Taiwan was among the countries with high prevalence of end stage renal disease (ESRD), and more than 90% of ESRD patients in Taiwan received hemodialysis. Thrombosis are the most common complications of hemodialysis vascular access, with an annual incidence of 30-65% for dialysis grafts. Although endovascular thrombectomy is effective and convenient, the recurrence rate was high, nearly 50% in three months.

The mechanisms of dialysis vascular access thrombosis were multi-factorial, including flow stasis, endothelial injury and hypercoagulability. Our recent study showed that 63% of patients with early thrombosis after angioplasty had at least one thrombophilic factor. Nonetheless, no antithrombotic regimen has been validated to be effective for prevention of thrombosis, either primary or secondary prevention. Novel oral anticoagulants (NOACs) have shown comparable efficacy as VKA with significant decrease in major bleeding. Furthermore, NOACs have the advantage of rapid onset without the need for titration, which should be more effective in the critical period early after thrombectomy. NOAC have almost replaced the role of VKA for the prevention of stroke in patients with atrial fibrillation. They also replaced oral and parenteral anticoagulants in the treatment and prevention of deep vein thrombosis. Among the 4 available NOACs today, only apixaban had received approval by the US Food and Drug Administration (FDA) to be used in patients with ESRD for stroke prevention in atrial fibrillation.

In consideration of the trade-off between thrombotic and bleeding risk, we aimed at secondary prevention for patients with a thrombosis event after a successful thrombectomy procedure. Apixaban would be used because it was approved by FDA for the use of hemodialysis patients, with a risk of major bleeding of 5% for 3 months. Furthermore, considering the ethnic (Asia population) and clinical (ESRD and high bleeding risk) background of our target population, 2.5 mg twice daily dose was chosen in this study to minimize the bleeding risk. This study is a multi-center, prospective, open-labeled, randomized trial with blinded evaluation of all outcomes (PROBE design). We anticipated to enroll 150 patients, with 1:1 randomization to apixaban and control group (no antithrombotic agent). The duration of therapy will be 3 months and the primary outcome is the time to recurrent thrombotic event. Secondary outcomes included frequency of thrombosis, repeat interventions, and bleeding events. We hypothesized that apixaban could prolong the thrombosis-free interval after a successful thrombectomy procedure of hemodialysis vascular access.

Inclusion Criteria:

• Age 20-99 years old
• End stage renal disease, patients on maintenance hemodialysis for at least one month
• Dialysis vascular access thrombosis, documented by angiograph, and thrombosis was salvaged by endovascular or surgical procedures successfully

Exclusion Criteria:

• History of intracranial hemorrhage
• Major bleeding in recent 3 months, which defined as Bleeding Academic Research Consortium (BARC) 2 criteria
• Concomitant use of dual antiplatelet agent (aspirin and clopidogrel)
• Concomitant use of ticagrelor
• Concomitant use of warfarin
• Planned to receive surgery in recent 3 months
• Planned to receive coronary stents in recent 3 months
• Hemoglobin < 7.0 g/dL or Platelet count < 80 K/uL
• Moderate or severe liver dysfunction, defined as Child-Pugh score > 6
• Patient received bioprosthetic or mechanical heart valve
• Cannot signed informed consent