June 22, 2020

Bedside Mechanical Assist Devices for Severe Heart Failure/Cardiogenic Shock:

When, Which One and How?

Nir Uriel MD, MSc, FACCProfessor of Medicine Columbia UniversityProfessor of Medicine Cornell University

Director of Heart Failure, Heart Transplantation & Mechanical Circulatory Support Programs

New York Presbyterian

Interagency Registry of Mechanically Assisted Circulatory Support (INTERMACS)

Traditional Definition of Cardiogenic Shock

Hypotension SBP < 90 mmHg for > 30 min Need for vasopressors, inotropes, or mechanical circulatory support to maintain SBP > 90 mmHg

Diminished cardiac output Cardiac index < 2.0 - 2.2 liters/min/m2

Elevated filling pressures Pulmonary capillary wedge pressure > 15 mmHgCentral venous pressure > 15 mmHg

Evidence of end organ dysfunction

New Pyramidal Scheme of CS

Performance of the SCAI Classification Scheme

• 10,004 patients who were admitted to the Mayo Clinic CCU from 2007 – 2015 (all comers)• 43.1% had acute coronary syndrome• 46.1% had heart failure• 12.1% with cardiac arrest

• SCAI class was assigned based on data from the hospitalization

• Each higher SCAI shock stage was associated with a step-wise increase in hospital mortality

Jentzer et al. JACC 2019

SCAI Classification Scheme Applied to Shock Cohort

• Study of 1007 consecutive patients presenting with cardiogenic shock or large MI between 2009 and 2017

• Higher SCAI classification was associated with lower 30-day survival

Schrage et al. Catheterization and Cardiovascular Interventions 2020

Cardiogenic Shock Pathophysiology

Reynolds H R , Hochman J S Circulation 2008;117:686-697Thiele H et al. Eur Heart J 2010;31:1828-1835

Goals of Treatment

1. Restore adequate end–organ perfusion– Break downward spiral– CO, MAP, oxygenation

2. Ventricular unloading– Treat/prevent pulmonary edema– Minimize myocardial o2 demand– Prevent/minimize remodeling

Hochman JS. Circulation. 2003;107:2998-3002.Rihal CS et al, J Am Coll Cardiol. 2015;65:e7-e26

• Degree of LV and/or RV compromise• Short term recoverability of LV/RV function• SVR and PVR• Volume Status• Background medical therapy• Duration of cardiomyopathy

Variable Hemodynamic Response To Therapies

Treatment Approaches: Pharmacologic

Van Diepen JACC 2018

Rising Use of Acute MCS

Stretch J, et al. J Am Coll Cardiol 2014

LV Unloading

§LV Unloading is defined as the reduction of total mechanical power expenditure (PVA·HR) of the ventricle which correlates with reductions in myocardial oxygen consumption and hemodynamic forces that lead to ventricular remodeling.

PVA = pressure-volume areaHR = heart rate

13

Pressure Volume Loop

Pressure Volume LoopESPVR – End Systolic Pressure Volume Relationship

right ventricle [RV], or whether it is a first event,with previously normal heart structure); 2) the de-gree of acute LV recovery following initiation ofMCS (e.g., potentially recoverable in some forms ofacute coronary syndrome, but less likely recover-able with idiopathic cardiomyopathy); 3) right-sidedfactors, such as RV systolic and diastolic functionand pulmonary vascular resistance; 4) the degree towhich baroreflexes are intact and can modulatevascular and ventricular properties; 5) concomitantmedications; and 6) metabolic factors, such aspH and pO2, which, if corrected, could result in

improved ventricular and vascular function. Finally,the characteristics of the pump (e.g., pulsatile,axial, or centrifugal flow) can also have an impacton several aspects of the hemodynamic responsesto MCS (14).

It is therefore important to understand anddistinguish between the primary hemodynamiceffects of a device (i.e., the expected effects onpressures and flow in the absence of any change innative heart or vascular properties) and the nethemodynamic effects observed after accounting forthe impact of secondary modulating factors invoked

FIGURE 1 Overview of PVLs and Relations

150

100

50

00 50 100 150

LV Volume (ml)

LV P

ress

ure

(mm

Hg)

150

100

50

00 50 100 150

LV Volume (ml)

LV P

ress

ure

(mm

Hg)

150

100

50

00 50 100 150

LV Volume (ml)

LV P

ress

ure

(mm

Hg)

Vo ESV EDV Vo EDV

Ees

Ea

ESPVR

EDPVR

Ea

ESPVR

TPR or HR

Baseline

TPR or HR

Preload

V120

EesBase

line

Ees

A B C

(A) Normal pressure–volume loop (PVL), is bounded by the end-systolic pressure–volume relationship (ESPVR) and end-diastolic pressure–volume rela-tionship (EDPVR). ESPVR is approximately linear with slope end-systolic elastance (Ees) and volume–axis intercept (Vo). Effective arterial elastance (Ea) isthe slope of the line extending from the end-diastolic volume (EDV) point on the volume axis through the end-systolic pressure–volume point of the loop.(B) Slope of the Ea line depends on total peripheral resistance (TPR) and heart rate (HR), and its position depends on EDV. (C) The ESPVR shifts with changesin ventricular contractility, which can be a combination of changes in Ees and Vo. Changes in contractility can be indexed by V120, the volume at which theESPVR intersects 120 mm Hg. ESV ¼ end-systolic volume; LV ¼ left ventricular.

