Samir R. Kapadia, MD On behalf of The PARTNER Trial Investigators Five-Year Outcomes of Transcatheter Aortic Valve Replacement (TAVR) in “Inoperable” Patients With Severe Aortic Stenosis: The PARTNER Trial Samir R. Kapadia, MD On behalf of The PARTNER Trial Investigators TCT 2014 | September 13, 2014

Background Transcatheter aortic valve replacement (TAVR) is the recommended treatment for “inoperable” patients with severe aortic stenosis (AS). Long term clinical benefit and valve performance in this population remain unknown. 2

Symptomatic Severe Aortic Stenosis ASSESSMENT: Transfemoral Access PARTNER Study Design Symptomatic Severe Aortic Stenosis Inoperable N = 358 Severe Symptomatic AS with AVA< 0.8 cm2 (EOA index < 0.5 cm2/m2), and mean gradient > 40 mmHg or jet velocity > 4.0 m/s Inoperable defined as risk of death or serious irreversible morbidity of AVR as assessed by cardiologist and two surgeons exceeding 50%. ASSESSMENT: Transfemoral Access 1:1 Randomization TF TAVR n = 179 Standard Therapy n = 179 VS Primary Endpoint: All-Cause Mortality Over Length of Trial (Superiority) Primary endpoint evaluated when all patients reached one year follow-up. After primary endpoint analysis reached, patients were allowed to cross-over to TAVR.

Key End-Points for 5 Year Analysis All-Cause Mortality Cardiac Mortality Re-hospitalization Stroke NYHA functional class Echo-derived valve areas, transvalvular gradients, and paravalvular leak. Mortality outcomes stratified by STS score, paravalvular leak and age. 4

Study Flow Inoperable Cohort Randomized Inoperable N = 179 Standard Therapy N = 179 TAVR 10 Patients Withdrew 85 / 85 patients 100% followed at 1 Yr 124 / 124 patients 100% followed at 1 Yr Cross Over 11 pts 19 / 19 patients 100% followed at 3 Yrs 81 / 83 patients 97.6% followed at 3 Yrs Cross Over 9 pts 6 / 6 patients 100% followed at 5 Yrs* 50 / 51 patients 98.0% followed at 5 Yrs* * ± 2 months follow-up window 5

Patient Characteristics TAVR N = 179 Standard Rx p-value Age – yr 83.1 ± 8.6 83.2 ± 8.3 0.95 Male sex (%) 45.8 46.9 0.92 STS Score 11.2 ± 5.8 12.1 ± 6.1 0.14 NYHA I or II (%) III or IV (%) 7.8 92.2 6.1 93.9 0.68 CAD (%) 67.6 74.3 0.20 COPD Any (%) O2 dependent (%) 41.3 21.2 52.5 25.7 0.04 0.38 Creatinine > 2 mg/dL (%) 5.6 9.6 0.23 Frailty (%) 18.1 28.0 0.09 Porcelain aorta (%) 19.0 11.2 0.05 Chest wall radiation (%) 8.9 8.4 1.00

All-Cause Mortality (ITT) Crossover Patients Censored at Crossover 71.8% 93.6% All-Cause Mortality (%) Months HR [95% CI] = 0.50 [0.39, 0.65] p (log rank) < 0.0001 Standard Rx (n = 179) TAVR (n = 179) 30.7% 50.8% 43.0% 68.0% 64.1% 87.5% 53.9% 80.9% * In an age and gender matched US population without comorbidities, the mortality at 5 years is 40.5%.

Median Survival 11.1 Months 29.7 Months Months p (log rank) < 0.0001

All-Cause Mortality (ITT) Landmark Analysis Standard Rx (n = 179) TAVR (n = 179) 30.7% 50.8% 33.4% 61.1% 38.9% 66.7% HR [95% CI] = 0.47 [0.24, 0.94] p (log rank) = 0.028 HR [95% CI] = 0.46 [0.32, 0.66] p (log rank) < 0.0001 HR [95% CI] = 0.50 [0.39, 0.65] Months 0-1 Year 3-5 Years 1-3 Years 6 12 18 24 30 36 42 48 54 60

Cardiovascular Mortality (%) Cardiovascular Mortality (ITT) Crossover Patients Censored at Crossover 57.3% 85.9% Cardiovascular Mortality (%) Months HR [95% CI] = 0.41 [0.31, 0.55] p (log rank) < 0.0001 Standard Rx (n = 179) TAVR (n = 179) 20.5% 44.6% 30.7% 62.4% 47.6% 80.6% 41.2% 74.5%

