Craig R. Smith, MD on behalf of The PARTNER Trial Investigators Transcatheter vs. Surgical Aortic Valve Replacement in High Risk Patients with Severe Aortic Stenosis: Results From The PARTNER Trial ACC 2011 | New Orleans | April 3, 2011

Craig R. Smith, MD PARTNER Trial sponsor (Edwards LifeSciences) reimburses customary travel and other expenses Presenter Disclosure Information for PARTNER at ACC April 3, 2011

Background Surgical aortic valve replacement (AVR) is the standard of care for treating patients with symptomatic aortic stenosis who are candidates for operation. Although transcatheter aortic valve replacement (TAVR) reduces mortality in patients who cannot have AVR, there have been no randomized trials comparing TAVR to AVR in patients who are at high-risk for operation.

Purpose To compare the safety and effectiveness of TAVR (either transfemoral or transapical) to surgical AVR in high-risk, operable patients with symptomatic, severe aortic stenosis.

N = 699 N = 358 High Risk Inoperable PARTNER Study Design Symptomatic Severe Aortic Stenosis ASSESSMENT: High-Risk AVR Candidate 3,105 Total Patients Screened ASSESSMENT: High-Risk AVR Candidate 3,105 Total Patients Screened Total = 1,057 patients 2 Parallel Trials: Individually Powered Standard Therapy Standard Therapy ASSESSMENT: Transfemoral Access Not In Study TF TAVR Primary Endpoint: All-Cause Mortality Over Length of Trial (Superiority) Co-Primary Endpoint: Composite of All-Cause Mortality and Repeat Hospitalization (Superiority) Primary Endpoint: All-Cause Mortality Over Length of Trial (Superiority) Co-Primary Endpoint: Composite of All-Cause Mortality and Repeat Hospitalization (Superiority) 1:1 Randomization VS Yes No N = 179

Inoperable PARTNER Cohort Primary Endpoint: All-Cause Mortality Numbers at Risk Numbers at Risk TAVI TAVI Standard Rx Standard Rx Standard Rx TAVI All-cause mortality (%) Months ∆ at 1 yr = 20.0% NNT = 5.0 pts 50.7% 30.7% HR [95% CI] = 0.54 [0.38, 0.78] P (log rank) < Leon et al, NEJM 2010; 363:

N = 179 N = 358 Inoperable Standard Therapy Standard Therapy ASSESSMENT: Transfemoral Access Not In Study TF TAVR Primary Endpoint: All-Cause Mortality Over Length of Trial (Superiority) Co-Primary Endpoint: Composite of All-Cause Mortality and Repeat Hospitalization (Superiority) Primary Endpoint: All-Cause Mortality Over Length of Trial (Superiority) Co-Primary Endpoint: Composite of All-Cause Mortality and Repeat Hospitalization (Superiority) 1:1 Randomization VS Yes No N = 179 TF TAVR AVR Primary Endpoint: All-Cause Mortality at 1 yr (Non-inferiority) TA TAVR AVR VS N = 248N = 104N = 103N = 244 PARTNER Study Design Symptomatic Severe Aortic Stenosis ASSESSMENT: High-Risk AVR Candidate 3,105 Total Patients Screened ASSESSMENT: High-Risk AVR Candidate 3,105 Total Patients Screened Total = 1,057 patients 2 Parallel Trials: Individually Powered N = 699 High Risk ASSESSMENT: Transfemoral Access Transapical (TA) Transfemoral (TF) 1:1 Randomization Yes No

Primary Endpoint All-cause mortality at one year Analysis by intent-to-treat Event rates by Kaplan-Meier estimates Crossovers permitted only when assigned therapy unsuccessful All patients followed for ≥ one year

Other Important Endpoints (1) Safety: Neurologic events –Prospective: Stroke and stroke plus TIA (all neuro events) –Retrospective: Major stroke (modified Rankin Score ≥ ≥ 30 days) Major vascular complications (VARC definition) Major bleeding (modified VARC definition) Repeat hospitalization New pacemakers and new-onset atrial fibrillation (ECG core lab) Procedural events (assigned therapy aborted or converted to AVR, multiple valves, etc.) Surgical complications (re-op for bleeding, sternal infection, etc.)