FIGURE 2 Characteristics of the EDPVR

40

30

20

10

00 50 100 150 200 250

LV Volume (ml)

LV P

ress

ure

(mm

Hg)

40

30

20

10

00 50 100 150 200 250

LV Volume (ml)

LV P

ress

ure

(mm

Hg)

P/VdP/dV

P/VdP/dV

V30

A B

(A) The EDPVR is nonlinear. Stiffness is indexed by the change in pressure divided by the change in volume (dP/dV), varies with pressure. P/V,the ratio of end-diastolic pressure to volume, also varies with pressure. The myocardial stiffness constant, (dP/dV)/(P/V), is considered a validmeasure of myocardial diastolic material properties. (B) One clinically useful index of diastolic properties is ventricular capacitance, which is thevolume at a specified pressure such as V30, the volume at 30 mm Hg. Abbreviations as in Figure 1.

Burkhoff et al. J A C C V O L . 6 6 , N O . 2 3 , 2 0 1 5

Hemodynamics of Circulatory Support D E C E M B E R 1 5 , 2 0 1 5 : 2 6 6 3 – 7 4

2666

Burkoff D /Sayer G/ Uriel N et al. JACC 2015

Pressure Volume Area

Burkoff D /Sayer G/ Uriel N et al. JACC 2015following initiation of MCS. Both components ofdevice effects will be discussed later.

Finally, use of the theories of ventricular me-chanics detailed earlier within the context of acomprehensive cardiovascular simulation (9,10)facilitates illustration and comparison of the hemo-dynamic effects of different forms of MCS. Thesimulation we used has been detailed, can be used tounderstand the physiology of MCS, and has beenvalidated to a certain degree pre-clinically (15). Otheraspects of validation and limitations of the simulationhave also been detailed previously (15–17). Note thatthe response of a given patient to MCS must accountfor baseline pre-load, afterload LV contractility, andthe flow rate of the MCS pump. For simplicity, sub-sequent comparisons keep these factors constant.Importantly, the basic principles to be discussedapply across a wide range of conditions.RA-TO-ARTERIAL CIRCULATORY SUPPORT. Extra-corporeal venoarterial membrane oxygenation(ECMO), also referred to as extracorporeal life sup-port, utilizes a pump with the capacity to assumeresponsibility for the entire cardiac output and a gasexchange unit for normalizing pCO2, pO2, and pH.However, strictly on a hemodynamic basis, the use ofthis circuit configuration can cause significant in-creases in LV pre-load and, in some cases, pulmonaryedema. This is illustrated in Figure 4A, which depictsPVLs in a case of cardiogenic shock due to profound,irreversible LV dysfunction. Baseline cardiogenicshock conditions (PVL in black) have a high LV EDP,low pressure generation, low SV, and low ejectionfraction. As ECMO flow is initiated and increasedstepwise from 1.5 to 3.0 to 4.5 l/min, the primary

hemodynamic effect is increased LV afterload pres-sure and effective Ea. If TPR and LV contractility arefixed, the only way for the LV to overcome theincreased afterload is via the Starling mechanism, andblood accumulates in the LV. Consequently, LV EDP,LA pressure, and pulmonary capillary wedge pressure(PCWP) increase, and the PVL becomes increasinglynarrow (decreased native LV SV) and taller (increasedafterload pressure), and shifts rightward and upwardalong the EDPVR. Because the EDPVR is nonlinear,large increases in LV EDP may cause only subtleincreases in LV EDV. An echocardiogram showing apersistently closed aortic valve during ECMO wouldalso signify a state of maximal LV loading and highPCWP. These increases in LV pre-load and PCWP aredetrimental to blood oxygen saturation coming fromthe lung and markedly increase myocardial oxygendemand (increased PVA), which can worsen LVfunction, especially in the setting of acute myocardialischemia or infarction.

These responses to ECMO can be modulated bysecondary regulatory factors that influence eitherTPR or LV contractility. TPR can be reduced naturallyby the baroreceptors, pharmacologically (e.g., nitro-prusside), or mechanically (e.g., by intra-aorticballoon pumping). As illustrated in Figure 4B, a 50%reduction in TPR during ECMO markedly blunts therise in LV EDP.

Short-term improvements in LV function can alsomodulate the rise in PCWP. LV function can beimproved during ECMO due to increased central aorticpressure, the improved coronary perfusion, normali-zation of blood oxygen content (improved oxygendelivery to the myocardium), and normalization of

FIGURE 3 Myocardial Energetics Assessed on the Pressure–Volume Diagram

150

100

50

00 50 100 150

LV Volume (ml)

LV P

ress

ure

(mm

Hg)

PVA=SW+PE0.20

0.15

0.10

0.05

0 5000 10000 15000PVA (mm Hg.ml)

MVO

2 (m

lO2/

beat

)

Oxygen for:Mechanical WorkCalcium CyclingBasal Metabolism

A B

PE

SW

(A) Pressure–volume area (PVA) is the sum of the stroke work (SW) and potential energy (PE). (B) Myocardial oxygen consumption (MVO2) islinearly correlated with PVA and is divided into 3 major components, as indicated in the figure. LV ¼ left ventricular.

