COAPT A Randomized Trial of Transcatheter Mitral Valve Leaflet Approximation in Patients - PowerPoint Presentation

1. COAPTA Randomized Trial of Transcatheter Mitral Valve Leaflet Approximation in Patients with Heart Failure and Secondary Mitral Regurgitation Gregg W. Stone, MDOn behalf of Michael Mack, William Abraham, JoAnn Lindenfeld and the COAPT Investigators

2. Gregg W. Stone MDConsulting fees from Neovasc, Valfix, GoreEquity/options in Ancora Institutional conflictColumbia University, my employer, receives royalties from Abbott for sale of the MitraClip Disclosure Statement

3. Background (i)Pts with heart failure (HF) in whom mitral regurgitation (MR) develops secondary to left ventricular dysfunction have a poor prognosis, with reduced quality-of-life, frequent hospitalizations for heart failure and decreased survival There are no proven therapies for secondary MR in HFGuideline-directed medical therapy (GDMT) and cardiac resynchronization therapy (CRT) may provide symptomatic relief in some ptsWhether correcting secondary MR improves the prognosis of pts with HF is unknown Surgery with a downsized annuloplasty ring has not been demonstrated to be beneficial for secondary MR, and has a high recurrence rate

4. Background (ii)By approximating the anterior and posterior mitral leaflets and forming a double-orifice valve, the MitraClip device reduces MR Registries have suggested that the MitraClip is safe and may provide symptomatic benefit to HF pts with secondary MRWe therefore performed the COAPT randomized trial to evaluate the safety and effectiveness of transcatheter mitral leaflet approximation in HF pts with secondary MR who remained symptomatic despite GDMT

5. The COAPT TrialCardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure Patients with Functional Mitral RegurgitationA parallel-controlled, open-label, multicenter trial in ~610 patients with heart failure and moderate-to-severe (3+) or severe (4+) secondary MR who remained symptomatic despite maximally-tolerated GDMTRandomize 1:1*GDMT aloneN=305MitraClip + GDMTN=305*Stratified by cardiomyopathy etiology (ischemic vs. non-ischemic) and site

6. Key Inclusion CriteriaIschemic or non-ischemic cardiomyopathy with LVEF 20%-50% and LVESD ≤70 mmModerate-to-severe (3+) or severe (4+) secondary MR confirmed by an independent echo core laboratory prior to enrollment (US ASE criteria)NYHA functional class II-IVa (ambulatory) despite a stable maximally-tolerated GDMT regimen and CRT (if appropriate) per societal guidelinesPt has had at least one HF hospitalization within 12 months and/or a BNP ≥300 pg/ml* or a NT-proBNP ≥1500 pg/ml* Not appropriate for mitral valve surgery by local heart team assessmentIC believes secondary MR can be successfully treated by the MitraClip Adjusted by a 4% reduction in the BNP or NT-proBNP cutoff for every increase of 1 kg/m2 in BMI >20 kg/m2

7. Key Exclusion CriteriaACC/AHA stage D HF, hemodynamic instability or cardiogenic shock Untreated clinically significant CAD requiring revascularizationCOPD requiring continuous home oxygen or chronic oral steroid use Severe pulmonary hypertension or moderate or severe right ventricular dysfunction Aortic or tricuspid valve disease requiring surgery or transcatheter interventionMitral valve orifice area <4.0 cm2 by site-assessed TTELife expectancy <12 months due to non-cardiac conditions

8. Central Echo Core Lab and Eligibility Committee ReviewA Central Echo Core Lab confirmed the presence of 3+ - 4+ secondary MR Potentially eligible pts were then presented by the local site investigators on weekly calls to a Central Eligibility Committee consisting of at a minimum a heart failure specialist and expert mitral valve surgeonThe CEC confirmed that all eligibility criteria were met, especially 1) use of maximally-tolerated GDMT for heart failure, and treatment with CRT, defibrillators and revascularization if appropriate, and that 2) mitral valve surgery was not considered appropriate at the treating center and would not be offered to the patient, even if randomized to controlPts not meeting these criteria were rejected, or in some cases were deferred and could be re-presented after suitable GDMT had been instituted if the pt remained symptomatic and repeat echo still showed 3+-4+ MR

9. Primary Endpoints*Analyzed when the last subject completes 12 months of follow-up; **Objective performance goalPrimary effectiveness endpoint: All HF hospitalizations through 24 months*Powered for superiority of the Device group compared with the Control groupPrimary safety endpoint: Freedom at 12 mos from device-related complications:- Single leaflet device attachment- Device embolization- Endocarditis requiring surgery- Echo core laboratory-confirmed mitral stenosis requiring surgery- Left ventricular assist device implant- Heart transplant- Any device-related complication requiring non-elective cardiovascular surgeryPowered for superiority of the Device group vs. a pre-specified OPG**

10. Primary Effectiveness EndpointAnalyzed using a joint frailty model to account for the competing risk of deathSample Size and Power AnalysisAssumptionsAnnualized HF hosp rates: 0.60 per pt-yr Control vs. 0.42 per pt-yr Device12-month mortality rates: 27% Control vs. 22% Device12-month attrition rate: 7.5%Power610 randomized pts provided 80% power at a 1-sided α of 0.05 to demonstrate superiority of the Device group compared with the Control group for the 24-month rate of all HF hospitalizationsPrimary Safety Endpoint 305 pts in the Device group provided >95% power to demonstrate that freedom from device-related complications at 12 months is more than a pre-specified objective performance goal of 88% at a one-sided α of 0.05

11. Powered Secondary Endpoints1All powered for superiority unless otherwise noted; 2Powered for noninferiority of the device vs. the control group; 3Powered for noninferiority against an objective performance goal- Tested in hierarchical order1 -MR grade 2+ at 12 months All-cause mortality at 12 months2 Death and all HF hospitalization through 24 months (Finkelstein-Schoenfeld and win ratio analysis)Change in QOL (KCCQ) from baseline to 12 monthsChange in 6MWD from baseline to 12 months All-cause hospitalizations through 24 months NYHA class I or II at 12 months Change in LVEDV from baseline to 12 months All-cause mortality at 24 monthsDeath, stroke, MI, or non-elective CV surgery for device-related compls at 30 days3

