NOACs for Primary and Secondary Stroke Prevention: From Clinical Trials to Real-World Data To Practical Considerations

NOACs for Primary and Secondary Stroke Prevention: From Clinical Trials to Real-World Data To Practical Considerations Mark J. Alberts, MD, FAHA Hartford HealthCare Hartford, CT USA AF confers an increased thromboembolic risk, notably in the brain AF confers a near 5-fold risk of stroke 1 It is estimated that 20% of all strokes are caused by AF 2 AF is often asymptomatic 3 The absence of symptoms (e.g. palpitations) does not imply a lower risk of thromboembolism 3 NVAF-related stroke is associated with increased severity and disability compared with non-nvaf-related stroke 4 1. Wolf et al. Stroke 1991;22:983-8. 2. Friedman et al. Circulation 1968;38:533-541. 3. Flaker et al. Am Heart J 2005;149:657-63. 4. Lin H-J et al. Stroke. 1996;27:1760-1764. 2 Selecting appropriate antithrombotic therapy for a patient is an important management decision in AF BALANCING RISK Stroke Bleeding Aim: reducing the risk of thrombotic events with an acceptable increase in bleeding complications Fang MC. Ann Intern Med 2011;155:636-7. 3 1

What are the NOACs? Two new classes of anticoagulants Direct thrombin inhibitors (DTI) Factor Xa inhibitors (FXa) Thin the blood, but unlike warfarin (coumadin) do not require monitoring, testing, have no common food or drug interactions Dabigatran Rivaroxaban Apixaban Edoxaban Recent meta-analysis all 4 major studies >42,000 patients Rx with a NOAC >29,000 Rx with warfarin RE-LY, ROCKET-AF, ARISTOTLE, ENGAGE (did not include AVERROES) Mean follow-up 1.8 to 2.8 years Median TTR 58% to 68% These data are from a pooled analysis These Ruff data et al. are Lancet. from a 2014;383:955-62 pooled analysis Ruff et al. Lancet. 2014;383:955-62 Meta-analysis: Efficacy of NOACs vs warfarin for stroke prevention in patients with NVAF 1 Stroke or SE Study NOAC (events) Warfarin (events) RR (95% CI) RR (95% CI) p value RE-LY* 134 / 6076 199 / 6022 0.66 (0.53 0.82) 0.0001 ROCKET AF 269 / 7081 306 / 7090 0.88 (0.75 1.03) 0.12 ARISTOTLE 212 / 9120 265 / 9081 0.80 (0.67 0.95) 0.012 ENGAGE AF-TIMI 48 296 / 7035 337 / 7036 0.88 (0.75 1.02) 0.10 Combined (random) 911 / 29,312 1107 / 29,229 0.81 (0.73 0.91) <0.0001 These are not head-to-head comparisons between the NOACs and should therefore be interpreted with caution 0.5 1.0 2.0 Favours NOACFavours warfarin Data are n/n, unless otherwise indicated. Heterogeneity: I 2 47%, p=0.13. *Dabigatran 150 mg BD; Rivaroxaban 20 mg OD; Apixaban 5 mg BD; Edoxaban 60 mg OD. BD: 1. twice Ruff daily; et al. Lancet CI: confidence 2014;383:955 962. interval; RR: risk ratio; SE: systemic embolism. Created from Ruff et al. 2014 1 2

Meta-analysis: safety of NOACs vs warfarin for stroke prevention in patients with NVAF 1 Major bleeding Study NOAC (events) Warfarin (events) RR (95% CI) RR (95% CI) p value RE-LY* 375 / 6076 397 / 6022 0.94 (0.82 1.07) 0.34 ROCKET AF 395 / 7111 386 / 7125 1.03 (0.90 1.18) 0.72 ARISTOTLE 327 / 9088 462 / 9052 0.71 (0.61 0.81) <0.0001 ENGAGE AF-TIMI 48 444 / 7012 557 / 7012 0.80 (0.71 0.90) 0.0002 Combined (random) 1514 / 29,287 1802 / 29,211 0.86 (0.73 1.00) 0.06 These are not head-to-head comparisons between the NOACs and should therefore be interpreted with caution 0.5 1.0 2.0 Favours NOAC Favours warfarin Data are n/n, unless otherwise indicated. Heterogeneity: I 2 83%, p=0.001. *Dabigatran 150 mg BD; Rivaroxaban 20 mg OD; Apixaban 5 mg BD; Edoxaban 60 mg OD. BD: 1. twice Ruff et daily; al. Lancet CI: confidence 2014;383:955 962. interval; RR: risk ratio; SE: systemic embolism. Created from Ruff et al. 2014 1 ASA is still widely used for stroke prevention in NVAF First cohort of GARFIELD registry: Antiplatelets as monotherapy used in approx. 25% of patients who should be treated with oral anticoagulation Stroke prevention therapy: CHA 2DS 2-VASc score 1. Kakkar et al. PLoS One. 2013;8:e64379. Low dose vs adjusted dose 3

