The use of vitamin K antagonists (VKAs) in hemodialysis patients with atrial fibrillation (AF) is controversial. No randomized trials are available and observational studies have yielded conflicting results, engendering a large clinical practice variability and physician uncertainty. An unresolved but highly relevant question is whether AF poses a true risk of ischemic stroke in hemodialysis and whether any form of oral anticoagulation is therefore warranted.

We conducted a systematic review of studies that compared the incidence of ischemic stroke and bleeding in hemodialysis patients with AF taking VKA and those not taking VKA. When hemodialysis patients had been pooled with peritoneal dialysis, kidney transplant, or stage V chronic kidney disease patients, unpublished outcome data of the hemodialysis subgroup were obtained through personal communication. The main outcome measures were ischemic stroke/thromboembolic events, all-cause mortality, major bleeding, and hemorrhagic stroke. Combined hazard ratios (HRs) and 95% CIs were calculated using a random-effects model.

Twelve prospective or retrospective cohort studies were included in the meta-analysis, totaling 17,380 hemodialysis patients of whom 4,010 (23.1%) received VKA. In VKA-treated patients, mean CHADS₂ or CHA₂DS₂VASc score was low (range 1.7-2.75) or a sizeable proportion of patients had scores <2 (range 2%-23%). Time in the therapeutic range or mean international normalized ratio was generally low. Treatment with VKA was associated with a nonsignificant 26% reduction of the risk of ischemic stroke (HR 0.74; 0.51-1.06), a 21% increase in total bleeding risk (HR 1.21; 1.03-1.43), and no effect on mortality (HR 1.00; 0.92-1.09). Vitamin K antagonist almost doubled the risk of hemorrhagic stroke, but this did not reach the limit of statistical significance (4 studies, n = 16.365; HR 1.93; 0.93-3.98).

Our meta-analysis revealed a trend for a reduction of the risk of ischemic stroke in hemodialysis patients with AF treated with VKA. The true protective effect may have been underestimated, owing to inclusion of low-risk patients not expected to benefit from anticoagulation and to suboptimal anticoagulation. However, assessment of the overall effect of VKA in hemodialysis patients should also take into account the increased risk of bleeding, in particular of hemorrhagic stroke. Whether new oral anticoagulants provide a better benefit-risk ratio in hemodialysis patients should be the subject of future trials.