12. Study LeadershipPrincipal InvestigatorsMichael J. Mack, MD, Baylor Scott & White Heart Hospital Plano, Plano, TX Gregg W. Stone, MD, Columbia University Medical Center, NY, NYHeart Failure Co-Principal InvestigatorsWilliam T. Abraham, MD, Ohio State University, Columbus, OH JoAnn Lindenfeld, MD, Vanderbilt Heart and Vascular Institute, Nashville, TN Steering CommitteeGregg W. Stone, Michael J. Mack, JoAnn Lindenfeld, William T. Abraham, Steven F. Bolling, Ted E. Feldman, Paul A. Grayburn, Samir R. Kapadia, Patrick M. McCarthyCentral Eligibility CommitteeGregg Stone, Paul Grayburn, Scott Lim, Michael Zile, James Udelson, William Abraham, JoAnn Lindenfeld, Rakesh Suri, James Gammie, Marc Gillinov, Steve Bolling, Patrick McCarthy, Donald Glower, David HeimansohnClinical Events CommitteeCardiovascular Research Foundation, New York, NY; Steven O. Marx, MD, chairEchocardiographic Core LaboratoryMedStar Health Research Institute, Hyattsville, MD; Neil J. Weissman, MD, directorCost-effectiveness and quality-of-life assessment Saint Luke’s Mid America Heart Institute, KC, MO; David J. Cohen, MD, directorSponsorAbbott, Santa Clara, CA

13. Study Flow and Follow-up1576 pts with HF and MR considered for enrollment between September 25th, 2012 and June 23th, 2017 at 89 centers in the US and Canada MitraClip + GDMTN=302GDMT aloneN=312Reasons for exclusionInadequate MR or DMR (n=244)Not treated with GDMT (n=79)All inclusion criteria not met (n=85)Exclusion criteria present (n=34)Echo criteria not met (n=255)Incomplete screeningor other (n=419)RandomizedN=614 at 78 sitesIneligibleN=911Roll-in casesN=51 at 34 sitesEligible for enrollment N=665

14. Study Flow and Follow-upN=293N=9N=1N=311Initial treatment MitraClipGDMT alone30-day follow-up1-year follow-up(primary endpoint minimum)2-year follow-up(eligible patients)N=298/302 (98.7%)N=309/312 (99.0%)Withdrew 3Lost to follow-up 13 Withdrew0 Lost to follow-up N=295/302 (97.7%)N=294/312 (94.2%)N=231/244 (94.7%)N=232/258 (89.9%)Withdrew 6Lost to follow-up 118 Withdrew0 Lost to follow-up Withdrew 10Lost to follow-up 323 Withdrew3 Lost to follow-up MitraClip + GDMTN=302GDMT aloneN=312

15. Top 10 Enrolling Sites1. Saibal KarCedars-Sinai Medical Center, Los Angeles, CA n=462. Scott LimUniversity of Virginia, Charlottesville, VA n=303. Jacob MishellKaiser Permanente, San Francisco, CA n=294. Brian WhisenantIntermountain Medical Center, Murray, UT n=265. Paul GrayburnBaylor Heart and Vascular Hospital, Dallas, TX n=256. Andreas BriekeUniversity Of Colorado Hospital, Aurora, CO n=176. Michael RinaldiCarolinas Medical Center, Charlotte, NCn=176. Samir KapadiaCleveland Clinic, Cleveland, OHn=176. Ian SarembockThe Christ Hospital, Cincinnati, OH n=176. Vivek RajagopalPiedmont Hospital, Atlanta, GA n=17

16. Baseline Characteristics (i)MitraClip +GDMT (N=302)GDMT alone(N=312)MitraClip + GDMT (N=302)GDMT alone(N=312)Age (years)71.7 ± 11.872.8 ± 10.5BMI (kg/m2)27.0 ± 5.827.1 ± 5.9Male 66.6% 61.5%CrCl (ml/min)50.9 ± 28.547.8 ± 25.0Diabetes 35.1% 39.4% - ≤60 ml/min 71.6% 75.2%Hypertension 80.5% 80.4%Anemia (WHO) 59.8% 62.7%Hyperchol. 55.0% 52.2% BNP (pg/mL)1015 ± 1086 1017 ± 1219 Prior MI 51.7% 51.3% NT-proBNP (pg/mL)5174 ± 6567 5944 ± 8438 Prior PCI 43.0% 49.0%STS replacement sc7.8 ± 5.58.5 ± 6.2Prior CABG 40.1% 40.4% - ≥8 41.7% 43.6%Prior stroke or TIA 18.5% 15.7%Surgical risk (central eligibility committee)PVD 17.2% 18.3% - High* 68.6% 69.9%COPD 23.5% 23.1% - Not-high 31.4% 30.1%H/o atrial fibr 57.3% 53.2%* STS repl score ≥8% or one or more factors present predicting extremely high surgical risk

17. Baseline Characteristics (ii)HF parametersMitraClip +GDMT (N=302)GDMT alone(N=312)Echo core labMitraClip + GDMT (N=302)GDMT alone(N=312)Etiology of HF  MR severity   - Ischemic60.9% 60.6% - Mod-to-sev (3+) 49.0% 55.3% - Non-ischemic 39.1% 39.4% - Severe (4+) 51.0% 44.7%NYHA class  EROA, cm20.41 ± 0.15 0.40 ± 0.15 - I 0.3% 0%LVESD, cm5.3 ± 0.9 5.3 ± 0.9 - II 42.7% 35.4%LVEDD, cm6.2 ± 0.7 6.2 ± 0.8 - III 51.0% 54.0%LVESV, mL135.5 ± 56.1 134.3 ± 60.3 - IV 6.0% 10.6%LVEDV, mL194.4 ± 69.2 191.0 ± 72.9 HF hosp w/i 1 year 58.3% 56.1%LVEF, %31.3 ± 9.1 31.3 ± 9.6 Prior CRT 38.1% 34.9% - 40%82.2%82.0%Prior defibrillator 30.1% 32.4%RVSP, mmHg44.0 ± 13.4 44.6 ± 14.0