Other Important Endpoints (2) Clinical Effectiveness and Valve Performance: NYHA symptoms Six-minute walk tests Quality-of-life measures and cost-effectiveness (core lab) Echocardiography assessment of valve performance (core lab) –Peak and mean gradients –Effective orifice area –Bioprosthetic valve regurgitation (esp. para-valvular) –Other: LV ejection fraction, MR, LV mass, evidence of structural valve deterioration

Co-Principal Investigators Martin B. Leon, Craig R. Smith Columbia University Medical Center Executive Committee Martin B. Leon, Michael Mack, D. Craig Miller, Jeffrey W. Moses, Craig R. Smith, Lars G. Svensson, E. Murat Tuzcu, John G. Webb Data & Safety Monitoring Board Chairman: Joseph P. Carrozza Tufts University School of Medicine Clinical Events Committee Chairman: John L. Petersen Duke University Medical Center Echo Core Laboratory Chairman: Pamela C. Douglas Duke University Medical Center Quality of Life and Cost-Effectiveness Chairman: David J. Cohen Mid America Heart Institute, Kansas City Independent Biostatistical Core Laboratory Stuart Pocock, Duolao Wang London School of Hygiene and Tropical Medicine William N. Anderson Publications Committee Co-Chairman: Jeffrey W. Moses Lars G. Svensson Sponsor Edwards Lifesciences: Jodi J. Akin Study Administration

Executive Committee Executive Committee John Webb Murat Tuzcu Marty Leon Jeff Moses Lars SvenssonCraig Miller Michael Mack Craig Smith

Participating Study Sites Intermountain Medical Center Salt Lake City, UT Emory University Atlanta, GA Univ. of Miami Miami, FL Univ. of Virginia Charlottesville, VA St. Luke’s Hospital Kansas City, MO Barnes-Jewish Hospital St. Louis, MO Medical City Dallas Dallas, TX St. Paul's Hospital Vancouver, Canada Univ. of Washington Seattle, WA Mayo Clinic Rochester, MN Stanford University Palo Alto, CA Hospital Laval Quebec City, Canada Ochsner Foundation New Orleans, LA Scripps Clinic La Jolla, CA Cedars-Sinai Medical Center Los Angeles, CA Cleveland Clinic Cleveland, OH Columbia University Cornell University New York, NY Washington Hosp. Center Wash., DC Univ. of Penn Phila., PA Brigham & Women’s Mass General Boston, MA Northwestern Univ. Chicago, IL Toronto Gen. Hospital Toronto, Canada Evanston Hospital Leipzig Heart Center Leipzig, Germany n = 1,057 patients 26 investigator sites 22 USA, 3 Canada, 1 Germany

High-Risk Enrollment by Site Cedars-Sinai Medical Ctr Los Angeles, CA G. Fontana, R. Makkar 116 Columbia University New York City, NY M. Leon, C. Smith 97 Medical City Dallas Dallas, TX D. Brown, T. Dewey 95 Emory University Atlanta, GA P. Block, R. Guyton 67 University of Pennsylvania Philadelphia, PA J. Bavaria, H. Herrmann 52 Cleveland Clinic Found Cleveland, OH L. Svensson, M. Tuzcu 47 Washington Hospital Ctr District of Columbia P. Corso, A. Pichard 40 University of Miami Miami, FL W. O’Neill, D. Williams 25 Barnes-Jewish Hospital St. Louis, MO R. Damiano, J, Lasala 24 Stanford University Palo Alto, CA C. Miller, A. Yeung 23 Northwestern University Chicago, IL C. Davidson, P. McCarthy 20 St. Paul's Hospital Vancouver, BC, Canada A. Cheung, J. Webb 19

