1. Impella Technology Elective Support Clinical Evidence and Investigations

2. PROTECT II O’Neill et al, Circulation. 2012;126(14):1717-27

3. PROTECT II Trial Design IMPELLA 2.5 + PCI IMPELLA 2.5 + PCI IABP + PCI IABP + PCI Primary Endpoint = 30-day Composite MAE* rate 1:1 R Patients Requiring Prophylactic Hemodynamic Support During Non-Emergent High Risk PCI on Unprotected LM/Last Patent Conduit and LVEF≤35% OR 3 Vessel Disease and LVEF≤30% Patients Requiring Prophylactic Hemodynamic Support During Non-Emergent High Risk PCI on Unprotected LM/Last Patent Conduit and LVEF≤35% OR 3 Vessel Disease and LVEF≤30% Follow-up of the Composite MAE* rate at 90 days *Major Adverse Events (MAE) : Death, MI (>3xULN CK-MB or Troponin), Stroke/TIA, Repeat Revasc, Cardiac or Vascular Operation or Vasc. Operation for limb ischemia, Acute Renal Dysfunction, Increase in Aortic insufficiency, Severe Hypotension, CPR/VT, Angio Failure O’Neill et al, Circulation. 2012;126(14):1717-27

4. 3 Baseline Patient Characteristics Patient Characteristics IABP () (N=223)Impella () (N=225)p-value Age67±1168±110.488 Gender-Male81.2%80.0%0.668 History of CHF83.4%91.1%0.014 Current NYHA (Class III / IV)64.6%67.0%0.632 Diabetes Mellitus50.7%52.0%0.779 Renal insufficiency30.2%23.1%0.091 Peripheral Vascular Disease26.5%25.7%0.851 Implantable Cardiac Defib.31.1%34.7%0.420 Prior CABG28.7%38.2%0.033 LVEF24.1±6.3%23.4±6.3%0.244 STS Mortality score6±7%6±6%0.809 Not Surgical Candidate64.6%63.6%0.822 SYNTAX score29.3±13.530.3±13.10.514 O’Neill et al, Circulation. 2012;126(14):1717-27

5. 4 Hemodynamic Support Effectiveness Cardiac Power Output Maximal Decrease in CPO on device Support from Baseline (in x0.01 Watts) IABP Impella N=138N=141 - 4.2 ± 24 - 14.2 ± 27 p=0.001 O’Neill et al, Circulation. 2012;126(14):1717-27

6. Procedural Characteristics IABP(N=223)Impella(N=225) p- value Use of Heparin83.3%93.3% <0.001 IIb/IIIa Inhibitors26.0%13.8% 0.001 Total Contrast Media (cc)241±114267±142 0.036 Rotational Atherectomy (RA)9.0%14.2% 0.083 Median # of RA Passes/lesion (IQ range) 1 (1-2)3 (2-5) 0.001 Median # of RA passes/pt (IQ range) 2.0 (2.0-4.0)5.0 (3.5-9.5) 0.003 Median RA time/lesion (IQ range sec) 40 (20-47)60 (40-118) 0.004 RA of Left Main Artery3.1%8.0% 0.024 Total Support Time (hours)8.4±21.81.9±2.7 <0.001 Discharge from Cath Lab on device36.7%5.9% <0.001 O’Neill et al, Circulation. 2012;126(14):1717-27

7. 6 LVEF and NYHA Improvement Post PCI LVEF (%) p<0.001Baseline 90 days 22% O’Neill et al, Circulation. 2012;126(14):1717-27 Baseline 90 days NYHA Class Distributionp<0.001 Class I Class II Class III Class IV 58% reduction in Class III,IV

8. 7 PROTECT II: Per Protocol MAE (N=427) IABP IMPELLA p=0.092 N=216 N=211p=0.023N=215N=210 21 (12/9) Patients Excluded to Due Not Meeting Inclusion/Exclusion Criteria Prior to Analysis 4 LVEF > 35% 3Active MI 4No Left Main or 3 V CAD 10 Other Exclusions Log rank test, p=0.048 IABP IMPELLA O’Neill et al, Circulation. 2012;126(14):1717-27

9. 8 Pre-Specified Analyses

10. 90 day MAE Relative Risk [95% CI] Group p-value Interaction p-value 0.79 [0.64, 0.97] 0.023 0.70 [0.55, 0.89] 0.003 1.19 [0.75, 1.91] 0.444 0.82 [0.53, 1.25] 0.351 0.78 [0.61, 0.99] 0.039 1.14 [0.75, 1.71] 0.540 0.71 [0.56, 0.91] 0.006 0.92 [0.62, 1.38] 0.697 0.74 [0.58, 0.95] 0.016 Pre-Specified Sub-group Analysis (PP) With Atherectomy (n=52) Without Atherectomy (n=373) STS ≥ 10 (n=71) STS < 10 (n=354) 1 st Impella/IABP Pt per site (n=116) After 1 st Impella/IABP Pt (n=309) ULM / Last conduit (n=101) 3VD (n=324) Anatomy PCI Procedure STS Mortality Score Roll in subject Overall – Per Protocol (n=425) Impella betterIABP better 0.00.51.01.52.0 0.087 0.845 0.092 0.348 PP= Per Protocol O’Neill et al, Circulation. 2012;126(14):1717-27

11. 10 O’Neill et al. - Summary The use of Impella during high risk PCI was safe and provided better hemodynamic support compared to IABP The outcomes in the Impella arm demonstrated a trend towards a reduction of Major Adverse Events (MAE) at 30 days and a significant reduction of the MAE rate at 90 day follow-up in the per protocol population. Additional upcoming evidence should provide more insight on the potential benefit of Impella