18. Medication Use at BaselineMaximally-tolerated dosesMitraClip + GDMT (n=302)GDMT alone (n=312) Beta-blocker91.1%89.7%ACEI, ARB or ARNI71.5% 62.8%Mineralocorticoid receptor antagonist50.7% 49.7%Nitrates6.3%8.0%Hydralazine16.6%17.6%Diuretic89.4%88.8%Chronic oral anticoagulant46.4% 40.1%Aspirin57.6% 64.7%P2Y12 receptor inhibitor25.2% 22.8%Statin62.6% 60.6%

19. Major Changes in HF Meds w/i 12 MonthsMitraClip + GDMT (n=302)GDMT alone(n=312)P valueACEI, ARB or ARNI    - ⇓ dose by >50% or discontinue6.6% 4.8%0.33 - ⇑ dose by >100% or new drug started7.6% 7.4%0.91 Beta-blocker    - ⇓ dose by >50% or discontinue5.3% 5.1%0.92 - ⇑ dose by >100% or new drug started8.6% 3.8%0.01 Mineralocorticoid receptor antagonist    - ⇓ dose by >50% or discontinue0.7%0.6%1.00 - ⇑ dose by >100% or new drug started5.3%2.6%0.08 Nitrates    - ⇓ dose by >50% or discontinue0.0%0.0%1.00 - ⇑ dose by >100% or new drug started1.0%1.9%0.51 Hydralazine    - ⇓ dose by >50% or discontinue1.0%0.0%0.12 - ⇑ dose by >100% or new drug started4.3%3.8%0.77

20. MitraClip Procedure (n=302)MitraClip procedure attempted293/302 (97.0%)Clip implanted (MitraClip procedure attempted)287/293 (98.0%)Clip implanted (all patients)287/302 (95.0%)Mean # of clips implanted1.7  0.7 (n=293) - 0 clips implanted6 (2.0%) - 1 clip implanted106 (36.2%) - 2 clips implanted157 (53.6%) - 3 clips implanted23 (7.9%) - 4 clips implanted1 (0.3%)Procedure duration (mins)162.9  118.1 - Device procedure time (mins)118.9  63.5 - Device time (mins)82.7  80.8 - Fluoroscopy time (mins)33.9  23.2TTE at discharge(n=260)

21. Primary Effectiveness EndpointAll Hospitalizations for HF within 24 monthsHR (95% CI] =0.53 [0.40-0.70]P<0.00103691215182124501001502002503000MitraClip + GDMTGDMT alone160in 92 pts283in 151 ptsCumulativeHF Hospitalizations (n)Time After Randomization (Months)MitraClipGDMT30228626925323619117816112431229427124521917614512188No. at Risk:Median [25%, 75%] FU= 19.1 [11.9, 24.0] mos

22. Primary Effectiveness EndpointHospitalizations for HF within 24 monthsAnnualized rates of HF hospitalization**Joint frailty modelGDMTaloneMitraClip+ GDMTHR (95% UCL] =0.53 [0.66]P<0.001160/446.5 pt-yrs 283/416.8 pt-yrs NNT (24 mo) = 3.1 [95% CI 1.9, 8.2]

23. Primary Safety EndpointFreedom from Device-related Complications within 12 monthsMitraClip procedure attemptedN=293Device-related complications9 (3.4%) - Single leaflet device attachment2 (0.7%) - Device embolization1 (0.3%) - Endocarditis requiring surgery0 (0.0%) - Mitral stenosis requiring surgery0 (0.0%) - Left ventricular assist device implant3 (1.2%) - Heart transplant2 (0.8%) - Any device-related complication requiring non-elective CV surgery1 (0.3%)*KM estimate; **Calculated from Z test with Greenwood’s method of estimated variance against a pre-specified objective performance goal of 88% 96.6%*88% OPC94.8% [95% LCL]P<0.001

24. Powered Secondary Endpoints1All powered for superiority unless otherwise noted; 2Powered for noninferiority of the device vs. the control group; 3Powered for noninferiority against an objective performance goal- Tested in hierarchical order1 -P-value1. MR grade 2+ at 12 months 2. All-cause mortality at 12 months2 3. Death and all HF hospitalization through 24 months (Finkelstein-Schoenfeld)4. Change in QOL (KCCQ) from baseline to 12 months5. Change in 6MWD from baseline to 12 months 6. All-cause hospitalizations through 24 months 7. NYHA class I or II at 12 months 8. Change in LVEDV from baseline to 12 months 9. All-cause mortality at 24 months10. Death, stroke, MI, or non-elective CV surgery for device-related compls at 30 days3

25. Powered Secondary Endpoints1All powered for superiority unless otherwise noted; 2Powered for noninferiority of the device vs. the control group; 3Powered for noninferiority against an objective performance goal- Tested in hierarchical order1 -P-value1. MR grade 2+ at 12 months <0.0012. All-cause mortality at 12 months2 <0.0013. Death and all HF hospitalization through 24 months (Finkelstein-Schoenfeld)<0.0014. Change in QOL (KCCQ) from baseline to 12 months<0.0015. Change in 6MWD from baseline to 12 months <0.0016. All-cause hospitalizations through 24 months 0.037. NYHA class I or II at 12 months <0.0018. Change in LVEDV from baseline to 12 months 0.0039. All-cause mortality at 24 months<0.00110. Death, stroke, MI, or non-elective CV surgery for device-related compls at 30 days3<0.001

26. All-cause MortalityAll-cause Mortality (%) 0% 20% 40% 60% 80% 100%Time After Randomization (Months)0369121518212446.1%29.1%HR [95% CI] = 0.62 [0.46-0.82]P<0.001MitraClip + GDMTGDMT alone30228626925323619117816112431229427124521917614512188No. at Risk:MitraClip + GDMTGDMT aloneNNT (24 mo) =5.9 [95% CI 3.9, 11.7]

27. MitraClip + GDMTGDMT aloneAll-cause Mortality orHF Hospitalization (%) 0% 20% 40% 60% 80% 100%Time After Randomization (Months)0369121518212467.9%45.7%MitraClip + GDMTGDMT alone30226423821519415414512697312244205174153117907555No. at Risk:HR [95% CI] = 0.57 [0.45-0.71]P<0.001NNT (24 mo) =4.5 [95% CI 3.3, 7.2] Death or HF Hospitalization