Mass General Hospital Boston, MA I. Palacios, G. Vlahakis 15 St. Luke’s Hospital Kansas City, MO K. Allen, D. Cohen 13 Universitaire de Quebec Laval, Quebec, CA D. Doyle, J. Rodes-Cabau 8 Scripps Clinic La Jolla, CA S. Brewster, P. Teirstein 7 Herzzentrum Leipzig Leipzig, Germany F. Mohr, G. Schuler 7 Mayo Clinic Rochester, MN C. Rihal, T. Sundt 6 Univ of Washington Seattle, WA M. Reisman, E. Verrier 5 Northshore Univ Health Sys Evanston, IL J. Alexander, T. Feldman 4 University of Virginia Charlottesville, VA I. Kron, S. Lim 3 Brigham & Women’s Hosp Boston, MA M. Davidson, A. Eisenhauer 2 Ochsner Foundation New Orleans, LA E. Parrino, S. Ramee 2 Intermountain Med Center Salt Lake City, UT K. Jones, B. Whisenant 1 Cornell University New York City, NY K. Krieger, C. Wong 1 Toronto General Hospital Toronto, Ontario, CA C. Feindel, E. Horlick 0 High-Risk Enrollment by Site

Study Devices Edwards SAPIEN THV 23 and 26 mm valves RetroFlex 22 and 24 F sheaths Ascendra 24 and 26 F sheaths

Transfemoral Transapical TAVR Transfemoral and Transapical

Inclusion Criteria Severe AS: Echo-derived AVA 40 mm Hg or peak jet velocity > 4.0 m/s Cardiac Symptoms: NYHA Functional Class ≥ II High surgical risk: Predicted risk of operative mortality ≥ 15% (determined by site surgeon and cardiologist); guideline = STS score ≥ 10

Key Exclusion Criteria (1) Bicuspid or non-calcified aortic valve Aortic annulus diameter (echo measurement) 25 mm Aortic dissection or iliac-femoral dimensions or disease precluding safe sheath insertion (esp. calcification) Severe LV dysfunction (LVEF < 20%) Untreated CAD requiring revascularization Severe AR or MR (> 3+) or prosthetic valve (any location) Anatomic:

Key Exclusion Criteria (2) Serum Cr > 3.0 mg/dL or dialysis dependent Acute MI within 1 monthAcute MI within 1 month Upper GI bleed within 3 monthsUpper GI bleed within 3 months CVA or TIA within 6 monthsCVA or TIA within 6 months Any cardiac procedure, other than BAV, within 1 month or within 6 months for DESAny cardiac procedure, other than BAV, within 1 month or within 6 months for DES Hemodynamic instability (e.g. requiring inotropic support)Hemodynamic instability (e.g. requiring inotropic support) Clinical:

Statistical Analysis Plan Primary hypothesis is non-inferiority of test (TAVR) vs. control (AVR) for all-cause mortality at 1 year Non-inferior if one-sided 95% upper confidence limit for the treatment difference is < 7.5% (α =0.05) Primary Endpoint: All TF and TA patients –Assuming true 1-year mortality 32% after AVR and 29% after TAVR –Intended sample size = 650 patients for ≥ 85% power Powered Secondary Endpoint: Only TF patients –Assuming true 1-year mortality 35% after AVR and 25% after TAVR –Intended sample size = 450 patients for ≥ 85% power

Study Methodology Preliminary eligibility determined by site investigators Every case reviewed by web-based conference call before enrollment Randomized to TF-TAVR vs. AVR, or TA-TAVR vs. AVR, to be treated within 2 weeks Intent-to-treat (ITT) analysis for the primary and most secondary endpoints; defined as the time of randomization As-treated (AT) analysis for some procedural endpoints and for echo assessments; defined as the time of procedural anesthesia induction

