1. REsynchronization reVErses Remodeling in Systolic left vEntricular dysfunction: Results of the REVERSE Trial Cecilia Linde, Stockholm, Sweden William T. Abraham, Columbus, U.S Michael R. Gold, Charleston, U.S. Jean-Claude Daubert, Rennes, France On Behalf of the REVERSE Investigators and Coordinators

2. Acknowledgments Steering Committee W. T. Abraham, J-C. Daubert (study initiator), C. Linde (coordinating clinical Investigator), M. Gold Echo Core Labs Ghio, S, St. John Sutton, MG Adverse Events Advisory Committee D. Böcker, J. P. Boehmer, J. G. F. Cleland, M. Gold, J. T. Heywood, A. Miller (chair) Data Monitoring Committee J. Aranda, J. Cohn (chair), P. Grambsch; M. KomajdaInvestigators Austria: H. Mayr, A. Teubl; Belgium: R. Willems; Canada: C. Simpson; Czech Republic: J. Lukl; Denmark: H. Eiskjær, C. Hassager, M. Møller, T. Vesterlund; France: E. Aliot, P. Chevalier, J-C. Daubert, J-M. Davy, P. Djiane, H. Le Marec; Germany: G. Groth, G. Klein, T. Lawo, C. Reithmann; Hungary: T. Forster, T. Szili-Török; Ireland: R. Sheahan; Italy: S. Lombroso, M. Lunati, L. Padeletti, M. Santini; Netherlands: B. Dijkman; Norway: S. Færestrand, F. T. Gjestvang; Spain: I. Fernandez Lozano, R. Muñoz Aguilera, A. Quesada Dorador; Sweden: C. Linde, F. Maru, K. Säfström; United Kingdom: G. Goode; United States: U. Birgersdotter-Green, J. Boehmer, E. Chung, S. Compton, J. Dinerman, D. Feldman, R. Fishel, G. J. Gallinghouse, M. Gold, S. Hankins, J. Herre, M. Hess, E. Horn, S. Hsu, S. Hustead, S. Jennison, E. Johnson, W. B. Johnson, G. Jones, R. Malik, A. Merliss, S. Mester, S. Moore, N. Nasir, F. Pelosi, Jr., D. Renlund, K. Rist, R. Sangrigoli, R. Silverman, D. Smull, K. Stein, L. Stevenson, J. Stone, N. Sweitzer, D. Venesy, L. Zaman.Sponsor Medtronic Inc.

3. Cecilia Linde, MD, PhD The following relationships exist related to this presentation: Presenter Disclosure Information Consulting Fees, Medtronic and St. Jude, moderate level Research Grants, Medtronic and the Sweden Heart and Lung Foundation, significant level

4. To determine the effects of CRT with or without an ICD on disease progression over 12 months in patients with asymptomatic and mildly symptomatic heart failure and ventricular dysynchrony Randomized, double-blind, parallel-controlled clinical trial Purpose and Design

5. NYHA Class II or I (previously symptomatic) QRS  120 ms; LVEF  40%; LVEDD  55 mm Optimal medical therapy (OMT) Without permanent cardiac pacing With or without an ICD indication Inclusion Criteria

6. Baseline Assessment Successful CRT Implant Randomized 1:2 CRT OFF (OMT ± ICD) CRT ON (OMT ± ICD) U.S., Canada: at 12 Months, all patients recommended CRT ON Europe: at 24 Months, all patients recommended CRT ON 1 2 12 Months Study Schematic

7. Primary: HF Clinical Composite Response, comparing the proportion of patients worsened in CRT OFF vs. CRT ON groups Composite includes: all-cause mortality, HF hospitalizations, crossover due to worsening HF, NYHA class, and the patient global assessment assessed in double blind manner Prospectively Powered Secondary: Left Ventricular End Systolic Volume Index (LVESVi) comparing CRT OFF vs. CRT ON subjects LVESVi is assessed by two core labs (1 in Europe, 1 in U.S) End Points

8. 684 Enrolled (2004-2006) 642 Implant Attempts 610 Patients Randomized U.S. 343 (56%); Europe 262 (43%); Canada 5 (<1%) CRT OFF 191 PatientsCRT ON 419 Patients - 594/598 completed 12 month follow-up - 12 deaths (2%) - 0 lost to follow-up, 0 exits -21 unsuccessful implant 621 Successful CRT Implants (97%) - 42 ineligible or withdrew -11 exits after successful implant Enrollment and Randomization