Criteria for Low Dose Regimen Several well known and tested criteria for using the lowdose protocol Advanced age Low body weight Elevated creatinine (poor renal function) These were all validated in large randomized trials But only 5-10% of patients received the low dose regimen NOAC Prescription Data In practice, an unexpectedly high proportion of prescriptions for apixaban are for a reduced dose of 2.5 mg. Similar patterns are seen with rivaroxaban and dabigatran. Apixaban Rivaroxaban Dabigatran Q4 2014 Q4 2014 Q4 2014 Country 2.5mg 5mg 10mg 15mg 20mg 75mg 110mg 150mg UNITED STATES 24% 76% 6% 21% 73% 16% 0% 84% JAPAN 58% 42% 55% 45% 0% 40% 60% 0% GERMANY 41% 59% 4% 34% 61% 2% 61% 37% CANADA 38% 62% 6% 26% 68% 1% 52% 47% AUSTRALIA 39% 61% 2% 30% 68% 0% 63% 37% UNITED KINGDOM 42% 58% 6% 22% 71% 3% 51% 46% SPAIN 37% 63% 5% 33% 63% 3% 60% 38% FRANCE 46% 54% 0% 0% 0% 0% 0% 0% BELGIUM 30% 70% 2% 42% 56% 0% 60% 40% ITALY 35% 65% 2% 37% 61% 0% 63% 36% Data source: IMS MIDAS and CSD GERS(France) data. Why do so Many Clinicians choose the low dose regimen? Voting with their feet They believe that current criteria and guidelines do not reflect their patient population Their clinical experience and judgment guides them to use a lower dose Influence for leaders Poor understanding of the data 4

A B C D Criteria A = Age B = Low body weight C = Creatinine (poor renal function) D = Disease state Prior cerebral hemorrhage Frail state, a lot of falls Diabetes Dementia Disabled Disturbed (psychiatric illness) Low Dose Regimens: Efficacy & Safety Outcomes Dabigatran 110 mg & Edoxaban 30 mg Risk Ratio (95% CI) Stroke or SEE 1.03 (0.84-1.27) p=0.74 Ischemic Stroke 1.28 (1.02-1.60) p=0.045 Hemorrhagic Stroke 0.33 (0.23-0.46) p<0.0001 All-Cause Mortality 0.89 (0.83-0.96) p=0.003 MI 1.25 (1.04-1.50) p=0.019 Major Bleeding 0.65 (0.43-1.00) p=0.05 ICH 0.31 (0.24-0.41) p<0.0001 GI Bleeding 0.89 (0.57-1.37) p=0.58 N=26,107 0.2 0.5 1 2 Favors Low Dose NOAC Favors Warfarin Heterogeneity These data are from a pooled analysis P=NS for outcomes except: Edoxaban approval status and label may vary from country to country Major Bleeding, p=<0.001 GI Bleeding, p=0.01 Ruff CT, et al. Lancet 2014;383:955-62 Low-Dose is Less Effective Stroke or SSE (%/yr) 2.5% 2.0% 1.5% 1.0% 0.5% RE-LY: subjects are randomized to receive 110mg & 150mg Dabigatran 1.5% 1.1% 1.7% ENGAGE-AF: subjects are randomized to receive 30/15mg & 60/30mg Edoxaban 2.0% 1.8% 1.6% 0.0% Dabigatran 110mg Dabigatran 150mg VKA Edoxaban Edoxaban 30/15mg 60/30mg n 6015 6076 6022 7034 7035 7036 Major bleeding (%/year) 2.7 3.1 3.4 1.6 2.8 3.4 VKA There are no head-to-head studies comparing NOACs. Edoxaban approval status and label may vary from country to country. Connolly SJ, et al. N Engl J Med 2009;361:1139-51. Giugliano RP, et al. N Engl J Med 2013;369:2093-104. 5