1 Year (189) Dead = 46 Withdrawal = 1 1 Year (189) Dead = 46 Withdrawal = 1 1 Year (168) Dead = 47 Withdrawal = 8 1 Year (168) Dead = 47 Withdrawal = 8 Study Flow AVR (248) 30 Days (236) Dead = 8 Withdrawal = 0 30 Days (236) Dead = 8 Withdrawal = 0 Randomized = 699 patients TF = 492 TA = 207 Transfemoral n = 492 TAVR (244) 30 Days (223) Dead = 15 Withdrawal = Days (223) Dead = 15 Withdrawal = 10 1 Year (73) Dead = 26 Withdrawal = 0 LTFU = 1 1 Year (73) Dead = 26 Withdrawal = 0 LTFU = 1 1 Year (68) Dead = 20 Withdrawal = 3 LTFU = 1 1 Year (68) Dead = 20 Withdrawal = 3 LTFU = 1 AVR (103) 30 Days (100) Dead = 4 Withdrawal = 0 30 Days (100) Dead = 4 Withdrawal = 0 Transapical n = 207 TAVR (104) 30 Days (92) Dead = 7 Withdrawal = 4 30 Days (92) Dead = 7 Withdrawal = 4 42 Patients not treated as assigned

Reason TAVR (N = 348) AVR (N = 351) Died before treatment - no. (%)2 (0.6)5 (1.4) Deterioration before treatment - no. (%)1 (0.3)5 (1.4) Refusal - no. (%)1 (0.3)17 (4.8) Withdrawal - no. (%)011 (3.1) Total – no. (%) 4 (1.1) 38 (10.8) Reasons for Non-treatment ITT = 699 patients │ AT = 657 patients NOTE: Time from randomization to treatment = TAVR 10.6 [SEM 0.7] days vs. AVR 15.6 [SEM 1.1] days; P <0.001

Characteristic TAVR (N = 348) AVR (N = 351) p-value Age (yr)83.6 ± ± Male sex - % STS Score11.8 ± ± Logistic EuroSCORE29.3 ± ± NYHA II - % III or IV - % CAD - % Previous MI - % Prior CV Intervention - % Prior CABG - % Prior PCI - % Prior BAV - % Patient Characteristics (1) Cerebrovascular disease - %

Characteristic TAVR (N = 348) AVR (N = 351) p-value Peripheral vascular disease - % COPD Any 43.4 Oxygen dependent Creatinine > 2mg/dL - % Atrial fibrillation - % Permanent pacemaker - % Pulmonary hypertension - % Frailty - % Porcelain aorta - % Chest wall radiation - % Liver disease - % Patient Characteristics (2)

Baseline Echocardiography Echo Findings TAVR (N = 348) AVR (N = 351) p-value AVA - cm ± ± AVG - mm Hg42.7 ± ± Mean LVEF - %52.5 ± ± Moderate or severe MR - %

Anesthesia time - min330 Total procedure time - min230 Aborted procedure - no. (%)0 Reoperation for bleeding - no. (%)12 (3.4) Intra-procedural death - no. (%)1 (0.3) Aortic perforation - no. (%)1 (0.3) Aortic dissection - no. (%)3 (0.9) Median ICU stay (days)5.0 Anesthesia time - min236 Total procedure time - min133 Aborted procedure - no. (%)7 (2.0) Reoperation for bleeding - no. (%)2 (0.6) Intra-procedural death - no. (%)3 (0.9) Aortic perforation - no. (%) 0 Aortic dissection - no. (%)3 (0.9) Median ICU stay (days)3.0 Procedural Outcomes - TAVR vs AVR *Converted to transapical TAVR due to porcelain aorta AVRTAVR 3 failed access 2 new TEE findings 2 died 3 failed access 2 new TEE findings 2 died Aborted procedure - no. (%)

Sternal wound infection - no. (%)7 (2.0) Total cross clamp time - min74 Pump time - min105 Access site infection - no. (%)7 (2.0) Fluoroscopy time - min31 Converted to AVR - no. (%)9 (2.6) Multiple (≥2) valves - no. (%)7 (2.0) Valve embolization - no. (%)9 (2.6) Procedural Outcomes - TAVR vs AVR *Converted to transapical TAVR due to porcelain aorta AVRTAVR 5 valve embolization 3 annulus size on TEE 1 large sigmoid septum 5 valve embolization 3 annulus size on TEE 1 large sigmoid septum Converted to AVR - no. (%) 5 converted to AVR 2 valve-in-valve 2 not treated 5 converted to AVR 2 valve-in-valve 2 not treated Valve embolization - no. (%)

