUCRA 0.53), rivaroxaban (SUCRA 0.51), and edoxaban60 (SUCRA 0.50) had the similar probability of ranking third. When referred to the safest drug, probability of ranking first for preventing major bleeding was edoxaban30 (SUCRA 0.98), followed by apixaban (SUCRA 0.85) and edoxaban60 (SUCRA 0.64). CONCLUSION: In patients with AF and renal impairment and for patients with normal renal function, dabigatran 110 mg (bid) might have a better effect on the clinical results. And it does not coincide with patients taking dabigatran 110 mg with dose reduction for other factors including aged ≥75 years, renal impairment (CrCL 30–50 mL/min), gastritis, esophagitis, or gastroesophageal reflux, receiving concomitant verapamil, and so on. For patients with mild renal impairment, apixaban 5 mg (bid) would be a better choice for preventing stroke or systemic embolism and major bleeding, while apixaban 5 mg (bid) and edoxaban 60 mg (qd) were recommended for patients with moderate renal impairment. However, considering the fact of no RCTs for the head-to-head comparison, caution should be exercised over selecting each of NOACs for patients.