PREVAIL: 5-Year Outcomes From a Randomized Trial of Left Atrial Appendage Closure vs Medical Therapy in Patients With Nonvalvular Atrial Fibrillation

PREVAIL: 5-Year Outcomes From a Randomized Trial of Left Atrial Appendage Closure vs Medical Therapy in Patients With Nonvalvular Atrial Fibrillation Dr. Saibal Kar, MD TCT 2017 Late Breaking Clinical Trial Denver, CO USA Authors: V Reddy, S Doshi; S Kar, D Gibson, M Price, K Huber, R Horton, M Buchbinder, P Neuzil, N Gordon, DR Holmes Jr on behalf of the PREVAIL and PROTECT-AF Investigators

Dr. Kar Disclosures Has received research grants from and served as a consultant for Abbott Vascular and Boston Scientific; Has served as a member of the advisory board for left atrial appendage closure; Is the national principal investigator of the Continuous Access Registries (CAP & CAP2); Has served as a proctor for Boston Scientific.

Background PROTECT AF and PREVAIL were RCTs comparing left atrial appendage closure (LAAC) with WATCHMAN to warfarin PROTECT-AF demonstrated similar stroke reduction to warfarin (at a mean of 3.8 years follow up) JAMA 312:1988-98 (2014) PREVAIL produced inconclusive findings due to a warfarin cohort with an implausibly low ischemic stroke rate, relatively few patients and relatively short follow up (~10 months) JACC 64:1-12 (2014) Both protocols specified 5 years of follow-up for all patients 5 year follow up completed in 2013 (final) 5 year follow up completed in 2017 (final)

Objectives To report the final, 5 year results of PREVAIL, both alone and as part of a patient-level meta-analysis with PROTECT-AF final 5 year data

Methods PREVAIL 1 st Primary efficacy: Comparison of rate ratios of 18- month event rates; Upper CI 1.75 for NI 2 nd primary efficacy: 1-tailed test, either the ratio or the difference between rates in the two arms satisfied the non-inferiority criteria, using 95% upper credible intervals (CrI) <2.0 and <0.0275, respectively, and posterior probability for non-inferiority 97.5%. No adjustment was made for multiple comparisons All analyses ITT; rates are events per 100 patient-years (indicated for simplicity by %).

Methods Patient-Level Meta-Analysis ITT: censoring data from patients without events at the time of the last known status Disabling Strokes: increase in the Modified Rankin Score (MRS) by at least 2 points Comparison of Event Rates: Cox proportional hazards model with confidence intervals (CIs) Stratified by study to account for differences in risk profiles Kaplan-Meier curves used for graphical assessment of timedependent events Frequentist statistics and 2-sided p-values nominally significant at p < 0.05, without adjustment for multiple comparisons

Demographics: Combined Cohort Characteristic Device N=732 Control N=382 p-value Age, years 72.6±8.4 73.5±8.6 0.09 Sex, male 69.4 71.7 0.42 CHADS 2 Score 2.3±1.1 2.4±1.2 0.06 CHA 2 DS 2 -Vasc Score 3.6 ± 1.4 3.9 ± 1.5 0.02 Risk Factors CHF 25.5 25.7 0.97 Hypertension 89.2 92.7 0.06 Age 75 years 40.4 43.2 0.38 Diabetes 27.9 29.6 0.55 Prior Stroke/TIA 22.1 23.6 0.59 AF Pattern Paroxysmal 45.2 44.5 0.82 Persistent 24.9 23.3 0.56 Permanent 27.6 30.1 0.38 Unknown 1.4 0.8 0.56 Paced 1.0 1.3 0.56