TAVR AVR Months No. at Risk TAVR AVR Primary Endpoint: All-Cause Mortality at 1 Year HR [95% CI] = 0.93 [0.71, 1.22] P (log rank) = 0.62

Primary Endpoint: All-Cause Mortality at 1 Year Non-inferior Upper one-sided 95% CI 8.0 % AVR (N = 351) 26.8% AVR (N = 351) 26.8% TAVR (N = 348) 24.2% TAVR (N = 348) 24.2% Difference-2.6% Upper 1-sided 95% CI3.0% Difference-2.6% Upper 1-sided 95% CI3.0% Primary Non-Inferiority Endpoint Met Non- inferiority P value = Non- inferiority P value = Zone of non-inferiority pre-specified margin = 7.5%

All-Cause Mortality Transfemoral (N=492) Months No. at Risk TAVR AVR HR [95% CI] = 0.83 [0.60, 1.15] P (log rank) = 0.25

Powered Secondary Endpoint (ITT): TF All-Cause Mortality at 1 Year Zone of non-inferiority pre-specified margin = 7.5% Non-inferior Upper one-sided 95% CI Secondary TF Non-Inferiority Endpoint Met 8.0 % Difference-4.2% Upper 1-sided 95% CI2.3% Difference-4.2% Upper 1-sided 95% CI2.3% Non- inferiority P value = Non- inferiority P value = AVR (N = 244) 26.4% AVR (N = 244) 26.4% TAVR (N = 248) 22.2% TAVR (N = 248) 22.2%

TAVR AVR MonthsNo. at Risk All-Cause Mortality Transapical (N=207) HR [95% CI] = 1.22 [0.75, 1.98] P (log rank) = 0.41

Surgical AVR Outcomes Using an established predictive risk model (STS), the expected (“E”) 30-day mortality after AVR was 11.8%. The observed (“O”) 30-day mortality in the as-treated AVR control group was 8.0%. O:E = 0.68 indicates better than predicted surgical outcomes in the control AVR patients. There were no significant site or surgeon differences.

All-Cause Mortality at 30 Days All Patients no. of patients ( %) TF Patients no. of patients ( %) TA Patients no. of patients ( %) TAVRAVRp-valueTAVRAVRp-valueTAVRAVRp-value ITT12 (3.4)22 (6.5)0.078 (3.3)15 (6.2)0.134 (3.8)7 (7.0)0.32 AT18 (5.2)25 (8.0)0.159 (3.7)18 (8.2)0.059 (8.7)7 (7.6)0.79 All-Cause Mortality at 1 Year All Patients no. of patients ( %) TF Patients no. of patients ( %) TA Patients no. of patients ( %) TAVRAVRp-valueTAVRAVRp-valueTAVRAVRp-value ITT84 (24.2)89 (26.8) (22.2)62 (26.4) (29.0)27 (27.9)0.85 AT81 (23.7)78 (25.2) (21.3)55 (25.2) (29.1)23 (25.3)0.55 All-Cause Mortality at 30 Days and 1 Year ITT and ATT by Subgroup

30 Days 1 YearOutcome TAVR (N = 348) AVR (N = 351) p-valueTAVR (N = 348) AVR (N = 351) p-value All mortality – no. (%) 12 (3.4)22 (6.5) (24.2)89 (26.8)0.44 Cardiac mortality – no. (%)11 (3.2)10 (3.0) (14.3)40 (13.0)0.63 Rehospitalization – no. (%)15 (4.4)12 (3.7) (18.2)45 (15.5)0.38 Death or rehosp – no. (%) 25 (7.2)33 (9.7) (34.6)119 (35.9)0.73 MI – no. (%)02 (0.6)0.161 (0.4)2 (0.6)0.69 Acute kidney inj* – no. (%)10 (2.9)10 (3.0) (5.4)20 (6.5)0.56 Clinical Outcomes at 30 Days and 1 Year All Patients (N=699) * Renal replacement therapy