Adjusted-Dose is not Low-Dose ARISTOTLE: use adjusted dose when meeting at least 2 of 3 of the following criteria Age >80 yrs, Cr >1.5 mg/dl, Wt <60 kg ROCKET-AF: use adjusted dose when CrCl 30 49 ml/min 4.0% 3.5% 3.3% Stroke or SSE (/yr) 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 1.7% 1.3% 1.5% 2.3% 2.8% 1.6% 2.0% 0.0% Apixban 2.5mg VKA Apixban VKA 5mg Xarelto 15mg VKA Xarelto 20mg VKA n 831 (5%) 17,370 2,950 (21%) 11,314 Major bleeding (%/year) 3.3% / 6.7% 2.1%/3.0% 4.5%/4.7% 3.4%/3.2% There are no head-to-head studies comparing NOACs. Fox KAA, et al. Eur Heart J 2011;32:2387-2394. Granger CB, et al. N Engl J Med 2011;365:981-92. Yao et al. J Am Heart Assoc. 2016;5:e003725. EFFECTIVENESS AND SAFETY OF DABIGATRAN, RIVAROXABAN, AND APIXABAN VERSUS WARFARIN IN NONVALVULARATRIAL FIBRILLATION An independent, retrospective US real-world analysis Objective: To assess the effectiveness and safety of apixaban, dabigatran, and rivaroxaban for stroke prevention in AF patients, in the largest US contemporary evaluation comparing NOACs and warfarin. Study endpoints Primary effectiveness outcome: stroke or systemic embolism, including ischaemic stroke, haemorrhagic stroke, and systemic embolism. Primary safety outcome: major bleeding, including gastrointestinal bleeding, intracranial bleeding, and bleeding from other sites. Yao et al. J Am Heart Assoc. 2016;5:e003725. 6

Effectiveness and Safety of Dabigatran, Rivaroxaban, and Apixaban Versus Warfarin in NonvalvularAtrial Fibrillation Three 1:1 propensity score matched cohorts Study population Use administrative claims data from OptumLabs Data Warehouse and Medicare Advantage Adult patients prescribed oral anticoagulants from 1 October 2010 30 June 2015* All patients were required to have at least 1 inpatient or outpatient AF diagnosis at either primary or secondary positions on the index date or at baseline apixaban vs. warfarin N=15,390 rivaroxaban vs. warfarin N=32,350 dabigatran vs. warfarin N=28,614 Apixaban N=2,402 Analysis Three matched cohorts were created (dabigatran vs. warfarin, rivaroxabanvs. warfarin, and apixaban vs. warfarin) using 1:1 propensity score matching Cox proportional hazards regression was used to compare outcomes in each of the propensity score matched cohorts, and to calculate HR Yao et al. J Am Heart Assoc. 2016;5:e003725. Baseline Characteristics in Propensity Score-Matched NOAC or Warfarin Users Apixaban (n=7695) Warfarin (n=7695) Dabigatran (n=14 307) Warfarin (n=14 307) Rivaroxaban (n=16 175) Warfarin (n=16 175) Age, yr Median (IQR) 73 (66 81) 73 (66 81) 70 (62 78) 70 (61 78) 72 (64 79) 72 (64 80) Female 46.9 46.8 39.7 40.4 43.2 43.7 Nonwhite race 20.2 20.4 18.9 19.3 19.9 20.4 Medical history Congestive heart failure 31.4 31.9 27.2 27.3 28.9 29.5 Hypertension 87.5 87.5 85.2 84.9 85.7 85.9 Diabetes mellitus 35.0 34.3 34.0 34.0 34.6 35.1 Stroke/TIA/SE 15.1 15.5 13.8 14.2 14.0 14.4 Vascular disease 28.3 28.4 23.1 23.4 26.9 27.5 Abnormal renal function 10.1 10.1 5.6 5.6 7.4 7.3 Bleeding history or predisposition 31.4 31.8 29.4 30.1 30.7 31.5 Yao et al. J Am Heart Assoc. 2016;5:e003725. IQR, interquartile range; NOAC, non-vitamin K antagonist oral anticoagulant; NSAID, nonsteroidal anti-inflammatory drug; SE, systemic stroke; TIA, transient ischemic attack; yr, year. Forest plot depicting the hazard ratio for each pairwise propensitymatched medication comparison for stroke and systemic embolism (S/SE), ischemic stroke, and hemorrhagic stroke Event Rate per 100 person-year Hazard Ratio (95% CI) P Value Apixaban vs Warfarin n=7695 n=7695 S/SE 1.33 1.66 0.67 (0.46 0.98) 0.04 Ischemic 1.03 1.05 0.83 (0.53 1.29) 0.40 Hemorrhagic 0.19 0.46 0.35 (0.14 0.88) 0.03 Dabigatran vs Warfarin n=14 307 n=14 307 S/SE 1.18 1.22 0.98 (0.76 1.26) 0.88 Ischemic 0.92 0.88 1.06 (0.79 1.42) 0.70 Hemorrhagic 0.16 0.29 0.56 (0.30 1.04) 0.07 Rivaroxaban vs Warfarin n=16 175 n=16 175 S/SE 1.26 1.29 0.93 (0.72 1.19) 0.56 Ischemic 0.95 0.88 1.01 (0.75 1.36) 0.95 Hemorrhagic 0.21 0.32 0.61 (0.35 1.07) 0.08 Yao et al. J Am Heart Assoc. 2016;5:e003725. Favor NOAC 1.0 Favor Warfarin CI, confidence interval; NOAC, non-vitamin K antagonist oral anticoagulant; S, stroke; SE, systemic embolism. 7