Demographics: Combined Cohort Characteristic Device N=732 Control N=382 p-value Age, years 72.6±8.4 73.5±8.6 0.09 Sex, male 69.4 71.7 0.42 CHADS 2 Score 2.3±1.1 2.4±1.2 0.06 CHA 2 DS 2 -Vasc Score 3.6 ± 1.4 3.9 ± 1.5 0.02 Risk Factors CHF 25.5 25.7 0.97 Hypertension 89.2 92.7 0.06 Age 75 years 40.4 43.2 0.38 Diabetes 27.9 29.6 0.55 Prior Stroke/TIA 22.1 23.6 0.59 AF Pattern Paroxysmal 45.2 44.5 0.82 Persistent 24.9 23.3 0.56 Permanent 27.6 30.1 0.38 Unknown 1.4 0.8 0.56 Paced 1.0 1.3 0.56

Follow-up Trial Number of Patients Device Control Mean Follow-up (months) Total Patient Years PROTECT AF 463 244 47.6 ± 21.3 2,717 PREVAIL-only* 269 138 47.9 ± 19.4 1,626 Total 732 382 4,343 * Does not include PROTECT AF informative prior from Bayesian model

RESULTS

Results PREVAIL Efficacy Dataset Device 18-Month Rate Control 18-Month Rate 18-Month Rate Ratio (95% CrI) First Primary Endpoint (ITT) - stroke, SE, and CV/Unexplained Death Primary analysis 1.07 0.064 0.063 (Jan 2013) (0.57, 1.89) 1 st Post Hoc Analysis 1.21 0.065 0.057 (Jun 2014) (0.69, 2.05) Final 5-Year Analysis 1.33 0.066 0.051 (Sep 2017) (0.78, 2.13) Dataset Device 18-Month Rate Control 18-Month Rate 18-Month Rate Ratio (95% CrI) Criteria Met? 95% CrI Upper Bound < 1.75 Posterior Probability Noninferiority (P NI 97.5%) No 93% No 93% No 88.4% Criteria Met? 95% CI Upper Bound < 2.0 18-Month Rate Difference (95% CrI) Second Primary Endpoint (ITT) ischemic stroke or SE >7 days Must meet at least one criteria Primary analysis 1.6 0.0053 0.0253 0.0200 No (Jan 2013) (0.5, 4.2) (-0.0190, 0.0273) 1 st Post Hoc Analysis 2.8 0.0163 0.0294 0.0131 No (Jun 2014) (0.9, 7.3) (-0.0023, 0.0342) Final 5-Year Analysis 2.2 0.0120 0.0255 0.0135 No (Sep 2017) (0.8, 4.9) (-0.0036, 0.0275) * Upper CrI is displayed as 0.0275, but if all significant digits were displayed, it is less than upper bound of 0.0275 Criteria Met? 95% CrI Upper Bound < 0.0275 Posterior Probability Noninferiority (P NI 97.5%) Yes 97.6 No 89.2 Yes 97.5%

Results PREVAIL 2 nd Endpoint Met Dataset Device 18-Month Rate Control 18-Month Rate 18-Month Rate Ratio (95% CrI) First Primary Endpoint (ITT) - stroke, SE, and CV/Unexplained Death Primary analysis 1.07 0.064 0.063 (Jan 2013) (0.57, 1.89) 1 st Post Hoc Analysis 1.21 0.065 0.057 (Jun 2014) (0.69, 2.05) Final 5-Year Analysis 1.33 0.066 0.051 (Sep 2017) (0.78, 2.13) Dataset Device 18-Month Rate Control 18-Month Rate 18-Month Rate Ratio (95% CrI) Criteria Met? 95% CrI Upper Bound < 1.75 Posterior Probability Noninferiority (P NI 97.5%) No 93% No 93% No 88.4% Criteria Met? 95% CI Upper Bound < 2.0 18-Month Rate Difference (95% CrI) Second Primary Endpoint (ITT) ischemic stroke or SE >7 days Must meet at least one criteria Primary analysis 1.6 0.0053 0.0253 0.0200 No (Jan 2013) (0.5, 4.2) (-0.0190, 0.0273) 1 st Post Hoc Analysis 2.8 0.0163 0.0294 0.0131 No (Jun 2014) (0.9, 7.3) (-0.0023, 0.0342) Final 5-Year Analysis 2.2 0.0120 0.0255 0.0135 No (Sep 2017) (0.8, 4.9) (-0.0036, 0.0275) * Upper CrI is displayed as 0.0275, but if all significant digits were displayed, it is less than upper bound of 0.0275 Criteria Met? 95% CrI Upper Bound < 0.0275 Posterior Probability Noninferiority (P NI 97.5%) Yes 97.6 No 89.2 Yes 97.5%