30 Days 1 YearOutcome TAVR (N = 348) AVR (N = 351) p-valueTAVR (N = 348) AVR (N = 351) p-value Vascular complications All – no. (%) 59 (17.0) 13 (3.8) < (18.0) 16 (4.8) <0.01 Major – no. (%) 38 (11.0) 11 (3.2) < (11.3) 12 (3.5) <0.01 Major bleeding – no. (%) 32 (9.3) 67 (19.5) < (14.7) 85 (25.7) <0.01 Endocarditis – no. (%) 0 (0.0) 1 (0.3) (0.6) 3 (1.0) 0.63 New AF – no. (%) 30 (8.6) 56 (16.0) < (12.1) 60 (17.1) 0.07 New PM – no. (%) 13 (3.8) 12 (3.6) (5.7) 16 (5.0) 0.68 Clinical Outcomes at 30 Days and 1 Year All Patients (N=699)

30 Days 1 YearOutcome TAVR (N = 348) AVR (N = 351) TAVR (N = 348) AVR (N = 351) All Stroke or TIA – no. (%) 19 (5.5) 8 (2.4) (8.3) 13 (4.3) 0.04 TIA – no. (%)3 (0.9)1 (0.3)0.337 (2.3)4 (1.5)0.47 All Stroke – no. (%)16 (4.6)8 (2.4) (6.0)10 (3.2)0.08 Major Stroke – no. (%) 13 (3.8) 7 (2.1) (5.1) 8 (2.4) 0.07 Minor Stroke – no. (%)3 (0.9)1 (0.3)0.343 (0.9)2 (0.7)0.84 Death/maj stroke – no. (%) 24 (6.9) 28 (8.2) (26.5) 93 (28.0) 0.68 Neurological Events at 30 Days and 1 Year All Patients (N=699) p-value p-value

Months No. at Risk TAVR AVR HR [95% CI] = 0.95 [0.73, 1.23] P (log rank) = 0.70 All-Cause Mortality or Stroke All Patients (N=699)

NYHA Functional Class Baseline1 Year6 Months30 Days Patients Surviving, % IIIIIIIV P = 1.00P < 0.001P = 0.05P = 0.75

Six-Minute Walk Test All Patients (N=699) Median Distance, meters P = 0.73P = 0.002P = 0.33P = 0.76

Subgroup TAVR (%) n=348 AVR (%) n=351 RR (95% CI) P-value for interaction Overall ( ) Age <85 > ( ) 1.03( ) 0.52 Sex Male Female ( ) BMI <26 > ( ) 0.99( ) 0.66 STS score <11 > ( ) 0.92( ) 0.87 LV ejection fraction <55 > ( ) 1.01( ) 0.80 Subgroup Analyses of Treatment Effect All-Cause Mortality at 1 Year TAVR better AVR better 0.512

Subgroup TAVR (%) n=348 AVR (%) n=351 RR (95% CI) P-value for interaction Pulmonary hypertension No Yes ( ) 0.92( ) 0.80 Mitral regurgitation No Yes ( ) 0.69( ) 0.12 Prior CABG No Yes ( ) 1.35( ) 0.02 Peripheral vasc disease No Yes ( ) 1.04( ) 0.57 Cohort TA TF ( ) 0.89( ) 0.43 TAVR better AVR better Subgroup Analyses of Treatment Effect All-Cause Mortality at 1 Year

Mean Gradient - AVR Mean Gradient - TAVR Peak Gradient - AVR Peak Gradient - TAVR Mean and Peak Gradient As-Treated Trial Arms (mmHg) Baseline30 Days6 Months1 Year TAVR n = 327 AVR n = 301 TAVR n = 287 AVR n = 231 TAVR n = 246 AVR n = 170 TAVR n = 227 AVR n = 159 Echo Findings Aortic Valve Gradients

Echo Findings Hemodynamic Assessments 30 Days1 Year Finding TAVRAVRTAVRAVR AVG – mmHg9.9 ± ± ± ± AVA - cm ± ± ± ± LVEF - %55.5 ± ± ± ± p-value