Forest plot depicting the hazard ratio for each pairwise propensity-matched medication comparison for major, intracranial, and gastrointestinal bleeding Event Rate per 100 person-year Hazard Ratio (95% CI) P Value Apixaban vs Warfarin n=7695 n=7695 Major Bleeding 2.33 4.46 0.45 (0.34 0.59) <0.001 Intracranial 0.29 1.06 0.24 (0.12 0.50) <0.001 Gastrointestinal 1.78 3.04 0.51 (0.37 0.70) <0.001 Dabigatran vs Warfarin 307 n=14 n=14 307 Major Bleeding 2.37 3.03 0.79 (0.67 0.94) <0.01 Intracranial 0.28 0.79 0.36 (0.23 0.56) <0.001 Gastrointestinal 1.97 1.95 1.03 (0.84 1.26) 0.78 Rivaroxaban vs Warfarin 175 n=16 n=16 175 Major Bleeding 4.04 3.64 1.04 (0.90 1.20) 0.60 Intracranial 0.44 0.79 0.51 (0.35 0.75) <0.001 Gastrointestinal 3.26 2.53 1.21 (1.02 1.43) 0.03 Yao et al. J Am Heart Assoc. 2016;5:e003725. Favor NOAC 1.0 Favor Warfarin CI, confidence interval; NOAC, non-vitamin K antagonist oral anticoagulant. Never started on it Reasons for Lack of Use of Anticoagulation Lack of knowledge No need for Rx Concerns about bleeding Started on AC-- then stopped Procedure Lack of funds Never restarted Possible Solutions Have EMRs programmed to trigger a BPA for patients with Afib on their problem list who qualify for AC Whenever AC gets stopped for a procedure, order a restart date (with proper controls in case of bleeding) Enroll patients in a registry or pharmacy tracking service to track AC refills Educate patients and family about need for staying on medications 8

Conclusions Compared to Warfarin, the NOACs are easier to use and safer for stroke prevention in patients with NVAF Too many patients are either on no anticoagulant therapy or just taking ASA There is more use of lower dose NOACs in practice than directed by the evidence and guidelines Anticoagulant efficacy was dose related, patient should be treated with the most effective doses and underdosing resulted in potentially preventable strokes But low-dose therapy is better than no therapy Using RWD is the new trend and tool for making drug-to-drug comparisons when clinical trial do not exist Alexandra et al. JAMA Cardiol. doi:10.1001/jamacardio.2016.1829 New Alteplase Guidelines Stroke, Feb, 2016 Use of IV Alteplase in patients taking NOACs Demaerschalk et al., Stroke, Feb 2016 9