Results PROTECT AF and PREVAIL Event Rates PROTECT-AF Subjects PREVAIL Subjects Primary Efficacy: Stroke/SE/CV Death Device (n=463) No. of Events Rate * Control (n=244) No. of Events Rate * p-value 2:1 Randomization Device (n=269) No. of Events Rate * Control (n=138) No. of Events Rate * p-value 40 / 1787.7 2.24 34 / 929.4 3.66 0.04 37 / 1038.3 3.65 15 / 530.4 2.94 0.47 All Stroke 26 / 1781.7 1.46 20 / 929.4 2.15 0.23 19 / 1042.4 1.97 7 / 530.4 1.29 0.32 Ischemic Stroke 24 / 1781.7 1.35 10 / 932.8 1.07 0.49 17 / 1043.1 1.68 4 / 533.3 0.73 0.13 Hemorrhagic Stroke 3 / 1837.7 0.16 10 / 945.6 1.06 0.005 2 / 1084.6 0.18 3 / 538.0 0.54 0.23 Systemic Embolism 3 / 1837.1 0.16 0 n/a n/a 1 / 1080.6 0.09 0 / 540.9 n/a n/a CV/Unexplained Death 19 / 1843.2 1.03 22 / 948.9 2.32 0.009 18 / 1084.7 1.79 10 / 540.9 1.98 0.76 * Events are per 100 patient-years

PROTECT AF Results WATCHMAN Superior Efficacy, Hemorrhagic Stroke, CV Death PROTECT-AF Subjects PREVAIL Subjects Primary Efficacy: Stroke/SE/CV Death Device (n=463) No. of Events Rate * Control (n=244) No. of Events Rate * p-value 2:1 Randomization Device (n=269) No. of Events Rate * Control (n=138) No. of Events Rate * p-value 40 / 1787.7 2.24 34 / 929.4 3.66 0.04 37 / 1038.3 3.65 15 / 530.4 2.94 0.47 All Stroke 26 / 1781.7 1.46 20 / 929.4 2.15 0.23 19 / 1042.4 1.97 7 / 530.4 1.29 0.32 Ischemic Stroke 24 / 1781.7 1.35 10 / 932.8 1.07 0.49 17 / 1043.1 1.68 4 / 533.3 0.73 0.13 Hemorrhagic Stroke 3 / 1837.7 0.16 10 / 945.6 1.06 0.005 2 / 1084.6 0.18 3 / 538.0 0.54 0.23 Systemic Embolism 3 / 1837.1 0.16 0 n/a n/a 1 / 1080.6 0.09 0 / 540.9 n/a n/a CV/Unexplained Death 19 / 1843.2 1.03 22 / 948.9 2.32 0.009 18 / 1084.7 1.79 10 / 540.9 1.98 0.76 * Events are per 100 patient-years