28. 24-Month Death or HF Hospitalization0.130.760.790.540.790.410.690.290.57 [0.45, 0.71]0.47 [0.33, 0.66]0.54 [0.41, 0.71]0.54 [0.37, 0.78]0.53 [0.39, 0.71]0.59 [0.40, 0.86]0.56 [0.28, 1.12]0.51 [0.37, 0.70]0.51 [0.33, 0.80]0.62 [0.45, 0.83]67.9% (191)65.3% (91)73.0% (125)65.2% (75)67.4% (122)67.8% (65)84.4% (26)65.0% (103)58.7% (51)71.4% (91)45.7% (129)37.8% (51)47.1% (90)41.1% (45)42.9% (74)47.6% (43)68.3% (12)39.2% (64)35.8% (32)53.4% (78)All patients0.310.50 [0.39, 0.65]71.9% (157)44.2% (96)0.320.46 [0.33, 0.64]77.8% (99)46.4% (56)0.420.48 [0.34, 0.67]69.5% (92)41.5% (54)All patientsAge (median)SexEtiology of cardiomyopathyPrior CRTHF hospitalization within the prior yearBaseline NYHA classSTS replacement scoreSurgical risk status*Baseline MR gradeBaseline LVEF0.65 [0.48, 0.88]70.2% (100)52.1% (78)≥74 years (n=317)<74 years (n=297)0.60 [0.40, 0.89]59.4% (66)43.2% (39)Female (n=221)Male (n=393)0.57 [0.43, 0.76]70.0% (116)48.1% (84)Ischemic (n=373)Non-ischemic (n=241)0.62 [0.44, 0.89]68.4% (69)50.2% (55)Yes (n=224)No (n=390)0.56 [0.42, 0.73]67.9% (126)44.7% (86)Yes (n=407)No (n=207)0.56 [0.39, 0.81]66.9% (65)41.1% (50)I or II (n=240)0.920.61 [0.44, 0.83]65.3% (99)46.6% (67)III (n=322)IV (n=51)0.64 [0.46, 0.88]71.4% (88)54.1% (65)≥8% (n=262)<8% (n=352)0.58 [0.45, 0.75]71.5% (140)49.7% (95)High (n=423)Not high (n=188)0.48 [0.34, 0.67]65.3% (100)37.5% (51)3+ (n=320)4+ (n=293)0.67 [0.38, 1.17]56.2% (27)49.7% (22)>40% (n=103)≤40% (n=472)0.60 [0.43, 0.84]61.2% (85)44.1% (62)≥30% (median; n=301)<30% (median; n=274)Baseline LVEDV (median)0.58 [0.42, 0.80]68.0% (92)48.9% (43)≥181 mL (n=288)<181 mL (n=287)P [Int]HR [95% CI]GDMT aloneMitraClip + GDMTSubgroupHR [95% CI]0.20.511.52.5Favors MitraClip + GDMT Favors GDMT aloneKM time-to-first event rates*Central eligibility committee assessment

29. 24-Month Event Rates (i)Kaplan-Meier time-to-first event ratesMitraClip + GDMT (n=302)GDMT alone (n=312)HR [95% CI]P-valueDeath, all-cause29.1% 46.1% 0.62 [0.46, 0.82] <0.001 - CV23.5% 38.2% 0.59 [0.43, 0.81] <0.001 - HF-related12.0% 25.9% 0.43 [0.27, 0.67] <0.001 - Non-HF-related13.1%16.6%0.86 [0.54, 1.38]0.53 - Non-CV7.3% 12.7% 0.73 [0.40, 1.34] 0.31 Hospitalization, all-cause69.6% 81.8% 0.77 [0.64, 0.93] 0.01 - CV51.9% 66.5% 0.68 [0.54, 0.85] <0.001 - HF-related35.7% 56.7% 0.52 [0.40, 0.67] <0.001 - Non-HF-related29.4% 31.0% 0.98 [0.71, 1.36] 0.92 - Non-CV48.2% 52.9% 0.91 [0.71, 1.17] 0.47 Death or HF hospitalization45.7% 67.9% 0.57 [0.45, 0.71] <0.001

30. 24-Month Event Rates (ii)*Unplanned. Kaplan-Meier time-to-first event ratesMitraClip + GDMT (n=302)GDMT alone (n=312)HR [95% CI]P-valueMV intervention or surgery*4.0% 9.0% 0.61 [0.27, 1.36] 0.23 - MitraClip3.7% 6.6% 0.99 [0.38, 2.58] 0.99 - Mitral valve surgery0.4% 2.5% 0.14 [0.02, 1.17] 0.07 PCI or CABG2.8% 4.3% 0.62 [0.24, 1.60] 0.32 Stroke4.4% 5.1% 0.96 [0.42, 2.22] 0.93 Myocardial infarction4.7% 6.5% 0.82 [0.38, 1.78] 0.62 New CRT implant2.9% 3.3% 0.85 [0.31, 2.34] 0.75 LVAD or heart transplant4.4% 9.5% 0.37 [0.17, 0.81] 0.01 - LVAD3.0% 7.1% 0.34 [0.13, 0.87] 0.02 - Heart transplant1.4% 3.6% 0.35 [0.09, 1.32] 0.12

31. 24-Month Event Rates (ii)*Unplanned. Kaplan-Meier time-to-first event ratesMitraClip + GDMT (n=302)GDMT alone (n=312)HR [95% CI]P-valueMV intervention or surgery*4.0% 9.0% 0.61 [0.27, 1.36] 0.23 - MitraClip3.7% 6.6% 0.99 [0.38, 2.58] 0.99 - Mitral valve surgery0.4% 2.5% 0.14 [0.02, 1.17] 0.07 PCI or CABG2.8% 4.3% 0.62 [0.24, 1.60] 0.32 Stroke4.4% 5.1% 0.96 [0.42, 2.22] 0.93 Myocardial infarction4.7% 6.5% 0.82 [0.38, 1.78] 0.62 New CRT implant2.9% 3.3% 0.85 [0.31, 2.34] 0.75 LVAD or heart transplant4.4% 9.5% 0.37 [0.17, 0.81] 0.01 - LVAD3.0% 7.1% 0.34 [0.13, 0.87] 0.02 - Heart transplant1.4% 3.6% 0.35 [0.09, 1.32] 0.12