Situations where anticoagulant reversal would be needed Emergency management: Need for Emergency Surgery Uncontrolled Bleeding (Brain, GI) Measures to overcome/reverse the anticoagulant effect of Warfarin and NOACs are available but have limitations Stop medication/wait, maintain diuresis Restore physiologic hemostasis Takes time Sufficient renal function needed Clotting factor concentrates (PCC, apcc, rfviia) Replenish coagulation factors (PCC, apcc) Trigger activation of clotting cascade (rfviia, apcc) Optimal dosing not established Potentially prothrombotic Dialysis Accelerates elimination of dabigatran Difficult if hemodynamically compromised Expertise, infrastructure, and time needed apcc, activated PCC; rfviia, recombinant activated Factor VII; PCC, prothrombin complex concentrate Pradaxa : EU SPC, 2016; Tran et al. Intern Med J 2014; Weitz et al. Circulation 2012; Akwaa & Spyropoulos. Curr Treat Options Cardiovasc Med 2013; Alikhan et al. Emerg Med J 2014 Coagulation Cascade 10

NOAC reversal agents: characteristics Idarucizumab 1 NOAC reversal agent Andexanet alfa (PRT064445) 1 Target Dabigatran FXa inhibitors Mechanism of action Regulatory approval status Humanized Fab: specifically binds dabigatran with high affinity 2 Recombinant modified FXa: competitive affinity for direct FXa inhibitors 6 Submitted to FDA; Approved by FDA, EMA, 7 and Health Canada 3 5 not currently approved in any country Idarucizumab is not approved in all countries. Please check your local prescribing information for details. Andexanet alfa is not approved in any country. Fab, fragment antigen binding; FXa, activated Factor X; 1. Greinacher et al. Thromb Haemost 2015; 2. Schiele et al. Blood 2013; 3. US FDA Press Release 16 Oct 2015; 4. European Commission Community Register of Medicinal Products for Human Use 20 November 2015; 5. Health Canada Regulatory Decision Summary 29 Apr 2016; 6. Lu et al. Nat Med 2013; 7. Portola Pharmaceuticals, Inc. Press Release 18 Dec 2015 No Specific Reversal Agent Is Approved to Manage FXa Inhibitor-Associated Bleeds Class Anticoagulant Reversal Agent Vitamin K Antagonist Direct Thrombin Inhibitor Indirect FXa Inhibitor Direct FXa Inhibitor Coumadin (warfarin) Pradaxa (dabigatran) Lovenox (enoxaparin) Xarelto (rivaroxaban) Eliquis (apixaban) Savaysa (edoxaban) Kcentra (4 Factor PCC) Vitamin K Praxbind (idarucizumab) Protamine No specific Reversal agent currently available 1. Kcentra [prescribing information]. Kankakee, IL: CSL Behring LLC. 2. Praxbind [prescribing information.] Ridgefield, CT. Boehringer Ingelheim Pharm aceuticals, Inc; 2015. 3. Weitz JI, Pollack CV. Thromb Haemost. 2015;114:1113-1126. 4. Mo Y, Yam FK. Pharmacotherapy. 2015;35:198-207. 3 2 Idarucizumab was designed as a specific reversal agent for the anticoagulant activity of dabigatran Thrombin Dabigatran Idarucizumab Humanized Fab fragment IV administration, immediate onset of action Binding affinity for dabigatran ~350 higher than dabigatran to thrombin No intrinsic procoagulant or anticoagulant activity Short half-life (initial t 1/2 ~45 min) Idarucizumab is not approved in all countries. Please check your local prescribing information for details. Adapted from Schiele et al. Blood 2013; Stangier et al. ISTH 2015; Pollack et al. Thromb Haemost 2015 11

