ECMO AND SHORT-TERM SUPPORT:

ECMO AND SHORT-TERM SUPPORT: UTILIZATION GUIDELINES AND IMPACT OF THE NEW HEART ALLOCATION SYSTEM Jeffrey Teuteberg Section Chief of Heart Failure, Cardiac Transplant and Mechanical Circulatory Support Stanford University

Disclosures: Medtronic/HeartWare (ad board, speaking), Abiomed (ad board, speaking), CareDx (ad board, speaking), Abbott/Thoratec (CEC) ECMO AND SHORT-TERM SUPPORT: UTILIZATION GUIDELINES AND IMPACT OF THE NEW HEART ALLOCATION SYSTEM Jeffrey Teuteberg Section Chief of Heart Failure, Cardiac Transplant and Mechanical Circulatory Support Stanford University

BACKGROUND Problems Too many candidates waiting as Status 1A (3x more likely to die on waiting list) Changing landscape of HF management LVAD usage Specific patient groups may be disenfranchised Inequities in access to organs because of artificial geographic boundaries Goals Reduce waiting list mortality Better stratify candidates based on medical urgency Expand access to donors for the most critically ill patients

PROPOSED NEW STATUSES: HIGH LEVEL Current Status 1A 1 Proposed Status 2 3 1B 4 2 5 6 Proposed statuses 1-3 are generally defined by current status 1A criteria Proposed status 4 is generally defined by current status 1B criteria Proposed status 5-6 are generally defined by current status 2 criteria 4

STATUSES 1-3 Status 1 2 3 Criteria ECMO Non-dischargeable (surgically implanted) non-endovascular BiVAD MCSD with life-threatening ventricular arrhythmia Non-dischargeable, surgically implanted, non-endovascular LVAD TAH BiVAD RVAD LVAD for single ventricle MCSD with device malfunction/mechanical failure Percutaneous endovascular MCS Intra-aortic balloon pump Ventricular tachycardia/ventricular fibrillation, mechanical support not required Dischargeable LVAD for up to 30 days Multiple inotropes or single high-dose inotropes with continuous hemodynamic monitoring MCSD with device infection MCSD with hemolysis MCSD with pump thrombosis MCSD with right heart failure MCSD with mucosal bleeding MCSD with aortic insufficiency 5

STATUSES 4-6 Status 4 Criteria Stable LVAD candidates not using 30 day discretionary period Inotropes without hemodynamic monitoring Diagnosis of congenital heart disease (CHD) Diagnosis of ischemic heart disease with intractable angina Diagnosis of hypertrophic cardiomyopathy Diagnosis of restrictive cardiomyopathy Diagnosis of amyloidosis Retransplant 5 Combined organ transplants 6 All remaining active candidates 6

% DEATH WITHIN 6 MONTHS A(i) =VAD for 30 days A(ii) = TAH A(iii) = IABP A(iv) = ECMO B(i) = Thromboembolism B(ii) = Device infection B(iii) = Device malfunction B(iv) = Life-threatening ventricular arrhythmia B(v) = Other device related complication Post-TX WL Status 2 Status 1B Status 1A-(E) Status 1A-(D) Status 1A-(C) Status 1A-(B) Status 1A-(A)(iii) Status 1A-(A) Status 1A-ALL ALL Status 2 Status 1B Status 1A-(E) Status 1A-(D) Status 1A-(C) Status 1A-(B) Status 1A-(B)(iv) Status 1A-(A)(ii) Status 1A-(A) Status 1A-ALL ALL Status 1A-(B)(iv) Status 1A-(A)(iii) Status 1A-(A)(iv) Status 1A-(A)(ii) TAH ECMO Status 1A-(A)(iv) 0% 5% 10% 15% 20% 25% 30% 35% * For WL analysis, time is computed from first entry into criteria/sub-criteria, rather than time since listing. ECMO Status 2 Status 1B Status 1A-(E) Status 1A-(D) Status 1A-(C) Status 1A-(B)(v) Status 1A-(B)(iv) Status 1A-(B)(iii) Status 1A-(B)(ii) Status 1A-(B)(i) Status 1A-(B) Status 1A-(A)(iv) Status 1A-(A)(iii) Status 1A-(A)(ii) Status 1A-(A)(i) Status 1A-(A) Status 1A-ALL ALL