J A C C V O L . 6 6 , N O . 2 3 , 2 0 1 5 Burkhoff et al.D E C E M B E R 1 5 , 2 0 1 5 : 2 6 6 3 – 7 4 Hemodynamics of Circulatory Support

2667

Most Common Devices Used in CS

• There are several devices for use in cardiogenic shock, each with a different risk profile and potential benefit to the patient

• The biggest factors in choosing a device for patients is how much circulatory support is required and whether LV, RV or BIV failure is present

Intraaortic Balloon Pump

Impact of IABP in CGSPressure-Volume

LVP and AoP

Small âPCWPSmall áCO

Lack of Hemodynamic Benefit of IABP in Acute MI with Cardiogenic Shock (IABP-Shock Trial)

Prondzinsky R et al. Shock 2012

N Engl J Med 2012;367:1287-96.

V-A ECMO

Keebler M JACC HF 2018 22

Impact of RAàAo MCS (ECMO) on Hemodynamics and Energetics

↑Afterload↑Preload

Pressure-Volume

AoP and LVP

↑AoP↑LVP

Impact of RAàAo MCS (ECMO) on Hemodynamics and Energetics

↑Afterload↑Preload

Pressure-Volume

PVA-MVO2 (ml O2/min)

↑PVA↑MVO2

ECMO – The Advantages

Easily deployed (at bedside if needed)

Significant hemodynamic support with flows of 5+L possible

Biventricular support Circulatory support is maintained independent of ventricular functionIn extreme examples, flow is maintained even in VF

Acute MCS device which supports gas exchange

Less hemolysis

ECMO – The Disadvantages

Inadequate ventricular unloading� ECMO maintains end organ perfusion but

may actually increase cardiac work

� This may cause pressure overload which can lead to LV distention and pulmonary edema

Large Cannula size (V=23F, A=15-17F)� Bleeding is common

� Ischemic complications cause major morbidity

Resource intensive

Percutaneous Pumps

Impella Family of Devices2.5/4.0/5.0/RP

Thoratec PHP, Tranvalvular Pump with Expandable Cage (Experimental)

TandemHeartLAàAo Pump

Transvalvular Transseptal

Impact of LAàAo MCS in CGSPressure-Volume

LVP and AoP

áBPáTotal COâPCWP

TandemHeart pVAD in refractory cardiogenic shock

§Baseline patient characteristics:–117 severe cardiogenic shock patients from 2003-08–All patients refractory to IABP and multiple vasopressors–48% of patients undergoing CPR at time of implant–68% ischemic, 32% non-ischemic cardiomyopathy

Kar B, et al. JACC 2011

Impact of LVàAo MCS on Hemodynamics and Energetics

Pressure-Volume

LVP and AoP

↓PeakLVP↓Preload

↑AoP↓LVPLV-Ao

Uncoupling

Impact of LVàAo MCS on Hemodynamics and Energetics

Pressure-Volume

PVA-MVO2 (ml O2/min)

↓PeakLVP↓Preload

↓PVA↓MVO2

32

LVAD Unloading and Decoupling

Uriel N / Sayer G / Burkhoff D JACC 2018

Dagmar M. Ouweneel, et all JACC, Volume 69, Issue 3, 2017, 278–287

Percutaneous Mechanical Circulatory Support Versus Intra-Aortic Balloon Pump in Cardiogenic Shock After Acute Myocardial Infarction - IMPRESS

(>20 or <20 min), lactate level >7.5 or <7.5 mmol/l,TIMI (Thrombolysis In Myocardial Infarction) flowpost-PCI, systolic blood pressure before IABP or pMCSplacement (>80 or <80 mm Hg), and the presence orabsence of traumatic injuries on admission.

RESULTS

PATIENTS. Between June 1, 2012, and September 15,2015, a total of 48 patients were randomly assigned toeither pMCS therapy (n ¼ 24) or IABP (n ¼ 24)(Figure 1). The baseline characteristics of the 2 groupswere well balanced (Table 1). The mean age was58 years, 79% were male, all patients were mechani-cally ventilated, 96% of the patients received cate-cholamines, and 92% had had a cardiac arrest beforerandomization.

TREATMENT. Randomization and placement of thepMCS or IABP took place after revascularizationexcept for 8 patients in whom IABP or the pMCS wasinitiated before revascularization (3 in the IABP groupand 5 in the pMCS group) (Table 2). The infarct-related artery was the left anterior descending (LAD)in the majority of the patients (65%) and 98% of thepatients were treated with stent placement. The me-dian duration of circulatory support was 48 h (IABP)and 49 h (pMCS). All patients were treated with cat-echolamines during admission to the intensive careunit, 31% received renal replacement therapy, and75% were treated with therapeutic hypothermia(Table 3).