Idarucizumab is being investigated in RE-VERSE AD : an ongoing, multicenter, single-arm, open-label, Phase III study Group A: Uncontrolled bleeding + dabigatran-treated Group B: Emergency surgery or procedure* + dabigatran-treated 2 2.5 g idarucizumab 0 15 minutes 90 days follow-up 2 2.5 g idarucizumab 0 15 minutes 90 days follow-up N~500 Hospital arrival 4 hrs Multiple blood samples 2 2.5 g idarucizumab: overwhelming dose of idarucizumab selected Primary endpoint: dabigatran reversal within 4 hrs (dtt or ECT) Idarucizumab is not approved in all countries. Please check your local prescribing information for details. *Other than for bleeding Pollack et al. Thromb Haemost 2015; Eikelboom et al. Circulation 2015 RE-VERSE AD interim analysis: baseline characteristics of the first 90 patients (1) Group A (n=51) Group B (n=39) Total (N=90) Male, n (%) 32 (63) 18 (46) 50 (56) Age, yrs,median (min, max) 77.0 (48, 93) 76.0 (56, 93) 76.5 (48, 93) CrCl (Cockcroft Gault), ml/min Median (min, max) 54 (16, 187) 60 (11, 171) 58 (11, 187) <30 5 7 12 30 <50 14 6 20 50 <80 16 11 27 80 6 9 15 Missing 10 6 16 Elevated dtt at baseline, n (%) 40 (78) 28 (72) 68 (76) Elevated ECT at baseline, n (%) 47 (92) 34 (87) 81 (90) Idarucizumab is not approved in all countries. Please check your local prescribing information for details. CrCl, creatinine clearance; dtt, diluted thrombin time; ECT, ecarin clotting time Pollack et al. N Engl J Med 2015 Patients were enrolled due to major bleeding events and a variety of emergency procedures Type of bleeding Group A (n=51) Intracranial 18 Trauma 9 Gastrointestinal 20 Other* 11 * Other bleeding types: urogenital, epistaxis, liver, aortic aneurism, and aortic dissection Reason for surgery Group B (n=39) Aortic dissection 1 Pericardial tamponade 1 Peritonitis 1 Acute mesenteric ischemia with sepsis 2 Bone fractures 8 Acute cholecystitis 5 Acute renal insufficiency, catheter placement 4 Acute appendicitis 3 Joint/wound infection 3 Abscess (suprapubic, scrotal) 2 Idarucizumab is not approved in all countries. Please check your local prescribing information for details. Pollack et al. N Engl J Med 2015 12

Group A interim results: reversal of dabigatran anticoagulation with idarucizumab based on dtt Group A: Uncontrolled bleeding 130 120 Primary endpoint: dabigatran reversal within 4 hours (dtt or ECT) Anticoagulation activity (dtt) 110 100 90 80 70 60 50 40 30 Idarucizumab 2 2.5 g Reversal sustained over 12 hours in 90% of patients No idarucizumab-related prothrombotic effects identified to date Assay upper limit of normal 20 0 Baseline Between 10 30 1 hrs 2 hrs 4 hrs 12 hrs 24 hrs vials min Time post-idarucizumab Idarucizumab is not approved in all countries. Please check your local prescribing information for details. Interim analysis includes data for the first 90 patients. dtt, diluted thrombin time Adapted from Pollack et al. N Engl J Med 2015 Group B interim results: reversal of dabigatran anticoagulation with idarucizumab based on dtt Group B: Emergency surgery or procedure 130 120 Primary end point: dabigatran reversal within 4 hours (dtt or ECT) Anticoagulation activity (dtt) 110 100 90 80 70 60 50 40 30 Idarucizumab 2 2.5 g Reversal sustained over 12 hours in 81% of patients Normal intraoperative hemostasis achieved in 92% who underwent procedures Assay upper limit of normal 20 0 Baseline Between 10 30 1 hrs 2 hrs 4 hrs 12 hrs 24 hrs vials min Time post-idarucizumab Idarucizumab is not approved in all countries. Please check your local prescribing information for details. Interim analysis includes data for the first 90 patients. dtt, diluted thrombin time Adapted from Pollack et al. N Engl J Med 2015 Mechanism of Action for Andexanet Alpha Shah et al., AMSRJ, Spring 2014AMSRJ 13