PERCUTANEOUS VAD Death rate Transplant rate

THE WAITLIST MORTALITY Stable VAD Patients have a 180-Day Waitlist Mortality=6.4% J Am Coll Cardiol 2012;60:36-43

WHAT IS THE INDICATION? Indication Ongoing cardiogenic shock that occurs immediately (<48 hours) following acute myocardial infarction (AMI) or open heart surgery as a result of isolated left ventricular failure that is not responsive to optimal medical management and conventional treatment measures with or without an intra-aortic balloon pump Impella 2.5 & Impella CP <4 days Impella 5.0 & Impella LD <6 days Impella RP up to 14 days Leaving beyond 4, 6, or 14 days due to unforeseen circumstances is a clinical decision

ALL ABOUT INCENTIVES CARROT OR STICK?

HOW WOULD YOU APPROACH THIS PATIENT? 50 year old with NICM with progressive dyspnea despite optimal medical therapy and CRT-D. Hemodynamics RA 6 PA 53/22 (32) W 16 CO/CI 3.2/1.8 Potential management options: A. Follow closely in clinic B. Try to adjust HF medications as an outpatient C. Perform CPET D. Place RHC, start two low-dose inotropes, admit for months long hospitalization in CCU, but allow to ambulate around the unit.

HOW COMMON/HOW RISKY? Tier 1 Tier 2 ECMO Ventilator No d/c VAD VAD & VT IABP VT VAD dysfxn TAH R/BiVAD Tier 3 LVAD 30 1Ae RHC LVAD comp LVAD inf Days Tx Tx rate Deaths Death rate 432 1086 404 1714 5263 11392 6996 4015 2076 31563 6863 28603 5769 31108 13 15 15 21 168 88 71 48 43 513 138 700 74 261 1099.1 504.5 1356.1 447.5 1165.9 282.1 370.7 436.7 756.5 593.6 734.4 893.9 468.5 306.4 18 6 2 2 15 4 6 0 0 15 7 15 2 6 1521.9 201.8 180.8 42.6 104.1 12.8 31.3 0 0 17.4 37.3 19.2 12.7 7.0

IMPELLA 5.0 47 pts, single-center 1/06-12/11 Shock CS: 32% Post-cardiotomy 68% Impella 5.0: 80% Lemaire et al. Ann Thorac Surg 2014;97:133-138.

SHOCK AND ACUTE SUPPORT Short-term VAD: 59% VA ECMO: 41% Impella 2.5 in 9pts LV vent in 1 pt Takayama et al. J Heart Lung Transplant 2013;32:106-111.

TECHNOLOGY IS MOVING QUICKLY

WE ARE ALREADY RESPONDING TO INCENTIVES Stevenson LW. J Heart Lung Transplant. 2013;32(9):861-7

WHO GETS THE HEARTS? Stevenson JAMA Int Med 2015;175:1406-9. Dardas et al. J Heart Lung Transplant 2017;36:666-672.

IS IT THE SAME EVERYWHERE? Nguyen et al. J Heart Lung Transplant 2016;35:986-994.

HEMODYNAMIC CRITERIA ACS IMPLANTATION Within 7 days, all met in within one 24 hour period SBP < 90 mmhg Cardiac index < 1.8 L/min/m 2 without inotropes < 2.0 L/min/m 2 with inotropes Wedge >15 mmhg No hemos in 7 days, then at least one, within 24 hours prior to ECMO CPR was performed Systolic blood pressure < 70 mmhg Arterial lactate > 4 mmol/l AST or ALT >1,000 U/L