Paravalvular Aortic Regurgitation P < Year6 Months30 Days Patients, % NoneTraceMildModerate Severe

Study Limitations 8% of the control (AVR) group withdrew or refused assigned therapy 5% of patients randomized to TAVR did not receive assigned therapy (procedure aborted or converted to AVR) Significantly longer interval between randomization and treatment in controls (AVR) An early version large TAVR delivery system was used Most sites had no previous TAVR experience - learning curve impact inherent in TAVR, but not in AVR Insufficient statistical power to compare TA to either AVR or TF Long-term follow-up not available to assess TAVR durability

Conclusions (1) The primary endpoint of the trial was met: –In patients with aortic stenosis at high risk for operation, TAVR was non-inferior to AVR for all-cause mortality at 1 year (24.2% vs. 26.8%, p=0.001 for non inferiority) –Transfemoral TAVR subgroup was also non-inferior to AVR (p=0.002 for non-inferiority) Death at 30 days was lower than expected in both arms of the trial: –TAVR mortality (3.4%) was the lowest reported in any series, despite an early generation device and limited previous operator experience –AVR mortality (6.5%) was lower than the expected operative mortality (11.8%)

Conclusions (2) Both TAVR and AVR were associated with important but different peri-procedural hazards: –Major strokes at 30 days (3.8 vs. 2.1%, p=0.20) and one year (5.1% vs. 2.4%, p=0.07) and major vascular complications were more frequent with TAVR (11.0% vs. 3.2%, p<0.001) –Major bleeding (9.3% vs. 19.5%, p<0.001) and new onset atrial fibrillation (8.6% vs. 16.0%, p<0.001) were more frequent with AVR TAVR and AVR are both acceptable therapies in these high-risk patients; differing peri-procedural hazards should influence case-based decision-making

Conclusions (3) Symptom improvement (NYHA class and 6-min walk distance) favored TAVR at 30 days and was similar to AVR at one year Echo findings indicate: –Small hemodynamic benefit with TAVR vs. AVR at 1 year (mean gradient p=0.008, AVA p=0.002) –Increased para-valvular regurgitation associated with TAVR (p<0.001) Preliminary subgroup analyses should be interpreted cautiously: –Possible TAVR benefit in women and patients without prior CABG

Implications A multidisciplinary valve team approach benefits patients and is recommended for all future valve centers. TAVR is already the standard-of-care for inoperable patients with severe aortic stenosis. These results indicate that TAVR is an acceptable alternative to AVR in selected high-risk operable patients. Future randomized studies should focus on lower risk patients who are candidates for operation.

Back-up Slides

PARTNER Comparison of Outcomes High-Risk vs. Inoperable Patients Per Cent (%)

TAVRAVRComplicationWithWithoutWithWithout Stroke or TIA All Patients – no Died ≤ 30 days – no. (%) 2 (6.5) 16 (5.1) 1 (6.3) 24 (8.1) ≤ 1 year – no. (%) Died ≤ 1 year – no. (%)10 (32.3)71 (22.8)3 (18.8)75 (25.6) Major Stroke All Patients – no Died ≤ 30 days – no. (%) 2 (11.1)16 (4.9)1 (9.1)24 (8.0) ≤ 1 year – no. (%) Died ≤ 1 year – no. (%)9 (50.0)72 (22.2)3 (27.3)75 (25.1) Neurological Events and Mortality at 30 Days and 1 Year (as treated)

TAVRAVRComplicationWithWithoutWithWithout Major Vascular All Patients – no Died ≤ 30 days – no. (%) 6 (15.8) 12 (3.9) 2 (19.2) 23 (7.6) ≤ 1 year – no. (%) Died ≤ 1 year – no. (%)14 (37.3)67 (22.0)5 (49.5)73 (24.4) Major Bleeding All Patients – no Died ≤ 30 days – no. (%) 2 (3.8)16 (5.5)15 (17.1)10 (4.5) ≤ 1 year – no. (%) Died ≤ 1 year – no. (%)16 (31.1)65 (22.3)36 (41.3)42 (18.9) Vascular/Bleeding Events and Mortality at 30 Days and 1 Year (as treated)