Of the patients in the IABP group, 1 patient sub-sequently received pMCS and was transferred toanother hospital for treatment with extracorporeallife support oxygenation. Two patients received analternative device, the Impella 5.0 (Abiomed, Aachen,Germany), after the IABP treatment, and 1 of themreceived subsequent extracorporeal life support andan LVAD at another hospital. Of the patients treatedwith the pMCS, 1 patient subsequently received theImpella 5.0. One patient was already on IABP supportbefore randomization (inserted before the start of theprimary PCI) and was randomized after the PCI topMCS treatment. Formally, this patient constitutes aprotocol violation, as IABP therapy before randomi-zation was an exclusion criterion. One patient did notreceive pMCS as the patient showed signs of recoveryafter randomization to receive device therapy.OUTCOMES. At 30 days, mortality was similar in pa-tients treated with IABP and pMCS therapy: 50% and46%, respectively (hazard ratio [HR] with pMCStherapy: 0.96; 95% confidence interval [CI]: 0.42 to2.18; p ¼ 0.92) (Table 4, Central Illustration). At 6months, the mortality rate was 50% in both groups

(HR: 1.04; 95% CI: 0.47 to 2.32; p ¼ 0.92). Only minordifferences were found in an analysis restricted to theper-protocol population from which 3 patientstreated with pMCS were excluded (Online Appendix).The Kaplan-Meier estimates for 6-month mortality inthe per-protocol population were 52% in the IABPgroup and 48% in the pMCS group (HR with pMCS:0.95; 95% CI: 0.41 to 2.21; p ¼ 0.91).

The primary cause of death at 6 months was braindamage (46% of the deceased patients; 6 of 12 in the

TABLE 1 Baseline Characteristics

pMCS(n ¼ 24)

IABP(n ¼ 24)

Age, yrs 58 " 9 59 " 11

Male 18/24 (75) 20/24 (83)

Body mass index, kg/m2 25 (23–26) 26 (25–27)

Cardiovascular risk factors

Current smoking 11/18 (61) 6/19 (32)

Hypertension 4/20 (20) 6/21 (29)

Hypercholesterolemia 4/20 (20) 5/21 (24)

Diabetes mellitus 2/22 (9) 3/23 (13)

Prior myocardial infarction 1/22 (5) 1/23 (4)

Prior stroke 0/22 (0) 1/23 (4)

Known peripheral arterial disease 2/23 (9) 0/23 (0)

Prior PCI or CABG 1/22 (5) 0/23 (0)

Hemodynamic variables before randomization

Heart rate, beats/min 81 " 21 83 " 28

Mean arterial pressure, mm Hg 66 " 15 66 " 15

Systolic blood pressure, mm Hg 81 " 17 84 " 19

Diastolic blood pressure, mm Hg 58 " 22 57 " 13

Medical therapy before randomization

Catecholamines or inotropes 24/24 (100) 22/24 (92)

Mechanical ventilation 24/24 (100) 24/24 (100)

Cardiac arrest before randomization 24/24 (100) 20/24 (83)

Witnessed arrest 22/24 (92) 17/20 (85)

First rhythm VT/VF 22/24 (92) 17/20 (85)

Time till return of spontaneouscirculation, min

21 (15–46) 27 (15–52)

Traumatic injuries at admission 5/24 (21) 2/24 (8)

Blood values on admission*

Lactate, mmol/l 7.5 " 3.2 8.9 " 6.6

Hemoglobin, mmol/l 8.6 " 1.2 8.6 " 1.2

Creatinine, mg/dl 96 " 29 102 " 22

Glucose, mmol/l 16.2 " 4.7 14.1 " 5.3

Arterial pH 7.14 " 0.14 7.17 " 0.17

Baseline echocardiography†

Estimated left ventricular ejection fraction

<20% 5/22 (23) 8/18 (44)

20%–40% 10/22 (46) 6/18 (33)

>40% 7/22 (32) 4/18 (22)

Values are mean " SD, n/N (%), or median (25th to 75th percentile). *Values are present for the followingnumber of patients: lactate (21 IABP and pMCS), hemoglobin (22 IABP and 21 pMCS), creatinine (23 IABP and 23pMCS), glucose (23 IABP and 20 pMCS), and pH (16 IABP and 20 pMCS). †First echocardiogram made duringadmission during the first day. In 20 patients, the echocardiography was done before pMCS or IABP placement,and in 21 patients, after pMCS or IABP placement (within 24 h).

CABG ¼ coronary artery bypass graft; IABP ¼ intra-aortic balloon pump; PCI ¼ percutaneous coronary inter-vention; pMCS ¼ percutaneous mechanical circulatory support; VF ¼ ventricular fibrillation; VT ¼ ventriculartachycardia.

