1. Initiating prevention after acute stroke in NVAF and beyond George Ntaios University of Thessaly, Larissa/Greece

2. Disclosures Scholarships: European Stroke Organization; Hellenic Society of Atherosclerosis Honoraria: Medtronic; Quintiles; Boehringer-Ingelheim Speaker fees: Sanofi; Boehringer-Ingelheim; Galenica; Elpen; Bayer; Winmedica; BMS/Pfizer Research support: European Union (Horizon 2020); BMS/Pfizer (ERISTA) Support to attend conferences: Bayer; Sanofi-Aventis; Pfizer; Lundbeck; Boehringer-Ingelheim; Galenica; Elpen; BMS/Pfizer

3. Anticoagulation after AF-related stroke: How soon (or late)?

4. Ntaios & Michel. Cerebrovasc Dis. 2011;32:246–253. Anticoagulation after AF-related stroke: How soon (or late)? 0 0 0.2 0.4 0.6 0.8 1.0 Recurrence risk (%) 20 Days after index stroke 15105

5. Anticoagulation after AF-related stroke: How soon (or late)? Anticoagulants for acute ischaemic stroke (review) Sandercock PAG, Counsell C, Kamal AK Treatment: unfractionated heparin (s.c. or i.v.), low molecular weight heparin, heparinoid (s.c. or i.v.), VKA, thrombin inhibitor or direct thrombin inhibitor (s.c.; for ICH outcome only) Sandercock, et al. Cochrane Database Syst Rev. 2015;3:CD000024. Treatment n Control n Odds-ratio Total10,86610,7390.76 [0.65, 0.88] Recurrent ischaemic or unknown stroke during treatment period Treatment n Control n Odds-ratio Total11,70111,2422.55 [1.95, 3.33] Symptomatic intracranial haemorrhage during treatment period Favours control Favours treatment 1.0 Favours control Favours treatment 1.0

6. Reasons to start early Low NIHSS Small/no brain infarction on MRI High recurrence risk e.g. thrombus on echo No haemorrhagic transformation Patient is clinically stable Young patient Blood pressure is controlled Reasons to start early Low NIHSS Small/no brain infarction on MRI High recurrence risk e.g. thrombus on echo No haemorrhagic transformation Patient is clinically stable Young patient Blood pressure is controlled Anticoagulation after AF-related stroke: How soon (or late)? Heidbuchel, et al. Europace. 2015;17:1467–1507. Reasons to wait High NIHSS Large/moderate brain infarction Haemorrhagic transformation Neurologically unstable Elderly patient Uncontrolled hypertension Reasons to wait High NIHSS Large/moderate brain infarction Haemorrhagic transformation Neurologically unstable Elderly patient Uncontrolled hypertension

7. TIA  1 day Small infarct  3 days Moderate infarct  6 days Large infarct  12 days Heidbuchel, et al. Europace. 2015;17:1467–1507. The 1–3–6–12 rule

8. Our non-AF patient 81 years old Fully dependent at 3 months Hypertensive, non-smoker, non-diabetic LDL: 104 mg/dL LA diameter: 42 mm 30% LICA stenosis 24 hrs ECG: - Echocardiography: -

9. ESUS: Embolic Strokes of Undetermined Source Hart, et al. Lancet Neurol. 2014;13:429–438.

10. ESUS: Diagnostic criteria Stroke detected by CT or MRI that is not lacunar. Absence of extracranial or intracranial atherosclerosis causing >50% luminal stenosis in arteries supplying the area of ischaemia. No major-risk cardioembolic source of embolism. (permanent or paroxysmal AF, sustained atrial flutter, intracardiac thrombus, prosthetic cardiac valve, atrial myxoma or other cardiac tumours, mitral stenosis, recent (<4 weeks) MI, LVEF <30%, valvular vegetations or infective endocarditis) No other specific cause of stroke identified. Hart, et al. Lancet Neurol. 2014;13:429–438.