ANNEXA-A and R (Healthy Volunteers, Age 50-75) In Healthy Volunteers, Andexanet Alfa Rapidly Reduced Anti-FXa Activity and Restored Thrombin Generation Inhibited by Apixaban 1,2 Apixaban-Treated Rivaroxaban Patients N=80 N=65 IN PATIENTS WHO RECEIVED BOLUS + INFUSION Reduction in Anti-FXa Activity Restoration of Thrombin Generation End of Bolus End of Infusion End of Bolus End of Infusion Anti-fXa (ng/ml) (%) 400 125 300 100 75 200 50 100 25 Placebo (n=13) (n=8) 800 mg bolus 960 mg x 2hr infusion (n=26) 400 mg bolus + 480 mg x 2hr infusion (n=23) ETP (nm.min) 2500 2000 1500 1000 500 Placebo (n=13) (n=8) 800 400 mg bolus + 960 480 mg x 2hr infusion (n=26) (n=23) Pre-anticoagulant Mean Baseline ± 1SD range 0 0.0 0.2 0.4 0.6 1 2 3 4 5 6 7 8 9 101112 0.0 0.2 0.4 0.6 1 2 3 4 5 6 7 8 9 101112 Time after bolus (hr) Time after bolus (hr) 0 Baseline 0.0 0.2 0.4 0.6 2 4 6 8 10121416182022 Time after bolus (hr) 1. Siegal DM et al. N Engl J Med. 2015;373:2413-2424. 2. Crowther M et al. Presented at ISTH 2015. 40 ANNEXA-A and R (Healthy Volunteers, Age 50-75) In Healthy Volunteers, Andexanet Alfa Rapidly Reduced Anti-FXa Activity and Restored Thrombin Generation Inhibited by Apixaban 1,2 Rivaroxaban Patients N=80 IN PATIENTS WHO RECEIVED BOLUS + INFUSION Reduction in Anti-FXa Activity Restoration of Thrombin Generation End of Bolus End of Infusion End of Bolus End of Infusion 400 Placebo (n=13) 2500 Placebo (n=13) Anti-fXa (ng/ml) 300 200 100 800 mg bolus + 960 mg x 2hr infusion (n=26) ETP (nm.min) 2000 1500 1000 800 mg bolus + 960 mg x 2hr infusion (n=26) Pre-anticoagulant Mean ± 1SD 500 0 0.0 0.2 0.4 0.6 1 2 3 4 5 6 7 8 9 101112 Time after bolus (hr) 0 Baseline 0.0 0.2 0.4 0.6 2 4 6 8 10121416182022 Time after bolus (hr) 1. Siegal DM et al. N Engl J Med. 2015;373:2413-2424. 2. Crowther M et al. Presented at ISTH 2015. 41 ANNEXA-R (Healthy Volunteers, Age 50-75) Overall Safety of Andexanet in Healthy Subjects Adverse Events No serious or severe adverse events reported in any subject Transient increases in D-dimer and F1+2 were observed in a subset of subjects, that generally returned to the normal range within 24-72 hours No thrombotic events No antibodies to Factor X or Factor Xa No neutralizing antibodies to andexanet Siegal DM et al. N Engl J Med. 2015;373:2413-2424. 42 14

ANNEXA-4 (Patients Taking FXa Inhibitors with Acute Major Bleeding) Andexanet Phase 3b Interim Analysis in Bleeding Patients Study Overview & Design Multicenter, prospective, open-label single-group study Patient Screening Andexanet Treatment Follow-Up Acute Major Bleeding FXa Inhibitor Within Last 18 hours IV Bolus 2-Hour IV Infusion After End of Infusion Study Assessments: Baseline End End 4 hr 8 hr 12 hr of Bolus of Infusion Day 1 Day 3 Day 30 Coprimary Outcomes % change in anti-fxa activity Rate of excellent or good hemostatic efficacy at 12 hours post-andexanet infusion Results presented are from a descriptive analysis of a preliminary cohort of 67 patients from an ongoing study Connolly SJ et al. N Eng J Med. 2016 vol375. Annexa 4 study population Connolly SJ et al. N Eng J Med. 2016 vol375. Annexa 4 Study Bleeding Site and Outcomes Bleeding Site Safety Population N = 67 Efficacy Population N = 47 GI 49% 53% Intracranial ICH 28/67 14/28 20/47 12/20 Subdural 11/28 7/20 SAH 3/28 1/20 Death 15% 15% Thromboembolic event 18% 15% Safety population = patient received any dose of Andexanet Efficacy population = anti FXa activity and qualifying bleed Connolly SJ et al. N Eng J Med. 2016 vol375. 15

ANNEXA-4 (Patients Taking FXa Inhibitors with Acute Major Bleeding) Hemostatic Efficacy in All Patients and by Bleed Site All Patients 79% All patients (n=47) 95% CI (64-89) 84% Gastrointestinal (n=25) 95% CI (64-96) Bleed Site 80% Intracranial (n=20) 95% CI (56-94) 0% Other (n=2) 0 25 50 75 100 Excellent or Good Hemostatic Efficacy (%) Siegal DM et al. N Engl J Med. 2015;373:2413-2424. 46 Bleeding Control Connolly SJ et al. N Eng J Med. 2016 vol375. 16