J A C C V O L . 6 9 , N O . 3 , 2 0 1 7 Ouweneel et al.J A N U A R Y 2 4 , 2 0 1 7 : 2 7 8 – 8 7 Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock

281

(>20 or <20 min), lactate level >7.5 or <7.5 mmol/l,TIMI (Thrombolysis In Myocardial Infarction) flowpost-PCI, systolic blood pressure before IABP or pMCSplacement (>80 or <80 mm Hg), and the presence orabsence of traumatic injuries on admission.

RESULTS

PATIENTS. Between June 1, 2012, and September 15,2015, a total of 48 patients were randomly assigned toeither pMCS therapy (n ¼ 24) or IABP (n ¼ 24)(Figure 1). The baseline characteristics of the 2 groupswere well balanced (Table 1). The mean age was58 years, 79% were male, all patients were mechani-cally ventilated, 96% of the patients received cate-cholamines, and 92% had had a cardiac arrest beforerandomization.

TREATMENT. Randomization and placement of thepMCS or IABP took place after revascularizationexcept for 8 patients in whom IABP or the pMCS wasinitiated before revascularization (3 in the IABP groupand 5 in the pMCS group) (Table 2). The infarct-related artery was the left anterior descending (LAD)in the majority of the patients (65%) and 98% of thepatients were treated with stent placement. The me-dian duration of circulatory support was 48 h (IABP)and 49 h (pMCS). All patients were treated with cat-echolamines during admission to the intensive careunit, 31% received renal replacement therapy, and75% were treated with therapeutic hypothermia(Table 3).

Of the patients in the IABP group, 1 patient sub-sequently received pMCS and was transferred toanother hospital for treatment with extracorporeallife support oxygenation. Two patients received analternative device, the Impella 5.0 (Abiomed, Aachen,Germany), after the IABP treatment, and 1 of themreceived subsequent extracorporeal life support andan LVAD at another hospital. Of the patients treatedwith the pMCS, 1 patient subsequently received theImpella 5.0. One patient was already on IABP supportbefore randomization (inserted before the start of theprimary PCI) and was randomized after the PCI topMCS treatment. Formally, this patient constitutes aprotocol violation, as IABP therapy before randomi-zation was an exclusion criterion. One patient did notreceive pMCS as the patient showed signs of recoveryafter randomization to receive device therapy.OUTCOMES. At 30 days, mortality was similar in pa-tients treated with IABP and pMCS therapy: 50% and46%, respectively (hazard ratio [HR] with pMCStherapy: 0.96; 95% confidence interval [CI]: 0.42 to2.18; p ¼ 0.92) (Table 4, Central Illustration). At 6months, the mortality rate was 50% in both groups

(HR: 1.04; 95% CI: 0.47 to 2.32; p ¼ 0.92). Only minordifferences were found in an analysis restricted to theper-protocol population from which 3 patientstreated with pMCS were excluded (Online Appendix).The Kaplan-Meier estimates for 6-month mortality inthe per-protocol population were 52% in the IABPgroup and 48% in the pMCS group (HR with pMCS:0.95; 95% CI: 0.41 to 2.21; p ¼ 0.91).

The primary cause of death at 6 months was braindamage (46% of the deceased patients; 6 of 12 in the

TABLE 1 Baseline Characteristics

pMCS(n ¼ 24)

IABP(n ¼ 24)

Age, yrs 58 " 9 59 " 11

Male 18/24 (75) 20/24 (83)

Body mass index, kg/m2 25 (23–26) 26 (25–27)

Cardiovascular risk factors

Current smoking 11/18 (61) 6/19 (32)

Hypertension 4/20 (20) 6/21 (29)

Hypercholesterolemia 4/20 (20) 5/21 (24)

Diabetes mellitus 2/22 (9) 3/23 (13)

Prior myocardial infarction 1/22 (5) 1/23 (4)

Prior stroke 0/22 (0) 1/23 (4)

Known peripheral arterial disease 2/23 (9) 0/23 (0)

Prior PCI or CABG 1/22 (5) 0/23 (0)

Hemodynamic variables before randomization

Heart rate, beats/min 81 " 21 83 " 28

Mean arterial pressure, mm Hg 66 " 15 66 " 15

Systolic blood pressure, mm Hg 81 " 17 84 " 19

Diastolic blood pressure, mm Hg 58 " 22 57 " 13

Medical therapy before randomization

Catecholamines or inotropes 24/24 (100) 22/24 (92)

Mechanical ventilation 24/24 (100) 24/24 (100)

Cardiac arrest before randomization 24/24 (100) 20/24 (83)

Witnessed arrest 22/24 (92) 17/20 (85)

First rhythm VT/VF 22/24 (92) 17/20 (85)

Time till return of spontaneouscirculation, min

21 (15–46) 27 (15–52)

Traumatic injuries at admission 5/24 (21) 2/24 (8)

Blood values on admission*

Lactate, mmol/l 7.5 " 3.2 8.9 " 6.6

Hemoglobin, mmol/l 8.6 " 1.2 8.6 " 1.2

Creatinine, mg/dl 96 " 29 102 " 22

Glucose, mmol/l 16.2 " 4.7 14.1 " 5.3

Arterial pH 7.14 " 0.14 7.17 " 0.17

Baseline echocardiography†

Estimated left ventricular ejection fraction

<20% 5/22 (23) 8/18 (44)

20%–40% 10/22 (46) 6/18 (33)

>40% 7/22 (32) 4/18 (22)

Values are mean " SD, n/N (%), or median (25th to 75th percentile). *Values are present for the followingnumber of patients: lactate (21 IABP and pMCS), hemoglobin (22 IABP and 21 pMCS), creatinine (23 IABP and 23pMCS), glucose (23 IABP and 20 pMCS), and pH (16 IABP and 20 pMCS). †First echocardiogram made duringadmission during the first day. In 20 patients, the echocardiography was done before pMCS or IABP placement,and in 21 patients, after pMCS or IABP placement (within 24 h).