PREVAIL Results WATCHMAN Comparable For Efficacy Control Group Continued To Overperform PROTECT-AF Subjects PREVAIL Subjects Primary Efficacy: Stroke/SE/CV Death Device (n=463) No. of Events Rate * Control (n=244) No. of Events Rate * p-value 2:1 Randomization Device (n=269) No. of Events Rate * Control (n=138) No. of Events Rate * p-value 40 / 1787.7 2.24 34 / 929.4 3.66 0.04 37 / 1038.3 3.65 15 / 530.4 2.94 0.47 All Stroke 26 / 1781.7 1.46 20 / 929.4 2.15 0.23 19 / 1042.4 1.97 7 / 530.4 1.29 0.32 Ischemic Stroke 24 / 1781.7 1.35 10 / 932.8 1.07 0.49 17 / 1043.1 1.68 4 / 533.3 0.73 0.13 Hemorrhagic Stroke 3 / 1837.7 0.16 10 / 945.6 1.06 0.005 2 / 1084.6 0.18 3 / 538.0 0.54 0.23 Systemic Embolism 3 / 1837.1 0.16 0 n/a n/a 1 / 1080.6 0.09 0 / 540.9 n/a n/a CV/Unexplained Death 19 / 1843.2 1.03 22 / 948.9 2.32 0.009 18 / 1084.7 1.79 10 / 540.9 1.98 0.76 * Events are per 100 patient-years 2:1 randomization Control Group continues to overperform Rate = 0.7%

Results WATCHMAN Comparable to Warfarin for Ischemic Stroke 10 8 Untreated AF Treated with Warfarin WATCHMAN Arm Ischemic Stroke Risk (events per 100 pt-yrs) 6 4 2 0 PROTECT AF 1.3 WASP 1.5 Baseline CHA 2 DS 2 -VASc Score CAP 1.3 CAP2 PREVAIL 2.3 1.7 EWOLUTION 1.1 1 2 3 4 5 EWOLUTION: Boersma Lva et al Heart Rhythm 2017;doi-10.1016/j.hrthm.2017.05.038; WASP: hilips K, et al.journal of Arrhythmia (in press).

Patient-Level Meta-Analysis PROTECT AF and PREVAIL 5 years HR p- value Efficacy 0.82 0.3 All stroke or SE 0.96 0.9 Ischemic stroke or SE 1.7 0.08 Hemorrhagic stroke 0.2 0.0022 Ischemic stroke or SE >7 days 1.4 0.3 CV/unexplained death 0.59 0.03 All-cause death 0.73 0.04 Major bleed, all 0.91 0.6 Major bleeding, non procedure-related 0.48 0.0003 Favors WATCHMAN Favors warfarin 0.01 0.1 1 10 Hazard Ratio (95% CI)

Patient-Level Meta-Analysis WATCHMAN Comparable To Warfarin For Ischemic Stroke HR p- value Efficacy 0.82 0.3 All stroke or SE 0.96 0.9 Ischemic stroke or SE 1.7 0.08 Hemorrhagic stroke 0.2 0.0022 Ischemic stroke or SE >7 days 1.4 0.3 CV/unexplained death 0.59 0.03 All-cause death 0.73 0.04 Major bleed, all 0.91 0.6 Major bleeding, non procedure-related 0.48 0.0003 Favors WATCHMAN Favors warfarin 0.01 0.1 1 10 Hazard Ratio (95% CI)

Patient-Level Meta-Analysis WATCHMAN Superior for Hemorrhagic Stroke, CV Death, All-Cause Death, Post-procedure Bleeding HR p- value Efficacy 0.82 0.3 All stroke or SE 0.96 0.9 Ischemic stroke or SE 1.7 0.08 Hemorrhagic stroke 0.2 0.0022 Ischemic stroke or SE >7 days 1.4 0.3 CV/unexplained death 0.59 0.03 All-cause death 0.73 0.04 Major bleed, all 0.91 0.6 Major bleeding, non procedure-related 0.48 0.0003 Favors WATCHMAN Favors warfarin 0.01 0.1 1 10 Hazard Ratio (95% CI)

PROTECT-AF / PREVAIL Combined 5-Year Kaplan-Meier Curves of the Major Efficacy Endpoints

Patient-Level Meta-Analysis WATCHMAN Superior Reduction in Disabling Strokes 2.00% 1.50% Disabling/Fatal Strokes Non-Disabling Strokes 1.00% 0.50% 0.00% WATCHMAN warfarin Disabling Stroke defined as MRS 2