32. Change in KCCQ from Baseline to 12 Months±23.3±22.7Adjusted change**Ancova±1.9±1.8P<0.001n=236n=228n=236n=228±32.0±28.6

33. Change in 6MWD from Baseline to 12 Months±127.1±125.3Adjusted change**Ancova±9.0P<0.001n=229n=237n=229n=237±166.7±157.7±9.1

34. NYHA Functional ClassNYHA classIIIIIIIVHF deathPtrend I or IIP-value BaselineMitraClip (n=302)0.3%42.7%51.0%6.0%--43.0%-GDMT (n=311)0%35.4%54.0%10.6%-35.4%30 daysMitraClip (n=283)15.5%60.8%19.4%3.5%0.7%<0.00176.3%<0.001GDMT (n=281)5.0%42.7%41.6%9.6%1.1%47.7%6 monthsMitraClip (n=263)19.4%52.9%21.3%2.7%3.8%<0.00172.2%<0.001GDMT (n=261)5.4%44.8%38.3%2.7%8.8%50.2%12 monthsMitraClip (n=237)16.9%55.3%17.7%2.5%7.6%<0.00172.2%<0.001GDMT (n=232)7.8%41.8%28.0%4.7%17.7%49.6%24 monthsMitraClip (n=157)12.1%42.7%21.7%5.7%17.8%<0.00154.8%<0.001GDMT (n=153)5.2%28.1%23.5%3.3%39.3%33.3%

35. MR Severity (Core Lab)MR grade≤1+2+3+4+Ptrend ≤2+P-value BaselineMitraClip (n=302)--49.0%51.0%---GDMT (n=311)--55.3%44.7%-30 daysMitraClip (n=273)72.9%19.8%5.9%1.5%<0.00192.7%<0.001GDMT (n=257)8.2%26.1%37.4%28.4%34.2%6 monthsMitraClip (n=240)66.7%27.1%4.6%1.7%<0.00193.8%<0.001GDMT (n=218)9.2%28.9%42.2%19.7%38.1%12 monthsMitraClip (n=210)69.1%25.7%4.3%1.0%<0.00194.8%<0.001GDMT (n=175)11.4%35.4%34.3%18.9%46.9%24 monthsMitraClip (n=114)77.2%21.9%0%0.9%<0.00199.1%<0.001GDMT (n=76)15.8%27.6%40.8%15.8%43.4%

36. MR Severity (Core Lab)MR grade≤1+2+3+4+Ptrend ≤2+P-value BaselineMitraClip (n=302)--49.0%51.0%---GDMT (n=311)--55.3%44.7%-30 daysMitraClip (n=273)72.9%19.8%5.9%1.5%<0.00192.7%<0.001GDMT (n=257)8.2%26.1%37.4%28.4%34.2%6 monthsMitraClip (n=240)66.7%27.1%4.6%1.7%<0.00193.8%<0.001GDMT (n=218)9.2%28.9%42.2%19.7%38.1%12 monthsMitraClip (n=210)69.1%25.7%4.3%1.0%<0.00194.8%<0.001GDMT (n=175)11.4%35.4%34.3%18.9%46.9%24 monthsMitraClip (n=114)77.2%21.9%0%0.9%<0.00199.1%<0.001GDMT (n=76)15.8%27.6%40.8%15.8%43.4%

37. MR Severity (Core Lab)MR grade≤1+2+3+4+Ptrend ≤2+P-value BaselineMitraClip (n=302)--49.0%51.0%---GDMT (n=311)--55.3%44.7%-30 daysMitraClip (n=273)72.9%19.8%5.9%1.5%<0.00192.7%<0.001GDMT (n=257)8.2%26.1%37.4%28.4%34.2%6 monthsMitraClip (n=240)66.7%27.1%4.6%1.7%<0.00193.8%<0.001GDMT (n=218)9.2%28.9%42.2%19.7%38.1%12 monthsMitraClip (n=210)69.1%25.7%4.3%1.0%<0.00194.8%<0.001GDMT (n=175)11.4%35.4%34.3%18.9%46.9%24 monthsMitraClip (n=114)77.2%21.9%0%0.9%<0.00199.1%<0.001GDMT (n=76)15.8%27.6%40.8%15.8%43.4%3+-4+6.3%5.3%0.9%7.4%

38. LimitationsBecause the MitraClip is visible on imaging tests, COAPT was unblindedBias was mitigated by GDMT standardization and use of independent CEC & ECLMedian FU duration was greater in the Device than the Control groupIn part due to improved survivalHowever, study withdrawals were more frequent in the Control groupResults were consistent using multiple imputation to account for missing dataThe present results apply to Rx of secondary MR with the MitraClipWhether other transcatheter or surgical approaches would have comparable results is uncertainPts were symptomatic (NYHA II - IVa) despite maximally-tolerated GDMT (with more than one-third having undergone CRT), had true moderate-to-severe or severe MR, LVEF 20%-50%, and frequent comorbiditiesWhether the MitraClip would be as safe and effective in more or less critically ill pts or those with lesser degrees of MR severity is unknown

39. Why are the COAPT Results so Different from MITRA-FR? Possible ReasonsMITRA-FR (n=304)COAPT (n=614)Severe MR entry criteriaSevere FMR by EU guidelines: EROA >20 mm2 or RV >30 mL/beatSevere FMR by US guidelines: EROA >30 mm2 or RV >45 mL/beatEROA (mean ± SD)31 ± 10 mm241 ± 15 mm2LVEDV (mean ± SD) 135 ± 35 mL/m2101 ± 34 mL/m2GDMT at baseline and FUReceiving HF meds at baseline – allowed variable adjustment in each group during follow-up per “real-world” practiceCEC confirmed pts were failing maximally-tolerated GDMT at baseline – few major changes during follow-up Acute results: No clip / ≥3+ MR 9% / 9%5% / 5%Procedural complications*14.6%8.5%12-mo MitraClip ≥3+ MR 17%5%*MITRA-FR defn: device implant failure, transf or vasc compl req surg, ASD, card shock, cardiac embolism/stroke, tamponade, urg card surg

40. ConclusionsIn pts with HF and moderate-to-severe or severe secondary MR who remained symptomatic despite maximally-tolerated GDMT, transcatheter mitral leaflet approximation with the MitraClip was safe, provided durable reduction in MR, reduced the rate of HF hospitalizations, and improved survival, quality-of-life and functional capacity during 24-month follow-upAs such, the MitraClip is the first therapy shown to improve the prognosis of patients with HF by reducing secondary MR due to LV dysfunction

41.