CABG ¼ coronary artery bypass graft; IABP ¼ intra-aortic balloon pump; PCI ¼ percutaneous coronary inter-vention; pMCS ¼ percutaneous mechanical circulatory support; VF ¼ ventricular fibrillation; VT ¼ ventriculartachycardia.

J A C C V O L . 6 9 , N O . 3 , 2 0 1 7 Ouweneel et al.J A N U A R Y 2 4 , 2 0 1 7 : 2 7 8 – 8 7 Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock

281

Dagmar M. Ouweneel, et all JACC, Volume 69, Issue 3, 2017, 278–287

Percutaneous Mechanical Circulatory Support Versus Intra-Aortic Balloon Pump in Cardiogenic Shock After Acute Myocardial Infarction - IMPRESS

bleeding in pMCS-treated patients compared withIABP-treated patients were also described in aregistry comparing Impella 2.5 and IABP in a post-cardiac arrest population (n ¼ 78), which foundsevere bleeding in 26% of the device patients versus6% of IABP patients (20). Two large multicenterImpella 2.5 registries describe the rates of bleedingrequiring transfusion of 24.2% and 17.5% and therates of hemolysis as 7.5% and 10.3% (21,22). Thesecomplication rates are comparable to the pMCS inour study (33.3% bleeding and 8.3% hemolysis).

The IABP-SHOCK II trial reports 20.7% bleeding in theIABP patients (and 20.8% in the control group). Thisis higher than the 8.2% bleeding in our IABP group.

Although discouraged, some crossovers and up-grades to other mechanical support therapy did takeplace: 3 in the IABP group and 1 in the pMCS group.Crossover or upgrading was solely at the discretion ofthe investigator. There was a trend toward moreupgrading/crossover in the IABP group.

Upon initiation of our study, IABP therapy was stillrecommended in the guidelines for CS, but wasdowngraded to a Class III recommendation in theEuropean guidelines and Class II in the Americanguidelines during the inclusion period of the study.After consultation with the institutional review boardand in the light of the severity of clinical conditionwith higher mortality rates than in the IABP-SHOCK IIstudy, the control therapy remained unchanged. Inaddition, after the interim analysis it was clear thestudy was underpowered to show a difference inmortality at 30 days, and the executive committeeallowed it to proceed for exploratory purposes.

Although not adequately powered, our trial sug-gests that in patients with CS without selection onage, ROSC times, and pre-procedural traumatic in-juries, no clear signal of superior outcome wasobserved in patients with pMCS support whencompared with the IABP.

There may be several reasons why the pMCS-treated patients did not show improved mortalityrates. A possible explanation is the unselective natureof the patients included in the study. In our study,92% of patients had resuscitated cardiac arrest, whichimplies a prevalence of post-anoxic neurologicaldamage present at the moment of randomization.Any kind of mechanical circulatory support may be oflimited clinical utility in these patients. Anotherexplanation might be that CS after AMI is not only amatter of low cardiac output. The shock syndromealso comprises an irreversible damage due to dimin-ished organ perfusion and inflammatory responses.Hence, providing mechanical hemodynamic supportmay not be enough to reverse the damage that hasalready occurred. Although the Impella CP can pro-vide up to 3.5 l/min of forward flow, it might still beinsufficient to reverse severe CS with advanced endorgan failure, especially as in clinical practice, long-term Impella CP support achieves <3.5 l/min hemo-dynamic support. In this trial, the main rationalefor using Impella CP instead of a device that canprovide even more hemodynamic support (e.g.,Impella 5.0), was the need for a surgical cut-downfor implantation. The Impella CP can be insertedpercutaneously, which enables quick insertion even

TABLE 4 Clinical and Functional Outcomes

pMCS(n ¼ 24)

IABP(n ¼ 24) p Value

Hazard Ratio WithpMCS (95% CI)

Mortality*

30-day all-cause mortality 11 (46) 12 (50) 0.92 0.96 (0.42–2.18)

6-month all-cause mortality 12 (50) 12 (50) 0.92 1.04 (0.47–2.32)

Clinical outcomes at 6 months

Cause of death

Refractory cardiogenic shock 4 (17) 3 (13)

Post-anoxic neurological death 5 (21) 6 (25)

Other reason 3 (13) 3 (13)

Stroke 1 (4) 1 (4)