ACS WEANING ECMO Every 7 days, must have both 1. Demonstrated contraindication to durable MCS 2. Within 48 hours of status expiring, failed wean, at least one: 1. MAP < 60 mmhg 2. CI < 2.0 L/min/m 2 3. Wedge > 15 mmhg 4. SvO 2 < 50% measured by central venous catheter Surgical nondis VAD, Perc VAD, IABP Every 14 days, must have both 1. Demonstrated contraindication to durable MCS 2. Within 48 hours of status expiring, failed wean, at least one: 1. MAP < 60 mmhg 2. CI < 2.0 L/min/m 2 3. Wedge > 15 mmhg 4. SvO 2 < 50% measured by central venous catheter After 14 days if extension not granted or not requested, then status 3

HEMODYNAMIC CRITERIA STICK NOT CARROT? Dobutamine 7.6 mcg/kg/min Milrinone 0.54 mcg/kg/min Dobutamine 3.5 mcg/kg/min Milrinone 0.36 mcg/kg/min Dopamine 2.4 mcg/kg/min Parker et al. J Heart Lung Transplant 2017;36:1013-17.

HEMODYNAMICS AND MORTALITY Parker et al. J Heart Lung Transplant 2017;36:1013-17.

ANTICIPATING BEHAVIOR?

LANDSCAPE OF RISK 1A was originally intended for patients who had less than 7 days left to live without transplant, now waiting 6-12 months BTT one year mortality is 12%, about 12-14% risk of CVA, and up to 30% have been removed from waiting list at one year due to complications VAD complications (thrombus, RHF, infection, bleeding, AI) only makes you status 3 In toto is this better or worse than waiting on ambulatory ACS? Stevenson et al. J Heart Lung Transplant 2016;35:547-549.

THE LAST WORD The modeling for the proposed new system also reflects considerable expertise and thought based on current listing practice. However, it cannot overcome the fundamental reality that future listing behavior cannot be modeled based on previous listing behavior if the rules are changing. Lynne Stevenson Stevenson et al. J Heart Lung Transplant 2016;35:547-549.

NONDISCHARGEABLE VAD Defined non-dischargeable non-endovascular Must remain hospitalized because device is not FDA-approved for out of hospital use Duration: 14 days Extension if both: 1. Contraindication to durable MCS 2. Within 48 hours of status expiring, failed weaning by at least one: 1. MAP < 60 mmhg 2. CI < 2.0 L/min/m2 3. Wedge > 15 mmhg 4. SvO 2 < 50% measured by central venous catheter

CURRENT ALLOCATION Local: Status 1A, Status 1B Zone A: Status 1A, Status 1B Local: Status 2 Zone B: Status 1A, Status 1B Zone A: Status 2 Zone B: Status 2 Etc Zone C 1500 Miles Zone B 1000 Miles Zone A 500 Miles

PROPOSED BROADER SHARING SEQUENCE Candidate Status Status 1 adult + Status 1A ped Status 1 adult + Status 1A ped Status 2 adult Status 2 adult Status 3 adult + Status 1B ped Status 4 adult Status 3 adult + Status 1B ped Location DSA + Zone A Zone B DSA + Zone A Zone B DSA DSA Zone A 30