42. Back-up slides

43. Baseline and Follow-up TestsClinical: Bl, 1 wk1, 1 mo, 6 mo, 12 mo, 18 mo, 2 yrs, 3 yrs, 4 yrs, 5 yrsLabs2: Bl, d/c1, 1 mo, 6 mo, 12 mo, 18 mo, 2 yrsTEE: BlTTE: Bl, d/c*, 1 mo, 6 mo, 12 mo, 18 mo, 2 yrs, 3 yrs, 4 yrs, 5 yrsNYHA3: Bl, 1 mo, 6 mo, 12 mo, 18 mo, 2 yrs, 3 yrs, 4 yrs, 5 yrsQOL (KCCQ and SF-36)3: Bl, 6 mo, 12 mo, 2 yrs6MWT3: Bl, 6 mo, 12 mo, 2 yrs1Device group; 2Includes BNP or NT-proBNP; 3By blinded study personnel. Bl = baseline; d/c = discharge

44. Definition of HF HospitalizationHospitalization with treatment in any in-patient unit or ward in the hospital for at least 24 hours, including emergency department stay*Clinical signs and/or symptoms of HF, including new or worsening dyspnea, orthopnea, paroxysmal nocturnal dyspnea, increasing fatigue, worsening functional capacity or activity intolerance, or signs and/or symptoms of volume overloadResults in IV (e.g., diuretic or vasoactive therapy) or invasive (e.g., ultrafiltration, IABP, mechanical assistance) treatment for HFArrival in emergency department with clinical presentation meeting the criteria of HF but dies in the emergency department before hospital admissionAdmission for an LVAD or heart transplant - Meets criteria A and B and C or D or E - Overnight stays at nursing home facilities, physical rehab or extended care facilities, including hospice, do not meet the protocol definition of hospitalization. All hospitalizations, including the index hospitalization for the MitraClip procedure, if complicated by acute worsening HF that would have prompted an admission to hospital for HF, AND requires intravenous or invasive treatment AND hospitalization is extended by 24 hours, as defined above, will also be considered a HF hospitalization. Elective HF “tune-ups” that occur following the MitraClip procedure and prolong hospitalization will not count as a HF hospitalization.

45. Origin of the MitraClip Primary SafetyEndpoint Objective Performance Goal of 88% The performance goal for the primary safety endpoint was developed in concert with FDA after determination of what composite risk of device-related complications might be acceptable given the expected effectiveness of the therapy. Based on prior MitraClip studies we anticipated an ~6% rate of these composite adverse events:Single leaflet device attachment - 4% Device embolization - <0.1% Endocarditis - <0.1% Mitral stenosis requiring surgery - <0.1% LVAD or heart transplant - 1% Other device-related complications that require non-elective CV surgery - <0.1% The objective performance goal of freedom from device-related complications was thus set at 88% which will be compared against the 95% CI of the lower limit of the observed safety endpoint

46. MitraClip Procedure (n=302)Reasons MitraClip procedure not attempted in 9 patients - Withdrew before procedure (n=5) - Expired before procedure (n=3) - Lost-to-follow-up before procedure (n=1) Reasons no Clip implanted in 6 attempted patients - Inability to complete the trans-septal procedure (n=4) - Inability to grasp the leaflets (n=1) - Pericardial effusion (n=1)

47. 30-Day Event Rates (i)Kaplan-Meier time-to-first event ratesMitraClip + GDMT (n=302)GDMT alone (n=312)HR [95% CI]P-valueDeath, all-cause2.3% 1.0% 2.43 [0.63, 9.40] 0.20 - CV2.3% 0.6%3.64 [0.76, 17.53] 0.11 - HF-related0.7%0.6% 1.05 [0.15, 7.42] 0.97 - Non-HF-related1.7%0.0% -- - Non-CV0.0% 0.3% - -Hospitalization, all-cause13.0% 12.6% 1.04 [0.67, 1.62] 0.86 - CV6.7% 7.7% 0.86 [0.48, 1.6] 0.62 - HF-related4.0% 6.8% 0.58 [0.29, 1.18] 0.14 - Non-HF-related3.0% 1.0% 3.17 [0.86, 11.7] 0.08 - Non-CV6.7% 5.5% 1.23 [0.64, 2.35] 0.53 Death or HF hospitalization5.6% 7.1% 0.79 [0.42, 1.48] 0.46

48. 30-Day Event Rates (ii)MitraClip + GDMT (n=302)GDMT alone (n=312)HR [95% CI]P-valueMV intervention or surgery*1.0% 0.3% 3.13 [0.33, 30.09] 0.32 - MitraClip1.0% 0.3% 3.13 [0.33, 30.09] 0.32 - Mitral valve surgery0.0% 0.0% - - PCI or CABG0.3% 0.0% - -Stroke0.7% 0.0% - -Myocardial infarction1.0% 0.0% - -New CRT implant0.0% 0.0% - -LVAD or heart transplant0.0% 1.0% - - - LVAD0.0% 0.0% - - - Heart transplant0.7% 0.0% - -*Unplanned. Kaplan-Meier time-to-first event rates

