Abstract
Patients with non-valvular atrial fibrillation (AF) and advanced chronic kidney disease (CKD), including those with end-stage kidney disease (ESKD) on dialysis, present a unique therapeutic challenge. This population is at an elevated risk for both thromboembolic events and severe bleeding complications. Because patients with creatinine clearance <25-30 mL/min were excluded from pivotal trials, the comparative safety and efficacy of direct oral anticoagulants (DOACs) versus traditional vitamin K antagonists (VKAs) in this cohort remain debated. A search of MEDLINE, Embase, and the Cochrane Central Register of Controlled Trials was performed from inception to the present to identify randomized controlled trials (RCTs) and adjusted observational studies comparing DOACs to VKAs in patients with AF and advanced CKD (stages 4-5 or dialysis). The primary outcomes were stroke or systemic embolism (SE) (efficacy) and major bleeding (safety). Hazard ratios (HRs) and 95% confidence intervals (CIs) were pooled using a random-effects (DerSimonian-Laird) model. Meta-regression, trial sequential analysis (TSA), and GRADE certainty assessments were applied. The final analysis included 21 studies (four RCTs, 17 observational cohorts) encompassing 184,136 participants. Compared to VKAs, DOACs were associated with a statistically significant 28% reduction in the risk of stroke or SE (heart rate (HR), 0.72; 95% confidence interval (CI), 0.60-0.86; p = 0.0004; moderate certainty). For safety, DOACs significantly reduced the risk of major bleeding by 26% compared to VKAs (HR, 0.74; 95% CI, 0.61-0.90; p = 0.0026; low to moderate certainty). However, substantial statistical heterogeneity was observed for the bleeding outcome (I(2) = 80.2%). Heterogeneity was primarily driven by agent-specific effects, with apixaban demonstrating the most favorable safety profile. Meta-regression confirmed that the specific DOAC agent was a significant moderator of bleeding risk (p < 0.0001), with apixaban (HR, 0.63) and rivaroxaban (HR, 0.75) driving the safety benefits, whereas dabigatran was associated with increased bleeding (HR, 1.48). TSA for stroke reduction indicated that while the cumulative Z-curve crossed the conventional benefit boundary, the information size remained below the heterogeneity-adjusted requirement. In patients with AF and advanced CKD or ESKD requiring dialysis, the use of DOACs, specifically apixaban and rivaroxaban, demonstrated superior efficacy in stroke prevention and a safer bleeding profile compared to VKAs. These findings suggest that factor Xa inhibitors, specifically apixaban and rivaroxaban, should be preferred over VKAs in this high-risk population; however, adequately powered RCTs are needed to refine specific dosing strategies.