11. Covert Atrial Fibrillation Cancer associated Covert non-bacterial thrombotic endocarditis Tumour emboli from occult cancer Arteriogenic emboli Aortic arch atherosclerotic plaques Cerebral artery non-stenotic plaques with ulceration Paradoxical embolism Patent foramen ovale Atrial septal defect Pulmonary arteriovenous fistula Minor-risk potential cardioembolic sources Mitral valve Myomatous valvulopathy with prolapse Mitral annular calcification Aortic valve Aortic valve stenosis Calcific aortic valve Non-AF atrial dysrhythmias and stasis Atrial asystole and sick-sinus syndrome Atrial high-rate episodes Atrial appendage stasis with reduced flow velocities or spontaneous echodensities Atrial structural abnormalities Atrial septal aneurysm Chiari network Left ventricle Moderate systolic or diastolic dysfunction (global or regional) Ventricular non-compaction Endomyocardial fibrosis ESUS: Potential causes Hart, et al. Lancet Neurol. 2014;13:429–438.

12. ESUS: Diagnostic algorithm Brain CT or MRI 12-lead ECG Precordial echocardiography Imaging of both extra- and intracranial arteries supplying the area of brain ischaemia Cardiac monitoring for ≥24 hours with automated rhythm detection Hart, et al. Lancet Neurol. 2014;13:429–438.

13. CRYSTAL-AF Sanna, et al. N Engl J Med. 2014;370:2478–2486.

14. CRYSTAL-AF: The more you look, the more you find Sanna. et al. N Engl J Med. 2014;370:2478–86. 0 10 Patients with AF detected (%) 36 Months since randomisation Hazard ratio 8.8 (95% CI 3.5, 22.2) p<0.001 by log-rank test 20 30 24181260 Control ICM

15. EMBRACE: The more you look, the more you find Gladstone, et al. N Engl J Med. 2014;370:2467–2477.

16. EMBRACE: The more you look, the more you find Gladstone. et al. N Engl J Med. 2014;370:2467–77. 0 Patients with AF detected (%) Duration of ECG monitoring 24 hours 5 10 15 20 1 week2 weeks3 weeks4 weeks 2.2 7.4 11.6 12.3 14.8

17. ESUS: Embolic Strokes of Undetermined Source Hart, et al. Lancet Neurol. 2014;13:429–438.

18. ESUS in the Athens Stroke Registry Ntaios, et al. Stroke. 2015;46:176–181. (Athens Stroke Registry)

19. Patients with acute first-ever ischaemic stroke (n= 2,735 ) ESUS (n= 275 ) n=2,731 n=2,109 n=1,612 n=743 n=641 Patients with missing data (n=4) Patients with lacunar stroke detected by CT or MRI (n=622) Presence of extracranial or intracranial atherosclerosis causing ≥50% luminal stenosis in arteries supplying the area of ischaemia (n=497) Major-risk cardioembolic source of embolism (n=869) Other/rare specific causes (n=102) Incomplete diagnostic work-up (or ≥2 causes identified) or non-visualised infarct (n=366)