49. Primary EndpointsPrimary effectiveness endpoint – All HF hospitalizations through 24 monthsPopulationMitraClip + GDMTGDMT aloneHR [upper 95% CL]P-valueIntention-to-treat*35.8% (160/446.5)167.9% (283/416.8)10.53 [0.66]<0.0012As-treated34.8% (154/442.0)170.6% (277/392.2)10.44 [0.56]<0.0012Per-protocol35.4% (145/409.4)170.0% (257/366.9)10.45 [0.58]<0.0012Primary safety endpoint – Freedom from device-related complications at 12 months PopulationMitraClip + GDMTGDMT aloneLower 95% CLP-value Safety analysis*96.6% (9)3-94.8%<0.0014As-treated97.5% (7)3-95.9%<0.0014Per-protocol97.7% (6)3-96.1%<0.0014The intention-to-treat population consists of all pts randomized in the trial, analyzed in the group randomized, regardless of the treatment actually received. The as-treated population consists of all randomized pts according to the treatment they received. Pts who experience a death or HF hospitalization prior to a MitraClip procedure are considered to be in the Control group regardless of their initial randomization. Pts who experience a death or HF hospitalization after (but not prior to) a MitraClip procedure are considered to be in the MitraClip group regardless of their initial randomization. For pts who do not experience a death or HF hospitalization at any time during follow-up, they will be assigned to the group that constituted >50% of their follow-up duration. The per-protocol population consists of all randomized pts who meet all major study inclusion criteria and none of the major exclusion criteria for the trial, are treated according to the randomized assignment, and followed consistent with all major study processes. Pts who are randomized to the Device group but do not have a MitraClip procedure attempted between 1-14 days of the randomization are excluded from the per-protocol population. For pts in the Control group who receive the MitraClip device, follow-up data after the date of the procedure will be excluded from analysis. For subjects in either group who undergo other major intervention for HF (e.g. mitral valve surgery, LVAD implant or heart transplant), follow-up data after the date of the intervention will be excluded from analysis. The safety analysis population consists of all randomized Device group subjects in whom a MitraClip procedure is attempted.1Annualized event rate (total number of events/patient-years of follow-up); 2Joint frailty model; 3Kaplan-Meier time-to-first event estimates (number of events); 4Calculated from Z test with Greenwood’s method of estimated variance against a pre-specified performance goal of 88%. *Population for the study primary endpoint.

50. Time to First HF HospitalizationHF Hospitalizations (%) 0% 20% 40% 60% 80% 100%Time After Randomization (Months)0369121518212456.7%35.7%HR [95% CI] = 0.52 [0.40-0.67]P<0.001MitraClip + GDMTGDMT alone30226423821519415414512697312244205174153117907555No. at Risk:MitraClip + GDMTGDMT aloneNNT (24 mo) =4.8 [95% CI 3.3, 8.2]

51. LVAD or Heart TransplantWithin 24 MonthsHR [95%CI] =0.37 [0.17, 0.81] P=0.01HR [95%CI] =0.34 [0.13, 0.87] P=0.02HR [95%CI] =0.35 [0.09, 1.32] P=0.12Stone GW et al. NEJM. 2018 Sept 23.

52. Change in LVEDV from Baseline to 12 Months±76.0±69.2Adjusted change**Ancova±5.1±5.1P<0.001n=175n=174n=175n=174±94.2±76.5

53. 24-Month All HF Hospitalizations0.290.05040.350.620.580.700.280.530.54 [0.41, 0.70]0.47 [0.31, 0.71]0.44 [0.32, 0.61]0.62 [0.41, 0.94]0.51 [0.36, 0.72]0.48 [0.30, 0.76]0.77 [0.34, 1.75]0.57 [0.40, 0.80]0.67 [0.41, 1.08]0.57 [0.40, 0.80]67.9% (283/416.8)74.2% (154/207.5)74.5% (185/248.4)70.7% (121/171.1)65.7% (176/268.1)65.7% (94/143.1)113.0% (40/35.4)59.8% (151/252.4)50.4% (68/135.0)80.1% (146/182.4)35.8% (160/446.5)33.9% (74/218.5)31.6% (94/297.5)44.0% (75/170.5)32.5% (92/283.2)30.8% (44/142.7)83.4% (20/24.0)33.5% (92/274.8)33.1% (51/153.9)44.6% (97/219.5)All patientsAge (median)SexEtiology of cardiomyopathyPrior CRTHF hospitalization within the prior yearBaseline NYHA classSTS replacement scoreSurgical risk statusBaseline MR gradeBaseline LVEF0.62 [0.44, 0.88]61.7% (129/209.2)37.7% (86/228.0)≥74 years (n=317)<74 years (n=297)0.77 [0.49, 1.21]58.2% (98/168.4)44.3% (66/149.0)Female (n=221)Male (n=393)0.48 [0.34, 0.68]66.0% (162/245.6)30.8% (85/276.0)Ischemic (n=373)Non-ischemic (n=241)0.58 [0.38, 0.89]72.0% (107/148.7)41.6% (68/163.3)Yes (n=224)No (n=390)0.56 [0.41, 0.78]69.1% (189/273.7)38.2% (116/303.8)Yes (n=407)No (n=207)0.59 [0.38, 0.92]53.4% (80/149.9)31.5% (62/197.1)I or II (n=240)0.660.52 [0.36, 0.75]68.9% (158/229.4)34.6% (78/225.5)III (n=322)IV (n=51)0.51 [0.34, 0.77]80.3% (132/164.4)39.6% (68/171.7)≥8% (n=262)<8% (n=352)0.49 [0.35, 0.68]76.3% (215/281.8)36.5% (106/290.4)High (n=423)Not high (n=188)0.48 [0.31, 0.74]58.8% (137/232.9)27.3% (62/227.0)3+ (n=320)4+ (n=293)0.340.47 [0.34, 0.64]75.0% (241/321.5)34.3% (116/338.3)0.65 [0.33, 1.27]51.0% (35/68.6)32.9% (26/79.0)>40% (n=103)≤40% (n=472)0.960.49 [0.32, 0.76]60.3% (123/204.0)29.0% (66/227.7)≥30% (median; n=301)0.50 [0.34, 0.72]82.2% (153/186.1)40.1% (76/189.6)<30% (median; n=274)0.440.56 [0.36, 0.85]60.4% (114/188.7)32.9% (69/209.4)Baseline LVEDV (median)0.44 [0.31, 0.64]80.4% (162/201.5)35.1% (73/207.9)≥181 mL (n=288)<181 mL (n=287)P [Int]HR [95% CI]GDMT aloneMitraClip + GDMTSubgroupHR [95% CI]0.20.511.52.5Favors MitraClip + GDMT Favors GDMT aloneAnnualized rates*Central eligibility committee assessment

