One-Year Interim Results Gore VIBRANT Clinical Study [GORE] VIaBahn® [Endoprosthesis] veRsus bAre Nitinol stenT in the Treatment of Long Lesion (≥ 8 cm) Superficial Femoral Artery Occlusive Disease Dr. Gary Ansel, MD, FACC For the VIBRANT Study Investigators

Gary M. Ansel, MD DISCLOSURES Royalty Cook Medical Consulting Fees AccessClosure, Inc., AngioSlide, Ltd., AtheroMed,Inc., Bacchus Vascular, Bard Peripheral Vascular, Inc., BioCardia, Inc., Boston Scientific Corporation, Cordis, a Johnson & Johnson company, ev3, Inc. Honoraria Abbott Vascular Ownership Interest (Stocks, Stock Options or Other Ownership Interest) ICON Interventional Systems, Inc.

Goals Leading to VIBRANT Real world fempop lesions Real world device use Stent overlaps allowed Core lab controlled, longer term results Clinical application of various duplex velocities Evaluate potential predictors of patency

Gore VIBRANT Clinical Study Investigators Co-PIs: Gary Ansel, MD; Columbus, Ohio Patrick Geraghty, MD; Washington University, St. Louis, Missouri Mark Mewissen, MD; St. Luke’s Hospital, Milwaukee, Wisconsin Core Lab: The Vascular Ultrasound Core Laboratory, Boston, MA Dr. Michael R. Jaff, DO, RPVI, FACP, FACC Site Investigators: David Lasorda, DO, Allegheny General Hospital, Pittsburgh, Pennsylvania Michael Rush, MD, Holy Cross Hospital, Ft Lauderdale, Florida Jeffrey Goldstein, MD, St. Johns, Springfield, Illinois Bob Smouse, MD, St. Francis, Peoria, Illinois Manju Kalra, MD, Mayo Clinic, Rochester, Minnesota Sean Lyden, MD, Cleveland Clinic, Cleveland, Ohio Rauh, MD, Midwest Heart Foundation, Lombard, Illinois Arun Chervu, MD, Vascular Surgical Associates, Austell, Georgia William Gray, MD, Columbia University Medical Center, New York, New York Romi Chopra, MD, Midwest Institute, Melrose Park, Illinois Sam Money, MD, Mayo Clinic, Scottsdale, Arizona David Mego, MD, Arkansas Heart, Little Rock, Arkansas Mark Bates, MD, Charleston Medical Center, West Virginia Richard Begg, MD, Heritage Valley Health System, Beaver Falls, Pennsylvania W. Charles Sternbergh, MD, Ochsner Clinic, New Orleans, Louisiana

Take Home Points VIBRANT: Multicenter; core lab controlled; real world trial of the highest femoropopliteal disease burden ever performed Treating claudicants with long diffuse disease is clinically safe and clinically effective at one year Stent grafts restenosed at edge; bare metal stents restenosed diffusely Primary patencies suboptimal; primary-assisted and secondary patencies are good even in diffuse lesions but strategies need to be refined Even helical stents appear to have high fracture rates in long lesions and these are associated with vessel restenosis Viabahn fracture is a rare event Surveillance appears to be important at achieving high assisted and secondary patency to over 80% at 1-year Longer term follow-up continues; primary end point will be completed 2011

Study Hypothesis: (3-year study) The use of GORE VIABAHN® Endoprosthesis will result in greater mid- and long-term patency (≥ 2-yr) of the treated arterial lesion compared to the use of bare nitinol stents in treating chronic SFA long lesion (≥ 8 cm) stenoses and occlusions

Study Design 148 randomized patients enrolled TEST GROUP: GORE VIABAHN® Endoprosthesis FDA approved for SFA indication, June 14, 2005 Does not include Bioactive Heparin Surface or contoured edge device n = 72 CONTROL GROUP: Bare Nitinol Stent Commercially available bare nitinol stent as determined by institutional standard of care when treating SFA occlusive disease and were not devices approved for SFA use n = 76

Study Endpoints Primary Endpoints Secondary Endpoints Three-year primary patency (duplex PSV ratio < 2.0 / core lab determination) Safety (30 day composite of major complications) Secondary Endpoints Primary assisted patency Secondary patency Technical success Target vessel revascularization, target lesion revascularization Clinical success Change in ABI from baseline Alternate peak systolic velocity ratios of 2.5 and 3.0 (core lab adjudicated) Stent fracture on plain X-ray (core lab adjudicated)

GORE VIABAHN® Endoprosthesis Patient Demographics GORE VIABAHN® Endoprosthesis Bare Nitinol Stent p-value NUMBER OF RANDOMIZED SUBJECTS 72 76 Gender (% male) 62.5% 64.5% 0.87 Average Age (years) 69 64 0.01 SMOKING STATUS 0.36 Current / Previous Smoker 81.9% 88.2% Never Smoked 18.1% 11.8% GENERAL MEDICAL HISTORY Diabetes 43.1% 44.7% Myocardial Infarction 22.2% 25.0% 0.70 Hypertension 87.5% 76.3% 0.09 Dyslipidemia 70.8% 77.6% Family History of CAD 60.0% 0.05 COPD 12.5% 7.9% 0.42 Previous PTA on Study Limb 6.9% 5.3% 0.74 Follow-up available for 120 patients (81%)