20. ESUS: Patient characteristics Ntaios, et al. Stroke. 2015;46:176–181. (Athens Stroke Registry) ESUS (n=275) Large-artery atherosclerotic (n=497) Cardioembolic (n=869) Lacunar (n=622) Undetermined other than ESUS (n=366) Other determined (n=102) Demographics Female sex, n (%)99 (36.0)114 (22.9)461 (53.0)173 (27.8)166 (45.4)49 (48.0) Median (IQR) age, years68.0 (58.0–76.0)67.0 (60.0–73.0)76.0 (70.0–82.0)69.0 (60.0–75.0)74.0 (67.0–81.0)56.0 (43.0–74.0) Comorbidities – risk factors, n (%) Hypertension178 (64.7)382 (76.9)631 (72.6)518 (83.3)259 (70.8)50 (49.0) Diabetes mellitus65 (23.6)163 (32.8)192 (22.1)181 (29.1)115 (31.4)17 (16.7) Smoking83 (30.2)251 (50.5)157 (18.1)235 (37.8)111 (30.3)39 (38.2) Previous TIA27 (9.8)102 (20.5)53 (6.1)59 (9.5)39 (10.7)17 (16.7) Heart failure22 (8.0)23 (4.6)139 (16.0)15 (2.4)31 (8.5)10 (9.8) Dyslipidemia140 (50.9)273 (55.3)266 (30.7)306 (49.4)159 (43.6)40 (39.2) Coronary artery disease65 (23.7)132 (26.8)169 (19.5)84 (13.6)86 (23.7)16 (15.7) AF0 (0.0)21 (4.2)774 (89.1)36 (5.8)41 (11.2)0 (0.0) Mode of onset, n (%) Maximal at onset204 (74.2)255 (51.3)713 (82.1)290 (46.6)219 (59.8)58 (56.9) Gradual worsening37 (13.5)112 (22.5)82 (9.4)99 (16.0)62 (16.9)15 (14.7) Shuttering/stepwise15 (5.5)66 (13.3)24 (2.8)132 (21.3)25 (6.8)9 (8.8) Flactuating4 (1.5)23 (4.6)10 (1.2)40 (6.5)11 (3.0)8 (7.8) Unknown or missing data15 (5.5)41 (8.2)39 (4.5)59 (9.5)49 (13.4)12 (11.8)

21. ESUS (n=275) Large-artery atherosclerotic (n=497) Cardioembolic (n=869) Lacunar (n=622) Undetermined other than ESUS (n=366) Other determined (n=102) Clinical and laboratory values Systolic blood pressure, mm Hg150 (130–160)150 (140–170)150 (130–170)160 (140–180)150 (135–170)140 (120–150) Diastolic blood pressure, mm Hg85 (80–90)90 (80–90)85 (80–90)90 (80–100)84 (80–90)80 (70–90) Glucose, mg/dL109 (93–141)111 (95–154)118 (98–153)105 (92–139)116 (98–163)100 (90–125) NIHSS score5 (2–14)5 (2–15)13 (4–22)2 (1–4)8 (3–18)4 (1–12) ESUS: Stroke severity Ntaios, et al. Stroke. 2015;46:176–181. (Athens Stroke Registry)

22. ESUS: Potential causes Ntaios, et al. Stroke. 2015;46:176–181. (Athens Stroke Registry) Mitral valve Myxomatous valvulopathy with prolapse 5 (1.8%) Mitral annular calcification 8 (2.9%) Aortic valve Aortic valve stenosis 3 (1.1%) Calcific aortic valve 12 (4.4%) Non-AF atrial dysrhythmias and stasis Atrial asystole and sick-sinus syndrome 3 (1.1%) Atrial high-rate episodes 7 (2.6%) Atrial appendage stasis with reduced flow velocities or spontaneous echodensities 6 (2.2%) Atrial structural abnormalities Atrial septal aneurysm 10 (3.6%) Chiari network 0 (0.0%) Left ventricle Moderate systolic or diastolic dysfunction (global or regional) 42 (15.4%) Ventricular non-compaction 12 (4.4%) Endomyocardial fibrosis 1 (0.4%) Covert paroxysmal AF (detected during follow-up) AF detected on stroke recurrence 30 (11.0%) AF detected on monitoring during follow-up 50 (18.3%) AF not confirmed but strongly suspected 38 (13.9%) Cancer associated Covert non-bacterial thrombotic endocarditis 1 (0.4%) Tumour emboli from occult cancer 2 (0.8%) Arteriogenic emboli Aortic arch atherosclerotic plaques 9 (3.3%) Cerebral artery non-stenotic plaques with ulceration 29 (10.6%) Paradoxical embolism Patent foramen ovale 11 (4.0%) Atrial septal defect 3 (1.1%)