Causes of Death 18% vs 34% 48% vs 33% n=179 Percent Patients

All-Cause Mortality Stratified by STS Score (ITT) Months STS < 5 STS 5-15 STS > 15 100% 55.9% 93.3% 73.7% 75.2% 93.4% p (log rank) = 0.0012 p (log rank) = 0.0002 p (log rank) = 0.0749 Standard Rx (n = 123) TAVR (n = 113) Standard Rx (n = 12) TAVR (n = 28) Standard Rx (n = 43) TAVR (n = 38)

Cardiovascular Mortality Stratified by STS Score (ITT) Months 100% 41.1% 91.8% 57.8% 61.6% 82.4% STS < 5 STS 5-15 STS > 15 p (log rank) < 0.0001 p (log rank) = 0.0098 Standard Rx (n = 123) TAVR (n = 113) Standard Rx (n = 12) TAVR (n = 28) Standard Rx (n = 43) TAVR (n = 38)

Repeat Hospitalization (ITT) 87.3% 47.6% Months Rehospitalization (%) Standard Rx (n = 179) TAVR (n = 179) HR [95% CI] = 0.40 [0.29, 0.55] p (log rank) < 0.0001 53.9% 27.0% 72.5% 34.9% 75.7% 43.1% 83.0% 46.3%

NYHA Class Over Time (ITT) Survivors Baseline 1 Year 3 Years 5 Years N = 14.3% 40.0% 30.0% 50.0% 23.7% 60.8% 92.2% 93.9% p = NS p < 0.0001

Competing Risks Analysis (ITT) Death and Stroke Months 14.6% 5.7% Incidence (%)

Paravalvular Leak (AT) Percent of Evaluable Echocardiograms N =

Mortality by Paravalvular Leak None-Mild (n = 142) Moderate-Severe (n = 23) 69.2% 78.3% p (log rank) = 0.510 p (log rank) = 0.043 All-Cause Mortality Cardiovascular Mortality 51.3% 74.6%

Mean Gradient & Valve Area (AT) Mean Gradient (mmHg) Valve Area (cm²) N = EOA Mean Gradient Error bars = ± 1 Std Dev 159 86 70 44 31 15 163 91 71 46 19

Mean Gradient & Valve Area (AT) Restricted to Patients with 5 Year Data Mean Gradient (mmHg) EOA Mean Gradient Valve Area (cm²) N = Error bars = ± 1 Std Dev 15 14 12 13 16 20

Subgroup Analysis All-Cause Mortality Hazard Ratio [95% CI] Interaction p-value Overall (N=358) 0.50 [0.39-0.65] Age < 85 (N=186) 0.46 [0.33-0.66] 0.40 Age ≥ 85 (N=172) 0.56 [0.39-0.79] Male (N=166) [0.32-0.66] 0.34 Female (N=192) 0.55 [0.40-0.78] BMI ≤ 25 (N=170) 0.58 [0.41-0.84] 0.71 BMI > 25 (N=188) 0.44 [0.31-0.63] STS ≤ 11 (N=170) 0.52 [0.36-0.76] 0.65 STS > 11 (N=187) 0.53 [0.37-0.74] EF ≤ 55 (N=173) 0.47 [0.33-0.67] 0.09 EF > 55 (N=171) 0.61 [0.42-0.88] Pulmonary Hypertension No (N=136) [0.37-0.85] 0.87 Yes (N=103) 0.51 [0.32-0.82] Mod / Sev MR No (N=261) [0.43-0.77] 0.03 Yes (N=77) 0.30 [0.17-0.53] Oxygen Dependent COPD No (N=270) [0.35-0.62] 0.14 Yes (N=88) 0.68 [0.42-1.10] Prior CABG or PCI No (N=182) [0.39-0.78] 0.27 Yes (N=176)

TAVR Mortality Stratified by Age (ITT) Months 96.0% 73.5% 91.8% 70.4%

Clinical Observations Mortality benefit was similar in elderly (>85 yr) patients compared to those ≤85 years. Cardiovascular mortality and all-cause mortality benefit was seen even in patients with high STS score. Patients with O2 dependent COPD may have less mortality benefit. Beyond early procedural risk of stroke there was no persistent risk over 5-year follow up. Moderate and severe paravalvular leak is associated with higher cardiovascular mortality particularly in patients with less comorbidities.

Main Conclusions At 5 years follow-up benefits of TAVR were sustained as measured by: All-Cause Mortality Cardiovascular Mortality Repeat Hospitalization Functional Status Valve durability was demonstrated with no increase in transvalvular gradient or attrition of valve area. 24

Thank You to the Dedicated Study Teams at All PARTNER Investigational Sites 25