Outcomes for 15,259 US Patients With Acute MI Cardiogenic Shock (AMICS) Supported With Impella William O’Neill, MD, FACC Medical Director Structural Heart Disease at Henry Ford Hospital, MI

High In-Hospital Mortality During AMI Cardiogenic Shock3

1.  Sandhu A, McCoy l, Negi S, et al. Use of Mechanical Circulatory Support in Patients Undergoing Percutaneous Coronary Intervention; Insights from the National Cardiovascular Data Registry. Circulation, 2015;132:1243-1251

2.  Acute Cardiac Assist Report, Health Research International – August 2015 3.  Jeger, et al. Ann Intern Med. 2008

N = 23,696

7435578954 78500 79823

8058582626

8669289923

2009 2010 2011 2012 2013 2014 2015 2016

US AMI/CGS cases per year1,2

AMI Shock Mortality Unchanged in > 20 years

Impella Quality (IQ) Database Methods •  Abiomed clinical personnel collecting real world data from

>98% of US cases since 2009; >50,000 patients

•  >15,000 patients with AMI-CGS

•  FDA Approval 2016, AMI/CGS therapy and heart recovery

•  Audited by Abiomed Heart Team (Cardiologists and CV Surgeon)

•  HIPAA compliant data collection, FDA Maude protocol compliant

•  “Exempt” status by Henry Ford Hospital IRB

•  Survival tracked to device explant

HRPCIElec+ve&Urgent

47%(n=1275)

CardiogenicShock40%

(n=1090)Other13%

(n=339)

IQ Program Data Resources

HRPCIElec+ve&Urgent

48%(n=22,678)

CardiogenicShock32%

(n=15,259)

Other19%

(n=9012)

cVAD Registry Data2N=2,704

Observational IQ Database IRB Registry Data

•  IRB Exempt / HIPAA Compliant •  1,010 US Impella Centers; 2009-2017 •  Abiomed Heart Team Physicians Audited •  All Devices, All Indications

•  IRB approval at all institutions (65) •  Retrospective (‘09 to ‘15); Prospective (‘16) •  FDA protocols and CEC Events Adjudication •  All Devices, All Patients Enrolled

Abiomed Impella Quality (IQ) Database1

N=46,949

1.  Abiomed Impella Quality (IQ) Database, Danvers MA 2.  cVAD Registry Data of Patients Undergoing PCI for Acute Myocardial Infarction Complicated by Cardiogenic Shock as of September 2015

AMI/CGS Impella Patient Demographics

•  Age –  Mean: 63.6 y/o –  Range: (19 – 99)

•  Gender –  73% Male

•  Duration Of Support –  Mean: 3.78 Days –  Median: 2.7 Days –  Max: 94 Days

•  Survival to Explant

N=3549 N=9693

IQ Database1 cVAD Registry2

•  Age –  Mean: 66.3 y/o –  Range: (19 – 95)

•  Gender –  76% Male

•  Duration Of Support –  Mean: 1.63 Days –  Median: 1.1 Days –  Max: 10.6 Days

•  Survival to Explant, Discharge & 30 days

1.  Abiomed Impella Quality (IQ) Database, Danvers MA 2.  cVAD Registry Data of Patients Undergoing PCI for Acute Myocardial Infarction Complicated by Cardiogenic Shock as of September 2015

Impella Utilization in AMI Shock

1.  Acute Cardiac Assist Report, Health Research International – August 2015 2.  Sandhu A, McCoy l, Negi S, et al. Use of Mechanical Circulatory Support in Patients Undergoing Percutaneous Coronary Intervention; Insights from the National Cardiovascular

Data Registry. Circulation, 2015;132:1243-1251 3.  Data on file. Abiomed Impella Quality (IQ) Database, US AMI/CGS Jan 2009 – Dec 2016. Danvers, MA: Abiomed.

74355 78954 78500 79823 80585 8262686692

89923

43% 45%42% 41% 39% 41% 42% 42%

0% 1% 1% 2% 2% 3% 4% 6%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

2009 2010 2011 2012 2013 2014 2015 2016

% Impella Supported Patients 3

% IABP Supported Patients 2

Total AMI/CGS US Patients 1,2

1stQuartile 0.41667Median 0.533333rdQuartile 0.64286Maximum 1.00000

0.51673 0.54190

0.51314 0.54545

0.17187 0.18970

A-Squared 1.60P-Value < 0.005

Mean 0.52931StDev 0.18034Variance 0.03252Skewness -0.147967Kurtosis 0.303483N 791

Minimum 0.00000

Anderson-Darling NormalityTest

95% ConfidenceIntervalforMean

95% ConfidenceIntervalforMedian

95% ConfidenceIntervalforStDev

90.0%75.0%60.0%45.0%30.0%15.0%0.0%

Median

Mean

55.0%54.0%53.0%52.0%51.0%

95%ConfidenceIntervals

SummaryReportforSurvivalVariation in Impella AMI/CGS Outcomes

1.  791 sites supporting >4 AMICS patients, 15,529 patients total. Data on file. Abiomed Impella Quality(IQ)Data, AMI/CGS Jan 2009 – Dec 2016. Danvers, MA: Abiomed. 2.  Top 20% performing sites have higher volume of Impella utilization 3.  Greater than 90% of survivors were explanted with native heart recovery in 2016 4.  Mean survival of 58% in 2016. Improvement of 14% (relative) since FDA approval