All-Cause Mortality (As Treated TAVR Trial Arm) Stratified by Major Stroke Months

All-Cause Mortality (As Treated TAVR Trial Arm) Stratified by Major Vascular Event Months

All-Cause Mortality (As Treated TAVR Trial Arm) Stratified by Major Bleed Months

All-Cause Mortality (As Treated AVR Trial Arm) Stratified by Major Stroke Months

All-Cause Mortality (As Treated AVR Trial Arm) Stratified by Major Bleed Months

All-Cause Mortality (As Treated AVR Trial Arm) Stratified by New Atrial fibrillation Months

All-Cause Mortality (As Treated) Pooled Implant Approaches (N= 657) Months No. at Risk TAVR AVR HR [95% CI] = 1.02 [0.77, 1.36] P (log rank) = 0.88

Primary Endpoint: (As Treated) All-Cause Mortality at 1 Year Zone of non-inferiority pre-specified margin = 7.5% Non-inferior Upper one-sided 95% CI Primary Non-Inferiority Endpoint Met 8.0 % Difference-1.6% Upper 1-sided 95% CI4.0% Difference-1.6% Upper 1-sided 95% CI4.0% Non- inferiority P value = Non- inferiority P value = AVR (N = 313) 25.2% AVR (N = 313) 25.2% TAVR (N = 344) 23.6% TAVR (N = 344) 23.6%

All-Cause Mortality (As Treated) Transfemoral (N=461) Months No. at Risk TAVR AVR HR [95% CI] = 0.90 [0.64, 1.26] P (log rank) = 0.53

Powered Secondary Endpoint (AT): TF All-Cause Mortality at 1 Year Zone of non-inferiority pre-specified margin = 7.5% Non-inferior Upper one-sided 95% CI Secondary TF Non-Inferiority Endpoint Met 8.0 % Difference-3.9% Upper 1-sided 95% CI2.6% Difference-3.9% Upper 1-sided 95% CI2.6% Non- inferiority P value = Non- inferiority P value = AVR (N = 221) 25.2% AVR (N = 221) 25.2% TAVR (N =240 ) 21.3% TAVR (N =240 ) 21.3%

All-Cause Mortality (As Treated) Transapical (N=196) Months No. at Risk TAVR AVR HR [95% CI] = 1.36 [0.82, 2.26] P (log rank) = 0.23

All-Cause Mortality Stratified by ITT Trial Arm and Patient Gender

All-Cause Mortality Stratified by ITT Trial Arm and Prior CABG

All-Cause Mortality Stratified by ITT trial arm and Moderate/Severe MR

Echo Findings Valvular Regurgitation 30 Days1 Year Finding – no. (%) TAVRAVRTAVRAVR Transvalv. Regurg. Mod/Severe 3 (1.0)2 (0.9)< (0.9)0 (0.0)<.0001 Paravalv. Regurg. Mod/Severe 35 (12.2) 2 (0.9) < (6.8) 3 (1.9) <.0001 All Regurg. Mod/Severe 21 (7.7)4 (1.7)< (5.5)3 (1.9)<0.001 p-valuep-value

Echo Findings Paravalvular Regurgitation 30 Days1 Year Finding – no. (%) TAVRAVRTAVRAVR None65 (22.6)168 (73.7)< (32.9)123 (77.8)<.0001 Trace/Mild 187 (65.2) 58 (25.4) < (60.4) 32 (20.3) <.0001 Mod/Severe 35 (12.2) 2 (0.9) < (6.8) 3 (1.9) <.0001 p-value p-value

AVR TAVR Valve Area, cm 2 Aortic Valve Area As Treated Trial Arms TAVR n = 319 AVR n = 297 TAVR n = 279 AVR n = 228 TAVR n = 235 AVR n = 165 TAVR n = 219 AVR n = 155