Hemorrhagic stroke 0 (0) 0 (0)

Ischemic stroke 1 (4) 1 (4)

Major vascular complication 1 (4) 0 (0)

Major bleeding 8 (33) 2 (8)

Device-related bleeding 3 (13) 1 (4)

Retroperitoneal 1 (4) 0 (0)

IABP/Impella puncture site 2 (8) 1 (4)

Nondevice-related bleeding 5 (21) 1 (4)

Gastro-intestinal bleeding 0 (0) 1 (4)

Bleeding at other puncture site 1 (4) 0 (0)

Other location 4 (17) 0 (0)

Hemolysis requiring extractionof the device

2 (8) 0 (0)

Device failure requiring extraction 0 (0) 0 (0)

Surgical LVAD placement 0 (0) 1 (4)

Heart transplantation 0 (0) 0 (0)

Other surgery 2 (8) 0 (0)

Myocardial (re)infarction 1 (4) 2 (8)

Repeat PCI 0 (0) 3 (13)

CABG 0 (0) 1 (4)

Rehospitalization 5 (21) 1 (4)

Cardiac 2 (8) 0 (0)

Noncardiac 3 (13) 1 (4)

Echocardiography at 6 months†

Left ventricular dimensions andsystolic function (n)

12 9‡

Ejection fraction (%) 46 " 11 49 " 9

End-diastolic volume (ml) 122 " 41 120 " 33

End-systolic volume (ml) 65 " 31 61 " 21

Values are frequencies (%) or mean " SD, unless otherwise indicated. Additional information about events can befound in the Online Appendix. *Mortality is shown as Kaplan-Meier estimates. †First available echocardiogramafter 2 months. Median follow-up time is 191 (176 to 297) days. ‡One patient with surgical LVAD, 1 patient lost tofollow-up, and 1 patient bedridden due to multiple sclerosis.

CI ¼ confidence interval; other abbreviations as in Tables 1 and 3.

Ouweneel et al. J A C C V O L . 6 9 , N O . 3 , 2 0 1 7

Percutaneous Mechanical Circulatory Support Versus IABP in Cardiogenic Shock J A N U A R Y 2 4 , 2 0 1 7 : 2 7 8 – 8 7

284

RV Support – Impella RP

Anderson MB et al. JHLT 2015

Impella Complications

Impella Complications

Percutaneous LVAD vs. IABP: Hemodynamic Effects

Cheng J M et al. Eur Heart J 2009;30:2102-2108

RAàAo MCS + LVàAo MCS(ECPELLA)

↑Afterload↑Preload

Pressure-Volume

AoPLVP

Centrimag BiVAD

Takayama et al Circ HF 2014

Eur J Cardiothorac Surg. 2017;52(6):1055-1061. doi:10.1093/ejcts/ezx189

Minimally invasive CentriMag integrated with ECMO

43

Early Outcome

rotational speed of 5500 rpm. Various types of cannulas can beconnected to the CentriMag system. These features allow easy in-sertion, flexible configuration, combination with an oxygenator,

mid-term use and easy postoperative maintenance. We have pre-viously reported our CentriMag VAD experience for various car-diogenic shock patients as a bridge to decision therapy.Outcome of our approach was favourable with 30-day survival of69% and 1-year survival of 49% [3].

Biventricular support is preferred for patients in cardiogenicshock. However, BiVAD implantation requires sternotomy andcardiopulmonary bypass, which limits its emergent use and maybe contraindicated in patients with severe coagulopathy orother relative contraindications to sternotomy. To overcomethese drawbacks, we previously developed a minimally invasiveapico-axillary CentriMag VAD insertion technique [7]. Saitoet al. reported an apico-ascending aortic temporary VAD usingECMO circuit through a minimally invasive approach [8].However, these techniques are reserved for patients who can bestabilized with isolated left VAD support. The use of percutan-eous biventricular support with 2 Impella pumps was reported,but this technique is obviously limited with regard to patientmobility and support duration [9]. The advantage of our Ec-VADtechnique is the ability to integrate right VAD with minimallyinvasive apico-axillary VAD. Moreover, once the patient is stabi-lized, we can easily convert Ec-VAD to apico-axillary VAD byeliminating venous limb and oxygenator (Fig. 1A and B). Then,the patient can participate in ambulatory rehabilitation andwait for a month for recovery.

Compared to the BiVAD cohort, Ec-VAD patients were olderand many patients had INTERMACS profile 1 [10]. Therefore, theEc-VAD cohort was likely sicker than BiVAD cohort. Furthermore,more patients had percutaneous ST-MCS support before receiv-ing CentriMag VAD (INTERMACS I TCS). This trend probablyreflects change of our practice to treat cardiogenic shock pa-tients during this study period. Percutaneous ST-MCS such asECMO has gained more popularity as a primary resuscitation de-vice for INTERMACS profile 1 patients. Currently, Ec-VAD is ourprimary mode of biventricular support for cardiogenic shockpatients who cannot be optimized by percutaneous ST-MCS andneed circulatory support longer than 1 week. We converted toEc-VAD after 3.2 days of percutaneous ST-MCS supports includ-ing femoral veno-arterial and/or Impella. We believe that earlyconversion is better to provide powerful ventricular unloadingeffect and prolonged circulatory support with ambulatoryrehabilitation.