23. ESUS & AF at follow-up: How much causality is there? Ntaios, et al. Manuscript submitted.

24. ESUS & AF at follow-up: How much causality is there? Ntaios, et al. Manuscript submitted. AF ESUS (n=80) Non-AF ESUS (n=195) p-value NIHSS score5 (2–13)5 (2–14)0.998

25. ESUS: 5-year functional outcome Ntaios, et al. Stroke. 2015;46:2087–2093. (Athens Stroke Registry) mRS = 0-1mRS = >1 ESUS Cardioembolic Large-artery atherosclerotic Lacunar Undetermined other than ESUS Miscellaneous

26. ESUS: 5-year stroke recurrence Ntaios, et al. Stroke. 2015;46:2087–2093. (Athens Stroke Registry) 0 ESUS 0 0.1 0.2 0.3 0.4 Risk of stroke recurrence (%) 1224364860 Months Cardioembolic Large-artery atherosclerotic Undetermined other than ESUS Lacunar Miscellaneous

27. So, how to treat my ESUS patient? Approach 1 Furie, et al. Stroke 2011;42:227–276.

28. So, how to treat my ESUS patient? Approach 2

29. So, how to treat my ESUS patient? Approach 3

30. NAVIGATE – ESUS Study number NCT02313909: https://clinicaltrials.gov/show/NCT02313909 Patients with recent ESUS (7 days to 6 months before randomisation) R N=7,060 Primary endpoints Time from randomisation to first occurrence of: Stroke (ischaemic, haemorrhagic or undefined) or TIA with positive neuroimaging or SE Major bleeding (ISTH criteria) Rivaroxaban 15 mg once daily Aspirin 100 mg once daily ~36 months

31. RE-SPECT ESUS *Dabigatran 110 mg twice daily in selected patients Study number NCT02239120: https://clinicaltrials.gov/show/NCT02239120 Primary endpoints Time to first recurrent stroke (ischaemic, haemorrhagic or unspecified) Patients with recent ESUS (up to 3 months before randomisation) R N=6,000 Dabigatran 150 mg* twice daily Aspirin 100 mg once daily Up to 36 months

32. ATTICUS *Apixaban 2.5 mg twice daily in selected patients Study number NCT02427126: https://clinicaltrials.gov/show/NCT02427126 Primary endpoint Occurrence of ≥1 new ischaemic lesion identified by MRI at 12 months when compared with baseline FLAIR/DWI MRI obtained at time of study drug initiation Patients with ESUS and ≥1 suggestive risk factor for cardiac embolism R N=500 Apixaban 5 mg* twice daily Aspirin 100 mg once daily 12 months

33. ESUS: Risk stratification for recurrence and mortality Ntaios. et al. Manuscript submitted. CHADS 2 (n) 3132982811315220 CHADS 2 – Recurrence (%) 0815.114.921.426.930 CHADS 2 – Mortality (%) 0412.118.519.828.835 0 01234 50 100 150 250 350 0 10 20 30 40 Percent (%) Absolute number (n) 200 300 5 & 6 0 012345678 50 100 150 200 250 0 10 20 30 40 50 Percent (%) Absolute number (n) CHA 2 DS 2 -VASc (n) 1522211632141939041156 CHA 2 DS 2 -VASc – Recurrence (%) 079.51616.41526.7242033 CHA 2 DS 2 -VASc – Mortality (%) 015.96.720.118.126.72746.733.3

34. Ntaios, Michel & Vemmos. Ongoing. Prediction of Atrial Fibrillation in patients with Embolic Stroke of Undetermined Source (AF-ESUS) Multicenter investigator- initiated study ESUS AF predictors Prognostic score Lausanne Larissa Athens

35. Take-home messages 1–3–6–12 rule Cryptogenic  ESUS ~10% of all stroke patients ESUS needs a complete (?) diagnostic work-up Covert AF is frequently detected in ESUS High recurrence rate Ongoing studies are evaluating the use of NOACs in ESUS