9. CRT OFF N=191 CRT ON N=419 P-value Age (mean) yrs 61.8 ± 11.662.9 ± 10.60.26 NYHA II 83%82%0.82 ICD 85%82%0.41 Beta-blockers94% 96%0.32 ACE-i/ ARB 97% 96%0.63 Diuretics 77% 81%0.33 EF26.4 ± 7.026.8 ± 7.00.50 LVEDD (mm)70 ± 9 69 ± 90.34 QRS (ms)154.4 ± 24.1152.8 ± 21.00.41 Ischemic51% 56%0.22 Baseline Characteristics of Randomized Cohort (n=610)

10. 40% 54% 39% 30% 16% 21% 0% 20% 40% 60% 80% 100% CRT OFFCRT ON Improved / Unchanged Pre-Specified Analysis Proportion Worsened Conventional Analysis Distribution Worsened/Unchanged /Improved Worsened Unchanged Improved P=0.004 Primary End Point: Clinical Composite Response 79% 84% 16% 21% 0% 20% 40% 60% 80% 100% CRT OFFCRT ON P=0.10 Worsened

11. 70 75 80 85 90 95 100 105 110 115 Baseline12 Months LVESVi (ml/m 2 ) CRT OFF  = -1.3 CRT ON  = -18.4 P<0.0001 n=487 Powered Secondary End Point: LVESVi (ml/m 2 )

12. 12 MonthsBaseline (ml/m 2 ) LVEDVi (ml/m 2 ) P<0.0001 LVEF (%) P<0.0001 12 MonthsBaseline CRT OFF ∆ = 0.6 CRT ON ∆ = 3.8 CRT OFF ∆ = -1.4 CRT ON ∆ = -20.5 n=487 20 22 24 26 28 30 32 34 90 100 110 120 130 140 150 Other Remodeling Parameters

13. MN LWHF P=0.26 6-Min Walk Test P=0.26 NYHA P=0.06 Other Secondary Endpoints

14. % of Patients Hospitalized for HF Number at Risk CRT OFF 191 187 181 176 119 CRT ON 419 415 411 409 251 P=0.03 Hazard Ratio=0.47 CRT OFF CRT ON Months Since Randomization Time to First HF Hospitalization

15. 97% implant success rate 9.5 % LV-lead related complications 66 in 59 / 621 successfully implanted patients LV lead dislodgements, diaphragmatic stimulation, subclavian vein thrombosis, etc. Safety

16. REVERSE is the first large, randomized, double-blind study to show that CRT in asymptomatic and mildly symptomatic heart failure patients on optimal medical therapy: Reverses LV remodeling Reduces the risk of heart failure hospitalization May improve clinical outcome as assessed by the clinical composite response measure Note: FDA has not yet reviewed the clinical data to determine whether or not CRT systems are safe and effective in this patient population. Conclusion

17. Backup Slides

18. *Note: Data in table is listed in hierarchical order (subjects are included only in one sub-category) Worsened Improved Clinical Composite Response Details

19. ComplicationNumber of Events LV lead dislodgement42 RV lead dislodgement15 Inappropriate device irritation of tissue14 RA lead dislodgement10 Failure to capture6 Implant site hematoma5 Atrial fibrillation5 Most Common Procedure Related Complications

20. Missing LVESVi CRT OFF (n=191) CRT ON (n=419) Total (n=610) No baseline, full echo not done No baseline, full echo not done 1 (1%) 2 (<1%) 3 (<1%) No baseline, LVESV on echo not readable No baseline, LVESV on echo not readable 11 (6%) 31 (7%) 42 (7%) Died before 12 months Died before 12 months 2 (1%) 9 (2%) 11 (2%) Missed 12-month follow-up visit Missed 12-month follow-up visit 0 (0%) 3 (1%) 3 (<1%) No 12-month, full echo not done No 12-month, full echo not done 1 (1%) --- 1 (<1%) No 12-month, limited echo not done No 12-month, limited echo not done--- 10 (2%) No 12-month, LVESV on echo not readable No 12-month, LVESV on echo not readable 13 (7%) 40 (10%) 53 (9%) Total 28 (15%) 95 (23%) 123 (20%) Note: Subjects are included in the first sub-category, reading down Reasons for Missing Paired LVESVi data

21. Causes of Death Group Days Post implantTypeCause of Death ON6 SCD Medical cause: unknown ON24 Other Gastrointestinal bleed ON100 HF Progressive heart failure ON107 SCD Tachy-arrhythmia (ventricular fibrillation) ON110 Other Medical cause: unknown ON196 SCD & HF Brady-arrhythmic event (asystole) ON300 HF Progressive heart failure, alcoholic cardiomyopathy ON343 Other Respiratory failure due to pulmonary fibrosis OFF254 Other Renal cancer OFF287 HF Progressive heart failure OFF344 SCD Electromechanical dissociation (preliminary)

22. LVESVi (ml/m 2 ) by NYHA Class