The investigators hypothesize that earlier initiation of edoxaban in AF-related stroke patients may significantly reduce the early recurrence of ischemic stroke, compared with conventional strategy of anticoagulation following 1-3-6-12 rule. To expedite the verification of the hypothesis, the investigators are planning to use diffusion weighted imaging (DWI), which has been reported to be a surrogate to predict both short-term and long-term prognosis after stroke, to detect the recurrent ischemic events. Because data on the early anticoagulation in patients with AF-related stroke are limited, the investigators decided to perform a pilot study before establishing an appropriate clinical trial protocol. This study will help estimate the efficacy and safety of early administration of edoxaban, and determine the sample size of a following clinical trial. To ensure the safety in this pilot exploration, the investigators will not include patients with severe ischemic strokes, who are often prone to experience hemorrhagic transformation in the acute post-stroke period.

In patients with ischemic stroke and atrial fibrillation (AF), the risk of stroke recurrence is high, especially shortly after the event. Because AF-related strokes are usually larger in their size and more fatal than other types of ischemic stroke, it is important to prevent recurrent cardioembolic strokes with adequate secondary prevention. However, as damaged brain tissues and vessels are prone to bleed, early anticoagulation may be harmful.

For this reason, urgent anticoagulation has not been recommended in stroke patients with AF, and the appropriate time point to start anticoagulation remains controversial. Guideline recommends 1-3-6-12 rule* in initiating anticoagulation. However, this rule is not derived from a scientifically proven study results. Furthermore, although the risk of intracranial hemorrhage may be reduced to some extent with this strategy, the risk of early recurrence of embolic stroke may outweigh the potential benefit of delayed anticoagulation.

Edoxaban, which selectively blocks factor Xa, has a lower risk of hemorrhage, but with a similar efficacy in preventing ischemic events in patients with AF compared with warfarin. Even compared with the other factor Xa inhibitors, it is considered to have a lower risk of bleeding. Therefore, edoxaban may be safely given in the early phase in patients with stroke associated with AF, while not significantly increasing the risk of hemorrhages.

Inclusion Criteria:

1. Acute ischemic strokes (< 48 h from symptom onset) showing ischemic lesions confirmed by DWI, which are attributable to atrial fibrillation
2. Evidence of persistent or paroxysmal atrial fibrillation (already known or newly detected)
3. Age ≥20 y
4. Patients who provided informed consent

Exclusion Criteria:

1. Transient ischemic attack with no DWI lesions or severe ischemic strokes (NIHSS >16)

2. Significant hemorrhagic transformation (parenchymal hematoma type I or type II by the ECASS definition or those accompanying with worsening of an existing focal neurological deficit [NIHSS ≥4])10 on baseline MRI

3. Mechanical heart valve, rheumatic heart valve disease, or any other conditions requiring strong anticoagulation such as vitamin K antagonist or heparin treatment

4. Concomitant significant atherosclerotic stenosis (>50%) in the proximal arteries, which are possibly responsible for stroke lesions

5. Recent (<3 months) history of cerebral bleeding

6. Active internal bleeding or clinically significant bleeding

7. Severe anemia (Hb <10 g/dL) or bleeding diathesis (platelet count <100,000/uL or PT-INR >1.7) (If there is no active bleeding sign, it is permitted to enroll Hb <9 g/dL , platelet count <70,000/uL)

8. Uncontrolled hypertension: persistent systolic pressure >180 mmHg or diastolic pressure >110 mmHg

9. Active, advanced medical diseases (liver, kidney, pulmonary disease or cancer) with a life expectancy <6 months

10. Renal impairment (CrCl <30 mL/min) or undergoing Hemodialysis (or Peritoneal Dialysis)

11. Treatment with a strong inducer of p-glycoprotein (carbamazepine, dexamethasone, doxorubicin, nefazodone, pentobarbital, phenobarbital, prazocin, rifampin, St.John's wort, tenofovir, tipranavir, trazodone, vinblastine)

12. Contraindication to MRI

13. Pregnancy, breast-feeding or having a plan to be pregnant

14. Participation in the other investigational drug trials simultaneously or within 3 months before the first administration of the study medication. Observational studies without an intervention (eg study medication) are allowed.

15. Any clinical conditions (eg abnormal lab tests) unsuitable for undergoing clinical trials at the discretion of the clinical investigators

16. Known hypersensitivity to the study drug (edoxaban), its ingredients, or formulation excipients

17. Patient with liver disease related to coagulation disorder and clinically significant bleeding risk

18. Severe Liver disease

19. Patient who has increase risk of bleeding due to the following disease

    • recent gastrointestinal ulcer history
    • carcinoma increased risk of bleeding
    • recent brain or spinal injury
    • recent brain, spinal or optical surgery histroy
    • esophageal varix
    • arteriovenous malformations
    • vascular aneurysms (over 3.5cm)
    • intra spinal or cerebral vascular disorder

20. Patient with other anticoagulants

21. intermitant or severe mitral stenosis

22. a pulmonary embolism patient who is hemodynamic unstabled or required thrombolytic therpy or pulmonary emnolectomy