Patients with Ec-VAD required less transfusion, despite signifi-cantly lower preoperative platelet count. This is attributed to theminimally invasive approach without using cardiopulmonary by-pass and sternotomy. Another interesting feature of Ec-VAD isthat it can facilitate a subsequent durable VAD surgery as shownin Table 4. From our experiences, in patients with BiVAD, weoften encountered dense adhesions around the cannula that cer-tainly increased the surgical complexity when BiVAD was con-verted to durable VAD. Subgroup analysis showed that Ec-VADsignificantly decreased the amount of transfusion and also car-diopulmonary bypass time. We found very minimal adhesion in-side the pericardium when we converted Ec-VAD to durableVAD. Although the number of Ec-VAD patients who reached dur-able VAD is still small, there was no mortality associated withbridge to durable VAD surgery.

The Ec-VAD provided lower 30-day mortality compared toBiVAD (13% vs 24%). Also, current study showed that Ec-VADoffers comparable mid-term survival to BiVAD. However, it failedto show significant survival benefit compared to BiVAD. This islikely related to baseline acuity of the cardiogenic shock cohort.

Table 2: Intraoperative variables

BiVAD(n = 90)

Ec-VAD(n = 22)

P-value

Use of cardiopulmonarybypass, n (%)

59 (65.6) 0 (0) 0.0001

Cardiopulmonary bypasstime (min)

85.3 ± 44.6 0 (0) 0.0000

Transfusion (unit)pRBC 3.25 ± 3.49 1.64 ± 2.11 0.0411FFP 4.09 ± 4.34 1.59 ± 1.94 0.0098Platelets 12.6 ± 9.72 7.64 ± 7.45 0.0279

Cannulation, n (%)LVAD inflow

LV through LV apex 59 (65.6) 22 (100)LA 23 (25.6) 0 (0)LV through LA 8 (8.89) 0 (0)

LVAD outflowAorta 88 (97.8) 0 (0)Axillary artery 2 (2.22) 22 (100)

RVAD inflowRA 90 (100) 0 (0)RA through femoral vein 0 (0) 22 (100)

RVAD outflowPulmonary artery 90 (100) 0 (0)

Mean systemic flow (l/min) 5.71 ± 1.08 5.50 ± 0.937 0.402

BiVAD: biventricular assist device; Ec-VAD: extracorporeal ventricularassist device; pRBC: packed red blood cells; FFP: fresh frozen plasma;LVAD: left ventricular assist device; LV: left ventricle; LA: left atrium;RVAD: right ventricular assist device; RA: right atrium; VAD: ventricularassist device.

Table 3: Early outcomes and next destinations

BiVAD(n = 90)

Ec-VAD(n = 22)

P-value

Support duration, days 24.2 ± 18.6 28.6 ± 14.0 0.299Major morbidity during

support, n (%)Major bleeding 59 (71.9) 7 (31.8) 0.0069Delayed sternal closure 34 (37.8) 0 (0) 0.0002Stroke 13 (14.4) 4 (18.2) 0.741

Mortality on device, n (%) 19 (21.1) 4 (18.2) 1.00Multi-organ failure/sepsis 12 1Stroke 7 2Other 0 1

Thirty-day mortality, n (%) 22 (24.4) 3 (13.6) 0.395Destination, n (%) 66 (73.3) 17 (77.3) 0.449

Recovery (weanedoff device)

25 (27.8) 6 (27.3)

Heart transplant 12 (13.3) 1 (4.55)Durable VAD 29 (32.2) 10 (45.5)Pulsatile LVAD 3 0Continuous-flow LVAD 24 10Continuous-flow BiVAD 2 0

BiVAD: biventricular assist device; Ec-VAD: extracorporeal ventricular assistdevice; LVAD: left ventricular assist device; VAD: ventricular assist device.

TX&

MCS

1059K. Takeda et al. / European Journal of Cardio-Thoracic Surgery

Downloaded from https://academ

ic.oup.com/ejcts/article-abstract/52/6/1055/3884560 by University of Chicago user on 03 Decem

ber 2018

Eur J Cardiothorac Surg. 2017;52(6):1055-1061. doi:10.1093/ejcts/ezx189

Team Based Care

TEAM – BASED is not limited to the choice of device!

Van Diepen S, et al. Circulation. 2017;136:e232–e268.

Basir MB, et al. Am J Cardiol 2017;119:845-851.

Paradigm Shift: Door to Support

Conclusions

• Mortality from CS remains unacceptably high• A new classification scheme for CS may provide important insights and a

framework for effective communication of severity of CS• The PA catheter remains an important tool in assessing and monitoring CS

patients • Acute MCS now forms the cornerstone of treatment of severe CS • The “Shock Team” approach is important to rapid recognition, triage and

treatment of CS patients and recent evidence supports its implementation can improve outcomes in CS

Thank you

476/22/20