Lesion Characteristics: Diffuse and complex GORE VIABAHN® Endoprosthesis Bare Nitinol Stent p-value TREATED OCCLUSIONS 59.7% 56.6% 0.74 TARGET LESION LENGTH (cm) 0.87 Mean (Std Dev) 19 (8) 18 (7) Median (Range) 20 (8 – 40) 16 (8 – 36) LESION CALCIFICATION 0.01 None – Mild 37.5% 57.9% Moderate – Severe 62.5% 42.1% TIBIAL RUNOFF 0.10 1 Vessel 15.3% 22.4% 2 Vessel 50.0% 32.9% 3 Vessel 34.7% 44.7%

GORE VIABAHN® Endoprosthesis SAFETY Study Endpoint 30-DAY SAFETY OUTCOMES GORE VIABAHN® Endoprosthesis Bare Nitinol Stent NUMBER OF ENROLLED SUBJECTS 72 76 PROCEDURE-RELATED SAFETY ISSUES Access / Treatment Site Complications 1* OTHER SAFETY EVENTS Death Myocardial Infarction Acute Renal Insufficiency Study Limb Amputation * The single access site complication was thrombosis in the non-treatment leg, that was resolved with subsequent treatment.

GORE VIABAHN® Endoprosthesis One-Year Interim Data INTERIM ONE-YEAR EFFICACY OUTCOMES GORE VIABAHN® Endoprosthesis Bare Nitinol Stent p-value Technical Success 97% 1.00 Primary Patency (PSVR 2.5) 53% 58% 0.58 Freedom from TLR 73% 69% 0.69 Assisted Primary Patency 84% 91% 0.41 Secondary Patency 93% 98% 0.19

GORE VIABAHN® Endoprosthesis Acute Limb Ischemia Acute Limb Ischemia GORE VIABAHN® Endoprosthesis Bare Nitinol Stent p-value DEVICE OCCLUSION 9 6 0.18 2 4 0.14 Chronic Limb Ischemia 7 Intervention < 24 hours 1 0.09 Intervention > 24 hours 0.04 Investigator defined

Clinical Success Quality of Life Questionnaires Significant improvement in IC and SF-36 physical summary scores at one year No significant difference between treatment groups

Patterns of Restenosis GORE VIABAHN® Endoprosthesis Focal edge stenosis comprised 87% GORE VIABAHN® Endoprosthesis failures 50% isolated proximal edge 30% both proximal and distal edges 6% isolated distal edge 14% undetermined Bare Nitinol Stent In-stent stenosis comprised 93% bare nitinol stent failures

Representative Images of Restenosis

Observations of Restenosis Patterns

Observations of Restenosis Patterns

Clinical Implications of Restenosis Patterns Flow Dynamics Calculations Assuming 80% focal (5 mm) versus diffuse (10 cm) restenosis pattern Pressure Drop in diffuse, in-stent stenosis is an order of magnitude higher than a focal edge stenosis Flow Rate through diffuse, in-stent stenosis is reduced by 50% as compared to flow through an edge stenosis

GORE VIABAHN® Endoprosthesis Stent Fracture GORE VIABAHN® Endoprosthesis Bare Nitinol Stent p-value STENT FRACTURE 1/47 (2%) 16/52 ( 30.8%) < 0.01 FRACTURE SEVERITY* Grade 1 1 7 Grade 2 Grade 3 3 Grade 4 2 * Not mutually exclusive

Stent Fracture Bare nitinol stents experienced significantly higher fracture rate Bare stent fractures independent of stent type 11/37 Tubular Stent fractures (30%) 6/17 Helical Stent fractures (35%) Fracture occurred predominantly in lesions longer than 15 cm In long lesions (> 15 cm), fractured stents were associated with a 17% lower patency rate and 25% higher TLR rate compared to unfractured stents (p = 0.40 and 0.17, respectively)

Representative Stent Fracture Images

Addressing the future VIPER (New Viabahn generation trial) Bare Stents Heparin Bioactive Surface “Contoured” end—potentially improved flow dynamics on the proximal end Bare Stents Is it back to the drawing board for stent designs for the treatment of longer lesions? We really do not have a bare metal solution for long SFA disease DOES A FRACTURE RATE OF > 30% MAKE BARE METAL STENTING A NON-STARTER, AT LEAST FOR LESIONS > 15 CM IN LENGTH?

Take Home Points VIBRANT: Multicenter; core lab controlled; real world trial of the highest femoropopliteal disease burden ever performed Treating claudicants with long diffuse disease is clinically safe and clinically effective at one year Stent grafts restenosed at edge; bare metal stents restenosed diffusely Primary patencies suboptimal; primary-assisted and secondary patencies are good even in diffuse lesions but strategies need to be refined Even helical stents appear to have high fracture rates in long lesions and these are associated with vessel restenosis Viabahn fracture is a rare event Surveillance appears to be important at achieving high assisted and secondary patency to over 80% at 1-year Longer term follow-up continues; primary end point will be completed 2011