Should percutaneous temporary VAD patients receive higher priority? The panel agreed that there was insufficient data with the Impella device and other percutaneous VADs with regard to waitlist and post-transplant mortality. It was believed that there appears to be a trend towards higher waitlist mortality in this cohort than those on intra-aortic balloon pumps (IABP). However, opinion was mixed regarding whether percutaneous temporary VAD patients should be of higher priority. A panel member contended that temporary VAD patients priority level should depend on the reason for implantation. For those patients who are clinically stable and walking with a temporary VAD, it was argued that these patients should have lower priority over a bedbound patient who is clinically worse. Such a policy, it was agreed, would have to be carefully worded to avoid misinterpretation and gaming. Overall, it was felt by this panel member that temporary VAD patients should be prioritized highly, likely in Tier 2, below ECMO and ideally above IABP (however, it was acknowledged that as part of Tier 2, time on the waiting list would be the sole differentiator between IABP patients and temporary VAD patients). Another attendee argued that Impella and other temporary VADs were by nature short-term devices, and that when they inevitably fail, they cause the patient to be extremely ill. Therefore, temporary VADs should not be considered stable and should be accorded one of the highest priority levels, likely Tier 1. ECMO patients could therefore be bridged to a percutaneous VAD (thus preventing an extra sternotomy) without losing higher priority in this scenario. There was some agreement with this, but the point was raised that if a patient was doing well on a temporary VAD, the incentive to transplant them urgently is lower. Status 2: Acute Circulatory Support (ACS) Device The Committee reviewed data to support placing ACS candidates in status 2. The cohort used in the TSAMs includes candidates registered for a heart transplant between mid-2009 to mid-2011. This cohort pre-dates the rapid growth of heart candidates supported by ACS, so the Committee reviewed data regarding waiting list and post-transplant outcomes for candidates supported by ACS (defined as balloon pump, Impella, and CardiacAssist TandemHeart) between 2011 and 2013. During this period, about 11,000 heart-alone candidates were registered for transplant, and approximately 4% were registered with ACS at listing, most commonly with a balloon pump. For all ACS candidates combined, the death/too sick rate was 31 per 100 patient years, compared with 34 per 100 patient years for all status 1A candidates and 21 per 100 patient years for candidates on inotropes. At time of transplant, 6% of candidates were supported by an ACS device, and the vast majority of those candidates were supported by a balloon pump alone. The two-year post-transplant survival rates for candidates transplanted while supported by an ACS was 84%, midway between BiVADs (82%) and LVADs (86%). When the Committee initially designed the straw man, it placed candidates with balloon pumps in status 2 based on supporting data and clinical experience, because those candidates are not as urgent as candidates supported by ECMO. The Committee was similarly wary of assigning ACS candidates to status 1, because the death rate for Impella and TandemHeart is closer to those candidates with balloon pump, and they do not appear to be as urgent as those candidates on ECMO. Additionally, the Committee does not want to create an inadvertent incentive for transplant teams to treat with ACS in order to place their candidates in status 1. To further avoid creating an inadvertent incentive to treat with ACS, the Committee proposes a requirement that the candidate be treated with ACS specifically for cardiogenic shock, and created a hemodynamic threshold of showing the candidate had a cardiac index of less than or equal to 2.2 L/min/m 2 prior to ACS implantation. This cardiac index value was adopted based on the Centers for Medicare and Medicaid Services (CMS) threshold for coverage for administration of home inotropes. To be consistent throughout policy, each time the Committee proposed a hemodynamic requirement in addition to therapy, the Committee included the proposed threshold of a cardiac index less than or equal to 2.2 L/min/m 2. UNOS will maintain a list of qualifying ACS devices which will be reviewed annually by the Committee. The current list of ACS devices is included in Exhibit C. Status 2: Intra-aortic balloon pump (IABP) The Committee discussed whether candidates supported by IABP should be in status 2, and whether they should be in the same status as those candidates treated with TAH. Though the waiting list mortality and post-transplant survival rates for candidates supported by IABP are worse than for those candidates supported by TAH, clinical practice led Committee members to believe that these two candidate groups are reasonable in the same status, and that IABP candidates are comparable to candidates supported by ACS devices. Like the ACS device patients, for this status the Committee also proposes a requirement that the candidate be treated with IABP specifically for cardiogenic shock, and created a hemodynamic threshold of showing the candidate had a cardiac index of less than or equal to 2.2 L/min/m 2 prior to ACS implantation.

Decreased utilization in donor hearts is an issue Khush et al. Am J Transplant. 2015; 15: 642 649

Waitlist candidates added/organ transplants performed Heart transplants performed vs waitlist candidates added, USA 1995-2014 5000 Transplants Performed Waitlist Candidates Added 4500 4000 3500 3000 2500 2000 1500 1000 500 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year From UNOS/SRTR data, available at https://optn.transplant.hrsa.gov/

ECMO PRIORITY Will ECMO in highest priority incentivize increased use of ECMO? If so, will post-transplant outcomes be worse? Is there potential for outcomes to be better if ECMO patients are transplanted quicker? Assessment of net transplant benefit 35