Patient-Level Meta-Analysis WATCHMAN Superior Reduction in Disabling Strokes 2.00% 1.50% Disabling/Fatal Strokes Non-Disabling Strokes 1.00% 0.50% 59% Reduction 0.00% WATCHMAN HR 0.41 (0.19 0.90) P=0.027 Disabling Stroke defined as MRS 2 warfarin

Patient-Level Meta-Analysis No Significant Difference In Outcomes By Patient Subset Subgroup Age Sex CHADS 2 Score CHA 2 DS 2 - VASc Score HAS-BLED History of TIA Stroke < 75 > 75 Female Male 3 >3 4 >4 2 >2 No Yes Favors WATCHMAN Favors warfarin Interaction p-value 0.98 0.94 0.47 0.13 0.054 0.35 0.1 1.0 10.0 Hazard Ratio

Patient-Level Meta-Analysis No Significant Difference In Outcomes By Patient Subset Subgroup Age Sex CHADS 2 Score CHA 2 DS 2 - VASc Score HAS-BLED History of TIA Stroke < 75 > 75 Female Male 3 >3 4 >4 2 >2 No Yes Favors WATCHMAN Favors warfarin Interaction p-value 0.98 0.94 0.47 0.13 0.054 0.35 0.1 1.0 10.0 Hazard Ratio

Summary PREVAIL 5 year follow-up demonstrates: 2 nd primary endpoint meets non-inferiority while the 1 st endpoint remains unchanged No significant differences between WATCHMAN and warfarin for primary efficacy measures despite an implausibly low rate of ischemic stroke (0.73%) in the control arm Meta-Analysis of PROTECT AF and PREVAIL with 5 year follow-up demonstrates: Comparable efficacy and stroke rates, with no significant differnce across subgroups No significant differences in ischemic stroke rates versus warfarin Significant, superior reductions in disabling strokes, nonprocedural bleeding, and mortality *Kaplan Meier Rate

Conclusion Long term 5-year outcomes of 2 RCTs demonstrate LAAC with the Watchman device provides stroke prevention in NVAF patients to a similar degree as oral anticoagulation By minimizing major bleeding, particularly hemorrhagic stroke, LAAC results in less disability or death than warfarin For patients who are poor candidates for long-term oral anticoagulation, left atrial appendage closure is a reasonable strategy for stroke prophylaxis

Conclusion Long term 5-year outcomes of 2 RCTs demonstrate LAAC with the Watchman device provides stroke prevention in NVAF patients to a similar degree as oral anticoagulation By minimizing major bleeding, particularly hemorrhagic stroke, LAAC results in less disability or death than warfarin WE HAVE PREVAILED! For patients who are poor candidates for long-term oral anticoagulation, left atrial appendage closure is a reasonable strategy for stroke prophylaxis