54. 24-Month All-cause Mortality0.340.940.840.280.790.190.250.790.61 [0.46, 0.82]0.51 [0.32, 0.82]0.59 [0.42, 0.82]0.59 [0.37, 0.95]0.54 [0.37, 0.78]0.65 [0.40, 1.07]0.64 [0.28, 1.46]0.49 [0.31, 0.76]0.67 [0.49, 0.92]0.43 [0.23, 0.82]0.62 [0.43, 0.90]46.1% (121)39.0% (50)53.7% (84)43.1% (46)45.3% (77)45.0% (40)62.9% (19)37.5% (54)50.5% (93)35.6% (28)54.5% (64)29.1% (80)21.3% (27)32.0% (60)26.5% (27)26.6% (44)29.2% (26)46.7% (8)19.1% (29)33.9% (64)17.5% (14)33.5% (48)All patients0.620.58 [0.42, 0.80]49.0% (99)28.8% (61)0.230.46 [0.30, 0.71]53.2% (63)25.4% (30)0.230.47 [0.31, 0.72]50.6% (63)26.1% (33)All patientsAge (median)SexEtiology of cardiomyopathyPrior CRTHF hospitalization within the prior yearBaseline NYHA classSTS replacement scoreSurgical risk status*Baseline MR gradeBaseline LVEF0.68 [0.48, 0.97]52.3% (71)35.9% (53)≥74 years (n=317)<74 years (n=297)0.61 [0.35, 1.05]34.0% (37)22.8% (20)Female (n=221)Male (n=393)0.63 [0.44, 0.89]48.1% (75)30.8% (53)Ischemic (n=373)Non-ischemic (n=241)0.75 [0.48, 1.16]47.5% (44)33.4% (36)Yes (n=224)No (n=390)0.60 [0.42, 0.85]46.6% (81)29.0% (54)Yes (n=407)No (n=207)0.58 [0.36, 0.92]45.3% (40)27.6% (31)I or II (n=240)IV (n=51)0.74 [0.51, 1.07]57.1% (67)42.6% (51)≥8% (n=262)<8% (n=352)High (n=423)Not high (n=188)0.58 [0.37, 0.89]39.1% (57)24.8% (32)3+ (n=320)4+ (n=293)0.47 [0.22, 0.98]43.6% (20)25.3% (11)>40% (n=103)≤40% (n=472)<30% (median; n=274)Baseline LVEDV (median)<181 mL (n=287)P [Int]HR [95% CI]GDMT aloneMitraClip + GDMTSubgroupHR [95% CI]0.20.511.52.50.67 [0.45, 1.01]45.1% (56)30.4% (39)≥181 mL (n=288)0.67 [0.45, 1.00]43.3% (56)30.4% (42)≥30% (median; n=301)0.910.68 [0.45, 1.00]43.7% (62)28.5% (41)III (n=322)Favors MitraClip + GDMT Favors GDMT aloneKM time-to-first event rates*Central eligibility committee assessment

55. COAPT vs. MITRA-FR: 12-Month Death or HF HospStone GW et al. NEJM. 2018 Sept 23.COAPTDeath or HF Hospitalization (%)Months100%90%80%60%20%0%50%40%30%10%Control GroupDevice GroupNo. at Risk:70%031230232442646205238917421512153194HR [95% CI]=0.63 [0.49–0.82]P<0.001MitraClip + GDMTGDMT alone33.9%46.5%MITRA-FRObadia JF et al. NEJM. 2018 Aug 27. doi: 10.1056/NEJMoa1805374Death or HF Hospitalization (%)Months100%90%80%60%20%0%50%40%30%10%Control GroupDevice GroupNo. at Risk:70%01521512123114410995694918868110807312736754.6%51.3%OR [95% CI]=1.16 [0.73–1.84]P=0.53MitraClip + MTMT alone

56. COAPT vs. MITRA-FR: MR, LV Volumes and FunctionCOAPT (n=614)MITRA-FR (n=304)EROA, mm2 (mean ± SD)41 ± 1531 ± 10 - <30 mm2 14% (80/591)52% (157/301) - 30 – 40 mm246% (270/591)32% (95/301) - >40 mm241% (241/591)16% (49/301)LVEF, % (mean ± SD)31 ± 933 ± 7LVEDV, mL/m2 (mean ± SD)101 ± 34135 ± 35Obadia JF et al. NEJM. 2018 Aug 27. doi: 10.1056/NEJMoa1805374; Stone GW et al. NEJM. 2018 Sept 23.

57. Impact of EROA and LVEDV: EROA >40 mm2 All-cause mortality or HF hospitalization through 12 months56LVEDVI >96 ml/m2 (N=130; 23.7%)LVEDVI ≤96 ml/m2 (N=92; 16.8%)

58. Impact of EROA and LVEDV: EROA >30-40 mm2 All-cause mortality or HF hospitalization through 12 months57LVEDVI >96 ml/m2 (N=88; 16.1%)LVEDVI ≤96 ml/m2 (N=131; 23.9%)

59. Impact of EROA and LVEDV: EROA ≤30 mm2 All-cause mortality or HF hospitalization through 12 months58LVEDVI >96 ml/m2 (N=56; 10.2%)LVEDVI ≤96 ml/m2 (N=51; 9.3%)

60. COAPT vs. MITRA-FR: MitraClip OutcomesCOAPT (n=302)MITRA-FR (n=152)MitraClip attempted293 (97.0%)144 (94.7%)≥1 Clip implanted287 (95.0%)138 (90.8%)Procedural complications25/293 (8.5%)21/144 (14.6%) - Device implant failure6 (2.0%)6 (4.2%) - Transfusion or vasc compl requiring surgery16 (5.5%)5 (3.5%) - ASD2 (0.7%)4 (2.8%) - Cardiogenic shock1 (0.3%)4 (2.8%) - Cardiac embolism/stroke1 (0.3%)2 (1.4%) - Tamponade1 (0.3%)2 (1.5%) - Urgent cardiac surgery 1 (0.3%)0 (0%)Acute result: MR ≥3+5%9%12-month result: MR ≥3+5%17%Stone GW et al. NEJM. 2018 Sept 23; Obadia JF et al. NEJM. 2018 Aug 27. doi: 10.1056/NEJMoa1805374