12. Dangas et al, Am. Journ of Cardiol. 2014 Jan 15;113(2):222-8

13. 12Background 1 Moussa et al. JACC 2013; 2 Ioannidis et al, JACC, 2003; 3 Kini et al, JACC, 1999; 4 Stone et al, Circulation, 2001 The significance of the clinical importance of small elevations of markers of myocardial injury post-PCI is subject of debateThe significance of the clinical importance of small elevations of markers of myocardial injury post-PCI is subject of debate Despite an overall statistical association, it has been recognized that small to medium level elevation has a benign clinical course 2,3Despite an overall statistical association, it has been recognized that small to medium level elevation has a benign clinical course 2,3 A prognostic relationship with subsequent mortality exists when new Q-wave infarction or a CK-MB elevation above eight times the upper normal value are detected 4A prognostic relationship with subsequent mortality exists when new Q-wave infarction or a CK-MB elevation above eight times the upper normal value are detected 4 We therefore investigated the performance of the Impella 2.5 vs the IABP in the PROTECT-II trial using prognostically important outcomes (periprocedural MI CKMB rise >8x ULN)We therefore investigated the performance of the Impella 2.5 vs the IABP in the PROTECT-II trial using prognostically important outcomes (periprocedural MI CKMB rise >8x ULN)

14. 13 PROTECT II MACCE* Log rank test, p=0.042 *Death, Stroke, Myocardial Infarction, Repeat revascularization IABP IMPELLA Dangas et al, Am. Journ of Cardiol. 2014 Jan 15;113(2):222-8

15. 14 PROTECT II Results Per-Protocol Treated Population Intention-To-Treat population Impella 2.5 (n=216) IABP (n=211) P- value Impella 2.5 (n=224) IABP (n=219) P- value 30-day event rates Major Adverse Events (MAE) 30%40%0.0431%38%0.11 Major Adverse Cardiac and Cerebrovascular Events (MACCE) 14%20%0.1015%19%0.23 Myocardial Infarction 5.6%7.1%0.515.8%6.8%0.67 90-day event rates Major Adverse Events (MAE) 37%49%0.0137%47%0.03 Major Adverse Cardiac and Cerebrovascular Events (MACCE) 22%31%0.0322%30%0.06 Myocardial Infarction 5.6%11%0.045.8%10.5%0.07 Dangas et al, Am. Journ of Cardiol. 2014 Jan 15;113(2):222-8

16. 15 Multivariate analysis predictors of MAE and MACCE at 90 days Odds Ratio Estimate 95% confidence intervalP-Value MAE Intention-To-treat Population Use of atherectomy during PCI 1.41 1.03 – 1.920.03 Renal Insufficiency 1.4 1.11 – 1.760.005 Device: IMPELLA 0.79 0.64 - 0.960.02 Per-Protocol Population Use of atherectomy during PCI 1.41 1.03 - 1.930.03 Renal Insufficiency 1.37 1.08-1.740.01 Device: IMPELLA 0.75 0.61 - 0.920.007 MACCE Intention-To-treat Population Device: IMPELLA 0.77 0.61 - 0.960.02 Per-Protocol Population Device: IMPELLA 0.8 0.64 - 0.990.042 Dangas et al, Am. Journ of Cardiol. 2014 Jan 15;113(2):222-8

17. 16 Dangas et al. - Summary Hemodynamic support with Impella in comparison to IABP during high risk PCI in the PROTECT-II trial resulted in improved event-free survival at 3-month follow-up This finding was further supported by multivariate analyses Dangas et al, Am. Journ of Cardiol. 2014 Jan 15;113(2):222-8

18. Cohen et al, Catheter Cardiovasc Interv. 2013 (In press)

19. 18 PROTECT II MAE Outcome Pre-specified High Risk PCI Without Atherectomy Group MAE= Major Adverse Event Rate ↓ 30% MAEp=0.003N=183N=190 p=0.01N=184N=191 IMPELLA IABP Log rank test, p=0.005 Per Protocol (N=374) Cohen et al, Catheter Cardiovasc Interv. 2013 (In press)

20. 19 Cohen et al. - Summary Operators tended to use RA to a larger extent in patients supported with Impella in comparison with IABP Accompanying this differential RA use we observed less frequent repeat revascularization counterbalanced by a higher frequency of cardiac enzyme elevation >3x ULN in Impella-treated patients Although our results suggest that a more aggressive RA use in patients treated with Impella is not associated with less favorable outcomes and potentially reduces the repeat revascularization rate in comparison with IABP Cohen et al, Catheter Cardiovasc Interv. 2013 (In press)

21. Henriques et al, Am Heart J. 2013 PROTECT II

22. 21 MAE= Major Adverse Event Rate N=83 N=81 N=64 N=63N=68N=65 IABP IMPELLA Study Device Learning Curve Effect Per Protocol Population 90day Outcome Henriques et al, Am Heart J. 2013

23. Study Device Learning Curve Effect Henriques et al, Am Heart J. 2013 First patient Remaining patients

24. 23 Henriques et al. - Summary Significantly lower 90-day rates of MAE were observed with use of the Impella 2.5 compared to the IABP after exclusion of the first patient at each site This prespecified analysis is suggestive of a learning curve associated with use of the Impella 2.5 during its initial introductory period affecting the outcome of the study This finding likely applies to other new medical devices. Clinical trials should therefore specifically address the training aspect of new devices Henriques et al, Am Heart J. 2013