ABBREVIATED STATEMENT (US) WATCHMAN TM Left Atrial Appendage Closure Device with Delivery System and WATCHMAN Access System INDICATIONS FOR USE The WATCHMAN Device is indicated to reduce the risk of thromboembolism from the left atrial appendage in patients with non-valvular atrial fibrillation who: Are at increased risk for stroke and systemic embolism based on CHADS 2 or CHA 2 DS 2 -VASc scores and are recommended for anticoagulation therapy; Are deemed by their physicians to be suitable for warfarin; and Have an appropriate rationale to seek a non-pharmacologic alternative to warfarin, taking into account the safety and effectiveness of the device compared to warfarin. The WATCHMAN Access System is intended to provide vascular and transseptal access for all WATCHMAN Left Atrial Appendage Closure Devices with Delivery Systems. CONTRAINDICATIONS Do not use the WATCHMAN Device if: Intracardiac thrombus is visualized by echocardiographic imaging. An atrial septal defect repair or closure device or a patent foramen ovale repair or closure device is present. The LAA anatomy will not accommodate a device. See Table 46 in the DFU. Any of the customary contraindications for other percutaneous catheterization procedures (e.g., patient size too small to accommodate TEE probe or required catheters) or conditions (e.g., active infection, bleeding disorder) are present. There are contraindications to the use of warfarin, aspirin, or clopidogrel. The patient has a known hypersensitivity to any portion of the device material or the individual components (see Device Description section) such that the use of the WATCHMAN Device is contraindicated. WARNINGS Device selection should be based on accurate LAA measurements obtained using fluoro and ultrasound guidance (TEE recommended) in multiple angles (e.g., 0º, 45º, 90º, 135º). Do not release the WATCHMAN Device from the core wire if the device does not meet all release criteria. If thrombus is observed on the device, warfarin therapy is recommended until resolution of thrombus is demonstrated by TEE. The potential for device embolization exists with cardioversion <30 days following device implantation. Verify device position post-cardioversion during this period. Administer appropriate endocarditis prophylaxis for 6 months following device implantation. The decision to continue endocarditis prophylaxis beyond 6 months is at physician discretion. For single use only. Do not reuse, reprocess, or resterilize. PRECAUTIONS The safety and effectiveness (and benefit-risk profile) of the WATCHMAN Device has not been established in patients for whom long-term anticoagulation is determined to be contraindicated. The LAA is a thin-walled structure. Use caution when accessing the LAA and deploying the device. Use caution when introducing the WATCHMAN Access System to prevent damage to cardiac structures. Use caution when introducing the Delivery System to prevent damage to cardiac structures. To prevent damage to the Delivery Catheter or device, do not allow the WATCHMAN Device to protrude beyond the distal tip of the Delivery Catheter when inserting the Delivery System into the Access Sheath. If using a power injector, the maximum pressure should not exceed 100 psi. In view of the concerns that were raised by the RE-ALIGN 1 study of dabigatran in the presence of prosthetic mechanical heart valves, caution should be used when prescribing oral anticoagulants other than warfarin in patients treated with the WATCHMAN Device. The WATCHMAN Device has only been evaluated with the use of warfarin post-device implantation. ADVERSE EVENTS Potential adverse events (in alphabetical order) which may be associated with the use of a left atrial appendage closure device or implantation procedure include but are not limited to: Air embolism, Airway trauma, Allergic reaction to contrast media/medications or device materials, Altered mental status, Anemia requiring transfusion, Anesthesia risks, Angina, Anoxic encephalopathy, Arrhythmias, Atrial septal defect, AV fistula, Bruising, hematoma or seroma, Cardiac perforation, Chest pain/discomfort, Confusion post procedure, Congestive heart failure, Contrast related nephropathy, Cranial bleed, Decreased hemoglobin, Deep vein thrombosis, Death, Device embolism, Device fracture, Device thrombosis, Edema, Excessive bleeding, Fever, Groin pain, Groin puncture bleed, Hematuria, Hemoptysis, Hypotension, Hypoxia, Improper wound healing, Inability to reposition, recapture, or retrieve the device, Infection / pneumonia, Interatrial septum thrombus, Intratracheal bleeding, Major bleeding requiring transfusion, Misplacement of the device / improper seal of the appendage / movement of device from appendage wall, Myocardia erosion, Nausea, Oral bleeding, Pericardial effusion / tamponade, Pleural effusion, Prolonged bleeding from a laceration, Pseudoaneurysm, Pulmonary edema, Renal failure, Respiratory insufficiency / failure, Surgical removal of the device, Stroke Ischemic, Stroke Hemorrhagic, Systemic embolism, TEE complications (throat pain, bleeding, esophageal trauma), Thrombocytopenia, Thrombosis, Transient ischemic attack (TIA), Valvular damage, Vasovagal reactions There may be other potential adverse events that are unforeseen at this time. CAUTION: Federal law (USA) restricts this device to sale by or on the order of a physician. Rx only. Prior to use, please see the complete Directions for Use for more information on Indications, Contraindications, Warnings, Precautions, Adverse Events, and Operator s Instructions. 2015 Boston Scientific Corporation or its affiliates. All rights reserved. 1 Eikelboom JW, Connolly SJ, Brueckmann M, et al. N Engl J Med 2013;369:1206-14.