Distribution of Impella Site Outcomes1

Top 20% of sites have mean survival of 76%2

Bottom 20% of sites have mean survival of 30%

SurvivaltoExplant3

#ofSites

20164

62%

67%

IABP/InotropesPre-PCI ImpellaPre-PCI

cVAD Registry2

N=164

52%

59%

IABP/InotropesPre-PCI ImpellaPre-PCI

IQ Database1

Impella Pre-PCI associated with Improved Survival in AMI/CGS

1.  Abiomed Impella Quality (IQ) Database, US AMI/CGS Apr 2009– Jan 2017. Survival to Explant. Danvers, MA: Abiomed. 2.  Basir M, Schreiber T, Grines C, et al. Effect of Early Initiation of Mechanical Circulatory Support on Survival in Cardiogenic Shock. Am. J. of Cardiology, 2016

P<0.001

N=121N=2450N=3121

P<0.001

Hemodynamic Monitoring associated with Improved Survival in AMI/CGS

68%76%

NoHemodynamicMonitoring

HemodynamicMonitoring

cVAD Registry2

49%

63%

NoHemodynamicMonitoring

HemodynamicMonitoring

IQ Database1

N=516N=634N=5217N=8767

P<0.0001

P=0.002

1.  Abiomed Impella Quality (IQ) Database, US AMI/CGS Apr 2009– Jan 2017. Survival to Explant. Danvers, MA: Abiomed. 2.  cVAD survival to explant 2009-2016

32%

54%65% 65%

74%

0 1 2 3 4+

Mortality and Number of Inotropes from cVAD Registry1 P<0.001 (N=287)

Number of Inotropes/Pressors

1. Basir M, Schreiber T, Grines C, et al. Effect of Early Initiation of Mechanical Circulatory Support on Survival in Cardiogenic Shock. Am. J. of Cardiology, 2016

Increased Inotrope Exposure is associated with Mortality in AMI/CGS

Samuels LE et al , J Card Surg. 1999

Mortality

Detroit Cardiogenic Shock Initiative

DETROIT CSI

Detroit CSI AMI/CGS Pilot Study

0.56WaRs

0.96WaRs

CPOPre-Impella CPOOnImpella

P<0.001

N=21 N=27

58%increase

•  July 2016 to February 2017 –  all sites performed >10 AMICS cases w/

Impella within last calendar year

•  Enrolled 37 patients –  Age 63 +/- 13 years (36-87)

•  Rapid Door to Unloading times (average 82 minutes)

•  62% supported w/ Impella Pre-PCI

•  RHC use 84%

•  86% of patients established TIMI III flow

•  Decrease Inotropic/Vasopressor use in 80% of cases

Hemodynamic Improvement On Support

Cardiac Power Output1 (CPO = MAP x CO)

1.  Fincke, et al., Cardiac Power Is the Strongest Hemodynamic Correlate of Mortality in Cardiogenic Shock: A Report From the SHOCK Trial Registry. JACC, Vol. 44, No. 2, 2004

51%

89% 84%

SurvivaltoExplantMetroDetroitBeforeStudy

SurvivaltoExplantDetroitCSI

SurvivaltoDischargeDetroitCSI

Outcomes

100% Native Heart Recovery in Survivors

100% Native Heart Recovery In surviving Patients (31/31)

1 2 2

1.  AbiomedImpellaQuality(IQ)Database,Jan2015toJuly2016forAggregateDTWMetroHospitalsAMI/CGSSurvivaltoExplant2.  FeasibilityofEarlyMechanicalSupportDuringMechanicalReperfusionofAcuteMyocardialInfarctCardiogenicShock;W.O’Neill,M.Basir,S.Dixon,KPatel,TSchreiber,S.Almany;InPressJACCIntervenXons

Conclusions •  AMI CGS mortality remains unchanged despite major advances in

cardiac care in past 20 years

•  Despite FDA PMA approval, Impella is used in ~5% of US AMI Shock Cases

•  There is a wide institutional variation in AMI CGS outcomes with Impella use

•  Key Observations Associated with Improved Outcomes: ‒  Increased institutional use of Impella ‒  Impella use prior to PCI ‒  Reduced exposure to high dose inotropes ‒  Protocol using hemodynamic monitoring to guide escalation and weaning

•  Prospective, systematic adoption of best practices (DCSI) markedly improves survival and native heart recovery

Thank You

EarlyIdenVficaVon

UnloadPriortoPCI(DTU)

HemodynamicMonitoring

ProtocolDrivenTreatment

EscalaVon

DetroitCSI@h]s.org